-------
Preface
This document was finalized in June 1989 and reviews
information from relevant studies of O3 health and welfare
effects and of exposure and risk analysis through early 1989.
The assessment contained in this staff paper reflects information
in the documents "Air Quality Criteria for Ozone and Other
Photochemical Oxidants" (EPA-600/8-84-020F) and "Summary of
Selected New .Information on Effects of Ozone on Health and
Vegetation: Supplement to Air Quality Criteria for Ozone and
Other Photochemical Oxidants" . (EPA-600/8-88/l-5a).
-------
-------
Acknowledgements
This staff paper is the product of the Office of Air Quality
Planning and Standards (OAQPS). Tables and Figures .not otherwise
cited are original to this report. The principal authors include
Dr. David J. McKee, Ms. Pamela M. Johnson, Mr. Thomas R. McCurdy,
and Mr. Harvey M. Richmond. This report has been improved by
comments from other staff within OAQPS, the Office of Research
and Development, the Office of Policy and Program Evaluation, and
the Office of General Counsel within EPA. Three drafts were '
formally reviewed by the clean Air Scientific Advisory Committee
and comments incorporated. Particularly important in the final
review of this staff paper was the technical and editorial
support provided by Ms; victoria Atw.ll and the clerical and
editorial support of Mrs. Patricia R. Crabtree and Mrs. Barbara
Miles.
Helpful comments and suggestions were also submitted by a
number of independent scientists, by officials from the state
environmental agencies of Illinois, Minnesota, California and
Texas, by the Department of the Havy, and the Department of
Energy, and by environmental and industrial groups including the
Natural Resources Defense Council, the American Lung Association,
the Chemical Manufacturers Association, the American Petroleum
institute, and the Motor Vehicle Manufacturers Association.
-------
Project Team For
Review of the National Ambient Air Quality Standards for Ozone
Dr. David J. McKee, Project Manager and Author of Chapters I
through III and VI through VI11
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Ms. Pamela M. Johnson, Author of Chapters IX through XI
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Mr. Thomas R. McCurdy, Author of Chapters IV and V and Appendix A
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Mr. Harvey M. Richmond, Author of Section VII.B,,
Ambient Standards Branch, Air Quality Management Division
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
-------
iii
U.S. Environment,il Protection Agency
Science Advisory Board
Clean Air Scientific Advisory Committee
-OH
Chairman
Dr. Roger O. McClellan
CUT
Post Office Box 12137
Research Triangle Park, NC 27709
Members
Dr. Eileen G. Brennan
Department of Plant Pathology
Martin Hall, Room 213
Lipman Drive
Cook College-NJAES, Rutgers Univ.
P.O. Box 231
New Brunswick, New Jersey 08903
Dr. Edward D. Crandall
Division of Pulmonary Medicine
Starr Pavilion 505
Cornell Medical College
1300 York Avenue
New York, New York 10021
Dr. James D. Crapo
Box 3177
Duke University Medical Center
Durham, North Carolina 27711
Dr. Robert Frank
Professor of Environmental Health
Sciences
School of Hygiene
615 N. Wolfe Street
Baltimore, Maryland 21205
Prof. A. Myrick Freeman, in
Department of Economics
Bowdoin College
Brunswick, Maine 04011
-------
IV
Dr. Jay S. Jacobson
Plant Physiologist
Boyce Thompson Institute
Tower Road
Ithaca, New York 14853
Dr. Jane Q. Koenig
Research Associate Professor
Department of Environmental
Health SC-34
University of Washington
Seattle, Washington 98195
Dr. Timothy Larson
Environmental Engineering and
Science Program
Department of Civil Engineerincr
FX-10
University of Washington
Seattle, Washington 98195
Dr. Morton Lippmann, Professor
Institute of Environmental Medj cine
NYU Medical Center
Tuxedo, New York 10987
Prof. M. Granger Morgan
Head, Department of Engineering
and Public Policy
Carnegie-Mellon University
Pittsburgh, Pennsylvania 15253
Dr. D. Warner North, Principal
Decision Focus, Inc.
Los Altos Office Center
Suite 200
4984 El Camino Real
Los Altos, California 94022
Dr. Gilbert S. Omenn,
Professor and Dean
School of Public Health and
Community Medicine SC-30
University of Washington
Seattle, Washington 98195
Dr. Robert D. Rowe
Energy and Resource Consultants
P.O. Drawer 0
Boulder, Colorado 80306
-------
Dr. Marc B. Schenker, Director
Occupational and Environmental
Health Unit
University of California
Davis, California 95616
Mr. Stephen Smallwood
Air Pollution Control Program
Manager
Bureau of Air Quality Management
Florida Department of Environmental
Regulation
Twin Towers Office Bldg.
2600 Blair Stone Road
Tallahassee, Florida 32301
Dr. George Taylor
Environmental Sciences Division
P.O. Box X
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37831
Dr. Mark J. Utell
Pulmonary Unit - Box 692
Strong Memorial Hospital
Rochester, New York 14642
Dr. Jerry Wesolowski
1176 Shattuck Avenue
Berkeley, California 94704
Dr. George T. Wolff
Senior Staff Research Scientist
General Motors Research Labs
Environmental Science Department
Warren, Michigan 48090
-------
vi
EPA Reviewers
Mr. Allen C. Basala (MD-12)
?TfeiC™?f Air QualitY Planning and Standards, OAR
U.S. EPA
RTF, NC 27711
Mr. Frank L. Bunyard (MD-12)
Office of Air Quality Planning and Standards, OAR
U * o * EPA
RTF, NC 27711
Dr. Thomas c. Curran (MD-14)
Office of Air Quality Planning and Standards, OAR
U.S. EPA
RTF, NC 27711
.Mr. Robert Fegley (PM-221)
Office of Policy Analysis, OPPE
U.S. EPA
Waterside Mall
401 M Street, SW
Washington, DC 20460
Mr. Lewis Felleisen
Air Programs & Engineering Branch
U.S. EPA, Region III
Curtis Building
6th & Walnut Streets
Philadelphia, PA 19106
Mr. Robert A. Flaak (A-107F)
Science Advisory Board, OA
U.S. EPA
Waterside Mall
401 M Street, SW
Washington, DC 20460
Dr. J.H.B. Garner (MD-52)
Environmental Criteria and Assessment Office, ORD
U.S. EPA
RTF, NC 27711
Mr. Gerald K. Gleason (LE-132A)
Office of General Counsel
U.S. EPA '
Waterside Mall
401 M Street, SW
Washington, DC 20460
-------
vii
Dr. Judith A. Graham (MD-52)
Environmental Criteria and Assessment office, ORD
RTF, NC 27711
Dr. Lester D. Grant (MD-52)
U?l!rEPASntal Criteria and Assessment Office, ORD
RTF, NC 27711
Dr. Carl G. Hayes (MD-55)
Sealt™i?ffects Research Laboratory, ORD
U. o. EPA
RTF, NC 27711
Dr. Donald H. Horstman (MD-58)
ne?lt:5™ffects Research Laboratory, ORD
u. b. EPA
RTF, NC 27711
Mr. William F. Hunt (MD-14)
SfI!CEPAf Alr QUality Planning and Standards, OAR
RTF, NC 27711
Mr. Michael H. Jones (MD-12)
S!b-!CEPAf Alr QUality Planning and Standards, OAR
RTP, NC 27711
Mr. Bruce c. Jordan (MD-12)
Sfs!CEPAf Alr QUality Panning and Standards, OAR
RTF, NC 27711
Mr. Bruce Madariaga (MD-12)
U?S*°SpAf Air Quality Planning and Standards, OAR
RTF, NC 27711
Dr. William F. McDonnell (MD-58)
U?S EPA"*0*3 Research Laboratory, ORD
RTF,.NC 27711
Mr. Thomas B. McMullen (MD-52)
Environmental Criteria and Assessment Office, ORD
RTF, NC 27711
-------
viii
Dr. Edwin L. Meyer (MD-14)
Office of Air Quality Planning and Standards,
U.S. EPA
RTP, NC 27711
Dr. John J. O'Neil (MD-58)
Health Effects Research Laboratory, ORD
U.S. EPA
RTP, NC 27711
Mr. Norman C. Possiel (MD-14)
Office of Air Quality Planning and Standards,
U.S. EPA
RTP, NC 27711
Mr. James A. Raub (MD-52)
Environmental Criteria and Assessment Office ORD
U.S. EPA
RTP, NC 27711
Mr. Robert Rose (ANR-443)
Office of Policy, Planning, and Evaluation
U.S. EPA
Waterside Mall
401 M Street, SW
Washington, DC 20460
Mr. Joel Scheraga (PM-221)
Office of Policy Analysis, OPPE
U.S. EPA
Waterside Mall
401 M Street, SW
Washington, DC 20460
Mr. William P. Smith (PM-223)
Office of Stds. & Regulations, OPPE
U.S. EPA
Waterside Mall
401 M Street, SW
Washington, DC 20460
Dr. Joseph Sommers
Emission Control Technology Division
Office of Mobile Sources, OAR
Ann Arbor, MI 48105
Ms. Beverly E. Tilton (MD-52)
Environmental Criteria and Assessment Office, ORD
U.S. EPA
RTP, NC 27711
-------
IX
Dr. Dave T. Tingey
Environmental Research
Laboratory—Corvallis/ORD
200 S.W. 35th Street
Corvallis, OR 97333
-------
-------
X
Table of Contents
•
Page
Acknowledgements
National -"«*Air
11
e* Scientific Advisory Committee Subcommittee
iii
EPA Reviewers
vi
Table of Contents
x
List of Figures
xv
List of Tables
xviii
Executive Summary
xxi
I. Purpose
1-1
II. Background....
II-l
III. Approach-
III-l
IV. Ambient Ozone Concentrations in Urban and Rural Areas, iv-l
A. Urban Areas....
IV-1
B. Non-MSA Areas...
IV-2
C. Natural Ozone Background Iv_
V. Ozone Exposure Analysis
A. Overview of the Ozone NAAQS Exposure Model V-l
B. Air Quality Concentrations in Microenvironments... v-2
C. Simulation of Population Movement v_4
°* Model.ArSaS M°deled in 020ne NAAQS Exposure
V-4
E. Exercise Modeling in Ozone NAAQS
Exposure Model...
V-7
F. Eight-Area Aggregated Estimates of Population
Exposure to Alternative Ozone Standards.??"? v_8
-------
XI
Page
G. Caveats and Limitations v-13
VI. Factors Relevant to Review of the Primary Standard(s)
for Ozone ' ^ vi-1
A. Ozone Absorption and Mechanisms of Effects .. vi-l
B. Factors Affecting Susceptibility to Ozone vi-3
1 • Age VI_4
2. Sex VI_5
3. Smoking Status VI-6
4. Nutritional Status VI-7
5. Environmental Stresses VI-8
6. Exercise VI-8
C. Potentially Susceptible Groups VI-9
1. Individuals Having Preexisting Disease.". VI-9
2. Exercising Individuals VI-13
VII. Assessment of Health Effects and Related Health Issues
Considered in Selecting Primary Standard(s) for
Ozone VII-i
A. Health Effects of Concern VII-l
1. Alterations in Pulmonary Function VII-2
2. Symptomatic Effects VII-15
3. Exercise Performance VII-20
4. Bronchial Reactivity and Inflammation VII-22
5. Aggravation of Existing Respiratory Disease VII-24
6. Morphological Effects „. viI-28
7. Effects of Ozone on Host Defense Mechanisms
in Experimental Animals VII-32
8. Extrapulmonary Effects VII-35
-------
xii
Page
B. Pulmonary Function and Symptom Health Risk*
Assessment
VII-37
1. Overview of Lung Function and Symptom
Risk Assessment . _ VII-37
2. Benchmark Risk Results Vli-40
3. Headcount Risk Results T7TT .c
•••••••• VXJ.—45
4. Caveats and Limitations Vll-so
C. Related Health Effects Issues VII-52
1. Adverse Respiratory Health Effects of
Acute Ozone Exposure VII-53
2. Attenuation of Acute Pulmonary Effects VII-56
3. Relationship Between Acute and Chronic
Effects
VII-58
4. Effects, of Other Photochemical Oxidants...1. viI-62
5. Interactions with other Pollutants! VII-63
VIII. staff Conclusions and Recommendations for Ozone
Primary Standard(s) . VIII-
A. Pollutant Indicator VIIl-i
B. Form of the Standard VIH-4
C. Averaging Time(s) VIIl-5
D. Level of the Primary Standard (s) VIII-9
E. Summary of Staff Recommendations
VIII-20
Review of the Secondary
IX-l
A. Mechanisms of Action for Vegetation.
1. Biochemical Response..,
2. Physiological Response.
IX-l
B. Factors Affecting Plant Response
IX-6
IX-5
1. Biological Factors....
-------
xiii
Page
a. Plant Genetics IX-6
b. Developmental Factors ..."....!!!!!!]**** ix-7
c. Pathogen and Pest Interactions
with Ozone IX-7
2. Physical Factors IX-8
3. Chemical Factors IX-9
a. Multiple Pollutants IX_9
b. Chemical Sprays !!!!!!"" ix-li
c. Heavy Metals '.'.'.'.'.'.'.'.'.I', ix-11
X. Assessment of Welfare Effects and Related Welfare
Issues Considered in Selecting Secondary Standard(s)
for Ozone v
************ *•* •••••••»••••• x—i
A. Vegetation Effects x_2
1. Types of Exposure Effects X-2
a. Visible Foliar Injury Effects x-3
b. Growth and Yield Effects ] X-6
1. Open Top Chamber Studies x-7
.2. Greenhouse and Controlled
Environment Studies X-13
3. Ambient Air Exposure Studies...... X-14
2. Related Vegetation Issues X-20
a. Empirical Models Used to Develop
Exposure Response Relationships X-20
b. Statistics Used to Characterize
Ozone Exposures X-21
c. Exposure and Response to
Peroxyacetyl Nitrate X-23
d. Economic Assessments of Agriculture.... X-24
B. Natural Ecosystem Effects X-26
Forest Ecosystems X-
27
a. Effects on Plant Processes X-29
b. Effects on Growth .* ] * ] x-31
c. Ecosystem Responses: The San
Bernardino Study , x-38
Interrelated Ecosystems X-40
a. Aquatic Ecosystems X-40
b. Agricultural Ecosystems '.'.'.'. X-40
-------
xiv
Page
C. Materials Damage ...............
***••* ™
1 . Elastomers
2 . Textile Fibers and Dyes
x_43
X-45
4. Conclusion ............ „ fc
...................... X-45
D. Effects on Personal Comfort and Well Being ....... X-45
E. Related Welfare Effects Information and Issues... x-46
1 . Air Quality Analyses ...................... x_49
2 . Crop Loss Estimates ........... v cn
**•*•••••••••••• A~*D(J
3 . Averaging Times ........
v _ .
X-54
a. NCLAN/CERL Reanalysis
b. New Studies ......... ;
4 . Forest Risk Assessment ...................... x_66
XI-1
A. Pollutant Indicator
•••••••••«•••••»..... X-L~"1
B. Form of the Standard and Averaging Time(s) XI-3
C. Level of Standard
••••«•••••••«......... XI—10
D. Summary of Conclusions. VT ,
••••••«•«..„..»„.... XI—16
Appendix A. Air Quality
• ••••••*••».»»., A**l
Appendix B. Glossary of Pulmonary Terms and Symbols B-l
Appendix C. CASAC Closure Letter
References
-------
-------
XV
List of Figures
Page
VII-i croup Mean Decrements in 1-sec Forced Expiratory
Volume During 2-hour Ozone Exposures with Different
Levels of Intermittent Exercise .. . viI-4
VII-2 Fraction of Heavily Exercising Population
Experiencing > 10% and > 20% change in ?-sec
Levels EXplratory Volume Due to Various Ozone
s * vii-n
Vll-3 Fraction of Heavily Exercising Population
Experiencing Mild and Moderate Symptoms
Due to Various Ozone Levels VII-17
VII~4 1™°^°^*^***™^ Population Ex-
Respiratory Symptoms Due to Various
VII-18
VII-5 Benchmark Risk in St. Louis for 1-sec Forced
Expiratory Volume Decrements of > 10% and
U?!/._Undet Heavy Exercise, for~Three
Kulle, and
VII-43
VII-6 Benchmark Risk in St. Louis for Chest Discomfort
bymptoms (any and moderate/severe), under Heavy
(*v^1S*:,^°r Th5Se ExP°sure-Response Data Sets
(Avol, Kulle, and McDonnell) VII-44
VII-7 Expected Headcount (pulmonary function)
S^™??"1?!:^ ?? 9 '3 Million Dumberlth *
responding during
1 VII-47
VH-8 Expected Headcount (chest discomfort) Aggregated
for Eight U.S. Urban Areas With a Total Popula-
tion of 9.3 Million (number of heavily P
eV^T-r"! 01 net «-^^_1_ v.— — — o- -• . JT
responding during the ozone
VII-48
X-l Examples of the Effects of Ozone on the Yield of
Soybean and Wheat Cultivars v
*******•"••••*•• X~9
X-2
C-10
X-3
Across Expert
X-73
-------
xvi
Page
A-l ' Correlations Among Short- and Long-Term Air Quality
Indicators in MSAs (Using 2nd High) A-10
A-2 Correlations Among Short- and Long-Term Air Quality
Indicators in MSAs (Using ExEx) A-12
A-3 Proportion (In Percent) of Urban Sites Exceeding
Expected Number of days with an 8-Hour Daily Maximum
Average > .08 ppm for Five 1-Hour Daily Maximum
Standards A-13
A-4 Proportion (In Percent) of Urban Sites Exceeding
Expected Number of Days with an 8-Hour Daily
Maximum Average > .06 ppm for Four 1-Hour Daily
Maximum Standards t m m A-14
A-5 Proportion (In Percent) of Urban Areas Exceeding
Expected Number of Days with an 8-Hour Daily
Maximum Average > .10 ppm for Three 1-Hour
Daily Maximum Standards A-15
A-6 Generalized Relationships of the Current Ozone
NAAQS and Three Alternative 8-Hour Averages A-19
A-7 Cumulative Frequency Distribution of Three Peak
Air Quality Indicators A-28
A-8 . Correlations Among Short-Term, Multiple-Peak,
and Longer-Term Air Quality Indicators in Non-
Urban Areas A-30
A-9 Proportion (In Percent) of Rural/Remote Sites
Exceeding Specified Expected Number of 8-Hour
Daily Maximum Averages > .08 ppm for Three 1-Hour
Daily Maximum Standards A-31
A-10 Proportion (In Percent) of Rural/Remote Sites
Exceeding Specified Maximum Monthly 1-Hour Daily
Maximum Values for Three 1-Hour Daily Maximum
NAAQS A_32
A-ll Proportion (In Percent) of Rural/Remote Sites
Exceeding Specified Three Month 8-Hour Averages
Daily Maximum Three Month 8-Hour Averages For
Three 1-Hour Daily Maximum NAAQS A-33
A-12 Proportion (In Percent) of Rural/Remote Sites
Exceeding Specified Second High 1-Hour Daily
Maximum Values for Three 8-Hour Daily Maximum
Averages > . 08 ppm Standards A-34
-------
xvii
Page
Pvo- ,. ~ *—-7 / of Rural/Remote Sites
Exceeding Specified Number of Second High 1-Hour
Daily Maximum Values fnr TK>-«« «,,.,• „_ _^, ,
Daily Maximum Values for Three Maximum Monthly Mean
1-Hour Daily Maximum Standards ?. . .V? A-35
-------
-------
xviii
List of Tables
Table Title
Page
V-l Study Areas Modeled in Ozone-National Exposure
Model V_5
v~2 Estimate of the Cumulative Number of Heavy
Exercisers in the 8-Area Aggregation Population
Exposed to One-Hour Average Ozone Concentration
During the Ozone Season at Heavy Exercise Under
Alternative Air Quality Scenarios V-12
V-3 Estimate of the Cumulative Number of Person-
Occurrences of Heavy Exercise in the 8-Area
Aggregation Population Exposed to One-Hour Average
Ozone During the Ozone Season at Very Heavy
Exercise Under Alternative Air Quality
Scenarios • V-14
VI-1 Estimated Values of Oxygen Consumption and
Minute Ventilation Associated with Representative
Types of Exercise Vl-io
Vl-2 Prevalence of Chronic Respiratory Conditions by
• Sex and Age for 1979 VI-12
Vll-i Key Human Studies Near the Current 1-Hour National
Ambient Air Quality Standard for Ozone VII-7
VII-2 Morphological Effects of Ozone in Experimental
Animals • VII—29
VII-3 Effects of Ozone on Host Defense Mechanisms
in Experimental Animals VII-34
VII-4 Percent of Heavy Exercisers Responding Under
Alternative Air Quality Scenarios VII-49
VII-5 Gradation of Response for Healthy Individuals
Acutely Exposed to Ozone VII-55
IX-1 Effect of Ozone on Photosynthesis IX-4
X-l Ozone Concentrations for Short-term Exposure that
Produce 5 or 20 Percent Injury to Vegetation
Growth Under Sensitive Conditions X-4
-------
XIX
LIST OF TABLES (continued)
Htla Pace
X-2 Summary of Ozone Concentrations Predicted to
Cause 10 Percent and 30 Percent Yield Losses
and Summary of Yield Losses Predicted to Occur
at 7-hour Seasonal Mean Ozone Concentrations of
0.04 and 0.06 ppm. X-ll
X-3 Ozone Concentrations at Which Significant Yield
Losses Have Been Noted for a Variety of Plant
Species Exposed Under Various Experimental
Conditions X-15
X-4 Effects of Ambient Air in Open-Top Chambers,
Outdoor CSTR Chambers, or Growth and Yield of
Selected Crops X-17
X-5 Effects of Ozone on Crop Yield as Determined
by the Use of Chemical Protectartts X-19
x~6 Continuum of Characteristic Ecosystem
Responses to Pollutant Stress X-28
X-7 Effects of Ozone Added to Filtered Air on the
Yield of Selected Tree Crops X-34
X-8 Potential Ambient Ozone Standards that would
Limit Soybean Crop Reduction to 5, 10, 15, or
20 Percent X-52
x~9 Percentiles and Mean Predicted Relative Yield
Losses Associated with Various Levels of the Four
Exposure Indices, HDM2, M7, SUM06,-and SUM07, for
the 16 NCLAN Cases J x_59
X-10 Exposure Levels Associated with Predicted
Relative Yield Losses of 5 to 30% for the Four
Exposure Indices, HDM2, M7, SUM06, and SUM07,
for the 16 NCLAN studies X-60
X-ll
Forest Response Experts X-71
XI-1 U.S. Agricultural Welfare Benefits from Reducing -
Rural Ambient Ozone (7-hr seasonal means) to 60,
45, and 30 ppb for Three Alternative Benefit
Measures XI-14
A-l Cumulative Frequency Descriptive Statistics
Associated with Peak and Multiple-Hour Ozone
Air Quality Indicators in Urban Areas A-5
A-2 Cumulative Frequency Descriptive Statistics
Associated with Various 8-Hour Ozone
Air Quality Indicators in Urban Areas A-7
-------
XX
LIST OF TABLES (continued)
Fre2uen°y Descriptive Statistics
Associated with Longer-Term Ozone Air Quality
Indicators in Urban Areas ......... uuanty
***•
Exceedin9 the Current Ozone NAAQS
Dailv Maximum
A-21
Episodes b*
A-6 Descriptive Cumulative Frequency Statistics
Associated with Peak Ozone Air Quality
Indicators .......
A-25
A-7 Descriptive Cumulative Frequency Statistics
Indicators
A-27
-------
-------
XXI
Executive Summary
This revised staff paper evaluates and interprets the
available scientific and technical information that the EPA staff
believe is most relevant to the review of primary (health) and
secondary (welfare) national ambient air quality standards
(NAAQS) for ozone (O3) and presents staff recommendations on
alternative approaches to revising the standards. Periodic
review of the NAAQS is a process instituted to ensure the
scientific adequacy of air quality standards and is required by
section 109 of the 1977 Clean Air Act Amendments. The assessment
in this staff paper is intended to help build a bridge between
the scientific review contained in the EPA O3 criteria document
(hereafter referred to as CD) (U.S. EPA, 1986)., and the CD
Supplement (hereafter referred'to as CDS) (U.S. EPA, 1988)
prepared by the Environmental Criteria and Assessment office
(ECAO) and the judgments required of the Administrator in setting
ambient standards for O3. Therefore, the staff paper is an
important element in the standards review process and provides an
opportunity for review by the clean Air Scientific Advisory
Committee (CASAC) and the general public on proposed staff
recommendations before they are presented to the Administrator.
This staff paper has been revised based upon comments received
from CASAC and the public and upon staff analyses which are
available for public review.
-------
xxii
Ozone is a trace constituent formed in the atmosphere as a
result of a series of complex chemical reactions involving both
anthropogenic and natural hydrocarbons and nitrogen oxides,
oxygen and sunlight. At ambient concentrations often measured
during warmer months, 03 can adversely affect human health,
agricultural crops, forests, ecosystems, and materials.
Interactions of 03 with nitrogen oxides and sulfur oxides may
also contribute to the formation of acidic vapors and aerosols
which might have direct effects on human health and welfare, as
well as indirect effects following their deposition on surfaces.
It should be noted that new evidence indicates that co-exposure
to acidic aerosols can potentiate response to O,.
• «J
Annual average background surface O3 concentrations in the
northern hemisphere generally range between 0.03 and 0.05 ppm but
are as low as 0.015 to 0.020 ppm in the tropics (U.S. EPA, 1986,
p. 3-80). Stratospheric intrusion is recognized as causing
locally high 03 levels for periods lasting from minutes to hours,
but these intrusions are usually worse in spring, fall, and
winter. In contrast, during the photochemically active summer
months intrusion is less common and less severe. Summertime
hourly O3 levels have recently been reported to be as high as
0.35 ppm in one of the nation's most heavily populated
metropolitan areas. Daily daylight seasonal averages of 03 in
some rural areas have been reported to be 0.06 ppm and higher.
-------
xxiii
Primary Standard
The staff reviewed scientific and technical information on
the known and potential health effects of O3 cited in the CD and
the CDS. The information includes studies of respiratory tract
absorption and deposition of O3, studies of mechanisms of 03
toxicity, and controlled human exposure, field, epidemiological
and animal toxicology studies of effects of exposure to O3 as
well as air quality information. On the basis of this review,
the staff derives the following conclusions.
1) Inhaled O3 may pose health risks as a result of (a)
penetration of 03 into various regions of the
respiratory tract and absorption of O3 in this tract
(b) provocation of pulmonary response resulting from
chemical interactions of O3 along the respiratory
tract, and (c) extrapulmonary effects caused indirectly
by reaction of O3 in the lungs.
2) The risks of adverse effects associated with absorption
of 03 in the tracheobronchial and alveolar regions of
the respiratory tract are much greater than for
absorption in the extrathoracic region (head).
increased exercise levels are generally associated with
higher ventilation rates and increased oronasal or oral
(mouth) breathing. Greater O3 penetration and exposure
of sensitive lung tissue occurs when individuals are
heavily exercising.
-------
xx iv
3) Factors which have been demonstrated to affect
susceptibility to 03 exposure are activity level and
environmental stress (e.g., humidity, high
temperature). Those factors which either have not been
adequately tested or remain uncertain include age, sex,
preexisting disease, nutrition, and smoking status.
4) Major subgroups of the population that may be at
greater risk to the effects of 03 include: (a) any
individual exercising heavily during exposure to 03,
particularly those who are otherwise healthy
individuals who may experience significantly greater
than group mean lung function response to O3 exposure,
and (b) individuals with preexisting respiratory
disease (e.g., asthmatics and persons with allergies).
The data base identifying exercising individuals as
being at greater risk to 03 exposure is much stronger
and more quantitative than that for individuals with
preexisting respiratory disease. This is due to the
large number of clinical studies investigating effects
of O3 on exercising persons.
5) The major effects categories of concern associated with
exposures to O3 include:
(a) alterations in pulmonary function
(b) symptomatic effects (e.g., cough, throat
irritation)
(c) effects on work or athletic performance
-------
XXV
(d)
aggravation of preexisting respiratory disease
(e) morphological effects (lung structure damage)
(f) altered host defense systems (e.g., increased
susceptibility to respiratory infection)
(g) extrapulmonary effects (e.g., effects on blood
enzymes, central nervous system, liver, endocrine
system).
6) An important source of applicable exposure-response
information for a short-term standard is controlled
human exposure and field studies, which provide
concentration-response relationships between
alterations in pulmonary function and O3 exposure
concentrations, other important sources of information
for standard setting are epidemiological and
toxicological studies. Epidemiology has provided
associations between ambient O3 exposures and lung
function decrements and aggravation of existing
respiratory disease, but with greater uncertainties
about the exposures involved than with controlled human
exposure and field studies. Animal toxicology data
provide acute and chronic exposure effects information
on increased susceptibility to respiratory infection,
lung structure damage, and extrapulmonary effects.
Although human exposure, epidemiology, and animal
toxicology studies all have limitations in assessing
adverse effects and risk, it is the weight of evidence
-------
xxvi
and integration of findings from all three disciplines which
should be used in assessing health effects associated with
exposure to 03.
Based on scientific and technical reviews, CASAC comments,
and policy considerations, the staff makes the following
recommendations with respect to primary O3 standards:
1) Ozone should remain as the surrogate for controlling
ambient concentrations of photochemical oxidants.
2) The existing form of the standard should be retained
(i.e., that the NAAQS is attained when the expected
- number of days per calendar year with maximum 1-hour
average concentrations above the level of the standard
is equal to or less than one).
3) The 1-hr averaging time of the standard should be
retained.
4) The range of 1-hour average 03 levels of concern for
standard-setting purposes is 0.08 to 0.12 ppm in
concordance with CASAC comments (CASAC, 1986, 1987,
1988) comments. This range is based solely on 1-2 hour
exposure data.
5) Because there is a good health effects data base
available on 1-2 hour exposures, the staff concurs with
the CASAC conclusion (McClellan, 1989) that review of
the scientific basis for the l-hr 03 primary standard
be closed out. With this portion of the review
complete, and after considering CASAC1s views on all
-------
xxvii
issues, the Administrator will be in a position to make
a regulatory decision on how and when to best act on
the 1-hour standard.
6) In response to suggestions made by CASAC (1986, 1987,
1988), staff investigated the potential need and basis
for a longer-term (6-8 hour) primary standard.
Although an emerging data base reporting significant
lung function decrements and symptoms in subjects
exposed to O3 for 6 to 8 hours has provided some
evidence of effects below 0.12 ppm 03, staff concurs
with CASAC.s conclusion that "... sucn information
can better be considered in the next review of the
ozone standards." (McClellan, 1989). 'it is recommended .
that EPA. continue review of scientific information on
health effects of prolonged exposure to 03. Once these
studies have been more completely evaluated during the
next CD review, the Administrator will be able to
assess the need for development of a longer-term O3
primary standard.
7) Further review and analysis also will be necessary
before fully assessing the need for a separate standard
to protect against chronic effects of 03. Data on
nasopharyngeal removal, dosimetry modeling and health
effects based on and chronic exposure of animals will
be used for future animal extrapolation and risk
assessment of chronic O3 exposures.
-------
xxviii
Secondary Stands -ret
The staff has reviewed the scientific and technical
information on the known and potential welfare effects of O3
cited in the CD and the CDS. This information includes impacts
on vegetation, natural ecosystems, materials, and symptomatic
effects on humans. Based on this review, the staff derives the
following conclusions:
1) The mechanisms by which O3 may injure plants and plant
communities include (a) absorption of O3 into leaf
through stomata, followed by diffusion through the cell
wall and membrane, (b) alteratipn of cell structure and
function as well as critical plant processes, resulting
from the chemical interaction' of O3 with cellular
components, and (c) occurrence of secondary effects
including reduced photosynthesis and growth and yield
and altered carbon allocation.
2) The magnitude of the 03-induced effects depends upon
the physical and chemical environment of the plant, as
well as on various biological factors (including
genetic potential, developmental age of plant, and
interaction with plant pests).
3) The weight of the recent evidence seems to suggest that
long-term averages, such as the 7-hour seasonal mean,
may not be adequate indicators for relating O3 exposure
and plant response.
-------
xxix
4) Repeated peak concentrations are the most critical
element in determining plant response. Exposure
indicators which emphasize peak concentrations and
accumulate concentrations over time probably provide
the best biological basis for standard setting (See
staff paper, p. x-50).
5) There is currently a lack of exposure-response
information on forest tree effects. in addition, there
is a broad range of uncertainty among scientists
regarding O3 effects on forest trees. Consequently
there is no consensus on the most important averaging
time for perennials or on the precise role of O3 vs.
other pollutants in causing forest decline. Therefore/
the staff concludes that a separate secondary standard
based on protection of forest trees is not warranted at
this time.
6) There appears to be no threshold level below which
materials damage will not occur; exposure of sensitive
materials to any non-zero concentration of O3
(including natural background levels) can produce
effects if the exposure duration is sufficiently long.
However, the slight acceleration of aging processes of
-------
XXX
materials which occurs at the level of the NAAQS is not
judged to be significant or adverse. Consequently, the
staff concludes that materials data should not be used
as a basis for adequately defining an averaging time or
concentration level for the secondary standard and that
the secondary standard should be based, on protection of
vegetation.
7) Effects on personal comfort and well-being, as defined
by human symptomatic effects, have been observed in
clinical studies at O3 levels in the range of 0.12-0.16
for 1-2 hour exposures and at somewhat lower levels in
extended exposure clinical and epidemiological studies.
CASAC recommended that these effects be considered
health effects in developing a basis for the primary
standard for 03.
Based on scientific and technical reviews, CASAC comments,
and policy considerations, the staff makes the following
recommendations with respect to secondary standards:
1) In consideration of the large base of welfare
information attributing effects to O3 exposure and the
limited evidence which demonstrates welfare effects
from exposure to ambient levels of non-O3 photochemical
oxidants, there appears to be little evidence to
suggest a change in chemical designation from 03 to
photochemical oxidants.
-------
xxxi
2) Given the lack of effects data on forests and the
preliminary nature of the Lee et al. (I988c) results
regarding selection of the appropriate exposure
statistic for crops, the EPA staff concludes that it
may be premature at this point in time to change the
form of the standard and the averaging time. it is our
judgment that a 1-hr averaging time standard in the
range of 0.06-0.12 ppm represents the best staff
recommendation that could be made to the Administrator
at this time to close out the review of the scientific
data. This is consistent with CASAC comments (CASAC,
1987, 1988) urging EPA to consider a l-hr averaging
time and to act on the existing state of science rather
than extend the review until a more exhaustive
assessment is made of alternative averaging times.
With this portion of the review complete, and after
considering CASAC's views on all issues, the
Administrator will be in a position to make a
regulatory decision on how and when to best act on the
1-hr standard.
Alternatively, EPA could continue the standard review until
the information on alternative exposure indicators has matured.
Additional time for review and revision of Lee et al. (M88c)
would allow the scientific community the opportunity to review
the alternative indicators and move toward a consensus regarding
selection of the most appropriate exposure indicator. The
-------
xxxii
liability of this alternative is that it postpones action on the
secondary standard and thus fails to utilize new and existing
information to assess the most appropriate exposure statistic or
the protection afforded by the current l-hr standard.
-------
1-1
Review of the National Ambient Air Quality Standards for Ozone
Assessment of Scientific and Technical Information
Staff Paper
I. Purpose
The purpose of this staff paper is to evaluate and interpret
scientific information contained in the CD and the CDS and to
identify critical elements to be considered in selecting primary
(health based) and secondary (welfare based) national ambient air
quality standards (NAAQS) for ozone (03). Staff conclusions and
recommendations will -integrate critical elements of the review of
standards with other factors such as averaging times, form of
standards, and margin of safety considerations.
-------
-------
II-l
II • Background
Since 1970 the Clean Air Act as amended has provided
authority and guidance for the listing of certain ambient air
pollutants which may endanger public health or welfare and the
setting and revising of NAAQS for those pollutants. Primary
standards must be based on health effects criteria and provide an
adeguate margin of safety to ensure protection of public health.
As several judicial decisions have made clear, the economic and
technological feasibility of attaining primary standards are not
to be considered in setting them, although such factors may be
considered in the development of state plans to implement the
standards (Lead Industries Association v. EPA, 1980; American
Petroleum Institute v. EPA, 1981). Further guidance provided in
the legislative history of the Act indicates that the standards
should be set at "the maximum permissible ambient air level .
which will protect the health of any (sensitive) group of the
population" (U.S. Senate, 1974). Also, margins of safety are to
be provided such that the standards will afford "a reasonable •
degree of protection . . . against hazards which research has not
yet identified" (U.S. Senate, 1974). In the final analysis, the
EPA Administrator must make a policy decision'in setting the
primary standard based on a judgment regarding the implications
of all the health effects evidence and the requirement that the
standards provide an adequate margin of safety.
Secondary ambient air quality standards must be based on
welfare effects and must be adequate to protect the public
welfare from any known or anticipated adverse effects associated
with the presence of a listed ambient air pollutant. Welfare
effects, which are defined in section 302(h) of the Act, include
effects on vegetation, visibility, water, crops, man-made
materials, animals, economic values and personal comfort and
well-being.
On April 30, 1971, the Environmental Protection Agency (EPA)
published in the Federal Register (36 FR 8186) primary and
-------
II-2
secondary national ambient air quality standards (NAAQS) for
photochemical oxidants. Both standards were set at an hourly
average level of 0.08 ppm not to be exceeded more than once per
year.
In accordance with the provisions of sections 108 and 109 of
the Clean Air Act as amended, EPA reviewed and revised the
criteria upon which the original photochemical oxidants NAAQS
were based. On February 8, 1979, revised standards were
published (44 FR 8202) with the following changes: (1) changing
the primary and secondary standards to 0.12 ppm, (2) changing the
chemical designation of the standards from photochemical oxidants
to ozone (03), (3) changing to a daily maximum standard rather
than an "all-hours" standard, and (4) changing the definition of
the point at which the primary and secondary standards are
attained to "when the expected number of days per calendar year
with maximum hourly average concentrations above 0.12 ppm is
equal to or less than one."
Several factors were cited as a basis for reivising the
primary standard for O3 in 1979. Among these were: (1) an
adverse health effect threshold for oxidants could not be
identified with certainty, (2) 03 is a pulmonary irritant that
affects respiratory mucous membranes as well as other lung tissue
and impairs respiratory function at levels as low as 0.15 ppm for
exercising persons (Delucia and Adams, 1977), (3) evidence
suggests an elevated number of asthma attacks when peak hourly
oxidant concentrations reach about 0.25 ppm (Schoettlin and
Landau, 1961), (4) several studies show increased susceptibility
to bacterial infection in laboratory animals exposed to 03 plus a
bacterial challenge, (5) premature aging symptoms reported in
animals and (6) apparent synergistic effects on pulmonary
function from exposure to 0.37 ppm O3 and 0.37 ppm sulfur
dioxide. EPA concluded that a primary standard of 0.12 ppm (one-
hour average) not to be exceeded more than one day per year on
-------
II-3
average would protect public health with an adequate margin of
safety.
EPA based its decision on the secondary standard on the
limited information that was available on growth and yield
reduction in commercially important crops and indigenous
vegetation exposed to O3 under field conditions. These studies
indicated that growth and yield responses were related to long-
term (growing season) exposure of plants. On the basis of this
information and available air quality data, EPA concluded that
there was no evidence indicating that a significant decrease in
growth and yield would result from long-term average
concentrations expected to occur when the primary standard was
attained. Consequently, a secondary standard more stringent than
the primary standard was deemed unnecessary on the basis of
03-related yield reduction effects on vegetation.
-------
-------
III-l
III. Approach
This is the final staff paper provided during current review
of the NAAQS for O3; judgments contained herein are based on
scientific evidence reviewed in the CD and in the CDS prepared by
the ECAO. This staff paper incorporates the results of a health
risk assessment and includes the results of an exposure analysis
for alternative NAAQS.
"Critical elements" have been identified which the staff
believe should be considered in a review of the primary and
secondary standards. Particular attention is drawn to those
judgments that must be based on careful interpretation of
incomplete or uncertain evidence. In such instances, the staff
paper states the staff's evaluation of evidence as it relates to
a specific judgment, sets forth alternatives that the staff
believe should be considered, and recommends a course of action.
To facilitate review, this paper is organized into sections as
outlined below.
Section IV provides an overview of ambient levels of O3
currently being experienced in various portions of the United
States. This section is intended to set the stage for the
remaining discussion by identifying the present air quality
situation so the reader can relate available health and welfare
information to O3 levels occurring in the real world.
Section V summarizes results of the 03 exposure analysis.
The NAAQS Exposure Model (NEM) was used to estimate nationwide
human exposure to O3 given attainment of alternative standards.
Section VI deals with elements related to the health effects
evidence examined in reaching conclusions regarding the primary
standards; these include the following:
• most probable mechanisms of toxicity by which health
effects occur,
-------
III-2
• discussion of factors potentially affecting
susceptibility to 03 exposure,
• description of the most sensitive population groups and
estimates of the size of those groups.
Section VII is a preliminary assessment of health effects
and related health issues. The section:
• identifies health effects which have been attributed to
O3 and other photochemical oxidant exposures,
• discusses health effects evidence used to develop staff
judgments concerning which effects are most important
for the Administrator to consider in reviewing and
setting primary standard(s),
• describes the health risk assessment for acute O3
exposures,
• discusses issues related to health effects attributed
to O3 and other photochemical oxidants.
Drawing on discussions in Sections IV through VII, Section
VIII identifies and assesses factors that the staff believes
should be considered in selecting a pollutant indicator,
averaging time, form, and level of primary standards, including
margin of safety considerations. Staff recommendations also are
presented in this section.
In Section IX, the effects of O3 and other photochemical
oxidants on vegetation, personal comfort, natural ecosystems, and
man-made materials are identified. The section:
• describes physiological and biochemical alterations
associated with welfare effects which result from
exposure to 03 and other photochemical oxidants,
• identifies welfare effects of O3 and other
photochemical oxidants,
• discusses factors affecting plant response.
Section X is a discussion of welfare effects to be
considered in selecting secondary standard(s). The section:
-------
III-3
• describes the existing scientific evidence on welfare
effects attributed to O3 and other photochemical
oxidant exposures,
• describes new studies and analyses related to the issue
of averaging times, and
identifies and evaluates scientific uncertainties
related to welfare effects evidence in addition to
staff judgments concerning which welfare effects are
important for the Administrator to consider in
reviewing and setting secondary standard(s).
Drawing on discussions in Sections IX and X, Section XI
identifies and assesses the factors that the staff believes
should be considered in selecting averaging time(s), level(s) and
form of secondary standards, staff recommendations also are
presented in this section.
-------
-------
IV- 1
IV.
and Rui
This Section provides a summary of ambient O3 air quality in
urban and non-urban areas. More information on the topic appears
in Appendix A. The data base used generally is 1985-1987 Storage
and Retrieval of Aerometric Data (SAROAD) air quality data,* but
data are presented for earlier time periods if 1985-1987 data are
not available. Urban areas are interpreted to be Metropolitan
Statistical Areas (MSAs, as defined by the U.S. Bureau of Census
T*as section a^o Briefly discusses the concept of "natural
03 background, - and provides a general estimate of its ambient
concentrations for different averaging times at ground-level.
A. Urban Areas
There are 331 MSAs in the 50 states. EPA staff in the
Ambient standards Branch has identified 224 MSAs (68%) as having
enough O3 air quality data to ascertain whether or not they
exceed the NAAQS. of these, 101 areas (45%, have more than one
expected exceedance per year of the "current 03 standard of o „
ppm. Thus, slightly less than then one-half of MSAs with
sufficient data exceed the standard. Approximately loo minion
people less than one-half of the total U.S. population, Uve in
these 101 metropolitan areas. (However, this does not .ean that
everyone ln these areas is exposed to O3 concentrations at or
above the standard. See Section V.)
About 10% of the MSAs with sufficient data have a
characteristic highest concentration (cue)' of 0.16 ppm o3 or
than
using more recent dat~fo
result in more exceedances and
here (and in Appendix *7 IS
incorporate these new data as
labelsTp?aced on
infrequently. o»
t™ coun*;1T Thus,
• period— would
^ " tha" resented
specific
-------
IV-2
higher; over 5% (11) of MSAs have a CHC above 0.18 ppm 03. There
is no clear temporal trend in 03 CHCs in most MSAs around the
country, although the 1984-1985 time period generally had lower
levels than previous years and 1987-1988 had higher levels. The
trend in expected exceedances in MSAs likewise is indefinite.
Over the 1980s, expected exceedances were relatively high in
1980, dropped in 1981-1982, and jumped radically in 1983.
Expected exceedances for 1984 and 1985 were like those in 1981-
1982, but increased again in 1987 and 1988.
Generally, maximum monthly mean concentrations for 1- and 8-
hour daily maximum averages are in the range of 0.050-0.085 ppm.
The maximum 3-month mean of 8-hour daily maximum averages is in
the range of 0.045-0.065 in most urban areas of-the country.
There are statistically significant relationships among peak and
longer-term mean indices of O3 air quality in urban areas. These
relationships can be used to estimate the impact that alternative
short- term peak O3 standards will have on long-term average
levels of O3.
•
B. Non-MSA Areas
Air quality data indicate that non-MSAs have a lower CHC
than do MSA areas and that the CHC often is associated with O3
transported into non-MSA areas after 4 pm. On average, a daily
represented by the second-highest 1-hour daily maximum O,
concentration monitored during an area's O3 season (generally
April through October). Another CHC is the characteristic
largest daily maximum (CLDM), usually represented by a
concentration from a mathematical distribution fitted to actual
air quality data. For 03 air quality data, a distribution often
used is the two-parameter Weibull. Generally, the CLDM is the
1/n value from the Weibull, where n = 03 season length (in days).
All three labels (CHC, design value, and CLDM) are used in
this report, especially in Appendix A. The CHC mentioned above
in the text is the "design value" version. See also the footnote
on p. V-4.
-------
IV-3
maximum 8-hour average of 0.08 ppm is exceeded on an average of
11 days during the O3 season in non-MSA areas. This
"exceedance rate" drops to 1 for a 0. 10 Ppm "outpoint '•
concentration and slightly less than 1 for a 0.12 ppm cutpoint
concentration.
Longer-term daily maximum O3 averages in non-MSAs usually
are lower than those seen in MSA areas. The maximum monthly
means for l- and 8-hour daily maximum averages generally are
0.067 ppm and 0.058 ppm, respectively. Means as high as o 118
pprn and 0.092 ppm for these two averaging times have been seen in
non-urban areas, however. High monthly means for either
averaging time occur most often in May, June, and July.
Longer-term (> month) air quality indicators are more
closely associated with each other than with short-term
indicators, in fact, correlations among long-term air quality
indicators are higher than correlations among short-term
indicators. This is to be expected, since long averaging time
statistical measures are relatively less variable than short
averaging time measures. (Figure A-8 in Appendix A provides
estimates of correlations among selected air quality indicators
in non-urban areas . )
Relationships can be developed among air quality indicators
xn non-MSA areas that can be used to estimate the impact that
attaining a standard for one averaging time will have on other
- these are presented in
C. Natural Ozone Background
Ozone is a trace constituent of the atmosphere. There is
controversy regarding ho* much of ambient 03 Monitored at ground-
level 1S natural and how much is produced from man-made
precursors. Estimates of the natural component of 03 vary widely
« the literature, and there is no standardized
-------
IV-4
terminology regarding the concept of natural O3 background.3
Even when a numerical estimate of background (however labeled) is
provided, rarely is the averaging time provided for the estimate.
Based on a thorough review of this literature, it is obvious
that natural 03 background .is a multidimensional and complex
concept. Natural O3 background concentrations vary by geographic
location, altitude and season.
A working definition of natural O3 background is:
n*-n,r-F?rna Par^iculaJ geographic area and averaging time,
natural 03 background is that constituent of "background o,»
(0-j that cannot be affected by manipulating anthropogenic
emissions in an area) which arises solely from
photochemically-reacted biogenic precursors and from
stratospheric O-j transported downward into the area (either
directly or indirectly).
Note that this definition is predicated upon the concept
that natural background O3 cannot be affected by manipulating
anthropogenic emission sources. Note also that the definition
does not include some constituents of natural background that are
considered by many to be background 03. Excluded constituents •
are (I) trapped anthropogenic-based O3 downwashed into an area
due to breakup of a morning inversion, (2) nocturnal O3 maxima
due to downward mixing of O3-rich air from above the inversion
layer, (3) transported O3 in an urban or stationary source plume,
and (4) anthropogenic-based O3 that is formed and/or stored in
the troposphere and subsequently downwashed.
u i. ^act' a survey of the available literature that mentions
background 03 or natural O3 background-approximately 50 mentl°ns
articles—did not uncover a single rigorous definition of either
term! Even the appellations used for the concepts vary greatly
in the relevant literature. Examples include: "baselinJ 0, "
"clean air background," "global background," "North American'
background," "Urban background," and "regional surface
background." In addition, twelve other labels were used—all
without being defined.
-------
IV-5
A reasonable estimate of the natural O3 background
concentration near sea-level in the U.S. for an annual average is
0.020-0.035 ppm. This includes a 0.010_0.015 ppm contribution
(averaged over time) from the stratosphere and a 0.01 ppm
contribution from photochemically-affected biogenic non-methane
hydrocarbons, m addition, another o.oi ppm is possible from the
photochemical reaction of biogenic methane.
A reasonable estimate of natural 03 background concentration
for a l-hour daily maximum at sea-level in the U.S. during the
summer is on the order of 0.03-0.05 ppm.
These estimates are synthesized from the available
literature, but rely most heavily on Altshuller (1986) and Kelly
et al. (1984) . *
-------
-------
V-l
v- Ozone EXPOSUT-O
Analysis of population exposure under alternative NAAQSs
requires that significant factors contributing to total human
exposure be taken into account. These factors include the
temporal and spatial distribution of people and 03 concentrations
throughout an urban area as people go through their typical daily
pattern of life. To the maximum extent possible with existing
data, this has been done in the NAAQS Exposure Model (HEM) a
simulation model designed to estimate human exposure in selected
urbanized areas under user- specified regulatory scenarios
(Biller et al., 1981). This chapter is a summary of information
provided in Paul et .al. (1986) , which has been used in
development of. the risk assessment described in chapter VII.
A. overview of the Ozone NAAQS Exposure Model
The 03 HEM model partitions all land within a selected urban
area into large "exposure districts" (Paul et al., 1986) . There
are between three and fifteen exposure districts identified in
the ten urban areas used in the 03 NEM analysis. The number of
districts identified is directly related to the number of
monitors having valid air quality data in a study area
People living within each exposure district, as estimated by
the U.S. census in 1980, are assigned to a single discrete point
the population centroid. The air guality !evel uithin each
exposure district is represented by air quality at the population
oentroid, which is estimated for each hour of the year l^^
monitoring data from nearby monitoring sites. Because pollutants
,n the ambient air are generally modified considerably when
''"
aluste -
adjusted to account for five different microenvironments :
indoors at home, indoors other (all other indoor locations)
inside a transportation vehicle, outdoors near a roadway and all
other outdoor locations. The air quality adjustments ar le by
us ng microenvironmental transformation factors, as explained
-------
V-2
Because degree of exposure and/or susceptibility to effects
of pollution may vary with age, occupation, and intensity of
exercise, the total population of each study area is divided into
age-occupation (A-O) groups. Each A-o group is further
subdivided into three or more subgroups. A typical pattern of
activity through the five inicroenvironments is established for
each subgroup and an exercise level (high, medium, or low) for
each is also specified.
Units of population analyzed by NEM are called cohorts.
Each cohort is identified by exposure district of residence, by
exposure district of employment, or school attendance, by A-O
group, and by activity-pattern subgroup. During each ten minute
period of the day, each cohort is assigned to a particular
exposure district and a particular microenvironment. The
assignment is based upon (1) data regarding human activity
patterns gathered by EPA and university "time budget"
researchers, and (2) home-to-work transportation data gathered by
the U.S. Census Bureau. Using these sources of information, NEM
simulates ten minute movements of cohorts through different
districts of the urban area and through different
microenvironments, combines the movements with hourly averaged
air quality data, and accumulates the resulting exposures over
the "O3 season."1
B. Air Quality Concentrations in Microenvironments
NEM requires that hourly air quality estimates be available
for each microenvironment that every cohort passes through. The
estimates are obtained by using a simple linear function relating
outdoor air quality concentrations to microenvironment levels via
a "transformation" relating outdoor-to-indoor levels and then
The 03 season is that part of the year for which o,
monitoring is undertaken as required by EPA regulations for
implementation of the O3 NAAQS (40 CFR 58, Appendix D). For
three urban areas that were modeled it is the entire year- for
the remaining 7 areas, it is April through October.
-------
V-3
adding to thxs the pollutant concentration due to sources located
in the mxcroenvironment itself. For 03/ the transformation is
essentially a multiplicative ratio derived from a review of the
03 exposure literature (Ferdo, 1985) .
The relationship is:
where Vt = air quality in microenvironment m during
hour t
am,t - hourly-averaged pollutant concentration'
due to sources located in the
microenvironment
bm * multiplicative ratio of the
microenvironment concentration value to the
monitored air quality value
xt ^ - monitor-derived air quality value for time
Because no significant sources of O3 were identified for any of
the mxcroenvironments, am,t = o for all environments. Estimated
al ^1985 °btalned fr°m the UteratUre are Ascribed in. Paul et
The xt values are actual monitored values for the current
(or "as is") situation. These monitored values are adjusted
n th *- ^l
in the study area Dust meets the 03 NAAQS being analyzed. By
Is tLT^3 NAAQVS attained when a11 monitors in -
less than one expected exceedance of the standard concentration
value (currently it is 0.12 Ppm) in a year. The
analyse ls based on a "just attains" scenario, where air quality
evels at the monitor currently having the highest number oT *
expected exceedances are reduced mathematically to where that
"
ta a ot ' «»
data at other monxtors in the study area are adjusted using a
non-lmear approach described in Paul et al. (1986) .
-------
V-4
C. Simulation of Population Movement
Population movement in NEM is based upon information
gathered by the U.S. Census Bureau regarding householders' home-
work commuting patterns (Bureau of the Census, 1982). The
information includes MSA- specific data on the census tract
level, which itself is based upon actual location information
regarding the sampled population's home and workplace. This
census tract information is aggregated for exposure districts
used in the NEM analysis to obtain district-to-district trip
information for those cohorts that work. (Other-wise, cohorts are
assumed to stay in their home districts.) Because of lack of
travel data for non-work activities, we assume that all shopping
is done in each cohort's residential district. The same
assumption also is made regarding school-related activities.
Three one-way commuting times are used to represent non-
household-worker commute times: 20, 30, and 40 minutes. Most
. workers fall into the 20 minute commute time (representing the
rather large interval of 0 to 24 minutes), since the average
commuting time in the United States is about. 20 minutes.
Housewives/househusbands are assumed to have no commuting time.
The number of different cohorts -explicitly modeled in an
area is equal to the product of 54 cohort groups times the square
of the number of districts identified in each study area. Thus,
the number of cohort groups explicitly modeled varies between 486
and 12,150 in the ten urban area sample. (See Table V-l.)
There is no trip information available from the U.S. Census
Bureau regarding weekend inter-district travel in SMSAs. (There
is information available regarding weekend recreational travel,
but it is not locationally specific nor is it systematic.)
Consequently, all cohorts are assigned to their home districts on
weekends in NEM analysis.
D. Study Areas Modeled in Ozone National Exposure Model
As mentioned, ten urban areas are used to model O3 exposures
explicitly. Results from these ten areas will be extrapolated to
-------
Table V-l
STUDY AREAS MODELED IN OZONE-NEM
Study
Area
Name
•"•' • i - —
Chicago
Denver
Houston
Los
Angeles
Miami
New York
Philadelphia
St. Louis
Tacoma
Washington,
DC
MSAs Included Study Area
(in whole or Population
ln part) Modeled (ii
Aurora-Elgin, IL
Chicago, IL
Gary. IN
Joliet, IL
Lake County. I L
Boulder, CO
Denver, CO
Houston, TX
Anaheim, CA
Los Angeles, CA
Riverside, CA
San Bernardino, CA
Fort Lauderdale. FL
Miami, FL
Middlesex, NJ
Nassau-Sulfolk, NY
Newark. NJ
New York, NY
Stamford. CT
Philadelphia. PA
St. Louis. MO
Tacoroa. WA
Washington, OC
7.48
1.54
2.54
10.22
1.55
13.62
4.36
2.20
0.58
2.88
Total
Population
of Included
JtSAs
1 -' —•—..
7.82
1.62
2.74
10.97
2.65
13.85
4.72
2.38
0.46
3.25
2nd Highest
Daily Maximum
03 Concentration
0.17
0.14
0.19
0.37
0.12
0.25
0.20
0.18
0.14
0.14
Number
of
Exposure
Districts
9
6
7
15
4
10
7
8
3
8
Number
of EPA
Cohorts Region
Modeled Number^
4.374 5
1.944 a
2.646 6
12.150 g
864 4
5.400 it2
2.646 3
3.456 5.7
486 10
3.456 3
r
-------
V-6
the nation's urban population as a whole to obtain aggregated
national exposure estimates. The study areas and their base
population, along with other pertinent area information, appear
in Table v-1.
The ten areas vary greatly in geographical location, 03
"design value",2 population size (both modeled and total MSA),
and number of exposure districts included. The areas were
selected to obtain as widely representative a modeling data base
as possible given the overall need for monitoring data
completeness in an area.
For instance, study area populations modeled vary from
Tacoma, with a population of almost 580,000, to New York, with a
population of about 13.6 million people. Design value
concentrations vary from the current standard level of 0.12 ppm
to 0.37 ppm—the highest design value included in EPA's
regulatory data base within the last five years.
The number of exposure districts in study areas varies
between 3—the minimum thought necessary to adequately capture
the variation in 03 air quality for an,area—and 15. The number
of exposure districts used in an area directly translates into
the number of cohort groups in the area that are tracked through
time and space in 03-NEM. Each one of these cohort groups
experiences a different pattern of air quality as its weekday
activities are simulated. Thus, in Los Angeles over 12,000
different patterns were modeled.
One limitation of the exposure assessment is that it does
not provide good coverage of the New England area (only Stamford,
CT is included). This is a limitation worthy of note since the
«
A design value is that measured air quality concentration
value in a MSA that must be reduced to the O3 standard level to
ensure that the area meets the current 03 NAAQS formulation of <1
expected exceedances of 0.12 ppm daily maximum 1-hour average.
The value shown in Table V-1 is the second-highest 1-hour daily
maximum concentration in the O3 air quality data base for the
year modeled. See the footnote on p. IV-2 for additional
information regarding design values (and other CHCs).
-------
V-7
CT is included). This is a limitation worthy of note since the
New England area has high population densities and high 03
levels. The staff was unable to adequately model the New England
area because of (1) the adequacy of monitoring data suited for
exposure assessment and (2) U.S. Census Bureau data—both
population and transportation—are difficult to use in New
England due to the political/spatial classification system used
there. No other MSA in New England could be explicitly modeled
for these two reasons.
E. Exercise Modeling in Ozone NAAQS Exposure Modeling
Because dose received by a person exposed to an air
pollutant is highly dependent upon her or his ventilation rate,
exercise level is an important consideration in exposure
modeling. In O3-NEM, four exercise levels are considered. They
are listed here, along with their associated ventilation rates
(in units of liters per minute, or 1/min):
1". low exercise, 25 L/min or less
2. medium exercise, 26-43 L/min
3. heavy exercise, 44-63 L/min
4. very heavy exercise, 64 1/min or higher
Two broad cohort group distinctions are made in 03-NEM with
respect to exercise: exercisers and non-exercisers. Exercisers
are further divided into those who exercise heavily and those who
exercise very heavily. Exercisers are those cohorts who
participate in any activity that requires a ventilation rate
greater than 43 L/min—i.e., the last two exercise categories.
Heavy exercisers participate in an activity level requiring more
than 63 L/min of ventilation.
Exercise level designations used in NEM are roughly adjusted
for age and body size. For example, a baby's "thrashing around"
is heavy exercise for the 0 to 6 month old cohort, but the same
type of activity in an adult cohort is low exercise. Exercising
cohorts are emphasized in 03-NEM because exercisers, per se, may
be "sensitive" (react more) to 03 exposure. In fact, 35 of the
-------
V-8
54 cohort groups used in NEM are exercisers. In most of the ten
study areas modeled, however, these 35 exercising cohorts
constitute less than 40% of a MSA's population. Twenty cohort
groups (included in the 35) participate in very heavy exercise,
but contain at maximum approximately 7% of any MSA's population.
Thus, both exercising and cohorts are over-sampled in O3-NEM.
The amount of time spent in exercise for any cohort is quite
small. The largest portion of any cohort group undertaking heavy
exercise is 6.2% for cohort #3, which consists of children 1 to 2
year old. (This cohort constitutes only about 1.0-1.5% of a
MSA's total population.) The proportion of time spent in heavy
exercise by people in cohorts that undertake such exercise is
less than 2%, on average.
The largest portion of any cohort group undertaking very
heavy exercise is 100% for cohorts #52-54, consisting of male
outdoor workers. Less than 2% of any study area's population is
included in this occupational grouping. The usual proportion of
exercising cohort population participating in very heavy exercise
is less than 1%.
The combination of small exercising fractions and small
amounts of time spent in heavy or very heavy exercise results in
a very small fraction of total population-time devoted to
exercise. For one modeled study area, New York, the amount of
total population-time spent in exercise is:
1. 0.9% for heavy exercise (undertaking exercise at a
level of 44-63 L/min).
2. 0.002% for very heavy exercise (undertaking exercise at
a level of 64 or higher L/min).
The other modeled MSAs have similar, but not identical, fractions
of total population-time spent in exercise.
F. Eight-Area Aggregated Estimates of Population Exposure to
Alternative Ozone Standards
Although ten urban areas were modeled using O3-NEM, only
eight areas were modeled for the "headcount" risk assessment that
-------
V-9
x. described in Section VII-B.3 The total population included
in the aggregation is 25.9 million people, and the total number
of one-hour exposure occurrences for this population is 125
billion person-hours annually.
Since the lung function/symptom health risk assessment
currently focuses on heavily exercising individuals as a
"sensitive population" with respect to O3 exposure, we present
exposure estimates for only that group. Thus, the exposure
estimates that follow are for people who are exercising at a
ventilation rate of between 44 and 63 L/min. Because only a
small portion of total population- time is spent in this exercise
category, the number of people in the aggregation drops to 9 3
million and the one-hour exposure occurrences drops to 2.6
billion person-hours. These are the "base numbers" for the
exposure estimates contained later in this chapter.
In any NEM analysis, three different indicators are used to
estimate exposure of people to various levels of air pollution
One unit is "occurrences of exposure:" the number of times a
given level of pollution is experienced by one individual. if 30
people experience a pollutant level of l ppm which remains steady
over a 3-hour period, population exposure can be expressed as 30
occurrences of exposure for a 3-hour averaging time or 90
occurrences for a 1-hour averaging time. A second indicator of
population exposure is "people-exposed." This is simply the
number of people who experience a given level of air pollution
or higher, at least one time during the time period of analysis.
The third indicator of exposure used in NEM is "people at peak
exposure." This is the number of people who experience their
-------
V-10
highest pollutant level within a given concentration interval.
Examples of the first two output measures will be provided in
this Section.
Four alternative air quality scenarios are modeled. The
first is the "as is" situation, which uses recent monitored air
quality data to represent O3 concentrations in the outdoor-other
microenvironment. (This concentration is then modified by the
microenvironmental factor to estimate 03 levels in the other
microenvironments.) Air quality for the "as is" case comes from
the 1983-1985 time period, using whichever year has the most
complete data base.
The remaining air quality scenarios represent the
hypothetical situation when a national ambient air quality
standard for 03 is just attained in an urban area. This
situation is simulated by adjusting current air quality data so
that the worst monitor in the urban area has a CHC4 equal to the
NAAQS concentration level. The adjustment procedure is complex
and nonlinear. (For instance, peak hourly concentrations are
adjusted more-absolutely and relatively—than those near the mean
of the "as is" distribution.) It utilizes regression analyses of
parameters of the Weibull distribution fit to each valid monitor
in the urban area. For more information regarding the air
quality adjustment procedure used to simulate a just-attaining
situation see Paul et al. (1986).
The CHC in this case is the characteristic largest dailv
maximum (CLDM) version. The worst monitor in most MSAs is
located in a downwind suburb of a central city. Most of the
people in the MSA never experience the high O3 levels seen at
that monitor, since they may not travel into that exposure
district as they go about their daily activities. Even people
living in that district do not experience the high concentrations
seen at the monitor, because (l) high O3 values often occur in
the mid- to late-afternoon when many district residents are at
work in other less polluted districts, and (2) most people spend
most of their time indoors, and do not directly experience hiah
outdoor 03 concentrations. When the worst monitor in an area
attains an alternative O3 NAAQS, all other monitors experience
better air quality bv definition. v
-------
V-il
Three alternative O3 NAAQS are simulated: 0.08, o.io and
0.12 ppm. The latter value, of course, is the current standard
level of the O3 NAAQS. It should be recognized that we are not
concerned in our exposure analyses about how or when an
alternative 03 NAAQS is attained. That is the concern of other
analyses which OAQPS and other EPA offices undertake: especially
the regulatory and benefits analyses. For O3 exposure analyses
purposes, it is sufficient to simulate the just-attaining
situation without being concerned about how, when, or even if
that situation will occur.
Estimates appear in Table V-2 of the cumulative numbers of
people (in millions) who experience increasing l-hour O3
exposures as they engage in heavy exercise. These estimates are
based upon using "best estimate" microenvironment factors The
first column shows the 03 concentration that is equaled or
exceeded, and the other four columns show cumulative exposure
distributions for the four air quality scenarios previously
discussed. Because all people experience the same number of
exposures at or above o.O ppm, the entire aggregated population
base of 9.3 million exercise is shown for all scenarios.
The overall pattern of data shown in Table V-2 is cordon to
most of the exposure results that follow. Basically, that
pattern is:
1. there is a long "tail" to the current situation
scenario, in that there are tabular entries for the
higher O3 concentration "cutpoints," even up to 0.34
ppm.
2. There is a short tail to the remaining scenario
distributions, with very small tabular entries at or
above the O3 concentration used for the scenario
standard.
What this pattern implies is that attainment of any alternative
NAAQS investigated results in a dranatic reduction in peak 0,
eZsTr ,F°VnStanCe' the -«»*• «* th. number of persons
exposed dunng heavy exercise to 03 levels 40.12 pp. drops to 0 1
-------
V-12
Table V-2
iS2X5nf BX!!, ^H!^II!(!._NUMBER OF HEAVY EXERCISERS
(millions of people)
03 Cone.
Equaled
or Exceeded
(ppm)
0.361
0.341
0.321
0.301
0.281
0.261
0.241
0.221
0.201
0.181
0.161
0.141
0.121
0.101
0.081
0.061
0.041 -
0.021
0.001
0.000
• *ii \4
The
"As Is"
Situation
n
\J
*
*
0.1
0.1
0.3
0.5
0.7
1.3
1.6
2.6
3.3
4.0
5.2
6.9
8.6
9.1
9.3
9.3
9.3
ua i ( ujr oueridrios
Alternative
0.12
0
0
0
0
0
0
0
0
0
0.1
0 3
NAAQS
0.10
0
0
0
0
0
0
0
0
0
0
0
0
0
+
w * *J
2.2 0.9
6.5 4
8.7 7
9.3 9
9.3 9
9.3 9
.9
.9
.3
.3
.3
(in ppm)
0.08
0
0
0
0
0
0
0
0
0
0
0
0
0
0
*
2.0
6.9
9.1
9.2
9.3
—— — i^— _ ___ _
*Fewer than 50,000 people.
-------
V-13
million with attainment of a 0.12 ppm NAAQS, as compared to 4.0
million people under "as is" O3 air quality conditions. Attaining
a tighter 03 NAAQS/ either 0.10 or 0.08 Ppin/ reduces the number
of people- occurrences >0.12 ppm to zero.
A different type of exposure statistic is depicted in Table
V-3. The subject of the Table is the estimated number of people-
occurrences of exposure to various O3 levels as they undertake
heavy exercise. The universe for this aggregation is 2.6 billion
person-occurrences. AS can be seen, the number of heavy
exercising person-occurrences of exposure >.12 ppm drops
radically from the "as is" situation with attainment of one of
the alternative standards, it should be realized, however, that
the relative proportion of heavy exercising person-occurrences
estimated for O3 exposure levels at or above 0.12 ppm is fairly
small, even for the "as is" situation. For example, it is 1.7%
of total heavy exercising person occurrences at >.i2 ppm and is
0.2% at >o.20 ppm. With attainment of' a 0.12 ppm 03 NAAQS in all
areas, the proportion of heavy exercising person occurrences of
03 exposure above that concentration would only be 0.003%.
G. Caveats and Limitations
A number of caveats must be acknowledged concerning the o -
NEM results. These must be recognized by the reader in her or
his interpretation of exposure estimates. Probably the most
important caveat is that there is considerable uncertainty
concerning a number of important inputs to 03-NEM, especially
those regarding human activity and exercise patterns
Uncertainty also exists regarding (!) predicted O3 concentrations
upon attainment of alternative O3 NAAQS, (2) microenvironment
factors relating outdoor-to-indoor concentrations, and (3)
extrapolation of urban area-specific exposure estimates to the
nation as a whole.
Some of these uncertainties are addressed by use of the
"low, "best,', and "high" estimates, but not all sources of
-------
Table V-14
Estimate of the Cumulative number of Person-Occurrences
of Heavy Exercise in the 8-Area Aggregation Population Exposed
to One-Hour Average Ozone During the Ozone Season at Verv
Heavy Exercise Under Alternative Air Quality Scenarios
(millions of person-occurrences)
O3 Cone
Air Quality Scenario
Equaled
or Exceeded
(ppm)
0.361
0.341
0.321
0.301
0.281
0.261
0.241
0.221
0.201
0.181
0.161
0.141
0.121
0.101
0.081
0.061
0.041
0.021
0.001
0.000
The
"As Is"
Situation
0
*
*
0.1
0.3
0.6
1.4
2.7
4.3
6.8
14.4
24.9.
45.2
67.4
160.9
343.5
758.0
1,622.4
2,348.3
2,610.3
Alternative NAAQS
0.12 0.10
0
0
0
0
0
0
0
0
0
0
0.1
0.7
10.5 1
77.8 39
439.2 352
1,655.7 1,617
2,348.3 2,348
2,610.3 2,610
0
0
0
0
0
0
0
0 '
0
0
Ci
*
.7
.0
.6
.0
.3
.3
(in ppm)
0.08
n
w
n
\J
0
o
n
w
n
\J
o
w
0
o
o
V
o
o
*
8.4
209.0
1,544.9
2,326.1
2,610.3
*Fewer than 50,000 person-occurrences
Source: Summarized from O3 NEM Results
-------
V-15
uncertainty are so addressed. The Ambient Standards Branch (ASB)
will attempt to explicitly analyze the uncertainties mentioned
above in future O3-NEM modeling efforts, if contract funds are
available for that purpose, in addition, due to increasing
interest in multiple-hour 03 exposures it is very likely that the
current O3-NEM will not be' used in the future as it only provides
hourly exposure estimates. ASB's plan is to revert back to the
CO-NEM and update it to handle 10-minute or "probabilistic"
activity patterns. Thus, future O3 exposure estimates may be
very different than those reported here.
-------
-------
VI-1
Factors Relevant to Review
Ozone
Of primary concern in this section are the health effects
associated with levels of O3 and other photochemical oxidants
that are observed in the ambient air of the United states, of
the photochemical oxidants, only 03 has been reported' to exist in
cities at sufficiently high concentrations to be of significant
concern for human health, since other photochemical oxidants
such as hydrogen peroxide (H2O2) and perdxyacetyl nitrate (PAN)
are known to produce health effects only at concentrations much
higher than those found in the ambient air, this section- will
focus on the mechanisms of toxicity for 03 and documented
evidence of pulmonary and extrapulmonary effects of O3 . This
approach will be based on both available human and animal data.
Mechanistic studies are used to support the possibility of human
health ffects
^
data. in addition, factors affecting susceptibility and
potentially susceptible groups will be discussed.
A. Ozone Absorption and Mechanisms of Effects
Ozone enters the human body through the respiratory system
During inhalation and exhalation, 03 is removed from the air by'
airway surfaces through a process of dissolution and chemical
reaction The rate, amount, and sites of O3 uptake and removal
largely determine the extent to which exposure of an individual
to a particular concentration of O3 will evoke adverse health
streets.
Numerous studies help to explain the quantity and sites of
03 uptake ln mammalian respiratory tracts. Nasopharyngeal
removal studies reviewed in the CD (p. 9.4) suggest that: !, the
fraction of 03 uptake depends inversely on the flow rate 2
tracheal and exposure chamber concentrations of 03 are positively
correlated (Yokoyama and Frank, 1972; Moorman, et al., 197 ; Ind
o >" f"d 3> ™*°*™^ S° percent of inh d
°3 0 72, for dogs and > 50% for rabbits) im removed in the
-------
VI-2
nasopharyngeal region of animals exposed to between 0.1 and 0.2
ppm 03, thereby indicating the role of the nasopharynx in
removing 03 before it reaches the more sensitive lung tissues.
Only one study measured 03 uptake in the lower respiratory tract
and reported 80 to 87% uptake by the lower respiratory tract of
dogs (Yokoyama and Frank, 1972). There are, however, no
published data currently available for human nasopharyngeal or
lower respiratory tract absorption.
Inhaled O3 not absorbed in the nasopharyngeal region can be
deposited along the entire respiratory tract. Penetration into
the lower respiratory tract increases in rats as inhaled O3
concentrations increase from 0.2 to 0.8 ppm (Dungworth et al.,
1975). Models have been developed to estimate lower respiratory
uptake (Miller et al., 1978b, 1985; McJilton et al. 1972). One O3
deposition model has predicted that similar patterns of O3
deposition occur in humans, rabbits, and guinea pigs and that the
junction of the conducting airways and gas exchange region
receives the maximal dose (Miller et al., I978b; 1985). This
prediction is consistent with location of O3-induced lesions as
indicated by pathology data from several species of 03-exposed
animals. Recently the Miller dosimetry model has been used to
estimate sensitivity to lower respiratory tract secretions and
exercise of the uptake of O3 in the human lung (Miller et al.,
1985). The Miller dosimetry model is also an important tool in
assessing animal toxicology data. Chapter 9 of the CD provides a
detailed discussion of the use and comparison of O3 dosimetry
models.
Biologically important functional groups which react
relatively rapidly with 03 include the alkenic groups (carbon-
carbon double bonds), amino groups, and sulfhydryl groups.
Alkenic bonds are found in essential fatty acids and
polyunsaturated fatty acids (PUFA) which are important components
of the lipids in cell membranes. Oxidation of amino and
sulfhydryl groups can result in protein denaturation with
-------
VI-3
concomitant loss of structural and functional integrity for the
affected protein.
Although there is general agreement that the oxidative
properties of O3 cause toxic effects, the precise molecular
mechanism of toxicity remains unclear. Several theories have
been proposed:
1. Oxidation of polyunsaturated lipids contained mainly in
cell membranes;
2. Oxidation of sulfhydryl, alcohol, aldehyde, or amine
groups in low molecular weight compounds or proteins;
3. Formation of toxic compounds (ozonides and peroxides)
through reaction with polyunsaturated lipids;
4. Formation of free radicals, either directly or
indirectly, through lipid peroxidation; and
5. injury mediated by some pharmacologic action, such as
via a neurohormonal mechanism or release of histamine (CD, p.
9"~ I'l 3) .
Molecular targets (e.g., carbon - carbon double bond
sulfhydryl groups) are shared across all species. Therefore if
03 initiates mechanisms such as those listed above in an animal
Which ultimately result in lung structura! damage, there is
reason for concern that the sane mechanisms may be activated in
humans with similar outcomes. This assumption is limited by
differences in human and animal dosimetry which produce different
doses of 03 for equivalent exposures. Also defense and repair
mechanisms are likeiy to provide quantitative differences in the
display of toxicity resulting from equivalent doses
Nonetheless, mechanistic information strongly supports the
hypothesis that equivalent effects may occur in humans and
animals, albeit not necessarily at the same concentrations.
B. Factors Affecting Susceptibility to Ozone
There are numerous factors which could affect susceptibility
to 03 exposure and to alter physiological responsiveness These
factors include such individual characteristics as age sex
-------
vr-4
smoking status, and nutritional status. In addition,
environmental stresses and exercise level during exposure can
influence the extent and level of an individual's response to O3
by increasing the volume of inhaled 03 and by promoting
penetration deeper into the lungs.
1. Age
It has been postulated that age-related changes in lung
growth and development are among the factors responsible for
altering individual susceptibility to 03; thus, extent of damage
may depend on the stage of lung development. Support for this
hypothesis has come from several human studies, but definitive
evidence is unavailable at this time.
Several clinical, field, and epidemiology studies have
provided evidence of pulmonary function and/or symptomatic
response in younger persons exposed to less than 0.15 ppm O3
.(McDonnell et al., I985b,c; Avol et al., I985a,b; Lebowitz et
al., 1982, 1983; Lebowitz, 1984; Lippmann et al., 1983; Lioy et
al., 1985; Bock et al.,'l985; Spektor et al., 19S8a). In
addition, symptoms have been reported for children exposed to
oxidant concentrations of 0.10 ppm and higher (Makino and
Mizoguchi, 1975). Although it has been suggested that older
individuals may be less susceptible to O3 exposure than younger
persons, age differences in response to O3 have not yet been
fully investigated in controlled human studies. Because children
tend to exercise outdoors in the summer more often than adults,
there may be more reason for concern about effects of O3 in
children.
Reisenauer et al. (1988) find that only female subjects show
a rise in total respiratory resistance (R,,) when exposed to 0.3
ppm 03 during exercise. Drechsler-Parks et al. (1984) also
report no significant differences between older males and females
exposed to higher O3 levels (0.45 ppm) for 2 hours during
intermittent exercise (25 L/min) and suggested that older
individuals may be less responsive to 03 than younger persons.
-------
VI-5
Bedi et al. (1988) substantiate these results in finding that
older subjects, as a group, are less responsive to 03 than
younger subjects. Thus, while there continues to be no
controlled human experimental evidence of additive effects from
exposure to other pollutants combined with O3 or of differential
response due to sex, preliminary evidence suggests possible age-
related differential response.
The few animal studies that have addressed the potential
susceptibility of the young have shown a few differences between
neonates and young adult animals. Generally, however, the very
young rats were not more responsive than the young adult rats
(CD, p. 12-36).
Conclusions regarding age as a susceptibility factor in '
humans will remain uncertain until the results of additional
research are available (CD, p. 12-35). while more information is
needed to fully understand age-related differences, it should be
noted that recent studies by Reisenauer et al. (1988) and
Drechsler-Parks et al. (1984) provide evidence of reduced
responsiveness in older adults.
2. Sex
Due to the small number of female subjects in controlled
human studies, no definitive conclusions can be drawn regarding
sex differences in response to o3 exposures. Those human studies
which have evaluated lung function differences report that forced
0 f°r 1— °
more
,
than in males for similar exposure concentrations and
exercise levels, but the results are not conclusive. Only four
studies gave enough information for limited comparative
evaluation (Horvath et al., 1979; diner et al., 1983; Delucia et
al., 1983, Launtzen and Adams, 1985). Results of these studies
suggest that lung function of women may be affected more than
et
et
'
report no differences in male and female response
-------
VI -6
when exposed to 0.3 ppm, 0.6 ppm, N02 separately or in
combination.
Most 03 research on animals has been performed using only
males. When females were used, sex comparisons were not usually
made. An exception is the study by Graham et al. (1981) which
reports increased pentobarbital-induced sleeping time in all
females but not in male mice or rats after 5 hour exposure to l.o
ppm 03. Although reasons for sex differences have not yet been
elucidated, these differences have been suggested in some
research and thus further investigation is needed (CD, p. 12-3?).
Because an inadequate data base prevents drawing definitive
conclusions about sex differences, staff assumes at this time
that there are no differences between sexes regarding response to
03 exposure.
3. Smoking Status
Results of studies comparing the effects of 03 on smokers
versus non-smokers remain somewhat inconclusive, although below
0.5 ppm 03 smokers appear to be somewhat less responsive. Two
studies reported greater pulmonary function changes for non-
smokers than for smokers at rest during exposure to 03 levels of
0.37 to 0.5 ppm; this finding was reversed for hdgher 03 levels
(0.75 ppm), with-smokers showing greater response (Hazucha et
al., 1973; Bates and Hazucha, L973). Greater pulmonary function
response was also reported in non-smokers than in smokers at 0.5
ppm (Kerr et al., 1975) and at 0.3 ppm (Delucia et al., 1983).
Effects of 03 were greater for non-smokers compared to smokers at
0.5 and at 0.3 ppm (Kagawa and Tsuru, I979a); in a subsequent
study exercising non-smokers showed a greater response (SGaw)
than exercising smokers at 0.15 ppm O3 (Kagawa, 1983).
Exercising smokers also showed slower and smaller spirometric
variable changes than non-smokers exposed 2 hours at 0.5 and 0.75
ppm 03 (Shephard et al., 1983). Although none of these studies
has examined the effects of different amounts of smoking,
available data supports the contention that smokers are less
-------
VI-7
responsive to O3 than non-smokers, at least for lower O3
exposures. Why smokers appear to be less responsive to O3 than
non-smokers is not clear, though it has been suggested that the
altered lung function and increased mucus production experienced
by smokers could influence O3 deposition in the lungs (CD, p 12.
37).
4. Nutritional status
Results of studies investigating nutritional status as a
factor affecting susceptibility to 03 are inconclusive. Some
inconsistencies between human and animal data are apparent but
might be explained based on different experimental approaches
Human subjects receiving 4 to 8 times the recommended daily
allowance of vitamin E showed no statistically significant
differences in blood biochemistry or agglutination compared to
unsupplemented subjects after 2-hour exposures to 0.5 ppm 0,
(Posin et. al., 1979; Hamburger et. al., I979a,b). Animal studies
show that vitamin E deficiency makes rats more susceptible to o -
induced enzymatic changes (Chow et. al., 1981; Piopper et. al '
1979; Chow and Tappel, 1972) and that vitamin E alters the rate
and extent of O3 toxicity but not the lesion in the centriacinar
region of the lungs (Stephens et al., 1974; Schwartz et al
1976). Lesions have been reported to be worse in vitamin E'
deflclent rats or in rats marginally supplemented with vitamin E
when compared to highly supplemented rats (Piopper et al., 1979-
Chow et al., 1981); this supports results of mortality (Donovan'
et al 1977> and biochemical studies suggesting that vitamin E
is protective in rats (CD, p. 12-38).
Since redistribution of vitamin E from extrapulmonary stores
in humans to the lungs is slow, it is possible that such I
protective effect does not occur over short-term exposures to C,
in human studies. Another possible explanation is that an effect
is only seen when subjects are vitamin E deficient. Animal
studies in which effects have been reported were conducted with a
vitamin E deficient group for comparison and for long-term
-------
VI-8
exposures (Sato et al., 1976, 1978, 1980), neither of which is
possible in human studies. Although laboratory animal studies of
vitamin E support the hypothesis that lipid peroxidation is
involved in 03 toxicity, benefits to be derived from dietary
vitamin E supplementation in connection with O3 exposure have not
yet been demonstrated in humans.
5. Environmental Stresses
Subjective symptoms and physiological impairment caused by
03 exposure can be substantially worsened by environmental
stresses such as heat and high relative humidity (rh). High
temperatures (31 to 40°C) and/or humid conditions (85% rh) when
combined with exercise during O3 exposure have been shown to
reduce FEV-L more than similar O3 exposures at a more moderate
temperature (25°C) and humidity (50% rh) (Folinsbee et. al.,
1977a,b; Gibbons and Adams, 1984). Enhanced effects of O3 caused.
by high temperatures and humidity may be the result of higher
ventilation which increases the volume of inhaled O3 and promotes
deeper penetration into the lungs. There may also be an
independent effect of elevated body temperature on pulmonary
function (CD, p. 12-35). Since, many urban areas around the U.S.
tend to experience the highest O3 levels during periods of high
temperatures and humidity, environmental stress should be
considered a factor of concern in assessing potential effects
from 03 exposure.
6. Exercise
Exercise stresses respiratory, cardiovascular, and
musculoskeletal systems and increases ventilation and, therefore,
total dose to the lungs. Exercise thus should be considered an
important susceptibility factor when 03 exposure occurs
concurrently. The most apparent and well-studied effects of O3
during exercise occur in the respiratory system. In particular,
pulmonary function decrements and respiratory symptoms caused by
03 exposure are increased by a greater work load, which is
-------
VI-9
characterized by increased frequency and depth of breathing
(Folinsbee et al., 1979, 1984; McDonnell et al., 1983, 1985a,b,c;
Avol et al., 1984, 1985a,b). Representative activities and
associated ventilation rates are summarized in Table Vl-i (CD p
10-14) . Higher ventilation increases total volume of O3 breathed
and could cause deeper penetration of O3 into peripheral areas of
the lung which tend to be most sensitive to injury. Because most
individuals engage in some type of outdoor exercise where 03
exposure is possible during summer months, exercise should be
considered a susceptibility factor in assessing health effects of
C. Potentially Susceptible Groups
The Clean Air Act requires EPA to set standards which
protect the^health of individuals who are potentially susceptible
to 03 exposure. 'This section identifies potentially susceptible
groups or subpopulations and provides a rationale for selecting
these groups.
susceptibility to any pollutant may depend upon many factors
such as those previously discussed (age, sex, smoking, nutrition
environment, and exercise, but also is influenced by individual '
sensitivity. Greater than normal sensitivity to a particular
pollutant »ay be conferred by numerous individual characteristics
including (1) predisposition to pulmonary infection, (2)
ar!^Stln! diSeaSe ^ nutritlonal Deficiency, (3) some aspect of
growth or dedme in lung development, (4, a prior infection or
immunological problem, (5) sensitization caused by prior
pollutant exposure or challenge with respiratory irritants and
(6) genetic variability in the population. These sensitivity and
susceptibility factors should be considered in identifying groups
or subpopulations. which May be susceptible to 03 exposure
1. individuals Having Preexisting Disease
The first major group identified in the CD (p. 12-73) as
appearing to be at particular ris* to O3 exposure is that group
-------
"nun.
Level of work
Light
Light
Moderate
Moderate
Moderate
Heavy
Heavy
Very heavy
Very heavy
Very heavy
Severe
Work
watts
25
50
75
100
125
150
175
200
225
250
300
-1
performed
kg-m/minb
•
150
300
vvU
icn
^9U .
fifln
uuu
750
/ wU
onn
y\j\M
ifl*ift
4U9U
1200
1350
* \/*JW
1500
A *JU\J
1800
4 U\J\J
,
02 consumption,
L/mln
._.
0.65
Of\£
.96
10 1~
.25
1C A
.54
1O1
.83
211
.12
24 ~i
.47
2.83
3 in
.19
3r r
.55
4OT
.27
.
Minute
ventilation
L/mln
•••
12-16
17-23
23-30
29-38
35-46
42-55
52-57
62-79
73-93
89-110
107-132
Representative activities
• •»•»»• •*^*i»ivivi Cd
Level walking at 2 mph; washing clothes
Level walking at 3 mph; bowling; scrubbing floors
Dancing; pushing wheelbarrow with 15-kg load-
simple construction; stacking firewooV
Easy cycling; pushing wheelbarrow with 75-ka
load; using sledgehammer 9
C 1 1 Alb*! nO «Ct" A i f*C • r\l a A *. .t
spade" ' l""jr
-------
VI-11
characterized as having preexisting disease. This includes
individuals with asthma, allergy, and chronic obstructive lung
disease (COLD) such as emphysema. Asthmatics, who experience
variable airway obstruction and/or reactivity and who may have
altered immunological states (e.g., atopy) or cellular function
(e.g., eosinophilia) may or may not be more sensitive to 03 than
non-asthmatics. They have not been shown to be more sensitive in
numerous studies, which have shown asthmatics to have equal or
lesser responsiveness. Asthma, however, is not a specific
homogeneous disease and efforts to precisely define asthma have
been unsuccessful. Susceptibility to 03 of patients with COLD
remains somewhat uncertain, depending on their clinical and
functional state. Table VI-2 provides estimates of the number of
individuals with these conditions (CD, p. 12-79).
Although several controlled human exposure studies suggest
that adults with preexisting respiratory disease may not be more
responsive to O3 than healthy, adults, two considerations place
these subgroups at potentially higher risk to 03 exposure
according to the CD (p. 12-88). First, due to concern for the
health of these persons, higher concentrations of O3 and higher
exercise levels have not been used during controlled exposure
studies, thus making comparisons with healthy subjects difficult.
A second consideration is that for individuals with already
compromised respiratory systems, any further decrement in
pulmonary function may be expected to impair the ability to
perform normal activities. The same decrement in a healthy
individual may go unnoticed. Increases in symptoms reported for
asthmatics and allergic individuals exposed to O3 may also be
expected to interfere with normal function. Because recent
studies (McDonnell et al., 1987; Eschenbacher et al., 1988; Kreit
et al., 1988) have reported that both asthmatics and allergic
subjects have a greater increase in airway resistance after 03
exposure than do healthy subjects, the CD Supplement (U.S. EPA,
1988) has concluded that the order of airway responsiveness to O3
is normal < allergic < asthmatic subjects.
-------
TABLE VI-2
PREVALENCE OF CHRONIC RESPIRATORY CONDITIONS BY SEX AND AGE FOR 1979a
Number of persons, in thousands
Condition
Chronic bronchitis
Emphysema
Asthma6
Hay fever and
other upper
respiratory
allergies
Total0
7474
2137
6402
15,620
Male
3289
1364
3113
7027
Female
4175
770
3293
8584
<17
years old
2458
12d
2225
3151
17-44
years old
2412
127d
2203
8278
45-64
years old
1547
1008
1482
3012
>65
years old
1060
990
488
1181
% of U.S.
population
3.5
1.0
3.0
7.2
aU.S. Department of Health and Human Services, 1981.
Classified by type, according to the Ninth Revision of the International Classific of Diseases (World Health
Organization, 1977).
""Reported as actual number in thousands; remaining subsets have been calculated from percentages and are rounded off.
Does not meet standards of reliability or precision set by the National Center for Health Statistics (more than 30%
relative standard error).
eWith or without hay fever.
fWithout asthma.
Source: 03 Criteria Document, U.S. EPA, 1986.
-------
VI -13
2. Exercising Individuals
A second susceptible group potentially at higher risk to o,
exposure is that group of. individuals including both normal
healthy persons and those with preexisting respiratory disease
whose regular outside activities cause increased minute
ventilation. As stated in the CD (p. 12-71) , ,,the most prominen
modxf.er of response to O3 in the general population is minute
ventilation, which increases proportionately with increases in
exercise workload." The increased proportion of oral breathing
with increased exercise also contributes to the increase in
effects since more O3 is delivered to the lungs. Although
individuals with preexisting respiratory or cardiovascular
disease may not exercise as heavily as healthy persons, any
increases in activity over resting levels will increase 0,
exposure and resultant effects.
Exercise has become recognized in numerous recent studies as
a factor which can predispose all individuals to 03 health
effects (McDonnell et al., 1983, 1985a,b,c; Avol et al., 1983
1984, 1985a,b; Folinsbee et al., 1978, 1984; Kulle 6t ^
Thus, activities which increase minute ventilation out-of-doors '
will also increase the risk associated with exposure to 0, of
exercising individuals.
Unusual responsiveness to O3 has been observed in some
individuals. These individuals, often referred to as
"responders," have not been characterized, as having any
particular medical problen but experience signif ioantly greater
pumonary function decrements than the average response of the
If0tlssecon7rTy studied during °3 exposure- zt is not *<>°™
if these individuals are a population subgroup with a specific
ri* factor or simply represent the upper 5 to 20 percent of the
03 response distribution. As yet there is no Means of
identifying these highly responsive individuals prior to 0
exposure. ' 3
reSP°nSiVeness *° °3 <* apparently healthy,
groups has been shown to vary widely. For
-------
VI-14
during heavy exercise (minute ventilation, Ve/ = 45-51 L/min) and
exposure to 0.4 ppm O3/ subjects showed FEY^ decrements ranging
from below 10% to as much as 40% of control values for those
subjects with an average of 26% (Haak et al., 1934; Silverman et
al., 1976). Mode (oral versus nasal) and pattern (rapid shallow
versus slow deep) of breathing may contribute to intersubject
variability but cannot fully explain it. Unsuccessful attempts
have been made to determine which factors are responsible for
modifying individual responsiveness (Hazucha, 1981). Factors
such as prior exposure to O3 or other pollutants, latent
infections, and nutritional deficiencies may also contribute to
differential response (CD, p. 12-22). The potential for these
contributions is biologically plausible, but these have yet to be
demonstrated in human subjects.
Although the specific factors which modify individual
response to 03 are not yet identified, it has been suggested that
intersubject variability in the magnitude of effects induced by
03 is caused by large differences in intrinsic responsiveness to
03 (McDonnell et al., I985a; Kulle et al., 1985). As measured by
FEVi and FVC, individual responses to O3 are highly reproducible
for at least the 10 month period of observation and for a given
individual differences tend to be much smaller than differences
between subjects (McDonnell et al., 1985a; Gliner et al., 1983).
Hence, some intrinsic factor appears to be responsible for
individual responsiveness.
Changes in FEVj_ of exercising subjects exposed to clean air
are small and normally distributed in the subject population. As
03 concentrations increase this distribution widens and becomes
" skewed toward larger FEVj. decrements, the largest changes
representing the most responsive subjects (McDonnell et al.,
1983; Kulle et al., 1985). Classification of subjects into'
"responders" and "non-responders" has been somewhat arbitrary.
Examples of retrospective classification include identifying
responders by using medical history and exposure test results
(Hackney et al. 1975) and selecting those with greater than 10%
-------
VI-15
post-exposure decrements (Horvath et al., 1981) or with FEV
decrements greater than 2 standard deviations above control (Haak
et al., 1984). while there are no clearly established criteria
for identifying or selecting "responders, '• it has been suggested
that between 5 and 20% of the healthy population may represent a
subgroup of the population which is more responsive and,
therefore, at higher risk to O3 exposures (CD, p. 12-22)'.
Further discussion of frequency distribution is in section
VII.A.I.
-------
-------
VII-1
VII. Assessment of H
Considered in S
This assessment of health effects attributed to O3 and
related health issues is based on the review and evaluation of
health effects research literature contained in Chapters 9 to 12
of the CD as well as newer information discussed in the CDS.
A. Health Effects of Concern
For purposes of this staff paper and review of the NAAQS for
03, the staff recommends that the following categories of health
effects attributed to 03 exposure be considered in developing the
basis for the primary 03 standard:
l. Alterations in pulmonary function
2. Symptomatic effects
3. Exercise performance
4. Bronchial Reactivity
5. Aggravation of existing respiratory disease
6. Morphological effects
7. Altered host defense systems
8. Extrapulmonary effects
Respiratory effects provide by far the strongest data base for
considering adverse health effects of o3. Extrapulmonary effects
such as behavioral, blood chemistry, chromosomal, cardiovascular
reproductive, teratological, liver metabolism, and endocrine '
system modifications also have been associated with O3 exposure
but most are considered to be of uncertain functional important
at this time and, thus, do not provide very strong evidence of
adverse effects associated with 03 exposures.
Evidence for the above effects has come from both human and
animal studies. The strongest and most quantifiable data are
provided by controlled human exposure studies, but these studies
are limited to acute (short-term) exposures. Field studies are
similarly limited, but permit investigation of the effects of
oxidants in ambient air and allow for better characterization of
exposure than epidemiological studies. Although use of most
-------
V1I-2
community studies has been hampered by the difficulty in
adequately characterizing exposure and numerous confounding
variables, these investigations provide important supporting
evidence for effects occurring in populations. Animal toxicology
studies have provided evidence of some acute and chronic (long-
term) exposure effects which can be detected only with invasive
procedures, but uncertainties inherent in dosimetry and species
sensitivity differences have limited quantitative extrapolation
to humans.
Assessment of health effects attributed to O3 requires
consideration of the data base in each of the above areas of
study. This section integrates research in each of these areas
to provide an indication of the strength of the data base for
each effect. A more thorough review and evaluation of individual
studies is available in chapters 9 to 12 of the CD.
1. Alterations in Pulmonary Function
The best documented and strongest evidence of human health
effects of 03 exposure are pulmonary function decrements.
Controlled human exposure, field, epidemiology, and animal
toxicology studies have provided evidence that exposure to O3 can
modify such pulmonary measurements as forced expiratory volume
(FEV), forced expiratory flow (FEF), forced vital capacity (FVC),
vital capacity (VC), tidal volume (VT) , peak expiratory flow rate
(PEFR), inspiratory capacity (1C), total lung capacity (TLC),
airway resistance (Raw) and breathing frequency (fB). These and
other terms are defined in Appendix B which is an abbreviated
version of the glossary found in the CD.
Early controlled experimental studies of resting human
subjects exposed to 03 levels up to 0.75 ppm for 2 hours
demonstrated little or no change in FVC (Silverman et al., 1976;
Folinsbee et al., 1975;), FEV1(, and FRC (Silverman et al., 1976).
Flow rate variables such as FEF 25% and FEF 50% showed up to 30%
decreases in some subjects exposed at rest to 0.75 ppm O3 (Bates
et al., 1972; Silverman et al., 1976), while only small increases
-------
VII-3
in R
aw (< 17%) were reported for > 0.5 ppm 03 exposures (Bates et
al.f 1972; Golden et al., 1978). More recent studies have
reported occurrence of FEV and FEF decrements during resting
exposures to > 0.5 ppm O3 (Folinsbee et al., 1978; Horvath et
al., 1979); however, no statistically significant changes in R
and only suggestive changes in RV and TLC have been reported for
similar exposures (Shephard et al., 1982). Airway resistance is
not generally affected in resting subjects at these 03 levels.
Changes in pulmonary function have not been observed in resting
subjects exposed to 03 levels between 0.12 ppm (Koenig et al
1985) and 0.3 ppm (Folinsbee et al., 1978), though some subjects
exhibit 03-induced pulmonary symptoms during resting exposures
(Konig et al., 1980; Golden et al., 1978). In general, however
because subjects were at rest in most of the older studies
significant respiratory effects were not reported even for higher
03 exposures.
Exercise, which causes increased minute ventilation (V )
enhances individual and group mean response to 03 exposure * As
discussed in section VLB. 6. of this staff paper, exercise
increases breathing frequency and depth of breathing resulting in
greater total dose of 03 inhaled and increased penetration to the
most sensitive lung tissue. As exercise levels increase to the
point where Ve exceeds approximately 35 L/min, oronasal or oral
breathing tends to predominate (Niinimaa et al., 1980); thus at
higher exercise levels a greater portion of the inhaled O3 will
bypass the nose and nasopharynx (Niinimaa et al., 1981)
individual variability will affect the Ve at which oral or
oronasal breathing predominates. The relationship between
exercise and magnitude of response is illustrated quite well by
Figure VII-i prepared as Figure 12-6 for the CD (p. 12-31)
showing group mean decrements in FEV, caused by exposure of
exercising subjects to various 03 levels based on results from 25
different studies. The curves clearly demonstrate that as
exercise levels increase for a given 03 exposure there is a
resultant increase in the group mean FEV, decrements.
-------
110
100
ui
5
> 90
cc
2 80
Q
UI
o
cc
o
£ 70
to
60
'••.. LIGHT EXERCISE
VERY HEAVY
EXERCISE
uc*t
HEAVY ^
EXERCISE
MODERATE
U.2
0.4
0.6
0.8
OZONE CONCENTRATION, ppm
FIGURE VI I-1. Group mean decrements in 1 -sec forced expiratory volume during 2-hr ozone
exposures with different levels of intermittent exercise: light (v"E ^23 L/min); moderate tip =
24-43L/mm); heavy (VE = 44-63 L/min); and very heavy (C'g ^ 64 L/min). Concentration-
response curves are taken from Figures 12-2 through 12-5.
Source: 03 Criteria Document, U.S. EPA, 1986
-------
VII-5
Pulmonary function decrements have been reported in
controlled exposure studies of healthy exercising human adults
exposed to 03 levels between 0.12 ppm and 0.24 ppm, a range for
which a concentration-response relationship has been established
During 2 hours of intermittent very heavy exercise (V = 64
L/min.), healthy subjects experienced group mean decrements in
FEVi of 4.5, 6.2, and 14.5% for O3 exposures of 0.12, 0.18 and
0.24 ppm, respectively (McDonnell et al., 1983). Positive
associations between group mean magnitude of FVC (3, 4/ and 12%)
and FEF (7.2, 12, and 23%) decrements and O3 exposures of (.12,
.18, and .24 ppm) also were reported in the McDonnell et al
(1983) study; statistically significant increases in SR and f
did not occur until O3 was > 0.24 ppm. In a separate study *
McDonnell et al. (I985b) reported a statistically significant
but small decline (3.4%) in group mean FEV, of children (8-11
years) after 2 hours of exposure to 0.12 ppm 03 during
intermittent heavy exercise (Ve = 39 L/min). Furthermore, there
was a suggestion that the small decrements in FEV, persisted for
16 to 20 hours after 03 exposure ended. Findings by Lioy et al
(1985) provide epidemiological evidence that lung function
decrements can persist in children up to a week following a smog
episode in which l-hour 03 peaks were 0.135 to 0.186 ppm.
Support for 03-induced pulmonary function changes also comes
from other controlled exposure studies. Avol et al. (1984)
report small but statistically significant decreases (6.1%) ln
FEVi, at 0.16 ppm O3 with a larger decrements (19.1%) at > o 24
PP* 03. Decrements have been reported in FEV,, FVC, and FEF for
distance runners and distance cyclists exposed to 0.20 ppm
and 0.21 ppm O3 respectively, during 1 hour of continuous, heavy
exercise (ve = 77.5 and 81 L/min) (Adams and Schelegle, 1983-
Folinsbee et al., 1984). Separate studies of continuously '
exercising males (Ve = 61.8 L/min) and females (Ve = 46 L/min)
exposed to 0.2 ppm O3 for about an hour showed no statistically
significant FEV, decrements, but these negative results can be
attributed in part to the small number of subjects (i e
-------
VII-6
inadequate power), 8 male and 6 female respectively (Adams et
al., 1981; Lauritzen and Adams, 1985). In. another study
involving 03 concentrations ranging between O.lo and 0.25 ppm/
exponential decreases in FVC, FEV^ FEF25_75/ SGaw/ IC/ and TLC
have been reported with exposure to increasing 03 concentration
during very heavy exercise (V(J = 68 1/min) ; time of exposure was
related to linear decreases in FVC and FEVX (Kulle et al., 1985).
Field studies, which contain elements of both controlled
human exposure and epidemiologic studies, provide the most
quantitatively useful human exposure data available for ambient
photochemical oxidants. Results from field studies are
consistent with pulmonary function decrements reported in
controlled O3 exposure studies. For a 1-hour exposure to mean O3
concentrations of 0.144 ppm, small significant group mean
decreases in FVC (3.3%), FEV0.75 (4.0%), FEVl (4.2%), MMFR (3.2%)
and PEFR (3.9%) relative to pre-exposure levels were reported in
59 healthy continuously exercising (Ve=32 L/min} adolescents
(Avol et al., 1985 a, b). Based on the comparison of ambient air
and control exposures in these studies, it appears that O3 was
the causative agent of the pulmonary function effects.
In a separate study of 50 healthy, adult, continuously
exercising (Ve = 53 L/min) bicyclists, mean 03 concentrations of
0.153 ppm for 1 hour produced statistically significant group
mean decreases in FEVl (5.3%) compared to pre-exposure (Avol et
al., 1984). This study showed that similar effects result in
subjects exposed to comparable O3 levels in ambient and
controlled 03 exposures. Small but statistically significant
decreases in FEVX and FVC were also reported in exercising
healthy and asthmatic adults during exposure to mean O3
concentrations of 0.165 and 0.174 ppm (Linn et al., 1980, 1983b;
Avol et al., 1983). Table VII-l provides a summary of group mean
% changes in FEVX for controlled exposure and field studies.
In summary, pulmonary function decrements have been reported
in healthy adult subjects (18 to 45 years old) after 1 to 3 hours
of exposure as a function of the level of exercise performed and
-------
Ozone
Concentration Measurement* 'b Exposure
Mtl/ir1 ppM Method Duration
0 0.00 UV, UV
157 0.08
0 0.00 UV, UV
157 0.08
196 0.10
235 0.12
0 0.00 UV. UV
157 0.08
196 0.10
0 0.00 CHEM. UV
157 0.08
196 0.10
235 0.12
0 0.00 UV, UV
196 0.10
0 0.00 CHEN, NBKI
196 0.10
0 0.00, UV, UV
216 O.llf
0 0.00 CHEM, UV
235 0.12
0 0.00 CHEM, UV
235 0.12
0 0.00 CHEM, UV
235 0.12
0 0.00 UV
235 0.12
0 0.00 UV
235 0.12
353 0.18
Source: Supplement to Air
1 hr
2 hr
2 hr
1 & 2 hr
of 6.6 hr
study
2 hr
2 hr
1 hr
2 hr
2 hr
1 i 2 hr
of 6.6 hr
study
1 hr
(mouthpiece)
40 mfn
(mouthpiece)
Quality Criteria
Activity0
Level (V£)
CE (57)
IE (68)
IE (68)
CE (40)
IE (68)
IE (67)
CE (22)
IE (68)
IE (39)
CE (40)
R
IE (33)
30 min
ft + 10 min
exercise
for 0
Percent .
Change in FEV?
+0.6
+1,7 (ns)
(26,4 i 6.9)
+1.0
+2.4 (ns)
+1.7 (ns)
+1.0
+2.4 (ns)
+1.7 (ns)
-1.5 (1 hr) -1.0 (2 hr)
-0.4 (ns) -1.1 (ns)
-1.3 (ns) -1.3 (ns)
-0.5 (ns) -2.7 (ns)
+1.5
+1.1 (nd)
range: +10 to -4
-0.3
-2.6 (ns)
' -2.7
-2.9 (ns)
-i.o
-4.5 (p = 0.016)9
range: +7 to -16
-0.5
-3.4 (p = 0.03)
range: +5 to -22
-0.2 (1 hr) -1.2 (2 hr)
-2.6 (ns) -3.8 (ns)
-1.1
0.0 (ns)
-1.0
+1.7 (ns)
-0.3 (ns)
Number, Sex, and
Age of Subjects
42 nale
6 female
24 male
(18-33 y'r)
24 Male
(18-33 yr)
21 male
O8-3J yr)
20 mate
(25.3 ±4.1 yr)
10 male
(18-28 yr)
33 male
33 female
(8-11 yr)
22 male
(22.3 t 3.1 yr)
23 male
(8-11 yr)
10 Male
(18-33 yr)
4 Male
6 female ,
(13-18 yr) .
5 male
7 females
(11-19 yr)
(continued on the
Reference6
Avol et al.
Linn et al.
Linn et al.
Horstman et
Kulle et al.
Foltnsbee et
Avol et al.
McDonnell et
McDonnell et
FoHnsbee et
Koenig et al.
Koenig et al.
following page)
(1984)
(1986)
(1986)
al. (1988)
(1985)
al. (1978)
(1987)
al. (1983)
al. (1985)
al. (1988)
(1985)
(1987)
-------
Table VII-1 (confd) KEY HUMAN STUDIES NEAR THE CURRENT 1-HR NAAQS FOR OZONE
Ozone
Concentration Measurement3'
ug/M3
0
235
0
235
0
235
0
235
274
0
274
0
294
0
294
0
294
0
314
0
314
0
314
0
333
0
333
PPM Method
0.00 UV
0.12
0.00 UV, UV
0.12
0.00 UV, UV
0.12
0.00 UV, UV
0.12
0.14
0.00, UV. UV
0.14r
0.00, uv, UV
0.15r
0.00 UV, UV
0.15
0.00 UV, UV
0.15
0.00 UV, UV
0.16
0.00 UV. UV
0.16
0.00, UV. NBKI
0.161
0.00, UV, NBKI
0.17r
0.00, UV. NBKI
0.17f
Exposure
Duration
1 hr
(Mouthpiece)
1 hr
(mouthpiece)
1 hr
2 hr
1 hr
1 hr
2 hr
1 hr
(mouthpiece)
1 hr
2 hr
1 hr
1 hr
2 hr
Activity0
Level (V£)
IE (33)
CE (86)
CE (89)
IE (6fl)
CE (31)
CE( 53)
IE (68)
CE (55)
CE (57)
IE (68)
CE (38)
CE (42)
IE (2XR)
Percent .
Change in FEV?
-2.4
-0.6 (ns)
+2.4
-1.8(ns)
+4.1
-5.6 (p <0.02)
range; +10 to -29
+1.0
+2.8 (ns)
+1.6 (ns)
' -0.5
-4.2 (p <0.01)
+0.6
-5.3 (p <0.05)
+ 1.5
-0.5 (nd)
range: +3 to -9
+0.6
-4.5 (ns)
range: +3.5 to -30.6
+0.6
-6.1 (p <0.05)
+ 1.0
-2.3 (p <0.05)
range: +8.9 to -35.8
-0.1
-0.8 (ns)
-0.4
-3.4 (p <0.006)
+0.6
-2.1 (p <0.05)
Number, Sex, and
Age of Subjects
5 male
8 females
(12-17 yr)
10 male
(19-29 yr)
15 male
2 female
(24 + 3 yr)
24 male
(18-33 yr)
46 male
13 female
(12-15 yr)
42 male
8 female
(26.4 ± 6.9 yr)
20 male
(25.3 ±4.1 yr)
10 female
(22.9 ± 2.5 yr)
42 male
8 female
'26. 4 ± 6 9 "r^
24 male
(18-33 yr)
27 male
21 female
(28 + 8 yr)
45 male
15, female
(30 + 11 yr)
14 male
20 female
(29 ± 8 yr)
Reference6
Koenig et al. (1988)
Schelegle and Adams (1986)
Gong et al. (1986)
Linn efal. (1986)
Avol et al. (1985)
Avol et al. (1984)
Kulle et al. (1985)
Gibbons and Adams (1984)
Avol et a). (1984)
Linn et al. (1986)
Linn et al. (1983);
Avol et al. (1983)
Linn et al. (1983);
Avol et al. (1983)
Linn et al. (1980, 1981)
(continued on the following page)
-------
KEY """*" STUD1ESNEARTHE CURRENT 1-HR
0 0.00
470 0.24
1 hr
CE (90)
+0.3
-3.1 (ns)
range: +6.0 to -16.6
+4.1
-21.6 (p <0.001)
range: +10 to -46
NAAQS FOR OZONE
Number, Sex. and
Age of Subjects
20 male
(23.3 ± 2.8 yr)
10 male
(19-29 yr)
...
20 male
(25.3 ±4.1 yr)
8 male
13 female
(18-31 yr)
15 male
2 female
(24 + 3 yr)
Reference6
'•
McDonnell et al. (1983)
— _—
Schelegle and Adams (1986)
——
Kulle et al. (1985)
— —
Gliner et al. (1983)
—— — .
Gong et a). (1986)
10 male
(24 + 4 yr)
Adams and Schelegle (1983)
-6.0 (p <0..05)
8 male
(22-46 yr)
Adams et al. (1981)
— —
Lauritzen and Adams (1985) ^
1 hr
(mouthpiece)
6 female
(22-29 yr)
Folinsbee et al. (1984)
UV, UV
1 hr
CE (60)
-1.0
-14.5 (p <0.005)
range: -1 to -36
___
+0.6
-19.1 (p <0.05)
20 male
(22.9 ± 2.9 yr)
42 male
8 female
(26.4 ± 6.9 yr)
McDonnell et al. (1983)
Avol et al. (1984)
Schelegle and Adams (1986)
—
Kulle et al. (1985)
u "
Measurement method: CHEM = gas phase chemilumin^n
1traviolet
NBKI = neutra)
^s , not significant; nd = not
^See U.S. Environmental Protection Agency (1986).
^Measured in ambient air (mobile laboratory)
"Suggested" significance based on Bonf.™, ,nequality
vent;)ation: IE =
"Bn.f.c.nc. based on difference between
03 and filtered air (0.0 ppm 03) exposures:
co,.tC,lon (p
-------
VII-10
the 03 concentration inhaled during exposure. Group mean data
pooled.from numerous controlled human exposure and field studies
and summarized in the CD (p. 12-80) indicate that,
on average, pulmonary function decrements occur at:
(a) > 0.5 ppm 03 when at rest (Ve . 5-10 L/min; e.g.,
sitting);
(b) > 0.37 ppm 03 with light exercise (Ve = 10 to 23 L/min;
e.g., slow walking); (c) > 0.30 ppm 03 with moderate
exercise (Ve = 23 to 43 L/min; e.g., brisk
walking);
(d) > 0.24 ppm 03 with heavy exercise (Ve = 44 to 63 L/min;
e.g., easy running);
(e) > 0.18 ppm 03 with very heavy exercise (Ve >64 L/min;
e.g., competitive running).
Although the group mean changes in lung function in the
.above studies are small, considerable intersubject variability in
the magnitude of individual pulmonary response exists, and some
subjects experienced responses which were quite large (See
Section VI.c.2). Controlled exposures to 0.12 ppm 03 during very
heavy exercise have resulted in individual pulmonary function
decrements up to 16% for adults (McDonnell et al., 1983) and up
to 22% for children (McDonnell et al., I985b). Individual
subject data from the Avol et al., 1984, McDonnell et al., 1983,
and Kulle et al., 1985 studies have been used to estimate the
fraction of population which experiences > 10% and > 20% FEV,
decrements due to 03 exposure (Figure VII-2). The exposure-
response relationships represented in Figure VII-2 suggest that
between 2 and 20% of the heavily exercising population might
exhibit > 10% decreases in FEV;L when exposed to 03 levels of
approximately 0.12 ppm. It is estimated that between 0 and 5% of
the heavily exercising population might experience FEVX
decrements of > 20% when O3 exposures are 0.12 ppm. These
estimates must be considered approximate considering: (i) the
amount of variability of response between the three studies
analyzed, (2) the small number of subjects tested in each study
-------
VII-11
I
I
I
D(FEV1) >= 10%
0-2 0.3
OZONE CONCENTRATION (ppm)
Avol
McDonnell
D(FEV1) >s 20%
0
OB
o.o
X"
4
a
[
I .•
f
• Avol
a Kuito
At A^n !•
McDonnell
OZONE CONCENTRATION (ppm)
0.4
>20* r POPULATIOH
-------
VII-12
(i.e., between 20-50 subjects at any given exposure level), (3)
the fact that the subjects in these three studies were not
derived from population based sampling, and (4) the use of a
fitted function based on empirical, rather than biological
grounds. Further discussion of this analysis can be found in
Section VII.B. of the staff paper and in Hayes et al. (1987).
Recent controlled human exposure, epidemiology, and animal
toxicology studies have increased concern about enhanced
respiratory effects associated with multi-hour (6-8 hour)
exposures to 03. Clinical research conducted since closure on
the CD has provided data which indicate lung function decrements,
symptoms, and even inflammation during prolonged 03 exposures
below the current 03 NAAQS level of 0.12 ppm. In addition, camp
and field studies offer evidence which corresponds well to
clinical data on lung function decrements and symptoms. New
animal toxicology research is helping to make a stronger case for
needing to consider time as an important variable along with
concentration and Ve in assessing health effects of O3.
Folinsbee et al. (1989) published data indicating that ten
healthy non-smoking males (18-33 years) exposed to 0.12 ppm O3
during intermittent moderate exercise. Exposures lasted 6.6
hours and consisted of six 50-ininute exercise Ve = 40 L/min)
periods interspersed with 10 minute rest breaks and a 35 minute
lunch break. Subjects experienced progressively decreasing FEY-,^
over the exposure period. At the end of O3 exposure, group mean
FEVi had decreased by 13.0%, FVC by 8.3%, and FEV25_75 by 17.4%
when compared to clean air exposures. One individual experienced
a 47.6% decrease in FEV^. There were also progressive increases
in symptom ratings of cough and pain on deep inspiration for
subjects exposed to 0.12 ppm 03 as well as a marked increase in
methacholine airway responsiveness.
In a continuation of the Folinsbee et al. (1989) work,
Horstman et al. (1988,1989) exposed each of 22 nonsmoking,
healthy, male subjects to 0.0, 0.08, 0.10, and 0.12 ppm O3 on
separate days. Exposure protocol was similar to that described
-------
VII-13
above for the Folinsbee et al. (1989) study. Substantial
pulmonary function decrements, respiratory symptoms, and
increases in nonspecific airway reactivity were reported for all
three O3 exposures when compared to filtered air exposures
Group mean decreases in FEV, were 7.0% at 0.08 ppm o3, 7.0% at
0.10 ppm 03/ and 12.3% at 0.12 ppm o3, all of which were
statistically significant (p < .01). Average ratings Qf ^ ^
deep inspiration were low but increased significantly (p < 05)
following exposure to all three 03 levels compared to filtered
air. Ratios (p < .005) of PD100 observed for filtered air
compared to 03 exposure were 1.56 at 0.08 ppm O3, 1.89 at O.io
PPin 03, and 1.89 at 0.12 ppra o3. These results led Horstman et
al. (1988, 1989) to conclude that clinically meaningful pulmonary
responses can be induced by 03 exposures and exercise levels
which simulate a 6- to 8-hour period of moderate to heavy work or
play outdoors during" the O3 season.
. Further support of the relationship between acute 03
exposure and pulmonary function decrements is provided by several
epidemiological studies of children and young adults (Lippmann et<
al., 1983; Lebowitz et al., I982a, 1983; Lebowitz. 1984; Bock et
al-, 1985; Lioy et al., 1985; Spektor et al., I988a,b). These
studies report decreased peak flow or increased airway resistance
for acute exposures to ambient O3 concentrations during the study
period. Lioy et al. (1985) reported that a persistent decrement
in lung function of children lasted as much as a week after the
end of a smog period of four days during which peak l-hr O3
levels were in the range of 0.135 to 0.186 ppm. The persistent
decrements suggested by Lioy et al. (1985) (i.e., altered
epithelial permeability and changes in airway secretion)
represent a potentially more important response than the more
transient effects found in controlled exposure studies; however
they recognize that elevated concentrations of inhalable
particulate matter associated with a large scale photochemical
smog episode also are associated with effects reported. Spektor
et al. (I988a) have followed up the work of Lioy et al. (1985)
-------
VII-14
with a summer camp study of children exposed to ambient 03 in an
area which did not exceed 0.12 ppm 03 during the study. Multiple
regression analyses indicate that the most explanatory
environmental variables for daily change in lung function were:
1) previous hour 03 levels, 2) cumulative daily O3 exposure
between 9 a.m. and the time of measurement of lung function, 3)
ambient temperature, and 4) humidity. Spektor et al. (1988a)
concludes that the l hour O3 levels have the strongest influence
on lung function and that regression slopes for FVC, FEV,, PEFR,
and MMEF predict average decrements of 4.9%, 7.7%, 17% and 11%,
respectively at 0.12 ppm 03. Upper decile decrements are
predicted to be 14%, 19%, 42%, and 33% for FVC, FEV^ PEFR, and
MMEF, respectively at 0.12 ppm O3. Lioy and Dyba (1988) have
proposed recently that the PEFR decrements reported in Lioy et
al. (1985) resulted from total O3 dose rather than persistence of
the effect from one day to the next.
Epidemiology studies comparing incidence of chronic lung
disease in communities have thus far been relatively unsuccessful
due to the lack of differences in pollutant levels, inadequate
control of covariables, and insufficient individual exposure data
(CD, p. 11-54). While it has been concluded in the CD (p. 11-53)
that not one of these epidemiological studies provides definitive
quantitative data by itself due to methodological problems and
confounding variables, the aggregation of studies provides
reasonably good qualitative evidence of association between
ambient photochemical oxidant exposure and acute pulmonary
effects. The association is strengthened by the consistency of
these epidemiological results with the findings of McDonnell et
al. (1985b) and Avol et al. (I985a,b) who reported small
decrements in pulmonary function for exercising children exposed
to 0.12 ppm 03 in purified air and adolescents exposed to 0.144
ppm O3 in ambient air, respectively.
In conclusion, the weight of currently available evidence
indicates that healthy, heavily exercising subjects can
experience pulmonary function decrements during controlled
-------
VII-15
exposures of > 0.12 ppm 03 for 2 hours. Although level of
exercise and individual responsiveness play a major role in
determining the extent of pulmonary function loss, staff
concludes that 0.12 ppm 03 is the lowest observed effects level
(LOEL) for this effect in controlled exposure studies of 2-hour
exposures. Effects have been reported as low as 0.08 ppm when
exposures lasted for 6.6 hours. Field and epidemiology data
provide added evidence of measurable functional decrements below
0.12 p», perhaps in part due to pollutant interactions
The.question of what degree of pulmonary function response
should be considered adverse is addressed in Section VII.c.l.
2. Symptomatic Effects
Respiratory symptoms have been associated with group mean
pulmonary function changes in adults acutely exposed in
controlled exposures to 03 and in ambient air containing 03 as
the predominant pollutant. Despite a close association observed
between changes in group mean FEV, and group mean respiratory
symptoms for 03 exposures (CD, p. 12-17), Hayes et al. (i987b)
report only a weak-to-moderate correlation between FEV, changes
and symptoms severity when the analysis is conducted using
individual data. It should be noted that symptoms reported are
inherently more subjective than FEV, decrements measured
in controlled O3 exposures, some heavily exercising (V > 65
L/min) adult subjects have experienced cough, shortness of " ""
breath, and pain on deep inspiration at 0.12 ppm 03, although the
group mean response was statistically significant for cough only
(McDonnell et al., 1983). Above 0.12 ppm o3, respiratory and
non-respiratory symptoms which have been reported include throat
dryness, chest tightness, substernal pain, cough, wheeze, pain on
deep inspiration, shortness of breath, dyspnea, lassitude
malaise, headache, and nausea (DeLucia and Adams, 1977; Kagawa
and Tsuru, I979a,b,c; McDonnell et al., 1983; Adams and
Schelegle, 1983; Avol et al., 1984; Gibbons and Adams, 1984;
Folinsbee et al., 1984; Kulle et al., 1985). At 0.2 ppm O3 and
-------
VII-16
controlled
exposure
.„ n n
(McDonnell et al., „
been anaiyzed
P0pulation
syn,ptoms
Figures VIM and VII.4
24 hours
r<*°«« .
et al. „ "
subjects
-creases. For exanple
., 1985; vol ^
the '
and
Varl°
t e.response
ng " thSSe studi«, the
*"** ** °'
, between o and 15* f
or chh:;tvidly 9 «•«*
level rises to o 2 nl ^K disc™fort. As the 0
the ««-ted fraction of the
cl>est
Population that night
increases to a
fact that the subjects in 7 giVe" aXposu" ^vel, (3
f - ^^cn based3sa:pl ;th:ndth;;: StU«- -re not ^ d
junction based on e*piricai, ^ther *H ^ US'3 °*
Section VH-B Of tnis *' r"her than biological
a .ore ^11
anaiysis. ot the respiratory sympto»
in
exercising (v. . 57
Purified air containing : ,
« -ich contained o.15 ppffl
°f
fieid studies, one
in
-------
= Mild Chest DiscomToT
0.0
0.0
0.0
0.0
01 02 0.3
OZONE CONCENTRATION (ppm)
01 0.2 0.3
OZONE CONCENTRATION (ppm)
0.4
FIGURE VII-3 FRA
1.0
0.8
>= Moder. Chest Discomfort
I "]
-------
VII-18
Lower Respir. Symp. (Avol)
a >. Mild
• >« Moderate
OZONE CONCENTRATION (ppm)
POPULATI™
-------
VII-19
significant differences, suggesting that increased symptoms
associated with lung function impairment were caused by O3 alone
(Avol et al., 1984). Several epidemiology studies have provided
evidence of qualitative associations between ambient oxidant
levels > 0.10 ppm and symptoms in children and young adults such
as throat irritation, chest discomfort, cough, and headache
(Hammer et al., 1974; Makino and Mizoguchi, 1975). Thus it can
be concluded that most symptoms reported in individuals exposed
to 03 m purified air are similar, but not identical to, those
found for ambient air exposures.
. An exception is eye irritation, a common symptom associated
with exposure to photochemical oxidants, which has not been
reported for controlled exposures to 03 alone. This appears to
hold even at 03 concentrations much higher than would be found in
the ambient air. it is widely accepted that other oxidants such
as aldehydes and peroxyacetyl nitrate (PAN) are primarily
responsible for eye irritation and are generally found in
atmospheres containing higher ambient O3 levels (Altshuller
1977; National Research Council, 1977; U.S. Environmental
Protection Agency, 1978; Okawada et al., 1979).
Pulmonary function decrements have been reported in studies
which did not report increases in symptoms. Children (age 8-11)
intermittently exercising (ve = 39 L/min) for 2.5 hours at 0 12
ppm 03 showed small, but statistically significant decreases in
FEV, but showed no changes in frequency or severity of cough
compared to control (McDonnell et al., 1985a,b). Similarly
adolescents (age 12-15) continuously exercising (V = 31-33'
L/min) during exposure to 0.144 ppm mean O3 in ambient air showed
no changes in symptoms despite statistically significant
decrements in group mean FEV, (4%, which persisted at least one
hour during resting post-exposure (Avol et al., I985a,b)
Because symptoms can be viewed as an early warning of related
lung function impairment by 03, the lack of symptoms in children
and adolescents during exposures which induce functional
decrements may be of concern. It is possible that children may
-------
VII-20
be at higher risk since, with no warning, they may not exhibit
avoidance behavior.
Symptoms when combined with objective measures of lung
function are considered useful adjuncts in assessing health
effects caused by O3 and photochemical oxidants. Group mean
symptoms have been shown to be closely associated with the time-
course and magnitude of group mean pulmonary function changes
associated with O3 exposures; however, only weak-to moderate
correlations exist when data are analyzed on an individual basis.
To the extent symptoms associated with exposure to 03 and other
photochemical oxidant exposures are associated with discomfort,
interfere with normal activity, and provide subjective evidence
of functional impairment, the staff in concordance with comments
made by CASAC (CASAC, 1986) recommends that they should be
considered adverse health effects and as such should be included
.in identifying a lowest observed effects level (LOEL).
3. Exercise Performance
A limited data base provides suggestive evidence of
decrements in exercise performance associated with O3 exposure.
A detailed discussion of those data can be found in the CD (pp.
10-60 to p. 10-65) and is summarized in Table 10-6 (CD, p. 10-
64).
Early epidemiological evidence on high school students
suggested that the percentage of track team members failing to
improve performance increased with increasing oxidant
concentrations the hour before a race (Wayne et al. 1967). The
authors concluded that the effects may have been related to
increased Raw or to associated discomfort which may have limited
motivation to run at maximal levels. Controlled exposure studies
of heavily exercising competitive runners have demonstrated
decreased max Ve at 0.3 ppm 03 (Savin and Adams, 1979) and
decreased FVC, FEF, and FEF at 0.20 ppm 03 (Adams and Schelegle,
1983). At 0.21 ppm 03, Folinsbee et al., (1984) reported
decreases in FVC, FEV, FEF, 1C, and MW at 75% max VO2 as well as
-------
VII-21
symptoms in 7 distance cyclists exercising heavily (V = 81
L/min). e
With regard to exercise performance studies, reductions in
group mean FEV, (-5.6%) and mild symptoms (cough, chest soreness
shortness of breath) are reported in 17 endurance cyclists
exposed at 31'c to 0.12 ppm 03 for l hour during very heavy (v =
88 L/min) continuous exercise (Gong et al., 1986). Although no
impairment of maximal performance is noted at 0.12 ppm, when Gong
et al. (1986) exposed the same subjects to 0.20 ppm 03 under the
same exercise conditions, he reports that both group mean FEV,
decrements (-21.6%) and symptoms (prominent cough and nausea) are
intensified and result in impairment of maximal performance
Similarly, Schelegle and Adams (1986) report that 10 highly
trained endurance athletes exposed to 0.12, 0.18, and 0.24 ppm 0,
while performing a 1-hour competitive simulation (Ve . 86 L/min)
on a bicycle ergometer show a significant (p < 0.05) increase in
the inability of subjects to complete the simulation. Even
though one subject was not able to complete the simulation at any
of the 03 levels tested, five were unable to complete at 0.18 ppm
03 and seven did not finish at 0.24 Ppm o3. Only at 0.18 and
0.24 ppm 03 are statistically significant changes reported in
PEVlf FVC, and subjective symptoms. Schelegle and Adams (1986)
report no significant changes in lung function or symptoms at
0.12 ppm 03. spektor et al. (i988b) offer substantiation of the
effects of 03 on respiratory function in populations engaging in
continuous exercise for short periods.
Although subjective statements by individuals engaged in
sports indicate possible voluntary curtailing of activities
during high-oxidant episodes, increased temperature and relative
humidity may be involved in provoking the symptoms and lung
function decrements observed in the studies above. Controlled
studies of 03 exposure have, however, demonstrated lung function
impairment and subjective symptoms which cause individuals to
0^ "" Perf°— at °'2 W °3 and above (Adams
, 1983; Folinsbee et al., 1984; Avol et al., 1984)
-------
VII-22
Because O3 may be implicated at least in part in reducing
exercise performance during periods of high oxidants and some
individuals experience this effect at levels near the 03 NAAQS,
the staff recommends that this effect of 03 be viewed as a matter
of potential public health concern and be considered in
developing a margin of safety.
4. Bronchial Reactivity and Inflammation
Bronchial reactivity is reportedly, associated with exposure
to 03 (CD, p. 10-30). Using airway responsiveness to the drugs
acetylcholine (ACh), methacholine, or histamine as a measure of
non-specific airway sensitivity, researchers have observed
increased bronchial reactivity in both healthy and asthmatic
subjects, exposed to 03 levels in the range of 0.32 to l.o ppm.
It should be noted that most of these studies were conducted on
resting or only moderately exercising individuals.
In both atopic and non-atopic subjects exposed, using a
noseclip, to 0.6 ppm O3 during intermittent exercise, bronchial
response to histamine and methacholine was greater when subjects
were exposed to O3 than to filtered air (Holtzman et al., 1979).
Symptoms of bronchial irritation in the Holtzman et al. (1979)
study were not detectable by the next day, which contrasts with
results of the Golden et al. (1979) study of healthy resting
subjects exposed to 0.6 ppm 03 reporting enhanced response to
histamine challenge for as much as 1 to 3 weeks after exposure.
Other studies of bronchial reactivity show statistically
significant enhancement when O3 exposures are as low as 0.32 ppm
but not at 0.20 ppm. Significant increases in bronchial
reactivity were reported with ACh challenge in healthy adult
subjects exposed to either 0.32 ppm or 1.0 ppm o3 (Konig et al.,
1980). Healthy subjects showed no alterations of bronchomotor
response to histamine aerosol when exposed to 0.2 ppm O3 for 2-
hr, although exposure to 0.4 ppm 03 did enhance bronchial
responsiveness to inhaled histamine (Dimeo et al., 1981). Kulle
et al. (I982b) reported significantly enhanced bronchial
-------
VII-23
reactivity to methacholine of healthy subjects exposed to 0.4 Ppm
03 as compared to filtered air.
McDonnell et al. (1987) report increased reactivity to
histamine following exposure of healthy subjects to 0.18 ppm o,
during intermittent very heavy exercise (64 L/min) for 2 hours
increased airway reactivity to methacholine following 2 hour
exposures of subjects to 0.4 and 0.6 ppm 03 during intermittent
moderate exercise is reported by Seltzer et al. (1986) to be
associated with neutrophil influx. Using bronchoalveolar lavage
(BAL), Keren et al. (1989a,b,c) observe increased inflammation
(8.2 fold increase in polymorphonuclear leukocytes, PMN's) at 18
hours post exposure in healthy adult males exposed for 2 hours to
0.4 ppm 03 during intermittent exercise, m addition to
confirmation of the Seltzer et al. (1986) findings, Koren et al
(1989a,b,c) provide evidence of stimulation of fibrogenic
processes and further suggest that the inflammatory process
initiated by 03 exposure is promptly initiated (Seltzer, et al
1986) and persists for at least 18 hours. Although the time
course of the inflammatory response has not been elucidated the
Koren et al. (I989a,bfc) research demonstrates that cells and
enzymes capable of causing damage to pulmonary tissues are
increased as a result of 03 exposure. Furthermore, proteins
which play a role in fibrotic and fibrinolytic processes are
elevated by exposure to o3. Studies by Kehrl et al. (1987/ 1989)
confirm that clearance of technetium labeled DTPA from airway and
alveoli into the bloodstream is accelerated after O3 exposure and
along with Koren et al. (1989a,b,c) suggest that this accelerated
change is due, in part, to increased epithelial permeability in
the lung. These permeability alterations are likely to be
associated with acute inflammation and may allow inhaled antigens
and other substances to more easily reach the submucosa
in an extension of a 2-hour exposure study, Koren et al.
(1989a,b,c) exposed ten healthy non-smoking males (18-35 yrs) to
0.1 ppm 03 for 6.6 hours while subjects were engaged in a
moderate to heavy (ve = 40 L/aln) interinittent exercise
-------
VII-24
similar to Folinsbee et al. (1989). Results indicated suggestive
evidence of an inflammatory ressponse in the lower airways as
evidenced by a 4.8-fold increase in percent of polymorphonuclear
leukocytes (PMN's) in the bronchoalveolar lavage fluid (p =
0.034). Importance of these preliminary results lies in the fact
that they represent evidence of inflammation in humans and
potential for damage in lower airways from O3 exposures which may
occur often in ambient air.
In conclusion, there is some evidence to suggest that
bronchial reactivity and inflammation occur at ambient
concentrations of O3. This is particularly true for subjects
exposed to O3 during exercise. Because the data base showing
effects at low concentrations of 03 is relatively new and small,
staff recommends that this information be used to develop a
margin of safety.
5. Aggravation of Existing Respiratory Disease
Some epidemiological studies suggest an association between
photochemical smog and aggravation of existing respiratory
disease. No clear evidence is available, however, from
controlled exposure or field studies to suggest that individuals
with asthma, chronic bronchitis, or emphysema have greater lung
function impairment caused by O3 or other photochemical oxidants
than healthy persons. However, individuals with preexisting
respiratory disease are considered to be especially "at risk" to
O3 exposure due to their already compromised respiratory systems
and concern that increased symptoms or pulmonary function
decrements may interfere with normal function (CD, p. 12-73).
In controlled human exposure studies, statistically
significant group mean decrements in pulmonary function were not
reported for adult asthmatics exposed for 2 hours at rest
(Silverman, 1979) or with intermittent light exercise (Linn et
al., 1978) to O3 concentrations of 0.25 ppm or lower. Similarly,
no statistically significant group mean changes in pulmonary
function or symptoms were found in adolescent, asthmatics exposed
-------
VII-25
to 0.12 ppm 03 for 1-hour at rest (Koenig et al., 1985).
Subjects with chronic obstructive lung disease (COLD) performing
light to moderate intermittent exercise show no statistically
significant group mean decrements in pulmonary function during 1-
and 2-hour exposures to < 0.30 ppm 03 (Linn et al., 1982, I983a-
Solic et al., 1982; Kehrl et al., 1983, 1985) and only small,
statistically significant group mean decrements in FEV, are
observed for 3-hour exposures of chronic bronchitics to 0.41 ppm
03 (Kulle et al., 1984). while these controlled exposure results
indicate that individuals with pre-existing respiratory disease
may not be more sensitive to 03 than healthy subjects,
experimental design considerations in these studies suggest that
the issues of sensitivity and aggravation of pre-existing
respiratory disease remain unresolved.
Several new studies compare effects of 03 exposure on
asthmatics with effects on healthy persons. Although no
differences in 03-induced symptoms are reported between 9 normals
and 9 asthmatics exposed to 0.4 ppm 03/ Kreit et al. (1988) do
report a more negative change in FEV,, FEV./FVC, and FEF2,• for
asthmatics than for normals. SRaw is not significantly increased
in normals but is in asthmatics, though it should be noted that
significant increase in SRaw occurs in asthmatics even for air
exposure. Koenig et al. (1987) compare the effects of 0.12 and
0.18 ppm 03 on 10 asthmatics and 10 nonasthmatics exposed for 30
mm. at rest followed by 10 min. exercise (32.5 L/min) . NO
significant effects are found at 0.12 ppm O3 while both groups
show increases in respiratory resistance (RT, at 0.18 ppm O3.
Koenig et al. (1987) conclude that-there are no significant
differences in response to 03 between asthmatics and
nonasthmatics. in a similar study, Koenig et al. (1988) find
that 12 healthy adolescents exposed to 0.12 ppm O3, 0.30 ppm NO2
or a combination during intermittent moderate exercise show no
significant response, while 12 asthmatic adolescents showed
marginal responses in pulmonary function for .12 ppm O3 and o 3
ppm N02 but not the combination. This leads the authors to
-------
VII-26
conclude that asthmatic subjects may be slightly more responsive.
However, replication of these observations will be required
before this suggestion can be substantiated.
Epidemiological studies do not provide a clear
concentration-response relationship between O3 and aggravation of
disease. One study (Whittemore and Korn, 1980) conducted in Los
Angeles reported increased daily asthma attack rates on cool days
and days with high oxidants and "particulates when median daily
maximum hourly oxidant levels were < 0.15 ppm; questionable
exposure assessment, lack of control for medication, pollen,
respiratory infections and other pollutants limits the use of
this study for developing quantitative dose-response
relationship.
Despite several exposure and variable control limitations,
the Houston Area Oxidants Study (Johnson et al., 1979; Javitz et
al., 1983) concluded that for the study period in which the daily
maximum hourly 03 concentrations were below- 0.21 ppm 03 near the
subjects' residence, 1) there was increased incidence of nasal
and respiratory symptoms and increased frequency of medication
use for asthmatics with increasing 03 levels; 2) FEV1 and FVC
decreased with increasing 03 and total oxidants; and 3) increased
incidence of chest discomfort, eye irritation, and malaise
occurred at high peroxyacetyl nitrate (PAN) concentrations. In a
subsequent related study, (Stock et al., 1983; Holguin et al.,
1985; Contant et al., 1985) increased probability of an asthma
attack was associated with the occurrence of a previous attack
and with exposure to increased O3 concentrations and temperature
when maximum 1-hour averages for 03 were between 0.001 and 0.127
ppm; however, other pollutants such as S02 and particulates may
have been involved.
In a series of studies conducted in a Tucson community,
adults with asthma, allergies, or airway obstructive disease
(AOD) were observed during an 11 month period in which 1-hour
daily maximum 03 concentrations were < 0.12 ppm (Lebowitz et al.,
1982, 1983; Lebowitz, 1984). After adjusting for covariables, O3
-------
VII-27
and TSP levels were significantly associated with decreases in
peak expiratory flow rate in adults with ADD, and the interaction
between O3 and temperature was significantly associated with
alterations in peak flow and symptoms in asthmatics. While these
results suggest an effect of 03 in individuals with preexisting
respiratory disease, interpretation is difficult due to the small
sample size in relation to the number of covariates and the fact
that individual exposure data were not available.
Bates and Sizto (1987) conducted a study in southern Ontario
and reported that there is a consistent summer relationship
between SO4/ 03 and temperature and hospital respiratory
admissions. The correlation between 03 and hospital respiratory
admissions was not affected by substituting the 8-hour average 03
• for mean hourly maximum 03 levels. They further suggest the
adverse, health effects studied may not be associated with either
O3 or sulfates but rather with some species not yet monitored
which they refer to as "acid summer haze." Bates and Sizto
(1988) failed to confirm association between excess hospital
admissions and 03 for June 1983, the month of the highest O3 of
any month in their analyses. In a separate study conducted in
Ontario, Raizenne and Spengler (1987) have preliminarily reported
that lung function decrements in children attending summer camp
are associated with maximum daily 03 levels as well as
temperature, fine particles, and sulfates. Over the course of a
41-day-study, Raizenne and Spengler (1988) continuously measured
03/ N02, S02, and acid aerosols (H2SO4) . In plotting individual
6-hour estimated dose for 03 and H2S04 (separately) versus
percent change in PEFR, they found a negative trend in lung
function as cumulative dose increases for both 03 and H2S04
although the slopes did not differ from zero (p > .10) when
dosimetry was used.
None of these controlled human exposure or epidemiology
studies alone demonstrates a clear relationship between exposure
to 03 and aggravation of preexisting respiratory disease. All of
the epidemiology studies cited report effects which may be
-------
VII-28
related to inhalable particle exposure and most have inadequate
characterization of exposure. However, the group of studies as a
whole supports the contention that exposures to ambient levels of
03 and other photochemical oxidants recently reported in many
cities (See Appendix A) may increase the rate of asthma attacks
for asthmatics. The staff believes that increased asthma attack
rate is an adverse effect but recognizes the limited and
uncertain data base relating this effect to 03 exposure. Thus,
in concordance with uncertainty expressed by CASAC (1986, 1987,
1988) it is recommended that data associating 03 exposure with
aggravation of existing respiratory disease be considered only in
the development of a margin of safety for the primary standard
unless further research provides more conclusive evidence.
6. Morphological Effects
Morphological effects of 03 have been reported primarily in
animal toxicology studies (See Table VII-2). For this reason it
is important to consider the differences in dosimetry and
sensitivity between humans and laboratory animals, which is
discussed in Chapter 9 of the CD. The following discussion
provides an indication of some of the structural changes which
might occur in human lungs as a result of repeated 'and long-term
exposures to O3.
Despite the differences in lung structure between humans,
dogs, monkeys, mice, rats, and guinea pigs, a characteristic
lesion occurs at the junction of the conducting airways and the
gaseous exchange tissues. In all species examined the effect is
typically damage to the ciliated and Type 1 cells and hyperplasia
of non-ciliated bronchiolar and Type 2 cells; an increase in
inflammatory cells is also observed (CD, p. 12-58). These
effects were reported after 7 days,- 8 hr/day exposures to 0.2 ppm
03 in monkeys (Dungworth et al., 1975; Castleman et al., 1977)
and in rats exposed for 7 days, 8 and 24 hrs/day (Schwartz et
al., 1976). Similar effects with different exposures were
-------
VI1-29
TABLE vn-2. MORPHOLOGICAL EFFECTS OF
Effect
OZONE IN EXPERIMENTAL ANIMALS
Ozone
concentration,ppn
Exposure
duration
Species Reference.
Damage to
ciliated and .
centriacinar
alveolar type 1
cells
J-2 8 hr/d x 7 d
°-2 8 hr/d x 7 d
0-25 12 hr/d x 6 wk
0-25 12 hr/d x 6 wk
0-5 2 hr
°-5 2 hr
Rat
Monkey
Rat
Rat
Rat
Rat
Schwartz.et al., 1976
Castleman et al., 1977
Barry et al., 1983
Crapo et al., 1934
Stephens, et al., I974a
Stephens, et al.,
Hyperplasia of
nonci Hated
bronchiolar and
alveolar type 2
cells
0.2
0.35
0.5
•0.5
0.8
0.8
8 hr/d x 7 d
continuous
for 1-8 d
8 hr/d x 90 d
24 hr/d x 180 d
continuous
for 7 d
24 hr/d x 10 d
Monkey
Rat
Monkey
Rat
Rat
Castleman et al., 197?
Evans et al., I976b
Eustis et al., 1981
Moore and Schwartz,
et al., 1973
Mouse Ibrahaaet al., 1980
Inflammation
0-2 8 hr/d x.7 d
0-2 8 hr/d x 7 d
0-5 2 hr
0.5 s hr
0.64 8 hr/d x
0.8 4 hr
1 yr
Rat
Monkey
Rat
Rat
Monkey
Monkey
Schwartz et al., 1976
Castleman et al., 1977
Stephens et al., I974a
Stephens et al., 19745
Fujinaka et al., 1985
Castleman et al., 1980
Continued
inflammation with
remodeling of
centriacinar airways
and increased collagen
0.5 24 hr/d x 14 d
0.5 8 hr/d x 90 d
0.5 24 hr/d x 180 d
0-64 8 hr/d x 361 d
1-0 8 hr/d x 18 mo
Rat
Rat
Rat
Monkey
Oog
Last et al., 1979
Boorman et al., 1980
Moore and Schwartz, 1881
Last et al., I984b
Freeman et ah, 1973
-------
VII-30
reported in rats (0.26 ppm, 6 hr, endotracheal tube, Boatman et
al., 1974), mice (0.5 ppm, 35 days, Zitnik et al., 1978), and
guinea pigs (0.5 ppm, 6 months, Cavender et al., 1978).
Inadequate data limit quantitative comparisons of human health
effects with these reported for monkeys and rats, but a rough
equivalency of responses has been observed under similar exposure
conditions between species. Because all species tested show
similar morphological responses to O3 exposure, there is no
reason to believe that humans exposed to O3 would not respond
similarly, although such effects may not necessarily occur at the
same O3 exposure concentrations or averaging times of exposure.
Changes in lung structure of monkeys and rats tend to
decrease after extended exposure to 03/ although structural
changes in the centriacinar region have been reported after long-
term exposures of rats (Boorman et al., 1980; Moore and Schwartz,
1981; Barry et al.', 1983; Crapo et al., 1984), monkeys (Eustis et
. al., 1981; Fujinaka, 1984; Fujinaka et al., 1985), and dogs
(Freeman et al., 1973). While cell repair begins within 18 hours
of exposure (Castleman et al., 1980; Evans et al., 1976 a,b,c;
Lum et al., 1978), cell damage continues throughout long-term
exposures but at a slower rate (CD, p; 12-47) .'
Increases of lung collagen content in the centriacinar
interalveolar septa may be indicative of lung fibrosis and, thus,
of structural damage. Thus morphological effect has been
associated with biochemical changes in activity of prolyl
hydroxylase and in hydroxyproline content, both related to lung
collagen content and fibrosis (Last et al., 1979; Bhatnagar et
al., 1983). Lung collagen content increased after exposures to
03 concentrations as low as 0.5 ppm (Last et al., 1979) and
continued to increase during long- term exposures (Last et al.,
1984b). Weanling and adult rats exposed for 6 and 13 weeks,
respectively, and young monkeys for one year to < i.o ppm 03 also
showed increased collagen content in the lungs (Last et al.,
1984b). In the Last et al., (l.984b) study, examination of some
of the exposed weanlings and controls at six weeks post-exposure
-------
VII-31
indicated a continued increase in lung collagen content, a result
demonstrating that damage continued to occur during the post-
exposure period. Also, there was little apparent difference in
collagen levels of rat lungs for those animals exposed
continuously vs. intermittently. This would suggest that
intermittent periods of clean air were insufficient for recovery.
The centriacinar inflammatory process also continues during
long-term 03 exposures and appears to be 'related to remodeling of
the centriacinar airways (Boorman et al., 1980; Moore and
Schwartz, 1981) and to increased lung collagen which appears
mainly in the centriacinar regions (Last et al., 1979; Boorman et
al., 1980; Moore and Schwartz, 1981; Last et al., I984b). in
addition there is morphometric (Fujinaka et al., 1985),
morphologic (Freeman et al., 1973), and functional (Costa et al
1983; wegner, 1982) evidence of distal airway narrowing.
Several new animal toxicology studies have focused on effect
of prolonged exposure and appropriate selection of an averaging
time associated with respiratory damage from 03 exposure
Rombout et al. (1989) conducted a study in which rats were
exposed to 03 levels of 0.13 to 2.0 ppm and time of exposure
ranged from l to 54 hours. With time course of protein and
albumin concentration as the endpoint, Rombout et al. (1989)
concluded that the response model indicates a strong influence of
time on the response, that increases with increasing
concentration. Van Bree et al. (1989) exposed rats to 0.4 ppm
for 12 hours on 3 consecutive nights and found that acute
pulmonary injury and inflammation, as measured by lung lavage
protein increase and neutrophil influx, were induced by the first
exposure and showed full reversibility in spite of continued
exposure. However, biochemical indices for cell proliferation
and lung tissue repair continued to show increases with
subsequent exposure. A polynomial model has been developed by
Costa et al. (1989) to depict lung injury from the interaction of
03 concentration (C) and exposure duration (T). The model was
based on lung fluid protein values in rats exposed to O3 levels
-------
Vi:C-32
between 0.1 and 0.8 ppm for 2, 4 or 8 hours and suggests that
impact of T on 03 toxicity appears to be C-dependent. Further
development of these models is expected as efforts continue to
improve understanding of c x T relationships.
While the distal airway narrowing and lesions at the
junction of the conducting airways and gaseous exchange region
are similar to the changes which have been found in lungs of
cigarette smokers (Niewohner et'al., 1974; Cosio et al., 1980;
Hale et al., 1980; Wright et al., 1983), there is no evidence of
emphysema in the lungs of animals exposed to only O3 based on the
currently accepted definition for emphysema cited in the CD (p.
9-49). Many of the lung structural changes that have been
reported for long-term exposures to < 1.0 ppm 03 in several
different species are considered adverse and relevant to standard
setting if demonstrated in humans at ambient exposure
concentrations of 03. Although the lowest O3 concentrations and
shortest averaging times which could produce structural changes
in human lungs is uncertain at this time, the staff concludes
that there is a need to protect the public from O3 exposures
which may induce such adverse effects. Further analysis is
needed to clarify how animal data should be used to estimate lung
structural changes in humans caused by O3 exposures. The staff
suggests that these data be used in developing a margin of safety
for the primary standard.
7. Effects of Ozone on Host Defense Mechanisms in
Experimental Animals
Respiratory systems of mammals are protected from bacterial
and viral infections by the closely interrelated particle removal
(both mucociliary and phagocytic) and immunological defense
systems. Numerous factors such as poor nutrition, preexisting
disease, and environmental stress may influence or alter these
host defense systems of individuals sufficiently to permit
development of respiratory infections. Animal studies have
-------
VII-33
indicated that exposure to O3 is one of those factors (See Table
VII-3). Both in viva (live animal) and in vitro (isolated cell)
studies have demonstrated that O3 can affect the ability of the
clearance and immune systems to defend against infection
increased susceptibility to bacterial infection has been reported
in mice at 0.08 to 0.10 ppm 03 for a single 3-hour exposure
(Coffin et al., 1967; Ehrlich.et al., 1977; Miller et al., I978a)
and at o.io ppm 03 for subchronic exposures (Aranyi et al
1983). several related alterations of the pulmonary defenses
caused by short-term and subchronic exposures to 03 include- i)
impaired ability to inactivate bacteria in rabbits and mice
(Coffin et al., 1968; Coffin and Gardner, 1972; Goldstein et al
1977; Ehrlich et al., 1979); 2) delayed mucociliary clearance
(Phalen et al., 1980; Frager et al., 1979; Kenoyer et al., 1981-
Abraham et al., 1980); 3) immunosuppression (Campbell and
Hilsenroth, 1976; Aranyi et al., 1983; Fujimaki et al., 1984)- 4,
significantly reduced number of pulmonary defense cells in
rabbits (Coffin et al., 1968; Alpert et al., 197!); and 5)
impaired macrophage phagocytic activity, less macrophage
mobility, more fragility and membrane alterations, and reduced
lysosomal enzymatic activity (Dowell et al., 1970; Hurst et al
1970; Hurst and Coffin, 1971; Goldstein et al., I971a,b;
Goldstein et al., 1977; Hadley et al., 1977; McAllen et'al
1981; witz et al., 1983; Amoruso et al., 1981). some of these
effects have been shown to occur in a variety of species
including mice, rats, rabbits, guinea pigs, dogs, sheep, and
monkeys.
Other studies indicate similar effects for short-term and
sub-chronic exposures of mice to 03 combined with pollutants such
as S02, N02, H2S04 and particles (Gardner et al., 1977; Aranyi et
al. 1983; Ehrlich, 1980, 1983; Grose et al., 1980, 1982; Phalen
et al., 1980; Goldstein et al., 1974). similar to human
pulmonary function response to 03, activity levels of mice
exposed to O3 has been shown to play a role in determining the
-------
711-34
TABLE VII-
EFFECTS OF'OZONE ON HOST DEFENSE MECHANISMS IN EXPERIMENTAL ANIMALS
_.. • ozone Exoosura
Effect concentration^ duration
Increased
susceptibility to
bacterial
respiratory
infection
Impaired ability
to inactivate
bacteria and
viruses
Delayed rauco-
ciliary clearance
Iiwnunosuppression
Impaired macrophage
function
0.08
0.08
0.1
0.1
0.1
0.5
0.5
0.5
0.4
0.3
1.0
1.2
0.1
0.59
0.8
0.25
0.25
0.5
0.5
0.5
0.62
0.99
1.0
3 hr
** III
3 hr
3 hr
5 hr/d x 103 d
3 hr/d x 3 mo
3 hr
** llf
3 hr
2 hr
4 hr
4 hr
2 hr
4 hr
5 hr/d x 90 d
continuous
for 36 d.
continuous
for 14 d
3 hr
3 hr
8 hr/d x 7 d
2 hr
3 hr
4 hr
17 hr
4 hr
Species
Mouse
Mouse
Mouse
Mouse
Mouse
Rabbit
Rabbit
Rat
Rat
Rat
Sheep
Rat
Mouse
Mouse
Mouse
Rabbit
Rabitt
Rabbit
Rat
Rabbit
Mouse
Mouse
Rat
Reference
Coffin et al., 1967
Miller et al., l978^
fhrlich et al., 1977
Aranyi et a?., 1983
Ehrlich et al.,
Coffin et al., 1968
Coffin and Gardner,
Goldstein et al., 1
Kenoyer et al.,
Phalen et al., 1980
Abraham et al., 1985
Frager et al., 197S
Aranyi et al., 1985
Campbell and Hilse?iru
1976
Fujimaki et al., 1984
Hurst et al., 1970
Hurst and Coffin, l§?i
Oowell et al., 1970
Goldstein et al., 13?7
Hadley et al., 1977
Goldstein et al., 1S71
Goldstein et al., 1971
McAllen et al., 1981
-------
VII-35
lowest effective concentration which alters the immune defenses
(Tiling et al., 1980).
Although this large body of evidence clearly demonstrates
that short- term and subchronic exposures to O3 can impair the
inunune defense systems of animals, technical and ethical
considerations have limited similar research on human subjects.
Thus inferences have been drawn and models developed to assess
the relevance of animal data to humans. For example, animal
endpoints other than increased mortality should be appropriately
compared to increased morbidity in humans, such as the increased
prevalence of respiratory illness in the community (CD, p. 12-
50) .
Based on current understanding of physiology, metabolism
and immune defenses, it is generally accepted that the basic '
mechanisms of action associated with defense against infectious
agents are similar in humans and animals. Similarities between
human and rodent antibacterial systems have been discussed in
detail (Green, 1984), but differences in lung structure and
biological response will inevitably cause differences'in
dosimetry, sensitivity, and endpoints of response. in addition
other factors such as preexisting disease, nutrition, presence of
other pollutants and environmental stresses (e.g. No2, S02 PM
high temperature or humidity) can influence the effect of
exposure to O3 and infectious agents. Despite the differences in
related factors which make precise estimation of human response
from animal data difficult, it is reasonable to hypothesize that
humans exposed to 03 could experience impairment of host defenses
(CD, p. 12-50). The staff recommends that these data be used in
developing a margin of safety.
8. Extrapulmonary Effects
Extrapulmonary effects which have been demonstrated in
hvomans or laboratory animals following exposure to O3 include
alterations in red blood cell morphology and enzyme activity
cytogenetic effects in circulating lymphocytes, and subjective
-------
VII-36
limitations in vigilance tasks. Additionally, animal toxicology
studies provide limited evidence for cardiovascular,
reproductive, teratological, endocrine system, and liver
metabolism effects. This wide variety of extrapulmonary effects
is probably caused by oxidative reaction products of O3. Due to
the high reactivity of 03 with biological tissue, mathematical
models predict that only a small fraction of 03 actually reaches
the circulatory system (Miller et al., 1985).
Of the extrapulmonary effects reported, cytogenetic and
mutational effects are probably the most controversial (CD, p.
12-51). Statistically significant increases in frequency of
sister chromatid exchanges (chromosomal alterations) have been
caused by in vitro exposures to 0.25 ppm for 1 hour (Guerrero et
al., 1979), suggesting a mutagenic response. However, in vitro
responses are not generally extrapolatable to in vivo responses,
and homeostatic mechanisms are not represented. Therefore,
isolated in vitro exposure studies have not been used to provide
estimations of risk for exposure to 03. In vivo animal studies
have shown significant increases in the number of chromosomal and
chromatid aberrations following 4 and 5 hour exposures to 0.2 and
0.43 ppm 03, respectively (Zelac et al., I971a,b,; Tice et al.,
1978). However, other animal studies (Gooch et al., 1976) and
controlled human exposures to O3 levels as high as 0.5 ppm have
shown no significant cytogenetic effects attributable to O3 (Merz
et al., 1975; McKenzie et al., 1977; McKenzie, 1982; Guerrero et
al., 1979), and epidemiology studies provide no evidence of
chromosomal changes induced by ambient 03 (Scott and Burkart,
1978; Magie et al., 1982). Thus, while the animal studies are
suggestive of possible cytogenetic effects from 03, human studies
have not demonstrated such effects for realistic ambient
exposures to O3.
Limited hematological and serum chemistry effects data
indicate that 03 or one of its reactive products can cross the
blood gas barrier and interfere with biochemical mechanisms in
human blood erythrocytes and sera (CD, p. 1-132). Much of the
-------
VII-37
data is In Yi£ro and useful primarily in studying mechanisms of
action. Ozone-induced effects such as pentobarbital- induced
sleep time alterations and hormone level changes have been
reported (Graham et al., 1981) and cause-effect hypotheses made
With regard to other extrapulmonary effects, exact mechanisms
remain to be elucidated and physiological significance for humans
remains uncertain. Nonetheless, the body of this evidence
suggests that 03 can cause effects distant from the lungs in
animals, and hence possibly in humans. it is the staff's
recommendation that these data be used in developing a margin of
safety.
B. Pulmonary Function and Symptom Health Risk Assessment
This section summarizes an assessment (Hayes et al., •
1987a,b; Whitfield, 1988) of risks for two categories of'effects
(pulmonary function and symptoms) associated with attainment of .
alternative 1-hr 03 NAAQS. These health risk estimates
characterize acute responses based on results of l- .and 2-hour
controlled human exposure studies reviewed in the CD. While
recent controlled human exposure research and field studies
suggest that longer exposures (6-8 hours) at less strenuous
levels of exercise may also pose a health threat, EPA is awaiting
completion of additional research prior to extending the risk
assessment to these subchronic exposures. EPA also plans to
pursue risk assessment for chronic, possibly irreversible,
effects on the lung observed in various animal toxicology'studies
upon completion of important dosimetry and lung injury research
currently underway.
1. Overview of Lung Function and Symptom Risk Assessment
The objective of the pulmonary function/symptom risk
assessment is to estimate the magnitude and extent of risk to the
most susceptible population for these effects (i.e., heavily
exercising individuals) while characterizing, as explicitly as
possible, the range and implications of uncertainties in the
-------
VI1-3 8
existing scientific data base. While the risk assessment
estimates should not be viewed as demonstrated health impacts
they do represent EPA's best estimate as to the possible
magnitude and extent of risk for these effects given the
available scientific information. Although it does not cover all
health effects caused by 03, the risk assessment is intended as a
tool that may, together with other information presented in this
staff paper and the CD and CDS aid the Administrator in judging
which alternative 03 NAAQS provides an adequate margin of safety.
Risk estimates for a number of urban areas, the aggregated total
for those areas, and the methodology used to generate these
estimates are described in detail in Hayes et al. (1987a,b) and
Whitfield (1988).
The three major types of inputs to the risk assessment are:
(1) exposure-response relationships used to characterize
pulmonary function and symptom effects of O3 exposure in heavily
exercising individuals developed from data obtained in three
controlled human exposure studies: Avol et al. (1984), Kulle et
al. (1985), and McDonnell et al. (1983);
(2) distributions of O3 hourly concentrations upon
attainment of alternative NAAQS and under the "as is" situation
obtained from the 03-NEM project (Paul et al., 1986); and
(3) distributions of population exposure, both in terms of
people exposed and occurrences of exposure, upon attainment of
alternative 03 NAAQS and under the "as is- situation obtained
from the O3-NEM analysis (Paul et al., 1986).
Chapter V of this staff paper presents an overview of the O3
exposure analysis methodology and summarizes exposure estimates
for the aggregated 8-urban area population for alternative air
quality scenarios. The pulmonary function and symptoms risk
assessment considers the same four scenarios used in the exposure
analysis: 0.08, O.io, and 0.12 ppm, daily maximum l-hour 03
NAAQS with an expected exceedance rate of once per year and the
"as is" situation.
-------
VII-39
Risk estimates have been developed for both pulmonary
function (as measured by FEVl decrement) and lower respiratory
symptoms (specifically, cough, chest discomfort, and as a group
of symptoms). A health effect, or endpoint, can be defined in
terms of a measure of biological response and the amount of
change in that measure thought to be of concern. To permit EPA
decision makers and the public to examine the implications of
alternative definitions of adversity, risk estimates are
presented for a range of different health endpoints:
(1) Pulmonary function. TWO different level of 03-induced
FEVi decrement: (a) at least 10 percent and; (b) at
least 20 percent.
(2) Lower respiratory symptoms. Cough, chest discomfort,
and lower respiratory symptoms as a group for two
different severity levels: (a) any symptoms (mild,
moderate, or severe) and (b) moderate or severe
symptoms.
Estimated exposure-response relationships derived from each
of the three health studies (Avol et al., 1984; Kulle et al
1985; and McDonnell et al., 1983) are used separately to derive
independent risk estimates. While the three studies are similar
in enough respects (e.g., health endpoints, young heavily
exercising healthy subjects, similar 1-2 hour 03 exposures) to
make useful comparisons, there are enough differences in
experimental protocol (e.g., i-hour continuous exposure in Avol
vs. 2-hour intermittent in Kulle and McDonnell and differences in
exact exercise level and exposure concentration) to make
statistical combination of these data bases undesirable
Estimated exposure-response relationships for FEV, decrement and
symptomatic effects used in the risk assessment are shown in
Figures VII-2 through VII-4 earlier in this staff paper. These
relationships were derived by fitting a four-parameter logistic
function to each data set and subtracting out the effect of
exercise alone based on the response at 0 ppm (see Chapter 3 of
Hayes et al., i987a for details).
-------
VI1-40
Uncertainty attributed to sampling error due to sample size
considerations in the exposure-response relationships is
reflected in the pulmonary function and symptom risk estimates.
This uncertainty was estimated using a Bayesian approach
described in Chapter 3 of Hayes et al. (I987a). other sources of
uncertainty due to differences in experimental protocol subject
population, measurement error, etc. have not been quantitatively
addressed. The calculation and presentation of separate risk
estimates for each of the three data sets provides a rough
picture of the degree of uncertainty due to these other factors.
Two distinct types of risk measures are provided by the O3
health risk assessment. The first measure, "benchmark risk,"
focuses on the probability or risk of unhealthful air. The
second measure, "headcount risk," focuses on the number of people
affected and number of incidences of a given health effect
considering individuals' personal exposures as they go about
their daily activities (e.g., from indoors to outdoors, moving
from place to place, and engaging in different exercise levels).
More specifically, benchmark risk is the probability that
pollutant levels will be sufficient to trigger a defined health
endpoint in at least a fraction of the sensitive population if
they were exposed. The second measure, headcount risk, assesses
the number of people or the percent of the sensitive population
that would be adversely affected given the normal movement and
activity patterns of the population of interest. Headcount risk
also provides estimates of the number of incidences of adverse
effects there would be. Staff believe that these risk measures
taken together capture two important perspectives that should be
considered in selecting an O3 primary standard that provides an
adequate margin of safety.
2. Benchmark Risk Results
For the 03 pulmonary function and symptom risk assessment
the benchmark risk is defined as the probability that, upon just
attaining a given 03 NAAQS or under "as is" conditions, the
-------
maximum hourly concentration will equal or exceed the level that
would cause i, 5, or 10% of the heavily exercising population to
exhibit particular health endpoints 1 or more times per year
The benchmark risk is estimated assuming the entire sensitive
population is exposed while heavily exercising (defined as Regime
3 in 03-NEM which is 3 or more 10 minute periods in an hour at a
> 44 L/min). As indicated in the risk report:
^
thi
Benchmark risk is calculated. by combining exposure-response
relationships and probability distributions of hourly O3 ambient
concentrations, either based on observed air quality data during
1981-1984 (the particular year depending on area) for the "as is"
case or based on conditions of exact attainment of alternative
NAAQS (0.08, 0.10, and 0.12 ppm, daily max hourly average 1
expected exceedance per year) . The benchmark risk model and more
detailed discussion of the inputs to the model are contained in
Hayes et al. (I987a,b) .
Benchmark risk estimates are calculated for the 10 urban
areas shown in Table V-i and are presented in Hayes et al
C1987a,b). Due to differences in the degree of non-attainment of
the 03 NAAQS among the ten areas (ranging from Los Angeles to
Tacoma) the benchmark risk associated with the "as is" scenario
varies significantly among the urban areas. However, benchmark
risk estimates for a particular endpoint, data base, and choice
of benchmark case are very similar for a given NAAQS -attainment
scenario across most of the 10 urban areas. Selected results for
-------
VII-42
a single urban area, St. Louis, are presented here which are
generally representative of benchmark risks for the other urban
areas.
Figure VII-5 presents benchmark risks for two different
degrees of FE^ decrement: at least 10% in the top chart and at
least 20% in the bottom chart. Risk estimates are shown as three
groups of bars, one group for each health data set used to
generate exposure-response relationships. Each group consists of
four bars: one bar for each of the three alternative 03 NAAQS
analyzed (0.08, 0.10, and 0.12 ppm, daily max hourly average, 1
expected exceedance per year) and one bar for the "as is"
situation. The height of the bar represents a 1%-benchmark case,
that is, "the probability that, at least r percent or more of the
heavy exercisers would respond with the specified health endpoint
than under background conditions, one or more times during the o'3
season" (Hayes et al., 1987a,b). The top of the vertically-
shaded portion corresponds to a 5%-benchmark case and the top of
the slant-shaded portion corresponds to a 10%-benchmark case.
As an example of how to read the benchmark risk figures,
consider the > 10 percent FEVl decrement case in Figure VII-5
(top):
Under exact attainment of the current 03 NAAQS (0.12 ppm) ,
the probability that at least a 10 percent FEVl decrement
would occur in at least 5 percent more heavy exercisers than
at background, at least once during the 03 season, is nearly
H 2 * V01 data Set' about °'1 for ^e Kulle data set,
and about 0.5 for the McDonnell data set (Hayes et al
1987a,b, p. 4-7). *
Using the same approach, Figure VII-6 presents benchmark
risk estimates for any chest discomfort (top chart) and for
moderate or severe chest discomfort (bottom chart). It should be
noted that risk estimates for the Avol data bases are for an
aggregate lower respiratory symptom score that reflects chest
discomfort, cough, and other lower respiratory symptoms.
Figures VII-5 and VII-6 illustrate the importance of certain
key decision parameters in interpreting risk estimates. Both
-------
VII-43
1.0
£ 0.6
jt
o
I 0.4
0.0
BENCHMARK RISK
=> 10%, Heavy Exercise —
.12
1
•
Avol
A«-ta
.tz.
.10
.10
M
1
1
Kulta McDonnell
Alternative Ozone Standards (ppm)
* IX Uer. R«*p«fMMng
than of Baekareund
(Total Bar
*5X Mora ft«tpondlna
(N»rt Un« Down)
*10X Uer.
BENCHMARK RISK
D(FEV1) => 20%; Heavy Exercise
0.8
jt
m
•c u.o
!
j...
0.2
0.0
48
-
-
„.
TTTT
.10
. s
^§§
I
i
1
Oo^
\s\^
SXSS
£ ^
.10
J>a ]
\ 1 1 1 1
.12
••^^H
[mil
A«-h
1 1|
Ul
^
JOB
1 — 1
1
.10
.12
(III
1
1
\ssv
NSN
ooo
\\V
n *1X Uor* Responding
than at Background
(Total Bar Height)
|Q] J5S More Responding
(Next Line Down)
J£3 *10S More Responding
Avol
Kulle
McDonnell
Alternative Ozone Standards (ppm)
St.Louls
FIGURE VII-5.Benchmark r1sk in St> Louis for FEV1 decrements Qf
^ 10% and > 20%, under heavy exercise, for three exposure-response
data sets (Avol, Kulle, and McDonnell).
Source: Hayes et al., 1987
-------
1.0
0.6
0.2
BENCHMARK RISK
Any Chest Discomfort. Heavy Exercise
_.t2 »«-*«
0.0"™
1
I
.12
.10
I
I
.12
.10
McDonnell
AlUmatfv* Ozone Standard* (ppm)
* 1X Mar* Responding
tho« of Boeicareund
CTetal Bar Height)
*9X Mere Responding
(Nert 4lne Down)
*10X Mere Responding
Aval data Is for lower
respiratory symptoms
BENCHMARK RISK
— Moderate or S«ver« Chest Discomfort, Heavy Exercise
1.0
0.6
I""
0.2-
0.0
.12
.10
jM
.12
.13
.10
1
%
Avol
Ku|to
McDonnell
Alternative Ozone Standards (ppm)
St. Louis
i * 1X More Responding
than at Background
(Total Bar Height)
*SX Uors Responding
(Went Un« Down)
klOX More Responding
Aval data Is for lower
respiratory symptoms
FIGURE vn-6.Benchmark risk in St. Louis for chest discomfort
symptoms (any and moderate/severe), under heavy exercise, for
three exposure-response data sets (Avol, Kulle, and McDonnell)
Source: Hayes et al., 1987
-------
VII-4 5
selection of endpoint definition (e.g., >io% or >20% decrement in
PEVi) and n%-benchmark case (e.g., at least 1, 5, or 10% of the
sensitive population which must experience the health effect)
significantly affect the risk estimates. In addition, comparing
risks across health data bases results in significant
differences. FEV^ risk estimates for a given benchmark
definition, say 1%, are highest for the Avol data base and lowest
for the Kulle-data base, in contrast, for moderate/severe chest
discomfort and lower respiratory symptoms, risk estimates are
highest for the McDonnell data base and lowest for the Avol data
base.
3. Headcount Risk Results "
For the 03 risk assessment headcpunt risk is characterized
by calculating- the expected headcount, that is, the expected
number of people experiencing a defined effect and the expected
number of incidenbes of that effect projected to occur during the
03 season, given that a particular NAAQS is just attained or
under "as is" air quality conditions. Headcount risk estimates
the risk posed to the sensitive population (in this case
individuals engaged in heavy exercise) as they, go about their
normal activities.
A major input to the headcount risk model is the series of
population exposure distributions for "as is" air quality and the
three alternative NAAQS analyzed by EPA. Using available
exposure estimates, headcount risk estimates were calculated for
eight of the ten urban areas listed in Table V-i. NO headcount
risk estimates are calculated for Chicago and New York, since EPA
is in the process of revising these estimates. Appendix C of
Hayes et al. (1987a,b) presents headcount risk estimates for each
of the 8 urban areas.
Risk estimates presented here are an aggregation of the
results from the 8 urban areas. The total population living in
the 8 urban areas is approximately 25.9 million. Of this
population, 9.3 million people (or 36%) are estimated to exercise
-------
VII-46
sufficiently heavily to reach NEM Exercise Regime 3 (3 or more
10-minute periods in an hour at heavy (Ve > 44 1/min) which could
include one 10-minute period at very heavy (Ve > 64 1/min)
exercise).
Figure VII-7 presents risk estimated for each of the three
health data sets for two levels of FEV1 response: at least 10%
decrement on top and at least 20% decrement on the bottom for the
three alternative O3 NAAQS analyzed (0.08, o.io, and 0.12 ppm,
daily max hourly average, 1 expected exceedance) and the "as is"
situation. Similarly, Figure VII-8 presents expected headcount
estimates for any chest discomfort (top chart) and moderate or
severe chest discomfort (bottom chart). Again, it should be
noted'that risk estimates for the Avol data base are for lower
respiratory symptoms which include chest discomfort as well as
cough and other symptoms. The complete headcount risk results
are contained in Hayes et al. (1987a,b), including risk estimates
for cough and for expected number of incidences of each of the 6
health endpoint definitions examined, as well as estimates for
each of the 8 urban areas examined.
In both Figures VII-7 and VII-8 the expected headcount in
millions of people is indicated by the filled-circles and
triangles and 90% credible intervals are displayed as vertical
bars. The credible interval reflects uncertainty in the expected
headcount attributed to uncertainty in exposure-response
relationships due to sample size considerations. The 90%
credible interval can be interpreted as a 0.9 probability that
the headcount lies within the interval, and a O.io probability
that it falls outside of the interval.
As with the benchmark risk results, for FEVX decrement the
Avol data base results in the highest headcount estimates and
Kulle shows the lowest estimate and for chest discomfort the
McDonnell data base results in the highest headcount estimate and
Avol shows the lowest estimate. One also observes a greater
reduction in headcount estimate going from the "as is" situation
to the 0.12 ppm NAAQS than in going from 0.12 ppm to a 0.10 or
-------
23
| 15.0
O
0
3: 10.0
T)
X
u
S.O
0.0
VII-47
HEADCOUNT RISK (People)
D(rEV1) => 10%. Heavy Exercise
Ai
A.-*
•*
T:
•
. mi
X
<
.:
«
4
0
•
<
-<•
Avcl »•"• McDonn.M
Alternative Ozon* Standards (ppm)
23
§ 13.0
O
O
X 10.0
O
4> S (
a 5-(
X
LJ
0.0
HEADCOUNT RISK (People)
— 0(FIV1) => 20%, Heavy Exercise
Avo1
A.-4,
A*
At-t.
.12
.10 r
Til r-?f
M T° f '
-ta
WcDonn.ll
Alternative Ozone Standards (ppm)
Eight City Aggregation
-------
HEADCOUNT RISK (People)
Any Chest Discomfort, Heavy Exercise
0 2S.O
X
| 20.0
O
0
T3 1S.O
O
«
I
•O 10.0
"o
0
0- 5.0
X
Ld
0.0
A*
A*
'
•
-te
A*
.1
JO '
.u -2?
.10 f T
TTl I-
i <
2 '
i
H«
'
.
' M
0
2
-ta .
Aval
Kulte
UcOonn«ll
Alternative Ozooe Standard* (ppm)
MX Crwdlbto
Interval
Avol data U for low«r
rwiplratofy •ymptom*
O
x
T>
»
T)
e>
Q.
X
HEADCOUNT RISK (People)
— Moderaia or Savere Chest Discomfort, Heavy Exercise —
23.0
Headcouni
8
b
5
b
0.0
.12
.10
A*-to
A.-*.
Av«<
McDonnell
Alternative Ozone Standards (ppm)
Elghi City Aggregation
Avol data l« for lower
•ymptoms
FIGURE vn-8. Expected headcount (chest discomfort) aggregated for
eight U.S. urban areas (number of heavily exercising people responding
during the ozone season).
Source: Hayes et al.; 1987
-------
VII-49
0.08 ppm NAAQS. However, one must note that the expected
headcount in Figures VII-7 and VII-8 for the "as is" case areas
dominated by LOS Angeles• contribution to headcount risk and
should not be viewed as representative for the nation. For
example, in Figure VII-7, the expected headcount for FEV
decrement > 10% based on the Avol et al. (1984) study is about
1.7 million people, of which 1.3 million people are from Los
Angeles. Headcount risk estimates can also be expressed as a
percentage of the heavily exercising population. Table VII-4
presents the range of mean percentages of heavy exercisers
responding 1 or more times per year upon attainment of 3
alternative NAAQS and for the "as is- situation for each of the 6
health endpoint definitions examined. The range is due both to
variation in the shape of air quality distributions in the 8
urban areas and to differences in exposure-response relationships
based on the three health data bases (Avol, Kulle, and
McDonnell). Again, it is important to recognize the influence of
the Los Angeles estimates when considering the »as-is» case
Removing Los Angeles from the "as-is" column would for example
reduce the upper end of the range from 36.3% to 14.0% for FEV >
i n& 1 —
10%.
Health
Bndpoint
TABLE VII-4 Percent of Heavy Exercisers Responding
Under Alternative Air Quality Scenarios
Mean Percentages of Heavy Exercisers
Responding Under NAAQS Attainment
Across Eight U.S. Urban Areas
0.12 ppm 0.10 oom 0^,08
Responding ,»,-!•.)
>20%
Cough Any
Mod/Sev
Chest Any
Discom- Mod/Sev
fort
1.1 to 36.3
0.5 to 20.7
3.1 to 41.8
0.2 to 27.7
2.7 to 20.4
0.2 to 18.7
0.5 to 4.2
0.2 to 2.1
0.9 to 18.8
0.1 to 4.8
0.8 to 8.5
0.1 to 7.1
0.1 to 3.0
0.0 to 1.3
0.2 to 15.2
0.0 to 2.9
0.1 to 7.0
0.0 to 5.8
0.0 to 1.9
0.0 to 0.7
0.1 to 10.1
0.0 to 1.4
0.1 to 4.8
0.0 to 3.8
-------
VII-50
4. Caveats and Limitations
A number of assumptions and limitations should be kept in
mind in interpreting results of the pulmonary function and
symptom risk assessment. Extrapolation of results of the Avol,
Kulle, and McDonnell data sets to the heavily exercising
population at large is affected by a number of considerations.
These include the following:
(1) Exercise group. The CD defines heavy exercise as lung
ventilation rates of 44-63 L/min and very heavy exercise as >64
L/min. The Avol study (57 L/min) corresponds to the mid-to-upper
portion of the heavy exercise range (54 L/min is the mid-point).
The McDonnell study (65 L/min) falls just in the lowest portion
of the very heavy exercise range, with the Kulle study (68 L/min)
some 5 percent higher. Exposure-response relationships obtained
from the three data sets have been assumed to represent exposure-
response in all heavy exercisers. Additional research is needed
to determine the extent to which exposure-response relationships
observed at 57, 64, and 65 L/min change at lower exercise levels
such as 44-56 L/min. To the extent that exercise rates among
many heavy exercisers are lower than in the subject studies, lung
function and symptom headcount risks may be overstated. The
extent to which this is so is unknown and must be regarded as an
additional risk assessment uncertainty (Hayes et al., 1987a,b).
Since the December 1987 CASAC meeting, an addendum to the
lung function and symptom risk assessment has been prepared
(Whitfield, 1988) which combines the estimated exposure-response
relationships from the same three studies discussed previously
with the NEM very heavy exercise exposure estimates to generate
expected headcount estimates. The resulting expected headcount
estimates for very heavy exercisers are approximately 100-150
times smaller in magnitude than the heavy exercise expected
headcount estimates. This is due principally to the fact that
the population of very heavy exercisers is much smaller (by about
two orders of magnitude) than the population of heavy exercisers.
-------
VII-51
(2) Interaction between o^ and other pollutants. The health
studies used in the risk assessment involved only 03 exposure.
It is assumed that the health effects of interest in the real
world where other pollutants are present are due solely to O3.
While controlled human exposure studies have not consistently
demonstrated enhancement of respiratory effects for O3 when
combined with SO2/ NO2, CO, H2SO4/ or other aerosols, there is
some animal toxicology research suggesting additive or possibly
synergistic effects. (See Section VII.C.5 for further discussion
of interactions with other pollutants).
(3) Reproducibility of Q^-induced response. It is assumed
that 03-induced respiratory responses are reproducible for
individuals. The CD cites both Gliner et al. (1983) and
McDonnell et al. (1985a) in concluding that respiratory effects
of 03 are highly reproducible. Analysis of the Avol and Kulle
data sets by Hayes et al. (l9-87b) also supports the
reproducibility of individual responses.
(4) Age. The risk assessment has been applied to all
heavily exercising individuals regardless of age. However,
controlled human exposure and recent field epidemiology studies
in children have reported pulmonary function, but not
symptomatic, effects for O3-exposures. Therefore, the headcount
symptomatic effect estimates which rely on population exposures
that include children may overstate symptom headcount estimates.
Pulmonary function risk estimates are not affected, and the lack
of apparent symptoms does not mean that biological processes
associated with O3 symptoms in adults are not also present in
children.
(5) Sex. There is some limited evidence that women may be
more responsive to O3 in terms of FEVj impairment than men. The
data sets used here to derive exposure-response relationships
involved mostly male subjects. To the extent that women are more
responsive than men, risk estimates may be understated.
(6) Smoking status. There is some limited evidence that
smokers may be less responsive to 03 than nonsmokers. The risk
-------
VII-52
assessment was applied to all heavy exercisers regardless of
smoking status. To the extent that smokers are less responsive
than nonsmokers, risk estimates may be overstated.
(7) Attenuation or enhancement-. Of response. The risk
assessment assumes that the 03 -induced response in any particular
hour is not affected by previous O3 exposure history. The extent
of attenuation and/ or enhancement of O3 -induced responses due to
previous O3 exposures cannot be addressed quantitatively and must
be regarded as an additional uncertainty in interpreting the risk
estimates.
Exposure and air quality estimates. A major input to
the headcount risk is the O3 exposure analysis estimates for the
heavily exercising population. Uncertainties about human
activity patterns and the procedures used to estimate 03
concentrations upon attainment of alternative standards, as well
as other uncertainties about the exposure analysis model and
inputs to the model, must be regarded as additional uncertainties
in interpreting the headcount estimates. As noted in Chapter V,
ASB plans to use a different version of O3-NEM in the future to
address multiple-hour O3 exposures and to more explicitly analyze
various exposure-related uncertainties. When new O3 NEM
estimates are available, ASB plans to revise the risk assessment
discussed in this section, assuming contract funds are available
for that purpose. The benchmark risk estimates are affected by
uncertainty in projecting O3 concentrations upon attainment of
alternative NAAQS at the "critical" monitor.
C. Related Health Effects Issues
The purpose of this section is to provide an overview of
important issues related to the health effects of O3 and other
photochemical oxidants. Some issues such as attenuation of
effects and the acute/ chronic relationship involve more than one
health effect, while other issues involve multiple pollutants.
Detailed discussion of the individual studies associated with
these issues can be found in chapters 9 to 12 of the CD.
-------
VII-53
1. Adverse Respiratory Health Effects of Acute Ozone
Exposure
What constitutes an adverse respiratory health effect for
acute exposure to 03 has been a matter of some controversy and
•diversity of opinion. Because the NAAQS for 03 is intended to
protect the population from exposure to 03 levels which might
produce adverse health effects, it is important to identify the
adverse effects of 03. Although incapacitating effects or
irreversible damage are widely accepted as being adverse, that
type of response has not been reported for acute exposures of
humans to O3 levels near the current standard of 0.12 ppm.
Public health organizations, such as the American Thoracic
Society (ATS), have developed general guidelines (ATS, 1985) as
to what constitutes an adverse respiratory health effect with
respect to interpretation of epidemiological studies. While
recognizing that perceptions of "medical significance" and
"normal activity" may differ among physicians and experimental
sub3ects, the ATS (1985) defines adverse respiratory health
effects as "... medically significant physiologic or pathologic
changes generally evidenced by one or more of the following- m
interference with the normal activity of the affected person or
persons, (2) episodic respiratory illness, (3) incapacitating
illness, (4) permanent respiratory injury, and/or (5) progressive
respiratory dysfunction." Although acute 03 exposures of human
subjects have been associated only with the first of the above
effects, animal studies suggest that longer-term O3 exposures may
cause more serious health effects as well.
The most commonly reported and well-established respiratory
health effect of acute 03 exposure, as discussed earlier in this
chapter, is reduction in lung function. In order to address the
adversity or clinical significance of this response, however, it
is important to consider: (1) the magnitude of a functional
change on a test-specific basis (e.g., FEV, FVC); (2) the
duration or persistence of the effect; (3) the respiratory
symptoms which are associated with functional changes; and (4)
-------
VII-54
limitation of activity which might result from functional losses
and respiratory discomfort.
In conjunction with health scientists of the Office of
Research and Development, OAQPS staff has developed Table VII-5
which shows a gradation of individual response to acute 03
exposure. While Table VII-5 is not intended to imply exact
quantitative relationships, it does categorize three variables as
being mild, moderate, severe, or incapacitating depending on
severity of response. Relationships among the response variables
not only assist in judging the severity of effects but can be
used as a general framework for assessing adversity of effects
for acute 03 exposure.
During the CASAC meeting of December 14-15, 1987 and
December 14-15, 1988, and in written comments received since, a
wide diversity of opinion has been expressed regarding what
should be considered an adverse respiratory^health effect. Most
comments indicated that Table'VII-5 was generally accepted as *
reflecting the range of effects associated with acute (1-2 hr)
exposures to O3. At the CASAC (1987) meeting some CASAC members
expressed the belief that either limitation of activity or
symptoms could be considered the primary determinant of adversity
while others believed the more objective spirometry measurements
were more appropriate. The point at which an effect can be
described adverse was discussed in some detail. Some CASAC
members felt that individuals would experience adverse effects
when O3 exposure induced any of the responses categorized as
moderate. Other CASAC members believed that adverse effects
would not be experienced until 03 induced severe effects. A
point of clarification was made that Table VII-5 refers only to
healthy individuals rather than to persons with respiratory
impairment. Individuals with preexisting respiratory disease may
experience adverse effects even when encountering responses
categorized as mild for healthy persons. Also, it was pointed
out at the CASAC (1988) meeting that because children show few,
if any, symptoms when exposed to ambient O3 levels, it would be
-------
TABLE VII-5.
HEALTHY I(IDIVIDUALS
RESPONSE
CHANGE IN
SPIHOMETRY
FEVj Q. FVC
MODERATE
10-20%
SEVERE
20-40%
INCAPACITING
>40jt
DURATION
OF
EFFECT
SYMPTOMS
LIMITATION
OF
ACTIVITY
COMPLETE
RECOVERY
IN <30 HIN
MILD TO MODEHATE
COUGH
NONE
COMPLETE
RECOVERY
IN <6 HR
Mlin TO MODERATE
COUGH. PAIN ON
DEEP INSPIRATION.
SHORTNESS OF
BREATH
FEW INDIVIDUALS
CHOOSE TO
DISCONTINUE
ACTIVITY
COMPLETE RECOVERY
IN 24 HOURS
REPEATED COUGH.
MODERATE TO SEVFRE
PAIN ON DEEP
INSPIRATION AND
SHORTNESS OF BREATH;
BREATHING DISTRESS
SOME INDIVIDUALS
CHOOSE TO
DISCONTINUE
ACTIVITY
RECOVERY IN
>24 HOURS
SEVERE COUGH.
PAIN ON DEEP
INSPIRATION. AND
SHORTNESS OF
BREATH; OBVIOUS
OISTRESS
MANY INDIVIDUALS
CHOOSE TO
DISCONTINUE
ACTIVITY
<
M
M
I
-------
VII-56
inappropriate to recommend that all categories of response be
seen in children before describing the effect as adverse.
Based on these comments of CASAC (1987, 1988), staff
recommends that the responses identified as "mild" for healthy,
adult response to 03 not be considered an adverse respiratory
health effect. Mild individual responses to O3 exposures are
viewed as involving only 5 to 10% decrements in spirometry which
last 30 minutes or less accompanied by only mild to moderate
cough and no limitation of activity. Mild response probably
would not be considered medically significant and should not
interfere with normal activity of most individuals. Staff
recommends, however, that moderate, severe, and incapacitating
responses be considered adverse respiratory health effects.
These categories of response are more likely to be considered
medically significant and could interfere with normal activities.
For example, some adults with preexisting respiratory disease or
heavily exercising healthy adults who experience a moderate
response (i.e., lung function loss of 10-20% which persists for
up to 6 hours accompanied by multiple symptoms) would tend to
curtail activity. Due to lack of a clear consensus and the wide
diversity of CASAC (1987,1988) opinions, staff recommends that
all criteria specified for moderate response be met for an effect
to be deemed adverse in healthy adults. The adverse nature of
longer-term (6 to 8 hour) and chronic effects of O3 will be
discussed following completion of future research which addresses
these effects.
2. Attenuation of Acute Pulmonary Effects
Attenuation of acute pulmonary response to 03 after repeated
daily exposures to 03 is a well-established and well-documented
phenomenon. Until recently, descriptive terms other than
attenuation have commonly been used to describe the response,
such as "adaptation" and "tolerance". These terms imply a
reduced impact of repeated exposure to 03 whereas recent evidence
suggests that lung injury continues during the process of
-------
VII-57
attenuation. A thorough review of the large body of supporting
evidence can be found in the CD (p. 10-47 to 10-60) and CDS
Functional decrements are generally greater on the second
exposure day; by the fourth or fifth day of exposure, small or no
decrements are observed (Farrell et al., 1979; Horvath et al
1981; Kulle et al., I982b; Linn et al., 1982). Attenuation of
functional response to a particular 03 level does not attenuate
response to higher O3 levels, nor is the attenuation process
permanent. Subjects repeatedly exposed to O.2 ppm 03 for three
consecutive days exhibited attenuation to that level of 03 but
showed no attenuation of response to higher (0.42 or 0.50 Ppm) o
levels (Gliner et al., 1983). once full attenuation is achieved
it does not continue for more than 3 to 7 days for most subjects
(Horvath et al., 1981; Kulle et al., I982b; Linn et al., 1982)
however, partial attenuation has been shown to persist for as '
long as 2 weeks (Horvath et al., 1981). Attenuation of symptoms
correlates with magnitude of functional response and was shown in
one isolated statistical result to last as long as 4 weeks (Linn
et al., 1982). Thus individuals living in areas with high O3
peaks may develop attenuated response to repeated lower O3 levels
and still respond to peak exposures, but after exposure to O3
ceases the attenuated response eventually will return to normal
Evidence reviewed in the CDS suggests the possibility that
ambient oxidant exposure during summer months produced an
"adaptation" response which persisted in human subjects for
several months (Hackney et al., 1988). The authors suggested
that allergy or atopy may be a risk factor for excess response
and, further, that nonresponders could be at increased risk for
chronic health effects of cumulative ambient O3 exposure since
they would be less likely to avoid such exposures due to lack of
symptomatic adaptation or attenuation of o3-induced deficits with
sustained but not worsening protein accumulation in lavage fluid
(Costa et al., 1989). However, histopathology of the animals
appeared to worsen and evolve from an acute to a more chronic
inflammatory pattern.
-------
VII-58
Attenuation of functional or symptomatic response does not
necessarily imply attenuation of morphological or biochemical
response. As indicated in the CD (p. 12-42),
Responses to 03/ whether functional, biochemical or
morphological, have the potential for undergoing changes
during repeated or continuous exposure. There is interplay
between tissue inflammation, hyperresponsiveness, ensuing
injury (damage), repair processes, and changes in response.
The initial response followed by its attenuation may be
viewed as sequential states in a continuing process of luna
injury and repair. . y
Public health implications of attenuation, therefore, are that
this response is not a protective mechanism but may in fact
result in an increased 03 dose to the deep lungs and potentially
cause greater tissue damage by permitting individuals to exercise
outdoors during elevated 03 episodes.
3. Relationship Between Acute and Chronic Effects
In order to better assess attenuation of pulmonary
responsiveness to 03 and other effects of O3 which may not be
reversible, it is important to understand the relationship
between acute and chronic effects. Although animal research
provides some evidence, little is known with certainty regarding
either long-term implications of repeated acute 03 exposures or
chronic effects of prolonged exposure of humans to 03. Human
chamber studies involving long-term exposures to 03 have not been
conducted due to concern for serious health effects which might
develop in subjects.
Various pulmonary effects demonstrated in animal studies
suggest that recovery from chronic exposure to O3 is not complete
even after an extended period. Monkeys exposed to 0.64 ppm O3
for as long as a year (8 hr/day, 7 days/wk) continued to show
either statistically significant or substantial decreases in lung
function (e.g. static lung compliance) even after a 3 month post-
exposure period (Wegner, 1982). This was interpreted as
suggesting recovery was not complete even after three months. In
-------
VII-59
several studies, increases in lung collagen content occurred
after short- and long-term exposures to 1.0 ppm 03 (Last et al.,
1979; Last and Greenberg, 1980; Last et al., 1981; Last et al.,
1984b). Adult and weanling rats exposed for 6 and 13 weeks and
monkeys exposed for one year to 03 below 1.0 ppm showed increased
lung collagen content, and even six weeks post-exposure weanling
rats continued to show increases in collagen (Last et al.,
1984b). This suggests that damage continued during the post-
exposure period.
Even though the above and numerous other animal studies show
structural changes caused by repeated short-term and long-term 03
exposures, there is no evidence of emphysema in animals exposed
to 03 according to the CD (p. 9-52). However, studies of
Bartlett et al. (1974) and Costa et al. (1983) provide evidence
that the recoiling force of the lungs of rodents is reduced by
chronic exposure to O3. Reevaluation of three studies (P'an et
al., 1972; Freeman et al., 1974; Stephens et al., 1976) cited in
the 1978 criteria document (U.S. EPA, 1978), which reported
emphysema in animals after prolonged exposure to
-------
VII-60
possible that the resting, nasal-breathing rats exposed to 0.7
ppm 03 may be receiving a lower O3 dose to the lungs than
exercising, oral breathing humans exposed to lower O3 levels.
Pickrell et al. (1987) observed that rats exposed to l.l ppm 03
for 19 hours per day for 11 days had a 40% increase in total lung
collagen and developed interstitial pneumonia, significant weight
loss, and pulmonary fibrosis based on examination 2 months after
exposure was initiated. A lower level of O3 (0.57 ppm) did not
produce these effects, thus indicating that these effects may be
concentration or dose dependent. Furthermore, chronic exposure
of rats and monkeys to > 0.40 ppm O3 causes a concentration-
dependent peribronchiolar inflammatory response (Barr et al.,
1988; Moffatt et al., 1987).
In a chronic exposure study, Gross et al. (1989) showed that
rats exposed for 12 months to an. episodic profile of 03 (2 hours
at 0.25 ppm 03 peak over 9 hours; 0.19 ppm 03 integrated
concentration) exhibited (1) functional lung changes indicative
of a stiffer lung; (2) biochemical changes suggestive of
increased antioxidant metabolism; and (3) no observable
immunological changes. Other recent studies indicate lower
levels of 03 may be responsible for epithelial injury. Barry et
al. (1985) found that rats exposed to both 0.12 and 0.25 ppm 03
for 12 hours per day over a six week period showed statistically
significant concentration dependent changes in alveolar type l
epithelial cells suggesting that low concentrations of 03 may
cause a chronic epithelial injury in the proximal alveolar
region. For chronic 03 exposures of 0.12 - 0.30 ppm, a lesion is
evident at the junction of the conducting airways and the gas
exchange regions of the lungs, characterized by cell population
shifts along with interstitial inflammation and thickening (Crapo
et al., 1.984; Barry and Crapo, 1985; Barry et al.,1985, 1988;
Sherwin and Richters, 1985). This occurs without increased lung
collagen (Filipowicz and McCauley, 1986; Wright et al., I988a)
unless exposure is intermittent (Tyler et al., 1988). The
increased lavagable lipids found in lungs of rats (Wright et al.,
-------
1988b) after chronic exposures to 0.15 - 0.30 ppm 03 are
consistent with cell population shifts and/or inflammation.
Exposure of monkeys by Harkema et al. (I987a,b) to 0.15 or
0.30 ppm 03 for 6 or 90 days (8 hours/day) resulted in
"quantitative changes in nasal transitional and respiratory
epithelium..., ciliated cell necrosis, shortened cilia, and
secretory cell hyperplasia.'• The authors conclude that ambient
levels of 03 can cause nasal epithelial lesions which may
compromise defense mechanisms of the upper respiratory tract
(Harkema et al., I987a,b). This did not occur in a controlled
exposure study of humans exposed to 0.4 ppm 03 for 4 hours
(Carson et al., 1985) despite increased neutrophils (Graham et
al., 1988). The tracheal region of monkeys shows similar acute
lesions as the nasal region but adaptation may occur (Hyde and
Plopper, 1988).
Since CD closure, new information h'as become available which
has improved the ability of EPA to extrapolate 03 toxicology data
to human health effects levels. An 03 dosimetry model has been
used to simulate local absorption of O3 in the lower respiratory
tracts of guinea pigs and rats (Overton et al., 1987). This
model along with information on respiratory system uptake of 03
in both rats and humans suggests that humans retain a greater
fraction (97%) of inhaled 03 than rodents (50%) (Wiester et al
1987, 1988; Gerrity and Wiester, 1987; Gerrity, 1987; Hatch and'
Aissa, 1987) and enhances EPA's ability to extrapolate animal
data to human health effects. According to Miller et al.
(1987a,b), application of an 03 dosimetry model to obtain intra-
and interspecies dose-response curves from collective assessments
of toxicological data will improve the overall risk assessment
process, particularly in its quantitative aspects, while this
new information indicates great strides have been made toward
understanding animal-to-human dosimetry relationships, large
uncertainties remain due to species sensitivity differences
(Hatch et al., 1986) as evidenced by significant variations in
lung txssue concentrations of antioxidant enzymes among mammalian
-------
V3I-62
species (Slade et al., 1985; Bryan and Jenkinson, 1987). These
differences will have to be dealt with in any animal to human
extrapolation. Furthermore, extrapolation across averaging times
or from high to low doses of 03 introduces additional
uncertainties into any extrapolation of these effects.
Until the degree of uncertainty associated with using animal
studies to estimate human effects and effect levels is
substantially reduced or methods are devised to study humans
during long-term exposures to O3, the relationship between acute
and chronic response to 03 will remain unclear.
4. Effects of Other Photochemical Oxidants
It has been postulated that oxidants such as peroxy acetyl
nitrate (PAN) and hydrogen peroxide (H202) may play a role in
producing health effects associated with photochemical oxidant
exposures. However, relatively few controlled human exposure or
animal toxicology studies routinely have investigated these
pollutants, and field and epidemiology studies evaluate mixtures
of pollutants thus making it difficult to judge which oxidant
caused effects.
No significant effects have been reported in controlled
exposures of intermittently exercising healthy young and middle-
aged males to PAN concentrations of 0.25 to 0.30 ppm (Drinkwater
et al., 1974; Raven et al., 1974a,b, 1976; Gliner et al., 1975).
Only one study suggested a possible simultaneous effect of PAN
(.3 ppm) and O3 (.45 ppm) (Drechsler-Parks et al., 1984), but as
with the other studies, this occurred at PAN concentrations much
higher than those reported for relatively high oxidant areas
(0.047 ppm). Except for an association of PAN with eye
irritation (Okawada et al., 1979; Altshuller, 1977; Javitz et
al., 1983; U.S. EPA, 1978; National Research Council, 1977) few
effects of PAN have been reported. Field and epidemiological
studies also have reported few relationships between health
effects and PAN, while animal toxicology studies suggest that
only very high PAN concentrations produce effects in animals,
-------
VII-63
such as significant alterations in host pulmonary defenses
(Thomas et al., 1979, 1981) . Review of ^ llterature on
ofMI can be found in the CD (pp. 9-205 to 9-306 and 10-80 to
Even fewer studies on the health effects of H20, are
available. No significant effects were observed in rats exposed
for 7 days to 0.5 ppm H202 in the presence of ammonium sulfate-
it is generally assumed to not penetrate into alveolar regions'
possibly due to the high solubility of H202 (Last et al., 1982)
Other studies are of little more than mechanistic value
in conclusion, the limited evidence available suggests that
at levels found in ambient air, PAH and H202 are not responsible
for adverse respiratory effects of photochemical air pollution
ozone „ considered to be chiefly responsible for the adverse
health effects of oxidants largely due to the relative abundance
compared to other oxidants (CD, pp. „.„ to 12.6S) . ndan°e
5. Interactions with other Pollutants
Although controlled human exposure studies have not
demonstrated consistently any enhancement of respiratory effects
for 03 when combined with S02 , N02, CO, and H2SO4 or other
particulate aerosols, animal toxicology research suggests
additive or possibly synergistic effects (CD, p. 12-67)
controlled exposure studies of animals provide evidence ' that 03
in combination with HO2 increases susceptibility to bacterial
infection (Ehrlich et al., 1977, 1979; Ehrlich, 1980) and
morphological lesions (Freeman et al., 1974). Exposure to O3 and
H2SO has been reported to produce additive or even synergistic
effects in host defense mechanisms (Gardner et a!., 1977, Last
SS'
t
.,
M
Mixtures of 03 and (NH4)2so4 have produced
" colaagen synthesis
-------
VII-64
polluted air or O3 in purified air (Avol et al., 1984). Results
of the study showed no differences thus suggesting that O3 is
largely responsible for respiratory effects in oxidant-polluted
air. However, it should be noted that combinations of oxidants
with 03 may contribute to decreased pulmonary function and
increased symptomatic effects in asthmatics (Whittemore and Korn,
1980; Linn et al., 1980, I983a; Lebowitz et al.,-1982, 1983,
1985; Lebowitz, 1984; Holguin et al., 1985) and in children'and
adolescents (Lippmann et al., 1983; Lebowitz et al., 1982, 1983,
1985; Bock et al., 1985; Lioy et al., 1985). Interactions
between 03 and total suspended particulate matter were reported
for decreased expiratory flow in children (Lebowitz et al., 1982,
1983, 1985; Lebowitz, 1984) and in adults with airway obstructive
disease (Lebowitz et al., 1982, 1983). Thus it appears
reasonable to conclude that 03 may cause most of the respiratory
effects in oxidant-polluted air, but effects may be exacerbated
by other pollutants. Because efforts to examine pollutant
interactions have been incomplete in clinical and epidemiological
studies thus far, the potential for pollutant interactions
warrants emphasis in the consideration of the margin of safety.
-------
VIII- 1
VIII. Staff Conclusions and Recommendations for Ozone Primary
Standardise
Drawing upon the evaluation of scientific information
contained in the CD and the CDS, which reviews new data made
available since CASAC closure on the CD, this section provides
preliminary staff conclusions and recommendations for
consideration by the Administrator in selecting a pollutant
indicator, averaging time, form, and level of the primary O3
standard. Staff conclusions and recommendations contained herein
are based upon the totality of scientific information and diverse
effects reported in clinical, epidemiology, and toxicology
research.
A. Pollutant Indicator
When the Environmental Protection Agency promulgated the
NAAQS for photochemical oxidants in the Federal Register (36 FR
8186) on April 30, 1971, the scientific data base for health
effects was very limited. The Schoettlin and Landau (1961)
study, which associated increased incidence of asthma attacks
with ambient photochemical oxidants, served an important role in
developing the basis for the primary photochemical oxidants
standard. Subsequent health research, however, indicated that O3
was the predominant oxidant of concern for public health, and on
February 8, 1979, the chemical designation of the primary
standard was changed from photochemical oxidants to O3 (44 FR
8202).
Since 1979, a substantial human health effects research base
has established 03 as being "...chiefly responsible for the
adverse effects of photochemical air pollutants, largely because
of its relative abundance compared to other photochemical
oxidants (CD, p. 12-65)." As discussed in section VII.B.3. of
the staff paper and section 12.6 of the CD, relatively few
controlled human studies have investigated the health
significance of peroxyacetyl nitrate (PAN) or hydrogen peroxide
-------
VIII- 2
(H202). Of the controlled human exposure studies of PAN, only
one (Drechsler-Parks et al.f 1984) suggested a possible
simultaneous effect of O3 and PAN. Other controlled studies have
reported no significant effects for PAN exposures of 0.25 to 0.30
ppm, much higher than PAN levels commonly reported in high
oxidant areas (CD, p. 12-65). Because H2O2 is highly soluble in
aqueous media, it is believed that H202 deposits on upper airway
surfaces rather than penetrating to the alveolar region (Last et
al., 1982). However, investigations of H202 effects in the
alveolar region have not yet been reported.
Regarding interactions with other pollutants, the CD (p. 12-
67) has concluded that O3 alone is considered responsible for
observed respiratory effects reported in controlled human
exposures of 03 with SO2, N02, CO, and H2SO4 or other particulate
aerosols. Animal toxicology studies, however, have produced
varied results, depending on the pollutant combination evaluated
and the variable measured. Additive and/or possibly synergistic
effects have-been described from exposure to 03 and NO2 (e.g.,
increased susceptibility to bacterial infection, morphological
lesions) and from exposure to O3 and H2SO4 (e.g., host defenses
effects and collagen synthesis). Although a controlled human O3
exposure vs. ambient oxidant exposure comparison study indicated
that 03 was the principle cause of respiratory effects, several
epidemiology studies suggest that combinations of oxidants may
contribute to such effects as decreased lung function and
exacerbation of symptoms in asthmatics and in children and young
adults (CD, p. 12-68).
Despite the apparent interactions of O3 with other
pollutants as reported in toxicology and epidemiology studies,
controlled human exposure and field studies have not consistently
demonstrated that respiratory effects observed in combined
exposure studies of 03 and other pollutants are caused by any
pollutant(s) other than 03. This divergence in the data base
does not reject the hypothesis that some portion of the human
respiratory effects associated with exposure to photochemical
-------
VIII- 3
oxidants may be attributed to pollutants other than o,. However
until further human research is performed to support or refute '
thls hypothesis, the staff beUeve the case for oontrolling 03 as
a surrogate for protecting public health fro* human exposure to
03 and other photochemical oxidants remains valid.
The question of whether 03 can serve as an abatement
surrogate for controlling other photochemical oxidants has been
addressed in the CD (p. 12-15, vith a quote from Altshuller
(1983, who concluded that "... ambient air measurements indicate
that 03 may serve directionally but cannot be expected to serve
quantitatively as a surrogate for the other products." This
conclusion appears to apply to the subset of pnotochelnical
products of concern - 03, PAN, PPN, and ^ . iaentified .„
CD, even though Altshuller (1983) examined the use of 0, as an
abatement surrogate for all photochemical products. Lack of a
quantitative, monotonic relationship between 03 and other
Photochemical oxidants is discussed in Chapter 5 of the CD and
demonstrated in Table 12-2 of the CD in which average PAN/03
ratios for different sites and years vary from 2 to 9 in
addition, it is emphasized in the CD (p. 12-17, that no single
measurement -thodology can quantitatively and reliably measure
« « ambient air
in spite of the above limitations, it is generally
recognised that control of ambient 03 levels currently provides
the best means of controlling photochemical oxidants of potential
health concern (o,, PAN PPN ana n n i ™. • potential
„,„, ,. 3 ' "' and H2°2>- This recognition along
with a controlled-exposure human health data base, which
implicates only 03 among the photochemical oxidants at levels
reported in ambient air, supports the recommendation that O3 be
retained as the pollutant indicator for controlling ambient
concentrations of photochemical oxidants. in addition, using c,
as a surrogate for oxidant control deemphasizes the need for
monitoring of PAN, H202 and other oxidants. Unless significant
additional evidence which demonstrates human health effects from
-------
VIII- 4
exposure to ambient levels of non-03 oxidants becomes available,
it is the staff's recommendation that 03 remain as the surrogate
for protection of public health from exposure to all
photochemical oxidants.
B. Form of the Standard
The current primary 03 NAAQS is expressed as an hourly
average which is the concentration not to be exceeded on more
than 1 day per year on average. During the last standard review
the decision was made to change from the deterministic to the
statistical form of the standard. The deterministic form, which
permitted only a single hourly exceedance of the standard level
in any given year, did not adequately deal with the inherent
variability in hourly 03 concentrations due to the stochastic
nature of meteorological factors affecting formation and
dispersion of 03 in the atmosphere. In addition, EPA further
modified the standard so that one expected exceedance would be
given a daily interpretation; that is, a day with two or more
hourly values over the standard level counts as one exceedance of
the standard level rather than two. These changes reduced the
magnitude and increase stability of the "design value" (or
characteristic highest concentration; see p. IV-2) used to
evaluate precontrol air quality (Hayes et al., 1984).
If it is desired to further increase stability of the design
value indicator by allowing more exceedances, the standard level
would have to be reduced to preserve equality of protection for a
multiple exceedances standard formulation. For example, going to
a 5 expected exceedances standard would require a reduction in
the standard level to the range of 0.09 to 0.11 ppm to maintain
the same level of protection as the current 0.12 ppm 03 NAAQS.
(A 5 exceedances standard has a design value 80% lower than a 1
exceedance standard, on average, but there is variability in this
relationship among urban areas.)
Since a multiple expected exceedances formulation for 0.12
ppm 03 does not provide as much protection against short-term
-------
VIII- 5
peak concentrations as does the current O3 NAAQS standard, it is
recommended that the 1 expected exceedances form of the current
O3 NAAQS be retained for a short-term primary 03 standard.
C. Averaging Time(s)
Exposure durations for studies reporting effects at or near
ambient O3 levels fall into three general categories—short-term
(1 to 3 hours), prolonged (6 to 8 hours), and chronic (months to
years). Controlled chamber and field studies of acute pulmonary
effects of 03 have reported statistically significant impairment
of group mean lung function at O3 concentrations < 0.20 ppm for
exposure durations of l to -2 hours (McDonnell et. al., 1983,
1985a,b,c; Kulle et al., 1985; Folinsbee et. al., 1984; Avol et
al., 1983, 1984, I985a,b; Linn et al., 1980, 1983a.,b; Adams and
Schelegle, 1983; Gong et al., 1986; Linn et al., 1986; Schelegle
and Adams, 1986). Prolonged and chronic exposure studies have
reported a variety of pulmonary and extrapulmonary effects for 03
exposure durations ranging from less than a day to more than a
year, but for ethical reasons chronic studies have been performed
using only animal subjects.
As was the case during the 1978 review of the ambient O3
standards, acute effects of O3 documented in controlled human
studies continue to provide the most quantitative and strongest
support for a short-term primary 03 standard (Table VII-i, staff
paper section VILA.). Although 03 exposures continue in these
studies for 1 to 2 hours, periods of intermittent exercise during
exposure at 15 to 30 minute intervals have caused effects to
occur at much lower O3 levels than for resting subjects. It has
been demonstrated that maximum impairment of lung function occurs
during or immediately after the exercise period (Folinsbee et.
al., 1977a,b). Although many individuals may exercise and work
outdoors for extended periods, most do not exercise heavily for
prolonged periods; therefore, the maximum impact of O3 for most
individuals probably occurs after a relatively short time period.
While consideration may be given to standards with averaging
-------
VIII- 6
times of less than one hour, such a standard does not appear to
be warranted. As shown in Section V, all of the alternative 03
standards investigated reduce the probability of experiencing a
high acute O3 exposure.
With regard to the need for a separate longer-term (6-8 hr)
primary O3 standard, recent controlled human exposure,
epidemiology and toxicology studies provide evidence of increased
respiratory impairment caused by multi-hour 03 exposures.
Preliminary results of controlled human exposure (6.6 hour)
studies indicate that 03 levels as low as 0.08 ppm produce
statistically significant lung function decrements and
respiratory symptoms (Folinsbee et al., 1988; Horstman et al.,
1988a,b). In a related study suggestive evidence of inflammation
was reported in humans exposed for 6.6 hours to 0.10 ppm (Koren
et al., I988a,b,c).
In the case of epidemiology studies interpretation of
exposure averaging times associated with this effect remains
uncertain. Associations have been reported between ambient
photochemical oxidant levels and both asthma attack rates
(Whittemore and Korn, 1980; HoLguin et al., 1985), increased
respiratory hospital admissions (Bates and Sizto, 1983, 1987),
and lung function decrements (Lebowitz et al., 1983; Lebowitz^
1984; Lippmann et al., 1983; Bock et al., 1985; Lioy et al.,
1985). in spite of the inherent exposure and environmental'
uncertainties of these studies, they provide evidence of more
serious health effects associated with longer-term exposures than
the apparently transitory effects reported for 1 to 2 hour
exposures in most controlled exposure studies. For example, Lioy
et al. (1985) have reported that healthy, active children (age 7-
13) experience "a persistent decrement in function lasting for as
much as a week after the end of a smog period of about four
days." Lioy and Dyba (1988) have proposed recently that the PEFR
decrements reported in Lioy et al. (1985) may result from total
03 dose rather than persistence of effect from one day to the
next. Finally, animal studies have provided collaborating
-------
VIII- 7
evidence for time of exposure as an important factor in
determining seriousness of effect. Until further evidence is
published which confirms exacerbation of FEV-L decrements from
extended 03 exposure or more serious effects such as inflammation
and persistence of lung function effects, the basis for a
separate 6 to 8 hour O3 standard will remain inadequate.
Much uncertainty continues to exist for extrapolation to
human health effects of levels and averaging times associated
with chronic exposure effects reported in animal toxicology
studies. Although the animal toxicology data base which
documents most of the reported chronic effects provides extensive
support for effects more serious than lung function decrements,
uncertainties regarding dosiitfetry and species sensitivity must be
addressed in any extrapolation to an effect level in humans.
Animal studies investigating continuous and intermittent exposure
to 03 lasting weeks to years have reported changes in lung
function (4 weeks to 1 year), morphology (i week to 18 months),
biochemistry (l week to l year), host defenses (1 week to 3
months) and various extrapulmonary effects (up to 1 year). with
respect to some of these alterations (e.g., lung function,
biochemistry) the postexposure period has been reported to be one
of continued damage rather than complete recovery (CD, p. 12-48).
New research, recent developments in animal extrapolation, and a
chronic effects risk assessment will provide a significantly
improved basis for addressing the need for a separate chronic
exposure standard in the future.
With respect to protection from longer-term and repeated
peak exposures afforded by a 1-hour standard, Appendix A
discusses relationships among air quality indicators in urban
areas. These relationships have been graphically presented in
Appendix A by Figures A-2 through A-4. Interpretation of these
figures suggests that if the current 0.12 ppm 1-hour daily
maximum 03 standard is met by all sites in the data set, then 10%
of all metropolitan statistical areas (MSAs) might have 17 or
more days with an 8-hour daily maximum > 0.08 ppm O3. Although
-------
VIII- 8
it is premature to suggest that 0.08 ppm O3 is the level of
concern for an 8-hour exposure, this does suggest that the
current standard does permit numerous potential 8-hour exposures
to 0.08 ppm 03. On the other hand, if 0.10 ppm 03 is determined
to be the level of concern for 8-hour exposures, the current O3
standard provides reasonable protection from multiple 8-hour
exposures in most areas. The scientific basis for judgment
regarding the level of concern for health effects has not been
determined for 8-hour or longer-term exposures. Until adequate
research has been completed and accepted for incorporation into
the CD, determination of a level of concern for health effects
arising from 8-hour or longer-term exposures to O3 is premature.
In conclusion, comments made by CASAC (1986, 1987, 1988)
support the need to protect the public from health effects caused
by acute exposures to O3. The staff concurs and recommends that
a primary O3 standard with a l--hour averaging time be maintained
to provide protection from acute exposures. The CASAC (1986,
1987, 1988) also discussed the need and basis for multi-hour O3
standards; however, no consensus was reached regarding the need
for alternative primary standards. New clinical, epidemiology,
and toxicology research has provided additional support for the
occurrence of more significant health effects resulting from 6 to
8 hour, multiple, and chronic exposures to O3 than result from
single l-hour O3 exposures. A CDS of new published research
which addresses both acute and longer-term O3 exposures has been
prepared by the Environmental Criteria and Assessment Office
(ECAO) and closure reached on the document. Staff recommends
that this information be used by the Administrator in assessing
the need for longer-term 03 primary standard(s) while carefully
considering the conclusions reached by CASAC in the closure
letter (McClellan, 1989). Although the CASAC recognized the
emerging data base on acute effects resulting from multi-hour
exposures, it was the CASAC view that it would be some time
before enough such information would be published and included in
a criteria document. Thus, "CASAC concluded such information can
-------
VIII- 9
better be considered in the next review of the ozone standards"
(McClellan, 1989). Also, CASAC reached "closure on the staff
position paper recommending a 1-hour standard ..." (CASAC,
1988; McClellan, 1989). With this portion of the review complete
and after considering CASAC's views on all issues, the
Administrator will be in a position to make a regulatory decision
on how and when to best act on the 1-hour standard.
D. Level of the Primary Standard(s)
In selecting the level of the primary NAAQS for O3, the
Administrator is faced with consideration of a large and diverse
health data base. Judgments regarding the scientific quality and
strength of that data base already have been made in Chapters 9
to 12 of the CD. The preliminary assessment of health effects
attributed to 03 presented in Section VII of this staff paper
presents a variety of health effects and discusses the
seriousness of these effects. In addition to establishing a
lowest observed effects level from these health data,
consideration must be given to the uncertain evidence which bears
on a margin of safety needed to protect public health. In
addition, the Administrator must judge which of the health
effects attributed to O3 should be considered adverse before a
final judgment on the level of the O3 primary standard can be
made. Although scientific literature supports the conclusion
that particular 03 concentrations and exposure patterns may pose
risks to human health, scientific data can only identify the
limits of a range within which a standard should be set.
Specific numeric standard levels, frequency of allowable
exceedances, and averaging times are largely a public policy
judgment.
This section draws conclusions regarding the scientific
literature used by staff to identify a lowest observed effects
level for short-term health effects attributed to 03. In
addition, the more uncertain or less quantifiable evidence, which
forms the basis for judgements about which standard provides an
-------
VIII- 10
adequate a margin of safety, is considered for use in
recommending short-term O3 standard options. Although concern
has been expressed for the more serious effects reported in
chronic animal exposure studies, the major uncertainties
associated with dosimetry and species sensitivity differences
limit direct quantitative extrapolation of effects in animals to
human health effects at this time.
The strongest evidence of human health effects from acute
exposure to 03 comes from controlled human exposure studies
because 03 exposures are known accurately, other pollutants are
not present, temperature and humidity are monitored, and human
subjects are used. Field studies also provide strong evidence
for similar reasons with the exception that other ambient
pollutants are present during exposure. This exception, however,
can provide support for effects occurring under real ambient
conditions. The ethical limitations of controlled human exposure
and field studies (e.g. restriction to short-term exposures, non-
invasive techniques, and limited use of infectious agents) have
prevented investigation in humans of the more serious chronic
exposure effects which have been reported in animal toxicology
studies. Field and epidemiology studies have the advantage of
providing associations between health effects and "real world"
ambient exposures to photochemical oxidant pollution; however,
available epidemiology studies provide less certain exposure-
response evidence for O3 than the controlled- exposure and field
studies. As a result, epidemiology data tend to be relied on
less quantitatively in establishing a lowest observed effects
level for 03. Epidemiology data must be given consideration in
developing the level of the NAAQS because they do provide
evidence of effects experienced in more realistic environments
than the artificial chamber exposure studies.
Staff observations, conclusions, and recommendations
regarding health effects of O3 are presented below. They are
based upon the scientific review in chapters 9 to 12 of the CD as
well as in the CDS and the preliminary assessment and analyses
-------
VIII- 11
discussed in Section VII of this staff paper.
are as follows:
These conclusions
Lults to n.
. m ., -- ~. .,. , .,—U.J.Q ppm. AS summarized
in Table VII-!, statistically significant FEV, decrements have
been reported in controlled exposure studies of intermittently
heavily exercising, healthy children and adults exposed for two
hours to 0.12 ppm O3 and of continuously heavily exercising
healthy adults exposed for one hour to 0.16 (McDonnell et al
1983; I985b; Avol et al., 1984,. Field studies have demonstrated
that statistically significant group mean FEV, decrements have
been induced in continuously heavily exercising, healthy
adolescents at mean ambient 03 levels of 0.144 ppm and in
continuously heavily exercising, healthy adults at mean ambient
03 levels of 0.153 and 0.165 ppm (Avol et al., I985a,b; Avol et
al., 1984; Avol et al., 1983; Linn et al., I983a)
experiencing > i n s-
*-
r.anqe from .1-0 i o Ifigur^ VTT-?). These estimates are based
upon an analysis (Hayes et al., 1987b) of data obtained from
individual subjects in three controlled exposure human studies
(Avol et al., 1984; Kulle et al., 1985; McDonnell et al., 1983)
Because a potentially large group of the population may
experience a substantial reduction (>lo*, of FEV, when exposed to
0-12 ppm 03, this 03 level should be considered as the lowest
observed effects level (LOEL, but not necessarily the lowest
observed adverse effects level (LOAEL) based on the staffs
recommended definition of adverse effects in section VII c 1
-------
VIII- 12
3• Despite a close association reported (cpf p
between changes in group mean FE^ and occurred of CTroup
respiratory symptoms — for acute o^ exposures of anmt-o i~
chambers, Haves et al. ri987t^ report only a
correlation between FE^ changes and symptoms severity when +-h*
analysis is conducted using individual data, one reason that
group mean analysis outcomes are different from results of
individual data analysis may be that symptoms reported are
inherently more subjective than FE^ decrements measured; also
some individuals may experience symptomatic effects without
notable changes in FEVX or vice versa due to differences in
mechanisms causing FEV^^ decrements versus symptomatic effects.
4 ' Estimates of the f-rac-hion of heavily exercising adul-hs
experiencing — moderate or severe chest discnmfnr-t- ranged from n
to 15% and those experiencing moderate or severe nnucrh ranged
'from 0-13% when exposed to 0.12 ppm P., (Figure VU-3K These
estimates are based upon an analysis (Hayes et al., I987b of data
obtained from individual subjects in two controlled human studies
(Kulle et al., 1985; McDonnell et al., 1983).
5- The pulmonary function and symptom risk assessment for
alternative o hourly NAAOS rHayas ot al.. I987b?
1988) is one of the factors that should be considered in
selecting an hourly primary NAaps that provides an adequate
margin of safety. From an air quality perspective without regard
to personal exposure patterns, the benchmark risk estimates show
probabilities from 0.5 to 1.0 that 03 concentrations will exceed
levels sufficient to cause > 10% decrements in FEVj^ or > mild
symptoms (cough and chest discomfort) in 1,5, or 10% of heavily
exercising individuals upon attainment of a 0.08, o.io, or 0.12
ppm O3 NAAQS. Benchmark risk estimates are < 0.2 for two of the
three health data bases used when the endpoint is >20% FEV,
decrements or moderate/ severe respiratory symptoms. When
personal exposure patterns are considered, mean expected
-------
VIII- 13
headcount estimates range from 0.5 to 18.8% of the heavily
exercising population for > 10% FEVX decrements or >'mild
respiratory symptoms depending upon health data base used and
urban area examined upon attainment of a 0.12 ppm 03 NAAQS. Mean
expected headcount estimates drop to 0.1 to 7.1% of the heavily
exercising population for > 20% FEVj^ decrements or
moderate/severe respiratory symptoms. The reader is referred to
Table VII-4 for a more detailed summary of expected headcount
estimates and to the discussion on pp. VII-41-43 of caveats and
limitations that should be considered in interpreting the risk
estimates.
6. Preliminary results indicate that prolonged (6 to 8-
hour) exposures to O^ mav cause enhanced lung function
decrements and respiratory symptoms as well as more persistent
respiratory effects than the luna function decrements produced
•
bv single one- and two-hour exposures to O3. Preliminary
evidence of statistically significant lung function decrements
and respiratory symptoms has been reported for exercising healthy
persons exposed to O3 levels of 0.08-0.12 ppm for 6.6 hours
(Folinsbee et al., 1988; Horstmann et al., 1988a,b). Similar
exposure protocols produced suggestive evidence of inflammation
at 0.10 ppm 03 (Koren et al., 1989a,b,c). Persistence of lung
function response in children, as measured by peak expiratory
flow rate (PEFR), has been reported to last for as long as a week
following a smog period of about four days duration in which the
03 concentration peaks were in the range of 0.120 to 0.185 ppm
(Lioy et al., 1985). While the respiratory responses reported
above appear to be primarily caused by O3/ the authors suggest
that other pollutants such as acid sulfates may have contributed
to the persistent lung function effects. Also, Lioy and Dyba
(1988) have proposed recently that the PEFR decrements reported
in Lioy et al. (1985) resulted from total 03 dose rather than
persistence of the effect from one day to the next. Although it
appears to be premature to draw firm conclusions regarding
-------
VIII- 14
persistent or enhanced respiratory effects associated with 6 to 8
hour or longer term exposures to ambient 03," the staff concludes
that there is reason for concern about enhanced respiratory
function changes beyond the transient pulmonary function
decrements thus far reported in most controlled human exposure
and field studies. The emerging data base of clinical and
epidemiology research suggests that standards with longer-term
averaging times may need to be considered as alternatives or
additions to the current 1-hour averaging time primary standard
after adeguate information is available to make an informed
judgment regarding need for such standards.
— - Work Performance eiirrPirMy appears tr» h* limited by
exposure to O3; however, too small a data base ja available -ho
quantify the magnitude of this impairment
this effect should be used only to develop an aH^i^te margin nf
safetv- Results of exposure to 03 during high exercise levels
indicate that discomfort may be an important factor in limiting
performance (Adams and Schelegle, 1983; Folinsbee et al., 1984;
Gong et al., 1986; Schelegle and Adams, 1986).
Subjective statements by individuals engaged in various
sport activities indicate that these individuals may
voluntarily limit strenuous exercise during high-oxidant
concentrations. However, increased ambient temperature and
relative humidity are also associated with episodes of hiqh-
oxidant concentrations, and these environmental conditions
may also enhance subjective symptoms and physiological
impairment during O3 exposure (CD, p. 10-65.)
While it may be difficult to differentiate performance effects
caused by 03 from those caused by other environmental conditions,
exacerbation of effects caused by 03, beyond those caused by
other conditions, may prevent or curtail normal activities and
should be viewed as adversely affecting individual performance;
however, limitations of the data base suggest using this
information only in margin of safety considerations.
-------
VIII- 15
8- Based on the CDS review of new clinical evidence, both
allergic and asthmatic subjects have a greater increase in airway
resistance after 0^ exposure than do healthy subjects (CDS, p.
MLs Epidemioloaical data provides additional qualitative
evidence of exacerbation of asthma in adults at ambient Q3
concentrations below those generally associated with symptoms or
functional changes in most healthy adults. Given the nature of
these data, they should be used in developing an adequate margin
of safety.
Whittemore and Korn (1980) and Holguin et al. (1985) found
small increases in the probability of asthma attacks
associated with previous attacks, decreased temperature, and
incremental increases in oxidant and 03 concentrations,
respectively. Lebowitz et al. (1982, 1983) and Lebowitz
(1984) also showed effects in asthmatics, such as decreased
peak expiratory flow and increased respiratory symptoms,
that were related to the interaction of O, and temperature.
. (CD, p. 12-55.) .
Uncertainties concerning individual exposure and confounding
environmental variables limit developing exposure-response
relationships with these epidemiological data at this time and
staff recommends that these data should be used to develop an
adequate margin of safety. However, if it is demonstrated that
an increased incidence of asthma attacks occurs as a function of
03 exposure, this would be an effect of great medical
significance and would require careful consideration.
9j Luna structure damage induced bv long-term exposures to
QT_ has been demonstrated in several animal species. These data
should be used in establishing an adequate margin of safety.
Significant morphological alterations in the centriacinar region
have been reported after long-term O3 exposures of rats (Barry et
al., 1983; Crapo et al., 1984), monkeys (Eustis et al., 1981;
Fujinaka et al., 1985) and dogs (Freeman et al., 1973). "There
is morphometric (Fujinaka et al., 1985), morphologic (Freeman et
-------
al
a1'' 19?3),
and
function -,
* SU
it is
al
s
possible that
-------
VIII- 17
humans exposed to 03 could experience decrements in host
defenses; but at'the present time, the exact concentration at
which effects might occur in man cannot be predicted, nor can the
severity of the effect" (CD, p. 12-50). The staff recommends
that these data be used in selecting a standard which provides an
adequate margin of safety.
11.
c, of o,^ reported in
addition to linjtftd -vMenea for- ri«^iovaBeil1jtT.
teratoloqical, mutaqpnin, endocrine svstem. *nr. Uv
effects- Until further analysis better establishes the
physiological significance of these effects on human health the
staff recommends that, these effects data be used in selecting a
standard which provides an adequate margin of safety.
susceptibilitY to o,. exposure are activity ieve]
environmental stress (..g. humidit
Those
factors which either have not been adequately tested or remain
uncertain include age, sex, nutrition, and smoking status.
The Administrator should consider taking them into account in
establishing an adequate margin of safety.
ion response
13 ' Attenuatior> of acute pulmo
by repeated daily exposes to o, is a
phenomenon, which should hi viaw^
increasing dosP of o^ to the lungs. Functional decrements are
greater on the second exposure day, with smaller decrements on
each successive day until the fourth or fifth day when small or
no decrements are observed (Farrell et al., 1979; Horvath et al
1981; Kulle et al., I982b; Linn et al., 1982). Attenuation of
functional response, however, does not necessarily imply
attenuation of morphological or biochemical response to 0
-------
VIII- 18
There is an interplay between tissue inflammation
hyperresponsiveness, ensuing injury (damage), repair
?o??«SSr^ *?? Chan9es in response. The initial response
followed by its attenuation may be viewed either as
sequential states in a continuing process of lung inlurv and
and
Attenuation of pulmonary function or symptomatic: effects,
therefore, should be seen as a human response of concern 'because
it may result in an increase in the total O3 dose reaching the
lungs by permitting individuals to exercise more heavily outdoors
when ambient O3 levels are high, thus resulting in continued, and
potentially greater, lung injury. However, the possibility
exists that this response may indicate the body's ability to
adapt to increased O3 levels. Attenuation should, therefore, .be
considered in developing an adequate margin of safety.
MJ - Exacerbation of rpgpi T^.torv effects by interaction of
other pollutants with o, has not been den.on.cH-r*^ in controlled
human exposure or field studio; however, epidemiology studies
suggest that respiratory effects reported near the n standard
may be caused by 0± in combination with other pn
Furthermore, animal studies -sugges* that combin
other pollutants may act additivelv or syn^rgi st Lcallv . depend i
on the pollutants and endooints chosen for study. While the lack
of individual exposure analysis may limit development of
quantitative exposure-response relationships in epidemiology
studies, the large body of evidence cited in the CD (p. 12-68)
supports the conclusion that ambient pollutants other than O3 may
interact with 03 to contribute to respiratory effects observed in
those studies and should be considered in developing an adequate
margin of safety.
•^ - Two groups have been iH^n^jfjed as being "potentially
at-risk" from exposure to O,; m that subgroup of the general
population characterized as having preexisting r-c.gpiratorv
-------
VIII- 19
disease, and (2) those individuals whose activities outdoors
result in increases in minute ventilation which includes
responsive individuals who experience significantly greater
decrements in luncr function from exposure to p., than the average
response of groups studied. rCD. pp. 12-88 to 12-89K
In conclusion, short-term controlled exposure studies
(McDonnell et al., 1983) suggest that a group mean lowest
observed effects level (LOEL) for pulmonary function and symptoms
for healthy, exercising subjects is 0.12 ppm O3 for an averaging
time of one to two hours. Effects somewhat below 0.12 ppm 03
have been reported in recent camp studies (Spektor et al.,
1988a,b), but it has been suggested that these effects may be due
in part to interactions with other pollutants such as acid
aerosols. Analysis of clinical studies indicates that a small
fraction of the population may experience measurable lung
function and symptomatic effects for O3 exposures even below 0.12
ppm (Hayes et al., I987a,b,) while some individuals show no
effects at 0.12 ppm (Linn et al., 1988). Although significant
lung function decrements and symptoms have been demonstrated in a
single study (Horstman et al., 1988, 1989) to occur in healthy
human subjects exposed for 6.6 hours to 03 levels as low as 0.08
ppm, staff believes a larger data base is required to establish a
LOEL for prolonged exposures. Uncertainty associated with
assessment of epidemiology and animal data prevents identifying a
lowest observed effects level for chronic exposures at this time.
Factors which staff believe should be considered in developing a
1-hour standard which provides an adequate margin of safety are:
(1) the possibility that individuals with pre-existing
respiratory disease may experience effects at levels below those
producing effects in healthy subjects;
(2) possible exacerbation of respiratory effects by other
pollutants in combination with O3 during ambient exposures;
(3) the possibility that attenuation of pulmonary function
decrements may increase 03 dose inhaled and result in more
serious effects;
-------
VIII- 20
(4) possible limitation of work performance by exposure to
°3;
(5) evidence of exacerbation of asthma at ambient O3
concentrations;
(6) evidence of bronchial reactivity in subjects exposed to
> 0.3 ppm 03;
(7) evidence indicating that repeated acute and chronic
exposures to O3 may cause lung structure damage;
(8) evidence of increased susceptibility to infection;
(9) extrapulmonary effects of O3'which are of uncertain
biological importance;
(10) potentially "at risk" groups that have not been
adequately tested; and
(11) factors which may affect susceptibility to O3.
E. Summary of Staff Recommendations
Based upon information and discussions contained in the CD
the CDS, and the staff conclusions drawn in Sections VIIIA-D, the
following staff recommendations; regarding the primary O3 standard
are as follows:
1. In consideration of the large base of health information
attributing effects to 03 exposure and the lack of evidence which
demonstrates human health effects from exposure to ambient levels
of photochemical oxidants other than 03, staff recommends that O3
remain as the surrogate for controlling ambient concentrations of
photochemical oxidants.
2. The current primary 03 NAAQS is attained when the
expected number of days per calendar year with maximum hourly
average concentrations above the level of the standard is equal
to or less than one (44 FR 8202). Staff recommends that these
attributes of the O3 standard be retained.
3. Because research has demonstrated that respiratory
effects of public health concern are associated with acute (1 and
-------
VIII- 21
2 hour) exposures it is recommended that the 1-hr averaging time
be retained for the 03 primary NAAQS. Although there are studies
which have identified health effects associated with O3 exposures
having longer than one-hour averaging times, there remains great
uncertainty about which exposure characteristics (e.g., level,
time, repeated peaks) are most important. Recent clinical
research provides evidence of enhanced respiratory effects of 03
levels below the current NAAQS level when subjects are exposed
for a 6.6 hour period (Folinsbee et al., 1988; Horstman et al.,
1988). Epidemiology research suggests persistent changes in lung
function, aggravation of respiratory disease, and possibly
increased hospital admissions, all of which are associated with
03 levels near the current NAAQS. In addition, data reported in
animal toxicology studies provide evidence of lung structure
damage, increased susceptibility to respiratory infection, and a
variety of extrapulmonary effects following O3 exposures lasting
from-a few hours to more than a year. The seriousness of these
effects indicates a need to protect public health from ambient 03
exposures which reasonably can be expected to induce such effects
in humans. This protection could be provided by setting separate
6 to 8-hour, monthly, seasonal, or annual standards, or by
setting a 1-hour standard which controls longer-term exposures of
concern.
Two types of data are necessary for considering the possible
need for a primary 03 standard with an averaging time of greater
than 1 hour: (l) health effects data and (2) aerometric data,
both empirical and statistical. Rombout et al. (1986, 1989)
published air monitoring data from the Netherlands and New Jersey
on relationships between the 1-hour and 8-hour averaging times.
The results of their particular data sets indicate that 1-hour
maximum 03 levels do not always predict 8-hour average O3
concentrations above 0.10 ppm and, therefore, that the daily
maximum hourly 03 standard currently in effect "... does not
adequately indicate the intensity of long duration and high
concentration exposures to ozone" (Rombout et al., 1986). They
-------
VIII- 22
concluded from their data that a longer-term averaging time,
between 7 and 10 hours, would be necessary to provide adequate
protection of public health and, further, that a standard should
be established for the highest daily 8-hour period based upon
running 8-hour averages. Consideration of the Rombout
et al. (1986, 1989) analyses and conclusions must be tempered by
recognition that a very limited data base was used and that
extrapolating the New Jersey data to all areas of the U.S. is
subject to large uncertainty.
The New Jersey data used by Rombout et al. (1986) were for
specific days, at the sites chosen, on which hourly O3
concentrations did not exceed 0.12 ppm, the level of the current
primary NAAQS. It should be noted, however, that those New
Jersey sites are located in a nonattainment area and that diurnal
03 curves at those sites, therefore, are not necessarily
representative of diurnal curves (i.e., the distribution of
hourly 03 concentrations over a day) observed in areas in which
the standard is actually attained.
Thus, it seems prudent to the staff that the relationships
among 1-hour O3 concentrations and other potentially appropriate
exposure statistics in attainment areas or under attainment
scenarios must be fully explored before adopting a longer-term
standard. Whether the current 1-hour primary standard can
effectively serve as a surrogate for one or more longer-term
averaging times depends upon concentrations of potential concern
associated with longer-term exposures. Those concentrations of
potential concern must first be identified before the question of
surrogacy can be resolved.
OAQPS has analyzed whether it is possible to set the short-
term standard at a level which reduces the probability of
experiencing an unacceptable long-term exposure. As discussed in
Appendix A, there is a marginally statistically significant
association between short- and longer-term O3 indices. (For
example, if the current 0.12 Ppm l-hour daily maximum O3 standard
-------
VIII- 23
is met by all sites in the data set, then 10% of all metropolitan
statistical areas might have 17 or more days with an 8-hour
daily maximum >0.08 ppm O3.)
Although the CASAC recognized the emerging data base on
acute effects resulting from multi-hour exposures, it was the
CASAC view that it would be some time before enough such
information would be published and included in a criteria
document. Thus, "CASAC concluded such information can better be
considered in the next review of the ozone standards" (McClellan,
1989). Also, CASAC reached "closure on the staff position paper
recommending a 1-hour standard ..." (CASAC, 1988; McClellan,
1989). With this portion of the review complete and after
considering CASAC?s views on all issues, the Administrator will
be in a position to make a regulatory decision on how and when to
best act on the 1-hour standard.
4. Based on the CD the CDS, the staff assessment of the
acute (1-2 hour) controlled exposure, field, epidemiology, and
animal toxicology data and recommendations of the Clean Air
Scientific Advisory Committee (CASAC, 1986, 1987, 1988), staff
recommends that the range of 1-hour 03 levels of concern for
standard-setting purposes should remain 0.08 ppm to 0.12 ppm.
The upper end of this range (0.12 ppm) represents lowest 03
exposure concentrations in controlled exposure studies in which
healthy, heavily exercising children, adolescents, and young
adults have experienced statistically significant, though small
group mean decrements in lung function. Analyses of individual
data from these studies indicates that larger individual
decrements have been observed in some subjects with cumulative
frequency distributions suggesting that some fraction of subjects
experience FEVj, decrements greater than 10% at 0.12 ppm O3. Mild
cough also has been reported for individuals in these studies at
the upper end of the range. On the other hand, recently
published acute exposure research suggests that 0.12 ppm and even
0.14 ppm 03 may be a no effects level in some individuals for
-------
VIII- 24
and symptoms. As discussed in Section VII.c.l., staff has
recommended that mild individual responses (i.e.., 5-10% FEV,
decrements, 30 minute recovery, mild cough, no limitation of
activity) should not be considered adverse health effects, if
this response is deemed to be not adverse or if 0.12 ppm is not
identified as a lowest observed effects level for O3, then it can
be argued that a standard level of 0.12 ppm does provide a margin
of safety. Some field and epidemiology studies provide
suggestive evidence of acute lung function impairment at similar
or lower exposure levels, though these studies are limited by
uncertainties concerning individual exposure levels and presence
of other pollutants contributing to or causing health effects
measured. Other epidemiology studies have provided evidence of
aggravation of asthma and airway obstructive disease, again
limited by uncertainties about individual exposure and other
pollutants. Given the difficulty of extrapolating these findings
to the U.S. population and the fact that the effects of O3 on
individuals with preexisting respiratory disease have not yet
been fully characterized, staff believes that the upper end of
the above range provides a relatively small if any margin of
safety. These uncertainties and the nature of potential effects
are important considerations in developing an adequate margin of
safety.
With regard to the lower end of the range, neither the
evidence provided by animal toxicology studies (e.g. increased
susceptibility to lung infection, lung structure and biochemical
changes, and extrapulmonary effects) nor other epidemiological
evidence suggesting respiratory effects in children and
susceptible groups provides scientific support for health effects
below 0.08 ppm for short-term (1-2 hour) exposures. These data
as well as estimates provided by exposure and risk analyses will
be considered in evaluating the. margin of safety provided by
alternative 1-hour standards in the range of 0.08 to 0.12 ppm.
Health effects information discussed in Chapter VII of the
staff paper and the CDS provides reason for concern about multi-
-------
VIII- 25
hour exposure (6-8 hours) of humans to O3. Recently published
clinical and epidemiology research suggests that extended
exposure of exercising individuals to 0.08 to 0.12 ppm O3 may
produce biologically important, as well as statistically
significant, lung function decrements, respiratory symptoms and
inflammation. Epidemiology evidence further indicates a possible
association of ambient 03 and other pollutants with increased
respiratory hospital admissions over an extended summertime
exposure. Animal toxicology data show reversible epithelial
injury in the lungs following sub-chronic exposure of animals to
03 levels as low as 0.12 ppm; structural alterations in the lung
and irreversible deposition of collagen in the lungs previously
have^been reported in animals chronically exposed to much higher
levels (0.61 ppm).
Based on the health data base discussed briefly above, EPA
has considered alternatives in addition to the existing 1-hour
primary O3 standard. One obvious alternative is to provide
public health protection from prolonged 03 exposures by setting a
separate longer-term primary st'andard or standards. Great
uncertainty exists at this time, however, regarding either
appropriate averaging times or levels of concern. Another
alternative under -consideration is to set a 1-hour 03 standard
which provides adequate protection from longer-term exposures.
The current 1-hour primary standard of 0.12 ppm O3 is
approximately equivalent to a 0.1 ppm, 8-hour standard. If it
should be determined that protection of public health requires
limiting 8-hour exposures to lower levels, then it would probably
be necessary to tighten the 1-hour standard, to be useful as a
surrogate for protection from longer exposures. Again, however,
this decision will continue to be premature until new research
has been fully evaluated and formally incorporated during the
next review cycle in a criteria document because such a change
would be based upon the need for protection against prolonged
exposure effects.
-------
VIII- 26
Because there is a good health effects data base available
on 1-2 hour exposures, the staff concurs with- CASAC that review
of this scientific information, including the staff paper and its
assessment of this information be closed out. With this portion
of the review complete, and after considering CASAC's views on
all issues, the Administrator will be in a position to make a
regulatory decision on how and when to best act on the 1-hour
standard.
-------
IX- 1
This section reviews and assesses research on welfare
effects attributed to 03 and other photochemical oxidants as
summarized in Air Ouaii+y cri^-ri* for Qg;nno ar^ ^.^
ghotochemic*] Orld.n^ (CD) . where possible, judgments are
provided which identify adverse effects and levels of these
effects for secondary-standard setting.
Important categories with respect to welfare include
agricultural vegetation effects (e.g., yield loss), natural
ecosystem effects, personal comfort and well-being effects and
materials damage. Because the possible indirect contributions of
the photochemical oxidants to visibility degradation, climatic
changes, and acidic deposition cannot at present be quantified
these atmospheric effects and phenomena are not addressed in this
document (CD, p. i-1N Tney have been addressed/ haw ^
other recent air quality documents (U.S. EPA, 1982a,b) The
approach taken in this staff paper is: 1) to describe each type
of 03 effect, 2)« to discuss what is known regarding exposure-
response relationships of 03 exposure, and 3) to evaluate the
factors which should be considered in selecting the level'
averaging time and form of the secondary NAAQS for O
Discussions of the status of the rural 03 exposure analysis, the
03 economic analyses, and the welfare risk assessment are
included.
A.
Mechanisms of Action for Vegetation
as the
of a sequence of physical, biochemical and
Physiological events. For ambient 03 to exert a phytotoxic
effect, lt must first diffuse into the plant, such an effect
will occur only if . sufficient amount of 03 reaches the
,v.^. Space throu9h the stomata, which can
exert some control on O3 uptake to the active sites within the
-------
IX- 2
leaf. Once within the leaf, O3 quickly dissolves in the aqueous
layer on the cells lining the air spaces. Ozone or its
decomposition products then diffuse through the cell wall and
membrane into the cell, where it may affect cellular or
organellar processes. Ozone injury will not occur if i) the rate
of 03 uptake is sufficiently small so that the plant is able to
detoxify O3 or its metabolites or 2) the plant is able to repair
or compensate for the 03 impacts (Tingey and Taylor, 1982). The
uptake and movement of 03 to the sensitive cellular sites are
subject to various physiological and biochemical controls (CD, p.
6-22).
At any point along this pathway, 03 or its decomposition
products may react with cellular components. Altered cell
structure and function may result in changes in membrane
permeability, carbon dioxide fixation, and many secondary
metabolic processes (Tingey and Taylor, 1982). The magnitude of
the 03-induced effects will depend upon the physical environment
and the chemical environment of the plant (including other
gaseous air pollutants and a wide variety of chemicals), and
biological factors (including genetic potential, developmental
age of the plant and interaction with plant pests). Cellular
injury manifests itself in a number of ways, including foliar
injury, premature senescence, reduced yield or growth or both,
reduced plant vigor, and sometimes death. Depending upon the
intended use of a plant species (for food, forage, fiber,
shelter, or amenity), any of these effects discussed above could
constitute a "welfare" effect, with potentially adverse effects
on society.
1. Biochemical Response
When 03 passes into the liquid phase, it undergoes
transformations that yield a variety of free radicals (e.g.,
superoxide and hydroxyl radicals). Whether these species result
from decomposition of 03 or reactions between 03 and biochemicals
in the extracellular fluid has not been determined. Ozone or its
-------
IX- 3
decomposition products, or both, will then react with cellular
components, resulting in structural and/or functional effects.
The potential for 03, directly or indirectly, to oxidize
biochemicals in vitro has been demonstrated. Ozone can oxidize
several classes of biochemicals including nucleotides, proteins,
some amino acids and various lipids. Data acquired from in vitro
studies are best utilized to demonstrate that many cellular
constituents are susceptible to oxidation by O3. New approaches
are needed to assess the full range of in vivo biochemical
changes caused by 03. (See CD, Section 6.3.1.3 for further
details.)
2. Physiological Responses
Physiological responses have been more useful than
biochemical changes in characterizing cell responses to oxidants
(CD, p. 6-26). Many ,consider membranes to be the primary site of
action of O3 (Heath, 1980; Tingey and Taylor, 1982). Whether the
plasma membrane or some organelle membrane is the primary site of
03 action is open to speculation (Tingey and Taylor, 1982). The-
alteration in plasma membrane function, however, is clearly an
early step in a series of 03-induced events that eventually leads
to leaf injury and subsequent yield loss.
The effect of 03 on key steps in photosynthesis has been
measured for several plant species as shown in Table IX-1 (CD, 6-
28). Reductions in photosynthesis may reflect the direct
impairment of chloroplast function or reduced CO2 uptake
resulting from O3-induced stomatal closure, or both. Regardless
of the mechanism, a sustained reduction in photosynthesis will
ultimately affect growth, yield, and vigor of plants. Some
examples of 03-induced reduction in apparent photosynthesis at
concentrations above and below 0.25 ppm are presented in Table
IX-1. These data highlight the potential of O3 to reduce primary
productivity at various air quality levels, some of which are
close to ambient.
-------
IX-4
TABLE IX-1. EFFECT OF OZONE ON PHOTOSYNTHESIS
Species
Loblolly pine
Slash pine
Bean
Alfalfa
Ponderosa pine
Eastern white pine
Eastern white pine
Sensitive
Intermediate
Bean
Black oak
Sugar maple
White pine
Sensitive
Tolerant
Poplar hybrid
Ponderosa pine
03
concentration,
ppm
0.05
0.05
0.072
0.10
0.20
0.15
0.30
0.15
0.10
0.20
0.30
0.10
0.20
0.30
0.30
0.50
0.50
0.7 or 0.9
0.70 to 0.95
0.90
450, 700
800 ppm-hr
Exposure duration
18 wk
continuously
18 wk
continuously
4 hr/day for 18 days
1 hr
1 hr
9 hr daily/
60 days
9 hr daily/
30 days
19 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr/day for 50 days
4 hr daily/50 days
4 hr daily/50 days
3 hr
4 hr daily/2 days
4 hr daily/2 days
3.0 or 10
10/30 days
1.5 hr
Cumulative
dose over
1,2,3 yr
X
inhibition
15b
9b
18b
4b
10b
•25C
67C
10C
24j
42?
sr
Not sig.
different
14b
20b
22C
30 1 10d
21 ± 10d
h
100b
ob
60e
90b
Reference
Barnes (1972a)
Barnes (1972a)
Coyne and Bingham (1978)
Bennett and Hilt (1974)
Miller et al. (1969)
Barnes (1972a)
Yang et al. (1983)
Pell and Brennan (1973)
Carlson (1979)
Carlson (1979)
Botkin et al. (1972)
Furukawa and Kadota (1C!/
Coyne and Bingham (1981)
1 ppra = 1960 ug/m.
bP < 0.05.
CP < 0.01.
Standard deviation.
eNo statistical information.
Source: 03 Criteria Document, U.S. EPA, 1986
-------
IX- 5
For example, Miller et al. (1969) (Table IX-1) found that 3-
year-old ponderosa pine seedlings sustained a 25-percent
reduction in apparent photosynthesis after a 60-day exposure to
an O3 concentration of 0.15 ppm for 9 hours a day. Yang et al.
(1983) exposed three clones of white pine to 03 concentrations of
0.10, 0.20 or 0.30 ppm for 4 hours per day for 50 days in CSTR
chambers. Net photosynthesis was reduced in the foliage of
sensitive and intermediate clones by 14 to 51 percent in direct
relation to O3 dose and clonal sensitivity. Coyne and Bingham
(1978) exposed field-grown snap beans to an 03 concentration of
0.072 ppm for 4 hours a day for 18 days. Apparent photosynthesis
was reduced 18 percent in plants treated with O3.
In addition to depressing photosynthesis in the foliage of
many plant species, 03 inhibits the allocation and translocation
of photosynthate (e.g., sucrose) from the shoots to other organs
(Tingey, 1974; Jacobson, 1982), often referred to as
partitioning. Tingey et al. (1971) found that, when radish
plants were exposed to 03 (0.05 ppm for 8 hrs a day, 5 days per
week for 5 weeks), hypocotyl (stem) growth was inhibited 50
percent while foliage growth was inhibited only 10 percent.
Ponderosa pine exposed to 0.10 ppm O3 for 6 hours per day for 20
weeks stored significantly less sugar and starch in their roots
than did control plants (Tingey, 1976). Several other studies
have measured this partitioning effect of O3 on photosynthate in
carrot, parsley, sweet corn, cotton and pepper (Oshima, 1973;
Bennett and Oshima, 1976; Oshima et al. 1978; Oshima et al.,
1979; Bennett et al., 1979). Plants were exposed to 03
concentrations of 0.12 to 0.25 ppm for 3 to 6 hours for 0.2
percent to 7 percent of the total growth period of all the
plants. In all species but pepper, root dry weight was depressed
much more than leaf dry weight. The reduction in photosynthate
translocation to roots and the resulting decrease in root size
indicate that the plant had fewer stored reserves, possibly
rendering it more sensitive to injury from cold, heat, or water
stress. When less carbohydrate is present in roots, less energy
-------
IX- 6
will- be available for root-related functions. An 03-induced
suppression of nitrogen fixation by root nodules could affect
total biomass and agricultural yield, especially in areas where
soil nitrogen is low.
Reproductive capacity (flowering and seed set) is reduced by
03 in ornamental plants, soybean, corn, wheat and some other
plants (Adedipe et al., 1972; Feder and Campbell, 1968; Heagle et
al., 1972, 1974; Shannon and Mulchi, 1974). These data suggest
that 03 impairs the fertilization process in plants. This
suggestion has been confirmed in tobacco and corn studies using
low concentrations (0.05 to 0.06 ppm) of O3 (Feder, 1968,
Mumford, 1972). In addition to these physiological effects,
which are directly related to productivity, there are many
secondary metabolic responses in a plant exposed to 03, such as
increase in stress ethylene, which may contribute to the
manifestation of foliar injury (CD, p. 6-31).
B. Factors Affecting Plant Response
The magnitude of plant response to O3 exposure can be
affected by biological, physical, and chemical variables. Each
plant exposed to a given concentration of O3 will respond to a
different extent, depending on such biological factors as
genetics, stage of development, and presence of pests or disease,
as well as such physical factors as temperature, humidity, light'
intensity, and soil moisture. Chemical factors in the
environment of plants can also modify plant response by either
magnifying effects, as with multiple pollutants, or by reducing
effects, as with antioxidant sprays. All of these determinants
of response can overlap, thus potentially creating a complex
myriad of causes for the effects observed.
1. Biological Factors
a. Plant Genetics
'Differential susceptibility of individual plants to 03
exposure is determined by genetic composition. Genetic variance
-------
IX- 7
in response to O3 is seen among species, cultivars, and
individuals within a population. Although foliar injury has been
used as the most common measure of comparative 03 sensitivity
other measures, such as yield and physiological effects
substantiate differential results based on genetic differences
In fact, the relative O3 sensitivity of cultivars within a
species can vary with dose and nature of response measured
(Tingey et al., 1972; Heagle, 1979b). Differences in relative
sensitivity of cultivars have been reported between controlled 0
exposures and ambient air field studies (Taylor, 1974; Meiners *
and Heggested, 1979; Hucl and Beyersdorf, 1982; DeVos et al
1983). The nature of inherited O3 sensitivity may help to "
explain this disparity, changes in gene expression during plant
development and due to environmental variations may explain in
part the potential variability in plant response.
b. Developmental Factors
The stage of plant development plays a role in determining
sensitivity to o3. Just prior to, or at maximum, leaf expansion,
Plant foliage appears to be most sensitive because stomata are
functional, intercellular spaces are expanded, and barriers to
gas exchange are minimal (CD, p. 6-34). stages of plant
development also can be affected by C, exposure. Premature aging
and leaf drop have been demonstrated in numerous field and
controlled studies (Menser and Street, 1962; Heggestad, 1973;
Hofstra et al., 1978; Pell et al., 1980; Reich, 1983; Mooi
1980). The premature leaf drop and senescence decrease the
amount of photosynthate that a leaf can contribute to plant
!enetc* " ^ ** C°nClUded that the effects of °3 on the
senescence process, whenever initiated, may be responsible for
many of the documented reductions in yield (CD, p. 6-34).
c. Pathogen and Pest Interactions with Ozone
Disease has been defined as the result of a complex
interaction between host plant, environment, and pathogen;
~s(:r:d r3rh rhogens and insects have — *«-
(CD, p. 6-35). ozone can affect the development of
-------
IX- 8
disease in plant populations. Laboratory evidence suggests that
03 (at ambient concentrations or greater for. 4 or more hours)
inhibits infection by pathogens and subsequent disease
development (Laurence, 1981; Heagle and Letchworth, 1982).
Increases, however, in diseases from "stress pathogens" have been
noted. For example, plants exposed to 03 were more readily
injured by Botrytis than plants not exposed to O3 (Manning et
al., 1970a,b; Wukasch and Hofstra, 1977a,b; Bisessar, 1982).
Both field and laboratory studies have confirmed that the roots
and cut stumps of 03-injured ponderosa and Jeffrey pines are more
readily colonized by a root rot (Heterobasidion annosus). The
degree of infection was correlated with the foliar injury (James
et al., 1980; Miller et al., 1982). Studies in the San
Bernardino National Forest showed that 03-injured trees were
predisposed to attack by bark beetles and that fewer bark beetles
were required to kill an 03-injured tree (Miller et al., 1982).
At any stage of the disease cycle 03 may alter the success of an
invading organism by direct effects on them or by modifying the
ability of the host-plants to defend against attack.
2. Physical Factors
Environmental conditions before, and during, plant exposure
are influential in determining plant.response. The influence of
environmental factors has been studied primarily under controlled
conditions, but field observations have substantiated the
results. Factors which can potentially influence response of
plants to 03 include light intensity, temperature, relative
humidity, soil moisture, and soil fertility. Variations in one
or more of these parameters can cause plants to be more or less
sensitive to O3 exposures. The CD has identified the following
generalizations concerning the influence of these factors on
plant response (CD, p. 6-244) :
a. Light conditions conducive to stomatal opening appear
to enhance 03 injury due to increased O3 absorption.
-------
IX- 9
b. No consistent relationship between temperature and
response to 03 has been reported; however, plants do
not appear to be as sensitive at extremely high or low
temperatures .
c. Plant injury tends to increase with increasing relative
humidity as a result of the effect of humidity on
stomatal opening. McLaughlin and Taylor (1980)
demonstrated that plants absorb significantly more O3
at high humidity than at low humidity.
d. Decreasing soil moisture increases plant water stress
causing a reduction in plant sensitivity to O3. The
reduced 03 sensitivity is apparently related to
stomatal closure, which reduces 03 uptake (U.S. EPA
19.78; Olszyk and Tibbitts, 1981; Tingey et al . , I9B'2}
Water stress, however, does not provide a permanent
tolerance to 03 (Tingey et al., 1982) .
3. Chemical Factors
The chemical environment of plants can include air
pollutants, herbicides, fungicides, insecticides, nematocides
InflT11^' T ^^ ^^^ Th- "ctor.. which may
influence plant response to 03, can be grouped into two areas-
multiple pollutants and chemical sprays.
a. Multiple Pollutants
Studies indicate that the joint action of 03 and sulfur
b * et al.,
,b, Beckerson and Hofstra, 1979; olszyk and Tibbitts 1981,
synergisn (injury enhancement, is most common at lot'
concentrations of each gas and when foliar injury induced by each
gas, md^duany, is small. At higher concentrations or when
the influence of S02 on
response to 03 at ambient and higher concentrations for
-------
IX- 10
several plant species - soybean (Heagle, et al., i983c; Reich and
Amundson, 1984), beans (Oshima, 1978; Heggestad and Bennett,
1981), and potatoes (Foster et al., 1983). Ozone altered plant
yield, but S02 had no significant effect and did not interact
with 03 to reduce plant yield unless the S02 exposures were much
greater than typically found in the ambient air in the U.S.
The applicability of the yield results from pollutant
combination studies to ambient conditions is not known. An
analysis of ambient air monitoring data (SAROAD, 1981; EPRI-SURE)
indicate that most sites wheres the two pollutants were co-
monitored had ten or fewer periods of co-occurrence occurred
during the growing season (Lefohn and Tingey, 1984). Co-
occurrence was defined as the simultaneous occurrence of hourly
averaged concentrations of 0.05 ppm or greater for both
pollutants. At this time, it appears that most of the studies of
the effects on pollutant combinations (03 and SO2) on plant yield
have used a longer exposure duration and a higher frequency of
pollutant co-occurrence than occur in the ambient air.
Preliminary studies using three pollutant mixtures (03, S02, N02)
have shown that the additions of SO2 and NO2 (at low
concentrations) caused a greater growth reduction than O3 alone
(CD, p. 1-79).
Although Lefohn and Tingey (1984) did not observe
significant co-occurrence of 03 and N02 or SO2 in their analysis
of ambient air monitoring data, there is some evidence to suggest
that 03 climatology in natural ecosystems may be correlated with
that of other anthropogenic pollutants (Taylor and Norby, 1985) .
Consequently, many forested regions of North America, where
monitors are scarce, may experience elevated levels of 03 in
combination with other gases (e.g., nitrogen oxides, sulfur
dioxide, nitric acid vapor, and organics) as well as various
pollutants deposited in rain or cloud water (e.g., trace
elements, hydrogen ion, nitrate and sulfate) (Taylor and Norby,
1985). Given the potential for these interactive effects to
occur (McLaughlin, 1985), a number of new studies examining the
-------
IX- 11
effects of 03 and other acidifying substances on forest trees are
H^L9 rrte? a-s part °f the Forsst Rera ^- »*«
NAPAP (National Acid Precipitation Assessment Program, . New data
will be available on this issue over the next two to five years
b. Chemical Sprays
Chemical sprays have long been used to protect agricultural
crops from pests and diseases. Fungicides, herbicides,
insecticides, and nematocides control damage caused by fungus
weeds, insects, and ne.atodes, respectively. They a!so have been
on :fait:r rsitivity °f piants *» •* >°"— • ^^
of the effects of pesticides on 03 sensitivity have shown
dUfenng results, with some chemicals (e.g., nematoxides,
Phenamlphos) increasing sensitivity and others (e.g., benomyl
=arboX1n) reducing sensitivity of plants to 03 Miner et aT
1976; sung and Moore, 1979). '
Antioxidants, which are commonly used to reduce rubber
cracKang and food sppiiage, have been reported to reduce
vegetation injury caused by 03 (Kendrick et al., „„,.- Tne
addition of antioxidants to insecticides, herbicides and
t vegetation
et al., 1980; Koiwai et al., 1977; Gilbert
Ethylenediurea ,„,. . widely ~
visible in3ury in bean plants exposed for „„ minutes to 0.8
can apparentiy protect
pract ca ^ tO be ~«ici«*
practical to be used solely for this purpose (CD, pp. 6-68 to
c. Heavy Metals
-------
IX- 12
fertilized with sludge, may have important effects on vegetation
and ecosystems. Heavy metals may penetrate the cuticle and cause
direct toxic effects on plants or penetrate the soil and affect
plant roots, which in turn may contaminate the food chain. The
effects of 03 in combination with heavy metals have been studied
in several plant species.. Zinc (Zn) and cadmium (Cd) reacted
synergistically with O3 (0.30 ppm for 6 hours) in producing
visible injury and loss of chlorophyll in garden cress and
lettuce (Czuba and Ormrod, 1974). Exposure to the combination of
Cd and O3 induced earlier development of necrosis and chlorosis
and the injury was observed at lower O3 plus Cd levels than for
the individual treatments (Czuba and Ormrod, 1981).
Low concentrations of Cd and nickel (Ni) have been shown to
enhance O3 phytotoxicity on the growth peas (Ormrod, 1977). The
interaction of Cd and 03 was influenced by both concentration and
environmental conditions. Tomato plants grown at- 0.25 and 0.75
mg Cd/ml developed only slight foliar injury when exposed to 03
(0.20 ppm for 3 hours) under cloudy skies; whereas the Cd
treatment alone had no significant effect (Harkov et al., 1979).
In full sun there was extensive 03 injury and the joint response
was synergistic. Quaking aspen treated with 10 jug Cd/ml for 30
days displayed significantly more foliar injury when exposed to
ambient air in New Jersey or exposed to 0.20 ppm O3 for 2.5 hours
(Clark and Brennan, 1980). When plants were exposed to 0.03 ppm
03, the Cd enhancement of injury was not apparent. Although it
is not possible at the present time to assess the risk from the
joint action of gaseous and heavy metal pollutants to vegetation,
the limited data available indicates that heavy metals can
increase the phytotoxic reactions of 03 (CD, p. 6-67).
-------
X- 1
X. Assessment of Welfare Effects and Related Welfare Issues
Considered in Selecting Secondary Standardrs^ for Ozone
Of the phytotoxic compounds commonly found in the ambient
air, 03 is the most prevalent, impairing crop production and
injuring native vegetation and ecosystems more than any other air
pollutant (Heck et al., 1980). Some of the effects of 03
reported in the literature occur at 03 levels at or below natural
background concentrations in many areas of the country (see
Section IV. for further discussion of background values). Ozone
has also been shown to damage elastomers, textile fibers and dyes
and certain types of paints. Other photochemical oxidants of
importance to effects on vegetation, ecosystems and materials are
nitrogen dioxide (NO2) and peroxyacetyl nitrates. Air Quality
Criteria for Oxides of Nitrogen (U.S. EPA, 1982) and Review of
the NAAOS for NO^; Assessment of Scientific and Technical
Information (U.S. EPA, 1984) previously assessed the phytoxicity
of N02, and thus NO2 will not be discussed in this staff paper.
In addition, while at a given dose the peroxyacetyl nitrates are
more phytotoxic than O3 (p. X-22), they generally occur at
significantly lower ambient concentrations. Because phytotoxic
concentrations of peroxyacetyl nitrates are less widely
distributed than those of 03 (CD, p. 6-1), the focus of this
staff paper will be on the effects of O3.
The objective of this section of the staff paper is to
assess the current basis for the 03 secondary NAAQS as contained
in Chapters 6, 7 and 8 of the CD. In addition, the section will
summarize new analyses that address key issues of concern for the
secondary standard: relationships of various air quality
indicators, crop loss estimates, averaging times and forest
response to 03. Key new studies that relate to the issue of
averaging time(s) will also be discussed to determine whether new
effects information suggests any change in existing secondary
NAAQS for 03.
-------
X- 2
A. Vegetation Effects
1. Types of Exposure Effects
Plant response to O3 exposure is quite varied and may be
expressed as biochemical, physiological, visible injury, growth,
yield, reproductive and ecosystem effects. When reviewing the
current data base for vegetation effects, it is apparent that the
bulk of the scientific research on crops that has been completed
since the last standard review focuses on reduction in growth and
yield from various long-term (days, months) exposures. In
contrast, the majority of short-term (few hours) exposure studies
focus on foliar injury or physiological changes as the response
measure. While biochemical and physiological alterations are the
basis of all subsequent effects as described in section IX A.l.
of this paper, visible foliar injury provides one of the earliest
manifestations of short-term effects of O3. However, these
effects are not always well correlated with reduction in growth
and yield (CD, p. 6-141).
This section will attempt to assess the data with regard to
foliar injury effects and reductions in growth and yield. For
purposes of this staff paper, greater emphasis will be placed on
damage or yield loss than on injury. Injury inc:ludes all plant
reactions, such as reversible changes in plant metabolism (e.g.,
altered photosynthesis), leaf necrosis, altered plant quality, or
reduced growth, which do not impair yield or intended use of a
plant (Guderian, 1977). Damage or yield loss, on the other hand,
involves any effect which reduces the quantity, use, or value
(e.g., aesthetic) of a plant or any impairment in the intended
use of a plant. Although foliar injury may have some limitations
in evaluating plant response, its presence is usually an
indicator of elevated O3 concentrations. Growth and yield losses
provide an important measure of the effects of 03 because such
losses impair the intended use of the plant and generally
constitute damage, whereas foliar injury may or may not be
considered damage. These growth and yield loss effects have
become the focus of most of the exposure response models and
-------
X- 3
assessments to be discussed later, thus providing the strongest
and best documented evidence of vegetation effects of o,
exposure.
a. Visible Foliar Injury Effects
The first documented observation of O3 injury to vegetation
in the field was by Richards et al. (1958), who described O3
stipple on grape vine leaves, other studies soon after confirmed
that foliar injury was being caused by O3 from nearby cities
(Heggestad and Middleton, 1959; Daines et al., i960). Numerous
subsequent studies have reported vegetation injury at rural sites
caused by O3 transported long distances from urban centers (Heck
et al., 1969; Kelleher and Feder, 1978; Skelly, 1980; Edinger et
al., 1972).
Of the many approaches taken to estimate the 03 levels and
exposure times required to induce these foliar injury effects
most have involved short-term exposures (less than one day) and
have measured visible injury as 'the response variable. Various
Plant species were exposed to a range of 03 levels and exposure
durations by Heck and Tingey (197!), who evaluated the resulting
data by regression analysis. Data for several species are
summarized in Table X-l (CD, p. 6-224) to illustrate the range of
03 levels and exposure times required to induce 5 and 20% foliar
injury on sensitive, intermediate, and less sensitive species
Limiting-value analysis is an alternative approach to
estimating 03 levels and exposure durations which induce foliar
inDury. such an analysis, which was performed on more than 100
studies of agricultural crops and 18 studies of tree species
yielded the following range of concentration and exposure '
durations that were likely to induce visible injury (Jacobson,
-i» y * i j •
1. Agricultural crops:
0.20 to 0.41 ppm for 0.5 hr
0.10 to 0.25 ppm for l.o hr
0.04 to 0.09 ppm for 4.0 hr
-------
X- 4
Table X-l. Ozone Concentrations for Short-Term Exposure that Produce 5 or
20 Percent Injury to Vegetation Growth Under Sensitive Conditions*
Ozone Concentrations that mav Produce 5% (20%) In-Juryi
Exposure Less
time, hr Sensitive plants Intermediate plants sensitive plants
8.0 0.02 to 0.04 0.07 to 0.12
°-5 0.35 to 0.50 0.55 to 0.70 > 0.70 (0.85)
(0.45 to 0.60) (0.65 to 0.85)
1-° 0.15 to 0.25 0.25 to 0.40 > 0.40 (0.55)
(0.20 to 0.35) (0.35 to 0.55)
2-° 0-09 to 0.15 0.15 to'0.25 > 0.30 (0.40)
(0.12 to 0.25) (0.25 to 0.35)
4.0 0.04 to 0.09 ^ 0.10 to 0.15 > 0.25 (0.35)
(0.10 to 0.15) (0.15 to 0.30)
> 0.20 (0.30)
aThe concentrations in parenthesis are for the 20% injury level. Table is
from U.S. Environmental Protection Agency (1986, p. 6-224).
Source: O3 Criteria Document, U.S. EPA, 1986
-------
X- 5
2. Trees and shrubs:
0.20 to 0.51 ppm for 1.0 hr
0.10 to 0.25 ppm for 2.0 hr
0.06 to 0.17 ppm for 4.0 hr
Foliar injury in sensitive plants provides a basis for a
good understanding of the occurrence of elevated 03
concentrations in a given area. Among the.plants that have been
identified as indicators of pollutants are milkweed (Duchelle and
Skelly, 1981), eastern white pine (Benoit et al., 1982),
ponderosa and Jeffrey pine (Miller, 1973) Bel W-3 tobacco
(Posthumus, 1976) and lichens (Sigal and Nash, 1983). Although
the presence of visible foliar symptoms on vegetation cannot be
directly related to effects on growth or yield, they do indicate
that elevated levels of O3 have occurred. The detection of
visible symptoms is an indication that additional studies should
be undertaken to determine if effects on growth and yield are
occurring (CD, pp. 6-84).
There are a few studies in which short-term exposures have
resulted in growth and yield reduction: exposure of cherry belle
radish to 0.25 ppm O3 for three hours resulted in a 38 percent
reduction in root dry weight (Adedipe and Ormrod, 1974) ; exposure
to 0.10 ppm O3 for two hours resulted in a 9 percent reduction in
the average of three growth responses of capri petunia (Adedipe,
1972) ; a 16 percent reduction in leaflet area growth rate was
reported in pinto bean exposed to 0.05 ppm O3 for 12 hours
(Evans, 1974). In addition, some cultivars of crops such as
spinach and tobacco experience yield losses due to extensive
foliar injury at 0.10 ppm for 2-hours (Menser and Hodges, 1972).
Thus, although a few studies relate short-term exposures and
yield loss, there is very little yield loss data based on
exposures that are easy to relate to the l-hr average of the
current secondary standard.
Despite the importance of visible symptoms, it must be
recognized that long-term exposure to low pollutant
-------
X- 6
concentrations may adversely affect plant health without
producing visible symptoms. Chronic injury from this type of
exposure may be represented by reductions in growth, and/or
yield, or premature senescence resulting from changes in
photosynthesis, respiration, chlorophyll content or other
processes (Dochinger et al., 1970; Feder, .1978; Heck, 1966;
Posthumus, 1976) .
Because plant growth and production depend on
photosynthetically functional leaves, various studies have been
conducted to assess the relationship between foliar injury and
yield for species in which the foliage is not part of the yield.
Some research has demonstrated significant yield loss with little
or no foliar injury (Tingey et al., 1971; Tingey and Reinert,
1975; Kress and Skelly, I985a,b,c; Adedipe et al., 1972), while
others have reported significant foliar injury without yield loss
(Heagle et al., 1974; Oshima et al., 1975). Relative
sensitivities of two potato, cultivars were reversed when judged
by foliar injury versus yield reductions (Pell et al., 1980).
Foliar injury in field corn was reported at lower O3
concentrations than yield loss., but" at higher 03 levels, yield
loss was increased to a greater extent than foliar injury;
similarly, it was reported that foliar injury was not a good
predictor of yield loss for wheat (Heagle et al., 1979a,b).
Therefore, while the presence of foliar injury is significant in
and of itself, no precise relationship exists between foliar
injury and yield loss for species of plants for which foliage is
not part of the yield (CD, p. 6-141).
b. Growth and Yield Effects
As stated previously, the bulk of the evidence since the
last standard review has focused on reduction in growth and yield
from various long-term exposures. The data base has been
summarized previously (Chapter 6, CD) and consists of open top
chamber studies, greenhouse and other controlled experiments, and
various ambient air exposures. This section will assess the
-------
X- 7
strengths and weaknesses of these various approaches. Most of
these studies have characterized the exposure-response
relationship in terms of the seasonal daily daylight mean 0,
concentration, although various averaging times have been used
For purposes of assessing exposure-response relationships
derived from the existing studies, yield loss is defined as an
impairment of, or decrease in, the value of the intended use of
the plant. This concept includes reduction in aesthetic values
changes in crop quality, and occurrence of foliar injury when '
foliage is the marketable part of the plant. The actual amount
of yield loss due to decreased aesthetic value or appearance may
be more difficult to quantify than yield loss in weight or bulk
but is extremely important for crops such as tobacco, spinach
and ornamentals. Such effects occur at concentrations as low'as
0.041 ppm for several weeks or o.io ppm for 2 hours, and can
constitute a yield loss when marketability of the plant is
decreased (Menser and Hodges, 1972).
Most of the recent studies of 03-induced yield loss have
measured effects on the weight gf the marketable plant organ.
These effects will be the primary focus of this section. studies
conducted to estimate the impact of 03 on the yield Of various
crop species have been grouped into two types, depending on the
experimental design and statistical methods used to analyze the
data: (i) studies that developed predictive equations relating
03 exposure to plant response and (2) studies comparing discrete
treatment levels to a control using analyses of variance. The
advantage of the regression approach is that it permits the
estimation of the O3 impact on plant yield over the range of
concentrations, not just at the treatment means as is the case
with the analysis of variance methods.
(1) Open-Top Chamber Studies
Data fro,, a series of studies conducted by the National Crop
Loss Assessment Network (NCLAN) have been analyzed to develop
predictive equations relating 7-hour seasonal nean 03 exposures
to crop y.eld loss. Exampies of the relationship between 03
-------
X- 8
concentration and plant yield are shown in Figure X-l (CD, p. 6-
229) and X-2 (CD, p. 6-230). These cultivars/species were
selected because they illustrate the kinds of exposure response
that occur and the type of year-to-year variation in plant
response to 03 that may occur. The derived regression equations
can be used to determine the concentrations that would be
predicted to cause a specific yield loss or to estimate the
predicted yield loss that would result from a specific O3
concentration. Both of the approaches cited above have been used
to summarize the data on crop response to 03 by use of the
Weibull (Rawlings and Cure, 1985) function.
As an example of response, the 03 concentrations that would
be predicted to cause a 10- to 30-percent yield loss have been
estimated (Table X-2 (CD, p. 6-232). These cutpoints were
selected to illustrate an effect and do not imply an effects
threshold. Other cutpoints could have been selected. A brief
review of these data in the table suggests that: (1) a X0% mean
yield loss is predicted for several species when the 7-hour
seasonal mean concentration of 03 exceeds 0.04-0.05 ppm; (2)
grain crops were generally less sensitive to 03 than other crops;
(3) sensitivity differences within a species may be as large as
differences between species. In addition to differences in
sensitivity among species and cultivars, the data in Figure X-l
and X-2 illustrate year to year variations in plant response to
Although linear regression equations have been used to
estimate yield loss, there appear to be systematic deviations
from the data for some species and cultivars even though the
equations had moderate to high coefficients of determination
(R2). The use of plateau-linear, polynomial equations and the
recently developed Weibull model (Heck et al., 1983) appeared to
fit the data better. On the basis of available data, it is
recommended that curvilinear exposure response functions be used
to describe and analyze plant response to 03 (CD, p. 7-106).
-------
X-9
6000
5000
i 4000
Jt
a
ut
2 3000
(0
2000
1000
(A)
SOYBEAN (OAVIS)
RALEIGH. 1981 AND 1982
1982 Uk£
Y * «831-
-------
S
6000
5600
SOOO
4500
4000
M 3500
Q
Z
3000
2500
2000
1500
COTTON (SJ-2)
SHAFTER. CA. 1981 AND 1982
1982(A)
.6872-KV0.088)2-1
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
O3 CONCENTRATION, ppm
a
34
33
32
31
30
29
M
S! 27
u.
26
25
24
23
,(8)
TOMATO (MURIETTA)
TRACY. CA. 1981 AND 1982
1982(A|
V02.3-«V°-082)3-06
,\
0 O.O2 0.04 0.06 0.08 0.1 0.12 0.14 0.16
O3 CONCENTRATION, ppm
_
a.
a>
X
(0
O
O
ac
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
TURNIP (TOKYO CROSS)
RALEIGH. 1979 AND 1980
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
O, CONCENTRATION, ppm
FIGURE X-2. Examples of the effects of ozone on the yield of cotton, tomato, and
turnip. The 03 concentrations are expressed as 7-hr seasonal mean concentrations.
The species were selected as examples of O3 effects and of year-to-year variations in
plant response to O3.
Source: Cotton and tomato data from Heck et al. (19846); turnip data from Heagle et
al. (1985).
-------
,« nm^ X~2' SUMMARY OF OZONE CONCENTRATIONS PREDICTED TO CAUSE
10 PERCENT AND 30 PERCENT YIELD LOSSES AND SUMMARY OF YIELD LOSSES PREDICT
TO OCCUR AT 7-hr SEASONABLE MEAN OZONE CONCENTRATIONS OF 0 04 and 0 06
Species
Legume crops
Soybean, Corsoy
Soybean, Davis (81)
Soybean, Davis (CA-82)
Soybean, Davis (PA-82)
Soybean, Essex
Soybean, Forrest
Soybean, Williams
Soybean, Hodgson
Bean, Kidney
Peanut, NC-6
Grain crops
Wheat, Abe
Wheat, Arthur 71
Wheat, Roland
Wheat, Vona
Wheat, Bluefaoy II
Wheat, Coker 47-27
Wheat, Holly
Wheat, Oasis
Corn, PAG 397
Corn, Pioneer 3780
Corn, Coker 16
Sorghum, DeKalb-28
Barley, Poco
Fiber crops
Cotton, Acala SJ-2 (81)
Cotton, Acala SJ-2 (82)
Cotton, Stoneville
Horticultural crops
Tomato, Murrieta (81)
Tomato, Murrieta (82)
Lettuce, Empire
Spinach, America
Spinach, Hybrid
Spinach, Viroflay
Spinach, Winter Bloom
Turnip, Just Right
Turnip, Pur Top W. G.
Turnip, Shogoin
Turnip, Tokyo Cross
==========—=—-__
03 concentrations, ppm,
predicted to cause
yield losses of:
lux
0.048
0.038
0.048
0.059
0.048
0.076
0.039 .
0.032
0.033
0.046
0.059
0.056
0.039
0.028
0.088
0.064
0.099
0.093
0.095
0.075
0.133
0.108
0.121
0.044
0.032
. 0.047
0.079
0.040
0.053
0.046
0.043
0.048
0.049
0.043
' 0.040
0.036
0.053
30%
0.082
0.071
0.081
0.081
0.099
0.118
0.093
0.066
0.063
0.073
0.095
0.094
0.067
0.041
0.127
0.107
0.127
0.135
0.126
0.111
0.175
0.186
0.161
0.096
0.055
0.075
0.108
0.059
0.075
0.082
0.082
0.080
0.080
0.064
0.064
0.060
0.072
Percent yield losses predicted
to occur at 7-hr seasonal
mean 03 concentration of:
0.04 ppm
6.4
11.5
6.4
2.0
7.2
1 7
10^4
15.4
14.9
6.4
3.3
4.1
10.3
28.8
n R
\j . j
2.2
0.0
0.4
0.3
1.4
0.0
0.0 '
0.0
8.3
16.1
4.6
0.8
10.3
0.0
6.8
2.6
6.0
5.8
7.7
10.1
13.0
3.3
0.06 ppm
16.6
24.1
16.5
10.4
14.3
5«*
. 3
18.1
18.4
28
19.4
10.4
11.7
24.5
51.2
2O
.8
8.4
0.9
2.4
1.5
5.1
0.3
2.7
0.5
16.2
35.1
16.2
3.7
31.2
16.8
17.2
9.2
16.7
16.5
24.9
26.5
29.7
15.6
... — from Weibull
in charcoal-filtered air.
Source: Derived from Heck et al. (1984b).
-------
X- 12
Although NCLAN provides valuable dose response information
on a variety of crops, the program has limitations that must also
be considered. The adequacy of the 7-hr seasonal mean has been
questioned on several grounds. Recent evidence has suggested
that the diurnal cycle of O3 in natural ecosystems may differ
markedly from that of urban airsheds where the highest O3
concentrations are in the mid-afternoon. In remote locations it
is likely that maximum O3 concentrations shift into the late
afternoon and evening hours (Taylor and Norby, 1985; McCurdy,
1987). In addition, the use of the seasonal mean as well as
other mean statistics in characterizing exposure implies that all
exposures over the course of the daylight period are equally
effective in eliciting a plant response and minimizes the
contribution of peak concentrations. Thus, while the 7-hr
seasonal mean may contain all hourly concentrations for the 7-hr
period, it treats all concentrations the same (CD, p. 6-10). An
infinite number of hourly distributions can be used to generate
the same 7-hr seasonal mean; some containing many peaks and
others containing none. In fact, Larsen and Heck (1984) have
emphasized that it is possible for two air sampling sites with
the same daytime arithmetic mean 03 concentration to have
different estimated crop reductions.
Several authors have reported the greater importance of
concentration compared to exposure duration in causing injury
(e.g., Heck et al., 1966; Bennett, 1979, Heck and Tingey, 1971,
Reinert and Nelson, 1979, Larsen and Heck, 1984). More recent
findings indicate that constant concentrations have less effect
on plant growth responses than variable or episodic exposures at
equivalent cumulative doses (Musselman et al., 1983; Hogsett et
al., 1985). Hogsett found that over the period of 3 cuttings
(133 days), alfalfa growth was reduced more when exposed to an
episodic 03 profile than to a regime of daily 03 peaks, both with
equivalent long-term means. Generally, assessment of human
physiological responses as well as vegetation responses to 03
have stressed the importance of episodic exposures (high
-------
X- 13
concentrations over short periods of tine) in eliciting a
biological response (Rogers, 1985).
There are several additional concerns regarding the NCLAN
data. First, there is the lack of validation of the model. That
is, however, a common deficiency among all models. A second
concern that has been identified is that the use of the Weibull
model might bias the estimated yield losses in the low dose range
of the curve, finally, the most serious limitation of the NCLAN
data stems from the inadequate sampling of environmental
conditions and in inadequate number of test sites. The problem
of inadequate sampling is an easy criticism to level at almost
any agricultural study that has the objective of making an
inference over time and space. The limited time and resources
meant that subjective judgments had to be made regarding which
species were to be studied. The primary criteria for these
decisions were the relative economic importance of the species
and their sensitivity to 03. While the limitations mentioned
above create uncertainties that should be considered in any
application of trie results, the staff concludes that with
appropriate caveats, the NCLAN data, as discussed, provide useful
information on crop loss due to O3 exposure. A more detailed
analysis of the NCLAN crop loss data which addresses some of the
key uncertainties, particularly regarding exposure dynamics, is
discussed on p. x-46.
(2) Greenhouse and Controlled Environment Studies
The effects of O3 on plant yield may be affected by a host
of genetic and environmental factors, m addition to the use of
regression approaches in the studies previously discussed
various other approaches have been used to- investigate the
effects of 03 on crop yield under more controlled (to various
degrees) conditions, as shown in Tables 6-22 and 6-23 of the CD
(CD P. 6-129 to -6-137,. These studies were designed to test
whether specific O3 treatments were different from the control
using analysis of variance. To summarize the data, the lowest'O
concentration that significantly reduced yield was taken fLT '
-------
X- 14
each study (Table X-3; CD, p. 6-235); this concentration was
frequently the lowest concentration used in the study. Although
it is difficult to estimate a "no-effect" exposure concentration,
the data generally seem to indicate that 03 concentrations of
0.10 ppm (frequently the lowest concentration used in the study)
for a few hours a day for several days to several weeks induced
yield losses of 10-55 percent.
One weakness of studies conducted under more controlled
conditions (greenhouse, growth chamber) is that it is difficult
to extrapolate data from the chamber to field conditions. The
more controlled chamber data, however, do serve to strengthen the
demonstration of 03 effects in the field. Concentrations of 0.10
ppm and above appear to cause yield reductions consistently,
although exceptions can be found. In studies which used
concentrations below 0.10 ppm, the response varied among species
(Table 6-22; CD, p. 6-130). Concentrations of 0..05 ppm in
extended or repeated exposures have been -shown to cause yield
reductions in some species or cultivars, no effects in others,
and increased yield in others (Table 6-23; CD, p., 6-134).
Although these studies seem to suggest that a higher 03
concentration was required to cause an effect than was estimated
from the regression studies, it should be noted that the
concentrations derived from the regression studies were based on
a 10% yield loss, while the studies using analysis of variance
indicate that 0.10 ppm frequently induced substantial losses (10
to 55 percent).
(3) Ambient Air Exposure Studies
Ambient air exposure studies conducted to date demonstrate
that 03 in many areas of the country can reduce plant yield.
Although the most severe effects appear to occur in areas with
the highest 03 concentrations such as the South Coast Air Basin
and the San Bernadino Mountains in California, other agricultural
areas in the country can be impacted as well. Recently, open-top
chamber studies evaluating the yield of plants grown in the
-------
Plant species
TABLE X-3. OZONE CONCENTRATIONS AT WHICH SIGNIFICANT YIELD LOSSES HAVE BEEN NOTED FOR
A VARIETY OF PLANT SPECIES EXPOSED UNDER VARIOUS EXPERIMENTAL CONDITIONS
Exposure duration
Yield reduction,
% of control
03 concentration,
ppm
Reference
Alfalfa
Alfalfa
Pasture grass
Lad i no clover
Soybean
Sweet corn
Sweet corn
Wheat
Radish
Beet
Potato
Pepper
Cotton
Carnation
Coleus
Begonia
Ponderosa pine
Western white
pine
Loblolly pine
Pitch pine
Poplar
Hybrid poplar
Hybrid poplar
Red maple
American
sycamore
Sweetgum
White ash
Green ash
Wi How oak
Sugar maple
7 hr/day, 70 days
2 hr/day, 21 day
4 hr/day, 5 days/wk, 5 wk
6 hr/day, 5 days
6 hr/day, 133 days
6 hr/day, 64 days
3 hr/day, 3 days/wk, 8 wk
4 hr/day, 7 day
3 hr
2 hr/day, 38 days
3 hr/day, every 2 wk,
120 days
3 hr/day, 3 days/wk, 11 wk
6 hr/day, 2 days/wk, 13 wk
24 hr/day, 12 days
2 hr
4 hr/day, once every 6 days
for a total of 4 times
6 hr/day, 126 days
6 hr/days, 126 days
6 hr/day, 28 days
6 hr/day, 28 days
12 hr/day, 5 mo
12 hr/day, 102 days
8 hr/day, 5 day/wk, 6 wk
8 hr/day, 6 wk
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
51, top dry wt
16, top dry wt
20, top dry wt
20, shoot dry wt
55, seed wt/plant
45, seed wt/plant
13, ear fresh wt
30, seed yield
33, root dry wt
40, storage root dry wt
25, tuber wt
19, fruit dry wt
62, fiber dry wt
74, no. of flower buds
21), flower no.
55, flower wt
21, stem dry wt
9, stem dry wt
• 18, height growth
13, height growth
+1333, leaf abscission
58, height growth
50, shoot dry wt
37, height growth
9, height growth
29, height growth
17, total dry wt
24, height growth
19, height growth
12, heigjit growth
—
0.10
0.10
0.09
0.10
0.10
0.10
0.20
0.20
0.25
0.20
0.20
0.12
0.25
0.05-0.09
0.20
0.25
0.10
0.10
0.05
0.10
0.041
0.15
0.15
0.25
0.05
0.10
0.15
0.10
0.15
0.15
Neely et al. (1977)
Hoffman et al. (1975)
Horsman et al. (1980)
Blum et al. (1982)
Heagle et al. (1974)
Heagle et al. (1972)
Oshima (1973)
Shannon and Mulchi (1974)
Adedipe and Ormrod (1974)
Ogata and Haas (1973)
Pell et al. (1980)
Bennett et al. (1979)
Oshima et al. (1979)
Feder and Campbell (1968)
Adedipe et al. (1972)
Reinert and Nelson (1979)
Wilhour and Neely (1977)
Wilhour and Neely (1977)
•Wilhour and Neely (1977)
Wilhour and Neely (1977)
Wilhour and Neely (1977)
Patton (1981)
Patton (1981)
Dochinger and Townsend (1979)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
Kress and Skelly (1982)
X
I
Source: 0 Criteria Document, U.S. EPA, 1986
-------
X- 16
presence of 03 (ambient air) versus charcoal-filtered air have
demonstrated losses in tomato (33 percent), bean (26 percent),
soybean (20 percent), snapbean (0 to 22 percent), sweet corn (9
percent), several tree species (12 to 67 percent), and forbes,
grasses and sedges (9 to 33 percent) (Table X-4; CD, pp. 6-145
and 6-146). Field studies in the San Bernadino Forest during the
last 30 years indicate that ambient 03 reduced the height growth
of ponderosa pine by 25 percent, annual radial growth by 37
percent and the total volume of wood produced by 84 percent
(Miller et al., 1982). Oxidant-induced changes in forest
ecosystems will be discussed in more detail in Section B.
Antioxidant chemical protectants appear to provide another
objective method of estimating the impact of ambient O3 on crop
production (Toivonen et al., 1982). Some limitations, however,
must be kept in mind. The chemical itself may alter plant growth
and therefore results must be interpreted carefully. Also, since
the chemical may not be effective against all concentrations of
all pollutants, an underestimation of yield loss may result
(Manning et al., 1974).
Ethylenediurea (EDU), developed as a chemical protectant to
prevent oxidative effects of 03/ has been used extensively to
reduce visible 03 injury in greenhouse and field studies and to
estimate O3-induced yield loss,. Several studies have reported
estimates of the impact of 03 on yield by comparing yield data
from plots with and without EDU treatment (Table X-5; CD, p. 6-
237). For a seven-week study in which ambient 03 exceeded 0.15
and 0.08 ppm on five separate days at each concentration, EDU
treatment reduced foliar injury of onions and increased yield by
37.8% (Wakasch and Hofstra, 1977). During a June-to-August study
in which ambient 03 exceeded 0.08 ppm on 15 days with a maximum
of 0.14 ppm, EDU treatment increased yield of the Tiny Tim tomato
by 30% but had no effect on the New Yorker cultivar (Legassicke
and Ormrod, 1981). Other studies with beans (Toivonen et al.,
1982), tobacco (Bissessar and Palmer, 1984), and potatoes
(Bissessar, 1982) provided evidence of EDU increasing yield by
-------
TABLE X-4
' — •""•" «•« «.n«notK». OH CiKtENHOUSES ON GROWTH AND VlFin (IF 0.05
0.052
0.051
0.035
>0.05
>0.05
>0.05
— ...
Exposure duration
__
99 day average (0600-2100)
43 day average (0600-2100)
3 no average (0900-2000)
31X of hr (8:00 a.m. to
10:00 p.m.) from late
June to mid-September
over three summers; SX
of the time the concen-
tration was above 0.08 ppm
1979, 8 hr/day average
1000-1800), April-
September
1980. 8 hr/day average
(1000-1800), April -
September
1981, 8 hr/day average
(1000-1800). April-
Sepbember
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
Average 170 hr over 60
days exposure (1972-1974)
(6 crops)
— ~ " •
Percent
reduction
fron control
j — - — • —
33 , fruit fresh
wt
26d, pod fresh wt;
24 , number of pods
1, pod wt
20d, seed wt; 10d.
wt/100 +2. X pro-.
tein content, 4X
oil content
,
32. total above
ground biomass
20, total above
ground biomass
21, total above
ground biomass
+5, pod fresh
wt
14d, pod fresh
wt
3, pod fresh
wt
„ „___. „. Mfekk** i bu vnwrj
Location
of study
"
New York
Maryland
Maryland
Virginia
Virginia
Maryland
Maryland
Maryland
Monitoring Calibration
nethod" method0
Mast NBKI
Mast NBKI
Not given Not given
Mast NBKI. known
Chem Known 03
source,
UV
Chen
Mast IX NBKI.
Chem
Mast IX NBKI ,
Chem
Mast IX NBKI.
Chem
Fumigation '
facility0 Reference
OT Maclean and
Schneider
(1976)
OT
OT Heggestad
and Bennett
(1981)
OT Howe)) et
al. (1979)
Howell and Rose
(1980)
OT Duchelle et
al. (1983)
OT. Heggestad et
al. (1980)
OT Heggestad et
al. (1980)
OT Heggestad et
al. (1980)
X
1
i — >
-vl
-------
TABLE X-4 (cont.)
EFFECTS OF AMBIENT AIR IN OPEN-TOP CHAMBERS, OUTDOOR CSTR CHAMBERS, OR GREENHOUSES ON GROWTH AND YIELD OF SELECTED CROPS
03 concn. ,
Plant species pp«
(Astro) >0.05
Snap bean >O.OS
(Gal latin 50)
(BBL 290) >O.OS
(BBL 274) >0.05
Sweet corn >0.08
(Bonanza)
(Monarch Advance) 0.08
Exposure duration
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
Average 160 hr over 60
days exposure (1975-1976)
(2 crops)
58% of hr (0600-2100)
between 1 July and
6 September
Percent
reduction Location Monitoring . Calibration Fumigation
from control of study Method method facility Reference
6, pod dry wt Maryland Mast
j
+1, pod dry wt Maryland Mast
10, pod dry wt Maryland Mast
22d, pod dry wt Maryland Mast
9 , ear fresh wt; California Hast
10 , no. seeds/ear
28^, ear fresh wt;
42 iK, no- seeds/ear
IX NBKI , OT
Chent
IX NBKI OT
Chem
IX NBKI OT
Chem
IX NBKI OT
Chen
UV OT
Heggestad et
al. (1980)
Heggestad et
al. (1980)
Heggestad et
al. (1980)
Heggestad et
al. (1980)
Thompson et
al.. 1976a
Chem = chemiluminescence; Mast = Mast oxidant meter (coulombmetrk); UV = ultraviolet spectrometry.
NBKI = neutral buffered potassium iodide; UV = ultraviolet spectrometry.
COT = open-top chamber; CSTR = continuous stirred tank reactor.
Significant at p = 0.05.
eTotal above ground biomass, 3 yr average; NF and open plot versus CF a significant at p = 0.05
Source: 0 Criteria Document, U.S. EPA, 1986
-------
X-19
TABLE X-5. EFFECTS OF OZONE ON CROP YIELD
AS DETERMINED BY THE USE OF CHEMICAL PROTECTANTS
Species
Yield reduction,
% of control
03 exposure,
ppm
Reference
Beans (green) 41
Onion 38
Tomato 30
Bean (dry) 24
Tobacco 18
Potato 36
Potato 25
>0.08 for total
of 27 hr over
3.5 months
>0.08 on 5 days out
of 48
>0.08 on 15 days
over 3 months
>0.08 on 11 days
(total of 34 hr)
over 3 months
>0.08 on 14 days
during the summer
>0.08 ppm on 18 days
(total of 68 hr)
over'3 months
Manning et al. (1974)
Wukasch and Hofstra (1977b)
Legassicke and Ormrod (1981)
Temple and Bisessar (1979)
Bisessar and Palmer (1984)
Bisessar (1982)
Clarke et al. (1983)
All the species were treated with" the antioxidant, EDU, except the bean study by
Manning et al. (1974) which used the systemic fungicide, benomyl.
Yield reduction was determined by comparing the yields of plants treated with
chemical protectants (control) to those that were not treated.
This study was run over 2 years when the 03 doses were 65 and 110 ppm-hr
respectively, but the yield loss was similar both years.
Source: C>3 Criteria Document, U.S. EPA, 1986
-------
X-20
. « — studies
and to enable o3-induced antl°Xidants to improve
-eral crop speoies/ ~ «°' »•• to be estimate.
concentrations occurring
-------
X- 21
threshold, are the plateau-linear model and Weibull model. The
plateau-linear model incorporates a threshold value but does not
allow curvature of any increase in yield followed by a decrease.
The Weibull model, however, can take on a plateau shape followed
by curved gradual decreases.
National Crop Loss Assessment Network (NCLAN) yield losses
have been modeled by Heck et al. (1983, 1984), who used a three
parameter Weibull function. The Weibull function chosen has a
flexible form, which covers the range of observed biological
responses, is biologically realistic, has parameters with clear
interpretations, and offers a method of summarizing species
responses by developing a common proportional model (Rawlings and
Cure, 1985). The Weibull model subsequently has been used with
NCLAN data previously modeled with other functional forms'(CD, p.
6-165).
b. Statistics Used to Characterize Ozone Exposures
Important factors needed to characterize 03 exposures
adequately in vegetation exposure studies include time,
concentration, and dynamic nature (i.e., constant or variable).
Which of the components of exposure actually causes plant
response, however, remains unclear. Although many studies tend
to report exposures as mean O3 concentrations, the exposure
statistic can be peak hourly, daily, weekly, monthly, or seasonal
means; number of hours above selected concentration/ or number of
hours above selected concentration intervals. None of these
statistics has adequately characterized the relationship among
concentration, exposure duration, interval between exposures and
plant responses. The lack of correlation between exposure
statistics and ambient air measurements has posed a major problem
for those trying to assess the effects of O3 exposure.
The implication inherent in the use of mean 03
concentrations is that all 03 concentrations are equally
effective in causing plant responses. Use of the mean statistic
minimizes the importance of peak concentrations by treating low-
level long-term exposures the same as high concentration short-
-------
X- 22
term exposures. Studies with beans and tobacco, however, have
shown that a given dose (concentration x time) distributed over a
short period induced more injury than did the same dose
distributed over a longer period (Heck et al., 1966).
Concentration was shown to be approximately twice as effective as
time of exposure at causing foliar injury in tobacco plants
(Tonneijck, 1984). The results of several other studies also
have supported the greater importance of concentration in
comparison to time in determining plant response (Bennett, 1979:
Heck and Tingey, 1971; Henderson and Reinert, 1979; Reinert and
Nelson, 1979; Amiro et al., 1934). Thus, a judgment must be made
as to whether greater protection of plants from O3 exposure is
provided by limiting short-term peak exposures, long-term average
exposures, or both.
Not only are concentration and time important but the
dynamic nature of the 03 exposure is also important; i.e.,
whether the exposure is at a constant or variable concentration.
Musselman et al. (1983) recently demonstrated that although
constant concentrations cause the same type of plant response as
variable concentrations at equivalent doses, the constant
concentrations had less effect on plant growth responses. This
finding has been confirmed by other studies. Ashmore found
significant yield reductions in radishes exposed to ambient O3
when the maximum O3 concentration exceeded 0.06 ppm at least 10%
of the days when the crop was growing. Initial studies have
compared the response of alfalfa to daily peak and episodic O3
exposure profiles which had the equivalent total O3 dose over the
growing season (Hogsett et al., 1985) alfalfa yield was reduced
to a greater extent in the episodic that the daily peak exposure.
The plants that displayed the greater growth reduction (in the
episodic exposure) have a significantly lower 7-hour seasonal
mean concentration, thus raising the possibility that the 7-hour
seasonal mean may not properly consider peak concentrations.
Studies with SO2 also showed that plants exposed to variable
concentrations exhibited a greater plant response than those
-------
X- 23
exposed to a constant concentration (McLaughlin et al. 1979)
in addition, the mean does not specifically include an exposure
duration component. Thus, it cannot distinguish between two
exposures to the same concentration, but of different durations.
c. Exposure and Response to Peroxyacetyl Nitrate
Response to .peroxyacetyl nitrate (PAN) exposure has been
assessed by use of the limiting-value method to estimate the
lowest PAN concentration and exposure duration required to
produce visible injury in plants (Jacobson, 1977) . The range of
PAN levels and exposure durations which this, analysis suggests
will produce foliar injury are: 1) 0.20 ppm for 0.5 hr; 2) O.lo
ppm for l.o hr; and 3) 0.035 ppm for 4.0 hr. (CD, p. 6-215) TO
reduce the likelihood of foliar injury to some plants, more
recent studies have suggested that PAN exposures may need to be
30 to 40% lower than the values cited above (Tonneijck, 1984)
Few studies are available which report the effects of PAN on
growth and yield. A study of the effects on non-vascular plants
indicates that growth, photosynthesis, and respiration of algae
can be adversely affected by exposure to PAN (Gross and Dugger
1969). Fumigation of lichens with 0.05 ppm and O.lo PPm PAN for
several days inhibited photosynthesis (sigal and Taylor 1979)
Vascular plants which have been studied to determine the effects
of PAN on growth and yield include radishes, oats, tomatoes,
Pinto beans, beets, and barley. These plants were exposed in
greenhouse studies to PAN concentrations up to 0.04 ppm for 4
hours per day, twice per week, from germination to maturity, with
no significant effects (Taylor et al., 1983, . In the same study
similar exposures of lettuce and Swiss chard produced yield
.losses of 13% and 23%, respectively, without visible foliar
«3«ry. Although it is possible for severe PAN exposure to make
some crops unmarketable, it is unlikely that PAN concentrations
th n cur in
a ^th 7ter.6XCePt ^^^ ln — «"• of California and
a few other localized areas.
-------
X- 24
d. Economic Assessments of Agriculture
Evidence from the plant science literature clearly
demonstrates that 03 at ambient levels will reduce the yields of
some crops (CD, p. 6-246). The fact that such reductions in U.S.
agriculture could adversely affect human welfare has resulted in
numerous attempts to assess, in monetary terms, the losses from
ambient O3 or the benefits of O3 control to agriculture. Most of
the past 1986 economic assessments focus on specific regions of
the country, primarily California and the Corn Belt (Illinois,
Indiana, Iowa, Ohio and Missouri). This regional emphasis may be
attributed to the relative availability of data on crop response
and air quality for selected regions, as well as the national
importance of these agricultural regions. Of the regional
assessments evaluated in the CD (1986, p. 6-247), Howitt et al.,
1984a,b; Rowe et al., 1984; Adams and McCarl, 1985; and Mjelde et
al., 1984 are judged to be adequate on the basis of the adequacy
of the plant science, aerometric and economic data and the
assumptions used in each assessment. Most regional studies
however abstract from the interdependencies that exist between
regions, which limits their utility in evaluating secondary NAAQs'
(CD, p. 6-252).
National-level studies can overcome this limitation of
regional analyses by accounting for economic linkages between
groups and regions. This requires additional data and more
complex models and frequently poses more difficult analytical
problems. Most of the national assessments conducted since 1978
suffer from either plant science and aerometric data problems,
incomplete economic models, or both. As a result of the
limitations, caution should be applied in using these estimates
to evaluate the efficiency of alternative secondary NAAQS. Two
of the studies, both based on NCLAN data, will be briefly
summarized here. Both are judged to be adequate in terms of the
three critical areas of data inputs (CD, p. 6-252) and together
they provide reasonably comprehensive estimates of the economic
consequences of changes in ambient levels of O3 on agriculture.
-------
X- 25
In the first of these studies, Kopp et al. (1984) measured
the national economic effects of changes in ambient 03 level on
the production of corn, cotton, soybean, wheat and peanuts. In
addition to accounting for price effects on producers and
consumers, the assessment methodology places emphasis on
developing producer level responses to O3 induced yield changes
(from NCLAN data) in 200 production regions. The results of the
Kopp et al. (1984) study indicated that a reduction in 03 from
1978 regional ambient levels to a seasonal 7-hr average of
approximately 0.04 ppm would result in a $1.2 billion net benefit
in 1978 dollars. On the other hand, an increase in 03 to an
assumed ambient concentration of 0.08 ppm (seasonal 7-hr average)
across all regions produced a net loss of approximately $3.0
billion (CD,.p. 1-87).
The second study, by Adams etal. (1984b)'isa component of
the NCLAN program. The results were from an economic model of
the U.S. agriculture sector individual farms models for 55
production regions. Using NCLAN data, the analysis examined
yield changes for six major crops (corn soybeans, wheat, cotton,
grain, sorghum and barley) that together account for over 75
percent of U.S. crop acreage. The estimated annual benefit (in
1980 dollars) from 03 adjustments are substantial, but make up a
relatively small percentage of total agricultural output (about 4
percent). In this analysis a 25 percent reduction in O3 from
1980 ambient levels, resulted in benefits of $1.7 billion. A 25
percent increase in O3 resulted in an annual loss (negative
benefit) of 2.363 billion. When adjusted for differences in
years and crop coverages, these estimates are quite close to the
Kopp et al. (1984) benefit estimates (CD, p. 1-88).
While the estimates from both of Kopp et al. (1984) and
Adams et al. (I984b) were derived from conceptually sound
economic models and the most defensible plant science and
aerometric data available at the time of CD closure, there are
several sources of uncertainty including the following:
-------
X- 26
• the issue of exposure dynamics and whether the 7-hr
seasonal mean is an appropriate exposure statistic,
• the lack of environmental interactions in the
experiments, particularly water stress,
• the lack of rural monitoring sites in the SAROAD system
of EPA, which is necessary for extensive validation of Kriging
data,
• the economic models themselves contain many potential
sources of uncertainty including the effects of benefits
estimates of market-distorting factors such as the Federal Farm
Programs. In addition to the revised NCLAN economic assessment,
which will be published in 1983, there have been other analyses
that explicitly consider the Federal Farm Programs (see SP, p.
XI-10).
Despite these uncertainties these two studies, in
combination with the underlying NCLAN data on yield effects,
provide the most comprehensive economic information to date on
which to base decisions regarding the economic efficiency of
alternative seqondary standards (CD, p. 1-88).
B. Natural Ecosystem Effects
The previous section discussed the responses of individual
species of agricultural plants, trees and other native vegetation
to 03. The responses, which are well documented, include: (1)
injury to foliage, (2) reductions in growth, (3) losses in yield,
(4) alterations in reproductive capacity, and (5) increased
susceptibility to pests and pathogens. This section discusses
the effects of 03 stress on simple and complex plant communities
to illustrate that such effects, because of the interconnections
and relationships among ecosystem components, can produce
perturbations in ecosystems. Stresses placed on plant
communities and the ecosystems of which they are a part can
produce changes that are long lasting and that may be
irreversible. No attempts have been made to examine the effects
-------
X- 27
of PAN on ecosystems since there are little data, and trees and
other woody plants appear to be resistant to PAN (CD, p. 7-1).
Evidence indicates that any impact of O3 on ecosystems will
depend on the responses to O3 of the "producer" community.
Producer species (trees and other green plants) are of particular
importance in maintaining the integrity of an ecosystem since
producers provide, through the process of photosynthesis, all of
the new organic matter (food, energy) added to an ecosystem.
Ozone-induced changes in photosynthesis influence energy
flow and mineral nutrient cycling. Alteration of these processes
in ecosystems can set the stage for changes in community
structure by influencing the nature and direction of successional
changes (Woodwell, 1970; Bormann, 1985), with possibly
irreversible consequences (see e.g., Odum, 1985, Bormann, 1985).
The sequence of responses outlined by Bormann is given in Table
X-6 to assist in understanding ecosystem effects (CD, p. 7-4).
1. Forest Ecosystems
Anthropogenic stress on forested ecosystems may result in
reductions in regional tree growth and a decline of stands in
more susceptible forest types (Johnson and Siccama, 1983).
Atmospheric pollutants, in particular oxidants and acid
deposition are important regional factors that are thought to
play a major role in the array of stresses affecting forests
(Smith 1981; Bormann 1982, 1985). Observations of naturally
occurring symptoms or reduced growth as a result of acidic
deposition are limited at the present time. New research results
on the effects of acidic deposition on forest are expected emerge
over the next two to five years.
While the nature and magnitude of the effects of atmospheric
pollution on North American forest are still relatively unknown,
there is evidence that some forest types are negatively affected
by ambient levels of O3. Among the more susceptible forested
areas are the mixed conifer forests of the San Gabriel and San
Bernardino mountain ranges east of Los Angeles, which have been
-------
X-28
TABLE X-6. CONTINUUM OF CHARACTERISTIC ECOSYSTEM
RESPONSES TO POLLUTANT STRESS
Pnase Response characteristics
0 No response occurs. Manmade pollutants are absent'or
constitute insignificant stress. Plant growth occurs
under natural conditions.
I Ecosystems serve as sinks for pollutants. Species
and/or ecosystem functions are relatively unaffected.
Self-repair occurs.
11 Sensitive species or individuals are subtly and
adversely affected. A reduction in photosynthesis,
a change in reproductive capacity, or a change in
predisposition to insect or fungus attack may occur.
111 Decline occurs in the populations with sensitive
species; some individuals will be lost. Their ef-
fectiveness as functional members of the ecosystem
diminishes. Ultimately, species could be lost from
the system.
IV Large plants, trees, and shrubs of all species die.
The basic structure of the forest ecosystem is changed.
Biotic regulation is affected as forest layers are
peeled off: first trees, tall shrubs, and, under
the most severe conditions, short shrubs and herbs.
The ecosystem is dominated by weedy species not
previously present and by small scattered shrubs
and herbs.
v The ecosystem collapses. The loss of species and
changes in ecosystem structure, nutrients, and soil
so damage the system that self-repair is impossible.
Source: Adapted from Bormann (1985).
-------
X- 29
exposed to oxidant pollution since the early 1950's (Miller,
1973). Ozone was first identified as the agent responsible for
the slow decline and death of ponderosa pine trees in these
forests in 1962 (Miller et al., 1963). Later, Jeffrey pine was
also found to be injured by O3 exposure. Oxidant injury of
eastern white pine has been observed for many years in the
eastern United States (although some genotypes of white pine
appear to be more tolerant). It was first reported as needle
blight in the early 1900s but in 1963 was shown to be the result
of acute and chronic ozone exposure (Berry and Ripperton, 1963).
More recently, oxidant injury of eastern white pine in the
Blue Ridge Mountains of Virginia has been reported by Hayes and
Skelly (1977), Skelly 1980, and Benoit et al. (1982) and on the
Cumberland Plateau of east Tennessee by Mann et al. (1980) and
McLaughlin et al. (1982). Ozone injury in natural plant
communities has been reported by Treshow and Stewart (1973) and
by Duchelle et al. (1983). In addition, there is evidence from
laboratory studies of visible injury, negative effects on
physiological function and reduced productivity as a result of
oxidant pollution (Guderian, 1977; Smith, 1981). However,
because of the disparity between effects observed in the
laboratory and those observed in the field, results from
controlled laboratory and greenhouse studies are not easily
extrapolated to field conditions and there is very little
experimental evidence directly linking ambient ozone
concentrations with decline of tree productivity in the field.
This section will discuss the effects of 03 on plant processes
and growth and the limited data available on ecosystem response.
a. Effects on Plant Processes
A discussion of individual tree response is the first step
in explaining ecosystem response. In forest ecosystems, trees
play a critical role. As producers, trees influence the
structure, energy flow and nutrient cycling of forest ecosystems
(Ehrlich and Mooney, 1983). According to the CD, while some of
the same plant processes are affected in trees and agricultural
-------
X- 30
crop species, perennial plants, because they live longer, must
cope with both short- and long-term stresses, the effects of
which can be cumulative, lasting over the years, or can be
delayed, not becoming apparent for many years. Likewise, effects
can possibly be mitigated through short-or long-term recovery or
replacements (CD, p. 7-23). Therefore, the permanent vegetation
in natural ecosystems receives much greater chronic exposure than
the short-lived vegetation that makes up agroecosystems. The
single agroecosystem has little resilience to pollutant stress;
the natural ecosystem is initially more resistent to pollutant
stress because of species diversity, but the longer chronic
exposures can disrupt the system. These differences between
natural ecosystems and agroecosystems raise a key issue in the
debate on the 03 secondary standard regarding the adequacy of the
current 1-hr standard and other exposure indicators under
consideration to protect trees as well as crops.
In regard to the effects of 03 on individual tree response,
inhibition or reduction in the rate of photosynthesis is possibly
the most significant effect of 03 entry into -the leaves of
sensitive plants, although other mechanisms such as increased
foliar leaching have been suggested (Taylor and Norby, 1985).
Ozone inhibits photosynthesis, decreases formation of organic •
compounds needed for plant growth, and can alter the transport
and allocation of the decreased products of photosynthesis so
that sugar storage and root growth are reduced. Trees in which
O3 has been shown to reduce photosynthesis are loblolly pine and
slash pine (Barnes, 1972), ponderosa pine (Miller et al., 1969;
Coyne and Bingham, 1981), eastern white pine, (Barnes, 1972, Yang
et al., 1983; Botkin et al., 1972) black oak, sugar maple
(Carlson, 1979), and a poplar hybrid (Furukawa and Kadota, 1975)
(Table IX-1, CD, p. 6-28 or SP, pg. IX-4).
Miller et al. (1969), Coyne and Bingham (1981), and Yang et
al. (1983) relate visible injury symptoms and reduced tree growth
to the effect of 03 on photosynthesis. Miller et al. (1969)
found that exposure of 3-year old ponderosa pine seedlings under
-------
X- 31
controlled conditions to concentrations of 0.15 and 0.30 ppm 9
hr/day for 30 days reduced photosynthesis by 10 and 70 percent,
respectively, in addition, it was noted that a reduction in
photosynthesis was accompanied by a decrease in the sugar content
of injured needles. Tingey et al. (1976) observed that the
amounts of soluble sugars, starch, and-phenols tended to increase
in the tops and decrease in the plant roots of ponderosa pine
seedlings exposed to 0.10 ppm 03 for 6 hours per day for 20
weeks. The sugars and starches stored in the tree roots were
significantly less when compared with the controls (CD, p. 7-14).
Coyne and Bingham (1981) measured photosynthesis and
stomatal conductance of attached ponderosa pine needles in
relation to cumulative 03 dose. The decline in photosynthesis
and stomatal function normally associated with aging was
accelerated as 03 injury symptoms increased. Premature
senescence and abscission of needles occurred soon after
photosynthesis reached its lowest level. In a study of white
pine, Yang et al. (1983) also observed that decrease in rates of
photosynthesis due to O3 exposure was closely associated with
visible needle injury, premature senescence and reduction in
biomass (CD, p. 7-13). Thus, the impact of O3 on important plant
processes such as photosynthesis does seem to be reflected in the
occurrence of other symptoms of 03 injury.
b. Effects on Growth
The limited observations of tree response to 03 in the field
and the chamber data that is currently available (Table X-7)
provide the strongest evidence to date of tree response to 03.
Studies made along the Blue Ridge Parkway support the view that
exposure to 03 reduces growth in sensitive trees (Benoit et al.,
1982). Eastern white pine located in experimental plots situated
along the Blue Ridge Parkway were studied to determine the radial
growth increment during the 1955 to 1978 period. Growth of trees
classified as sensitive was 25 percent less than in tolerant
trees; growth was 15 percent less in trees classified as
intermediate in sensitivity. Mean radial increments for all
-------
X- 32
trees during the last 10 years of the study were smaller than for
the previous 24 years. During the period of the study,
concentrations of 0.05 to 0.07 ppm of 03 were recorded on a
recurring basis, with episodic peaks of 0.12 ppm or higher
occurring (Benoit et al., 1982).
Hayes and Skelly (1977) monitored total oxidants and
recorded oxidant related injury on eastern white pine in three
rural Virginia sites between April 1975 and March 1976. Injury
was associated with total oxidant peaks of 0.08 ppm for 1-hr or
more. Peak 1-hr ozone concentrations of 0.17 have been measured
in the Blue Ridge Mountains (Skelly, 1980). Monthly 8-hr average
O3 concentrations ranged from 0.035 to 0.065 ppm during the
oxidant season (April - October) and peak hourly concentrations
ranged from 0.08 to 0.13 ppm (Skelly et al., 1984). Increased
injury symptoms were observed on pine trees previously classified
as sensitive or intermediately sensitive after an O3 episode.
Hayes and Skelly (1977) suggested that continued exposure of
white pine to acute and chronic oxidant concentrations could
influence their vigor and reproductive ability, ultimately
resulting in replacement by tolerant species.
A steady decline in annual ring increments was also noted on
the Cumberland Plateau of east Tennessee during the years 1962 to
1979 (McLaughlin et al., 1982). A reduction of 70 percent in
average annual growth and 90 percent in average bole growth was
observed in sensitive white pine when compared to both trees
classified as tolerant to O3 and trees of intermediate
sensitivity. Decline was attributed primarily to chronic
exposure to 03, which frequently occurred at phytotoxic
concentrations in the area. For the years 1975-1979 the
incidence rates for hourly concentrations >0.08 ppm ranged from
129-339 hours above 0.08 ppm. Maximum 1-hr concentrations range
from 0.12 to 0.2 ppm during this time period. The reduction in
growth on the Blue Ridge Parkway and on the Cumberland Plateau,
as in the case of the San Bernardino Mountains, was correlated
with the predisposing symptoms of chronic decline, which includes
-------
X- 33
the following sequence of events and conditions: (1) premature
senescence and loss of older needles; (2) reduced storage
capacity of carbohydrates in the fall and resupply capacity in
the spring to support new needle growth; (3) shorter new needles,
resulting in lower gross photosynthetic productivity; (4) reduced
availability of photosynthate for external usage (including
repair of chronically stressed tissues of older needles); and (5)
premature casting of older needles (McLaughlin et al., 1982).
Degeneration of feeder roots and mycorrhizae usually precedes the
onset of above ground symptoms (Manion, 1981). Decreases in
nutrient and water uptake may also occur. These changes produce
weakened trees. Weakened trees are, in turn, predisposed to
attack by root rot fungi, to defoliation by insects, and to
attack by the pine beetle. Growth reductions in trees (Table X-
7) grown under controlled conditions have also been observed
(Mooi, 1980; Kress et al., I982a,b,c; Kress and Skelly, 1982;
Jensen, 1979; McClenahen, 1979; Jensen and Dochinger, 1974;
Jensen and Masters, 1975) (CD, p. 6-134).
Injury by 03 to native herbaceous vegetation growing in the
Virginia mountains was also observed (Duchelle et al., 1983).
Ambient O'3 concentrations were shown to reduce growth and
productivity of graminoid and forb vegetation in the Shenendoah
National Park. For each year of the study, biomass production
was greatest for vegetation grown in filtered-air chambers. The
total 3-year cumulative dry weight of plants in filtered chambers
was significantly greater than that in non-filtered and open-air
plots. Common milkweed and common blackberry were the only two
native species to develop foliar injury. Ozone episodes occurred
several times during the period of the study. Peak hourly
concentrations ranged from 0.08 to 0.12 ppm; however, monthly
hourly average concentrations ranged from 0.03 to 0.06 ppm.
-------
Poplar
(Oorskamp)
(Zeeland)
(16-SVC-23)
(16-SYC-23)
Sweetgum
African Sycamore
White ash
Green ash
Exposure duration
12 hr/day , 5 mo
0.041
0.05
0.05
0.05
0.05
0.05
0.10
0.15
0.05
0.10
0.15
0.05
0.10
0.15
0.05
0.10
0.15
6 hr/day. 28 days
6 hr/day, 28 day
6 hr/day, 28 days
6 hr/day, 28 day
6 hr/day, 28 days
9 • height growth
2. height growth
U. height growth
a
9 . height growth
6 hr/day, 28 days +* , .
y 27* 9ht 9rowth
6 hr/day, 28 days
6 hr/day, 28 days
Mooi (1980)
Chem
Pu__
WIICHI
Chem
Chem
Chem
1* NBKI
Constant
source,
NBKI. UV
Constant
source.
NBKI-, UV
Constant
source,
NBKI. UV
Constant
source,
NBKI, UV
LH Kress et
«)• (19826)
•
CSTR Kress et
a>- (19826)
CSTR Kress and
Skelly
(1982)
CSTR
CSTR Kress and
Skelly
(1982)
CST* Kress and
Skelly
(1982)
-------
TABLE X-7. (cont.) tFFECTS OF OZONE ADDED TO FILTERED AIR ON YIELD OF SELECTED TREE CROPS
Plant species
Willow oak
Sugar maple
Yellow poplar
Yellow poplar
Cottonwood
White ash
White ash
Black cherry
Hybrid poplar
(NS 207 + NE 211)
03
concn. ,
ppm
0.05
0.10
0.05
0.10
0.15
0.05
0.10
0.15
0.10
0.10
0.10
0.10
0.20
0.30
0.40
0.10
0.20
0.30
0.40
0.15
Exposure duration
6 hr/day, 28 days
6 hr/day, 28 days
6 hr/day, 28 days
12 hr/day, 48 days
4 hr/day, 1 day/wk,
9 wk
8 hr/day, 5 days/wk,
6 wk
Yield, •% Monitoring
from control method
1, height growth; 2, total dry wt Chem
4, height growth; 11,, total dry wt
5, height growth; 2, total dry wt Chem
+8* height growth; 7, total dry wt
12 , height growth; 41 , total dry wt
+60e, height growth; +41, total dry wt Chem
+8, height growth; +5, total dry wt
12, height growth; +18, total dry wt
19ng, relative growth rate Chem
59n9, relative growth rate
no significant effects
+13, total height; +7, shoot dry wt Not given
0, total height; +5, shoot dry wt
0, total height; 11, shoot dry wt
0, total height; 14, shoot dry wt
+16, total height; +15, shoot dry wt Not given
+5, total height; 4, shoot dry wt
+3, total height; 4, shoot dry wt
28 , total height; 15. shoot dry wt
p
50 , dry wt new shoots from terminal Not given
cuttings
62 , dry wt new shoots from basal
Calibration Fumigation
methodc facility0
Constant CSTR
source,
NBKI, UV
CSTR
CSTR
Not given CSTR
Not given Not given
Not given Not given
Not given GH-CH
Reference
Kress and
Skelly
(1982)
Kress and
Skelly
(1982)
Jensen
(1981)
McClenahen,
(1979)
McClenahen,
(1979)
Jensen and
Dochinger
(1974)
cuttings
-------
TABLE X-7. (cont.)
EFFECTS OF OZONE ADDED TO FILTERED AIR ON YIELD OF SELECTED CROPS
Plant species
Hybrid poplar
(207)
Yellow birch
White birch
Bigtooth aspen
Eastern cottonwood
Red maple (163 ME)
(167 NB)
(128 OH)
Loblolly pine
(4-5 x 523)
(14-5 x 517)
Loblolly pine
Pitch pine
03
concn. ,
ppra
0.20
0.20
0.25
0.25
0.25
0.25
0.25
0.05
0.05
0.05
0.10
0.15
0.05
0.10
0.15
Yield, % Monitoring Calibration Fumigatiog
Exposure duration from control3 method method facility
7.5 hr/day, 5 day/wk, 5, height Not given Not given CH
6 wk 8. height
8 hr/day, 5 day/wk, 9, height MAST NBKI GH-CH
15 wk
34, height
+7, height
8 hr/day, 6 wk 18, height MAST « NBKI CH
32, height
37e, height
6 hr/day, 28 days 6, height growth Chem 1* NBKI CH
6 hr/day, 28 days 18e, height growth; 14 total dry wt Chem Constant CSTR
27e, height growth; 22 , total dry wt source,
41e, height growth; 28e, total dry wt NBKI, UV
6 hr/day, 28 days 4 height growth; 8, total dry wt
13 , height growth; 19. total dry wt
26e, height growth; 24 , total dry wt
Reference
Jensen
(1979)
Jensen and
Masters
(1975)
Dochinger
and Town-
send (1979)
Kress et
al. (1982a)
Kress and
Skelly
(1982)
-------
Plant species
•
Virginia pine
White spruce
Japanese larch
TABLE X-7. (cont.)
EFFECTS OF. OZONE ADDED TO FILTERED AIR ON YIELD OF SELECTED CROPS
0.05
0.10
0.15
0.25
Exposure duration
•
6 hr/day, 28 days
8 hr/day. 5 day/wk,
15 wk
Yield. %
from control3
5. height growth; +2, total dry wt
11. height growth; 3. total dry wt
14, height growth; 13. total dry wt
Mm*Jh°$n9 Ca>ibratcon legation
method methodC facility" Reference
Chem
Mast
Constant
source,
NBKI, UV
NBKI
CSTR
GH-CH
GH-CH
Source: 0;| Criteria Document, U.S. EPA, 1986
Kress and
Skelly (1982)
Jensen and
Masters
(1975)
-------
X- 38
Treshow and Stewart (1973) conducted one of the few studies
investigating the impact of air pollution on natural plant
communities. Grassland, oak, aspen, and conifer communities in
Salt Lake Valley, Utah were studied. Two dominant species,
cheatgrass and aspen, considered key to community integrity were
found to be sensitive. In both cases single 2-hr exposures to
0.15 ppm 03 caused severe injury. Removal of the dominant
species from plant communities could result in a shift to another
species. In a companion study conducted in portable plastic
chambers, 03 exposures of 0.15 to 0.3 ppm for 3-hr per day 5
days/week throughout the growing season reduced root and top
growth'and fewer seeds were produced (Harward and Treshow, 1973).
c. Ecosystem Responses: The San Bernardino Study
The interdisciplinary study of the decline of the pine and
mixed conifer forest in the San Bernardino Mountains is the most
comprehensive and best documented report available on the effects
of oxidants on an ecosystem (Miller et al., 1982). While San
Bernardino may be regarded as a worst case scenario due to the
high'levels of 03 found in Southern California, it still provides
valuable information on the potential consequences of ecosystem
exposure to O3.
Many of the major ecosystem processes were shown by the
study to be affected directly or indirectly. Ozone associated
stress on trees decreased photosynthesis, affected directly or
indirectly translocation of carbon (energy), mineral nutrients,
and water, and reduced trunk diameter, tree height and seed
production in ponderosa and Jeffrey pine (Miller et al., 1982).
Foliar injury of O3 sensitive ponderosa and Jeffrey pines was
observed when 24-hr 03 concentrations ranged from 0.05 to 0.06 '
ppm. During the period of the study, average 24-hr O3
concentrations during the months of May through September ranged
from a background of 0.03 ppm to a maximum of 0.10 to 0.12 ppm.
A comparison of radial growth of ponderosa pine from 1910 to
1940, a period of low pollution (<0.03 ppm), with the years 1941
to 1971, a period of high pollution (0.03 to 0.12 ppm), indicated
-------
X- 39
that 03 exposure reduced the average annual radial growth by
approximately 40 percent, height by 25 percent, and wood volume
by 84 percent in trees less than 40 years of age. The marketable
volume of trees 30 years of age was reduced by 83 percent in the
areas with the highest 03 concentrations (Miller and Elderman,
1977). In addition, stressed pines also became more susceptible
to root rot and pine beetle as a result of weakening by
photochemical oxidants (Stark and Cobb, 1969).
Ozone-induced stress on sensitive ponderosa and Jeffrey pine
and to a lesser extent on sensitive white fir, black oak,
increase cedar and sugar pine, accompanied by fire, brought about
the removal of the pine forest overstory. A shift in dominance
to self-perpetuating fire adapted, O3 tolerant shrub and oak
species resulted. These mixtures provide fewer commodity and
amenity values than the former pine forest (Miller et al., 1982).
The case of San Bernardino suggests that a potential
consequence of 03 stress is a change in'the composition and
successional patterns of some plant communities (Woodwell, 1974).
With regard to forests in eastern North America, both the extent
of the decline and the causal mechanisms remain controversial
(Taylor and Norby, 1985). Changes in the growth patterns of
eastern white pine have been attributed to stress resulting from
03 exposure that began 15 to 20 years earlier (Miller and
Elderman, 1977; Miller et al., 1982; McLaughlin et al., 1982).
More recently, dendroecological studies of the decline of red
spruce in the northeast (Johnson and Siccama, 1983) and of
reduced growth rates of red spruce, balsam fir and frasier fir in
central West Virginia and western Virginia also provide further
evidence that the reductions in growth and mortality measurable
today probably began at least 20 years ago. In addition,
reductions in growth rates of loblolly and short leaf pine
have been reported in the piedmont regions of the southeastern
U.S. (McLaughlin, 1985).
In regard to these most recent declines, there is currently
no agreement as to the trigger factor that precipitated the
-------
X- 40
dieback, mortality and decreased growth. A number of stresses
have been identified including natural processes and air
pollution (Johnson and Siccama, 1983). Given the regional
distribution of 03 in North America, the frequent occurrence of
elevated O3 concentrations, and the recognized effects of 03 on
agricultural productivity in the region, the potential influence
of 03 on forest ecosystems should be given high priority (Taylor
and Norby, 1985).
2. Interrelated Ecosystems
a. Aquatic Ecosystems
It is extremely important to consider that an adverse impact
on a forest or agricultural ecosystem may in turn adversely
affect adjacent aquatic systems. A variety of linkages for
energy and nutrient exchange exist. Disruptions induced by air
pollution stress on terrestrial ecosystems often trigger
dysfunctions in neighboring aquatic ecosystems, such as streams,
lakes, and reservoirs. Sediments resulting from erosion can
change the physical character of stream channels, causing changes
in bottom deposits, erosion of channel banks, obstruction of
flow, and increased flooding. They can fill in natural ponds and
reservoirs. Finer sediments can reduce water quality, affecting
public and industrial water supplies and recreational areas.
Turbidity caused by increased erosion can also reduce the
penetration of light into natural waters. This, in turn, can
reduce plant photosynthesis and lower supplies of dissolved
oxygen, leading to changes in the natural flora and fauna
(Bormann and Smith, 1980). Significant forest alterations,
therefore, may have a regional impact on nutrient cycling, soil
stabilization, sedimentation, and eutrophication of adjacent or
nearby aquatic systems. Interfacing areas, such as wetlands and
bogs, may be especially vulnerable to impact (CD, p. 7-47).
b. Agricultural Ecosystems
Natural and agricultural ecosystems possess the same basic
functional components, require energy flow and mineral cycling
-------
X-41
for maintenance, and
claand »»« influences of
simnT ? SUbStrate- »«t»»l «osyste»s vary l» diversity fro,
SS
simn
spec eT W SW SPe°iSS t0 °°"'PleX ^sterns «*" -ny
spec.es Their populations also vary in genetic composition
peLt JPe°leS diVerSity> *"•* a" -"-"^atin* and se f-
perpetuat^g. Agroecosystems , on the other hand, are usually
highly manipulated monocultures of similar genetic and age
composition and are unable" to maintain themselves without the
If any of the species, varieties, or cultivars is
very sensitive to 03, its market value can be destroyed. When
this occurs, efforts are made to find a resistant cultivar, as
With tobacco, or to grow a crop less sensitive to 03 stress
C. Materials. Damage
th t oSSrrCh °VSr a Peri°d °f I°°re than tW° decades ^s shown
that 03 has the ability to react with both manmade and natural
totaT:^ rne some research has been d°ne °- ^ •««*•
total oxidants on materials, the only components of total
sTite^ tT "^ ^ "Udled ^^^ •» 03 and Ho
" ro^o \ reSear°h f°CUS °n °3' h°WeVer' the «"
from 03 to actual in-house materials remains poorly
characterized.
The materials known to be most susceptible to 03 attack are
™
and
and amount of 03
mage functions
*«WJ.wlia. ine economic impact of o
related damage could then be estimated by using accelerated
-------
X-42
repair and replacement costs. Because little recent work has
been reported on the effects on nonbiological materials, however,
it is necessary to rely on older studies which, in many cases, do
not provide sufficient information to assess the amount and costs
of oxidant related damage.
1. Elastomers
The effects of 03 on elastomers are the best documented.
Natural rubber and synthetic polymers/copolymers of butadiene,
isoprene, and styrene account for the bulk of elastomer
production for products such as automobile tires and protective
electrical coverings. The mechanism of 03 damage on elastomers
shares similarities and differences with simple oxidation from
atmospheric oxygen. Ozone damage, usually in the form of
cracking, tends to be more of a surface phenomenon than simple
oxidation. It is greatly accelerated by mechanical stress, which
produces fresh surface areas at crack boundaries. Simple
oxidation, on the other hand, is slower; it occurs more in the
bulk of a material, and it is less affected by the degree of
stress (Mueller and Stickney, 1970). At pollutant concentrations
and stress levels normally encountered outdoors (and in many
indoor environments), the-elastomer hardens or becomes brittle
and cracked, losing its physical integrity. High humidity and
mechanical stress greatly affect the formation, depth of
cracking, and, in automotive tires, the adhesion piles (Davies,
1979; Wenghoefer, 1974) (CD, p. 8-3).
03 affects natural rubber and other elastomers in a dose
related fashion. Dose is defined in materials research as the
product of concentration and duration of exposure. The
importance of O3 dose was demonstrated by Bradley and Haagen-Smit
(1951), who used a specially formulated 03 sensitive natural
rubber (NR). Samples exposed to 20,000 ppm cracked almost
instantaneously, and those exposed to lower concentrations took a
proportionately longer time to crack. Cracking occurred at a
-------
X- 43
rate of 0.02 to 0.03 ppm-hr over the entire range of
concentrations.
Haynie et al. (1976) exposed samples of a tire sidewall to
O3 at concentrations of 0.08 and 0.5 ppm for 250 to 1000 hr under
10-20 percent strain. From these and other data, they estimated
that at the 03 standard of the time (0.08 ppm, 1-hr average), and
at the annual NOX standard of 0.05 ppm, it would take 2-5 years
for a crack to penetrate cord depth. In addition to stress,
factors affecting the cracking rate include atmospheric pressure,
humidity, sunlight, and other atmospheric pollutants. Veith and
Evans (1980) found a 16-percent difference in cracking rates
reported from laboratories located at various geographic
locations (CD, p. 8-48).
Ozone has been found to affect the adhesion of piles
(rubber-layered strips) in tire manufacturing. Exposures to 03
concentrations of 0.05 to 0.15 for a few hours significantly
decreased adhesion in an NR/SBR blend, causing a 30 percent
reduction at the highest 03 level. This adhesion problem
worsened at higher relative humidities. Wenghoefer (1974) showed
that 03 (up to 0.15 ppm), especially in combination with high
relative humidity (up to 90 percent), caused greater adhesion
losses than heat and NO2 did, with or without high relative
humidity (CD, p. 8-49).
2. Textile Fibers and Dyes
The effects of 03 on dyes have been known for nearly
three decades. In 1955, Salvin and Walker exposed certain red
and blue anthraquinone dyes to 0.1 ppm 03 and noted fading which
until that time had been thought to be caused by NO2. Subsequent
work by Schmitt (i960, 1962) confirmed the fading action caused
by 03 and the importance of relative humidity in the absorption
and reaction of vulnerable dyes. Later Beloin (1972, 1973) noted
the acceleration in fading of certain dyes at an 03 concentration
of 0.05 ppm and a relative humidity of 90 percent (CD, p. 8-49).
-------
X- 44
Both the type of dye and the material in which it is
incorporated are important factors in the resistance a fabric has
to 03. Haylock and Rush (1976, 1978) showed that anthraquinone
dyes on nylon fibers were sensitive to fading from O3 at a
concentration of 0.2 ppm for 16 hrs, while Haynie et al. (1976)
and Upham et al. (1976) found no effects from .O3 concentrations
of 0.1 to 0.5 on royal blue acetate, red rayon-acetate, or plum
cotton. Field studies by Nipe (1981) and laboratory work by
Kamath et al. 1982 showed a positive association between O3
levels and dye fading of nylon materials at an 03 concentration
of 0.2 ppm and at various relative humidities. In summary, dye
fading is a complex function of 03 concentration, relative
humidity, and the presence of other gaseous pollutants. At
present, the available research is insufficient for quantifying
the amount of damage to fibrous materials attributable to O3
alone. Anthraquinone dyes incorporated into cotton and nylon
fibers appear to be the most sensitive to 03 damage (CD, p.. 8-
50) .
The degradation of fibers from exposure to 03 is poorly
characterized. In general, most synthetic fibers such as
modacrylic and polyester are relatively resistant; and cotton,
nylon, and acrylic fibers show variable sensitivities to the gas.
Ozone reduces the breaking strength of these fibers, and the
degree of reduction depends on the amount of moisture present.
Under laboratory conditions, Bogaty et al. (1952) found a 20
percent loss in breaking strength in cotton textiles under high-
moisture conditions after exposure to a 0.06 ppm concentration of
03 for 50 days. They equated these conditions to a 500- to 600-
day exposure under natural conditions. Kerr et al. (1969) found
a net loss of 9 percent in breaking strength of moist cotton
fibers exposed to 03 at a concentration of 1.0 ppm for 60 days.
The limited research in this area indicates that 03 in ambient
air may have a minimal effect on textile fibers, but additional
research is needed to verify this conclusion (CD, p. 8-50).
-------
X-45
3. Paints
The effects of O3 on paint are small in comparison with
those of other factors (Campbell st al., 1974) . Past studies
have shown that, of various paints, only vinyl and acrylic coil
coatings are affected (Haynie et a!., 1976) , and that this impact
has a negllgible effect on the useful life of the material
coated. Preliminary results of current studies have indicated a
statically significant effect of 03 and relative humidity on
latex house paint, but final results are needed before
conclusions can be drawn.
Pigments in artists' paints have also been tested under
controlled conditions for 3 .onths at an average exposure level
of 0.4 ppm of o3. While fading occurred in anthraquinone-based
Pigments, no quantitative information on dose-response
relationships is available.
4. Conclusion
Among the various materials studied, research has narrowed
the type of materials most likely to affect the ec from
increased 03 exposure. These include easterners and textile
ro, ' «• -
probably the most economically important. Hhile the limitations
of McCarthy et al. (1983) preclude the ^^ ^^^ °
the
d > gures ndicate the magnitude °f p«
damage from exposures to 03 in ambient air (CD, p. a-46) .
D. Effects on Personal Comfort and Well-Being
The Clean Air Act requires that secondary NAAQS for a
pollutant specify a leve! of air quality which is adequate to
protect public welfare from any known or anticipated adverse
effects, section 302 (h) includes personal comfort and well-beina
,n referring to effects on welfare. Those effects of a
on humans which are not identified as being adverse health
effects but do affect personal comfort and well-being are covered
by th» Provlslon. symptoms are defined generally as subjective
-------
X- 46
evidence of disease or physical disturbance but are not
necessarily adverse health effects.
Symptomatic effects associated with human exposure to O3 and
other photochemical oxidants may contribute to a reduction in
personal comfort and well-being. Similar, but not identical,
symptoms have been reported for clinical 03 and community-
photochemical oxidant exposures. Eye irritation, for example, is
commonly associated with ambient photochemical oxidant levels of
about 0.10 ppm but does not occur during controlled O3 exposures
at much higher levels than found in ambient air. Symptoms such
as nose and throat irritation, chest discomfort, cough and
headache have been reported at > 0.10 ppm O3 in epidemiology
studies (Hammer et al., 1974; Makino and Mizoguchi, 1975; Okawda
et al., 1979) and at > 0.12 ppm in controlled studies (McDonnell
et al., 1983; Avol et al., 1984; Kulle et al., 1985). Impairment
of athletic performance in high school students may have been
caused by irritative symptoms associated with 0.12 ppm oxidants
(Wayne et. al., 1967). Other symptoms commonly reported in
clinical 03 studies are throat dryness, difficulty or pain during
deep inspiration, chest tightness, substernal soreness or pain,
wheezing, lassitude, malaise, and nausea. CASAC recommended that
these effects be considered health effects in developing a basis
for the primary standard for O-,.
E. Related Welfare Effects Information and Issues
In its public meetings of April 20-21, 1986, December 14-15,
1987, and December 14-15, 1988 on the CD, staff paper, and NAAQS-
related analyses, the Clean Air Scientific Advisory Committee
(CASAC) discussed preliminary new information relevant to the
possible need for additional standard(s) to protect public health
from exposure to 03. At the time of the meetings much of the
information was unpublished and not incorporated into the CD.
Subsequently this new information base has been reviewed and
largely incorporated by ECAO into the CDS. For a more detailed
-------
X- 47
discussion and review of individual studies the reader is
referred to that document.
The purpose of this section is to provide an overview of the
key issues of concern for the secondary standard: relationships
of various air quality indicators, crop loss estimates, averaging
times and forest response to 03. This section will summarize new
analyses that attempt to address these issues and discuss key new
studies that relate to the issue of averaging time(s). The
review of new data in this section and in ECAO's Summary of
Selected New Information on Effects of Ozone on Health and
Vegetation will provide input to staff conclusions and
recommendations in Chapter XI. For convenience of readers,
citations for new research reviewed in this section are included
in as part of the reference list at the end of this staff paper.
The CDS provides a more current reference list and discussion of
03 welfare studies published since closure of the CD.
Past research, as well as current scientific opinion
indicate that the current air quality standard is not fully
protective of all vegetation (Reich and Amundson, 1984; CASAC,
1986). Yet because of uncertainties in the data base for crops,
and the paucity of data regarding forest response to O3/ the
averaging time, the level of the standard and even the need for a
separate secondary standard remain difficult issues to resolve.
The issue of averaging times and form of the standard has
proven to be one of the most difficult issues associated with the
review of the O3 secondary standard. EPA's previous
recommendation (CD) that serious consideration be given to
setting both a 1-hr and a longer-term secondary standard was
based on two findings:
• That forest trees, because they are perennials,
experience chronic 03 stress due to exposure to long-
term O3 concentrations
• That the relationship between peak values and seasonal
averages is generally not predictable with any degree
of confidence; therefore the 1-hr standard may not
-------
X- 48
adequately reduce the probability of a high chronic
exposure.
While CASAC members (CASAC, 1987) strongly endorsed the
judgment that repeated peaks are critical in eliciting plant
response, the Committee's views regarding the need for a long-
term standard were less clear. Although some members expressed
concern that long-term, low level 03 exposures could adversely
affect vegetation, there was reluctance to suggest a long-term
standard in the range suggested by the data (0.04-0.06 ppm)
because of the obvious conflict with background levels in many
areas of the country. The CASAC challenged EPA to identify a
single standard formulation that would offer protection from both
repeated peaks of concern and Long-term exposures. While monthly
standard and was specifically mentioned for consideration, it was
acknowledged that this exposure statistic would not be ideal in
terms of fully capturing extreme situations or biological
response.
In response to these CASAC comments several analyses were
undertaken and were presented at the December, 1987 CASAC
meeting. While CASAC found the results of the new analyses
interesting and encouraged EPA to pursue the analysis of
alternative exposure indicators; with Corvallis, they rejected
EPA's recommendation to continue the standard review until the
analyses on alternative exposure indicators is complete. Rather,
they endorsed the idea of making a decision on retaining the 1-hr
standard with the information currently available. There was
some difference of opinion among the committee members as to
whether the current standard or a lower level would be adequate
to protect vegetation. EPA has recently completed an initial
draft (Lee et al., I988c) of the analysis of alternative exposure
indicators requested by CASAC. These results are considered
preliminary and will be undergoing peer-review. In addition to
the analysis, a few new studies have been completed which look at
various aspects of the exposure dynamics issue. The analysis and
-------
X- 49
the new studies, as well as CASAC's response to them at the
December 1988 meeting, are discussed in Section E.3.
1. Air Quality Analyses
Air quality analyses and, in particular, relationships
between alternative air quality indicators, are critical in the
selection of an averaging time for the secondary standard. While
previous analyses have indicated that relationships between peak
and mean statistics are generally not stable or predictable with
any degree of confidence (Johnson et al., 1986; Lefohn, 1984;
Heck et al., 1984a, 1984b; Larsen and Heck, 1983) none of the
analyses specifically examined the type of compromise indicator
that was suggested by CASAC (CASAC, 1986). An analysis by
McCurdy (1987) is the latest effort to examine relationships
between various types of air quality indicators. These results
have been provided in separate reports and summarized in Appendix
A.
For purposes of this discussion, it is interesting to note
that after examining several short-term peak, multiple peak, and
long-term average indicators, McCurdy found that all three
exposure patterns of interest are highly correlated with two
monthly forms of the standard: maximum monthly mean of the 1-hr
daily maximums and maximum monthly mean of the 8-hr daily
maximums (SP, p. A-22). The first indicator performed better
than the second, but because they are highly correlated with each
other (r = .95), either one could be used. It appears from this
analysis that if a single secondary 03 NAAQS which provides
reasonable protection from both repeated peaks and long-term
exposures of concern is desired, it should be the maximum monthly
mean of the daily maximum 1-hr averages (max monthly mean). It
should be noted that this exposure statistic is largely based on
purely statistical concerns, as there is little or no data for a
one month exposure period. The same is likely to be true for
other similar compromise exposure statistics.
-------
X- 50
2. Crop Loss Estimates
As stated above, the max monthly mean appears to be an
appropriate statistic in terms of the protection it affords from
repeated peaks and long-term average concentrations of concern;
however, there is little or no effects data for a one month
exposure period. This makes it difficult to arrive at crop loss
estimates for a monthly indicator.
In an effort to arrive at crop loss estimates for a monthly
indicator, Larsen has analyzed new exposure indicators in his
existing crop loss model. This model (Larsen and Heck, 1984) was
adapted from the Lar sen-Heck lognormal plant, injury model (Larsen
and Heck, 1976) to estimate the impact of O3 on crop yield. This
model uses the "effective mean" 03 concentration (defined in the
next paragraph) to adjust for the greater effect of peak O3
concentrations. While calculation of the effective mean is '
somewhat cumbersome for standard-setting, the Larsen-Heck crop
reduction model does allow one to look at other exposure
indicators and the crop loss estimates associated with them.
The "effective mean" 03 concentration can be defined by
comparing it with the arithmetic mean. Estimates of crop loss as
a function of an arithmetic mean 03 concentration essentially
assume that crop loss is proportional to the summation of all
(seasonal daytime) 1-hr average O3 concentrations. Some studies
suggest (Larsen and Heck, 1984) that peak 1-hr average 03
concentrations cause a much greater than proportional effect,
suggesting that effects may be proportional to a new dose
parameter called air pollution "impact," a parameter that is the
summation of all (seasonal daytime) 1-hr average 03
concentrations raised to the 2.66 power. The "effective mean" is
the 2.66th root of impact divided by the number of hourly O3
concentrations that are exponentiated:
me = [(Ech 2'66)/n]
-------
X- 51
The arithmetic mean can be calculated by merely replacing 2.66
with 1 in this equation, thus showing the parallel construction
of the arithmetic and effective means.
Larsen et al. (1988) discussed several crop reduction dose-
response mathematical models and concluded that the lognormal,
Weibull, and Box-Tidwell models are all closely related. When
fitted to data from the National Crop Loss Assessment Network
(NCLAN), all three models produce similar dose-response plots.
The Larsen-Heck dose-response mathematical model was used to
estimate percent crop reduction for soybeans: the above
"effective mean" O3 concentration was the "dose" and the
lognormal model was used to estimate the "response." This model
was used to estimate soybean percent crop reduction at each
agricultural site in the National Aerometric Data Bank (NADB) for
each year of years 1981-1985.
Larsen et al. (1988) calculated fourteen O3 concentration
parameters for each site year of data. The potential ambient
standards that would limit crop reduction to 5, 10, 15 or 20
percent -of agricultural sites are summarized in Table X-8. The
three of these parameters that correlate best with crop reduction
are the effective mean O3 concentration '(I percent of the
variance unexplained), an arithmetic mean O3 concentration (4
percent unexplained), and the maximum l month mean of daily
maximum 1-hr 03 concentrations (15 percent unexplained).
Potential O3 standards can be selected to achieve a desired
response. For instance, if no more than 15 percent crop
reduction were desired at any agricultural NADB site, a second
highest daily maximum 1-hr 03 standard of 0.099 ppm could be
used. A maximum monthly mean standard (of l-hr daily maximum) of
0.072 would be required to achieve the same protection. Table X-
8 indicates that 65% of agricultural NADB sites achieved this
potential standard in 1981-1985 as opposed to 40% of sites that
achieved the second high daily max standard of 0.099 ppm.
-------
Table X-8. Potential Ambient Ozone Standards that Would Llimit Soybean Crop Reduction to 5, 10, 15, or 20 Percent
Ambient ozone standard that Percentage of agricultural NADB Percentage of variance
would limit soybean crop sites that achieved this potential unexplained by
reduction to stated percentage standard in years 1981-1985 regression of soybean
croo reduction attains*:
Ozone Parameter 5% 10% 15% 20% 5% 10% 15% 20%
Second highest daily maximum .059 .079 .099 .119 2 10 40 74
l-h ozone cone (ppm)
Maximum 1-month mean of daily .048 .060 .072 .085 5 26 65 91
maximum l-h ozone cone (ppm)
Maximum 1-month mean of daily .042 .052 .062 .073 5 26 64 89
maximum 8-h ozone cone (ppm)
Maximum 3-month mean of daily .034 .044 .054 .064 3 15 48 85
maximum 8-h ozone cone (ppm)
Number of days/yr that u-h 6 25 - - 17 61
daily max. cone exceeds
.08 ppm
Summer daytime arithmetic mean .035 .045 .055 .065 9 36 70 93
ozone cone (ppm)
Summer daytime effective mean .042 .052 .063 .073 14 45 77 94
ozone cone (ppm)
this ozone parameter
42
15
17
24
27
4
1
><.
u
-------
X- 53
As mentioned earlier the "effective mean" 03 concentration
adjusts for the greater effect of peak O3 concentrations.
Several sites may have identical arithmetic mean values, but the
higher the peak concentration at one of these sites, the higher
will be the effective mean value and the expected effects at that
site.
The exponent of 2.66 on the concentration term for the
effective mean was derived from injury studies and then applied
to yield studies without validating its applicability to yield
(CD, 6-159). Larsen et al. (1988) did an additional analysis on
this issue and concluded that crop reduction is probably closely
analogous to growth suppression, leaf dry weight reduction, and
percent leaf injury; thus, the average exponent of 2.66 for these
three effects was assumed to be a good approximation to use in
this model for crop reduction. Also, because of the low
variability of 03 concentrations measured at agricultural sites,
estimated effects would not change much if an exponent far from
2.66 was used.
The approach taken by Larsen et al. (1988) differs from that
of Lee et al. (I987a,b; 1988a,b,c). Larsen's estimated crop
reduction was calculated from a lognormal model that expresses O3
exposure as a function of the effective mean times the exposure
duration, i.e., the total impact (Larsen and Heck, 1985). By
definition, the estimated crop loss is totally determined by the
effective mean once the exposed duration is fixed. Since there
is no biological variation in the data, (as in Lee et al.,
1987a,b; 1988a,b) correlations between the estimated crop loss
and the exposure indices are, in fact, measures of association
between the (transformed) effective mean and the other exposure
indices. Thus, it is difficult to tell whether mean indices are
better correlated with plant response than other exposure
indicators based on Larsen et al (1988).
-------
X-54
3. Averaging Times
a. NCLAN/CERL Reanalysis
The discrepancy between the seasonal mean exposure indicator
used in the NCLAN studies and the repeated peak exposures
identified in the CD as being most important for plant response
make it difficult to evaluate the vegetation effects data base
for 03. Accordingly, OAQPS requested that the 03 team at
Corvallis Environmental Research Laboratory (CERL) conduct a more
detailed analysis of the crop yield response data from the
National Crop Loss Assessment Network (NCLAN) and from the
Corvallis laboratory.
Lee et al., (1987a,b) reanalyzed several NCLAN data sets
(soybean, wheat, cotton, and alfalfa) in an effort to determine
which exposure index produces the most accurate exposure response
relationship. Although the most widely used exposure index to
assess O3 effects on crops is the 7-hr seasonal mean, current
information about agricultural crops suggests that short-term,
high concentration exposures appear to be more detrimental to
crop production than long-term, low concentration exposures. The
seasonal mean ignores many factors known to affect vegetation
growth including phenological response, peak concentrations,
length of episodes and days between peaks (CD, 1986).
Lee et al. (1987a,b) examined several exposure statistics
that emphasize peak concentrations more than lower concentrations
as alternatives to the seasonal mean. These include: 1)
summation of all concentrations above a cut-off level, 2) number
of hourly concentrations above a cutoff level, 3) summation of
all concentrations raised to a power greater than 1. in
conjunction with 3), an exponential weighting scheme over time
was used to account for differences in response at various stages
of phenological development; this is described as phenologically
weighted cumulative impact statistics (PWCI). Exposure indices
that included all the data performed (24 hr) performed better
than the indices that used only 7 hours of data. The 7-hr
seasonal mean was never "best" and was near optimal in only 5 of
-------
X-55
14 cases. The exposure indices that emphasized peaks performed
better than those that gave equal weighting to all
concentrations; indices that accumulated the exposures performed
better than those that averaged the exposures.
Because of the extreme importance of this type of analysis
in resolving the issue of averaging time, and form of the
standard for the secondary standard, Corvallis has conducted
increasingly more detailed assessments of alternative exposure
statistics using various weighting schemes for relating exposure
NCi^T/eSP°nSe' ^ a m°re eXtensive ^respective analysis of
NCLAN data, Lee et al., (.I988a,b) fit 24 common and 589 general
phenologically weight cumulative impact (GPWCI) functions to the
response data from seven crop studies. The criteria established
for determining "best" exposure indices were those that displayed
the smallest residual sums of square error when the yield
response data were regressed on the various 03 exposure indices
using the Box-Tidwell model.
The "best" exposure index was a GPWCI with sigmoid
weighting. Cumulative indices (with concentration thresholds)
performed as well as the GPWCIs while mean indices did not
perform as well. The authors concluded that,
"While no single index was deemed "best" in relating o,
exposure to plant response, the top performing indices were
those indices that (1) cumulate the hourly O3 concentrations
over time, (2) emphasize concentrations of 0.06 ppm and
higher either by continuous sigmoid weights or by discrete
0-1 weights of the threshold indices and 3) phenologically
weight the exposure such that greatest weight occurs during
the plant growth stage. When assessing the impact of o, on
Plant growth, these findings illustrate the importance of
repeated peaks, and the time of increased sensitivity in
assessing the impact of O3 on plant growth."
Although peak concentrations should be given greater weight, the
authors suggested that lower concentrations should also L
included in the calculation of an exposure index
-------
X- 56
Tingey et al. (1988) is essentially a condensation of the
paper by Lee et al. (I988a,b) and therefore the conclusions are
basically the same. However, the paper does show the importance
of exposure duration and the limitation of the seasonal mean to
specifically incorporate duration. For example,, the mean does
not distinguish among exposures to the same concentration (mean)
but of different durations (e.g., 10, 50, or 100 days).
The most recent analyses by Lee et al. (I988c), developed in
response to a request by CASAC (December, 1987), evaluates
selected ambient O3 air quality indicators and estimates the
exposure levels associated with agricultural losses. The results
of this initial draft analysis are discussed in detail in Lee et
al. (1988c). This section will highlight key points regarding
the methodology and the results. Ideally, the selection of the
most appropriate exposure indices would be derived from
scientific principles and be validated by experimental data
specifically designed to identify the most appropriate exposure
indices. Such an approach, however, would delay resolving the
selection of exposure indices for years until the necessary
experiments were designed and conducted. Since the likelihood of
the Agency funding another major crops research program in the
near future is small, EPA has developed an alternative approach
of conducting a retrospective analysis of existing plant-response
data. While recognizing the limitations of the original NCLAN
data, CASAC has requested further analysis of the exposure
dynamics issue through retrospective analyses. The major
limitation of such analyses is that the specific studies were not
designed.for developing exposure indices or testing exposure
hypotheses; consequently, the usability of the data is somewhat
limited.
Plant yield and hourly O3 monitoring data were obtained for
twelve NCLAN field experiments (soybean, winter wheat, corn,
sorghum and cotton). The twelve NCLAN studies represented five
species with differing cultivars for a total of 17 individual
cases. Crops were grown according to standard agricultural
-------
X-57
practices and exposed to a range of concentration in open-top
chambers according to NCLAN protocols. These crop yield data
ml
model (defined in Appendix A of Lee et al., I988b)
Selection criteria for determining the "best" exposure index
was based on the minimum residual sum of squares (Rss, for each
of the 17 cases. Ranking and selection of exposure indices was
performed separately for each case. For each individual case
the relative performance of an index was measured as the ratio of
its RSS to the minimum Rss. For overall oomparlson
and cultivars, the exposure indices were evaluated according to
three criteria: (l, the average score calculated as the mean
relative RSS.s; and (3, percent variation (i.e., range/mean score
Based on the criterion of minimum RSS, no single exposure
index performed "best" for all seventeen cases. The best overall
fits are obtained using indices that:
1. cumulate hourly concentrations overtime,
2. place greater weight on concentrations of 0.06 ppm or
higher either by continuous slgmoid weight or by
discrete o-l weights, and
3. weight hourly concentrations according to the
phenological stage of plant development.
The results indicate that while the generalized phenological
cumulative impact (GPWCI, indices best related plant respons! to
the
there
Uded 3 •1»»"— ***- integrated index
as an at "
indices th, 9U3llt nar' c™ul"-e censored
indices that integrated concentrations of 0.06 (or 0.07, ppm or
higher (SUM06 and SUM07, also performed well, suggesting that
ambient 03 levels below O.OS ppm are important in triggering
a"
b r Lefohn et al.
.b) and Lee et al. (I987a,b; I988b, who used NCLAK data and
-------
X- 58
cumulation indices with sigmoid and allometric weights in
demonstrating the importance of peak concentrations in
determining plant response.
The integrated exposure indices (SUMO6 and SUM07) used to
characterize experimental exposures are functions of exposure
duration and concentration. Correspondingly the exposure levels
associated with various yield losses calculated from experimental
data are functionally related to exposure "seasons." An implicit
assumption here is that exposures of varying duration, with equal
values of SUM06 (or SUM07) cause the same biological response.
According to Lee et al. (1988c) experiments replicated in time
and/or space, differing in exposure duration, should produce
identical predicted relative yield losses using SUM06 or SUM07.
Therefore, the discrepancy between exposure "seasons" and between
experimental and ambient averaging times are accounted for in the
integrated indices. Thus, these results indicate the integrated
indices capture the key components of exposure, are adequate
descriptors of plant response, and are simple and easy to
implement from a regulatory perspective.
The magnitude of the O3-induced yield losses varied among
sites, years and species/cultivars. The NCLAN studies used in
this analysis represent both 0;! sensitive and tolerant
species/cultivars, and multiple years, and experimental sites.
Crop yield losses for 16 NCLAN studies are contained in Table
3.3. The predicted relative yield losses for the four exposure
indices across the 16 cases are shown for the mean, the 25th,
50th, 75th, and 85th percentiles.
As can be seen in Table x~9, median crop losses of 18.9% are
expected in agricultural sites experiencing 03 exposures with a
HDM2 value of 0.12 ppm. Lowering the concentration to 0.10 ppm
would result in an estimated median crop loss of 12.8%, whereas
going to 0.08 ppm would result in an estimated median crop loss
of 7.9%. Table X-10 presents the exposure levels associated with
mean and percentiles of predicted relative yield loss of 5 to 30%
-------
TABLE X-9
rPRvo a?d mean Predicted relative yield losses
(PRYLs) associated with various levels of the four
ras
HDM2
Percentiles nfL
50th
1.7%
4.2%
7.9%
12.8%
18.9%
32.6%
48.7%
MEAN
6.3%
9.9%
13.7%
17.8%
22.3%
33.0%
48.6%
- 75th
9.8%
15.9%
22.1%
28.8%
34.8%
46.0%
58.3%
15.0%
21.7%
28.0%
33.8%
39.7%.
50.5%
60.4%
Source: Lee et al.
.(1988c)
-------
"ABLE X-10.
I
Exposure levels associated with predicted relative yield losses
(PRYLs) of 5 to 30% for the four exposure indices, HDM2, M7,
SUM06, and SUM07, for the 16 NCLAN studies. Separate
calculations are performed using the 50th, 75th, and 85th
percentiles, and the mean PRYLs for the 16 NCLAN studies.
A) Exposure levels with Mean PRYLs of 5 to 30%.
Mean Predicted Relative Yield Loss
HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUM07 rDDnrt
5%.
0.032
0.027
9.7
7.5
10%
0.061
0.045
17.1
14.4
15%
0.087
0.059
23.7
20.9
20%
0.110
0.070
29.8
26.8
25%
0.131
0.080
34.3
32.3
30%
0.149
0.089
38.8
37.2
B) Exposure levels with SOtli Percentile PRYLs of 5 to 30%,
50th
HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUM07 ronun
_5%
0.065
0.042
14.1
11.6
10%
0.089
0.057
21.3
17.9
15%
0.107
0.068
28.0
- 24.3
20%
0.123
0.077
33.1
30.5
25%
0.137
0.085
37.5
34.9
»-t j-n-ijj
30%
0.152
0.093
42.2
39.2
C) Exposure levels with 75th Percentile PRYLs of 5 to 30%.
75th Percentile Predi
HDM2 (ppm)
M7 (ppm)
SUMO 6 (ppm)
SUMO 7 fDDirO
5%
0.023
0.028
10.2
6.9
10%
0.041
0.039
16.6
13.0
15%
0.057
0.048
22.4
18.5
cted, Relative Yield To-**:
20%
0.073
0.056
27.6
23.7
25%
0.089
0.066
31.3
28.8
30%
0.104
0.075
33.7
32.6
D) Exposure levels with 85th Percentile PRYLs of 5 to 30%.
85th Percent!lg Predicted Relative Yield Loss
53 10% 15% 20% 25% 30%
HDM2 (ppm) <0.020 0.026 0.040 0.055 0.070 0.087
M7 (ppm) 0.026 0.035 0.045 0.054 0.062 0.069
STJM06 (ppm) 3.7 7.5 10.9 13.8 16.7 19 6
SUMQ7 fppm) 2^2 S^J 8.7 11.7 14.7 17.R
Source: Lee et al. (1988c)
-------
X- 61
for the four indicators across all sites, if EPA
-------
X- 62
In regard to the exposure period of interest, an air quality
analysis of 83 non-urban site-years of ambient O3 data indicate
an averaging time of five months from May to September (153 days)
should be used to capture 90% of the hourly concentrations of
0.06 ppm or higher. This period includes the EPA-defined O3
season for most sites, represents the period when the second
highest daily maximum 1-hr concentration generally occurs, and
coincides with the agricultural season for the majority of crops
grown in the United States.
b. New Studies
Although only a limited number of studies have been
conducted with the specific objective of developing or evaluating
various exposure indices, several studies have attempted to use
existing exposure response data for evaluating a range of
exposure indices. The results of these retrospective analyses
have provided useful concepts and the conclusions are in general
agreement. However, because the studies were not specifically
designed to evaluate various indices, the differences in and
among exposure treatments may be relatively small. As a result,
the power of these studies is Less than desirable.
In a small retrospective analysis, Lefohn et al. (I988a) has
come to similar conclusions as Lee et al. (I987a,b) using
different NCLAN data sets with the Weibull and linear models. As
in Lee et al. (I987a,b), he is comparing the use of several
indices of exposure in describing the relationship between O3 and
reduction in agricultural crop yield. Prior to these studies, no
attempt had been made to determine which exposure response models
best fit the data sets examined. Lefohn, et al. (I988a) used
hourly mean concentration data based on 2-3 measurements per hour
to develop indices of exposure from soybean and winter wheat
studies conducted in open-top chambers at the Boyce Thompson
field site. The efficiency of using seasonal mean and cumulative
indices (i.e., number of occurrences equal to or above specific
hourly mean concentrations, sums of all hourly mean
-------
X-63
concentrations equal to or above a selected level, and the
weighted sum of all hourly mean concentrations, to describe the
relationship between exposure to 03 and reductions in the yield
of agricultural crops has been evaluated. In most cases the
Weibull model, a functional form used extensively by the NCLAN
program, tends to fit the data based on cumulative indices better
than on seasonal mean indices. This conclusion is similar to
what Lee et al. (1987a,b) found with different models and data
sets, in addition, two other recent papers (Hogsett et al., 1988
and Musselman et al., 1988) support the same conclusion that 1)
mean indices are not among the best exposure indicators and 2)
the preferred (yield best statistical fit to the data, exposure
indices cumulate the exposure impact over the growing season and
preferentially weight the peak concentrations. While the
exposure indicators developed in these analyses can be
potentially complex, Hogsett et al. (1988) has recently found
that there are other simple indices such as summation of all
concentrations above 0.06 (SUM06, and 0.08 ppm (STO08) which are
almost as good as the complex PWd statistics. These potential
exposure statistics were evaiuated in greater depth in a new
analysis (Lee et al., I988b) previously discussed on p. x-48
abou/th ^7 ^ Lef°hn " a1' (1988a> haS ^-ated Discussion
about the interpretation of the data (Runeckles, 1988; Parry and
Day, 1,8.,. Both commenters felt that the data presented was
insufficient to support the conclusion that peak-weight exposure
did point out that peak-weighted indices
t as wel1 as mean indioes- The «—•*«•
the compilation of two years of wheat data with
markedly different exposure durations into a single model. i»
response, Lefohn et al. (lM8b, stated that the wheat data
support the need to include a cumulative component in an exposure
" "
that
<., . _, *«-=* ...» more relevant to use in
the standard setting process than seasonal means, which ignore
the length of the exposure period."
-------
X- 64
These conclusions on the importance of cumulative peak
concentrations in causing plant response are consistent with the
data presented in the CD as well as other recent studies. Wang
et al. (1986) characterized 03 exposure of three tree species as
the number of daily occurrences above 0.08 ppm and 0.12 ppm over
a four month growing season. The authors concluded that O3
significantly impaired the growth of hybrid poplar in the absence
of visible- injury. In a three year study with quaking aspen,
Wang et al. (1986) found that plant growth was reduced 12 to 24%
although the current national air quality standard was exceeded
in only one of those years. Adomait et al. (1987) characterized
O3 exposure of white bean as the cumulative 03 concentration
above a threshold of 0.08 ppm for a one month period. In
addition, Reich and Amundson (1986) stated that the O3 induced
decrease in growth of several tree species was directly related
to reduced photosynthesis, which was impaired by the cumulative
O3 dose.
Although the CD and the weight of the new evidence since the
CD review (with the exception of papers reporting results from
NCLAN) seem to suggest the importance of peaks, another issue of
concern is how to treat the peaks in the development of exposure
response indicators. While Lefohn et al. (1988a) recommended
that a weighting scheme which gives greater weight to higher
concentrations be used in developing exposure indicators, others
support the removal of some dose from each hourly 03 value so
that low levels which are less likely to effect plant growth are
eliminated (McCool et al., 1986). Basically, three weighting
approaches have been used: 1) a concentration threshold in which
the concentrations, number of occurrences or hours above the
threshold cumulated; 2) an exponential weighting in which all
concentrations are raised to a specific exponential power and 3)
a sigmoid weighting in which all concentrations are weighted with
a multiplicative weighting factor (which depends on
concentration). The functional weighting approach using either
exponential or sigmoid weighting seems preferable to the
-------
X- 65
threshold approach because it considers the contribution of all
concentrations in eliciting a response.
While several studies (Lee et al., I987a,b and 1988b;
Hogsett et al., 1988; Lefohn et al., I988a; Musselman et al.
1988) have focused on developing new exposure statistics, there
have been several attempts to evaluate the statistic used most
often in the current 03 data base - the 7-hr seasonal mean.
According to smith et al. (1987) the open.top cnMber method
which predictive models such as NCLAN are based do not reflect
the episodic nature of O3 pollution or some of the more important
environmental conditions that influence plant growth. The
authors concluded that the failure of the NCLAN predictive model
to account for such conditions might explain why they found no
effect of ambient 03 on yield of field grown soybeans in New
Jersey while the NCLAN model predicted losses of around 19%
Heagle et al. (1986, 1987) examined the seasonal mean
statistic in studies on soybean and tobacco. Both studies added
03 in constant or in variable amounts which were proportiona! to
ambient 03 concentrations, mterestingly enough, while there
were greater fluctuations and higher peak concentrations with the
proportional method than with the constant addition method the
two methods of addition gave similar seasonal mean
concentrations. Regression of yield response on 0,
concentrations showed no signlficant differences between the two
fc e * the
fact that the 7-hr seasonal mean, even with the proportional
addition, may fail to accurately reflect the elevated exposures.
*""
***" "»
seted «
selected is because it was less sensitive to variations in 03
patterns (cure, 1986) . Therefore, the conclusion in these
studies that differences in peak 03 concentrations do not
Rawlings et al. (1988) conducted additional analyses of the
soybean (Heagle et al. a986, and tobacco (Heagle et al. , 1987)
-------
X- 66
and
data suggested that the
contest, the 7
and that
limited. «*«-»«, the -best" exposure index is
<1985) COmP«ed various
rr • The
additiona! $2.225 billion
The information reviewed in this section suggests a
^w • exposure that the plant
-
4. Forest Risk Assessment
-------
X- 67
for 03. While the potential contribution of O3 and other
atmospheric pollutants has been receiving increasing scientific
and regulatory interest, the numerous and complex interactions
between 03, environmental factors, and health and productivity of
forests malce it difficult to distinguish 03 effects from natural
forest processes (Peterson and Violette, 1986). Although the
nature and magnitude of the effects and the precise contribution
of various atmospheric pollutants is not well understood there
is evidence that some sensitive forest types are negatively
affected by O3. The effects data, summarized in Chapter X of
this staff paper, is somewhat scarce at this point in time and
tends to focus more on controlled studies of seedlings or
saplings. The results of these controlled laboratory or
greenhouse studies are not easily extrapolated to field
conditions, thus making current extrapolation (e.g., seedlings to
mature trees, stands to populations) difficult.
a. Overview of Forest Risk Assessment
While significant new research is underway to improve our
knowledge of the exposure-response relationships of various tree
species and 03 and to develop our extrapolation modeling
capabilities, at this point in the standard review we must assess
a data base that has considerable uncertainty associated with it
Standard statistical procedures cannot quantify this type of
uncertainty because of the absence of valid statistical saxnples
across the relevant populations. This type of uncertainty,
however, can be represented using appropriate decision analytic
procedures to elicit judgmental probabilities (Walsten and
Whitfield, 1986; Peterson and Violette, 1986). EPA decided to
use this approach as a tool to better quantify the range of
scientific judgment and uncertainties regarding 03-caused risks
to forest resources.
This section win briefly describe the methods and
prenminary results of . study whioh ^^^ ^ ^^
around estates of forest-tree growth decrement and foliar
due to 03 exposures. This information is meant to
-------
X- 68
supplement scientific information and data by presenting an
interpretation of current scientific knowledge regarding O3
caused forest effects. It is extremely important to recognize
that the results presented here are not a substitute for
continued scientific research; clearly such research is needed
and the results of this work confirm that fact. However, because
the magnitude of forest damage caused by O3 is likely to remain
uncertain in the near term, policy analysts and regulators may
find information regarding the uncertainty among forest experts
useful in formulating policy options and better focusing research
efforts.
b. Protocol Summary
For judgmental probabilities regarding forest damage to be
useful to policy makers they must satisfy three criteria
(Peterson and Sueker, 1987):
1) Judgments should be made over the appropriate forest
response parameters for policy assessment - in this case the
endpoints of concern are growth decrement and foliar injury,
2) Judgments should be obtained from experts who are likely
to possess interpretations of O3 associated forest effects that
span the range of respected opinion, and
3) Judgments should be relatively stable over the course of
the study.
To satisfy these criteria, a protocol for probability
encoding was developed by the investigators and reviewed by EPA.
While detailed information on the protocol is contained in the
report (Petersen et al., 1987) it is useful to summarize several
of the basic assumptions made in the protocol:
1) Total above ground growth decrement and foliar injury
were selected as response parameters. Several other response
parameters were considered, but rejected as being beyond the
scope of this study.
2) While numerous O3-exposure measures were reviewed, two
were selected: the maximum 3-month mean for 24- and 7-hour (0900-
1600) daily averages. The selection of exposure statistics for
-------
X- 69
this analysis was extremely difficult. While new data indicate
the importance of repeated peaks in eliciting plant response (at
least for crops), several experts expressed concern about their
ability to reiate their judgments to a repeated peak statistic
not commonly .used in the effects literature. Rather, they seemed
more comfortable with a mean statistic, which was defined for
purposes of the study as including the episodic as well as the
low-level exposures seen by the trees. These very real concerns
strongly influenced the selection of the exposure statistics for
this analysis.
3) Trees in each O3 treatment "population" were assumed to
be exposed to 03 exposures represented by the 0.035, 0.055 and
0.085 03 maximum 3-month means for the 24- and 7-hour daily
averages. The concentrations were chosen to express the range of
°3 exposures over which growth decrement or foliar injury could
be detected in o.s. forests. Trees in the "control" population
were assumed to be exposed to concentrations ranging between
0.020 and 0.030 ppm 03 to represent naturally occurring
background conditions.
4) The following assumptions were made about the conditions
of the forests in the hypothetical studies:
a. The forests were mixed/monoculture type and of
uneven/even aged composition.
b. Forest trees were exposed to one of the 03 exposures
from germination to the time growth measurements were
taken.
c. Ozone sensitive genotypes were distributed in forests
according to the experts' beliefs.
d. Except in the use of plantation trees, trees were
entirely wild type genetically.
e. Site conditions, including slope, aspect, elevation,
soil type, air temperature and humidity, and pest and
pathogen populations were distributed across the
control and experimental populations as the experts
believed they were in the real world.
-------
X- 70
f. Growth effects were measured on all trees (no sampling
problems) at all locations. Measurements were assumed
to reveal effects on all age classes within a true
population.
g. Effects concerned only the species in question and not
second-order effects involving other species.
c. Scientific -Experts
In order for the judgments to be useful to decision makers,
they must be obtained from knowledgeable experts in the area of
03 effects on forests. The experts were selected to provide
judgments regarding O3 associated forestry effects spanning the
range of respected opinion. Establishing the range of opinion
and selecting the experts to be encoded may appear to be very
subjective; however, the field was narrowed considerably by
focusing on those individuals whose careers had been devoted to
conducting air pollution research (with a focus on 03) in the
area of forest effects. Recent reviews of 03 associated forest
.effects and references to the published literature also
facilitated the selection of experts. Several offices within EPA
(OAQPS, OPPE, ORD) 'collaborated, to develop a set of potential
candidates from which the five experts were selected. Names of
participating experts are provided in Table X-ll to facilitate
the interpretation of the range of responses by other forest
scientist and policy makers. All experts selected by EPA agreed
to participate with the understanding that their judgments would
remain anonymous. Particular responses or judgments for experts
A through E are identified only by randomly assigned code
letters.
d. Credible Judgments
To ensure that an expert's judgments were relatively stable
and accurately represented his uncertainty, a number of steps
were followed. First, a draft of the protocol was developed and
reviewed by the staff at EPA and the consultants prior to the
encoding session. Following the revision of the protocol, two
-------
X- 71
Table X-ll
Forest Response Experts
Expert
Sam Linzon
Samuel McLaughlin
Paul Miller
John Skelly
David Tingey
_.-
Ontario Ministry of Environment
Oak Ridge National Laboratory
USDA Forest Service
Pennsylvania State University
U.S. EPA, Corvallis Environmental
Research Laboratory
Souce: Peterson et al., 1987
-------
X- 72
separate meetings to encode the judgmental exposure-response
relationships were held with each expert. The first meetings
were conducted during November-December 1986; the second during
February-March 1987. During the first meeting, which lasted 6 to
8 hours, experts reviewed the protocol, discussed relevant
literature and research into factors determining exposure-
response functions and estimated probability distributions for
selected tree species. At the second meeting, the results of the
first session, which had been forwarded to the experts by mail,
were carefully reviewed and an additional probability encoding'
session was completed. Draft final results were distributed to
the experts and discussed in a telephone conference with each
expert.
e. Results
The findings suggest a wide range of opinion regarding tree
sensitivity to 03 among credible forest scientists. Three
different sets of results were developed as a result of the
analysis. The first set, discussed in sections 3.1 through 3.6
of the report, present 152 cumulative probability distributions
for growth decrement estimated by Experts A through E. These
judgments dimension the uncertainty over growth decrement in 19
different tree species due to six different 03 exposures.
Cumulative probability distributions across tree species and
experts reflect different underlying probability models. Some
cumulative probability distributions are virtually linear, others
reflect the normal cumulative probability model, and others a
"hockey stick" distribution. This variability prevented fitting
a single probability model to the cumulative probability
distributions developed in this study. For purposes of
description, individual judgments within any given distribution
were linked using cubic spline interpolation.
Figure X-3 summarizes these judgements across experts for
eastern white pine, deemed to be the most sensitive species by
Experts A, B, C and E at all 03 exposures. The vertical axes in
the figure represent percent growth decrement relevant to
-------
. fc
o
c
1-1
o
n
rt
n
<-i
01
o
3
10 —
f) f*)
•6-8
s!
S ^
c °
I"
c 3
• S
ft £.
5 |
V 3
| I 1 h
1
**
p
u •
in j
fo <
?] ^
PERCENT GROWTH DECREMENT
* S 3 8 S - s *
9-
s
«* §
p
b
o
I
r>
O
a.
O ^
b ^
tn ~
%
A
»O
O
S 8 S g g g
OQ
C
1
-------
X- 74
background conditions. The horizontal axes represent the six
different O3 exposure levels. The vertical barss at each 03
concentration represent each expert's maximum credible confidence
interval, assumed to be 98%, for growth decrement. The symbols
within each bar indicate respectively the .01, ,,25, .50, .75, and
.99 probability fractiles. Thus, in Figure X-3,, the maximum
credible interval for Expert C's judgment of growth decline at
exposures characterized by the 0.035 ppm O3 7-hour average
statistic ranges between a lower bound (F=0.01) growth decrement
of 0% to an upper bound (F=0.99) growth decrement of 15%. The
three intermediate symbols represent growth decrement at
cumulative probability fractiles .25, .50, and .75. Figure X-3
illustrates a number of important points:
• First, there is a wide range of uncertainty expressed
in the experts' judgments. A and E are consistently
the most confident, c the least.
• Second, uncertainty tends to increase as O3 exposures
increase.
• Third, there is no agreement across experts that growth
decrement due to O3 will occur (within the maximum
credible interval of 98%) until O3 exposures associated
with the 0.085 ppm 7 hour average statistic are
reached.
• Fourth, 2 of 4 experts believe that there will be
little, if any, growth decrement at the 0.035 7-hr
average.
The second set of results presents exposure-response
functions for O3 caused growth decrements over the range of 03
exposures considered in the study. The experts were unanimous in
their belief that exposures of 0.035 ppm 03 7-hr average would
not cause measurable growth decrement in the tree species
considered. These exposure-response functions are summarized in
Figures 3.15 through 3.30 of the report. Figures 3.31 and 3.32
(Peterson et al., 1987) presents upper and lower bound judgments
across experts, providing the largest possible credible interval
-------
X- 75
and the broadest interpretation of uncertainty. As can be seen
for eastern white pine, this interval is often quite large. For
example, extreme risk estimates range from 0 to 65% growth
decrement at 0.055 ppm 03, 24-hr average.
The third set of results focus on the judgements on foliar
injury to forest trees, Although the experts indicated
considerable uncertainty with respect to growth and yield
effects, their lack of confidence in risk estimates for foliar
injury were even more pronounced.
Peterson et al. (1987) was designed to explicitly present
the scientific uncertainty in estimates of 03 induced forest
decline. it is intended to augment scientific research by
reframing effects data so that non-scientists can better evaluate
X
does h •• *
does not, however, replace effects research as it did not
generate any new hypotheses or data. What it does is confirm
what many felt was the case - that there .is a wide range of
f°rest
regarding the role
and forestry effects. Further scientific research is
needed before confident estimates of growth decrement can be
attributed to o3. Much of this research will be conducted during
the next several to five years under NAPAP.s Forest Response
Program and EPA.s research program, the Effects of TropLpheric
Ozone on Forest Types. F"*pneric
-------
-------
XI- 1
XI. Staff Conriln.
Secondary .Standard
Drawing upon the evaluation of scientific information
contained in the CD and CDS, this section provides preliminary
staff conclusions that should be considered by the Administrator
in selecting the pollutant indicator, form, averaging time and
level of the O3 secondary standard. The primary focus of this
discussion will be vegetation and ecosystem effects, as this data
base provides the best support for the secondary standard for o,
Materials damage and effects on personal comfort and well-being
will be covered in the final section. Risk and benefit analyses
coverzng vegetation effects are currently underway as part of a
continuing program to assess secondary standard effects.
A. Pollutant Indicator
On February 8, 1979, the chemical designation of the 0
primary and secondary standards was changed from photochemical
oxidants to 03 (44 FR 8202). EPA changed the designation of the -
standard to 03 since the Federal Reference Method (FRM) for
determining compliance specifically measured 03 as a surrogate
for total oxidants, and because a substantial vegetation effects
research base has established 03 as being chiefly responsible for
the adverse effects of photochemical air pollutants, largely
because of its relative abundance compared to other photochemical
oxidants.
The toxicities of the peroxyaoetyl nitrates, of hydrogen
peroxide (H202,, and of formic acid are less well documented than
the toxioity of 03. These oxidants have been the focus of
considerably less research because the levels at which they occur
in the ambient air, even in urban areas, appear to warrant much
less concern (CD, p. 7-1).
Peroxyacetyl nitrate, perhaps the best studied of the non-0,
oxxdants, is produced by photochemical reactions similar to those
that produce o3. controlled exposures of plants to PAN have
indicated that PAN can cause foliar injury and yield losses in
-------
XI- 2
sensitive cultivars of leafy vegetable crops (p., X-22) . However,
a comparison of PAN concentrations likely to cause either visible
injury or reduced yield with measured ambient concentrations (CD,
Chapter 5) indicates that it is unlikely
that PAN related effects occur in plants in the U.S. except in
some areas of California and possibly in a few other localized
areas. Because phytotoxic concentrations of PAN and the other
photochemical oxidants generally occur at significantly lower
ambient concentrations and are less widely distributed than O3,
the focus of this standard review will be the effects of 03.
The question of whether 03 can serve as an abatement
surrogate for controlling other photochemical oxidants is
addressed by Altshuller (1983) who concluded that "the ambient
air measurements indicate that O3 may serve directionally but
cannot be expected to serve quantitatively as a surrogate for the
other products." This conclusion-appears to apply to the subset
of photochemical products of. concern - 03, PAN, PPN, and H2O2 -
identified in the CD, even though Altshuller (1983) examined the
use of O3 as an abatement surrogate for all photochemical
products. Lack of a quantitative, monotonic relationship between
03 and other photochemical oxidants is discussed in Chapter 5 of
the CD in which average PAN/03 ratios for different sites and
years vary from 9 to 3. In addition, no single measurement
methodology can quantitatively and reliably measure O3 and other
photochemical oxidants, either individually or in ambient air
mixtures.
In spite of the above limitations, it is generally
recognized that control of ambient 03 levels currently provides
the best means of controlling photochemical oxidants of potential
welfare concern (03/ PAN, PPN, and H202) . This recognition along
with a controiled-exposure, welfare data base which implicates
only O3 among the photochemical oxidants at levels commonly
reported in ambient air, supports the recommendation that 03 be
retained as the pollutant indicator for controlling ambient
concentrations of photochemical oxidants. Unless significant
-------
XI- 3
additional evidence which demonstrates welfare effects from
exposure to ambient levels of non-03 oxidants becomes available
it ls the staffs conclusion that 03 remains a reasonable
surrogate for protection of public health from exposure to
photochemical oxidants.
B. Form of the Standard and Averaging Time(s)
The current secondary 03 NAAQS is expressed as an hourly
average which is the concentration not to be exceeded on more
than 1 day per year on average. During the last standard review
the decision was made to change from the deterministic to the
statistical form of the standard. The deterministic form, which
permitted only a single hourly exceedance of the standard level
in any given year, did not adequately deal with the variations in
03 concentrations which are largely due to the random nature of
meteorological factors affecting formation and dispersion of o
in the atmosphere, in addition, EPA further modified the
standard so that one expected exceedance would be given a daily
interpretation; that is, a day with two hourly values over the
standard level counts as one exceedance of the standard level
rather than two. it is recommended that the statistical form
(i.e., a number of expected exceedances allowed per year) be
retained for the secondary standard.
"" """* aPPr°priate "eraging time to protect
and ecosystems from exposure to 03, questions
immediately arise concerning which components of an exposure are
most important in causing vegetation responses. The studies
conducted thus far offer some useful information in assessing the
answer to this question, but to date research has not yet cllarly
defined which components of exposure are the most critical in
eliciting plant responses (CD, p. 6-7) . in lignt of these
uncertainties, the selection of an averaging time which
correlates well with the effects of concern is a difficult tart.
to the m" "t Ut"e C°nSenSUS " ^ S0ienti"= ~-»ity as
to the most appropriate summary statistic, but researchers are
-------
XI- 4
currently using many different exposure statistics in their
studies, thus making comparisons between studies; extremely
difficult. This section assesses the information available on
which to base a judgment regarding the averaging time for the
secondary standard.
Much of the research over the last three years has been
driven by these concerns related to the issue of averaging
time(s). In particular, the discrepancy between the seasonal
mean exposure indicator used in the NCLAN studies and the
repeated peak exposures identified as being most important for
plant response have motivated several retrospective analyses
which have attempted to investigate alternative exposure indices
using existing experimental data. Based on the weight of the new
evidence from analyses of plant response data from NCLAN and
other research programs, EPA concludes that exposure indices that
"best" related to plant response were those indices that
cumulated hourly concentrations over time and gave greater
weights to higher concentrations (Lee et al., 1987a,b; 1988b;
Hogsett et al., 1988; Lefohn et al., 1988). These results
support the conclusions of the CD and the CDS on the importance
of peak concentrations and the cumulative impact of O3 in
relating plant response to exposure.
EPA has responded directly to CASAC's challenge (CASAC,
1986) to identify a single standard formulation that would offer
protection from both repeated peaks and long-term exposures of
concern in a series of analyses designed to identify the most
appropriate exposure indicator. The most recent analysis (Lee,
et al., 1988c) proceeded along two lines: (l) regression
analysis of plant response data against various exposure indices
to determine indices that "best" depict biological response, and
(2) ambient air quality analysis of ambient 03 data to identify
indices that correlate well with various exposure patterns.
While the results of these analyses have been discussed in some
detail previously (SP, p. x-47), the key conclusions also provide
valuable insight regarding the issue of averaging times.
-------
XI- S
Among the indicators Lee et al. (19880, analyzed was the
second highest daily maximum l-hr concentration (HDM2) because it
10SS
th h aPPr°Xlmation °f the =^ent air quality standard and
the 7-hr seasonal mean (M7) , because it is the statistic most
often used in the current data base for crop loss. According to
Lee et al. (19Mo, , neither the single pea)c ^^ ^ ^ »
term averages are biologically relevant. The HDM2 and M7 indices
do not adequately describe the temporal variations in exposure
and do not relate well to plant response. Thus, a major
' n0nCUnUlatiVe indices is «>e exclusion of exposure
' The importance of this factor in determining piant response
*s readily apparent when examining yield loss results for the
wheat cultivar VONA, which is extremely sensitive to 03. m the
VOKA wheat results, there is greater variation in the redicted
as "°<~*tive indices such
as HDM2 and
(F.gures 3.1, 3.2, Lee et al., 1988c) were used
than when yield loss was estimated using the SUM06 and SUM07
3.3, 3.4, Lee et al., 1988c) exposure indices. The lower
YTr VariabUity in PRYLS *« the cumulative indices
and 07, are used are iargely explained by differences in
ortn
or the M7. Larsen et al. (1983) also points to the
l^tations of the present !-hr O3 NAAQS in controlling crop
reductxon because it forces some sites to control more than is
necessary ln order to protect against a given level of crop lL
S^lany, cure et al. (19M) have alluded to some limitations of
the M7 such as the fact that the seasonal mean characterizations
of O, exposure were much less sensitive than the l-hr max to
yearly concentration variations in O3 patterns. This latter
pomt supports the conclusion of Lee et al. (I988c) „.,„ hh
speoifioally ^^ expQsure ^ £.£ - he
concentrations to the
-------
XI- 6
The results of Lee et al. (1988c) indicate that while the
generalized phenological cumulative impact (GPWCI) indices best
related plant response to 03 exposure, there were other indices
that were near optimal. These indices included sigmoid-weighted
integrated index (SIGMOID centered at 0.062 ppm) which are much
too complex to implement in the air quality management system of
air pollution control. The cumulative censored indices that
integrated concentrations of 0.06 (or 0.07) ppm or higher (SUM06
and SUM07) also performed well, suggesting that ambient O3 levels
below 0.08 ppm are important in triggering plant; response and
should be included in an exposure index. The cumulative censored
indices are adequate descriptors of exposure which relate well to
biological response, and are simple and easy to implement from a
regulatory perspective. As stated earlier, these results support
the conclusions reached by Lefohn et al. (1988) and Lee et al.
(1987a,b; 1988b) who used NCLAN data and cumulative indices with
sigmoid and allometric weights in demonstrating the importance of
peak concentrations in determining plant response. In addition,
the results indicate that fair to strong associations exist
between the cumulative censored indices (SUM06 and 07) and the
peak and mean indices. The integrated indices SUM06 AND SUM07,
are strongly related to M7, and somewhat less related to HDM2.
Thus, the relationship between SUM07 and HDM2 falls just below
the level defined by the authors (Lee et al., 1988c) as
indicative of a strong association.
These results suggest the cumulative indices SUM 06 and SUM
07 correlate well with a long-term index (M7) and a short-term
index (HDM2) and relate well to biological response. Thus these
indicators have potential for setting a standard that protects
against adverse effects from repeated peak and long-term
exposures. While these indicators appear to be promising, EPA
staff concludes that the preliminary nature of the methodology
and results make a decision to change the averaging time and form
of the standard premature at this point in time. EPA further
believes that the analysis should be reviewed further by the
-------
XI- 7
d^"""!!!^1117 and CASAC before the resuits are used ^
standard form and averaging time
The analysis, although preliminary, does provide CASAC with
an additional piece of information regarding the crop loss
expected at alternative 1-hr standard levels. As pointed out
previously, these results are derived differently than Larsen et
«1. (1988, in that actual air quality and plant response data
were used in deriving exposure-response functions. Nevertheless,
the two analyses represent different approaches to estimating
crop !oss for alternative exposure indicators and provide EPA
With information that was not available previously, nilu the
form of the 1-hr standard has definite limitations, EPA staff
concludes that establishing a 1-hr averaging time standard in an
appropriate range (level of the standard is discussed in the next
section, represents the best staff recommendation that could be
made to the Administrator at this time to close out the review of
the scientific data. »ith this portion of the review complete?
and after considering CASAC
-------
XI- 8
economic losses because of the cost and time involved in such an
effort. The views of CASAC and the Administrator regarding the
need for such an analysis would have to be taken into
consideration.
Secondly, the option of continuing the review may appeal to
those who do not find the current form of the standard to be a
suitable air quality control target and would prefer to use one
of the cumulative indicators suggested by Lee et al. (I988c) once
the analysis has been finalized. Additional time for review and
revision of this analysis would allow the scientific community an
opportunity to review the alternative indicators and move toward
a consensus regarding selection of the most appropriate exposure
indicator.• The liability of this alternative is that it
postpones action on the secondary standard and thus, fails to
utilize new and existing data to assess the most appropriate
exposure indicator or the protection afforded by the current 1-hr
standard.
While evidence exists to support the hypothesis that for
agricultural crops, cumulative, repeated peak exposures appear to
be more important than long-term, low level concentrations, there
is still considerable uncertainty regarding the O3 distribution
patterns that affect trees. Some CASAC members objected to the
attempt in the previous staff paper to draw similarities between
the responses of trees and crops to 03 and to the suggestion that
the exposure pattern of interest for crops might be the same for
trees - that is, repeated peaks over time. Perhaps there are no
lessons from the years of research on crops that can be applied
to forestry. On the other hand, it seems incumbent upon staff to
examine the total data base, look at similarities and differences
between crops and trees, and lay out plausible hypotheses that
will spur debate on the key issues which science has not yet
resolved. It is with this in mind that staff examines the data
on exposure of trees to 03 and notes a good many studies that
report the exposure period in terms of hours above given
thresholds.
-------
XI- 9
Mclaughlin et .1. (19aa) reported that patterns of annual
growth in which pine along the Cumberland Plateau reflected a
loss in capacity for short-term recovery and eventual loss of
tree vigor. The primary cause of the observed decline was
chronic exposure to elevated concentrations of O3, possibly
accompanied by low levels of S02 and other pollutants. Reports
of oxidant injury on white pine in rural Virginia (Hayes and
SXelly, 1977, and on the Cumberland Plateau (Mann et al. 1980)
indicate that the injury appears to be associated with total '
oxidants concentrations of 0.08 ppm or higher. Hayes and SXelly
(1977) reported concentrations which equalled or exceeded 0 08
ppm for 104 hours at one of the rural sites they monitored in
western Virginia, skelly and Johnston (1978, also reported that
.03 concentrations during July of 1977 were above 0.08 ppm 30
percent of the time at the same location. In addition
relatively new evidence with slash pine Hogsett et al. (1985)
suggests that regimes containing peaX 03 concentrations elicit a
greater response than regimes containing mostly lower
concentrations over similar time periods. Also, Wang et al
(1936, conceded that the 20 days when 03 exceeded concentrations
of 0.08 ppm and the 1 day when the concentration exceeded 0 12
ppm was responsible for significantly impairing the growth of
hybrid poplar.
While assessment of the above evidence suggests the
otrees °
of trees to 03, further research is needed to confirm or deny
this hypothesis. As Peterson et al. (1987, point out, there is
currently considerable uncertainty regarding the effe^s
forest trees, in addition, there is considerable uncertainty
regarding the O3 distribution patterns that effect trees
consequently, EPA staff recommend that until these scientific
uncertainties can be resolved, no separate secondary standard
'~ '— **-
d ~ — **- - that the secondary
should be based on protection of vegetation
-------
XI- 10
C. Level of Standard
A critical factor in determining the level of the standard
is deciding which effects are to be considered adverse. There
are two important dimensions to this question. The first is what
should be measured. The second is how big does the change in the
selected measure have to be to be considered adverse. Answers to
the question of what should be measured fall, broadly speaking,
into one of two groups, physical measures and economic measures.
The most commonly proposed physical measure for commercial crops
is a change in yield. In addition, the bulk of the data
available since the last standard review uses growth and yield
loss as the response measure. While a change in yield loss is
the most commonly proposed physical effect measure for commercial
crops, a broader range of effects could be considered adverse.
It is common to segregate the effects of air pollutants on
vegetation, into "injury" and "damage" categories (Guderian et
al., 1985). Injury encompasses all measurable plant reactons,
such as reversible changes in metabolism, reduced photosynthesis,
leaf necrosis, leaf drop, altered quality or reduced growth that
does not influence all effects that reduce the intended use or
value of the plant or plant system. Value or intended use must
be viewed from the perspectives of both; (l) value to society or
human welfare and (2) value to the organism itself (i.e.,
reproduction) or the surrounding community. From this
perspective damage is equivalent to an "adverse" effect. In this
context, yield loss is easily identified as an adverse effect.
Concerning the magnitude of the change that is of concern,
one could pick some absolute value (e.g., tons) or percentage
(e.g., l%, 10%); but there is an element of arbitrariness in the
selection of any particular numerical value. Clearly, the answer
to the question of what constitutes an adverse effect is a policy
judgment. Previous attempts to answer this question by imposing
a specified percentage yield decrement to be avoided were
criticized by members of the scientific community at the 03 CASAC
meeting of April 1986. Some CASAC members argued that 10 percent
-------
XI- 11
loss in yield may not be a meaningful criterion for all crops and
that economic significance needs to be considered when
determining the adversity of a given effect. For example, a 10
percent decrement on a low valued crop does not have the same
welfare effect as a 10 percent decrement on a high valued crop
An economic measure of adversity can serve to weight yield
effects according to their relative impact on the well-being of
agricultural producers and consumers.
For these reasons, a measure of adversity such as percent or
absolute decrement in crop value is useful in addition to a
physical effect measure such as percent yield and growth
decrement. However, the range of economic effects resulting from
03 reductions will not be uniform across agricultural producers
consumers, income groups, and geographic areas. Thus, a single'
economic measure of adversity may be less appropriate
than a more extensive analysis that considers the'distribution of
economic effects.
Clearly, the answer to the question of what constitutes an
adverse effect is a policy judgment. The definition of "adverse"
ultimately involves an interplay among scientists and society
represented by the Administrator, while the role of the
scientist is to identify or describe the effect and explain why
it is adverse, the Administrator must determine if he is willing
to accept the impact/consequence of the effect or take action
(i.e., expend money and resources) to limit the consequences of
the adverse effect. Information on both physical effect measures
and measures of economic welfare should be useful in making this
policy judgment. y
Two national models currently exist that have used NCLAN
crop-response data, along with adequate aerometric and economic
information, to estimate the magnitude and distribution of
agricultural economic benefits associated with the control of o,
(Adams et al., 1984 and Kopp et al., 1985). Both of these models
:tfits as the change in producer and —
after O3 control. These models can be used to
-------
XI- 12
calculate economic measures of adversity (such as percent change
in farmer profits), as well as to describe the distribution of
economic impacts.
A major shortcoming of the Adams et al. (1984) and Kopp et
al. (1985) models is that neither model accounted for on-going
U.S. agricultural subsidy policies. Agricultural subsidies tend
to distort prices such that they do not reflect the true value of
additional production and thus make the value of increased
production difficult to assess. The forum of agricultural price
support policy can have a substantial impact on the benefits
realized from any reduction in 03. Measures of economic benefits
must be based on realistic assumptions concerning the features of
agricultural policy and benefit measures must be interpreted as
being conditional upon the existence of the policy being assumed.
The original Kopp-Adams measures of benefits may be
misleading because they ignore crop surpluses caused by
agricultural subsidies. McGartland (1987) argued that increased
production due to decreased 03 could exacerbate crop surplus
problems, and that the Adams et al., 1988 and Kopp et al., 1988
models substantially overestimated the economic benefits from 03
control. Madariaga (1988) argued that additional surpluses are
not likely to occur under current provisions in the law. The
Food Security Act of 1985 allows for subsidy reductions and
requires acreage restriction increases to counter additional
surplus production. However, despite the provisions of this Act,
future farm policies and the resulting economic effects on
agriculture are uncertain.
Because of CASAC's interest in economic measures of
adversity, EPA conducted an illustrative analysis (Table XI-l)
which presents national benefit estimates for nine crops, three
air quality scenarios, and three measures of benefits using a
modified version of the Kopp et al., 1988 model. The nine crops
that have been studied were selected based on economic importance
and data availability. The three crops soybeans, corn, and wheat
alone account for more than one-half of the value of all U.S.
-------
XI- 13
sales of agricultural crops. The estimates in Table xi-i
represent benefits from reducing seasonal ambient 03 levels (7-hr
means) to 60, 45, and 30 ppb under three different assumptions
regarding agricultural policies, in response to CASAC's
recommendations, three separate measures of economic benefits are
presented. Measure (i), the original Adazas/Kopp approach
assumes no agricultural market distortions. Measure (2) the
approach suggested by McGartland, accounts for market distortions
but assumes no acreage restriction increases or other policy
adjustments to counter additional surplus production. Measure
(3), the approach suggested by Madariaga, is a cost savings
measure of benefits that takes both market distortions and policy
adjustments to increased surpluses into account. The benefit
estimates for soybeans, peanuts, barley, sorghum, oats, and
alfalfa were calculated as the simple change in producer and
consumer surpluses after 03 control, since the prices of these
crops are not supported.
Given present agricultural policy, the benefits of
controlling o3 are partly dissipated by the costs of the
additional excess production of farm crops. This can be seen by -
comparing measures l and 2 in Table xi-i. An alternative policy
that effectively prevented an increase in excess production would
permit the realization of greater net economic benefits from o,
control. This can be seen by comparing measure 3 with measure 2
in the table.
The total benefit estimates presented in Table XI-l can be
used as measures of the agricultural welfare impacts resulting
from glven 03 changes. Adversity may better charactered using
reduced economic value and physical effect decrements.
Nevertheless, the choice of a specific threshold level of
economic benefits that is to be considered as significant, is
still somevhat arbitrary and open to subjective judgment.
Further, distributional concerns may be of importance. Results
fron the modified Kopp et al. model indicate that the
distribution! impacts on agriculturaZ producers and consumers
-------
XI- 14
from reducing 03 will be non-neutral and will depend on farm
policies. For example, corn, wheat, and cotton producers would
benefit substantially from O3 reductions if producer prices for
these crops were fixed by agricultural policies, in contrast, it
was estimated that soybean producers would actually lose profits
after O3 reductions since increased production would cause large
price declines. The big gainers in this case would be soybean
consumers. Any use of economic information to assess the welfare
impacts from changing 03 levels should take these considerations
Table XI-1. U.S. Agricultural Welfare Benefits from Reducing
Rural Ambient Ozone (7-hr seasonal means) to 60,
45, and 30 ppb for Three Alternative Benefit Measures
(millions of 1986 $)
Crop Benefits
Ozone
•
Reduction Benefit
(PPb)
60
60
60
45-
45
45
30
30
30
Measure
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
Soybeans
4
4
4
248
248
248
927
927
Corn
0
0
8
77
30
314
415
105
927 1019
Wheat
0
0
3
169
101
435
426
259
1000
Cotton
6
6
9
639
617
642
1351
1250
1107
Other
Crops*
N/A
N/A
N/A
N/A
N/A
N/A
103
103
103
•
Total
10
10
24
1133
996
1639
3222
2542
4156
NA = Not Available
*0ther Crops include: peanuts, barley, sorghum, oats, and alfalfa.
into account. In part, because the views of CASAC regarding the
use of economic information in determining adversity are so
divergent, EPA plans to provide the Administrator with
information on physical effects and economic measures of
-------
XI- 15
adversity to use at his discretion in reaching a final decision
on the O3 secondary NAAQS.
During the last SP review the only approach available for
analyzing estimated crop loss for alternative averaging times was
that of Larsen et al. (1988). The preliminary results of Lee et
al. (1988C) provide an alternative approach for evaluating
alternative exposures indices and the crop losses associated with
them. It should be noted that while Lee et al. (I988c) is
confmed to a subset of the NCLAN data, three of the crops
studied (corn, soybeans, and wheat) account for more than one-
half of the value of agricultural production in the U.S. In
addition, the cultivars selected were the dominant cultivars in
the crop growing regions at the time of the study. Therefore
these results provide the information necessary to assess the
appropriate range for a l-hr secondary standard.
CASAC's past comments regarding the range for the l-hr
standard reflect a division of opinion among the Committee
inembers. At the April 1986 meeting the Committee indicated that
the current standard was not protective of vegetation, but no
alternative level was suggested. At the December 1987 meeting
there was support expressed for the current 0.12 ppm standard as
well as for 0.10 ppm. Based on CASAC's written comments as well
as comments made at the December 1987 meeting, a range of 0.08 -
0-12 ppm appears to capture the range of opinion expressed.
that °f ^ St a1' (1988C>' " •
that the upper-end of this range offers very little protection
for vegetation. As can be seen in Table X-9 (SP, p X-52)
median crop losses of 18.9% are expected in agricultural sites
h^rTlln9 °3 eXP°SUreS With a HDM2 value of 0.12 ppm. Thus,
half of the crops in this study experienced losses of 18 9% or
greater upon attainment of the current standard, if EPA decides
to use a more protective strategy such as the 75th percentile
rather than the median PH^s, the crop loss in agricultural
"lie xTT ^ °'12 ^ ^^ C°Uld g° - ^ - 34.8%
(Table X-9, SP, p. x-52). in addition, a clear majority of CASAC
-------
XI- 16
members at the December 1988 meeting felt that the current
standard was not protective of vegetation. Therefore, staff
concludes that 0.06 ppm - 0.10 ppm is an appropriate range for
the 1-hr standard.
Lowering the HDM2 to 0.10 ppm would result in an estimated
median crop loss of 12.8%, which could go as high as 28.8% when
using the 75th percentile of PRYL's (Table X-9, SP, p. x-52).
Similarly, lowering the HMD2 to 0.08 would result in an estimated
median crop loss of 7.9% which would go as high as 22.1% when
using the 75th percentile of PRYL's. EPA concludes that in order
to be protective of anything other than median crop losses, the
bottom end of the range needs to be lowered to 0.06 ppm. This
would allow for a 15.9% or greater yield loss in 25% of the crops
at 0.06 ppm as compared to 22.1% or greater loss in 25% of the
crops at 0.08 ppm.
While these results are preliminary, they do provide
additional information on which to base a decision on the level
of a 1-hr standard. CASAC comment on the methodology and the
results of this analysis help determine how these results should
be factored into the range. In the judgment of EPA staff, a
range of 0.06-0.10 ppm takes into consideration the results of
Lee et al. (I988c) as well as CASAC comments regarding the upper
end of the range.
D. Staff Conclusions
Based upon the assessment in Section XI.A-E, the staff
conclusions regarding the welfare effects of O3 are as follows:
1. In consideration of the large base of welfare
information attributing effects to 03 exposure and the limited
evidence which demonstrates welfare effects from exposure to
ambient levels of non-O3 photochemical oxidants, there appears to
be little evidence to suggest a. change in chemical designation
from 03 to photochemical oxidants.
2. Because the bulk of the data available since the last
standard review uses growth and yield loss as the response
-------
XI'17 luce
measure, this parameter could be utitt»«d to evaluate different
levels of the secondary standard. Alternatively a broader range
of effects could be considered adverse-including: foliar injury
premature senescence, reduced photosynthesis, altered carbon '
allocation and reduced plant vigor, ^Although there is no precise
relationship between these effects ana-growth and yield
reduction, these more subtle effects'^ often early warning
signals of potentially harmful effe«:*=on plant vigor. In
addition, economic measures of advtesffcy such as changes in
producer and consumer surpluses resting afterO3 control may be
useful in evaluating alternative levels of the secondary
standard . ve
3. The weight of the recent evidence seems to suggest that
long-term averages, such as the 7-ho«*l seasonal mean, may not be
adequate indicators for relating exposure and plant response
4. Repeated peak concentrations Accumulated over time are
the most critical element in determining plant response for
agricultural crops. Exposure indicates which emphasize peak
concentrations and accumulate concentrations over time probably
provide the best biological basis fofostandard setting
Unfortunately, the analysis currently available on such
indicators is preliminary. EPA ..«, CASAC advice and comment on
the methodology as well as the suggested exposure indicators
(SUM06 and 07) . . ,a .,
5. There is currently a laokrsrcsxposure.response
information on forest tree .M.ct.rodM addition, there is a
broad range of uncertainty among statists regarding o, effects
imoorTt treSS'. COnSe9Uent1^ «»»**• •» -nsensus on the most
important averaging time for pereftnNls or on the precise role of
03 vs. other pollutants in causing --fSMst decline. Therefore
offt
of forest trees is not warranted, 'hat
«. Given the lack of effects. data on forests and the
preliminary nature of the Lee et al. (19e8c) results regarding
selection of the appropriate exposure statistic for crops th!
-------
XI- 18
EPA staff concludes that it may be premature at this point in
time to change the form of the standard and averaging time, it
is our judgement that a 1-hr averaging time standard in the range
of 0.06 - 0.10 ppm represents the best staff recommendation that
could be made to the Administrator at this time to close out the
review- of the scientific data. With this portion of the review
complete, and after considering CASAC's views on all issues, the
Administrator will be in a position to make a regulatory decision
on how and when to best act on the 1-hr standard. This is
consistent with CASAC comments (December 1987 and December 1988
meetings) urging EPA to consider a 1-hr averaging time and act on
the existing state of science rather than extend the review until
a more exhaustive assessment is made of alternative averaging
times.
Alternatively, EPA could continue the standard review.
Additional time for review of information on alternative exposure
indicators would allow the scientific community the opportunity
to move toward a consensus regarding selection of the most
appropriate exposure indicator. The liability of this
alternative is that it postpones action on the secondary standard
and thus fails to utilize new and existing data to assess the
most appropriate exposure indicator or the protection afforded by
the current 1-hr standard.
7. There appears to be no threshold level below which
materials damage will not occur; exposure of sensitive materials
to any non-zero concentration of O3 (including natural background
levels) can produce effects if the exposure duration is
sufficiently long. However, the slight acceleration of aging
processes of materials which occurs at the level of the NAAQS is
not judged to be significant or adverse. Consequently, the staff
concludes that materials data should not be used as a basis for
defining an averaging time and concentration for the secondary
standard and that the secondary standard be based on protection
of vegetation.
-------
XI- 19
8. Effects on personal comfort and well-being, as defined
by human symptomatic effects, have been observed in clinical
studies at 03 levels in the range of 0.12 - 0.16 pp. and at
somewhat lower levels in extended exposure epidemiological
studies. These effects include nose, and throat irritation,
chest discomfort, cough, and headache. In addition, cough, chest
pam on deep inspiration, chest tightness, wheezing, lassitude
mala.se, and nausea have been reported in controlled O3 exposure
stud.es. CASAC recommended that these effects be considered
health effects in developing a basis for the primary standard for
-------
-------
A- 1
APPENDIX A: AIR QUALITY
,.°i
- °8-
suss s-.i'Ss's.s.'gr.ss's •a? %»
EStap'S^Srsa^asia-
researchers in investioatino o ?™~ long-term mean used by NCLAN
(Heck and Tingey? Jlvi? All of ?h!fS -°2 .a9Jlcu""ral crops
-------
A- 2
-Air Quality), but data
of me F1rst ivo Air
The standard is attained when the expected number of days
y0^ of exceedances is determined by a
and ™« J0rmula explained in Appendix H of 40 CFR 50
and repeated here in simplified form to remove a term:
1-Hour ExEx = y *
n
where: v = number of measured 1-hour daily maximum values
>o . 12 ppm
N = number of days in the 03 season
2 = number of days assumed to have a 1-hour dailv
maximum <0.12 ppm
n = number of days having a valid 1-hour daily
maximum O3 value
This formula, and its application, is described in an EPA
guidance document (Curran, 1979).
The second air quality indicator of interest here is one of
the variant approaches to defining the characteristic highes?
concentration (CHC) . For our purposes, the CHC is that daiS
San^n/i? <3Uality ValU! tJ?3t is exPe<*ed to be exceeded iomore
than one time per year (during the -03 season/' actually) on
?ro™ T; ^ ar%a "UmbSr °f Ways ^° choose the CHC, ranging
from using fitted statistical distributions to a tabular "look-
ai^XS0?? (°HrHan' i979)' A modified version of the look-up
used^v «i!t lf\ ™ beCaifSe - it; is the simPle«t method and isP
used by most states in analyzing O3 air quality fAMTB 19811
The CHC under this approach is theVl highest" ob^ved daily
maximum value where n equals the number of years of valid data.2
qualityedatae f°°tnote on p' IV~1 regarding more recent O3 air
2A valid year of data is one that contains a valid dailv
maximum one-hour average for 75% of all the days in the state-
defined 03 season. A valid day has data for 75% of the hours
-------
A- 3
ui £ta,sthe CHC is the.4th-
of data, it is the second-highest value? " ***** 1S °ne
B.
50 stages? neauutv'Sata*!5"031 *"" (MSAs> in «»
with
for a MSA. This «as dne in MSAS * available
"
1- Expected Exceedances of the Existing ozone NAAQS
had .ore ?han onflxpectefexcLf^ <45'1% °f th°Se With
Thus, a signifiSanrro Par "* ™* °* NAAQS
between 9 am and 9
number of states, lt "6"^ ln
because exceedances are rare in
experience concen ™ " thSSS MSAs
on
daily lives, plople ivino in thi« S?^ 3? they g° about their
doing so, but *any people ?iiin« In evf*3 *?* Jh? fiatsntiai for
areas never experienceo3 le"^?! greate? SSf^o ,%y polluted °3
discussion of this, see Paul, et al (!|86? PP™' F°r a
-------
A- 4
and 1987. over 18 million people live in MSAs that had between
10 and 20 expected exceedances. Between
H,* ?art °f the cumulative frequency distribution of l-hour ExEx
^ JJ! °™ ln ??ble.A"*' along with ^lar information for
two other air quality indicators that will be discussed later
The median number of l-hour expected exceedances is 1.1; however
there is a lot of variability in the data. Although not SSSrtly
observable in Table A-l, 37% of the MSAs investigated4 have 0
exceedances. The mean of the sample, almost 5 exceedances, is
much greater than the median due to the very high number of
expected exceedances in the worst areas. The worst MSA is Los
Angeles, with about 125 exceedances.
Expected exceedances data for the 1979-1987 time period were
investigated in a sample of the 224 MSAs to determine if a
temporal trend existed in the l-hour ExEx indicator. A MSA was
included in this trend analysis if it had 3 or more valid years
of data in the 1979-1987 period. Under this criterion, 128 MSAs
had sufficient data for trend analysis. An area was considered
to have a trend in its expected exceedance index if the slope
coefficient for the temporal variable was significant at p< 10 5
Of the 128 areas, only l (
-------
A-5
Table A-l
CUMULATIVE FREQUENCY DESCRIPTIVE STATISTS ^cnr
WITH PEAK AND MULTIPLE-HOW of J£ A?OUALm
INDICATORS IN URBAN AREAS
Statistic
Mean
Std. Dev.
Minimum
Median
75th Percent.
90th Percent.
95th Percent.
Maximum
Sample Size
Expected
Exceedances
of 0.12 ppm
1-Hour Daily
^Maximum
0.0
1.1
4.0
10.0
14.8
125.3
224
2nd-High
1-Hour
Maximum
m]
.039
.119
.140
.167
.180
.340
224
Observed
Exceedances
of 0.08 ppm
8-Hour Daily
Maximum
0
9
18
29
38
150
224
Note: Std. Dev. = Standard Deviation
Percent. = Percentile
Source: McCurdy and Atherton (1988).
-------
A- 6
quite bad in 1988 over much of the United States, and using 1988
data in the regression analyses might change many of the
significant down-trending results.
Additional information on trend in the l-hour ExEx indicator
is available. EPA's Monitoring and Data Analysis Division (MDAD)
undertakes air quality trends analyses of all NAAQS pollutants on
an annual basis (MDAD, 1988).
MDAD's trend analyses used 242 sites to investigate if a
trend existed in the number of expected exceedances over the ten-
year time period 1977-1986. The composite average of the expected
number of exceedances decreased 38% during that time. 1979
1980, and 1983 values are higher than the other years.
2. Second-Highest 1-hour and 8-hour Daily Maximum Air
Quality Indicators
Cumulative frequency descriptive data for these two air
quality indicators appear in Table A-l, previously introduced
The Table indicates that the median 2nd-high 1-hour design value
for the sample of 224 MSAs is slightly lower than the current 0,
NAAQS level. The mean is higher than the current standard, 0.126
ppm, because of very high design values seen in a few MSAs. For
instance, 10% of the MSAs have a 2nd-high daily max >0.167 ppm.
The worst 5% have a design value >0.180 ppm. The highest design
value sites are located in MSAs near or adjacent to the New York,
Los Angeles, and Chicago major metropolitan areas.
Looking at the temporal trend in the 2nd-high daily max
indicator, the composite average of this indicator decreased 13%
between 1979 and 1986 (MDAD, 1988). This decrease is modest in
the post-1983 time period, and 1983 2nd-High values are higher
than adjacent years' values. Thus, the trend in this indicator
is not monotonic. If trend in individual MSAs is investigated,
rather than a composite average, it is obvious that there is a
cyclical pattern in 2nd-high daily max at most sites.
As seen in Table A-l, there are many expected exceedances of
an 8-hour daily maximum >0.08 ppm. There also is wide
variability in their number, from 0 for 63 MSAs (28% of the
sample) to 150 for the worst MSA. The mean is about 15, and the
standard deviation of the sample is larger than the mean.
3. Alternate 8-Hour Daily Maximum Air Quality Indicators
In previous 03 Staff Papers, the ShrDM indicator of interest
focused on a concentration level of 0.08 ppm. (See Table A-l.)
Recent interest has focused upon alternative concentration
levels, such as 0.06, 0.10, and 0.12 ppm. Descriptive
statistical data for these concentration levels appear in Table
A-2.
-------
A-7
Table A-2
Statistic
Observed
Exceedances
of 0.06 ppm
8-hour Daily
Maximum
Observed
Exceedances
of 0.08 ppm
8-hour Daily
Maximum
Observed
Exceedances
of 0.10 ppm
8-hour Daily
M ax imum
Observed
Exceedances
of 0.12 ppm
8-hour Daily
Mean
Std. Dev.
Minimum
Median
75th Percent.
90th Percent.
95th Percent.
Maximum
Sample Size
No. of MSAs >1
% of Sample
47.4
24.9
0
48
61
77
87
183
204*
197
96.6
14.9
20.9
18
f\f\
29
38
' 150
224
161
71.9
. -.
4.2
11.8
0
1
4
10
14
111
224
49
21.9
IIUA i mum
1.4
6 8
\j % %j
n
\j
0
o
2
5-
74
224
30
13.4
NOTE: Std7~Dev7
Percent.
Standard Deviation
Percentile
greatly affect the statistics shown ?or
Source: McCurdy and Atherton (1988).
indicator
W°uld
-------
A- 8
Note the great differences among the alternative 8-hour
indicators in the mean number of days above the various
concentration "outpoints." The average MSA has 3 times as many
days with an 8-hour daily maximum average over 0.10 ppm as it
does over 0.12 ppm. It also has about 3.5 times as Sany days
above 0.08 ppm as it does over 0.10 ppm.
Finally, Table A-2 shows that the average MSA has about 3
HowverS ?h?Y ^^ W*th an 8"hOUr °M >'06 PPm than da^S >'08 PP*-
However, this ratio is suspect because 8hrDM>.06 data are missing
for many California MSAs. Where the data are available, the mean
value for 8hrDM>.06 indicator would be much higher, as would the
ratioo (The median statistics are messed up for the same
reason.) Relationships among the various 8-hour air quality
indicators are more fully discussed in subsequent sections of
this Appendix.
Note the last two lines of Table A-2. They indicate the
number and relative portion of MSAs having one or more 8-hour DM
average greater than the cutpoint concentration shown. Only 22%
of the sample has one or more 8hrDM>.10 ppm, for instance, while
72% has one or more 8hrDM>.08 ppm.
4. Longer-term Air Quality Indicators
»
Two longer-term indicators of interest are the max monthly
and 3-month mean indices. The daily maximum averaging time used
for these indicators is 1-hour and 8-hour, respectively.
Cumulative frequency descriptive information for these two
indicators is shown in Table A-3.
There is moderate sample variability in the two indicators
as evidenced by the small difference between the mean and median
values and the small standard deviations relative to their means.
The large 3-month means of the top 5% MSAs in the sample should
be noted. The median value is 0.055 ppm. The worst area has a
maximum 3-month mean of 8-hour daily maximum averages of 0.132
ppm—higher than the current 1-hour daily maximum NAAQS! (The
area is Los Angeles.)
5. Relationships Among Air Quality Indicators in Urban
Areas
As an introduction to relationship analyses among air
quality indicators, we consider Figure A-l. Depicted are linear
and non-parametric correlation coefficients among the air quality
-------
A-9
Table A-3
^
Statistic
Mean
Std. Dev.
Minimum
Median
75th Percent.
90th Percent.
95th Percent.
Maximum
Sample Size
Maximum
Monthly
Mean for
1-Hour
Daily
Ma xi mums
.071
.017
.025
.069
.079
.086
.099
.178
224
Maximum
3-Month
Mean for
8-Hour
Daily
Maximums
.056
.012
.016
.055
.061
.067
.072
.132
222*
Note: Std. Dev. = Standard Deviation
Percent. = Percentile
Source: McCurdy and Atherton (1988).
-------
A-10
Figure A-l
CORRELATIONS AMONG SHORT- ANO LONG-TERM AIR QUALITY
INDICATORS IN MSAs (USING 2nd HIGH) Q
Exposure Pattern of Interest
Peak,
Short-Term
Multi-Peak,
Short-Term
Long-term
Average
.89
1 —
1
Expected
Exceed, of
1-Hour
Daily Max.
I
\&
2nd-
1-Ho
Da 11
Maxi
'
High
ur ^r
Tium
^
firt
.80
on
. oU
.70
Number
v Days> .
•^ 8-Hour
Maximum
/*"
,88
.88
\
of fi,
08 ppm *oc
Daily 4. "85 ;
— ^ Maximum
Month. Mean
_v of 1-Hour
Maximum "
3-Month Mean
» Of 8-Hour
Daily Max.
s
s
\
.95
.94
^ 1
Daily Max.
Source: McCurdy and Atherton (1988).
are depicted. A dotted line is used when the coefficient I Tell than'
|./5|. All coefficients are significant at p <.05.
-------
A- 11
depict
Kf • ,,
-
stratification is descred n ordy (if 8?^ USed *°r the
All coefficien dpte in « F-
Significantly different than 2e«"t p< of 9Ure A'2) are
-------
A- 12
Figure A- 2
AIR QUALITV
Exposure Pattern of Interest
Peak,
Short-Term
Multi-Peak,
Short-Term
Long-term
Average
.70
.74
2nd-High
1-Hour
Daily
Maximum
V
Expected .71
Number of
8-Hour Daily
Maximum
.88
.88
.80
.75
y
Maximum "
3-Month Mean
Of 8-Hour
Daily Max.
.95
.94
Maximum
Month. Mean
of 1-Hour
Daily Max.
Source: McCurdy and Atherton (1988).
-------
Proportion (*) Of Sites Exceeding Value
on Horiz. Axis
*» m TJ
< X 3D
ss°
» n
mm
o
3D
Vg2
o x _
OJ CD I-H
m 2
u jo
•D -O
3 O m
O O
jj ^>
rn
m n TJ
T3S
X J> CD
02^
3D m
0 X 3D
**• o m
n c: *>
r~ 33 tn
^55
•xPS
t« -< m
(a
c:
T
CD
I
CO
in
3D
D
t/D
3> 2
X G1
-------
Proportion (X) of Sites Exceeding Value on Horiz. Axis
ru
o
t_>
>• m T
< x 3 :
CD T3 M. I <
•> CD O 4. *J
Q> O ^^
to it-
(D CD
(0 Q.
i
i—•
4*
-------
Proportion (%) of Sitea Exceeding Value on Horiz. Axia
. M.
M.O
O
—J^^
IX)
o
(I)
o
Ik
o
Ul
o
0)
o
1
VI
o
GO
o
(O
o
-
o
o
c
-------
A- 16
„>. . The. Fel^t:L?nfhiPf am°ng certain air quality indicators are
graphically depicted in Figures A-3 through A-5. Figure A-3
relates the number of 8-hour daily maximum averages >.08 ppm (x
axis) to five alternative 1-hour standards: 0.08, 0.09, o 10
0.11 and 0.12 ppm. since the x-axis variable is an interval-
scale expected exceedance variable in this case, the curves
follow a negative exponential shape.
An example or two may aid the reader in reading Figure A-3
and subsequent curves in this Section. The Figure indicates that
if a 0.12 ppm 1-hour daily maximum standard is met by all sites
in the data set, 10% of all MSAs might have 13 or more days with
an 8-hour daily maximum >.08 ppm, on average. At least one area
might have as many as 30 days. Fifty percent of the MSAs might
have 4 or more days with an 8-hour ExEx indicator >.08 ppm, on
average. If the l-hour daily maximum air standard was lowered to
0.10 ppm and attained by all areas in the data set, 10% of the
MSAs may have 6 or more days over ShrDM > .08 ppm, on average.
If a 0.08 ppm daily max 1-hour standard were attained, no area
would be expected to have any day with an 8-hour ExEx value > 08
ppm.
Figure A-4 is like Figure A-3 except that the 8-hour average
cutoff concentration is now.0.06 ppm. Note that there are many
more expected exceedances of 0.06 ppm than 0.08 ppm for any l-
hour daily maximum standard (compare A-4 with A-3). The ratio is
anywhere between 8:1 and 20:1 for the various l-hour standards
analyzed.
It should be noted that there is no 0.11 ppm l-hour daily
max standard depicted in Figure A-4. It so closely followed the
0.10 ppm curve that it was deleted for the sake of clarity.
Thus, it can be said that there is no practical difference
between 0.10 and 0.11 ppm 1-hour NAAQS with respect to reducing
the number of days with a 8-hour daily maximum >.06 ppm.
In order to reduce the expected number of days with an 8-
hour 03 daily max >.06 ppm down to, say, 20 per year in the worst
10% of urban areas would require a 1-hour NAAQS of about 0.09
ppm. To further reduce these days to 10 per year may require a
1-hour 03 NAAQS of 0.08 ppm.
Figure A-5 depicts the relationship between alternative 1-
hour daily max standards and the number of days with an 8-hour
daily max average > .10 ppm. If the current 03 standard would be
attained everywhere, the 10% worst area probably would have only
slightly more than 1 day per year with an 8-hour daily max >.10
ppm, on average. There would be less than 1 day per year on
average in 98% of all areas if an 03 NAAQS of 0.11 ppm or lower
were attained everywhere.
-------
A- 17
tested.) y ^xmum and' cannot be
^covered that *
°'12 PE"° ^ «y maximum
2. are + 42-129% for the o.io pprc standard.
3- ^^pofs L'r^fpp^!"9 an intervai
4- u^Tthr^re't^^^^na^Lr
tS
Thus, there may be no
Say,
'- - "' '"
us-'"
-------
A- 18
nn !:h°Ur daily max >'°* PPm- Additional work is
underway to refine these uncertainty estimates,.
Numerous ratio analyses were undertaken among the urban O,
air quality indicators. For instance, the mean ratio of the 2nd-
highest 8-hour daily max to the 2nd highest l-hour daily max is
0.811 (s.d.=.082), with a range for the 224 area data set of
0.467 to 0.940. These statistics are similar to those found bv
other investigators. *
Other relational analyses focused upon exceedances of
various 8-hour outpoints vis-a-vis exceedances of the current O,
NAAQS. One such analysis fitted curves having the general form
8hr>(c) = a * (l ExEx ** b), where (c) represents the cutpoint
concentration of interest (e.g., .08 ppm, etc). The fitted curves
for R (c >.06) are depicted in Figure A-6. While the curves show'
the general relationship between the two air quality indices for
alternative cutpoints, they are misleading at ExEx=0 (at x=0)
At ExEx=0, the predicted number of exceedances of the alternative
8-hour cutpoints as depicted in Figure A-6 are all 0. In
reality, that is only true for R(.12); at an ExEx or x of 0, the
observed data indicate the following exceedances:
Rf.06) Rf .081 Rf ;iO) R ( . 1-2 )
Minimum o o o o
Median 37 4 o o
Mean 36.1 5.0 0.3 o
Maximum 102 23 6 o
Thus, the equations in general underestimate the number of 8-hour
exceedances for the 0.06, 0.08, and 0.10 ppm cutpoints for 0
exceedances of the 1-hour daily maximum NAAQS.8
While a statistic such as R2 is not readily available for
non-linear equations using the PC statistical package available
to us, one index of goodness-of-fit of the equation might be to
divide the "explained" sum-of-squares due to the regression
equation by the total sum-of-squares. Using this index, the
following values are obtained:
R(.06): 66% R(.08): 83% R(.IO): 96% R(.12): 98%
Another measure of the fitting ability of the equations is to
investigate the 95% confidence intervals for the two unknown
parameters, a and b. The predicted intervals were compared to
the estimated a and b values to give the following relative
confidence intervals for a and b, respectively:
R(.06): ±19%, ±51% R(.08): ±16%, ±7% R(.IO): +13%, +3%
R(.12): ± 21%, ±3%
-------
Figure A-6
«»«
0)
>
c.
3
o
c
o
0)
D
r-1
(0
C.
o
0)
0)
o
c
10
•o
0>
01
o
X
LJ
T3
0)
+J
u
0)
a
x
UJ
O
CO
200
160
160
140
120
100
80
60
40
20
30 60 go 120
1-hour Expected Exceedances
150
-------
A- 20
alt«rJ£??Vo J* J *? that you can have many exceedances of
alternative 8-hour daily maximum averages without equaling or
exceeding the current NAAQS. Attaining the 1-hour NAAQS is no
nr?tSn n*at altefnativ* 8-hour daily maximum, in the r^nge of
£v * H^°-?8HPpm Y11^ be atta^ed. This conclusion is reinforced
an Ah analysis of specific days in 21 MSAs having l-hour
JnH f^°UJ eX?a!dfnCeS °f the cutP°ints mentioned above (McCurdy
and Atherton 1988). Results of this analysis appear in Table A-
If a relative frequency viewpoint of probability is taken the
results indicate that if a day exceeds the current NAAQS
concentration level of 0.12 ppm then it is very likely (> 95%) to
exceed a 0.06 or 0.08 ppm 8-hour daily maximum' (DM) cutpoint. it
On thoth - -ur cupn.
On the other hand, a day can experience an exceedance of various
8-hour DM outpoints without exceeding the 0.12 ppm 1-hour DM
i™ i* Y ^ °f days exceedin9 an 8-hour DM cutpoint of 0.06
ppm also exceed a 1-hour 0.12 ppm concentration. This increases
to 51-s for a 0.08 ppm cutpoint and to 91% for a 0.10 ppm
cutpoint. Thus, it is very possible to have days exceeding the
** without e
6. Consecutive Daily Exceedances of the Current 1-Hour
Standard Level
•v nA number of CASAC members were interested in knowing how
likely it is to have two or more days in a row with a violation
of the current 1-hour O3 standard level of 0.12 ppm. An analysis
or tnis multiple-day phenomenon was undertaken for most of the
MSAs in our sample of 224. See Table A-5. The total number of
days in the sample with a 1-hour DM >.12 ppm is 1,002. About 58%
of these "violations" are single-day episodes. Almost 17% are
two-day episodes, while about 12% are three-day episodes. Four-
day episodes constitute about 7% of violation-days, and greater-
than-four-day episodes are 6% or so of violation-days. Thus
almost 75% of days having a 1-hour DM >.12 ppm occur singly or in
pairs. J
Perhaps a brief explanation of the data appearing in Table
A-5 is in order. The table provides information on both episode
length (the left-hand column) and how often that episode occurs
in an individual MSA (the second column) . In addition it
provides data on how many MSAs had a particular episode-
length/frequency combination. For instance, look at the 2 -day
episode length portion of the table. The first row indicates
that 28 MSAs only had one 2-day episode. The next row indicates
that 11 MSAs had two 2-day episodes, and so forth. The fourth
column simply is the product of the number of MSAs times the
episode-length/frequency combination. (For example, for the first
row of 2-day episodes: 2 x 1 x 28 = 56.)
-------
A- 21
Table A-4
Percentage of days that are:
1. >.12 ppm 1-hour DM* &
>8-hour outpoint shown
2. >8-hour outpoint shown .&
>.12 ppm 1-hour DM
3. >8-hour outpoint shown &
<.12 ppm 1-hour DM
Alternative 8-Hour Average Daily
Maximum Outpoints (in ppm)
0.06
99.7
26.3
73.7
96.7
51.3
48.7
58.6
91.0
9.0
Ua7ly Maximum
-------
A- 22
Taole A-5
ESTIMATED FREQUENCY OF DAILY OZONE EPISODES BY
LENGTH OF THE EPISODES
Length of Frequency
Episodes of Episodes
(Days) in an MSA
1
2
2
3
4
7
9
3
1
2
3
4
5
6
4
1
3
4
>4 (All)
Number of
Occurrences of
That Frequency
579
28
11
3
2
1
1
13
5
1
1
1
1
7
1
2
8
Total Percent
Episode- of Total
Pays Episode Davs
579 57.5
16.6
56
44
18
16
14
18
12.3
39
30
9
12
15
18
7.2
28
12
32
62 6.2
TOTAL
1002
100.0
-------
A-23
the >4 category
detail for all years of datl indi^^n* 4 ePlsodes in more
the country have episodes o? tSat ?ena^h ^nly a few areas of
South coast Air Basin olcaUfS?nia^;>, h°SS areas are the
region. The South Coast recency Ls h J greatef New ^ork City
to 30 days, and lengthlc^S™ days are rellS^T^* °f Up
^^
different states
s
valid ambient air quality dat stuctY' 88 monitoring sites with
.the following SAROAD land use codes':ldentlfled as having one of
32 - rural agricultural
41 - strictly remote
45 - other remote
are located hia sCesSsureauV^0^^ °f
Statistical Area (MSA) Thi£ fL? 5«!2 "de£lned Metropolitan
use coding classification SrS ar2al%r ^C°n5raVene the land
as are forested areas • tru 1 v^L? f°Und within an MSA,
except in unusLlclrc^llJncll T^T.f' howeYer' are not'
urban analyses reported in ?h?f ' o ?® S Used for the non~
(1987). see ?ha?P?eport ?or idd???oia? ^ described ^ "cCurdy
included, year of analvsif r addltlonal information on sites
highest i-hour and s^hoSr avlrages?"""96 by State' arid month of
short-term peak indicaors" * r lscussl°" Begins with
-------
A- 24
1. 1-Hour ExEx Indicator
A one expected exceedance indicator was calculated for each
nonurban monitoring site. For the 86 sites:
• 70% attained the 0.12 ppm 03 standard in the years
analyzed. J
• 76% had fewer than 1.5 expected exceedances of a 0 12
ppm daily maximum one-hour value.
• the largest number of expected exceedances at a site
was 27.3 days.9
Cumulative frequency distribution statistics for the one expected
exceedances air quality indicator is depicted in Table A-6. (The
other indicators shown in the Table will be discussed below ) As
can be seen, there is a lot of variability in the data.
The non-urban distribution is very different than the urban
distribution (compare Tables A-l and A-6). There are many fewer
expected exceedances of the current 03 standard in rural/remote
areas than in urban areas. This is to be expected, given the
urban origin of precursor pollutants that ultimately result in
high ambient O3 concentrations.
2. Second-High Daily Maximum
Cumulative frequency distribution data for the 2nd-high
daily max indicator in rural/remote areas is presented in Table
A-6. There is not much variability in the data, especially when
compared to that seen for the same indicator in urban areas.
While non-urban areas have similar, but somewhat lower, mean 2nd-
high "design values," there is a large difference in this air
quality indicator for the worst areas. The maximum 2nd-high
indicator seen in rural/remote areas is 0.16 ppm. It is 0 37 for
urban areas.
This site is Kern County, CA and it is very different than
most of the other rural/remote sites. It is the "outlier," so to
speak, for many of the air quality indicators investigated.
However, the County is definitely an agricultural area. Johnson
et al. (1986) indicates that it has the second-highest amount of
crop acreage and harvested acreage of the 84 counties in the
sample (two counties have two sites). Thus, the site belongs in
the sample and it indicates how much variability in air quality
is possible at rural sites.
-------
Table A-6
Statistic
^ — "™ ""^ """^ •""-"^—^•••••™^1
Mean
Std. Dev.
Minimum
Median
75th Percentile
90th Percentile
95th Percentile
Maximum
Sample Size
Expected
Exceedances
of 0.12 ppm
1-Hour Daily
Maximum
1 0
i .0
3 4
w «T
o n
U . VJ
0 0
\J 9\J
1.4
* * «
3.5
6 8
*s • W
27.3
86
2nd-High
1-Hour
Daily Max.
m"~ 1
... j
™"™^^~"**^^'~"^^-^^i*^^^.^^M
.11
/"»*•»
.02
.06
.11
T O
.12
.14
1 C
.15
.16
o^r
00
Expected
Number
of Daily
Maxs >
f\f>
.08 ppm
21.1
18.1
0
17
29
48
58
132
85*
°ne
Source: McCurdy (1987)
value for
-------
A- 26
3. 8-Hour Expected Exceedances
Cumulative frequency data for this indicator are seen in
Table A-6. There is relatively a lot of variability in this
indicator. Note the particularly large number of expected
exceedances for the worst site (Kern County, CA).
The data indicate that it is possible for rural and remote
areas to experience many days with a maximum 8-hour average >0 08
ppm. The median area, for instance, has 17 days over that
cutpoint concentration.
Although not shown here, the number of 8-hour expected
exceedances drop dramatically as the cutpoint concentration
increases. For instance, the median non-urban area has 2 days
over a 0.10 ppm daily max and no days over a 0.12 ppm cutpoint
(McCurdy, 1987). The mean values for these two indicators is 5.7
and 1.4, respectively.
4. Longer-term Air Quality Indicators
Data for the max monthly and 3-month mean indicators appear
in Table A-7. The 1-hour monthly mean is significantly higher
than the 8-hour 3-month mean for any percentile level chosen.
The two distributions are statistically significantly different
at p<.05, using a paired t-test.
5. Discussion of the 3-Month Mean and Other Multiple-Hour
Indicators
The 3-month indicator used here is different from that often
used to characterize O3 exposure to agricultural crops. Most of
these long-term studies, such as those conducted by the National
Crop Loss Assessment Network (NCLAN), use a "growing season"
fixed 7-hour (9am - 4pm) daily average exposure statistic.
This would be a difficult standard formulation to implement
because of variability in growing.season across the country.
While a three month period seems biologically relevant because it
approximates the average growing season for most crops in the
U.S., there is considerable uncertainty as to whether the fixed
daily daylight period captures the time interval of greatest
plant sensitivity. It certainly does not capture the time period
of maximum 03 in a day for many rural and remote sites (McCurdy,
1984). This is depicted in Figure A-7, which compares cumulative
frequency distributions of the 8-hour daily maximum, 7-hour fixed
time period, and 1-hour daily maximum peak air quality
-------
A-27
Table A-7
Statistic
Mean
Std. Dev.
Minimum
Median
75th Percentile
90th Percentile
95th Percentile
Maximum
Sample Size
Maximum
Monthly
Mean for
1-hr Daily
Max. Ave.
.066
.012
.048
.067
.073
.079
.086
.118
86
es: 5td. Dev. = standard Devi at ion
*tt!e
Maximum
3-Month
Mean for
8-hr Daily
.Max. Ave.
.054
.009
.029
.054
.059
.063
.067
.086
85*
°"e Slte h» • ->««1n« value for
Source: McCurdy (1987)
-------
A-28
Figure A-7
CUMULATIVE FREQUENCY DISTRIBUTION OF THREE PEAK
AIR QUALITY INDICATORS
• 9ain-4pm Fixed Time
Period
-•8-Hour Daily Max.
••••1-Hour Daily Max.
n ~ 86
.04
.10
.12
I
.14
I
.16
Ozone Concentration Level ( ppm )
-------
A- 29
ere
peak ai:
parohe sn sss £est r°ng an «»»^«
means are significant? dilfSSnt Tat po.80)T ' ' indicators
80)
fro* S
values respectively a;o9%46% d «n,° °-°f3 Ppm- These
%
hour onSr1 m"ns- T"ird-guarter 7
Mccurdy (1985). The Sean ?w ?h»"rf afeas.a« analyzed in
Undoubtedly the dif4rnc arS , ?• S^ndfrd deviation value.
iLTt
uncover a better exposure statistir oer indicators may
perspective .(Lee et af? 1988)^ 3 Plant dose-
6' AJeai10"^1^ Am°ng Mr Qualit* indicators in Non-Urban
the current standard in non^urbfn Jr-« Jndlcat^s that attaining
-
n r-«
8-hour daily maximums win no be ^IriTc^r^T^^ that
these "attaining areas, » as many as^L? JJSS^' ^ct in 10% of
hour daily maximum > o 08 ppm This n™2£ S °°2Ur With an 8~
0.10 ppm l-hour daily maxiX' standard * ^°PS tO 8 days for a
-------
A-30
Figure A-8
CORRELATIONS AMONG SHORT-TERM, MULTIPLE-PEAK AND
LONGER-TERM AIR QUALITY INDICATORS IN NON-URBAN AREAS
Exposure Patterns of Interest
Peak,
Short-
Term
Multiple-Peak
Short-Term
.63
V
2nd-High
Expected
Number of
^?ur £..:7.L..\ Days .08 ppm
Daily \ X fl-Ho
Daily
Maximum
8-Hour Daily
Maximum Ave.
.85
Longer-Term
Average
.76
y
Maximum
v 3-Month Mean
' of 8-Hour
Daily Max.
,76
Maximum
Monthly Mean
of 1-Hour
Daily Max.
.90
Source: McCurdy (1987a).
The sample size for all indicators is 86. A dotted line is used when
jne correlation coefficient between the two variables is less than
10.75.1 All coefficients are significant at p<.05.
-------
n>
Q.
fu
r+
Cu
to
3
<-i-
ZT
It)
Q.
r+
Cu
ro
Proportion |%) Of Sites Exceeding Value on Horiz. Axis
09
c
-J
n>
-------
O>
Q.
O)
£U
rt-
0>
fD
Q.
Q)
fu
fD
Proportion (%) of Sites Exceeding Value on Horiz. Axis
£
(0
-j
O)
I
t-«
o
>
N)
-------
O>
Q.
Cu
Proportion <*> of Sites Exceeding Value on Horiz. Axis
rt>
1/1
n>
QL
H-
Cu
O)
r*-
D 2
0) Q)
M- X
I—> M-
«< 3
3C 3
Q)
X Q)
M- |
3 Z
C O
~J (D
Q) Ol
(O D
(D
10 O
*D CD
C
•3
-------
fD
Q.
CU
c+
ft)
fD
l/l
O)
Q.
Ql
CU
t/1
n>
Proportion (%) of Sites Exceeding Value on Horiz. Axis
>
(Q
n>
-------
ct-
O;
rt>
O.
Oi
fD
00
0)
CL
O>
Qi
in
fD
nj
QL
Proportion <*) of Sites Exceeding
n
to
c
-J
I
I—»
CO
-------
A- 36
•
The curves shown in Figure A-9 are almost identical to those
shown in Figure A-3. This may indicate that air quality
distributions are practically the same for urban and non-urban
areas if the characteristic highest concentrations10 are
similar.
The proportion of non-urban sites exceeding various- max
monthly means given attainment of alternative 1-hour standards is
depicted in Figure A-10. It indicates that 10% of non-urban sites
attaining the 0.12 ppm 03 standard could have a max monthly mean
of 0.073 ppm or higher. If a 0.10 ppm 1-hour standard were
attained, the mean for the highest 10% of non-urban sites may be
>0.068 ppm. J
Figure A-ll depicts the relationship between attaining
alternative 1-hour daily max standards and the maximum 3-month 8-
hour daily max indicator in non-urban areas. Attaining a 0 12
ppm 1-hour 03 standard might keep all but 10% of non-urban areas
under a 0.06 ppm 3-month mean. The worst 10% 3-month mean drops
to 0.056 ppm for a 0.10 ppm standard and to 0.052 ppm for a 0.08
ppm standard.
Figure A-12 indicates what may happen to short-term peaks
when exceedances of a 8-hour daily maximum average are regulated.
Attempting to reduce short-term peaks by reducing the number of
days with an 8-hour daily maximum average >0.08 ppm is
inefficient. It takes a large reduction in expected exceedances
of a 0.08 ppm daily max to effectuate a small reduction in the 1-
hour design value.
A slightly more efficient relationship is shown in Figure A-
13. This Figure relates alternative max monthly mean standards
to 1-hour design values. Going from a 0.08 ppm to a 0.07 ppm max
monthly standard may reduce the design value from 0.125 to 0.118
ppm in the worst 10% of non-urban sites. A drop to a 0.06 ppm
max monthly standard might result in a very large reduction in l-
hour daily maximum CHC. For the worst 10% of non-urban sites
the CHC could be as low as 0.102 ppm.
10CHC; in this case, the 2nd-highest 1-hour daily maximum
design value.
-------
A- 37
Summary
86 (or urban and non-urban ar^l, r*speCtl"liy.*iySSS "" 22" and
-------
-------
B-l
APPENDIX B: GLOSSARY OF PULMONARY TERMS AND SYMBOLS**
Air
aiveoiar
Airway conductance (Gaw) , Reciprocal of ai
. T0
airway resistance. Gas
mouth and the alveoli
to airflow
opening at the
contrasted with AIR SPACE
or
bronchioles. To be
of allegyor
idiosyncratic hypersensitiviti4s
cells lining the
tissue, and a layer
a
State
assoclated with
o£ epithelial
the lungs. The
Adapted from Appendix A, Volume v of the CD
-------
B-2
Asthma: A disease characterized by an increased responsiveness
of the airways to various stimuli and manifested by siowina
of forced expiration which changes in severity either
spontaneously or as a result of therapy. The tlrm asthma
may be modified by words or phrases indica^ng III Biology
factors provoking attacks, or its duration. etioiogy,
Breathing pattern: A general term designating the
characteristics of the ventilatory activity, e.g. tidal
volume, frequency of breathing, and shape of th2 volume" time
Bronchiole: One of the finer subdivisions of the airwavs less
than l mm in diameter, and having no cartilage fni?s wSl.
Bronchiolitis: Inflammation of the bronchioles which may be
Jf^6 °r ?5ronic' If the eti°l°9Y is known, it should be
stated. If permanent occlusion of the lumens is present
the term bronchiolitis obliterans may be used.
Bronchitis: A non-neoplastic disorder of structure or function
or the bronchi resulting from infectious or noninfectious
irritation. The term bronchitis should be modified bv
appropriate words or phrases to .indicate its etioloqv its
chronicity, the presence of associated airways dysfunction .
or type of anatomic change. The term chronic bronchitis
when unqualified, refers to a condition associated with
prolonged exposure to nonspecific bronchial irritants and
accompanied by mucous hypersecretion and certain structural
alterations in the bronchi. Anatomic changes may include
hypertrophy of the mucous-secreting apparatus and epithelial
metaplasia, as well as more classic evidences of epltnellal
inflammation. In epidemiologic studies, the presence of
cough or sputum production on most days for at least three
months of the year has sometimes been accepted as a
criterion for the diagnosis.
Bronchoconstrictor: An agent that causes a reduction in the
caliber (diameter) of airways.
Bronchodilator: An agent that causes an increase in the caliber
(diameter) of airways.
Bronchus: One of the subdivisions of the trachea serving to
£?J!hJy,air1t5Land-fr2m the.lun
-------
B-3
Carboxyhemoglobin (COHb): Hemoglobin in which the iron is
associated with carbon monoxide. The affinity of hemoglobin
for CO is about 300 times greater than for O2f y-"»in
Chronic obstructive lung disease (COLD): This term refers to
diseases of uncertain etiology characterized by persistent
slowing of airflow during forced expiration. It is
recommended that a more specific term, such as chronic
obstructive bronchitis or chronic obstructive emphysema, be
used whenever possible. Synonymous with chronic obstructive
pulmonary disease (COPD).
Closing capacity (CC): Closing volume plus residual volume
often expressed as a ratio of TLC, i.e. (CC/TLC%).
Closing volume (CV): The volume exhaled after the expired qas
concentration is inflected from an alveolar plateau during a
controlled breathing maneuver. Since the value obtained is
dependent on the specific test technique, the method used
must be designated in the text, and when necessary
specified by a qualifying symbol, closing volume is often
expressed as a ratio of the VC, i.e. (CV/VC%).
Conductance (G): The reciprocal o£ RESISTANCE. See AIRWAY
FEVt/FVC: A ratio of timed (t = 0.5, 1, 2, 3 s) forced
expiratory volume (FEVt) to forced vital capacity (FVC).
The ratio is often expressed in percent 100 x FEV^/FVC It
is an index of airway obstruction.
Forced expiratory flow (FEFx): Related to some portion of the
FVC curve. Modifiers refer to the amount of the FVC already
exhaled when the measurement is made. For example:
FEF75% = instantaneous forced expiratory flow after
75% of the FVC has been exhaled.
FEF200-i200 = mean forced expiratory flow between 200
ml and 1200 ml of the FVC (formerly
called the maximum expiratory flow rate
(MEFR).
FEF25-75% = mean forced expiratory flow during the
middle half of the FVC [formerly called
the maximum mid-expiratory flow rate
(MMFR)].
FEFmax = tne maximal forced expiratory flow achieved
during an FVC.
-------
B-4
Forced expiratory volume (FEV) : Denotes the volume of gas which
capacity, e.g. (FEV^/VC) X 100. Forced
Forced vital capacity (FVC) : Vital capacity performed with a
maximally forced expiratory effort. *°nnea witn a
Functional residual capacity (FRC) : The sum of RV and ERV (the
Gas exchange: Movement of oxygen from the alveoli into the
pulmonary capillary blood as carbon dioxide enters the
^r'i Jr0m th? bl°°d- In broader terms' ^e exchange of
gases between alveoli and lung capillaries. •
H TraPPin
-------
B-5
Lung
«-
Maxina.y
synonymous with nxi
expiratory f low rate
rate (MMFR or
(V , : volume of air ,reathed in one minut£_
(fa) - sie VENTIWTION ( T) a"d breathing frequency
various inorganic salt sue atr
the
ease?Y VirUS' ni— 9-ism, or etiologic agent causing
action of
ingredient of photchecal so"3 a" N°" in the air' •»
-------
B-6
Pulmonary edema: An accumulation of excessive amounts of fluid
in the lung extravascular tissue and air spaces.
Pulmonary emphysema: An abnormal, permanent enlargement of the
air spaces distal to the terminal nonrespiratory bronchiole
accompanied by destructive changes of the alveolar £a?ls" and
™H??"VHV10US/ibr0sis- The term emphysema may be
™J« * by words or phrases to indicate Its etiology, its
anatomic subtype, or any associated airways dysfunction
Residual volume (RV) : That volume of air remaining in the lungs
be fndTS^1'6^1^1011- The meth°d of measurement should
< * appropriate
Resistance flow (R) : The ratio of the flow-resistive components
.of pressure to simultaneous flow, in cm H20/liter pi? ?ec
Flow-resistive components of pressure are' obtained by
subtracting any elastic or inertial components, proportional
respectively to volume and volume acceleration ?Mos? f low
w??iS^nCeS 1^thS resPirat°ry Astern are nonlinear, varying
with the magnitude and direction of flow, with lung volume
and lung volume history, and possibly with volume
acceleration. Accordingly, careful specification of the
conditions of measurement is necessary; see AIRWAY
Respiratory frequency (fR) : The number of breathing cycles per
unit of time. Synonymous with breathing frequency (fB).
Specific airway conductance (SGaw) : Airway conductance divided
TGV * Volume at which ifc was measured. SRaw = Raw x
C°lorless
-------
B-7
The method of
lung tissue (p, ^t^S-" RreS=iSRtan°°f
^* ~~
(chest, Whereit
ventilatory flow rate
is the product of SS
conditions
designates
V°lume>
V =
V =
ExPired volume per minute (BT.PS) ,
Inspired volume per minute (BTPS) .
ALVEOLAR) .
to
r
gas. Alveolar ventilton is
because when a total volSme of
spaces, the last part does no J
but occupies the dead space to
inspiration. Thus the ?o?Sme of
expelled completely
volume of the
(see VENTILATION,
»
reP.lac^d with fresh
tOtal ventilation
the alve°lar
SXelled fro™ the body
with the next
actually
from the
-------
Appendix C
UNITED STATES ENV,RONMENTAL PROTECT,ON AGENCV
WASHINGTON. D.C. 20460
May 1. 1989
Office OF
K-
Protectlon
Washington. DC 20460
Dear Mr. .Reilly:
»
CASAC did not r«.»rh
position paper recommendation tha?" the' r^l ,°n endo^ment of the staff
-------
-2-
margin of safety Is intended to provWe orSo "^ *« you are aware.
have not yet been uncovered by research , «E * ? *?tn* adverse effects which
a matter of disagreement. ^^11^^!^ ^T medlcal ^cancels
development of a standard wfth . ^ItSS^? * f* subcommi«ee favored
annual distribution of ozone conLSS)i?SSilly»,iIObu8t1. Upper bound °" the
expected exceedance form of the s?anda d Wht th r" ld^ncg On thc curr«"t
adv,ce on what form the Agency should consijlr Comrr,!ttee offers no further
which alters the
staff position paper. Within CASAC Y
felt that healthy individuals ^
any of the responses c^^
nuld to moderate respiratory symptoms) in thp
members believed that adverse S wou?d
mduced more severe effects fle
severe respiratory symptoms. The
influenced by recognition {h»t
represents a blending of scientific and
appropriate to inform^u of
are or are not adverse
presentat"°n of this issue in the
°p!nF°n; SOme me^ers
T™ ^^ !nduced
decrement in FEV ™
™ Paper' whlle alfe^
^ w:Perience^ until ozone
m°derate "'
health effect !ssue
we fee, it
potentlal for effects arising from
" -
based on recent controlled human
phalSUCh, Pr°lon8ed «posures
Further, for people exposed to
possibility that chronic Sfevers
changes have not been demonstrated
concern
a"d . toxlcol°gy studies.
«sp'ratory impairment.
°ver a lifetime, the
C°ncern' aIthouS"
and
on various measures of
illnesses or exacerbation o
Have obvious .imitations in
cite
Supplemen'
h .the effects of
osP|tall«tons for respiratory
-------
-3-
>-<-
r
not r«.a/-h ,
Pff on Paper recSSTm^Swi " of^n'hLr^11100 - ^ endors^nt of the staff
th« tPhPm>a"d 1Ve favored a" W ^.tarf oT1 a" UP?-lr value of '«* *»
data base is very large and adequate fo *y? |3,/ew hours) to ozone. The
-------
-4-
aan w -
seasonal and lifetime exposure* o ozone iTlS^ ^ * mU't!ple hou''
expanded research effort. This musfbe done S th^T^ *" accele«ted and
ozone standards will derive from a stronger sdent.4 base. " COnsIderatlo"s <>f
t. set
nature While th Comittee I to to^T °f 3 Stf?Ctly sdentlflc
standards we see no need, in view o7 the f alrLdv extLl ** Y°U °" thc ozone
rev,ew the proposed ozone standards ono/ to t£?^IpCOI?meiltf ProvWed. to
g^tej;. In this instance, the public comln. P"bllcai°n in the Federal
opportunity for the Committee tc ^provide anHS tirST1 Wf" Pr°vWe
ma be nece P any addlt«onal comments or
e c prove an tir
may be necessary. P any addlt«onal comments or review that
to contact
Sincerely,
Roger O. McClellan. D.V.M.
Chairman, Clean Air Scientific Advisory
Committee
ROM:ewb
-------
United States
Environmental Protection
Agency
Office of the Administrator
science Advisory Board
Washington, DC 20460
EPA-SAB-CASAC-89-019
May 1989
EPA
Report of the
Clean Air Scientific
Advisory Committee
(CASAC)
_
f°LQzone: Closure on
the OAQPS Staff
Paper (1988) and the
Criteria Document
Supplement (1988)
-------
ABSTRACT
Scientific and Technical InforLtion (19881°
ss^yirer ssss sHTf -
adequate scientific and fa^iSal^»2i *~ ™°fu**nt"J Provide an
pri-a^ and secondary
-------
NOTICE
sclent--i -P-?/^ •m.^.j. Provide a balancer* CV^^^L. **'d'=*»»-y. Tne
jSs^ ^^^^^^^^
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
LEAN AIR SCTENTTFTC ADVTSDPV /•»nmTTTfT
(STATUTORY MEMBEPfi) —
Chairman
Dr.
Members
"'
Executive Secret-.ar-y
Scientist, Science Advisory
, U.S. Environmental Protection Aaencv
M Street, SW, Washington, D.C. 20460 Agencv'
-------
0. S.
Clean
^iron,..nt.i Protection Agency
Chairman
Dr-
CASAC Q7n^ Pmn? mm.,,.,..
Brunswick, New Jersey
"'
Maryland
s University, New
Health sciences,
and Public Health,
Futura, Bashington,
Xnstitute,
Boyce
POCUS me., Los altos,
-------
Dr-
President-
Dr-
school of
Dr. Jerome J. Wesolowski, chief Air
«., CaUfornia Oepart^l 'O
Dr. Georae T Wni F-F
Laboratorie"' '
Michigan
^^
«°tors Research
Department, Warren,
Executive
'
-------
REFERENCES
Abraham, W. M.; Januszkiewicz, A. J.; Mingle, M.; Welker, M.;
Wanner, A.; Sackner, M. A. (1980) Sensitivity of
bronchoprovocation and tracheal mucous velocity in detecting
airway responses to 03. J. Appl. Physiol. Respir. Environ.
Exercise Physiol. 48: 789-793.
Adams, W. C.; Schelegle, E. S. (1983) Ozone and high ventilation
effects on pulmonary function and endurance performance. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 55: 805-812.
Adams, R. M.; McCarl, B. A. (1985) Assessing the benefits of
alternative oxidant standards on agriculture: the role of
response information. J. Environ. Econ. Manage.: 12: 264-276.
Adams, R. M.; Hamilton, S. A.; McCarl, B. A. (1985) An assessment
of the economic effects of ozone on U.S. agriculture. J. Air
Pollut. Control Assoc.: 35: 938-943.
Adams, R. M.; McCarl, B. A. (1985) Assessing the benefits of"
alternative ozone standards on agriculture: The role of response
information. J. Environ. Econ. and Mgmt. 12: 264-276.
Adams, R.M.; Hamilton, S.A.; McCarl, B.A. (1984) The economic
effects of ozone on agriculture. Corvallis, OR: U.S.
Environmental Protection Agency; EPA report no. EPA-600/3-84-090.
Available from: NTIS, Springfield, VA; PB85-168441/XAB.
Adams, W.C.; (1987) Effects of ozone exposure at ambient air
pollution episode levels on exercise performance. Sports med.
(Auckland) 4: 395-424.
Adedipe, N. 0.; Barrett, R. E.; Ormrod, D. P. (1972)
Phytotoxicity and growth responses of ornamental bedding plants
to ozone and sulfur dioxide. J. Am. Soc. Hortic. Sci. 97: 341-
345.
Adedipe, N. 0.; Ormrod, D. P. (1974) Ozone induced growth
suppression in radish plants in relation to pre- and post-
fumigation temperatures. Z. Pflanzenphysiol. 71: 281-287.
Adomait, E. J.; Ensing, J.; Hofstra, G. (1987) A dose-response
function for the impact of O3 on Ontario-grown white bean and an
estimate of economic loss. Can. J. Plant Sci. 67: 131-136.
Air Management Technology Branch (AMTB, 1981) Guideline for use
of city-specific EKMA in preparing ozone SIPs. Research Triangle
Park, N.C.: U.S. Environmental Protection Agency (EPA-450/480-
U £• I •
-------
Alpert, S. M.; Gardner, D. E.; Hurst, D. J.; Lewis, T. R. ;
mfohi^'J?* L; £97^} Effects of exposure to ozone on defensive
mechanisms of the lung. J. Appl. Physiol. 31: 247-252.
Altshuller, A. P. (1977) Eye irritation as an effect of
air pollution> j^ Air Pollute control Assoc. 27:
Altshuller, A. P. (1983) Measurements of the products of
atmospheric photochemical reactions in laboratory studies and in
ambient air-relationships between ozone and other products
Atmos. Environ. 17: 2383-2427.
Altshuller, A. P. (1986) Review paper: The role of nitrogen
oxides in nonurban ozone formation in the planetary boundary
layer over N America, W Europe, and adjacent areas of ocean!
Atmospheric Environment 20: 245-268.
American Thoracic Society (1985) Guidelines as to what
constitutes an adverse respiratory health effect, with special
reference to epidemiologic studies of air pollution. • Am. Rev
Respir. Dis. 131: 666-668. '
Amiro, B. D.; Gillespie, T. J. ; Thurtell, G. W. (1984) Injury
response of Phaseolus vulgar! s to ozone flux density. Atmos
T?Titri >-/••> T> 1Q.1->n-7_i">ic J "wiuwo.
Environ. 18: 1207-1215.
Amoruso, M.A. ; Witz, G.; Goldstein, B. D. (1981) Decreased
superoxide anion radical production by rat alveolar macrophages
following inhalation of ozone or nitrogen dioxide. Life Sci. 28:
Aranyi, c. ; Vana, S. c. ; Thomas, P. T. ; Bradof, J. N. ; Fenters
J. D.; Graham, J. A.; Miller, F. J. (1983) Effects of subchronic
exposure to a mixture of 03, S02, and (NH4)2 SO4 on host defenses
of mice. J. Toxicol. Environ. Health 12: 55-71.
Astrand, P.O.; Rodahl, K. (1977) Textbook of work physiology. New
York, N.Y.: McGraw-Hill, Inc.
Avol, E. L. ; Linn, W. S.; Shamoo, D. A.; Spier, C. E. ; Valencia,
L. M.; Venet, T. G. ; Trim, S. C. ; Hackney, J. D. (1987) Short-
term respiratory effects of photochemical oxidant exposure in
exercising children. J. Air Pollut. Control Assoc. 37: 158-162.
Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Hackney, J. D. (1988)
Seasonal ozone reactivity in Los Angeles residents. Presented
at: 8 1st annual meeting of the Air Pollution Control Association-
June; Dallas, TX. Pittsburgh, PA: Air Pollution Control
Association; paper no. 88-122.6.
-------
.• "**»•*.
ro -soc
Am. Rev. Respir D!S 131:^19-622? exercisi^ adolescents
f
air pollutionin adolescents. roSeSjnai off e^S,°f a^ient
specialty conference on the evaluation 5 55 °f .an international
-
-------
Barrv' B- E:'*.Mlller' F. J.; Crapo, J. D. (1983) Alveolar
rSi 5 S ^ury caused by inhalation of 0.25 ppm of ozone. In:
Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related
photochemical oxidants; March 1982; Pinehurst, NC. Princeton
NJ: Princeton Scientific Publishers, Inc; pp. 299-309. (Advances
in modern environmental toxicology: v. 5).
Barry, B. E. (1983) Morphometric and morphologic studies of the
effects of asbestos, oxygen, and ozone on the lung [Ph.D.
thesis]. Durham, NC: Duke University.
Bartlett, D., Jr.; Faulkner, C. S. II; Cook, K. (1974) Effect of
chronic ozone exposure on lung elasticity in younq rats J
Appl. Physiol. 37: 92-96. *
Bates, D. V.; Bell, G. M.; Burnham, C. D.; Hazucha, M.; Mantha,
J.; Pengelly, L. D.; Silverman, F. (1972) Short-term effects of
ozone on the lung. J. Appl. Physiol. 32: 176-181.
Bates, D. V.; Hazucha, M. (1973) The short-term effects of ozone
on the human lung. In: Proceedings of the conference on health
effects of air pollutants; October; Washington,, DC. Washington,
DC: -U.S Senate, Committee on Public Works; pp. 507-540- serial
no. 93—15.
Bates, D. V.; Sizto, R. (1983) Relationships between air
. pollutant levels and hospital admissions in Southern Ontario
Can. J. Public Health 74: 117-122.
Bates, D. V.; Sizto, R. (1987) Air pollution arid hospital
admissions in southern Ontario: the acid summer haze effect
Environ. Res. 43: 317-331.
Bates/ p. V.; Sizto, R. (1989) The Ontario air pollution study:
identification of the causative agent. In: International
symposium on the health effects of acid aerosols: addressing
obstacles in an emerging data base; October 1987; Research
Triangle Park, NC. EHP Environ. Health Perspect. 79: 69-72.
Beckerson, D. W.; Hofstra, G. (1979) Response of leaf diffusive
resistance of radish, cucumber, and soybean to O3 and SO, singly
or in combination. Atmos. Environ. 13: 1263-1268".
Bedi, J. F.; Folinsbee, L. J.; Horvath, S. M.; Ebenstein, R. S.
(1979) Human exposure to sulfur dioxide and ozone: absence of a
synergistic effect. Arch. Environ. Health 34: 233-239.
Bedi, J. F.; Horvath, S. M.; Drechsler-Parks, D. M. (1988)
Reproducibility of the pulmonary function response of older men
and women to a 2-hour ozone exposure. JAPCA J. Air Pollut
Control Assoc. 38: 1016-1019.
-------
Sis
of dyed
pollutants-
Bennett, j. p.- oshima, R. j. ;
-------
6
iM?; T"' Tinge*' D- T- <:L977) A study of the potential ways
in which ozone could reduce root growth and nodulation of
soybean. Atmos. Environ. 11: 737-739.
Boatman E. S.; Sato, S.; Frank, R. (1974) Acute effects of ozone
on cat lungs. II. Structural., Am. Rev. Respir. Dis. 110: 157-169.
B°ck' N.; Lippman, M'; Lioy' P-' Mun°Zf A.; Speizer, F. (1985)
The effects of ozone on the pulmonary function of children. In-
Proceedings of an inter-national specialty conference on the
evaluation of the scientific basis for an ozone/oxidant standard-
November 1984; Houston, TX. Pittsburgh, PA; Air Pollution
Control Association; pp. 297-308. (APCA international specialty
conference transactions: v. 4). «-j.aj.uy
Bogaty H. ; Campbell, K. S.; Appel, W. D. (1952) The oxidation
of cellulose by ozone in small concentrations. Text. Res. J. 22:
8 1—83 .
Boorman, G. A.; Schwartz, L. W. ; Dungworth, D. L. (1980)
Pulmonary effects of prolonged ozone insult in rats: morphometric
evaluation of the central acinus. Lab. Invest. 43: 108-115.
Bormann, F. H. (1985) Air pollution and forests: an ecosystem
perspective. Bio Science 35: Q34-441.
Bormann, F. H. ; Smith, W. H. (1980) Effects of air pollution on
forest ecosystems. In: Energy and the fate of ecosystems.
Washington, DC: National Academy of Sciences; pp. 308-318.
Botkin, D. B.; Smith, W. H.; Carlson, R. W. ; Smith, T. L. (1972)
Effects of ozone on white pine saplings: variation in inhibition
and recovery of net photosynthesis. Environ. Pollut. 3: 273-289.
Bradley, C. E. ; Haagen-Smit, A. J. (1951) The application of
rubber in the quantitative determination of ozone. Rubber Chem
Technol. 24: 750-775.
Bryan, C. L. ; Jenkinson, S. G. (1987) Species variation in lung
antioxidant enzyme activities. J. Appl. Physiol. 63: 597-602.
Bureau of the Census (1982) The journey to work. Washington,
D.C.: U.S. Department of Commerce (Special studies, Series P-
23 , No. 122) .
Butler, L. K.; Tibbitts, T. W. (1979) Stomatal mechanisms
determining genetic resistance to ozone in Phaseolus vulgaris L
J. Am. Soc. Hortic. Sci. 104: 213-216.
Cameron, J. W. (1975) Inheritance in sweet corn for resistance to
acute ozone injury. J. Am. Soc. Hortic. Sci. 100: 577-579.
-------
3S sS318 — "
-------
Clarke B.B.; Henninger, M. R.; Brennan, E. (1983) An assessment
of potato losses caused by oxidant air pollution in New Jersey.
Phytopathology 73: 104-108. *
Clean Air Scientific Advisory Committee Meeting, held in Research
Triangle Park, N.C. on April 22 and 23, 1986.
Clean Air Scientific Advisory Committee Meeting, held in
Washington, D.c. on December 14-15, 1987.
Clean Air Scientific Advisory Committee Meeting, held in Research
Triangle Park, December 14-15, 1988.
Code of Federal Regulations (CFR); various titles and sections.
Coffin, D. L.; Gardner, D. E. (1972) Interaction of biological
agents and chemical air pollutants. Ann. Occup. Hyg. 15: 219-
235.
Coffin, D. L.; Gardner, D. E.; Holzman, R. S.; Wolock, F." J.
(1968) Influence of ozone on pulmonary cells. Arch. Environ
Health 16: 633-636.
Coffin, D. L.; Blommer, E. J.; Garnder,-D. E.; Holzman, R. (1967)
Effect of air pollution on alteration of susceptibility to
pulmonary infection. In: Proceedings of the third annual
conference on atmospheric contamination in confined spaces; May
Dayton, OH. Wright-Patterson Air Force Base, OH: Aerospace '
Medical Research Laboratories; pp.'71-80; report no. AMRL-TR-67-
200. Available from: NTIS, Springfield, VA; AD-835008.
Contant, C. F., Jr.; Gehan, B. M.; Stock, T. H.; Holguin, A. H.;
Buffl-er, P. A. (1985) Estimation of individual ozone exposures
using microenvironment measures. In: Proceedings of an
international specialty conference on the- evaluation of the
scientific basis for an ozone/oxidant standard; November 1984;
Houston, TX. Pittsburgh, PA: Air Pollution Control Association;
pp. 250-261. (APCA international specialty conference
transactions: v. 4).
Cosio, M. G.; Hale, K. A.; Niewoehner, D. E. (1980) Morphologic
and morphometric effects of prolonged cigarette smoking on the
small airways. Am. Rev. Respir. Dis. 122: 265-271.
Costa, p. L. (1989) Pulmonary function studies in the rat
addressing concentration versus time relationships of ozone.
Presented at: Third U.S.Dutch international symposium on
atmospheric ozone research and its policy implications; May;
Nijmegen, The Netherlands. Amsterdam, The Netherlands: Elsevier
Science Publishers; p. 733-743 (Studies in environmental Science
35) .
-------
-------
10
De Vos, N. E.; Hill, R. R., Jr.; Pell, E. J.; Cole
CYM3»i-i+- -i ^a^ •! »»»» j..v._._.!4._ f- , — • w • , WVS.LC ,
De Vos N. E.; Pell, E. J.; Hill, R. R. , Jr. . Cole R. H-
of potato
S?aA A* J"' ^i***^' J' A- Bryant, L. R. (1983) Effects of
combined exposure to ozone and carbon monoxide in humans. In:
Mehlman, M. A.; Lee,S. D.; Mustafa, M. G., eds. International
symposium on the biomedical effects of ozone and related
photochemical oxidants; March 1982; Pinehurst, NC. Princeton
NJ; Princeton Scientific Publishers, Inc. p. 145-149 (Advances in
modern environmental toxicology, V.5). (Advances in
DeLucia, A.J.; Adams, W.C. (1977). Effects of ozone inhalation
during exercise on pulmonary function and blood biochemistry. J
Appl. Physiol.: Respir. Environ. Exercise Physiol. 43: 75-81.
Dimeo, M. J.; Glenn, M. G'.; Holtzman, M. J.; Sheller, J. R. ;
Nadel, J. A.; Boushey, H. A. (1981) Threshold concentration of
ozone causing an increase in bronchial reactivity in humans and
adaptation with repeated exposures. Am. Rev. Respir. Dis. 124:
^4O^^4o •
Dochinger, L. s.;- Bender, F. W. ; Fox, F. L. ; Heck, W. W. (1970)
Chlorotic dwarf of eastern white pine caused by an ozone and
sulphur dioxide interaction. Nature (London) 225: 476.
Dochinger, L. S.; Townsend, A. M. (1979) Effects of roadside
deicer salts and ozone on red maple progenies. Environ. Pollut.
19 t 229~237.
Donovan, D. H. ; Williams, S. J. ; Charles, J. M. ; Menzel, D. B.
(1977) Ozone toxicity: effect of dietary vitamin E and
polyunsaturated fatty acids. Toxicol. Lett. 1: 135-139.
Dowell, A. R.; Lohrbauer, L. A.; Hurst, D.; Lee, S. D. (1970)
Rabbit alveolar macrophage damage caused by in vivo ozone
inhalation. Arch. Environ. Health 21: 121-127.
Drechsler-Parks, D. M. ; Bedi, J. F. ; Horvath, S. M. (1984)
Interaction of peroxyacetyl nitrate 'and ozone on pulmonary
functions. Am. Rev. Respir. Dis. 130: 1033-1037.
Drinkwater, B. L. ; Raven, P. B.; Horvath, S. M. ; Gliner, J. A. ;
Ruhling, R. W. ; Bolduan, N. W., (1974) Air pollution, exercise and
heat stress. Arch. Environ. Health 28: 177-181.
-------
11
2STi &.'"
Ltional
- p-
-
. Health 5: 631-642.
Toxicol. Environ. HeSlS^eJf-fls? tO °Z°ne exp°sure' J-
-------
12
- R- L-; Kreit' J- w- ,
ozone-induced lung function changes in normal and
TMru1! "SSJ?" ^ *?? 6ffeCt °f in<*omethacin. Presented at:
Third U.S. -Dutch international symposium on atmospheric ozone
research and its policy implications. May; Nijmegen, The
?S« her*an ds: Amsterdam: Elsevier Science Publishers; p. 493-
499. (Studies in environmental science 35).
f?5aJf'r^' L-; fchwar^2' L- w->' K°sch, P. c.; Dungworth, D. L.
(1981) Chronic bronchiolitis in nonhuman primates after prolonged
ozone exposure. Am.J.Pathol. 105: 121-13?T prolonged
Evans, L. S.; Ting, I. P. (1974) Ozone sensitivity of leaves:
.ni?™?^**, leaf.wfter content, gas transfer resistance, and
anatomical characteristics. Am. J. Bot. 61: 592-597.
.M; •?.•'' Johnson, L. V.; Stephens, R. j. ; Freeman, G.
(1976b) Cell renewal in the lungs of rats exposed to low levels
of ozone. Exp. Mol. Pathol. 24: 70-83.
Evans, M. J.; Stephens, R. J. ; Freeman, G. (1976c) Renewal of
pulmonary epithelium following oxldant injury. In: Bouhuys A
ed. Lung cells in disease: proceedings of a Brook Lodge ' "'
conference; April; Augusta, MI. New York, NY: Elsevier/North-
Holland Biomedical Press; pp. 165-178.
Evans, M. J.; Johnson, L. V.; Stephens, R. J. ; Freeman G
(1976a) Renewal of the terminal bronchiolar epithelium in the rat
following exposure to N02 or O3. Lab. Invest. 35: 246-257.
. T' J-; Sauder' L- R"*
Adaptation in human subjects to the effects of
inhaled ozone after repeated exposure. Am. Rev. Respir. Dis. 119:
™ / j U •
Feder, W. A. (1978) Plants as bioassay systems for monitoring
atmospheric pollutants. EHP Environ. Health Perspect. 27: 139-
147.
Feder, W. A.; Campbell, F. J. (1968) Influence of low levels of
ozone on flowering of carnations. Phytopathology 58: 1038-1039.
Ferdo, A. (1985) Ozone micr©environment factors (Memo). Durham
N.C.: PEI Associates. '
Filipowicz, C.; McCauley, R. (1986) The effects of chronic ozone
exposure on pulmonary collagen content and collagen synthesis in
rats. J. Appl. Toxicol. 6: 87-90.
-------
13
Folinsbee, L. J.; Hazucha, M. J. (1989) Persistence of ozone-
induced changes in lung function and airway responsiveness.
Presented at: Third U.S.-Dutch international symposium on
atmospheric ozone research and its policy implications; May 1988;
Nijmegen, The Netherlands. Amsterdam: Elsevier Science
Publishers; p. 483-492. (Studies in environmental science 35).
Folinsbee, L. J.; McDonnell, w. F.; Horstman, D.H. (1988)
Pulmonary function and symptom responses after a 6.6 hr exposure
to 0.12 ppm ozone with moderate exercise. J. Air Pollut. Control
Assoc. 38: 28-35.
Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1980) Respiratory
responses in humans repeatedly exposed to low concentrations of
ozone. Am. Rev. Respir. Dis. 121: 43.1-439.
Folinsbee, L. J.; Silverman, F.; Shephard, R. J. (1977a) Decrease
of maximum work performance following ozone exposure. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 42: 531-536.
Folinsbee, L. J.; Horvath, S. M.; Raven, P. B.; Bedi, J. F.;
Morton, A. R.; Drinkwater, B. L.; Bolduan, N. W.; Gliner, J. A.
(1977b) Influence of exercise and heat stress on pulmonary
function during ozone exposure. J. Appl. Physiol.: Respir.
Environ. Exercise Physiol. 43: 409-413.
Folinsbee, L. J.; Silverman, F.; Shephard, R'. J. (1975) Exercise
responses following ozone exposure. J. Appl. Physiol. 38: 996-
1001.
Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J. F.; Horvath, S. M.
(1978) The influence of exercise on the pulmonary changes due to
exposure to low concentrations of ozone. In: Folinsbee, L. J.;
Wagner, J. A.; Borgia, J. F.; Drinkwater, B.L.; Gliner, J.A.;
Bedi, J.F., eds. Environmental stress: individual human
adaptations. New York, NY: Academic Press; pp. 125-145.
Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1984) Pulmonary
function changes in trained athletes following 1-hour continuous
heavy exercise while breathing 0.21 ppm ozone. J. Appl. Physiol.:
Respir. Environ. Exercise Physiol. 57: 984-988.
Foster, K. W.; Timm, H.; Labanauskas, C. K.; Oshima, R. J. (1983)
Effects of ozone and sulfur dioxide on tuber yield and quality of
potatoes. J. Environ. Qual. 12: 75-80.
Frager, N. B.; Phalen, R. F.; Kenoyer, J. L. (1979) Adaptations
to ozone in reference to mucociliary clearance. Arch. Environ.
Health 34: 51-57.
-------
14
Freeman, G.; Juhos, L. T.; Fxiriosi, N. J.; Mussenden, R.;
Stephens, R. j.; Evans, M. J., (1974) Pathology of pulmonary
disease from exposure to interdependent ambient gases (nitrogen
dioxide and ozone). Arch. Environ. Health 29: 203-210.
Freeman, G.; Stephens, R. J.; Coffin, D. L.; Stara, J. F. (1973)
Changes in dogs' lungs after long-term exposure to ozone: light
and electron microscopy. Arch. Environ. Health 26: 209-216.
Fujimaki, H.; Ozawa, M.; Imai, T.; Shimizu, F. (1984) Effect of
short-term exposure to 03 on antibody response in mice. Environ.
Res. 35: 490-496.
Fujinaka, L. E. (1984) Respiratory bronchiolitis following long-
term ozone exposure in Bonnet, monkeys: a morphometric study
[master's thesis], Davis, CA: University of California.
Fujinaka, L. E.; Hyde, D. M.; Plopper, C. G.; Tyler, w. S.;
Dungworth, D. L.; Lollini, L. O. (1985) Respiratory bronchiolitis
following long-term ozone exposure in Bonnet monkeys: a
morphometric study. Exp. Lung Res. 8: 167-190.
Fukuda, H.; Terakado, K. (1974) On the damage of plants due to
peroxyacyl nitrates (PAN). Tokyo Nogyo Shikenjo Sokuho: 35-36
[English abstract].
Furukawa, A.; Kadota, M. (1975) Effect of ozone on photosynthesis
and respiration in poplar leaves. Seibutsu Kankyo Chosetsu 13: 1-
/ •
Gardner, D. E.; Miller, F. J.; Illing, J. W.; Kirtz, J. M. (1977)
Increased infectivity with exposure to ozone arid sulfuric acid
Toxicol. Lett. 1: 59-64.
Gerrity, T. R.; McDonnell, W. F. (1989) Do functional changes in
humans correlate with airway removal efficiency of ozone?
Proceedings of: Third U.S.-Dutch international symposium on
atmospheric ozone research and its policy implications; May 1988;
Nijmegen, The Netherlands. Amsterdam: Elsevier Science
Publishers; p. 293-300. (Studies in environmental science 35).
Gerrity, T. R.; Wiester, M. J. (1987) Experimental measurements
of the uptake of ozone in rats and human subjects. Presented at:
80th annual meeting of the Air Pollution Control Association;
June 21-26, 1987; New York, NY. Pittsburgh, PA: Air Pollution
Control Association; paper no. 87-99.3.
Gerrity, T. R. (1987) Nasopharyngeal uptake of ozone in humans
and animals. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Health Effects Research Laboratory; report no.
EPA/600/D-87/357. Available from: NTIS, Springfield, VA; PB88-
140561.
-------
15
S. I.; Adams, W. c.
and ambient heat on
57: 450-456.
SSBJMS..
ssar-s,
-------
16
™« ?' (1984> Similarities of host defense mechanisms
against pulmonary disease in animals and man. J. Toxicol
Environ. Health 13: 471-478.
Grose, E. C. ; Richards, j. H. ; Illing, j. w. ; Miller, F. J. •
SaVJe3'-D' W*; Graham' J- A"- Gardner, D. E. (1982) Pulmonary
host defense responses to inhalation of sulfuric acid and ozone
J. Toxicol. Environ. Health 10: 351-362.
Grose, E. C.; Costa, D. L. ; Hatch, G. E. ; Miller, F. J. ; Graham,
J. A.; Crapo, J. D.; Chang, L.-Y; Stevens, M. A. J Jaskot, R. H •
Tepper, J. s. (1989) Chronic exposure to ozone restrictive luno;
disease. Research Triangle Park, NC: U.S. Enviornmental
A
-------
17
Atmospheric ozone research and '+• ' i« • *' 9ran^' ^* ^* / eds
environmental science 5
Hackney, J. D. ; Linn, W. S.; Law D C •
sss.-sssr srs^ssr as
Health 7: 38?-390 ^ ln huinans- J- Toxicol. Environ.
. (Advances in modern environmental toxicology" v'
Available fro,; Srin?"??! 09^ . "'
-------
18
ohnAo • '* Nlewoenner' D- E.; Cosio, M. G. (1980) Morphologic
2X? in.muscuiar P^^onary arteries: relationship to cigarette
**™** dlS6ase' and emPhysema. Am. Rev. Resir. D
2273-18™** lS6ase' and emPhysema. Am. Rev. Respir. Dis
Hamburger, S. J. ; Goldstein, B. D. (1979a) Effect of ozone on the
°the
.
of erythrocytes by concanavalin A. II: study of human subjects
receiving supplemental vitamin E. Environ. Res. 19: 299-305.
Hammer, D. I.; Hasselblad, V.; Portnoy, B.; Wehrle, P. F. (1974)
Los Angeles student nurse study: daily symptom reporting and
photochemical oxidants.Arch. Environ. Health 23: 255-260.
Hanson, G. P.; Thome, L. ; Addis, D. H. (1975) The ozone
sensitivity of Petunia hybrida Vilm. as related to physiological
age. J. Am. Soc. Hortic. Sci. 100: 188-190.
Harkema, J. R, ; Plopper, C. G.'; Hyde, D. M. ; St. George, J. A. ;
Dungworth, D.L. (I987a) Effects of an ambient level of ozone on
primate nasal epithelial mucosubstances . Quantitative
histochemistry. Am. J. Pathol. 127: 90-96.
Harkema, J. R. ; Plopper, C. G.; Hyde, D. M. St., George, J. A, ;
Wilson, D w.; Dungworth, D. L. (1987b) Response of the macaque
nasal epithelium to ambient levels of ozone. A morphometric Study
of the transitional and respiratory epithelium. Am. J. Pathol.
128: 29-44.
Harkov, R. ; Brennan, E. (1979) An ecophysiological analysis of
the response to oxidant pollution. J. Air Poliut. Control Assoc.
Vol. 29: 157—161.
Harkov, R.; Clarke, B.; Brennan, E. (1979) Cadmium contamination
may modify response of tomato to atmospheric ozone. J. Air
pollut. Control Assoc. 29: 1247-1249.
Harvard, M. R. ; Treshow, M. (1973) Impact of ozone on the growth
and reproduction of understory plants in the aspen zone of
western U.S.A. Environ. Conserv. 2: 17-23.
Hatch, G. E.; Slade, R. ; stead, A. G. ; Graham, J. A. (1986)
Species comparison of acute inhalation toxicity of zone and
phosgene. J. Toxicol. Environ. Health 19: 43-53.
-------
19
and
: Air Pollution Contro ssociatio;
soiene
science 3S) .
on
P 5H-5W . Elsevier
(Studies ln environmental
ssr
achr d
-------
20
A ? ' Sllverman' P-; Parent, C. ; Field, S.; Bates, D. V.
(1973) Pulmonary function in man after short-term exposure to
ozone. Arch. Environ. Health 27: 183-188.
Hazucha, M. (1981) Assessment of ozone -induced hyperreactivitv bv
histamine in normal healthy subjects. In: Proceedings of the
Jes®ar°V*a?ning W°rkshop on health effects of oxidants; January
1980; Raleigh, NC. Research Triangle Park, NC: U.S. Environmental
Protection Agency; pp. 314-327; EPA report no. EPA-600/9-81-001
Available from: NTIS, Springfield, VA; PB81-178832.
Heagle, A. S.; Spencer, S.; Letchworth, M. B. (1979c) Yield
Wheat t0 chronic doses of ozoiM. Can. J. Bot .
Heagle, A. S.; Heck, W. W. ; Lesser, V. M. ; Rawlings, J. Q. (1987)
Effects of daily ozone exposure duration and concentration
fluctuation on yield of tobacco. Phytopath. 77: 856-862.
Heagle, A. S.; Lesser, V. M. ; Rawlings, J. 0.; Heck, W. W. ;
Philbeck, R. B. (1986) Response of soybeans to chronic doses of
ozone applied as constant or proportional additions to ambient
air. Phytopath. 76: 51-56.
Heagle, A. S.; Body, D. E. ; Pounds, E. K. (1972) Effect of ozone
on yield of sweet corn. Phytopathology 62: 683-687.
Heagle, A. S.; Body, D. E.; Neely, G. E. (1974) Injury and 'yield
responses of soybean to chronic doses of ozone and sulfur dioxide
in the field. Phytopathology 64: 132-136.
Heagle, A.S.; Letchworth, M.B. (1982) Relationships among injury
growth, and yield responses of soybean cultivars exposed to ozon4
at different light intensities, J. Environ. Qual. 11: 690-694.
Heagle, A. S.; Philbeck, R. B.; Knott, W. M. (1979a) Thresholds
for injury growth, and yield loss caused by ozone on field corn
hybrids. Phytopathology 69: 21-26.
Heagle, A. S.; Heck, W. W. ; Rawlings, J. 0.; Philbeck, R. B.
(1983) Effects of chronic doses of ozone and sulfur dioxide on
injury and yield of soybeans in open-top field chambers. Cron
Sci. 23: 1184-1191. v
Heagle, A. S.; Philbeck, R. B.; Letchworth, M. B. (1979b) Injury
and yield response of spinach cultivars to chronic doses of ozone
in open-top field chambers. J. Environ. Qual. 8: 368-273.
Heath, R. L. (1980) Initial events in injury to plants by air
pollutants. Annu. Rev. Plant Physiol. 31: 395-431.
-------
21
*•
- c-
Heck, W.w. Personal communication. March 6, 1986
-------
22
Heggestad, H. E. (1973) Photochemical air pollution injury to
potatoes in the Atlantic coastal states. Am. Potato J. 50: 315-
328.
Henderson, W. R.; Reinert, R. A. (1979) Yield response of four
fresh market tomato cultivars after acute ozone exposure in the
seedling stage. J. Am. Soc. Hortic. Sci. 104: 754-759.
Hoffman, G. J.; Maas, E. V.; Rawlins, s. L. (1975) Salinity-ozone
interactive effects on alfalfa yield and water relations. J
Environ. Qual.4: 326-331.
Hofstra, G.; Littlejohns, D. A.; Wukasch, R. T. (1978) The
efficacy of the antioxidant ethylene-diurea (EDU) compared to
carboxin and benomyl in reducing yield losses from ozone in navy
bean. Plant. Dis. Rep.62: 350-352.
Hogsett, W. E.; Tingey, D. T.; Lee, E. H. (1983) Ozone exposure
indices: Concepts for development and evaluation of their use.
In: Assessment of Crop Loss from Air Pollutants, pp. 107-138.*
W.W. Heck, O.C. Taylor and D.T. Tingey (eds.). - Elsevier Applied
Science, New York.
Hogsett, W. E.; Tingey, D. T.; Holman, S. R. (1985) A
programmable exposure control system for determination of the
effects of pollutant exposure regimes on plant growth. Atmos.
Environ. 19: 1135-1145.
Holguin, A. H.; Buffler, P. A.; Contant, C. F., Jr.; Stock, T,
H.; Kotchmar,D. J.; Hsi, B. P.; Jenkins, D. E.; Gehan, B. M.;
Noel, L. M.; Mei, M. (1985) The effects of ozone on asthmatics in
the Houston area. In: Proceedings of an intermrtional specialty
conference on the evaluation of the scientific basis for an
ozone/oxidant standard; November 1984; Houston, TX. Pittsburgh,
PA: Air Pollution Control Association; pp. 262-280. (APCA
international specialty conference transactions: v. 4).
Holtzman, M. I.; Cunningham, J. H.; Sheller, J. R.; Irsigler, G.
B.; Nadel, J. A.; Boushey, H. A. (1979) Effect of ozone on
bronchial reactivity inatopic and nonatopic subjects. Am. Rev.
Respir. Dis. 120: 1059-1067.
Holtzman, M. J.; Fabbri, L. M.; Skoogh, B.E.; O'Byrne, P. M.;
Walters, E. H.;Aizawa, H.; Nadel, J. A. (I983a) Time course of
airway hyperresponsiveness induced by ozone in dogs. J. Appl.
Physiol.: Respir. Environ. Exercise Physiol. 55: 1232-1236.
Holtzman, M. J.; Fabbri, L. M.; O'Byrne, P. M.; Gold, B. D.;
Aizawa, H.; Walters, E. H.; Alpert, S. E.; Nadel, J. A. (1983b)
Importance of airway inflammation for hyperresponsiveness induced
by ozone. Am. Rev. Respir. Dis. 127 686-690.
-------
23
Horsman, D. C.; Nicholls, A. 0.; Calder, D. M. (1980) Growth
responses of Dactylis glomerata, Lolium perenne and Phalaris
aquatica to chronic ozone exposure. Aust. J. Plant Physiol. 7:
511—517.
Horstman, D.; McDonnell, W.; Abdul-Salaam, S.; Folinsbee, L.;
Ives, P. (1988) Current US EPA research concerning more
pronounced human exposures to low ozone concentrations.
Presented at: 81st annual meeting of the Air Pollution Control
Association; June; Dallas, TX. Pittsburgh, PA: Air Pollution
Control Association; paper no. 88-122.5.
Horstman, D.; McDonnell, W.; Folinsbee, L.; Abdul-Salaam, S.;
Ives, P. (1989) Changes in pulmonary function and airway
reactivity due to prolonged exposure to typical ambient ozone
levels. In: Schneider, T., Lee, S. D.; Wolters, G. J. R.;
Grant, L. D. eds. Atmospheric ozone research and its policy
implications: Proceedings of: The Third U.S.-Dutch
international symposium on atmospheric ozone research and its
policy implications; May; Nijmegen, The Netherlands. Amsterdam
The Netherlands: Elsevier Science Publishers; pp. 755-762.
(Studies in environmental science 35).
Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation
to ozone duration of effect. Am. Rev. Respir. Dis. 123: 496-499.
Horvath, S. M.; Gliner, J. A.; Matsen-Twisdale, J. A. (1979)
Pulmonary function and maximum exercise responses following acute
ozone exposure. Aviat. Space Environ. Med. 50: 901-905.
Howell, R. K.;'Koch, E. J.; Rose, L. P., Jr. (1979) Field
assessment of air pollution-induced soybean yield losses. Agron
J. 71: 285-288.
Howell, R. K.; Rose, L. P., Jr. (1980) Residual air pollution
effects on soy-bean seed quality. Plant Dis. 64: 385-386.
Howitt, R. E.; Gossard, T. E.; Adams, R. M. (1984a) Effects of
alternative ozone levels and response data on economic
assessments: The case of California crops. J. Air pollut. Control
Assoc. 34: 1122-1127.
Howitt, R. E.; Gossard, T. E.; Adams, R. M. (1984b) Applied
Welfare Economics and Public Policy. New york: Prentice-Hall.
Hucl, P.; Beyersdorf, W. D. (1982) The inheritance of ozone
sensitivity in selected Phaseolus vulgaris L. populations. Can
J. Plant Sci. 62: 861-865.
Hunt, W. F. ed. (1985) National air quality and emissions trends
report, 1983. Research Triangle Park, N.C.: U.S. Environmental
Protection Agency (EPA 450/4-84-029).
-------
24
Hunt, W. F.; Curran, T.; and Freas, W. P. (1987) National and
regional trends in ambient ozone and nitrogen dioxide
measurements, 1976-1985. Durham, N.C.: U.S. Environmental
Protection Agency.
Hurst, D. J.; Gardner, D. E.; Coffin, D. L. (1970) Effect of
ozone on acid hydrolases of the pulmonary alveolar macrophage. J.
Reticuloendothel. Soc. 8: 288-300.
Hurst, D. J.; Coffin, D. L. (1971) Ozone effect on lysosomal
hydrolases of alveolar macrophages in vitro. Arch. Intern. Med.
127: 1059-1063.
Hyde, D. M.; Plopper, C. G. (1989) Ozone-induced structural
changes in monkey respiratory system. Proceedings of: The Third
U.S.-Dutch international symposium on atmospheric ozone research
and its policy implications; May; Nijmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; p. 523-
532. (Studies in environmental science 35).
Illing, J. W.; Miller, F. J.; Gardner, D. E. (1980) Decreased
resistance to infection in exercised mice exposed to NO 2 and O
3 . J. Toxicol. Environ. Health 6: 843-851.
Jacobson, J. S. (1977) The effects of photochemical oxidants on
vegetation. In: Ozon und Begleitsubstanzen im photochemischen
Smog. [Ozone and other substances in photochemical smog]: VDI
colloguium; 1976; Dusseldorf, West Germany: Verein deutscher
Ingenieure (VDI) GmbH; pp. 163-173. (VDI-Berichte no. 270).
Jacobson, J. S. (1982) Ozone and the growth and productivity of
agricultural crops. In: Unsworth, M. H.; Ormrod, D. P., eds.
Effects of gaseous air pollution in agriculture and horticulture.
London, United Kingdom: Butterworth Scientific; pp. 293-304.
James, C. (1980) Economic, social and political implications of
crop losses: a holistic framework for loss assessment in
agricultural systems. In: Crop loss assessment:; proceedings of E.
C. Stakman commemorative symposium; miscellaneous publication no.
7; pp. 10-16.
Javitz, H. S.; Kransnow, R.; Thompson, C.; Patton, K. M.;
Berthiaume, D. E.; Palmer, A. (1983) Ambient oxidant
concentrations in Houston and acute health symptoms in subjects
with chronic obstructive pulmonary disease: a reanalysis of the
HAOS health study. In: Lee, S.D.; Mustafa, M..G.; Mehlman, M.A. ,
eds. International symposium on the biomedical effects of ozone
and related photochemical oxidants; March 1982; Pinehurst, NC.
Princeton, NJ: Princeton Scientific Publishers,, Inc.; pp. 227-
256. (Advances in modern toxicology: v. 5).
-------
25
SwRl project no. 01-4902 Southwest Research Institute.
2SS:
.
v. 5). (Advances ln modern environmental toxicology:
elferti on;reioy rcton6"0? ar P°"^ion= its
Arch. Environ PHeal?S 30? ?l"?2ff elenentalT sch°°l children.
ren ed toauonn
-------
26
Kagawa, J.; Tsuru, K. (1979b) Respiratory effects of 2-hour
exposure to ozone and nitrogen dioxide alone and in combination
in normal subjects performing intermittent exercise. Nippon Kyobu
Shikkan Gakkai Zasshi 17: 765-774. ^yoou
Kagawa, J.; Tsuru, K. (1979a) Effects of ozone and smoking alone
and in combination on bronchial reactivity to inhaled
acetylcholine. Nippon KyobuShikkan Gakkai Zasshi 17: 703-709.
Kamath, Y. K. ; Ruetsch, S. B.; Weigmann, H. D. (1982)
Microspectrophotometric study of ozone fading of disperse dves in
nylon. Text. Res. J. : 53: 391-402.
' H* R*; vincent' L- M.; Kowalsky, R. J. ; Horstman, D. H. ;
O'Neil, J. j.; McCartney, W. H. ; Bromberg, P. A. (1987) Ozone
exposure increases respiratory epithelial permeability in humans.
Am. Rev. Respir. Dis. 135: 1124-1128.
Kehrl, H. R. ; Hazucha, M. J. ; Solic, J. j.; Bromberg, P. A.
(1985) Responses of subjects with chronic pulmonary disease after
exposures to 0.3 ppm ozone. Am. Rev. Respir. Dis. 131: 719-724.
Kehrl, H. R. ; Vincent, L. M. ; Kowalsky, R. J. ; Horstman, D. H. ;
O'Neil, J. j.; McCartney, W. H. ; Bromberg, P. A. (1989) Ozone-
induced changes in the pulmonary clearance of 99mTc-DTPA in man
In: Schneider, T. ; Lee, S. D. ; Wolters, G. J. R. ; Grant, L. D.'
eds. Atmospheric ozone research and its policy implications:
Proceedings of: The Third U.S. -Dutch international symposium;
May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands-
Elsevier Science Publishers; pp. 343-351. (Studies in
environmental science 35) .
Kehrl, H. R. ; Hazucha, M. J.; Solic, J. J.; Bromberg, P. A.
(1983) The acute effects of 0.2 and 0.3 ppm ozone in persons with
chronic obstructive lung disease (COLD). In: Lee, S. D. ; Mustafa,
M. G.^Mehlman, M. A., eds. International symposium on the
biomedical effects of ozone and related photochemical oxidants;
March 1982; Pinehurst, NC. Princeton, NJ: Princeton Scientific
Publishers, Inc.; pp. 213-225. (Advances in modern environmental
toxicology: v. 5) .
Kelleher, T. J. ; Feder, W. A. (1978) Phytotoxic concentrations of
ozone on Nantucket Island: long range transport from the middle
Atlantic states over the open ocean confirmed by bioassay with
ozone-sensitive tobacco plants. Environ. Pollut. 17: 187-194.
Kelly, N. A.; Wolff, G. T.; and Ferman, M. A. (1984) Sources and
sinks of ozone in rural areas. Atmospheric Environment 18: 1251-
1266 .
-------
27
-
Respir. 2: i-l. er* ceara*ce. Exp. Lung
Health 4: 521-532
«
Rev. Reir. is s s"* asthmatic -dole-cents.
2 S5S3-
-------
28
Kopp, R.J.; Vaughan, W.J.; Hazilla, M. (1984) Agricultural
sector benefits analysis for ozone: methods evaluation and
demonstration. Research Triangle Park, NC: U.S. Environmental
Protection Agency, OAQPS; EPA report no. EPA-450/5-84-003
Available from: NTIS, Springfield, VA; PB85-119477/XAB.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; Horstman,
,D. H. (1988a) Cellular and biochemical changes in the lower
airways of subjects exposed to ozone. Research Triangle Park
NC: U. S. Environmental Protection Agency, Health Effects
Research Laboratory; report no. EPA/600/D-88/031. Available from-
NTIS; Springfield, VA; PB88-170048.
Koren, H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; Horstman,
D. H.; Kozumbo, W. J.; Becker, S.; McDonnell, W. F. (1988b)
Cellular and biochemical changes in the lower airways of subjects
exposed to ozone. In: Sorg, C., ed. The alveolar macrophage.
Local Immunity 4: 36-49.
Koren; H. S.; Devlin, R. B.; Graham, D. E.; Mann, R.; McDonnel,
.W. F. (1989) The inflammatory response in human lungs exposed to
ambient levels of ozone. In: Schneider, T.; Lee, S. D.;
Wolters, G. J. R.; Grant, L. D., eds. Atmospheric ozone research
and its policy implications: Proceedings of: The Third U.S.-
Dutch international symposium; May 1988; Nijmegen, The
Netherlands. Amsterdam, The Netherlands: Elsevier Science
Publishers, pp. 745-753. (Studies in environmental science 35).
< '
Kreit, J. W.; Gross, K. B.; Moore, T. B.; Lorenzen, T. J.;
D'Arcy, J.; Eschenbacher, W. L. (1988) Ozone-induced changes in
pulmonary function and bronchial responsiveness in asthmatics.
J. Appl. Physiol. 66: 217-222.
Kress, L. W.; Skelly, J. M. (1982a) Response of several eastern
forest tree species to chronic doses of ozone and nitrogen
dioxide. Plant Dis. 66: 1149-1152.
Kress, L. W.; Skelly, J. M.; Hinkelmann, K. H. (1982b) Growth
impact of O3, N02, and/or SO2 on Pinus taeda. Environ. Monit.
Assess. 1: 229-239.
Kress, L. W.; Skelly, J. M.; Hinkelmann, K. H. (1982c) Growth
impact of O3, NO2, and/or SO2 on Platanus occidentalis. Agric.
Environ. 7: 265-274.
Kress, L. W.; Skelly, I. M.; Hinkelmann, K. H. (1982d) Relative
sensitivity of 18 full-sib families of Pinus toeda to 0 3 . Can.
J. For. Res. 12: 203-209.
Kress, L. W.; Miller, J. E.; Smith, H. J. (I985a) Impact of ozone
on winter wheat yield. Environ. Exp. Bot. 25: 211-228.
-------
29
«««• « «•" com
needed source reductions: Part ? An o -n T
mathematical model, j. Air Pollut coJ^ A lni?Ury
AJ.IT i'oj.iut:. Control Assoc. 33: 198-207.
ESS? Slnt^n^:/^" *°*™» «°P «*«*ion. ,. Mr
-------
30
Last, J. A.; Greenberg, D. B.; Castleman, W. L. (1979) Ozone-
induced alterations in collagen metabolism of rat lungs. Toxicol.
Appl. Pharmacol. 51: 247-258. •
Last, J. A.; Greenberg, D. B. (1980) Ozone-induced alterations in
collagen metabolism of rat lungs. II. Long-term exposure.
Toxicol. Appl. Pharmacol. 55: 108-114.
Last, J. A.; Hesterberg, T. W.; Reiser, K. M.; Cross, C. E. ;
^?J!i'tTA C"* ?U2n' C"' Steffev' E- *•; Grandy, J.; Henrickson, R.
(1981) Ozone-induced alterations in collagen metabolism of monkey
lungs: use of biopsy-obtained lung tissue. Toxicol. Apol
Pharmacol. 60: 579-585.
Last, J. A.; Gerriets, J. E.; Hyde, D. M. (1983) Synergistic
effects on rat lungs of mixture of oxidant air pollutants (ozone
or nitrogen dioxide)and respirable aerosols. Am. Rev. Resnir
Dis. 128: 539-544. P
Last, J. A.; Hyde, D. M.; Chang, D. P. Y. (1984a) A mechanism of
synergistic lung damage by ozone and a respirable aerosol EXD
Lung Res. 7: 223-235. F*
Last, J. A.; Reiser, K. M.; Tyler, W. S.; Rucker, R. B. (1984b)
Long-term consequences of exposure to ozone; I. lung collagen
content. Toxicol. Appl. Pharmacol. 72: 111-118.
Laurence, J. A.; Wood, F. A. (1978) Effects of ozone on infection
of soybean by Pseudomonas glycinea. Phytopathology 68: 441-445.
Lauritzen, S. K. ; Adams, W, C,. (1985) Ozone inhalation effects
consequent to continuous exercise in females: comparison to
males. J. Appl. Physiol. 59: 1601-1606.
Lebowitz, M. D.; O'Rourke, M. K.; Dodge, R.; Holberg, C. J. ;
Gorman, G.; Hoshaw, R. W.; Pinnas, J. L.; Barbee, R. A.; Sneller,
M. R. (1982) The adverse health effects of biological aerosols,
other aerosols, and indoor microclimate on asthmatics and
nonasthmatics. Environ. Int. 8: 375-380.
Lebowitz, M. D.; Holberg, C. J.; Dodge, R. R. (1983) Respiratory
effects on populations from low level exposures to ozone.
Presented at: 34th annual meeting of the Air Pollution Control
Association; June; Atlanta, GA. Pittsburgh, PA: Air Pollution
Control Association; paper no. 83-12.5.
Lebowitz, M. D. (1984) The effects of environmental tobacco smoke
exposure and gas stoves on daily peak flow rates in asthmatic and
non-asthmatic families. Eur. J. Respir. Dis. 65(suppl. 133): 90-
97«
-------
31
Lebowitz, M. D.; Holberg, C. j. ; Boyer, B.; Hayes
production and for estimating agricultural losses
"
Lee, E. H . ; Tingey, D. T. , and Hogsett, W. E. (1988c)
Interrelationship of Experimental Exposure and Ambient Air
Quality Data for Comparison of Ozone Exposure
Lefohn A. s.; Tingey, D. T. (1984) The co-occurrence of
-------
32
N P' • Fif;L*U n T ^ V Spier' C- E'; Ble«sey, R. L. ; Jones,
M. P., Fischer, D. A.; Hackney, J. D. (1978) Health effects of
ozone exposure in asthmatics. Am. Rev. Respir. Dis. 117: 835-843.
voETwM'vP*; Bachmayer' E- *•; Spier, C. E.; Mazur,
™1'*-1-; Hackney, a. D. (1980) Short-term respiratory
OV^H P°llute<* .fir: a laboratory study of volunteers in a
-oxidant community. Am. Rev. Respir. Dis. 121: 243-252.
W; S*; Medwav' D- A-; Anzar, u. T. ; Valencia, L. M. ; Spier
; Tsao, F. S-0. ; Fischer, D. A.; Hackney, J. D. (1982)
Persistence of adaptation to ozone in volunteers exposed
repeatedly over six weeks. An. Rev. Respir. Dis, 1251 491-495.
Linn, W. S.; Avol, E. L. ; Hackney, J. D. (1983a) Effects of
ambient oxidant pollutants on humans: a movable environmental
chamber study In: Lee, S. D. ; Mustafa, M. G. ; Mehlman, M. A.,
eds. International symposium on the biomedical effects of ozone
and related photochemical oxidants; March 1982; Pinehurst NC
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. ' 125-
137.. (Advances in modern toxicology: v. 5) .
Linn, W. S.; Shamoo, D. A.; Venet, T. G. ; Spier, C. E. ; Valencia,
L. M.; Anzar, U. T. ; Hackney, J. D. (1983b) Response to ozone in
volunteers with chronic obstructive pulmonary disease. Arch
Environ. Health 38: 278-283.
Linn, W s.; Avol, E. L. ; Shamoo, D. A.; Spier, C. E. ; Valencia,
L. M.; Venet, T. G. ; Fischer, D. A.; Hackney, J. D. (1986) A dose
response study of healthy heavily exercising mean exposed to
ozone at concentrations near the ambient air quality standard
Toxicol. Ind. Health 2: 99-112.
Linn, W. S.; Avol, E. L. ; Shamoo, D. A.; Peng, R.-C.; Valencia,
L. M.; Little, D. E. ; Hackney, J. D. (1988) Repeated laboratory
ozone exposures of volunteer Los Angeles residents: apparent
seasonal variation in response. Toxicol. .Ind. Health 4: 505-520.
Lioy, P. j.; Vollmuth, T. A.; Lippmann, M. (1935) Persistence of
peak flow decrement in children following ozone exposures
exceeding the National Ambient Air Quality Standard. J. Air
Pollut. Control Assoc. 35: 1068-1071.
Lioy, P. J.; Dyba, R. V. (1989) The dynamics of human exposure to
tropospheric ozone. In: Schneider, T. ; Lee, S. D. ; Wolters, G.
J. R. ; Grant, L. D. , eds. Atmospheric ozone research and its
policy implications: Proceedings of: The Third U.S. -Dutch
international symposium; May 1988; Nigmegen, The Netherlands.
Amsterdam, The Netherlands: Elsevier Science Publishers; pp.
711-721. (Studies in environmental science 35) .
dose-
-------
33
, D.;
me.,. »
p- to
' "6
ser.es: 43,. Available from:
SiS - -
by
wed^ Si .
stress. Environ. Exp.Bot. 20: 99-iJS? chronic air pollution
-------
34
Mauderly, J. L. (1984) Respiratory function responses of animals
and man to oxidant gases and with pulmonary emphysema. J.
Toxicol. Environ. Health 13: 345-361.
McAllen, S. J.; chiu, S. P.; Phalen, R. F.; Rasmussen, R. E.
(1981) Effect of in vivo ozone exposure on in vitro pulmonary
alveolar macrophage mobility. J. Toxicol. Environ. Health 7: 373-
3 o J» •
McCarthy, E. F.; Stankunas, A. R.; Yocum, J. E.,; Rae, D. (1983)
Damage cost models for pollutant effects on material. East
Hartford, CT: TRC Environmental Consultants, Inc.; TRC project
no. 1574-J80. U.S. Environmental Protection Agency; EPA report
no. 600/3-84-012.
McClellan, R. O. (1989) Letter from Chairman of Clean Air
Scientific Advisory Committee to the EPA Administrator concerning
"closure" on the Ozone Criteria Document Supplement and the Ozone
Staff Paper, dated May 1, 1989.
McClenahen, J. R. (1979) Effects of ethylene diurea and ozone on
the growth of the tree seedlings. Plant Dis. Rep. 63: 320-323.
McCool, P. M.; Menge, J. A.; Taylor, O. C. (1979) Effects of
ozone and HC1 gas on the development of the mycorrhizal fungus
Glomus faciculatus and growth o of 'Troyer' citrange. J. Am. Soc.
Hortic. Sci. 104: 151-154.
McCool, P. M.; Musselman, R. C.; Teso, R. R.; Oshima, R. J.
(1986) Determining crop yield losses from air pollutants. Calif.
Agric. July-August: 9-10.
McCool, P. M.; Musselman, R. C.; Teso, R. R. (1987) Air pollutant
yield->loss assessment for four vegetable crops. Ag. Ecosys. and
Environ. 20: 11-21.
McCurdy, T. (1984) Miscellaneous analyses of alternative O3
NAAQS formulations and their impact on O3 non-attainment and cost
analyses procedures. Research Triangle Park, N.C.: U.S.
Environmental Protection Agency.
McCurdy, T. (1987) Descriptive statistical analyses of ozone air
quality indicators in rural/remote and metropolitan areas.
Research Triangle Park, N.C.: U.S. Environmental Protection
Agency.
McCurdy, T.; Atherton, R.A. (1988) Variability of ozone air
quality indicators in selected metropolitan statistical areas.
Research Triangle Park, N.C.: U.S. Environmental Protection
Agency.
-------
35
' I' n"* ?°fstmann' D- H. ; Hazucha, M. J. ; Seal, E
t-j=r-
-•
-
-------
36
McLaughlin, S. B.; McConathy, R. K.; Duvick, D,,; Mann L K
(1982) Effects of chronic air pollution stress on photosynthesis
carbon allocation, and growth of white pine trees. For. Sci. 28:
60—70.
McLaughlin, S.B. (1985) Effects of air pollution on forest: A
critical review. J. Air Pollut. Assoc. 35: 512-534.
Meiners, J. p.; Heggestad, H. E. (1979) Evaluation of snap bean
cultivars for resistance to ambient oxidants in field plots and
to ozone in chambers. Plant Dis. Rep. 63: 273-277.
Menser, H. A.; Street, 0. E. (1962) Effects of air pollution
nitrogen levels, supplemental irrigation, and plant spacing on
weather fleck and leaf losses of Maryland tobacco. Tobacco 155:
192—19 6.
Menser, H. A.; Hodges, G. H. (1972) Oxidant injury to shade
tobacco cultivars developed in Connecticut for weather fleck
resistance. Agron. J. 64: 189-192.
Merz, T.; Bender, M. A.; Kerr, H. D.; Kulle, T. J. (1975)
Observations of aberrations in chromosomes of lymphocytes from
human subjects exposed to ozone at a concentration of 0.5 ppm for
6 and 10 hours. Mutat. Res. 31: 299-302. -
Miller, P. R.; Parmeter, J. R., Jr.; Taylor, O. C.; Cardiff, E.
A. (1963) Ozone injury to the foliage of Pinus ponderosa
Phytopathology 53: 1072-1076.
Miller, P. R.; Parmeter, J. R., Jr.; Flick, B. H.; Martinez, C.
W. (1969) Ozone dosage response of ponderosa pine seedlings. J.
Air Pollut. Control Assoc. 19: 435^-438.
Miller, P. R. (1973) Oxidant"induced community change in a mixed
conifer forest. In: Naegle, J. A., ed. Air pollution damage to
vegetation. Washington, DC: American Chemical Society; pp. 101-
117. (Advances in Chemistry series no. 122).
Miller, P. M.; Tomlinson, H. ; Taylor, G. S. (1976) Reducing
severity of ozone damage to tobacco and beans by combining
benomyl or carboxin with contactnematicides. Plant Dis. Rep. 60-
433-436.
Miller, P. R.; Elderman, M. J., eds. (1977) Photochemical oxidant
air pollutant effects on a mixed conifer forest ecosystem: a
progress report, 1976. Corvallis, OR: U.S. Environmental
Protection Agency; EPA report no. EPA-600/3-77-104. Available
from: NTIS, Springfield, VA; PB-274531.
-------
37
lung to lower respiratory tract 2™??? of ozone in the human
Toxicol. Appl. PharmacolT 7^11-?? a°d t0 exerci^-
S^i-SS? ^
8*^^
2' *'
Toxicol. Lett. 2: .-mcuced respiratory infections.
bten manaabratoryaniS^fin' D' L" (19?8b)
ozone. Environ. Res? "7° 84^101 re9lon^ deposition of
EPA report no. EPA-eSo/D-SV/oJo Effects Research Laboratory;
EPA report no. EPA-600/D-87/1?! Efe^s Research Laboratory;
N.C.: u.S.
of bonnet monkeys. Exp. Lung
National air
-------
38
Monitoring and Data Analysis Division (MDAD, 1987) Metropolitan
statistical areas (MSA) regulatory analysis air quality data
base, 1983-1985." Durham, N.C.: U.S. EPA.
Mooi, J. (1980) Influence of ozone on the growth of two ponlar
cultivars. Plant Dis. 64: 772-773. .
Moore, P. F.; Schwartz, L. W. (1981) Morphological effects of
prolonged exposure to ozone and sulfuric acid aerosol on the rat
lung. Exp. Mol. Pathol. 35: 108-123.
Moore, G. S.; Calabrese, E. J.; Schulz, E. (1931) Effect of in
vivo ozone exposure to Dorset sheep, an animal model with low"
levels of erythrocyte glucose-6-phosphate dehydrogenase activity.
Bull. Environ. Contam. Toxicol. 26: 273-280.
Moorman, W. J.; Chmiel, J. J.; Stara, J. F.; Lewis, T. R. (1973)
Comparative decomposition of ozone in the nasopharynx of beagles:
acute vs. chronic exposure. Arch. Environ. Health 26: 153-155.
Mueller, W. J.; Stickney, P. B. (1970) A survey and economic
assessment of the effects of air pollution on elastomers.
Columbus, OH: Battelle Memorial Institute; NAPCA contract no.
CPA-22-69-146.
Mumford, R. A.; Lipke, H.; Laufer, D. A.; Feder, W. A. (1972)
Ozone induced changes in corn pollen. Environ. Sci. Technol. 6-
427-430.
Musselman, R. C.; Huerta, A. J.; McCool, P. M.; Oshima, R. J.
(1986) Response of beans to simulated ambient and uniform ozone
distributions with- equal peak concentrations. J. Amer. Soc.
Horticult. Sci. ill: 470-473.
Musselman, R. C.; McCool, P. M.; Younglove, T. (1988) Selecting
ozone exposure statistics for determining crop yield loss from
air pollutants. Environ. Pollut. 53: 63-78.
Musselman, R. C.; Oshima, R. J.; Gallavan, R. E. (1983)
Significance of pollutant concentration distribution in the
response of 'red kidney1 beans to ozone. J. Am. Soc. Hortic. Sci.
108: 347-351.
Mustafa, M. G.; Tierney, D. F. (1978) Biochemical and metabolic
changes in the lung with oxygen, ozone, and nitrogen dioxide
toxicity. Am. Rev. Respir. Dis. 118: 1061-1090.
National Research Council. (1977) Toxicology. In: Ozone and other
photochemical oxidants. Washington, DC: National Academy of
Sciences, Committee on Medical and Biologic Effects of
Environmental Pollutants; pp. 323-387.
-------
39
quality and N-fi
of
NC.
Engl. j. Med. 291: 755-758.
518-520.
NC
Springfield, VA; PB-26423
of -one^
Soc. Hortic. Sci.103: 348-350
°n yield'
from: NTIS,
Y°Ung ci?arette smokers. N
--
fading. Text. cheln.
ecosystem.
beet. J. Environ. Qual. 2:
a:
in parsley, j. Am.
-------
40
97E ' P. K-; Baldwin, D. W.
(1975) Effect of ozone on the yield and plant biomass of a
commercial variety of tomato. J. Environ. Qual. 4: 463-464.
Oshima, R J. (1978) The impact of sulfur dioxide on vegetation:
a sulfur dioxide-ozone response model. Sacramento, CA: California
Air Resources Board; final report; agreement no. A6-162-30.
Overton, J. H. ; Graham, R. c. ; Miller, F. J. (1987) A model of
the regional uptake of gaseous pollutants in the lung. II The
sensitivity of ozone uptake in laboratory animal lungs to
aat
o ventilatory parameters. Toxicol. Appl. Pharmacol.
88 '. 418—432 .
P'an, A. Y. S.; Beland, J. ; Jegier, Z. (1972) Ozone-induced
arterial lesions. Arch. Environ. Health 24: 229-232.
Parry, M. A.. J. ; Day, W. (1988) A comparison of indices that
describe the relationship between exposure to ozone and reduction
in the yield of agricultural crops [Comments on article by A S
Lefohn, J.A. Laurence and R.J. Kohut] . Atmos. Environ. 22: (in*
press) .
Patton, R. L. (1981) Effects of ozone and sulfur dioxide on
height and stem specific gravity of Populus hybrids. Broomall
PA: Northeastern Forest Experiment Station; FSRP-NE-471
Available from: NTIS, Springfield, VA; PB81-17 1274.
Paul, R, A.; Johnson, T.; Pope, A.; Ferdo, A.; and Biller, W. F.
(1986) National estimates of exposure to ozone under alternative
national standards. Durham, N.C.: PEI Associates, Inc.
Pell, E. J.; Weissberger, W. C. ; Speroni, J. J. (1980) Impact of
ozone on quantity and quality of greenhouse-grown potato plants
Environ. Sci. Technol. 14: 568-571.
Peterson, D.C.; Sueker, J.K.; Wallsten, T.S. (1987). Risks of
Forest Tree Response Due to Ambient Ozone (Draft Final Report) .
Energy and Resource Consultants, Inc. for U.S., Environmental
Protection Agency. Research Triangle Park, NC,.
Peterson, D.C. and D. Violette. 1986. Dimensioning Uncertainty
in Estimates of Regional Fish Population Damage Caused by
Acidification in Adirondak Ponded Waters, U.S. EPA, 200-12-86-
022.
Peterson, D.C. and J.K. Sueker. 1987. "Risks of forest response
due to ambient ozone," APCA 87-36.3.
-------
41
Environ. Health 6
978lO
P. 11. '
environmental science 35)
(Studies in
Pollution centre sSociation; paper°no' S?^
2SS-
' PA= Air
29: 207-214.
-------
42
Raub, J. A.; Miller, F. J.; Graham, J. A. (1983) Effects of low-
level ozone exposure on pulmonary function in adult and neonatal
rats. In: Lee,S. D.; Mustafa, M. G.; Mehlman, M. A., eds.
International symposium on the biomedical effects of ozone and
related photochemical oxidants; March 1982; Pinehurst NC
Princeton, NJ: Princeton Scientific Publishers, Inc.; pp."363-
367. (Advances in modern environmental toxicology: v. 5).
Raven, P. B., Drinkwater, B. L.; Horvath, S. M.; Ruhling, R. o.;
Gliner, J.A.; Sutton, J. C.; Bolduan, N. W. (1974a) 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.
Raven, P. B., Drinkwater, B. L.; Ruhling, R. o.; Bolduan, N •
Taguchi, S.; Gliner, J. A.; Horvath, S. M. (1974b) Effect of
carbon monoxide and peroxyacetyl nitrate on man's maximal aerobic
capacity. J. Appl. Physiol. 36: 288-293.
Rawlings, J. o.; Lesser, V. M.; Heagle, A. S.; Heck W. W. (1988)
Alternative ozone dose metrics to characterize ozone impact on
crop yield loss. J. Environ. Qual. 17: 285-291.
Rawlings, J. O.; Cure, W. W. (1985) The Weibull function as a
doseresponse model to describe ozone effects on crop yields. Crop
Science (in press).
Reich, P. B. (1983) Effects of low concentrations of 0, on net
photosynthesis, dark respiration, and chlorophyll contents in
aging hybrid poplar leaves. Plant Physiol. 73:: 291-296.
Reich, P. B.; Amundson, R. G. (1986) Ambient levels of ozone
reduce net photosynthesis in tree and crop species. Science 230:
566—570.
Reich, P. B.; Amundson, R. G. (1984) Low level 03 and/or SO,
exposure causes a linear decline in soybean yield. Environ.
Pollut. (Series A) 34: 345-355.
Reinert, R. A.; Nelson, P. V. (1979) Sensitivity and growth of
twelve elatior begonia cultivars to ozone. Hort Science 14: 747-
748.
Reinert, R. A.; Gray, T. N..(1981) The response of radish to
nitrogen dioxide, sulfur dioxide, and ozone, alone and in
combination. J. Environ. Qual. 10: 240-243.
Reisenauer, C. S.; Koenig, J. Q.; McManus, M. S.; Smith, M. S.;
Kusic, G.; Pierson, W. E. (1988) Pulmonary responses to ozone
exposures in healthy individuals aged 55 years or greater. J. Air
Pollut. Control Assoc. 38: 51-55.
-------
43
2
Appl. Pharmacol? sJT jJJ-L 9 C°llagen of m°^eys. Toxicol
Richards, B. L. ; Middleton, J. T. ; Hewitt w
"'
913-917.
Atnosphirio pc ...
environmental science 35). (Studies in
a sss
-------
44
!*"d?f!4
-------
45
Shannon . j . G . : Mu 1 r h i r T / 1 n •? ^ \ ^
varieties anthisis crip sci! 1" ^5^ daMge tO Wheat
Sherwin, R. P.; Richters, V. (1985) Effect
S"~
J- M' (1980) Photochemical oxidant impact
on
°
-------
46
Skelly, J. M.; Yaw-Shing Yang; Chevone, B. J.; Long, S. J. (1984)
Ozone concentrations and their influence on forest species in the
Blue Ridge Mountains of Virginia. In: Davis, D. D.; Millen, A.A.;
Dochinger, L., eds. Air pollution and the productivity of the
forest: proceedings of the symposium; October 1983; Washington
D.C. Arlington, VA: Izaak Walton League of America; pp.143-159.
Slade, R.; Stead, A. G.; Graham, J. A.; Hatch, G. E. (1985)
Comparison of lung antioxidant levels in humans and laboratory
animals. Am. Rev. Respir. Dis. 131: 742-746.
Smith, G.; Greenhalgh, B.; Brennan, E.; Justin, J. (1987) Soybean
yield in New Jersey relative to ozone pollution and antioxidant
application. Plant Dis. 71: 121-125.
Solic, J. J.; Hazucha, M. J.; Bromberg, P. A. (1982) The acute
effects of 0.2 ppm ozone in patients with chronic obstructive
pulmonary disease. Am. Rev. Respir. Dis. 125: 664-669.
Spektor, D. M.; Lippmann, M. ; Lioy, p. J.; Thurston, G. D.;
Citak, K.; James, D. J.; Bock, N.; Speizer, F. E.; Hayes, C.
(1988a) Effects of ambient ozone on respiratory function in
active normal children. Am. Rev. Respir. Dis. 137: 313-320.
Spektor, S. M.; Lippmann, M.; Thurston, G. D.; Lioy, P. J.;
Stecko, J.; O'Connor, G.; Garshick, E.; Speizer, F. E.; Hayes, C.
(1988b) Effects of ambient ozone on respiratory function in
healthy adults exercising outdoors. Am. Rev. Respir. Dis. 138:
821-828.
Stark, R. w.; Miller, P. R.; Cobb, D. L., Jr.; Wood, D. L.;
Parmeter, J. R., Jr.(1968) Photochemical oxidant injury and bark
beetle (Coleoptera: Scolyticlea) infestation of ponderosa pine: I.
Incidence of bark beetle infestation of ponderosa pine: I.
Incidence of bark butter infestation in injured trees. Hilqardia
39: 121-126.
Stark, R. W.; Cobb, F. W., Jr. (1969) Smog injury, root diseases
and bark beetle damage in ponderosa pine. Calif. Agric. 23: 13-
JLO •
Stephens, R. J.; Sloan, M. F.; Evans, M. J.; Freeman, G. (1974)
Early response of lung to low levels of ozone. Am. J. Pathol. 74-
31-58.
Stephens, R. J.; Sloan, M. F.; Groth, D. G. (1976) Effects of
long-term, low level exposure of NO2 or O3 on rat lungs. EHP
Environ. Health Perspect. 16: 178-179.
Stephens, R. J.; Sloan, M. F.; Groth, D. G.; Negi, D. S.; Lunan,
K. D. (1978) Cytologic responses of postnatal rat lungs to 0, or
N02 exposure. Am. J. Pathol. 93: 183-200.
-------
47
the biomedical effects of ozone and reStS"^1?11^ sYmP°sium °*
ovirJan+'c.. M,—i. ,««_ _T «*«-»*'« ana related photochemi r?ai
the biomedical' effec of ozone SS'reSSTSh1?"^ 8Y»Po«iu»
oxidants; March 1982- Pin^h,,?=J SS r®1?ted Photochemical
Scientific Publisher^ inc
-------
48
Thompson, C. R. ; Kats, G. ; Cameron, J. w. (1976) Effects of
ambient photochemical air pollutants on growth, yield, and ear
character of two sweet corn hybrids. J. Environ. Qua 1. 5? 4?0-
4 Xb e
Thurlbeck, W. M. (1984) The pathobiology .and epidemiology of
human emphysema. In: Miller, F.J. and MenzelT D.B., eds.
SS?SI2S 5 ?f . e^rapolation modeling of inhaled toxicants.
Hemisphere Publishing Corporation; pp. 323-343.
Tice, R. R.; Bender, M. A.; Ivett, J. L. ; Drew, R. T. (1978)
Cytogenetic effects of inhaled ozone. Mutat. Res. 58: 293-304.
Tingey, D. T. ; Wilhour, R. G. ; Standley, E. (1976) The effect of
chronic ozone exposures on the metabolite content of ponderosa
pine seedlings. For. Sci. 22: 234-241. ponaerosa
Tingey, D. T.; Reinert, R. A.; Carter, H. B. C1972) Soybean
cultivars: acute -foliar response to ozone. Crop Sci. 12: 268-270.
7™ «??'«?' ?"' ?hUtt: f* L:; GumPertz' M- !••; Hogsett, W. E.
(1982) Plant water status influences ozone sensitivity of bean
plants. Agric. Environ. 7: 243-254.
Tingey, D. T. ; Taylor, G. E. , Jr. (1982) Variation in plant
response to ozone: a conceptual model of physiological events.
In: Unsworth, M. H. ; Ormrod, D. P., eds. Effects of gaseous air
pollution in agriculture and horticulture. London, United
Kingdom Butterworth Scientific; pp. 113-138.
Tingey, p. T. ; Reinert, R. A. (1975) The effect of ozone and
sulfur dioxide singly and in combination on plant growth
Environ. Pollut. 9: 117-125.
Tingey, D. T.; Hogsett, W. E. ; Lee, E. H. (1989) Analysis of crop
loss for alternative ozone exposure indices. In: Atmospheric
Ozone Research and its Policy Implications. T. Schneider G
Wolters, s. Lee and L. Grant (eds.). Elsevier Science Publishers
B.V., Amsterdam, The Netherlands. (Studies in environmental
science 35). pp. 219-227.
Tingey, D. T.; HecJc, W. W. ; Reinert, R. A. (1971) Effect of low
concentrations of ozone and sulfur dioxide on foliage, growth and
yield of radish. J. Am.Soc. Hortic. Sci. 96: 369-371.
Tingey, D. T. (1974) Ozone-induced alterations in the metabolite
pools and enzyme activities of plants. In: Dugger, M. , ed. Air
pollution effects on plant growth. Washington, DC: American
Chemical Society; pp. 40-57.
-------
49
Toivonen, P. M. A.; Hofstra, G.; Wukasch, R. T. (1982) Assessment
of yield losses in white bean due to ozone using the antioxidant
EDU. Can. J. Plant Path. 4: 381-386.
Tonneijck, A. E. G. (1984) Effects of peroxyacetyl nitrate (PAN)
and ozone on some plant species. In: Proceedings of the OECD
workshop on ozone; Gothenburg, Sweden.
Treshow, M. (1980) Pollution effects on plant distribution.
Environ. Conserv. 7: 279-286.
Treshow, M.; Stewart, D. (1973) Ozone sensitivity of plants in
natural communities. Biol. Conserv. 5: 205-214.
Tyler, W. S.; Tyler, N. K.; Last, J.A.; Gillespie, M.J.; Barstow,
T.J. (1988) Comparison of daily and seasonal exposures of young
monkeys to ozone. Toxicology 50: 131-144.
Tyson, C. A.; Lunan, K. D.; Stephens, R. J. (1982) Age-related
differences in GSH-shuttle enzymes in NO 2- or O3-exposed rat
lungs. Arch. Environ. Health 37: 167-176.
U.S. Court of Appeals for the District of Columbia, American
Petroleum Institute vs. EPA. No. 79-1104 and 79-1201, September
1981.
U.S. Court of Appeals for the District of Columbia Circuit, Lead
Industries Association. Inc. vs. EPAf "No. 78-2201 and 78-2220.
June 27, 1980.
U.S. Department of Health and Human Services (1981) Current
estimates from the National Health Interview Survey: United
States, 1979. Hyattsville, MD: Public Health Service, Office of
Health Research, Statistics and Technology, National Center for
Health Statistics; DHHS publication no. (PHS) 81-1564. (Vital and
health statistics: series 10, no. 136).
U.S. Environmental Protection Agency (1978) Air quality criteria
for ozone and photochemical oxidants. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Environmental Criteria and*
Assessment Office; EPA-600/8-78-004. Available from: NTIS
Springfield, VA; PB80-124753.
U.S. Environmental Protection Agency (1980) Air quality data—
1979 annual statistics including summaries with reference to
standards. Research Triangle Park, NC: Office of Air Quality
Planning and Standards; EPA report no. EPA-450/4-80-014.
Available from: NTIS, Springfield, VA; PB81-123093.
-------
50
U.S. Environmental Protection Agency (1981) Air quality data—
1980 annual statistics including summaries with reference to
standards. Research Triangle Park, NC: Monitoring and Data
Analysis Division; EPA report no. EPA-450/4-81--027. Available
from: NTIS, Springfield, VA; PB82-106501.
U.S. Environmental Protection Agency (1982a) Air quality criteria
for oxides of nitrogen. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Environmental Criteria and
Assessment Office; EPA report no. EPA-600/8-82--026F. Available
from: NTIS, Springfield, VA; PB83-163337.
U.S. Environmental Protection Agency (1982b) Air quality
criteria for particulate matter and sulfur oxides. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/8-82-029a,b,c,d.
U.S. Environmental Protection Agency (1986) Air quality criteria
for ozone and other photochemical oxidants. Draft Final.
Environmental Criteria and Assessment Office; EPA Report No. EPA-
600/8-84/020a to 020e. Available from: NTIS, Springfield, VA- PB
87-142949.
U.S. Environmental Protection Agency (1988) Summary of Selected
New Information on Effects of Ozone on Health and Vegetation:
Draft Supplement to Air Quality Criteria for Ozone and Other
Photochemical Oxidants. Research Triangle Park, NC: Office of
Health and Environmental Assessment, Environmental Criteria and
Assessment .Office; EPA/600/8-88/105A. Available from NTIS
Springfield, VA.
U.S. Senate (1979) 93rd Congress, 2nd Session Committee on
Environment and Public Works (Comm. Print 1974). A legislative
history of the clean air act amendments of 1970, Vol. l, p. 410.
Upham, J. B.; Haynie, F. H.; Spence, J. W. (1976) Fading of
selected drapery fabrics by air pollutants. J. Air Pollut.
Control Assoc. 26: 790-792.
van Bree, L.; Rombout , P. J. A.; Rietjens, I. M. C. M.; Dormans,
J. A. M. A.; Marra, M. (1989) Pathobiochemical effects in rat
lung related to episodic ozone exposure. In: Schneider, T.;
lee, S. D.; Wolters, G. J. R.; Grant, L. D., eds. Atmospheric
ozone research and its policy implications: Proceedings of: The
Third U.S.-Dutch international symposium; May 1988; Nijmegen, The
Netherlands. Amsterdam, The Netherlands: Elsevier Science
Publishers; pp. 723-732. (Studies in environmental science 35).
Veith, A. G.; Evans, R. L. (1980) Effect of atmospheric
pressure on ozone cracking of rubber. Polym. Testing 1: 27-38.
-------
51
on
adhesion
function and
Assessment of
Toxiool. Appl. Phar. «9H137
' - "
Sprague-Dawley Rats
Wilhour, R. G.; Neely G E (1Q77)
^ternatLnl/confe^encrS^photochemi J?^^*?lade8' °*"* «*•
its control: proceedings, vol. "-. ^-_. ^ox_ ant Pollution and
Park, NC. U.S. Environmental*
report no. EPA-600/3-77-ooib
Springfield, VA; PB-264233
: NTIS,
635-645; EPA
-------
52
Witz G.; Amoruso, M. A.; Goldstein, B. D. (1983) Effect of ozone
on alveolar macrophage function: membrane dynamic properties. In-
Lee, S. D.; Mustafa,M. G.; Mehlman, M. A., eds. International
symposium on the biomedical effects of ozone and related
photochemical oxidants; March 1982; Pinehurst, NC. Princeton
NJ: Princeton Scientific Publishers, Inc,; pp. 263-272. (AdvaAces
in modern enviromental toxicology: v. 5).
Woodwell, G. M. (1970) Effects of pollution on the structure and
physiology of ecosystems. Science (Washington, DC)- 168: 429-433.
Woodwell, G. M. (1974) Success, succession, and Adam Smith.
BioScience 24: 81-87.
World Health Organization (1977) Manual of the international
statistical classification of diseases, injuries, and causes of
death. Geneva, Switzerland: World Health Organization.
Wright, J. L.; Lawson, L. M.; Pare, P. D.; Wiggs, B.'J.; Kennedy,
S.; Hogg, J.C. (1983) Morphology of peripheral airways in current
smokers and ex-smokers. Am. Rev. Respir. Dis. 127: 474-477.
Wright, E. S.; Kehrer, J. P.; White, D. M.; Smiler, K. L. (1988a)
Effects of chronic exposure to ozone on collagen in rat lunq
Toxicol. Appl. Pharm. 92: 445-452.
Wukasch, R. T.; Hofstra, G.' (I977a) Ozone and Botrvtis
interactions in onion leaf dieback: open-top chamber studies
Phytopathology 67: 1080-1084.
Wukasch, R. T.; Hofstra, G. (1977b) Ozone and Botrvtis spp.
interaction in onion leaf dieback: field studies. J. Am. Soc
Hortic. Sci. 102: 543-546.
Yang, Y.S.; Skelly, J. M.; Chevone, B. I.; Birch, J. B. (1983)
Effects of long-term ozone exposure on photosynthesis and dark
respiration of eastern white pine. Environ. Sci. Technol. 17-
371-373.
Yokoyama, E.; Frank, R. (1972) Respiratory uptake of ozone in
dogs. Arch. Environ. Health 25: 132-138.
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E., Jr.; Dunavant, B. G.;
Bevis, H. A. (1971b) Inhaled ozone as a mutageri. II. Effect on
the frequency of chromosome aberrations observed in irradiated
Chinese hamsters. Environ. Res. 4: 325-342.
Zelac, R. E.; Cromroy, H. L.; Bolch, W. E., Jr.; Dunavant, B. G.;
Bevis, H. A. (I97la) Inhaled ozone as a mutagen. I. Chromosome
aberrations induced in Chinese hamster lymphocytes. Environ. Res.
4: 262-282.
-------
53
-376. . . nviron. Pathol. Toxicol
-------
-------
TECHNICAL REPORT DATA
.Please read Instructions on the reverse be/ore completing)
1. REPORT NO.
EPA-450/a-92-OQl
[<*. TITLE AND SUBTITLE
Review of the National Ambient Air Quality Standards
tor Ozone-Assessment of Scientific and Technical
Information: QAQPS Staff Paper
7. AUTHORIS) —
McKee, D. J.; Johnson, P. M.; McCurdy, T M •
Richmond, H. M. ''
9. PERFORMING ORGANIZATION NAME AND ADDRESS '
Office of Air and Radiation
Office of Air Quality Planning and Standards
u. S. Environmental Protection Agency
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS '
15. SUPPLEMENTARY NOTES
5. REPORT DATE
June 1989
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO?
11. CONTRACT/GRANT NO.
3. TYPE OF REPORT AND PERIOD COVERED
Final
4. SPONSORING AGENCY CODE
scientific and technical information
ambient air quality standardTfor^onrasUwen IT^ °5 ****!*** ^^ nati°nal
ozone. This staff paper provides staff eon ? secondary (welfare) standards for
oxidants; (2) The current fora of the ozone con"ntrations of photochemical
standard is attained when expected number"of''davs/vfar8!!?^ bVeta*ned (i>e-'
concentrations above the level of the standard HT maximum hourly
1-hour standards should be retained althh ls.e^ual or greater than one; (3) Th
-future development of alte^e fl^f b^th^p^im^rfa^"1^ ^^ b' S^
range of consideration for orimarv c;fanHa^^ u TJ v n
r \ j f *•***• t'j.-Lujdry scanaard should be 0
^ppm; and for welfare atanrfarrf .^ range should be 0>Q6
the
The
17.
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
Ozone
Photochemcal Oxidants
Air Pollution
Health Effects
Welfare Effects
18. DISTRIBUTION STATEMENT
Release to Public
Form 2220-1 (R.v. 4-77) PREVIOUS EDITION is OBSOLETE
b.lDENTIFIERS/OPEN ENDED TERMS
Air Quality Standards
19. SECURITY CLASS (Tins Report,
Unclassified
c. COSATI Field/Group
20 SECURITY CLASS (This pagei
21 NO. OF PAGES
22 PRICE
-------
INSTRUCTIONS
1 REPORT NUMBER
Insert the 1 PA report number as it appears on the cover ot the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title snould indicate clearly and briefly the subject coverage ol the report, and be displayed promiucniK .VI siilniile. i! used. ,n smaller
type or otherwise subordinate U to mam title. When a report is prepared in more than one volume, rcpr.it il:c prmi.irv Mile, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate Hie lusis on which ii «.is selected /<• v . date <>/ mi/c, dale of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doc. J Robert Doc. etc I List author's allilialion il it fillers irr.in the pertofinint! organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign (his number
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels ol an orujiii/juunjl lurearUiv
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report *as ore-pared. Subordmalc numbers cn.i> be indiuud m p.iieniltcscs
11. CONTRACT/GRANT NUMBER
Insert contract or giant number under which report was prepaied.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGtNCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NOTES ,,,„,,.
Enter information not included elsewhere but useful, such as Prepared m cooperation with. Ir.inslation <•!. 1 rc-suitci! .it COHIUCIU.C »i
To be published in, Supersedes, Supplements, etc.
Include a brief (200 words or lessl factual summarv ot the mo>t Mgnihcant mtormaimn contained in ilu upori- II ilie icpoi i loiiums j
significant bibliography or literature survey, mention it here.
17 KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS • Select from the Thesaurus of Engineering and Suentilic lerms ihc proper julluiri/cU terms Hui identity the nuiur
concept of the research and are sufficiently specitic and precise to be used js index entries tor cataloging
(b) 1DENTITIERS AND OPEN-ENDED TERMS - Use identities for project names, code names, equipment designators, etc L se open-
ended terms written in descriptor form for those subjects for which no descriptor exists
(ciCOSATl 1 1ILDGROLP- 1 icld and group assignments are to be taken trom the 1965 ( OSA I I Sulnect ( JUT.TS I ist. Since the ma-
jority of documents are multidisciplinary in nature, the Primary I itld/Gioup assignments w,ll be MXXIIK d,s,,plmc .-tea ol mman
endeavor, or type of physical object. The applications) will be cross-relerenced with sex umlary I leUI '< -r.u.p -issiuiiincnts that « ill K.iinw.
the primary postine(s)
18. DISTRIBUTION STATEMENT ,. .i.u.iiu',,
Denote releasabtlin lo the public or umiiation lor reasons other than sccuritv tor example Relejs,. \ ninmKii. n, .m\ jvni.ini.nv
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to tnc National Technical Information service
Insert the total number ot pages, including this one and unnumbered pages, hut exclude distribution list, n any
22 PRICE
Insert the price set by the National technical Intormation S,r\^e ur the Government I'rmtmg Ullu e n known
EPA Form 2220-1 iRev. 4-77) (Reverse.
------- |