r-Wv^Ui
                       *"?&
                                   ••*>••:< ?2:&W%&%
                                   '?:t.v&$J&&%&%
 .t      . .•.•.   !! ..  ..;.:....    v.
^liM^lr>*::; • '^;:-                       ;^i^lf ^vllK

 ilBll^fiS^^^f^11^^^^
"


 '

-------
                                   DISCLAIMER
      This report has been reviewed by the Office of Air Quality Planning and Standards
(OAQPS), U. S. Environmental Protection Agency (EPA), and approved for publication.
This OAQPS Staff Paper contains the findings and conclusions of the staff of the OAQPS
and does not necessarily represent those of the EPA.  Mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use.

-------
                                  Acknowledgments
      The primary authors of this OAQPS Staff Paper include Dr. David J. McKee (Project
Manager, health effects), Ms. Victoria V. Atwell (welfare effects), Mr. Harvey M Richmond
(exposure and health risk analyses), Mr. Warren P. Freas (air  quality and form of the
standard),  and Ms. Rosalina M. Rodriguez (economic analyses).  This document has been
improved by comments and support from Dr.  Karen M. Martin,  Mr. Eric O. Ginsburg, and
Mr. John H. Haines (all of OAQPS), as well as staff of the Office of Research and
Development (particularly staff of the National Center for Environmental  Assessment and the
National Health and Environmental  Effects Laboratory-Western Ecology Division), the
Office of Policy and Program Evaluation, and the Office of General Counsel, all within the
U.S. Environmental Protection Agency (EPA).  Of particular importance  in the review of
this document has been the technical and editorial support provided by Ms.  Patricia R.
Crabtree and Ms.  Barbara  Miles.
      On three different occasions, draft chapters of this document were formally reviewed
by the Clean Air Scientific Advisory Committee (CASAC).  Helpful comments and
suggestions were also submitted by  a number  of independent scientists and by environmental
and industrial groups.

-------
                                         1

                              TABLE OF CONTENTS


List of Figures	v

List of Tables  	viii

I.     PURPOSE	  1

II.    BACKGROUND	  2
      A.     Legislative Requirements	  2
      B.     History of NAAQS Reviews	  4
             1.     Establishment of NAAQS for Photochemical Oxidants  	  4
             2.     Review and Revision of NAAQS for Photochemical
                   Oxidants  	  4
             3.     Subsequent Review of Ozone NAAQS	  5
             4.     Current Review of Ozone NAAQS  	  7

III.    APPROACH	  9
      A.     Bases for Initial Analytic Assessments	  9
      B.     Organization of Document	   10

IV.  AIR QUALITY CHARACTERIZATION  	12
      A.     Air Quality Trends  	12
      B.     Air Quality Distributions	13
      C.     Ozone Background  	18

V.    SCIENTIFIC AND TECHNICAL BASIS FOR PRIMARY NAAQS	22
      A.     Introduction	22
      B.     Mechanisms of Toxicity  	22
      C.     Health Effects of Ozone  	23
             1.     Pulmonary Function Responses   	24
             2.     Respiratory Symptoms and Effects on Exercise
                   Performance	33
             3.     Increased Airway Responsiveness	34
             4.     Impairment of Host Defenses	36
             5.     Hospital Admissions and Emergency Room Visits	38
             6.     Daily Mortality  	41
             7.     Acute Inflammation and Respiratory Cell Damage	42
             8.     Chronic Respiratory Damage	45
             9.     Genotoxicity and Carcinogenicity  	52
      D.     Factors Modifying Acute Human Response  to Ozone	53
             1.     Exertion and Ventilation	53
             2.     Preexisting Disease	55

-------
            3.     Age, Gender, Ethnic, and Tobacco Smoke Factors   	55
            4.     Interactions with Other Pollutants	56
      E.     Sensitive Population Groups	57
            1.     Active ("Exercising") Individuals	57
            2.     Individuals with Preexisting Respiratory Disease	58
            3.     Other Population Groups  	58
      F.     Adverse Respiratory Effects of Ozone Exposures	59
            1.     Permanent Respiratory Injury and/or Progressive
                   Dysfunction   	60
            2.     Episodic and Incapacitating Illness in Persons with
                   Impaired Respiratory  Systems  	62
            3      Interference with Normal Activity	66
      G.    Ozone Exposure Analysis  	73
            1.     Overview	73
            2.     Exposure Modeling Methodology	77
            3.     Population Exposure Estimates Upon Attainment of
                   Alternative Ozone Standards  	90
            4.     Caxeats and Limitations	   101
      H.    Ozone Health  Risk Assessment	   104
             1.     Overview	104
            2.     Exposure-Response Relationships 	   105
            3.     Benchmark Risk Results	   116
            4.     Population ("Headcount") Risk Results	   117
            5.     Excess Respiratory-Related Hospital Admissions	   122
            6.     Assumptions and Limitations Associated with the
                   Health Risk Assessment	   129
      I.     Alternative Forms of the  Primary NAAQS	   134
             1.     Form of the Current  Standard  	   134
            2.     Issues  Associated  with Consideration of Alternative
                    Forms 	  134
             3.     Alternative NAAQS Statistics  	   136
             4.     Alternative Attainment Test Criteria  .	   141
             5.     Alternatives for Treatment of Missing Values	   141

VI.   STAFF CONCLUSIONS AND  RECOMMENDATIONS ON  PRIMARY
      NAAQS  	144
      A.    Pollutant Indicator	  145
      B.    Averaging Times   	  145
              1.     Short-Term and Prolonged (1 to 8 hours)   	   145
             2.     Long-Term	"	148
      C.    Form of the Standard	  148

-------
                                         Ill

      D.     Level of the Standard	  149
             1.     General Conclusions	  149
             2.     Margin of Safety Considerations Based on Quantitative
                   Exposure and Risk Assessment	  154
      E.     Summary of Staff Recommendations	  164
             1.     Pollutant Indicator 	  164
             2.     Averaging Times	  164
             3.     Form of the Standard  	  165
             4.     Level of the Standard  	  166

VII.   SCIENTIFIC AND TECHNICAL BASIS FOR SECONDARY NAAQS  ....    169
      A.     Introduction	169
      B.     Plant Response/Mode of Action	    170
             1.     Ozone  Uptake	    170
             2.     Extracellular Effects	    171
             3.     Intracellular Effects  	    172
             4.     Resistance and Compensation Mechanisms  	    173
             5.     Physiological Effects	    174
      C.     Environmental Factors Affecting Plant Response  	  177
             1.     Biological Factors  	  178
             2.     Physical Factors	  181
             3.     Chemical Factors	  184
      D.     Ozone Effects on Crops and  Other Vegetation	  187
             1.     Visible Foliar Injury	  189
             2.     Growth/Yield Reductions in Annual Crops  	  193
             3.     Growth Reductions in Tree Seedings and  Mature Trees	  202
             4.     Forest  and Ecosystems Effects	211
      E.     Biologically Relevant  Measures  of Ozone Exposure   	  215
             1.     Biological Considerations	215
             2.     Alternative Forms of the Secondary NAAQS	  223
      F.     Considerations in Characterizing Adverse Welfare Effects  	  227
             1.     Exposure Characterization   	,	227
             2.     Assessment of Risks to Vegetation   	241
             3.     Economic Benefits Assessment	253

VIII.  STAFF CONCLUSIONS AND RECOMMENDATIONS ON SECONDARY
      NAAQS   	277
      A.     Pollutani  Indicator	277
      B.     Averaging Times  	277
      C.     Form of the Standard	279
      D.     Level of the Standard	283
      E.     Recommendation	285

References   	R-l

-------
                                          IV

Appendix  A  Air Quality Assessment	A-l
Appendix  B  Selected Exposure Analysis Results  	  B-I
Appendix  C  Health Risk Assessment:  Selected Results	  C-l
Appendix  D  List of Species in Selected National Parks
                    and Associated Ozone Response	D-l
Appendix  E  Predicted  Yield and Biomass Loss for Crop
             and Tree Seedlings, Based on Growth Regions
             and CIS Predictions of 1990 National Air Quality
             Using the W126 Index  	E-l
Appendix  F  Selected Ambient Ozone Air Quality Distributions
             for NCLAN, Rural  (Class I) and Urban Sites in
             Terms of  Different Exposure Indices	  F-l
Appendix  G  Letters of Closure on the Criteria Document
             and Staff Paper from the Chairman of the
             Clean Air Scientific Advisory Committee to
             the Administrator of the EPA	G-l

-------
                                           V

                                  LIST OF FIGURES
IV-1         Metropolitan area O^ trends adjusted for meteorological
             variability, 1984-93	14

IV-2         Spatial distribution of counties with 1-hour daily maximum,
             1 expected exceedance design values greater than 0.12 ppm
             based on  1991-93 air quality data	15

IV-3         Spatial distribution of counties with 8-hour daily maximum,
             1 expected exceedance design values greater than 0.08 ppm
             based on  1991-93 air quality data	16

IV-4         Spatial distribution of counties with highest 3-month Sum06
             exposure  index values greater than 25 ppm-hours in 1990
             (based on 8:00 am - 8:00 pm LST hours only)	        17

1V-5         Hourly frequency distributions for the maximum 3-month O^ period    ...  19

V-l          Mean predicted changes in  forced expiratory volume in  1 sec following 2-hr
             exposures to ozone  with increasing levels of intermittent exercise	31

V-2A,2B     Predicted mean decrements in forced expiratory volume in 1 sec
             for 1- and 2-hr exposures to ozone with intermittent heavy exercise
             (a) and 6.6-h exposures with moderate prolonged exercise  (b)  	32

V-3          A summary of morphologic lesions found in the terminal bronchioles and the
             centriacinar region of the lung following exposure of laboratory rats to filtered
             air or a simulated ambient pattern of O-^  	47

V-4          Schematic comparison of the duration-response,  profiles for epithelial
             hyperplasia. bronchioloalveolar exudation, and interstitial fibrosis  in
             the centriacinar region of the  king exposed to a  constant low concentration of
             ozone   	48

V-5          Major components of PNEM/O-^ model and associated  health risk assessment
             procedures	74

V-6          New York urban area monitoring sites used in PNEM analyses  	79

-------
                                          VI

V-7          Eight-hour maximum dose exposure distributions for outdoor children exposed
               •-                                                                T
             on one or more days under moderate exertion (EVR 13-27 liters/min/-m^) in
             Philadelphia, PA	92

V-8          Eight-hour maximum dose exposure distributions of total occurrences for
             outdoor children exposure under moderate exertion (EVR 13-27 liters/min/-
             m2) in Philadelphia, PA  	92

V-9          Mean percent of outdoor children exposed on one or more days to ozone levels
              >_ 0.08 ppm for eight hours while engaged in moderate exertion (8-hour
             average EVR in range 13-27 liters/min-m  ) in nine urban areas (alternative  1-
             hour standards)	97

V-10        Mean percent of outdoor children exposed on one or more days to ozone levels
             _>_ 0.08 ppm for eight hours while engaged in moderate exertion (8-hour
             average EVR in range 13-27 liters/min-irr) in nine urban areas (alternative  1-
             expected exceedance 8-hour standards)  	98

V-l 1        Mean percent of outdoor children exposed on one or more days to ozone levels
             _>_ 0.08 ppm for eight hours while engaged in moderate exertion (8-hour
             average EVR m range  13-27 liters/min-m") in nine urban areas (alternative 1-
             and  5-expected exceedance 8-hour standards)  	99

V-12        Steps used to develop probabilistic exposure-response relationships  ....   115

V-l3         Probability that the benchmark response for the  eight-hour, moderate exertion
              health endpoint FEVj decrement >  20%  will be exceeded  five or more times
              in an o^one season  	  118

V-l4         Representative  risk distributions for alternative air quality scenarios
              (FEVj decrements > 10% and  > 20%, Philadelphia, outdoor children, 8 hr
              exposures,  moderate exertion)	  120

V-l5         Representative  risk distributions for alternative air quality scenarios (FEVj
              decrements  >  15%, Philadelphia, outdoor children, 8 hr exposures,  moderate
              exertion)	121

V-l6         Concentration-response  relationship for daily hospital admissions of asthmatics
              in New  York Cit\ area	  128

V-17         Excess annual hospital admissions of asthmatics attributable to
              ozone exposure tor alternative air quality  scenarios	  128

-------
                                         Vll

V-18         Frequency distribution of the average and maximum number of exceedances of
             0.08 ppm 8-hour daily maximum concentrations for sites just attaining an
             average annual second highest daily maximum standard of 0.08 ppm . . .   138

V-19         Frequency distribution of the average and maximum number of exceedances of
             0.08 ppm 8-hour daily maximum concentrations for sites just attaining an
             average annual fifth highest daily maximum standard of 0.08 ppm   ....   139

VI-1         Eight-hour maximum dose  exposure distributions for outdoor children exposed
             on one or more days under moderate exertion (EVR 13-27 liters/min-m^) in
             Philadelphia, PA	159

VI-2         Eight-hour maximum dose  exposure distributions of total occurrences for
             outdoor children exposure  under moderate exertion (EVR 13-27 liters/min-m2)
             in Philadelphia, PA	159

VH-la       Variability in  NCLAN Crop Yield  Sensitivities for 10% Yield Loss  ...   199

VII-Ib       Variability in  NCLAN Crop Yield  Sensitivities for 30% Yield Loss  ...   199

VII-2         Median Crop  Yield Loss from NCLAN Crops   	   200

VII-3         Median Biomass Loss from Seedlings	207

VI1-4         Diagram of the Propagation Pathway of Ozone Effects from Plants to
             Ecosystems	212

VII-5         Diurnal Ozone Patterns at Rural  Sites	222

VII-6a       1991 Rural  Ozone Monitoring Site Locations  	   729

VII-6b       1991 Rural  Ozone Monitoring Site Locations	   230

VJI-7a       1991-1993 Air Quality Relationships Using Alternative Primary
             (1 exceedence)	231

VII-7b       1991-1993 Air Quality Relationships Using Alternative Primary
             (5 exceedcnces)  	  232

-------
                                       Vlll
VII-8         Ozone Frequency Distributions from June to September 1992   	   235

VII-9         The Comparative Ranking of Ozone Monitoring Sites from April to October
             Using the 24 hr W126 Exposure Index (1988-1992)  	   237

VII-10       Estimated 12-Hour SUM06 Ozone Exposure - Max 3-months for 1990 . .   242

Vll-lla      Estimated 12-Hour SUM06 Ozone Exposure - Max 3-months for
             1990 for Current Primary	244

VII-1 Ib      Estimated 12-Hour SUM06 Ozone Exposure - Max 3-months for
             1990 for Alternative Primary (1 exceedence)	  245

VII-llc      Estimated 12-Hour SUM06 Ozone Exposure - Max 3-months for
             1990 for Alternative Primary (5 exceedences)	  246

-------
                                           IX
                                   LIST OF TABLES
V-l          Key health studies supporting the current 1-hr national ambient air quality
             primary standard for ozone  	25

V-2          Key health studies published since the last review of the primary national
             ambient air quality standard  for ozone   	28

V-3          Estimated minute ventilation rates and representative activities associated
             with varying levels of exertion   	54

V-4          Gradation of individual responses to short-term ozone exposure in persons
             with impaired respiratory systems  	64

V-5          Gradation of individual responses to short-term ozone exposure in healthy
             persons   	70

V-6          Characteristics of ozone study areas used in PNEM/O^ analyses	76

V-7          Population estimates for ozone study areas used in PNEM/O^ analyses  . .  80

V-8          Characteristics of human activity studies used in  outdoor worker exposure
             analysis   	83

V-9          Characteristics of human activity studies providing data  for outdoor
             children exposure analysis	84

V-10        Estimates of one-hour maximum dosage exposures experienced by outdoor
             children in Philadelphia during which ozone concentration  exceeded
             0.12 ppm and EVRa equaled or  exceeded 30  liters- min  •  m~^  	94

V-l 1         Estimates of eight-hour maximum dosage exposures experienced  by outdoor
             children in Philadelphia during which ozone concentration  exceeded 0.08
             ppm and EVRa was  in the range 13-27 liters- min" • m   	95

V-!2        Basis for acute health endpoints  addressed by risk assessment  	  106

V-l3        Summary of studies  used in  developing 1-hour exposure-response
             relationships for populations engaged in heavy exertion  	  108

V-14        Summar\  of the studs used to de\elop ! -hour exposure-response relationships
             for populations engaged in modeiate exertion   	  109

-------
V-15         Summary of studies used to develop 8-hour exposure response
             relationships   	110

V-16         Summary of effect estimates for ozone in recent studies of respiratory
             hospital admissions  	113

V-17         Range of median percent of outdoor children responding across nine U.S.
             urban areas upon attaining alternative air quality standards  	   123

V-18         Range of median percent of outdoor children responding across nine U.S.
             urban areas upon attaining alternative air quality  	   124

V-19         Range of median percent of outdoor children responding across nine U.S.
             urban areas upon attaining alternative air quality standards  	   125

V-20         Admissions of New York City asthmatics  	  130

V-21         Criteria  for O^ conducive conditions for the eastern U.S  	   143

VI-1         Percent of outdoor children estimated to experience various health effects  1 or
             more times per year associated with 1-  or 8-hour ozone exposures upon
             attaining alternative standards	  156

VI-2         Admissions of New York City asthmatics - with a comparison relative to
             meeting the current standard	  158

VII-1        Summary of Ozone Exposure Indices Calculated for 3-Month Growing Seasons
              from 1982 to 1991	 188

VI1-2         Comparison  of Exposure-Response Values Calculated Using  the 3-Month. 12-
              Hour SUM06 and W126 Exposure Indices for 54 NCLAN Cases	  196

 VII-3         Exposure-Response Values that Relate Total Biomass (Foliage, Stem, and
              Root) to 12-H SUM06 Exposures!*) Adjusted to 92 Days (ppm-h/year) .  204

 VII-4a       Alternative Standard Scenarios Evaluated Using NHEERL-WED
              CIS for SUM06   	257

 \'ll-4b       Alternative Standard Scenarios Evaluated  Using NHEERL-WED
              CIS for W126  	258

 VIl-5a       RMF Changes in  Economic Surplus for Alternative Primary and
              SUM06 Secondan, Standards (1  Expected Exceedance)  	 261

-------
                                          XI

VII-5b       RMF Changes in Economic Surplus for Alternative Primary and
             SUM06 Secondary Standards (5 Expected Exceedances)	  263

VII-5c       RMF Changes in Economic Surplus for Alternative Primary and
             W126 Secondary Standards  (1 Expected Exceedance)  	  264

VII-5d       RMF Changes in Economic Surplus for Alternative Primary and
             W126 Secondary Standards  (5 Expected Exceedances)	  265

VII-6        Illustration of Incremental Changes in Economic Surplus for Alternative
             Primary and SUM06 Secondary Standards	267

VII-7        Changes in Economic Surpluses for California Crops under Alternative
             Primary and Incremental Secondary Ozone Standards  	  270

VII-8        Summary of Welfare Benefits Estimates Associated with Various Primary and
             Incremental Secondary Regulatory Options	  273

-------
       REVIEW OF THE NATIONAL AMBIENT AIR QUALITY STANDARDS
                                   FOR OZONE
         ASSESSMENT OF SCIENTIFIC  AND TECHNICAL INFORMATION
I.      PURPOSE
       The purpose of this Office of Air Quality Planning and Standards (OAQPS) Staff
Paper is to evaluate the key scientific information contained in the EPA document, "Air
Quality Criteria for Ozone and Related Photochemical Oxidants" (U.S. EPA, 1996a;
henceforth referred to as CD), and identify the critical elements that the EPA staff believes
should be considered in the reviev, of the national ambient air quality standards (NAAQS) for
ozone (O-}).  This Staff Paper includes factors relevant to the evaluation of current primary
(health) and secondary (welfare)  NAAQS. as well as staff conclusions and recommendations
regarding the  most appropriate alternative primary and secondary NAAQS based on current
evaluation of scientific and technical information contained in the CD and this Staff Paper.

-------
II.    BACKGROUND
A.    Legislative Requirements
    Two sections of the Clean Air Act (Act) govern the establishment and revision of
NAAQS.  Section 108 (42 U.S.C. 7408) directs the Administrator to identify pollutants
which "may reasonably be anticipated to endanger public health and welfare" and to issue air
quality criteria for them.  These air quality criteria are  intended to "accurately reflect  the
latest scientific knowledge useful in indicating the kind  and extent of all identifiable effects
on public health  or welfare which may be expected from the presence of [a]  pollutant  in the
ambient air .  ..."
       Section 109 (42 U.S.C.  7409)  directs the  Administrator to propose and promulgate
"primary"  and "secondary" NAAQS for pollutants identified under section  108.  Section
109(b)(l) defines a primary standard as one "the  attainment and maintenance of which, in the
judgment of the  Administrator,  based  on the criteria and allowing an adequate margin of
safety, [is] requisite to protect the public health."   A secondary standard, as defined  in
section 109(b)(2), must "specify a level of air  quality the attainment and  maintenance  of
which, in the judgment of the Administrator, based on  [the] criteria, is requisite  to protect
the public  welfare from any known  or anticipated adverse effects associated with the presence
of [the] pollutant in  the ambient air."   Welfare effects as defined in section 302(h) [42
U.S.C. 7602(h)]  include,  but  are not limited to, "effects on soils, water, crops, vegetation,
manmade materials, animals,  wildlife, weather, visibility and  climate, damage to and
deterioration  of property,  and hazards to transportation, as well as effects on economic
values and on personal comfort and well-being."
       The U.S.  Court of Appeals for the District of Columbia Circuit has held  that the
"margin of safety"  requirement for  primary standards uas intended to address uncertainties
   The legislative history of section 109 indicates that a primary standard is to be set at "the maximum permissible
ambient air level . .  . which will protect the health of any IseiiMtive) group of the population," and that for this
purpo-e "reterein.e vlnn.M he- made to .1 rel'>u-'-.ent.i'i\c vnnple <>f persons <.'»:rpnsing ;he sensiti\e group rather than
to ., .singL- person in suJi ., group  " S. Rep  No  9 I -I 196. 9 1st Cong.. 2d S^s. 10 (1970)  The legislative history
specifically identifies bronchia! asthmatics as A sensitive  group to be protected. Id.

-------
                                            3
associated with inconclusive scientific and  technical information available at the time of
standard setting.  It was also intended to provide a reasonable degree of protection against
hazards that research  has not yet identified.  Lead Industries Association v. EPA. 647 F.2d
1130,  1154 (D.C. Cir.  1980), cert, denied, 101 S. Ct. 621  (1980h American Petroleum
Institute v. Costle. 665 F.2d 1176, 1177 (D.C.  Cir. 1981), cert, denied, 102  S. Ct. 1737
(1982).  Both kinds of uncertainties are components of the risk associated with pollution  at
levels  below  those at  which human health effects can be said to occur with reasonable
scientific certainty. Thus, by selecting  primary standards that provide an adequate margin of
safety, the Administrator is seeking not only to  prevent pollution levels that have been
demonstrated  to be harmful but also to  prevent  lower pollutant levels that she finds may pose
an unacceptable risk of harm, even if the risk is not precisely identified as to nature or
degree.
       In selecting a  margin of safety,  the EPA considers such factors as the nature and
severity of the health effects involved,  the size  of the sensitive population(s) at risk, and  the
kind and degree of the uncertainties that must be addressed. Given that the margin of safety
requirement by definition only comes into play  at levels where there is no conclusive
showing of adverse effects, such factors, which involve  unknown or only partially quantified
risks,  have their  inherent limits as guides to action. The selection of a particular approach to
providing an adequate margin of safety is  a policy choice left specifically to the
Administrator's judgment.  Lead Industries Association v. EPA, supra. 647 F.2d at 1161-62.
        Section 109(d)(l) of the Act (enacted in 1977) requires that "not later than December
31,  1980,  and at 5-year intervals  thereafter, the Administrator'shall complete a thorough
review of the criteria published under section 108 and the national ambient air quality
standards . .  . and shall make such revisions in such criteria and  standards and promulgate
such new standards as  may be appropriate . . .  ."  Section 109(d)(2) requires that an
independent  scientific review committee he appointed and provides that at corresponding
intervals the committee "shall complete a  review of the  criteria .  . . and the national primary
and secondary ambient air quality standards . .  . and shall  recommend to the Administrator
any new . .  . standards and  revisions of existing criteria and standards as  may be appropriate

-------
                                          4
.  . .  ."  Since the early 1980's, this independent review function has been performed by the
Clean Air Scientific Advisory Committee (CASAC) of EPA's Science Advisory Board.
B.     History of NAAOS Reviews
       1.     Establishment of NAAQS for Photochemical Oxidants
       On April 30, 1971, the EPA promulgated NAAQS for photochemical oxidants under
section 109 of the Act (36 FR 8186).  Identical primary and secondary NAAQS were set at
an hourly average of 0.08 parts per million (ppm) total photochemical oxidants not to be
exceeded more than 1  hr per year.  Scientific and  technical bases for these NAAQS were
provided in the document, Air Quality Criteria for Photochemical Oxidants (U.S. DHEW,
1970). The primary standard was based in part on several epidemiology studies (Schoettlin
and Landau, 1961; Motley et al., 1959;  Rokaw and Massey, 1962) conducted in Los
Angeles, which reported a relationship between ambient oxidant levels and aggravation of
respiratory  disease.  The secondary standard was based on evidence of acute and chronic
vegetation injury and physiological effects, including growth alterations, reduced yields, and
changes in  the quality  of plant products (U.S. DHEW, 1970, p. 6-18).
       2.     Review  and Revision of NAAQS for Photochemical Oxidants
       In 1977, the EPA  announced (42 FR 20493) that it was  reviewing the 1970 Criteria
Document in accordance with section  109(d)(l) of the  Act and, in 1978, published a revised
Criteria Document (U.S.  EPA, 1978). Based on the revised Criteria Document, EPA
published proposed revisions to the original NAAQS in 1978 (43 FR  16962) and final
revisions in 1979  (44 FR  8202).  The primary standard was revised from 0.08 ppm to 0.12
ppm; the secondary standard was set identical to the primary standard; the chemical
designation of the standards was changed from photochemical oxidants to O^; and the form
of the standards was revised  from a deterministic  form to a statistical  form, which defined
attainment of the standards as occurring when the  expected number  of days per calendar year
\\ith maximum hourly average concentrations greater than  0.12 ppm is equal to or less than
one.  The revised standards were upheld on judicial appeal. American Petroleum Institute v.
Costle, supra.

-------
                                          5
       3.     Subsequent Review of Ozone NAAQS
       In  1982  (47 FR 11561), the EPA announced plans to revise the  1978 Criteria
Document.  In  1983, the EPA announced (48 FR 38009) that review of primary and
secondary standards for O^ had been  initiated.  The EPA subsequently  provided a number of
opportunities for public review and comment on drafts of the Criteria Document and
associated Staff Paper (U.S. EPA,  1989).  After reviewing the draft Criteria Document in
1985 and  1986, the CASAC sent to the Administrator a "closure letter" outlining key issues
and recommendations indicating that it was satisfied with the final draft of the  1986 Criteria
Document (U.S. EPA, 1986).
       Following closure,  a number of scientific articles and abstracts were published or
accepted for publication that appeared to be of sufficient importance concerning potential
health  and welfare effects of O^  to warrant preparation of a Supplement to the 1986 Criteria
Document (U.S. EPA, 1992).  The CASAC, having already reviewed two drafts of the Staff
Paper  in 1986 and 1987, concluded that sufficient new information existed to recommend
incorporation of relevant new information  into a third draft of the Staff Paper.
       The  CASAC  held a public meeting in 1988 to review a draft Supplement and the third
draft Staff Paper.  Major issues included the definition of adverse health effects of Oy, the
significance of health studies suggesting that exercising individuals exposed for 6 to 8  hours
to O^  levels at or below 0.12 ppin may experience lung inflammation and transient decreases
in pulmonary function; the possibility that chronic irreversible effects may result from long-
term exposures to elevated levels of O^; and the importance of analyses indicating that
agricultural  crop damage may be better defined by a cumulative seasonal average  than by a
 1-hr peak level of O3.  In  its closure letter of 1989 (58 FR 13018), the CASAC indicated
that the draft Supplement and draft Staff Paper "provide an adequate scientific basis for the
EPA to retain  or revise pnmar\  and  secondary standards for ozone."   With regard to the
emerging database on  exposures of 6 hours or more,  CASAC concluded  that such
information could better be considered in  the next review of the ozone NAAQS.
       On October 22. 1991, the American Lung Association (ALA) and  other plaintiffs filed
 suit under section 304 of the Act to compel the  EPA  to complete its rexiew of the ciiteria
 and standards  for Ov   The U.S. District Court for the Eastern District of New York

-------
                                           6
subsequently issued an order requiring the Administrator to sign a Federal Register notice
announcing its proposed decision on whether to revise the standards for Ch by August 1,
1992 and to sign a Federal Register notice announcing EPA's final decision by March 1,
1993.
       On August  10, 1992 (57 FR 35542), the EPA published a proposed decision under
section 109(d)(l) that revisions to the existing primary and secondary standards were not
appropriate at that  time.  The notice explained  (see 57 FR 35546) that the proposed decision
would complete the EPA's review of information on health and welfare effects of O^
assembled over a 7-year period and contained in the  1986 Criteria Document and  its
Supplement.  The notice indicated that the Administrator had not taken into account more
recent studies on the health and welfare effects of O^ because these studies had not been
assessed in the 1986 Criteria Document or its Supplement, nor had they collectively
undergone the rigorous,  integrative review process (including CASAC review) necessary to
incorporate them into a new criteria document. Because that process and other necessary
steps could not, in  EPA's view, be completed in time to meet the March  1993 deadline  for a
final decision, the proposed decision was based on EPA's evaluation of key information
published through early  1989, as contained in the  1986  Criteria Document and its
Supplement; the 1989  Staff Paper assessment of the most relevant information  in these
documents; and the advice and recommendations of the  CASAC as presented both in the
discussion of these documents at public meetings  and in the  CASAC's  1986 and 1989 closure
letters.
       In view of the potential significance of the more recent scientific  papers, as well as
ongoing research on the health and welfare effects of O-^. the August 10, 1992 notice also
announced  the EPA's intention to proceed as rapidly as  possible with the next  review of the
air quality criteria and standards  for O^. Shortly thereafter, the EPA's Environmental
Criteria and Assessment Office (ECAO) formallv initiated action to update the 1986 Criteria
Document and its Supplement (57 FR 38832).
       On March 9, 1993 (58 FR 13008), the EPA published a final decision concluding that
revisions to the current priinan and secondary N'AAQS for  O^ were not  appropriate at that
time.  Given the potential importance of the ne\\  studies and the EPA's continuing concern

-------
                                          7
about the health and welfare effects of 63. the March 9,  1993 notice emphasized the
Administrator's intention to complete the next review of the NAAQS as rapidly as possible
and, if appropriate, to propose revisions of the standards  at the earliest possible date.  The
Administrator subsequently adopted a substantially accelerated schedule for the next review
(59 FR 5164).
       The ALA sought judicial review of the March  1993 decision under section 307(b) of
the Act.  Noting that the Administrator intended to reconsider that decision as rapidly as
possible in light of the more recent scientific information, EPA sought and was subsequently
granted a voluntary remand of ALA's petition for review.
       4.     Current Review of Ozone NAAQS
       As indicated above, ECAO initiated action  to update the air quality criteria document
for O^ in August 1992 (57 FR 38832). A series of peer-review workshops  was held on draft
chapters of the revised  Criteria Document in July  1993 (58 FR 35454) and September  1993
(59 FR 48063). and a first external review draft was made available for CASAC and public
review on January 31,  1994 (59 FR 4278).
       On November 18,  1993, ECAO and OAQPS discussed with CASAC (58 FR 59034)
EPA's accelerated schedule for completing the O^ NAAQS review, formally published  on
February 3, 1994 (59 FR 5164).  In December 1993, OAQPS completed  an Ozone NAAQS
Development Project Plan, which identified key issues to be addressed in  this Staff Paper and
the basis for the initial scientific and  technical assessments planned to address the issues.
OAQPS also met with a subcommittee of the CASAC in  December 1993  (58 FR 59034) and
March 1994 to discuss methodologies used in the exposure and risk assessments summarized
in this Staff Paper.
       The CASAC  reviewed  the first external review draft of the revised Criteria Document
(CD) at a public meeting held on July 20-21. 1994 and made recommendations for revisions.
At a public meeting  held on March 21-22. 1995. the CASAC reviewed a  second external
review draft of the CD and a first external review draft of the basis for the primary standard
contained in  this Staff Paper.  Following  revisions of both the CD and the Staff Paper, an
exiernai  review draft of the entire Staff Paper and Chapter 5  of the CD were reviewed  at a
public meeting held  on September  19-20. 1995.  Following that meeting,  letters were

-------
                                         8
forwarded by the Chairman of CASAC to the Administrator of EPA which came to closure
on the draft CD and on the primary portion of the draft Staff Paper. These letters dated
November 28, 1995 and November 30, 1995, respectively, are reproduced in Appendix G of
this Staff Paper.  Finally, at a public meeting held on March 21,  1996, the majority of the
CASAC members came to closure  on the secondary standard portion of the draft Staff Paper.
The closure letter sent from  the CASAC Chairman to the EPA Administrator dated April 4r
1996 is reproduced in Appendix G  of this Staff Paper.

-------
                                           9
III.    APPROACH
       This Staff Paper is based on the scientific evidence in the CD.  Quantitative
assessments of human exposure and health risks, vegetation exposure, risk, and economic
benefits, and  air quality comparisons provide additional information considered by the EPA
staff in evaluating the appropriateness of revising the current primary and secondary NAAQS
and in assessing potential alternative NAAQS.
       Critical elements are identified in this Staff Paper which  the staff believes should be
considered  in this review of the 0^ NAAQS.  Attention is drawn to judgments that must be
based on careful interpretation of incomplete or uncertain evidence.  In  such  instances, the
Staff Paper provides the staffs evaluation, sets forth alternatives the staff believes should  be
considered, and recommends a course of action.
A.     Bases  for Analytic Assessments
       To meet the accelerated schedule established by the Administrator for this review of
the ozone NAAQS, the OAQPS Ozone NAAQS Development Project Plan identified several
alternative primary and secondary standards to provide a basis for various initial analytic
assessments of air quality, human exposure and health risks, and crop yield loss.   In  so
doing, the staff recognized that additional alternatives  might need to be  analyzed as the
review process continues; e.g., as a result of CASAC and public reviews of the CD and
Staff Paper drafts.
       The Plan identified the following alternative primary standards for use in initial
analytic assessments:
       •      The current 1-hr standard at a level of 0.12 ppm, with a  maximum  expected
              exceedance rate of one per year (averaged over 3 years).
       •      An 8-hr standard in  the range of 0.08-0.10 ppm,  with a maximum expected
              exceedance rate of one per year (averaged over 3 years).
       •      An 8-hr standard m  the range of 0.06-0.08 ppm.  with a maximum expected
              exceedance rate of fi\e per year (averaged  over 3 years).
The following alternative standard  was subsequently added:
       •      An 8-hr standard at a level of 0.07 ppm. with a maximum  expected
              exceedance rate of one per year (a\eraged o\er 3 years).

-------
                                           10
Alternative concentration-based forms (e.g.. the 2nd to 5th highest 8-hr daily maximum
concentration, averaged over 3 years) for various alternative standards have also been
assessed, as discussed in Section V.I.
       Similarly, the Plan identified the following alternative secondary standards for use in
initial analytic assessments:
       •      A standard with a form that is seasonal, cumulative, and peak-weighted.
              Specifically:
                    a SUM06 standard (which sums all hourly Ch concentrations of 0.06
                    ppm and higher over a specified period of time) in the range of 16.5-
                    26.4 ppm-hrs for the maximum 3 calendar-month period.
                    a SUM08 standard (which sums all hourly O-^ concentrations of 0.08
                    ppm and higher over a specified period of time) at a level equivalent in
                    crop protection to the range of SUM06 options.
       •      An  8-hr secondary standard equivalent to any 8-hr primary standard that may
              be established.
Additional seasonal, cumulative, peak-weighted forms that incorporate peak-weighting
functions other than the SUMxx form have also been assessed, as discussed in Section VII of
this Staff Paper.
B.     Organization of Document
       This Staff Paper is organized into  sections as outlined below.  Section IV provides a
summary of air quality trends, air quality distributions, and a characterization  of ozone
background concentrations.
       Section V presents discussions of mechanisms of human  toxicity, factors which
modify responses, effects of concern and  effect levels, populations potentially at risk, and
exposure and  risk analyses.   Staff judgments are  made concerning which effects are
important for the Administrator to consider in  selecting appropriate primary standard(s).
       Section VI discusses factors important in selecting primary standard(s) including
alternative averaging times and forms of the standard. Drawing on these factors and on

-------
                                           11
information contained in Section V, staff conclusions and recommendations are presented for
the Administrator to consider in selecting appropriate primary O^ NAAQS.
       In a similar approach for selecting appropriate secondary standard(s), Section VII of
provides information on mode of vegetation response, factors that modify plant response,
effects on vegetation and natural ecosystems, exposure indices,  and exposure, risk, and
economic benefits assessments.  Based on this information, Section  VIII discusses alternative
forms, averaging times, and levels for the secondary NAAQS, and offers staff conclusions
and recommendations for the Administrator to consider in selecting appropriate secondary O^
NAAQS.

-------
IV.  AIR QUALITY CHARACTERIZATION

       This section provides summaries of O^ air quality trends and the spatial and temporal
distribution of O^  air quality concentrations.  The concept of O^ background is also
presented, together with estimates of background concentrations at ground-level  for various
averaging times.

A.     Air Quality Trends
       States and local air pollution control agencies  measured ground level hourly O^
concentrations at 925 monitoring stations throughout  the nation during 1993.  Most of these
monitoring sites are  located in urban and suburban area locations, with far less frequent
measurement in rural areas.  These  data constitute the ambient data base used in  this staff
paper to assess O-^ air quality trends, as well as to compare selected alternative standards.
       The interpretation of recent O^  trends is difficult due to the large temporal variation
that results from the confounding factors of meteorology and emissions changes.  Peak O^
concentrations typically occur during hot, dry, stagnant summertime conditions.   Thus,
Summer  1988,  as the third hottest summer on record since 1931, was  highly conducive to Oi
formation with  peak O^ levels comparable to those recorded  in the earlier peak year of 1983.
       Meteorological conditions in 1991 and 1993 were also highly conducive to O-j
formation, especially in the eastern  half of the country, although the magnitude and
frequency of exceedances were significantly less than those recorded in 1988. In contrast,
the years  1989  and 1992 saw meteorological conditions that were generally  not as conducive
to O-} formation.  These changes in meteorological conditions have led to large year-to year
differences in peak O^ concentrations.  In response to the National Academy of Sciences
recommendations (NAS, 1991), EPA has developed a statistical model (Cox and  Chu, 1993)
that adjusts for meteorological variability to detect the underlying O-, trend.

-------
                                           13
       Figure IV-1  presents the meteorologically adjusted, and unadjusted, ten year O^
trends in 43 metropolitan areas.  The 99th percentile daily maximum 1-hour concentration
declined 1 percent per year or 12 percent since 1984. The national trend in the composite
mean of the annual second highest daily maximum 1-hour concentration at 509 sites is shown
for comparison, which coincidently,  also declined by 12 percent between 1984 and 1993.
The large year-to-year fluctuations in peak 03 levels introduces a measure of instability  in
nonattainment statistics.  Appendix A contains an expanded discussion of both O-^ air quality
trends,  variability in nonattainment status, and the relationship among alternative averaging
times, air quality statistics, and standards.

B.      Air Quality  Distributions
        This section provides a brief  overview of how both 1-hour and 8-hour O^
concentrations vary across the country  and  among differing monitoring environments.  Figure
IV-2 displays a map of those counties with 1-hour daily maximum, 1 expected exceedance
O^ design values greater than 0.12 ppm based on 1991-93 air quality monitoring data.   The
bar chart to the right of the map indicates the number of people living  in the corresponding
shaded  counties. Figure IV-3 shows the spatial distribution of counties with 8-hour daily
maximum. 1 expected exceedance design values greater than  0.08 ppm, based on 1991-93
data also.  Figure IV-4 depicts those counties with a 3-month SUM06 exposure index value
greater than  25 ppm-hours.  These SUM06 values are based only on the daylight hours, 8:00
am - 8:00 pm Local Standard Time  (LST) in  1990.  In each of these maps,  the county air
quality status was determined by the peak design value site in each county.  For the one
exceedance standard options, the design value is simply the fourth highest concentration
 measured during 1991-93. since if the  fourth highest value is reduced to the level of the
 standard, there will be only  three days above the level of the standard,  or 1 exceedance per
 year.   Similarly, the SUM06 exposure index design  value is simply the index value itself.
 Additional air quality comparisons are presented  in Appendix A,  including an examination of

-------
 Figure IV-1.  Metropolitan area O^ trends adjusted for meteorological variability, 1984-93.
Concentration, ppm
u. 10
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
1
Met Adjusted Trend-43 MSA's
^^x^^\. (99th percentile daily max 1-hr cone.)
...••••.. Unadjusted Ozone Trend-43 MS /V^-v.
.•••'"'. '••. (99th percentile daily max 1-hr cone.)
	 t
National Composite Mean Ozone Trend "'••
(Annual 2nd Daily Max 1-hr)



Actual (43 MSA's) Met Adjusted (43 MSA's) National (509 sites)
(99th Percentile) (99th Percentile) (2nd Daily Max 1-hr)

i i i i i I i
984 1985 1986 1987 1988 1989 1990 1991
-~^~~^



I i
1992 1993

-------
        Figure IV-2.  Spatial distribution of counties with 1-hour daily maximum,  1 expected
                      exceedance design values greater than 0.12 ppm  based on  1991-93 air quality
                      data.
20 -

-------
Figure IV-3.  Spatial distribution of counties with 8-hour daily maximum, 1  expected
             exceedance design values greater than 0.08 ppm based on 1991-93 air quality
             data.

-------
IB"


ir-o


MO


1,10


1PO -


1 10 -
            Figure IV-4.  Spatial distribution of counties with highest 3-month Sum06 exposure index
                         values greater than 25 ppm-hours in 1990 (based on 8:00 am - 8:00 pm LST
                         hours only).

-------
                                            18
alternative levels, forms and interrelationships among alternative standards.  These
comparisons are summarized on a county, metropolitan area,  and nonattainment area basis.
       The staff has also examined the differences in air quality concentration distributions
among differing monitoring  site location environments, particularly for the secondary
standard comparisons.  Figure IV-5 presents histograms of the hourly O3 concentrations for
the peak 3-month summer period at urban and downwind sites in Chicago, a site downwind
of Atlanta, and  a site at higher  elevation in Albuquerque.  All hourly concentrations equal to,
or greater than, 0.06 ppm are displayed with darker shading.   The values of three alternative
exposure indices for alternative secondary' standards have been computed and are displayed
for each site.  There are distinct differences among these sites, with  downwind sites
exhibiting a greater frequency of higher concentrations.

C.     Ozone Background
       Ozone is a naturally occurring, trace constituent of the atmosphere.  There is
controversy regarding how much of ambient O3 monitored at  ground-level is natural and how
much is produced from man-made precursors.  Estimates of the natural  component of O3
van' widely in the  literature,  and there has historically been no standardized terminology
regarding the concept  of 03  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  review  of the available literature, it is obvious that "natural"  O,
background is a  multidimensional and complex concept.  Background O, concentrations vary
by geographic location, altitude and season.   For the purposes of this document,  background
ozone is defined as the ozone concentrations  that would be observed  in the U.S.  in the
 In fact, a survey of the available literature that mentions background ozone or natural ozone background—
approximate!) 50 articles—did not uncover a single rigorous definition of either term!  Even the appellations used for
the concepts van greatly in the relevant literature. Examples include: "baseline ozone." "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.

-------
                                               19

Figure IV-5.  Hourly frequency distributions for the maximum 3-month O3 period.
      Chicago - Cook Co., IL
           MM 3-Month Period
  (SumOe * 32.8; W126*26.4; AOT06*6.S)
 Number of hours
 600
 500
 300
 200
               IL
Chicago • Downwind Kenosha Site
          MurMfent/i Period
  (Sum06m49J; W126-41.4;
Number of hours
600
                                                500
                                                400
                                                300
                                                200
100
   J
     0   .02   04  .06  08   10  12   14  16
      01  03   05  07   09  .11   13  .15  17
    0  .02  04  06  08  1C  12   14  16
     .01  .03   05  .07   09  11   13  15   17
           Atlanta, GA
          Max 3-Month Period
 (Sum06 = 49.8; W126 = 41.S; AOT06 = 12.9;
Number of hou'S
400
300
200
 100
        Albuquerque, NM
          Max 3-Month Period
  (Sum06 = 29.1; W126 = 19.8;  AOT06 = 2.1)
 Number of hours
                                                 400
                                                 300
                                                 200
                                                 100
                                                                I
    C   32  04   06  08   10  .12   14  1C
      01   03  05  07  09  11  13  IS  17
     0   .02  .04  .06  .06   10  12   14  16
      .01   03  05  07  OS   11  13   1J  17

-------
absence of anthropogenic or biogenic emissions of VOCs and NOX in North America.
During the summertime 63 season in the U.S., daily 1-hr, maximum background 63 is
typically between 0.03  to 0.05 ppm.  Part of this background is due to natural sources and
part of it is due to long-range transport of anthropogenic or biogenic emissions.
       The natural component of the background originates from three sources:   stratospheric
O^ which  is transported down to the troposphere, 63 formed from the photochemically-
initiated oxidation of biogenic and geogenic methane and carbon monoxide, and the
photochemically-initiated oxidation of biogenic VOCs.  The magnitude of this natural part
cannot be  precisely determined for two reasons.   First, the part due to long-range transport
of anthropogenic precursor emissions is not known. Second, NOX plays an important role in
the oxidation of methane, carbon monoxide and  the biogenic VOCs and it  is not possible to
determine amount of O^ that  would  have been formed  just due to natural NOX emissions.
However,  some estimates can be made.
       On the basis of  03 data from isolated monitoring sites (CD, Ch. 4), a reasonable
estimate of the O^ background concentration near sea level in the U.S. for an annual average
is 0.020 to 0.035 ppm.  This  estimate includes a 0.005 to 0.015 ppm contribution (averaged
over time) from stratospheric  intrusions into the  troposphere and  a 0.01 ppm contribution
from photochemically-initiated oxidation of methane and carbon monoxide. The remainder is
due to the photochemically-initiated  oxidation of biogenic VOCs and long-range transport.
       Similarly,  a reasonable estimate of the background  O^ concentration near sea level  in
the U.S. for a 1-hour daily maximum during the summer is usually in the  range of 0.03 to
0.05 ppm.  At clean  sites in the Western U.S., the  maximum annual hourly values are in the
range of 0.060 to 0.075 ppm.  Such elevated O^ levels may be occurring at higher altitudes
due to stratospheric O^ intrusion.  Summertime  daily maxima of less than  0.03  are also
observed due to precipitation  sca\engin«.   These estimates are synthesized from the available
literature,  but rely most heavily on Altshuller (1986), Kelly et a!. (1982, 1984), and  Lefohn

-------
                                          21
and Foley (1992).  Based on the diurnal profiles presented for O^ at rural sites in Kelly et al.
(1982, 1984), it is reasonable to estimate that the 8-hour daily maximum 63 during the
summer is also in the range of 0.03 to 0.05 ppm.

-------
                                           22
V.     SCIENTIFIC AND TECHNICAL BASIS FOR PRIMARY NAAQS

A.     Introduction
       This section presents critical information for the review of the primary NAAQS for
O3.  This information includes identification of:  (1) the principal mechanisms of toxicity
which help to establish a link between O3 exposure and resultant health effects; (2) specific
health effects associated with O3 exposure and estimates of lowest observed effects levels; (3)
factors which  may modify  the extent and nature of responses to O3 experienced by
individuals; (4) a qualitative discussion of populations potentially at risk to O3 exposures; and
(5)  which effects may be of public health concern  (i.e., "adverse" effects). Further, this
section presents quantitative estimates of exposure  and risk to  help inform judgments as to
which primary standard(s)  for O3 would protect public health with an  adequate margin of
safety.  Finally, alternative forms of primary standards are discussed.
B.     Mechanisms of Toxicity
       Ozone enters the human body  through the respiratory tract where it reacts with
polyunsaturated fatty acids  (PUFAs),  various electron donors (e.g.,  ascorbate and vitamin E).
and the thiol, aldehyde, and amine groups of low molecular weight  biochemicals and
proteins.  Mechanisms which explain  biochemical and physiological effects of human
exposure to O3 are complex and often involve the direct action of O3 on macromolecules in
the  lungs.  However, they  also can involve the reaction of secondary biochemical products
resulting from the generation of free radical-precursor molecules, the  release  of endogenous
mediators of physiological  response,  and the reactive oxygen intermediates and proteinases
associated with the activities of inflammatory cells that subsequently infiltrate into O3-
damaged lungs.
       One hypothesis, based on the high reactivity of O,. suggests  that O3 does not
penetrate beyond the surface lining fluids of the lungs except in those terminal airway regions
with minimal thickness of the lining, where epithelial  cells might be unprotected by either
mucus or  surfactant (Pryor, 1992).  In a  review  of pathological effects of O-,  (Pryor. 1991).

-------
                                           23
it was suggested that O3-induced cellular damage is more likely to result from reactions of
the more stable, less reactive ozonide, aldehyde,  and hydroperoxide reaction products of O3
with surface-lining fluid components than from direct reaction of O3 with intracellular
components.  A wide variety of lexicological effects has been linked to O3 exposure,
including lung inflammation, effects on host defense mechanisms, morphological effects,
pulmonary function decrements, changes in lung  biochemistry, and genotoxicity and
carcinogenicity.  Although these effects may have different physiological mechanisms, each
effect is initiated by the preliminary interactions of O3 and O3-reaction products with fluids
and epithelial cells in the respiratory tract.
       Mechanisms leading to  O3-induced lung function decrements and symptoms are
probably the best  understood of the mechanisms of O3 toxicity in humans.  The CD (Sec.
7.2.1.1) identifies several such mechanisms, including:  (1) O3 delivery  to the tissue (i.e.,
inhaled concentration of O3, breathing pattern, airway geometry); (2) O3 reactions with the
airway  lining fluid and/or epithelial cell  membranes: (3) local tissue responses, including
injury and inflammation; and (4) stimulation of neural afferents (bronchial C- fibers) and the
resulting reflex responses and symptoms.  The cyclooxygenase inhibitors block production of
prostaglandin E2 (PGEj) and interleukin-6  (IL-6), as well as  reduce lung volume responses;
however, cyclocxygenase inhibitors don't reduce neutrophilic inflammation and levels of cell
damage markers such  as lactate dehydrogenase.  More detailed discussions of biochemical
targets  of O3 interaction and of the mechanisms of acute pulmonary response can be found in
the CD in Chapters 6  and Chapter 7,  respectively.
C.     Health Effects  of Ozone
        The following  discussion  of O3 health effects is presented as a summary of the most
important conclusions and is based on the review and evaluation of health effects research
literature which has been discussed in much greater detail in Chapters 6 through 9 of the
 CD.  This  section of the Staff Paper integrates information from human clinical.
epidemiological, and animal lexicological  studies, as appropriate, within each subsection on
effects.  Furthermore, the effects on healthy individuals and on individuals wilh impaired
 respiratory systems (e.g., asthmatics) are discussed in. the context of acute, prolonged, and.
 finally, chronic exposures within  the following subsections.

-------
                                           24
       A wide array of health effects has been attributed to short-term (1 to 3 hrs),
prolonged (6 to 8 hrs), and long-term (months to years) exposures to O3. Those acute health
effects induced by short-term exposures to O3 concentrations as low as 0.12 ppm, generally
occur while subjects are engaged in heavy (e.g., running) exercise, include:  transient
pulmonary function responses, transient respiratory symptoms and effects on exercise
performance, increased  airway responsiveness, transient pulmonary inflammation, and
increased hospital admissions and emergency room visits for respiratory causes.  Similar
health effects have been observed following  prolonged exposures to O3, at concentrations of
O3 as low as 0.08 ppm and at lower levels of exercise than for short-term exposures.
Although chronic effects such as structural damage to pulmonary tissue and impaired host
defense mechanisms have been established in a substantial number of laboratory animal
studies, there remains little of no evidence of association between ambient O3 exposures and
carcinogenicity and/or genotoxicity at this time.
       Prior to completion of the previous review  of scientific criteria  in 1989, there was a
substantial data base defining the health effects of O3.  Key human and laboratory animal
studies in that database are listed in Table V-l.  Since 1989 numerous  new studies have
greatly expanded  the information on  O3 health effects,  particularly on  prolonged exposures of
6- to 8-hrs.  A selected  group of recent key  human and laboratory animal studies has been
summarized in Table V-2.  Each of these tables  includes only a small  fraction of the total
data base linking  O3 exposure with health effects in humans.  Inclusion criteria for these
studies are mainly the adequacy  of scientific credibility, as determined and  discussed in  the
CD, and the relevance to regulatory  decision making, as determined by staff, the CASAC,
and public review.
       1.      Pulmonary Function Responses
       A variety of pulmonary function responses  has been observed in healthy and impaired
humans acutely exposed to O3.  These responses include reductions in forced vital  capacity
(FVC),  forced expiratory volume in  1 sec (FEV,), and forced expiratory flow at 25 to 75%of
FVC (FEF;5.7<%), which are usually measured by  having an individual exhale forcefully into a
spirometer designed to measure expiratory flow rates.  Another acute  response to O,. airwa\
resistance  (RaJ, is typically  measured in a body  plethysmograph.

-------
                   TABLE V-1. KEY HEALTH STUDIES SUPPORTING THE CURRENT 1-HOUR
                     NATIONAL AMBIENT AIR QUALITY PRIMARY STANDARD FOR OZONE
       03 Concentration,
             ppm
             Health Effect
              Reference
Ambient air containing
0.01 -0.14 daily 1 -hr max over
days  to weeks
Decrements in lung function in children,
adolescents and adults exercising outdoors
Berry et al. (1991)
Socket al. (1985)
Higgins et al. (1990)
Kinney et al. (1989)
Lioy and Dyba (1989)
Lioy et al. (1985)
Lippmann et al. (1983)
Raizenne et al. (1987, 1989)
Spektor et al. (1988a,b; 1991]
>0.12 (1-3 hr) or
^0.08 (6.6 hr)
(chamber exposures)
Decrements in lung function (reduced
ability to take a deep breath), increased
respiratory symptoms (cough, shortness of
breath, pain upon deep inspiration),
increased airway responsiveness and
increased airway inflammation in heavily
exercising adults
Adams et al. (1981)
Avolet al. (1983, 1984)
Devlin et al. (1991)
Folinsbee and Horvath (1986)
Folinsbee et al. (1978, 1984, 1988)
Gibbons and Adams (1984)
Gliner et al. (1983)
Horstman et a\. (1990)
Koren et al. (1989a,b, 1991)
Kulleet al. (1985)
Lauritzen and Adams (1985)
Linnetal. (1980, 1983a,b, 1986, 1988)
McDonnell et al.  (1983, 1991)
Seltzer et al. (1986)

-------
        03 Concentration,
              ppm
              Health Effect
               Reference
2iO 12 (1-3 hr)
(chamber exposures)
Decrements in lung function in heavily
exercising children and adolescents
Avolet al. (1985a,b,c, 1987)
Koeniget al. (1987, 1988)
McDonnell et al. (1985)
>:0 12 (1-3 hr)
(chamber exposures)
Effects are similar in individuals with
preexisting disease except for a greater
increase in airway responsiveness for
asthmatic and allergic subjects
Koenig et al. (1985, 1987, 1988)
Kreit et al. (1989)
McDonnell et al. (1987)
>_0.12 (1-3 hr)
(chamber exposures)
Older subjects (>50 yr old! have smaller
and less reproducible changes in lung
function
Bed! and Horvath (1987)
Bedi et al. (1988,  1989)
Drechsler-Parks et al. (1987,  1989, 1990)
Reisenauer et al. (1988)
> 0.18 (1-3 hr)
(chamber exposures)
Reduced exercise performance in heavily
exercising adults
Adams and Schelegle (1983)
Folinsbee et al. (1984)
Gong et al. (1986)
Under et al. (1988)
Schelegle and Adams (1986)
>_ 0.12 (1-3 hr)
(chamber exposures)
Attenuation of lung function response with
repeated exposure
Avol et al. (1988)
Farrell et al. (1979)
Hackney et al. (1976, 1989)
Horvath et al. (1981)
Kulle et al. (1982)
Linn et al. (1982, 1988)

-------
        03 Concentration,
              ppm
              Health Effect
               Reference
 >0.12 with chronic, repeated
exposure (chamber exposures)
Changes in lung structure, function, and
biochemistry in laboratory animals that
are indicative of airway irritation and
inflammation with possible development
of chronic lung disease
Amdur et al. (1978)
Barry et al. (1983, 1985, 1988)
Boorman et al. (1980)
Castleman et al. (1977,  1980)
Chowet al. (1981)
Costa et al. (1983)
Crapo et al. (1984)
Eustis et al. (1981)
Filipowicz and McCauley (1986a,b)
Fujinaka et al. (1985)
Grose et al. (1989)
Lastetal. (1979,1984)
Moore and Schwartz  (1981)
Mustafa et al. (1985)
Plopper et al.  (1979)
Rao et al. (1985a,b)
Schwartz et al. (1976)
Sherwin and Richters (1985)
Tyler et  al. (1988)
Wegner  (1982)
Wright et al. (1988)
_>_ 0.08 (3 hr| or
_>_ 0 10 with chronic repeated
exposure (chamber exposures)
Increased susceptibility to bacterial
respiratory infections in laboratory animals
Coffin et al. (1972)
Ehrlich et al. (1977)
Miller et al. (1978)
Aranyi  et al. (1983)

-------
       TABLE V-2.  KEY HEALTH STUDIES PUBLISHED SINCE THE LAST REVIEW OF THE PRIMARY
                         NATIONAL AMBIENT AIR QUALITY STANDARD FOR OZONE
      03 Concentration,
            ppm
                Health Effect
                 Reference
Ambient air containing 0.01-
0.1 54 daily 1 -hr max over
days to weeks
Decrements in lung function (FEV,) in children,
adolescents, and adults exposed to 03 outdoors
Avolet al. (1990, 1991)
Braun-Fahrlander et al. (1994)
Castillejos et al. (1992)
Hoeket al. (1993a,b)
Kilburnet al. (1992)
Krzyzanowski et al. (1992)
Lebowitz  et al. (1991)
Raizenne et al. (1987, 1989)
Raizenne and Spengler (1989)
Schmitzberger et al. (1993)
Schwartz et al. (1994a,b,c)
Spektor et al. (1991)
Spektor and Lippmann (1991)
Stern et al. (1989,  1994)
Thurston  et al. (1995)
                             Exacerbation of respiratory symptoms (e.g.,
                             cough, chest pain) in individuals with preexisting
                             disease (e.g., asthma) with low ambient
                             exposure, decreased temperature, and other
                             environmental factors resulting in increased
                             summertime hospital admissions and emergency
                             department visits for respiratory causes.
                                              Burnett et al. (1994)
                                              Cody et al. (1992)
                                              Thurston et al. (1992,
                                              Weisel et al. (1995)
                                              White et al. (1994)
                                               00
                    1994, 1995)
3:0.12 (1-3 hr)
s:0 08 (6.6 hr)
(chamber exposures)
Decrements in lung function
(reduced ability to take a
deep breath), increased
respiratory symptoms (cough,
shortness of breath, pain
upon deep inspiration),
increased airway responsive-
ness and increased airway
inflammation in exercising adults
Devlin et al. (1991, 1990)
Folinsbee et al. (1991,  1994,  1995)
Frampton et al. (1993)
Gross et al. (1991)
Hazucha et al. (1992, 1987)
Horstman et al. (1990, 1995)
Koren et al. (1991)
McKittrick et al. (1995)
McDonnell et al. (1991)
Linn et al.  (1994)
>0 12 (1-3 hr)
>0 08 (6.6 hr)
(chamber exposures)
Effects are similar in individuals with preexisting
disease except for a greater increase in airway
responsiveness for asthmatic and allergic
subjects
Koenig et al. (1988)
Kreit et al. (1989)

-------
      O3 Concentration,
            ppm
                 Health Effect
                  Reference
>0.12 (1-3 hr)
>0.08 (6.6 hr)
(chamber exposures)
Older subjects (>50 yr old) have smaller and
less reproducible changes in lung function.
Drechsler-Parks et al. (1990)
Horvath et al. (1991)
Seal et al. (1993, 1994)
                              Attenuation of response with repeated exposure
                                               Hackney et al. (1989)
                                               van Bree et al. (1994)
>0.12 with prolonged,
repeated exposure
(chamber exposures)
Changes in lung structure, function, elasticity,
and biochemistry in laboratory animals that are
indicative of airway irritation and inflammation
with possible development of chronic lung
disease
Catalano et al. (1995a,b)
Chang et al., (1991, 1992, 1995)
Costa et al., (1995)
Harkemaetal. (1989, 1993, 1994)
Harkema and Mauderly (1994)
Hiroshima et al. (1989)
Hotchkiss et al. (1989a,b)
Hyde et al. (1992)
Last et al. (1993a,b, 1994)
National Toxicology Program/Health Effects
Institute (1995)
Parks and Roby (1994)
Pinkerton et al.  (1992,  1993, 1995)
Pino et al. (1992a,b,&c)
Plopperet al. (1991, 1994a,b)
Radhakrishnamurthy (1994)
Schultheis et al. (1991)
S?arek (1994)
Tan et al.  (1992)
Tepper et al. (1994, 1995)
Tyler et al. (1991)
Van Bree et al.  (1992)
                                                                                                                             ho
                                                                                                                             vo
                              Increased susceptibility to bacterial respiratory
                              infections in laboratory animals
                                               Gilmour et al. (1993a,b)
                                               Jakab and Bassett (1990)
                                               Jakab et al. (1995)
                                               Belgrade et al. (1990)

-------
                                           30
       Neurogenic inhibition of maximal inspiration, possibly caused by stimulation of C-
fiber afferents, is believed to be the cause of decreased FVC and inspiratory capacity in
humans (CD,  Sec. 9.3.1.1).  Although a large body of evidence generally suggests that
healthy individuals and those with impaired respiratory systems have similar functional
responses, one recent study (Horstman et al.  (1995)), not yet replicated, reported that
asthmatics had a greater change in lung function than healthy individuals.
       The strongest and most quantifiable exposure-response data on pulmonary function
responses to O3 have come from controlled human exposure studies.  The magnitude and
time course of spirometry responses to O3 depend upon the O3  concentration (C), the
"exercise" level (minute ventilation, Vg), and the duration of exposure (T). One of the best
demonstrations of the impact of various "exercise" levels and O3 concentrations on group
mean FEV, following 2-hr exposures is summarized by Hazucha (1987) in Figure V-l (CD,
Figure 9-1).  This figure clearly shows that FEV, decrements are enhanced by increased
levels of "exercise" and/or increased levels of O3 exposure.
       In experimental studies, increased ventilation rates are brought about by having the
subjects engage in activities typically identified as "exercise."   This  exercise is meant to
simulate  any type of activity involving exertion that increases the ventilation rate.  Thus,
while experimental studies typically report "exercise" levels, the broader term  "exertion" will
be used throughout this Staff Paper when referring to the types of normal activities in which
people engage that result in similar increased ventilation rates.   The staff intends that the
term exertion  be understood to  encompass a much broader class of activities than is typically
associated with the term exercise, as discussed in Section V.D.I of  this Staff Paper.
       Numerous experimental studies of exercising adults have demonstrated decrements in
lung function for  exposures of 1-3 hrs at _>_0.12 ppm O3 (Adams et al., 1981;  Avol et al.,
1983, 1984; Folinsbee and Horvath,  1986; Folinsbee et al., 1978, 1984.  1988; Gibbons and
Adams, 1984; Gliner et al..  1983:  Kulle  ei al., 1985; Lauritzen and Adams. 1985; Linn et
al., 1980, 1983, 1986, 1988; McDonnell et al.,  1983; Seltzer et al.,  1986) and tor exposures
of 6.6 hrs at  >:0.08 ppm O3 (Folinsbee et al.. 1988, 1991,  1994; Hazucha et al.,  1992:
Horstman et al., 1990, 1995; Horvath et a!..  1991; Keren et al., 1988.  1989. 1991; Linn et

-------
                                          31
                110
                 60
                                                                        0.8
                                0.2          QA          0.6
                                    Ozone Concentration (ppm)
Figure V-l. Mean predicted changes in forced expiratory  volume in  1 sec following 2-hr
            exposures to ozone with increasing levels of intermittent exercise.
Source:  Hazucha (1987); CD, Figure 9-1, p. 9-16.
ah, 1994; McDonnell et al., 1991). These studies provide conclusive evidence that O3
levels commonly monitored in the ambient air induce FEV, decrements  in exercising adults.
       For short-term exposures of 1 to 2 hr, subjects exposed to higher O3 concentrations
(e.g., > 0.25 ppm) during intermittent heavy exertion tend to experience rapid responses
indicative of a plateau (See Figure V-2.A). In contrast,  lower O3 concentrations with lighter
exertion tend to induce responses which progress slowly and may not reach a plateau during
the period of exposure.  McDonnell and Smith (1994) plotted predicted  mean decrements in
FEV, vs.  time, with intermittent moderate exertion during a 6.6 hr exposure; they found no
response plateaus at 0.08, 0.10,  or 0.12 ppm O3 during the first 3 hr but did show plateaus
developing at each concentration during the latter portion of the exposure (See Figure V-2B).
       Summer camp studies have provided the most extensive and reliable data base on
acute lung function responses to ambient O3 and other pollutants in children and adolescents
living in the northeastern U.S. (Bock et al., 1985;  Spektor et al., 1988a,b, 1991; Spektor and
Lippmann,  1991; Lippmann ei ai., 1983; Lioy et al., 1985; Lioy and Dyba, 1989; Kmnev et
al., 1989; Berry et al., 1991; Thurston et al., 1995), southern California (Higgins et al..
1990; Avol  et al., 1990,  1991),  and southeastern Canada (Raizenne et al., 1987,  1989;

-------
                                  32
-5
 O.CXi
               0.10   0.20   0.30    0.40   0.50       012345
                  Ozone (ppm)                             Time (hours)
Figure V-2. Predicted mean decrements in forced expiratory volume in 1 s for 1- and 2-h
            exposures to ozone with intermittent heavy exercise  (A) and 6.6-h exposures
            with moderate prolonged exercise (B).
Source:  McDonnell and Smith (1994); CD, Figure 9-2, p.  9-17.
Raizenne and Spengler,  1989).  Lung function changes reported at low O3 concentrations are
comparable to those reported in children and  adults exposed under controlled experimental
conditions, although direct comparisons are difficult to make because of differences in
experimental design and analytical approach.  Even though exposures at the summer camps
occurred over periods of many hours to days, a key calculation  made for many of the studies
is the slope of the relationship between  FEV, and the  O3 concentration measured during the
previous hour, without consideration of the background levels.  The average slope from six
of the camp studies (Spektor et  al., 1988a, 1991;  Spektor and Lippmann,  1991;  Raizenne et
ah, 1987, 1989; Higgins et ah,  1990; Avol et ah, 1990, 1991) was  -0.50 mh/ppb O3, within
a concentration range  of 0.01 to 0.16 ppm (CD, Sec.  7.4.1.2).  The slope corresponds to a
decrease in FEV, of 60  mh at 0.12 ppm from a base  level of 2000 to 2500 mL  or roughly a
2.4 to 3.0% decrease  in FEV,.  This is comparable to the 3.4% decrease in FEV, reported
by McDonnell et al. (1985) for boys (8 to 11 years old) exposed to  0.12 ppm O3 during   .
heavy exercise under controlled experimental conditions.  Although  outdoor studies (Spektor
et ah, 1988b;  Selwyn et ah, 1985; Brunekreef et  ah,  1994) of exercising adults have shown
similar associations between spirometric changes  and  increasing O3 concentrations (CD.

-------
                                           33
9.3.1), "daily life studies" (Kinney et al.,  1989; Castillejos et al.,  1992; Hoek et al., 1993b;
Krzyzanowski et al., 1989) are difficult to interpret due to the role of seasonal factors (e.g.,
pollens, epidemics of respiratory infection, changes in activity patterns) and the
preponderance of time  spent indoors by subjects (CD, Sec. 7.4.1.2).
       2.  Respiratory  Symptoms and Effects on Exercise Performance
       Various human  respiratory symptoms, including cough, throat irritation, chest pain on
deep inspiration, nausea, and shortness of breath, have been  induced by O3 exposures of
healthy and impaired individuals. As is the case for spirometric lung function decrements,
O3- exposure data do not support enhanced sensitivity  to symptoms of individuals with
asthma.  Although eye irritation is a symptom commonly associated with exposure to ambient
oxidant mixtures, which include such oxidants as  O3 and peroxyacyl  nitrates, controlled
human exposure studies of O3 have demonstrated  that  at concentrations reported in the
ambient air, O3 alone does not induce eye irritation.
       A potential linkage between  changes in spirometry and at least one symptom  may be
explained in part by the mechanism which induces cough.  The receptors responsible for
cough may be unmyelinated C-fibers or rapidly adapting receptors located in the larynx and
the largest conducting  airways (CD, Sec. 9.3.1.1).  Field and epidemiology studies (Ostro et
al., 1993; Krupnick et al., 1990) which have reported spirometry changes associated with
ambient O3 levels also  have indicated associations between hourly  or daily ambient O3  levels
and presence  of symptoms such as cough, particularly in asthmatic children.
       Respiratory  symptom responses to O3 exposure follow a monotonic exposure-response
relationship that has a  similar form to that for spirometry responses.  Increasing exposure
levels elicit increasingly more severe symptoms that persist for longer periods.  Symptom
and spirometry responses follow a similar time course during an acute exposure and the
subsequent recovery, as well as over the course of several days in a repeated exposure  study.
Furthermore, medication interventions  that block  or reduce spirometry responses have a
similar effect on symptom responses.  As with spirometry responses, symptom responses
vary  considerably among  subjects, although  the individual correlations between spirometry
and symptom responses are relatively low.

-------
                                           34
       Ozone-induced interference with exercise performance, either by reducing maximal
sustainable levels of activity or reducing the duration of activity that can be tolerated at a
particular work level, is likely related to symptoms.  In several heavy or severe exercise
studies (Schelegle and Adams, 1986; Gong et al., 1986; Adams and Schelegle, 1983) of
athletes exposed to O3, the discomfort associated with the respiratory symptoms caused by
O3 concentrations in excess of 0.18 ppm was of sufficient severity that the athletes reported
that they would have been  unable to perform maximally if the conditions of the exposure
were present during  athletic competition.  In workers or active people exposed to ambient
O3, respiratory symptoms may cause reduced productivity or may curb the ability or desire to
engage in normal activities.
       3.      Increased Airway Responsiveness
       Increased airway responsiveness is an indication that the airways are predisposed to
bronchoconstriction which  can be induced by a wide variety of external stimuli (e.g.,
pollens, dust,  cold air, SO2, etc.).  A high level of bronchial responsiveness  is characteristic
of asthma (CD, 7.2.3). Ozone exposure causes increased responsiveness of the pulmonary
airways to subsequent challenge with bronchoconstrictor drugs such as histamine or
methacholine. Airway responsiveness is usually measured by having an individual exhale
forcefully into a spirometer designed to measure expiratory flow rates (e.g.,  FEV,) or by
measuring ainvay resistance (R,,w) in a body plethysmograph.   Measurements of FEV, are
taken before and after small amounts of an aerosolized bronchoconstrictor are administered.
and the dose is increased until a predetermined  degree  of airway response has been
measured. The provocative dose that produced a 20%  drop in FEV, would be referred to as
"PD20" and  the provocative dose that produced  a 100% increase in Raw would be referred to
as the "PD100."
       Increased airway responsiveness  is seen  even after recovery from spirometric
changes, but this effect typically disappears after 24 hrs (CD. Sec. 9.3.1.3).  Although
changes in airway responsiveness tend to resolve somewhat more slowly than spirometric
changes and appear to be less likely to attenuate with repeated exposure, the evidence for a
persistent  increase in  responsiveness from animal  studies is inconsistent.  Changes in airway
responsiveness in rats and guinea pigs tend to occur at  higher O3 concentrations and, as in

-------
                                           35
humans, tend to be most pronounced shortly after the exposure and less so 24-hr
postexposure.  Changes in airway responsiveness appear to occur independently of changes in
pulmonary function.  This response does not appear to be due to airway inflammation (at
least the influx of polymorphonuclear leukocytes [PMN's]  into the airways) or to the release
of arachidonic acid metabolites, but it may be due to epithelial damage and the consequent
increased access of these chemicals to smooth muscle in the airways or to the receptors in the
airways responsible for reflex bronchoconstriction. The clinical relevance  of this observation
is that, after O3 exposure,  human airways may be more susceptible to  a variety of stimuli,
including antigens, chemicals, and particles.
       Healthy subjects have experienced small increases in nonspecific bronchial
responsiveness, which resolve within 24 hrs, after being exposed to O3 concentrations as  low
as 0.20 ppm for  1 hr (Gong et al, 1986) and 0.08 to 0.12  ppm for 6.6 hr (Horstman et al.,
1990; Folinsbee et al.,  1988).  Asthmatic subjects typically have increased airway
responsiveness at baseline, and differences in baseline bronchial responsiveness between
healthy individuals and sensitive asthmatics may be as much  as 100-fold.   Changes induced
by O3 exposure,  however, are usually only  2- to 4-fold. Only one published study (Molfino
et al.,  1991) suggested  an O3-induced increase in specific (i.e., allergen-induced) airway
reactivity.  This effect was reported after a 1-hr resting exposure of atopic  asthmatics to 0.12
ppm O3, and thus provided a plausible linkage between ambient O3 exposure and increased
hospital admissions.  However, the study had experimental design flaws and has not yet been
replicated.  With such a limited and uncertain data base on O3-induced airway
responsiveness,  it appears to be premature to draw conclusions regarding this health endpoint
at this point in time.
        Ongoing studies of O3-induced increases in airway  responsiveness will need to be
evaluated in order to determine the exposure-response relationship for alterations in responses
to inhaled antigens,  especially  with regard to sensitive asthmatics.  Enhanced  response to
antigens in asthmatics could lead to increased morbidity (i.e., medical treatment, emergency
room visits, hospital admissions) or to more persistent alterations  in airway responsiveness
 (CD, 9.3.1.3).

-------
                                           36
       4.      Impairment of Host Defenses
       As discussed below, the mammalian respiratory tract has numerous closely integrated
defense mechanisms that provide protection from the adverse effects of a wide variety of
inhaled particles and microbes, if they function normally.  However, when these defense
mechanisms break down or are impaired by O3, there can be an increase in susceptibility to
respiratory infection and related respiratory disfunction (CD, Sec. 9.3.3.2).
       Mucociliary Clearance of Inhaled Particles.  Impaired mucociliary clearance can result
in unwanted accumulation of cellular secretions and increased numbers of particles and
microorganisms in the lung, leading to increased risk  of respiratory infection and bronchitis.
Animal studies show that clearance of inhaled insoluble particles is  slowed after acute
exposure to O3.  Ozone-induced damage to cilia and increased mucus secretion  likely
contribute to a slowing of mucociliary transport rates.  In one study investigating the effects
of longer-term alveolarbronchiolar clearance, Pinkerton et a). (1993) exposed rats to an urban
pattern of O3 (continuous 0.06 ppm, 7 days/week with a slow rise to a peak of 0.25  ppm and
subsequent decrease to 0.06 ppm over a 9-hr period for 5 days/week) for 6 weeks.  The rats
were exposed  3 days later to asbestos, which can cause pulmonary fibrosis and tumor
formation.  Although O3 did not affect the deposition  of asbestos at the site of maximal
deposition of both O3 and asbestos, thirty days later the lungs of the O3-exposed animals had
twice the number and  mass of asbestos  fibers as the air-exposed rats (CD, Sec. 6.2.3.3).  In
general, however, the  CD  (Sec. 9.3.3.2) notes that retarded mucociliary clearance is not
observed in animals exposed repeatedly to O3.
       The effects of O3 on mucociliary clearance in humans have not been well studied, and
the results are somewhat conflicting.  One study (Foster et al.,  1987) reports an O3-induced
increase in particle clearance in  subjects exposed to 0.4 ppm O3 for 2 h,  while  another study
(Gerrity et al., 1993) reports no O3-induced change in particle clearance with a similar
exposure regimen.  The discrepancy between these two studies  may be explained by
differences in  exposure protocol, time of particle inhalation, or time of clearance
measurement,  or by the presence of cough immediately following O3 exposure, which may
have accelerated clearance  in the first study (CD, Sec. 7.2.4.7).

-------
                                           37
       Alveolar Macrophage Function.  Macrophages represent the first line of defense
against inhaled microorganisms and particles that reach the lower airways and alveoli (CD,
Sec. 9.3.3.2).  Studies in both humans and animals have shown that there is an immediate
decrease in the number of macrophages following O3 exposure.  Alveolar macrophages also
have been shown to be crucial to the clearance of certain gram-positive bacteria from the
lung. Several studies in  both humans and laboratory animals also have shown that O3
impairs the phagocytic capacity of alveolar macrophages, and some studies suggest that mice
may be more impaired than rats (Gilmour and Selgrade, 1993; Costing et al., 1991a). The
production of superoxide anion (an oxygen radical  used  in bacterial killing) by alveolar
macrophages also is depressed in both humans and animals (Ryer-Powder et al., 1988;
Costing et al.,  199Ib) exposed to O3, and the ability of alveolar macrophages to kill bacteria
directly is impaired.  Decrements in alveolar macrophage function have been observed in
moderately exercising humans exposed to the lowest concentration tested, 0.08 ppm O3 for
6.6 hrs (Devlin et al.,  1991).
       Interaction with Infectious Agents.  Concern about  the effect of O3 on susceptibility to
respiratory infection derives primarily from animal studies in which  O3-exposed mice die
following a subsequent challenge with aerosolized bacteria (CD, Sec. 9.3.3.2).  Increased
mortality of experimental laboratory animals has been shown to be concentration-dependent,
and exposure to as little as 0.08 ppm O, for 3 hours (Coffin et al., 1967; Coffin and
Gardner, 1972; Miller et al., 1978) can  increase mortality of mice to a subsequent challenge
with streptococcus bacteria. In addition, younger mice are more susceptible to infection than
older mice (Gilmour et al., 1991, 1993a,b; Miller et al,  1978); this has been related to
increased PGE, production in these animals, which likely decreases alveolar  macrophage
activity.
       It has been suggested that impaired alveolar macrophage function is the mechanism
likely responsible for enhanced susceptibility to bacteria.  However,  mortality is not observed
with other rodent species, raising the question of whether this phenomenon is restricted to
mice.  Although both  mice and rats show impaired macrophage killing of inhaled  bacteria
following O3 exposure, rats mount a faster PMN response  to O, to compensate for the deficit
in alveolar macrophage function.  The resulting slower clearance time in mice allows the

-------
                                           38
streptococcus strain to persist in lung tissue and, subsequently, to elaborate a number of
virulence factors that evade secondary host defense and lead to bacterial multiplication and
death of the host.  Although increased mortality in laboratory animals is not directly relevant
to humans, laboratory animals and humans share many host defense mechanisms being
measured by mortality in the mouse  model.  Thus, this category of effect (i.e., decrement in
antibacterial defenses) can be qualitatively extrapolated to humans (CD, 9.3.3.2).
       With regard to antiviral defenses, a study of experimental rhinovirus infection in
susceptible volunteers failed to show any effect of 5 consecutive days of 03 exposure (0.3
ppm, 6 hrs/day) on the clinical outcome or on host response (Henderson et al.,  1993).
Studies in which O3-exposed mice were challenged with influenza virus report conflicting
results: some studies show increased mortality, some show decreased mortality, and still
others show no change at all.  However, even when increased mortality was demonstrated,
there was no difference in viral liters in the lung, suggesting virus-specific immune functions
were not altered.  One animal study  (Jakab and Bassett, 1990) reported that even though a
120-day exposure to 0.5 ppm O3 did not affect the acute course of a  viral infection from
influenza virus administered immediately before O3 exposure began,  it did enhance
postinfluenzal alveolitis and lung parenchymal changes.
       Although there is no single experimental human or animal study or group of studies
which proves that respiratory infection is worsened by exposure  to O3,  taken as a whole,  the
data suggest that acute O3 exposures  can impair the host defense capability of both  humans
and animals, primarily by depressing alveolar  macrophage function and perhaps also by
decreasing  mucociliary clearance of inhaled particles and microorganisms.  This suggests that
humans exposed to  O3 may be predisposed to bacterial infections in the lower respiratory
tract.  The  seriousness of such infections may  depend on how quickly bacteria develop
virulence factors and  how rapidly neutrophils are mobilized to compensate for the deficit  in
alveolar macrophage function (CD, Sec. 9.3.3.2).
       5.      Hospital Admissions and Emergency Room Visits
       People with  preexisting pulmonary disease may be at increased risk  to responses
associated with short-term O3 exposures leading to increased hospital  admissions  and
emergency  room visits. Furthermore, some  individuals with pulmonary disease may have an

-------
                                          39
inherently greater sensitivity to O3 (CD, Sec. 9.3.2). Asthmatics characteristically have
greater baseline bronchial responsiveness, but,  depending on the severity of their disease and
clinical status, their FEVj can be within the normal  range (100 ±  20% predicted) or may be
less than 50% predicted.  Patients with chronic obstructive pulmonary disease (COPD) can
have FEVs ranging from 30 to 80% of predicted, again depending on disease severity.
       Because of their depressed functional state, small absolute changes in lung function of
individuals with preexisting pulmonary disease have a larger relative impact than for healthy
individuals.  For example, a 500-mL FEV, decrease in a healthy young man with an FEV, of
4,000 mL causes only a 12% decline.  In a 55-year-old COPD patient with an FEV, that is
50% of predicted,  or about 1,670 mL, a 500-mL decline in FEV,  would result in a 30%
decline in FEV,.  Asthmatics with depressed baseline function would have similarly
magnified relative  responses and, because of increased bronchial responsiveness, may also
experience larger changes in airway resistance.  Evaluating the intersection of risk factors
and exposures is more  complex.  However, an individual with more severe  lung disease is
unlikely to engage in heavy exertion and, thus, would be less likely  to encounter an effective
exposure to O3.
       About  12 million people in the United States (approximately 5% of the population)
are estimated to have asthma (National Institutes of  Health, 1991). The prevalence is higher
among African Americans, older (8- to  11-year-old) children, and urban residents.  The
annual incidence of hospitalization for all asthmatic  individuals is  estimated to be about 45
per 1000 (National Institutes of Health,  1991).  Although death due  to asthma is a relatively
infrequent event (i.e., on an annual basis,  about  one death occurs  per 10,000 asthmatic
individuals), over 4000 deaths are attributed to asthma each year.  Mortality rates are higher
among males and are at least 100% higher among nonwhites. In  two large urban centers
(New York and Chicago), mortality rates from asthma among nonwhites may exceed  the city
average by up to  fivefold (Sly, 1988;  National Institutes of Health. 1991; Weiss and
Wagener, 1990; Carr et al., 1992). Although  some innercity areas may have lower O3
concentrations than some suburban areas, O3 concentrations are much higher than those in
most rural areas.   The impact of ambient O, on  asthma morbidity and mortality in this
apparently susceptible population is not well understood.  Those epidemiological studies

-------
                                           40
which have been conducted to date are subject to confounding factors and have rarely
focused on innercity nonwhite asthmatics.  Furthermore, controlled human exposure studies
of asthmatics typically include mild to moderate asthmatics and also have not dealt
specifically with nonwhite asthmatics.
       A number of epidemiological studies have shown a consistent relationship between
ambient oxidant exposure and acute respiratory morbidity in the population.  Decreased lung
function and increased respiratory symptoms, including exacerbation of asthma,  occur with
increasing  ambient O3, especially in children.  Modifying factors, such as ambient
temperature, aeroallergens, and  other copollutants (e.g., panicles) also can contribute to this
relationship. Ozone air pollution can account for a portion of summertime hospital
admissions and emergency room visits for respiratory causes.  Studies conducted in various
locations in the eastern United States (Cody et al.,  1992; Thurston et al., 1992,  1994; White
et al., 1994; Schwartz 1994a,b,c) and Canada (Bates et al.,  1990; Lipfert and Hammerstrom,
1992; Burnett et al., 1994; Delfino et al., 1994a,b) consistently have shown a relationship
with increased incidence of visits and admissions, even after controlling  for modifying
factors, as  well  as when  considering only concentrations <0.12 ppm O3. It has been
estimated from these studies that O3 may account for roughly one to three excess
summertime respiratory hospital admissions per hundred parts per billion O3, per million
persons.  In Section V-H on ozone health risk assessment, Figure V-17 summarizes the
excess annual  hospital admissions of asthmatics attributable to O3 exposure for alternative air
quality scenarios and provides "effect size" and "relative risk" estimates.
       The association between  elevated ambient O3 concentrations  during the summer
months and increased hospital visits and admissions for respiratory causes has a plausible
biologic basis  in  the physiologic, symptomatic,  and field study evidence discussed earlier.
Specifically, increased  airway resistance, bronchial  responsiveness,  susceptibility to
respiratory infection, airway permeability, and incidence of asthma  attacks and airway
inflammation suggest that ambient O3 exposure  could be a cause of the increased hospital
admissions, particularly for asthmatics  (CD, Sec. 9.3.2).

-------
                                           41
       6.      Daily Mortality
       Several studies published during the 1950's (California Dept. of Public Health, 1955,
1956, 1957; Mills,  1957a,b), the 1960's (Tucker, 1962; Massey et al., 1961; Mills 1960;
Hechter and Goldsmith, 1961), and the 1970's (Biersteker and Evendijk,  1976) have
suggested a possible association of O3 or oxidants with human mortality.  Most of these
studies were conducted using data from Los Angeles, CA, and all were flawed in some way,
which prevented drawing any definitive conclusions in earlier criteria documents.
       Several daily mortality studies published more recently have provided additional,
though limited, evidence of the association between O3 and daily mortality.  The Shumway et
al. (1988) analysis of 1970 to  1979 LA County mortality data indicated that disease factors
and other pollutants dominate the seasonal cycles in mortality in LA.  However, the Kinney
and Ozkaynak (1991) reanalysis of the Shumway et al. (1988) data concluded that O3
explained a small, but statistically significant, portion of day-to-day variations in total
mortality in that city over a 10-year period.  The authors of the reanalysis did recognize that
the possible mechanism linking O3 with mortality is speculation based  on  known acute
pulmonary effects.  They further emphasize  that, although statistically significant  associations
have been detected  among mortality and environmental variables, one  can not conclude with
complete confidence that such associations are causal  based on results  from an observational
study.
       Total daily human mortality data in Detroit,  MI during the period from  1973 to 1982
were analyzed by Schwartz (1991) to investigate the effects of paniculate matter on mortality
and concluded that  O3 was "highly insignificant as a predictor of daily mortality." The CD
(Sec. 7.4.1.3) concluded that the poor documentation  of the mortality-O,  modeling,
especially regarding the lack of model specification  details or model coefficient
intercorrelations, makes the author's statement very difficult to evaluate.  Finally, Dockery et
al. (1992) conducted an analysis of total daily mortality in St. Louis, MO and Kingston-
Harriman, TN during the period September  1985 to August 1986, also with the intention of
assessing the effects of particulate matter on mortality.  Although the Dockery et  al. (1992)
stud\  showed  no association between  O3 and  mortality,  this may have  been in part a result  of
the particular methodological and exposure characteristics of the study vis-a-vis identification

-------
                                           42
of O3 health effects.  Therefore, the CD (Sec. 9.6) concludes that although an association
between ambient O3 exposure in areas with very high O3 levels and daily mortality has been
suggested, the strength of any such association remains unclear at this time.
       7.     Acute Inflammation and Respiratory Cell Damage
       Ozone has the potential to induce inflammatory responses throughout the respiratory
tract, including the nasopharyngeal region and the lungs.  Humans and laboratory animals
exposed to O3 can develop inflammation and  increased permeability in the nasal passages. A
positive correlation was reported between nasal inflammation in children and measured
ambient O3 concentrations.  Experimental studies of rats suggest a potential competing
mechanism between the nose and lung, with inflammation occurring preferentially in the nose
at lower O3 concentrations and shifting to the lung at higher concentrations.  It is unclear if
this represents a peculiarity of rats or is a more general phenomenon (CD, Sec.  9.3.3.1).
       In general, respiratory inflammation can be considered to be a host response to injury
and indicators of inflammation as evidence that respiratory cell damage  has occurred.
Inflammation induced by exposure of humans to O3 can have several potential outcomes: (1)
inflammation induced by a single exposure (or even several exposures over the course of a
season) can resolve entirely; (2) repeated acute inflammation  can develop  into a chronic
inflammatory state; (3) continued inflammation can alter the structure and  function of other
pulmonary tissue, leading to disease processes such as  fibrosis; (4) inflammation can interfere
with the body's host defense response to particles and inhaled microorganisms, particularly in
potentially vulnerable populations such as children and older individuals; and (5)
inflammation can interfere with the lung's response to other agents such as allergens or
toxins.  Except for outcome (1), the possible  chronic responses have not been  demonstrated
with inflammation induced by exposure of humans to O3.  It is also possible that the profile
of response can be altered in persons with preexisting pulmonary disease (e.g., asthma,
COPD) or smokers.
       The recent use of bronchoalveolar lavage (BAL) as a research tool  in humans has
afforded the opportunity to sample cells and fluid from the lung and lower airways of
humans exposed to O3 and to ascertain the extent and course of inflammation and its
constitutive elements. Several studies (Arts et ai., 1993a,b; Schelegle et a).,  1991; Koren el

-------
                                           43
al., 1989a,b; Devlin and Koren, 1990; McGee et al.,  1990; Koren et al., 1991; Devlin et
al., 1995; Hazucha et al., 1995; Seltzer et al., 1986) have shown that humans exposed  for
short-term periods (1- to 3- hr) to 0.2 to 0.6 ppm O3 had O3-induced inflammation, cell
damage, and altered permeability of epithelial cells lining the respiratory tract (allowing
components from plasma to enter the lung). The lowest concentration of O3 tested,
0.08 ppm for 6.6 hours with  moderate exercise, also induced small but statistically
significant increases in these  endpoints (Devlin et al.,  1990,  1991; Koren et al., 1991).
       Polymorphonuclear leukocytes (PMNs),  generally considered to be the hallmark of
inflammation, make up 8 to 10% of recovered BAL cells in  individuals exposed for 2 hrs to
0.4 to 0.6 ppm O3 (Seltzer et al., 1986). This is a 5- to 8-fold increase in PMNs compared
to similar individuals exposed to clean air,  who typically have 1 to 2% PMNs in their BAL
fluid.  Asthmatic individuals  generally have baseline levels of PMNs which do not differ
significantly from  those of healthy individuals, but PMN levels can increase following
allergen bronchoprovocation  (CD, Sec.  9.3.3.1).
        Exposures of animals to O3 for periods _<_ 8hr also result in cell damage,
inflammation, and altered permeability, although, in general, higher O3 concentrations are
required to  elicit a response equivalent to that of humans.  Because humans were  exposed to
O3 while exercising  and most animal studies were done at rest, differences in ventilation
likely play a significant role in the different response of humans and  rodents to the same
O3 concentration.  Studies  in which laboratory animals were exposed at night (during their
active period) or in  which  ventilation  was increased with CO2 tend to support this idea (CD,
Sec. 9.3.3.1).
        Studies utilizing BAL techniques sample only free or loosely adherent cells in the
lung; thus,  it is possible that cellular changes have occurred in the interstitium that are not
reflected in BAL studies, or  that BAL changes  exist in the absence of interstitial changes.
However, morphometric analyses of inflammatory cells present in lung and airway tissue
sections of  animals exposed to O3 are in general agreement with BAL studies.   Ozone
exposures of <_ 8 hrs  cause  similar types of alterations in lung morphology in all laboratory
animal  species studied. The  most affected  cells are the ciliated epithelial cells of the airways
and  Type 1 cells in  the alveolar region.  The centriacinar  region (CAR),  the junction of the

-------
                                           44
conducting airways and gas exchange region, is a primary target in all species studied,
possibly because it receives the greatest dose of O3 delivered to the lower respiratory tract.
Sloughing of ciliated epithelial and Type 1 cells occurs within 2 to 4 hrs of exposure of rats
to 0.5 ppm O3.
       Findings from human and animal studies show that the O3-induced inflammatory
response occurs rapidly and persists for at least 24 hrs.  Increased levels of neutrophils and
protein are observed  in the BAL fluid within 1 hr following a 2-hr exposure of humans to
O3 and continue for at least 20 hrs.  The kinetics of response during this time have not been
well studied in humans, although a single study shows that neutrophil levels are higher at
6 hours postexposure than at 1 or  20 hrs in different individuals.  Several animal studies
suggest that neutrophil and BAL protein levels peak 12 to  16 hrs after an acute O3 exposure
and begin to decline by 24 hrs, although some  studies report detectable BAL neutrophils even
36 hrs after exposure. It is also clear that in humans the pattern of response differs for
different inflammatory mediators.  Mediators of acute inflammation, such as interleukin-6
(IL-6) and prostaglandin-E^ (PGE^), are more elevated immediately after exposure;  whereas
mediators that potentially could play a role in resolving inflammation, such  as flbronectin and
plasminogen activator, are preferentially elevated 18 hours after exposure.   The rapidity with
which cellular and biochemical mediators  are induced by O3 makes it conceivable that some
of them may play a role in O3-induced changes in lung function.  Indeed, there is j>ome
evidence that BAL  PGE? levels are correlated  with decrements  in FEV,, and anti-
inflammatory medications that block PGEj production also reduce or block the spirometric
responses to O3.  Although earlier studies suggested that O3-induced PMN influx might
contribute to the observed  increase in airway hyperreactivity, animal studies show that when
neutrophils are prevented from entering the lung, Gyinduced hyperreactivity or increases  in
many inflammatory mediators still occur.  In addition, studies in which anti-inflammatory
drugs are used  to block O3-induced lung function decrements still show increases in
neutrophils and most other inflammatory mediators (although PGE^ is not increased) (CD,
Sec.  9.3.3.1).
       It is the view of staff and of medical experts consulted that the repeated acute
inflammatory response and morphological changes discussed above is potentially a matter of

-------
                                           45
public health concern; however, it is also recognized that most, if not all, of these effects
have begun to resolve in most individuals within 24 hrs if the exposure to O3 is not repeated.
Of possibly greater public health concern is the potential for chronic respiratory damage
which could be the result of repeated 03 exposures occurring over a season  or a lifetime.
The evidence for these chronic effects is discussed in the following section.
       8.  Chronic Respiratory Damage
       To  evaluate the impact on the human respiratory system of long-term exposures to O3,
it has been necessary for researchers to utilize the results from both epidemiology and animal
toxicology studies.  There are clear limitations  in using these approaches which cannot be
fully overcome when compared to controlled-exposure  human experimental  studies.
Epidemiology studies do not provide clear causal relationships due to the presence of
confounding variables (e.g., heat,  humidity, other pollutants); however, the results can
provide associations which may suggest causal  relationships. Animal toxicology studies,
though limited by species sensitivity and dosimetry differences between humans and
experimental animals, can offer controlled experimental conditions for chronic exposures and
thereby provide evidence of causal relationships.  Dosimetric extrapolation techniques  have
improved dose-target tissue  relationships, but lack of a full understanding of species
sensitivity  differences between humans and animals limits the extent to which results of
toxicology data can be extrapolated to human health effects.
       Epidemiologic studies (Abbey et al., 1993; Detels et al.,  1991; Euler et al., 1988;
Hodgkin et al., 1984; Schwartz, 1989; Stern et ai., 1989, 1994; Portney and Mullahy, 1990;
Schmitzberger et  al., 1993)  that have investigated potential associations between long-term O,
exposures and chronic respiratory  effects in humans thus far have provided  only suggestive
evidence that such a relationship exists.  Most studies investigating  this association have  been
cross-sectional in design and have  been compromised by incomplete control of  confounding
variables and inadequate exposure  information.  Other  studies have attempted to follow
variably exposed  groups prospectively.  Studies have been conducted in Southern California
(Detels et al., 1991) and in Canada (Stern et al.,  1989, 1994) designed to compare lung
function changes  over several years between populations living in communities  with high and
low oxidant ambient air levels.  While recognizing that pollution levels have improved
markedly in Southern California during the  past several decades, the findings still suggest

-------
                                             46
small, but consistent, decrements in lung function among inhabitants of the more
communities which have been highly polluted; however, associations between O3 and other
copollutants and problems with study population loss have reduced the level of confidence in
these conclusions.  Another study (Abbey et al., 1993), reporting  associations between O3
and the incidence and severity of asthma in Seventh Day Adventists over a decade, had
similar results,  but were even less suggestive due to the colinearity of O3 with other air
pollutants.  This is largely due to the difficulty of partitioning effects between O3 and
particles.  Nevertheless,  in all of the studies assessing lung function, the pattern  of
dysfunction associated with the long-term O3 exposure has been consistent with the functional
and structural abnormalities seen in laboratory animals, as discussed in  the CD (Sec. 9.4.2
and Sec. 6.5.3.3).
       The advantage of laboratory animal studies is the ability to examine closely the
distribution and intensity of the O3-induced morphologic changes that have been  identified
throughout the respiratory tract (CD, Sec. 6.2.4  and Sec. 9.4.2).  Indeed, cells of the nose,
like the distal lung, clearly are affected by O3.  Perhaps of greater health concern are the
"lesions3"  that occur in the small airways and in the centriacinar regions (CAR)  of the lung
where the  alveoli meet the distal airways, as pictured in Figure V-3 (CD,  Figure 9-12).
Altered function of the distal airways, the proximal conduits of air to the gas-exchange
regions, can result in reduced communication of fresh air with the alveoli and air-trapping.
In fact, "lesions" found in animals following chronic O3 exposures are reminiscent of the
earliest "lesions"  found in respiratory bronchiolitis, some of which may progress to fibroiic
lung disease (Kuhn et al., 1989; King, 1993).
       "Lesions" in the CAR are one of the  hallmarks of O3 toxicity. having been  well
established.  The study of Chang et al. (1992) exposed rats to an urban pattern of O, (13 hr
0.06 ppm background, 7 days/week, on  which were superimposed 9-hr peaks, 5  days/week,
slowly rising to 0.25 ppm)  for 78 weeks and made periodic examinations of the  CAR
* During the March 1995 CAS AC meeting, and in subsequent written comments, substantial disagreement was,
expressed among Panel members regarding the use of the term "lesion."  Some believe use of the term implies more
serious damage than has been observed for O-, exposures, while other Panel members believe "lesion" i\ an
appropriate term to describe O3-mduced morphological abnormalities.  The CD (Sec. 6.2.4.1) describes and discusses
these degenerative changes, referred to as "lesions" for purposes of the CD and this Staff Paper.

-------
                       47
   Air
             Ozone
                    >'
£
             --•.*•-
                          •
                          VJ
                          s :
FIGURE V-3.  A SUMMARY OF MORPHOLOGIC LESIONS FOUND IN THE
TERMINAL BRONCHIOLES AND THE CENTRIAC1NAR REGION (CAR) OF
THE LUNG FOLLOWING EXPOSURE OF LABORATORY RATS TO FILTERED
AIR OR A SIMULATED AMBIENT PATTERN OF O-, FOR UP TO 78 WEEKS
IN THE TERMINAL BRONCHIOLE, SIZES OF THE DOME OF CLARA CELLS
BECAME SMALLER WITH 03 EXPOSURE, AND THE NUMBER OF CILIA IS
REDUCED (ARROWS). IN THE CAR, THE EPITHELIUM BECOMES THICKER.
AND ACCUMULATION OF COLLAGEN FIBERS OCCURS (ARROW HEADS)
Source:  Chang et al. (1992); CD, Figure 9-12.

-------
                                        48
tracheobronchial and proximal alveoli by TEM morphometry during and after exposure.  In
general, Chang et al. (1992) found (1) changes in Type 1 and Type 2 cell volume which
returned to normal  17 weeks post exposure, (2) epithelial and endothelial basement
membrane were thickened and accompanied by increased collagen fibers at  17 weeks post
exposure, and (3) in the tracheobronchial regions, surface areas of ciliated and nonciliated
cells decreased during exposure.  In  a related study of identically exposed groups of rats,
Tepper et al. (1994) reported (1) increases in expiratory resistance suggesting central airway
narrowing after 78  weeks exposure,  (2)  tidal volumes reduced at all evaluation times during
the exposure, and (3) breathing frequency reduced though no single evaluation time was
significant.  In another related study  with similar protocol, Costa et al. (1995) reported
reduced lung volume, which is consistent with a "stiffer" lung (i.e., restrictive lung disease).
             As shown in Figure V-4 (CD, Figure 9-13), the temporal pattern of effects
during and after a chronic exposure is complex. During the early days of exposure, the end-
            0)
            v>
            I
            V)
            
            TO
                            Epithelial hyperplasia
                            Bronchioloalveolar exudate
                            Interstitial fibrosis
            Exposure
                      6 1 mo
   Postexposure
                 days
6 mo
                                      Time
12 mo
             FIGURE V-4.  SCHEMATIC COMPARISON OF THE DURATION-
             RESPONSE PROFILES FOR EPITHELIAL  HYPERPLASIA,
             BRONCHIOLOALVEOLAR EXUDATION, AND INTERSTITIAL
             FIBROSIS IN THE CENTRIACINAR REGION OF LUNG EXPOSED TO A
             CONSTANT LOW CONCENTRATION OF OZONE.
Source:  Dungworth (1989). CD, Figure 9-13.

-------
                                           49
airway lumenal and interstitial inflammation peaks, and, thereafter, appears to subside at a
lower plateau of activity sometimes referred to as a "smoldering lesion".  Several cytokines
remain elevated beyond the apparent adaptation phase of the response and may be linked
conceptually to the development of chronic "lesions" in the distal lung. However, a clear
association of these BAL-derived mediators and cells with long-term toxicity has yet to be
demonstrated (CD, Sec. 9.4.2).
       A multicenter chronic study, supported by the National Toxicology Program (NTP)
and the Health Effects Institute (HEI), involved numerous researchers and laboratories (Last
et al.,  1994; Szarek, 1994; Radharkrishnamurthy, 1994; Parks and Roby, 1994; Harkema
and Mauderly,  1994; Harkema et al., 1994; Chang et al., 1995; Pinkerton et al.,  1995;
Catalano et al., 1995a,b).  This NTP (1995) study further illustrates some of the complex
interrelationships among the structural, functional, and biochemical effects.   These three
health  endpoints were evaluated in a collaborative project using rats exposed 6 hrs/day, 5
days/week for 20 months to 0.12, 0.50, 1.00 ppm O3.  Although lung biochemistry and
structure were affected at 0.5 ppm and 1.00 ppm but not at 0.12 ppm  O3, there were no
observed effects on pulmonary function at any exposure level.
       Combined analyses of the NTP (1995) collaborative studies showed that 0.5 ppm and
1.00 ppm O3 caused a variety  of structural and biochemical effects.  Exposures to 0.12 ppm
O3 caused  no major effects, although a few specific endpoints were altered.  Hallmarks of
chronic rhinitis (e.g., inflammation,  mucous cell hyperplasia, decreased mucous flow) were
observed in focal regions of the nasal cavity.  Structural and  biochemical changes included
some,  but  not many, hallmarks of airway  disease.  Typical Cyinduced changes (e.g.,
bronchiolarization, increased interstitial matrix) observed in the tracheobronchial region and
in the  CAR were characteristic of centriacinar fibrosis; however, diffuse pulmonary fibrosis
was not observed.
       Trends for centriacinar fibrosis. airway disease, and chronic rhinitis were  examined
by Catalano et al. (1995a) for 10, 18, and 3 endpoints, respectively, from  the individual NTP
(1995) studies.  A statistically significant trend was noted for the association between chronic
rhinitis and increasing O, concentration.  The differences between control and exposed rats
were statistically significant at 0.50 ppm and 1.00 ppm O3.  Marginally significant and

-------
                                           50
significant trends were found for the association between centriacinar fibrosis or airway
disease and increasing O3 concentration; however, no statistically significant differences were
found between control and O3-exposed rats (CD, Sec. 6.5.3.3).
       Studies of prolonged O3 exposures in monkeys and rats reveal generally similar
morphologic responses, although it appears that the monkey exhibits somewhat more tissue
injury than does the rat under roughly similar exposure conditions (CD, Chapter 8).
Interspecies comparisons of dosimetric data indicate that  the monkey, with its similarity to
the human in distal airway structure,  provides data that may best reflect the potential effects
of O3 in humans exercising out of doors.  As such, monkeys exposed to O3 at 0.15 ppm for 8
h each day for 6 to 90 days exhibit significant distal airway remodeling.  Rats show similar
but more modest changes at 0.25 ppm O3 after exposures of longer duration, up to 18  mo
and beyond (near-lifetime). The chronic distal lung and  airway  alterations appear consistent
with incipient peribronchiolar fibrogenesis within the interstitium.  Attempts to correlate
functional deficits have  been variable, perhaps due in part to the degree and distribution of
the "lesions"  and the  general insensitivity of most measures of the distal lung  function. The
interstitial changes may progress,  however.  Moreover, one recent primate study revealed
evidence that intermittent challenge with a pattern of O3  exposure more reflective of seasonal
episodes, with extended periods of clean air in between extended periods of O3, actually
leads to greater injury.  The reasons for this are unclear  but may relate to  the known loss of
tolerance that occurs  in both humans  and  animal test species with removal  of the oxidant
burden.
       Probably the most provocative, albeit preliminary, evidence of possible pollutant
effects in the population is offered by Sherwin  (1991) and Sherwin and Richters (1991).
They performed a pathological evaluation of the lungs from 107 Los Angeles  County
residents (15 to 25 years of age),  who had a  sudden death without disease or lung trauma.
Sherwin (1991) reported that the odds ratio for severe CAR disease  (defined as the extension
of a respiratory bronchiolitis into  the proximal  acinar structures) in subjects living in
metropolitan  Los Angeles versus those living in other cities in Los  Angeles County was 4.0
(95% confidence limit,  1.4 to 11.3).  Unfortunately, no  exposure data or  lifetime residence
data, no smoking  histories, no cotinine results, nor occupational histories  were available.

-------
                                           51
The lack of a smoking history on subjects is of critical importance because respiratory
bronchiolotis has been shown to be an early pathologic change found in the pulmonary
airways of young smokers (CD, Sec.  7.4.2.2.) Furthermore, the subjects were mainly of
low socioeconomic status and only  10 were female, and the observation is limited by a lack
of quantitative morphometry of lung specimens and by the lack of a control group from an
ambient environment with low oxidant pollution.  (Many of these limitations should be
addressed in research which is currently being planned by the USEPA National Health and
Environmental Effects Research Laboratory.)  Therefore, although the Sherwin (1991)
observation is of great interest, particularly with regard to other primate data which show O3-
associated effects in the CAR, the results are not of particular value in determining  human
exposure levels for O3 which might induce chronic respiratory disease, nor do  they establish
a causal relationship between the oxidant environment found in  metropolitan Los Angeles and
the pathologic effects observed  by Sherwin (1991) (CD, Sec. 7.4.2.2).
       In summary, the collective data on chronic exposure to O3 garnered in animal
exposure and human population studies have many ambiguities.  It is clear that the
distribution of the O3 "lesions"  is roughly similar across species (e.g., monkeys, rats, mice).
These responses are concentration dependent (and perhaps time or exposure-pattern
dependent).  Under certain conditions, some of these structural  changes may become
irreversible.  It is unclear whether ambient exposure scenarios encountered by  humans result
in similar "lesions."  Furthermore,  it is highly uncertain whether there are resultant
functional or impaired health outcomes in humans chronically exposed to O3, particularly
because the human exposure scenario involves much longer-term exposures than can be
investigated in the laboratory.   The epidemiology studies of lung function change generally
parallel those of the animal studies, but they lack good information on individual O3 exposure
and are frequently confounded by personal or copollutant variables (CD,  Sec. 9.4.2).   In
summary,  the animal toxicology data discussed above, and  in the CD  (Chapter  6), provides a
biologically plausible basis for considering the possibility that repeated inflammation
associated  with exposure to O3  over a lifetime may result in sufficient damage  to respiratory
tissue such that individuals later in  life may experience a reduced quality of life, although
such relationships remain highly uncertain.

-------
                                          52
       9.   Genotoxicity and Carcinogenicity
       Numerous in vitro exposure studies suggest that O3 has either weak or no potential to
cause mutagenic, cytogenetic, or cellular transformation effects.  Most of these experiments
utilized high concentrations of Oj (>5.0 ppm).  Because of the exposure systems used, there
are unknowns about uncertainties regarding the formation of artifacts and  the dose of O3.
Therefore, these studies are not very useful in health assessment.  Cytogenetic effects have
been observed in some, but not all, laboratory animal and human studies of short-term
O3 exposure.   However, well-designed human clinical cytogenetic  studies  were negative.
       Until recently, in vivo exposure studies of carcinogenicity,  with and without
co-exposure to known carcinogens, were either negative or ambiguous.  A well-designed
cancer bioassay study has recently been  completed  by the National Toxicology Program
(NTP, 1995)  using  male and female Fischer 344/N rats and B6C3F, mice.  Animals were
exposed for 2 years to 0.12, 0.5, and 1.0 ppm O3 (6 h/day, 5 days/week).  A similar lifetime
exposure was conducted, but 0.12 ppm was not used.  The NTP (1995) evaluated the weight-
of-evidence for this study; they found "no evidence" of carcinogenicity in rats but reported
"equivocal evidence" of carcinogenicity  in  O3-exposed male mice and  "some evidence" of
carcinogenic activity in one strain of O3-exposed  female mice. The increases in adenomas
and carcinomas were observed only in the  lungs.  There was no concentration response.  In
the male mice, the incidence of neoplasms in the 2-year study was not elevated significantly
by O3 and was within the range of historical controls.  Also, the lifetime exposure did  not
increase significantly the  incidence of neoplasms, even though the  incidence of carcinomas
was increased.  In the female mice, a 2-year (but not lifetime) exposure to 1.0 ppm O3 only
increased the incidence of animals with neoplasms.  When the female  mouse data from the
two exposure regimens (at 1.0 ppm)  were combined, there was a statistically significant
increase (almost double) in neoplasms.  In  a companion study, male rats were treated with  a
tobacco carcinogen  and exposed  for 2 years to 0.5  ppm O3. Ozone did not affect the
response and,  therefore, had no tumor promoting activity.
       In summary, only long-term exposure to a high concentration of O3 (1.0 ppm) has
been shown to evoke a limited degree of carcinogenic activity in B6C3F! mice.  Rats were
unaffected. Furthermore, there was no concentration response, and there  is inadequate

-------
                                            53
information from other research to provide mechanistic support for the finding in mice.  For
these reasons, the staff believes it is inappropriate to extrapolate these mouse data to humans.

D.     Factors Modifying Acute Human Response to Ozone
       There are several factors which have been identified as potentially affecting human
susceptibility to O3 exposure by altering acute physiological susceptibility.  The more
significant of these factors are exertion (e.g., exercise, manual labor),  preexisting disease,
age, gender, ethnicity/race, smoking status, environmental  factors.  Although  most of these
factors have not been addressed adequately in clinical studies in  order to draw definitive
conclusions, preliminary observations have been made regarding each of these potential
modifiers of response.  A thorough discussion is presented  in Section 7.2 of the CD.
       1.   Exertion and Ventilation
       Exertion resulting in an increased minute ventilation (V^ is a factor which increases
O3 sensitivity of most humans at any elevated O3 concentration.  This is in part due to the
fact  that at higher VE there is an increase in O3  dose received by the  lungs.  It is  also due to
the deeper penetration of O3 into more peripheral regions of the lungs, which  are more
sensitive to acute O3 response and injury.  This provides general support for the hypothesis
that  increasing the level of exertion for most  individuals increases  the impact of a given
concentration of O3.  Furthermore, research has shown that respiratory effects are observed
at lower O3 concentrations if the level of exertion is increased and/or the duration of exertion
is extended.  An increased level of exertion can cause an individual,  who has  a respiratory
system which  is highly responsive to O3, to experience lung function impairment and
symptoms sufficient to  curtail activity, even though  the individual is  otherwise healthy.
       Representative activities and associated ventilation rates are summarized in Table V-3
for varying levels of exertion. While the table  identifies only a  few  of the many activities in
which individuals engage, it is intended to provide the reader a sense of the relationship
between level  of exertion, ventilation rate, and  type of activity.

-------
    TABLE V-3. ESTIMATED MINUTE VENTILATION RATES AND REPRESENTATIVE ACTIVITIES
                     ASSOCIATED WITH VARYING LEVELS OF EXERTION1
Level of Exertion
Light
Light
Moderate
Moderate
Moderate
Heavy
Heavy
Very Heavy
Very Heavy
Very Heavy
Severe
Minute
Ventilation
L/min
12-16
17-23
23-30
29-38
35-46
42-55
52-57
62-79
73-93
89-110
107-132
Representative Activities'"
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 firewood
Easy cycling; pushing wheelbarrow with 75-kg
load; using sledgehammer
Climbing stairs; playing tennis; digging with
spade
Cycling at 13 mph; walking on snow; digging
trenches
Cross-country skiing; rock climbing; stair
climbing with load; playing squash and
handball; chopping with axe
Level running at 10 mph; competitive cycling
Competitive long distance running; cross-
country skiing
'See text of Criteria Document (U.S. EPA. 1986, pp. 10-13 to 10-15) for discussion.
hAdapted  from Astrand and Rodah! (1977).

-------
                                           55
       2.   Preexisting Disease
       Controlled studies on mild asthmatics suggest that they have similar lung volume
responses but greater airway resistance changes to O3 than nonasthmatics.  Furthermore,
limited data from studies of moderate asthmatics  suggest that this group may have greater
lung volume responses than nonasthmatics.  Daily life studies reporting an exacerbation of
asthma and decrease in peak expiratory flow rates, particularly in asthmatic children, appear
to support the controlled studies; however, those  studies are confounded by temperature,
particle or aeroallergen exposure,  and asthma severity of the subjects or their medication  use.
In addition, field studies of summertime daily hospital admissions for respiratory causes show
a consistent relationship between hospital admissions for asthmatics and ambient levels of O,
in various locations in the Northeastern United States, even after controlling for independent
contributing factors.
       Other population groups  with preexisting limitations in pulmonary  function and
exercise  capacity (e.g., chronic obstructive pulmonary disease, ischemic heart disease) would
be of primary concern in evaluating the health effects of O3.  Unfortunately, not enough is
known about the responses of these individuals to make definitive conclusions regarding their
relative sensitivity to Ov  Indeed, functional effects  in these individuals with reduced lung
function  may have greater clinical significance than comparable changes in healthy persons.
       3.   Age, Gender, Ethnic, and Tobacco Smoke Factors
       Age Factors.  Age differences as a factor  influencing response to O3 are not yet fully
understood.   This is in part due  to the fact that most of the O3 controlled-exposure studies
have been conducted with young adults rather than with children or older subjects.
However, there is a growing body of evidence, including clinical,  field, and epidemiology
studies, which  suggests that age  plays a role in determining sensitivity to Ov  Based on the
available data,  it appears that children  respond to low-level O3 exposures in a manner
comparable to that of young  adults, albeit without symptoms, while older persons  exhibit  a
decreased sensitivity relative to young  adults (CD, Sec. 9.6).  The lack of symptoms in
children  and  reduced sensitivity  in the  elderly could  lead to an increased risk of an individual
receiving a higher O< dose.   This  increased  risk of O3 exposure and dose is a direct result of
children  and  the elderly not taking mitigating behavior to avoid exposure because they do not

-------
                                           56
experience respiratory symptoms; however, this hypothesis has not been tested and has not
been demonstrated at this time.
       Gender Factors.  During the previous review of O3 NAAQS, so few human or animal
studies had been conducted using any female subjects in controlled-exposure studies of O3
that few substantive conclusions could be drawn regarding gender differences.  Although
there are more data on female subjects than was the case previously, these new data have not
yet provided conclusive evidence that men and women respond differently to O3.  Thus, the
question as to whether there is a difference between males and females in respiratory
susceptibility to O3 remains unresolved. Furthermore, it can be stated that if gender
differences do exist for respiratory susceptibility to  O3, they are not based on hormonal
changes,  differences in lung volume, or ratio of forced vital capacity (FVC)  to ^E (CD, Sec.
1.7).
       Ethnicity Factors.  In  studies  ihus far  conducted, the lung function decrements in
African-Americans were not statistically significantly greater than in other groups at all
concentrations tested. Even though these results can be considered suggestive of ethnic
differences, further research,  particularly on non-white asthmatics, must be conducted before
ethnicity  can be established as a clear factor in determining pulmonary  responsiveness to O,
(CD, Sec. 7.2.1.3).
       Tobacco Smoke.  Results of several early studies, which compare sensitivity of
individuals voluntarily exposed to tobacco smoke (i.e, smokers) versus  sensitivity of those
who have not been exposed to tobacco smoke, suggest that smokers are less  responsive to O,
than nonsmokers.  Although data on  O3 susceptibility of both active and passive smokers
remains limited, recent studies indicate that cessation of exposure to tobacco smoke leads to
improved baseline pulmonary  function and possibly  a return to O3 susceptibility (CD,  Sec.
7.2.1.3).
       4.    Interactions with Other Pollutants
       In  genera], controlled  human  studies of O3 mixed with other pollutants show no more
than an additive response with symptoms or spirometry as an endpoint.  This applies to O3 in
combination with nitrogen dioxide (NO:). SO2.  sulfuric acid (H:SO4). nitric acid  (HNO-,). or
carbon monoxide (CO).  Indeed, at the levels of copollutants used in human  exposure

-------
                                           57
studies, the responses can be attributed primarily to O3.  In one study, exposure to
O3 increased  airway responsiveness to SO2 in asthmatics.  Similarly, other pollutants (e.g.,
paniculate matter) that may increase airway responsiveness could augment the effect of O3 on
airway responsiveness.
       The relatively large number of animal studies of Q in mixture with NO2 and
H2SO4 shows that additivity, synergism,  and antagonism  can result, depending on  the
exposure regimen and the endpoint studied.  The numerous observations of synergism are of
concern, but the interpretation of most of these studies relative to the real  world is
confounded by unrealistic exposure designs.  For example,  ambient concentrations of
O3 often were combined with levels of copollutants substantially higher than  ambient,
creating the possibility that mechanisms of toxicity unlikely in the real world contributed  to
               f
the experimental  outcome.  Nevertheless, the data support a hypothesis that coexposure to
pollutants, each at innocuous or low-effect levels,  may result in effects of  significance.

E.     Sensitive Population Groups
       Several characteristics which influence the  extent to which  an  individual or population
group may show  increased sensitivity to  O3 have been discussed in the CD (p. 9-41).  These
individual or group characteristics are based on:   1) biological  responses to O3; 2)
physiological status; 3)  activity patterns;  4) exposure  history; and 5) personal factors such as
age, gender,  social, ethnic,  cultural, and nutritional status.
       1.   Active ("Exercising") Individuals
       One large group of individuals at risk to O3 exposure consists of those healthy
children, adolescents, and adults who  engage in outdoor activities  involving  exertion (i.e.,
"exercising"  individuals) during summer daylight hours.   This  conclusion is  based on a large
number of controlled O3-exposure human experimental studies, which have been conducted
on healthy, non-smoking, exercising adults and children  (ages 8 to 45).  These studies also
have demonstrated  wide variability among subjects in sensitivity to O3, although factors
contributing  to this variability are  not  well understood.

-------
                                            58
       2.   Individuals with Preexisting Respiratory Disease
       There is limited evidence from human controlled exposure studies to suggest that mild
asthmatics have greater changes in airway resistance following O3 exposure than
nonasthmatics but have similar lung volume responses; however, moderate asthmatics appear
to have greater lung volume responses than nonasthmatics.  Support for considering
asthmatics to be at increased risk to O3 exposure also comes from studies of hospital
admissions for respiratory causes which show a consistent relationship between asthma and
ambient O3 levels in the northeastern U.S., even after controlling for independent
contributing factors.  Studies of asthmatic children which report exacerbation of asthma and
decreased peak expiratory flow rates seem to  provide some further evidence of asthmatics
being at risk, but these studies are confounded by variables such as temperature, particle or
aero allergen exposure, severity of asthma, and medication  use (CD, Sec.  9.6).
       Although there are limited data on individuals with preexisting respiratory disease or
other limitations on their pulmonary function and exercise capacity (e.g., those with chronic
obstructive pulmonary disease, ischemic heart disease), there is insufficient information at
this time to draw any clear conclusions about their susceptibility to O3 relative to other
individuals. The major reason individuals with preexisting  respiratory disease may be of
concern is the likelihood that decrements in lung function or exercise capacity may have
greater clinical importance to the individual than similar changes in healthy persons.
       3.    Other Population Groups
       Several  population groups identified in the CD (Sec. 9.6) as not providing compelling
evidence to suggest that  they are more responsive than  the normal, healthy population
include:  the young and elderly, males and females, ethnic and racial groups, and individuals
with vitamin E deficiency or other nutritional  deficiencies.  Thus, in addition to the more
speculative at-risk status of those individuals with respiratory disease or pulmonary
deficiency, the  CD (Sec. 9.6) identifies only "exercising" or active health)  and asthmatic-
individuals, including children, adolescents, and adults  as having demonstrated  susceptibility
to O,.

-------
                                           59
F.     Adverse Respiratory Effects of Ozone Exposures
       As discussed in Chapter II of this Staff Paper, setting a primary O3 NAAQS involves
assessing protection of public health based on consideration  of sensitive populations at risk,
including factors such as the nature, severity, and frequency of O3-induced health effects
involved.  This section focuses on the nature, severity, and  frequency of specific O3-induced
health effects in order to provide a  basis for judgments regarding physiological changes that
become sufficiently severe to adversely  affect the health status of those individuals
experiencing such effects.  In considering populations at risk, staff recognizes that there is
wide variability in the severity of response to O3 among both healthy individuals and those
with impaired respiratory systems.  Individual sensitivity of healthy persons  to O3 and  the
extent to which impaired respiratory systems amplify the impact of various effects in
individuals with asthma and chronic obstructive pulmonary disease (COPD)  should be  taken
into account in making judgments about the adversity of Q,  effects. These judgments about
individual adverse effects are put into broader context in  the following sections on exposure
and risk analysis.  This broader context includes consideration, to the extent possible, of size
of the sensitive populations potentially at risk for various effects, and the kind and degree of
uncertainties inherent in assessing such  risks  in order to form judgments  about the various
levels of risk and adequacy of public  health protection afforded by alternative NAAQS.
       In this section, staff has attempted to identify and characterize the current
understanding as well as the divergence of opinion within the scientific community as to what
effects and degrees of response might be regarded as adverse health effects associated  with
exposure to O3.  This section presents staffs views on this issue, taking  into account
information in the CD, opinions that have been expressed by members of the  CASAC Ozone
Review Panel (Panel) at  its meetings and in written comments,  and opinions of other  health
and medical respiratory experts provided during the current  and previous O3 NAAQS
reviews.
       In 1985, the American Thoracic Society (ATS) published general  guidelines
describing what constitutes an adverse respiratory health  effect.  While recognizing that
perceptions of "medical significance"  and "normal activity"  mav differ among physicians.
lung physiologists,  and experimental subjects, the ATS (1985) defined adverse respiratory

-------
                                            60
health effects as "medically significant physiologic or pathologic changes generally evidenced
by one or more of the following:  (1) 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."  Staff believes this
definition provides a reasonable framework for present purposes.  As discussed above in
section V.C., human health effects for which clear, causal relationships  with exposure to G,
have been demonstrated fall  into the first category listed in the ATS definition.  Human
health effects for which statistically significant associations have been reported in
epidemiology studies fall into the second and third categories.  These effects include
respiratory illness that may require medication (e.g., asthma), but not necessarily
hospitalization, as well as emergency room visits and hospital admissions for acute
occurrences of respiratory  morbidity. Human health effects for which associations have been
suggested but not  conclusively demonstrated fall primarily into the last two categories.
Those health endpoints are based  on studies  of effects in laboratory animals and to a lesser
extent on human epidemiological  studies and can be extrapolated to human health effects only
with great uncertainty.
       1.  Permanent Respiratory Injury and/or Progressive Dysfunction
       An increase in daily mortality associated  with O3 exposure is unquestionably the most
adverse health effect for which only  limited,  suggestive evidence exists.  As discussed in
section V.C., causal  relationships have been reported in animal infectivity  studies of O3
(Coffin et al., 1967;  Coffin and Gardner,  1972;  Miller et  al., 1978).  However, only one
published epidemiological study (Kinney and Ozkaynak, 1991) has provided statistically
significant evidence of an association with daily  mortality  even at the very high levels of O,
found in Los Angeles.  In  that study, the authors state that although statistically significant
associations were  found between daily mortality and environmental variables,  one can not
conclude with complete confidence that  such  associations  are causal.  Also, other pollutants
(e.g., paniculate mattei) have been found to be significant contributors   to daily mortality,
but it is hard to determine the relative contributions of various pollutants (U.S.  EPA, 1996b).
No other human studies cited in the CD have reported statistically significant associations
between  O3 and mortality.

-------
                                           61
       Other adverse effects, such as scarring of lung tissue, reduced elasticity of the lungs,
and accelerated reductions in lung function have been clearly demonstrated only in laboratory
animal studies. These effects may be the result of repeated  pulmonary inflammation.
Although any association between ambient O3 exposures and permanent structural change in
the lung tissue in humans remains largely hypothetical at this time, indicators of acute
pulmonary inflammation following short-term O3 exposures  have been reported in several
human experimental studies.  There appears to be general agreement that a single exposure to
O3 that induces an inflammatory response has little or no health significance, just as a single,
short-term  exposure to the sun, sufficient to result in sunburn, would have little health
significance for most individuals.  However, it is well documented that long-term, repeated
exposures to the sun can damage the skin irreversibly. Analogously, some health scientists
have cautioned that if O3 exposures are repeated over many  months or years, the highly
irritating nature of O3 could induce chronic inflammatory responses in  humans, which may
culminate in irreversible lung tissue damage.
       Morphological abnormalities in the centriacinar region of the lungs, also referred to as
"lesions" by some researchers (as discussed in Sec. V.C.8 of this Staff Paper), are among
the most investigated chronic O3 effects in laboratory animal studies.  If these repeated acute
responses do in fact lead to  similar chronic effects in humans as have been observed in
laboratory  animals, it is possible that such effects could accelerate the loss of lung function
and the ability of elderly individuals to engage in activities which require  exertion later in
life. This  could impair their quality of life and could shorten longevity of affected
individuals.   Several efforts have been made to find associations between long-term O3
exposure and chronic respiratory dysfunction and disease (Detels et al., 1991; Abbey et al..
1993;  Schwartz, 1989).  Taken as a whole, these studies suggest that it is not possible to
conclude if there is an effect of O3 on the health effects studied, in part due to limitations
introduced by loss of subjects during the studies and by confounding variables  such as
coexposure to paniculate matter. Thus, the appropriate conclusion to be drawn at this time
is that associations  between  O3 exposure and chronic health  impacts  have  not been
sufficiently demonstrated in humans.  Some Panel  members expressed  views at the March
1995 and September  1995 CASAC meetings, and in subsequent written comments, that the

-------
                                            62
chronic health effects discussed above pose a sufficiently important public health threat as to
warrant serious consideration in this review of the O3 NAAQS.  Other Panel members
expressed the opinion that such health outcomes are too uncertain to be considered at this
time.  In consideration of the potential seriousness to public health of possible chronic health
effects of O3, staff agrees with the position taken  in the CASAC closure letter (Wolff, 1995b)
recommending that research  efforts continue on the  chronic health effects of O3 to reduce the
uncertainties before the next review of the O3 NAAQS.
       2.   Episodic and  Incapacitating Illness in Persons with Impaired Respiratory Systems
       The most significant episodic health effects that have been associated with short-term
O3 exposures are increased hospital admissions and  emergency room visits due to respiratory
causes. Health effects related to increased respiratory hospital admissions and emergency
room visits include respiratory infections (e.g.,  pneumonia), asthma attacks, and exacerbation
of other respiratory diseases  (e.g., COPD).  There exists a substantial, and growing, data
base which suggests an association between O3 and  increased respiratory hospital  admissions
for individuals with asthma and other  impaired respiratory systems. By analogy, while it is
plausible that healthy individuals,  particularly the  individuals who have lost  a significant
amount of their lung reserve capacity, could be adversely affected at very high O, levels
sufficient  to require hospitalization, there is no evidence to show this  is occurring with
ambient O3.
       Although  little controversy exists regarding the adversity  to the individual for
responses  that lead to being admitted to the hospital or to visiting an emergency  room, there
is still debate over the extent to which exposure to O3 is directly responsible for  these
adverse responses relative to other environmental  factors (e.g., exposure to  other air
pollutants, heat, humidity, allergens),  which could confound the  association  with Ov  In
assessing the significance of  other effects of short-term O, exposures that have been
demonstrated in controlled human exposure studies  (e.g.. decreased lung function, respirators
symptoms),  it is important to consider the magnitude of such individual changes  in persons
with impaired respiratory systems (e.g., asthmatics) who already have reduced lung  function.
A comparable change in lung function could have greater impact on the health status.
whether illness or interference with normal activity, of an individual with a  preexisting

-------
                                          63
respiratory disease, such as asthma, chronic bronchitis, emphysema, or serious allergies, than
on a healthy individual with normal lung function and reserve capacity. Any change in lung
function that causes these individuals with  impaired respiratory systems to drop below 40 to
50 percent of predicted values would be considered clinically adverse.  For example, O3-
induced changes in SR,W, a measure of airway narrowing, are small and of minimal clinical
significance in nonasthmatic individuals.  Asthmatics,  however, often have baseline airway
narrowing  and experience larger changes in SR,W on exposure to O3 than do nonasthmatics.
Because of these baseline differences, the clinical significance of increases in SR,W depends
both on percent change  from baseline and  on absolute increases in SR^ (CD, p. 9-23).
       Individuals with  asthma represent a population subgroup which has been examined
extensively in experimental and epidemiological studies of O3.  Asthmatic individuals have
been found to exhibit O3-induced airway responses that are slightly more pronounced than
those found in non-asthmatic persons.  It is important to understand asthma as a disease and
place the effects reported in controlled human exposure studies into proper context.  This
involves careful definition of asthma,  classification of asthma by severity of disease,
discussion  of medication use, and description of the  nature and time course of response.
These considerations of asthma have been  addressed previously by the EPA and are
presented  in the SO2 Staff Paper Addendum (USEPA, 1994a, pp. 11-33) and Criteria
Document Addendum (USEPA,  1994b).  The definition of asthma contained in those
documents and taken from the Expert Panel Report from the National  Asthma Education
Program of the National Heart, Lung, and Blood Institute (NIH,  1991) is:
       Asthma is a lung disease with  the following characteristics:  (1) airway obstruction
       that is reversible (but not completely so in some patients) either spontaneously or with
       treatment, (2) airway inflammation, and (3) increased airway responsiveness to a
       variety of stimuli.
       Working with scientists in EPA's ORD, staff de\eloped Tables V-4a. V-4b. and  V-4c
(Table 9-2 in the CD), which categorize acute respiratory responses to O3 in individuals with
impaired respiratory systems according to type and severity of response.  These tables are
based on a similar categorization for healths individuals developed by staff as Table VII-5

-------
                                           64
        Tables V-4a, V-4b, and V-4c.  Gradation of Individual Responses to
     Short-Term Ozone Exposure in Persons with Impaired Respiratory Systems"
Table V-4a
Functional Response
FEV, change
Nonspecific
bronchial responsiveness11
Airway resistance
(SR.J
Duration of response
None
Decrements of
<3%
Within normal
range
Within normal
range (±20%)
None
Small
Decrements of 3 %
to £10%
Increases of < 100%
SR,^ increased
<100%
<4hr
Moderate
Decrements of
> 10% but
<20%
Increases of
:£300%
SR,U increased up
to 200% or up to
15cm H:O/s
>4 but <24 hr
Large
Decrements of
S20%
Increases of
>300%
SR^ increased
>200% or more
than 15 cm H:0/s
>24hr
 Table V-4b
Symptomatic Response
Wheeze
Cough
Chest pain
Duration of response
Normal
None
Infrequent
cough
None
None
Mild
With otherwise
normal breathing
Cough with deep
breath
Discomfort just
noticeable on
exercise or deep
breath
<4hr
Moderate
With shortness of
breath
Frequent
spontaneous
cough
Marked
discomfort on
exercise or deep
breath
>4 but <24 hr
Severe
Persistent with
shortness of
breath
Persistent
uncontrollable
cough
Severe
discomfort on
exercise or deep
breath
>24 hr
 Table V-4c
Impact of Various
Functional and/or
Symptomatic Responses
Interference with normal
activity
Medical treatment/Self
Medication
Normal
Functional
and /or
Symptomatic
Responses
None
Na change
Small Functional
and/or Mild
Symptomatic
Responses
Few individuals
likely to limit
activity
Normal medication
as needed
Moderate
Functional and -or
Symptomatic
Responses
Many individuals
likely to limit
activity
Increased
frequency or
additional
medication use
Large Functional
and or Se\ere
Symptomatic
Responses
Most individuals
likely to limit
activity
Increased
likelihood of
physician or ER
visit
'See text for discussion, abbreviations and acronyms.
bAn increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PD^, or PD,,
(see Chapter 7, Section 7.2.3 of the CD for a more complete discussion).

-------
                                           65
in the previous Ozone Staff Paper (U.S. EPA, 1989) produced during the last O3 NAAQS
review.
       In addition to the health status of the individual, the clinical  significance of individual
responses to O3 depends on the magnitude of changes in pulmonary function, the severity of
respiratory symptoms, and the duration of response.  Tables V-4a and V-4b categorize
individual functional and symptomatic responses to O3  exposure as either normal (i.e., none)
or with graded levels of increasing severity in individuals with impaired respiratory systems,
similar to Tables V-5a and V-5b  for healthy individuals discussed later in this section.
Pulmonary function responses are represented in these tables by changes in spirometry (e.g.,
FEV,), SRaw, and non-specific bronchial responsiveness. Respiratory symptom responses
include cough, pain on  deep inspiration, and  wheeze.  The predominant changes in
spirometry discussed  in this Staff Paper are O3-induced decrements  in FEV, because they are
more easily quantified,  have a continuous distribution, and have been used  to provide most of
the exposure-response relationships described in the CD and in the  exposure and risk
analyses. The combined impacts of both functional and symptomatic responses are presented
for individuals with impaired respiratory systems in Tables V-4c and for healthy individuals
in Tables V-5c as interference with normal activity and as changes  in medical treatment
and/or self medication.   (See Tables 9-1 and  9-2 and the discussion in CD  starting on p. 9-23).
       It is staffs judgment that  responses of individuals with impaired respiratory systems.
categorized in Table V-4a as "large" for functional responses or categorized in Table V-4b  as
"severe" for symptomatic responses, would result in the potential for episodic or
incapacitating illness.  Those responses would include the more quantifiable responses such
as a significant increase in nonspecific bronchial responsiveness (i.e., dose is <25%  of
baseline), an increase in nonspecific airway resistance (SR.J of >200% or more than 15 cm
H20/s, and a decrease in FEV, of _>. 20% from baseline.  The less quantifiable, but
potentially incapacitating effects, to  the individual with impaired respiratory systems include
persistent wheeze, uncontrollable cough, severe discomfort on exercise or deep breath, and
 multiple bronchodilator usage giving only partial relief.  Since these "severe" symptomatic
 and "large" functional responses  for individuals with impaired respiratory systems could limit
 activity and increase  the likelihood of physician or emergency room (ER) visits as well as

-------
                                           66
hospital admissions for some affected individuals, staff recommends that they be
characterized as adverse.  Because "small" and "moderate"  functional response and  "mild"
and "moderate" symptomatic responses would not be likely  to result in impacts comparable
to episodic or incapacitating illness, they are discussed in the following section on
interference with normal activity.
       3.  Interference with Normal Activity
       For both healthy individuals and for individuals with impaired respiratory systems,
there has been a great deal of controversy regarding the  extent to which other acute
responses that have been associated with short-term and prolonged O3 exposures should be
considered adverse.  The previous  section contains a discussion  of functional effects
categorized as  "large" and  symptomatic effects as "severe"  for individuals with impaired
respiratory systems.   Those effects would be more likely to lead to episodic or incapacitating
illness in asthmatic individuals as discussed  above, whereas  the  "small" and "moderate"
functional effects and "mild" or "moderate" symptomatic effects discussed below are more
likely to be limited at most to interference with normal activity of either asthmatic or healthy
individuals.   "Moderate" functional and/or symptomatic responses in  either healthy or
asthmatic individuals  are most problematic with regard to judging adversity because they are
not serious enough to be clearly described as adverse but may still interfere with the ability
of some individuals to perform normal activity and, therefore, have the potential for
adversity in some  sensitive individuals.
       Asthmatic Individuals.  The response of asthmatic individuals  to O3 and other irritants
can be highly variable depending on the severity  of disease  in the individual.  A normal
range of change in specific airway  resistance (SR,,W) is within ±2Q% with little or no change
in nonspecific bronchial responsiveness.  Ozone-induced  decrements in FEV, of <3%,
increases in SR,a and nonspecific bronchial responsiveness approaching 100%, which last
less than 4 hr,  are categorized as "small" in Table V-4a.  Even  in conjunction with
symptomatic  effects categorized as  "mild" (i.e., wheeze with otherwise normal breathing,
cough with deep breath, and discomfort just noticeable on exercise or deep breath) lasting
less than 4 hr,  staff believes that these effects should not be considered adverse  for asthmatic
individuals.  Although a few individuals are likely to limit activity, these responses are not

-------
                                           67
likely to significantly interfere with the ability of most asthmatics to conduct normal activity
or change normal medication usage.
       Ozone-induced decrements in FEVj of > 10% but <20%, increases in SR,W of up to
200%, and increases in nonspecific bronchial responsiveness of up to 300%, which last for
>4 hr but <24 hr, have been categorized for asthmatic individuals  as "moderate" functional
responses  in Table V-4a.  These responses are typically accompanied by "moderate"
symptomatic responses, including wheeze with shortness of breath,  frequent spontaneous
cough, and marked discomfort on exercise or deep breath, which last for >4 hr but <24 hr.
Based on discussions with medical experts who  have worked with asthmatics, staff concluded
that single O3 exposure events which result in these responses are not likely to  interfere with
the normal activity of many asthmatic individuals nor to result in the increased frequency of
medication use or the use of additional medications.  Complete recovery could  result from a
single use of a bronchodilator.  However, because repeated exposure of asthmatics to O, over
periods of several days could result in exacerbation of the underlying inflammation and a
buildup of mucus in the respiratory system, medical experts who were consulted expressed
concern that the small airways, including the bronchioles and alveoli, may be more adversely
affected than effects induced by a single, acute  exposure.  Staff believes that multiple
exposures to O3 that induce repeated "moderate" responses in asthmatics could result in
increased  frequency or additional medication  usage, mucus buildup,  exacerbation of
inflammation, and an  increased likelihood of many asthmatic individuals to limit normal
activity. Therefore,  staff recommends that "moderate" functional and/or symptomatic
responses, when repeated,  should be considered to be adverse health effects. These health
endpoints are a matter of public health concern  in light of the increasing asthma morbidity
and mortality which has been occurring in the U.S. during the past decade.
       Healthy Individuals. During the previous  O3 NAAQS review, a wide range of
opinion was expressed regarding the adversity of lung function decrements, increases in the
severity of respiratory symptoms and increases  in nonspecific bronchial responsiveness  in
healthy individuals.  In particular, the focus of  debate was on the degree of response for
acute respiratory effects that should be considered adverse for purposes of setting  NAAQS.
A table was presented in the previous O3 Staff Paper (Table VII-5, p. VII-55, USEPA, 1989)

-------
                                          68
which categorized these acute respiratory responses of healthy individuals to O3 according to
type and severity of response.  Several specific aspects of such responses were characterized
including:  (1) the magnitude of lung function decrements on a test-specific basis (e.g.,
FEV,); (2) the presence of respiratory symptoms (e.g., cough, pain on deep inspiration,
shortness of breath);  (3) the duration of individual response; and (4) the extent to which
activity  is curtailed due to O3 exposures.
      At the December 1987 CAS AC meeting, some members of the CAS AC Ozone
Review  Panel expressed the belief that either limitation of activity or increased respiratory
symptoms could be considered the primary determinant of adversity, while  others believed
that the  more objective spirometry measurements were most appropriate. Some Panel
members felt that healthy individuals would experience adverse effects when O, exposure
induced  any of the responses categorized in the  1989  table as "moderate" (i.e., FEV,
decrement of 10-20%; mild to  moderate cough,  pain on deep inspiration, shortness of breath;
complete recovery in <6 hours;  and few sensitive individuals likely to discontinue activity).
Other Panel members believed that adverse effects would not result unless a healthy
individual encountered 03-induced effects categorized as severe (i.e., FEV,  decrement of 20-
40%; repeated cough, moderate to severe pain on deep inspiration  and breathing distress;
complete recovery in 24 hours, and  some sensitive individuals likely to discontinue activity).
      One of the Panel members pointed out at the December 1988 CASAC meeting that
because children report few, if any,  symptoms when exposed to O, concentrations likely to
induce symptoms in adults,  it may be inappropriate to recommend  that all categories  of
response be experienced by children before describing the effects as adverse. This is due to
concern that by not experiencing the "early warning signals"  (i.e.,  respiratory  symptoms)
children would be more likely to continue high levels of exertion during periods of exposure
to O3 levels that could potentially induce substantial pulmonary function changes and  repeated
acute inflammatory responses.  In written comments following the  March 1995 CASAC
meeting, another Panel member expressed the opinion that the lack of a symptom response
should not be considered a risk factor for children.  This divergence of opinion by  the  Panel
members regarding lack of a symptom response in children possibly introducing increased.

-------
                                          69
risk of exposure to O3 and resultant adverse consequences was discussed at the September
1995 CAS AC meeting.
       Taking into account both previous and current expert opinion, staff worked with
scientists in EPA's ORD to develop Tables V-5a, V-5b, and V-5c (Table 9-1 in the CD)
showing a gradation  of responses  to short-term O3 exposures for healthy individuals.
Consistent with ATS guidelines, current and past CASAC views, and the judgments of two
EPA Administrators  in the previous O3 NAAQS  rulemaking, staff recommends that
functional responses  categorized as "small" in Table V-5a not be considered adverse
respiratory effects for healthy individuals. Individual "small" responses to O3 exposures are
characterized by 3% to  <_ 10% decrements in spirometry and <  100%  increases in
nonspecific  bronchial responsiveness, which  last  less than 4 hours. These are often
accompanied by respiratory symptoms categorized in Table V-5b as "mild," such as cough
only during deep inspiration or during lung function tests.   "Small" functional responses and
"mild"  symptomatic  responses would not generally be considered  medically significant and
would not be expected to interfere with  normal activity of healthy individuals.
       Staff also recommends that any of the individual functional responses categorized as
"large" in Table V-5a or symptomatic responses  categorized as "severe" in Table V-5b
should be considered adverse respiratory effects, in  and of themselves,  for healthy
individuals.  Staff believes that such responses (e.g., FEV,  decrements  > 20%. increases in
nonspecific bronchial responsiveness >  300%, and  uncontrollable, persistent cough, and/or
chest pain  lasting 24 hours and longer) are medically significant under the ATS guidelines.
Such responses would likely cause many individuals to halt normal activities involving
physical exertion. Furthermore,  individuals experiencing such effects would most hkelv
judge that  they were being adversely affected at  least for the duration of the response.  As
discussed by Panel members at the March 1995  CASAC meeting, such effects might be
 similar to those experienced by an individual with acute bronchitis.  Staff believes that it is
 more likely that responses of this degree could be associated with exposures that may be
 linked to more serious, but not subjectively  noticeable, responses (e.g., respiratory
 inflammation, lung tissue damage) that individuals would not perceive were occurring.

-------
                                           70
        Tables V-5a, V-5b, and V5-c.  Gradation of Individual Responses to
                            Short-Term  Ozone Exposure in
                                    Healthy Persons'
Table V-5a
Functional Response
FEV,
Nonspecific
bronchial responsiveness'"
Duration of
response
None
Within normal
range (±3%)
Within normal
range
None
Small
Decrements of 3 %
to £10%
Increases of
<100%
<4hr
Moderate
Decrements of
>10% but <20%
Increases of
<300%
>4 but <24 hr
Large
Decrements of
£20%
Increases of
>300%
>24 hr
 Table V-5b
Symptomatic Response
Cough
Chesl pain
Duration of response
Normal
Infrequent
cough
None
None
Mild
Cough with
breath

deep
Discomfort just
noticeable on
exercise or deep
breath
<4hr
Moderate
Frequent
spontaneous

cough
Marked discomfort
on exercise or deep
breath
>4 bui <24 hr
Severe
Persistent
uncontrollable
cough


Severe discomfort
on exercise or deep
breath
>24 hr

 Table V-5c
Impact of Various
Functional and/or
Symptomatic Responses
Interference with normal
activity
Normal
Functional and/or
Symptomatic
Responses
None
Small Functional
and/or Mild
Symptomatic
Responses
None
Moderate
Functional and/or
Symptomatic
Responses
A fev, sensitive
individuals likely to
limit activity
Large Functional
and/or Sex ere
Symptomatic
Responses
Many sensitive
individuals likely
to limit activity
'See text for discussion, abbreviations and acronyms.
bAn increase in nonspecific bronchial responsiveness of 100% is equivalent to a 50% decrease in PDX or PD,,
(see Chapter 7, Section 7.2.3 of the CD for a more complete discussion).

-------
                                           71
       Establishing specific staff recommendations with regard to effects in the "moderate"
categories in Tables V-5a,  V-5b, and V-5c is more problematic.  The effects in this category
could interfere with the activities of a few sensitive, healthy individuals, particularly for the
symptom responses alone or when symptoms are accompanied by lung function decrements.
Those sensitive individuals who experience a combination of "moderate" functional responses
(i.e., lung function loss of > 10%  but  <20% and increased nonspecific bronchial
responsiveness of  <300%  lasting from 4 to 24 hr) accompanied by "moderate" respiratory
symptoms (i.e.,  marked discomfort and frequent cough persisting from 4 up to 24  hr) are
likely to limit activity and may perceive that  they had been affected adversely.  It is unlikely.
however, that these individuals would seek medical treatment or use self medication.
       Lung function decrements  at the "moderate" level, which may be a more likely
response in  children, may not be noticed  by the individuals affected due to a lack of
respiratory symptoms.  In such cases, the extent to which such responses should be judged as
adverse may depend on the likelihood that exposures causing such  moderate decreases in lung
function are associated with more serious effects, as to which there is substantial uncertainty
as discussed above.  A  further complication is that the likelihood of such exposures is related
to the attenuation of effects that is typically observed after repeated exposures.  For example.
it is well established that lung function and symptom responses in individuals exposed to O:,
on  consecutive days will attenuate until absent (CD, Sec. 7.2.1.4).  Most individuals initial!)
experience larger decrements in FEV,  on the second day but by  the third or fourth  da\ will
experience  disappearance of FEV, decrements when exposed to O3.  This attenuation of
response can last for as much as 1 to 2 weeks,  thus reducing the self-protective behavior that
might otherwise tend to limit ongoing exposure. Without respirator)1  symptoms or altered
lung function  as "early  warning signals," some individuals may be more likely to expose
their lungs repeatedly to O3 levels potentially associated with more serious effects, such as
pulmonary inflammation, although, as discussed above, the extent  to which these more
serious effects are linked to "moderate" lung function changes is uncertain.
        During the March 1995 CASAC meeting, there was considerable discussion regarding
the adverse nature of these acute  "moderate" health effects of Ov  One of the Panel members
indicated that experiencing these effects on a single occasion might be considered by the

-------
                                           72
individual to be a "nuisance."  There also would be little likelihood of attenuation of
response following a single O3 exposure.  However, if these same exposures were repeated
on multiple occasions they might become a matter of public health concern, particularly if
large segments of the population experienced "moderate" effects repeatedly.  In written
comments by one of the Panel members following the March 1995 CAS AC meeting, it was
suggested that "moderate" symptoms (e.g., frequent spontaneous cough) represent significant
inflammation of airways which is an important indicator of simple chronic bronchitis, and
therefore should be considered adverse only if they occur repeatedly.  Similarly, with regard
to marked discomfort on exercise or FVC test, this category should be considered adverse if
repeated but not for a single event.   These comments underscore the consistency of CASAC
opinion that single,  acute, Cyinduced health effects described above as "moderate"  for
healthy individuals should not be considered adverse.  However, as one Panel  member stated
in written comments,  "a series of peaks could well set  the stage for serious illness."  Because
there appears to be a greater consensus of opinion regarding the adverse nature of repeated
health effects of multiple O3 exposure,  it is the recommendation of staff that the number of
O3 exposures resulting in "moderate" health effects should be considered as a factor in
characterizing adversity for healthy  individuals.  EPA staff are concerned that  multiple
exposures to O3 could induce adverse effects in healthy individuals if they  are  particularly
sensitive and could result in limitation activity or self medication due to G, exposure.  In
summary, the degree of adversity of repeated "moderate" responses in healthy individuals is
likely to increase with  the increasing number of occurrences and with the combination of
different responses.

-------
                                          73
G.     Ozone Exposure Analysis
       1.     Overview
        This section discusses a series of analyses designed to provide estimates of O3
exposure for the general population and two subpopulations (i.e., "outdoor workers" and
"outdoor children") living in 9 U.S. urban areas under the conditions that various alternative
1- and 8-hour, 1- and 5-expected exceedance NAAQS are just attained.  To provide some
perspective, exposure estimates are also provided for a recent year (either  1990 or 1991) for
each of the 9 urban areas.  The exposure estimates summarized in this section  also are an
important input to the "headcount" health risk assessment described  in Section  V.H.
       The regulatory scenarios examined in the exposure analysis are limited  to 1- and 5-
expected exceedance alternative standards and are  based on use of a single year of data.
However, one can use these estimates to roughly bound the exposure and health risks for
other forms of the standard under consideration (e.g., average of the 2nd daily maximum 8-
hr average over a 3-year period)  by using air quality analyses that compare alternative forms
of the primary standard.  In analyzing the exposures and health risks for any of the forms of
the standard that are based on an average concentration  or expected  number of exceedances
over a multiple year period, including the current  1-expected  exceedance. 1-hr standard,  the
exposure and risk estimates reflect what would be expected in a typical or  average year in  an
area just attaining a given standard.  An area just attaining a  standard might have annual
exposures and health risks somewhat  lower or higher than the average estimates over the
multiple year period used to define attainment of the standard.
       Figure V-5 illustrates the  various  components of the exposure model and how the
exposure assessment relates to  risk assessment.  Four versions of the probabilistic NAAQS
exposure model for O3 (pNEM/O3) were used to estimate population exposure under
alternative 1- and 8-hr standards.  The pNEM/O3 exposure model builds on earlier
deterministic versions of NEM by modeling random  processes within the exposure
simulation. A brief summary of  the pNEM/O3 model is provided below.  A more detailed
description of pNEM/03 and its application to the  general population, outdoor workers,
outdoor children,  and  a single summer camp in California can be found elsewhere in a

-------
                                74

FIGURE V-5. MAJOR COMPONENTS OF PNEM/O3 MODEL AND ASSOCIATED

               HEALTH RISK ASSESSMENT PROCEDURES
Dose - Response
Relationships


k
>
Exposure
Modeling
j
^ 1 Micro-
k

*
A
k.


indoor Air Factors
^
r
Risl; Modeling

k I

r]
                   environmental
                   Data
                            w~i
and
Parameter

-------
                                            75
collection of exposure support documents (Johnson, 1994; Johnson et al., 1996a,b,c;
McCurdy, 1994a).  The pNEM/O3 model has been designed to take  into account the most
significant factors contributing to total human O3 exposure.  These factors include the
temporal and spatial distribution of people and O3 concentrations throughout an urban area,
the variation of O3 levels within each microenvironment, and the effects of exertion
(increased ventilation) on O3 uptake in exposed individuals.
       Three versions of the pNEM/O3 model have been run (general population, outdoor
workers,  and outdoor children) and applied to the same nine major urban areas.  The fourth
version was  applied to a specific summer camp in Pine Springs, California  (see Johnson,
1994).  The  nine urban areas used in the general population, outdoor worker, and outdoor
children versions of the model vary greatly in geographical location, O3 "design  value",4
population size  (both modeled and total MSA), and number of exposure districts included.
The areas were selected to obtain as widely representative a modeling domain as possible
given the overall need  for monitoring data completeness in an area.
       For instance,  urban study area populations modeled vary from Denver, with a
population of 1.5 million, to the New York area, with a population  of about 10.7 million
people. Information about the study area population, number of exposure districts, year and
O3 season modeled, and summary air quality statistics  for the nine study areas are presented
in Table  V-6.
       The total population included in the 9 urban study areas covered by  the exposure
analysis is 41.7 million people.  Given the considerable additional uncertainty that would be
introduced,  OAQPS  has chosen not to extrapolate the exposure estimates from the 9 urban
areas to obtain  national exposure estimates.  The 9  urban areas represent a  significant
fraction of the  U.S.  urban population and include the largest areas with major O3
nonattainment problems (e.g., Los Angeles,  Chicago,  New York, and Houston).
 4 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 O3 NAAQS formulation of _<_ 1 expected exceedances of 0.12 ppm
 daily maximum 1-hour average. The design value shown in Table V-l is the second-highest 1-hour cl.iih  maximum
 concentration in the O3 air quality data base for the year modeled.

-------
          TABLE V-6. CHARACTERISTICS OF OZONE STUDY AREAS USED IN PNEM/O, ANALYSES
Study Area




Chicago
Denver
Houston
Los Angeles
Miami
New York City
Philadelphia
St. Louis
Washington,
D.C.
1990 Population
in Study Area
(millions)


6.2
1.5
2.4
10.4
1.9
10.7
3.8
1.7
3.1

1990
MSA or
CMS A
Population
(millions)
8.1
3.8
3.7
13.8
1.9
18
6
2.4
3.9

Number of
Exposure
Districts


12
7
11
16
6
12
10
11
11

Exposure Period

Year Months


1991
1990
1990
1991
1991
1991
1991
1990
1991

Apr-Oct
Mar-Sep
Jan -Dec
Jan -Dec
Jan -Dec
Apr-Oct
Apr-Oct
Apr-Oct
Apr-Oct

Daily Max.
Hourly Design
Value (ppm)'


0.129
0.115
0.23
0.31
0.123
0.175
0.156
0.13
0.174

"The design value listed here is the second-highest  1 hour daily maximum concentration in the O3 air quality data base for the
year modeled.

-------
                                           77
       2.      Exposure Modeling Methodology
       The pNEM/O3 model consists of two principal parts, the cohort exposure program
and the exposure extrapolation program.  The cohort exposure program estimates the
sequence of O3 exposures experienced by defined population groups.  The exposure
extrapolation program estimates the number of persons within a particular study that are
represented by each cohort and then combines this information with cohort exposure
sequences to estimate the distribution of exposures over a defined population of interest.
       The pNEM/O, methodology consists of the following five  steps:
       (1)     define a study area,  a population  of interest,  appropriate subdivisions of the
              study area, and  an exposure period,
       (2)     divide the population of interest into an exhaustive  set  of cohorts,
       (3)     develop an exposure event sequence for each cohort for the exposure period,
       (4)     estimate the pollutant concentration and ventilation  rate associated with each
              exposure event, and
       (5)     extrapolate cohort exposures to the population of interest.

Each of these steps is described in more detail in the following discussion.
       Define the Study Area. Subdivisions of the Study Area,  the Exposure Period and the
Population of Interest.  The study area is defined as an aggregation of exposure districts.
Each exposure district is defined as a contiguous set of geographical  census units (GCU).
Each GCU consists of one or more census  tracts as defined by the 1990 census.  All GCUs
assigned to a particular exposure district are located within a specified radius (15 km) of a
fixed-site O3 monitor.
       As indicated previously, the nine urban areas  used in the general population, outdoor
worker, and outdoor children versions of the model vary greatly in geographical location,  O3
"design value", population size (both modeled and total MSA), and number of exposure
districts included. Each urban area was divided into large exposure  districts, varying from 6
to 16 in the nine areas modeled, corresponding  to the number of air  quality monitors having
valid air quality data in a study area.  Most of the urban areas had 10 or more districts

-------
                                           78
within their boundaries.  The study areas included urban and suburban counties next to the
central city for which the study areas are named.  To illustrate this point, Figure V-6 shows
the New York consolidated metropolitan statistical area and the monitors used in  the New
York study area for purposes of the pNEM/O3 analysis are indicated by the circles on the
map.
       From 3 to  16 monitoring sites were selected  to represent the spatial variation of O3
levels in each of the 9 study areas.  The number of  monitors chosen for each area depends
upon a data completeness criterion (i.e., data are at  least 75% complete) for each monitor
                                              •
and the availability of home-to-work commuting data for a district. For Los Angeles, 30
possible monitors were pared down to 16 because of limitations on computational resources.
This paring down  was accomplished by removing one of nearby pairs of mojiitors that has
similar O3  air quality distributions.  In the New York study area, one site, the World Trade
Center, was removed because the monitor is placed  on the top  of the building and is not
considered representative for population exposure.  The nearest New York City monitor is
used to represent the World Trade Center district. Otherwise,  all  available  monitors meeting
the above criteria were used in each of the  study areas.
       The exposure period is defined as a series of months within a particular calendar year
corresponding to the designated O3 monitoring season specified for the urban area by the
U.S.  EPA.  For six of the nine urban areas the season is nine months long,  while three areas
(Los Angeles, Houston, and Miami) have a 12 month season.
       The CD identifies outdoor workers and children as  two  population groups particularly
at risk for  experiencing O3-related health effects. These two groups were identified based on
the increased time they spend outdoors engaged in moderate and heavy exertion which
increases the likelihood of experiencing O3-induced health effects.  While children and
workers were included in  the general population version of pNEM/O3,  EPA analysts felt that
the procedures used did not adequately represent exposures for workers or children that
spend considerable time outdoors on a regular basis.  Therefore, special versions  of
pNEM/O3 were developed to estimate population exposures for outdoor workers and outdoor
children.  Table V-7 lists the 1990 population estimates for the general population, outdoor
workers, and outdoor children in each of the nine urban areas.

-------
                        79
FIGURE V-6. New York Urban Area Monitoring Sites Used in pNEM Analyses.
          NY
    PA
       NJ
                            = Monitoring Site
                         Population in 1990
                          pNEM Study Area
                          = 10,660,000

                          NY CMSA
                          = 17,950,000

-------
                            80
TABLE V-7. POPULATION ESTIMATES FOR OZONE STUDY AREAS USED IN
                     pNEM/O3 ANALYSES
Study Area
Chicago
Denver
Houston
Los Angeles
Miami
New York City
Philadelphia
St. Louis
Washington,
D.C.
1990
General
Population in
Study Area
(millions)
6.2
1.5
2.4
10.4
1.9
10.7
3.8
1.7
3.1
1990
Outdoor
Worker
Population in
Study Area
(thousands)
141
36
72
294
47
196
99
41
76
1990
Outdoor
Children
Population
in Study
Area
(thousands)
473
107
201
798
133
783
275
128
199

-------
                                           81
       Divide the Population-of-Interest Into an Exhaustive Set of Cohorts.  The population
of interest, in each version of the model, is divided into a  set of cohorts such that each
person is assigned to only one cohort. Each cohort is assumed to contain persons with
identical exposures during the specified exposure period.  Cohort exposure is typically
assumed to be a function of (1) demographic group, (2) location of residence, and (3)
location of work place. Specifying the home and work district of each cohort provides a
means of linking cohort exposure to ambient pollutant concentrations.  Specifying the
demographic groups provides a means of linking cohort exposure to activity patterns that
vary with age, work status,  and other demographic variables.
       Because both the intake dose received and susceptibility to O3 health effects may vary
with age,  occupation, and intensity of exertion, the total population of each study area was
divided into 9 age-occupation (A-O) groups. Each A-O group was further subdivided into
cohorts depending upon (1)  the type of air conditioning system present in the home, if any,
(2) home district, and  (3) work district.
       Develop  an Exposure Event Sequence for Each Cohort for the  Exposure Period.  The
exposure of each cohort is determined by an exposure event  sequence  (EES) specific to  the
cohort. Each EES consists  of a series of events with durations from 1  to 60 minutes.  To
determine average exposures for specific clock hours, exposure events are defined such  that
no event falls within more than one clock hour.  Each exposure event assigns the cohort to a
particular combination of geographic area and  microenvironment.  Each event also provides
an indication of breathing rate.  The breathing  rates are classified  as sleeping, slow, medium,
and fast.
       EESs are determined by assembling activity diary records relating to  individual 24-
hour periods into a series of records spanning the O3 season  of the study area.  Because each
subject of the activity  pattern diary studies provides data for one to three days,  the
construction of a multi-month EES requires either repetition  of data from one subject or the
use of data from multiple subjects.  The latter  approach is used in all three PNEM versions
discussed here.
       The use  of activity data from multiple persons to construct a multi-month EES for
each cohort is believed to better represent  the variability of exposure that is expected to

-------
                                          82
occur among persons included in each cohort.  However, the fact that a multi-month EES is
constructed by using data from multiple subjects  means there is more uncertainty in the
persons exposure measure and, in particular, in the estimates of how many times any
individual is exposed to a given concentration.  The PNEM/O3 model probably
underestimates the frequency of exposures for those individuals in the population that engage
in moderate or heavy exertion on a regular basis.
       For the general population version  of the  model, activity diary data were obtained
from the Cincinnati Activity Diary Study (CADS) (Johnson, 1987).  The CADS data base
includes over 900 subjects who completed three-day activity diaries.
       For the outdoor worker version of  PNEM/O3, additional data from six other
time/activity studies were combined with the CADS database and processed to provide a
unified time/activity database representative of outdoor workers.  These studies are
summarized in Table V-8.  The activity data selected to represent outdoor workers were
based on selecting data from  subjects that  spent at least four hours at work and spent at least
50 percent of their work time outdoors. The final pool contained 89 outdoor workers with
136 person-days of diary data. City-specific outdoor worker estimates were derived based on
city-specific 1990 Census data and judgments by  a panel of researchers about the percentage
of outdoor workers in each of 37 Census occupation groups.   Section 6 of Johnson et al.
(1996c) provides a detailed description of  the procedures used to  develop the outdoor worker
time/activity data base  and population extrapolation.
       For the outdoor children version of PNEM/O3, additional  data from six other
time/activity studies were combined with the CADS database and processed to provide a
unified time/activity database representative of outdoor children.  These studies are
summarized in Table V-9.  The pool of activity patterns used  to  represent outdoor children
was based on selecting children that met the following conditions:
       (1)    during a "non-summer" weekday the child had at least one diary day where
             he/she spent two hours or more outdoors, or
       (2)    during a "non-summer" weekend the child had at least one diary day where
             he/she spent three hours or  more outdoors, or

-------
TABLE V-8. CHARACTERISTICS OF HUMAN ACTIVITY STUDIES USED IN OUTDOOR WORKER EXPOSURE

                                     ANALYSIS.
Database name
California - 12
and over
Cincinnati
Denver
Los Angeles -
construction
Los Angeles -
outdoor worker
Valdc/
Washington
Total
Reference
Wiley ct a).,
1991b
Johnson, 1987
Johnson, 1984
Linn et al.,
1993
Shamoo et al ,
1991
Goldstein cl
al , 1992
Hartwell ct al ,
1984

Characteristics
of subjects
Apes 12 to 94
Ages 0 to 86
Ages 18 to 70
Construction
workers
(ages 23 (o 42)
Adult outdoor
workers
(ages 19 to 50)
Ages 10 to 72
Ages 18 to 70

Number of
subject -
days in
original
study
1762
2800
859
19
60
405
705

Number of
subject-days
used in outdoor
worker PNEM
analysis
156
105
41
19
29
25
33
408
Study
calendar
periods
Oct. 1987 -
July 1988
March and
August 1985
Nov. 1982 -
Feb. 1983
July - Nov.
1991
Summer 1989
Nov 1990-
Oet 1991
Nov. 1982 -
Fcb 1983

Diary type
Retrospective
Real-time"
Real-time
Real-time'
Real-time*
Retrospective
Real-time

Diary time
period
Midnight to
midnight
Midnight to
midnight
7 p.m to
7 p.m.
(nominal)
Subject wakeup
to subject
returns home
from work
Midnight to
midnight
Varying
24-h period
7pm. to
7 p.m.
(nominal)

Breathing
rates
reported?
No
Yes
No
Yes
Yes
No
No

"Study employed the Cincinnati diary lonnat
                                                                                     oo
                                                                                     UJ

-------
     TABLE V-9. CHARACTERISTICS OF HUMAN ACTIVITY STUDIES PROVIDING DATA FOR OUTDOOR CHILDREN EXPOSURE ANALYSIS.


Database
name

California - 1 1
and under
California - 12
and over
Cincinnati

Los Angeles -
elementary
school
Los Angeles -
high school

Valdcz

Washington


Total

Reference



Wiley et al ,
1991a
Wiley et al.,
199lh
Johnson, 1987

Spier et al.,
1992; Linn et
al., 1992
Spier el al ,
1992; Linn ct
al . 1992
Goldstein, ct al ,
1992
Hartwcll ct
al., 1984



Characlcrislics
of subjects


Children ages 1 to
11
Ages 12 to 94

Ages 0 to 86

Elementary school
students, 10 to 12
years
High school
students, 13 to 17
years
Ages 10 to 72

Ages 18 to 70



Number of
subject-
days in original
study

1200

1762

2800

58


66


405

705



Number of
subject days used
in outdoor
children pNEM
analysis
257

54

384

38


47


9

3


792
Study
calendar
periods


April 1989 -
Feb. 1990
Oct 1987-
July 1988
March and
August 1985
Oct 1989


Sept and Oct.
1990

Nov 1990 -
Oct 1991
Nov 1982 -
Fcb 1983




Diary type


Retrospective

Retrospective

Real-time

Real-time"


Real-time-


Re! respective

Real-time




Diary time
period


Midnight to
midnight
Midnight to
midnight
Midnight to
midnight
Midnight to
midnight

Midnight to
midnight

Retrospect!
ve
7 p m. to 7
p.m.
(nominal)

Breathing
rates
reported7


No

No

Yes

Yes


Yes


No

No



                                                                                                                           00
"Study employed the Cincinnati diary formal.

-------
                                           85
       (3)    during a "summer" weekday or weekend the child had at least one diary day
             where he/she spent 4 l/i hours or more outdoors.
       For this analysis "summer" was defined  as June through August and "nonsummer" as
all other months. This procedure produced a pool containing 479 outdoor children with  792
person-days of activity diary data.  Outdoor children included in the analysis were in 2
demographic groups: children ages 6 to  13 ("preteenagers") and children ages 14 to 18
("teenagers").  The city-specific percentages of  outdoor children were derived based on city-
specific 1990 Census data for the two demographic groups and the percentages of outdoor
pre-teenager and teenager subjects in three of the time/activity studies conducted in
Cincinnati and  California (Johnson, 1987; Wiley et ah, 1991a,b) that employed  a random
selection procedure to enroll subjects. About 47 percent of preteens and 31 percent of
teenagers were judged to meet the selection criteria for outdoor children.
       A distinct EES is developed for each cohort. The exposure event within an EES  is
defined by  the district, the  microenvironment, and the breathing rate associated  with the
activity being undertaken by the sampled individual.
       The district is defined as being either the home or work district associated with the
cohort.  For children, it is  assumed that their school district is the same as their home
district.  Population movement  in pNEM/O3 is based upon information gathered by the U.S.
Census Bureau regarding householders'  home-work commuting patterns (Bureau of the
Census,  1990).  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
pNEM/Oj analysis to obtain district-to-district trip information for those cohorts that work.
Otherwise, cohorts are assumed to stay in their  home districts.
       The seven microenvironments used in all three versions of pNEM  are: 1) indoors-
residence with  a central air conditioning system, 2) indoors-residence with window air
conditioning units, 3) indoors-residence  with no air conditioning system, 4) indoors-
nonresidential locations, 5) outdoors near a road, 6) outdoors - other locations, and 7) in-
vehicle.

-------
                                          86
       Estimate the Pollutant Concentration and Ventilation Rate Associated With Each
Exposure Event.  Pollutant concentrations associated with each exposure event depend on the
03 air quality within each exposure district, which is estimated using ambient  data from
monitoring sites in each exposure district, and the microenvironment where the event
occurred.  The general population,  outdoor worker, and outdoor children versions of
pNEM/O3 examine nine air quality  scenarios.  All of the regulatory scenarios  are on a daily
maximum basis with either  1 or 5 expected exceedances allowed per year. The  scenarios
(and a  short-hand label for each listed in parentheses) are:
       (1)     1990 or  1991 air quality-the "as is" or baseline scenario (As Is);
       (2)    Just attaining a 1-hr, 0.12 ppm, 1 expected exceedance standard-the current
             standard (1H1EX-0.12);
       (3)    Just attaining a 1-hr, 0.10 ppm, 1 expected exceedance standard (1H1EX-
             0.10);
       (4)    Just attaining an 8-hr, 0.10 ppm, 1 expected exceedance standard  (8H1EX-
             0.10);
       (5)    Just attaining an 8-hr, 0.09 ppm, 1 expected exceedance standard  (8H1EX-
             0.09);
       (6)    Just attaining an 8-hr, 0.08 ppm, 1 expected exceedance standard  (8H1EX-
             0.08);
       (7)    Just attaining an 8-hr, 0.07 ppm, 1 expected exceedance standard  (8H1EX-
             0.07);
       (8)    Just attaining an 8-hr, 0.09 ppm, 5 expected exceedances standard (8H5EX-
             0.09); and
       (9)    Just attaining an 8-hr, 0.08 ppm, 5 expected exceedances standard (8H5EX-
             0.08).
       For all of the indoor and in-vehicle microenvironments the season-long sequence of
hourly O3 values is estimated using a mass balance algorithm.  The  mass-balance model used
in pNEM/O3 is a simplified version of the generalized Nagda,  Rector, and Koontz model
(Nagda et al., 1987).  This  model was revised to incorporate the assumption that indoor
decay rate is proportional to indoor O3 concentration. This algorithm estimates  the hourly

-------
                                           87
average indoor O3 concentrations during hour h as a function of:  indoor O3 concentration at
the end of the preceding hour, the outdoor O3 concentration during hour h, air exchange rate
during hour h, and an O3 decay rate.  Values for the air exchange rate and the O3 decay
factor are sampled from an appropriate distribution, based on the available scientific
literature on these parameters, on a daily basis. Air exchange rate is permitted to change
hourly in the three residential microenvironments depending  on  whether windows are
assigned a status of open or closed. This assignment is determined through use of a
probabilistic model in which the status during each clock hour is assumed to be a function of
air conditioning system temperature range and window status during the previous clock hour.
In each of the pNEM/O3 simulations,  the O3 concentration in a particular microenvironmerit
during a particular clock hour is assumed to be constant.
       For the two outdoor microenvironments and as an input to  the mass balance algorithm
for the indoor and in-vehicle microenvironments, representative ambient air quality data is
required for each district in the form of a time series of hourly values for the specified O,
season. The outdoor O3 concentration associated with microenvironment  m in district d
during hour h was determined by an expression having the general form
                    Co,,, (m,d,t,s)  =  1.056 x Cmon (d,t,s) + e (t),     (equation V-l)
where C^, (m,d,t,s)  is the outdoor (or ambient) O3 concentration in microenvironment m in
exposure district d at time t under regulatory scenario s, Cmon (d,t,s) is the O, concentration
estimated  to occur at the monitor representing district d at time  t under regulatory scenario s,
and e (t) is a random normal variable  with mean = 0 and standard deviation  = 0.0053 ppm.
The factor of 1.056 and the value of the standard  deviation for e (t) were derived based on
regression analyses relating personal exposure data and fixed site monitors obtained from the
Houston Asthmatic Oxidant Study (Stock et al., 1985). The  derivation of these parameters is
described in more detail in Chapter 2  of Johnson et al. (1996b,c).
       To represent  ambient O3 air quality concentrations in  the nine  urban areas, monitored
values are adjusted mathematically to  represent a future regulatory scenario (s) when  air
quality in the study area just meets the O3 NAAQS being analyzed.  It should be recognized
that we are not concerned in our exposure analyses about how  or  when an alternative O3
NAAQS is attained.  That is the concern of other analyses which OAQPS and other EPA

-------
offices undertake: especially the regulatory and benefits analyses.  For the O3 exposure
analyses conducted to support decisions on the NAAQS; it is sufficient to simulate the just-
attaining situation without being concerned about how, when, or even if that situation will
occur.
       By definition, a NAAQS is attained when all monitors in an area have less than one
(or five for some of the alternative  8-hr standards analyzed) expected exceedance of the
standard concentration value (e.g., 0.12 ppm for the current 1-hr standard) in a year.  The
exposure analysis is based on  a "just attains" scenario, where air quality levels at the monitor
currently having the highest number of expected exceedances are reduced mathematically to
where that monitor just  attains the standard being analyzed. The adjustment procedure used
for six of the nine urban areas 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.
       The adjustment procedures were developed by comparing the O, data reported  by a
site in a high year with  O? data reported by the same site in a low O3 year.  Therefore, these
procedures are expected to perform best when used to simulate a significant reduction in the
O3 levels at a site.  These procedures may produce unrealistic data sets for areas that involve
either a small reduction or an increase in O,  levels  to simulate just attaining certain
regulatory scenarios.  Therefore, for three of the urban areas (all regulatory scenarios for
Miami, Denver, and Chicago) a simpler adjustment procedure involving proportional rollback
(or rollup in some cases) was used  because the design values for the baseline year (i.e.,  1990
or 1991) were relatively close to or, in some cases, even below the levels required to just
attain the  alternative standards being examined.  For more information regarding the air
quality adjustment procedure used to simulate a just-attaining situation see Chapter 5 in
Johnson et al. (1996b,c).
       An analysis evaluating the air quality  adjustment procedure used to simulate
attainment conditions  was recently completed (Johnson, 1995).   This analysis examined six of
the nine urban areas (Chicago, Washington, D.C.,  Houston, Los Angeles, New York, and
Philadelphia). O3 levels for the baseline ("as is") year used in  the pNEM analyses (either

-------
                                           89
1990 or 1991) were adjusted using the same procedures as in the pNEM analyses to just meet
the levels in a "lower"  year (using data from the 1992 to 1994 time period).  Comparisons
were then made between the adjusted data set and the observed data set for the "lower" year
for each site in these six urban areas at various cutpoints of the  1-hr and 8-hr air quality
distributions (50, 90, 95, 99, 99.5, 99.75 percentiles, and sixth largest value, second largest
value,  and largest value). The conclusion was that "the air quality adjustment procedures
perform adequately in  the upper-tail region (90th percentile and  above) of the distribution,
the region that determines the O3 exposures of most concern in pNEM/ analyses" (Johnson,
1995).  The adjustment procedures tended to overestimate  1-hr O3 concentrations around the
50th percentile, however, because O3 data follows a skewed distribution, the midrange value
is typically  closer to the 90th percentile than the 50th percentile. This limited evaluation
indicates that the air quality adjustment procedures  used in pNEM/O3 do reasonably mimic
changes in O3 levels that have occurred in the past  in the  six urban areas evaluated.  This
does not guarantee that the adjustment procedure will accurately reflect future changes which
depend on several variables including how controls influence the VOC/NOX ratio, spatial
patterns of  growth, and  transport of pollutants from other urban areas.
       Because dose received by a person exposed to an air pollutant is highly dependent
upon ventilation rate,  exertion level is an important consideration in exposure modeling.  The
exposure  model includes an algorithm that assigns the equivalent ventilation  rate (EVR)
associated with each exposure event.  Clinical research by EPA suggests that there is less
variability in EVR than  in ventilation rate for a given  level of exertion.  The outdoor
children version of pNEM/O3 employs an EVR-generator module that uses one of  four
 Monte Carlo models to  generate an EVR value for each exposure event associated  with  a
 given cohort.  The Monte Carlo models were developed through an analysis of data from two
 studies that measured  heart rate of elementary and high school students while engaged in
 various typical daily activities (Spier et al., 1992; Linn et al., 1992).  These studies then
 measured heart rate and ventilation rates simultaneously in a clinical setting  to obtain a
 "calibration curve" for each subject relating heart rate to ventilation rate on  a minute-by-
 minute basis.  A regression  analysis was applied to the Spier and Linn data  bases to relate

-------
                                           90
various predictor variables contained in the diary data (e.g., daytime activities, day of week,
breathing rate) that were considered likely to influence the EVR values.
       The EVR-generator module also contained an algorithm which established an upper
limit for the EVR value assigned to each exposure event.  This limit was set at a level
estimated to be achievable by members of the cohort who (1) exercised regularly, (2) were
motivated to attain high exertion levels, and (3) were not professional athletes.  Individuals
engaged in amateur sports (e.g., jogging, tennis) would be included, but professional athletes
would  not be included.  The EVR limit was derived based on a subset of the specified cohort
(e.g., data from males aged 11  were used for the preteen group aged 6-13 and data from
males aged 15 were used  for the teenage group aged 14-18) and applied to the all members
of the demographic group (either preteens or teenagers).  Because the selected  subset used in
each case is likely to be higher than the average EVR limit for each demographic group, the
pNEM/O3 simulation will tend to overpredict the occurrence of high EVR values within each
demographic group.
       Extrapolate the Cohort Exposures to the Population of Interest.  The cohort-specific
exposure estimates were extrapolated to the general population, outdoor children, and
outdoor workers by estimating the size of each cohort based on 1990 Bureau of Census
(Bureau of Census, 1990) data files that list population data for age groups by  census unit.
The population in each census  unit was multiplied by an air conditioning fraction (based on
1980 Bureau of Census data) in the specific census unit to provide an estimate of the number
of outdoor children (or outdoor workers) in each air conditioning category. These air
conditioning specific values were then  summed over each exposure district to derive
estimates for the entire study area.
       3.     Population Exposure Estimates Upon Attainment of Alternative Ozone
              Standards
       The pNEM/O3 contains a large  number of stochastic variables and, therefore,
exposure estimates will vary from run to run.  For the general population, outdoor worker,
and outdoor children exposure analyses, 10 simulations of pNEM were done for each
regulatory scenario in each of the 9 urban areas to  better characterize the uncertainty in  the
exposure estimates. Based on a previous analysis of sets of 10-run results versus a 108-run

-------
                                          91
result, McCurdy (1994b) has shown that results from only 10 runs of the model adequately
predict the mean and variance observed in  100 or more runs of pNEM/O3.  Additional runs
of the model would, however, increase the range of possible outcomes, but limited resources
preclude undertaking more runs.
      In any pNEM analysis, several different indicators are used to estimate exposure of
people to various levels of air pollution.  One 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.  Another indicator
is "occurrences of exposure:" the number of times a given level of pollution is experienced
by the population of interest.
      The exposure model provides exposure estimates in terms of both highest
concentrations (exposures) or highest dose.  The exposure estimates summarized here pertain
to "daily maximum dose (MAXD)," where dose is defined as the product of O3 concentration
and ventilation  rate over a defined period.  The daily maximum dose  does not necessarily
occur during the time period of maximum O3 concentration in a given urban area.  The dail\
maximum dose indicator was selected because it is a better surrogate  for the number of 03
molecules that enter the oral-nasal cavities per unit time period, and therefore,  is more likely
to be more relevant from a health  risk viewpoint than maximum exposure.
       It should be stressed that the exposure model produces exposure estimates for the
entire range of concentrations and that the health risk assessment,  described in  the next
section, makes  use of all exposures at a given exertion level that exceed an estimated
background  level of 0.04 ppm.  Figure V-7 shows the exposure distributions for outdoor
children living  in the Philadelphia area experiencing daily maximum dose 8-hr  exposures on
one or more days while engaged in moderate exertion (EVR in the range 13-27 1/min-m2).
Attaining any of the alternative standards reduces the number of children experiencing daily
maximum dose 8-hr exposures exceeding 0.1 ppm  to less than 2,000 persons.  At  the 0.08
ppm level the number of children  estimated to be exposed ranges from near 0 for  the 0.07
ppm, 1  expected exceedance, 8-hr standard  (8H1EX-0.07) to about 69,000 children for the
0.10 ppm, 1 expected exceedance,  1-hr standard (8H1EX-0.10).  And at the 0.06  ppm level
the number  of children estimated to be exposed under moderate exertion for an  8-hr average

-------
                                       92
   FIGURE V-7. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
  OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
          EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA
   300
                                                         ASIS

                                                       1H1EX-0.12

                                                       8H1EX-0.09

                                                       8H1EX-0.08
                                                          -B-
                                                       8H1EX-0.10
                                                          -e-
                                                       8H5EX-0.08
                                                       8H1EX-0.07

                                                       1H1EX-0.10

                                                       8H5EX-0.09
      i   0.02  0.04  0.06  0.08   0.1  0.12  0.14  0.16  0.18  0.2
                     CONCENTRATION, PPM

    FIGURE V-8. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
          EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA
    16,000
~10,000
HI

z  8,000
a:

g  6,000
O
§  4,000
O
V)
8]  2,000
a.
                                                           ASIS

                                                        1H1EX-0.12

                                                        8H1EX-0.09

                                                        8H1EX-0.08

                                                        8H1E)U).10
                                                            -£-
                                                        8H5EX-0.08
                                                            -*-
                                                        8H1EX-0.07

                                                        1H1EX-0.10

                                                        8H5EX-0.09
         0   0.02  0.04  0.06  0.08  0.1  0.12 0.14  0.16 0.18  0.2
                       CONCENTRATION, PPM

-------
                                           93
ranges from about 32,000 under the 8H1EX-0.07 standard to about 210,000 for the 8H1EX-
0-.10 standard.  Another exposure indicator of interest is the total number of person-
occurrences at various concentration levels.  Figure V-8 displays the person-occurrences for
outdoor children experiencing daily maximum dose 8-hr exposures while engaged in
moderate exertion in Philadelphia. It is estimated that there are roughly 14 million person
occurrences of outdoor children engaged in moderate exertion for an 8-hr averaging time in
Philadelphia over the 7-month O3 season.  Attaining any of the alternative standards reduces
total person occurrences  exceeding 0.06 ppm to less than 1.5 million.  One can calculate an
average number of exposures per person to a given level by dividing the total person
occurrences at  a given 63 level by the number  of children experiencing a given level one or
more times per season.  For exposures exceeding 0.06 ppm, the average number of
occurrences per person for the alternative standards analyzed ranges from about 1.5 to 5.
For exposures  at or above 0.08 ppm the average number of occurrences per person drops to
a range of 1.0  to 1.5.   Similar figures showing persons and person-occurrences outdoor
children living in three other urban areas (Houston, Washington, D.C., and New York)  are
included in Appendix B.
       Tables  V-10 and V-ll provide summary exposure estimates for the Philadelphia
study area for outdoor children for 2  particular exposure indicators. Similar tables are
available for outdoor children, outdoor workers, and the general population for the 9 urban
areas in the exposure support documents cited at the beginning of this section.  Table V-10 is
for 1-hr MAXD exposure estimates where the 63 concentration exceeded 0.12 ppm and EVR
equaled or exceeded 30 1/min-m"2, while Table V-ll  is for exposures where the 03
concentration exceeded 0.08 ppm and EVR in the range of 13-27 1/min-m"2.  These
indicators  were selected  because they correspond to the lowest concentration levels at which
                                                                                   2
effects were observed in the 1 to 2 hour clinical studies at heavy exertion (_> 30 1/min-irr)
and 6 to 8 hour studies at moderate exertion  (13-27 1/min-m2).  Use of any single cutpoint in
the exposure distribution to compare  alternative standards must be done with  caution.  Using
any single cutpoint does not adequately represent the differences in the entire exposure
distribution between alternative standards.

-------
TABLE V-JO.  ESTIMATES OF ONE HOUR  MAXIMUM DOSAGE  EXPOSURES EXPERIENCED BV OUTDOOR CHILDREN IN
Ant"! PHI A niiDixir- t\ IM/-II r\-n\\iir f viMfi.'NTU ATiniv |."YfM<"F m~M n t") „„.,. AMM L-VDr 10 runs
Mean Estimate ol Person-Occurrences
Percent ol Total Person Occurrences
Range in (his percentage fur 10 runs
Statistic1'
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage for 10 runs
Mean Estimate of Person Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Regulatory scenarios - l-llour Standards
Baseline*
0.700
3 5
1 3 S 8
10.100
002
001 003
IIIIKX-0.12
ppm
81
0 03
0 00 0 2
81
d
d
1II1KX-O.IO
ppm
0
000
-
0
0 00
-

Regulatory scenarios - 8-IIour Standards
Baseline0
9.700
3 5
1 3-5 8
10,100
002
001-0.03
8IIIKX-
0.10
ppm
140
0.05
000-.2I
140
d
d
8IIIKX-
0.09 ppm
0
0.00
-
0
0.00
-
8II1KX-
0.08
ppm
0
000
-
0
000
-
8IIIEX-
0.07
ppm
0
0.00
-
0
0.00
-
8II5KX-
0.09
PPM
0
0.00

0
0.00
-
8II5KX-
0.08
ppm
0
0.00
-
0
0.00
-
          Equivalent ventilation r.ilc — (venlilnlion iato)/(l>ody MII|;ICV! 
-------

TABLF. V-ll. F.STIMATI S OF F.K.II 1 -HOUR MAXIMUM DOSAC.F, KXI'OSURKS F.XI'KRIKNCKI)
BY OUTDOOR CHILDRF.N IN I'HII.ADFl.l'IHA DURING WHICH O/.ONK CONCENTRATION
F,X( i;i;i)i:i) turn ppm AND KVR:I WAS IN nil RAN<;F. 13-2? I.IIFRS- Miiv1 • ivr2
Statistic1'
Mean Estimate of the Number ol Outdoor ChiMren
Percent of Total Outdoor Children Population
Range in this percentage lor 10 nitis
Mean Estimate of Person Occurrences
Percent of Total Person Occurrences
Range in this percentage lor 10 runs
Statistic*1
Mean Estimate of the Number of Outdoor Children
Percent of Total Outdoor Children Population
Range in this percentage lor 10 runs
Mean Estimate of Person-Occurrences
Percent of Total Person-Occurrences
Range in this percentage for 10 runs
Regulatory scenarios - 1-Hour Standards
Baseline*
186.000
67 7
65 6 70 2
580.000
0 98
089 1 1
IIIIF.X-0.12
ppm
58,400
21 2
1 8 7 24 3
79,300
0 13
0.12-0.17
IIMKX-O.IO
ppm
7,500
2.7
1 1-44
7,700
001
001-002



Regulatory scenarios - 8-Hour Standards
Baseline*"
1 86,000
67 7
65 6702
580,000
1 0
089 -1 1
8IIIFX-
0.10 ppm
68,800
25 0
20 9 30.2
98,100
0 17
0 140 20
SIIIKX-
0.09 ppm
18.000
65
4 3-10.7
20,300
003
0.02-0 06
8IIIF.X-
0.08
ppm
1,630
0.6
0 1 2.3
1.630
c
0.00-
001
811 IKX-
0.07
ppm
0
0.0
-
0
0.0
-
8II5KX-
0.09
PPM
50,300
18.3
15.1-
21.8
65,900
0.11
0.08
0.14
8II5EX-
0.08
ppm
6,200
2.3
0.9-4 8
7,100
001
0 0-0.02
'"Equivalent ventilation rale ~ (ventilation r,ite)/(bodv surlace area)
 Mean or range for  10 runs ol pNF.M/O^
cRaselinc scenario is based on 1991 ambient air quality  levels
llLcss than 0 01 percent
                                                                                                                                                            Ul

-------
                                           96
       In order to get a feel for how exposure estimates vary across the 9 urban study areas,
Figures V-9, V-10, and V-ll show the mean percent of outdoor children exposed  on 1 or
more days to the 8-hr MAXD indicators of interest for the various regulatory scenarios
analyzed. Figure V-9 presents these estimates for the "as is" situation, the current 1-hr
standard, and a 0.10 ppm,  1-hr, 1-expected exceedance standard.  The considerable
variability in baseline O-^ levels across the urban areas analyzed results in large variation in
"as is" exposure estimates. Except for Miami and Denver,  which were either in or near
attainment of the current 1-hr standard, exposures _>. 0.08 ppm at moderate exertion would
generally be significantly reduced upon attainment of the current 1-hr standard.  Attaining the
0.10 ppm, 1 expected exceedance, 1-hr standard would further reduce exposures for this
indicator. Figure V-10 presents similar estimates for the alternative 8-hr, 1-expected
exceedance standards.  Finally, Figure V-ll shows estimates for the 1- and 5-expected
exceedance standards set at 0.08 and 0.09 ppm.
       Some summary observations from the exposure analyses of the general population,
outdoor workers, and outdoor children are listed below:
       (1)    For 8-hr exposures at moderate exertion (EVR in the range 13-27 I/
             min-m^), outdoor children appear to have the highest percentage and number
             of individuals exposed to levels exceeding 0.08 ppm.
                                                              ^
       (2)    For 1-hr exposures at heavy exertion (> 30 1/min-m ), it depends on the
             urban area whether outdoor children or outdoor workers have the highest Og
             exposures exceeding 0.12 ppm.
       (3)    While not shown in this Section, the exposure estimates for exceeding 0.12
             ppm  at any exertion level are considerably higher than the numbers  and
             percentages presented in Tables V-10 and V-ll.  For example,  the model
             predicts 269,000 (97.8 percent) outdoor children exceeded 0.12 ppm under
             baseline (1991 air quality) compared to only 9,700 or 3.5 percent of outdoor
             children living in the Philadelphia study area  for this same baseline  air quality
                                          n
             when EVR was _>_ 30 1/min-m .  Thus, exertion level and its associated
             ventilation rate at maximum dose significantly affects MAXD exposure
             estimates.

-------
Air Quality Scenarios
 Mean Percent of Population Exposed

-------
Air Quality Scenarios
  Mean Percent of Population Exposed
                        86

-------
Air Qyality Scenarios
            00
            o   oi   o  en
Mean Percent of Population Exposed
                                         > "0
                       66

-------
                                   100
(4)     On both an absolute number and percentage basis, exposure estimates are
       greater for the 8-hr, _>_ 0.08 ppm indicator at moderate exertion than the  1-hr,
       >_ 0.12 ppm indicator at heavy  exertion.  This is not surprising since more
       people normally engage in moderate exertion than in heavy exertion activities.
(5)     Based on the general population, outdoor worker,  and outdoor children
       estimates, the 0.10 ppm, 1 exceedance, 8-hr regulatory option usually provides
       the least protection when judged by either 1- or 8-hr MAXD indicators and in
       most study areas results in greater exposures than  the current  1-hr standard.
(6)     There are relatively small differences in comparing the distributions of daily
       maximum dose 8-hr exposure estimates for outdoor children associated with  1-
       and 5-expected exceedance standards set at the same concentration level.
       However, if one selects a particular cutpoint on the distribution, such as
       exposures exceeding 0.08 ppm under moderate exertion, the differences
       between these  two forms of the  standard can appear to be more significant in
       some urban areas.  In the vast majority of the nine study areas the 5-expected
       exceedance form of the  standard results in greater exposures.  However,  in at
       least two study areas (i.e., Denver for the 8-hr daily max dose and St. Louis
       for the 1-hr daily max dose), the estimated exposures are somewhat greater for
       the 1-expected exceedance standard than the 5-expected exceedance  standard
       set at 0.09 ppm.  This seemingly illogical result occurs because the  design
       value monitor  for the 1-expected exceedance standard  is different than the
       design  value monitor for the 5-expected exceedance standard.  This  can occur
       because the site with the maximum second highest 8-hr daily  maximum
       concentration  is not always the same as the site with the maximum sixth
       highest 8-hr daily maximum  value.  The combination of significant differences
       in population  sizes assigned to different districts in an urban study area and
       differences in the impact of the  air quality adjustment  procedure due to the
       design  monitor being in different exposure districts for these two forms of the
       standard can lead to exposures being higher for a  1-expected exceedance  form
       than the 5-expected exceedance  form.

-------
                                           101
       4.      Caveats and Limitations
              A number of caveats must be acknowledged concerning the pNEM/O3 results.
Probably the most important caveat is that there is considerable uncertainty concerning a
number of important inputs to the model. Listed below are the most important caveats and
limitations in the current version of the exposure model.
       (1)     The subjects who contributed to the human activity database may not provide a
              balanced representation of U.S. outdoor children or outdoor workers.  The
              majority of subjects resided in either the State of California or in Cincinnati.
              Although the algorithm which constructs exposure event sequences attempts to
              account for effects of local climate on activity, it is unlikely that this
              adjustment procedure corrects for all inter-city differences in children's or
              outdoor workers' activities. Time/activity patterns are likely to be affected  by
              a variety of local factors, including topography, land-use,  traffic patterns,  mass
              transit systems, and recreational opportunities.
       (2)     As discussed previously, the average subject provided less than two days of
              diary data.  For this reason, the construction of each season-long exposure
              event sequence  required either the repetition of data from  one  subject or the
              use of data from multiple subjects.  The latter approach was used in the
              outdoor children and outdoor worker pNEM/O3 analyses to better represent the
              variability of exposure expected to occur among the children included in each
              cohort.   The principal deficiency  of this approach is that it may not adequately
              account for the day-to-day repetition of activities common to individual
              children.  Consequently, pNEM/O3 may tend to under-estimate the  number  of
              people who experience multiple occurrences of high exposure  while engaged in
              moderate or heavy exertion. For example, the outdoor children analysis does
              not adequately reflect exposures for children attending residential summer
              camps  because  this type of activity pattern is not included in human activity
              pattern  data base used in the  outdoor children exposure analysis.
        (3)    Exposure estimates have been presented separately for outdoor children.
              outdoor workers, and the general population and have not been aggregated.

-------
                                    102
       Any aggregation would have to adjust the general population exposure
       estimates to avoid double counting exposures for workers and children.
(4)    The algorithm that assigns the  EVR associated  with each exposure is based on
       an analysis of data from several studies conducted by Dr. Hackney and his
       associates in Los Angeles.  Because of the small sample sizes (e.g., 39
       children and 36 outdoor workers) in these studies and the lack of subjects
       below age 10 or above the age of 50 there is uncertainty which cannot be
       quantified about these EVR estimates. The pNEM/O3 model also employs an
       EVR limiting algorithm that determines the maximum EVR that can be
       maintained for a given duration by an individual that exercised regularly and
       was motivated to reach a high  ventilation rate.   In general, the EVR limiting
       algorithm tends to allow more  high EVR values to occur than would occur in
       the total population of interest  (i.e., outdoor children, outdoor workers, or
       general population).
(5)    The air quality adjustment procedures used to simulate just attaining alternative
       NAAQS were based on statistical analyses  of O3 data from sites that
       experienced moderate reductions in O3 levels during the  1980's. These
       procedures assume that (1) the  Weibull distribution provides a good fit to most
       O3 data and (2) the parameters  of the Weibull distribution fitting data from a
       particular  monitoring site will change over time in a predictable fashion.
       Because of the empirical basis  for the adjustment procedure, there is less
       confidence in the predicted air  quality levels for.just attaining  alternative
       standards in Los Angeles where significant reductions would have to take place
       to attain any of the alternative  standards analyzed.  The adjustment procedure
       was developed and tested with  a focus on the tail of the 1-hr and 8-hr air
       quality distributions.  Therefore, there is more  uncertainty about how well the
       adjustment procedure characterizes longer averaging times (i.e., seasonal 8-hr
       averages) and 1-hr and 8-hr daily maximum values that are in the middle of
       the distribution. A limited evaluation of the adjustment procedure (Johnson,
       1995)  suggests that the approach does a reasonable job of estimating the upper

-------
                                    103
       10% of the distribution of hourly  O3 values based on an empirical analysis of
       six of the nine urban areas included in the exposure analysis.  Further research
       and analysis is needed to better characterize uncertainty about possible changes
       in the spatial pattern and shape of O3 air quality distributions associated with
       control strategies adopted to attain the O3 NAAQS in the future.
(6)     The pNEM/O3 model uses a mass balance model to estimate O3 levels in
       residential buildings (windows open), residential  buildings  (windows closed,
       nonresidential buildings, and inside motor vehicles.  For some of these
       microenvironments the data base on air exchange rates (AER) and O3 decay
       rates, which are key inputs to the mass balance model,  is rather sparse.  For
       example, the AER and O3 decay rate for  motor vehicles is a point estimate
       based on data for a single vehicle. In contrast, data on AER  values for
       residential buildings with closed windows are based on  a lognormal
       distribution fit to AER data from  312 residences  across the U.S.  It  should be
       noted that the uncertainties about  O3 levels in these  "indoor"
       microenvironments should not have a significant  effect on  exposure estimates
       at moderate and high exertion  where  exposure levels exceed 0.08 ppm. since
       these are likely to be due to outdoor exposures.

-------
                                          104
H.     Ozone Health Risk Assessment
       1.     Overview
       This section summarizes an assessment of risks for several categories of respiratory
effects associated with attainment of alternative  1-and 8-hr O3 NAAQS.  This risk assessment
builds upon the earlier O3 NAAQS health risk assessment described in detail in Hayes et al.
(1987) and summarized in the previous O3 OAQPS Staff Paper (U.S.  EPA,  1989).  The O3
health risk assessment considers the same alternative air quality scenarios examined in the
exposure analysis described in Section V.G.
       The objective of the risk assessment is to estimate the magnitude of risk to the most
susceptible populations (i.e., outdoor workers and outdoor children) while characterizing, as
explicitly as possible, the range and implications of uncertainties  in the existing scientific
data base.  While the risk assessment estimates  should not be viewed  as demonstrated health
impacts, they do represent EPA's estimate as to the possible extent of risk for these effects
given the available scientific information.  Although it does not cover all health effects
caused by O3, the risk assessment is intended as a tool that may,  together with  other
information presented in this Staff Paper and in  the CD, aid the Administrator in judging
which alternative O3 NAAQS provides an adequate margin of safety.   Risk estimates for nine
urban areas and the methodology used to generate these estimates are described in detail in
Whitfield et al.  (1996).
       The three  major types of inputs to the risk assessment are:
       (1)    concentration-response or exposure-response relationships used to characterize
             various respiratory effects of O3 exposure;
       (2)    distributions of O3 1-hr and  8-hr daily maximum concentrations upon
             attainment  of alternative NAAQS obtained  from  the pNEM/O, analyses
             described in Section V.G.; and
       (3)    distributions of population exposure, in terms of the general population,
             outdoor workers, and outdoor children exposed and occurrences  of exposure,
             upon attainment of alternative O3  NAAQS obtained from the O3 exposure
             analyses.

-------
                                           105
       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 activities at different exertion levels).
       More specifically, benchmark risk is the  probability that a time-averaged O3
concentration will exceed a given benchmark concentration k or more times in a given period
at some location within a geographic area. The  benchmark concentration is the time-
averaged O3 concentration that will cause the occurrence of a specific health effect or
response in up to a given percentage of a sensitive population (e.g., outdoor children) under
given conditions of exposure.   Benchmark risk,  which is calculated assuming that all
members of the sensitive population are exposed outdoors under identical exposure
conditions,  is a measure  of the hazard posed  by  elevated ambient O3 levels.
       The second measure, headcount  risk,  is a population risk measure that assesses
number of people or percent of the sensitive  population that would be adversely affected
given normal movement  and activity patterns of the population of  interest.  Headcount risk
also provides estimates of the  number of occurrences 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 standard that provides an adequate margin of safety.
       2.     Exposure-Response Relationships
       Risk estimates  have been developed for a variety of respiratory effects reported to be
associated  with O3 exposure.  Table V-12 summarizes the effect categories covered by the
risk assessment that are summarized in  this section of the Staff Paper. Each of the effects is
associated  with a particular averaging time and for most of the acute (1 to 8-hr) responses
effects also are estimated separately for specific EVR ranges that correspond to the EVR
ranges measured in the health studies used to derive exposure-response relationships. An
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.  Risk estimates are summarized

-------
           TABLE V-12. BASIS FOR ACUTE HEALTH ENDPOINTS ADDRESSED BY RISK ASSESSMENT
Health
Endpoint
FEV1 decrement > 10%, 15%,
20%
Lower respiratory (moderate/severe)
symptoms
Cough (moderate/ severe)
Chest pain on deep inspiration
(moderate/ severe)
Acute excess respiratory-related
hospital admissions for asthmatics
Exposure Time (Exertion Level)
1-hr
(heavy)"
Avol et al., 1984
Kulleet al., 1983
McDonnell et al., 1985
Avol et al., 1984
Kulleet al., 1985
McDonnell et al., 1983
Kullc et al., 1985
McDonnell et al., 1983

1-hr
(moderate)b
Seal et al.,
1993

Seal et al.,
1993
Seal et al.,
1993
Thurston et al.,
1992d ,
8-hr
(moderate)0
Folinsbeeetal., 1988
Horstman etal., 1989
McDonnell etal., 1991

Folinsbeeetal., 1988
Horstman et al., 1989
McDonnell etal., 1991
Folinsbeeetal., 1988
Horstman et al., 1989
McDonnell etal., 1991

  'Equivalent Ventilation Rate (EVR) s 30 I mm'  m

  "EVR 2;  16 and £ 301 mm1 m:

  'EVR 2  13 and s 27 I min' nv
  •"There In no exertion level associated with the lOhct-ntration response relationship lor this endpomi because it is based on epideminlogical data   that
does not provide this information.

-------
                                           107
in this section for a variety of acute health endpoints. For lung function decrements
estimates are provided for the lower end,  midpoint, and upper end of the range of response
that might be considered an adverse health effect (i.e., >_ 10, 15, or 20% FEV, decrements).
For acute symptomatic effects, this section focuses on responses that the staff recommends be
considered as adverse effects (i.e., moderate or severe cough, moderate or severe pain on
deep inspiration (PDI)).
       Risk estimates have been calculated for each acute effect separately.  Consequently,
no risk estimates are available for multiple effects such as the joint probability of having
moderate or severe cough and a FEV, decrement _>. 15%. Preparing such joint effects risk
estimates would be very difficult given the existing  data base. The basis for these staff
recommendations was discussed in Section V-F of this Staff Paper.
       For the 1-hr, heavy exertion cases, exposure-response relationships were derived
separately based on three controlled chamber studies (Avol et al., 1984; Kulle et ah, 1985:
and McDonnell et al., 1983) and used to develop independent risk estimates. Table V-13
summarizes the studies  used to estimate 1-hr exposure-response relationships for populations
engaged in heavy exertion.  While the three  studies are similar in enough respects (e.g.,
health endpoints, young heavily exercising healthy subjects, similar 1-2 hour O-, exposures)
to make useful comparisons, there are enough differences in  experimental protocol  (e.g.,  1-
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.
       For the 1-hr, moderate exertion cases, exposure-response relationships were derived
based on a single, relatively large controlled human chamber study (Seal et al.,  1993).  This
study is summarized in Table V-14.
       A pooled data set based on three controlled  human exposure studies (Folinsbee et al..
 1988; Horstman et al.,  1989, McDonnell  et al., 1991) served as the basis for developing
exposure-response relationships for the 8-hr, moderate exertion cases.   These studies are
 summarized  in Table V-15.  It was felt that these data sets could be pooled because the
 studies were performed  in  the same location using essentially identical experimental protocols

-------
                                               108
TABLE V-13.  SUMMARY OF STUDIES USED IN DEVELOPING 1-HOUR EXPOSURE-
RESPONSE  RELATIONSHIPS FOR POPULATIONS ENGAGED IN HEAVY EXERTION
Study Protocol
Number of subjects
Avol et al (1984)
50 bicyclists: 42 male
and 8 female; complete
data were available for
48 of the subjects
Kulleetal (1985)
20 healthy males;
8 of 20 subjects
exposed to
0.30 ppm'
McDonnell et al. (1983)
135 healthy males;
complete data were
available for 132 of the
subjects
 Exposure concentrations
 (ppm)

 Ventilation rateb
 (L/mm)

 EVRC (L/min/m:)

 Exercise pattern
 Exercise duration (heavy)
 (mm)

 Exposure duration (h)

 Subject exposures
0.00, 0.08, 0.16, 0.24,
and 0 32

57.6 ± 12.5
30.3

Continuous (10-min
warm-up, 60 min of
continuous exercise. 10-
min cooldown and
measurement)

60
1 33

Exposed to all
concentrations
0.00,0.10, 0.15, 0.20,
0.25, and 0.30

67 8  ± 8 2
35 7

Intermittent (4 cycles
of 14 mm of exercise
alternated with 16 mm
of rest)
                                                   56
000, 0 12, 0 18. 0.24,
0 30, and 0 40

65 6 + 1 4
34 3  ± 3 1

Intermittent (4 cycles of 15
min of exercise and 15 mm
of rest)
                       60
Exposed to all
concentrations.
Di\ided ahou; equalK ink
6 groups, each exposed tu
a single concentration
 '  Data for exposures of 0.30 ppm not reported in Kulle et al. (1985) were taken from Hayes el al  (1987)
 b  Mean ± standard deviation; averages of group (based on ozone concentration) means
 c  Estimated for Avol et al. and Kulle et al  by dividing ventilation rate by 1.9 m:, the approximate human body
   surface area, to obtain equivalent liters per minute; calculated for McDonnell et al from available data
 11  Includes a final 30-min period during which subjects rested, and spirometnc and symptoms measurements were
   made.

-------
                                        109
   TABLE V-14.  SUMMARY OF THE STUDY USED TO DEVELOP 1-HOUR
EXPOSURE-RESPONSE RELATIONSHIPS FOR POPULATIONS ENGAGED IN
                            MODERATE EXERTION
                      Study Protocol
      Seal et al. (1993)
                   Number of subjects


                   Exposure
                   concentration (ppm)

                   Mean ventilation rate"
                   (L/min)

                   EVRb (L/min/nr)

                   Exercise pattern


                   Exercise duration (h)

                   Exposure duration (h)

                   Subject exposures
372 African-American and
White males and females

0.00, 0.12, 0.18, 0.24, 0.30, or
0.40

45
23.8 ± 2.8

Intermittent (4 periods of
15-min exercise, 15-min rest)
1
2.33C

About 60 subjects exposed at
each level; each subject exposed
to onlv 1 concentration level
                   '  Calculated from mean EVR by multiplying by 1.9 nr,
                     the approximate body surface area.
                   b  Mean ± standard deviation; averages of group means.
                   c  Includes a final 20-min period during which subjects
                     rested, and spirometnc and symptom measurements
                     were made.

-------
                                             110
TABLE V-15. SUMMARY OF STUDIES USED TO DEVELOP 8-HOUR EXPOSURE
                             RESPONSE RELATIONSHIPS
   Study Protocol
 Folinsbee et al. (1988)     Horstman et al. (1990)     McDonnell et al. (1991)
Number of subjects

Exposure
concentration (ppm)

Ventilation rate
(L/min)

EVR (L/min/nr)
Exercise pattern
10 nonsmoking males

0.00 or 0.12


39-42'


20.5-22. lc
50 rrun of exercise,
10 min of rest for each
hour, and 35 min of rest
after third hour
Exercise duration (h)   5

Exposure duration (h)   6.6
22 nonsmoking males

0.00, 0.08, 0.10, or
0.12

37-41'
19.5-21.6C


50 rrun of exercise,
10 min of rest for each
hour and 35 mm of rest
after third hour

5

6.6
38 nonsmoking males

0.00,0.08, or 0.10
(40.3, 40.5, 39.6) ±  (4.3,
4.3, 6.3)b

(20.1, 20.2, 19.9) ±  (1.8.
1.8, 2.3)b

50 min of exercise, 10 min
of rest for each hour and
35 mm of rest after third
hour

5

6.6
Subject exposures
Exposed to all
concentrations
Exposed to all
concentrations, except
for 1 subject who
experienced respiratory
problems at 0. 10 ppm
28 subjects exposed to 0.00
and 0.08 ppm;
10 subjects exposed to
0.00. 0.08. and 0.10 ppm
'  Range of group means.

b  Means  ± standard deviation for 0. 0.08, and 0.10 ppm. respectively.

c  Range of group means estimated by dividing the ventilation rate by 1.9 nr, the approximate human body
  surface area, to obtain equivalent liters per minute.

-------
                                          Ill
and population groups.  Several other controlled human exposure studies (summarized in
Table 7-9 of the CD) have reported lung function decrements and symptoms in healthy and
asthmatic subjects due to O3 exposures lasting 6.6 to 8.0 hours (Horvath et ah, 1991;
Hazucha et ah, 1992; and Linn et ah, 1994; Horstman et ah, 1995). These additional
studies were not included in developing the estimated exposure-response relationships because
they each involved a single exposure level and differences in study protocols precluded
pooling the data from these studies with the Chapel Hill studies.  The magnitude of the
responses was somewhat lower in some of these studies, specifically the Linn et ah (1994)
and Horvath et ah (1991), compared to the three Chapel Hill studies used in the risk
assessment.  However, this may have been due to use of a lower ventilation rate and
attenuation due to previous exposures in Los Angeles for the Linn et ah (1994) study and the
use of older, less sensitive subjects in the case of the Horvath et ah (1991)  study.  The
responses of asthmatics in the Linn et ah (1994) and the Horstman et ah (1995) studies is
more comparable to the level of responses seen in the three Chapel  Hill studies used in the
risk assessment.
       The acute exposure-response relationships developed based on the chamber studies
referenced in Table V-12 were applied to "outdoor children," "outdoor workers," and the
general population.  While these specific chamber studies  only included adults aged 18-35.
findings from other chamber studies  (McDonnell et ah,  1985) and summer  camp field studies
in at least six different locations in the northeast United  States, Canada, and Southern
California indicate changes in lung function in healthy children similar  to those observed in
healthy adults exposed to O3 under controlled chamber conditions (CD, Section 9.3.1.2).  As
stated in the CD, "although direct comparisons cannot be  made because of incompatible
differences in experimental design and analytical approach," the range of response in the
summer camp studies "is comparable to the range of response seen  in chamber studies at low
O3 concentrations."
        As discussed earlier in this SP, there is a growing  data base of epidemiological
studies reporting associations between increased acute respiratory-related hospital admissions
and elevated O3 levels during the summertime.   In reviewing the studies reporting increased
hospital admissions associated with elevated O3 levels, only those studies which adequately

-------
                                           112
addressed statistical confounding by long-wave cycles in respiratory hospital admissions were
considered due to concern that hospital admissions are clearly dominated by other causes
(e.g., spring pollen, fall respiratory infection). Table V-16 (Table 7-23 from the CD)
summarizes those studies that met this criterion.  The concentration-response relationships for
acute respiratory-related hospital admissions for asthmatics were derived based on one of
these epidemiological studies (Thurston et al., 1992) that examined several New York cities.
Only the data  for the New York City population has been analyzed for the risk assessment.
The choice of New York City was driven by the availability of O-^ hourly values for the
entire O^ season upon attainment of alternative standards which was produced for each of the
nine urban areas as part of effort to develop exposure estimates.  As shown in Table V-16,
the effect size observed in other cities and studies that used the same O^ indicator and
general  approach ranged from  1.4 to  3.1 admissions/100 ppb O-^/day/10" persons.
       Concentration-response relationships are available for total  excess respiratory-related
admissions or  excess  respiratory-related admissions for asthmatics only.  In this section,
excess hospital admissions for asthmatics are summarized.   Additional estimates, including
total respiratory hospital admissions, are included in Appendix  C and Whitfield et al. (1996).
The Schwartz  et al. (1994a.b.c) studies focus  on  only a subset of the total respiratory-related
hospital admissions and were not analyzed  in this risk  assessment.
       Given the lack of experimental human  data.  EPA sponsored an effort in  1990-1991 to
develop a chronic lung injury risk assessment  based on experts' judgments (Winkler et al.,
1995).  In the 1990-1991 assessment, the experts were explicitly told that their judgments
were not being used as part  of the NAAQS review process, but rather to gain a more general
insight into  the potential for chronic  effects in areas with significantly elevated O^ levels.
Based on the age of this analysis and advice from the CAS AC O^ Review Panel, OAQPS is
not considering the results from this  chronic risk assessment in the current NAAQS review.

-------
      Table V-16.  Summary of Effect Estimates for Ozone in Recent Studies of Respiratory Hospital Admissions"
Location
New York City, NYC
Buffalo, NY'
Ontario, Canada0
Toronto, Canada0
Montreal, Canadad
Birmingham, ALe
Birmingham, AL'
Detroit, Mle
Detroit, MIC
Minneapolis, MNC
Minneapolis, MNC
Reference
Thurston et al. (1 992)
Thurston et al. (1 992)
Burnett et al. (1 994)
Thurston et al. (1 994)
Delfinoet al. (I994a)
Schwartz (1994a)
Schwartz (I994a)
Schwartz (I994h)
Schwartz (1994H)
Schwartz (I994c)
Schwartz (I994c)
Respiratory Admission Effect Size (± SE) Relative Risk (95% CJ)b
Category [ Admissions/ 1 00 ppb O3/day/ 1 0* persons] [RR of 100 ppb Oj, 1-h max]
All
All
All
All
All
Pneumonia in elderly
COPD in elderly
Pneumonia in elderly
COPD in elderly
Pneumonia in elderly
COPD in elderly
1.4 (± 0.5)
3.1 (± 1.6)
1.4 (± 0.3)
2.1 (±0.8)
1.4 (± 0.5)
0.73 (± 0.54)
0.83 (± 0.33)
0.82 (± 0.26)
0.90 (± 0.41)
0.41 (+ 0.19)
r
1.14 (1.06 to
1.25 (1.04 to
1.10 (1.06 to
1.36 (1.13 to
1.22 (1.09 to
1.11 (0.97 to
1.13 (0.92 to
1.22 (1.12 to
1.25 (1.07 to
1.117(1.03 to
r
1.22)
1.46)
1.14)
1.59)
1.35)
1.26)
1.39)
1.35)
1.45)
1.39)

"See Appendix A in the Ozone Criteria Document for abbreviations and acronyms.
hOne-way (0 ± 1.65 SE).
cl-h daily maximum ozone data employed in analysis.
d8-h daily maximum ozone data employed in analysis.
e24-h daily average ozone data employed in analysis.  (I h/24 h avg ratio = 2.5 assumed to compute effects and RR estimates).
'Not reported (nonsignificant).

-------
                                            114
       Methodology for Developing Probabilistic Exposure-Response Relationships.  A brief
 summary of the methods used to derive probabilistic exposure- and concentration-response
 relationships is described below.  A more detailed description of the methodology can be
 found in Whitfield et al.  (1996).
       The development of exposure-response relationships for acute endpoints is a 3-step
 process.  The starting point is data  from the laboratory experiments described above.  Before
 developing the needed probabilistic exposure-response relationships.  The data were corrected
 for exercise in clean air in an effort to remove any systematic bias that might be present in
 the data attributable to an exercise effect. Generally, this correction for exercise in clean air
 was small relative to the total effects measured in the O3-exposed cases.  These data become
 the "observations" shown in  Fig. V-12 and indicated by a 1 inside a circle to denote step 1.
       Step 2 is to fit a  function to the data via regression techniques.  This step is necessary
 because of the need to estimate response rates at O3 concentrations that differ from those at
 which laboratory data are available.
       Step 3 is to develop, for example, the 90% credible interval about the  fitted
 (predicted) response rate at O3 concentrations needed for the risk assessment calculations
 (i.e., those used in pNEM).  This step characterizes uncertainty attributed to  sampling error
 based on sample size considerations.  This uncertainty was estimated using a Bayesian
 approach  involving the application of the inverse beta function with parameters X and N - X,
 where X is the predicted response rate at a particular O3 concentration, and N is  the number
 of subjects associated with the chosen O3 concentration. The 90% CI is defined by the 0.05
 and 0.95 fractiles.
       For the risk assessment, response rates were calculated for 21 fractiles (for
 cumulative probabilities  from 0.05 to 0.95 in steps of 0.05, plus probabilities  of 0.01 and
 0.99) at a  number of O3 concentrations that depended on the health endpoint.  A  function that
 "best fit" the data was chosen subject to the constraint that linear functions were  favored,
especially when the number of observation points (i.e.,O3 concentrations at which laboratory
data are available) was small.  There are as few as  2 useable observation points  and as many
as 6 observation points for the endpoints examined.

-------
                                115
FIGURE V-12. STEPS USED TO DEVELOP PROBABILISTIC EXPOSURE-
                    RESPONSE RELATIONSHIPS.
        Derived Exposure/Response Distributions
   Horst-Foll-McDon, FEVl Decrement 2:10%, 6.6 Hr Expo, Moderate Exercise
        100


         80
         60
         40
         20
                                FU Function (vit
                                Regression)
                                                      Legend:
                          .95 frac
                          .75 frac
                          .50 frac
                          .25 frac
                          .05 frac
0.00
0.04
0.08
0.12
                                            0.16
0.20
                    O3 Concentration (ppm)
                                               CP108 - M9-95 - 15:48

-------
                                          116
       Some illustrative concentration- and exposure-response relationships for some of the
effects examined in the risk assessment are displayed in Appendix C.  A table listing the
functional form and parameters for all of the concentration-response and exposure-response
relationships included in the risk assessment also is contained in Appendix C and discussed in
more detail  in Whitfield et al.  (1996).
        Other sources of uncertainty due to differences in experimental protocol, subject
population,  measurement error, etc. have not been quantitatively addressed for these acute
health endpoints.  The calculation and presentation of separate risk estimates for each of the
three heavy  exertion data sets provides a rough picture of the degree of uncertainty due to
these other factors because this health endpoint was examined in more than one study.
       3.      Benchmark  Risk Results
       For the O3 health risk assessment the benchmark risk is defined as  the probability
that, upon just attaining a  given O3 NAAQS, the daily maximum 1-hr (or 8-hr)  concentration
will equal or exceed the level that would cause 5 or 10% of the population of interest
(e.g.,outdoor workers, outdoor children, or the general population) to exhibit particular
health endpoints 1  or more times per year.  The benchmark risk is estimated assuming the
entire sensitive population is exposed under exertion levels associated with a particular effect.
       Benchmark risk is  measured in excess of that  which would occur under background
conditions because (a) only O3 levels above background are amenable to human control and
(b) it is difficult to reliably estimate the very small, hypersensitive  fraction of the population
engaged in moderate or heavy exertion that might respond at O3 levels at or below
background.  While background O3 levels  can  vary during the day  and  from day to day, for
the purposes of this risk assessment, 0.04 ppm is  used as a reasonable estimate  of the
background  level for both  1- and 8-hr daily maximum concentrations experienced on a
typical O3 season day.
       Benchmark risk is calculated by combining exposure-response relationships and
probability distributions of daily maximum  1- or 8-hr O3 ambient concentrations, based on
conditions of exact attainment of alternative NAAQS. The benchmark risk model and more
detailed discussion of the inputs to the model are contained in Whitfield et al. (1996).

-------
                                          117
       Benchmark risk estimates are calculated for the 9 urban areas shown in Table V-7.
Figure V-13 shows the benchmark risk estimates for Philadelphia for the 8-hr, moderate
exertion health endpoint defined as FEV, decrements _>. 20%.  The solid vertical bars
indicate the probability that upon attaining a given alternative standard O3 levels will be
exceeded five or more times in a season that would result in 5 % of the population
experiencing  this endpoint if they  were exposed while engaged  in moderate exertion. The
dashed vertical bars represent a similar measure, but for 10% of the population experiencing
the specified  endpoint. For example, attaining the 8H1EX-0.09 ppm standard results in a
benchmark risk probability of around 0.95 that 5% of the population would experience an
FEV, decrement _>_ 20% and the probability is about 0.7 that 10% of the population would
experience this same health response, if the population were exposed to these O, levels while
engaged in moderate exertion.
       Benchmark risk estimates for selected health endpoints are presented in Appendix C.
Additional health endpoints are included in Whitfield et al. (1996).
       4.      Population ("Headcount") Risk Results
       Population, or "headcount," risk is characterized by calculating the number of people
experiencing  a defined effect and  the expected number of incidences of that effect projected
to occur during the O3 season, given that a particular NAAQS  is just attained.  Risk
estimates  have been developed for the general population and for two groups expected  to be
at greater risk due to  their increased activity or exertion level outdoors during the O, season:
outdoor workers and outdoor children.
       A  major input to the headcount risk model is the series  of population exposure
distributions  for the alternative NAAQS analyzed by EPA.  Using available exposure
estimates, risk estimates were  calculated for the nine urban areas listed in Table V-7.  For 8-
hr exposures under moderate exertion, outdoor children represent the population group
experiencing the greatest exposure, and, therefore,  this population also has the highest  risk
estimates  in terms of the percent of the population estimated to respond.   Therefore, the
remainder of this section focuses on the risk estimates for  outdoor children.  Whitfield et  al.
(1996) presents a more complete summary of the headcount risk estimates for each  of the
nine urban areas for outdoor children and outdoor workers.  To illustrate the type of risk

-------
FIGURE V-13.  PROBABILITY THAT THE BENCHMARK RESPONSE FOR THE EIGHT-HOUR, MODERATE EXERTION
HEALTH ENDPOINT FEV, DECREMENT 2: 20% WILL BE EXCEEDED FIVE OR MORE TIMES IN AN OZONE SEASON.
1.0
S 0.8
• v_4
od
2
& 0.6
^
2
J^
H 0.4
cd
•g
ffl 0.2
0.0









|
1
I
i
I
|












I









:
|
t
r




I




|
|
§
:
:
1
|




:



:




I








|
ZJ BFCAHDG
Benchmarks:

	 o.io
Scenarios:
Z-As-Is
J-8H5EX-90
B-8H1EX-90
F=8H5EX-80
C-8H1EX-80
A-1H1EX-120
H-1H1EX-100
D-8H1EX-100
G=8H1EX-70
Philadelphia
                                                                                                 CD

-------
                                           119
assessment output that is available, Figure V-14 shows cumulative probability distributions
corresponding to just attaining alternative O3 standards  for two of the 8-hr, moderate
exertion cases: percent of outdoor children estimated to have 1 or more occurrences of FEV,
decrement _>. 10% and percent of outdoor children estimated to have 1 or more occurrences
of FEY, decrement _>. 20% in Philadelphia.  This figure shows, for example, that just
attaining the 8H1EX-0.09 ppm standard in  Philadelphia results in a median (0.5) probability
that about 34,000 outdoor children would experience FEV, decrements _>_ 10%.  When the
health endpoint of interest is defined as FEV, decrement >, 20%, the median probability is
that about 12,000 outdoor children would experience this effect.  These risk estimates  are for
the number  of children experiencing O3-induced occurrences in excess of estimated
background levels during a single O3 season for the Philadelphia urban area.   The variation
due to the 10 different pNEM/O3 runs  for each alternative standard has been collapsed into a
single representative distribution in order to better examine the differences  between
alternative standards.5
       The  top diagram in Figure V-15 shows outdoor children living in Philadelphia
estimated  to experience lung function decrements _>_ 15% one or more times in  an  O3  season
under moderate exertion for an 8-hr averaging time. Since any individual may experience
multiple occurrences  of an effect, risk  estimates also have been developed for total
occurrences of a specified effect. The bottom diagram in Figure V-15 displays  total
occurrences of this same response among outdoor children living in Philadelphia.  As an
example,  Figure V-15 shows for that just attaining the 8H1EX-0.09 ppm standard in
Philadelphia results in a  median (0.5) probability that about 30,000 outdoor children would
experience FEV, decrements _>_ 15% and a median probability that there would be 240,000
total occurrences of this effect.  This results in an estimated average number of occurrences
of eight per outdoor child for this endpoint.
       The  ratios for the mean number of occurrences and mean  numbers of outdoor children
responding  have been calculated for three health endpoints (FEV, decrements > 15%  and >
20%  for 8-hour exposures under moderate exertion and moderate or severe PD1 for 1-hour
  The representative distributions are obtained by effectively integrating across all 10 distributions for a given
 standard simultaneously and normalizing the result by dividing by 10. This calculation assumes that the distributions
 are perfectly correlated.

-------
Cumulative Probability
Cumulative Probability

                                 e-
                                 ffl
                                           p
                                           k>
      PPP
      A      i>     oo
                                CfQ
                                 a
                                   K>
                                   O
                                 I
                                 §
                                 IV
                                 I—'
                                 o
                                 \M I

                                 S
O  00 ^



     <

                                                                        o

                              IV
                                                                        30 w 3
                                                                        H ., O
                              > g
                              M >
                              r H

                              II

                                l
                                                                                 to
                                                                                 o

-------
                                121
FIGURE V-15  REPRESENTATIVE RISK DISTRIBUTIONS FOR ALTERNATIVE AIR QUALITY

SCENARIOS (FEV, DECREMENTS  ^  15%, PHILADELPHIA,  OUTDOOR CHILDREN, 8 HR

EXPOSURES, MODERATE EXERTION)
£»
3

I

I

I
   u
                    Illlf/l
          o.o
            0        20        40        60         80         100


           Children Having FEVl Decrements >15% (thousands)
    03

    i
J2
3
          0.4 -
          0.2
          0.0
            0        200        400        600       800        1000


            Incidences of FEVl Decrements ;>15% (thousands)

-------
                                           122
exposures under moderate exertion) across the nine urban areas for five air quality scenarios
(8H1EX-0.08 ppm, 8H5EX-0.08 ppm, 8H1EX-0.09 ppm, 8H5EX-0.09 ppm, 1H1EX-0.12
ppm). Figures C.24  through C.26 in  Appendix C display these ratios.  These ratios provide
an estimate of the average number of times that a responder would experience the specified
effect during an O3 season.  For the two 8-hr moderate exertion endpoints, the ratio ranges
from about 4  to 8.7 for FEV, decrements > 15% and from about 2 to 4.7 for FEV,
decrements > 20% across the nine urban areas.  For the 1-hr moderate exertion endpoint
defined as moderate-to-severe PDI, the ratio ranges from about 8 to 20 occurrences per
responder across the  nine urban areas.  There is no consistent ordering or pattern in the
ratios as one compares alternative scenarios across the different urban areas.  In addition,
there is no consistent pattern to the ratios among the alternative scenarios examined.
       In order to facilitate comparison of risk estimates across the 9 urban areas, a central
tendency risk  estimate (the median values) for outdoor children for the acute 1- and 8-hr
moderate exertion and 1-hr heavy exertion6 health endpoints have been included in Tables  V-
17, V-18, and V-19 respectively.  The range of risk estimates in each cell in these tables
indicates the variability in risk that occurs as one compares the nine different urban areas.
The  range of risk estimates  may be due to differences in the shape of the Q, air quality
distributions among the nine urban areas, differences  in exposure due to different levels of
air conditioning use in each urban area, or other differences in the exposure estimates such
as the spatial pattern  of population residences.
       5.     Excess Respiratory-Related Hospital Admissions.
       As discussed earlier in this  section, several epidemiology studies, mainly conducted in
northeastern U.S. and southeastern Canada, have reported excess daily respiratory-related
hospital admissions being associated with elevated O3 levels during  the O3 season (see
Table V-16).  To gain insight into  the possible impact of just attaining alternative 1- and 8-hr
O3 standards,  OAQPS and ANL have developed a risk model for this endpoint for the New
York City population. The  model  is based on the regression coefficients (and the
corresponding standard errors) developed by Thurston et al.  (1992) for New York City and
  For the 1-hr, heavy exertion case, only the risk estimates based on the exposure-response relationships derived
from McDonnell et al. (3983) are presented. Risk estimates based on Kulle et al. (1985) and Avol. (1984) are
presented in Whitfield et al. (1995).

-------
TABLE V-17.  RANGE OF MEDIAN PERCENT OF OUTDOOR CHILDREN RESPONDING ACROSS NINE U.S.
URBAN AREAS UPON ATTAINING ALTERNATIVE AIR QUALITY STANDARDS
l-hr Health Kndpoinls Under
Moderate Kvertion"
FEV| decrement _>_ 10%
FEVj decrement _> 15%
FEV| decrement _>_ 20%
Moderate or Severe Conph
Moderate or Severe Pain on Deep
Inspiration
Avis
2 9-22 0
12129
0681
0076
0 68 2
Alternative 1- Mr Daily M;i\imiiin
NAAOS
Mill X-
0.12 ppmh
2461
0028
0 51 5
0004
0516
inn \-
0.10 ppm
0 51 9
0107
01-03
00
01-04
Alternative 8-Hr
Daily Maximum NAAQS
811 II X-
0.10 ppm
4088
1 84 3
1 0-2 4
0 II 2
1 0-2 6
8IIIKX-
0.09 ppm
2.7-6 3
1129
0.6-1 6
0 0-0.5
0.6-1.7
81 III, X-
0.08 ppm
1 86 3
0.7-2.9
04-1.6
0 0-0.5
0.4-1.7
Kill KX-
0.07 ppm
1 1-3.1
0.4-1.3
0.2-0.6
00
0.2-0.7
                                                                                                       N>
                                                                                                       LO
  ''Equivalent ventilation rate J>. 16 I min  m  and _<^ 30 I min m .
  ''Current O3 NAAQS.

-------
 TABU: v-is. RANGE OF MEDIAN PERCENT 01 OUTDOOR CHILDREN RESPONDING ACROSS NINE u.s. URBAN
                        AREAS UPON ATTAINING ALTERNATIVE AIR QUALITY
8-hr Health EiulpoinLs Under
Moderate Exertion3
FEV, decrement > 10%
1 ™
FEV, decrement _>. 15%
FEV| decrement J>. 20%
Moderate or Severe Cough
Moderate or Severe Pain on
Deep Inspiration
As-ls
5.7-29.6
4.2-27.1
0.7-13.7
0.0-5.6
0.0-7.8
AltiTiiativi1 Mir Daily
Maximum NAAQS
1IJIEX-
0.12 p|)inb
6.0-15 7
46-13.7
1 1-5.9
0 ()-() 6
0.0-0 3
IIIIEX-
0.10 ppm
3.6-11.4
2.6-9.4
0.4-3.5
0 0-0 1
0.0
Alternative 8-Hr Daily
Maximum NAAQS
8IIIEX-
0.10 ppm
8.4-18.2
6.9-16.1
2.4-7.3
0.2-1.3
0 i-1.1
8IIIEX-
0.09 ppm
6.2-13.9
4.9-11.9
1.5-4.8
0.0-0.5
0 0-0.2
8HIEX-
0.08 ppm
4.3-102
3.3-8.3
0.9-2.8
0.0-0.1
0.0
8IIIEX-
0.07 ppm
7 4-6.8
1 7-5.1
02-1.2
0.0
0.0
aEi|iiivaleiit ventilation rate _>_ 13 and _<_ 27 I miif  in" .

''Current O^ NAAQS.

-------
TABLE V-19. RANGE OF MEDIAN PERCENT OF OUTDOOR CHILDREN RESPONDING ACROSS NINE U.S.
             URBAN AREAS UPON ATTAINING ALTERNATIVE AIR QUALITY STANDARDS'
1-hr Health Endpoints Under
Heavy Exertion1*
FEV, decrement _>_ 10%
FEV, decrement _>_ 15%
FEV, decrement >_ 20%
Moderate or Severe Cough
Moderate or Severe Pain on
Deep Inspiration
As-ls
0.1-4.4
0.0-1.8
0.0-1.3
1.0-7.0
0.5-5.4
Alternative 1-Hr Daily
Maximum NAAQS
1H1EX-
0.12 ppmc
0.1-0.5
0.0
0.0
1.4-3.7
0.7-2.4
IHIEX-
0.10 ppm
0.0-0.2
0.0
0.0
0.9-2.6
0.4-1.5
Alternative 8-Hr Daily
Maximum NAAQS
8HIEX-
0.10 ppm
0.2-0.7
0.0
0.0
1.9-3.9
1.1-2.7
8HIEX-
0.09 ppm
0.1-0.5
0.0
0.0
1.5-3.3
0.8-2.2
8H1EX-
0.08 ppm
0.1-0.2
0.0
0.0
1.0-2.5
0.6-1.4
8HIEX-
0.07 ppm
0.0-0.1
0.0
0.0
0.7-1.9
0.3-1.0
                                                                                                       Un
'Based on exposure-response relationships derived from McDonnell et al. (1983).
""Equivalent ventilation rate >_ 30 I min"1 m:.
'Current O, NAAQS.

-------
                                           126
 estimated daily maximum hourly average O3 levels over an entire season at various monitors
 in New York City upon attainment of alternative standards developed by IT-AQS for the
-pNEM/O3 analyses.  Since the original published analysis only examined the relationship of
 excess hospital admissions and daily maximum 1-hr O3 concentrations, we are unable to
 address at this time the relationship between 8-hr daily maximum O3 concentrations and
 excess hospital admissions for the New York area.  However, Delfino et al. (1994) report a
 similar effect size in their study involving excess hospital admissions where they used the 8-
 hr daily maximum 03 concentration on the day prior to admission.  Thurston et al. (1992)
 developed regression coefficients for two types of respiratory admissions: (1) for asthmatics
 only and (2) for total respiratory-related admissions.  Since the results are  fairly similar,
 only the risk estimates for asthma admissions are presented here.  The regression coefficient
 (11.7 admissions/ppm O3/106 people) for excess hospital admissions for asthmatics and  its
 standard error (4.7 admissions/ppm O3/106 people) were used to  define a probabilistic
 concentration-response relationship. A 1-day lag is associated with  O, exposure and the
 subsequent admissions of asthmatics.  The model is described in  more detail in  Section  6 of
 Whitfield et al. (1996).
        One hour daily maximum 03 concentrations for one O3 season under various
 alternative air  quality standards were used to estimate the number of excess (i.e.,
 attributable to  O3 concentrations higher than  background) respiratory-related admissions of
 asthmatics.  Risk estimates have been  prepared using 4 different  monitors in the New York
 City area (Queens-monitor 9, Greenwich-monitor  1, White Plains-monitor  11. and Babylon-
 monitor 12).  The O3 concentration-response relationship developed  by Thurston et al.
 (1992) was based  on air quality data from the Queens monitor.  Therefore, the  risk
 estimates based on the Queens County monitor most closely  represent the air quality index
 used in the original study. In each analysis, the air quality was adjusted to just attaining a
 particular standard at the monitor with the highest O3 levels for the New York area (i.e., the
 Babylon monitor)  and the O3 levels were adjusted  at the other monitors using the procedures
 described in Johnson et al. (1996b).
       The median estimate of the concentration-response relationship and  the 0.05 and 0.95
 fractiles estimates for O3-induced, excess respiratory admissions for  asthmatics  in New

-------
                                           127
York City are shown in Figure V-16.  Figure V-17 displays the cumulative probability
function for excess annual hospital admissions attributable to O3 for asthmatics,  corrected
for background O3 of 0.04 ppm, for each of nine air quality scenarios.4  Similar estimates
also  are available for excess total respiratory-related admissions attributable to O3  exposure
and are included in Whitfield et al. (1996).
       The hospital admissions risk model produces a median estimate of nearly 390 excess
annual admissions (corrected for background) for the 1991 "As Is" scenario using the
Queens monitor and 214 day O3 season.  Thurston et al. (1992) examined  unscheduled
admissions during a 3-month period  in  1988.  When the hospital admissions risk model is
limited to the same time period used in the original study and with background  set at 0 ppm
(the  approach used in Thurston et al. (1992)), nearly identical excess daily admissions
estimates are obtained (5.9 per day,  s.d. 2.5).
       To examine the impact of using alternative monitors to serve as the basis for an O3
index for the New York City area population, the risk estimates for asthmatics and total
respiratory hospital admissions also have been calculated  using other monitors and are
included in Whitfield et al. (1996).
       Focusing on the estimates based on the Queens County monitoring  site,  the median
estimate for Cyinduced hospital admissions for asthmatics in the New York City area is
about 210 (90% credible  interval (C.I.) = 70-344) upon attaining the current  1H1EX-0.12
ppm standard. This represents a nearly 50% decrease in O3-induced admissions due to
concentrations in excess of an estimated 0.04 ppm background compared to the As Is
scenario.  It is estimated  that attaining an 8H5EX-0.08 ppm standard would reduce O3-
induced excess hospital admissions to about  120 (90% C.I. =  41-199) which  represents a
70% decrease in O3-induced admissions from the As Is scenario due to concentrations in
excess  of the estimated 0.04 ppm background.
        There is very little difference in the risk estimates for excess  hospital  admissions
between the 1- and 5-expected exceedance standards  set at the same concentration level.
 7 The population used in the hospital admissions analysis is smaller (i.e., 7.3 million people) than the New York
 urban area population used in the pNEM/O3 analysis and risk estimates for the other health endpomts (i.e.,  10.7
 million people) and represents the same geographic area as in the Thurston et al. (1992) study.

-------
                              128

FIGURE V-16. CONCENTRATION-RESPONSE RELATIONSHIP FOR DAILY
  HOSPITAL ADMISSIONS OF ASTHMATICS IN NEW YORK CITY AREA.
         IDaily Hospital Admissions of Asthmatics
        Probabilistic Concentration-Response Relationships
                              0.20
           One-Hr Daily Maximum Ozone Level Cppm)
                                              bo«a«l - 1-26-93 - 1O:O«
FIGURE V-17.  EXCESS ANNUAL HOSPITAL ADMISSIONS OF ASTHMATICS
ATTRIBUTABLE TO OZONE EXPOSURE FOR ALTERNATIVE AIR QUALITY
                          SCENARIOS.
  •8
  I
  I
  I
  u
        i.o
        0.8
0.6
0.4
        0.2
        0.0
         7
                       s
                      if?
           V?  //
          C5 f t*y
          i   4 •/ -f
                   t? /•

/
        100    200    300    400    500    600

              Excess Annual Admissions
                                                     700
                       800

-------
                                          129
For example, the median risk estimate is 115  (90% C.I.  = 39-191) for the 8H1EX-0.08 ppm
standard and 120 (90% C.I. = 41-199) for the 8H5EX-0.08 ppm standard.
       It should be recognized that estimated O3-induced hospital admissions represents only
a small portion of the overall respiratory-related hospital admissions for asthmatics from all
causes. Another way to examine the risk results which highlights this point is presented in
Table V-20.  The excess admissions come from the hospital admissions risk model. The
estimates for asthmatic respiratory-related hospital admissions due to all causes are based on
(1) the 14-16 thousand admissions per O3 season estimates provided by Thurston (1995) and
(2) excess admissions attributable to exposures at O3 levels >  0.04 ppm.  As expected as the
population base for comparison increases the percentage change relative to admissions
associated with the current 1-hr standard decreases substantially.  For example, the excess
admissions associated with concentrations exceeding a 0.04 background results in a 42%
reduction  in admissions for the 8H5EX-0.08 ppm standard relative to just attaining the
current 1H1EX-0.12 ppm standard.  However, this represents  only a 12% reduction in O3-
induced excess hospital admissions when the contribution  of all O3 concentrations are
considered (i.e., background is set equal to 0 ppm).   Finally, if the comparison is made in
terms of all respiratory related admissions during the O3 season, the reduction associated
with attaining the 8H5EX-0.08 ppm standard relative to the current 1H1EX-0.12 ppm
standard is only 0.6 percent.
       6.      Assumptions and Limitations Associated with the Health Risk Assessment
       This section briefly summarizes a number of assumptions and limitations-should be
kept in mind in interpreting results of the O3 health risk assessment.  A  fuller discussion of
the assumptions and limitations is contained in Whitfield et al.  (1996).  These assumptions
and limitations include the following:
        (1)    Extrapolation of Exposure-Response Relationships.  In developing the
              probabilistic exposure-response relationships for the 1- and 8-hr health
              endpoints based on controlled human exposure studies it was necessary to
              extrapolate below the lowest exposure  level used in these studies (i.e., 0.08
              ppm for the moderate exertion studies  used to represent 8-hr exposure-
              response relationships, 0.08 or 0.12 ppm for the heavy and moderate exertion

-------
                                                  130
 TABLE V-20. Admissions of New York City Asthmatics — With a Comparison Relative to Meeting the
 Current Standard (1 h, 1 expected exceedance, 0.12 ppm)
Air Quality Scenarios
Issue
Excess Admissions1'
(background = 0.04 ppm)
% Change from Current
Standard'
Excess Admissions1'
(background = 0 ppm)
% Change from Current
Standard'
All Admissions'
(thousands)
% Change from Current
Standard*1
1H1EX-0.12
ppm
(Scenario A)
207C
(70, 344)d
0

909
(308, 1,509)
0

14.8
(13.8-15.8)
0

8H1EX-0.08
ppm
(Scenario C)
115
(39, 191)
-44

804
(273, 1,336)
-12

14.7
(13.7-15.7)
-0.6
(-0.2, 1.1)
8H5EX-0.08
ppm
(Scenario F)
120
(41, 199)
-42

797
(270, 1,324)
-12

14.7
(13.7-15.7)
-0.6
(-0.2, -1.1)
As-Is
(Scenario Z)
388
(132, 644)
87

1,065
(361, 1.768)
17

\y
(14-16?
1.2
(-0.4, 2.2)
'  Expected exceedance.

b  Admissions of asthmatics attributable to exposure to ozone.

c  Median estimate.

d  90 5c credible interval (about the median).

e  Because of the necessary assumption that results across scenarios are highly correlated (i.e..  if admissions are
  high for one scenario, they are high for all scenanos), there is very little variation in the percentage change
  from the current standard.

f  Admissions of asthmatics for any respiratory-related reason; for scenano /, based on estimates of all
  admissions and excess admissions attributable to ozone levels  >0.04 ppm for As-Is scenario, and  estimate ot
  excess  admissions attributable to ozone levels >0.04 ppm for scenano ; (e.g.. for scenano 1H1EX-0.12 ppm:
  14,800 * 15,000 - 388  -I- 207).

*  Admissions of New York City asthmatics  for any respiratory-related reason during the  1988-90 ozone seasons
  (Thurston, 1995).

h  Variation in  these results is attributable to the different numbers of admissions of New York  City asthmatics
  for  any respiratory-related reason during the 1988-90 ozone seasons (Thurston, 1995).

-------
                                    131
       studies used to develop 1-hr exposure-response relationships).  Based on an
       initial sensitivity analysis, a significant portion of the estimated risks are due to
       exposures between the estimated background value of 0.04 ppm and the lowest
       measured values in the various controlled human exposure studies relied upon
       in this risk analysis.  The CASAC O3 Exposure and Risk Subcommittee
       generally supported the extrapolation of modeled exposure-response
       relationships when they reviewed the proposed risk assessment methodology  in
       March 1994.
(2)     Exposure and air quality estimates.   A major  input to the headcount risk
       estimates for the general population, outdoor workers, and outdoor  children is
       the O3 exposure  analysis estimates for these populations.  Uncertainties about
       human activity patterns and the procedures used to estimate O3 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. There  are
       uncertainties about the appropriate O3 monitor to use in  applying the excess
       respiratory-related  hospital  admissions  risk model to the New  York City area
       and in the procedures used  to adjust 03 levels  to just attaining alternative air
       quality standards in the New York area.  Benchmark risk  estimates for the
       other  acute health endpoints are affected by uncertainty  in projecting O,
       concentrations upon attainment of alternative NAAQS at the design value
       monitor. In addition, the values selected as representing reasonable estimates
       of background concentrations for 1- and 8-hr daily maximum  levels are subject
       to uncertainty.  Since all of the risk estimates  presented here are calculated  as
       O3-induced risk in  excess of background, alternative values for background
       could potentially alter the risk estimates.   Argonne National Labs has
       conducted some  limited sensitivity analyses in  two of  the urban areas to
       examine the influence of different background assumptions on selected
       headcount risk estimates.  Generally, the results indicate that O3 exposures in
       the range  of 0.04-0.06 ppm contribute little to the total  headcount risk

-------
                                   132
       estimates.  Therefore, alternative values for background in the range 0.03-0.06
       ppm are likely to have little impact on the overall risk estimates.  The results
       of this limited sensitivity analysis is described in  more detail in Whitfield et al.
       (1996).
(3)     Age.  The risk assessment has been applied to the general population, outdoor
       workers, and outdoor children.  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 O, symptoms in  adults
       are not also present in children.
(4)     Attenuation or enhancement of response.  For the acute health endpoints, the
       risk assessment assumes that the O3-induced response in any particular hour is
       not affected by  previous 0, 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.
(5)     Interaction between O3 and other pollutants.  The controlled  human exposure
       studies used in  the risk assessment  involved only O3 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.
       Analysis of lung function decrement data from several field studies of children
       at a variety of summer camps in the northeast has found similar O3-induced
       lung function changes as observed in the controlled human exposure studies
       (see pp.9-7 and 9-8 of the CD).

-------
                                    133
(6)     Smoking status.  There is some limited evidence that smokers may be less
       responsive to O3 than nonsmokers.  The risk assessment was applied to the
       general population, outdoor workers, and outdoor children regardless of
       smoking  status.  To the extent that smokers are less responsive than
       nonsmokers, risk estimates may be overstated.
(7)     Selection of Averaging Time.  In developing risk estimates for 1-hr exposures
       at moderate and  heavy  exertion,  data from 1- to 2 '/2-hr controlled human
       exposure studies were used and matched with 1-hr exposures at moderate or
       heavy exertion.  The studies that ran longer than an hour were conducted with
       intermittent exercise periods.  McKittrick and Adams (1994) has reported that
       lung function responses were very similar for subjects exposed either
       continuously exercising for 1  hour or exposed for 2 hours with intermittent
       exercise.  In matching  the 1-hr exposure-relationships to 1-hr exposure
       estimates, the EVR range selected for the exposure estimates was selected to
       match the hourly average EVR in the health effects  studies.  The 8-hr,
       moderate exertion risk  estimates were developed based on three controlled
       human exposure studies that were conducted using a 6.6-hr exposure period.
       Since the lung function response appears  to level off after 4-6 hours of
       exposure, it is unlikely that the exposure-response relationships  would have
       been appreciably different, even if the studies had been conducted for 8 hours.
(8)    Reproducibility of O?-induced response.  It is assumed that O3-mduced
       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 O3 are highly  reproducible.  Analysis of Avol and Kulle
       data sets by Hayes et al.  (1987b) also supports  the reproducibility of individual
       responses.

-------
                                          134
I.      Alternative Forms of the Primary NAAOS
       1.     Form of the Current Standard
       The current primary O3 NAAQS has a level of 0.12 ppm, an averaging time of 1
hour, and is expressed in a "1 expected exceedance" form.  That is, the standard is
formulated on the basis of the expected number of days on which the level is exceeded.
More specifically, the attainment test specifies that the expected number of days per year on
which the level is exceeded be equal to or less than 1.0 (values less than 1.05 are rounded
down), averaged over a three-year period, and that specific adjustments be made for missing
monitoring data.  The standard is applied on a site-by-site basis; data from multiple sites are
not combined.  These procedures have remained unchanged since the original  promulgation
in 1979.
       2.     Issues Associated with Consideration of Alternative Forms
       As the above description of the current  standard illustrates, the "standard" is defined
by more than just its level.  The following elements have been used in the formulation of air
quality standards:
             - the level, e.g., 0.12 ppm,
             - the averaging time,  e.g.,  1  hour,
             - the NAAQS statistic, e.g.. the number of exceedances.
             - the attainment test criteria, e.g.. expected number of  exceedances equal to or
               less than  1.0,
             - the length of the compliance period, e.g., 3 years,  and
             - data handling conventions, e.g., adjustments for missing monitoring data.
       The staff is considering alternatives for the NAAQS statistic, the attainment test
criteria, and the data handling conventions which address some of the concerns about the
stability of the attainment test and the missing data adjustment procedure which  were  raised
during public review of the Clean Air Act O3 Design Value Study (EPA,  1994).  State
agency and industry representatives  expressed concern about:
(1)    the probability of misclassification between attainment and nonattainment,
(2)    the possibility of areas moving in and out of attainment ("flip/flops") with each
       additional year of data,
(3)    the significant impact of year to year variability in meteorological  conditions
       conducive to O3 formation,

-------
                                           135
(4)     the need for a more "robust" test statistic (e.g., more exceedances or a lower
       percentile statistic)
(5)     the impact of using only a single monitor within a large network of monitoring
       stations,
(6)     the lack of consideration of population exposure,
(7)     the impact of the adjustment for missing data (e.g.,  areas with only three observed
       exceedances can fail to meet the standard if they have less than 95 percent data
       completeness), and
(8)     the impact of O-^ transport on downwind areas.
Some of these  concerns will be addressed  in this standard review,  while others will be
considered in the development of associated new implementation strategies.
       As part of this standard review, the staff is evaluating alternative approaches to
specifying the  NAAQS statistic and attainment test criteria  that are designed to 1) better
reflect the relationship between air quality and human exposure and risk; 2) increase the
stability and, hence, reduce the likelihood of attainment/nonattainment flip-flops; and 3)
address missing data  issues.   The approaches currently under consideration by the staff
include:
•     alternative, less variable NAAQS  statistics,
•     alternative attainment  test criteria,  including the use of a range, rather than a bright-
       line  standard, and
•     alternatives for the treatment of missing data.
       Of foremost consideration in evaluating alternative forms for the primary standard is
an  assessment  of the adequacy of the health protection provided.  The staff is considering
whether alternative forms address concerns raised with regard to the current standard without
introducing  other problems of equal, or greater concern.  The staff is also giving major
consideration to the feasibility of implementation and the infrastructure needed to implement
alternative forms, such as the adequacy of the current ambient monitoring network.
 Public policy issues associated with these various alternative approaches that the staff is
 considering  include consistency with Clean Air Act requirements,  environmental equity

-------
                                           136
considerations, and the ability to effectively communicate any change in the standard to the
public at large.
       The staff has determined that it is more appropriate to consider  misclassification and
transport issues within a new implementation strategy framework.  For example, the
California Air Resources Board (CARB) has already implemented an approach to address
misclassification by defining a  nonattainment "transitional" category.  Although this new
category helps to reduce the probability of misclassifying borderline sites, there are important
implementation considerations if such an approach were to be introduced.  These include
how long an area can remain in this category, and what emissions reductions, if any, would
be required.  The level  of health protection intended by the NAAQS is not altered by this
approach as long as the area eventually comes into attainment with the NAAQS.
       3.      Alternative NAAQS  Statistics
       The NAAQS statistic for the current standard is the annual expected exceedance rate.
The staff has considered the use of the design value, which is a concentration-based statistic,
as an alternative NAAQS statistic.  Use of a concentration-based statistic is consistent with
health concerns, and is  also consistent with the form of the current standard, in that the level
of the standard necessary to provide a given degree of  health protection would be the same
for a design value NAAQS  statistic standard as an annual expected exceedance form.  The
primary reasons for considering such a change are that  (1) the  design value has greater
temporal stability than expected exceedances,  (2) it is more directly related to the database
characterizing health and welfare effects, and  (3) the dichotomy with currently exists between
the expected exceedances attainment test and the design values is eliminated because with a
concentration based standard the attainment test and the design value  are the same statistic.
       One approach to increasing  stability in the air quality management process is to
specify a less variable NAAQS statistic.  Use of a less  variable NAAQS statistic would result
in a reduction in the number of borderline sites with relatively high misclassification
probabilities, and thus there will be fewer  reversals in compliance status ("flip-flops"), at
least until some of the other nonattainment sites are brought close to attainment through
emission reductions.  However, use of a less variable statistic cannot reduce misclassification
probabilities for sites on the borderline between attainment and nonattainment, and for those

-------
                                          137
sites there will likely be frequent classification changes if the emissions remain fairly
constant from year to year.
       Chock and Nance (Chock et al., 1993) suggest the use of the mean or median of the
three annual second daily maxima (AvgMax2) as the NAAQS statistic. The staff has
examined the relationship between the AvgMax2 statistic and a one exceedance NAAQS.
The AvgMax2 statistic  is about 6 percent lower, on average, than the  design value
concentration for a 1-hour 1  exceedance NAAQS (i.e., the fourth highest daily maximum
concentration over three years), and 10 percent lower for an 8-hour 1  exceedance alternative
based on ambient monitoring data from the last ten years.  On the basis of exceedances, the
AvgMax2 lies between the one exceedance and 5 exceedances alternatives.  The top-half of
Figure V-18 presents a histogram of the average number of daily maximum exceedances for
277 sites just attaining  an  AvgMax2 standard of 0.08 ppm based on 1991-93 data.  Thirty
percent of the sites had average exceedance rates greater than 1, with  two sites having an
average of 5 exceedances per year.  The bottom half of Figure V-18 presents the maximum
number of exceedances in the worst year during 1991-93 for the same comparison.  Half of
the sites have two or fewer exceedances in the peak year.  Four  sites  had  10 or more
exceedances in the worst year.  Figure V-19 repeats this presentation  for a fifth highest daily
maximum 8-hour standard formulation  (AvgMax5). To summarize  in terms of exceedances.
on  average,  sites meeting  an average annual  2nd highest daily maximum 0.08 ppm standard
have 1.2 exceedances per  year, and 2.3 exceedances in  the worst year of three, while sites
meeting an average annual 5th highest daily  maximum 0.08 ppm  standard have 3.0
exceedances per year,  and 5.4 exceedances in the worst year of three. Also, in the worst
year of three, 95 % of sites meeting  the average annual 2nd highest daily maximum 0.08
ppm standard  have 7 or fewer exceedances, while 95  %  of sites  meeting  an average annual
5th highest daily maximum 0.08 ppm standard  have 12  or fewer exceedances.
       Chock has also suggested the use of the  95th, or other percentiles as the test statistic.
Various parametric approaches can be used to estimate the concentration  percentiles by fitting
a distribution to the daily  maxima.  Of particular importance is the  tail exponential
distribution, both because of certain statistical asymptotic properties of that distribution, and
because it often fits the observed concentrations.  One approach  (Breiman et al., 1978) fits

-------
                                    138

Figure V-18. Frequency distribution of the average and maximum number of exceedances of
           0.08 ppm 8-hour daily maximum concentrations for sites just attaining an
           average annual second highest daily maximum standard of 0.08 ppm.

             Average Number of Exceedances for Sites
           Just Attaining an Average Second Max 8-hour
         Standard Equal to 0.08 ppm based on 1991-93 Data
Number of Sites
          1    2   3   4    5    6    7   8   9   10   11   12   13
           Number of Exceedances of 0.08 ppm Daily Max 8-hr
          Number of Exceedances in Worst Year for Sites
           Just Attaining an Average Second Max 8-hour
         Standard Equal to 0.08 ppm based on 1991-93 Data
Number of Sites
          1    2   3   4    5    6    7   8   9   10   11   12
           Number of Exceedances of 0.08 ppm Daily Max 8-hr
13

-------
                                139

Figure V-19. Frequency distribution of the average and maximum number of exceedances of
          0.08 ppm 8-hour daily maximum concentrations for sites just attaining an
          average annual fifth highest daily maximum standard of 0.08 ppm.

            Average Number of Exceedances for Sites

            Just Attaining an Average Fifth Max 8-hour

      Standard Equal to 0.08 ppm based on 1991-1993 Data

   Number of Sites
   120
       0  1  2 3  4  5  6  7  8  9  10 11 12 13 14  15 16 17  18 19 20 21 22

              Number of Exceedances of 0.08 ppm Daily Max 8-hr
            Number of Exceedances in Worst Year Sites
            Just Attaining an Average Fifth Max 8-hour
       Standard Equal to 0.08 ppm based on 1991-1993 Data
   Number of Sites
   100
          1  2  3  4 5  6 7  8  9 10 11 12 13  14 15 16 17 18 19 20 21 22

              Number of Exceedances of 0.08 ppm Daily Max 8-hr

-------
                                           140
the exponential distribution to the top 5 or 10 percent of the data.  Use of the "tail-
exponential" approach  can yield a more stable NAAQS statistic than the current expected
exceedance statistic provided that the exponential assumption is applicable.  For example,
Larsen used a version of Breiman's approach (CARB,  1992) to derive estimates of the one in
one year percentile based on three years of date, and the California Air Resources Board
(CARB)  uses this approach when determining attainment of the California O^ standard.
       Fairley and  Blanchard have proposed  several other alternatives (Fairley et al., 1991)
including the use of (1) averages of annual high percentiles (such as the mean of three annual
95th percentiles), (2) averages of a certain number of the highest O-^  concentrations, and (3)
spatial averages  (across sites).  Using averages  of a certain number of the highest O^
concentrations is very similar to using  the tail exponential approach since  the tail exponential
design value is a high percentile plus a multiple of the average excess above that percentile.
Most  of these  forms can be  viewed as  multiple  exceedance standards.  For the typical O^
monitoring season,  the 95th percentile  translates into an average of 10 exceedances per year
of the specified  standard level.   As noted  previously, the staff is currently not considering
alternatives with an average of more than  5 exceedances per year.
       The staff is  also addressing how concerns about the spatial  representativeness of
monitoring sites and population exposure  might be incorporated into the form of the
standard.  However, any consideration of some form of spatial averaging  or population
weighting across monitoring  sites raises issues  about environmental equity, the adequacy of
the current monitoring network, and the specificity  of  monitor siting  requirements.  On the
other hand, such a  conceptual approach may  better  reflect population exposure and risk.  The
staff is also considering whether concerns about population exposure  might also be addressed
by monitor siting guidance and control strategy assessments as part of the implementation
process.
       For alternative NAAQS statistics other than  the design value,  such as those discussed
above, the level of the standard  would need to  be established as a function of the NAAQS
statistic.  Comparisons of the level of  health  protection can not be specified exactly for
alternative NAAQS statistics, since health protection is a function  of  the entire distribution of
ambient concentrations that  would exist when a standard is attained, and the resultant impacts

-------
                                           141
on that distribution will be different for different forms.  Alternative forms can, however, be
compared on average, although the spatial distribution of protection (in terms of reduced risk
of adverse health effects) would vary from one form of the standard to another.
       4.     Alternative Attainment Test Criteria
       The current NAAQS is a "bright-line" standard.  This means that a site is either
attainment or nonattainment.  Specifying a standard in this way conceptually relates best to
pollutants that exhibit health effects thresholds.  Staff is  now considering specification of a
range, rather than a bright-line standard.  Such an approach is intended to recognize the
absence  of discernible health effects thresholds and the projections that population risk varies
little with small changes in air quality.  Use of a range for the specification of a standard is
conceptually a way to recognize the continuum of risk associated with varying levels of O^
exposure and, thus, O^ air quality.  Within this context, the staff is considering whether the
specification of a range rather than a bright-line standard would help to facilitate individual
and/or regulatory agency efforts to provide additional safeguards against responses that may,
in a small number of particularly sensitive individuals, occur at levels even below the level
of a standard that protects public health with an adequate margin of safety.
       5.    Alternatives for Treatment of Missing Values
       The formulation  of the current standard includes procedures  for dealing with
 incomplete monitoring data.  These missing data procedures assume that missing O^ values
 during the O^ season follow the same patterns as the non-missing values.  A missing day
 during the O^ season is assumed to be less than the level of the  NAAQS only if it is a single
 missing  value that occurs between  two valid daily maximum  O^ measurements that are less
 than  75  percent of the level of the NAAQS.  Because of the  rounding convention established
 for estimated exceedances for the current standard, the missing data adjustment only becomes
 a factor when more  than 5 percent of the days are missing, i.e., more than  10 days for the
 typical O-} season.
        The  staff is exploring two different conceptual approaches to  the treatment of missing
 data:  (1) requiring a greater degree of data completeness to  demonstrate compliance with  the
 standard than noncompliance and (2) the use of simple  statistical procedures to account or
 adjust for missing data.

-------
                                          142
       In the first case, the staff is considering an average data completeness requirement
that monitoring sites would have to meet to demonstrate attainment of the standard.  Based
on a review of the current monitoring network,  more than 80 to 90 percent of all sites
achieve better than 90 percent data completeness.
       Secondly, the staff is evaluating whether information on meteorological conditions
could be used to provide an objective procedure for judging if meteorological conditions on a
missing day were not conducive to exceedances of the O^ NAAQS,  then that day could be
assumed to be less than the level of the  NAAQS.  The literature contains numerous
references to the use of meteorological data to define O^ conducive days for peak 1-hour
concentrations (Jones.  1985; Kolaz et al., 1990; Jones, 1992). Table V-21  adapted from Chu
(Chu. 1995) presents a set of criteria for O^ conducive conditions for  the eastern United
States.  These criteria represent, to a great extent, the necessary conditions for daily
maximum 1-hour O^ concentrations to exceed 0.12  ppm. While these are conditions
necessary for high  O^. they are not necessarily  sufficient conditions.  Other factors may also
be important.  Additional analyses would be needed to define the necessary conditions for
peak 8-hour concentrations in  the ranges of concern.  The first step of such an approach
might also include  a review of ambient concentrations recorded at other monitoring sites in
the area for possible exceedances  of the NAAQS and the historical relationships among
nearby monitoring  sites.  The  meteorological data would not be viewed as a surrogate for O^
monitoring data. Such an approach presumes the availability  of meteorological data, such  as
a nearby National Weather Service Station.  It also  means that the attainment status of sites
near the level of the standard with a  large number of missing days could not be determined
directly from the ambient data base, but rather would  require an analysis  of the
meteorological data as  well.

-------
                                        143
     TABLE V-21.  CRITERIA FOR O3 CONDUCIVE CONDITIONS FOR THE
                                  EASTERN U.S.
1. T _> 26.5°C for cities north of 40°N,
  T >. 29°C  for cities between 35°N and 40°N,
  T >L 32°C  for cities south of 35°N.

2. Wa.m. _<_ 5 m/s for cities in transport regions (i.e., Midwest and Northeast),
  Wa m  <_ 4 m/s for cities outside transport regions.

3. W  m _<_ 7.5 m/s for cities in transport regions,
  Vv'  m  _<. 6 m/s for cities outside transport regions,
  WD m  <. 5 m/s for Gulf Coast cities and  Florida.

4. RH  <_ 75% for coastal cities north of 40°N,
  RH _<_ 65% for inland cities between 30°N and 40°N,
  RH <_ 70% for all cities south of 30CN.

Adapted from Chu (1995).

-------
                                          144
VI.  STAFF CONCLUSIONS AND RECOMMENDATIONS ON PRIMARY NAAQS
       This section presents staff conclusions and recommendations for consideration by the
Administrator in selecting a pollutant indicator, averaging time, form, and level of the
primary O3 NAAQS.  In developing these conclusions and recommendations, the staff has
drawn upon the scientific and technical information contained in the CD and summarized in
Section V of this Staff Paper, the exposure and risk analyses presented in Section V, and
comments provided by the CASAC and the public on drafts of this Staff Paper.
       The staff has attempted to integrate information on acute and chronic health effects of
O3, the expert judgments on the adversity of such effects, and,  when possible, quantitative
assessments of the risk of experiencing such  effects into a basis for conclusions and
recommendations on the primary O3 NAAQS.  This approach recognizes that for most of the
health effects associated  with O3, no population threshold can be clearly identified.  Thus, the
approach taken here uses assessments of exposure and risk, when possible, to provide
additional insight and  to inform judgments about  the protection of public health with an
adequate  margin  of safety.
       As discussed in Section V.H, quantitative risk assessments have been premised on
extrapolating exposure-response functions from lowest observed effect  levels down to
background  levels.  Thus, these assessments  reflect a continuum consisting  of levels at which
health effects are certain  through levels at which  scientists generally agree that health effects
may occur but the likelihood and magnitude of the response is more uncertain.  The
integrated approach taken in this Staff Paper  links risk assessments, which preside estimates
of how many people are likely to experience various effects, with consideration of the degree
of severity of the effects as bases for judgments about the point at which risks have been
reduced sufficiently to achieve protection  of public  health with an adequate margin of safety.
       In recommending a range of options for the Administrator to consider,  the staff notes
that the final decision  is  largely a public health policy judgment. A final decision must draw
upon scientific information  about health effects and risks,  as well as a series of judgments:
1) about  when physiological effects become adverse from a public health perspective, as
discussed in Section V.F of this Staff Paper,  2) the relative severity of various effects with
estimates of the expected incidence of those effects, and 3) how to deal with the range of

-------
                                          145
uncertainties that are inherent in the evidence and assessments.  This approach is consistent
with the requirements of the NAAQS provisions of the Clean Air Act (Act) and with how the
EPA and the courts have historically interpreted the Act.  These provisions do not require the
Administrator  to establish a NAAQS at a zero-risk level but rather at a level that avoids
unacceptable risks  and, thus, protects public health with an adequate margin of safety.
       The following staff conclusions and recommendations are based  primarily upon those
analyses discussed  above and in Section V of the Staff Paper, staff judgment regarding  those
analyses, and the comments provided by the CASAC and the public.
A.     Pollutant Indicator
       The staff believes that the conclusions on the appropriate indicator for the primary O3
NAAQS that were presented in the previous Staff Paper (USEPA, 1989) remain valid today.
As indicated in the previous Staff Paper, it is generally recognized that control of ambient  O3
levels provides the best  means of controlling photochemical oxidants of potential health
concern.  Further,  among the photochemical oxidants, the acute-exposure chamber,  field, and
epidemiological human health data base raises concern only for O3 at levels of photochemical
oxidants commonly reported in ambient air. Thus,  the staff recommends  that O3 remain as
the pollutant indicator for protection of public health from exposure  to all photochemical
oxidants found in the ambient air.
B.     Averaging Times
       1.  Short-Term and Prolonged (1 to 8 hours)
       The current primary O3 NAAQS was set in 1979 with a 1-hr averaging time. This
was intended to protect  the public against  the health effects associated with 1- to 3-hr
exposures to O3 in addition to the health effects potentially associated with longer-term  O3
exposures which were not as well documented at that time.
       Since 1979, a numerous researchers have investigated the health effects associated
with short-term (1- to 3-hr) and prolonged acute (6- to 8-hr) exposures  to O3.  Numerous
controlled-exposure studies of human subjects, who engaged in activities involving  heavy and
moderate exertion, provide a basis for quantitative concentration-response relationships
between 1- to 3-hr O3 exposures and a variety of lung function parameters and respiratory
symptoms.  Also,  field and epidemiological studies now provide additional  evidence of

-------
                                          146
associations between 1-hr ambient O3 levels and health effects ranging from respiratory
symptoms and lung function decrements reported at summer camp studies to increased
hospital admissions for respiratory causes.  However, the field and epidemiological studies
have not been analyzed sufficiently as yet to determine whether the observed effects correlate
as well or better with 6-  to 8-hr exposures as with the 1- to 3-hr exposures.  More recent
controlled-exposure studies have been conducted providing  evidence that  the same respiratory
effects (i.e., lung function decrements and respiratory symptoms) occur when human subjects
are exposed to O3 while engaging in activities involving intermittent, moderate exertion for
prolonged exposure periods of 6 to 8 hrs. These effects occur at lower concentrations of O3
and at less severe exertion levels than for 1- to  3-hr exposures.   Other effects, such as the
presence of biochemical indicators of inflammation and reductions in pulmonary defense
mechanisms leading to increased susceptibility to infection, have also been reported for
prolonged exposures and, in some cases, for short-term exposures.
       This brief summary of the averaging times associated  with various acute health effects
highlights that averaging  times of 1 to 3 hrs and of 6  to 8 hrs have both been associated with
a wide range of observed respiratory effects caused by O3 exposure.  The current 1-hr
averaging time is judged  to be most appropriate to address  acute health effects associated
with 1- to 3-hr exposures because these effects  typically occur within the first hour of
exposure, during  moderate and heavy exertion.   On the other hand, an 8-hr averaging time is
judged to be more appropriate for addressing similar health effects associated with 6- to 8-hr
exposures, since health effects typically build up over time  in moderately exercising subjects.
approaching a plateau somewhat beyond the 6.6 hr exposure periods for which most of the
prolonged exposure studies have been conducted.  Furthermore,  it is generally convenient to
assess air quality  and  exposure patterns in 8-hr  time periods.
       In selecting an averaging time or times for the primary O3 NAAQS, questions arise as
to whether both a 1-hr NAAQS and an 8-hr NAAQS  are necessary and appropriate to protect
public health, and, if both are not needed, which averaging time is more  appropriate.
       The primary way  in which these questions have been addressed in this draft Staff
Paper is through the quantitative risk analyses presented in  section V-H.  These analyses
produce estimates of the  reduction in the estimated risks of both  1- and 8-hr effects

-------
                                           147
associated with attaining the current 1-hr standard and several alternative 8-hr standards.
Attainment of any particular 1-hr or 8-hr standard was modeled by projecting a change in the
entire air quality distribution sufficient to just attain the  standard.   The resulting air quality
distribution  is then analyzed in terms of both 1-hr and 8-hr average concentrations to develop
risk estimates for certain health endpoints associated with 1-hr and 8-hr exposures.  These
results show that attaining either the current 1-hr standard or a 0.10 ppm, 1-hr standard
reduces the  risk of experiencing  health effects associated with either 1-hr or 8-hr O3
exposures in areas  that do not currently  attain  the 0.12 ppm, 1-hr  standard.  Likewise,
attaining most of the alternative 8-hr standards examined reduces  the risk of experiencing
health  effects associated with either 8-hr or 1-hr O3 exposures in areas currently exceeding
the 0.12 ppm,  1-hr standard. Based on  these analyses,  the staff believes that adequate
reductions in risks  from both 1-hr and 8-hr effects can be achieved through a primary
standard with an averaging time  of either 1 or 8 hrs.  Staff judges that the  8-hr averaging
time is most directly associated with health effects of concern at the lowest concentration of
O3.  As a result, the staff concludes that the establishment of both 1-hr and 8-hr standards is
not necessary to reduce the risks associated with the range of acute effects considered in
these analyses.
        The  staff also has given consideration to the question of whether both a  1- and 8-hr
standard are appropriate.  In the staff s judgment, two short-term  standards could be
appropriate if the combination of two such  standards were determined to be a more efficient
way to provide public health protection  than through a single  standard with an averaging time
of either 1 or 8 hours.  A combination of standards may be more  effective if, for example, a
 1-hr standard would need  to be set at a  significantly lower level to provide adequate
protection from 8-hr effects than would  otherwise be necessary to provide adequate
protection from the 1-hr effects, and if such a standard  would, in  effect, represent
overcontrol in some geographic areas as a result of varying  air quality patterns.  In looking
at risks associated  with changes  in lung  function and the areas that would be impacted  by
alternative standards,  staff judges that neither  the 1-hr nor 8-hr averaging times appear to

-------
                                             148
  have any advantage in efficiency.  Based on the above, relative to the 1-hr alternatives,  staff
  concludes that establishment of an 8-hr standard would likely be more directly related to
  providing public health protection, increased stability, and, by taking into account more air
  quality data, would be a more robust standard.
         2.     Long-Term
         There is a very large animal  toxicology data base providing clear  evidence of lung
  tissue damage, with additional evidence of reduced lung elasticity and loss of lung function,
  caused by exposures to higher levels of simulated ambient O3, though not at or below 0.12
  ppm O3, lasting from a few months  to years.  Although there have been substantial  recent
'  advancements in dosimetry  extrapolation from animals to humans (see CD, Chapter 8),
  further research in the area  of species sensitivity must be conducted before quantitative
  linkages to specific health effects in  human could  be established with known uncertainty.
         Further,  since there  is considerable uncertainty regarding the temporal patterns and
  levels of exposure that might be most directly associated with any such chronic effects,
  should they occur, in humans (i.e.,  the importance of the occurrence and pattern of repeated
  short-term  and/or prolonged peaks relative to cumulative total exposure), it is not possible to
  evaluate the extent to which either a 1- or  8-hr standard would contribute to protecting
  against any such effects.  On the other  hand, it is likely that the alternative 1-  and 8-hr
  standards under consideration would directionally  provide protection against  such  effects
  should they occur, in that such alternatives would result in  both lower short-term  and
  prolonged peaks as well as  lower overall O3 concentration distributions which would reduce
  cumulative long-term exposures.  Thus, until additional research and related analyses have
  been conducted, the staff believes that consideration of a separate long-term  O3 NAAQS is
  not appropriate.
  C.     Form of the Standard
         Based upon  information contained in the CD and sections IV and  V.I of this Staff
  Paper, and on discussions and comments received during the review of the Clean  Air Act
  Ozone Design Value Study, the staff has reached  the following conclusions on the form of
  the standard.  Staff concludes that several test statistics should be considered in specifying
  the form of any new or revised primary standards.  Such test statistics include the expected

-------
                                           149
exceedance rate, including both the 1-expected exceedance (the basis for the current standard)
and multiple exceedances (up to 5 expected exceedances) per year averaged over three years.
as well as concentration-based  test statistics, including in particular the average second to the
fifth highest daily maximum 8-hr concentration averaged over three years.  In addition,
specifying the standard in terms of a range of air quality values (e.g., the second  to the fifth
highest daily maximum 8-hr average concentration, averaged over a 3-year period) should
also be considered.  In conjunction with such alternative statistics, some form of spatial
averaging or population weighting across monitors may also warrant consideration.
D.     Level of the Standard
       As discussed at the beginning of this section, the staffs approach to formulating
recommendations with regard to an appropriate range of standard levels focuses on general
conclusions regarding  lowest observed effect levels and a qualitative assessment of evidence
regarding health effects for which no quantitative estimates of risks  were developed,  together
with quantitative risk assessments for selected health effects to provide additional  input into
consideration of an adequate margin of safety.  The staff's conclusions presented  in this Staff
Paper are informed by qualitative evidence discussed in  section V.D, judgments about
adversity discussed in  section V.F, and exposure and risk estimates for selected health
endpoints for sensitive population groups summarized in sections V.G and V.H and
Appendices B and C.  Consistent with the above conclusions on averaging times,  the
following discussions and conclusions are primarily directed toward identifying a  range of
levels associated with  alternative 8-hr standards for consideration by the Administrator in
selecting a standard(s) that, in  her judgment, would  reduce risks  to public  health sufficient!)
to protect public health with an adequate margin of safety.
       1.     General Conclusions
       Taking into account information on health effects, sensitive and at-risk populations,
and adversity of effects contained in the CD and in section V of this Staff Paper,  the  staff
has drawn the following conclusions with regard to effects that the staff judges are of
particular importance in considering the need for new or revised  primary O, NAAQS.
•     In controlled-exposure human studies, the lowest range within  which 1- to 3-hr
       exposures to O3 at heavy exertion have induced group mean  statistically significant

-------
                                           150
       lung function decrements is 0.12 to 0.16 ppm, and the lowest range within which 6-
       to 8-hr exposures to O3 at moderate exertion have induced group mean statistically
       significant lung function decrements is 0.08 to 0.12 ppm.  In epidemiology studies,
       similar effects have been associated with short-term ambient O3 exposures below 0.12
       ppm when subjects were engaged in physical activity.
 •     In controlled-exposure human studies, the lowest range within which 1- to 3-hr
       exposures to O3 at heavy exertion have induced group mean statistically significant
       respiratory symptoms, including cough and pain on deep inspiration, is 0.16 to 0.18
       ppm, and the lowest range within which  6- to 8-hr exposures to  O3 at moderate
       exertion have induced group  mean statistically significant respiratory symptoms is
       0.08 to 0.12 ppm.
       These lung function and symptoms effects are based on numerous controlled human
exposure and field studies of both healthy and respiratory-impaired  (e.g., asthmatic) subjects.
As discussed in section V.F, the staff concludes that these effects are adverse to healthy
individuals and those with  impaired respiratory systems experiencing these effects at levels
characterized as severe in Table V-5, and that the adversity of effects levels categorized as
moderate for healthy individuals is a function of the  number of times an affected individual
would experience such effects, and is a matter for the Administrator's judgment.  The staff
also  concludes that  the population  group at greatest risk for experiencing lung  function effects
is active outdoor children (i.e., children who typically play outdoors during  summer when  O3
levels are highest) engaged in physical activity,  with outdoor workers engaged in physical
labor also being at increased risk relative to the  entire population.  Both outdoor workers and
outdoor children are at increased  risk for experiencing symptoms, based on exposure
estimates, although children do not typically report symptoms to the same degree as adults.
These groups engage in activities requiring exertion at levels that are associated  with
significant lung function decrements  and symptoms.
       In  making judgments about the level at which public health protection from these
effects incorporates an adequate margin of safety, the staff believes that it is important, when
possible, to  consider (1) the extent to which at-risk groups  are likely to  be exposed to
ambient concentrations associated with such adverse effects, (2) the mechanisms by which

-------
                                           151
they occur, and (3) the resulting risk of experiencing adverse effects predicted for these at-
risk groups.  The exposure and risk analyses that have been done to further inform this
decision take into account the significant variability in responses that have been observed in
these studies, in that some individuals experience lung function decrements and symptoms
both greater than and less than  the group mean.  Furthermore, these analyses recognize that
there is no indication that a threshold exists at the lower end of these ranges.  Drawing from
the results of the risk analysis presented in section V.H, the next section summarizes risk
information that the staff believes is relevant to the Administrator's consideration of an
adequate margin of safety with regard to lung function and symptom effects associated with
both  1- to 3-hr and 6- to 8-hr exposures  to O3.
•     The lowest observed effects level at which 1- to 3-hr exposures to O3 at very heavy
       exercise have induced increases in nonspecific bronchial responsiveness in  healthy
       adults is 0.18 ppm, and the range at which 6-  to 8-hr exposures to O3 at moderate
       exertion have induced such increases  in nonspecific bronchial responsiveness is 0.08
       to 0.12 ppm.  Exercising asthmatic individuals experience larger increases in
       nonspecific bronchial responsiveness at lower O3 exposures, but evidence is too
       limited to draw quantitative conclusions at this time.
       Nonspecific bronchial responsiveness is an indication of an individual's susceptibility
to stimuli such as antigens, chemicals, and particles, has been demonstrated in both human
and animal  studies.  Staff believes that increases  in nonspecific bronchial  responsiveness have
the potential to aggravate asthma and other types of preexisting respiratory impairment, and,
thus, staff concludes that this effect may be adverse for some  exercising individuals with
 significantly impaired respiratory symptoms  at levels  characterized  as moderate in Table V-5.
 •     The lowest level at which 1- to 3-hr exposures to O3 of healthy adults engaged in
       activities  involving very heavy exertion have been tested and have induced
       biochemical  indicators of pulmonary  inflammation is 0.20 ppm, and the range for
       which 6-  to 8-hr exposures to O3 of healthy adults engaged  in moderate exertion have
       induced this effect is 0.08  to 0.10 ppm.
       These indicators of inflammation have been observed in both controlled human
 exposure studies and in experimental animal studies.  While there is divergent opinion as to

-------
                                           152
the clinical significance of a singular occurrence of acute pulmonary inflammation, the staff
believes that based on scientific evidence repeated occurrences of acute pulmonary
inflammation over periods lasting months to years have the potential to result in structural
changes in the lungs for which there is suggestive evidence of an association with permanent
respiratory injury and/or progressive dysfunction.  This view is supported both by the
linkages that have been demonstrated between the  biochemical indicators of inflammation
identified  in the fluids extracted from the lungs of humans after short-term and prolonged
exposures to O3 and the structural damage which has been reported in  laboratory animals as a
result of long-term exposures to O3.
       In  making judgments about the standard level at which public health is protected
against indicators of acute pulmonary inflammation with an adequate margin of safety, the
staff believes that the exposure analysis presented  in Section V.G provides useful
information.  Such information is summarized in the next section under margin of safety
considerations.   No risk assessment has been conducted for the effect due to (1) the limited
amount of data which is insufficient to derive exposure-response relationships and (2) the
uncertainties which remain in the dosimetric extrapolation of animal to human data (as
discussed  in Chapter 8 of the CD), as well as large observed differences in species sensitivity
between humans and laboratory animals that is yet to be adequately understood.
•      Evidence from animal toxicology studies suggest that acute exposures to O, in the
       range of 0.08 to 0.10 ppm can induce pulmonary changes that decrease  the
       effectiveness of the lung's defenses against bacterial lung infections.
       The staff concludes that there is adequate evidence to  reasonably anticipate that such
reductions in the human defense mechanisms could result  in increased  susceptibility to
pulmonary infection.  This conclusion is based in large part on the existence of a substantial
animal toxicology data base  which indicates that O3 increases susceptibility of experimental
animals to respiratory infection.  Although few controlled human exposure studies have been
conducted to assess the impact of exposing human subjects to O3 and a bacterial challenge,
those that  have been conducted have provided insufficient evidence of  increased susceptibility
to  infection caused by O3. This may well be a result of the extremely cautious manner in
which these studies must be conducted when  using human subjects with infectious material.

-------
                                           153
       However, despite limited human data, a biologically plausible case can be made for
prolonged exposures to O3 increasing human susceptibility to respiratory infection.  There
exist many similarities between laboratory animals and humans with regard to many of the
host defense mechanisms used to defend against infections of the lung.  When the ability of
the lungs of either humans or animals to destroy invading microbes or to remove inhaled
paniculate  matter is adversely affected by inhaled O3, it is reasonable to anticipate that there
will be an increased risk of developing respiratory infection. Depending on the level of O3
exposure, the period of time or number of times exposed, and the susceptibility of the
individual exposed, the resulting respirator}' infection could be relatively minor or result in
the need for hospitalization.
•      Exposures to O3, as currently experienced in several cities in the eastern United States
       and Canada, are associated with excess hospital admissions and emergency room
       visits for respiratory causes,  with evidence of this effect  occurring to some extent
       even when hourly O3 concentrations are as low as 0.08 to 0.10 ppm.
       In a number of epidemiological studies, linear, nonthreshold associations have been
reported between the  daily maximum hourly O3 concentration on the day prior to admissions
and an increase in hospital admissions and emergency room visits for respiratory causes.
These  effects have  been attributed primarily to O3 exposures, since the effects of copollutants
and other confounding factors were judged in the CD to be adequately accounted for in these
analyses.  The  biological plausibility of O3-related increases in hospital  admissions is further
supported by the controlled human exposure data showing Cyinduced increases in
nonspecific bronchial responsiveness and the animal toxicology, data noted above with respect
to increased susceptibility to respiratory infection.
       In order to provide for more informed judgments regarding which standards would
reduce risks to public health sufficiently to protect public health with an adequate margin of
safety, the staff has conducted a quantitative risk assessment, which is discussed in Section
V.H.  That risk assessment,  which uses air quality monitoring data and  hospital admissions
data from New York City, is summarized in the next section under margin of safety
considerations.  Further, the staff believes that it is not yet possible to assess any association
between either excess hospital admissions or emergency room visits and 6- to 8-hr exposures

-------
                                           154
because most of the basic data have not been analyzed in terms of 8-hr average
concentrations at this time, although such an association with 8-hr exposures has been
reported.
•      An association between daily mortality and exposure to O3 in an area with very high
       O3 levels (i.e., Los Angeles) has been suggested, although the magnitude of such an
       effect remains unclear at this time.
       Reanalysis of  1970's data from Los Angeles County suggests that O3 exposures are
associated with a small, but statistically significant, portion of day-to-day variations in  total
daily mortality in that city, where hourly  O3 concentrations >0.20 ppm occur, over a  10-
year period.  However, the researchers who conducted the reanalysis emphasized that since
statistically significant associations have been detected among both  mortality and
environmental variables, one can not conclude with confidence that an association with
mortality is causal based on results from their observational study.  In another epidemiology
study no such association was seen where hourly O, concentrations were <0.15 ppm.  Other
studies of this potential  effect have been confounded by copollutants. especially particulate
matter, and by inadequate methods to characterize exposure or to account for other
confounding factors.  Based on the available published evidence, the staff believes that
protection against this potential effect would likely result from  any O, standard that is
protective of other effects discussed above,  such as increased hospital admissions and
susceptibility to pulmonary infection.
       2.      Margin of Safety  Considerations Based on Quantitative Exposure and Risk
              Assessment
       The following discussion presents  summary results and  observations of exposure and
risk drawn from the quantitative exposure and risk assessments presented in sections V.G.
and V.H.  and Appendices B. and  C..  This information  is intended to provide additional
insight about the extent to which at-risk populations  may experience the specific health
effects addressed in these analyses when various alternative standards are just attained.  The
staff believes that such information, when available, is useful to inform judgments about
which standards would reduce risks  to public health  sufficiently to protect public health with
an adequate margin of safety.

-------
                                           155
       The staff believes that the exposure and risk assessment methods used in these
analyses represent the state of the art at the present time, and that these analyses provide
reasonable estimates for the purposes intended.  The staff cautions, however, that in  light of
the many  sources of uncertainty inherent in such analyses, the results should not be
interpreted as precise measures of exposure and risk.  Some important  uncertainties inherent
in the analyses include (1) the air quality adjustment procedures used to simulate just
attaining the  alternative standards, (2) the specification of activity patterns and associated
exertion levels for the population groups of interest based on limited activity diary data, (3)
the extrapolation  of exposure-response functions below the lowest observed effects levels to
an estimated  background level, and (4) the inability to account for factors which are  known
to affect the  exposure-response relationships (e.g., assigning children the same symptomatic
response rate as has been observed for adults and not  adjusting response rates to reflect the
increase and  attenuation of responses that have been observed in  studies of lung function and
symptoms upon repeated exposures).  Sections V.G. and V.H., Appendices B. and C., and
the associated support documents (Johnson et al., 1996 a,b,c; Whitfield et al., 1996) include
a more complete  discussion of uncertainties inherent in these analyses and 90% credible
intervals are  presented for all risk estimates.
       Summary  Results. The following  information  draws from key analyses discussed in
Sections V.G and V.H  and described in more detail in several technical support documents
(Johnson  et al., 1996a,b,c and Whitfield et al.,  1996).
       Table VI-1 presents a summary of risk estimates  for  1-hr  or 8-hr health endpoints for
outdoor children  upon attainment of alternative  8-hr,  1-expected  exceedance standards and
the current 0.12 ppm, 1-hr standard.  The risk estimates in Table VI-1 are for effects
associated with exposure under moderate  exertion.  These risk estimates are provided only
for outdoor children and only for levels of lung function decrement and symptoms that the

-------
                                 156
TABLE VM.  PERCENT OF OUTDOOR CHILDREN ESTIMATED TO
EXPERIENCE VARIOUS HEALTH EFFECTS 1 OR MORE TIMES PER YEAR
ASSOCIATED WITH 1- OR 8-HOUR OZONE EXPOSURES UPON ATTAINING
ALTERNATIVE STANDARDS8
Alternative Standards
(1 expected exceedance
per
year)
0.07 ppm, 8-hr
0.08 ppm, 8-hr
0.09 ppm, 8-hr
0.12 ppm, 1-hr
Pulmonary
Function
Decrements,
FEV, > 15%
Associated with
8-hr Exposures
3.0
(1.0-6.6)b
5.1
(2.2-9.6)
7.7
(3.3-13.3)
8.3
(8.2-14,2)
Pulmonary
Function
Decrements,
FEV, > 20%
Associated with
8-hr Exposures
0.4
(0.1-1.8)
1.4
(0.5-3.7)
2.7
(1.0-6.1)
3.0
(1.1-6.6)
Moderate or
Severe Pain on
Deep Inspiration
Associated with
1-hr Exposures
0.3
(0.01-1.9)
0.6
(0.05-2.7)
0.9
(0.1-3.5)
1.0
(0.1-3.6)
"Estimates represent aggregate results for 9 urban areas examined.  The total number of
outdoor children residing in the 9 urban areas was 3.1 million.

b90% credible interval

-------
                                           157
staff believes are of most importance in addressing potentially adverse effects.  Lung function
decrements associated with 6- to 8-hour exposures at moderate exertion and symptoms
associated with 1- to 2-hour exposures at either moderate or heavy exertion were found to be
the effects of most concern among the full range of lung function and respiratory symptom
effects evaluated in the risk assessment.  These risk estimates represent an aggregate estimate
for the nine urban areas examined; an aggregate estimate is presented since there is
significant variability in this risk measure across the areas. The uncertainty in these risk
estimates associated with  sample size considerations is characterized by the 90 percentile
credible intervals.
       Since exposure estimates for  outdoor children are higher for most exposure indicators
and alternative standards than exposure estimates for outdoor workers, the risk estimates
summarized here are likely to be among  the highest for the populations being analyzed (i.e.,
general population, outdoor workers, and outdoor children).  The staff has chosen to focus
on the percentage, rather  than the number, of individuals responding in order to reduce
confusion which  might result from the use of numbers of people given the differences in
population size across the nine urban areas.
       Table VI-2 summarizes estimates  of excess hospital admissions for  asthmatics in  the
New York City area associated with just attaining a range of alternative O, standards. These
excess admissions only include those associated with  O3 levels exceeding an estimated
background  O3 level of 0.04 ppm for an  hourly average.
       The staff believes  the following observations on exposure and risk, based  in  part on
the information summarized in Figures VI-1 and  VI-2 and Tables VI-1 and VI-2 are useful in
formulating recommendations about  levels  of alternative  standards that reduce risks  to public
health sufficiently to protect public health with an adequate margin  of safety.
       Exposure Observations.
       (1)     Children who are active outdoors (representing approximately 7% of  the
              population  in the study areas) appear to be the at-risk population group
              examined with the highest percentage and  number of individuals exposed  to O3
              concentrations at and  above  which there is evidence of health effects.

-------
                                                 158
TABLE VI.2  Admissions of New York City Asthmatics — With a Comparison Relative to Meeting the
Current Standard (1 h, 1 expected exceedance, 0.12 ppm)
Air Quality Scenarios
Issue

Excess Admissions'"
(background = 0.04 ppm)
% Change from Current
Standard"
Excess Admissions1'
(background = 0 ppm)
% Change from Current
Standard"
All Admissions'
(thousands)
% Change from Current
Standard11
1H1EX-0.12
ppm'
(Scenario A)
207C
(70, 344)"
0
909
(308, 1,510)
0
14.8
(13.8-15.8)
0
8H1EX-0.08
ppm
(Scenario C)
115
(39, 191)
-44
804
(273, 1,340)
-12
14.7
(13.7-15.7)
-0.6
(-0.2, 1.1)
8H5EX-0.08
ppm
(Scenario F)
120
(41, 199)
-42
797
(270, 1,320)
-12
14.7
(13.7-15.7)
-0.6
(-0.2, -I.I)
As-Is
(Scenario Z)
388
(132, 644)
87
1,070
(361, 1,770)
17
15«
(14-16)'
1.2
(-0.4, 2.2)
       EX stands for expected exceedance.

       Admissions of asthmatics attributable to exposure to ozone.

       Median estimate.
       90% credible interval (about the median).
       Because of the necessary assumption that results across scenarios are highly correlated (i.e., if
       admissions are high for one scenario, they are high for all scenarios), there is very little variation in the
       percentage change from the current standard.
       Admissions of asthmatics for any respiratory-related reason; for scenario i, based on estimates of all
       admissions and excess admissions attributable to ozone levels >0.04 ppm for As-Is scenario, and
       estimate of excess admissions attributable to ozone levels >0.04 ppm for scenario / (e.g., for scenario
       1112:  14,800 « 15,000 - 388 + 207).

       Admissions of New York City asthmatics for any respiratory-related reason during the 1988-90 ozone
       seasons (Thurston, 1995).

       Variation in these results is attributable to the different numbers of admissions of New York City
       asthmatics for any respiratory-related reason during the 1988-90 ozone seasons (Thurston, 1995).

-------
                                 159
   FIGURE VI-1. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
  OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
          EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA
    300
                                                          ASIS

                                                        1H1EX-0.12

                                                        8H1EX-0.09

                                                        8H1EX-0.08
                                                        8H1EX-0.10

                                                       8H5EX-0.08

                                                       8H1EX-0.07
                                                        1H1EX-0.10

                                                        8H5EX-0.09
       0   0.02  0.04  0.06 0.08  0.1  0.12  0.14  0.16 0.18  0.2
                     CONCENTRATION, PPM

   FIGURE VI-2.. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
TOTAL OCCURRENCES FOR OUTDOOR CHILDREN  EXPOSURE UNDER MODERATE
         EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA.
    16,000
  Q 14,000
  ASIS

1H1EX-0.12

8H1EX-0.09

8H1EX-0.08
                                                       8H1EX-0.10
                                                           -e_
                                                       8H5EX-0.08
                                                           -*-
                                                       8H1EX-0.07
                                                       1H1EX-0.10

                                                       8H5EX-0.09
         0  0.02  0.04 0.06  0.08  0.1  0.12  0.14  0.16 0.18  0.2
                      CONCENTRATION, PPM

-------
                                   160
       particularly for 8-hr average exposures at moderate exertion to O3
       concentrations _>_0.08 ppm.
(2)     On both an absolute number and a percentage basis, exposure estimates are
       higher for the 8-hr average effects level of 0.08 ppm at moderate exertion than
       for the 1-hr average effects level of 0.12 ppm at heavy exertion.
(3)     Estimated exposures above these effects curpoints, even on a percentage basis,
       vary significantly across the urban areas examined in this analysis.  However,
       general patterns of exposure can be seen in comparing the current NAAQS and
       alternative standards, particularly in looking at the seven current nonattainment
       areas examined.  For example, for estimates  of the mean percent of outdoor
       children exposed to 8-hr average O3 concentrations _>_  0.08 ppm while at
       moderate exertion, the following patterns are seen:  the range of estimates
       associated with the current 1-hr NAAQS is approximately 3-21%, dropping to
       approximately  <3%  for a 0.10 ppm 1-hr standard.  For alternative 8-hr
       standards (of the same 1-expected-exceedance form as  the current NAAQS),
       the estimated ranges of mean percentages of outdoor children exposed are
       approximately 3-7% for a 0.09 ppm standard, 0-1.3%  for a 0.08 ppm
       standard, and from essentially 0 in  most areas to <0.1 % for a 0.07 ppm
       standard.
(4)     In general, there are relatively  small differences in comparing the distributions
       of 8-hr exposure estimates for outdoor children  associated with 1- and 5-
       expected exceedance forms of any given  alternative standard, although at
       particular cutpoints on the distribution, differences between these two forms
       can  appear to be significant in some areas.
(5)     Based on comparisons of air quality distributions, estimated  exposures are
       generally comparable between 8-hr  standards with 5-expected-exceedance or
       5th highest daily maximum concentration forms. In either case, exposure
       estimates for the worst year of a 3-year compliance period would be higher
       than for the average or typical year, with the magnitude  of the  difference
       varying across areas.   For example, for an 8-hr, 0.08  ppm standard of either

-------
                                           161
              form, about 95% of current nonattainment areas would have 10 or fewer
              exceedances of the 0.08 ppm  level in the worst year, compared to an average
              of less than 5 exceedances in  the typical year. Exposures estimated  for a year
              in which there were 10 exceedances would be roughly comparable to the
              exposures estimated to occur  upon attainment in a typical year of a 0.09 ppm,
              8-hr  standard,  with 1- to 5-expected-exceedance forms.
       In taking these  observations into account, the staff recognizes the uncertainties and
limitations associated with such analyses,  including the considerable, but unquantifiable,
degree of uncertainty associated with a number of important inputs to the exposure model.  A
key uncertainty in model inputs results from the availability of only a limited human activity
database, both with regard to the number of subjects who contributed daily activity diary data
and the short  time period over which each subject recorded their daily activity  patterns.
These limitations may  not adequately account for day-to-day repetition of activities common
to children,  such that the number of people  who experience multiple occurrences of high
exposure levels may be underestimated.  Small sample size also limits the extent to which
ventilation rates associated with various activities may be representative of the  population
group to which they are applied in the model.  In addition, the air quality adjustment
procedure used to simulate air quality distributions associated with attaining alternative
standards, while based on statistical  analyses of empirical data, incorporates significant
uncertainty, especially when  applied to areas requiring very large  reductions in air quality to
attain the alternative standards examined or to areas that are now  in attainment with the
current NAAQS.  A more complete discussion of these uncertainties and limitations is
presented in Section V.G. of this Staff Paper and in the technical  support documents
(Johnson et al., 1996a,b,c).
       Risk Observations.
       (1)    On both an absolute number  and  percentage basis, risk estimates are higher for
              effects associated with 8-hr exposures under moderate exertion than  for effects
              associated with 1-hr exposures under heavy exertion.
       (2)    Reflecting a continuum  of risk, there is a decreasing trend in the median
              estimates of the  population estimated to experience  the lung function and

-------
                                   162
       symptomatic responses as one moves along the range of alternative 8-hr
       average,  1-expected exceedance standards under consideration.  For example,
       based on the aggregate risk estimates summarized in Table VI-1, the median
       percent of outdoor children estimated to experience FEV, decrements greater
       than 15 percent is reduced from about 7.7 percent for a 0.09 ppm, 8-hr
       standard to about 6.8 percent for a 0.08 ppm, 8-hr standard.  Attaining a 0.07
       ppm, 8-hr standard results in a further reduction to about 3.0 percent of
       outdoor children estimated to experience this effect.
(3)    In general,  the differences in risk estimates for outdoor children associated
       with 1- and 5-expected exceedance standards set at the same standard level are
       relatively modest within the continuum of risk.  For example, the risk
       estimates for lung function decrements _>_ 15 percent associated  with a 5-
       expected exceedance standard set at 0.08 ppm fall between the risk estimates
       for the 0.08 and 0.09 ppm, 1-expected exceedance, 8-hr standards. Similarly.
       the risk estimates for a 5-expected exceedance standard set at 0.09 ppm fall
       between the risk estimates for the 0.09 and 0.10 ppm, 1-expected exceedance,
       8-hr standards.
(4)    Multiple occurrences of lung function decrements _>_  15 percent and _>. 20
       percent associated with 8-hr exposures under moderate exertion  are estimated
       to occur for outdoor children upon  attainment of any  of the alternative 1- or 8-
       hr standards analyzed.  The average seasonal numbers of occurrences per
       responder across the urban areas included in  the analysis range from four to
       about nine for lung function decrements _>. 15 percent and from two to about
       five for lung function decrements _>, 20 percent.  Some individuals will
       experience more frequent occurrences  of effects during the O3 season,  whereas
       others will experience fewer occurrences than the average in any given area.
(5)    Based on comparisons of air quality distributions, risk estimates are generally
       comparable between 8-hr standards with 5-expected exceedances or 5th highest
       daily maximum  concentration forms.  As noted in the previous discussion of
       the exposure estimates, for either form the worst year of a 3-year compliance

-------
                                          163
             period would be higher than for the average or typical year.  For example,
             about 95 percent of current nonattainment areas meeting either form of an 8-
             hr, 0.08 ppm standard would have 10 or fewer exceedances in the worst year,
             compared to an average of less than five exceedances in a typical year.  Risk
             estimates for a year in which there were 10 exceedances of 0.08 ppm, 8-hr
             average vary from urban area to urban area but fall between the risk estimates
             for a 5-expected exceedance standard of 0.08 ppm and a 5-expected
             exceedance standard set at 0.09 ppm.
       (6)    Risk estimates for excess  hospital admissions for asthmatics attributable to O3
             exposures in excess of an estimated background level of 0.04 ppm are
             projected to be significantly reduced (44 percent)  under a 0.08 ppm,  8-hr, 1-
             expected exceedance standard compared to the current 1-hr NAAQS  (see Table
             VI-2).
       (7)    The excess hospital admissions risk estimates associated with 1-  and 5-
             expected exceedance standards set at 0.08 ppm are very similar.
       (8)    When viewed from the perspective of respiratory-related admissions for
             asthmatics  due to all causes,  the excess hospital admissions attributable to O,
             exposures in excess of an estimated background concentration of 0.04 ppm
             constitute a relatively small portion of  total admissions. For example,
             comparing the risk estimates associated with the current 1 -hr NAAQS and a
             0.08 ppm,  8-hr, 1-expected exceedance standard results in only a 0.6 percent
             reduction in respiratory hospital admissions for asthmatics due to all  causes.
       The staff believe,  and the CASAC concurred, that the models selected to estimate
exposure and risk are appropriate and that the methods used  to conduct the health risk
assessment represent the state of the art.  Nevertheless, there are many sources of
uncertainties inherent in such analyses.  Some of the  most important caveats and limitations
concerning the health risk assessment for lung function and respiratory symptom endpoints
include:  (1) the uncertainties and limitations associated with the exposure analyses discussed
above, (2) the extrapolation of exposure-response functions below the lowest observed effects
levels to an estimated background level of 0.04 ppm, and (3) the inability  to account for

-------
                                           164
some factors which are known to affect the exposure-response relationships (e.g., assigning
children the same symptomatic response rates as observed for adults and not adjusting
response rates to reflect the increase and attenuation of responses that have been  observed in
studies of lung function and symptoms upon repeated exposures).
       Similarly, there are uncertainties and limitations associated with the hospital admission
risk assessment.  These include: (1) the inability at this time to quantitatively extrapolate the
risk estimates for the New York City area to other urban areas,  (2) uncertainty associated
with the underlying epidemiological study that served as  the basis for developing the
concentration-response relationship used in the analysis, and (3)  uncertainties associated with
the air quality adjustment procedure used  to simulate attainment  of alternative standards for
the New  York City area.  A more complete discussion of these  uncertainties and limitations
is presented in the technical support document (Whitfield et al.,  1996).
E.     Summary of Staff Recommendations
       Drawing on the staff conclusions and observations on margin of safety considerations
presented above, together with consideration of the information  in the CD  and section V of
this Staff Paper, the staff offers the following recommendations  on the primary 0, standard.
       1.      Pollutant indicator
       Staff recommends that O3  remain as the indicator for controlling ambient
concentrations of photochemical oxidants.  This  recommendation-is based on the large base
of health effects information attributing effects to O3 exposure and the  lack of convincing
evidence demonstrating effects from exposure to ambient levels  of photochemical oxidants
other than O3.
       2.      Averaging times
       Staff recommends that further  consideration be given at this time only to short-term
averaging times  associated with acute effects.  No further consideration of a  long-term
standard is recommended by staff in this O3 NAAQS review cycle.  The staff offers the
following additional specific recommendations:
       •      Staff recommends that principle consideration be given  to an  8-hr averaging
              time be considered for a new O3 primary standard.

-------
                                           165
       This recommendation is based on 1) the health effects studies reporting a number of
health effects associated with 6- to 8-hr exposures at and below the level of the current  1-hr
standard; 2) the staffs judgments that the 6- to 8-hr effects at moderate exertion are of
greater public health concern at lower O3 levels than similar 1-hr effects at heavy exertion, 3)
the staffs judgments that  these effects are within the range that the Administrator might
consider to be adverse; and 4) the exposure assessments and the quantitative risk assessments
for some of the effects showing that reductions in the risks associated with these 6- to 8-hr
effects can be achieved by attaining alternative 8-hr standards.
       Although staff recommends that principle  consideration be given to a standard with an
8-hr averaging time, staff recognizes  that a standard with a 1-hr averaging time could be set
at a level that would provide roughly equivalent health protection to that provided by an 8-hr
standard.
       3.      Form of the Standard
       Based on the discussion in section V.I and the conclusions presented above, staff
offers the following recommendations with regard to form  of the standard  and attainment test
issues:
       •      Staff recommends consideration be given to  the current expected exceedance
              form, ranging from  1-  to 5-expected exceedances, averaged  over 3 years, as
              well  as to a concentration-based form, ranging from the second  to the fifth
              highest 8-hr daily maximum concentration, averaged over 3  years.
       •      Staff also recommends consideration of defining the standard in terms of a
              range of air quality values.
       Risk analyses discussed  above and in  Section V.H indicate that for  most of the health
endpoints analyzed  there is little difference in health risk, at a given level of the standard.
within the ranges of 1- to 5-expected-exceedances and  the second to the fifth highest 8-hr
daily maximum concentration forms of the O3 primary standard.  On average,  the 1-expected
exceedance form provides  the greatest exposure and health risk  protection  but onh slight!)
greater than that provided by the second  to the fifth highest 8-hr daily maximum
concentration form.  There is also  not much  difference between the fifth 8-hr daily maximum
concentration form  and a 5-expected exceedance form, which on average are  roughh
equivalent for any given level of the primary standard  selected.  Based  on  these analyses.

-------
                                           166
therefore, it is the level of the standard which mainly determines the degree of public health
protection afforded by an 8-hr primary NAAQS for O3 within those alternatives considered
above.
       4.     Level of the Standard
       In making recommendations, staff notes that the decision ultimately made by the
Administrator regarding  level of the primary O3 NAAQS will be based on a policy judgment
as to the degree of risk reduction that is necessary to protect public health with an adequate
margin of safety.  The following recommendations on level address the staff recommendation
that consideration be given to a  standard with an 8-hr averaging time and a form of 1- to 5-
expected-exceedances or a second to the fifth highest 8-hr daily maximum concentration
form.
       The following staff recommendations suggest a range of levels based on considering:
1) protection against health effects directly associated with both 1-  to 3-hr and 6- to 8-hr
exposures (e.g., lung function decrements, respiratory symptoms, nonspecific bronchial
responsiveness, acute pulmonary inflammation, and  increased susceptibility to  infection), as
well as against the effect of increased  hospital admissions; 2) quantitative risk  assessments
which  provide insight as to the degree of protection afforded by alternative 8-hr standards for
some of these effects, and 3) protection against the effects of repeated inflammatory
responses that could lead over time to chronic respiratory illness.
       •      Staff recommends that the upper end of the range of consideration for an
              8-hour primary O3 NAAQS be 0.09 ppm.
       As discussed in the general conclusions and margin of safety considerations presented
above, the primary range of lowest effects levels relevant to all the effects of concern
identified above is 0.08 to 0.10  pm.  As previously discussed,  the staff believes that this
range of effects levels does not necessarily reflect a  threshold below which effects do not
occur,  but rather may reflect levels at  which studies finding statistically significant effects of
concern have been conducted.  Thus, the staff believes that in assessing the adequacy of
health  protection  afforded by alternative standards levels it is also important to consider:  (1)
the severity  and variability of these effects, (2) the extent to which sensitive or at-risk
populations are likely to  experience exposures associated with these effects, and (3)

-------
                                          167
quantitative estimates, when available, of the risk to sensitive and at-risk populations in terms
of the estimated numbers or percentages of the populations groups that are likely to
experience adverse levels of these effects.
       Based on consideration of the above factors, the staff recommends that 0.09 pm is the
highest level of an 8-hour standard that would reduce estimated exposures of the at-risk
populations sufficiently to provide some margin of safety against pulmonary inflammation
and increased susceptibility to pulmonary infection.  Further, the staff recommends that 0.09
pm is the highest 8-hr level that would reduce the estimated risk to the at-risk populations of
experiencing increased hospital admissions and emergency room visits, as well as
experiencing adverse levels of lung function decrements, respiratory symptoms, and
nonspecific bronchial responsiveness  sufficiently to provide some margin of safety  against
these effects.
       These staff recommendations also reflect consideration of previous advice from the
CASAC  during the last review of the O3 NAAQS.  In its previous review, the CASAC
(McClellan,  1989) concluded that the existing 1-hr primary standard provided "little, if any,
margin of safety," and that the upper end of the range of consideration for the 1-hr primary
standard  should be 0.12 ppm.  Several members of the CASAC Ozone NAAQS Review
Panel felt that consideration should be given to a 1-hr standard  level of 0.10 ppm in  order to
provide for an adequate margin of safety and to offer some health  protection against  8-hr
exposures of concern.  This advice provides support for considering 0.09 ppm rather than
0.10 ppm as the upper end of the range for an 8-hr standard, in that exposures associated
with the  8-hr effects and risks for respiratory symptoms are greater when a 0.10 ppm 8-hr
standard  is just attained than when a  0.12 ppm 1-hr standard is just attained.
       •     Staff recommends that the lower end of the range of consideration for the
              primary 8-hr O3 NAAQS be 0.07 ppm.
       In conducting exposure and risk analyses of the 0.06 ppm level, staff concluded that
the risk of health effects of concern occurring was extremely low, approaching zero  in most
cases.  Considering both the nature of the health effects involved and the very small
percentage of the population that would be affected, staff believes  that a  primary 8-hr
standard  with a level of 0.07 ppm could be judged  to provide public health  protection with an

-------
                                           168
adequate margin of safety for these effects of concern.  A standard set at this level would be
more precautionary than a standard set at the upper end of the range,  in that it would provide
increased protection from long-term exposures that may be associated  with potentially more
serious but more uncertain chronic effects.

-------

-------
                                           169
 VII.   SCIENTIFIC AND TECHNICAL BASIS FOR SECONDARY NAAQS
 A.     Introduction
        This section presents critical information for the review of the secondary NAAQS for
 O3.  Welfare effects addressed by a secondary NAAQS include, but are not limited to,
 effects on soils, water, crops, vegetation, man-made materials, animals, wildlife, weather,
 visibility and climate, damage to and deterioration of property, and hazards to transportation,
 as well as effects on economic values and on personal comfort and well-being.  Of these
 welfare effects categories, the effects of O3 on crops, vegetation and  ecosystems are of most
 concern at concentrations typically occurring in the U.S.  As stated in the previous CD and
 SP, "of the phytotoxic compounds commonly found in the ambient air, O3  is the most
 prevalent, impairing crop production and injuring native vegetation and ecosystems more
- than any  other air pollutant" (U.S. EPA, 1989).  By affecting crops and native  vegetation, O3
 also directly and indirectly affects natural ecosystem components such as soils,  water,
 animals,  and wildlife and ultimately the ecosystem itself.  Some of these impacts have direct.
 quantifiable  economic value, while others are currently not quantifiable.  Thus, the staff
 infers that increasing protection for crops and vegetation would also improve the protection
 afforded  to these other related public welfare categories.
        Ozone damages certain manmade materials (e.g., elastomers,  textile fibers, dyes,
 paints, and pigments).  The amount of damage to actual ir.-use materials  and the economic
 consequences of that damage are poorly  characterized, however, and the scientific  literature
 contains very little new information to adequately quantify estimates of materials damage
 from photochemical oxidant exposure (CD,  1996).  Effects on personal comfort and well-
 being have already been addressed under the section of the Staff Paper on human health.
 Therefore, these effects categories  will not be reviewed in this portion of the Staff Paper, and
..the reader is referred to the last Staff Paper (U.S. EPA, 1989) for a discussion  of these
 effects categories.
        The remainder of this chapter focuses on O3 effects on crops, native vegetation and
 ecosystems,  drawing upon the most relevant information contained in  the CD.   This
 information includes:  (1) plant response and mode of action of O, on vegetation; (2)
 environmental factors affecting plant response; (3) relevant research on O3  effects on crops

-------
                                           170
and native vegetation; (4) considerations and criteria for selecting an appropriate measure of
O3 exposure that can meaningfully relate O3 air quality to plant response under varying O3
regimes; and (5) other policy relevant considerations that would assist  the Administrator in
judging the need for a new secondary standard, including analyses of air quality patterns, the
relationships between primary and secondary standard options, national exposures, risks, and
economic values.
B.     Plant Response/Mode of Action
       The first observation of O3 injury to vegetation in the field (O3  stipple on grape
leaves) was reported  in the 1950's (Richards et al., 1958).  Since that  time, a substantial
amount of research has been done on the effects of C^ on plants that has increased scientific
understanding of the  mechanisms of action, factors that modify plant response to O3, and
relative sensitivities of various  species and cultivars to O3 concentrations found currently in
the U.S.
       1.     Ozone Uptake
       The primary  site of O3 uptake into the plant is the leaf. The leaf is the site of gas
exchange for the plant. To cause injury, O3 must diffuse in the gas-phase from the
atmosphere surrounding the leaves through the stomata into the air spaces, dissolve  in the
water coating the cell walls and, then, its reaction products, diffuse through or react with the
membrane of the cell.  Once inside the  cell,  they can react with cellular components and
affect metabolic processes (CD, 1996).
       The movement of O3, as well as other gases, into and  out of leaves is controlled
primarily through the stomata.   The aperture of the stomata are controlled by guard  cells,
which are affected by a variety  of internal species-specific factors and  external environmental
factors such as light, humidity,  CO2 concentration, soil fertility and nutrient availability,
water status of the plant and, in some cases, the presence of air pollutants, including O, (See
U.S. EPA, 1986: Zeiger  et al., 1987; Schulze and Hall, 1985; Beadle et al., 1985a  and b;
Kearns and Assmann, 1993).  To investigate variability in diurnal gas  exchange, Tenhunen et
al. (1980) evaluated diurnal photosynthesis for apricot measured between July and September
 1976. While there is a general  pattern  of increase in the morning and decrease in the
evening, the path of photosynthesis (and conductance)  are quite different among the  days.

-------
                                           171
 For example, mid-day stomatal closure is frequently observed under conditions of high
 -temperature and low water availability (Tenhunen et al., 1980; Weber and Gates,  1990).
        However, uptake measured for one leaf may not be the same for any of the other
 leaves on the plant.  Leaves exist in a complex three dimensional environment called a
 ;canopy. Each leaf has an unique orientation within that canopy and receives a different
.-exposure to the ambient air.  In addition, a plant may be located within a stand of other
 plants which further  modifies ambient air exchange with individual leaves. This makes it
 difficult to  extrapolate from the uptake measured in a single leaf to that of an entire plant or
 canopy.
        Within the canopy, due to its high reactivity, O3 may be scavenged or absorbed by
 other environmental components and surfaces before it ever reaches the leaf itself.   Though
.- models of stomatal conductance for canopies and stands have been developed to account  for
 some of this complexity, these models require the use of several assumptions that at this
 time, have  not been adequately tested or validated by direct measurements.  One particular
.area that needs further study is the relative importance of cumulative  uptake versus the rate
 of uptake (CD, 1996).
        Because O3 flux incorporates environmental factors and physiological processes,
 several authors (Grandjean, et al., 1992a,b; Fuhrer et al., 1992) have suggested that it be the
 relevant measure for use in relating exposure  to plant response (CD,  1996).  However,
 measurement of flux for an entire plant or canopy is very complex.  Therefore,  most
 research has been done to  try to develop appropriate surrogate  measures for uptake
 (discussed below in section VILE).
        2.     Extracellular Effects
        Once within the leaf, O3 quickly dissolves in the aqueous layer on  the cells lining the
 ^air spaces.  When O3 passes into the liquid phase, it undergoes reactions that yield a variety
 of free radicals (e.g., superoxide and hydroxyl radicals).
        Many consider membranes to be the primary site of action  of O3 (Heath, 1988;
 Tingey and Taylor,  1982).  Whether the plasma membrane or some organelle membrane is
 the primary site of O3 action is open to speculation (Tingey and Taylor. 1982).   The

-------
                                           172
alteration in plasma membrane function, however, is clearly an early step in a series of O3-
induced events that eventually leads to leaf injury.
       3.     Intracellular Effects
       Once O3 reaction products diffuse through the cell wall and interact with or diffuse
through the cell membrane they may affect cellular or organellar processes. Altered cell
structure and function may result in changes in membrane permeability, carbon dioxide
fixation, and many secondary metabolic processes (Tingey and Taylor, 1982).
        In addition to the disruption and alteration of a number of membrane-dependent
functions associated with photosynthesis, O3 can interfere with the biochemical aspects of the
photosynthetic process itself.  For example, the enzyme that catalyzes CO2 fixation during
photosynthesis, RuBP, can be inhibited. Nakamura and Saka (1978) reported  reduced
activity of RuBP carboxylase in rice after exposure to 0.12 ppm O3 for only 2 hours.  Pell
and Pearson (1983) observed 36, 68, and 80 percent decreases, respectively, in the
concentration of RuBP carboxylase in the foliage of three alfalfa cultivars that had been
exposed to an O3 concentration of 0.25  ppm for 2 hours.  These observations were made  48
hours after exposure on leaves that did  not exhibit macroscopic injury symptoms (US EPA,
1986), showing that there is no clear connection between foliar injury symptoms and
biochemical changes within the leaf.
       The  potential for O3, directly or indirectly, to oxidize several other classes of
biochemicals including nucleotides, proteins, some amino acids and various lipids has been
demonstrated in several in vitro studies. New approaches are needed to assess the full range
of in vivo biochemical changes caused by O3 (US EPA,  1986).
       Finally, changes in the  in vivo concentrations of various  growth  regulators or
hormones such as ethylene have been shown in a few studies to  be correlated with O3
sensitivity.  Because ethylene production also occurs with the ripening of fruit, during
periods of stress, and during leaf senescence (Abeles et al., 1992), increased levels of
ethylene in  the leaves could play a role in the early senescence of foliage.  However, because
the relationship is complex, the use of ethylene production as an index of sensitivity is still
problematic (Pell, 1988).

-------
                                            173
        4.      Resistance and Compensation Mechanisms
        Plant stress from O3 occurs when the atmospheric concentrations exceed the limits of
 plant tolerance.  In the leaves, O3 injury will not occur if the rate of uptake is sufficiently
 small so that the plant is able to detoxify O3 or its metabolites.  Leaves may physically
 exclude O3 from sensitive tissues.  A few studies have documented a direct stomatal closure
 or restriction in response to the presence of O3.  In studies  at O3 concentrations > 0.30 ppm
 stomatal response was rapid (Moldau et al., 1990). In other studies, reduction in
 conductance in response to O3 required  hours to days  of exposure (Dann and  Pell, 1989;
 Weber etal.,  1993).
        Additionally, plants may have a  biochemical defense in the production of antioxidants.
 Two general kinds of antioxidants have  been  reported in plants: I) reductants and 2)
 enzymes.  In either case excess oxidizing power is dissipated  in a controlled manner,
 effectively protecting the tissue against damage.  For detoxification to occur,  oxidant and
 antioxidant must occur proximately  and  the rate  of production of antioxidant must at least
 balance the rate of oxidant entry into the system.  Because  potential rates of detoxification
 for given tissues and the sites in which they occur are not yet known, the effectiveness of
 these systems  in protecting plant tissue from damage to O3  cannot be determined (CD.  1996).
        Once O5 injury has  occurred in leaf tissue,  some plants are able to  repair or
 compensate for the O3 impacts (Tingey and Taylor, 1982).  In general, plants  have  a variety
 of compensatory mechanisms for low levels of environmental stress,  of which O3 is one.
 Since these mechanisms are genetically  determined, not all  plants have the same complement
 of defensive tools or degree of O3 tolerance, nor are all stages in a plant's development
 equally sensitive to O3.
        A wide range of compensatory responses have been  identified, including reallocation
-of resources, changes in root/shoot  ratio, production of new tissue, and/or biochemical shifts.
 such as increased photosynthetic capacity in new foliage,  and changes in respiration rates
 indicating possible repair or replacement of damaged membranes or enzymes.  For  example,
 replacement of injured leaf tissue has been reported for some species after exposure to O?
 (Held et al., 1991; Temple et al.. -1993) and  increased photosynthetic capacity of new
 needles in O3 treatments compared to controls.  Additionally, ponderosa pine  has been  shown

-------
                                           174
to recover from decreased photosynthetic rates in O3-treated needles.  In one case injured
needles were able to regain the photosynthetic rate of controls after 40-50 days (Weber et al.,
1993).
       While these systems potentially provide protection against O3 alteration to tissue
physiology, it is not yet known to what degree or how  the use of plant resources for repair
processes affects the overall carbohydrate budget or subsequent plant response to O3 or other
stresses (CD, 1996).
       5.     Physiological Effects
       The effects of O3 injury at the cellular level in the ways described above, when they
have accumulated sufficiently, will be propagated to the level of the whole leaf or plant.
These larger scale effects can include: (1) visible foliar injury; (2) premature needle/leaf
senescence; (3) reduced photosynthesis;  (4) reduced carbohydrate production and allocation;
(5) reduced plant  vigor; and (6) reduced growth or reproduction or both (Miller et al., 1982;
McLaughlin et al., 1982; Skelly et al., 1984; U.S. EPA, 1986).
       Visible Foliar Injury.  Although O3 can significantly alter cellular and photosynthetic
processes without resulting in changes in leaf appearance,  cellular injury can and often does
become visible.  For coniferous trees, two visibly recognizable syndromes have been
associated with oxidant injury.  One, emergent tipburn, is noted most often on eastern white
pine.  This injury is characterized as a tip dieback of newly elongating needles.  Silvery or
chlorotic (absence or deficiency of green pigment) flecks,  chlorotic mottling, and tip necrosis
(tissue death) of needles may also be present.  The other O3 injury response, chlorotic
decline, results in the loss of all but the current season's needles and was first noted on
ponderosa pine.  Yellow  mottling and a reduction in the number and size of the remaining
needles may also  occur.  In non-coniferous species, acute O3 injury usually results in cell
destruction (bifacial necrosis) due to  the disruption of normal cell  structure and  processes and
the subsequent loss of water and salts from the cell (U.S.  EPA, 1978).
       Acute injury usually appears within 24 hours after exposure to O3 and, depending on
the species, can occur under a range of exposures and  durations.  For example, a summary
of limiting values for  visible injury showed effects occurring across a range of exposures

-------
                                            175
 from 0.04 ppm for a period of 4 h to 0.41 ppm  for 0.5 h for crops,  and 0.06 ppm for 4 h to
 0.51 ppm for  1 h for trees and shrubs (U.S. EPA, 1986).
        Chronic injury may be mild or severe, and is associated with long-term or multiple
 exposures to elevated O3 levels.  Under chronic  exposures, disruption of normal cellular
 activity occurs, leading to chlorotic mottling, bleached or unpigmented lesions (flecks) or
 pigmentation (stippling). Though cell death may eventually result, depending on dose and
 environmental conditions, membrane  permeability may be restored and cell recovery occur
 (U.S. EPA, 1978). Chronic O3 injury patterns may be confused with symptoms resulting
 from normal senescence, biotic pathogens, including insects, nutritional disorders, or other
 environmental stresses.  These patterns may appear as premature leaf senescence.
        The significance of O3 injury  at the leaf level depends on how much of the total leaf
 area of the plant has been affected, as well as the plant's age and size, developmental stage,
 and degree of functional redundancy  among the existing leaf area. As a result, it is  not
 presently possible to determine with consistency across species and environments what degree
- of injury at the leaf level has significance to the vigor of the whole plant.
        Premature Needle/Leaf Senescence.  Ozone has been shown to affect  needle or leaf
 retention in loblolly pine (Stow et al., 1992; Kress et al.,  1992), slash pine (Byres et al.,
 1992), aspen (Keller, 1988; Matyssek et  al.,  1993a,b) and apple (Wiltshire et al., 1993), as
 well as other species.  Leaf replacement  may be part  of the normal growth strategy employed
 by the plant to maintain photosynthetic production.  However, leaves that have to be replaced
 more frequently drain energy from plant  reserves and those that  are irreplaceably lost
 represent a net loss of photosynthetic  capacity that can have significant effects on plant vigor.
        Impaired Photosynthesis: Changed Carbohydrate Production and Allocation.
 Photosynthesis, the process by which  plants produce energy-rich compounds  (e.g., )  for use
-in growth, maintenance, reproduction or  storage, can  be impaired by Ov  This impairment
 may result from the direct impact of O3 on chloroplast function or from O?-induced  stomatal
 closure resulting in reduced CO2 uptake,  or both.  As discussed above, this can occur
 without any  macroscopic visible injury.
        If total plant photosynthesis is  sufficiently reduced, the plant will respond by
 reallocating the remaining carbohydrate at the level of the whole organism.  Since the roots
 are often the largest source of stored  carbohydrate, they (Andersen et al., 1991;  Andersen

-------
                                           176
and Rygiewicz, 1991) and associated mycorrhizal fungi (Adams and O'Neil, 1991; Edwards
and Kelly, 1992;  McQuattie and Schier, 1992; Meier et al., 1990; Taylor and  Davies,  1990)
become especially susceptible to reduced carbohydrate  availability, and quite frequently show
the greatest decline in growth.  Cooley and Manning (1987) reviewed the literature on
carbohydrate partitioning and noted that "storage organs ... are most affected by Oj-induced
partitioning changes when O3 concentrations are in the  range commonly observed in polluted
ambient air."
       When less carbohydrates are present in roots, less energy will  be available for root-
related functions such as the acquisition of water and nutrients. Mycorrhizal fungi, which
invade the roots of terrestrial plants, are of great importance for vegetational growth (U.S.
EPA, 1978).  These fungi increase the solubility of minerals, improve the uptake of nutrients
for host plants, protect host roots against pathogens, produce plant growth hormones, and
may transport carbohydrate from one plant to another (CD,  1996). Ozone has the capability
of disrupting  the association between the mycorrhizal fungi and host plants by  inhibiting
photosynthesis and the amount of sugars available for transfer to the roots.  In one example,
Berry (1961)  examined the roots of eastern white pine  injured by O3 and observed that
healthy trees  had  almost twice the percentage of living  feeder roots as trees with O3 injury.
Primary roots of affected trees have even been shown to die after repeated needle injury
(U.S. EPA, 1978).
       Unlike root systems,  effects on leaf and needle  carbohydrate content under conditions
of O3 stress have  ranged  from a reduction  (Barnes et al., 1990; Miller et al., 1989), to  no
effect (Alscher  et al.,  1989), to an increase (Luethy-Krause and Landolt, 1990).  Friend and
Tomlinson (1992) found that O3 exposure increased retention of C14-labelled photosynthate in
needles of loblolly pine, and modified its distribution among starch, lipids, and organic acids
(Edwards  et al., 1992; Friend et al., 1992).  These  responses have been measured in
ponderosa pine seedlings exposed to O3  concentrations  of 0.10 ppm for 6 hr/day for 20
weeks (Tingey et al.,  1976).
       Reduced Plant Vigor.  There is no  evidence  to suggest that O3  levels over most  of the
U.S. are high enough  to kill vegetation directly.  However, at current ambient levels that
occur during  O3 episodes, the ability of  many sensitive  species or  genotypes within species to

-------
                                           177
 adapt to other environmental stresses, including competition for available resources, can be
 sufficiently compromised such that the end results prove fatal for some plants.  For example,
 McLaughlin et al. (1982) observed that the reduced availability of carbohydrates  associated
 with O3 exposure resulted in enhanced susceptibility of trees to root disease and influenced
..the success of pest infestations (Hain, 1987;  Lechowiez, 1987). Fincher (1992) and Davison
 et al. (1988) found that O3 also can decrease the ability of trees to withstand winter injury
 caused by exposure to freezing temperatures.
        Reduced Growth and/or Reproduction.  As  discussed above, O3 exposure  can reduce
 carbohydrate production or storage in plants. In annual species this affects plant growth,
 flowering, and seed development.  Unlike annuals, deciduous perennials that must  survive
 more than one year and develop new leaves  each year after a penod of dormancy depend on
-long-term storage of carbohydrates to get them through unfavorable growth periods.  Thus,
 Avhile no O3 effects on growth may be observed above ground during a year of elevated O3
 levels, the following year  may show a decrease in  root growth or new biomass production.
        Coniferous species also must maintain foliage from one year to the next, and may, in
 some spruce species, retain as many as 10 years of needles at any point in time,  and continue
 to produce carbohydrates even during winter months.  Therefore, injury to or early loss of
 needles can result in a greater shift in remaining carbohydrates to repair and replacement of
 needles, thus potentially reducing biomass production.   When storage carbohydrates are
 limited, older needles may become the source of photosynthate for  new needle growth in the
 spring and storage sinks in the fall (McLaughlin et al., 1982). Thus, O3 impacts may be felt
 over multiple years.  These "carry-over"  effects have been documented in  the growth of tree
 seedlings (Hogsett et al., 1989; Sasek et al.,  1991; Temple et al., 1993) and in regrowth of
 roots (Andersen et al.,  1991).  Controlled exposures, however, have been  for the most pan
-only 2-3 years in duration so that data on the cumulative effects of multiple years of O-,
 exposure are extremely limited (CD,  1996).
 C.     Environmental Factors Affecting Plant Response
        Plant response to O3 exposure is a function  of the plant's ongoing integration of
 biological, physical and chemical factors both within and external to the plant.  The corollarx
 is also true: O3 exposure can modify the plant's subsequent integrated response to other

-------
                                           178
environmental factors. Thus, there are inherent multiple sources of uncertainty which must
be recognized in relating plant response under one set of growing conditions to responses
under other conditions.  Additionally, the numerous methodologies used in vegetation
research have been designed to address particular sets of uncertainties and answer particular
type of questions.  When discussing the results of a study, the uncertainties introduced by the
type of methodology used should also be recognized.
       1.     Biological Factors
       Genetics.  The genetic code of each plant contains a gene or genes that govern its
response to O3.  Even within the genomes of a particular plant species, there is wide
variability  in O3 sensitivity. This has been amply demonstrated through inter- and
intraspecific comparisons which have shown that it is not uncommon to have a species with
genotypes  that vary by as much as 50% in  the same  study. These findings have significant
implications for predicting plant response to O3. First, an exposure response relationship
generated for a single genotype or small  group of genotypes may not adequately represent the
response of the species as a whole (Temple,  1990).  Further, a study that uses only the most
sensitive genotypes within a species might overestimate the injury being done to the species
as a whole.
       Secondly, this variability in response means that O3 can impose a  selective force
favoring tolerant genotypes over sensitive ones.  For example, sensitive species are unable to
compete for the required water and nutrients, or may not be able to reproduce (Roose et al..
1982; Treshow, 1980).
       Numerous  studies show that it is likely that sensitive genotypes are being lost from
natural  ecosystems at current ambient O3 exposures in some parts of the U.S.  Berrang  et al.
(1986,  1989,  1991) have presented evidence  for population change in trembling aspen
(Populus tremuloides L.) by showing that aspen clones from polluted areas were visibly
injured to  a lesser degree than those taken  from unpolluted areas.  Additionally. Karnosky
(1981,  1989)  studied the O3 symptom expression and survival of over 1,500 eastern white
pine trees  growing in southern  Wisconsin and found that O3-sensitive genotypes had a ten-
times-higher rate of mortality than  did the  O,-~esistam genotypes over a 15-year study.
During the 1970's, significant numbers of sensitive white pine were lost  from the

-------
                                           179
Cumberland Plateau in Tennessee.  Heagle et al. (1991) found a population change in O3
sensitivity in white clover (Trifolium repens L) after two years of O3 exposures in open  top
chambers, and Gillespie and Winner (1989) found O3 to be a strong and rapid selective force
with radish cultivar "Cherry Belle."
       Limited evidence also suggests that O3 may affect the reproductive success of O3
sensitive species.  Studies on the effects of O3 on pollen germination and tube elongation of
some Scott's pine, eastern white pine, corn, petunia, and tomato generally found a negative
impact of O3 on this critical element of reproduction (CD,  1996). Reduced flowering as the
result of prolonged fumigation  with O3 has been shown in Bladder campion, geranium, and
carnation.  This effect reduces  the fitness of the affected genotypes, and may result in the
eventual loss of these genetic units from O3-stressed environments.
       Plant breeders working  in locations with high O3 concentrations have developed
varieties more tolerant to O3 than those developed under low O3 conditions  for such  species
as alfalfa, potato, cotton, and sugar beet (CD, 1996).  Likewise, nursery owners, Christmas
tree growers, and seed orchard managers have all routinely discarded pollution-sensitive
chlorotic dwarf and tipburned white pine trees because of their slow growth in areas with
high O, (Umbach and Davis, 1984), and thus, contributed to the selection of more tolerant
commercial forests.
       In natural ecosystems, the loss  of genetic diversity is considered an adverse impact by
Federal Land Managers of Federal  Class I areas who have been given the charge to  preserve
for future generations the genetic resources within their borders.  In addition, such loss may
have economic implications for commercially important species if the remaining populations
are made up of O3 resistant plants that are less adaptable to subsequent change, or if the O,
tolerant trait is linked to other traits such as slower growth and productivity.
       Pollutant/Plant/Pest/Pathogen Interactions.  Significant research has been done  on this
topic since the  1986 Criteria Document.  Several recent studies on the effects of O3 on the
feeding  preference of herbivorous insects, and on their growth, fecundity, and survival have
reported that O3-induced changes in the host plants frequently result in increased  feeding
preference of a range of insect  species.  For example,  Chappelka et al. (1988)  found that O,
enhanced the feeding preference and larval growth of the Mexican bean beetle on soybean.

-------
                                           180
leading to increased defoliation.  Similarly, stimulatory responses were observed with
pinworm on tomato, with aphid and weevil on European beech, with the monarch butterfly
on milkweed, and with infestation by the willow leaf beetle on cotton wood (CD, 1996).
These  data do not provide, however, a consistent relationship between different levels or
patterns of O3 exposure and insect growth  response.  Additionally, the reports of Cyinsect-
plant interactions only represent a small  fraction of the interactions that exist, making
generalization to other combinations uncertain.
       With respect to the interaction between O3 and plant disease, new information since
the 1986 Criteria Document changes the earlier conclusion from  it is "impossible to
generalize and predict effects in particular  situations" (U.S. EPA, 1986) to the conclusion in
the CD that "pathogens which can benefit  from injured host cells or from disordered
transport mechanisms (facultive) are enhanced by pollution insult to their hosts, whereas
those that require  a healthy mature host  for successful invasion and development (obligate)
are depressed by pollutant stress to their host." In a few studies, infection of the plant with
obligate bacteria or pathogens or nematodes tended to reduce  the impact of O3.  The majority
of these studies  have been conducted in laboratories or greenhouses, which raises the
question of relevance under field conditions.  Much more study is needed, and with a wider
range  of species, to quantify the magnitude of the interactive  effects to different levels of O3
exposure.
        Pollutant/Plant/Plant Interactions.  While vegetation  literature is replete with
experimental studies associating O3 exposure with  observed effects on  plants, any attempt to
extrapolate these results to field conditions must recognize that other factors  such as
competition with other plant species for  limited resources such as light, water, nuirients and
space  can effect the degree of injury observed.  Several studies have reported that
environmental and site conditions often explain the patterns of O3 injury for  a given species
more than  the actual O,  concentration levels.  For example, it has been reported that canopy
trees can be more affected than understory, and that ponderosa pines growing at the top of
ridges or on  dry sites experienced greater  foliar injury than those grown elsewhere (Second
Progress Report of FOREST, 1994). Though very few studies have been conducted to
evaluate the effects of O3 on competition between  species, it is clear that the implication  of

-------
                                            181
 the known effects of O3 described in the previous section tend to impair a plant's ability to
 compete with other species.  For example, a shift in allocation of carbohydrates away from
 roots to leaves and  shoots results in a compromised root system which limits the plant's
 ability  to explore the soil for water and nutrients, and injury and/or loss of leaves would
slimit the plant's ability to take advantage of available light.  A prime example of O3-induced
.shifts in species dominance is that observed in studies of the San Bernardino Forests, as
 discussed  below in  Section VII.D.
        Competition may be either between plants of different (inter-)  or the same (intra-)
 species.  The planting densities and row spacings adopted for agricultural  crops represent
 compromises between maximizing the number of plants per unit area and  the adverse effects
 of intra-species competition.  Though weeds are typical inter-species competitors,  no studies
- appear to  have been conducted on the effects of O3  pollution on such  competition.  Inter-
- species competition also occurs in mixed plantings such as  grass-clover forage and pasture
 plantings, and is an important feature of natural  ecosystems.  A  consistent finding with grass-
-clover  mixtures has been a significant shift in the mixture biomass in  favor of the grass
 species (CD, 1996).
        Recently, the development of field exposure systems have permitted some studies of
 crop species to  be conducted  in the field.  Because the crops were planted at normal planting
 densities,  inter-species  competition was incorporated as an environmental factor.  On the
 other hand, most forest tree studies have tended  to be "artificial" in their use of individual
 seedlings or saplings or spaced  trees, even when exposed to open-air  systems (McLeod et al.,
 1992).   The significance of the effects of competitive interactions on the O3 response of the
 competing species is largely unknown, and leads to considerable uncertainty when
.extrapolating from effects on individual species to managed  and  natural ecosystems.
        2.     Physical Factors
        The physical components of a plant's aerial environment  are light,  temperature,
 humidity,  wind velocity and surface wetness, while the physical, edaphic components
 affecting the plant roots are temperature and soil moisture and salinity.  Since the effects of
 the physical climatic factors on  plant growth are major determinants of the geographic
 distribution of the earth's natural  vegetation and  of the distribution of agricultural lands and

-------
                                           182
the suitability of the crops grown on them, it is important to note when experimental
conditions vary from those which are normally found in the field.
       Light.  In most species, light plays a major role in the opening and closing of
stomata, thus dictating to some extent when O3 can be taken up by foliage from the ambient
air.  Because many studies are done under reduced light intensities, it is important to note the
general conclusion reported previously (U.S. EPA, 1986) that susceptibility to foliar injury is
increased by low light intensities and short photoperiods.  Reduced light intensities have been
measured in open-top chambers  in the field, resulting from the build-up of dust on the walls.
However, Heagle and Letchworth  (1982) could detect no significant effects on soybean
growth and yield in  a comparison of plants grown in  unshaded open-top chambers and
chambers to which  shading cloth was applied.
       Temperature. An important O3-temperature interaction affecting trees and other
woody perennials is winter hardiness.  Several studies have shown that exposures to O3 at
realistic levels may reduce the cold- or frost-hardiness of plants, as reviewed by Davison et
al. (1988).  It is the temperature within the plant tissues that is important, because it affects
almost all physical and chemical processes within the plant. However, in addition to air
temperature, the  temperature of the leaf is determined by  the absorption of infra-red radiation
and the loss of water vapor through  transpiration.  Further, temperature within the leaf has
been  shown to rise with  the closing of stomata (Matsushima et al., 1985;  Temple and Benoit.
 1988). Therefore,  since vapor pressure deficit and degree of stomatal closure control the
rate of evapotranspiration,  the effects of temperature  are unavoidably confounded with these
other factors.
        Water Usage and Availability.  Water is essential to plant survival, growth and
reproduction.   Because  different regions of the country have different water regimes, plants
growing in each  of these regions are those adapted to the fluctuating water supplies from
 season to season.  These differences among species make it difficult to draw firm conclusions
about the nature  of the relationship between soil moisture deficit (SMD) and O3 effects.  For
 example, Bytnerowicz et al. (1988)  found no interaction  between SMD and O3 effects in 18
 desert annual  species.  On the other hand, in the more mesic environment of the mid-Ohio
 River Valley, a field survey of  milkweed  revealed much less foliar injury attributable to O,

-------
                                           183
 in 1988 (a dry year), even with O3 concentrations reaching 0.2 ppm, than in 1989 (a year
 with ample precipitation), and a maximum O3 concentration of only 0.12 ppm (Showman,
 1991).  Though, in the latter case, SMD seemed to confer some degree of O3 resistance, this
 cannot be extrapolated to other species which have not been studied. Further, the
-^relationship between SMD and O3 may also change throughout the life of a plant or growing
 season, as a plant's sensitivity to  water stress varies with stage of plant development (Moser
 etal.,  1988).
        Recognizing the possibility of an interaction of drought with O3 on the yield of
 agricultural crops, the National Crop Loss Assessment Network (NCLAN) studies conducted
 several experiments to examine this relationship.  Out of eleven studies (six soybean, three
 cotton, and one each of alfalfa and clover-fescue),  only half (three soybean, two cotton, and
Uhe alfalfa) showed significant interactions between SMD and  O3.  In some cases, the lack of
- a significant response to O3 reflects a decreased range of yield response under SMD within
 which  an O3 effect could be ascertained.  Unfortunately, because different measures of SMD
-or SMD-induced stress were  used in different studies, it is not possible to quantify  the
 relationship between the suppression of the O3 response and the level of drought stress.
 Additionally, soil conditions and the depth of the water table at different sites appear to
 influence the O3 response as well  (Heggestad et al., 1988).
        Trees have been the subject of several recent studies on the interaction between SMD
 and 03. Though there is no consistency among the studies in the  treatments used or the
 measurements made, these studies do provide some support for the view that drought stress
 may reduce the impact of O3. For example, beech, poplar, arid loblolly pine seedlings have
 clearly demonstrated significant interactions.  The CD reports that drought has reduced O3_
 Jnduced foliar injury to poplar, ponderosa pine, and loblolly pine.  This work with trees.
-however, is not yet at the point to allow quantification of the  O3-drought interaction.
        The bulk of the evidence supports the view that drought stress may reduce the  impact
 of O3 on plants.  However, it must be emphasized  that, in terms of growth  and  productivity,
 any "protective" benefit will be offset by the effects of SMD per se, as noted in the previous
 criteria document (U.S. EPA, 1986).

-------
                                          184
       The Oj-water interaction is not confined to the effects of SMD on direct plant
response to O3.  Some studies have shown that O3 may affect various other aspects of plant
water status,  including water use efficiency (WUE).  However, WUE is a complex resultant
of both stomatal conductance and the activity of the photosynthetic system, both of which
may be independently affected by O3. Only one study has been performed on trees, and this
was done at high O3 concentrations (Johnson and Taylor, 1989). Though  the foliage of
loblolly pine seedlings at these higher levels adapted to a more efficient use of water,  more
study will be needed before it will be possible to generalize about the implications of this
effect and its importance for mature trees and  forest ecosystems.
       Finally, the relative humidity (RH) of the ambient air can significantly influence O3
uptake.  In one study using pinto beans, O3 uptake increased  fourfold at an O3 concentration
of 0.079 ppm when the relative humidity was  increased  from 35%  to 73% (McLaughlin  and
Taylor, 1981). However, stomatal responses to O3 show considerable variability among
species and even  among cultivars of the same  species (Elkiey, et al., 1979).  The influence
of RH on plant sensitivity may explain important variations in plant response under field
conditions (U.S. EPA,  1986).
       3.     Chemical Factors
       Nutritional Factors.  Plants require a supply  of mineral nutrients such as nitrogen,
potassium, phosphorus, sulfur, magnesium and calcium  for growth.  For optimal growth,
these supplies must be balanced.  A number of studies have examined the relationship
between nutrient  status  and  plant response to O3 exposure.  Heagle (1979) found that injury
and growth reductions tended to be greatest at normal levels of fertility, though the effects
also were dependent on the rooting medium  used.  It has also been reported that increased
levels  of phosphorus, potassium, and sulfur have resulted in a decrease in sensitivity to O3
(CD, 1996).   On the other hand,  with respect  to nitrogen (the area where  the most of the
nutritional research  has been done), the results have been mixed.
       It has been suggested that the relationship between O3 sensitivity and nutrient
condition could be better characterized if studies began with a knowledge  of actual plant
tissue  nutrient levels at the time of exposure to O,.  Cowling and Koziol (1982) indicate that

-------
                                           185
differences in sensitivity are ultimately linked to changes in the status of soluble
carbohydrates in the plant tissues.
       Since these nutritional studies used different combinations of nutrients, species, and
experimental conditions, the results cannot be integrated to develop a general relationship
between soil fertility and sensitivity to O3.  In view of the vast number of possible
permutations and combination of nutrient elements and their levels that may exert effects on
O3 response, a concerted effort by researchers to use standardized protocols will  have to be
made if the uncertainties associated with the role of nutritional status on O3 sensitivity is to
be better understood.
       Interactions with Other Pollutants.  The concurrent or sequential exposure of
vegetation to different gaseous air  pollutants has been found to modify the  magnitude and
nature of the response to individual pollutants (U.S. EPA,  1986).  Lefohn  and Tingey (1984)
and Lefohn et al. (1987) reviewed the patterns of co-occurrence of O3, SO2, and  NO2 in
urban, rural, and remote sites in the U.S. for the years 1978 to  1982 and found that co-
occurrences were usually of short  duration and occurred infrequently.  The most frequent
types of co-occurrence were either purely sequential or a combination of sequential and
overlapping exposures of short duration. This discussion will focus only on those studies
which use exposure patterns or levels that are typical of ambient air.
       An  exception is the co-occurrence of PAN and  O3, which are both constituents of
photochemical oxidant.   The few  studies that have been done on this combination, reviewed
in the 1986 Criteria  Document (U.S.  EPA, 1986). show that the two gases tend  to  act
antagonistically in both  concurrent and sequential exposures.  At the present, no  studies have
looked at the interactions between  O3 and hydrogen peroxide (H2O2), which is another
constituent of photochemically polluted atmospheres.
       Despite the fact that the photochemical formation of O3 involves a complex  series of
reactions in which NO,  NO2 and HNO3 participate as intermediate reaction products, and that
in many areas daily peak O3 levels are followed by increasing NO2 levels,  only a few studies
have been done to explore  possible interactive effects with  O3 and are confined to the
nitrogen species,  NO2 (CD, 1996).  These studies  have reported both antagonistic  and

-------
                                           186
synergistic or additive interactions between O3 and NO2 even with the same species, but,
with such limited information, it is not possible to generalize the response at this time.
       A large number of studies have examined the relationship between O3 and SO2 (CD,
1996).  These studies have used a wide variety of species, exposure regimes, and
experimental conditions.  Because of the contradictory nature of the results from these
studies, all that can be concluded is that the type of  interaction, and whether or not one
exists, is probably highly dependent upon species and cultivar, and possibly other
environmental variables.  The available evidence is insufficient to be able to decide in which
way and to what extent SO2 exposure will influence  the effects of O3 on a particular species
or cultivar at a particular location.
       The recognition of the damaging effects of acid rain on various terrestrial and aquatic
systems has led to numerous studies of the combined effects of O, and simulated acid rain
(SAR) or acid fog.  Due to concern over the possible role of exposures to acid rain or  acid
fog and O3 in the forest decline syndrome, several of the more recent studies have focused
on forest tree species.  Of over 80 recent reports of studies on over 30 species, more than
75% indicated no significant interactions between O3 and SAR or acid fog (CD, 1996).
However, in other studies, statistically significant interactions have been reported for several
species.  In most cases where significant interactions on  growth or physiology  have been
reported, the interactions were mostly antagonistic.  Overall, it appears that exposure to
acidic precipitation is unlikely to result  in significant enhancement of the effects of O3 in
most species. In the few cases of antagonistic interactions, the suggestion was made that
these may  have reflected  a beneficial fertilizer effect due to the nitrate and sulfate present in
the SAR applied.
       Only a few studies have been published on CO2/O3 interactions.  CO2 alone has been
found to increase leaf area and stimulate photosynthetic rates, which can have the secondary
effects of inducing stomatal closure, reducing transpiration, and increasing leaf temperature.
When applied together, CO2 countered  the negative  effects of O3 on photosynthesis, shoot
growth rate, leaf area,  and water use efficiency for  radish and soybean.  Because these
studies were conducted in growth chambers or open-top  field chambers, uncertainties due to
variable environmental conditions would be introduced when applying these results to the

-------
                                            187
 open field.  It is further unclear whether such CO2-induced reductions of the impact of O3
 also apply to the long term growth of trees, and how increased CO2 will affect the impact of
 O3 on ecosystems.
        A limited database  exists for studies involving mixtures of O3 with two or more
 -pollutants.  Due to the small number of studies, it is difficult to draw any firm conclusions.
 In general,  the consequences of such exposures appear to be largely dictated by the dominant
 individual two-way interaction (CD, 1996).
        Agricultural Chemicals.  Several categories of compounds (commercial fungicides,
 herbicides and growth regulators)  that are routinely applied to agricultural plants have been
 found in some cases to protect against O3 injury. Though no comprehensive and  systematic
 studies have been reported, the existing data indicate that certain fungicides are consistent in
 providing protection (CD,  1996).  Most of the effective fungicides have been carbamates and
 have also been used as antioxidants in other applications such as rubber formulations.  Other
 compounds used as growth regulators and herbicides have also been reported  to protect some
-plants against O3 injury. However, these results appear to be species- or cukivar-dependent
 (CD, 1996).  Other than noting the general efficacy of the carbamate fungicides, knowledge
 of the interactions of these different types of agricultural chemicals with O, is still too
 fragmentary' to be able to draw any general conclusions.   Thus, it is considered premature to
 recommend their use specifically for  protecting crops from the adverse effects of O3, rather
 than for their primary purpose (CD,  1996).
 D.    Ozone Effects on Crops and Other Vegetation
        This section presents information on vegetation effects associated with exposures to
 O3.  Effects discussed include: 1)  visible foliar injury, 2)  growth reductions and yield loss in
 annual crops and other species, 3) growth reductions in tree  seedlings and mature trees, and
"4) effects that can have impacts at the forest and ecosystem levels. The section highlights
 results from observational  and controlled studies, together with the limitations and
 uncertainties associated with the studies.
        The results presented in this section are in terms of a number of different air  quality
 index forms.  Table VII-1, which  presents the 10-year summary  of yearly average air quality
 monitored at U.S. sites for the years  1982 to 1991 for three  selected  forms, is included here

-------
                             188
 Table  VII-l.   SUMMARY OF OZONE EXPOSURE INDICES CALCULATED FOR
            3-MONTH GROWING SEASONS-FROM 1982 to 1991

Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
Among

No.
Sites
99
102
104
117
123
121
139
171
188
199
Years


M7
SUM06
ppm
Mean CV
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
052
056
052
052
052
055
060
051
053
054
054
18.
21.
18.
17.
19.
17.
17.
17.
18.
18.
10.
7%
9%
2%
1%
1%
6%
8%
5%
3%
4%
0%
ppni1
Mean
26.8
34.5
27.7
27.4
27.7
31.2
45.2
24.8
25.8
28.3
29.5
•h
CV
68.
58.
58.
59.
65.
56.
46.
78.
76.
74.
42.
8%
1%
4%
6%
0%
4%
8%
7%
2%
2%
1%
SIGMOID
PPIH'
Mean
26.3
33.0
27.4
27.4
27.7
30.4
42.9
25.8
26.6
28.9
29.4
•h
CV
56.
52.
47.
47.
51.
46.
42.
59.
59.
59.
31.
7%
3%
9%
6%
8%
8%
4%
4%
2%
5%
0%
Modified from Table 5-20, U.S.  EPA/1996

-------
                                            189
  to provide a context in which to consider the reported vegetation effects.  These forms
  include the M7 (seven hour seasonal mean), the SUM06 (all hourly O3 concentrations equal
  to or above 0.06 ppm summed over 3 months), and an example of a sigmoidal form,
* SIGMOID, (all hourly concentrations weighted by a specific sigmoidal weighting function
  and summed over 3 months). Another sigmoidal form used in  this chapter, W126, has an
  inflection point at 0.067 ppm and gives equal weight to values above 0.10 ppm.
         1.     Visible Foliar Injury
        Visible foliar injury can be an effect of concern either when it directly represents  loss
  in the intended use of the plant, ranging from reduced  yield and marketability to impairment
  of the aesthetic value of individual plants or natural landscapes, or when it serves as an
  indicator of the presence of concentrations  of O3 in the ambient air which are associated with
- more serious effects. Because visible foliar injury was the first effect of O3  to be observed.
'•  the database associated with it is large and  covers a wide variety of species.   However, much
  of this database is incomplete in terms of characterizing the O3  concentrations and exposure
-" regimes that were experienced by plants in the field, or was produced under  unrealistically
  high or low O3 exposure levels in artificial  growing conditions.  Studies conducted more
  recently have begun to remedy those limitations.
        Reduced Yield or Marketability.  Loss of use may occur when changes in quality
  and/or physical appearance result in reduced yield or marketability of leafy crops (e.g.,
  spinach, lettuce,  cabbage) and ornamental plants.  Unfortunately, little research has been
  done to describe the relationship between O3 concentrations and changes  in visible responses
  on leafy crops and ornamentals. Heck et al. (1984b) summarize O3 effects on a variety of
  vegetables.  Four varieties of spinach are shown to incur 10% yield loss and 30% yield loss
  over the ranges of 0.043 to 0.049 ppm and 0.08 to 0.082 ppm  (7 h seasonal  means),
  respectively (U.S. EPA, 1986).  Additionally, Empire  lettuce was reported to experience a
  10% and 30% yield loss at the 7 hr seasonal mean concentrations of 0.053 ppm  and  0.075
  ppm, respectively. Temple et al. (1986) reported a 35% reduction in lettuce head weight at
  0.128 ppm (7 h mean over 52 days), while Olszyk et al. (1986) found no effects at a 7 week
  mean of 0.034 ppm.  Earlier studies, cited  in the 1978 CD, reported that to  prevent visible
  foliar symptoms for crops, concentrations in the range  of 0.10 to 0.25 ppm for a duration of

-------
                                           190
1 hour were identified as a limiting value, which decreased to 0.04 ppm to 0.09 ppm when
duration was increased to 4 hours.  For trees, the ranges of concentration  were slightly
higher, including 0.06 to 0.17 ppm at the 4 hour duration.  These limiting values are still
considered relevant today, although it is recognized that the studies available at the time often
used experimental protocols that were unrealistic with respect to the natural  growing
environment of the plants (CD, 1996).
       Foliar symptoms that can decrease the value of ornamentals including turf grasses,
floral foliage, and ornamental trees and shrubs have also been reported. For example, when
petunia, geranium, and poinsettia were exposed  to O3 for 6 h/day for 9 days (petunia), 8
days (geranium), and 50 days (poinsettia), flower size was  significantly reduced in all three
species at a concentration of 0.10 to 0.12 ppm, and flower color was reduced  at the  same or
lower concentrations.  All of these changes in flower appearance occurred without visible
injury to the plant leaves. Ozone concentrations of 0.10 ppm  for 3.5 h/day for 5 days or
0.20 ppm for 2 h  were high enough to elicit  injury in most turf grasses (U.S.  EPA,  1986).
       Impairment of Aesthetic Value.  On a larger scale, foliar injury currently occurring
on native  vegetation in national parks, forests, and wilderness areas in  some cases may be
degrading the aesthetic quality of the natural  landscape,  a resource important to public
welfare.  The first concerted effort to relate total oxidant concentrations in national forests to
injury in white pine began in  1975.  Injury was  observed in the Jefferson  and George
Washington National Forests and throughout the Blue Ridge Mountains, including areas of
the Shenandoah National Park (Hayes and Skelly,  1977;  Skelly et al.,  1984).  Taylor and
Norby (1985) report that there were an average  of five episodes (i.e., any day with a 1 h
concentration  >  0.08 ppm) during the growing  season in this area, with episodes  lasting
from 1 to 3 consecutive days.
       In the Great Smoky Mountains National  Park, surveys made in  the summers  from
 1987 through 1990 found 95 plant species, including herbaceous and woody plants, exhibited
foliar injury symptoms consistent with those  thought to be caused by O3 (Neufeld, et al.,
 1992).  At the  same time, O3 monitoring data indicated  that there were both elevated
concentrations and prolonged  exposures to O?, especially at the higher  elevation sites \\hich
could experience as much as 2 times the levels experienced at lower elevation sites.  In order

-------
                                           191
 to verify that O3 produced these symptoms, 28 species that had shown foliar injury symptoms
 in the field were fumigated with O3 in open-top chambers.  Twenty-five of the 28 showed
 foliar injury symptoms like those found in the field in response to O3 (Neufeld, et al. 1992).
        In a similar survey, Chappelka et al. (1992) examined black cherry, yellow poplar,
. sassafras, and white ash in the Shenandoah and Great Smoky Mountains National Parks.
 Black cherry exhibited foliar injury symptoms in both parks, with the percentage of leaves
 injured in 1991 ranging from  18 to 40% and from 8 to 29% in the two parks, respectively.
 Black cherry also exhibited the highest percentage of symptomatic trees (97%).
        The western U.S. contains the largest forested area in the world documented to  have
 visible injury from high O3 exposures, the Sierra Nevada Mountains, an area approximately
 300 miles long (Peterson and  Arbaugh, 1992).  Foliar O3 injury-to ponderosa and Jeffrey
 pine was first documented there in the early 1970's (Miller and Millecan,  1971).
 Monitoring of visible injury to ponderosa pine  on National  Forest land in the western Sierra
 Nevadas, however, was not begun until 1975 (Duriscoe and Stolte,  1989).   Results of the
 monitoring in the Sierra Nevada and Sequoia National .Forests showed that there was an
 increase in chlorotic mottle of pines in the plots from approximately 20% in 1977 to 55% in
 1988, and an increase in severity of injury as well.  Sequoia National Forest and the
 Sequoia-Kings Canyon  National Park,  the southernmost federal administrative units, have the
 highest O3 levels, with  mean hourly averages ranging from 0.018 to 0.076 ppm, and annual
 hourly maxima of 0.11 to 0.17 ppm for 1987.
        Since 1991, there  has been an  annual survey of the amount  of crown injury by O, to
 the same trees in approximately 33 sample plots located in several National Parks and
 Forests in the Sierra Nevada Mountains.  Dominant tree species in the area are ponderosa
 and Jeffrey pine, white fir, sugar pine, incense cedar, Douglas fir,  California black oak, and
 the giant sequoia (Peterson and Arbaugh,  1992).  Big cone  Douglas fir is usually rated  as
 less sensitive than ponderosa or Jeffrey pine; however, injury symptoms resulting from
 elevated O3 have been seen (Peterson  et al., 1995). Based on their study,  the authors
 conclude that while O3  does not have the same  level of impact on this tree as on ponderosa
 and Jeffrey pine, reduced needle retention and lower recent growth rates could indicate
 increased O3 stress (or  O3  stress mediated by climate) in big cone Douglas fir (CD,  1996).

-------
                                          192
       Visible Injury as an Ozone Indicator. Though visible foliar injury cannot at present
serve as a surrogate measure for other O3-related vegetation effects, it can be a useful
indicator that phytotoxic concentrations of O3 are present in the ambient air.  It can thus
serve as a warning that other O3 impacts may be taking place on the injured plant or other
nearby vegetation.
       Several field studies have been conducted which successfully used  sensitive species as
bioindicators of O3 concentration including studies of morning glory (Nouchi and Aoki,
1979), milkweed  (Douchelle and Skelly, 1981), and pinto bean (Oshima,  1975). The value
of deploying networks of bioindicators for the early detection of developing regional oxidant
pollution problems, identification of trends in pollutant occurrence, and the supplementation
of physical monitoring networks has been  demonstrated (Laurence, 1984). This use of plants
as bioindicators is an important element of the  Environmental Monitoring and  Assessment
Program (EMAP), which seeks to identify and  document associations between selected
indicators of natural and anthropogenic stresses  and the condition of ecological resources.
This information  can then be used to track national trends in pollution and provide sound
data on which to  base environmental risk management decisions (U.S. EPA, 1993).
Surveys have been made in Class I areas in New Hampshire and  Vermont during the years
1988 to 1990 (Manning et al., 1991). Ozone injury was extensive on vegetation growing in
open-top and ambient air experimental plots in  both states in 1988 when O3 was unusually
high. The incidence and intensity of O3 injury symptoms was considerably less in both 1989
and  1990, though some degree of symptoms were evident on all plants.  Based on the
studies, it was determined that black cherry, milkweed, white ash, white pine  and two
species of blackberry were all  reliable biological indicators of ambient O3 exposure (Manning
etal., 1991).
       Concurrently, a regional initiative,  the Forest Health Monitoring Program (FHM),
which began in 1990 as a cooperative program between the USDA, EPA, and EMAP.
monitors the condition of forests in the Northeastern United States. In 1992, bioindicator
evaluation was conducted on 39 of 222 forested plots.  Sensitive plant species in the
Northeast include blackberry, milkweed, black cherry, white ash,  and white pine.  Of the 39

-------
                                          193
plots, 28% included plants which showed some visible symptoms of O3 injury (U.S.D.A.,
1993, Summary report).
       2.      Growth/Yield Reductions in Annual Crops
       As was discussed in section VII.B, O3 can interfere with carbon gain (photosynthesis)
and allocation of carbon with or without the presence of visible foliar injury.  As a result,
the carbohydrates that remain may be allocated  to sites of injured tissue or employed in other
repair or compensatory processes, thus reducing plant growth and/or yield.  Growth
reductions indicate that plant vigor is being compromised which can lead  to yield reductions
in commercial crops.
       Agricultural Crop Studies.  Annuals tend to be fast growing, have no need for long-
term storage of carbohydrates, and, in the case  of well-fertilized crops, have less need for
extensive root development.  Instead,  most resources go toward producing seeds for the
following year, making  fruit or seed production the most significant of the processes sensitive
to a reduction in carbohydrate production occurring as a result of O3 exposure.   Changes in
susceptibility to insect damage is likely to be of greater concern than for perennials which
may  have the chance to recover the following year.
       The largest body of research to date on the growth and yield effects of O3 on annuals
is that  for agricultural crop species. The majority of this research occurred before 1986, and
includes the National Crop Loss Assessment Network (NCLAN) studies which remain the
largest, most uniform database for crops available today.  The NCLAN project  which began
in 1980 was originally undertaken to quantify the relationships between O3 exposure and
yields of major agricultural crops.  The project  was intended to provide data necessary for a
credible evaluation of the economic effects of ambient O3 on U.S. agriculture, and for input
into the review of the O3 NAAQS (Preston and  Tingey,  1988).
       The NCLAN  protocol was designed to produce crop exposure-response data
representative of the  areas in which the  crops were typically grown.  The United States was
divided into 5 regions over which a network of field sites was established.  In total, 15
species (corn, soybean,  wheat, hay (alfalfa, clover, and fescue), tobacco, sorghum, cotton,
barley, peanuts, dry beans, potato, lettuce,  and  turnip), were studied.  The  first 13 of these
15 listed  species,  accounted for greater than 85% of US  agricultural acreage planted (Preston

-------
                                          194
and Tingey, 1988). These 13 species, which included 38 different cultivars,  were studied
under a variety of unique combinations of sites, water regimes, and exposure conditions,
producing a total of 54 separate cases.  These studies were a tremendous improvement over
earlier studies because crops were grown 1)  using typical farm practices and  2) using open-
top chambers, which produce the least amount of environmental modification of any outdoor
chamber (CD, 1996). Another major advantage of the NCLAN approach is that it used a
wide range of exposure levels (e.g., charcoal filtered, ambient and modified ambient),
allowing for the use of regression analyses to develop exposure-response functions, which
could be used to predict yield  loss as a function of O3 exposure levels across the range of
treatment levels, cultivars,  and growing conditions used in the studies.
       Some plant scientists continue to express concern that in the case of NCLAN,
experiments using OTC's were designed and conducted in a way that results  in
overestimation of O3 impacts.  For example, the modified ambient treatments contained
numerous high peaks  (O3 concentrations equal  to or above 0.10 ppm), occurring more
frequently than in typical ambient air quality distributions. Such  exposure patterns have
raised  questions among  some researchers as  to how  much of the plant's response was a result
of having an excessive number of high peaks versus a cumulation of more moderate
exposures.  Exposure durations were species dependent but typically went from stand
establishment to harvest (on average 78 days).   Some crops were grown in more than one
geographical region and repeated over years. In addition, the charcoal  filtered chambers
used to establish baseline crop yield loss were  exposed to approximately 0.025 ppm  O3,
which is lower than the range of 0.03 to 0.05 ppm identified in chapter 4 of the staff paper
as the value for seasonal background O3 levels. The result of using this lower level  of 0.025
ppm is an overestimation of yield loss relative to that expected using 0.03 to 0.05 ppm.
These aspects of the NCLAN protocols contribute to the various  types of uncertainty
inherent in extrapolating controlled field study results of percentage yield  reductions to non-
chambered ambient field conditions and crop regions having  different O3 air quality
distributions.
         Based on regression of the NCLAN analyses, at least 50% of the species/cultivars
tested  were predicted to exhibit a 10%  yield loss at  a 7-hour seasonal mean O, concentration

-------
                                          195
of 0.05 ppm or more (CD, 1996).  These findings produced by the NCLAN project have
also been reported in terms of various cumulative exposure indices,  such as the 3-month, 12
hour SUM06 and W126, and are shown in Table VII-2 (derived from Tables 5-21 and 5-22
in the CD). Review of the NCLAN data indicates that differences in O3 sensitivity
within species may be as great as differences between species with substantial variation in
sensitivity from year to year. Figure VII-la and b show how many  of the 54 NCLAN cases
experience a  10 or 30% yield loss, respectively, for each 10 ppm-hr change in O3 exposure
level, expressed in terms of the 12 h, 3 month W126 index. In Figure VII-la, 40% of the
cases will experience 10% yield loss at 40  ppm-hrs.  In contrast, only half that amount,
20%, will experience 30% yield loss  at 40 ppm-hrs.  This suggests that the variability in
sensitivity increases as  O3 exposures increase.  Additionally, Figure  VII-2 (taken  from Figure
5-29 in  the CD) shows that as 24 hour SUM06 levels increase  the range of variability in
relative yield  loss between the 50th and 75th percentiles among NCLAN cases increase, from
a 2 % difference in yield loss at 10 ppm-hr to a 27 % difference at 60 ppm-hrs, thus
showing a disproportional increase in impact on sensitive species as  O3 exposure levels
increase.
       In a re-analysis  of NCLAN data, Lesser et al., (1990) predicted relative yield losses
for a number of crops species or groups of species (compared to an  assumed background
concentration of 0.025  ppm) of 10 to 20%  at 12 h  seasonal means of 0.045 to 0.06 ppm.
respectively.  Most significantly, based on  the NCLAN results,  it can be seen that several
economically  important crop species are sensitive to O3 levels typical of those found in the
U.S.
       Other  studies (on beans, potatoes, tomatoes, onion, and tobacco) examined effects of
O3 under ambient conditions by using the chemical protectant, ethylene diurea (EDU).
Though there has  been  some concern  that EDU might itself alter plant growth, it  is generally
considered  an objective method for evaluating O, effects (U.S.  EPA, 1986).  Estimates of
yield loss were provided by comparing yield from plants grown in ambient air that were
protected with EDU to  those that were not.  Studies indicated that yields were reduced by 18
to 41 <%  relative to the chemically protected controls when ambient O, concentrations
exceeded 0.08 ppm during the day for 5-18 days over the growing season (U.S. EPA,  1978).

-------
                                196
  Table VII-2.   COMPARISON OF  EXPOSURE-RESPONSE  VALUES CALCULI
                   USING THE 3-MONTH,  12-HOUR SUM06 AND  W126  EXPC
                   INDICES FOR 54  NCLAN CASES
Species
Cultivar
Moisture*
                                        3 mo 12-h         3 mo 12-h
                                         SUM06 c              W126d
                                    Values for Yield   Values for Yield
                                       Losses of          Losses of
                                      10%     30%       10%     30%
Barley (Linear)
Baricy (Linear)
Cora (L) b
Corn(L)
Cotton (L)
Cotton (L)
Cotton (L)
Cotton (L)
Cotton (L, Linear)
Cotton (L, Linear)
Cotton
Cotton
Cotton
Kidney Bean
Kidney Bean (L)
Lettuce CO b
Peanut (L)
Potato
Potato
Sorghum
Soybean
Soybean
CM-72 DRY
CM-72 WET
PIO
PAG
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
ACALA DRY
ACALA WET
STONEVILLE
MCNAIR DRY
MCNAIR WET
CAL LT RED
CAL LT RED
EMPIRE
NC-<5
NORCHIP
NORCHIP
DEKALB
CORSOY
CORSOY
173.1
250.0
41.6
55. *8
35.7
23.1
24.8
14.0
63.2
60.0
30.9
73.4
26.6
15.2
17.7
36.5
3 6-. 2
9.9
20.3
67.6
15.3
42.0
250.C
250.0
64.1
.74.1
59.8
42.5
48.0
35.5
103.5
203.2
56.6
114.0
59.3
20.9
28. B
45.5
62.7
33.5
42.5
114.2
21.2
53.0
117.2
1382.7
38.6
55. 0
28.3
16.4
18.8
9.1
61.6
62.0
22.2
68.5
22.6
15.0
15.6
34.6
29.4
10.1
20. 1
65.0
' 12.2
35.5
250.0
3329.4
62.2
73.6
53.0
35.0
41.6
28.4
107, -}
210.0
48.4
113.4
54.2
20.9
27.1
44.3
56.1
34.3
si.;;
121.2
28.5
48.2

-------
                              197
Table VII-2.   (Cont'd.).   COMPARISON OF EXPOSURE-RESPONSE VALUES
                  CALCULATED USING THE 3-MONTH,  12-HOUR SUMO6
                  AND W126 EXPOSURE INDICES  FOR  54  NCLAN CASES
                                      3 mo 12-h         3 mo 12-h
                                      SUH06C              W126d
                                  Values  for Yield  Values for Yield
                                     Losses of         Losses of
Species
Soybean

Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Soybean
Tobacco (IJ
Turnip (T)
Turnip (T)
Turnip (T)
Culiivar
AMSOY

PELLA
WILLIAMS
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
CORSOY
WILLIAMS
WILLIAMS
HODGSON
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
DAVIS
YOUNG
YOUNG
MCNAIR
JUST RIGHT
PURPLE TOP
SHOGOIN
Moisture*




DRY
WET
DRY
WET
DRY
WET
DRY
WET



DRY
WET
DRY
WET
DRY
WET




1U%
32.8
18.2

15.5 .
71.2""
70.0
89.1
62.2
10.2
11.8
21.1
14.8
8.4
13.8
23.4
57.1
35.2
45,9
24.1
38.8
25.0
24.4
7.4
5.9
6.6
JU%
51.4
61.6

52.4
81.2
154.0
96.4
87.6
34.7
29.9
48.8
36.6
28.4
46.9
47.2
193.4
79.6
66.1
55.7
90.1
(,4.9
70.3
14.5
13.4
14.0
1U%
24.6
16.9

14. S
66.8
63.1
91.5
56.7
10.0
8.9
17.2
11.4
8.0
13.7
19.2
56.9
26.8
40.2
18.9
32.8
20.4
18.9
5.2
4.0
4.1
JO%
46.1
57.2

49.2
76.7
188.7
101.7
86.5
34.0
26.3
46.1
33.0
27.0
46.4
43.6
192.6
73.5
62.3
50.8
87.4
60.0
63.9
12.0
11.6
11.7

-------
                                198
       Table VII-2.
(Cont'd.).   COMPARISON OF EXPOSURE-RESPONSE VAL1
 CALCULATED USING THE 3-MONTH,  12-HOUR SUMO6
 AND W126 EXPOSURE INDICES FOR  54  NCLAN CASES
                                         3 mo 12-h
                                         SUM06 c
                                     Values for Yield
                                        Losses of
                                   3 mo 12-h
                                     W126d
                               Values for Yield
                                  Losses of
Species
Turnip (T)
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Wheat
Cuhivar
TOKYO
CROSS
ABE
ARTHUR
ROLAND
ABE
ARTHUR
VON A
VONA
Moisture1 10%
9/3
25.1
21.3
7.4
34.8
27.7
2.9
7.7
30%
IS. 9
37.5
37.5
21.3
40.9
46.5
9.7
16.5
10%
7.2
22.1
17.3
5.4
32.3
25.4
2.6
6.0
30%
14.7
35-4
35.0
18.1
40.0
46.2
8.8
14.6
*Wet  refers to experiments conducted under well-watered
 conditions  while dry refers to experiment conducted under  some
 controlled  level of drought stress.
 "For those studies whose species name is followed by "(Linear)" a
 linear model was  fit.  A Weibull model  was  fit  to all  other
 studies.  For those  studies whose  species name  is followed by
 " (L) "  a log transformation was  used to  stabilize the variance.
 For  those crops whose name is followed  by "(T)" the yield is
 expressed as either  g/plant or  g/m.
 CT,hne0 12rh  S1™06  val^e (ppm-h)  that was predicted to cause a  10  or
 30-g yield loss  (compared to zero  SUM06) .
 o*!fo 12~h  W126 Value (PP1"-11)  that was predicted to cause a  10  or
 30-s yield loss  (compared to zero  W126) .

 Modified from CD  Tables 5-21 and 5-22.

-------
                                 199
FIGURE  VI1-1.  Variability in NCLAN Crop Yield Sensitivities
    A.    i"
          O

          £



          03

          (0
          O
Ld
                           ,
                              t   f
                        •0 TO K 90 1OO 110 190 1*0 140 IK 1«D 1TO 1«0 1«C «DO »10 230 MO 94O

                        Ozone Concentration, ppm-hr (3 MO,, 12HR, W126)
    B.
        Graph Shows Different Ozone  Levels Required  to
         Produce  10% Yield  Loss in 54 Different Cases
g 12
O
"o 10
.2
O
CO
-C 6
i
CO "
CO
CO
3 '
"o



_



-

-


I
1






1





i




.
i

1

t
[


i " "" " "
I _
i •{ f
1 1 r j 	 	 *
III ,111 	 Ill
^ ii «c to 40 v «o ?» *> to  no i» i«o i« i«o t« ijo iw i« n nc no a*> »w an
                        Ozone Concentration, ppnvhf (3 MO., 12HR, W126)
        Graph Shows Different Ozone  Levels Required to
         Produce  30% Yield  Loss in 54  Different Cases

-------
                          200
Figure VII-2.  Median Crop  Yield  Loss from NCLAN Crops
                                                  50th p«fO»ntH«
                                                  25th poroentHe
                                               60
                      24-h Sum06 (ppnvh)
           Box-plot distribution of yield loss predictions
           from Weibull  and linear exposure-response models
           that relate biomass and ozone exposure as
           characterized by the 24-h SUM06 statistic using
           data from 31  crop studies from the National TJrop
           Loss Assessment Network (NCLAN) program.  Separate
           regressions were calculated for studies with
           multiple harvests and/or cultivars resulting in a
           total of 54 individuals equations from the 31
           NCLAN studies.  Each equation was used to
           calculate the predicted relative yield or biomass
           loss at 10, 20, 30, 40, 50 and 60 ppm-h and the
           distributions of the resulting losses plotted.
           The solid line is the calculated Weibull fit at
           the 50th percentile.  From Hogsett et al. (1995) .

-------
                                         201
       Taken together, the studies discussed above as well as others (e.g., Heagle et al.,
(1988b), Miller et al., (1988), and Temple et al., (1988) continue to provide a basis for the
conclusion presented in the last two Criteria Documents that O3 concentrations at ambient
levels in the U.S. are sufficiently elevated in several parts of the country to impair the
growth and yield of commercial crops.
       Caveats and Uncertainties.   In order to isolate and measure a plant's response to O3
from the plant's response to other environmental variables, many study designs employ some
type of exposure chamber.   A chamber allows the researcher to create a variety of O3
regimes while all other variables are kept constant or their conditions well-characterized.
Though there are numerous fumigation  systems, the  most widely utilized design has been the
open-top chamber (OTC).  This review of the standard relies heavily on agricultural crop
exposure-response functions developed in open top chambers during the National Crop Loss
Assessment Network (NCLAN) from 1980 to 1988.  Other types of exposure methods are
also discussed below.  For a more detailed discussion of methods and uncertainties, see the
discussion in  the 1996 CD.
       The main advantage of the OTC is that it provides the least amount of environmental
modification of any  outdoor chamber.  However, the open-top chamber does alter ambient
microclimate conditions such as light intensity, wind velocity, rainfall,  dew formation and
persistence, air temperature and relative humidity (CD, 1996).  As a result, exhaustive
comparisons between plants grown in carbon filtered chambers (CF), non-filtered chambers
(NF), and similarly  sized and located ambient air plots (AA) have been conducted. These
comparisons demonstrate that the only consistently observed effect is that chamber plants
were taller than those grown in the ambient air (Heagle et al., 1988a).  Though plant yield in
ambient air can be greater than, less than or equal to that in NF chambers, there is no
evidence that there is a large effect of chambers on  plant response to O3 (Heck et al., 1994:
Heagle et al.,  19S8a; Olszyk et al., 1992).
       An additional concern in open-top chambers is the addition of trace pollutants (N205
and NO) in chambers receiving O3 generated from dry air and NO2 in chambers receiving
ambient air.   Generated O3 treatments have been shown to be more phytotoxic than ambient
O3 treatments.  To avoid this problem, OTC  studies  have used filtered  vs.  non-filtered
ambient O3.  The drawback to this approach is that  low ambient O3 levels make detecting

-------
                                          202
differences in plant response between filtered and non-filtered chambers difficult, thus
requiring a high number of replications for statistical reasons.
       Several other exposure methodologies have been employed in more natural
environmental conditions.  One method involves using chemicals, specifically EDU, to
protect plants from the effects of O3 in the field.  This technique is cheaper and easier to
apply to large areas than open top chambers, eliminates the uncertainties associated with
chamber effects, and reduces uncertainties associated with scaling up from small to larger
areas.
       However, field protocols for the use of EDU have not been well established.
Frequency and rate of application that protects plants vary with species and edaphic and
atmospheric conditions.  Two-treatment studies of EDU and  plant response to O^  (Kostka-
Rick and Manning, 1992a,b) indicate that protection is variable, suggesting that the
experimental system under investigation (soil,  plant, climate) would have to be extremely
well characterized and understood for interpretation of results.  The mechanism by  which
EDU protects plants, beyond being a systematic antioxidant,  is unknown;  understanding this
mechanism has the potential to contribute to the broader understanding of the mechanisms of
O3 injury at the cellular/metabolic level of the plant.
       A third  method uses open air field fumigation systems, such as the  zonal air pollution
system.  Such systems have the capability to fumigate plants  with diurnally varying patterns
of concentrations typical of ambient O3 under  field conditions.  However,  studies which use
such systems are the least controllable and repeatable.  Another method is  the  ambient
gradient system.  This method is structured  to take into account the preexisting gradient of
pollutant concentrations over a given area where a species is  grown.  For ambient gradient
studies  to be interpretable, good characterization of site parameters (rainfall, temperature.
radiation, soil type, etc.) is needed,  as well as knowledge of how these factors should be
used to adjust apparent  plant response.  At this time, however, the  relationships are still
incompletely understood, and therefore confound interpretation of the results.
       3.     Growth Reductions in Tree Seedings and Mature Trees
       Since the preparation  of the 1986 CD,  a number of new studies have been published
relating O3 exposure to effects on deciduous and evergreen seedlings and mature trees.

-------
                                           203
  These studies partially address a significant gap in O3 effects data identified by CASAC in
  the last review of the O3 NAAQS.
        The relationship between responses to O3 exposure in seedlings and mature trees is
  still not well understood.  Several studies describe a number of potential differences between
, seedlings and mature trees (Cregg et ah, 1989; Pye, 1988) including stomata number,
  photosynthetic rate, water use efficiency, nutritional needs, recycling capacities, and canopy
  effects (e.g., sun vs.  shade, wind speed, CO2 concentrations).  Limited experimental
  evidence  shows no consistent relationship between the sensitivities of seedlings and mature
  trees.  For example, Samuelson and Edwards (1993) found that while the canopy weight of a
  mature 30 year old northern red oak experienced  a 41 % reduction when exposed to  a 7 h
  seasonal mean of 0.069 ppm as compared to a very low exposure level of 0.015 7 hr
  seasonal mean, two year old seedlings were not affected at similar exposures.  Thus, because
  tree seedling studies can not at this time be extrapolated to quantify  responses to O3  in
  mature trees,  they will be discussed separately below.
        Deciduous And Evergreen Seedlings.  Growth and productivity has been reported  to
  be affected  by O3  for numerous species in the Blue Ridge  Mountains of Virginia.  In the
- Shenandoah National Park, Duchelle et al.  (1982, 1983) compared the growth of seedlings
  and  productivity of herbaceous vegetation grown  in charcoal-filtered air in open-top  chambers
  to that in open plots,  and found that tulip poplar,  green ash, sweet gum, black locust, as  well
  as several evergreen species (e.g., Eastern hemlock, Table Mountain pine, pitch pine, and
  Virginia pine), common milkweed, and common blackberry all  demonstrated growth
  suppression.  Except  for the last two species mentioned, almost no visible injury symptoms
  accompanied the growth reductions.
        Between 1989 and 1992, the EPA's National Health and Environmental Effects
  Research  Laboratory-Western Ecology Division (NHEERL-WED) in Corvallis sponsored a
  research program  to address the effects of tropospheric O, on forest trees.  Using the same
  open top  chamber methodology as NCLAN, this program  has developed exposure-response
  functions for six deciduous species, including aspen, red alder, black cherry, red maple,
  sugar  maple, and tulip poplar and five evergreen  species,  including douglas fir. ponderosa
  pine, loblolly pine, eastern white  pine, and Virginia pine.   Table VII-3 presents the
  individual results for all cases evaluated in this  study (Table 5-28 from the CD).

-------
                              204
Table VII-3.
 Exposure-Response Values that Relate Total  Biomass
 (Foliage/ Stem, and Root)  to 12-H SUMO6
Exposures(*)  Adjusted to 92 Days (ppm-h/year)
Rate of
Growth Habit Study Species
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
FAST
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
SLOW
FAST
FAST
FAST
FAST
D
D
D
D
D
D
D
D
D
D
D
D
D
D
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
D
D
D
D
1
1
2
2
3
3
4
4
4
4
5
5
5
6
7
7
7
7
8
8
8
9
9
10
10
10
10
11
11
11
12
13
14
15
15
16
Aspen - wild
Aspen - wild
Aspen - wild
Apsen - wild
Apsen - wild
Aspen - wild
Aspen 216
Aspen 253
Aspen 259
Aspen 271
Aspen 216
Aspen 259
Aspen 271
Aspen - Wild
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Douglas Fir
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Ponderosa Pine
Red Alder
Red Alder
Red Alder
Red Alder
SUMO 6 for Loss
of
10% 30%
19.08
15.83
43.72 .
55.90
55.44
18.66
14.70
8.09
4.69
13.28
9.52
5.18
29.64
14.99
89.31
250.00
90.84
94.44
72.03
70.82
63.03
17.86
26.30
18.53
27.09
11.29
21.64
19.47
14.86
27.85
55.18
43.42
32.05
17.87
79.04
35.84
43.14
53.60
63.67
70.81
66.49
45.76
37.78
27.38
15.87
24.58
32.21
17.56
37.71
50.73
160.46
250.00
109.85
250.00
73.89
70.94
69.23
60.45
89.03
55.96
85.87
38.23
73.25
64.72
50.29
69.15
86.23
146.93
68. 80
60.51
95. 77
121.31

-------
                                205
 Table VII-3. (cont.)
Exposure-Response Values that Relate Total Biomass
(Foliage,  Stem,  and Root)  to 12-H SUM06
Exposures(*)  Adjusted to 92  Days (ppm-h/year)
Rate of
Growth Habit Study
FAST
FAST
FAST
FAST
SLOW
FAST
FAST
FAST
FAST
FAST
SLOW
SLOW
SLOW
SLOW
SLOW
D
D
D
D
D
D
D
D
E
E
D
D
E
E
E
16
17
18
19
20
21
21
22
23
23
24
24
25
25
26
SUM06 for Loss
of
Species 10% 30%
Red Alder
Red Alder
Black Cherry
Black Cherry
Red Maple
Tulip Poplar
Tulip Poplar
Tulip Poplar
Loblolly GAKR 15-91
Loblolly GAKR 15-23
Sugar Maple
Sugar Maple
E. White Pine
E. White Pine
Virginia Pine
250.00
21.81
6.59
4.37
71.74
23.44
19.85
14.66
71.03
212.08
25.29
23.81
21.60
31.51
191.24
250.00
67.19
20.31
12.60
147.00
28.69
67.19
25.25
240.44
250.00
30.26
29.14
28.28
106.68
250.00
Source:  Hogsett  et  al.  (1995).

*Note:  Seeding studies  were not all of equal duration.  To
compare the results  from seedling studies of varying exposure
duration, the  SUM06  value is calculated for an exposure of a
fixed period of 92 days  per year.  The calculation assumes that
.exposures can  be  scaled  up or down in uniform fashion.

-------
                                          206
       When the distribution of the relative yield losses for various percentiles of the
deciduous and evergreen seedling studies are aggregated, (Figure VII-3), a 24 hr. SUM06
exposure of 33.3 ppm-h over 92 days is associated with less than 10% biomass reduction in
50% of the seedling cases studied.  When the exposure-response functions for just the
deciduous seedling cases are combined, the results show that a lower 24 hr. SUM06
exposure of 31.5 ppm-h over 92 days is associated with less than 10% biomass reduction in
50% of the deciduous cases (Table 5-29, CD).  For the evergreen seedlings, a 3 month, 24
hr.  SUM06 exposure of 42.6 ppm-h was predicted to result in less than  a 10% biomass loss
in 50% of the evergreen cases studied.  Thus, the evergreen seedlings studied, on average,
exhibited less sensitivity to O3  than the  deciduous seedlings studied.
       As with crops, these studies also showed that there was significant variability in
sensitivity to O3 among species and genotypes within  species.  For example, red alder
seedlings showed  substantial variability, with a 10%  reduction in biomass observed over the
range of 21.7 to 95.8 ppm-h (24 h SUM06), and a 30% biomass loss observed  over the
range of 73.5 to 250.0 ppm-h  (24 h  SUM06).  This variation in  growth  response to O3
exposure can also result from different  climates and growing environments (e.g., drought,
nutrient level),  pest/pathogen interactions, exposure intensity and dynamics, and genetics
(Hogsett et al., 1995). Because it is not known whether the genome(s) that were studied
represent the complete range of sensitivities within a given species, the results from these
studies should be used with caution.
       Some studies further  showed that several deciduous species as seedlings are extremely
sensitive to O3  with respect to  biomass  loss.  For example, Davis and Skelly (1992a, b) and
Simimi et al., (1992)  describe  black cherry seedlings  as very sensitive, with 24 hour SUM06
exposures as low as 12.9 ppm-h over 92 days predicted to  cause a 10%  biomass loss
(Hogsett et al., 1995). Two different Aspen clones showed 10% biomass loss at 24 hour
SUM06 exposures of 10.96 and 9.49 ppin-h, respectively.  Given the mean 3 month 24 hour
SUM06 value over the 10 year period 1982-1991 of 29.5 ppm-h (from Table V1I-1), the
potential for biomass  loss in such sensitive seedling species could be significant.

-------
                           207
Figure VII-3.   Median Biomass Loss  From Seedlings
100% -3
         Tree Seedlings
 70%-j

 60%-f

 50% \

 40%-=
                                             ^^_ 75th peroentile
 30%-4                                        ^~
                                                  50th pe
-------
                                          208
       Other studies have been performed on Douglas fir, Jeffrey, Lodgepole, Monterey,
ponderosa, shore, sugar, and western white pine and Sitka spruce by Wilhour and Neely
(1977).  In closed-top field chambers, O3 was added to charcoal-filtered air at the constant
exposure of 0.10 ppm for 6 h/day for 126 days.  Across species,  observed  reductions ranged
from 0 to 11 % for height and 0 to 21 % for stem  dry weight.  In another study, hybrid poplar
was exposed to 0.15 ppm for  12 hrs/day for 102 days in open-top chambers, with observed
reductions ranging from 3% to 58% for height and  1 to 14% for stem specific gravity
(Patton, 1981).  Hogsett et al. (1985a) noted growth reductions in height, diameter, and root
systems in two varieties of slash pine seedlings under chronic episodic exposure regimes
typical of the southeastern U.S. Both varieties of slash  pine exhibited an increasing
reduction  in growth with increasing O3 concentration, with the most pronounced change
observed in the growth  of roots.  The significance of these findings is not yet understood.
Because trees are perennials, the effect of even a 1-2% per year loss in seedling biomass
(versus 10 to 20% yield loss in crops), if compounded over multiple years  under natural field
conditions of competition  for resources, could be severe.  Furthermore, given the variability
in meteorology and O3 concentrations between years, it  is not known to what degree
seedlings  would recover given periods of more favorable conditions. Because of these
uncertainties, the staff caution against treating equal percentages of yield loss in annuals and
biomass loss in perennials as representing the same  degree of adversity.
       Uncertainties and Limitations In Seedling Studies.  In order to more accurately
understand the results of the seedling studies presented above, there are  several important
caveats and limitations to  keep in mind.  The 11 species selected were grown in open top
chambers (see discussion of caveats and uncertainties for OTC above).  The influence of
multiple environmental  factors were not taken into account as the seedlings were grown
under  optimal growing  conditions and few experiments included multiple year exposures.
These  facts make it problematic when trying to predict effects on  perennial species growing
in an ecosystem context (Hogsett,  et al., 1995).   The parameter used to measure O3 effects
on seedlings, total biomass loss, is measured, against biomass at an O3 concentration level of
SUM06 equal to 0 ppm-hrs.   However, the database shows that there is no distinguishable
threshold  between concentrations  that produce an effect  and those that do not. and biomass

-------
                                          209
loss occurring at exposure levels below 0.06 ppm may be significant for some sensitive
species.  Thus, the data are limited to the conditions under which the experiments were
conducted.
       Deciduous and Evergreen Trees.  Many important field observations of mature
evergreen trees occurred prior to 1986, and were discussed in the 1986 CD.  One such study
by Mann et al. (1980) reported a reduction in radial growth of sensitive white pine
individuals of as much as 30-50%  annually over a period of 15 to 20 years  on  the
Cumberland Plateau in Tennessee.  Field studies in the San Bernardino National Forest
indicated that over a period of 30 years, O3 may have reduced the growth in height of
ponderosa pine by as much as 25%, radial growth by 37%, and total volume of wood
produced by 84% (Miller et al.,  1982).  Because these observations were  made in the field
with numerous uncontrolled environmental factors,  the extent to which the observed growth
effects can be attributed to O3 is uncertain.  It is reported, however,  that O3 was a significant
contributor that potentially exacerbated the effects of the other environmental  stresses.
       Several field studies indicate that injury associated with exposure to O3 and other
oxidants has been occurring in the Appalachian Mountains for many  years. Benoit et al.
(1982) conducted studies in the Blue Ridge Mountains  of Virginia to evaluate the long term
effects of oxidants on  growth in eastern white pine  of reproducing age.  By comparing
growth rates from the period  1955-1959 with those in  1974-1978, decreases of 26, 37, and
51% were reported for tree species characterized as tolerant, intermediate, and sensitive,
respectively.  Because no significant change in seasonal precipitation occurred over the same
time period, the effects on growth were attributed to O3, which during the latter  time period
reached peaks frequently in excess of 0.12 ppm and monthly averages of 0.05-0.07 ppm on a
recurring basis (U.S. EPA, 1986).   Duchelle et al.  (1982), monitoring in the  same area,
reported peak hourly averages >0.08  ppm for the months of April through September in
1979 and 1980.   As early as 1979,  Skelly et al.  (1984) concluded that the most  sensitive
eastern white pines were so severely injured by oxidant exposure that they were  probably
being removed from the population.
       In 1985, to evaluate growth changes in O3-stressed ponderosa and Jeffery pine.
Peterson and  his coworkers conducted the largest investigation of regional tree growth in the

-------
                                         210
western U.S. (Peterson et al., 1987; Peterson and Arbaugh,  1988, 1992; Peterson et al..
1991).  Using cores to determine whether growth reductions had occurred, they randomly
sampled both trees with  visible O3 injury symptoms and asymptomatic trees.  Major
decreases in growth occurred for both symptomatic and asymptomatic trees during the 1950's
and 1960's. The percentage of trees exhibiting growth decreases at any given site never
exceeded 25% in a given decade (Peterson et al., 1991). Mean annual radial increment in
trees with visible symptoms of O3 injury was 11% less than trees at sites without O3 injury.
Trees larger than 40 cm diameter and trees older than 100 years showed greater decreases in
growth  than smaller and younger trees. Again, the significance of these effects on  the above
and below ground forest ecosystem is unknown.
       The response of a number of fruit and nut trees to O3 exposure has been reported
(McCool and Musselman,  1990;  Retzlaff et al.,  1991, 1992a, b).  Almond has  been
identified as the most  sensitive, but peach,  apricot, pear, and plum have also been affected
under study conditions.  Net growth of almond,  as well as stem diameter of peach and the
stem diameter and number of shoots produced on apricot were reduced by  four months of
once-weekly exposure to 0.25 ppm-h O3 for 4 h (a high level of exposure generally found
with fruit and nut trees only in California).  A few studies have measured  O3 effects on citrus
or avocado.  Valencia orange trees (during a production year) exposed to a seasonal 12 h
mean of 0.04 and 0.075 ppm had 11 and 31% lower yields than trees grown in  filtered air at
with a very low O3  concentration of 0.012  ppm.  In contrast, growth  of Ruby Red grapefruit
was not affected by concentrations 3 times ambient (CD, 1996).  Avocado growth was
reduced by 20 or 60% by exposure to 12 h seasonal means of 0.068  and 0.096 ppm during
two growing seasons.
       The methodologies employed in mature tree studies often differ from those used with
seedlings.  In only a few cases have OTCs large enough for mature trees been  used, which
were built at great expense (Mandi et al.. 1989).  Other exposure methods that have been
used with large trees include branch and leaf chambers. Though these chambers have many
advantages, it is not yet known whether the branch or leaf being studied is responding the
same as other parts  of the plant which are  experiencing different environmental  conditions.

-------
                                          211
Thus, estimating total tree response from branch or leaf chamber studies introduces a high
degree of uncertainty.
       In light of the greater difficulties associated with conducting studies on mature trees,
the scientific community has developed process-level models to simulate growth and stand
dynamics over time under various O3 levels.  These models "utilize the large base of data in
tree physiology and forest ecology, watershed chemistry, and atmosphere-forest canopy
meteorology to develop models of tree physiology and growth and  to subsequently  scale these
investigations to the levels of forest stands and landscapes" (Taylor, et al., 1994).  TREGRO
models whole tree or seedling growth and simulates multiyear O3 effects to either seedlings
or mature trees under selected climates or soil conditions. ZELIG  considers the competition
for resources that occurs between four individuals of the same or different species in a stand.
Such modeling studies  are expected to lead to a better understanding of O3 effects on mature
trees and forests in the future.
       4.      Forest and Ecosystems Effects
        Plant populations can be affected by O3 exposure particularly when they contain
many sensitive individuals. Changes within sensitive populations, or stands, if they are severe
enough, ultimately can change community and ecosystem structure.  This progression of
effects is depicted  in Figure VII-4 (Figure 5-30 from the CD).  Structural changes that alter
the ecosystem functions of energy flow and nutrient cycling can arrest or reverse ecosystem
development.
       The only known example of the above sequence of events occurring in which O3 has
been a fundamental stressor, is the San Bernardino Forest ecosystem.  This ecosystem  has
experienced chronic O3 exposures over a period of 50 or more years.  From  1968 to 1972
the average  daily maximum  for total oxidants for  each month was measured at Rim  Forest
(5,640 ft.), in southern California, where the highest concentrations are usually recorded.
For  the months of May through August, the average daily maximum for total oxidants went
from a low of 0.14 ppm in 1969 to approximately 0.28 ppm in 1971, with concentrations
rarely going below 0.05 ppm at night at this elevation. For the same period  the total number
of hours/month exceeding 0.10 ppm  varied from around 4 to  15. Ozone concentrations
exhibited a cyclic diurnal pattern, with the monthly average of hourly values ranging from

-------
                                     212
   Figure VII-4
 Diagram of the Propagation Pathway of  Ozone
Effects  From Plants  to Ecosystems
Atmospheric Processes
  Canopy Processes
   Leaf Processes/
    Ozone Uptake
    Leaf Processes/
    Mode of Action
    Plant Response
         I
 Population Response
           Reduced
         Carbohydrate
          Production
                          Carbohydrate
                            Allocation
  Ecosystem Response
                                        Myoorrhizae Formation
    Effects of ozone on plant function and growth. Reduced carbohydrate
    production decreases allocation and resources needed for plant growth
    processes.  Individual plant responses must be propagated hierarchically
    through the more integrative levels of population and community to
    produce an ecosystem response.  Solid black  arrows indicate the effects of
    ozone absorption; stippled arrows indicate affects on plant functions.
    Double border indicates site of response; darkened border indicates site of
    impact.

-------
                                           213
 0.07 to 0.10 ppm at 10:00 am and from 0.15 to 0.22 ppm at 4:00 pm.  The primary effect of
 O3 at these high levels was on the more susceptible members of the forest community,
 individuals of ponderosa and Jeffrey pine, that could no longer compete effectively for
 essential nutrients, water, light and space.  As a consequence, there was a decline in the
^sensitive species, permitting the enhanced growth of more tolerant species (Miller et al.,
 1982; US EPA, 1978, 1986).
        Follow-up studies of the San Bernardino  forest ecosystem done from 1973 to 1978
 reported that the major changes in the ecosystem began with injury to ponderosa and Jeffrey
 pine. Foliar injury, premature senescence, and needle fall decreased the photosynthetic
 capacity of stressed pines and reduced the production of carbohydrates needed for growth and
 reproduction by the trees.  Decreased carbohydrate production resulted in  a decrease in radial
 growth  and  in height of stressed trees. Numerous other organisms and processes were also
 affected either directly or indirectly, including successional patterns of fungal microflora and
 their relationship to the decomposer community.  Nutrient availability was  influenced by the
• carbon and mineral nutrients accumulated in the heavy litter and thick needle layer under
 stands with  the most severe needle injury and defoliation.  A comparison of species of
 -lichens  found  on conifers during the years 1976 to 1979 with collections from the early
 1900's indicated a 50% reduction in species in the more recent period.
 The sequence of events occurring in the San Bernardino forest confirm that adequate
 protection of vegetation will  have indirect benefits for soils, wildlife and other welfare
 categories.
        Studies on the combined effect of O3 and nitrogen together, typical of the  conditions
 in the San Bernardino forest, have shown  that the effect of O3 for  some species is greater at
 higher levels of nitrogen than at low levels of nitrogen. Because nitrogen and sulfur
 compounds  also occur in the pollutant mixture to which the  mountains downwind of Los,
 Angeles are exposed (Bytnerowicz et al., 1987 a, b; Solomon et al., 1992), this finding
 suggests that plants grown with a high nitrogen supply are more sensitive  to chronic O3 stress
 in terms of biomass reduction (Tjoelker and Luxmoore, 1991).
        Since the period when these earlier studies were conducted, air quality in  the San
 Bernardino region  has improved.  For example,  the total number of days/year with

-------
                                          214
concentrations greater than 0.12 ppm was as high as 159 in 1978, whereas only  105  such
days occurred in  1990 (Davidson,  1993).  For the period 1974 to 1988, there was an
improvement shown in the injury index used to describe chronic injury to crowns of
ponderosa and Jeffrey pines in  13 of 15 plots located on the gradient of decreasing O3
exposure in the San Bernardino Mountains (Miller et al., 1989). Two exceptions were noted
in plots located at the highest exposure end of the gradient, where the basal area increase of
ponderosa pine was generally less than competing species.  Ponderosa and Jeffery pines in
plots with slight to  severe crown injury lost basal area in relation to competing species that
are more tolerant to O3, namely white fir, incense cedar, sugar pine and California black
oak.  In effect, stand development was reversed, and the development of the normal  fire
climax mixture dominated by ponderosa and Jeffery pine was altered. This allowed  the
formation of a fuel  ladder that could jeopardize the remaining overstory trees in the event of
a catastrophic  fire.  Continued monitoring of this system  is needed to determine if declining
O3 would eventually allow  ponderosa and Jeffery pine to resume dominance in basal  area.
       An example of the consequences of losing a dominant species in eastern  forests is the
elimination of the American chestnut from eastern deciduous forests in  North  America during
the first half of this century (Taylor and Norby,  1985).  Before the blight, American  chestnut
comprised 20 to 25% of the canopy in eastern Tennessee.  A full half century later recovery
patterns are complex, with  six distinct successional forest types  occupying former American
chestnut sites.  Thus, it  would appear that the nature of community dynamics, particularly  in
mixed species, uneven aged stands, indicates that subtle  long-term forest responses (e.g.,
shifts in species composition) to elevated levels of a chronic stress like exposure to O3 are
more likely  than wide-spread community degradation (Shaver et al.,  1994).
       Dieback of the spruce-fir forests in the Appalachian mountains has been  attributed to
many causes, with O3 sometimes listed.  Though this high elevation forest is exposed to a
broad range of air pollution stresses, the main  culprit of the dieback  for Mt. Mitchell is
always stated as the balsam wooly adelgid, an  insect.  There have been no  studies done to
show insect preference for  O3 damaged trees in this system or that O3 had weakened  trees
attacked by  the insect.  However, given that O3 can predispose some plants to insect  attack.

-------
                                             215
  and considering the example of such a connection in the San Bernardino Forest in California,
  staff believes this possibility deserves further study.
  E.     Biologically Relevant Measures of Ozone Exposure
         The CD lists a number of exposure indices that have been used in vegetation research.
  These measures vary considerably in their ability to capture biologically relevant aspects of
  O3 exposure that have been shown or theorized to have the greatest potential to influence
  plant response, thus making it difficult to compare study results that are expressed using
  different indices.  Therefore, a discussion of the current scientific understanding of exposure
  dynamics and how different index forms capture these exposure features is described below.
  This information, in combination with the information on welfare effects discussed in the
  preceding section,  permits the evaluation of plant response relatrve to O3 exposure  levels,
  patterns, and duration.
         1.     Biological Considerations
         The information  discussed in the previous sections describes the plant processes that
  can be  impacted by O3 once it enters the leaf, and the ways a plant can protect itself in some
  cases from O3  injury through stomatal control, antioxidant production, or compensation.
  Additionally, numerous  sources of variability/uncertainty (e.g., biological, chemical,
  physical, and experimental) were presented that must  be considered when explaining study
  results  or comparing one study scenario to other studies done  under a different set of
  conditions.   Thus,  in the discussions below of the features of O3 exposure (e.g.,
  concentration,  duration, timing, and exposure pattern) that influence plant uptake, it is
  important to remember that the magnitude of the  plant response will be modified by the
  environmental  and  biological context in which these exposures occur.
         One measure, O3 uptake,  accounts for both the biological and air quality features.
  Of all the available exposure measures identified, O3 uptake most closely relates to O3 dose
  which is the concentration and duration of the O3 exposure that is taken up by the plant and
-_ actually reaches the target tissue.  However, because uptake depends on  so many species-
  and situation-specific variables, it is very difficult to measure and not particularly useful as a
  basis for standard  setting. As  a result, researchers have  focused their research on identifying
  suitable surrogate exposure indices for plant response  to O3.

-------
                                          216
       Exposure Duration. In comparisons of replicate studies of varying duration in which
O3 was determined to be the primary cause of variation in plant response,  greater reductions
in yield (reproduction) or growth occurred to plants exposed for the longer duration (Lee et
al., 1991; Olszyk et al.,  1993; Adaros et al., 1991).  Likewise, with respect to foliar injury,
Jacobson identified limiting values for crops and trees which showed that as  time of exposure
was extended, less O3 was needed to produce the same response (Jacobson, 1977).
Additionally, this relationship between duration of O3 exposure and plant response has also
been supported based  on statistical analyses, in which indices that accounted  for the length of
exposure were better able to rank vegetation effects than those which did not take the
duration of exposure into account (Lee et al., 1989). Lefohn (1988) and Lefohn et al.
(1988b) conclude that duration has value in explaining variation in plant response and that a
cumulative-type index is preferred over a mean or peak index based on statistical fit.  Thus.
an index that cumulates O3 exposures over the period of plant sensitivity is desirable.
       Despite differences in the period  of plant sensitivity across species, a constant
duration over which to cumulate exposure must be defined in order  to set a nationally
consistent standard. Under the current secondary NAAQS, the timeframe of  concern  (season
of highest O3 production) varies  from state to state and may consist of anywhere from 4 to
12 months.  Lee et al. (1989) analyzed air quality from 82  non-urban site-years of ambient
03 data across the U.S.  Exposure patterns for the 82  non-urban site-years have similar long-
term averages but differ widely in how the O3 concentrations are distributed  within the O3
season. When the same fixed consecutive three month measurement period during the O3
season was used for every area,  less than 63% of the concentrations of 0.06 ppm or  higher
were captured. By using instead a floating maximum consecutive three- or four-month time
period at each site within the O3 season, up to 73% and  83% of the peak concentrations were
captured,  respectively.
       Though for most agricultural species the growing season is only approximately 3
months, it has been suggested that trees  or other perennial species may require cumulation of
exposure over a longer growing  season (e.g., 5 months) in order to include the most relevant
exposures.  However,  given that the importance of a longer seasonal period  for trees has not
yet been adequately investit..ied, and the maximum consecutive three month period captures

-------
                                           217
 around 3/4ths of exposures to concentrations above 0.06 ppm, the staff judges that the
 maximum 3-month period is an appropriate surrogate for the entire O3 season.
        An additional aspect of duration involves consideration of the year-to-year
 meteorological variability which can produce widely varying O3 levels.  This aspect can be
-addressed by considering averaging seasonal exposures over multiple years.  Because annuals
 go through their entire life cycle within a period of one year, an appropriate secondary
 standard for annuals must be able to provide protection  on a year-by-year basis.  In addition,
 recent seedling research has shown that some perennials also show significant growth effects
 from O3 within an annual timeframe.  Likewise, foliar injury to sensitive trees in Class I
 areas  often results from short-term acute annual  exposures.  Additional analyses would  be
 needed to explore the impact of meteorological variability on seasonal exposure  indices
 before a multi-year average could be evaluated with respect  to its ability to provide year-by-
-year protection.
        Seasonal patterns of exposure.  The sensitivity of annual species to  O3 can  vary  within
 the  same growing season due to changes in phenology (plant developmental stage).  For
 NCLAN crops, Lee et al. (1989) tested various phenologically  weighted functions for their
-fit to NCLAN data.  Out of two exponential and 16 gamma phenological functions, the
 gamma  function which had the  highest weight  during the time period between 20-40 days
 before harvest resulted in the best  fit.  These  statistics  mirror the biological finding that O,
 can negatively effect different aspects of plant  reproduction  (Venne et al.,  1989; Feder,
 1986; Krause et al., 1975; Ernst et al., 1985;  Houston and Dochinger,  1977).  Increased
 sensitivity of reproductive processes has implications for yield outputs, genetic  success, and
 aesthetics (flower production) for ornamentals. This increase in reproductive sensitivity has
 not  been found in all tested cases,  indicating that phenology is species-dependent.  More
 information on  a  wider range of species would be needed before a phenologically-based index
 suitable for national standard setting could be established. Based on the available
-information,  the staff judges that the maximum consecutive three month time period within
 existing O3 seasons will likely include the sensitive  phenological stages in annuals in most
 cases.

-------
                                          218
       Exposure Concentrations.  As has been stated earlier, there is no threshold O3
concentration or seasonal exposure level above which effects occur and below which they do
not for all species.  Over the years, many studies have shown that, depending on the duration
of exposures and sensitivity of the plants, injury  to crops and other vegetation could occur
when exposed to O3 concentrations that ranged from 0.04 to 0.4 ppm, with higher
concentrations usually causing injury in the shortest period of time (CD, 1996).  This is due
to the known mode of action of O3 described in section VII-B that vegetation effects occur
when the amount of pollutant absorbed exceeds the ability of the plant to detoxify O3 or
repair the initial impact (Tingey and Taylor, 1982).  Because many  factors influence the
amount of O3 that is absorbed by a plant at any given time, it is impossible to state with
certainty that a given concentration will have a known impact on the plant  unless all other
factors have been accounted for.
       Over the last several  decades, research  has continued to advance the understanding of
the complexity of interaction  between O3  air quality and  exposure dynamics and timing  of
plant sensitivities.   Until the early 1980's, seasonal means were commonly used to
characterize the 03 exposure believed to be relevant to plant response.  However, Larsen  and
Heck (1984) mathematically showed that  it was possible for two air sampling sites with the
same daytime arithmetic mean O3 concentration to experience different estimated crop
reductions.  At about the same time, concerns  about using  the long-term average to
summarize O3 exposures began appearing in the literature,  specifically that peak
concentrations, which some believed might be  the most important in determining plant
response,  were not adequately accounted  for by a mean exposure index (CD, 1996).
       Since that time  several studies have attempted to  relate O3 exposure to plant response.
Unfortunately, no two  studies have exposed plants  in the same  manner or under similar
conditions so that the data from each study is unique.  Exposure methods,  concentrations  and
durations  used, age of  plants at exposure,  length of exposure, the plants exposed and the
media in which they were grown all differ across experiments. Some exposures were in
chambers  in the greenhouse,  others in open-top chambers and others in the ambient air.
Though the  results in general have been inconclusive, two  different viewpoints have
emerged.

-------
                                          219
       One set of studies, including Musselman et al., (1983, 1986, 1994) and Hogsett et al.
(1985b) find evidence to support the view that peak concentrations are important in defining
an exposure index.  Musselman et al.,  1983 and Hogsett et al. (19855) were among the first
to demonstrate that variable concentrations produced greater effect on plant growth than fixed
or set diurnal patterns of exposure of equal total exposure with lower peak concentrations.
Subsequently, Musselman et al. (1986) and Musselman et  al. (1994) found that patterns with
higher peak concentrations or longer duration of higher concentrations produced significantly
greater effect on top dry weight (kidney bean) than the square wave pattern.  These results
provide evidence that 1) total exposure (i.e.  SUMOO), being unable to differentiate among the
exposure patterns,  is a poor predictor of plant response; 2) the peak concentrations or
sequence of peak concentrations (> 0.16 ppm) are important in determining plant  response;
and 3) greater weight should be given to higher concentrations when describing exposure.
Though the Hogsett et al., 1985b  study contained  both  "mid-range" and "peak"
concentrations, Musselman, et al. (1983, 1986, and 1994), used exposure levels containing
peaks that greatly exceed those in any of the other exposure studies or in ambient air,
possibly influencing their findings on the importance of peaks in  determining  plant response.
       The above findings were further investigated for tree seedlings.  Hogsett and Tingey
(1990) exposed ponderosa pine and aspen  seedlings  to three different exposure regimes and
observed greater growth reductions in the episodic exposure pattern, which had the largest
peak concentrations of the three patterns.  The smallest growth reductions in both species
were observed  with the more constant high elevation pattern that had peak concentrations
below  0.10 ppm. The authors concluded that  temporal pattern and concentration were
important in influencing long-term growth response  of tree seedlings, just as in crops, and,
consequently,  should be considered in measures of exposure.
       Other publications that have been cited as evidence supporting the importance of high
concentrations in eliciting plant response are a series of retrospective studies reporting
regression results using data from the NCLAN program (Lee et al., 1987, 1988; Lefohn et
al., 1988a; Tingey et al.,  1989).  These studies were in general agreement and consistently
favored the use of cumulative peak-weighted exposure indices.  However. Lefohn and Foley

-------
                                          220
(1992) note that the high number of hourly O3 concentrations above 0.10 ppm in the NCLAN
protocol may prevent generalization of these findings to other types of exposure regimes.
       On  the other hand, Tonneijck and Bugter (1991), Tonneijck (1994), and  Krupa et al.,
(1993, 1994, 1995) stress the view that mid-level (0.05 to 0.09 ppm) concentrations are as
equally important as higher hourly concentrations in affecting vegetation.   Krupa (1995)
states that in their analyses of European data, "concentrations >  90 ppb (0.09 ppm) appeared
to be of little importance because such concentrations in general appear to occur during
atmospheric conditions which did not facilitate ... uptake."  Most of the studies listed  above
base their conclusions on the foliar injury response of Bel  W-3 tobacco, which has been
noted in numerous earlier studies to have an inconsistent relationship with O3 air quality.
The authors Tonneijck and Bugter (1991) and Tonneijck (1994) also acknowledge this
deficiency  and  state that "foliar injury on tobacco Bel W3  was poorly related to the ambient
O3 in the Netherlands." Further complicating the interpretation of their results, the papers do
not cite the actual O3 concentrations to which the plants were exposed,  except as mean
values. The 1996 CD concludes that based on the known  inconsistency of Bel W3, the
conclusions presented by Krupa et al., 1995 needed to be verified.
       As  stated earlier under section VII-B-1, stomatal conductance is influential in
determining the dose  received by the plant.  The information presented above suggests  that
the range of concentrations of concern for any particular plant can change depending on the
combination of other  stresses acting on the plant, the phenological stage of the plant, the
duration of the exposure and stomatal conductance.
       Diurnal Patterns of Exposure.  Most plants are believed to open their stomata during
the day and close them at night.  It has been reported,  however,  that some plants have
stomata that stay open well after the sun  has set.  For example, it has been reported (U.S.
EPA,  1978) that white pine keeps its  stomata open all night, while Tobiessen (1982) reported
that many  early successional trees, such as black cherry, big-toothed aspen and  white ash
open their stomata at or before dawn, whereas late successional species tend not to show
such a pattern.  This  suggests that if environmental conditions favor stomatal opening,  and
O3 levels are elevated, plants may take up O3 in the predawn hours (Neufeld et  al.. 1992).

-------
                                           221
        Air quality information has already shown that diurnal patterns in rural sites may vary
  significantly  from the typical urban pattern of an early afternoon peak, due to long-range
  transport or elevational effects.  Figure VII-5 shows a range of diurnal patterns of Oj
  concentrations for U.S. rural sites.  Long-range transport processes bringing O3 from more
  distant urban areas can change the timing of rural O3 peaks from afternoon to evening to
  early morning hours.  Furthermore, at mountainous sites which are above the nocturnal
  inversion layer, more constant levels occur because there is little or no nighttime scavaging
  of O3 by nitric oxide.  Urban areas, on the other hand, typically experience a wide range of
  O3 concentrations that build to a peak in the daytime and then fall to near negligible levels at
  night from scavenging. On reviewing the AIRS database for 1990 to 1992 for those
  agricultural and forested site types experiencing a 3-month, 24 hour SUM06 greater than or
* equal to 26.4 ppm-hours in  the U.S., it was  found that 70% of the sites experienced at least
  50% of the occurrences of >_ 0.10 ppm O3 peaks during the 7 hour period 9 am to 4 pm
  (CD, 1996).
        The published literature shows that the most major effects associated with O3 exposure
..are linked to the disruption of the photosynthetic process.  Since photosynthesis occurs only
  when certain minimum light conditions are met, the staff believe that the diurnal time period
  of greatest concern for vegetation, therefore,  can be defined as  the daylight hours.   Using  a
  daily exposure period longer than the traditional 7  hour  window of 9 am to 4 pm is
  supported by a study done by Heagle et al. (1987) which compared the effect of increasing
  the time frame of exposure  from 7 to 12 h/day.  They found that plants receiving exposures
  for an additional 5 h/day (i.e,  12 h) showed  10%  greater yield loss  than those exposed for  7
  h/day.
        Though the length of daylight hours can range from close to 12 hours to as much as
  16 hours in different parts of the U.S. during the  summertime,  maximum O3 months, the
  staff feel that the percent of concentrations occurring outside the 12 hour daylight window
- are sufficiently small and outside the period of greatest sensitivity for most studied
  vegetation, as to make the 12 hour daylight (8 am  to 8 pm) an appropriate timeframe over
  which to cumulate exposure.

-------
                                222
  Figure VII-5.  Diurnal Ozone Patterns  at Rural Sites
               50

               40

               30

               20
(a)
- AZ
r /"

i i
' MT

s"' 	 __msw^,
.-"'" \>R ;
i i i
                       4      8      12     16     20
                              Time of Day (h)
                Diurnal  behavior of ozone at rural sites in  the
                United States in July.  Sites are identified by
                the state in which they are located.   (a) Western
                National Air Pollution Background Network  (NAPBN);
                (b)  Whiteface Mountain (WFM) located at 1.5  km
                above sea level; (c) eastern NAPBN sites; and  (d)
                site  selected from the Electric Power Research
                Institute's Sulfate Regional Air Quality study.
                IN (R) refers to Rockport.
Source:  Logan  (1989)

-------
                                           223
        Intra-episodic Patterns of Exposure.  Other factors, including a predisposition effect
 of early O3 exposures on plant sensitivity to later  O3 episodes (McCool et al.,  1988) can
 contribute to variations in biological response.  Consideration of this factor suggests the need
 for weighing O3 exposures to account for predisposition time.  However, the role of
 predisposition in influencing plant response varies with species and environmental conditions
-and is not understood well enough to allow specification of a weighing function for use in
 characterizing plant exposure.
        Summary.  In summary, in spite of the complexity inherent in vegetative systems,
 research developed over the last several decades and most recently since  1988, has produced
 information  on exposure dynamics and their role in producing plant response.   Data in
 published literature still supports the conclusion that cumulative seasonal  exposure and higher
-concentrations are important features of exposure  for both crops and trees.   Ideally, an
 exposure index would account for all of the variation in vegetation effects that are associated
 with exposure to O3.  A second, more practical objective is the  specification of an exposure
 index that is applicable in the ambient air quality  standard setting process.  An exposure
-index for a NAAQS  should be easy to develop and  applicable to a wide range  of species and
 environmental/exposure conditions.  The attainment of these criteria, however, necessarily
 represents a compromise  in the features included  in the formulation  of the best exposure
 index (Lee,  et al., 1989).  The section  that follows  highlights those types of forms which
 meet these criteria.
        2.      Alternative Forms of the Secondary NAAQS
        The discussion of exposure indices presented in the CD groups indices  into several
 generalized forms  including one event, mean, cumulative, concentration weighing, and
 multicomponent.  These general forms  are discussed below with regard to their biological
 relevance and their suitability as a basis for standard setting.
        One Event, Mean, and  Cumulative Forms. The one event form, which includes the
 current form of the NAAQS, measures only one or a very limited number of peak events out
 of a plant's entire growing season.  If O3 concentrations never achieve the level of the peak
 of concern, the assumption is that no growth or yield effects of concern are occurring. It
 also does not distinguish among exposures of different  durations.  This index form does not

-------
                                          224
account for many exposures and patterns of exposure that are associated with vegetation
effects. The mean form, such as the seasonal mean of the 7-hr daily means, averages all
selected concentrations equally and does not address the varying patterns of exposure.
Larsen and Heck (1984) demonstrated that for two air sampling sites with the same daytime
arithmetic mean O3 concentration very different estimated crop reductions could occur.  The
cumulative form sums all hourly concentrations across a season (SUMOO) and,  thus, contains
an exposure duration component, but still weights all concentrations equally. Using air
quality data,  Lefohn et al., 1989 showed that the magnitude of the SUMOO exposure index
was largely determined by the lower hourly average concentrations.  In a similar  study,
Lefohn et al. (1992) noted that the magnitude of the SUMOO index did not adequately
account for the occurrence of the higher hourly average concentrations in the ambient
treatments.  The coupling  of the air quality considerations as described by Lefohn et al.
(1989, 1992), with the biological findings reported by Mussel man et al. (1983, 1994) and
Hogsett et al. (1985b), builds a consistent picture that the SUMOO index does not adequately
account for the occurrence of peak hourly concentrations.  The one event, mean, and
cumulative forms do not take into account the many features of O3 exposure regimes that
influence plant response and are, therefore, limited in their usefulness as predictors of O,
injury.
       Multicomponent Forms. In contrast to  the relatively simple forms discussed  above,
multicompartment forms have been developed which attempt to take into account  a range of
factors that have been associated with vegetation effects.  For example, Lee et  al. (1989)
analyzed crop yield data for 17 individual NCLAN studies, using regression analysis to
evaluate a total of 614 indices, including 589 variations of the Generalized Phenological
Weighted Cumulative Index (GPWCI) using the Weibull model.  The exposure indices were
evaluated on the basis of statistical fit along three different criteria and an average score
given.  Though no single exposure index performed best for  all  17 cases, there was a group
of indices that were always at or near the top ranking scores. The  100 top performing
indices were all GPWCI indices, with overall best index being the GPWCI index  PWCI485.
with a sigmoid weight centered at 0.062 ppm and a phenological  gamma weighting scheme
placing maximum weight 30 days prior to maturity.  Although such multicomponent indices

-------
                                           225
 may well take into account many relevant factors, including plant phenology, predisposition
 effects of early O3 exposures to later ones, and relative impacts of different concentration
 ranges, such forms have not been advanced as being applicable for standard setting due to
 being species-specific and highly complex.
        Concentration-Weighted Forms.  In the same analysis by Lee et al. (1989), 25 general
 index forms were also evaluated.  Several threshold (SUM06, SUM07, SUMO8, AOTO8)
 and sigmoidally weighted cumulative indices were nearly as optimal in fitting the NCLAN
 database as was the multicomponant GPWCI index. The sigmoidal weighting function that
 performed the best increases monotonically from 0 to 1 and assigns a weight of 0.5 at 0.062
 ppm (designated SIGMOID in  later documents).   In a separate study,  Lefohn et al. (1988)
 compared two sigmoidal functions, W126 (which has an inflection point at 0.067 ppm and
 gives equal weight to values above 0.10 ppm) and W95 (which gives greater weight to values
 above 0.10 ppm).  Though the W126 performed better than the W95, it has not been  directly
 compared to the SIGMOID to determine if one better represents plant response. The W126
"form does,  however, give less  weight to the lowest range of concentrations (i.e., those that
-fall within typical background levels) than the SIGMOID.
        Several threshold cumulative forms also attempt to take into account evidence that
 peaks produce a disproportionate response relative to  lower concentrations by selecting a
 "threshold" value below which the O3 concentrations are not counted.  These forms assigned
 a weight of 0 to concentrations below the "threshold" and concentrations above the
 "threshold" are assigned a weight of  1.  The hourly concentrations which  fall above the
 "threshold" level (i.e., the SUMXX forms) or the difference between  the  concentration and
 "threshold" levels (i.e., the AOTXX forms) are then added together to give a cumulative
 seasonal total exposure.  The establishment of a "threshold" value  is somewhat arbitrary  and
 is not based on any evidence of a discernible threshold for  vegetation  effects in general.  The
 "threshold" levels identified in  previous research  have sometimes been set to factor in other
 considerations,  such as the level of background concentrations or the ability of models to
 predict air quality concentrations below certain concentrations.
        Staff notes that the results of the retrospective  analyses of exposure indices have
 received several critiques which point out the artificiality of the O, exposure regimes used in

-------
                                          226
the NCLAN studies to elicit vegetation responses (Lefohn and Foley,  1992).  The concern is
that the use of an exposure regime which contains unrealistically large numbers of high peaks
may have exaggerated the vegetation response.  Thus, it is not clear at this time whether the
exposure indices identified in Lee et al. (1989) as the best predictors of plant response in the
NCLAN studies would have been ranked in the same order under a different exposure
profile.  Others have critiqued the analyses because they do not directly address biological
factors, but rather reflect only statistical associations.
       Additionally, this same NCLAN database has been used to calculate levels of O3
exposure for different indices that would provide "equivalent" protection for crops from
certain defined  yield loss percentages.  Lefohn and Foley (1992) suggest it is important that
any exposure index that sets a level of protection based  on the response of plants  in the
NCLAN experiments  should recognize the peakiness of the exposure regimes used when
attempting to predict biological responses over the range of ambient O3 exposure regimes.
       Though  variations of three selected exposure indices (SUM06, AOT06, and W126)
being considered in this review performed equally well  in statistically predicting plant
response from the NCLAN data  in Lee, et al. (1989) which had high  numbers of peak
concentrations, questions have been raised as to how these indices would compare under
different exposure scenarios.  In  attempting to address these questions, staff compared how
these selected exposure indices differentially weight peak, mid-range and low O3
concentrations for various ambient air quality scenarios  from several NCLAN studies and air
quality distributions produced from the AIRS database for a variety of selected monitored
locations (Appendix F).  These comparisons were done  on the'basis of the percentage of the
total value of the index contributed by selected portions of the range of Q, concentrations.
For example, using the O3 ambient air quality distribution for one NCLAN study  (R85CO)
conducted  in Raleigh  with cotton in 1985 to compare SUM06, AOT06, and W126,  the
following percentages were observed:
                   < 60        60-80         80-100    >  iOQ (ppb)
       SUM06       0            60           30          10
       AOT06       0            38           42          20
       W126         15           42           31          12

-------
                                           227
       From such comparisons staff observed that for any given representative ambient air
quality distribution, the percentage of the total index value contributed by the upper mid-
range and peak concentrations was highest for the AOT06 index, indicating that AOT06
effectively weighted the upper mid-range and peak concentrations relatively more than did
either the W126 or the  SUM06 indices.  Furthermore, unlike the AOT06 and SUM06 indices
which  do not include concentrations  below 0.06 ppm, low concentrations (i.e., <0.06 ppm)
can account for a significant percent of the total for the W126 index while concentrations
between  0.06 and 0.08  ppm tend to represent the greatest percentage of the total value in the
SUM06.
F-     Considerations in Characterizing Adverse Welfare Effects
       Though exposure-response functions of O3  on vegetation have been  developed  from
studies described in the sections above, several additional pieces of information (e.g.,
national O3 exposure patterns; location of O3  sensitive species) are added in this  section to
put this information into a national context that can more fully inform the Administrator as to
the need  for additional protection for vegetation  against 03-induced adverse welfare effects
and to better characterize  the residual risks and benefits associated with various policy
decisions.
       As part of this effort, the  staff took into account the possibility of a new primary
standard  consistent with the options presented in section VI.  Recognizing that attainment of
any of these alternatives would generally lower O3 exposures nationally,  the staff performed
several comparisons using both monitored and projected air quality to evaluate the impact  of
attainment of various primary options on O3 concentrations of concern to vegetation.
Benefit and risk  reductions associated with reductions in O3 exposure from  the attainment of
a separate secondary standard are therefore considered incremental to those achieved by
attaining any of the primary standard options.
       1.    Exposure Characterization
       National Monitoring  Network.  Sparse air quality monitoring has  always constrained
the characterization of national rural  and remote air quality.   The first national network of
air monitoring stations designed to measure levels  of O. in remote areas (100 or more miles
from any major urban area)  within the 48 contiguous states was established in  1976 and ran

-------
                                          228
through 1983 as a cooperative effort between the U.S. EPA and the U.S. Forest Service.  By
1980, this National Air Pollution Background Network (NAPBN) consisted of eight sites
operating in selected National Forests, although only six of these sites operated  year-round
(Evans et al., 1983; Evans, 1985). As of 1987, less than  2% of the O3 monitoring sites in
the AIRS database, were classified as forested.   The remaining  sites were divided between
residential, commercial or industrial, 79%, agricultural, 17%, and desert or mobile, 3%
(U.S.EPA, 1987).
       Though  the rural monitoring network has grown to now  include approximately 80
monitors in Class I areas, the majority of AIRS sites are still located in urban or near urban
areas.  Even many of the monitors classified as rural occur within cities or Census
Metropolitan Statistical Areas (CMSAs), and often show O3 air  quality patterns  typical of
urban areas (e.g., low nighttime O3 due to scavenging (Stasiuk  and Coffey, 1974). with high
diurnal peaks, often including occurrences of hourly average concentrations above 0.10
ppm).  The 1991 monitoring network for both urban and rural U.S. monitors still show large
sections of the country with little or no monitor coverage  (Figures VII-6a and 6b). Some of
these non-monitored areas are significant for a  variety of crops  such as wheat, barley, corn,
sorghum,  soybean and kidney bean production, as well as for the tree species black cherry,
sugar maple, red maple, aspen,  red alder, white pine, Douglas  fir, and ponderosa pine.
       Using 1991 to 1993 AIRS  monitoring data, the staff examined the question whether
attaining a proposed  primary standard  could provide sufficient protection to vegetation with
respect to cumulative seasonal exposures that have been shown  to injure plants. The specific
example chosen for comparison was the primary option (0.08 ppm, 8  hr.,  I or 5 expected
exceedences) and the secondary option (SUM06, 12 hr., 25 to 38 ppm-hrs.).  Staff examined
the 8-hour daily maximum and  12-hour SUM06 design values for 581 counties  (those having
sufficient  monitoring data in AIRS for the period 1991-1993).   Figures VII-7a and b show
the associations between  the 8-hour values and  the SUM06 values. These  figures show, for
example,  that almost all areas that are within or above a SUM06 range of 25-38 ppm-hr
would also fall above an 8-hour design value of 0.08 ppm.  Thus, those monitored sites that
would likely be of most concern for effects on  vegetation  would also be addressed by an 8-
hour primary standard set a. a 0.08 ppm level.

-------
                             230
Figure VII-6b.   1991 Rural Ozone Monitoring Site Locations
   1991   Rural  Ozone  Monitoring   Site  Locations
                                             ° Cities over  50,000
                                             • Sites  classified  as
                                               forest, agriculture,
                                                 barren, mobile
  Shaded  counties  are Bureau  of Census Metropolitan Statistical  Areas
   d teloc/iheaj -w»l-r .an)

-------
                                229


  Figure VII-6a.   1991 Rural Ozone Monitoroing Site  Locations
     1991  Rural  Ozone   Monitoring   Site   Locations
                                                ° Cities over 50,000
                                                • Sites  classified as
                                                   forest, agriculture,
                                                    barren,  mobile
    Shaded counties are Bureau  of Census Metropolitan Statistical Areas
//(total 4dy/im»ctoleloc/titdoc-r.cml

-------
                Figure VII-7a.  1991-1993 Air Quality Relationships
8 HR, 0.08 ppm (1 exex)
 0.2
                      County Design Values
0.15



 0.1

 0.08

0.05



   0
n B nnFXB i
                   P8  59 H





                     05  n
    0
     20 25  38 40       60       80
            12 HR SUM06 (ppm-hrs)
100
                                                            Counties
                                                                        KJ
                                                                        CO
120

-------
               Figure VII-7b.  1991-1993 Air Quality Relationships
8 HR, 0.08 ppm (5 exex)
                     County Design Values
u.^
0.15
0.1
0.08
0.05
n

— B (3

anm
g
I

S ffitTOR
BEED raw;

irasn a
	 i 	

n
a a
ttira n R is n ra n
ZB QSVB HWKMHQB EHfU ffl E9 B B)

1,1,1.1,
Counties
H

    0
2025   3840       60      80
       12HRSUM06(ppm-hrs)
100
                                                                     CA)
                                                                     N)
120

-------
                                           233
        However, given that the greatest proportion of vegetation in the U.S.  is found
 growing outside of urban centers where there are few or no monitors and where different
 environmental and elevational factors interact with O3 precursors or O3 transported into the
 site, there is considerable uncertainty as to the  exact nature and  strength of the relationship
 between urban 03 air quality and distributions that occur in rural or remote areas.
        In recent years, more and more researchers have undertaken studies to better
 characterize O3 air quality in rural, remote or "clean" sites, (Lefohn and Jones, 1986; Lefohn
 and Foley, 1992; Logan, 1989; Bohm, 1992).   Though often extremely limited in scope,
 these studies continue to refine and build on  the work by Evans  et al., 1983 and Evans, 1985
 and can generally be divided between  western and eastern U.S.
        Western Air Quality.  Logan (1989)  corroborated an  earlier finding from the three
 western sites (Apache, AZ; Ochoco, OR; Custer,  MT)  in the original  NAPBN that G, hourly
.^average concentrations above 0.08 ppm rarely are exceeded at remote  western sites except in
 some areas in California.  For example, outside of California, even near urban sites, O,
.-.concentrations remained low, with growing season  means ranging from 0.012 to 0.022 ppm,
.0.028 to 0.037 ppm, and 0.032 to 0.058 ppm in Washington, Utah and Colorado,
 respectively.  Not unexpectedly, there is little evidence  of O3 injury at these sites.  On the
 other hand, Bohm (1992) reports that  Yosemite and Sequoia  National Parks,  which receive
 pollutants  transported from highly urbanized  areas, had  24 h  means ranging from 0.036 to
 0.085 ppm on 75%  of summer days, whereas Lake Gregory  had a growing season mean of
 0.073 ppm.  During 49% of the summer days,  means of diurnal patterns ranged from 0.085
 to 0.100 ppm, decreasing with altitude and distance from the source (Bohm,  1992).  These
 levels, as mentioned in the earlier effects  section,  have  been  associated with growth
 decreases and  foliar injury in some species.
        An ongoing study,  the Sierra Cooperative Ozone Impact  Assessment Study (SCOIAS),
 co-operated by the U.S. Forest Service, California Air  Resources Board, and the University
 of California at Davis, is documenting exposure of sensitive pine species in the Sierran
 forests to O3 and the amount of injury the trees exhibit.  As part of this project, six sites,
 ranging in elevation between 3550 and 6000  feet above  mean sea level, were selected to

-------
                                          234
monitor O3 concentrations and meteorological conditions (temperature, humidity, wind speed,
wind direction and solar radiation).
       After completing 3 to 4 years of data collection (1990, 1991 to 1993) the researchers
found that O3 concentrations were typically highest in the afternoon, and increased as one
moved toward the southern end of the network,  which also was in the direction of increasing
elevation.  Stations located on well defined steep slopes show a very strong diurnal variation
in O3 concentrations  and meteorological conditions.  Air quality distributions of hourly data
from 1992  for these  sites are shown in Figure VII-8. Hourly peak O3 concentrations from
June through September were greater than 0.06 ppm at all sites nearly every day, in excess
of 0.08 ppm at most sites more than half the days  and in excess  of 0.10 ppm  at least a few
days a month.  The most impacted sites, Mountain Home (6000  ft.) and Shaver Lake (5650
ft.) were the southernmost units and  had concentrations  above 0.10 ppm for nearly  half of
the days during the monitoring period.  At the two sites in  the middle of the network
(Jerseydale (3750 ft.) and Five-Mile  Learning  Center (4000 ft.)) the diurnal variations in O3
were not very well pronounced and nighttime values remained relatively high.  At White
Cloud  (4350 ft.), the northernmost unit in the  study, the highest  concentrations occurred at
night,  when winds were from the NNE (Sierra Cooperative  Ozone Impact Assessment Study.
1993).  At the time of this interim report, foliar injury information at these sites is  still
incompletely characterized.  However, these O3  exposures are well above those
concentrations associated with vegetation injury.  According to the  report,  "available
literature indicates that  needle injury occurs from exposure to O3 concentrations of 0.06 ppm,
and is  significant at and above 0.08 ppm .... The recorded  data suggest that serious to
severe exposure (>  0.08 ppm) of pines to ozone is likely"  (Sierra Cooperative Ozone Impact
Assessment Study, 1993).
       Eastern Monitored Air Quality.   Questions have been raised as  to whether O, air
quality distributions  experienced at eastern sites  are representative of sites in the western
United States.  Differences in biogenic precursors and more gradual changes in elevation that
never  reach the high elevations found in the west complicate efforts to define an O3 air
quality pattern that consistently applies to all high  elevation  sites  nationally.  In eastern
locations, the relatively flat air quality patterns considered typical of a high elevation site

-------
                                235
Figure VII-8.
Ozone Frequency Distributions of Hourly O3
Concentrations from June to September 1992
Mountain Home
25 Elev. 6000 ft.

i. 20
t 15 |.-
iio ||l||
1: JUL
o o o o o o o o
ppm

Jerseydale
Elev. 3750 ft.
£25
€0
t 15
o
£ 10 i
1 5 1
,o ! ||
^ n - »,»•.«,



,.
^— O1} tQ f*» O5 T~ fO UO
OOOOOT--*-*-
O 0 O O O O O O
(
pprr. .

i
Sly Park
Elev. 3550 ft.
25 !
en I
.2 20 -

! 20
% of observat
— k _v
o in o th

..I

I,

OOOOC3OCDO
ppm

-------
                                          236
(including an area being above the nocturnal inversion layer and being affected by long
distance or regional transport) occur at different heights.
       In a study by Winner et al. (1989) an effort was made to relate O3 exposure patterns
to elevation.  Several sites were monitored in western Virginia from May to December 1982,
ranging in elevation from 457 m to 1067 m.  In general,  the high elevation site, Big
Meadows, in the Shenandoah National Park, had higher monthly O3 concentrations than the
lower elevation sites.  However, the number of peak O3 occurrences (> 0.10 pprn) did not
necessarily increase with altitude.  Instead, higher monthly averages seem to be associated
more with a lack of nighttime scavenging than with a large number of peak hourly
concentrations.  Additional comparisons with the following years  1983-1985 showed that  the
pattern was consistent across years.
       Lefohn et al.  (1994)  compared several sites located  in West Virginia, Virginia, and
Pennsylvania for the years 1988 through 1992 in terms of the seasonal (April to October) 24
h W126 exposure index, as  shown in  Figure VII-9.  Of the 11 sites with data for all 5  years,
the 6 sites with the highest exposures  were also the higher  elevation sites (>  500 m),  while
those sites with the lower exposures were all  below 500 m in elevation.  The highest
elevation sites were also observed to have a large number of O3 episodes, with the number of
hourly peaks _>. 0.10 ppm ranging from  only a few in  1992 to over 100 in  1988.
       Though these  values  were collected over a 7 month period, in 1988  all 11  sites
exceeded the 3 month W126 level (21.0  ppm-hr) identified by Hogsett et al. (1995) as
protecting 50% of the tree seedling cases studied from greater than 10% biomass loss,  while
only two sites were below this level in 1991.  In the other years,  except for 1992, more  than
half the  sites exceeded this level (Figure VII-9). Such patterns are consistent with results of
recent monitoring in the Great Smokies National Park described by Neufeld et al. (1992).
However, with only 2 permanent and 1 seasonal monitoring stations, O3 levels are not  well
characterized over much of the park.
       Though O3 concentrations in the east may not reach high concentrations as frequently
as in the west, the above studies indicate that O3 air quality.in some forested areas in the east
contain both peak and cumulative, seasonal exposures known to have caused both foliar
injury and growth reduction  in some vegetation.

-------
                                                       237
Figure  VII-9.
The Comparative  Ranking  of  Ozone Monitoring  Sites
From  April to  October  Using the  24  hr W126
Exposure Index  (1988-1992)
        *Sites  with  less  than  5 years  of  data
                Morton Station, VA
  Shenandoah N.P. (Dickey Ridge), VA
               Greenbrier Co., WV
                 « Smyth Co., VA
  Shenandoah Hf. (Big Meadows), VA
                   Laurel Hill, PA
   Shenandoah N.P. (Sawmill  Run), VA
                     Panom, WV
                 Cedar Creek, WV
                     Vienna, WV
                    » Marion, VA
                   Wheeling. WV
                     Weirton, WV
                                          100
                                                ISO
                              W126 Exposure Index (ppm-h)
Horton Station, VA-
Shenandoah N.P. (Dickey Ridge), VA-
Laurel Hill, PA-
Parsons, WV-
Shenandoah N.P. (Big Meadows), VA-
Vienna, WV-
» Marion, VA-
Greenbrier Co., WV-
Cedar Creek, WV-
Wheelinfl, WV-
Shenandoah N.P. (Sawmill Run), VA-
Weirton, WV-

•••••70.6
••••44.5
•••29.8
•••29.4
••129.3
••27.8
•124.7
•123. i
• 21.9
• 20.0
• u s
• 14 3












                                                                                              50
                                                                                                    100
                                                                                                           150
                                                                W126 Exposure Index (ppm-h)
        The comparative ranking of 03 monitoring sites in April to October
1988 using the W126 exposure index
                                          The comparative ranking of 03 monitoring sites in April to October
                                  1989 using the W126 exposure index
                Horton Station, VA
  Shenandoah Hf. (Big Meadows), VA
  Shenandoah N.P. (Dickey Ridge), VA
                 » Wythe Co., VA
                   Laurel Hill, PA
                    Parsons, WV
                     Vienna, WV
   Shenandoah N.P. (Sawmill Run), VA
               Greenbrier Co., WV
                 Cedar Creek, WV
                   Wheeling, WV
                     Weirton, WV
                •> Cumberland, MD-Jl
                                          100
                                                150
                              W126 Exposure Index (ppm-h)
                                                  Horton Station, VA
                                    Shenandoah N.P. (Big Meadows), VA
                                                 Greenbrier Co., WV
                                      SJieaandoih N.P. (Dickey Ridge), WV
                                                     Laurel Hill, PA
                                                      Parsons, WV
                                                       Vienna, WV
                                                   Cedar Creek, WV
                                       StieiuiHloah N.P.  (Sawmill  Run), VA
                                                     Wheeling, WV
                                                   «> Wythe Co., VA
                                                       Weirton. WV
                                                 *  Cumberland, MD
                                                                                                    100
                                                                                                           150
                                                                 W126 Exposure Index (ppm-h)
        The comparative ranking of 03 monitoring sites in April to October
 1990 using the W126 exposure index.
                                          The comparative ranking of 03 monitoring sites in April to October
                                   1991 using the W126 exposure index.

Shenandoah N.P.
»
Shenandoah N.P



Horton Station, VA-
(Blg Meadows), VA-
Bearden Knob. WV-
(Oickey Ridge), VA-
Greenbrier Co., WV-
* Wythe Co., VA-
Laurel Hill, PA-
Shendandoah N.P. (Sawmill Run), VA-







Parsons, WV-
Wheeling, WV-
Cedar Creek, WV-
Vienna, WV-
Weirton, WV-

•••• 45 <
•••••••V 4S 0
•BBVBBl 33 '
••i 26.8
•• 22.C
•• 21 C
••V 19. 5
•• 16 2
•1 17 4
•i 16 6
•Vise
• 14 ;
• 11.5
0 50 100
W126 Errosu'e Inflex













160
(ppm-ri)
                               The comparative ranking of 63 monitoring sites in April to October
                       1992 using the W126 exposure index

-------
                                           238
       Uncertainties in Air Quality Relationships.  As has been shown earlier in this
document (Figure VII-5), diurnal patterns of O^ can differ significantly between urban and
rural areas.  The staffs evaluation of monitoring data indicates that there is considerable
overlap between areas exceeding a 0.08 ppm, 8 hr. standard and those exceeding the 25 to
38 ppm-hr. range of concern for vegetation,  suggesting that improvements in national air
quality from attaining an 8-hour primary standard within the recommended range of levels
would  also reduce levels below those of concern for vegetation in  those same areas.
However, there remains  concern whether urban control strategies would result in attainment
of either 8-hr or cumulative standards in downwind rural areas.  Therefore, in order to
develop a more complete understanding of rural and remote O-^ air quality and its
relationship to attainment of various primary options, staff examined various techniques and
methods that  have been or are used to produce spatial estimations  of national O^ exposure.
       Spatial Estimation of Ozone Exposure.  At the time of the  NCLAN research project
discussed in the previous section, national air quality typical of agricultural crop growing
areas was unknown.  To estimate Ch exposure over non-monitored areas across the country,
NCLAN used the spatial interpolation technique of kriging. Kriging, like all spatial
interpolation  techniques, is based on the common observation  that, on average, points close
together in space are  more likely to have similar values of a property than points further
away.   As such, spatial interpolators generate a generally smooth gradient of values from one
monitored site  to another.  Three key issues  affect the reliability and appropriateness of using
spatial interpolation to estimate O^ exposure values at non-monitored locations.  These are:
1) the  spatial representativeness of the monitored sites;  2) the-sampling density of the
monitored sites; and 3) the spatial variation of O^.  Using spatial  interpolation to estimate O-^
exposure at non-monitored locations is appropriate only if the monitored sites are
representative of all possible locations to which an estimate is to be made and if the
monitored sites occur at a sufficient sampling density to adequately represent the
non-monitored areas and capture the spatial variation in O-^ exposure in these non-monitored
areas.   That is, the monitored sites must be able to capture the frequency and amplitude of
exposure "peaks and  valleys".  The existing  monitoring network is inadequate to make
estimates of O^ exposure owr large sections of the country (see section VII-F.l).

-------
                                          239
    More recently other techniques have been used to estimate O3 exposure over
non-monitored areas.  These include computer simulation or modeling approaches such as the
regional oxidant model (ROM).  Such modeling approaches predict O3 air quality by
simulating the many chemical and physical factors and relationships that influence O3
formation, transport and attenuation.  Obtaining the input data to run  such  models and
identifying the mathematical relationships between all the many variables is an expensive and
time consuming process.  Modeling the spatial variability of O3 air quality  from these
relationships can easily become a task on the same magnitude as  modeling  weather and
climate-change, requiring massive computational power.
    The NHEERL-WED is using a geographical information system (GIS)  as a tool to
develop an estimation technique that has characteristics of both interpolation and modeling
(Hogsett et ah, 1995).  GIS is a formalized computer tool that allows one to integrate and
manipulate spatial data.  A GIS has characteristics similar to other computer programs such
as computerized mapping, database managers,  and spread sheets  and, in fact, GIS has
borrowed heavily from these technologies.
    The objective of the GIS-based technique is to improve the estimation of O3 air quality in
non-monitored areas by using information on factors that influence O3 formation and
transport. These factors include sources of O3-forming precursors, wind direction,
temperature, cloud cover,  elevation, and distance from emissions sources.  The factors are
used to identify the expected trends and patterns  of O3 air quality one could logically expect
to observe between  the sparse and distant monitored sites. The assumption is that areas
experiencing a great number of days with elevated temperatures and low cloud cover and  that
are down-wind of sites having large emissions of Oj-forming precursors will have a greater
potential for experiencing elevated O3 concentrations than areas not situated in such high
potential locations.
    To make estimations of O3 air quality in non-monitored areas the GIS is used to create
what is referred to as a potential exposure surface (PES) using information about  the
influencing factors.  The form of the PES is a 10 km grid superimposed across the country.
Each 10  km cell of the PES receives a value representing the  sum of all factors influencing
O3 formation and dispersal.  Hence cells in  locations with high temperatures and generally

-------
                                           240
cloud-free skies and down-wind from large amounts of O^-forming precursors will have high
values.  Once the PES is created, the relationship between the PES values and monitored Ch
values at each monitored site are used to calibrate the PES so that it can be used as a
surrogate for O^ air quality in non-monitored areas. Rather than assuming a smoothly
changing surface between the monitored sites as traditional interpolation techniques do, the
CIS-based approach predicts more structure between monitored sites based on factors  known
to influence 0*3 air quality.  Although this integrated CIS-based approach is not as complex
or sophisticated as a true computer model it has the advantage of using data that is readily
available across the entire country, is relatively inexpensive to run, and allows one to  quickly
produce estimates  of any exposure index for multiple months or years.
       Uncertainties Associated with GIS-Based Ozone Air Quality Estimation Technique.
As with any  spatial interpolation technique that must rely on sparse data representative of
urban or near-urban areas uncertainty  is great and non-quantifiable.  Intuitively  the estimates
made from the CIS-based approach  make more sense than those made using a traditional
spatial interpolation technique such as kriging or inverse distance weighting. This is
especially true in the western U.S. where monitoring sites are many hundreds of kilometers
apart.  The CIS-based approach predicts large areas of the West as having  low  O^ air quality
due to their remoteness to sources of O ^-forming precursors whereas traditional interpolators
predict a generally smooth gradient  of elevated air quality between such distant  monitors as
San Francisco,  Salt Lake City, and  Seattle.   The CIS-based technique makes use of the data
values at each monitored  site and honors the monitored value at each site.  Assuming  the
PES is a reliable indicator of the patterns and variation of O^ air quality between monitored
sites and since the density of the 10 km grid of PES values is much greater than the density
of monitored O^ sites, which can be hundreds of kilometers apart, the ability to capture
trends and variation of O^ air quality between monitored sites is theoretically improved.
However.  EPA recognizes that  uncertainties inherent in methods for estimating air quality
and for adjusting air quality to consider different attainment scenarios remain.   Moreover,
while the results of applying the CIS  technique correspond well with other methods for
assessing and displaying air quality data from the same year, there has been no  formal
external peer review or penormance evaluation of this national air quality extrapolation.

-------
                                           241
        2.      Assessment of Risks to Vegetation
        For over twenty years EPA has been developing, defending and enforcing risk
 assessment-based regulation (U.S. EPA, 1995). To make its risk assessment approach more
 transparent to those outside EPA, however, the Agency has begun to establish Agency-wide
 guidance and principles for risk assessment, risk characterization and risk management.   As
 part of this effort, the Agency's 1992 Guidance on Risk Characterization for Risk Managers
 and Risk Assessors  was updated in February, 1995, and on March 21,  1995 the
 Administrator of EPA signed a "Policy for Risk Characterization".  Though both of these
 documents were designed specifically to address human health, guidance specific to
 ecological risk will be available in the future.  In the meantime, many of the steps involved
 in a human health risk assessment have parallels in environmentalaassessments (e.g.,  hazard
 identification; dose-response assessment; exposure assessment). Thus, uncertainties
 regarding human health risk assessment can also be found in techniques for assessing risks to
 vegetation,  e.g., extrapolating  from exposure-response functions established in clinical
 studies or, in the case of vegetation, from open-top chambers.
       This section uses and builds upon effects and exposure-response information drawn
 from the CD  with additional information on location of sensitive species and GIS-projected
 national O3 air quality to identify potential areas of residual risk to vegetation for crop yield
 and/or tree  seedling biomass loss, under various alternative primary and secondary attainment
 scenarios.  Additionally, this section discusses potential qualitativelrisks to vegetation
 including impacts on biodiversity, long-term health of forests and forested ecosystems,
 aesthetic values,  and on the value of vegetation as  habitats for birds and other species.
       In order to evaluate the present risk to vegetation under ambient air quality, the CIS
 was used to project seasonal O3 air quality for a base year,  selected as 1990, in terms of the
 3 month, 12 h SUM06 exposure index (Figure  VII-10). The map for the same scenario
 shown in Figure VII-10 can be found in Appendix  E expressed in  terms ,x>f W126. Though
 the uncertainty associated with  these air quality projections cannot be quantified, as stated
 above, in the  absence of more complete monitoring data it serves as a useful tool for
identifying areas  across  the country where exposure levels would be expected to exceed those
known to produce yield or biomass loss at given levels for crops and trees, respectively.

-------
   Estimated  12-Hour SUM06  Ozone  Exposure -  Max  3-months  for  1990
          Estimated SUM06 (
           m 0-4.9
           Ms.O - 14.9
           • 15.0 • 25.4
             25.5 - 38.4 .
             > 38.4
                                                                                          ere
                                                                                          c
                                                                                               to
                                                                                               -p>
                                                                                               to
//dotoJ/ondih/lOEJmo/ltlnt/
                     10/31/95

-------
                                          243
 Staff judges that a visual comparison of Figure VII-10 with the 1990 map of monitored
 SUM06 values greater than 25 ppm-hrs by county (Figure IV-4) in chapter IV shows a good
 match, increasing staffs confidence in these air quality projections.
        Figure VII-10 suggests that under the base year (1990) air quality, a large portion of
 California and a few localized areas in North Carolina and Georgia have seasonal O3 levels
 above that have been reported to produce greater than 20% yield loss in 50%  of NCLAN
 crops and greater than 17% biomass loss in 50%  of studied tree seedling species.  A broader
 multistate region in the east is estimated to have air quality sufficient to allow up to  20%
 yield loss in 50% of NCLAN crops and 17% biomass loss in seedlings, while at least a third
 of the country, again mostly in the eastern  U.S. most likely has seasonal exposure levels
 which could allow up to 10%. yield loss in  50% of NCLAN cropland studied seedlings.
 Thus,  the staff concludes that current air quality is resulting in  significant impacts to
 vegetation.
       Maps were also generated for selected "just attain" scenarios (Figures VII-lla,b,c) by
 analytically adjusting air quality distributions to reflect attainment of various alternative
 primary standard options (see Horst, R. and M. Duff, 1995).  These maps, used in
 estimating benefits of control, can also depict areas which might experience residual  risk
 after attainment of the standard.
       When  1990 air quality is rolled back to attaining the current 0.12 ppm, 1-hour
                                                            -^B
 primary and secondary NAAQS, the overall seasonal 12 hr SUM06 exposures improve, but
 not dramatically (Figure VII-1 la).  The areas in North Carolina and Georgia previously
 having exposure levels in the highest part of the range now drop down to the second highest
 level (25.5 to  38.4 ppm-hr).  The area in California showing exposures in the highest part of
 the range shrinks, with more areas now in the second highest level.  A few other areas,
including areas in Texas, Louisiana, Oklahoma, Maryland, New Jersey and Delaware also
dropped to the third highest range (15.0 to  25.5 ppm-hr) from the second. However, under
this attainment scenario, there are still areas of the country judged to have seasonal O3 levels
sufficient to cause greater than (California)  or equal to (multistate region in east) 20% and
 17% yield or biomass loss in crops and tree seedlings, respectively.  Thus, the staff

-------
    Estimated  12-Hour  SUM06  Ozone  Exposure  -  Max  3-months  for  1990
   Monitored sites rolled  back to just attain the 1-hr daily max >0.12 ppm current primary standard  with  1 exceedence
           Estimated SUM06 (ppm-hr)
            CD 0-4.9 	
              I 5.0 - 14.9
              I 15.0 - 25.4
              I 25.5 - 38.4
              I > 38.4
                                                                                                        a
                                                                                                        CTQ
                                                                                                        I
                                                                                                            K)
//
-------
   Estimated  12-Hour SUM06  Ozone  Exposure  -  Max  3-months  for  1990
Monitored  sites rolled back to just attain  the 8-hr daily  mean >0.08 ppm alternative primary standard with 5 exceedences
           Estimated SUM06 (ppm-hr)
               o - 4.9
              5.0 - 14.9
              15.0 - 25.4
              25.5 • 38.4
               > 38.4
                                                                                                       cro
                                                                                                        I
                                                                                                        K   ^
//doloJ/cnd
-------
    Estimated  12-Hour  SUM06  Ozone  Exposure  -  Max  3-months  for  1990
 Monitored sites  rolled  back to  just attain the 8-hr daily mean >0.08 ppm alternative primary standard  with 1  exceedence
           ,
stimated SUM06 (ppm-hr)
fD 0-4.9
  5.0 - 14.9
  15.0 - 25.4
  25.5 - 38.4
   > 38.4
                                                                                                     Crq
                                                                                                           N)
                                                                                                           ^
                                                                                                           0>
                                                                                                     o
//(olo3/ond»fc/il263ino/!elol/hl«5 ami

-------
                                          247
concludes that attaining the current (primary and secondary) NAAQS does not provide
adequate protection of vegetation.
       To illustrate the impact of recommended alternative 8-hour primary standards,
exposure maps have been generated for the 0.08 ppm, 1- and 5-expected-exceedance
alternatives.  These maps (Figures VII-llb,c) show a markedly improved picture of O3 air
quality compared to Figure VII-1 la, with only slight improvements achieved by moving from
the 5-to 1-expected-exceedance form.  The only state still shown to exhibit seasonal
exposures high enough to result in 20% yield loss for crops is California, while the majority
of the southeast has been estimated to drop to exposure levels that could allow up to 10%
yield and biomass loss in 50% of NCLAN crops and tested tree seedlings, respectively.
Thus, the staff concludes that a primary standard of 0.08 ppm,.8.-lir, 1- to 5-expected
exceedance, when attained at all locations, would be expected to provide significantly
improved protection of vegetation from seasonal O3 exposures of concern.
       It remains uncertain, however, the extent to which air quality improvements designed
to reduce 8-hour 03 concentrations would reduce all of the biologically relevant O3 exposures
measured by a SUM06 index.  Thus, because 1) the biological database stresses the
importance of cumulative, seasonal exposures in determining plant response, 2) plants have
not been specifically tested for the importance of the daily maximum 8-hour O3
concentrations in relation to plant response, and 3) the effects of attainment of an 8-hour
standard on seasonal O3 air quality distributions is still uncertain, these uncertainties should
be considered when evaluating the vegetation benefits associated with attaining alternative
primary standards.
       Quantifiable Risks: Commodity Crops and Fruits/Vegetables.  For eight of the
NCLAN crops for which exposure-response functions were described in the CD, the 1992
U.S.  Department of Agriculture national crop production statistics  were used to derive the
location and type of commodity crops growing in the U.S.  EPA's NHEERL-WED at
Corvallis,  using the GIS, combined this information with projections of air quality based  on
1990  monitoring data as discussed above, to produce 8 maps showing estimated yield  loss

-------
                                          248
individually for each of the 8 NCLAN crops (Appendix E).  Yield losses greater than  10%
(relative to the baseline of yield at O3 levels of 0.025 ppm used in the NCLAN project) are
estimated  to occur in a few areas for soybean, kidney bean, wheat, cotton, and peanut, with
lower yield losses estimated for barley, corn, and sorghum.  Building on these quantitative
estimates, and also considering fruits and vegetables grown in California, section VI1-F-3
estimates  the economic benefits of reductions in O3 exposure associated with alternative
standards.
       Quantifiable Risks: Tree Seedlings.  In a process similar to that used  for crops  above,
information on exposure-response functions for biomass loss for seedlings of eleven tree
species taken from the CD and information on tree growing regions derived  from the U.S.
Department of Agriculture's Atlas of United States Trees (Little, 1971) was combined  with
projections of air quality based on 1990 monitoring data, to produce  11 maps showing
estimated  biomass loss individually for each of the 11 seedling tree species.  These maps
(Appendix E) show significant variability in projected seedling biomass loss under 1990 air
quality conditions across the species studied.  For example, for the most sensitive species
studied, black cherry, seedling  biomass loss is projected to be greater than 30% for over half
of its geographic range, though it can reach as high as 44% in as  much as 10% of its
growing area. Tulip poplar seedlings are predicted to have a median area weighted biomass
loss of 9%, though more  sensitive strains growing in 10% of its geographic range may
experience up to 18% biomass  loss.  The less  sensitive white pine and aspen seedlings reach
a projected biomass loss up to  10% for 10% of the growing region but only  2-3% over 50%
of their mapped  area.  Sugar maple and ponderosa pine seedling losses range from a
maximum of 3-4% in approximately 10% of their growing region though on average they
experience only  1-2% biomass  loss.  Biomass  loss estimates for the least sensitive species
studied, red alder, Douglas fir, Virginia pine,  and red maple seedlings, are projected to be
less than 2% in  all areas.   These findings again show the wide variability in  seedling
sensitivities,  both inter- and intraspecific.  Though these maps show the geographical range
for each species, they do  not indicate that each species will be found at every point  within its
range.  It should also be recognized that the production of these maps incorporates several
separate sources of uncertainty, beginning with the exposure-response functions produced for

-------
                                           249
 seedlings in open-top chambers to the uncertainties associated with the inputs to the GIS and
 projecting to the national level.  Furthermore, percent biomass  loss in tree seedlings is not
 intended to provide any information on expected biomass loss in mature trees of the same
 species, nor can it be considered  comparable to percent yield loss in agricultural crops.   The
 latter is because a 1-2% biomass  loss per year in perennial species, if compounded over
 multiple years of exposure could  become severe,  while the same percentage yield loss in
 crops annually would not be significant.
        Uncertainties in Quantitative Risk Assessment The uncertainties in exposure-response
 functions, experimental procedures, air quality data,  and the use of GIS technology are  all
 highlighted above.  The combination of these elements in the risk assessment process itself,
 however, produces additional uncertainties  that are discussed here.  First, map interpretation
 can be  a major source of error in perception because maps showing the geographical range
 for each species do not indicate that this  species will be found at every point within its range.
 For example, maps projecting estimated  biomass reduction for individual tree species may be
 interpreted by some as suggesting a homogeneously dense mono-culture forest,  which is not
 the case.   Furthermore,  maps of species  ranges do not reflect where a particular species is a
"key component in  an ecological system, or where a particular species is economically
 important  to the region.  The AVHRR satellite data was used to mask out urban and
 agricultural areas,  but potentially other factors influence the presence/absence of a species
 within  the range defined by Little (Little, 1971).  These include islands of high elevation
 which  might not be conducive to  that particular species or mono-culture tree plantations of
 different species (Hogsett, et al.,  1995).  However, given  these uncertainties, the estimated
 seedling biomass losses represent potential  risks that species may experience, affecting
 seedling establishment, reforestation or natural regeneration.
        Qualitative Risks: Crops/Fruits/Vegetables/Urban Ornamentals. Beyond yield loss
 effects, commercial crops,  including fruit and nut species, may be at risk  of other indirect
 effects  from O^  , such as shifts in their relationships  with pests and pathogens and reduced
 biodiversity.  For example, breeding programs  designed to improve yield  could inadvertently
 be selecting plants with either greater or  reduced  OT tolerance.  Several cases cited in the
 CD show that cultivars grown in  areas with high  O  levels are  more tolerant than  their

-------
                                           250
counterparts developed elsewhere (Reinert et al., 1982, Roose et al, 1982).  Velissariou et al.
(1992) suggested that selection for higher yields (and higher stomatal conductance) had
resulted in a higher O3 uptake for modern spring wheat cultivars, contributing to their
increased O3 sensitivity. It is therefore possible that the persistence of O3 in crop growing
regions may result in a reduction in genetic diversity of crop cultivars available, together
with the loss of other beneficial traits that may be linked genetically with O, sensitivity.
Tolerance mechanisms  based on reduced stomatal conductivity would likely reduce growth of
tolerant plants, while tolerance based on the production of antioxidant compounds would
likely shunt plant resources away from  growth to the production of the defense compounds.
Such indirect effects may also occur in  plants used in urban landscaping  and gardens, both
for aesthetics and for other purposes such as shade/cooling, animal/songbird habitat, soil
retention, and commercial florists.
       Qualitative Risks: Trees/Commercial Forests/Forested Ecosystems/Class  I Areas.  As
previously discussed, methods have not at this time been adequately developed to scale
biomass loss effects in  seedlings to effects in mature trees.  However, field  observations of
seedling health and mortality can provide information relevant to assessing risks to trees and
forests.  For example, field plot observations of seedling health and mortality  in natural giant
sequoia groves over a 4-year period showed that  seedling numbers were  reduced drastically
from drought and other abiotic factors.  Ozone injury symptoms were also observed in the
weeks following germination.  These observations suggest that O3 could  be  stressing
seedlings sufficiently to reduce root growth immediately after germination, thus  increasing
vulnerability to late summer drought.
       The above observations have a counterpart in commercial tree nurseries where the
importance of root development as an indicator of plant health has long been known.
Tourney and Korstian  (1947) stated that "in judging quality of nursery stock much greater
emphasis should be placed on number,  size, extent, and conditions  of roots  than on
appearance above ground." Several more recent studies conducted  at nurseries have
investigated the relationship between the number of first order lateral roots/seedling (FOLK)
and seedling quality, competitiveness, and survivability. For example, nursery studies of
sweetgum and  loblolly pine seedlings suggested that very early adverse impacts on root

-------
                                          251
growth that could result in pan from O3-related effects, can have significant impacts in
growth and competition in forest environments in subsequent years (Kormanik, 1986;
Kormanik et al., 1990).  These more recent studies lend support to the suggestive earlier
observations.
       The importance of below-ground effects on trees, forests,  and ecosystems has often
been overlooked when evaluating responses to O3 exposure.  The below ground system  is
dependant on the above ground system for inputs of energy containing substrates.
Mycorrhizal fungi  invade the roots of the vast majority of terrestrial  plants and assist the host
plant in the uptake of nutrients and water, protect the roots against pathogens, produce plant
growth hormones,  and transport carbohydrate from one plant to another.  In exchange,  the
roots of the host plant provide the fungi with simple sugars.  This symbiotic relationship is
especially beneficial  to plants growing on nutrient poor soils, and contributes substantially to
ecosystem  function.  As discussed in section VII.B, O3  stress inhibits photosynthesis and
reduces the amount of sugars available for transfer to the roots.  Reduction in available
sugars  in the roots can alter mycorrhizal  colonization and compatibility, reducing mycorrhizal
formation and root growth.  Berry (1961) examined the roots of eastern white pine injured by
O3 and observed that healthy trees had almost twice the percentage of living feeder roots as
trees with O3 injury.  In the San Bernardino forest in  California, Parmeter et al. (1962)
observed that the feeder roots system of ponderosa pine exposed to O, showed marked
deterioration (US EPA, 1986).  Numerous other studies cited in the  CD have  documented the
reallocation of carbohydrates away from  roots to photosynthetically active portions of the
plant as a result of O3.  For example, Spence et al. (1990)  found a reduction in transport of
photosynthates to roots in Cytreated loblolly pine, and Edwards (1991)  reported reduced root
and soil respiration.
       Beyond biomass loss and impacts  on root systems,  other risks to vegetation associated
with O3 include shifts in the relationship between  tree species and insects or pathogens which
can result in imbalances within communities that may have long-term effects.  Significant
risks could result,  such as has been observed in the San Bernadino forest as discussed in
section VII-D.  Ozone effects can also reduce the ability of affected  areas to provide habitats
to endangered species.  For example, two listed endangered plant species, the spreading aven

-------
                                           252
and Roan Mountain bluet, are currently found at a small number of sites in eastern
Tennessee and western North Carolina ~ forested areas where O3-related injury is of
concern.  In addition, O3-related effects on individual O3-sensitive species that provide unique
support to other species can have broader impacts.  For example, one such species is the
common milkweed, long  known for its sensitivity to O3 and usefulness as an  indicator species
of elevated O3 levels, as well as being the sole food of the monarch butterfly larvae.  Thus, a
major risk associated with of the loss of milkweed foliage for a season is  that it might have
significant indirect effects on the monarch butterfly population.
       A large number of studies have shown that O3-sensitive vegetation exists over much
of the U.S., with  many native species located in forests and Class I areas, which are
federally  mandated to preserve certain air quality related values. A list of species found in
the Great Smoky Mountains National Park and in Acadia National  Park and their associated
O3 response are included  in Appendix D.  In managing federal Class I areas, the National
Park  Service within the U.S. Department of Interior considers air pollution effects on
resources in Class I areas to constitute unacceptable adverse impacts if such effects diminish
the national significance of the area, impair the quality of the visitor's experience, and/or
impair the structure and functioning of ecosystems.  Factors that are considered in  the
determination of whether an  effect is unacceptable, and therefore adverse, include the
projected frequency, magnitude, duration, location, and reversibility of the impact.  Based on
monitoring data, review of scientific literature, and field reconnaissance, O3 has been
identified as a priority  pollutant of concern to biological resources,  and the National Park
Service Report to  Congress (1985) indicated that  sensitive vegetation is being injured by O3
transported into the parks.
       Thus,  since the state of the science does not yet permit adequate characterization of
the chronic effects of long-term exposures at the above levels to perennial species,  the staff
conclude that additional protection to vegetation  may be warranted and recommend that
research be continued into the long-term chronic effects of O3 on perennial species.

-------
                                          253
       3.     Economic Benefits Assessment
       As discussed above, ambient O3 levels have been associated with a number of welfare
effects on vegetation, including reductions in the yield of some important commercial crops.
This section presents results of quantitative analyses of the economic benefits associated with
attaining alternative standards.  Adequate data are currently available to assess  economic
benefits for the commodity crops studied in the NCLAN project (discussed in section VII-
D.2) and for fruits  and vegetables grown in California.  Urban ornamentals and commercial
forests are also addressed below, although no quantitative benefits assessment is presented for
these effects due primarily to the lack of exposure-response functions (relating  O3 exposure
levels to losses in yield, function, or aesthetic value) and economic welfare models.
       Commodity  Crops.  The economic value associated with varying levels  of yield  loss
for commodity crops was analyzed using a revised and updated (Mathtech, 1994) Regional
Model Farm (RMF) (Kopp et al., 1985) agricultural benefits model.  The RMF is an
agricultural benefits model for crops  that account for about 75% of all U.S.  sales of
agricultural crops (Mathtech, 1994).  RMF is an economic model that explicitly incorporates
exposure-response functions into microeconomic models of agricultural producer behavior.
The model uses the theory of applied welfare economics to value changes  in ambient O3
concentrations brought about by particular policy actions such as attaining  alternative ambient
air quality standards.
       The measure of welfare  calculated by the model is the net change in consumers' and
producers'  surplus from baseline O3 concentrations to the O3 concentrations resulting from
attainment of alternative standards. To calculate this change in. surplus, the model
determines how the supply function for specific agricultural products shifts in response  to
changes in  O3  concentrations. This information is combined with estimates of agricultural
commodity demand to determine the surplus gain or loss.
       To calculate the change in surplus,  the model requires five pieces of information:
       •       baseline costs of production
       •       baseline O3 concentrations
       •       demand elasticity for each crop

-------
                                          254
       •     post-control O3 concentrations (associated with attaining alternative standards)
       •     exposure-response functions
       Using the baseline and post-control equilibriums, the model calculates the change in
net consumers'  and producers' surplus on a crop-by-crop basis.  This dollar value  represents
a measure of the change in social welfare associated with the scenario policy action. The
model includes three alternative welfare measures to reflect different assumptions about the
role of agricultural policy measures.
       Since the original development of RMF in the mid-1980's, there has been discussion
in the literature concerning the appropriate measure of welfare change in agricultural sector
markets (McGartland, 1987; Madariaga, 1988).  At issue in these discussions has been the
proper consideration of price and income support measures, acreage constraints, and other
programs put in place to address potential market failures due to the inherent instability of
agricultural production.  Changes in surplus based on the assumption of perfect competition
do not properly characterize markets supported by government policy actions.  When govern-
ment  establishes loan rates and target prices and accumulates surplus production, additional
production due to O3 improvements could have oniy marginal economic value.
       The three alternative measures of welfare change calculated by the RMF are discussed
in the RMF User Manual (Mathtech, 1994) and listed below:
       •     Welfare change under conditions of perfect competition.
       •     Welfare change with price supports but no agricultural policy adjustments, or
              "present agricultural policy", (McGartland, 1987).
       •     Welfare change with price supports and agricultural policy adjustments to
             affect production levels,  or "modified agricultural policy", (Madariaga, 1988).
       The RMF model is capable of providing benefit estimates for eleven major field
crops:  soybeans, corn,  spring wheat, winter wheat, cotton, peanuts, barley,  sorghum, oats,
alfalfa and hay.  The model employs biological exposure-response information derived (Lee
et ah, 1996) from controlled experiments conducted by the NCLAN, and  a simple, yet
detailed, microeconomic model of producer behavior.  Inherent uncertainties associated with
the NCLAN data are discussed  in Section VII.D.2.  Uncertainties associated with  the use of
the RMF are discussed in Mathtech 1994 and include: assumption  of O, neutrality, no

-------
                                          255
 behavioral adjustments, and differences in estimated benefits from alternative measures of
 welfare.
       An evaluation of the effects of changes in ambient concentrations of O, on agricultural
 crops requires O3 data which reflect exposure conditions pre- and post-control strategy.
 These data must coincide in both spatial and temporal dimensions with the exposure-response
 relationships which characterize the sensitivity of selected crops to O3. Given the non-linear
 relationship between exposure and response, both the baseline level of concentrations as well
 as the change in concentrations are important data for a proper assessment of the effects of
 O3 on crop yield.
       The RMF allows the user to choose between EPA's Regional Oxidant Model (ROM)
 or the Aerometric Information Retrieval System (AIRS) as the source of baseline data.  In
 addition,  the most recent update of the RMF completed in June 1995 (Mathtech, 1995) uses
 concentration values generated from the national air quality GIS projections developed by
 EPA's NHEERL-WED (Lee et al.,1996), as discussed  above in section  VII.F.I.
       The NHEERL-WED CIS-based  approach (Lee et al., 1996 and Herstrom et al., 1995)
 has characteristics of both interpolation and modeling.  Although this integrated GIS-based
 approach  is not  as complex as a true computer model,  it has the advantage  of using data that
 is readily  available across the entire country, it is very  inexpensive lo run, and it allows one
 to quickly produce exposure estimates of any exposure index for multiple months or years.
 For RMF applications,  the most important advantage of the approach is its  ability to produce
 a dense grid  of  0, exposure values across the U.S.  The major limitation associated with the
 use of the GIS approach is the lack of monitored air quality data in rural areas.
       To use the NHEERL-WED  GIS-based approach, the distribution of  hourly O3
 concentrations at a monitor is first evaluated to produce a particular O3 statistic for  that
.monitor.  The monitor-specific statistics for all monitors are then passed to  the GIS-based
 estimation program which calculates an O3 exposure value at each  10 km cell in  a grid
 covering the  U.S.  Across the United States, a 10 kilometer grid has 289 rows and  461
 columns,  for a total of about 135,000 cells.  Next, the  program aggregates  all  10 km cells
 within each county  and  determines the mean of those values.  This aggregation results in a
 monthly baseline O3 value (for the year 1990 for purposes of this  analysis) for over 3,000

-------
                                          256
counties, thus defining the baseline O3 conditions  in each county.  As with any spatial
interpolation technique that must rely on sparse data representative of urban or near-urban
areas, the uncertainty of the GIS-based O3 estimation technique is great and non-quantifiable.
However, the ability of the GIS-based technique to capture trends and variations of O3
exposure between monitored sites is theoretically  improved over other available methods.
       If baseline concentrations came from AIRS (monitored) data,  there are several
algorithmic (i.e., functional) options for creating post-control O3 conditions.  If the ROM
(modeling) baseline is used, the post-control conditions come from applying control strategies
such as "least cost", NOX/VOC ratios, etc.  The NHEERL-WED baseline, which is based on
the extrapolation of monitored data, is closer in concept to AIRS data when it comes to
creating post control O3 conditions.  NHEERL-WED post-control O3 conditions are generated
from the NHEERL-WED  baseline O3 conditions using a functional relationship.  Each
NHEERL-WED post-control O3 scenario represents a change in the profile of hourly
concentrations consistent with  the attainment of a  specific  standard.
       In  this analysis, a quadratic roll-back procedure (Horst,  1995a) was used to generate
post-control conditions for the W126 statistic (allowing 1 and 5 exceedances) and for the
SUM06 statistic (allowing 1 and 5 exceedances).  This procedure  produces a distribution of
O3 concentrations that reflect projected air quality when an alternative standard  is "just
attained".   The "just-attained"  air quality distribution is not directly linked to an emissions
control strategy.  For the  analyses of alternative O3  secondary standards,  the quadratic
rollback procedure has been applied to the NHEERL-WED baseline to create the different
alternative standard scenarios.  Each scenario is equivalent to the  attainment of a different
ambient air quality standard.
       Specifically, 4 different primary  standards  are analyzed individually and with
alternative 12 h W126 and SUM06 secondary standards of 3 stringency levels.  The
secondary  standards  are analyzed for incremental  improvement beyond that achieved by each
individual primary alternative.  A total of 16 scenarios were analyzed for the W126 option: 4
primary regulatory scenarios and  12 incremental secondary regulatory scenarios. Another 16
scenarios  were analyzed for the SUM06 option: the same  4 primary regulatory  scenarios and
12 incremental secondary  regulatory scenarios.  A summary of these scenarios is presented m
Tables Vll-4a and VII-4b.

-------
                     257
                 Table VII-4a
ALTERNATIVE STANDARD SCENARIOS EVALUATED
       USING NHEERL-WED CIS for SUM06
SCENARIO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
BASELINE
1990 Baseline
1990 Baseline
1990 Baseline
1990 Baseline
0.12 ppm 1-hour
0.12 ppm 1-hour
0.12 ppm 1-hour
0.07 ppm 8-hour
0.07 ppm 8-hour
0.07 ppm 8-hour
0.08 ppm 8-hour
0.08 ppm 8-hour
0.08 ppm 8-hour
0.09 ppm 8-hour
0.09 ppm 8-hour
0.09 ppm 8-hour
STANDARD*
0.12 ppm 1-hour
0.07 ppm 8-hour
0.08 ppm 8-hour
0.09 ppm 8-hour
25.4 SUM06 (10%)
38.4 SUM06 (20%)
49.6 SUM06 (30%)
25.4SUM06(10%)
38.4 SUM06 (20%)
49.6 SUM06 (30%)
25.4 SUM06 (10%)
38.4 SUM06 (20%)
49.6 SUM06 (30%)
25.4 SUM06 (10%)
38.4 SUM06 (20%)
49.6 SUM06 (30%)
* All SUM06 secondary standards are defined for a 12-hour averaging time, 8AM-
8PM, for the contiguous 3 month period with highest O3 in the calendar year.
Percents indicate stringency in terms of protection against yield loss for 50% of
crops as observed in NCLAN experiments.

-------
                    258
                 Table VII-4b
ALTERNATIVE STANDARD SCENARIOS EVALUATED
        USING NHEERL-WED CIS for W126
SCENARIO
1
3
3
4
5
6
7
8
9
10
11
12
13
14
15
16
BASELINE
1990 Baseline
1990 Baseline
1990 Baseline
1990 Baseline
0.12 ppm 1-hour
0.12 ppm 1-hour
0.12 ppm 1-hour
0.07 ppm 8-hour
0.07 ppm 8-hour
0.07 ppm 8-hour
0.08 ppm 8-hour
0.08 ppm 8-hour
0.08 ppm 8-hour
0.09 ppm 8-hour
0.09 ppm 8-hour
0.09 ppm 8-hour
STANDARD*
0.12 ppm 1-hour
0.07 ppm 8-hour
0.08 ppm 8-hour
0.09 ppm 8-hour
21.0W126(10%)
34.9 W126 (20%)
47.9 W126 (30%)
21.0 W126 (10%)
34.9 W126(20%)
47.9 W126 (30%)
21.0 W126 (10%)
34.9 W126 (20%)
47.9 W126 (30%)
21.0 W126 (10%)
34.9 W126 (20%)
47.9 W126 (30%)
* All W126 secondary standards are defined for a 12-hour averaging time, 8AM-
8PM, for the contiguous 3 month period with highest O3 in the calendar year.
Percents indicate stringency in terms of protection against yield loss for 50% of
crops as observed in NCLAN experiments.

-------
                                          259
       This analysis used exposure response functions estimated by NHEERL-WED (Lee et
al., 1996) for the cultivars examined in the  NCLAN experiments for corn, cotton, peanuts,
sorghum, soybean, and winter wheat.  Two functions were chosen for each crop.  One
function represents the cultivar which experienced the minimum yield impact when subject to
O3 exposure in the NCLAN  studies  (reflecting relatively low sensitivity to O3) and the other
function represents the cultivar which experienced the maximum yield impact when subject to
O3 exposure in the NCLAN  studies  (reflecting relatively high sensitivity to O3). For peanuts
and sorghum, only a single NCLAN exposure-response function was available.  Although
RMF is designed to handle regional  choices of cultivars, it does not attempt  to assign the
cultivar specific exposure-response functions to different geographic regions. Instead,
analyses are performed as if each cultivar were grown throughout the country.
       The functional form used to estimate exposure-response  functions for all crops in this
analysis is a Weibull specification transformed to predict relative yield loss.  In this analysis,
12-hr (8:00am - 8:00pm) SUM06 and W126 exposure indices were used.  Each index  is
calculated over a calendar period representing the median  exposure period for each specific
crop in the NCLAN experiments. The  median exposure period is the median of all exposure
dates  for all NCLAN experiments across all cultivars of a crop.
       Calendar periods used for computing SUM06 and W126 statistics are based on  the
harvest date and the duration of exposure for a particular crop.   Calculations are done on a
State-specific basis. This allows for geographic  variation which permits the  model  to better
reflect actual O3 exposure during the true growing period of the crop. The first day of the
month in which the crop is harvested is used as the anchor point for computing the calendar
period of exposure duration.  The start  date is computed as the  harvest date  minus median
exposure days.  The first day of the  corresponding month  is used as the start date.  Duration
is therefore the difference between the harvest date and the start date. Data  used for these
calculations were obtained from USD A, Usual Planting and Harvesting Dates for U.S.  Field
Crops (1984) and the NCLAN experimental data.
       Among the crops originally included in RMF, barley, oats,  hay, alfalfa, and spring
wheat have not been included in this analysis. The reasons for  not including these crops
vary.   The model farm data included in  RMF focus on fall barley,  while the NCLAN

-------
                                          260
experiments focus on spring barley.  Therefore,  the exposure-response functions do not
correspond to the RMF production information.  Similarly, NCLAN experiments focused
exclusively on winter wheat with no experiments related to spring wheat.  Therefore,
exposure-response functions exist for winter wheat only.  Finally, there are no exposure-
response functions that represent oats, hay, or alfalfa.  Therefore, there is no basis for
explicitly applying the RMF model  to these crops.  The crops included in the analysis are:
corn, cotton, peanuts, sorghum, soybean, and winter wheat.
       It is important to note that the range of results from this analysis represents impacts
associated only with available NCLAN experimental data.  Not all cultivars of a crop have
been subjected to exposure-response experiments.  Furthermore, the mix of cultivars actually
grown in 1990 is not represented by any single cultivar examined during the NCLAN study.
The RMF benefits results associated with either  the "minimum" or  "maximum" exposure-
response function simply represent impacts that span a range of possible results that could be
obtained with available experimental information.
       The results from this analysis measure the national economic effects of changes in
ambient O^ levels on the production of a subset of important commodity crops.  The results
indicate that  a reduction in O^ from 1990 ambient levels to a controlled  level results in a
monetary net benefit.  Conversely,  an increase in O^ to an assumed ambient concentration
across all regions produces a net loss.
        It is important to restate and summarize the uncertainties associated with the  results of
the RMF analysis presented below.  Uncertainties are  introduced by: (1) the extrapolation of
limited monitored air quality data to national air quality distributions; (2) the application of
exposure-response functions from NCLAN open-top chamber studies extrapolated to 1990
ambient air exposure patterns and crop  production; (3) the use of a quadratic rollback
 methodology to project the "just attain" air quality  distributions without a direct link to an
 emissions control strategy; and (4)  the use of the RMF with the inherent uncertainties of an
 economic model and with  differences in estimated benefits from the various economic
 measures (perfect competition, price supports with no agricultural policy adjustments, and
 price supports with  agricultural policy adjustments).  In addition, estimated yield losses are
 directional!v overestimated because  they are  based  on  NCLAN-derived function^ which use a

-------
                                           261
  background concentration of 0.025 ppm instead of a background level of 0.04 ppm (which is
  the midpoint of the range assumed in Section IV of this Staff Paper).  Overestimation of
  yield losses results in higher benefits.  Quantification of the individual or compounded effects
  of these uncertainties,  or of the directional bias resulting from the above assumptions, is
  infeasible at this time.  While the results presented should be viewed as rough
- approximations, they provide useful  insights for comparing the relative benefits obtained as a
  result of attaining alternative regulatory scenarios.
        Tables VII-5a through VII-5d present the results from applying the  RMF to determine
  commodity crop benefits based on four alternative  primary standards (shaded rows) and on
  three alternative levels of stringency for a particular secondary standard. The four primary
  standards are: 0.12 ppm Ihr; 0.07 ppm 8hr; 0.08 ppm 8hr; aricTtJ.09 ppm  8hr.  The two
  alternative secondary standards are: SUM06 and W126, 12hr, 3 month at various  stringency
  levels.  For the SUM06 index, the levels 25.4 ppm-hr, 38.4 ppm-hr, and 49.6 ppm-hr are
 •associated  with  10%, 20%, and 30% yield loss prevention, respectively, in 50% of crops
  studied in the NCLAN experiments.   For the W126 index, the levels 21.0  ppm-hr, 34.9
  ppm-hr, and 47.9 ppm-hr are associated with 10%, 20%,  and 30% yield loss prevention,
_ respectively, in  50% of crops  studied in the  NCLAN experiments.  Tables 5a and 5b
  correspond  to the  SUM06 index based on primary  standards  that allou 1 and 5 exceedances,
  respectively.  Tables 5c and 5d correspond to the W126 index based  on primary standards
  that allow  1 and 5 exceedances, respectively. The 1- and  5-exceedance cases are  calculated
  at the primary standard level for the 8-hour options; only the 1-exceedance case is calculated
  for the present 0.12 ppm, 1-hour standard.
        As  stated above, the baseline for the welfare benefits calculated in this analysis was
  the 1990 air quality.  The 1990 air quality baseline was chosen  for consistency reasons: it is
  the baseline used for modeling, implementation, and other analyses related to the O, standard
  review.  As evidenced by the results presented in tables VII-5a through 5d, it is estimated
  that positive benefits result from attaining the present standard based  on the 1990  air quality
  baseline.  Additional incremental benefits are estimated for also attaining alternative
  secondary standards.

-------
                                               262

                                            Table Vll-5a
                                     Regional Model Farm (RMF)*:
                              Changes in Economic Surplus for Alternative
                       Primary and Sum06 Secondary Standards (millions $ in 1990)
                                       (1 Expected Exceedance)

                          	ALTERNATIVE WELFARE MEASURES
        Standard
                    , xs
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
1 Exoaedaftce Primary:
Secondary
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6SUM06 12-HR
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
1 Exceedance Pnmary:
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
130
50
20
 0
 0
 0
10
 1
 0
60
10
 1
 0
 0
 0
30
 3
 0
150
30
 3
            100
            40
            10
 0
 0
 0
10
 1
 0
40
10
 1
            210
            70
            20
 0
 0
 0
20
 1
 0
90
20
2
            160
            60
            20
 0
 0
 0
20
 1
 0
70
20
2
                                                           Maximum
                                                                $??&'
 0
 0
 0
40
 3
 0
160
30
 4
* Includes estimates for corn, cotton, peanuts, sorghum, soybean, and winter wheat

NOTE: The first two significant figures were retained in arriving at the above values.

-------
                                                263
                                            Table Vl!-5b
                                     Regional Model Farm (RMF) *:
                              Changes in Economic Surplus for Alternative
                        Primary and Sum06 Secondary Standards (millions $ in 1990)
                                       (5 Expected Exceedances)
                                 ALTERNATIVE WELFARE MEASURES
        Standard
                  j-^v«
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
Secondary.
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
5^ Eatceedance Primary;
Secondary:
  25.4 SUM06 12-HR
  38.4 SUM06 12-HR
  49.6 SUM06 12-HR
130
50
20
 0
 0
 0
10
 1
 0
80
10
 1
320
110
40
 0
 0
 0
40
 3
 0
200
30
 2
100
40
10
 0
 0
 0
10
 1
 0
60
10
 1
210
 70
 20
 0
 0
 0
20
 1
 0
130
20
 1
 0
 0
 0
20
 1
 0
100
20
 1
            380
            120
             40
                                                                                       ;
 0
 0
 0
40
 3
230
30
 2
* Includes estimates for corn, cotton, peanuts, sorghum, soybean, and winter wheat

NOTE: The first two significant figures were retained in arriving at the above values

-------
                                                     264




                                                Table Vll-5c
Regional Model Farm (RMF) *:
Changes in Economic Surplus for Alternative
Primary and W126 Secondary Standards (millions $ in 1990)
(1 Expected Exceedance)
ALTERNATIVE WELFARE MEASURES
Standard
tExceed *>«
ti&wmVte **"
Secondary.
21. OW126 12-HR
34.9 W126 12-HR
47.9 W1 26 12-HR
t Exceedance Primary:
0,07ppm 84U-
Secondary
21.0W126 12-HR
349W126 12-HR
479W126 12-HR
1 Exceedaoce Prtmary: - ...
OXte pprw iNtr -:.. •.;::.:/::'•.;:.
Secondary
21 OW126 12-HR
34.9 W126 12-HR
47.9 W1 26 12-HR
1 Exceedance Primary:
0.09 ppm 8-hr
Secondary
21.0W126 12-HR
349W126 12-HR
47.9 W126 12-HR
^Competitive gq^Iflfctftsinii
Minimum
f»' 5

$80
$20
$2

4(340

$0
$0
$0
•.;;::.;:.:.;.:•::....•
•P:.:;$3J40,. .. .:

$4
$0
,.. $0

$130

$30
$2
$0
Maximum
j^t&o^l

$190
$40
$3

$1/fQ0 ,

$0
$0
$0

ST30

$10
$0
$0

$3flQ ' :'. ;

$60
$4
$0
Iti^ftowwtaw P^fqr"
Minimum
'Swi!'\ '

$60
$10
$1

$260

$0
$0
$0

$180

$3
$0
$0

'"-. ?too i"

$20
$1
so
Maximum
'- SJ(^; ^

$130
$30
$2

, $830

$0
$0
$0

$$40

$10
$0
$0

$270

$40
$2
$0
ModmedAsPoWcy |
Minimum
'fo^ '|

$100
$20
$2

$390

$0
$0
$0

$280

$10
$0
$0

"' • $ii$o

$30
$2
$0
Maximum
';-)$$!$, 'J

$230
$50
$4

$1,400 ' ]-

$0
$0
$0
% 1
$920

$10
$0
$0

$470 ;

$80
S4
$0
* Includes estimates for corn, cotton, peanuts, sorghum, soybean, and winter wheat



NOTE. The first two significant figures were retained m arriving at the above values

-------
                                                 265

                                             Table Vll-5d
                                     Regional Model Farm (RMF) *:
                              Changes in Economic Surplus for Alternative
                        Primary and W126 Secondary Standards (millions $ in 1990)
                                       (5 Expected Exceedances)
                                  ALTERNATIVE WELFARE MEASURES
        Standard
Secondary:
  21.0 W126 12-HR
  34.9 W126 12-HR
  47.9 W126 12-HR
$ Exieedam* Primary;
Secondary
  21. OW126 12-HR
  34.9 W126 12-HR
  47.9 W126 12-HR
Secondary.
  21.0 W126 12-HR
  34.9 W126 12-HR
  47.9W126.12-HR
5:Ej«pedance:PriFrtary;
Secondary:
  21.OW126 12-HR
  34.9 W126 12-HR
  47.9 W126 12-HR
80
20
 2
 0
 0
 0
4
0
0
30
2
0
190
40
 3
 0
 0
 0
10
 0
 0
90
 4
 0
60
10
 1
 0
 0
 0
 3
 0
 0
30
 1
 0
130
30
 2
 0
 0
 0
10
 0
 0
60
 2
 0
100
20
 2
 0
 0
 0
10
 0
 0
50
 2
 0
230
 50
 0
 0
 0
 10
 0
 0
                                                                                         ..'-•
110
 4
 0
* Includes estimates for corn, cotton, peanuts, sorghum, soybean, and winter wheat

NOTE: The first two significant figures were retained in arriving at the above values.

-------
                                          266
       Results are presented in annual 1990 dollars for three different economic measures
(competitive equilibrium, present agricultural policy, and modified agricultural policy) and
for two endpoints (minimum and maximum impacts) as discussed above.  The shaded rows
present results for the alternative primary standard options, with incremental benefits from
the alternative secondary standards shown below each primary standard alternative.  Tables
detailing results for combined crops and individual crops are presented in Horst (1995b).
       In summary, this analysis estimates a range of benefits using three different economic
measures for several alternative primary and incremental secondary standards.  Based on the
SUM06 analysis and the economic model that incorporates price supports and agricultural
policy adjustments (the modified agricultural policy), estimated annual benefits relative to a
1990 baseline associated with just attaining the current (primary and secondary) NAAQS
(0.12 ppm, 1-hr average) range from approximately $80 - $230 million, with incremental
annual benefits associated with simultaneously attaining  the most stringent secondary standard
alternative ranging from approximately $160 - $380 million. Estimated annual benefits
associated with attaining an alternative primary  standard (0.08 ppm, 8 h average,  1  expected
exceedance) range from approximately  $350 million - $1.2 billion, with small ($20-40
million) incremental benefits associated with a simultaneous most stringent secondary
standard. The annual benefits from attaining the 0.08 ppm, 8 h alternative  primary standard
alone are estimated to be greater than the combined  benefits associated with attaining the
current NAAQS and the most stringent secondary standard alternative considered  in this
analysis.
        For illustration purposes, Table VII-6 presents the incremental benefits of  moving
from the present  standard to a different standard.  This  illustration is an alternative
representation of selected information from Table VH-5a. It shows incremental benefits for
alternative primary and  SUM06 secondary standards, assuming one expected exceedance, and
using maximum results  from  the modified agricultural policy welfare measure.

-------
                                      267
                                  Table VII-6
  An Illustration of Incremental Changes in Economic Surplus for Alternative Primary and
SUM06 Secondary Standards Using Regional Model Farm (RMF) Results from Table VII-5a
                               (Millions $ in 1990)
                             (1 Expected Exceedance)
                Standard
Modified Agricultural Policy Welfare
Measure and Maximum Yield Impact
Functions
 Primary 0.12 ppm 1-hr
 Secondary 25.4 SUM06 12-hr
$230
$380
 Primary 0.09 ppm 8-hr
 Secondary 25.4 SUM06 12-hr
$360
$160
 Primary 0.08 ppm 8-hr
 Secondary 25.4 SUM06 12-hr
$970
 $40
 Primary 0.07 ppm 8-hr
 Secondary 25.4 SUM06 12-hr
$1,600
  $0

-------
                                          268
       Benefits from economic studies reported in the CD range from SI.3 billion to $2.5
billion. Two of the cited studies were national in scope and judged adequate in terms of data
inputs.  Kopp et al. (1985), using the RMF including corn, soybeans, wheat, cotton, and
peanuts, reported $1.3 billion (1980 dollars) in benefits from the reduction of ambient O3
from 53 to 40 ppb. Adams et al. (1986) included corn, soybeans, cotton, wheat, sorghum,
and barley, and reported $1.7 billion (1980 dollars) in benefits from a 25% reduction in
ambient O3 from  1980 levels.  These results are not directly comparable to  the results from
this RMF analysis because they are based on higher baseline air quality (1980 versus 1990 in
this analysis), and they are based on different post-control average air quality levels that
cannot be related to the alternative standards  used in this analysis.
       Fruit and Vegetable Crops.  There are currently no  national-level economic models
that incorporate fruits and vegetables, although such efforts are underway.  A regional
model, the California Agricultural  Resources Model (CARM), has been developed and used
by the California Air Resources Board (Howitt, 1995a, b).  This model was used to analyze
the benefits of reducing ambient O3 on the sensitive crops grown in California (Abt, 1995).
Among these crops are the economically important fruits and vegetables endemic of
California and other states  with similar climate, such as Florida.  In 1990, California crops
accounted for almost 50% of the U.S. fruit and vegetable production.
       The CARM is a nonlinear optimization model of California agriculture which assumes
that producers maximize farm profit subject to land, water,  and other agronomic constraints.
The model maximizes total economic surplus and predicts producers' shifts in acreage
planted to different crops due to changing market conditions or resources.  (The RMF  model
discussed above does not have  this crop substitution predictive capability.)  The CARM uses
constant elasticity of substitution (CFJS) production functions, which permits the  substitution
of variable inputs in response  to a change in  conditions such as those due to changes in yield
associated with O3 exposure.  The CARM is disaggregated into nine production regions and
it includes 38 different cropping activities. It does not have a livestock sector, although the
demand function for some crops reflects livestock feed demand over time.  The version of
the CARM used for this analysis was calibrated to 1990 production and price data.

-------
                                           269
        The results from this CARM analysis measure the regional economic effects of
 changes in ambient O3 levels on the production of almonds, apricots, avocados, cantaloupes,
 broccoli, citrus, grapes, plums, tomatoes, and dry beans.  The results, as summarized in
 Table VII-7, indicate that reductions in O3 from 1990 levels to controlled levels result in a
 monetary net benefit.  As with the RMF analysis discussed above, two levels of response
 were evaluated for each standard to reflect the minimum response and the maximum
 response.  The CARM analysis provides an  indication of the range of benefits that can be
 estimated based on the available data.
        As  with the uncertainties of the RMF analysis discussed above, it is important to
 restate and summarize the uncertainties associated with the results of the California fruits and
 vegetables analysis presented in Table VII-7. Uncertainties are introduced by: (1) the
 extrapolation of limited monitored air quality data to national air quality distributions; (2) the
 application of exposure-response functions from NCLAN open-top chamber studies
 extrapolated to 1990 ambient air exposure patterns and crop production; (3) the use of
^alternative non-NCLAN exposure-response functions for a variety of fruits and vegetables not
- included in the NCLAN studies; (4) the use  of exposure-response functions in a non-W126
 form for non-NCLAN crops for which a W126 function could not be derived; (5) the use of
 a quadratic rollback methodology to project  the "just attain" air quality distributions without
 a direct link to an emissions control strategy; and (6) the use of the CARM with the inherent
 uncertainties  of an economic model. In addition, estimated  yield  losses based on  functions
 that include a background concentration level of 0.025 ppm instead of a background level of
 0.04 ppm (which is the midpoint of the range assumed in chapter 4) are directionally
 overestimated. Overestimation of yield losses will result in higher benefits.  Quantification
 of the individual or compounded effects of these uncertainties, or of the directional  bias
 resulting from the above assumptions, is  infeasible at this time. While the results presented
 should be viewed as  rough approximations, they provide useful insights for comparing the
 relative benefits obtained as a  result of attaining alternative regulatory scenarios.
        This analysis  estimates a range of benefits for similar alternative primary  and
 incremental secondary standards as  were  analyzed using the RMF for commodity crops.
 (The CARM  analysis adjusted air quality to  allow for no exceedances.)  Specifically, 4

-------
                                                  270

                                              Table VII-7
Changes in Economic Surpluses for California Crops
Under Alternative Primary and Incremental Secondary Ozone Standards ($ million, 1990)
Standard
Primary:/
Clt2 ppm 1-4w
Secondary.
21. OW126 12-HR
34.9 W126 12-HR
47.9 W126 12-HR
Primary::
fct£ppm «, t*K»Jacw
Minimum
% "" i
f*r

$110
$40
$30

*m

$10
$10
$10

$200

$20
($10)
($10)

•i-:-$130'-; • ':|

$50
$20
$10

$130 \

$60
$20
$10
"•&tt&W9^& \ %:
Maximum
• "•
$3®

$50
$4
$10

$100

$10
$10
$10

. '•: *8Q

$10
$0
$0

" "-$4e: : %:

$20
$10
$4

:34G

$20
$10
$4
NS^^KS^OBWll
Minimum
•^ •'vvt %% :
"-"$»& \ '

$140
$40
$40

' f98fr :

$10
$10
$10

$250

$20
($20)
($20)

"••I|MO-V -

$70
$40
$30

$140

$70
$40
$30
i^l^fel*,^,- , i
Maximum
'••,<•.•' ';
$110 I

$200
$70
$50

$480 1

$30
$30
$30

•-$350 i

$40
$0
$0

• : $220 :

$110
$40
$10

$220

$110
$40
$10
* Minimum estimates exceed maximum estimates because no change in the ozone index was estimated for the
 maximum function, but the index for the minimum function did vary between scenarios.

NOTE The first two significant figures were retained in arriving at the above values

-------
                                            271
  different primary standards are analyzed individually and with alternative 12-hr W126
  secondary standards of 3 stringency levels.  The secondary standards are analyzed for
  incremental improvement beyond that achieved by each individual primary alternative.  A
  total of 16 scenarios were analyzed for  the W126 option: 4 primary regulatory scenarios and
  12 incremental secondary regulatory scenarios. Estimated annual benefits in California
  relative to a 1990 baseline associated with just attaining a 1-hr, 0.12 ppm NAAQS range
  from approximately $80  - $110 million, with incremental annual benefits associated  with
  attaining the most stringent secondary standard alternative ranging from approximately $140 -
  $200 million.  Estimated annual benefits associated with just attaining the 0.08 ppm  8 h
  primary standard ranges  from approximately $250 - $350 million, with   incremental benefits
  associated with just attaining the most stringent alternative secondary standard ranging from
-„ approximately $20 - $40 million.
         Ornamentals and  Commercial Forests.  Based on the discussion of effects in  section
  VII.D, ornamentals and commercial forests are additional vegetation categories that  represent
  large economic sectors which are likely to  experience some degree of effects associated with
  exposure to ambient O^  levels.  However,  in the absence of adequate exposure-response
  functions and economic damage functions for the potential range of effects relevant to these
  types of vegetation, no quantitative economic benefits analysis has been conducted in these
  areas as part of this staff assessment.  However, as discussed briefly below, significant
  economic benefits could  potentially result upon attaining the alternative  standards analyzed
  above for commercial crops and fruits and  vegetables.
         Ornamentals used in the urban and  suburban landscape include vegetation such as
  shrubs,  trees,  grasses,  and  flowers.  The types of economic losses that could potentially
  result from  effects that have  been associated with O^ exposure include:  1) reduction in
  aesthetic services over the realized lifetime of a plant,  2) the loss of aesthetic services
  resulting from the premature death (or early replacement) of an injured  plant. 3) the cost
  associated with removing the injured plant  and replacing it with a new plant, 4) reduced
  seedling survivability, and 5) any additional costs incurred over the lifetime of the injured
  plant to mitigate the effects of Oyinduced  injury.  Billions of dollars are spent annually on
  landscaping using ornamentals (Abt, 1995), both  by private property owners/tenants and by
  governmental  units  responsible for public areas, making this a potentially important  welfare

-------
                                           272
effects category  However,  information and valuation methods are not available to allow for
credible estimates of the percentage of these expenditures that may be related to impacts
associated with O3 exposure.
       It is  recognized that commercial markets for turf, flowers, and woody plants actively
develop, test, and select new plant cultivars for use in the urban landscape.  Breeders and
nurseries select for hardy cultivars that are resistant to the environmental stresses present in
the urban landscape.  Though breeding programs may not select specifically for O3-resistant
strains, they probably indirectly result in selection  for O3-resistance.  Many of the breeders
and nurseries are located in high O3 areas, and plant varieties developed in breeding
programs that do not initially do well at the nursery generally are discarded.  Likewise, plant
species that either do poorly at the nursery or in the market areas served by the nursery are
not carried at the wholesale level.  Thus, potential  economic losses may be limited to a
significant degree by commercial markets in which such vegetation is bred and  marketed.
       Any attempt to estimate economic benefits for commercial forests associated with
attaining alternative O3 standards is constrained by  a lack of exposure-response  functions for
the commercially important mature trees.  As previously discussed, although exposure-
response functions have been developed for seedlings for a number of important tree species,
these seedling functions  cannot be extrapolated to mature trees based on current knowledge.
Recognizing this limitation, a study by Pye et al. (1988a, b) used a method involving expert
judgments about the effect of O3 levels on percent  growth change to  develop estimates of O,-
related economic losses  for forest products.  However, until additional effects information
becomes available allowing for exposure-response  functions for mature trees to be
extrapolated from seedling effects studies, the uncertainties associated with this method limit
its usefulness in assessing the benefits of attaining  alternative air quality standards.
       Summary.  Table VII-8 summarizes the effects categories discussed above  and
presents the monetized  and non-monetized benefits associated with selected regulatory
scenarios.  Monetized benefits were calculated for  commodity  crops  studied in  the NCLAN
project and for fruits and vegetables grown in  California.  Monetized benefits for  other
crops, urban ornamentals, tree seedlings, mature forests, commercial forests. Class I areas.
and other ecosystems are not presented because of the lack of exposure-response functions

-------
                                                                   Table VII-8


                     Summary of Welfare Benefit* Estimate* Associated With Various Primary and Incremental Secondary Regulatory Options
                                                           (Rounded Value, in Millions $ in 1990)
Alternative
Standards
Benefits
Categories
COMMODITY CROPS -
Economic Vnlue
(75% of Market )'•'
FRUITS 6 VEGETABLES -
Economic Value
(50* of Market)'-0
COMMODITY CROPS &
FRUITS & VEGETABLES -
Non-economic Value
URBAN ORNAMENTALS
TREE SEEDLINGS
MATURE TREES
FORESTED ECOSYSTEMS
CLASS I AREAS
COMMERCIAL FORESTS
Lowest
Secondary Total
Current PLUS Standard • Estimated
NAAQS Analyzed Benefit
$80-230 $160-380 $240-610
$80-110 $140-200 $220-310
8-hr, 0.08 ppra Lowest
Primary Secondary Total
Standard PLUS Standard - Estimated
(5 xx) Analyzed Benefit
$320-1000 $20-40 $340-1040

8-hr, 0.08 ppm Lowest
Primary Secondary Total
Standard PLUS Standard - Estimated
(1 xx) Analyzed Benefit
$350-1180 $20-40 $360-1220
$250-350 $20-40 $270-390
\
/
Directionally Increasing Benefits, Including:
Protection from Biomass Loss (e.g., commercial value of ornamentals, forests)
Protection of Functional Values (e.g., ecosystems, ornamentals)
Preservation of Biodiversity
Protection of Aesthetic Values
Preservation of Habitat
Existence Value (e.g., Class I areas)
10
-J
<_•->
'Using  range min-max benefits under "Modified Agricultural  Policy"  measure of welfare.
*SUM06  -  25.4 ppm,  12hr is the lowest secondary standard analyzed.
'Using  range min-max benefits based on total consumer and producer  surplus.
*W126 -21  ppm,  12hr is the lowest secondary standard analyzed.
*The CA fruit and vegetable analysis was done toe. a no expected  exceedance case;  estimated values for attaining current  standard with 1 expected  exceedance
 and 'for attaining alternative 8-hr standards with 1 and 5 expected exceedances would be somewhat (between 5% and 20%) lower than values shown here.

-------
                                          274
and appropriate economic valuation models. Information presented  in preceding sections,
however, suggests that reductions in ambient 63 levels will yield benefits from these
categories.
       Table VII-8 presents a range of benefits for the commodity crops and for the
California fruits and  vegetables.  The lower end of the range is based on the cultivars which
experienced the minimum yield impact when subject to 63 exposure in the  NCLAN studies
(reflecting relatively  low sensitivity to 63).  The upper end of the range reflects the cultivars
which experienced the maximum yield impact  when subject to 63 exposure in  the NCLAN
studies (reflecting relatively high sensitivity to 03).  These results are based on the
"modified agricultural policy" measure which combines price supports and agricultural policy
adjustments to affect production levels (Madariaga,  1988).  The commodity crops analyzed
represent 75% of the U.S. sales of agricultural crops,  and the  California fruits and vegetables
analyzed represent approximately  50% of the fruit and vegetable markets.
       The information in Table VII-8 provides, for example,  the following comparisons
between the estimated benefits associated with  attaining the  current NAAQS and a new 8-
hour, 0.08  ppm primary standard,  as well as the incremental benefits associated  with
attaining the lowest seasonal secondary analyzed.  Combining the benefits estimates for
commodity crops and fruits and vegetables represents  only an approximation since there were
small differences in the forms of the alternative standards analyzed (as noted in the  Table
VII-8 footnotes).
•      Total estimated annual benefits associated with attaining the current  NAAQS include
       approximately $160-$340 M in  monetized  benefits from the commodity crops and
       California fruits and vegetables analyzed, as  well as some level of benefits from the
       other benefits categories shown, for which no quantitative estimates  could be made.
•      Total estimated annual benefits associated with attaining a new 8-hour primary
       standard of 0.08 ppm, 1-expected-exceedance,  include approximately $600-$!,550 M
       in monetized  benefits  from the commodity crops and California fruits and vegetables
       analyzed, as well as some level of benefits from the other benefits categories shown.
       for which no  quantitative estimates could be  made, although  directionally these
       benefits would be expected to be greater than those associated with attaining the
       current NAAQS.

-------
                                           275
 •      Incremental annual benefits associated with attaining the lowest seasonal secondary
        standards analyzed include approximately $300-$580 M in monetized benefits relative
        to the current NAAQS, compared to approximately $40-$80 M relative to a new 8-
        hour, 0.08 ppm, 1-expected-exceedance standard.  Additional incremental benefits
        would be obtained for the other benefits categories shown, although no quantitative
        estimates of these additional benefits could be made.

        To project monetized benefits nationwide, the above reported estimates would need to
 be scaled upward,  since the commodity crops included in the analyses account for only 75%
 of the U.S. sales of all agricultural crops and the California fruits and vegetables include
 only approximately 50%  of the nation's fruit and vegetable markets.  A rough approximation
 of a national estimate can be calculated by proportionately scaling the monetized estimates to
 the entire market.  It is recognized, however, that factors such as the sensitivity to O, of
 crops and fruits and vegetables not formally analyzed, regional air quality, and regional
 economics introduce  considerable  uncertainty to any such approach  to developing a national
 estimate. Applying such a scaling approach to the ranges given  above results in the
 following rough approximations to national monetized benefits associated with the categories
'of commodity crops and fruits and vegetables:
 •      National approximation of annual monetized benefits associated with attaining the
        current NAAQS:  $270-$530 M.
 •      National approximation of annual monetized benefits associated with attaining a new
        8-hour primary standard of 0.08 ppm, 1-expected-exceedance:  $970-$2,300 M.
 •      National approximation of incremental annual monetized  benefits associated with
        attaining the lowest seasonal secondary standards  analyzed:  $490-$910 M relative to
        the current  NAAQS, compared to approximately $70-$ 130 M relative to a new 8-
        hour, 0.08 ppm, 1-expected-exceedance standard.
        An examination of the monetized benefits reported in Table  Vli-8 indicates that most
 of the benefits accrue from the attainment of the 8-hr primary standard alternatives with a
 small incremental improvement obtained by the addition  of a seasonal secondary standard.
 The projected national approximations for commodity crops and  fruits and vegetables  suggest
 benefits on  the order of 1 to over  2 billion dollars for a new 8-hour, 0.08 ppm primary
 standard alone or in combination with a seasonal secondary  standard.  The qualitative

-------
                                          276
information presented in the table also suggests that the monetized benefits alone understate
the public welfare benefits obtained from the adoption of an 8-hour primary standard alone or
in combination with a new seasonal secondary standard.

-------
                                          277
VIII.  STAFF CONCLUSIONS AND RECOMMENDATIONS ON SECONDARY
       NAAQS
       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 secondary O3 NAAQS.  In developing these conclusions and
recommendations, the staff has drawn upon the scientific and technical information contained
in the CD and summarized in this Staff Paper.  The staff has attempted to integrate scientific
information on vegetation effects of O3,  assessments of air quality, exposure, and the risks of
such effects occurring, and both quantitative and qualitative assessments of the benefits
associated with protection of commercial crops, forest tree species ana ecosystems  into a
basis for conclusions and recommendations for a secondary NAAQS.
       In addition to scientific and technical considerations, staff has included other, policy-
relevant factors, such as the impact of the  primary standard alternatives recommended in
section VI.E, in order to place the scientific information into a policy framework for
standard setting.  The staff recognize that a decision on the standard will require a  blending
of scientific and policy considerations.
A.     Pollutant Indicator
       The staff believes that the conclusions on the appropriate indicator for the secondary
O, NAAQS presented in the previous Staff Paper  (U.S. EPA, 1989) remain valid today.   As
indicated in the previous  Staff Paper, it is  generally recognized that control of ambient O,
levels provides the best means of controlling photochemical oxidants of potential welfare
concern.  Further, among the photochemical oxidants, the database for vegetation effects
from controlled-exposure, field,  and observational sitings only raises concern at levels found
in the ambient air for O3.  Thus, the staff recommends that O3 remain as the pollutant
indicator for protection of public welfare from exposure  to all photochemical oxidants.
B.     Averaging Times
       Staff has drawn upon the considerations of averaging times developed during the last
O3 NAAQS  review as context for this current review.  The research reviewed in the 1986
CD,  1989 Staff Paper, and  the 1992 Supplement to the 1986 CD most often used a 7-hour
seasonal mean to  relate air quality to plant effects.   This index was usually preferred over

-------
                                          278
the 1-hour peak due to its greater stability as an air quality measure.  At the same time,
however,  the limitations of both the current 1-hour form of the standard and the 7-hour
seasonal mean for relating air quality to observed vegetation response were also becoming
more generally recognized.
       Based on the information available at the time, the 1989 SP concluded that 1) long-
term averages,  such as the 7-hour seasonal mean, may not be adequate indices  for relating O3
exposure and plant response, 2) repeated peak concentrations accumulated over time are the
most critical element in determining plant response for agricultural crops, and therefore
exposure indices which emphasize peak concentrations and accumulate concentrations over
time probably provide  the best biological basis for standard setting,  and 3) there is currently
a lack of exposure-response  information on forest tree effects as well as no consensus on the
most important averaging time for perennials.  At that time, the analyses available on
cumulative, peak-weighted indices was considered  preliminary.
       As the basis for the current review, the information in the 1996 CD builds on that
history and supports the earlier findings.  Many of the most recent studies have specifically
selected exposure indices that take into account cumulative impact of effects throughout the
growing season when measuring growth and yield  impacts.  Though these studies still
support the above  findings, there is growing evidence that other aspects of exposure such as
the timing of peaks, predisposition effects, and  the length of respite times between peak
exposures also play a role in determining plant response.  Insufficient information exists at
this time to incorporate these features into an appropriate averaging time for a secondary
standard,  but may become important refinements in the future.'
       Based on the information presented in Section  VII and the above considerations, the
staff continues to believe that the selection of an appropriate averaging time should  take into
account the cumulative impact from repeated peaks over  an entire growing season.  The staff
notes that there is  significant variability in growth patterns  and  lengths of growing seasons
among the wide range  of vegetation species that may experience adverse effects associated
with O3 exposure,  such that  the selection of any single averaging time for a national standard
will necessarily be a compromise relative to the  range of growing seasons for all vegetation
species of concern.  Based on the information in the CD and in Section VII.D,  staff

-------
                                            279
 concludes that the maximum consecutive three months is a reasonable compromise among the
 various periods of plant sensitivity to O3 identified by vegetation effects researchers.  In most
 cases this maximum three month averaging time would most likely cover the periods of
 greatest plant sensitivity.
        Another feature related to specifying an averaging time is the diurnal window (i.e.,
 the hours during  the day) over which concentrations are cumulated in computing a seasonal
 index.  While studies have shown that by increasing the diurnal window from 7 to 12 or 24
 hours the index captures more of the peak O3 concentrations that occur in some
 environments, the associated reductions in growth or yield and increases in foliar injury have
 not been seen to  increase proportionally with increasing diurnal period.  This observation is
 consistent with the fact that growth and yield reductions are in large part the result of
 decreases in carbohydrate production through photosynthesis which only occurs during
 daylight hours.  Because photosynthesis occurs only during daylight hours, and the majority
 of plants, although not all, have significantly reduced stomatal conductance at night, the staff
-_ believe that the potential for significant impacts from nighttime O3 exposures is very low.
        In considering this aspect of averaging time,  the staff has taken into account studies
 that have evaluated the association between various diurnal windows and  vegetation
 responses,  information about diurnal patterns in stomatal conductance and air quality
 concentrations, and consideration of the relative importance of photosynthetic effects.  Based
 on  these considerations, the staff concludes that a 12-hr diurnal window,  including the
 daylight hours from 8:00am to 8:00pm, is an appropriate diurnal window to use  as a basis
 for a national air quality standard  designed to protect a wide range of vegetation  growing in
 environmental conditions  found across the U.S.
 C.     Form of the Standard
        Defining a mathematical relationship between vegetation response  and  O,  exposure is
 significantly complicated by the many variables that influence the uptake of O3 b\ the plant
 on  any given day at any given hour.  In spite of the large number of studies that have been
 conducted to evaluate effects on vegetation associated with exposure to O3, only a few studies
 have been conducted that  can be used directly to  evaluate the differential effects of specific
 ranges or patterns of O3 concentrations on plant responses.

-------
                                          280
       On the basis of statistical reanalyses of the NCLAN data (which used exposure
regimes with numerous high peaks) to identify exposure indices that best captured the
relationship between O3 concentration and crop responses, several studies concluded that the
cumulative indices that weighted peak concentrations proportionately more than mid-range
concentrations were most effective.  Other studies that employed exposure regimes similar to
NCLAN (numerous high peaks), or encountered them in observational field  studies, have
typically observed a similar relationship between high numbers of peak concentrations and
the presence of foliar injury and/or growth or yield reduction.  On the other hand, effects
have been reported for a variety of plant species when concentrations  never  exceeded those
considered  mid-range (0.05 to 0.09 ppm).
       Since plant response to O3 exposure depends on a multitude of other  factors
controlling stomatal conductance and other regulators  of plant sensitivity, it  is clear that the
relative importance of peak and mid-range concentrations in predicting plant response
depends on  timing.  At present it  can best be said that no one concentration-weighted
exposure index best accounts for the complex relationships between O3 concentrations and
plant responses across a wide range of species.   It is unlikely that this  issue  will be resolved
until new studies are designed and  conducted, or new investigative approaches are used,  to
explicitly evaluate the relative importance of various O3 concentrations in producing effects
across  a  range of species in a variety of environments.
       In light of these unresolved issues, and based on the biological effects information
described in the CD and summarized  in Section VII, the staff draws the following
conclusions:
•      The  staff concludes  that peak  O3 concentrations less than 0.12 ppm but _>_ 0.10 ppm
       can  be phytotoxic to a large number of plant species, and can produce acute foliar
       injury responses, crop yield loss and reduced biomass production.  Thus, staff further
       concludes that the current secondary standard,  which only limits 1-hour peak
       concentrations >  0.12 ppm, is not adequately  protective of vegetation.
Peak O3 concentrations  >_ 0.10 ppm  result from anthropogenic emissions of NOX and VOCs
combined with conducive meteorology, either from emission sources within  an area or from

-------
                                            281
  long-range transport from distant urban centers, and are not attributable to background
  sources.
  •     The staff also concludes that mid-range O3 concentrations (0.05 to 0.09 ppm) have the
        potential over a longer duration of creating chronic stresses on vegetation that can
        lead to effects of concern such as reduced plant growth and yield, shifts in
        competitive advantages  in mixed populations, and decreased vigor leading to
        diminished resistance to pests,  pathogens, and injury from other environmental
        stresses. Furthermore,  some sensitive species can experience foliar injury and growth
        and yield effects even when concentrations never get above the upper end of this
        range.
  Though concentrations  in the upper mid-range also generally result from anthropogenic
- emissions in combination with  appropriate meteorology, it is unclear to what extent
. concentrations in the lower  mid-range result from or are significantly influenced by
  background sources of  NO, and VOCs, and, thus, would be more appropriately regarded as
ithe natural environmental growing conditions for vegetation in some areas.
  •     The staff recognizes that concentrations below 0.05 ppm are  estimated to be within
        the range of background O3 concentrations in the U.S., although a few studies have
        reported growth or yield effects in sensitive species at  or below 0.05 ppm.
  For example, at sea level, annual average background values are estimated to be between
  0.02 and 0.035 ppm, with persistent and episodic natural sources contributing to background
  hourly O3 concentrations in the range of 0.03 - 0.05 ppm. Further,  predictive air quality
  models do not currently have the capability to simulate the air quality conditions that are
  typically related to O3 concentrations in this range.
        With regard to the three specific cumulative, peak-weighted exposure indices that staff
^has  focused on in  this review,  W126,  AOT06, and SUM06, all three indices were  about
  equally good as exposure measures to predict exposure-response relationships reported in the
  NCLAN crop studies.   Since science is currently unable to provide a basis for selecting  one
  form from the other, the staff recommend that the Administrator take into account  additional
  policy considerations in comparing these indices for use as the basis of a  national standard.

-------
                                          282
The staff draws the following conclusions in comparing these indices, based on the
information in Sections IV and VII and building on the general conclusions listed above:
•      The W126 exposure index incorporates a weighting function which gives increasing
       value to all concentrations between 0.00 ppm and 0.10 ppm, with a weight of 1
       applied to all concentrations  > 0.10 ppm.  The staff  concludes that there is
       insufficient evidence at this time to judge the biological relevance of this weighting
       function, especially at the lower concentrations. Further, staff observes that
       concentrations below 0.05 ppm, are within the range  of background O3
       concentrations. Thus, staff concludes that the W126  index is not the preferred index
       form for use in establishing a national standard which is intended to provide an
       appropriate attainment target for air quality management programs designed to reduce
       anthropogenic sources contributing to O3 formation.
•      The AOT06 form equally weights the difference between all concentrations above
       0.06 ppm and the lower boundary concentration of 0.06 ppm.   Though there  is no
       biological evidence of an effects threshold, this index  will not  be influenced by
       background  O3 concentrations and concentrations near 0.06 ppm will contribute  only a
       relatively small proportion of the total index value under many typical  air quality
       distributions.  The functional form of this index inherently results in the upper mid-
       range and peak concentrations contributing proportionately more to the seasonal  value
       of index than for either of the other two indices.  However, precisely because of its
       proportionately greater influence by peak concentrations, this form is somewhat  closer
       to the peak forms of the alternative primary standards recommended in Section VI,
       and, thus, would be more likely to be duplicative of the protection afforded by any of
       the recommended alternative primary standards, while being relatively  less responsive
       to those mid-range concentrations which have also been shown to influence vegetative
       damage. Thus, the  staff conclude that AOT06 is not  the preferred index.
 •     The SUMOb form, like the AOT06, incorporates mid-range and peak concentrations
       above the lower boundary of 0.06 ppm. In contrast to the AOT06 index, however,
       because it cumulates the  full value of all concentrations above 0.06 ppm.  the seasonal
       value of the index is more strongly influenced by the  mid-range concentrations jus! at

-------
                                           283
       and above 0.06 ppm. This would likely result in less duplicative protection from that
       afforded by the recommended alternative primary standards.  Also, this index will not
       be influenced by background O3 concentrations under many typical air quality
       distributions, in contrast to the W126 index.   Thus, the staff concludes that, while
       there is no biological evidence of an effects threshold, the SUM06 index has been
       shown to adequately relate O3 exposures to plant response, and has several policy-
       relevant advantages over both the AOT06 and the W126 as a basis for the form of a
       secondary standard.
       Further, the staff recognizes that any of the recommended alternative primary
standards, when attained at all  locations, would have the effect of lowering not only the peak
O3 concentrations, but also, in  some fashion, the entire air quality distribution.  Thus, taking
into account the above observations and conclusions, the staff further concludes that the
SUM06 index would likely provide a better complement to any of the alternative primary
standards, by better accounting for the vegetation effects associated with exposures within the
mid-range concentrations, without being influenced by background O3 concentrations.
D.     Level of the Standard
       The  level at which a seasonal  secondary standard should be set depends on policy
judgments by the Administrator as 10 the level  of air quality the attainment and maintenance
of which is  requisite to protect the public welfare from any  known  or anticipated adverse
effects associated with the pollutant in the ambient air.  Among the range of vegetation
effects of concern,  some are quantifiable, either economically or non-economically,  while
many can still only be discussed qualitatively.  The following conclusions are based on
consideration of the quantitative economic benefits associated with protection of agricultural
commodity  crops and California fruits and vegetables.  Additionally, consideration is given
to benefits derived from protection of tree seedlings and commercial, Class I, and other
forested areas in urban, rural, and remote environments, as  well as potential impacts on
forested ecosystems.
       Based on studies of agricultural crops, as presented in Section VII.F, a significant
degree of vegetation protection is estimated  to be afforded by the recommended alternative
primary standards, ranging,  for example, from  roughly $1 billion to over $2 billion annually

-------
                                           284
for commodity crops and fruits and vegetables for the 8-hour, 0.08 ppm alternatives.
Additional protection from alternative secondary standards beyond that estimated for the
alternative primary standards is generally estimated to be relatively small, with the degree of
incremental  protection  decreasing as the stringency of the alternative primary standards
increase.  Further,  of course, this incremental protection varies with the stringency of the
alternative secondary standards,  ranging to close to zero in many cases especially for the
least stringent alternative considered.  Thus, staff concludes that, when considered in
conjunction  with the protection provided by both primary  and secondary standards,
consideration of the level for a secondary  standard should focus on the more stringent
alternatives  where some  degree of incremental protection would more likely  be expected.
       More specifically, staff concludes that a secondary standard with  a 3-month, 12-hour,
SUM06 form, set at a  level within the range of approximately 38 - 25 ppm-hr (corresponding
to the 20%  and 10% yield loss protection  levels for 50% of the NCLAN crops,
respectively),  would provide substantial protection against vegetation effects. Though these
levels would not be expected to protect the most sensitive species or individuals within a
species, when considered in conjunction with the recommended alternative primary
standards, such a secondary standard would be expected to provide some degree  of
incremental  protection  beyond that provided by recommended alternative primary standards.
Although  staff judges  that this degree of incremental protection may be  small at the national
level, staff believes that it could be potentially significant  at regional or local levels.
       In  addition to the more extensively studied  effects associated with O3  exposure  such as
visible foliar injury and decreased growth or yield in  short-lived species, the available
information  further points to more subtle impacts of O3  acting in synergy with other natural
and man-made stressors to adversely affect individual plants, populations and whole systems.
By disrupting the photosynthetic process, decreasing carbohydrate storage in  roots, increasing
early senescence of leaves and affecting water use efficiency in trees, O3 exposure can
disrupt or change the nutrient and water flow of an entire system.  Weakened trees can
become susceptible to pest and pathogen outbreaks, loss of competitive advantage and
decreased reproductive (seedling survivability) success,  perhaps resulting in reduced genetic
variability within the species or entire ecosystem.  However, staff concludes  that there  is

-------
                                           285
insufficient information to estimate the severity of these impacts as a function of the levels of
alternative secondary standards, and,  thus, no quantitative estimate has been made in the
context of this review of the potential benefits associated  with alternative standard levels.
This information, however, should be weighed in considering the extent to which a
secondary standard should  be precautionary in nature in protecting  against effects for which
scientific studies have not adequately  quantified  exposure-response  relationships.
E.     Recommendation
       Based on the conclusion that the current  secondary standard does not provide adequate
protection for vegetation, the staff recommends  that the Administrator consider the following
factors in determining appropriate revisions to the secondary standard:  1) the varying
degrees of protection afforded by  the alternative primary  standards recommended in  Section
VI; 2) the incremental protection associated with alternative cumulative, seasonal secondary
standards under consideration;  and 3) the value of establishing a  seasonal form  for the
secondary standard that is more representative of biologically relevant exposures.  Additional
consideration should be given  to the possibility of ozone impacts acting in synergy with other
natural and manmade stressors and the extent to which  a  secondary standard should be
precautionary in nature  against such effects, particularly given their potential  significance  at a
regional scale and in Class I areas.
       If the Administrator determines that additional protection  is  needed beyond that
provided by the alternative primary standards recommended in Section  VI (or that no
revisions to the current  primary standard are warranted) and/or that establishing a seasonal
form  for the secondary  standard is justified, staff recommends consideration of a new
secondary standard in the form of a 3-month, 12-hr, SUM06 exposure  index, set at a level
within the range of approximately 38 to 25 ppm-hrs. The upper end of this range focuses on
providing additional protection against effects to a wide range of commercial  crops and tree
species  that  can be  most clearly attributed  to ambient O3 concentrations above background
levels.  The lower end of this range would be expected to provide  further incremental
protection for commercial crop and tree species,  while  directionally providing increased
protection against effects to vegetation and ecosystem resources in  Class 1 and other areas.

-------

-------
                                         R-l

                                   REFERENCES

Abbey, D.E.; Petersen, F.; Mills, P.K.; Beeson, W.L. (1993)  Long-term ambient
      concentrations of total suspended participates, ozone, and sulfur dioxide and
      respiratory symptioms in a nonsmoking population.  Arch. Environ. Health 48: 33-
      46.

Abeles, F. B.; Morgan, P. W.; Saltveit, M. E., Jr.  (1992) Ethylene in plant biology. 2nd
      ed. San Diego, CA: Academic Press; pp. 414.

Abt Associates, Inc. (1995). Ozone NAAQS benefits analysis: California crops.  Report to
      U.S.  EPA, July.

Abt Associates, Inc. (1993) Urban ornamental plants:  sensitivity to ozone and potential
      economic losses.  Report to U.S. EPA, December

Adams, M. B.; O'Neill, E. G. (1991) Effects of ozone and acidic deposition on carbon
      allocation and mycorrhizal colonization of Pinus meda L. seedlings. For.  Sci. 37:
      5-16.

Adams, R. M. (1986) Agriculture, forestry, and related benefits of air pollution Control: A
      Review and some observations. Am. J.  Agric. Econ. 68:464-472.

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, W. C.; Savin, W. M.; Christo, A. E. (1981) Detection of ozone toxicity during
      continuous exercise via the effective dose concept. J. Appl. Physiol: Respir.  Environ.
      Exercise Physiol. 51: 415-422.

Adaros, G.;  Weigel, H. J.; Jager, H. J. (1991) Die Wirkung von Ozon auf Wachstums- und
      Ertragsparameter zweier Sommerweizensorten (Triticum aestivum L.) [Impact of
      ozone on growth and yield parameters of two spring wheat cultivars (Triticum
      aestivum L.)]. Z. Pflanzenkr. Pflanzenschutz 98: 113-124.

Allegra, L.;  Moavero, N.  E.; Rampoldi, C. (1991)  Ozone-induced  impairment of
      mucociliary transport and its prevention with N-acetylcysteine. Am. J. Med.  91(suppl.
      3C):  67S-71S.

Alscher, R.  G.; Amundson, R. G.; Gumming, J. R.; Fellows,  S.; Fincher, J.: Rubin, G.;
      van Leuken, P.; Weinstein,  L. H. (1989) Seasonal changes in the pigments,
      carbohydrates and growth of red  spruce as affected  by ozone. New Phytol. 113:
      211-223.

-------
                                         R-2

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.

Amdur, M. O.; Ugro, V.; Underbill, D. W. (1978) Respiratory response of guinea pigs to
       ozone alone and with sulfur dioxide. Am. Ind. Hyg. Assoc. J. 39: 958-961.

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.

Andersen, C. P.; Hogsett, W. E.; Wessling, R.; Plocher, M. (1991) Ozone decreases spring
       root growth and root carbohydrate content in ponderosa pine the year  following
       exposure. Can. J. For. Res. 21:1288-1291.

Andersen, C. P.; Rygiewicz, P. T. (1991) Stress interactions and mycorrhizal plant
       response: understanding carbon allocation priorities.  Environ. Pollut. 73: 217-244.

Aranyi, C.; Vana,  S. C.; Thomas, P. T.; Bradof, J.  N.; Fenters, J. D.; Graham, J. A.;
       Miller,  F. J. (1983) Effects of subchronic exposure to a mixture of O3, SO2, and
       (NH4)2SO4 on host defenses of mice. J. Toxicol. Environ. Health 12: 55-71.

Aris, R.M.; Christian, D.; Hearne, P.Q.; Kerr,  K.; Finkbeiner, W.E.;  Balmes, J.R. (1993a)
       Ozone-induced airway inflammation in  human subjects as determined by airway
       lavage and biopsy. Am. Rev. Respir. Dis.  148: 1363-1372.

Aris, R.; Christian, D.;  Tager, I.; Ngo,  I.; Finkbeiner, W.E.; Balmes,  J.R. (1993b)  Effects
       of nitric acid gas alone or in combination with ozone on healthy volunteers. Am.
       Rev.  Respir. Dis. 148: 965-973.

Astrand, P.O.; Rodahl, K. (1977) Textbook of work physiology. New York, N.Y.:
       McGraw-Hill, Inc.

Avol, E. L.; Trim, S. C.; Little, D.  E.;  Spier, C.  E.; Smith, M. N.; Peng, R. C.; Linn, W.
       S.; Hackney, J. D. (1991) Ozone exposure  and lung function:  a southern California
       summer camp study. In:  Berglund, R. L.; Lawson, D. R.; McKee, D. J., eds.
       Tropospheric ozone and the environment: papers from an international conference;
       March  1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
       Association; pp. 90-99.  (A&WMA transaction series no. TR-19).

-------
                                          R-3

 Avol, E. L.; Trim, S. C.; Little, D. E.; Spier, C. E.; Smith, M. N.; Peng, R.-C.; Linn, W.
       S.; Hackney, J. D.; Gross,  K.  B.; D'Arcy, J. B.; Gibbons, D.; Higgins, I. T. T.
       (1990) Ozone exposure and lung function in children attending a southern California
       summer camp. Presented at: 83rd annual meeting and exhibition of the Air & Waste
       Management Association; June; Pittsburgh, PA. Pittsburgh, PA: Air & Waste
       Management Association; paper no. 90-150.3.

 Avol, E.L.; Linn, W.S.; Shamoo,  D.A.; Hackney, J.D.  (1988)  Seasonal ozone activity in
       Los Angeles residents.  Presented at:  81st annual  meeting of the Air Pollution Control
       Association; June; Dallas, TX.   Pittsburgh, PA: Air Pollution Control Association;
       paper no.  88-122.6.

 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Spier, C. E.; Valencia, L. M.; Venet, T. G.;
       Trim, S. C.; Hackney,  J. D. (1987) Short-term respiratory effects of photochemical
       oxidant exposure in exercising  children. JAPCA 37: 158-162.

 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L.  M.; Anzar, U. T.; Venet. T. G.;
       Hackney, J. D.  (1985a) Respiratory effects of photochemical oxidant air pollution in
       exercising  adolescents.  Am. Rev. Respir. Dis. 132: 619-622.

 Avol, E. L.; Linn, W. S.; Shamoo, D. A.; Valencia, L.  M.; Anzar, U. T.; Hackney, J. D.
       (1985b) Short-term health effects of ambient air pollution in adolescents. In: Lee, S.
       D., ed. Evaluation of the scientific basis  for ozone/oxidants standards: transactions of
       an APCA  international  specialty conference; November 1984; Houston, TX.
       Pittsburgh, PA: Air Pollution Control Association; pp. 329-336. (APCA  international
       specialty conference transactions: TR-4).

 Avol, E. L.; Linn, W.  S.; Venet, T. G.; Shamoo, D. A.; Spier, C. E.; Hackney, J.  D.
       (1985c) Comparative effects of laboratory generated ozone and ambient oxidant
       exposure in continuously exercising subjects. In: Lee, S. D., ed. Evaluation of the
       scientific basis for ozone/oxidants standards: proceedings of an APCA international
       specialty conference; November 1984; Houston, TX.  Pittsburgh, PA: Air Pollution
       Control Association; pp. 216-225. (APCA international specialty conference
       transactions: TR-4).

 Avol, E. L.; Linn, W.  S.; Venet,  T. G.; Shamoo, D. A.; Hackney, J. D.  (1984)
       Comparative respiratory effects of ozone and ambient oxidant pollution exposure
       during heavy exercise. J. Air Pollut. Control Assoc. 34: 804-809.

Avol, E. L.;  Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Hackney, J. D.  (1983) Acute
       respiratory effects of Los Angeles smog in continuously exercising adults. J. Air
       Pollut. Control Assoc.  33: 1055-1060.

-------
                                          R-4

Barnes, J. D.; Eamus, D.; Brown, K. A. (1990) The influence of ozone, acid mist and soil
       nutrient status on Norway spruce [Picea abies (L.) Karst.]: II. photosynthesis, dark
       respiration and soluble carbohydrates  of trees during late autumn.  New Phytol. 115:
       149-156.

Barr, B. C.; Hyde,  D. M.; Plopper, C. G.; Dungworth, D. L. (1990) A  comparison of
       terminal airway remodeling in chronic daily versus episodic ozone exposure. Toxicol.
       Appl. Pharmacol. 106: 384-407.

Barr, B. C.; Hyde,  D. M.; Plopper, C. G.; Dungworth, D. L. (1988) Distal airway
       remodeling in rats chronically exposed to ozone. Am. Rev. Respir. Dis.  137: 924-
       938.

Barry, B. E.;  Mercer, R. R.; Miller, F. J.; Crapo, J. D. (1988) Effects of inhalation of 0.25
       ppm ozone on the terminal bronchioles of juvenile and adult rats.  Exp. Lung Res. 14:
       225-245.

Barry, B. E.;  Miller, F. J.; Crapo, J. D. (1985) Effects of inhalation of 0.12 and 0.25 parts
       per million ozone on the proximal alveolar region of juvenile and  adult rats. Lab.
       Invest. 53: 692-704.

Barry, B. E.;  Miller, F. J.; Crapo, J. D. (1983) Alveolar epithelial injury 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).

Bates, D.V.; Baker-Anderson, M.; Sizto, R. (1990)  Asthma attack periodicity: a study of
       hospital emergency visits in Vancouver.  Environ. Res. 51: 51-70.

Beadle, C. L.; Jarvis, P. G.; Talbot, H.; Neilson, R. E. (1985a) Stomatal conductance and
       photosynthesis in a mature Scots pine forest: II.  Dependence on environmental
       variables of single shoots.  J.  Appl. Ecol. 22: 573-586.

Beadle, C. L.; Neilson, R. E.;  Talbot, H.; Jarvis, P. G. (1985b) Stomatal  conductance and
       photosynthesis in a mature Scots pine forest: I. Diurnal, seasonal  and spatial variation
       in shoots. J. Appl.  Ecol. 22:557-571.

Bedi, J. F.; Horvath,  S. M.; Drechsler-Parks, D.  M. (1989) Adaptation  by older individuals
       repeatedly exposed to 0.45 parts per million ozone for two hours. JAPCA 39:  194-
        199.

-------
                                         R-5

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 38: 1016-1019.

Benoit, L. F.; Skelly, J. M.; Moore, L. D.; Dochinger, L.  S. (1982) Radial growth
       reductions in Pinus strobus L. correlated with foliar ozone sensitivity as an indicator
       of ozone-induced losses in eastern forests. Can. J. For. Res. 12:673-678.

Berrang, P.; Karnosky, D. F.; Bennett, J. P. (1991) Natural selection for ozone tolerance in
       Populus tremuloides: an evaluation of nationwide trends.  Can. J. For. Res. 21:
       1091-1097.

Berrang, P.; Karnosky, D. F.; Bennett, J. P. (1989) Natural selection for ozone tolerance in
       Populus tremuloides: field verification. Can. J. For.  Res. 19: 519-522.

Berrang, P.; Karnosky, D. F.; Mickler, R. A.;  Bennet, J. P. (1986) Natural selection for
       ozone tolerance in Populus tremuloides. Can. J. For. Res. 16: 1214-1216.

Berry, C. R.  (1961) White pine emergence tipburn, a physiogenic disturbance. Asheville,
       NC: U.S. Department  of  Agriculture-Forest Service, Southeastern Forest Experiment
       Station, station paper no.  130.

Berry, M.; Lioy, P. J.; Gelperin, K.;  Buckler, G.; Klotz, J. (1991) Accumulated exposure to
       ozone and measurement of health effects in children and counselors at two summer
       camps. Environ. Res. 54: 135-150.

Bhalla, D. K.; Young, C. (1992) Effects of acute exposure to O3 on rats: sequence of
       epithelial and inflammatory changes in  the distal airways. Inhalation Toxicol. 4: 17-
       31.

Biersteker,  K.; Evendijk, J.E. (1976)  Ozone, temperature, and mortality in  Rotterdam in the
       summers of 1974 and 1975.  Environ.  Res.  12: 214-217.

Bleavins, M.  R.; Dziedzic,  D. (1990)  An immunofluorescence study of T and B lymphocytes
       in ozone-induced pulmonary  lesions in  the mouse. Toxicol. Appl. Pharmacol. 105:
       93-102.

Bock,  N.; Lippman, M.; Lioy, P.; Munoz, 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).

-------
                                         R-6

Bohm, M. (1992) Air quality and deposition. In: Olson, R. K.; Binkley, D.; Bohm, M., eds.
       The responses of western forests to air pollution. New York, NY: Springer Verlag;
       pp. 63-152 (Ecological studies no. 97).

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.

Braun-Fahrlander, C.; Kiinzli, N.; Domenighetti, G.; Carell, C. F.; Ackermann-Liebrich,
       U. (1994) Acute effects of ambient ozone on respiratory function of Swiss
       schoolchildren after a 10-minute heavy exercise. Pediatr. Pulmonol. 17:  169-177.

Brunekreef,  B.; Hoek, G.; Breugelmans, O.; Leentvaar, M. (1994)  Respiratory effects of
       low-level photochemical air pollution in amateur cyclists.  Am. J. Respir. Crit. Care
       Med.  150: 962-966.

Bureau of Census,  1990,  "1990 Census of Population and Housing", Summary  Tape File 3A,
       Washington, D.C.

Burleson, G. R.; Keyes, L. L.; Stutzman, J.  D.  (1989) Immunosuppression of pulmonary
       natural killer activity by exposure to ozone. Immunopharmacol. Immunotoxicol. 11:
       715-735.

Burnett, R.T.; Dales, R.E.; Raizenne, M.E.; Krewski,  D.; Summers, P.W.; Roberts, G.R.;
       Raad-Young, M.;  Dann,T.; Brook, J. (1994) Effects of low ambient levels of ozone
       and sulfates on the frequency of respiratory admissions to Ontario hospitals.  Environ.
       Res.  65: 172-194.

Byres, D. P.; Dean, T. J.; Johnson, J. D. (1992) Long-term effects of ozone and simulated
       acid rain on the foliage dynamics of slash pine (Pinus elliotrii var elliottii Engelm.).
       New Phytol. 120:  61-67.

Bytnerowicz, A.; Olszyk, D. M.; Fox,  C. A.; Dawson, P.  J.; Kats, G.; Morrison,  C.  L.;
       Wolf, J. (1988) Responses of desert annual plants to ozone and water stress  in an in
       situ experiment. JAPCA 38: 1145-1151.

Bytnerowicz, A.; Miller,  P. R.; Olszyk, D.  M.; Dawson, P. J.; Fox, C. A. (1987a)
       Gaseous and paniculate air pollution in the San Gabriel Mountains of southern
       California.  Atmos. Environ. 21: 1805-1814.

Bytnerowicz, A.; Miller,  P. R.; Olszyk, D.  M. (1987b) Dry deposition of nitrate,
       ammonium and sulfate to a Ceanothus crassifolius canopy and surrogate  surfaces.
       Atmos. Environ.  21: 1749-1757.

-------
                                          R-7

California Department of Public Health. (1957) Report III: a progress report of California's
       fight against air pollution. Berkley, CA: State of California, Department of Public
       Health.

California Department of Public Health. (1956)  Clean air for California: second report of
       the Air Pollution Study Project.  Berkley, CA: State of California, Department of
       Public Health.

California Department of Public Health. (1955)  Clean air for California: initial report of the
       Air Pollution Study Project. San Francisco, CA: State of California, Department of
       Public Health.

Carr, W.; Zeitel, L.; Weiss, K. (1992) Variations in asthma hospitalizations and deaths in
       New York City.  Am. J. Public Health 82: 59-65.

Castillejos, M.; Gold, D. R.; Dockery, D.; Tosteson, T.; Baum, T.; Speizer, F. E. (1992)
       Effects of ambient ozone on respiratory function and symptoms in Mexico City
       schoolchildren. Am. Rev. Respir.  Dis. 145: 276-282.

Castleman, W. L,; Dungworth, D. L.;  Schwartz, L. W.; Tyler, W. S. (1980) Acute
       respiratory bronchiolitis: an ultrastructural and autoradiographic study of epithelial
       cell injury and renewal  in rhesus monkeys exposed to ozone. Am. J. Pathol.  98: 811-
       840.

Castleman, W. L.; Tyler, W. S.; Dungworth, D. L. (1977) Lesions in respiratory
       bronchioles and  conducting airways of monkeys exposed to ambient levels of ozone.
       Exp. Mol. Pathol. 26: 384-400.

Catalano, P. J.; Rogus, J.; Ryan,  L. M. (1995a) Consequences of prolonged inhalation of
       ozone on  F344/N rats: collaborative studies. Part X: robust composite scores based on
       median polish analysis.  Cambridge, MA: Health Effects Institute; research report no
       65.

Catalano, P. J.; Chang, L.-Y.; Harkema, J. R.; Kaden, D.  A.; Last, J. A.; Mellick, P. W.:
       Parks, W. C.; Pinkerton, K. E.; Radhakrishnamurthy, B.; Ryan, L. M.; Szarek, J.
       L.  (1995b) Consequences of prolonged inhalation of ozone on F344/N rats:
       collaborative studies.  Part XI:  integrative summary. Cambridge, MA: Health Effects
       Institute; research report no. 65.

Chang, L.-Y.; Stockstill,  B.  L.; Menache, M. G.; Mercer, R. R.; Crapo, J. D. (1995)
       Consequences of prolonged inhalation of ozone on F344/N rats: collaborative studies.
       Part VIII. Morphometric analysis of structural alterations in alveolar regions.
       Cambridge, MA: Health Effects Institute; research  report no. 65.

-------
                                         R-8

Chang, L.-Y.; Huang, Y.; Stockstill, B.  L.; Graham, J. A.; Grose, E. C.; Menache, M. G.;
       Miller, F. J.; Costa, D. L.; Crapo, J. D. (1992) Epithelial injury and interstitial
       fibrosis in the proximal alveolar regions of rats chronically exposed to a simulated
       pattern of urban ambient ozone. Toxicol. Appl. Pharmacol.  115: 241-252.

Chang, L.; Miller, F. J.; Ultman, J.; Huang, Y.; Stockstill, B. L.; Grose, E.;  Graham, J.
       A.; Ospital, J. J.; Crapo, J. D. (1991) Alveolar epithelial cell injuries by subchronic
       exposure to low concentrations of ozone correlate with cumulative exposure. Toxicol.
       Appl. Pharmacol. 109: 219-234.

Chappelka, A. H.; Hildebrand,  E.; Skelly, J. M.; Mangis, D.; Renfro, J. R. (1992) Effects
       of ambient ozone concentrations on mature eastern hardwood trees growing in Great
       Smoky Mountains National Park and Shenandoah National Park. Presented at: 85th
       annual meeting of the Air and Waste Management Association; June; Kansas City,
       MO. Pittsburgh,  PA: Air and Waste Management Association; paper no. 92-150.04.

Chappelka, A. H.; Kraemer, M. E.; Mebrahtu, T.; Rangappa,  M.; Benepal, P. S. (1988)
       Effects of ozone  on  soybean  resistance to the Mexican bean beetle (Epilachna
       varivestis Mulsant).  Environ. Exp. Bot. 28: 53-60.

Cherniak, R.M.; Banks, D.E.; Bell, D.Y.; Davis, G.S.; Hughes, J.M.;  King, T.E., Jr.
       (1990) Bronchoalveolar lavage constituents in healthy individuals, idiopathic
       pulmonary fibrosis,  and  selected comparison groups. Am. Rev. Respir.  Dis.
       141(suppl.):S169-S202.

Chow, C. K.; Plopper,  C.  G.; Chiu, M.; Dungworth, D. L. (1981) Dietary vitamin E and
       pulmonary biochemical and morphological alterations of rats exposed to 0.1  ppm
       ozone. Environ.  Res. 24: 315-324.

Chu, S.H. (1995) Meteorological considerations in siting photochemical  pollutant monitors
       Atmospheric Environment, Vol. 29, No. 21,  2905-2913.

Cody, R. P.; Weisel, C. P.; Birnbaum,  G.; Lioy, P. J. (1992) The effect of ozone associated
       with summertime photochemical  smog on the frequency of asthma visits to hospital
       emergency departments. Environ. Res. 58: 184-194.

Coffin, D.L.; Gardner,  D.E. (1972) Interaction  of biological  agents and chemical air
       pollutants.  Ann. Occup. Hyg. 15:219-234.

-------
                                         R-9

 Coffin, D. L.; Blommer, E. J.; Gardner, D. E.; Holzman, R. (1967) Effect of air pollution
       on alteration of susceptibility to pulmonary  infection. In: Proceedings of the 3rd
       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.

 Cooley, D.  R.; Manning, W. J. (1987) The impact of ozone on assimilate partitioning in
       plants: a review. Environ. Pollut. 47: 95-113.

 Costa, D. L.; Tepper, J. S.; Stevens, M. A.; Watkinson, W. P.; Doerfler, D. L.;
       Gelzleichter, T. R.;  Last, J. A. (1995) Restrictive lung disease in rats exposed
       chronically to an urban profile of ozone (O3).  Am. J. Respir. Crit. Care Med. 151:
       1512-1518.

 Costa, D. L.; Kutzman, R.  S.; Lehmann, J. R.; Popenoe, E. A.; Drew, R.  T. (1983) A
       subchronic multidose ozone  study in rats. In: Lee, S. D.; Mustafa, M.  G.; Mehlman,
       M. A., eds.  Internationa] symposium on the biomedical effects of ozone and related
       photochemical oxidants; March 1982; Pinehurst, NC. Princeton, NJ:  Princeton
       Scientific Publishers, Inc.; pp. 369-393. (Advances in modern environmental
       toxicology: v. 5).

 Cowling, D. W.;  Koziol, M. J. (1982) Mineral nutrition and plant response  to air pollutants.
       In: Unsworth, M. H.; Ormrod, D. P., eds.  Effects of gaseous air pollution on
       agriculture and horticulture.  London, United Kingdom: Butterworth Scientific; pp.
       349-375.

 Cox, W.C. and S.H. Chu. (1993) Meteorologically adjusted ozone trends in urban areas:  a
       probabilistic  approach.  Atmospheric Environment, Vol.  27b, No. 4,  Pergamon Press,
       Great Britain, p. 425-434.

 Crapo, J. D.; Barry, B. E.; Chang, L.-Y.; Mercer, R.  R. (1984) Alterations in lung
       structure caused by inhalation of oxidants. J. Toxicol. Environ.  Health  13: 301-321.

 Cregg, B. M.; Halpin, J. E.; Dougherty, P. M.; Testey, R. O.  (1989) Comparative
       physiology and morphology of seedling and  mature forest trees. In: Noble, R. D.;
       Martin, J.  L.; Jensen, K. F., eds. Air pollution effects on vegetation  including forest
       ecosystems. Washington, DC: U.S. Department  of Agriculture, Northeastern
       Experiment Station; pp. 111-118.

Cueto, R.; Squadrito, G. L.; Bermudez, E.; Pryor, W.  A. (1992) Identification of heptana!
       and nonanal in bronchoalveolar lavage from rats exposed to low levels of ozone.
       Biochem.  Biophys. Res. Commun. 188:  129-134.

-------
                                         R-10

Dann, M. S.; Pell, E. J.  (1989) Decline of activity and quantity of ribulose bisphosphate
      carboxylase/oxygenase and net photosynthesis in ozone-treated potato foliage. Plant
      Physiol. 91: 427-432.

Davidson, A. (1993) Update of ozone trends in California's South Coast Air Basin. Air
      Waste 43: 226-227.

Davis, D. D.; Skelly, J. M. (1992a) Growth response of four species of eastern hardwood
       tree seedlings exposed to ozone, acidic precipitation, and sulfur dioxide. J. Air Waste
       Manage.  Assoc. 42: 309-311.

Davis, D. D.; Skelly, J. M. (1992b) Foliar sensitivity of eight eastern hardwood tree species
       to ozone. Water Air Soil Pollut. 62: 269-277.

Davison, A. W.; Barnes, J. D.; Renner, C. J. (1988) Interactions between air pollutants and
       cold stress. In: Schulte-Hostede, S.; Darrall, N.  M.; Blank, L. W.; Wellburn, A. R..
       eds. Air pollution and plant metabolism. London, United Kingdom: Elsevier Applied
       Science; pp. 307-328.

Delfino, R.J.; Becklake, M.R.; Hanley, J.A. (1994a) The relationship of urgent hospital
       admissions for respiratory illnesses to photochemical air pollution levels in Montreal.
       Environ.  Res. 67: 1-19.

Delfino, R.J.; Becklake, M.R.; Hanley, J.A. (1994b)  Estimation  of unmeasured paniculate
       air pollution data for an epidemiological study of daily respiratory morbidity.
       Environ.  Res. 67: 20-38.

Detels,  R.; Tashkin,  D.P.; Sayre, J.W.; Rokaw,  S.N.;  Massey, F.J., Jr.;  Coulson, A.H.;
       Wegman, D.H. (1991) The UCLA population studies of CORD: X.  a cohort stud\  of
       changes in respiratory function associated  with chronic exposure to SOX, NOX, and
       hydrocarbons.  Am. J. Public Health 81: 350-359.

Devlin, R.B.; Folinsbee, L.; Becker, S.; Madden,  M.;  Biscardi, F.; Koren, H.S. (1995)
       Attenuation of cellular and biochemical changes  in the lungs of humans exposed  to
       ozone for five consecutive days.  Manuscript in  preparation.

Devlin, R. B.; McDonnell, W. F.; Mann, R.; Becker,  S.; House, D. E.;  Schreinemachers,
       D.; Koren, H. S. (1991) Exposure of humans to ambient levels of  ozone for 6.6
       hours causes cellullar and biochemical changes in the lung. Am. J. Respir.  Cell Mol.
       Biol. 4: 72-81.

Devlin, R.B.; Koren, H.S. (1990) The use of quantitative two-dimensional gel
       electrophoresis to analyze changes in alveolar macrophage  proteins in  humans exposed
       to ozone.  Am. J. Respir.  Cell Mol. Biol. 2:281-288.

-------
                                         R-ll

Devlin, R.B.; McDonnell, W.F.; Koren, H.S.; Becker, S. (1990) Prolonged exposure of
       humans to 0.10 and 0.08 ppm ozone results in inflammation in the lung.  Presented
       at: 83rd annual meeting of the Air & Waste Management Association; June;
       Pittsburgh, PA. Pittsburgh, PA: Air & Waste Management Association; paper no. 90-
       150.2.

Dockery, D.W.; Schwartz, J.; Spengler, J.D. (1992) Air pollution and daily mortality:
       associations with  particulates and acid aerosols. Environ. Res. 59:362-373.

Drechsler-Parks, D. M.; Horvath, S. M.; Bedi, J. F.  (1990) The "effective dose" concept in
       older adults exposed to ozone. Exp. Gerontol.  25: 107-115.

Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M.  (1989) Pulmonary  function responses
       of young  and older adults to mixtures of O3,  NO3 and PAN. Toxicol. Ind. Health 5:
       505-517.

Drechsler-Parks, D. M.; Bedi, J. F.; Horvath, S. M.  (1987) Pulmonary  function responses
       of older men and women to ozone exposure. Exp. Gerontol. 22: 91-101.

Duchelle, S. F.;  Skelly,  J. M.; Sharick, T. L.; Chevone, B. 1.; Yang, Y.; Nellessen, J.  E.
       (1983) Effects of ozone on the productivity of natural vegetation in  a high meadow of
       the Shenandoah National Park of Virginia. J of Env. Manage. 17:299-308.

Duchelle, S. F.;  Skelly,  J. M.; Chevone,  B. I. (1982) Oxidant effects on forest  tree seedling
       growth in the Appalachian Mountains. J Water, Air, Soil Pollut.  18:363-373.

Duchelle, S. F.;  Skelly,  J. M. (1981) The response of common milkweed to oxidant air
       pollution in the Shenandoah National Park of Virginia.  PI. Dis. 65:661-663.

Dugger, W. M., Jr.; Taylor,  O. C.; Cardiff, E.;  Thompson, C. R. (1962) Relationship
       between carbohydrate content and susceptibility of pinto bean plants to ozone damage.
       Proc. Am. Soc. Hort.  Sci.  81: 304-314.

Duriscoe, D. M.; Stolte, K. W. (1989) Photochemical oxidant injury to ponderosa (Pinus
       ponderosa Dougl. ex Laws) and Jeffrey pine (Pinus jeffreyi Grev. and Balf.) in the
       national parks of the Sierra Nevada of California.  In: Olson,  R. K.; Lefohn,  A. S.,
       eds. Effects of  air pollution on western forests. Pittsburgh, PA: Air and Waste
       Management Association; pp. 261-278 (Transactions series, no. 16).

Dziedzic, D.; Wright,  E. S.;  Sargent, N.  E.  (1990) Pulmonary response to ozone: reaction
       of bronchus-associated lymphoid tissue and lymph node lymphocytes in the rat.
       Environ. Res.  51: 194-208.

-------
                                         R-12

Edwards, G. S.; Kelly, J. M. (1992) Ectomycorrhizal colonization of loblolly pine seedlings
       during three growing seasons in response to ozone, acidic precipitation, and soil Mg
       status. Environ. Pollut. 76: 71-77.

Edwards, G. S.; Friend,  A.  L.; O'Neill, E. G.; Tomlinson, P. T.  (1992) Seasonal patterns
       of biomass accumulation and carbon allocation in Pinus taeda seedlings exposed to
       ozone, acidic precipitation, and reduced soil Mg.  Can.  J. For. Res. 22: 640-646.

Edwards, N. T.  (1991) Root and soil respiration responses to ozone in Pinus taeda L.
       seedlings. New Phytol. 118: 315-321.

Ehrlich, R.; Findlay, J. C.;  Fenters, J. D.; Gardner, D.  E. (1977) Health effects of short-
       term inhalation of nitrogen dioxide and ozone mixtures. Environ. Res. 14:  223-231.

Eissenstat,  D. M.;  Syvertsen, J. P.; Dean, T. J.; Yelenosky, G.; Johnson, J.  D. (1991)
       Sensitivity of frost resistance and growth in citrus and avocado to chronic ozone
       exposure.  New Phytol. 118: 139-146.

Elkiey, T.; Ormrod, D.  P. Pelletier, R. L. (1979) Stomatal and leaf surface features as
       related to the ozone sensitivity of petunia cultivars. J.  Am. Soc. Horde. Sci. 104:
       510-514.

Ernst,  W. H. O.; Tonneijck, A. E. C.; Pasman, F. J. M. (1985) Ecotypic response of Silene
       cucubalus to air pollutants (SO2, O3). J. Plant Physiol.  118: 439-450.

Eschenbacher, W.  L.; Moore, T. B.; Lorenzen, T. J.; Weg,  J. G.; Gross. K.  B. (1992)
       Pulmonary responses  of asthmatic and normal subjects  to different temperature and
       humidity  conditions in an environmental chamber. Lung 170:  51-62.

Euler,  G.L.; Abbey, D.E.; Hodgkin, I.E.; Magie, A.R.  (1988) Chronic obstructive
       pulmonary disease symptom effects of long-term cumulative exposure to ambient
       levels of total oxidants and nitrogen dioxide in California Seventh-day  Adventist
       residents.  Arch. Environ. Health 43: 279-285.

Eustis, S. L.; Schwartz, L. W.; Kosch, P. C.; Dungworth, D. L. (1981) Chronic
       bronchiolitis in nonhuman primates after prolonged ozone exposure. Am. J. Pathol.
       105: 121-137.

Evans, G.  F. (1985) The National Air Pollution Background Network Final  Project Report:
       Project Summary.  Research Triangle Park, NC: Office of Research and
       Development, Environmental Monitoring Systems Laboratory; EPA report  no.
       EPA/600/S4-85/038.

-------
                                         R-13

Evans, G. F.; Finkelstein, P. L.; Martin, B. E.; Possiel, N.C. (1983) Ozone Measurements
       from a Network of Remote Sites. J. Air Pollut. Control Association 33:4.

Farrell, B.P.; Kerr, H.D.; Kulle, T.J.; Sauder, L.R.; Young, J.L. (1979) Adaptation in
       human subjects to the effects of inhaled ozone after repeated exposure. Am. Rev.
       Respir. Dis. 119:  725-730.

Feagans, T.B.; Biller, W.F. (1981)  Describing the protection provided by ambient air
       quality standards.  Env. Prof. 3: 235-247.

Feder, W. A. (1986) Using pollen to predict species response to ozone stress.
       Phytopathology 76: 1097.

Filipowicz, C.; McCauley, R. (1986a) The effect of chronic ozone exposure on  lung
       benzo(a)pyrene oxidase,  benzphetamine demethylase and monoamine oxidase.  Res.
       Commun. Chem. Pathol. Pharmacol. 51: 289-296.

Filipowicz, C.; McCauley, R. (1986b) The effects of chronic ozone exposure on pulmonary
       collagen content and collagen synthesis in rats. J.  Appl. Toxicol. 6: 87-90.

Fincher, J.  (1992)  Comparison of structural  changes in red spruce (Picea rubens Sarg.)
       during cold hardening in mature trees and in seedlings used in pollutant exposure
       studies. For. Ecol. Manage. 51: 105-113.

Folinsbee, L. J.; Devlin,  R.  B.; Biscardi, R.; Robbins, M.; Abdul-Salaam, S.; Koren, H. S.
       (1995) Changes in pulmonary function and breathing pattern with repeated ozone
       exposure in humans. Manuscript in preparation.

Folinsbee, L. J.; Horstman,  D. H.; Kehr). H. R.; Harder. S. D.;  Abdul-Salaam, S.: Ives. P.
       J. (1994) Respiratory responses to repeated prolonged exposure to 0.12 ppm ozone.
       Am. J. Respir. Crit. Care Med. 149: 98-105.

Folinsbee, L. J.; Horstman,  D. H.; Kehrl, H. R.; McDonnell, W. F.; Gerrity, T.  R.; Seal,
       E.; Larson, R.; Hazucha, M. J.; Abdul-Salaam, S.; Faucette,  B.; Ives, P. J. (1991)
       Effects of single and repeated prolonged low-level ozone exposure in man.  Presented
       at: annual meeting of the Society for  Occupational and Environmental Health; March;
       Washington, DC.

Folinsbee, L. J.; McDonnell, W. F.; Horstman,  D. H. (1988) Pulmonary function and
       symptom responses after 6.6-hour exposure to 0.12 ppm ozone with moderate
       exercise. JAPCA 38:  28-35.

Folinsbee, L. J.; Horvath, S. M. (1986) Persistence of the acute effects of ozone exposure.
       Aviat. Space Environ. Med. 57: 1136-1143.

-------
                                        R-14

Folinsbee, L. J.; Bedi, J. F.; Horvath, S. M. (1984) Pulmonary function changes after 1 h
       continuous heavy exercise in 0.21 ppm ozone. J. Appl. Physiol.: Respir. Environ.
       Exercise Physiol. 57: 984-988.

Folinsbee, L. J.; Drinkwater, B. L.; Bedi, J.  F.; Horvath, S. M. (1978) The influence of
       exercise on the pulmonary function changes due to exposure to low concentrations of
       ozone. In: 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.

Foster, W. M.; Costa, D. L.; Langenback, E. G. (1987) Ozone exposure alters
       tracheobronchial mucociliary function in humans.  J. Appl. Physiol. 63: 996-1002.

Frampton, M. W.; Morrow, P. E.; Gibb, F.  R.; Speers, D. M.; Gavras, J.  B.; Utell, M. J.
       (1993) Airways responsiveness to ozone in smokers and nonsmokers. Am. Rev.
       Respir. Dis.  147: A636.

Friend, A.  L.; Tomlinson, P. T. (1992)  Mild ozone exposure alters 14C dynamics in foliage
       of Pinus taeda L. Tree Physiol. 11:215-227.

Friend, A.  L.; Tomlinson, P. T.; Dickson, R. E.;  O'Neill, E. G.; Edwards,  N. T.; Taylor,
       G. E., Jr.  (1992)  Biochemical composition of loblolly pine reflects pollutant
       exposure.  Tree Physiol.  11:35-47.

Fiihrer, J.; Grimm,  A. G.; Tschannen, W.; Shariat-Madari, H. (1992) The response of
       spring wheat (Triticum aesrivum L.) to ozone at higher elevations: II. changes in
       yield, yield components and grain quality in response to ozone flux.  New Phytol.
       121: 211-219.

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.

Gerrity, T.R.; Bennett, W.D.; Kehrl, H.; DeWitt,  P.J. (1993)  Mucociliary  clearance of
       inhaled panicles measured at 2 h after ozone exposure in humans. J. Appl. Physiol.
       74:  2984-2989.

Gibbons, S. I.; Adams, W.  C. (1984) Combined effects of ozone exposure and ambient heat
       on exercising females. J.  Appl. Physiol.: Respir. Environ. Exercise Physiol. 57: 450-
       456.

Gillespie, C. T.; Winner, W. E. (1989) Development of lines of radish  differing in
       resistance to O3 and SO2. New Phytol. 112: 353-361.

-------
                                         R-15

Gilmour, M. I.; Park, P.; Selgrade, M. K. (I993a) Ozone-enhanced pulmonary infection
       with Streptococcus zooepidemicus in  mice. Am. Rev. Respir. Dis. 147: 753-760.

Gilmour, M. I.; Park, P.; Doerfler, D.; Selgrade,  M. K. (1993b) Factors that influence the
       suppression of pulmonary antibacterial defenses in mice exposed to ozone. Exp. Lung
       Res. 19: 299-314.

Gilmour, M.I.; Selgrade, M.K. (1993)  A comparison of the pulmonary defenses against
       streptococcal infection in rats and mice following O3 exposure:  differences in disease
       susceptibility and neutrophil recruitment.  Toxicol. Appl. Pharmacol.  123:211-218.

Gilmour, M. I.; Hmieleski, R. R.; Stafford, E. A.; Jakab, G. J.  (1991) Suppression and
       recovery of the alveolar macrophage  phagocytic system during continuous exposure to
       0.5 ppm ozone.  Exp. Lung Res. 17:  547-558.

Gliner, J. A.; Horvath, S. M.; Folinsbee, L. J. (1983) Preexposure to low ozone
       concentrations does not diminish the pulmonary function response on exposure to
       higher ozone concentrations. Am. Rev. Respir. Dis. 127:  51-55.

Goldstein, B.D. et a].,  1992, "Valdez Air Health Study," Alyeska Pipeline Service
       Company,  Anchorage, Alaska, June.

Gong, H., Jr. (1987) Relationship between air quality and the respiratory status  of asthmatics
       in an area  of high oxidant pollution in Los Angeles County  [final report]. Sacramento,
       CA: California Air Resources Board; contract  nos. Al-151-33 and A4-135-33.

Gong, H., Jr.; Bradley, P. W.; Simmons, M.  S.; Tashkin, D.  P. (1986) Impaired exercise
       performance and pulmonary function in elite cyclists during low-level ozone exposure
       in a hot environment. Am. Rev. Respir. Dis. 134: 726-733.

Grandjean,  A.; Grimm, A. G.; Fiihrer, J. (1992a)  The response of spring wheat (Triticum
       aestivum L.) to ozone at higher elevations: I. Measurement of ozone and carbon
       dioxide fluxes in open-top field chambers. New Phytol. 121: 201-210.

Grandjean,  A.; Grimm, A. G.; Fiihrer, J. (1992b)  The response of spring wheat (Triticum
       aestivum L.) to ozone at higher elevations: III. Responses of leaf and  canopy gas
       exchange,  and chlorophyll fluorescence to ozone flux. New Phytol. 122: 321-328.

-------
                                         R-16

Grose, E. C.; Stevens, M. A.; Hatch, G. E.; Jaskot, R. H.; Selgrade, M. J. K.; Stead, A.
       G.; Costa, D. L.; Graham, J. A. (1989) The impact of a 12-month exposure to a
       diurnal pattern of ozone on pulmonary function, antioxidant biochemistry and
       immunology. In: Schneider, T.; Lee, S.  D.; Wolters, G. J. R.; Grant, L. D., eds.
       Atmospheric ozone research and its policy implications: proceedings of the 3rd
       US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
       Amsterdam, The Netherlands: Elsevier Science Publishers;  pp. 535-544. (Studies in
       environmental science 35).

Gross, K. B.; White, H. J.; Sargent, N. E. (1991) The effect of ozone inhalation on
       metabolic functioning of vascular endothelium and on ventilatory function. Toxicol.
       Appl. Pharmacol. 109: 336-351.

Hackney, J. D.; Linn,  W. S.; Shamoo, D. A.;  Avol, E. L. (1989) Responses of selected
       reactive and nonreactive volunteers to ozone exposure in high and low pollution
       seasons. In: Schneider, T.; Lee, S. D.; Wolters, G.  J. R.;  Grant, L. D., eds.
       Atmospheric ozone research and its policy implications: proceedings  of the 3rd  US-
       Dutch international symposium; May 1988;  Nijmegen, The Netherlands. Amsterdam.
       The Netherlands: Elsevier Science Publishers; pp. 311-318. (Studies  in environmental
       science 35).

Hackney, J. D.; Linn,  W. S.; Mohler, J. G.; Collier, C. R. (1977a) Adaptation to  short-
       term respiratory effects of ozone in men exposed repeatedly. J. Appl. Physiol.:
       Respir. Environ. Exercise Physiol. 43: 82-85.

Hackney, J. D.; Linn, W. S.; Karuza, S. K.; Buckley, R. D.; Law. D. C.; Bates,  D.  V.;
       Hazucha,  M.; Pengelly,  L. D.; Silverman, F. (1977b) Effects of ozone exposure in
       Canadians and southern Californians: evidence for adaptation? Arch. Environ. Health
       32: 110-116.

Hackney, J. D.; Linn, W. S.; Buckley,  R.  D.;  Hislop, H. J. (1976) Studies in  adaptation to
       ambient oxidant air pollution: effects of ozone exposure in  Los Angeles residents vs.
       new arrivals. Environ. Health Perspect.  18: 141-146.

Hain, F.  P. (1987) Interactions of insects, trees and air pollutants. Tree Physiol. 3:  93-102.

Harkema, J. R.; Mauderly,  J. L. (1994) Consequences of prolonged inhalation of ozone on
       F344/N rats: collaborative studies, part V: effects on pulmonary function. Cambridge,
       MA:  Health Effects Institute; research report  no. 65.

Harkema, J. R.; Morgan, K. T.; Gross, E. A.; Catalano, P. J.; Griffith, W. C. (1994)
       Consequences of prolonged inhalation of ozone  on F344/N rats: collaborative studies.
       Part VII:  effects on the nasal mucociliary apparatus. Cambridge, MA: Health Effects
       Institute; research report no. 65.

-------
                                        R-17

Harkema, J. R.; Plopper, C. G.; Hyde, D. M.; St. George, J. A.; Wilson, D. W.;
       Dungworth, D. L. (1993) Response of macaque bronchiolar epithelium to ambient
       concentrations of ozone.  Am. J. Pathol.  143: 857-866.

Harkema, J. R.; Hotchkiss, J. A.;  Henderson,  R. F. (1989) Effects of 0.12 and 0.80 ppm
       ozone on rat nasal and nasopharyngeal epithelial mucosubstances: quantitative
       histochemistry. Toxicol. Pathol. 17: 525-535.

Hartwell, T.D.  et al.,  1984, "Study of Carbon Monoxide Exposure of Residents of
       Washington, D.C., and Denver, Colorado" EPA-600/54-84-031, U.S. Environmental
       Protection Agency, Research Triangle Park, North Carolina, March.

Hayes, E. M.; Skelly, J. M. (1977) Transport  of ozone from the northeast U.S. into Virginia
       and its effect on eastern white pines. Plant Dis. Rep. 61: 778-782.

Hayes, S. R.; Rosenbaum, A. S.; Wallsten, T. S.;  Whitfield, R. G.; Winkler,  R.L.;
       Richmond, H. (1989) A health risk assessment  for use in setting the U.S. primary
       ozone standard. In: Schneider, T.; Lee, S. D.;  Wolters, G. J. R.;  Grant, L. D.: eds.
       Atmospheric ozone research and its policy implications: proceedings of the 3rd  US-
       Dutch international symposium; May 1988; Nijmegen, The Netherlands. Amsterdam,
       The Netherlands: Elsevier Science Publishers; pp. 851-867. (Studies in environmental
       science 35).

Hayes, S.R.; Rosenbaum, A.S.; Wallsten, T.S.; Whitfield, R.G.; Winkler, R.L. (1987)
       Assessment of lung function and symptom health risks associated with attainment  of
       alternative ozone NAAQS, Systems Applications, Inc., San Rafael, CA, draft final
       report No. SYSAPP-87-171.  (For copies, contact Harvey  M. Richmond, U.S.
       Environmental Protection Agency, OAQPS, MD-15, Research Triangle Park, N.C.
       27711, (919)541-5271.)

Hazucha, M.J.; Madden, M.; Pape, G.; Becker, S.; Devlin, R.B.; Koren, H.S.; Kehrl, H.:
       Bromberg, P. A. (1995) Effects of cyclooxygenase inhibition on ozone-induced
       respiratory inflammation  and lung function changes.  Eur.  J. Appl. Physio). Occup.
       Med.: in press.

Hazucha, M. J.; Folinsbee, L. J.; Seal, E., Jr. (1992) Effects of steady-state and variable
       ozone concentration profiles on pulmonary function. Am. Rev. Respir. Dis. 146:
       1487-1493.

Hazucha, M. J. (1987) Relationship between ozone exposure and pulmonary function
       changes. J. Appl. Physiol. 62:  1671-1680.

Heagle, A. S.; McLaughlin, M.  R.; Miller, J.  E.; Joyner,  R. L.; Spruill, S. E. (1991)
       Adaptation of a white clover population to ozone stress. New Phytol. 119:  61-68.

-------
                                         R-18

Heagle, A. S.; Kress, L. W.; Temple, P. J.; Kohut, R. J.; Miller, J. E.; Heggestad,  H. E.
       (1988a) Factors influencing ozone dose-yield response relationships in open-top field
       chamber studies. In: Heck,  W. W.; Taylor, O. C.; Tingey, T. D. eds. Assessment of
       crop loss from air pollutants: proceedings of an international conference; October,
       1987, Raleigh, NC. New York, NY: Elsevier Applied Science; pp. 141-179.

Heagle, A. S.; Miller, J. E.; Heck, W. W.; Patterson, R. P. (1988b) Injury and yield
       response of cotton to chronic doses of ozone and soil moisture deficit. J. Environ.
       Qual. 17: 627-635.

Heagle, A. S.; Heck, W. W.; Lesser, V. M.; Rawlings, J. O. (1987) Effects of daily ozone
       exposure duration and concentration fluctuation on yield of tobacco. Phytopathology
       77: 856-862.

Heagle, A. S.; Letchworth, M. B. (1982) Relationships among injury, growth,  and yield
       responses of soybean cultivars  exposed to ozone at different light intensities. J.
       Environ.  Qual.  11:690-694.

Heagle, A. A.; (1979)  Effects of growth media, fertilizer rate and hour and season of
       exposure on sensitivity of four soybean cultivars to ozone.  Environ. Pollut.  18: 313-
       322.

Heath, R. L. (1988) Biochemical mechanisms of pollutant stress.  In: Heck, W. W.;  Taylor,
       O. C.; Tingey, D.  T. eds. Assessment of crop loss from air pollutants:  proceedings
       of an international conference;  October,  1987, Raleigh, NC. New York, NY:  Elsevier
       Applied Science; pp. 259-286.

Hechter, H.H.; Goldsmith, J.R. (1961) Air pollution and daily mortality. Am.  J.  Med. Sci.
       241: 581-588.

Heck, W. W.; Heagle, J.  E.; Miller,  J. E.  (1994)  Factors affecting plant response to ozone:
       an overview.  In: Kuttler, W.; Jochimsen, M.; eds. Immissionsokologische  Forschung
       im Wandel der Seit Festschrift fur Robert Guderian. Westarp Wissenschaften,
       Uhlichstrasse; pp. 113-130.

Heck, W. W.; Cure, W. W.; Rawlings, J. O.;  Zaragoza, L. J.; Heagle, A. S.; Heggestad,
       H. E.; Kohut, R. J.; Kress, L. W.; Temple, P. J. (1984) Assessing impacts of ozone
       on agricultural crops:  II. crop  yield functions and  alternative exposure statistics. J.
       Air Pollut. Control Assoc.  34: 810-817.

Heggestad,  H. E.; Anderson, E. L.; Gish, T. J.; Lee, E. H. (1988) Effects of ozone and soil
       water deficit  on roots and shoots of field-grown soybeans.  Environ. Pollut. 50:
       259-278.

-------
                                         R-19

Held, A. A.; Mooney, H. A.;  Gorham, J. N.  (1991) Acclimation to ozone stress in radish:
       leaf demography and photosynthesis.  New Phytol. 118:417-423.

Henderson, R. F.; Hotchkiss, J. A.; Chang, I. Y.; Scott, B. R.; Harkema, J.  R.  (1993)
       Effect of cumulative exposure on nasal response to ozone. Toxicol. Appl.
       Pharmacol. 119: 59-65.

Herstrom, A., W. Hogsett, D. Tingey,  E. Lee, and D. Phillips (1995).  Using a
       geographical information system to estimate ozone exposure over forests across the
       United States (Submitted for publication).

Higgins, I. T. T.;  D'Arcy, J. B.; Gibbons, D. I.; Avol, E. L.; Gross, K. B. (1990) Effect
       of exposures to ambient ozone on ventilatory lung function in children.  Am. Rev.
       Respir. Dis. 141:  1136-1146.

Highfill, J. W.; Costa, D. L.  (1994) Statistical response-models for ozone exposure: their
       generality, when applied to human spirometric and animal permeability  functions of
       the lung. J. Air Waste Manage. Assoc. 45: 95-102.

Highfill, J. W.; Hatch, G. E.; Slade, R.; Crissman, K. M.; Norwood, J.; Devlin, R. B.;
       Costa, D. L. (1992) Concentration-time models for the effects of ozone on
       bronchoalveolar lavage fluid protein from rats and guinea pigs.  Inhalation Toxicol. 4:
       1-16.

Hiroshima. K.; Kohno, T.; Ohwada, H.; Hayashi, Y. (1989) Morphological study of the
       effects of ozone on rat lung: II. long-term exposure. Exp. Mol. Pathol.  50: 270-280.

Hodgkin, J.E.; Abbey, D.E.; Euler, G.L.; Magie. A.R. (1984) COPD prevalence in
       nonsmokers in  high and low photochemical air pollution areas. Chest 86: 830-838.

Hoek,  G.; Brunekreef, B.; Kosterink, P.; Van den Berg,  R.; Hofschreuder, P. (1993a)
       Effect of ambient ozone on peak expiratory flow of exercising children  in the
       Netherlands. Arch. Environ. Health 48: 27-32.

Hoek,  G.; Fischer, P.; Brunekreef, B.;  Lebret, E.; Hofschreuder, P.; Mennen, M. G.
       (1993b) Acute  effects of ambient ozone on pulmonary function of children in the
       Netherlands. Am.  Rev.  Respir. Dis.  147: 111-117.

Hogsett, W. E.; Herstrom, A.  A.; Laurence, J. A.; Lee, E. H.; Weber, J. E.; Tingey, D.
       T. (1995)  Risk characterization of tropospheric ozone to forests: in Comparative
       Risk Analysis and  Priority Setting for Air Pollution Issues.  Proceedings of the 4th
       U.S.-Dutch International Symposium.  Pittsburgh, PA. Air and Waste Management
       Association. 119-145.

-------
                                        R-20

Hogsett, W. E.; Tingey, D. T.; Hendricks, C.; Rossi, D. (1989) Sensitivity of western
       conifers to SO2 and seasonal interaction of acid fog and ozone. In: Olson, R. K.;
       Lefohn, A.  S., eds. Transactions: effects of air pollution on western forests; June;
       Anaheim, CA. Pittsburgh, PA: Air Pollution Control Associations; pp. 469-491.

Hogsett, W. E.; Plocher, M.; Wildman, V.; Tingey, D. T.; Bennett, J. P. (1985a) Growth
       response of two varieties of slash pine seedlings to chronic ozone exposures. Can. J.
       Bot. 63: 2369-2376.

Hogsett, W. E.; Tingey, D. T.; Holman, S. R. (1985b) 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.; 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: Lee, S. D., ed.  Evaluation of the scientific
       basis for ozone/oxidants standards: proceedings of an APCA international specialty
       conference; November 1984;  Houston, TX. Pittsburgh, PA: Air Pollution Control
       Association; pp. 262-280. (APCA international specialty conference transactions: TR-
       4).

Horst, R.; Duff, M.  (1995a).   Concentration data transformation and the quadratic rollback
       methodology (Round 2, Revised).  Unpublished memorandum to R. Rodriguez, U.S.
       EPA,  June 8.

Horst, R.  (1995b).  Transmission of full set of data for RMF  analysis.  Unpublished
       memorandum to R. Rodriguez, U.S. EPA, November  6.

Horstman, D.H.;  Ball, B.A.;  Folinsbee, L.J.; Brown, J.; Gerrity, T. (1995) Comparison of
       pulmonary responses of asthmatic and nonasthmatic subjects performing light exercise
       while exposed to a low level  of ozone.  Toxicol. Ind. Health: in press.

Horstman, D.H.;  Folinsbee, L.J.; Ives, P.J.; Abdul-Salaam, S.; McDonnell, W.F. (1990)
       Ozone concentration and pulmonary response relationships for 6.6-hour exposures
       with five hours of moderate exercise to 0.08, 0.10, and 0.12 ppm. Am. Rev.  Respir.
       Dis.  142: 1158-1163.

Horvath, S. M.; Bedi, J. F.; Drechsler-Parks, D. M.; Williams, R. E.  (1991) Alterations in
       pulmonary function parameters during exposure to 80 ppb ozone for 6.6 hours in
       healthy middle aged individuals. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J.,
       eds. Tropospheric ozone and  the environment: papers from an international
       conference; March 1990; Los Angeles, CA.  Pittsburgh, PA: Air & Waste
       Management Association; pp. 59-70. (A&WMA  transaction series no. TR-19).

-------
                                         R-21

Horvath, S. M.; Gliner, J. A.; Folinsbee, L. J. (1981) Adaptation to ozone: duration of
       effect. Am. Rev. Respir. Dis. 123: 496-499.

Hotchkiss,  J. A.; Harkema, J. R.; Sun, J. D.; Henderson, R. F. (1989a) Comparison of
       acute ozone-induced nasal and pulmonary inflammatory responses in rats. Toxicol.
       Appl. Pharmacol. 98: 289-302.

Hotchkiss,  J. A.; Harkema, J. R.; Kirkpatrick, D. T.; Henderson, R. F. (1989b) Response
       of rat alveolar macrophages to ozone: quantitative assessment of population size,
       morphology, and proliferation following acute exposure. Exp. Lung Res. 15:  1-16.

Houston, D. B.; Dochinger, L. S. (1977) Effects of ambient air pollution on cone, seed,
       and pollen characteristics in eastern white and red pines.  Environ. Pollut. 12: 1-5.

Howitt, R.  E. (1995a).  Calibration methods for agricultural economic production models.
       Journal of Agricultural Economics.  Vol 46(2): 147-159.

Howitt, R.  E. (1995b).  Positive mathematical programming. American Journal of
       Agricultural Economics.  Vol 77(2) starts p. 329.

Hyde,  D. M.; Hubbard, W.  C.; Wong, V.;  Wu, R.; Pinkerton, K.; Plopper, C. G. (1992)
       Ozone-induced acute tracheobronchial epithelial injury: relationship to granulocyte
       emigration in  the lung. Am. J. Respir.  Cell Mol. Biol. 6: 481-497.

Jacobson, J. S.  (1977) The effects of photochemical oxidants on vegetation. In: Ozon  und
       Begleitsubstanzen im photochemischen  Smog: das Kolloquium [Ozone and related
       substances in photochemical smog: the colloquium]; September 1976; Dusseldorf,
       Federal Republic of Germany. Dusseldorf, Federal Republic of Germany:  VDI-Verlag
       GmbH; pp. 163-173.  (VDI-Berichte nr. 270).

Jakab,  G. J.; Spannhake, E.  W.;  Canning, B.  J.;  Kleeberger, S. R.; Gilmour, M.  I. (1995)
       The effects of ozone on immune function.  Environ. Health Perspect.  103 (suppl. 2):
       77-89).

Jakab,  G. J.; Bassett, D. J.  P. (1990) Influenza virus infection, ozone exposure, and
       fibrogenesis. Am. Rev. Respir. Dis.  141: 1307-1315.

Johnson, D. W.  ; Taylor, G. E. (1989) Role of air pollution in forest decline in eastern
       North America.   Water Air Soil Pollut. 48: 21-43.

Johnson, T.R. (1995)  Letter Report:  Evaluation of air quality adjustment procedure for
       pNEM/Oj.  Prepared  by IT/Air Quality Services for U.S. EPA,  OAQPS; Research
       Triangle Park, NC, August 4.   (For copies, contact Harvey M. Richmond, OAQPS,
       MD-15,  Research Triangle Park, NC 27711, (919) 541-5271).

-------
                                        R-22

Johnson, T. R. (1994) Letter Report: Enhancements to the pNEM summer camp
       methodology. Prepared by IT/Air Quality Services for U.S. EPA, OAQPS; Research
       Triangle Park, NC, March 21. (For copies, contact Harvey M. Richmond, U.S.
       Environmental Protection Agency, OAQPS, MD-15, Research Triangle Park, N.C.
       27711, (919)541-5271.)

Johnson, T.R. (1987)  A study of human activity patterns in Cincinnati, Ohio, prepared by
       PEI Associates, Inc. for Electric Power Research Institute, Palo Alto, available from
       Ted Johnson, IT Corporation, 3710 University Drive, Durham, North  Carolina
       27707.

Johnson, T.R. (1984)  A Study of Personal Exposure to Carbon Monoxide in Denver,
       Colorado," EPA-600/54-84-014, U.S. Environmental Protection Agency,  Research
       Triangle Park, North Carolina, March.

Johnson, T; Capel, J.; Mozier,  J.; McCoy, M. (1996a).  Estimation of ozone exposures
       experienced by outdoor children in nine urban areas using a probabilistic version of
       NEM. Prepared by IT/Air Quality Services for U.S. EPA, OAQPS; Research
       Triangle Park, NC, August.

Johnson, T.; Capel, J.; McCoy, M.; Mozier, J. (1996b) Estimation of ozone  exposures
       experienced by outdoor workers in nine urban areas using a probabilistic version of
       NEM. Prepared by IT/Air Quality Services for U.S. EPA, OAQPS; Research
       Triangle Park, NC, August.

Johnson, T.; Capel, J.; McCoy, M. (1994).  Estimation of ozone exposures experienced  by
       urban residents using a probabilistic version of NEM and  1990 population data.
       Prepared by IT/Air Quality Services for U.S. EPA, OAQPS; Research Triangle Park,
       NC, December 1994.

Jones,  K.H. (1992) "The  1990/91/92 Ozone data Base and Its Implications Relative to
       Currently Designated Ozone Nonattainment Area Regulatory Programs," in
       Transactions of Tropospheric Ozone: Nonattainment and Design Value Issues. Boston,
       MA.

Jones,  K.H. (1985) "Urban Air Quality,"  in Environmental Quality 1984.  Council on
       Environmental Quality, Washington,  D.C.

Karnosky, D. F. (1989) Air-pollution induced population  changes in North American forests.
       In: Bucher, J.  B.;  Bucher-Wallin, I., eds. Air pollution and forest decline; v. I; pp.
       315-317.

Karnosky, D. F. (1981) Changes in eastern white pine stands related  to ozone stress. Mitt.
       Forstl. Bundesversuchsanst. Wien  137: 41-45.

-------
                                         R-23

Kearns, E. V.; Assmann, S.M. (1993) The guard cell-environment connection. Plant
       Physiol. 102:711-715.

Keller, T. (1988) Growth and premature leaf fall in American aspen as bioindications for
       ozone.  Environ. Pollut. 52:  183-192.

Kelly, N.A.; Wolff, G.T.; and Ferman, M.A. (1984) Sources and sinks of ozone in rural
       areas.  Atmospheric Environment 18:  1251-1266.

Kelly, N.A.; Wolff, G.T.; Ferman, M.A. (1982)  Background pollutant measurements in air
       masses  affecting the eastern  half of the United States-I. air masses arriving  from the
       northwest. Atmos. Environ.  16:  1077-1088.

Kilburn,  K. H.; Warshaw, R. H.; Thornton,  J. C. (1992) Expiratory flows decreased in Los
       Angeles Children from  1984 to 1987:  is this evidence of effects of air pollution?
       Environ. Res. 59: 150-158.

King, T.E., Jr. (1993)  Bronchiolitis.  In: Schwarz, M.I.; King, T.E., Jr., eds.  Interstitial
       lung disease.  2nd ed. St. Louis, MO: Mosby Year Book; pp. 463-495.

Kinney, P. L.;  Ozkaynak,  H. (1991) Associations of daily mortality and air pollution in Los
       Angeles County.  Environ. Res. 54: 99-120.

Kinney, P. L.;  Ware, J. H.; Spengler, J. D.; Dockery, D. W.; Speizer, F. E.; Ferris, B.
       G., Jr.  (1989) Short-term pulmonary function change in association with ozone levels.
       Am. Rev.  Respir. Dis.  139:  56-61.

Koenig, J.  Q.; Covert, D.  S.; Smith,  M. S.;  Van Belle, G.; Pierson, W. E. (1988) The
      pulmonary effects of ozone and nitrogen dioxide alone and combined in healthy and
       asthmatic adolescent subjects. Toxicol. Ind. Health 4: 521-532.

Koenig, J.  Q.; Covert, D.  S.; Marshall, S. G.; Van Belle, G.;. Pierson, W.  E. (1987) The
       effects of ozone and nitrogen dioxide on pulmonary function in healthy and in
      asthmatic adolescents. Am. Rev. Respir. Dis. 136:  1152-1157.

Koenig, J.  Q.; Covert, D.  S.; Morgan, M. S.; Horike, M.; Horike, N.; Marshall, S.  G.;
      Pierson, W. E. (1985) Acute effects of 0.12 ppm ozone or 0.12 ppm nitrogen dioxide
      on pulmonary function in healthy and asthmatic adolescents. Am. Rev. Respir. Dis.
       132: 648-651.

Kolaz, D.J. and R.L. Swinford (1990) "How to Remove the Influence of Meteorology  from
      the Chicago Area Ozone Trend," presented  at the Air and Waste Management
      Association 83rd  Annual Meeting, Pittsburgh, PA.

-------
                                        R-24

Kopp, R.; Vaughn, W.; Hazilla, M.; Carson, R. (1985) Implications of environmental policy
      for U.S. agriculture: the Case of ambient ozone standards.  J. Environ. Econ.
      Manage. 20:321-31.

Koren, H. S.; Devlin, R. B.; Becker, S.; Perez, R.; McDonnell, W. F. (1991) Time-
      dependent changes of markers associated with inflammation in the lungs of humans
      exposed to ambient levels of ozone.  Toxicol. Pathol. 19:  406-411.

Koren, H. S.; Devlin, R. B.; Graham, D. E.;  Mann, R.;  McGee, M. P.; Horstman, D. H.;
      Kozumbo, W. J.; Becker, S.; House, D. E.; McDonnell, W. F.; Bromberg,  P.  A.
      (1989a) Ozone-induced inflammation in the lower airways of human subjects. Am.
      Rev. Respir. Dis.  139: 407-415.

Koren, H. S.; Devlin, R. B.; Graham, D. E.;  Mann, R.;  McDonnell, W.  F.  (19895) The
      inflammatory response in human lung 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 3rd US-Dutch  international
       symposium; May 1988; Nijmegen, The Netherlands. Amsterdam, The Netherlands:
       Elsevier Science Publishers; pp. 745-753. (Studies in environmental science 35).

Kormanik, P. P.; Ruehle, J. L.; Muse, H.  D.  (1990)  Frequency distribution and heritability
       of first-order lateral roots in loblolly pine seedlings.  Forest  Sci. 36(3):802-814.

Kormanik, P. P. (1986)  Lateral root morphology as an expression  of sweetgum  seedling
       quality. Forest Sci. 32(3) 595-604.

Kostka-Rick, R.; Manning, W. J. (1992a) Partitioning of biomass and carbohydrates in field-
       grown radish under ambient concentrations of ozone, and treated with the anti-ozonant
       ethylene-diurea (EDU). New Phytol. 121: 187-200.

Kostka-Rick, R.; Manning, W. J. (19925) Effects and interactions of ozone and the anti-
       ozonant EDU at different stages of radish  (Raphanus sativus L.) development. J. Exp
       Bot43: ;1621-1631.

Krause,  G.  H. M.; Riley, W. D.; Feder, W. A. (1975) Effects  of ozone on petunia and
       tomato pollen tube elongation in vivo. Proc.  Am. Phytopathol. Soc. 2: 100.

 Kreit, J. W.; Gross, K.  B.; Moore, T. B.; Lorenzen, T.  J.; D'Arcy, J.; Eschenbacher, W.
       L. (1989) Ozone-induced changes in pulmonary function and bronchial responsiveness
       in asthmatics. J.  Appl. Physiol. 66: 217-222.

 Kress, L. W.; Allen,  H. L.; Mudano, J. E.; Stow, T. K. (1992) Impact of ozone on loblolly
       pine seedling foliage production and retention. Environ. Toxicol. Chem.  11:
        1115-1128.

-------
                                         R-25

Krupa, S. V.; Grunhage, L.; Jager, H. J.; Nosal,  M.; Manning, W. J.; Legge, A. H,;
       Hanewald, K. (1995) Ambient ozone (O3) and adverse crop response: a unified view
       of cause and effect.  Environ. Pollut. 87:119-126.

Krupa, S. V.; Nosal, M.; Legge, A. H. (1994) Ambient ozone and crop loss: establishing a
       cause-effect relationship.  Environ. Pollut.  83:269-276.

Krupa, S. V.; Manning, W. J.; Nosal,  M. (1993)  Use of tobacco cultivars as biological
       indicators of ambient ozone pollution:  An  analysis of exposure-response
       relationships.  Environ. Pollut. 81:(2) 137-146.

Krupnik, A.J.; Harrington,  W.; Ostro,  B. (1990)  Ambient ozone and acute health effects:
       evidence  from daily  data. J. Environ. Econ. Manage.  18: 1-18.

Krzyzanowski, M.; Quackenboss, J. J.; Lebowitz, M. D.  (1992) Relation  of peak expiratory
       flow rates and symptoms to ambient ozone. Arch. Environ. Health  47: 107-115.

Kuhn, C., Ill; Boldt, J.; King, T.E., Jr.;  Crouch,  E.;  Vartio, T.; McDonald, J.A. (1989)
       An immunohistochemical study of architectural remodeling and connective tissue
       synthesis  in pulmonary  fibrosis.  Am. Rev. Respir. Dis. 140: 1693-1703.

Kulle, T. J.;  Sauder, L. R.; Hebel, J. R.; Chatham, M. D. (1985) Ozone  response
       relationships in healthy  nonsmokers. Am. Rev. Respir. Dis.  132: 36-41.

Kulle, T. J.;  Sauder, L. R.; Kerr, H. D.; Farrell,  B. P.; Bermel, M. S.; Smith, D. M.
       (1982) Duration of pulmonary function adaptation to ozone  in humans. Am. Ind.
       Hyg. Assoc. J. 43: 832-837.

Larsen, R.  L; Heck, W.  W. (1984) An air quality data analysis system for interrelating
       effects, standards, and needed source reductions: part 8. an effective mean O3 crop
       reduction mathematical  model. J. Air Pollut.  Control.  Assoc. 34: 1023-1034.

Last, J. A.; Gelzleichter, T. R.;  Harkema, J.; Hawk, S. (1994) Consequences of prolonged
       inhalation of ozone on Fischer-344/N rats:  collaborative studies. Part I. Content and
       cross-linking of lung collagen. Cambridge,  MA: Health  Effects Institute; research
       report no. 65.

Last, J. A.; Gelzleichter, T.; Harkema, J.; Parks,  W.  C.; Mellick,  P. (1993a) Effects of 20
       months of ozone exposure on lung collagen in Fischer 344 rats. Toxicology 84: 83-
       102.

Last, J.  A.; Gelzleichter, T. R.;  Pinkerton, K.  E.; Walker, R.  M.;  Witschi,  H. (1993b) A
       new model of progressive pulmonary fibrosis in rats. Am. Rev. Respir. Dis. 148:
      487-494.

-------
                                         R-26

Last, J. A.; Reiser,  K.  M.; Tyler, W.  S.; Rucker, R. B.  (1984) Long-term consequences of
       exposure to ozone:  I. lung collagen content. Toxicol. Appl. Pharmacol. 72: 111-118.

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.

Laurence, J. A. (1984) Development of a biological air quality indexing system: a report to
       the Minnesota Environmental Quality Board, St. Paul, MN: Minnesota Environmental
       Quality Board; SE0184B

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.; Quackenboss, J.  J.;  Krzyzanowski, M. (1991) Acute respiratory effects of
       prolonged  ambient ozone. In: Berglund, R. L.; Lawson, D. R.; McKee, D. J.,  eds.
       Tropospheric ozone and  the environment: papers from an international conference;
       March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
       Association;  pp. 111-119. (A&WMA transaction series no. TR-19).

Lechowicz, M. J. (1987) Resource  allocation by plants under air pollution stress:
       implications  for plant-pest-pathogen interactions. Bot. Rev. 53: 281-300.

Lee, E. H.; Hogsett, W. E. (1996) Methodology for calculating inputs for ozone secondary
       standard benefits analysis: part  II.  Report prepared for Office of Air  Quality Planning
       and Standards, Air  Quality Strategies and Standards  Division, U.S. Environmental
       Protection Agency, Research Triangle Park, N.C., March.

Lee, E. H.; Hogsett, W. E.; Tingey, D. T.  (1991) Efficacy of ozone exposure  indices in the
       standard setting process. In: Berglund.  R.  L.; Lawson, D. R.; McKee, D. J.. eds.
       Tropospheric ozone and  the environment: papers from an international conference;
       March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
       Association; pp. 255-271. (A&WMA transaction series no. TR-19).

Lee, E. H.; Tingey, D. T.; Hogsett, W. E.  (1989) Interrelation of experimental exposure
       and ambient air quality data for comparison of ozone exposure indices and estimating
       agricultural losses.  Corvallis, OR: U.S. Environmental Protection Agency,
       Environmental Research Laboratory; EPA report no. EPA-600/3-89-047. Available
       from: NTIS, Springfield, VA; PB89-195036.

Lee, E. H.; Tingey, D. T.; Hogsett, W. E. (1988) Evaluation of ozone exposure indices in
       exposure-response  modeling. Environ. Pollut. 53:43-62.

-------
                                         R-27

Lee, E. H.; Tingey, D. T.; Hogsett, W. E. (1987) Selection of the best exposure-response
       model using various 7-hour ozone exposure statistics. Research Triangle Park, NC:
       U.S. Environmental Protection Agency, Office of Air Quality Planning and
       Standards.

Lefohn, A. S.; Edwards, P. J.; Adams, M. B. (1994) The characterization of ozone
       exposures in rural West Virginia and Virginia. J. Air Waste Manage. Assoc. 44:
       1276-1283.

Lefohn, A. S.; Foley, J. K. (1992) NCLAN results and their application to the
       standard-setting process: protecting  vegetation  from surface ozone exposures. J.  Air
       Waste Manage. Assoc.  42: 1046-1052.

Lefohn, A. S.; Shadwick,  D. S.; Somerville, M. C.;  Chappelka, A. H ; Lockaby. B. G.;
       Meldahl, R. S. (1992) The characterization and comparison of ozone exposure indices
       used in assessing the response of loblolly pine  to ozone. Atmos. Environ.  Part A 26:
       287-298.

Lefohn, A. S.; Runeckles, V. C.; Krupa, S. V.; Shadwick,  D. S. (1989) Important
       considerations for establishing a secondary  ozone standard  to protect vegetation.
       JAPCA  39: 1039-1045.

Lefohn, A. S. (1988) Author's reply  [a response to comments by Parry and Day,  1988].
       Atmos. Environ. 22: 2058-2060.

Lefohn, A. S.;  Laurence, J. A.; Kohut, R. J.  (1988a) A comparison of indices that describe
       the relationship between exposure to ozone and reduction in the yield of agricultural
       crops. Atmos. Environ.  22: 1229-1240.

Lefohn, A. S.;  Laurence, J. A.; Kohut, R. J.  (1988b) A comparison of indices that describe
       the relationship between exposure to ozone and reduction in the yield of agricultural
       crops [response to comments by V.  C.  RuneckJes]. Atmos.  Environ. 22: 1242-1243.

Lefohn, A. S.;  Davis, C. E.; Jones, C. K.; Tingey, D. T.; Hogsett, W. E. (1987)
       Co-occurrence patterns  of gaseous air pollutant pairs  at  different minimum
       concentrations in the United States.  Atmos. Environ.  21: 2435-2444.

Lefohn, A. S.;  Jones, C. K. (1986) The characterization of ozone and sulfur dioxide air
       quality data for assessing possible vegetation effects.  J.  Air Pollut. Control Assoc. 36'
       1123-1129.

Lefohn, A. S.;  Tingey, D. T. (1984) The co-occurrence of potentially phytotoxic
       concentrations of various gaseous  air pollutants. Atmos. Environ. 18: 2521-2526.

-------
                                         R-28

Lesser, V. M.; Rawlings, J. O.; Spruill, S.  E.; Somerville, M.C. (1990) Ozone effects on
      agricultural crops: statistical methodologies and estimated  dose-response relationships.
      Crop Sci. 30: 148-155.

Linder, J.; Herren, D.; Monn, C.; Wanner, H.-U. (1988) Die Wirkung von Ozon auf die
      koerperliche Leistungsfaehigkeit [The effect of ozone on physical activity].  Schweiz
      Z. Sportmed. 36: 5-10.

Linn, W. S.; Shamoo, D. A.; Anderson, K. R.; Peng,  R.-C.; Avol, E. L.; Hackney, J. D.
      (1994) Effects of prolonged, repeated exposure to ozone,  sulfuric acid, and their
      combination in healthy and asthmatic volunteers. Am. J. Respir. Crit. Care Med.
       150:  431-440.

Linn, W.S., C.E. Spier, and  J.D. Hackney, 1993, "Activity Patterns in Ozone-Exposed
      Construction Workers," Journal of Occupational Medicine and Toxicology. Vol. 2.
      pp. 1-14.

Linn, W.S., D. Shamoo, and J. Hackney, 1992, "Documentation of Activity Patterns in
      High-Risk Groups Exposed to Ozone in the Los Angeles Area," Tropospheric  Ozone
      in the Environment II. editor,  R. Berglund, Air and Waste Management Association,
      Pittsburgh, PA.

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: an apparent  seasonal variation in response. Toxicol. Ind.  Health 4: 505-
      520.

Linn, W. S.; Avol, E. L.; Shamoo, D. A.;  Spier, C. E.; Valencia, L. M.; Venet, T. G.;
       Fischer, D. A.; Hackney, J. D.  (1986) A dose-response study of healthy, heavily
       exercising men exposed to ozone at concentrations near the ambient air quality
       standard. Toxicol. Ind. Health 2: 99-112.

Linn, W. S.; Shamoo, D. A.; Venet, T. G.; Spier, C.  E.;  Valencia,  L. M.;  Anzar, U. T.:
       Hackney, J. D. (1983a) Response to  ozone in volunteers with chronic obstructive
       pulmonary disease. Arch. Environ. Health 38: 278-283.

Linn, W. S.; Avol, E. L.; Hackney,  J. D. (1983b) 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
       environmental toxicology:  v. 5).

-------
                                         R-29

Linn, W. S.; Medway, D. A.; Anzar, U. T.; Valencia, L. M.; Spier, C. E.; Tsao, F. S.-D.
       Fischer, D. A.; Hackney, J. D. (1982) Persistence of adaptation to ozone in
       volunteers exposed repeatedly  for six weeks. Am. Rev. Respir. Dis. 125: 491-495.

Linn, W. S.; Jones, M. P.; Bachmayer, E. A.; Spier, C. E.; Mazur, S.  F.; Avol, E. L.;
       Hackney, J. D. (1980) Short-term respiratory effects of polluted ambient air:  a
       laboratory study of volunteers  in a high-oxidant community. An1.  Rev. Respir. Dis.
       121: 243-252.

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).

Lioy, P.J.; Vollmuth,  T.A.;  Lippmann, M.  (1985)  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.

Lipfert, F.W.; Hamrnerstrom, T. (1992) Temporal patterns in air pollution and hospital
       admissions. Environ.  Res. 59:  374-399.

Lippmann,  M.; Lioy,  P.J.; Liekauf, G.; Green, K.B.; Baxter,  D.; Morandi, M.; Pasternack,
       B.S. (1983)  Effects of ozone on the  pulmonary functin of children.  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. 423-446.  (Advances in
       modern technology: v. 5).

Little,  E. L., Jr. (1971) Atlas of United States Trees.  U.S. Department of Agriculture.
       Forest Service; Miscellaneous  Publication No. 1146.

Logan, J. A. (1989) Ozone in rural areas of the United States. J. Geophys. Res.  94:
       8511-8532.

Luethy-Krause, B.; Landolt,  W. (1990) Effects of ozone on starch accumulation in Norway
       spruce (Picea abies).  Trees 4:  107-110.

Madariaga, B. (1988).  Ambient air quality standards for U.S. agriculture: The correct
       welfare measure revisited.  Journal of Environmental  Management, 24,  139-146.

-------
                                        R-30

Mandl, R. H.; Laurence, J. A.; Kohut, R. J.  (1989) Development and testing of open-top
      chamber for the exposure of plants to air pollutants in the field.  J.  Environ. Qual.
      2:371-376.

Mann, L. K.; McLaughlin, S. B.; Shriner, D. S. (1980) Seasonal physiological responses of
      white pine under chronic air pollution stress. Environ. Exp. Bot. 20: 99-105.

Manning, W. J.; Bergman, J. R.; O'Brien, J. T. (1991) Ozone injury on native vegetation in
      Class I wilderness areas in New Hampshire and  Vermont.  Presented at:  84th annual
      meeting & exhibition of the Air & Waste Management Association; June; Vancouver,
      BC, Canada. Pittsburgh, PA: Air  and Waste Management Association; paper no. 91-
       144.5.

Massey, F.J., Jr.; Landau, E.; Deane, M. (1961) Air pollution and  mortality in two areas of
       Los Angeles County. Presented at: the joint meeting of the American Statistical
       Association and the Biometric Society  (ENAR);  December, New York, NY.

Mathtech, Inc.  (1995). Addendum  to the regional model farm (RMF) user's guide version
       3.0  for personal computers.  Report to U.S. EPA, June.

Mathtech, Inc.  (1994). The regional model farm (RMF): an  agricultural sector benefits
       assessment model, version 3.0 for personal computers.  Report to U.S. EPA,
       September.

Matsushima, J.; Yonemori, K.; Iwao, K. (1985) Sensitivity of satsuma mandarin to ozone as
       related to stomatal function indicated by  transpiration  rate, change of stem  diameter.
       and leaf temperature.  J. Am. Soc. Hortic. Sci. 110: 106-108.

Matyssek,  R.; Gunthardt-Goerg, M. S.; Landolt, W.; Keller, T. (1993a)   Whole-plant
       growth and leaf formation in ozonated hybrid poplar (Populus x euramericana).
       Environ. Pollut. 81:207-212.

Matyssek,  R.; Keller, T.;  Koike, T. (1993b)  Branch growth  and leaf gas exchange of
       Populus iremula exposed to low ozone concentrations throughout two growing
       seasons. Environ. Pollut. 79: 1-7.

McCool, P. M.; Musselman,  R. C.  (1990) Impact of ozone on growth of peach, apricot, and
       almond. Hortscience 25: 1384-1385.

McCool, P. M.; Musselman,  R. C.; Younglove, T.; Teso, R. R. (1988) Response of kidney
       bean to sequential ozone exposures. Environ. Exp. Bot. 28:  307-313.

McCurdy,  T.R. (1994a) Human exposure to ambient ozone,  pp. 85-127 in: D.J. McKee, ed.
       Tropospheric Ozone:  Human Health and Agricultural Impacts; Lewis Publishers.

-------
                                        R-31

 McCurdy, T.R. (1994b) Estimating ozone exposures for the general population in 8 urban
       areas using a probabilistic version of NEM. U.S. EPA, OAQPS; Research Triangle
       Park, NC.  (For copies, contact Harvey  M. Richmond, U.S. Environmental
       Protection Agency, OAQPS, MD-15, Research Triangle Park, N.C. 27711, (919)
       541-5271.)

 McCurdy, T. R. (1994c) Testing pNEM/O3 runs to determine if a set of 10 Runs is
       "representative", EPA Memorandum; April 19.

 McDonnell, W. F.; Smith, M. V. (1994) Prediction of acute ozone response as a function of
       exposure rate and total inhaled dose. J.  Appl. Physiol. 76: 2776-2784.

 McDonnell, W. F.; Andreoni, S.; Smith, M. V. (1994) Proportion of moderately exercising
       individuals responding to low-level multihour ozone exposure.  Am. J. Respir. Crit.
       Care Med.: in press.

 McDonnell, W. F.; Muller, K. E.; Bromberg,  P. A.; Shy, C. M.  (1993) Predictors of
       individual differences in acute response  to ozone exposure.  Am. Rev. Respir. Dis.
       147: 818-825.

 McDonnell, W. F.; Kehrl, H. R.; Abdul-Salaam, S.; Ives, P. J.;  Folinsbee, L. J.; Devlin.
       R. B.; O'Neil, J. J.; Horstman, D. H. (1991) Respiratory  response of humans
       exposed  to low levels of ozone for 6.6 hours. Arch. Environ.  Health 46: 145-150.

 McDonnell, W. F.; Horstman, D. H.; Abdul-Salaam, S.; Raggio,  L. J.; Green, J. A. (1987)
       The respirator)' responses  of subjects with allergic rhinitis to ozone exposure and their
       relationship to nonspecific airway reactivity. Toxicol.  Ind.  Health 3: 507-517.

 McDonnell, W.F., III;  Chapman, R.S.; Leigh, M.W.; Strope, G.L.: Collier. A.M. (1985)
       Respiratory responses of vigorously exercising children to 0.12 ppm ozone exposure.
       Am. Rev.  Respir.  Dis. 132:875-879.

 McDonnell, W.F.; Horstman, D.H.; Hazucha, M.J.; Seal, E., Jr.; Haak, E.D.; Salaam,
       S.A.; House, D.E. (1983) Pulmonary effects of ozone exposure during exercise:
       dose-response characteristics. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol
       54: 1345-1352.

McGartland, A. M. (1987) The implications of ambient ozone standards for U.S. agriculture:
       a comment and some further evidence. J.  Environ. Manage. 24: 139-146.

McKittrick, T.;  Adams, W. C. (1995) Pulmonary function response to equivalent doses of
       ozone consequent to intermittent and continuous exercise. Arch. Environ. Health  50"
       153-158.

-------
                                         R-32

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.; Taylor, G. E. (1981) Relative humidity: important modifier of pollutant
       uptake by plants. Science (Washington, DC) 211: 167-169.

McLeod, A. R.; Shaw, P. J. A.; Holland, M. R. (1992) The Liphook Forest Fumigation
       Project:  studies of sulphur dioxide and ozone effects on coniferous trees. For. Ecol.
       Manage. 51: 121-127.

McQuattie,  C. J.; Schier, G. A.  (1992) Effect of ozone and aluminum on pitch pine (Pinus
       rigidd) seedlings: anatomy of mycorrhizae. Can. J. For. Res. 22: 1901-1916.

Meier, S.; Grand,  L. F.; Schoeneberger,  M. M.; Reinert, R. A.; Bruck, R. I. (1990)
       Growth, ectomycorrhizae and nonstructural carbohydrates  of loblolly pine seedlings
       exposed to ozone and soil water deficit. Environ. Pollut. 64: 11-27.

Miller, F. J.; Illing, J.  W.; Gardner, D. E. (1978) Effect of urban ozone levels on
       laboratory-induced respiratory infections. Toxicol. Lett.  2: 163-169.

Miller, J. E.; Patterson, R. P.; Heagle, A.  S.; Pursley, W. A.; Heck, W. W. (1988) Growth
       of cotton under chronic ozone stress at two levels of soil moisture. J. Environ. Qua!.
       17: 635-643.

Miller, P. R.; McBride, J. R.; Schilling,  S. L.; Gomez, A. P.  (1989) Trend of ozone
       damage  to conifer forests between 1974 and 1988 in the San Bernardino  mountains in
       southern California. In: Olson, R. K.; Lefohn,  A. S., eds. Effects of air pollution on
       western  forests. Pittsburgh.  PA: Air and Waste Management Association: pp. 309-324
       (Transactions series no. 16).

Miller, P. R.; Taylor, O. C.; Wilhour, R. G. (1982) Oxidant.air pollution effects on a
       western  coniferous forest ecosystem. Corvallis,  OR:  U.S.  Environmental  Protection
       Agency, Environmental Research Laboratory; EPA report no. EPA-600/D-82-276.
       Available from: NTIS, Springfield, VA; PB83-189308.

Miller, P. R.; Millecan, A. A. (1971)  Extent of oxidant air pollution damage to some pines
       and  other conifers in  California.  Plant. Dis. Rep. 55:555-559.

Mills, C.A.  (1960) Respiratory  and cardiac deaths in Los Angeles smogs during 1956,  1957
       and  1958.  Am. J.  Med. Sci. 239: 307-315.

Mills, C.A.  (1957a) Do smogs threaten community health?  Cincinnati J. Med.  38: 259-261.

-------
                                        R-33

Mills, C.A. (1957b) Respiratory and cardiac deaths in Los Angeles smogs. Am. J. Med. Sci.
      233: 379-386.

Moldau, H.; Sober, J.; Sober, A. (1990) Differential sensitivity of stomata and mesophyll to
      sudden exposure of bean  shoots to ozone. Photosynthetica 24: 446-458.

Molfmo, N. A.; Wright, S. C.;  Katz, I.; Tarlo, S.; Silverman, F.; McClean, P.  A.; Szalai,
      J. P.; Raizenne, M.; Slutsky, A. S.; Zamel, N. (1991) Effect of low  concentrations
      of ozone on inhaled allergen responses in asthmatic subjects. Lancet 338(8761):  199-
      203.

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.

Moser, T. J.; Tingey, D. T.; Rodecap, K. D.; Rossi, D.  J.; Clark, C. S. (1988) Drought
      stress applied during the  reproductive phase reduced ozone-induced effects in bush
      bean. In: Heck, W. W,;  Taylor, O. C.; Tingey, D. T., eds. Assessment of crop loss
      from air pollutants. London, United Kingdom: Elsevier Applied Science; pp.
      345-364.

Motley,  H.L.; R.H. Smart, and  C.I. Leftwich.  Effect of polluted Los Angeles air (SMOG)
      on lung volume measurements.  J. Am. Med.  Assoc. 171:1469-1477,  1959.

Musselman, R.C.;  Younglove, T.; McCool,  P.M. (1994)  Response of Phascolus vulgaris L.
      to differing  ozone regimes having identical total exposure.   Atmos. Environ. 28:
      2727-2731.

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 concentration.
      J. Am. Soc. Hortic.  Sci.  Ill: 470-473.

Musselman, R.  C.; Oshima, R.  J.; Gallavan, R. E. (1983) Significance of pollutant
      concentration distribution in the response of 'red kidney' beans to ozone. J. Am. Soc.
      Hortic.  Sci. 108: 347-351.

Mustafa, M. G.; Elsayed, N. M.; Lee, S. D. (1985)  The combined effects of O3 and NO2 on
      pulmonary biochemistry.  In: Lee,  S. D., ed. Evaluation of the scientific basis for
      ozone/oxidants standards: transactions of an APCA specialty conference; November
      1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association; pp. 337-347.

Nagda, N.L.; Rector,  H.E.; Koontz, M.D.  (1987) Guidelines for Monitoring Indoor Air
      Quality.  Washington, DC: Hemisphere Pub. Co..

-------
                                        R-34

Nakamura, H.; H.  Saka, (1978) Photochemical oxidants injury in rice plants.  III. Effect of
       ozone on physiological activities in rice plants. Nippon Sakumotsu Gakkai Kiji 47:
       707-714 [English summary]

NAS. 1991. Rethinking the Ozone Problem in Urban and Regional Air Pollution. National
       Research Council, National Academy of Sciences, National Academy Press,
       Washington, DC.

National Institutes of Health. (1991) Guidelines for the diagnosis and management of asthma.
       Bethesda, MD; U.S. Department of Health and Human Services, National Heart,
       Lung, and Blood Institute, National Asthma Education Program; publication no. 91-
       3042.

National Toxicology Program. (1995) NTP technical  report on the toxicology and
       carcinogenesis studies of ozone (CAS no.  10028-15-6) and ozone/NNK (CAS  no.
       10028-15-6/64091-91-4) in F344/N rats and B6C3F, mice (inhalation studies).
       Research Triangle Park, NC: U.S. Department of Health and Human Services,
       National Institutes of Health; in press. (National Toxicology Program technical report
       series no. 440).

Neufeld, H. S.; Renfro, J. R.; Hacker, W. D.; Silsbee, D. (1992)  Ozone in Great Smoky
       Mountains National Park: dynamics and effects on plants.  In:  Troposphenc ozone
       and the environment II - effects, modeling and control: papers from an international
       specialty conference; November; Atlanta,  GA, Pittsburgh, PA: Air & Waste
       Management Association; pp. 594-617. (A&WMA transactions series:  TR-20).

Nouchi, I.; Aoki, K. (1979) Morning glory as a  photochemical oxidant indicator. Environ.
       Pollut. 18: 289-303.

Olszyk, D.; Bytnerowicz, A.; Kats, G.; Reagan, C.;  Hake, S.; Kerby, T.; Millhouse, D.;
       Roberts, B.; Anderson, C.; Lee, E. H. (1993) Cotton yield losses and ambient ozone
       concentrations in California's San Joaquin Valley. J. Environ.  Qual.: in press.

Olszyk, D. M.; Takemoto, B. K.; Ktas, G.; Dawson, P. J.; Morrison, C.; Preston, J W.;
       Thompson,  C. R.  (1992) Effects of open-top chambers on 'Valencia' orange trees. J.
       Environ. Qual. 21: 128-134.

Olszyk, D. M.; Bytnerowicz, A.; Kats, G.; Dawson, P. J.; Wolf, J.; Thompson, C.  R.
       (1986) Effects of sulfur dioxide and ambient ozone on winter wheat and lettuce. J.
       Environ. Qual.  15: 363-369.

Costing, R.S.; Van Golde, L.M.G.; Verhoef, J.; Van Bree, L. (1991a) Species differences
       in  impairment and recovery  of alveolar macrophage functions following single and
       repeated ozone exposures.  Toxicol. Appl. Pharmacol.  110:  170-178.

-------
                                         R-35

Costing, R.S.; Van Rees-Verhoef, M.; Verhoef, J.; Van Golde, L.M.G.; Van Bree, L.
       (1991b) Effects of ozone on cellular ATP levels in rat and mouse alveolar
       macrophages. Toxicology 70: 195-202.

Oshima, R. J. (1975) Development of a system for evaluating and reporting economic crop
       losses caused by air pollution in California.  III. Ozone dosage-crop loss conversion
       function—alfalfa, sweet corn. Ill A. Procedures for production, ozone effects on alfalfa
       and sweet corn, and evaluation of these systems.  Sacramento, CA: California Air
       Resources Board; report no. ARB-R-3-690-75-35.  Available  from: NTIS,
       Springfield, VA; PB-272994.

Ostro, B.D.;  Lipsett, M.J.; Mann, J.K.; Krupnick, A.; Harrington, W. (1993)  Air pollution
       and respiratory morbidity among adults  in Southern California.  Am. J. Epidemiol.
       137: 691-700.

Parks, W. C.; Roby, J. D. (1994) Consequences of prolonged inhalation of ozone on F344/N
       rats: collaborative studies.  Part IV.  Effects  on expression of extracellular  matrix
       genes. Cambridge, MA: Health Effects  Institute; research report no.  65.

Parmeter, J. R.,  Jr.; Bega, R. V.; Neff, T. (1962)  A chlorotic decline of ponderosa pine  in
       southern California. Plant Dis. Rep. 46:269-273.

Patton, R. L. (1981) Effects of ozone and sulfur dioxide on height and stem specific gravity
       of Populus hybrids. Washington, DC: USD A Forest Service research paper; paper
       no. NE-471.

Pell, E. J. (1988)  Secondary  metabolism and air pollutants.  In: Gaseous air pollution and
       plant metabolism: proceedings of the 2nd international symposium; pp. 222-237.

Pell, E. J.; Pearson, N. S.  (1983) Ozone-induced reduction in quantity of
       ribulose-l,5-bisphosphate carboxylase in alfalfa foliage.  Plant Physiol. 73: 185-187

Peterson, D.  L.; Silsbee, D. G.; Poth, M.;  Arbaugh, M. J.; Biles, F. E. (1995) Growth
       response of bigcone Douglas fir (Pseudotsuga macrocarpa) to long-term ozone
       exposure  in southern California. J. Air Waste Manage. Assoc. 45: 36-45.

Peterson, D.  L.; Arbaugh, M. J. (1992) Coniferous forests of the Colorado  front range. Pan
       B: ponderosa pine second-growth stands. In: Olson, R. K.; Binkley, D.; Boehm, M.,
       eds. The response of western forests to  air pollution. New York, NY:
       Springer-Verlag; pp. 365 and 385-401.  (Billings, W. D.; Golley,  F.; Lange, O. L.;
       Olson, J.  S.; Remmert, H. Ecological studies; analysis and synthesis v. 97).

-------
                                         R-36

Peterson, D. L.; Arbaugh, M. J.; Linday, J. R. (1991) Regional growth changes in
       ozone-stressed ponderosa pine (Pinus ponderosa) in the Sierra Nevada, California,
       USA. Holocene 1: 50-61.

Peterson, D. L.; Arbaugh, M. J. (1988) An evaluation of the effects of ozone injury on
       radial growth of ponderosa pine (Pinus ponderosa) in the southern Sierra Nevada.
       JAPCA 38: 921-927.

Peterson, D. L.; Arbaugh, M. J.; Wakefield, V. A.; Miller, P. R. (1987) Evidence of
       growth reduction in ozone-injured Jeffrey pine (Pinus Jeffreyi Grev. and Balf.) in
       Sequoia and Kings Canyon National  Parks.  JAPCA 37: 906-912.

Pinkerton, K. E.; Menache, M. G.;  Plopper, C. G. (1995) Consequences of prolonged
       inhalation of ozone on F344/N rats: collaborative  studies. Pan IX. Changes in the
       tracheobronchial epithelium, pulmonary acinus, and lung antioxidant enzyme activity.
       Cambridge, MA: Health Effects Institute; pp. 41-98;  research report number 65.

Pinkerton, K. E.; Dodge, D.  E.; Cederdahl-Demmler, J.; Wong, V. J.;  Peake, J.; Haselton,
       C. J.; Mellick,  P. W.; Singh, G.; Plopper, C. G. (1993) Differentiated bronchiolar
       epithelium in alveolar  ducts of rats exposed to ozone for 20.months. Am.  J. Pathol.
       142: 947-956.

Pinkerton, K. E.; Mercer, R. R.; Plopper, C. G.;  Crapo, J. D. (1992) Distribution of injury
       and microdosimetry of ozone in the ventilatory unit of the rat. J.  Appl. Physiol. 73:
       817-824.

Pino, M. V.; McDonald, R. J.; Berry, J. D.; Joad, J. P.; Tarkington, B. K.; Hyde, D.  M.
       (1992a) Functional and morphologic  changes caused by acute ozone exposure in the
       isolated and perfused rat lung. Am. Rev. Respir.  Dis. 145: 882-889.

Pino, M. V.; Stovall,  M. Y.; Levin, J. R.;  Devlin, R. B.; Koren, H.  S.: Hyde,  D. M.
       (19925) Acute  ozone-induced lung injury in neutrophil-depleted rats. Toxicol. Appl.
       Pharmacol. 114: 268-276.

Pino, M. V.; Levin, J. R.; Stovall, M. Y.;  Hyde,  D. M. (1992c) Pulmonary inflammation
       and epithelial injury in response to acute ozone exposure in the rat. Toxicol. Appl.
       Pharmacol. 112: 64-72.

Plopper, C. G.; Chu,  F.-P.; Haselton, C. J.; Peake, J.; Wu, J.; Pinkerton, K. E. (1994a)
       Dose-dependent tolerance to ozone: I. tracheobronchial epithelial reorganization in
       rats after 20 months' exposure. Am. J. Pathol. 144: 404-420.

-------
                                         R-37

Plopper, C. G.; Duan, X.; Buckpitt, A. R.; Pinkerton, K. E. (1994b) Dose-dependent
       tolerance to ozone. IV. Site-specific elevation in antioxidant enzymes in the lungs of
       rats exposed for 90 days or 20 months. Toxicol. Appl. Pharmacol.  127: 124-131.

Plopper, C. G.; Harkema, J.  R.; Last, J.  A.; Pinkerton, K. E.; Tyler, W. S.; St. George, J.
       A.; Wong, V.  J.;  Nishio, S. J.; Weir, A. S.; Dungworth, D.  L.; Barry, B. E.; Hyde,
       D. M.  (1991) The respiratory system of nonhuman primates responds more to
       ambient concentrations of ozone than does that of rats. In: Berglund, R. L.; Lawson,
       D. R.;  McKee, D. J., eds. Tropospheric ozone and the environment: papers from an
       international conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air &
       Waste Management Association; pp.  137-150. (A&WMA transaction series no. TR-
       19).

Plopper, C. G.; Dungworth, D. L.; Tyler, W. S.; Chow,  C. K. (1979) Pulmonary
       alterations  in rats exposed to 0.2 and 0.1 ppm ozone: a correlated morphological and
       biochemical study. Arch. Environ. Health 34: 390-395.

Ponka, A. (1991)  Asthma and low level air pollution in Helsinki. Arch. Environ. Health 46:
       262-270.

Portney, P.R.; Mullahy, J. (1990) Urban air quality and chronic respiratory' disease.   Reg.
       Sci. Urban Econ. 20: 407-418.

Preston, E. M.; Tingey, D. T. (1988)  The NCLAN program for crop loss assessment. In:
       Heck, W.  W.;  Taylor, O. C.; Tingey, D. T. eds.  Assessment of crop loss from air
       pollutants:  proceedings of an international conference; October, 1987, Raleigh, NC.
       New York, NY: Elsevier Applied Science; pp. 45-62.

Pryor, W.A. (1992) How far does ozone penetrate into the pulmonary air/tissue boundary
       before it reacts? Free Radic. Biol. Med. 12: 83-88.

Pryor, W.A. (1991) Can  vitamin E protect humans against the  pathological effects of ozone
       in smog?   Am. J.  Clin. Nutr. 53:  702-722.

Pye, J. M. (1988) Impact of ozone on the growth and yield of trees: a review.  J.  Environ.
       Qual. 17: 347-360.

Pye, J. M., J.  E.  deSteiger, and C. Love  (1988).  Expert opinion survey on the impact of air
       pollutants on forests of the USA.  Proceedings of air pollution and forest decline,
       Interlaken, Switzerland, Oct. 2-8.
                                           t
Radhakrishnamurthy, B. (1994) Consequences of prolonged inhalation of ozone on F344/N
       rats:  collaborative  studies. Part III. Effects on complex carbohydrates of lung
       connective  tissue. Cambridge, MA: Health Effects Institute; research report no. 65.

-------
                                         R-38

Raizenne, M. E.; Burnett, R. T.; Stern, B.; Franklin, C.  A.; Spengler, J. D. (1989) Acute
       lung function responses to ambient acid aerosol exposures in children. Environ.
       Health Perspect. 79: 179-185.

Raizenne, M.E.; Spengler, J.D.  (1989)  Dosimetric model of acute health effects of ozone
       and acid aerosols in children.  In: Schneider, T.; Lee, S.D.; Wolters, G.J.R.; Grant,
       L.D., eds. Atmospheric ozone research and its policy implications:  proceedings of
       the 3rd US-Dutch international symposium; May 1988; Nijmegen, The Netherlands.
       Amsterdam, The Netherlands: Elsevier Science Publishers; pp. 319-329. (Studies in
       environmental science 35).

Raizenne, M.; Stern, B.; Burnett, R.; Spengler, J. (1987) Acute respiratory function and
       transported air pollutants: observational studies. Presented at: 80th annual meeting of
       the Air Pollution Control Association;  June; New York, NY. Pittsburgh, PA: Air
       Pollution Control Association; paper no. 87-32.6.

Rao, G.  A.; Larkin, E. C.; Harkema, J. R.; Dungworth,  D. L.  (1985a) Changes  in the
       levels of polyunsaturated  fatty acids in the  lung and lecithin cholesterol acyl
       transferase activity in plasma of monkeys exposed  to ambient levels of ozone.
       Toxicol. Lett. 24:  125-129.

Rao, G.  A.; Larkin, E. C.; Harkema, J. R.; Dungworth,  D. L.  (1985b) Changes in lipids of
       lung lavage in monkeys after chronic exposure to ambient levels of ozone. Toxicol.
       Lett. 29: 207-214.

Reinert,  R. A.; Heggestad, H. E.; Heck, W.  W.  (1982) Response and genetic modification
       of plants for  tolerance to air pollutants. In: Christiansen,  M. L.; Lewis, C. F., eds.
       Breeding plants for less favorable environments. New York, NY: John Wilev and
       Sons; pp.  259-292.

Reisenauer, C. S.; Koenig, J.  Q.; McManus,  M.  S.; Smith, M. S.; Kusic, G.; Pierson, W.
       E.  (1988) Pulmonary response to ozone exposures in healthy individuals aged 55
       years or greater. JAPCA 38: 51-55.

Retzlaff, W. A.;  DeJong, T. M.; Williams, L. E. (1992a) Photosynthesis and growth
       response of almond to increased  atmospheric ozone partial pressures. J.  Environ.
       Qual. 21: 208-216.

Retzlaff, W. A.; Williams, L. E.; Dejong, T. M. (1992b) Photosynthesis, growth, and yield
       response of 'Casselman'  plum to various ozone partial  pressures during orchard
       establishment. J. Am. Soc. Hortic. Sci. 117: 703-710.

-------
                                         R-39

Retzlaff, W. A.; Williams, L. E.; DeJong, T. M.  (1991) The effect of different atmospheric
       ozone partial pressures on photosynthesis and growth of nine fruit and nut tree
       species.  Tree Physiol. 8: 93-105.

Richards, B. L.; Middleton, J. T.; Hewitt, W. B.  (1958) Air pollution with relation to
       agronomic crops: V. oxidant stipple of grape. Agron.  J. 50: 559-561.

Roemer, W.; Hoek, G.; Brunekreef, B. (1993) Effect of ambient winter air pollution on
       respiratory health of children with chronic respiratory  symptoms. Am. Rev. Respir.
       Dis. 147: 118-124.

Rokaw, S.N. and F. Massey. Air pollution and chronic respiratory disease.  Am. Rev.
       Respir. Dis. 86:703-704. 1962.

Roose, M. L.; Bradshaw,  A. D.; Roberts, T. M. (1982) Evolution of resistance to gaseous
       air pollutants. In: Unsworth, M. H.; Ormrod, D. P., eds. Effects of gaseous air
       pollution in agriculture and horticulture. London, United Kingdom: Butterworth
       Scientific; pp. 379-409.

Ryer-Powder, J.E.; Amoruso, M.A.; Czerniecki, B.; Witz, G.; Goldstein, B.D. (1988)
       Inhalation of ozone produces a decrease in superoxide anion radical production in
       mouse alveolar macrophages. Am. Rev. Respir. Dis. 138: 1129-1133.

Samuelson, L. J.; Edwards, G. S. (1993) A comparison of sensitivity  to ozone in  seedlings
       and trees of Quercus rubra L. New Phytol. 12:  373-379.

Sasek, T. W.; Richardson,  C. J.;  Fendick, E. A.;  Bevington, S. R.; Kress, L. W. (1991)
       Carryover effects of acid rain and  ozone on the physiology of multiple flushes of
       loblolly pine seedlings. For. Sci. 37: 1078-1098.

Schelegle, E.S.; Siefkin, A.D.; McDonald,  R.J. (1991) Time course  of ozone-induced
       neutrophilia in  normal humans.  Am. Rev.  Respir.  Dis.  143:1353-1358.

Schelegle, E. S.; Adams, W. C. (1986) Reduced exercise  time in competitive simulations
       consequent to low level ozone exposure. Med. Sci. Sports Exercise 18: 408-414.

Schmitzberger,  R.; Rhomberg, K.; Buchele, H.; et al.  (1993) Effects of air pollution on the
       respiratory systems of children. Pediatr. Pulmonol. 15: 68-74.

Schoettlin, C.E. and E. Landau. Air pollution and asthmatic attacks in the Los Angeles area.
       Public Health Rep.  76:545-548.  1961.

Schultheis, A. H.; Bassett,  D. J. P. (1991) Inflammatory cell influx into ozone-exposed
       guinea pig lung interstitial and airways spaces. Agents Actions 34: 270-273.

-------
                                         R-40

Schulze, E. D.; Hall, A. E. (1985)  Stomatal responses, water loss and CO2 assimilation
       rates of plants in contrasting environments. In: Lange, O. L.; Nobel, P. S.; Osmond,
       C.  B.;  Ziegler, H., eds. Physiological plant ecology II. pp. 181-230. (Encyclopedia
       of plant physiology new series: 12B).

Schwartz,  J. (1994a) Air pollution and hospital admissions for the elderly in Birmingham,
       Alabama. Am. J.  Epidemiol. 139: 589-598.

Schwartz,  J. (1994b) Air pollution and hospital admissions for the elderly in Detroit, MI.
       Am. J. Respir. Crit. Care Med. 150: 648-655.

Schwartz,  J. (1994c) PM10, ozone, and hospital admissions for the elderly  in Minneapolis,
       MN. Arch. Environ. Health. 49:  366-374.

Schwartz,  J. (1991) Paniculate air pollution and daily mortality in Detroit.  Environ. Res.
       56:  204-213.

Schwartz,  J. (1989) Lung function and chronic exposure to air pollution: a cross-sectional
       analysis of NHANES II. Environ. Res.  50: 309-321.

Schwartz,  L.  W.; Dungworth, D. L.; Mustafa, M. G.; Tarkington,  B. K.;  Tyler, W. S.
       (1976)  Pulmonary responses of rats to ambient levels of ozone: effects of 7-day
       intermittent or continuous exposure.  Lab. Invest. 34:  565-578.

Seal,  E., Jr.; McDonnell, W. F.; House, D. E. (1994) The effect of age, socioeconomic
       status,  and menstrua] cycle on the pulmonary response to ozone. Research  Triangle
       Park, NC: U.S. Environmental Protection Agency, Health Effects Research
       Laboratory: manuscript submitted  for publication.

Seal,  E., Jr.; McDonnell, W. F.; House, D. E.; Salaam, S. A.; Dewitt, P. J.; Butler, S. O.:
       Green, J.; Raggio, L.  (1993) The pulmonary response of white and  black adults to six
       concentrations of ozone. Am. Rev. Respir. Dis.  147: 804-810.

Second Progress Report of FOREST, An Interagency Cooperative Project:  Forest Ozone
       Response STudy (Draft) Feb. 1994.  Prepared by the FOREST Data Archiving and
       Analysis Group.

Selgrade,  M. K.; Daniels, M. J.; Grose,  E. C. (1990) Acute, subchronic, and  chronic
       exposure to a simulated urban profile of ozone: effects on extrapulmonary natural
       killer cell activity and  lymphocyte mitogenic responses. Inhalation Toxicol. 2:  375-
       389.

-------
                                         R-41

Seltzer, J.; Bigby, B. G.; Stulbarg, M.; Holtzman, M. J.; Nadel, J. A.; Ueki, I. F.;
       Leikauf, G. D.; Goetzl, E. J.; Boushey, H. A.  (1986) Gyinduced change in bronchial
       reactivity to methacholine and airway inflammation in humans. J. Appl. Physiol. 60:
       1321-1326.

Selwyn, B.J.; Stock, T.H.; Hardy, R.J.; Chan, F.A.; Jenkins, D.E.; Kotchmar, D.J.;
       Chapman, R.S.  (1985)  Health effects of ambient ozone exposure in vigorously
       exercising adults.  In: Lee, S.D., ed. Evaluation of the scientific basis for
       ozone/oxidants standards:  proceedings of an APCA international specialty conference;
       November 1984; Houston, TX. Pittsburgh, PA: Air Pollution Control Association;
       pp. 281-296. (APCA international specialty conference  transactions: TR-4).

Shaver, C. L.; Tonnessen, K. A.; Maniero, T. G. (1994)  Clearing the air at Great Smoky
       Mountains National Park.  Ecol. App. 4: 690-701.

Sherwin, R.P. (1991)  Air pollution: the pathobiologic issues. J.  Toxicol. Clin. Toxicol.  29:
       385-400.

Sherwin, R.P.; Richters, V. (1991) Centriacinar region (CAR) disease in the lungs of young
       adults: a preliminary report. In: Berglund, R.L.; Lawson, D.R.; McKee, D.J., eds.
       Tropospheric ozone and the environment: papers from an  international conference;
       March 1990; Los Angeles, CA. Pittsburgh, PA: Air &  Waste Management
       Association; pp. 178-196.  (A&WMA transaction series  no. TR-19).

Sherwin, R. P.; Richters, V. (1985) Effect of 0.3 ppm ozone exposure on  type II cells and
       alveolar walls of newborn  mice:  an image-analysis quantitation. J. Toxicol. Environ.
       Health 16: 535-546.

Showman, R. E. (1991) A comparison of ozone injury to vegetation during moist and
       drought years. J. Air Waste Manage. Assoc. 41: 63-64.

Shumway, R.H.; Azari, A.S.; Pawitan, Y. (1988) Modeling mortality fluctuations in Los
       Angeles as functions of pollution and weather effects. Environ. Res. 45:  224-241.

Sierra Cooperative Ozone Impact Assessment Study:  Year 3, Volume 1. California
       Environmental Protection Agency, Air Resources Board. Contract No. A-132-188,
       Final Report, October 1993.

Simini, M.;  Skelly, J. M.; Davis, D. D.; Savage, J.  E.; Comrie, A. C.  (1992) Sensitivity of
       four hardwood species to ambient ozone in north central Pennsylvania.  Can. J. For.
       Res. 22: 1789-1799.

-------
                                        R-42

Skelly, J. M.; Yang, Y. S.; Chevone, B. I.; Long, S. J.; Nellessen, J. E.; Winner, W.E.
       (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, DC. Arlington, VA; Izaak Walton League of America
       Endowment; pp.  143-159.

Sly, M.D. (1988) Mortality from asthma in children 1979-1984. Ann. Allergy 60: 433-443.

Solomon, P. A.; Salmon, L. G.; Fall, T.; Cass, G.  R. (1992)  Spatial and temporal
       distribution of atmospheric nitric acid and paniculate nitrate concentrations in the Los
       Angeles area. Environ. Sci. Technol. 26: 1596-1601.

Spektor, D. M.;  Lippmann, M. (1991) Health effects of ambient ozone on healthy children at
       a summer camp.  In: Berglund, R. L.;  Lawson, D. R.;  McKee, D. J., eds.
       Tropospheric  ozone and the environment: papers from an international conference;
       March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste Management
       Association; pp.  83-89.  (A&WMA transaction series no. TR-19).

Spektor, D. M.;  Thurston, G. D.; Mao, J.; He, D.; Hayes, C.; Lippmann, M. (1991)
       Effects of single- and multiday ozone exposures on respiratory function in active
       normal children.  Environ. Res.  55:  107-122.

Spektor, D. M.;  Lippman, M.; Lioy, P. J.; Thurston, G. D.;s 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, D. 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."

Spence, R. D.; Rykiel,  E. J., Jr.; Sharpe, P. J.  H. (1990) Ozone alters carbon allocation in
       loblolly pine: assessment with carbon-11  labeling. Environ.  Pollut. 64: 93-106.

Spier, C.E. et al., 1992, "Activity Patterns in Elementary and High School  Students Exposed
       to Oxidant  Pollution," Journal of Exposure Analysis and Environmental
       Epidemiology. Vol. 2, pp. 277-293.

Stasiuk, W. N.,  Jr.;  Coffey, P. E. (1974) Rural and urban ozone relationships in New York
       State. J. Air Pollut. Control Assoc. 24: 564-568.

-------
                                         R-43

Stern, B. R.; Raizenne, M. E.; Burnett, R. T.; Jones, L.;  Kearney, J.; Franklin, C. A.
       (1994) Air pollution and childhood respiratory health: exposure to sulfate and ozone
       in 10 Canadian rural communities. Environ. Res. 66: 125-142.

Stern, B.; Jones, L.; Raizenne, M.; Burnett, R.; Meranger, J. C.; Franklin, C. A.  (1989)
       Respiratory health effects associated with ambient sulfates and ozone in two  rural
       Canadian communities. Environ. Research  49: 20-39.

Stock,  T.H.;  Kotchman, D.J.; Contant, C.F.; et ah, 1985, "The Estimation of Personal
       Exposures to Air Pollutants for a Community-Based Study of Health Effects in
       Asthmatics - Design and Results of Air Monitoring," Journal of the Air Pollution
       Control  Association, Vol. 35, p.  1266.

Stow, T. K.; Allen, H. L.; Kress,  L. W. (1992) Ozone impacts on seasonal foliage dynamics
       of young loblolly pine. For. Sci. 38: 102-119.

Szarek, J. L. (1994) Consequences of prolonged inhalation of ozone on F344/N rats:
       collaborative studies. Part II. Mechanical properties, responses to bronchoactive
       stimuli,  and eicosanoid release in isolated large and small airways. Cambridge, MA:
       Health Effects  Institute; research report no. 65.

Tan, W. C.; Bethel, R. A. (1992)  The  effect of platelet activating factor antagonist on
       ozone-induced  airway  inflammation and bronchial hyperresponsiveness in guinea pigs.
       Am. Rev.  Respir. Dis. 146: 916-922.

Taylor, G. E.,  Jr.; Johnson, D. W.;  Andersen, C. P. (1994) Air pollution and forest
       ecosystems: a regional  to global perspective.  Ecol.  Appl. 4: 662-689.

Taylor, G.; Davies, W. J. (1990) Root growth of Fagus sylvarica: impact of air quality and
       drought  at a site in southern Britain. New Phytol. 116: 457-464.

Taylor, G. E.,  Jr.; Norby, R. J. (1985) The significance of elevated levels of ozone on
       natural ecosystems of North America. In: Lee, S. D., ed. Evaluation of the  scientific
       basis for ozone/oxidants standards: proceedings of an APCA international  specialty
       conference; November 1984; Houston, TX. Pittsburgh, PA: Air Pollution  Control
       Association; pp. 152-175. (APCA transactions: TR-4).

Temple, P. J.;  Riechers,  G. H.; Miller, P. R.; Lennox, R. W. (1993) Growth responses of
       ponderosa pine to long-term exposure to ozone, wet and dry acidic deposition, and
       drought. Can. J. For.  Res.  23: 59-66.

Temple, P. J. (1990) Growth and yield responses  of processing tomato (Lycopersicon
       esculenrum Mill.) cultivars to  ozone. Environ. Exp. Bot. 30: 283-291.

-------
                                         R-44

Temple, P. J.; Benoit, L. F. (1988) Effects of ozone and water stress on canopy
       temperature, water use, and water use efficiency of alfalfa. Agron. J. 80: 439-447.

Temple, P. J.; Kupper, R. S.; Lennox, R. W.; Rohr, K. (1988) Injury and yield responses
       of differentially irrigated cotton to ozone. Agron. J. 80: 751-755.

Temple, P. J.; Taylor, O. C.; Benoit, L. F. (1986) Yield  response of head lettuce  (Lactuca
       sativa) to ozone.  Environ. Exp. Bot. 26: 53-58.

Tenhunen, J. D.; Meyer, A.; Lange, O. L.; Gates, D. M. (1980) Development of a
       photosynthesis model with an emphasis on ecological applications.  V. Test of
       applicability of a steady-state model to description of net photosynthesis of Prunus
       armeniaca under field conditions.  Oecologia 45: 147-155.

Tepper, J. S.; et al. (1995) Time course of inflammation in the rat exposed to ozone for
       short and prolonged periods with intermittent hyperventilation.  Research  Triangle
       Park,  NC: U.S. Environmental Protection Agency,  National Health & Environmental
       Effects Research  Laboratory, Pulmonary Toxicology Branch.

Tepper, J. S.; Costa, D. L.; Doerfler, D.; Lehmann, J. R.; Stevens, M. A.;  Madden, M. C.
       (1994) The relative contribution of concentration and duration to ozone (O3)-induced
       lung injury. Fundam. Appl. Toxicol.: submitted.

Thurston, G.D.; Kinney, P.I . (1995)  Air pollution epidemiology: considerations in time-
       series modeling.  Inhalation Toxicol. 7:71-83.

Thurston, G.D.; Ito, K.;  Hayes, C.G.; Bates, D.V.; Lippmann, M. (1994) Respiratory
       hospital admissions and summertime haze air pollution in Toronto, Ontario:
       consideration of the  role of acid aerosols. Environ.  Res. 65: 271-290.

Thurston, G.D.; Ito, K.; Kinney, P.L.; Lippmann, M. (1992) A multi-year study  of air
       pollution  and respiratory hospital  admissions in three New York State  metropolitan
       areas: results for 1988 and 1989 summers.  J. Exposure Anal.  Environ. Epidemiol.
       2:429-450.

Tingey, D. T.;  Hogsett, W. E.; Lee, E. H. (1989)  Analysis of crop loss for alternative
       ozone exposure indices. In:  Schneider, T.; Lee, S.D.; Wolters, G.J.R.; Grant, L.D.,
       eds. Atmospheric  ozone research  and its policy implications: proceedings of the 3rd
       US-Dutch international symposium; May 1988; Nijmegen, The  Netherlands.
       Amsterdam, The Netherlands: Elsevier Science Publishers; pp.  219-227. (Studies in
       environmental science 35).

-------
                                         R-45

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, D. T.; Wilhour,  R. G.; Standley, C. (1976)  The effect of chronic ozone exposures
       on the metabolite content of ponderosa pine seedlings.  For. Sci.  22: 234-241.

Tjoelker, M. G.; Luxmoore, R. J. (1991) Soil nitrogen and chronic ozone stress influence
       physiology, growth and nutrient status of Pinus taeda L. and Liriodendron lulipifera
       L. seedlings. New Phytol.  119: 69-81.

Tobiessen, P. (1982) Dark opening of stomata in successional trees. Oecologia. 52: 356.

Tonneijck, A.  E. G.  (1994) Use of several plant species as indicators of ambient ozone:
       exposure-response relationships. In: Fuhrer, J.; Achermann, B., eds. Critical levels
       for ozone: a UN-ECE  workshop report; November 1993; Bern, Switzerland.
       Liebefeld-Bern, Switzerland:  Swiss Federal Research Station for Agricultural
       Chemistry and Environmental Hygiene; pp. 288-292.

Tonneijck, A.  E. G.; Bugter,  R.  J. F. (1991)  Biological monitoring of ozone effects on
       indicator plants in the Netherlands: initial research on exposure-response functions.
       VDI Ber. (901): 613-624.

Tourney, J.  W.; Korstian, C.  F. (1947) Seedling and planting in the practice of forestry. Ed.
       3. Wiley, New York.

Treshow, M. (1980) Pollution effects on plant distribution.  Environ. Conserv.  7: 279-286.

Tucker, H.G. (1962)  Effects  of air  pollution and temperature on residents of nursing homes
       in the Los Angeles area.  Berkeley, CA:  California State Public Health.

Tyler,  W. S.; Tyler,  N. K.; Magliano,  D. J.; Hinds,  D. M.; Tarkington, B.; Julian, M. D.;
       Hyde, D. M.; Plopper, C. G.; Dungworth, D. L. (1991) Effects  of ozone inhalation
       on lungs of juvenile monkeys. Morphometry after a 12 month exposure and following
       a 6  month post-exposure period. In: Berglund, R.  L.; Lawson, D. R.; McKee, D. J.,
       eds. Tropospheric ozone and  the environment: papers from  an international
       conference; March 1990; Los Angeles, CA. Pittsburgh, PA: Air & Waste
       Management Association;  pp. 151-160. (A&WMA transaction series no. TR-19).

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.

-------
                                        R-46

Umbach, D. M.; Davis, D. D. (1984) Severity and frequency of SO2-induced leaf necrosis
      on seedlings of 57 tree species. For. Sci. 30: 587-596.

U.S. Department of Agriculture. (1993) Forest Health Monitoring: Summary Report: New
      England/Mid-Atlantic, 1992. USDA Forest Service. NE/NA-INF-115-R93.

U.S. Department of Agriculture. (1984) Usual Planting and Harvesting Dates for U.S. Field
      Crops.  Statistical Reporting Service Agricultural Handbook No. 628.

U.S. Department of Health, Education, and Welfare. (1970) Air quality criteria for
      photochemical oxidants.  Washington, DC:  National Air Pollution Control
      Administration; publication  no. AP-63.  Available from:  NTIS, Springfield, VA; PB-
      190262/BA.

U.S. Department of Health and Human Services. (1994) Vital and  health statistics:  current
      estimates from the national  health interview survey,  1992.  Hyattsville, MD: Public
      Health Service, National Center for Health  Statistics; DHHS publication no. (PHS)
      94-1517.

U.S. Environmental Protection Agency. (1996a) Air quality criteria for ozone and related
      photochemical oxidants. Research Triangle  Park, NC: Office of Health and
      Environmental Assessment,  Environmental Criteria and Assessment Office;  EPA
      report nos. EPA/600/P-93/004aF-cF.

U.S. Environmental Protection Agency. (1996b) Air quality criteria for paniculate  matter.
      Research Triangle Park, NC: Office of Health and Environmental Assessment.
      Environmental Criteria and  Assessment  Office; EPA report nos.  EPA/600/P-
      95/OOlcF.

U.S. Environmental Protection Agency.  (1993) EMAP: The Environmental  Monitoring and
      Assessment Program.  Washington, DC: Office of Research and Development.  EPA
      report no: EPA/620/N-93/001, July 1993.

U.S. Environmental Protection Agency.  (1992) Summary of selected  new information on
      effects of ozone on health and vegetation: supplement to  1986 air quality criteria for
      ozone and other photochemical oxidants. Research Triangle Park, NC: Office of
      Health and Environmental Assessment, Environmental Criteria and Assessment
      Office;  EPA report no. EPA/600/8-88/105F. Available from: NTIS, Springfield,
      VA; PB92-235670.

-------
                                        R-47

U.S. Environmental Protection Agency. (1989) Review of the national ambient air quality
       standards for ozone: assessment of scientific and technical information. OAQPS staff
       paper. Research Triangle Park,  NC: Office of Air Quality Planning and Standards;
       EPA report no. EPA-450/2-92/001. Available from: NTIS, Springfield, VA;
       PB92-190446.

U.S. Environmental Protection Agency. (1987) Aerometric Information Retrieval Systems
       User's Guide, Vol. 7.  U.S. Environmental Protection Agency, Office of Air Quality
       Planning and Standards, Technical Support Division, National Air Data Branch:
       Research Triangle Park, North Carolina.

U.S. Environmental Protection Agency. (1986) 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
       report nos. EPA-600/8-84-020aF-eF. 5v. Available from: NTIS, Springfield,  VA;
       PB87-142949.

U.S. Environmental Protection Agency. (1978) 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
       report no. EPA-600/8-78-004. Available from: NTIS, Springfield, VA:  PB80-124753.

Van Bree, L.; Marra, M.; Rombout, P. J. A. (1992) Differences in pulmonary biochemical
       and inflammatory responses of rats and guinea pigs  resulting from daytime or
       nighttime, single and repeated exposure to ozone. Toxicol. Appl.  Pharmacol.  116:
       209-216.

Velissariou, D.; Barnes, J. D.; Davison, A. W. (1992) Has inadvertent selection by  plant
       breeders affected the O3 sensitivity of modern Greek cultivars of spring wheat?  Agric.
       Ecosyst. Environ.  38: 79-89.

Venne, H.; Scholz, F.; Vornweg, A. (1989) Effects of air pollutants on reproductive
       processes of poplar (Populus spp.) and Scots pine (Pinus sylvestris L.).  In: Scholz,
       F.; Gregorius, H.-R.; Rudin, D., eds. Genetic effects of air pollutants in forest tree
       populations: proceedings of the joint meeting of the  IUFRO working parties;  August
       1987; Grosshansdorf, Federal Republic of Germany. Berlin, Federal Republic of
       Germany: Springer-Verlag; pp.  89-103.

Weber, J. A.; Clark, C. S.; Hogsett, W. E. (1993) Analysis of the relationship(s) among O,
       uptake,  conductance, and photosynthesis in needles of Pinus ponderosa. Tree Physiol.
       13:  157-172.

-------
                                        R-48

Weber, J. A.; Gates, D. M.; (1990) Gas Exchange in Quercus rubra (Northern Red Oak)
      during a drought: Analysis of relations among photosynthesis, transpiration, and leaf
      conductance.  Tree Physio. 7:215-225.

Wegner,  C. D.  (1982) Characterization of dynamic respiratory mechanics by measuring
      pulmonary and respiratory system impedances in adult bonnet monkeys (Macaca
      radiata): including the effects of long-term exposure to low-level ozone [Ph.D.
      dissertation]. Davis, CA: University of California, Graduate Division. Available
      from: University Microfilms  International, Ann Arbor, MI; publication no.  8227900.

Weisel,  C. P.;  Cody, R.  P.; Lioy, P. J.  (1995) Relationship between summertime  ambient
      ozone levels and emergency department visits for asthma in central New Jersey.
      Environ. Health Perspect. 103 (suppl. 2):97-102.

Weiss, K.B.; Wagener, D.K. (1990) Changing patterns of asthma mortality: identifying
      target populations at high risk. JAMA J. Am. Med.  Assoc. 264: 1683-1687.

White, M. C.;  Etzel, R. A.; Wilcox, W. D.; Lloyd, C. (1994) Exacerbations of childhood
       asthma and ozone pollution in Atlanta. Environ. Res. 65: 56-68.

Whitfield, R.G.; Biller, W.F.; Jusko, M.J.; Keisler, J.M. (1996) A probabilistic assessment
       of health risks associated  with short-term exposure to tropospheric ozone.   Report
       prepared for U.S. EPA, OAQPS.  Argonne National Laboratory; Argonne, IL,
       August. (For copies, contact Harvey  M. Richmond. U.S. Environmental Protection
       Agency, OAQPS, MD-15, Research Triangle Park,  N.C. 27711, (919) 541-5271.)

Whitfield, R.G.; Richmond, H.M.;  Hayes, S.R.; Rosenbaum, A.S.; Wallsten. T.S.;
       Winkler, R.L.;  Absil, M.J.G.; and Narducci, P. (1994)  Health risk assessment of
       ozone, pp.  129-173 in: D.J.  McKee, ed.  Tropospheric Ozone: Human Health and
       Agricultural Impacts; Lewis  Publishers.

Whittemore, A. S.; Korn, E. L. (1980) Asthma and air pollution in the Los Angeles area.
       Am. J.  Public Health 70: 687-696.

Wiley, J.A.  et al.  (1991a) Study of children's activity patterns, Research Division, California
       Air Resources Board, Sacramento, California, September.

Wiley, J.A.  et al.  (1991b) Activity patterns of California residents, Research Division,
       California Air Resources Board, Sacramento, California, May.

-------
                                        R-49

Wilhour, R. G.; Neely, G. E. (1977) Growth response of conifer seedlings to low ozone
       concentrations. In: Dimitriades,  B., ed. International conference on photochemical
       oxidant pollution and its control:  proceedings, Vol.  II; January; Research Triangle
       Park, NC.  Research Triangle Park, NC: U.S. Environmental Protection Agency; pp.
       635-645; EPA report no. EPA-600/3-77-0015. Available from: NTIS, Springfield,
       VA; PB-264233.

Wiltshire, J. J.  J.;  Wright, C. J.; Unsworth, M. H.;  Craigon, J. (1993) The effects of ozone
       episodes on autumn leaf fall in apple. New Phytol 124: 433-437.

Winkler, R.L.;  Wallsten,  T.S.; Whitfield, R.G.; Richmond, H.M.; Hayes, S.R.; Rosenbaum
       A.S. (1995) An assessment of the risk of chronic lung injury atributable to long-term
       ozone exposure. J. Op. Res. 43,  No. 1:  19-28.

Winner, W.E.;  Lefohn, A.S.; Cotter, I.S.; Greiter, C.S.; Nellessen, J.; McEvoy, L.R.. Jr.;
       Olson, R.L.; Atkinson, C.J.; Moore, L.D. (1989)  Plant responses to elevational
       gradients of O3 exposures in Virginia.  Proc. Natl. Acad. Sci. USA. 86:8828-8832.

Wright, E. S.; Kehrer, J.  P.; White, D. M.; Smiler,  K.  L. (1988) Effects of chronic
       exposure to ozone  on collagen in rat lung. Toxicol.  Appl. Pharmacol. 92: 445-452.

Zeiger, E.; Farquhar, G.  D.; Cowan, I. R. (1987) Stomatal physiology. Stanford, CA:
       Stanford University Press.

-------

-------
      APPENDIX A




Air Quality Assessment

-------
               APPENDIX A:  AIR QUALITY ASSESSMENT

     This Appendix characterizes 03  air  quality  status  and  trends
for the current 1-hour O3  National Ambient Air Quality  Standard
(NAAQS).   Air quality patterns and data handing conventions for
possible new alternative O3  primary  and  secondary  NAAQS are also
presented.  Emphasis is placed on air quality comparisons among
alternative forms of 03  standards, especially within  the current
range of levels of concern.

O3  Trends
     In 1993, hourly O3  measurements made at  925 ambient air
quality monitoring sites were reported to EPA's Aerometric
Information Retrieval System (AIRS).  Most of these sites are
located within, or near, metropolitan areas.   To account for the
seasonal  03  pattern  and  to accommodate  differences in local
climates,  the EPA has designated specific "O, seasons"  in each
state consisting of a contiguous set of months during which
minimal ambient air quality monitoring requirements must be met.
In southern locales, the O3  season spans all  12  months,  while  in
northern states such as Montana, the monitoring season  spans only
the summer months June through September.
          Figure A-l displays the 10-year trend,  1984-93,  in the
composite average and the inter-site variability of the  annual
second highest daily maximum 1-hour concentration at 532 trend
sites.  Only those sites with at least 75 percent data
completeness during the O3 season for at least  8 of  the 10  years
were selected as trend sites.  The  1993 composite average  for
these 532 sites is 12 percent lower than the 1984 level.  The
1993 composite average is higher than the 1992  level, which was
the lowest composite average of the past 10 years.  The  1993
composite average is still the second lowest level during the
past ten years.  The entire concentration distribution  in  1992  is
also lower than any other year.  The increase in the composite
average between 1992 and 1993 is not statistically significant.

-------
                                A-2
Figure A-l.
National  trend in the  annual second  highest daily
maximum 1-hour concentration at 532 sites, 1984-93.
   0.25
       CONCENTRATION, PPM
   0.20-
   0.15-
   0.10-
   0.05-
   0.00
                                                       532 SITES
                                                              NAAQS
             I     I     1     I     I      I     1      I     I     I
           1984   1985  1986  1987  1988  1989   1990  1991  1992  1993

-------
  Figure A-2
                    A-3

    National trend  in the  annual  maximum Sum06  index
    value by county, 1987-93.
    ppm-hrs
        oo

No. of Counties
                Trend in Maximum Sum06 by County
       80.0 -    <705
       700 -
       600 •
       500 -
       40.0 •
       300 -
       200 -
       100
 87      88
421      451
 89
447
 90
469
 91
504
 92
516
                                                           1540   (mu)
 93
543

-------
                               A-4
Figure A-2 presents the trend in the annual three-month maximum
Sum06 value by county for the years 1987-93.  Trends in exposure
indices under consideration for possible secondary standards,
such as the Sum06 index, exhibit temporal patterns similar to the
annual second maximum hourly concentrations described above.
     The interpretation of recent O3 trends  is  difficult  due to
the confounding factors of meteorology and emission changes.
Just as the increase in 1988 is attributed in part to
meteorological conditions that were more conducive to 03
formation than prior years, the 1992 decrease is due in part, to
meteorological conditions being less favorable for O,  formation
in 1992 than in other recent years (Brown et. al., 1992)  .
Meteorological conditions in 1993 were once again more favorable
to 03 formation,  especially in  the east  and  southeastern  areas of
the country (Brown, 1993) although the magnitude and frequency of
O3 concentrations above  the NAAQS  levels were significantly less
than 1988.  Also, since the peak year of 1988, the volatility of
gasoline has been reduced by new regulations which lowered
national average summertime Reid Vapor Pressure  (RVP)  in regular
unleaded gasoline  (EPA, 1989).
     Year-to-year meteorological fluctuations and long-term
trends in the frequency and magnitude of peak O3 concentrations
can have a significant influence on an area's compliance status.
Table A-l presents the number of areas not meeting the current O-,
NAAQS for three compliance periods. The  first compliance period,
1987-89, contains the 1988 peak O, year  and  corresponds  to the
initial nonattainment area designations  under the Clean Air  Act
Amendments of 1990.  By 1990-1992, both  as a result of more
favorable meteorology  (i.e., less conducive  for  O, formation) and
reductions in emissions, the number of areas not meeting the O3
NAAQS has decreased significantly.

-------
                               A-5
Table A-l.   Number of areas not meeting the current O3  NAAQS,
Compliance Period
Number of
Areas not
meeting the
03 NAAQS
1987-89 98
1990-92 52
1991-93 43
1992-94 33
 Data Handling Conventions
      If the current 1-hour standard is replaced with  a standard
 based  on an  8-hour  averaging time,  it will  be  necessary to
 specify  some  initial  data handling conventions for the  8-hour
 averages.  03 data are reported on an hourly basis,  so that it
 is possible to compute a  running 8-hour average O3 concentration
 for  each  hour of the year  (in the  case of complete or  nearly
 complete  data).   For  the analyses presented in this Appendix,
 daily maximum 8-hour averages were considered valid if at least
 18 hourly values  were present during  the day.
      Figure A-3 shows  a hypothetical example of the  relationship
 between the 1-hour data and  the 8-hour averages. The  individual
 bars  correspond to the  hourly  values and  the line shows the
 corresponding 8-hour averages.  There are a few technical  points
 worth  noting.   When an  8-hour average  is  computed,  it  can be
 associated with the start hour,  or the end hour of  the  8-hour
 period,  or some  intermediate  hour.   The  convention  used in
 Figure A-3 is to have  the 8-hour average identified by the start
 hour of the 8-hour period.   For example, the 8-hour  average from
 4  p.m.  to midnight would be  plotted at 4  p.m.   It should be
 noted that the next 8-hour period, from 5 p.m.  to 1 a.m., would
 be plotted  at  5  p.m.  but it contains  hourly  values from two
 different days.   If the 8-hour daily maximum  is selected from

-------
                             A-6
all 24 8-hour averages starting within the day,  it is possible
to have daily  maximums  from adjacent days that  actually have
some hourly values in common.   This is unlikely for the typical
urban diurnal pattern shown in Figure A-3 but it can occur for
sites with less  pronounced diurnal patterns,  such  as  rural
mountaintop sites.
     Another point to note with 8-hour averages is that the 8-
hour daily maximum for a particular day can actually be higher
than the 1-hour maximum.   Again, this would be uncommon but can
happen  because  the 8-hour  average  could contain up  to  seven
hourly  values  from an adjacent day.  Figure  A-4  shows 8-hour
averages for three days in July 1988 at a  rural mountaintop site
in  southeast  Virginia illustrating  these points.   The  daily
maximum for July  7 is higher than any of the 1-hour values for
that day because  it is driven by the high hourly values in the
early hours  of July 8.   Again,  it should be  noted that this
particular rural  site departs significantly from the classical
03 diurnal pattern in urban areas.

-------
                              A-7
Figure A-3.
1-hour 03 data and 8-hour averages.
urban area, three days  in  July.
      Hypothetical
   0.2
  0.15
   0.1
  0.05
             July 6
              July?
July8
1-HOUR
 DATA
                                             8-HOUR
                                            AVERAGE

                                           OVERLAPPING
                                            PERIODS
      MDN 6AM NOON 6PM MDN 6AM NOON 6PM MDN 6AM NOON 6PM MDN

-------
Figure A-4.
                              A-8
1-hour  O3  data  and  8-hour  averages
County, Virginia, July  6-8,  1988.
                                                           Smyth
         0.15 -
         0.05 -
             MDN 6AM NOON 6PM MDN 6AM NOON 6PM  MDN 6AM NOON 6PM MDN

-------
                             A-9
Monitoring Considerations
     Implementation of an 8-hour or cumulative index standard
would require the monitoring community to examine appropriate
siting criteria for the new standard forms.  The current O3
monitoring network was designed to address a 1-hour standard.
While extension to an 8-hour standard may require few changes,
the current network may not be adequate for a cumulative index
standard, particularly if there is a need for more rural
monitoring stations.  A preliminary analysis of 192 urban
areas (Consolidated Metropolitan Statistical Areas and
Metropolitan Statistical Areas, CMSAs/MSAs) for the 1990-92
period found that the 1-hour and 8-hour "design value" sites
differed about 17 percent of the time.  The peak SUM06 site
differed from the peak 1-hour site about 28 percent of the
time.

Distributions of hourly 03  data
     The distributions of hourly 03 concentrations  were
examined at different monitoring environments.  Sites with
relatively complete data located in rural/agriculture,
background, urban, and high elevation environments were
selected.  The data were analyzed using the MAXFIT program
(Fitz-Simons et al., 1979)  which fits 8 distributions: the
normal,  the 3-parameter log-normal, the Box-Cox distribution,
the Johnson SB,  the 3-parameter gamma,  the. 4-parameter beta,
the 3-parameter Weibull, and the extreme value distribution.
Figures A-5 through A-8 display the frequency distributions  of
the hourly concentrations and best-fit distributions for an
example site in each category.
     The urban site shown in Figure A-5 is located at Taft
High School in Chicago.  For the hourly data, the normal and
extreme value distributions gave the best fit.
     The high elevation site shown in Figure A-6 is in
Albuquerque, New Mexico.  The normal distribution did about  as

-------
                             A-10
well as any for these data.  The data at the lower values
almost exhibited a uniform distribution.
     The background environment site shown in Figure A-7 is
located at Clemson University in South Carolina.  The hourly
data histograms resembled the hourly data for the Albuquerque
site.  The normal distribution produced the best
approximations or fit for these data.
     The Percy Priest Lake Visitor Center near Nashville
Tennessee shown in Figure A-8 represents a site in a
rural/agriculture-forest environment.  The hourly data which
are skewed, and have a long tail toward the higher
concentrations, were fit best by the normal and the Box-Cox
transformation.
     It is clear from the above results that there are
differences across monitoring environments in the distribution
of hourly O,  concentrations.   Nevertheless,  some interesting
observations come to light from this exercise. One, the normal
distribution displayed good fitting characteristics more often
than one would think since the data are usually considered to
be too skewed  for the normal distribution.  Two, the extreme
value distribution did not consistently yield the best fit on
the  1-hour and 8-hour daily max data.  Finally, the sites
selected through the process described above did not record
very large 03 concentrations  during recent years.   One  would
expect sites with high values of O3 to exhibit data with a
more pronounced coefficient of skewness.

-------
                                    A-ll
       Figure A-5.
   Hourly 0, concentration distribution at  an urban
   monitoring site.
                     Chicago, Taft H.S. 1993 Hourly Data
Normal Distribution
      -(X-H)
f(X) = e
= 0.016370
= 0.000203
     /H
                 \

-------
                                  A-12
    Figure A-6.
   Hourly O3  concentration  distribution  at  a high
   elevation  monitoring site.
Normal Distribution
                      Albuquerque, NM 1991 Hourly Data
f(X) = e
H   = 0.024857
a2  =0.(

-------
                                     A-13
       Figure A-7.
Hourly 03 concentration distribution at an urban
background monitoring site.
                      Clemson University 1992 Hourly Data
Normal Distribution
       ~^T     n   =59.945717
f(X) = e	    02  =1607.694448

-------
                                 A-14
   Figure  A-8.
Hourly 03 concentration  distribution at a  rural

monitoring site.
                 Percy Priest Lake Vis. Center 1991 Hourly Data
Box-Cox Distribution
         x°-i
           5-
f(X) =£
 ^i  =-1.889480

 a2 = 0.025174
 a  =0.429649
       \_
                 \

-------
                             A-15
Air Quality Data Base for Standards Comparisons
     Design values consistent with each of the alternative
standards were calculated for each county with sufficient
monitoring data.  Design values are used in these comparisons
since, although they are based on the form of the standard,
they are independent of the level of the standard.  The design
value is that concentration that when reduced to the level of
the standard ensures compliance with the standard.  For
example, the design value for a daily maximum 1 exceedance per
year standard is the fourth highest daily maximum
concentration given a three year compliance period.  If the
fourth highest day during three years is reduced to the level
of the standard, then there will be exactly three days above
the level of the standard, or one day per year on average.
Similarly, if the standard were expressed as the average
annual 2nd highest daily maximum concentration over three
years, the design value is just that same average annual
second highest daily maximum concentration.  Thus, design
values enable one to map complex standard forms into to single
number for ease of comparison.
     These design value estimates used all hourly 0,
concentration data available on AIRS for the years 1987-93.
One-hour data were processed according to the conventions for
the current 03  standard.   That  is,  the  design  value  for  a  1
exceedance standard is the (n* +1)  largest value,  where "n" is
the number of years meeting the annual 75 percent data
completeness requirement.  Eight-hour averages were computed
from the 1-hour data and associated with the start hour.  An
8-hour average had to have at least six hourly values.  The
daily maximum 8-hour average is the highest of the 24 possible
8-hour averages starting within the calendar day.   These 8-
hour daily maximums were then processed using the same
conventions analogous to those for the 1-hour data.  For the 5
exceedance option, these basic principles were extended so

-------
                             A-16
that the sixth highest value would be used if only one year of
data were available, the llth if two years were available, and
the 16th if all three years were valid.
     The design values for the three cumulative index
alternatives, Sum06, W126, and AOT06, were based upon the
highest consecutive three-month value for any calendar year in
the 1991-93 period.  A index total was computed for each day
and then totaled for the month.  Daily index values were
considered valid if the day had at least 18 hourly values.  A
month had to have at least 50 percent valid days for the month
to be considered valid.

Missing Data Considerations for Secondary Standards
     A missing value adjustment was used to scale up index
values for a month by simply multiplying the observed index by
the ratio of days in the month divided by days with valid
data.  A preliminary analysis of the impact of missing data on
index values was conducted.  Sites were selected for four
groups - rural/forest and rural/agriculture, high elevation,
background, and urban and city center.  These sites were
selected from AIRS based on the location-setting code, land
use code, elevation, and monitoring objective code.
Rural/forest and rural/agriculture were selected including
areas classified as desert or blighted areas.  However, most
sites are forest and agricultural.  The high elevation sites
were selected by limiting the site elevation to above 1000
meters.  The background sites were selected on the basis  of
their monitoring objective.
     All subsets of sites exhibited the same patterns of
missing data.  All groups had a large  number of single missing
hours.  These are mostly the missing hour each day for
calibration purposes.  For longer gaps, the frequency falls
off  exponentially as gap  length grows  larger.  The starting
hour for a series of missing data seems to have no pattern  in
all  groups.  There  are some hours in the background sites that

-------
                             A-17
seem to have a larger percentage, although this is likely due
to chance since there a much smaller number of sites falling
in this category.  Site maintenance seems to be standard
across different groups of sites and we can expect to see the
same patterns of missing data in rural areas as in urban
areas.
     A site-year combination was selected that had a very high
data capture (Tucson, AZ).  This site then had gaps of data
systematically produced and all secondary standard alternative
indicators were calculated and adjusted for missing data.
These were compared to the same indicators calculated without
any data taken out and the relative error was calculated as:
                         RE =
SA - SU
  SU
where RE is the relative error, SA is the standard calculated
using data with gaps and SU is the standard calculated with
all data.  The relative error was averaged over several data
sets where gaps were shifted across the critical months used
to calculate alternative indicators for the secondary
standard.  The results appear below.
                  Table A-2.  Relative Errors
Gap Size
10
50
100
200
300
SumOG
0. 0118
0.0201
0.0221
0.0571
0.0842
AOT06
0. 0030
0.0014
-0.0048
0.0360
0.0818
W126
0. 0086
0.0124
0. 0120
0.0316
0.0508
The AOT06 indicator option seems to have a lower average

-------
                             A-18
relative error until the gaps become larger than 100 hours in
length.  The SUM06 indicator seems to be the most sensitive at
all levels.

Multi-year Compliance Considerations for Secondary Standards
     In the air quality comparisons that follow, it is
important to note that use of the maximum year in the three
year compliance period interpretation for the secondary
standard index options is a mixture of the old "once per year"
standard and the newer "expected exceedance" standards.  These
maximum cumulative exposure indices do not represent a multi-
year average so it differs from the form of current O,  NAAQS.
A single year compliance test is consistent with providing
protection for annual species.  However, if only a single year
is used then, for all practical purposes, by the time the
decision is made to declare an area nonattainment, the data
from the next year are almost complete and might reverse the
decision.  To avoid this situation, earlier attainment
decisions  for these types of standards reguired eight
consecutive guarters of data to show attainment  (EPA, 1978).
Three  calendar years were selected for the compliance period
for consistency with the primary standard.  The 1991-93 air
guality data were used to compare the difference between the
maximum three-month index value and the three year average of
the annual maximum three month index values.  The Sum06 value
in the maximum year was at least 24 percent larger than the
average index value across the three years for half of the
counties with monitoring data. The maximum was greater than
the average by more than 22 percent for W126, and 29 percent
for AOT06,  for half of the counties with data.

Air Quality Comparisons
     Using the ambient air quality data base and the
alternative primary and secondary NAAQS, the staff estimated

-------
                             A-19
the number of counties and metropolitan areas which would
meet, or fail to meet, these standards based on data for 1987-
89 and 1991-93.  These two 3-year periods contrast the impact
of 1988, and varying weather patterns with the current
compliance period.
     This assessment used design values, rather than expected
exceedances, to facilitate the comparisons among varying
standard levels.  For the one exceedance standard options, the
design value is simply the fourth highest concentration during
the three year period.  Because one exceedance is allowed for
each year,  if the fourth highest value is greater than the
specified standard level, then the county has failed to meet
the alternative NAAQS.  Similarly, for the five exceedance
alternative, the design value is the sixteenth largest daily
maximum 1-hour concentration.  Once the design value is
calculated,  comparisons with alternative standard levels can
be made directly without having to recompute the number of
exceedances.  Table A-3 presents the air quality data
comparisons on a county basis.

-------
                             A-20
Table A-3.
Number of counties not meeting selected
analytic options based on design values for
1987-89 and 1991-93.
ANALYTIC OPTIONS
Primary Standard
1-hour, 1 exceedance, 0.12 ppm
, avg annual 2nd max, 0.12
, 1 exceedance, 0.10 ppm
, avg annual 2nd max, 0.10
8-hour, 1 exceedance, 0.10 ppm
, 1 exceedance, 0.09 ppm
, avg annual 2nd max, 0.09
, 3 exceedance, 0.09 ppm
, 5 exceedance, 0.09 ppm
, 1 exceedance, 0.08 ppm
, 3 exceedance, 0.08 ppm
, 5 exceedance, 0.08 ppm
, avg annual 2nd max, 0.08 ppm
, avg annual 3rd max, 0.08 ppm
, avg annual 4th max, 0.08 ppm
, avg annual 5th max, 0.08 ppm
, 1 exceedance, 0.07 ppm
Total Number of Counties with monitors
Total number of Counties in the U.S.
1987
-89

224
168
384
338
236
332
279
263
190
419
370
321
379
352
333
305
467
506
3142
1991
-93

104
72
310
243
103
221
160
113
67
394
274
194
329
274
227
187
514
581
3142

-------
                             A-21
The impact of using the average annual second maximum
(AvgMax2) concentration as the NAAQS statistic on the number
of counties not meeting selected standard alternatives is
shown in Table A-3.  For each alternative, there are fewer
non-compliant counties for the AvgMax2 statistic (about 20-30%
less) than the current one exceedance form.   Table A-4
presents a comparison of the impact of the selected analytic
options on the compliance status of the original 98 areas
designated nonattainment for O3  under  the  Clean  Air  Act
Amendments of 1990.  A detailed listing of design values for
each alternative by nonattainment area is provided in Table A-
13 at the end of this Appendix.
Table A-4.
Number of original nonattainment areas not
meeting selected analytic options based on
1991-93 air quality monitoring design values.
      ANALYTIC OPTIONS
           1-hour,    1 exceedance,  0.12 ppm
                 ,    avg annual 2nd max,  0.12
                 ,    1 exceedance,  0.10 ppm
                 ,    avg annual 2nd max,  0.10
           8-hour,    1 exceedance,  0.09 ppm
                 ,    avg annual 2nd max,  0.09

                 ,    1 exceedance,  0.08 ppm
                 ,    avg annual 2nd max,  0.08
                 ,    avg annual 5nd max,  0.08
                 ,    5 exceedance,  0.08 ppm
                 ,    1 exceedance,  0.07 ppm
      Number of Areas with monitors.
                                  1991-93
                                     39
                                     34

                                     87
                                     69
                                     72
                                     52

                                     93
                                     85
                                     64
                                     63
                                     97
                                     97

-------
                             A-22
Table A-5 compares the number of counties in both compliance
periods with design values in excess of the current 1-hour
standard with selected 8-hour one exceedance alternatives.
Despite the larger number of counties with data in 1991-1993,
the impact of O3 data  from the  summer  of  1988  is  readily
apparent in the earlier compliance period.  The number of
counties not meeting an 8-hour, one expected exceedance
standard with a level of 0.10 ppm is quite close to the number
under the current 1-hour 0.12 ppm standard.  Tables A-6
through A-8 contrast the current 1-hour standard with the
levels of concern for the Sum06, W126 and AOT06 secondary
standard options.  Small differences in the total number of
counties with monitors among these tables and Table A-5 result
from the data completeness requirements for the secondary
standard alternatives.  For the levels of concern, the Sum06
standard yields almost four times as many non-compliance areas
as the current  1-hour primary standard.  Table A-9 contrasts
the one and five exceedance options for an 8-hour standard.
For both compliance periods, an 8-hour five exceedance
standard of 0.08 ppm is comparable to an 8-hour one exceedance
standard of 0.09 ppm in terms of the number of non-complying
areas.  Table A-10 contrasts the alternative 8-hour one
exceedance options with the three peak-weighted indices.
Finally, Tables A-ll and A-12 repeat the 1-hour and 8-hour
comparisons on  a metropolitan area, rather than a county,
basis.
     Another way to view the information presented above  is  in
terms  of ratios among design values for alternative standards.
For  example, the average  ratio  of 8-hour to 1-hour design
values for a 1  exceedance  per year standard is 0.86 based  on
1991-93  data.   This ratio  has been fairly  stable  over time
increasing slightly from  0.81  in 1980 to  0.86  in  1993.  The
ratio  is also  fairly  consistent across EPA Regions, with  the
median ratio ranging  from  0.80  to 0.88.   The median 8-hour/1-

-------
                             A-23
hour 1 exceedance ratio for nine remote O, monitors in AIRS
was slightly higher, 0.90, which would be expected for sites
with less pronounced diurnal patterns.  As noted above in
Table A-5, the number of counties not meeting a 0.10 ppm 8-
hour 1-exceedance standard is almost identical to the number
not meeting a 0.12 ppm 1-hour standard.  This is consistent
with the 0.86 ratio because the ratio of the corresponding
exceedance levels for the two standards (0.105 ppm/0.125 ppm)
is equal to 0.84.
     Design value comparisons based on 1991-93 data show that,
on average, ozone design values for an 8-hour, 0.08 ppm, 1
exceedance standard are about 15% lower than for the current
1-hour standard, while design values for both an 8-hour, 0.08
ppm, 5 exceedance, and average annual 5th highest daily
maximum standard are about 25% lower than the current 1-hour
standard.  On average,  design values for an average annual 2nd
highest daily maximum 8-hour standard are 18% lower than the
current 1-hour standard.
     In terms of exceedances, on average,  sites meeting an
average annual 2nd highest daily maximum 0.08 ppm standard
have 1.2 exceedances per year, and 2.3 exceedances in the
worst year of three, while sites meeting an average annual 5th
highest daily maximum 0.08 ppm standard have 3.0 exceedances
per year, and 5.4 exceedances in the worst year of three.
Also,  in the worst year of three, 95 % of sites meeting the
average annual 2nd highest daily maximum 0.08 ppm standard
have 7 or fewer exceedances,  while 95 % of sites meeting an
average annual 5th highest daily maximum 0.08 ppm standard
have 12 or fewer exceedances.
     Table A-13 lists design values based on 1991-93
monitoring data for each of the 1-hour and 8-hour standard
alternatives in the original 98 areas designated nonattainment
in 1991.   The maximum Sum06 index value for 1991-93 is also
provided in the last column.

-------
                             A-24

                          Table A-5.

Comparison of design values for the current .12 ppm 1 hour,
1 exceedance standard  vs. an 8 hour, 1 exceedance standard,
by county.

                          1987  -  1989
NAAQS
. 12 ppm
1 Exc


DV <= .12
DV > .12
Total
8 Hour, 1 Exceedance NAAQS

DV
<=
.07
39
0
39

.07 <
DV
<=.08
48
0
48

.08 <
DV
<=.09
84
3
87

.09 <
DV
<=. 10
759
21
96

DV
' >
.10
36
200
236
Number
of
Counties


282
224
506
                          1991  - 1993
NAAQS
. 12 ppm
1 Exc


DV <= .12
DV > .12
Total
8 Hour, 1 Exceedance NAAQS

DV
<=
.07
67
0
67

.07 <
DV
<=.08
120
0
120

.08 <
DV
<=.09
171
2
173

.09 <
DV
<=. 10
100
18
US

DV
>
. 10
19
84
103
Number
of
Counties


477
104
581

-------
                             A-25

                          Table A-6.

Comparison of design values for the current .12 ppm 1 hour,
1 exceedance standard vs. max Sum06 design values, by county.

                          1987  -  1989
NAAQS
. 1 2 ppm
DV <= . 12
DV > .12
Total
Sum06 Max Year NAAQS
DV <= 16.5
57
5
62
16.5 < DV <=26.4
44
10
54
DV > 26.4
175
209
384
Number
of
Counties
276
224
500
                          1991  -  1993
NAAQS
.12 ppm
1 Exc
DV <= .12
DV > .12
Total
Sum06 Max Year NAAQS
DV <= 16.5
137
2
139
16.5 < DV <=26.4
126
16
142
DV > 26.4
211
86
297
Number
of
Counties
474
104
578

-------
                             A-26

                          Table A-7.

Comparison of design values for the current .12 ppm 1 hour,
1 exceedance standard vs. max W126 design values,  by county.

                          1991  -  1993
NAAQS
. 12 ppm
1 Exc
DV <= .12
DV > .12
Total
W126 Max Year NAAQS
DV <= 13.8
144
1
145
13.8 < DV <=22.4
146
16
162
DV > 22.4
184
87
271
Number
of
Counties
474
104
578
                          Table A-8.

Comparison of design values for the current .12 ppm 1 hour,
1 exceedance standard vs. max AOT06 design values, by county.

                          1991  -  1993
NAAQS
. 12 ppm
1 Exc
DV <= .12
DV > .12
Total
AOT06 Max Year NAAQS
DV <= 4.8
285
9
294
4.9 < DV <= 7.5
106
21
127
DV > 7.5
83
74
157
Number
of
Counties
474
104
578

-------
                             A-27

                          Table A-9.

Comparison of design values for an 8 hour, I exceedance
standard vs. design values for an 8 hour, 5 exceedance
standard, by county.

                          1987  -  1989
8 Hour
1 Exc
NAAQS
DV <= .07
.07 <= .08
.08 <= .09
.09 <= .10
DV > .10
Total
8 Hour 5 Exc NAAQS
DV <=.08
38
48
71
25
2
184
DV > .08
0
0
16
71
234
321
Number
of
Counties
38
48
87
96
236
505
                         1991 - 1993
8 Hour
1 Exc
NAAQS
DV <= .07
.07 <= .08
.08 <= .09
.09 <= .10
DV > .10
Total
8 Hour 5 Exc NAAQS
DV <=.08
67
120
159
39
2
387
DV > .'08
0
0
14
79
101
194
Number
of
Counties
67
120
173
118
103
581

-------
                             A-28

                          Table  A-10.

Comparison of design values for an 8 hour,  1 exceedance standard
vs. alternative secondary standard exposure indices, by county.
                          1991  -  1993
8 Hour
1 Exc
NAAQS
DV <= .07
.07 <= .08
.08 <= .09
.09 <= .10
DV > .10
Total
No.
of
Counties
64
120
173
118
103
578
Alternative Secondary NAAQS
Exposure Indices
10% Crop Loss
Sum 06
> 26.4
2
14
92
96
93
297
-W126
> 22.4
1
7
76
93
94
269
AOT06
> 7.5
0
0
18
55
84
157

-------
                             A-29

                          Table A-ll.
Comparison of design values for the current  .12 ppm  l hour,
1 exceedance standard  vs.  an  8  hour,  1 exceedance  standard, by
metropolitan area (CMSA/MSA).

                          1987 -• 1989
NAAQS
. 12 ppm
1 EXC


DV <= . 12
DV > .12
Total
8 Hour, 1 Exceedance NAAQS


DV
<=
.07
8
0
8

.07 <
DV
<=.08
19
0
19

.08 <
DV
<=.09
34
1
35

.09 <
DV
<=.10
30
8
38

DV
>
.10
14
72
86
Number
of Areas
(CMSAs
& MSAs)

106
80
186
                          1991 - 1993
NAAQS
. 12 ppm
1 Exc


DV <= .12
DV > .12
Total
8 Hour, 1 Exceedance NAAQS

DV
<=
.07
14
0
14
.07 <
DV
<=.08
38
0
38
.08 <
DV
<=. 09
63
2
65
.09 <
DV
<=. 10
30
4'
34
DV
>
. 10
8
36
44
Number
of Areas
( CMSAs
4 MSAS)

153
42
195

-------
                             A-30

                          Table  A-12.
Comparison of design values for an 8 hour,  I exceedance standard
vs. design  values  for  an 8  hour,  5 exceedance  standard,  by
metropolitan area (CMSA/MSA).
                          1987  -  1989
8 Hour
1 Exc
NAAQS
DV <= .07
.07 <= .08
.08 <= .09
.09 <= .10
DV > .10
Total
8 Hour 5 Exc NAAQS
DV <=.08
8
19
29
10
1
67
DV > .08
0
0
6
28
85
119
Number
of Areas
(CMS As
& MSAs)
8
19
35
38
86
186
                          1991 - 1993
8 Hour
1 Exc
NAAQS
DV <= .07
.07 <= .08
.08 <= .09
.09 <= .10
DV > .10
Total
8 Hour 5 Exc NAAQS
DV <=.08
12
38
61
8
0
121
DV > .08
0
0
4
26
44
Number
of Areas
(CMSAs
& MSAs)
14
38
65
34
44
74 1 195

-------
Table A-13.
   Nonattainment
   Area Name
                      Comparison  of design values for alternative 03 air quality standards
                      for the original 98 classified nonattainment areas.
   Albany-Schenectady-Troy NA  Area
   Allentown-Bethlehem-Easton  NA
   Altoona NA Area
   Atlanta NA Area
   Atlantic City NA Area
   Baltimore NA Area
   Baton Rouge NA Area
   Beaumont-Port Arthur  NA Area
   Birmingham NA Area
   Boston-Lawrence-Worcester NA Area
   Buffalo-Niagara Falls NA Area
   Canton NA Area
   Charleston NA Area
   Charlotte-Gastonia NA Area
   Cherokee Co NA Area
   Chicago-Gary-Lake County NA Area
   Cincinnati-HamiIton NA Area
   Cleveland-Akron-Lorain NA Area
   Columbus NA Area
   Dal las-Fort Worth NA  Area
   Dayton-Springfield NA Area
   Detroit-Ann Arbor NA  Area
   Door  Co NA  Area
   Edmonson Co NA Area
   El  Paso NA  Area
   Erie  NA Area
   Essex Co NA Area
   EvansviIle  NA Area
   Grands Rapids NA Area
   Greater Connecticut NA Area
   Greenbrier  NA Area
   Greensboro-Winston-Sal em-High
   Hancock Co  and Waldo  Co NA Area
   Harrisburg-Lebanon-Carlisle NA
   Houston-Galveston-Brazoria NA
   Huntington-Ashland NA Area
   Indianapolis NA  Area
   Jefferson Co NA  Area
   Jersey Co NA Area
   Johnstown NA Area
   Kansas City NA Area
   Kent  County and  Queen Anne's County
   Kewaunee  Co NA Area
   Knox  Co and Lincoln Co NA Area
   KnoxviIle NA Area
   Lake  Charles NA  Area
   Lancaster NA Area
   Lewiston  -  Auburn NA  Area
   Lexington-Fayette NA  Area
   Los Angeles South  Coa^t  Air Bnsin
Clean Air Act
Class if icat ion
Marginal
Marginal
Marginal
Serious
Moderate
Severe 15
Serious
Serious
Marginal
Serious
Marginal
Attainment
Attainment
Attainment
Attainment
Severe 17
Moderate
Moderate
Attainment
Moderate
Attainment
Attainment
Marginal
Attainment
Serious
Marginal
Marginal
Marginal
Moderate
Severe 17
Attainment
Attainment
Marginal
Marginal
Severe 17
Attainment
Attainment
Marginal
Attainment
Marginal
Attainment
y Marginal
Moderate
Moderate
Attainment
Marginal
Marginal
Moderate
At tainmont
E x 1 1 emo
D.V.
1h, lex
0.10
0.12
0.11
0.15
0.12
0.15
O.K
0.13
0.12
O.K
0.11
0.11
0.11
0.12
0.10
0.15
0.13
0.14
0.12
0.14
0.11
0.12
0.13
0.09
O.K
0.11
0.12
0.11
0.15
0.16
0.10
0.11
0.11
0.11
0.20
0.12
0.10
0.11
0.11
0.11
0.11
0.13
0.11
0.13
0.12
0.12
0.12
0.11
0.10
0.30
D.V.
8h, 1ex
0.09
0.10
0.09
0.13
0.11
0.12
0.11
0.11
0.10
0.11
0.10
0.10
0.09
0.10
0.09
0.12
0.11
0.12
0.10
0.12
0.10
0.11
0.10
0.08
0.09
0.09
0.10
0.10
0.13
0.12
0.09
0.10
0.10
0.10
0.13
0.10
0.10
0.11
0.09
0.10
0.09
0.11
0.09
0.12
0.10
0.09
0.10
0.09
0.09
0.19
Avg Max2
8 hr
0.09
0.10
0.09
0.11
0.11
0.12
0.11
0.11
0.10
0.12
0.09
0.09
0.07
0.10
0.09
0.11
0.10
0.13
0.09
0.11
0.09
0.10
0.09
0.08
0.09
0.09
0.10
0.09
0.13
0.12
0.08
0.09
0.10
0.10
0.12
0.09
0.09
0.10
0.09
0.09
0.09
0.11
0.09
0.11
0.09
0.09
0.10
0.09
0.08
0.19
Avg Max5
8 hr
0.08
0.09
0.08
0.10
0.10
0.11
0.09
0.10
0.09
0.09
0.08
0.09
0.07
0.09
0.08
0.10
0.09
0.11
0.08
0.09
0.09
0.09
0.09
0.07
0.08
0.08
0.09
0.09
0.11
0.11
0.08
0.09
0.08
0.09
0.10
0.09
0.09
0.09
0.08
0.08
0.08
0.10
0.08
0.10
0.09
0.09
0.09
0.08
0.08
0.18
D.V.
8h, Sex
0.08
0.09
0.08
0.10
0.10
0.10
0.09
0.09
0.09
0.10
0.08
0.09
0.07
0.09
0.08
0.10
0.09
0.10
0.08
0.09
0.08
0.09
0.08
0.07
0.08
0.08
0.09
0.09
0.11
0.11
0.08
0.09
0.08
0.09
0.10
0.09
0.09
0.08
0.08
0.08
0.08
0.10
0.08
O.C9
0.09
0.09
0.09
0.08
0.08
0.17
Sum06
8am- 8pm
15.8
28.7
30.0
50.0
47.9
55.9
23.0
26.8
25.1
29.0
22.3
45.5
29.7
50.6
32.0
38.9
46.1
45.6
31.0
39.2
35.0
30.1
19.5
17.1
15.7
30.8
23.0
40.5
40.3
35.9
22.9
48.3
13.2
36.5
30.0
35.2
38.3
18.4
26.9
29.5
32.3
48.9
18.1
20.6
42.7
20.4
38.0
13.4
29.4
111.0

-------
Table A-13.  (cont.)
   Nonattainment
   Area Name
Comparison of design values  for  alternative 03 air quality standards
for the original  98 classified nonattainment areas.
   LouisviIle NA Area
   Manchester MA Area
   Hanitowoc Co NA Area
   Memphis  NA Area
   Miami-Fort Lauderdale-W.  Palm Beach
   Milwaukee-Racine NA Area
   Monterey Bay Unified MA Area
   Muskegon NA Area
   Nashville NA Area
   New York-N. New Jersey-Long  Island
   Norfolk-Virginia Beach-Newport
   Owensboro NA Area
   Paducah  NA Area
   Parkersburg NA Area
   Philadelphia-WiImington-Trenton
   Phoenix
   Pittsburgh-Beaver Valley  NA Area
   Portland  NA Area
   Port land-Vancouver AQMA NA Area
   Portsmouth-Dover-Rochester, NH
   Poughkeepsie NA Area
   Providence (all  of RI)  NA Area
   Raleigh-Durham NA Area
   Reading NA Area
   Reno
   Richmond-Petersburg NA  Area
   Sacramento Metro NA Area
   Salt  Lake City-Ogden NA Area
   San Diego NA Area
   San Francisco-Bay NA Area
   San Joaquin Valley NA Area
   Santa Barbara - Santa Maria  - •
   Scranton-WiIkes-Barre NA  Area
   Seattle  -  Tacoma NA Area
   Sheboygan NA Area
   Smyth Co NA Area
   South Bend-Elkhart NA Area
   Southeast Desert Modified AOMD
   Springfield (W. Mass) NA  Area
   St.  Louis NA Area
   Sussex Co NA Area
   Tampa-St.  Petersburg-Clearwater
   Toledo NA Area
   Ventura  Co NA Area
   Walworth Co NA Area
   Washington NA Area
   York  NA  Area
   Youngstown-Uarren-Sharon  NA  Area
Clean Air Act
Classification
Moderate
Marginal
Moderate
Attainment
:h Attainment
Severe 17
Moderate
Moderate
Moderate
I Severe 17
Marginal
Attainment
Attainment
Attainment
Severe 15
Moderate
Moderate
Moderate
Marginal
Serious
Moderate
Serious
Attainment
Moderate
Marginal
Moderate
Serious
Moderate
Severe 15
Attainment
Serious
Moderate
Marginal
Marginal
Moderate
Marginal
Attainment
Severe 17
Serious
Moderate
Marginal
Attainment
Attainment
Severe 15
Marginal
Serious
Marginal
Marqi nal
O.V.
1h, 1ex
0.13
0.09
0.13
0.12
0.11
0.15
0.11
O.H
0.12
0.15
0.13
0.10
0.11
0.12
0.16
0.15
0.12
0.13
0.11
0.13
0.13
0.15
0.12
0.12
0.09
0.13
0.15
0.11
0.15
0.12
0.16
0.12
0.12
0.11
O.H
ND
0.10
0.20
O.H
0.13
0.12
0.11
0.12
0.15
0.12
O.H
0.11
0.11
D.V.
8h, 1ex
0.11
0.08
0.10
0.10
0.09
0.12
0.09
0.11
0.11
0.13
0.10
0.09
0.09
0.10
0.13
0.11
0.11
0.11
0.10
0.11
0.10
0.12
0.10
0.11
0.08
0.10
0.12
0.09
0.12
0.09
0.12
0.10
0.11
0.09
0.11
ND
0.09
0.15
0.11
0.11
0.11
0.08
0.10
0.13
0.10
0.11
0. 10
0.10
Avg Max2
8 hr
0.10
0.08
0.10
0.09
0.08
0.11
0.09
0.11
0.10
0.12
0.10
0.08
0.09
0.09
0.12
0.09
0.10
0.11
0.08
0.11
0.10
0.11
0.09
0.10
0.07
0.10
0.12
0.08
0.12
0.09
0.12
0.10
0.10
0.08
0.09
ND
0.09
0.15
0.11
0.11
0.11
0.08
0.09
0.13
0.09
0.11
0.09
0.10
Avg Max5
8 hr
0.09
0.07
0.08
0.09
0.08
0.10
0.08
0.11
0.09
0.11
0.09
0.08
0.08
0.09
0.12
0.09
0.09
0.10
0.07
0.09
0.10
0.10
0.09
0.09
0.07
0.09
0.11
0.08
0.11
0.08
0.11
0.09
0.09
0.08
0.08
ND
0.08
0.14
0.10
0.09
0.09
0.08
0.08
0.11
0.08
0.10
0.09
0.09
D.V.
8h, 5ex
0.09
0.07
0.08
0.09
0.07
0.10
0.08
0.10
0.09
0.11
0.09
0.08
0.08
0.09
0.12
0.09
0.09
0.09
0.07
0.09
0.09
0.10
0.09
0.09
0.07
0.09
0.11
0.08
0.11
0.08
0.11
0.09
0.09
0.08
0.09
ND
0.08
O.H
0.10
0.09
0.09
0.08
0.08
0.12
0.09
0.10
0.09
0.09
Sum06
Sam- 8pm
49.5
8.9
25.6
37.9
14.3
38.9
32.9
36.5
44.5
45.1
46.4
24.1
34.8
37.1
58.6
39.1
38.3
21.2
14.5
17.1
26.0
29.8
59.9
34.4
23.9
47.3
65.8
25.9
54.9
19.9
83.2
44.3
30.0
12.3
29.2
ND
38.8
103.9
24.6
42.4
41.6
22.0
21.7
72.9
35.8
50.1
40.2
37.7
                                                                                                                            U)
                                                                                                                            i-o

-------
                             A-33

REFERENCES

Brown, W. 0. ;  R .R. Heim  (1992)  Climate  variations bulletin,
Vol. 4, No.  8.   Available  from National Climatic Data Center,
NOAA, Asheville, N.C.

Brown, W.  O.  (1993)  Climate variations bulletin, Vol. 5, No. 8.
Available from National Climatic Data Center, NOAA, Asheville,
N.C.

U.  S.  Environmental   Protection  Agency  (1989)   Volatility
regulations  for  gasoline  and alcohol blends  sold  in calendar
years 1989 and beyond. Federal Register 54: 11868.

U. S.  Environmental  Protection Agency  (1978) Section 107 states
attainment status.  Federal Register 43:  8962.

Fitz-Simons,  T. ;  D.M. Holland  (1979)  The  maximum likelihood
approach to  probabilistic  modeling  of air quality  data.   EPA-
600/4-79-044. U.S.  Environmental Protection Agency.

-------

-------
                       APPENDIX B

     8-Hr Daily Maximum Dose Exposure Distributions
For Outdoor Children Under Various Air Quality Scenarios

-------

-------
                                    APPENDIX B
       Figures B.I through B.8 are graphs from Johnson et al. (1996b) showing 8-hour daily
maximum dose exposures for outdoor children under various air quality scenarios. Two graphs
are provided for each of four study areas (Houston, New York, Philadelphia, and Washington,
D.C.).  The graphs use two indicators  to characterize O3 exposure:

              Number of children experiencing 8-hour daily maximum dose exposures on one
              or more days under moderate exertion conditions,

              Number of total occurrences in which a child experiences a daily maximum dose
              exposure under moderate exertion conditions.

Moderate exertion conditions are defined as an EVR level between 13 and 27 1 min"' m"2.

       Figure  B. 1 presents results for the first indicator (number of outdoor children) based on
applications of pNEM/O3 to Houston.   Nine distributions are plotted on the  graph:  one for
baseline ("As Is") conditions; two for 1-hour, 1-exceedance standards (1H1EX-0.12 and 1H1EX-
0.10);  four for 8-hour,  1 exceedance  standards (8H1EX-0.10, 8H1EX-0.09,  8H1EX-0.08,
8H1EX-0.07); and two for 8-hour, 5-exceedance standards (8H5EX-0.08, 8H5EX-0.09). For
example, 8H5EX-0.08 indicates an 8-hour, 5-exceedances standard with O3 concentration set at
8 pphm or  0.08 ppm.

       The ordinate (y coordinate) of each point on the graph shows the number of children with
one or more daily maximum dose exposures equal to or above the O3 concentration indicated by
the point's  abscissa (x coordinate).  In Figure B.I, the  "As  Is" curve is  associated with the
highest number of children exposed when the specified O3 concentrations fall between 0.05 and
0.16 ppm.   The nine  curves tend to converge  at lower and  higher 03  concentrations.  In  a
similar manner, the 8H1EX-0.07 standard  is associated with  the lowest number of children
exposed when  the specified O3 concentration falls between 0.03 and 0.08 ppm.

       Appendix E of Johnson et al. (1996b) provides similar graphs for two other EVR ranges
of interest:  16-30 1 min"1 m'2 and  > 30 1 min"1 m"2 for 1-hour  exposures.

-------
                                             B-2
          FIGURE B.1. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
         OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
                  EXERTION (EVR 13-27 LITERS/MIN-M2) IN HOUSTON, TX.
           250
                                                                  ASIS

                                                               1H1EX-0.12

                                                               8H1EX-0.09

                                                               8H1EX-0.08
                                                               8H1EX-0.10
                                                                   (**.

                                                               8H5EX-0.08
                                                               8H1EX-0.07
                                                               1H1EX-0.10

                                                               8H5EX-0.09
            0 "
                  0.02  0.04  0.06  0.08   0.1   0.12  0.14 0.16  0.18  0.2
                              CONCENTRATION, PPM
           FIGURE B.2. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
       TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
                  EXERTION (EVR 13-27 LITERS/MIN-M2) IN HOUSTON, TX.
           25,000
                                                                  ASIS

                                                               1H1EX-0.12

                                                               8H1EX-0.09

                                                               8H1EX-0.08
                                                               8H1EX-0.10

                                                               8H5EX-0.08

                                                               8H1EX-0.07
                                                                1H1EX-0.10

                                                                8H5EX-0.09
                                                                   \ /
                   0.02  0.04 0.06  0.08  0.1  0.12 0.14  0.16  0.18   0.2
                               CONCENTRATION, PPM
appdxb.pre

-------
                                               B-3

           FIGURE B.3. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
          OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
                   EXERTION (EVR 13-27 LITERS/MIN-M2) IN NEW YORK, NY.
            1,000
             800
   ASIS

1H1EX-0.12

8H1EX-0.09

8H1EX-0.08
                                                                8H5EX-0.08
                                                                    *
                                                                8H1EX-0.07
                                                                 1H1EX-0.10

                                                                 8H5EX-0.09
                   0.02  0.04  0.06  0.08  0.1   0.12  0.14  0.16  0.18  0.2
                               CONCENTRATION, PPM
           FIGURE B.4. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
       TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
                  EXERTION (EVR 13-27 LITERS/MIN-M2) IN NEW YORK, NY.
           50,000
         <0 10,000
         cc
         111
         Q.
                                                                   ASIS

                                                                1H1EX-0.12

                                                                8H1EX-0.09

                                                                8H1EX-0.08

                                                                8H1DWJ.10
                                                                    -e-
                                                                8H5EX-0.08
                                                                8H1EX-0.07
1H1EX-0.10

8H5EX-0.09
                0   0.02  0.04  0.06 0.08  0.1   0.12  0.14  0.16  0.18  0.2
                               CONCENTRATION, PPM
appdxb.pre

-------
                                              B-4
          FIGURE B.5. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
         OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
                EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA.
           300
                                                                  ASIS

                                                                1H1EX-0.12

                                                                8H1EX-0.09

                                                                8H1EX-0.08
                                                                8H1EX-0.10
                                                                8H5EX^0.08
                                                                8H1EX-0.07

                                                                1H1EX-0.10

                                                                8H5EX-0.09
              0    0.02  0.04  0.06  0.08  0.1  0.12  0.14  0.16  0.18   0.2
                              CONCENTRATION, PPM
           FIGURE B.6. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
       TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
                 EXERTION (EVR 13-27 LITERS/MIN-M2) IN PHILADELPHIA, PA.
           16,000
                                                                  ASIS

                                                                1H1EX-0.12

                                                                8H1EX-0.09

                                                                8H1EX-0.08
                                                                8H1EX-0.10
                                                                    s\

                                                                8H5EXX-0.08
                                                                    )/
                                                                8H1EX-0.07
                                                                1H1EX-0.10

                                                                8H5EX-0.09
                    0.02  0.04  0.06  0.08  0.1  0.12  0.14 0.16  0.18  0.2
                               CONCENTRATION, PPM
appdxb.pre

-------
                                              B-5

          FIGURE B.7. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS FOR
          OUTDOOR CHILDREN EXPOSED ON ONE OR MORE DAYS UNDER MODERATE
                EXERTION (EVR 13-27 LITERS/MIN-M2) IN WASHINGTON, D.C.
           250
                                                                  ASIS

                                                                 1H1EX-0.12

                                                                8H1EX-0.09

                                                                8H1EX-0.08

                                                                8H1EX-0.10
                                                                   -0-
                                                                8H5EX-0.08
                                                                   -*-
                                                                8H1EX-0.07
                                                                1H1EX-0.10

                                                                8H5EX-0.09
                  0.02  0.04  0.06  0.08  0.1  0.12  0.14  0.16  0.18   0.2
                              CONCENTRATION, PPM
           FIGURE B.8. EIGHT-HOUR MAXIMUM DOSE EXPOSURE DISTRIBUTIONS OF
       TOTAL OCCURRENCES FOR OUTDOOR CHILDREN EXPOSURE UNDER MODERATE
                 EXERTION (EVR 13-27 LITERS/MIN-M2) IN WASHINGTON, D.C.
           12,000
                                                                  ASIS

                                                                1H1EX-0.12

                                                                8H1EX-0.09

                                                                8H1EX-0.08
                                                                   -e-
                                                                8H5EX-0.08
                                                                8H1EX-0.07
                                                                1H1EX-0.10

                                                                8H5EX-0.09
                    0.02  0.04  0.06  0.08  0.1  0.12  0.14  0.16  0.18   0.2
                               CONCENTRATION, PPM
appdxb pre

-------

-------
                  APPENDIX  C

Probabilistic Exposure-Response Relationships
      for FEV, Decrements >15% and >20%

-------

-------
                                     APPENDIX C
 C.I PROBABILISTIC EXPOSURE/RESPONSE RELATIONSHIPS

         Figures C.I and C.2 (which are Figures B.28 and B.29, respectively, from Whitfield et
 al., 1996) show the probabilistic exposure-response relationships for FEVj decrement > 15% and
 >20%, derived from Folinsbee et al. (1988), Horstman et al. (1990), and McDonnell et al. (1991)
-for 8 hr exposures and individuals engaged in moderate exertion.  The small squares indicate the
 response rates at the ozone concentrations  at which subjects were exposed in these studies (0.08,
 0.10, and 0.12 ppm), and the short horizontal  line segments above and below the data point,
 which are connected by a line segment, indicate the 90% Credible Interval (CI).  For example,
 for FEVj decrement >15%, the response rate at 0.08 ppm was about 18%, and the 90% CI about
 this value is about 10-28%; the derived 90% CI compares favorably to this range. The "derived"
 90% CI  and the "experimental" CIs compare  less favorably at 0.12 ppm because the linear
 regression used to fit the data does not capture the nonlinear characteristics of the data.  The
 derived  and experimental  CIs  for other endpoints compare more favorably  at all  ozone
 concentrations for the FEVj >20% endpoint.  Figures for 31 other endpoints are given in
 Appendix B of Whitfield et al.  (1996). Characteristics (i.e., parameters of functions fit to the
 data, and regression r  values) of functions representing exposure-response relationships for 33
 endpoints are listed  in Table C.I and C.2 (Tables 3.7 and 3.8, respectively, from Whitfield et al.,
 1996).

-------
                                     C-2
 cs

       100
        80
        60
        40
        20
                                                             Legend:
                                                             I	
                                                                    H 90% O
                                                                     a  data
                                                                      .95 free
                                                                    • .75frac
                                                                      .50 frac
                                                                    - .25frac
                                                                    - .05frac
         0.00
                                                           0.20
                   0.04       0.08       0.12       0.16
                      O3 Concentration (ppm)
FIGURE C.I  Probabilistic Exposure-Response Relationship for FEVj Decrement
^ 15% Derived from Folinsbee et al. (1988), Horstman et al. (1990), and McDonnell
et al. (1991) [for 8 hr exposures, heavy exertion; includes data, medians, and 90%
CIs for data and relationship]
 03
C*
 0
 I
       100
        80
        60
        40
        20
         0
         0.00
                                                             Legend:
          .90% O
           data
           .95 frac
           .75frac
           .50 frac
           .25 frac
           .05 frac
                   0.04       0.08       0.12       0.16
                     O3  Concentration (ppm)
0.20
FIGURE C.2  Probabilistic Exposure-Response Relationship for FEVj Decrement
^ 20% Derived from Folinsbee et al. (1988), Horstman et al. (1990), and McDonnell
et al. (1991) [for 8 hr exposures, heavy exertion; includes data, medians, and 90%
CIs for data and relationship]

-------
TABLE C.I Summary of Functions Fit to Experimental Data — FEVj Decrement — Corrected for Exercise in Clean Air
Study
Avol etal., 1984
(1.33 hrs)
Kulleetal., 1985
(2 hrs)
McDonnell et al.,
1983 (2.5 hrs)
Seal et al., 1993
(2.33 hrs)
Folinsbee et al.,
1988; Horstman
et al., 1990;
McDonnell et al.,
1991 (8 hrs)
Endpointa
DFEV^10%
DFEV,>15%
DFEV,>20%
DFEV,>10%
DFEV,>15%
DFEV,^20%
DFEV,>10%
DFEV,>15%
DFEV,>20%
DFEV,>10%
DFEV,;>15%
DFEVj>20%
DFEV,>10%
DFEV,>15%
DFEV,>20%
Function
Linear
Linear
Linear
Linear
Linear
Linear
Logistic
Logistic
Logistic
Probit
Probit
Probit
Linear
Linear
Linear
a
-0.2395
-0.2400
-0.2395
-0.3225
-0.2600
-0.2375

-1.0276
-0.6639
-0.3259
-0.0980
-0.2087
-0.1462
b
3.4388
2.9713
2.6825
2.3500
1.6000
1.2500
0.6420
0.4968
0.3347
0.7917
0.8401
0.9192
5.0000
4.9000
2.9250
d


5.5996
9.4948
12.0073


e


-27.2927
-45.3838
-60.4547


r2
0.98
0.99
0.99
0.95
0.93
0.89
0.99
1.00
1.00
0.99
0.99
0.97
1.00
1.00
0.98
aDFEVj means forced expiratory volume (in 1 sec.) decrement.
o

-------
 TABLE C.2 Summary of Functions Fit to Experimental Data — Symptoms — Corrected for Exercise in Clean Air
Study
Avol etal., 1984
(1.33 hrs)
Kulle et al., 1985
(2 hrs)


McDonnell et al.,
1983 (2.5 hrs)


Seal et al., 1993
(2.33 hrs)


Folinsbee et al.,
1988; Horstman
et al 1990"
McDonnell et al.,
1991 (8 hrs)
Endpoint3
Any lower respiratory
M/S lower resp
Any cough
Any PDI
M/S cough
M/S PDI
Any cough
Any PDI
M/S cough
M/S PDI
Any cough
Any PDI
M/S cough
M/S PDI
Any cough
Any PDI

M/S cough
M/S PDI
Function
Linear
Linear
Linear
Linear
Linear
Linear
Probit
Probit
Linear
Linear
Lognormal
Lognormal
Linear
Probit
Linear
Linear

Linear
Linear
a
-0.2084
-0.0902
-0.2650
-0.4550
-0.1626
-0.5250
-2.0954
-1.6071
0.0062
-0.0427
0.2469
0.2464
-0.1445
-0.3209
-0.2928
0.7372

-0.1747
-0.3087
b
2.6824
0.5206
3.0000
3.8000
0.8675
3.0000
1.2098
1.5124
1.2604
1.1512
1.9248
2.3641
1.3704
0.9317
5.0750
10.1750

2.3000
3.7000
d


















e


















r2
0.99
0.94
0.97
0.79
-0.33b
0.72
0.99
0.96
0.70
0.96
0.97
0.99
0.97
0.96
0.54
1.00

0.88
0.93
                                                                                                                                         o
Initializations: M/S means moderate or severe, PDI means pain on deep inspiration.
bThe data do not support a meaningful exposure-response relationship for this health endpoint.  The negative r value flags this situation.

-------
                                            C-5

 C.2  HEADCOUNT RISK RESULTS

         Risk results for each endpoint are available in the form of 10 probability distributions
 for each air quality scenario.  Since there are nine scenarios, it is not practical to plot all of the
 distributions on one  figure because the figure would be quite messy.  The nine air quality
 scenarios include:  one for baseline ("As Is") conditions; two for 1-hour, 1-exceedance standards
 (1112 and 1110); four for 8-hour, 1 exceedance standards (8110, 8109, 8108, 8107); and two for
 8-hour, 5-exceedance standards (8508, 8509).   The first digit in the code for each standard
 indicates  the averaging tune, the second digit specifies the number of exceedances, and the last
 two digits specify  the standard expressed in pphm.  For example, 8508 indicates an 8-hour, 5-
 exceedances standard with Oj concentration set at 8 pphm or 0.08 ppm.

         To gain insight about the  risk implications of the ah- quality scenarios, we developed
 "representative distributions" and "Box plots."  These are shown in Figures C.3-20 and C.21-23,
 respectively. Figures C.24-26 estimate the number of times that responders (i.e., outdoor children
 who experience a specific condition, such as having FEVj decrements >15%) respond.


 C.2.1  Representative Risk Distributions

        Figure C.3, which contains representative distributions, shows two sets of nine plots for
 FEVj  decrement >15%, 8 hr exposures, moderate exertion, Chicago, and children.  The top half
 of the figure  shows representative  distributions over the  number of children experiencing the
 effect one  or  more  times  (i.e., persons  basis), and the bottom half  shows representative
 distributions over the  number of times any child experiences the effect (i.e., person-occurrences
 basis).  Each representative distribution gives some idea of the range of results among each set
 of 10 distributions.  There is one representative distribution for each of nine air quality scenarios.
 Each plot is "representative" of the 10 distributions for a particular scenario.  Since there are only
 9 plots instead of 90, it is easier to see  patterns.  Each  of these plots is a valid cumulative
 probability  distribution.   The plot indicates, for example, that the median number of children
 in Chicago who may experience FEVj decrements >15%  under as-is air quality is around
 65 thousand.  When the most stringent standard  (8107) is just attained, the median estimate is
 about  15 thousand  children.
C.2.2  Box Plots for Risk Results

         Box plots provide another perspective about risk results.  Each Box plot displays the
ranges of the medians (or 0.5 fractile), 0.05 fractiles, and 0.95 fractiles of 10 risk distributions
that result from the 10  pNEM exposure  distributions that are available.   These ranges are
represented by rectangles in the figures (unless there is no range, hi which case the rectangle
   lrThe representative distribution is obtained by computing the average cumulative probability at selected points
along the X-axis.  This  calculation, like the risk calculations described earlier, implicitly assumes  that  the
distributions are  perfectly correlated.  It may be argued that perfect correlation, while not correct,  is more
reasonable that perfect independence, and there is no basis for choosing any other degree of correlation  between
these two extremes.

-------
                                           C-6

"collapses" into a horizontal line). There are 3 rectangles above the code letter for each standard.
The top rectangle represents the range of the 0.95 fractiles, the middle rectangle represents the
range of the medians, and the bottom rectangle represents the range of the 0.05 fractiles. A line
connects the bottom of the 0.95-fractile rectangle and the top of the 0.05-fractile rectangle and
passes through the 0.5-fractile rectangle.  With this format, results for 81 scenarios (9 scenarios
for each of 9 urban areas) can be displayed in one figure. For these plots, however, we switched
from numbers of persons or person-occurrences to percentage of persons responding.  As shown
in Figure C.3, under as-is conditions in Chicago,  the median risk estimates for the percentage of
children having FEVj decrements >15% vary from 13-14%, the 0.95 fractiles vary from 21-22%,
and the 0.05 fractiles vary from 7-8%. Box plots have the following  characteristics.  If the risk
distributions for a particular air quality scenario are quite "similar," the rectangles will be small.
If the  variance of a risk distribution is  small,  the rectangles will be close together.  If the
distributions are spaced far enough apart (indicative of widely varying  risk estimates for different
pNEM runs), rectangles will overlap.

-------
%%%*


-------
            Cumulative Probability
                                                            Cumulative Probability
o
o
s
o
8!
s,

1
I—*
o
o
i


1
IV
I—1
Lft
^

f
     p
     b
                                   p

                                   bo
   l/i
   o
&
     oo H- oo

      £2  ^ P Q
                  oo oo oo oo

                  o o o o
g
I—1

I
ffi
p
                                               00
     p

     b
p
to
p

cr\
p

bo
                                                ss
o
CD
o  t«
l-t  ^



I


£  8

IV
   K>
   tn
                                                                                              o

                                                                                              oo
      > 00 t— OO K-
      r^ i—» H-* >—* t—*
                                                                 oo oo oo oo

-------
Cumulative Probability
Cumulative Probability
                                                                            o
                                       ^j i_» i—•  o i—' o o o o
                                       ry3oN>35ov5oooo~j

-------
o 3

g o

B g
o
5? O
iv bs
  2
               Cumulative Probability
Cumulative Probability
OK)
  a
So

il
O &*
s §•
s 2'
II

  f
§
  t/3
a
il-
  I
                                                                                   O

-------
  3
  B
           Cumulative Probability
                                                           Cumulative Probability
  n
IV
I" 8
si
N
B- »•
— 3

if
".I
W "«
If
re S'
3 I
«* f>
» §
w »
    p
    b
                                    p
                                    bo
     1
     s,
1
I1
   »


f8
o
B
  a
  »

  o'
&
      >OO>—OOi--OOOOOOOO
      r/« ^*^ f^* *••* ^™* l^t l«/l l™* t—•
      j2»—ii—>O»—»OOOO
      O3ON>VDO*OOOOO^J
                                                 O
                                            I
                                            ffi
                                            So

                                           I
p
b
                                                                                p
                                                                                bo
                                                1
                                                1
                                                IV
                                                   g
                                                   u>
                                                   o
                                                          »— OO
                                                                   C»OOOO

-------
                  Cumulative Probability
                                                                   Cumulative Probability
   -—

3  2
2  H
D

ff  p

IV  oo
is
?8
§ g


ft 3.
er o"
••• (•
& a


i §
B w
•a a

S S
e a
S-o
     o
     o
     p

     b
O       p      O      p

|O       -£t      ON      00
3 g   2
ff K  55
I
2,

a
;ii ^
~^^u c5
i—*
D




i,
IV *

               *
j> oo ^- oo H-
                        OO 00 OO OO
       J2i-«i-'O»-*OOOO
       -  ^»  «. •> \oovooooo~j
o
ET   P
E:   b
                                                       p

                                                       to
p

bo
                                                       i__l K)


                                                       ffi "
                                                       OQ
                                                       a
                                                          -

                                                       I «
                                                       IV
                                                        ">OOi-'OOl—
                                                           (_.i_.t_ii_.
                                                                    OO
                                                                    t_i

                                                                    o
                                                                         OOOOOOOO
                                                                         ol^l_.l_.

                                                                         O O O O
                                                                                                       o

-------
PJ *5^
n g
q o
^ '-_j
9 
                              O

                              bo
»  O

00
•o g.

t/1 <;'
c n
•n
    5?

    IV
s^ ^
&ff §
2 K S
1^ g
W o5 ~

i»
o 3.
eg
           f
           i
           i

           i
           00
                                                  Cumulative Probability
g
»—>
i
ffi
p
b
p
bo
                                           to
                                           o
                                         OQ

                                         a

                                         5?

                                         IV
                                           s
                                           00
                                           o
                                         \fj


                                         S 8
                                                                             n

                                                                             UJ
              OO H- 00 OO OO OO

-------
n
3
S
fD
a
S
IV
  n
?v
*l
fg
11
Ft
O"
I?
O 9?
° 5
P|
1^
l|
00 »
Wo*
T 1
W p».
M £•
•o f?
S 3
  I.
  ^
Ife
Q. -I

It
WJF
M ^
S «>
a o
o g
a g
s«x P
  i'
a
     o
     o
     I
     8
          p
          b
                Cumulative Probability
P      p      p      p
N)      -^      CTN      OO
                                                              Cumulative Probability
     o
     CD
     O ~
     Ht Wl
     CD °
        to
        o
        Ui
        8
                                                                                             O

-------
  a
  8
           Cumulative Probability
5
2
B
     O
     O
  n
|V b
E    fcs "
8|  I


11  8' s

f I  S,

     W K>
     < ^
    p
    b
                                     p
                                     bo
11  o
S s1  8  «
"1 a  «  K>
ii  §  -
"
I fe
b^

&  &
IV  W
*  00 t- 00 i—
                      00 OO OO OO
                                                             ^OOv— OOOOOOOO

-------
\

-------
Cumulative Probability
Cumulative Probability
o
B*  °
E:  o
                                 1
                                 IV
                                            p
                                            to
                     O

                     bo
                                 §
                                 OQ


                                 1
                                 I—*

                                 O
                                                                            n
                                              1


                                              }
                                           i_> QO
                                                h-OOOOOOOO

-------
      Cumulative Probability
> 00 H- 00 h-
^ t—* t—• t—* I—»
           OO OO OO OO
Cumulative Probability
                                          O
                                          B:
p
N>
p
bo
                                           *  o
                                         OQ

                                         1
                                             K>
                                             O

                                             U)
                                             o
                                          IV
                                          to

                                             -• OO H-
      OO OO OO OO

      Q O O O
                                                                                        n
                                                                                        i—»
                                                                                        00

-------
Cumulative Probability
                                                Cumulative Probability
9  -
P«  fc>
                                                               p
                                                               b\
                                   p
                                   bo
 §
 ffi
 Co
 p-
Crq g

 1
 (—*
 O
 2 -


   '
 I
 IV
                                    s

                                                                                  n
                                             » i  i
       >>oo>->ooi-*
       ffl I— '"— 'I— '•— '
                                                      oooooooo
                                                      t/lV»l— «!-»
                                                      O O O O

-------
Cumulative Probability
Cumulative Probability
                                                                             o

                                                                             K>
                                                                             O
                                       J«H->t-4Ot-'OOOO
                                       rxJONJVOOVOOOOO^J

-------
Cumulative Probability
      Cumulative Probability
                              i
                              r
                              ffi
p
'o
p
to
p
bs
p
bo
                                N>
                                

                                   O
                                   to

-------
Cumulative Probability
Cumulative Probability
                                                                      n
                                                                      to

-------
Cumulative Probability
Cumulative Probability
                                                                               o
                                                                               to
                                        >>.OOH~OOl->
                                        ri.»-*l— 'I— 'I— >
                                                   OOOOOOOO

-------
Cumulative Probability
Cumulative Probability
                                                                      O

-------
FIGURE C.21 Headcount Risk Results for the Percentage of Children Responding for the Eight-Hr, Moderate Exertion, FEV, Decrements
>15% Endpoint
\j " 	 - 	 • 	 -•
ZEKBHJPOG ZDAEHJKG ZBVEHJFOG ZEttEHJFCG ZErtEHJFOG ZDABHIPOG ZEABHJFCG 2DAEHJFQ3 ZDABHJEQ5
Chicago Denver Houston Los Ang Miami New York Phila St Louis WashDC
% Responding, Persons
H-* i— ' ts> ts> OO U> 4
DUiO<-ftO<-ftO<-ft C
a = 	 — 	 =


— =— — =-=•





° ™







= 	 "^^ 	 ^T3 ,
^S. 	 =

= 	 ~" 	 =
a a en "*** i I ~J

, , 	 , ,
a , ^ ۥ J

_, = 	 — 	 ; 	 ; 	 =
0 	 ^ 	 C=3

O •• t ITS

a =T" •_' ' i — i 1




=_S 	 	 	 ==, ' '

LD mm ^^
1 1 •• C^


O Ctt vO «S H H '' OQ II II II II J| O
O O Q
£ £: ^
§* s

-------
% Responding, Persons
IE C.22 Headcount Risk Results for the Percentage of Childi
Endpoint
ren Responding for the Eight-Hr, Moderate Exertion, FEVj Decreme
\j • 	 	
ZEKBHJFOG ZEKHHJFOG ZI^UBHJFCG 2n^HJP03 ZEttEHJFOG ZTVSEHJFCX} ZTVEHIFOG ZQ^BHJPOQ ZEKEHJEQ3
Chicago Denver Houston Los Ang Miami New York Phila St Louis WashDC
I-1 I— Ni h
D <-* O <-n O <-






D ' ' C* ^~ • 	 • ' '


	
° ™ o "

M 	 a
1^ °


"^ 	 =
a HI i j — -i


o 	 Ka 	 *~ " •
a 	 a 	 1=3
	 	 a

• 	 ^ 	 ™ 	 ^=3
a 	 
	 	 o

OM /~N t?^ >^^>l__ie_.>_lbdl^ l_^k.i ^0
p . ?» O r^Tiypjrn^rpisj^
o n vo ^2 u n n JL n n n ii n CD
B S- ^ ° i*
S S

-------
FI
Se
JURE C.23 Headcount Risk Results for the Percentage of Children Responding for the One-Hr, Moderate Exertion, Moderate-1
ere Pain on Deep Inspiration Endpoint
yj •
ZrttEHJFCG ZDAEHJKG aiHEHJPCG ZDABHJPOG 2DAEHJPOG ZEWEHJPO3 ZEAEHJFOT Za^HJEOS ZnABHJBQ3
Chicago Denver Houston Los Ang Miami New York Phila St Louis WashDC
% Responding, Persons
t-* H
D <-n o <-


° T* tirV^
—


° ° f -^ '' *


= "Li - =.

°


II iM


1=1 	 	 ' 	 *•** 	 '

*"" L






""






" ^* ( 	 " "

a ••• ^^^
, 	 ea 	 =

CD 1 Illl 1 II
~ *
i a ciJ

a 1 	


°


1 "" ( 	




*


— " —
C~3
• ' 	 '


— r=l





I • C3

a •§ rirjj




& li S
s s .

-------
                                         C-28

C.23  Ratios of Mean Numbers of Occurrences and Mean Numbers of Responders

        Figures C.24-26 show the ratios for the mean numbers of occurrences and mean numbers
of outdoor children responding for three endpoints, nine urban areas, and five air quality scenarios.
The endpoints are FEVj decrements £15% and ^20% for 8 hr exposures at moderate exertion, and
moderate-to-severe pain on deep inspiration for 1 hr exposures at moderate exertion. The following
letter codes were used to identify the urban areas: CH = Chicago, DE = Denver, HO = Houston,
LA = Los Angeles, MI = Miami, NY = New York City, PH = Philadelphia, SL = St Louis, and DC
= Washington, D.C.

        The ratios were computed in the following way.  For a specific endpoint, available risk
results include 10 probability distributions (one for each of 10 pNEM runs) over the  number of
persons who respond one  or more times, and  10 probability distributions over  the number of
person-occurrences (which  allows for the possibility that an individual may respond more than one
time).  The ratio of interest here is the sum of the expected values  of the person-occurrences
distributions divided by the sum of the expected values of the persons distributions. The ratio is,
in a sense, an estimate of the average number of times that a responder responds during an ozone
season.

-------
                          C-29
  FIGURE C.24 RATIOS OF MEAN NUMBER OF OCCURRENCES AND
MEAN NUMBER OF RESPONDERS (FEVj DECREMENTS Ssl5%, 9 URBAN
    AREAS, OUTDOOR CHILDREN, 8 HR EXPOSURES, MODERATE
                       EXERTION)
        CH   DE
LA  MI   NY
Urban Area
                                    PH
SL
DC
  FIGURE C.25 RATIOS OF MEAN NUMBER OF OCCURRENCES AND
MEAN NUMBER OF RESPONDERS (FEVj DECREMENTS ^20%, 9 URBAN
   AREAS, OUTDOOR CHILDREN, 8 HR EXPOSURES, MODERATE
                       EXERTION)
        CH   DE   HO   LA   MI   NY
                      Urban Area
             PH
SL
DC

-------
                         C-30
 FIGURE C.26 RATIOS OF MEAN NUMBER OF OCCURRENCES AND
MEAN NUMBER OF RESPONDERS (MODERATE-TO-SEVERE PAIN ON
 DEEP INSPIRATION, 9 URBAN AREAS, OUTDOOR CHILDREN, 1 HR
            EXPOSURES, MODERATE EXERTION)
      20
      18
      16
      14
      12
      10
       CH   DE   HO   LA   MI   NY
                     Urban Area
PH
SL
DC

-------
                                         C-31

C.3 BENCHMARK RISK

        Figures C.27-29 illustrate benchmark risk results for the probability that the benchmark
response will be exceeded 5 or more times in an ozone season.  Benchmark response is r, the
fraction of the  population who experience the specified health effect upon exposure to ozone.
Benchmark risk is defined as the probability that the benchmark response is >r n or more times
in a given period of time (1 ozone season) at some location within a geographic region, given
a specific condition of air quality (e.g., that standard 1112 is just attained).  In this report, we use
r values of 0.05 and 0.1 (sometimes referred to as 0.05 and 0.1  benchmarks, or 5%  and 10%
benchmarks).

        Figure C.27 shows the graphical format used to display benchmark risk results.  The
figure  is for the probability that the benchmark  response for the FEVj decrement £15%, 8 hr
exposure, moderate exertion endpoint will be exceeded 5 or more times in an ozone season for
0.05 and 0.1 benchmarks.

        This figure includes results for 9 urban areas and 9 air quality scenarios for each urban
area.  The air quality scenarios are indicated by a letter code above the name of each area.  The
letter code is explained on the right side of the  figure.  There are 2 vertical lines for each air
quality scenario: the one on the left (a solid line) is for the 0.05 benchmark, and the one on the
right (a dotted line) is for the 0.1 benchmark. The height of the line  indicates the benchmark
risk. The benchmark risk for the 0.05 benchmark is, logically, < the benchmark risk for the 0.1
benchmark.

        For example, for Miami, scenario G,  daily maximum 8-hr-running-average ozone
concentrations, and using the distribution for the highest ozone concentration, the benchmark risk
for the 0.05 benchmark is about 0.52; and the benchmark risk for the 0.1 benchmark is about 0.2.
In other  words, if standard 8107  is just attained in Miami, then the  benchmark risk  (i.e.,
probability) is 0.52 that >5% of the population will experience FEVj decrements >15% 5 or more
times in an ozone season.

-------
£3
                 Benchmark Risk (probability)
a.
°3
3 cr
8 gf
N 3
O o

5 °"
« g
        o
        •

        O
                 O

                 io
p

bo
  s
  •o
  o
  o
  •1
  W

  ota"
  a
  o
  a.

  §
  W
  x
  »
  g.
  s*
  IV
  a
  a.
  f+

  3
l!U!ll!il!li!iiT77!T
Mi.iiiiiiiiiiii.iiii.imiini


mii'iii'iii'mii'iiii'mii'ii'iii'ii'i"'""""!'

HHiliiiiiiiliiililiiiiliiliiiii-


"iii" 	 "' 	 "i 	




'triTTttiTTtrtrtrtr


iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii-
itBttiiiMummititi'ititnitiitiitiitiim"]1"""""18"'




N
it
5?

00
i
s
o*
" J=
h-
c


5 C
5 5
w
S
g.
i 1

-------
£2
w 2
o ^^
I-
Benchmark Risk (probability)

5£
               0
               to
                          0
                          bo
»
N 3
2 «
5 ^
« g
(/} »

i*
§ s

  S"

  «
  2
  ff
  w
  %

  S"
  p
  S?
  3
  i
  r+
  IV

ii!!!i!!l!!!!









iiiiiiiiiiii!

ilimiiillllllll 	 	 = 	








^•••' 	







iiiiiiiiiiiiiiiiiiiiii!!!1 	 =

                                     »

                                         o  o
                                         s  s

-------


^
.1? 0.8
r^
•8
JD
O
& 0.6
GO
s
HI °-4
1
o
d
S
QQ 0.2
r^^^
no



































in

















i

















i 1

















j

















i

















i



































nil

















,

















i



































1

















11

















Ilil,

















i

















Id

















i



































III

















i



































ih

















i



































in

















1


















Benchmarks:


	 	 o.io
Scenarios:

Z=AsIs
J=8509
B=8109
F=8508
C=8108
A=1112
H=1110

D=8110

G=8107

                                                                                                                      n
                 ZJBPQ\HDQ ZJBPCAHX} ZJBPCAHX} ZJHPCAttC ZJHPCAIEG ZJBPCAH3G ZJBPCAHDG ZJBPCAHDO ZJBPCAtPG
                 Chicago Denver  Houston Los Ang  Miami New York Philadel St Louis Wash DC

FIGURE C.29  Probability that the Benchmark Response for the One-Hr, Moderate Exertion, Moderate-to-Severe  Pain on Deep
Inspiration Endpoint Will Be Exceeded 5 or More Times in an Ozone Season

-------
                APPENDIX D

List of Species in Selected National Parks
      and Associated Ozone Response

-------
                     D-l
               Table  D-l
tixde* exhibiting tc>li«  injury ind>c»tm of ctpoturt ta coat to CRSl
Sfxciw
Ater mbrum
Acer taccharunt
Aconitum uneinaittm
Atfculus oeiandn
Ajntlonchter Lstvis
Antmone qui/ujmfolio
Antmone virginiana
Aptxynum cann'dnnum
Anuolochia dtirior
Ascitpias txaiiata
Axlepias incarnate
Atflepiaa quadrifolia
Atctfpias ryriaca
Aatr acuniinatus
Alter cstrt&u
Aaer drvasicatiu
Aatr infinnuj
Afttr pnrticeui
BttuUt lutea
Bid era Jrondcxo
Cecelia atrrpScifolie
Cofttlia ruftlia
CefyctSllltiui florvla
Cephatonihus octnitntala
Cereij certadmsx
Clematis w%m>ana
Chtoria maneme
Comtu florule
Croiaeguj spp
DimtHe nailifvlin
Diospytot virgi/iinna
Eupatonum n/gatNin
FraxmiLs amrnana
Froxinuj ptnnsytvoniia
Geybttsacia baccara
Goyiussacta nrjinn
Gcum nduttum
GiEtnia irifotiaw
Gtetria nubigftta
Helena Carolina
Hemamtlii vifginiana
iltluinthuj fiaticoptrythu
Htlitnlhiu micnxtplinhu
Htxestybt ttrifoLa
Hitrocitvn panicnlatutn
Jmpaoetis copinsu
Imp&iitru pa&da
Lindire bttuoin
Lujuulambar tnrttciflua
Common Name
Red Maple
Sugar Maple
Wfld Moakriood
Yelkw Bucke)'C
Allegheny Strvicrt^rry
Wood AooiK»[nc
TUmblrweod
luUi iD-h«mp
Duichmin'* Pipe
Tad Milkweed
WborJ«d Milkweed
Swamp Milku^f d
Comrooo Milkweed
VV'horleJ-wood Aster
Grrtis'i Asler
Wbitc-wooO Alter
Entire-leaved A.M«I
Purpk-5l«»imcd A$ler
Ydk>* Birch
fkgjar-ttcki
filt Indion-plaaUiD
Rwsel's raewat
Swect-ihrub
BunoDbu^h
Ra*ten> Redbud
Virgia'*-bo*«:r
Oauerfly Pea
Flowtricj Dogwood
Hawihcni
Bash Hoc^'suclde
Conunou Pcr«>0iBion
White Snakerooi
White A*b
Otc«n Aih
Black Muckt«bcrry
Bear HuckWxrry
Mountain Avcns
BowmanVroot
Smoky Mouoiain Munna Crass
SirverbcU
Wittb-Hazrt
Wkii«4eaf Sunflower
Small Wood Sunftown
PlcarUeaf
Panicled lJa\vlc*eed
Spoiled Jtfwel«-e«d
Y«Uow J«weh»-ced
jjpic«bttsh
S*ect{;um
Ufdonn
Tit*
t>««
Herb
TIP*
Troe*
K«rb
Herb
Herb
Viae
Herb
Herb
Herb
Herb
Herb
Herb
Herb
Herb
Herb
Tf*e
Herb
Hcrta
Herb
Shnib
Shrub
Tiee
Vice
H«(h
Tree
Tw
Shrub
Tre*
Herb
Tree
Trtc
Shrvb
Shrub
Herb
Herb
Grass
Tr«*
Shrub
Herb
Herb
Herb
Herb
Herb
Herb
Shrub
Tree
Code
1.X5
I

1,24





1.2,3



13

6

12
U


Ii2


\2.


\2-

6

1,2


6
6
1

U








6
1A3

-------
                                           D-2
                            Table  D-l  (Continued)
Species
i.v\oatnAron luitfijtra
Lyonia lifiutrina
Mtgnelia frcstrt
MfruUsia pU
Rhantnus elm/alia
Kicdodfntlrvn CfOktn
K\0do£ehdron caltndnli>crnn<
Rhododendron eait\*'tHii(st
R}U4S topallina
Riua radiCAfis
Mania ptfudoacacic
Rutui canadtJisif
Kutm idntus
Rudtxtha hrta
Ruiibtcht loan talc
Sambucus ctvitdenfis
Swikuciu pubfns
Se&efras nlbidtim
SUphiwn asrrnsatf
J/ni/Ar glance
Smdax romndifoUa
S&bdefo roanerua
Stewrna cvnia
Stacf Crape
Frcsi Gripe
Lifefem Code
lft« IJJ,4l5.o
Shrub
••Tree
Herb
Tw«
Tree 6
Viu«
Tr««
Tree 1^3
Tree 1,2,3
Tree 5,6
Tr«e 1Z3.4.6
Tree 123.6
Herb
Tree
Tree 1,24.5,6
Fern
Tree 1,6
Tree 1,6
Shrub
Shrub
SI) rub
Shnib
•Tree 12,3
Vine
TT«* U.3
Shrub IJ,J
Shrub U,3
Herb U,3
Herb 1^3
Shrub
Sbrub
Tr««. 1^3,5
Herb
ViLC
• Vine
Herb
Tre<
Herb
Tree
Herb
Tree U
Shrwb
Herb U.3
Vine
Vine
'Spcctct ini>< fuvr tx«n fumlpifd Jl T>in Ctectt
^Speow thc*ifl{ lot'*r injur) *flti fuia^ciKXi

-------
               Table D-2.

Vascular Plants of Acadia National Park

  That Have Been Studied for O, Impacts
GENUS SPECKS
AM» talnmn
Ae« oefundo
Acar pttUnoid««
Aotr nibnim
Anr ncchuioum
Ac*r Mccbinim
Achille* mitlefolium
Afrortiialbt
AfrodH tanvlf
COMMON NAMI
0*1 um fir
Bnelder
Nocway maple
Radmiphi
Silwrmipla
Suftr miplo
Comma yurow
Bkcktwrt
RoifhfeM
POU.VTAKT
0,
0,
0,
0,
0,
0, + SO,
0, + Add
precip.
0,
o,
0,
o,
LEVEL
25.0 pphra
15.0; 25.0 30.0: 40.0 pptn
O.S ppa tlo» ind wiih 1 .0 ppin SO,
O.JSpptn
30.0ppbm±
25.0 to 35.0 pptun
0; O.I;0.2;0.)ppB
.05: 0.1; 0.2; ppffl O, «loo» ud with O.I
ppmSO,
.0); .06: -09; .12 ul/1 wjlh pH 3.6; 4.0; 3.0
(1.25 hn 2 UnM/wnk)
0; 0*5'; O.I 0.15 ppm
15.0; 25.0; 30.0; 40.0 pphra
5.0-7.0; 15.0; 30.0 pphra
15.0 pphm
OJppi,
DWAtlON
tboun
2houn
rinfU 7 IH hour
•XpMUf*
Ihoun
1 hn/dij foe I
ooaOu
Ikn/di7
5d.yi/«*f(x
iBoalhs
l2hn/d*yDpto«0
«0
-Uyi
7bn/d«7 5d.y«/
wk fot 6 to 10 wti
«bn/d«yfor2t
d.y.
Si«»U2-kr«po«»
3 hn/d«y 5 
•od tafinlgbl id Adlaed with tocnMtaj O,
Oraodi nh wl •ffacfed X O.OS «d 0.1 ppak
fata ndmd by 0.2 pps iod SO, Iralml
O, awed rifalHaal dwUnx b M*
pboCocyvQMitt* no nlMcttcw wilh Kid pp4

pm.SmmMdtfO.IS ppm.
IVvkoU for Iqhjr, VM om 30.0 ppha
VUM. tajvy: plw top> nd itxX w*|fai
docnwd M 15.0 ppfcB, ProAiMd Howm but
wimbcldlbmb).
Bh«dt«f •< woolf of bnM. 40» UT
ttwbjmd .
Modnk to M»M I4wjr
REFIKENCE
DnU. M •!., 1772
Trabow, « d , 1973
tUraoAy, 1911
Rotwli, M •!., 1985
Jean, 1973
Ivan, 1973
J«MB, 1912
I«d«a. I9S3
Rekb. «. •!., 1996
KIM, •) •!., 1912
Trabow, « •!., 1973
tUrwud, «•]., 1971
Hill, X •!.. 1974
YoMpr, « •! , 1910
                                                                  D
                                                                  I
                                                                  u

-------
.9
  I

      f

          3
           I

1!
                 a.;
                 3f
                 »*



                 B
IT










I
                                , QJ



                                II
                                ft M
                                g W
          Hi  H-

          2 ^c

            ft»
          Pj-
          •6
                                en

-------
i
f
      i*


      I
J
                                1
                                    If
                                   Hi
I!
 I
 9

                                               CU
                                                 Q)
                                                 en
                                               H-
                                                 o
                          73
                           rt
                           M
                 g-a

-------
a.
          ^i
»».

I1
Jl
                    PS
                    a
                  H-J
                  s*
If
p
h
E:
-C
                              ^
                                ii
                                 I
                                     i
                                                  o
                                                H 3
85

"g-
                  9-a

-------
          Table  D-2  (continued)

Vascular Plants of Acadia National Park

  That Have Been Studied for 03  Impacts
OiNuiSrtcn
Plum tMoblma
Piniuri(idi
Pimn qrlveatrii
Plantago lannolau
PtanUfo major
Poaannua
Poa pratcnsit
Poljfonlun aviculara
Popului f randideaUU


COMMON NAMI
Mr*"*
Pilch fiat
Scotch pin*
Eailuhpluuua
Oreal plantala
Annual blMf ma
KaQtucfcjr bluegraal
Yudknotwml
Bl|-uxxh aapca
Quaking atpta
POUVTAMT
0,
7 l 2 wk»
1 kn/diy > 5 d.yW
wt for 4.5 wt.
3 hn/dar « 5 '••X"'
wfc QlfWfb |fOWIA|
i dar^wk for 74
dtyt
< hn/oriu wdl
^Niof floe 2 vpftaff
Brner
JJRofplMttKBttTdM.

•ju-mt««tii»«- 1 nil 1 1 • 	 tnw
pofnuMKNu; nyvj WM tow

P,

(n» not: rfuol ratio
Lov lohnaoii vUM* du»|»
SlfiifioM ntedoi ta me« nUtiv. (rowih
nu nd root:riioot ndo
0.73±.4«XI«fmut7
7.3Ji4.04*l»ffcv«iT
lUdiiMl lop «d raw mi|hl «nd w«i
productk««30.0pphm
No diaHical (ffiKl 01 M|W powth
Natural «al«otloi fcr eaoM Whnaw* nay
hiv« cccomd !• «on» ifmftrim
(UFCUNCt
Onto. fid.. 1972
Aramtaoo, •! al., 19r7
Scbarar, •( al.. 1919
Km, «( al., 1912
Divif, dal., 1972
Skarby, 1917
Railinc, 1992
lUUiBf. 1992
Richard., el >!.. 1910
Roiling, 1992
Martin, el al.. 1911
Hirward. « al., 1971
Jena). * al., 1975
tmot. <* al.. 1916
                                                                   O
                                                                   I
                                                                   -J

-------
          Table  D-2  (continued)

Vascular Plants of Acadia National Park

  That Have Been Studied for 03  Impacts
GEmnSnem
Pniflut pcnuyivttKft
Pmnuf wrotuu
Querctti rubra

RhiMtyphiM


Rama crifpw
Rom.1 obftrifolhif
Sunburnt cnxfcwh
SilenoctniMui


Syrinfi votfuif
Thuji oeeUaiblif
Tmlotxlcnihai ndteuu
TrUolluni prataaM


COMMON NAMI
, Flrecberrjr
BlMkckorr
Northern red oA

SUf-bora mme
Comma Aim wmwl

Dock
Blttaraock
AmecktttUer
MiMai-i leui


Commltte
CMlCB ftlt)UTVIt>0
Polfaaivjr
R<«) clovtr


POU-UTAKT
0,
0,
O, + Add
Precip.
o,
o,
rt

o,
o,
o,
•
o>
o,


Livn.
0.10; 0.20; 0.10; 0.40
0.20 ppa for 5 boon
0.02; 0.07: 0.13 ul/1
pH 5.0; 4.0; 1.0 (1.25 hn 1 HUM «*)
0; 0.075; 0. 15 .1/1
0.25pp.


25.0 pptu.
70.0 nil)
0.25 ppn
0, 60.0 «t/m'
SO, 50; 100 Uf la? 6 wlu Um 4 wki


0.35; 0.55; 0.75; 0.95 ppn
NatMRpOftod
15.0; 25.0; JO.O-, 40.0 pphra
0.5 aflaf mi (niter


DURATION
I 4»« whh
wy H|bt KNiv t^jtwy
Nvy ntfai( by tovrf: 1.0; 1.5; 5 J; 7.5 M •
OtolOicakofkOorr

Ivjwr tkmftold onr JO.Opptu.
in)«Ti»i««i'OJn'">


KVRINCI
Pjr*. 1*11
Dtvb, el «!.. 1911
ReKk, M •!.. I9M
Dtvii, •) •)., 1992
Dnh, M •!.. 1976

IMUif, rt •!., 1992
Tn*ow, rt «!.. 1973
RBlfaf, 0 «!., 1992
Dtvlt, «t «!., I97«
Dock. M il., I9B6

Divli, rt «!.. 1976
OAey. I?t2
Hutov. •) «!., 1912
TiMfcow, at •!., 1973
UAjr-Knw, M •!.. 1919


                                                                  a
                                                                   i
                                                                  00

-------
          Table  D-2  (continued)
Vascular Plants of Acadia National Park
  That Have Been Studied for O3  Impacts
OfKVf SrtCIM
Tri folium repeal
Ttiif* catudcasii
Ulfflus Bnericuu
Urlica dioica
Urtic* tncltfs
VioUftdwica
COMMON NAME
Y/hito clover
Eastern hemlock
American cltn
SUnftaf nrtlle
Slmftai Mtilg
Hooktpar viol*
POUOTAKT
0,
0.
0,
0,
o,
0,
Uvo.
t20lol40v|/m>
O.OM nl/1
0:0.05; O.IO;0. 15 ppm
0.05 ul^l
0.07 nl/1
Not reported
0.9 ppm
70.0 nlfl
I5.0-. 25.0; 30.0; 40.0 pphm
IS.O; 25 0; JO.O; 40.0 pphm
DUKATTON
1 ttr mu for nvenl
oonoartlvt di)r<
11 hiVr-500d>r>

5 boon for 9 diyt
7 hn/ifay for 2 wb
2 boun
2houn
BaM^M*
BfVH,!
NtcrotknoA.
Orowih rednoloa; t»* yMil mhiclloa
lUAiwd pint mifh whfc bcnued coma.
I3« rxfadtoa in he((M powth
40* fwfuctraB to onfM f rowu
CoaMmd miibnt 
-------
                            Appendix E

  Predicted  Yield and Biomass Loss  for Crops and Tree Seedlings
          Based on Growth Regions and CIS  Predictions of
          1990 National Air Quality Using  the  W126  Index
*These maps have been generated using data which contain
unquantif iable  levels of uncertainties with respect to
extrapolating exposure-response functions generated in open-top
chambers to the field  (section D.2 and D.3)  and use of limited
meteorological  and O3 precursor  emissions data to develop
national projections using the CIS (section F.I).   The impact and
the interaction of these uncertainties on these national
projections is  not known (section F.2).

-------
 Estimated  W126  Ozone  Exposure  (MAX 3  months  1990)
Estimated W126
Q  o - 4.9
™ 5.0 - 9.9
  10.0 - 19.9
  19.0 - 29.9
   > 30.0
          /05/95

-------
                                          Estimated  Yield   Loss

                                                 Soybean  -  1990
                                                                                                                               w
                                                                                                                                I
                                                                                                                               NJ
     Shaded areas represent major soybean growing regions

     Weibi|l model:     PRBL =  1 - exp[  -(w126/B) ** C]

     Weibull parameter:   B = 110.2

     Weibull parameter:   C = 1.359

     Based on estimates ol 1990 maximum 3 month 12 hour W126 ozone exposure
                                                                              •
Note:  NCLAN exposure-response  functions used in projections of yield  loss.

This map  shows the geographic range for this species.   It does not  indicate
that this species will be found  at , every, point within  its range.

-------
                                             Estimated   Yield   Loss
                                                 Kidney  Bean  -   1990
     Shaded areas represent major dry bean growing regions
     Weibijll model:     PRBL = 1  - exp[  -(W126/B) ** C]
     Weibull parameter:   B = 43.1
     Welbull parameter:   C = 2.219
     Based on estimates of 1990 maximum 3 month 12  hour W126 ozone exposure
Note:   NOLAN exposure-response functions used in projections of yield loss.
This  map shows  the geographic  range for this species.   It does not  indicate
that  this species will be  found at every point within  its range. _
//dotoJ/ondvh/bio Ioss3/:olol/ moo.oral
                              06/05/95

-------
                                               Estimated   Yield   Loss
                                                      Wheat  -  1990
     Shaded areas  represent  major wheat growing regions
     Weibull model:     PRBL = 1 -  exp[ -(w126/B)  ** C]
     Weibull parameter:   B  = 53.4
     Weibull parameter:   C  = 2.367
     Based on estimates.of  1990  maxmum 3 month  12 hour W126 ozone  exposure                  •
Note:   NCLAN exposure-response functions  used in projections of yield lojas.
This map shows the  geographic  range for this species.   It does not  indicate
     t-hif? snorM on will ho -FniinH at- ovamr  n/->i n<- ».ii*-Vi
-------
Estimated   Yield  Loss
        Cotton  -  1990
                                 Percent  Loss
                                  Si o - 1.9
                                    2.0  - 3.9
                                    4.0  - 5.9
                                    6.0  • 9.9
     Shaded areas represent major cotton growing regions
     Weibull model:      PRBL = 1 -  exp[ -{w126/B) **  C]
     Weibull parameter:   B = 96.1
     Weibull parameter:   C = 1.482
     Based on estimates of 1990 maximum 3  month 12 hour W126 ozone exposure
Note:  NCLAN exposure-response functions  used  in projections of yield loss.
This map shows the geographic range  for this species.   It does not indicate
that this  species will be found  at every  point within  its range.
                                                                                    M
                                                                                    (Ji

-------
                                             Estimated  Yield  Loss
                                                    Peanut -  1990
     Shaded areas  represent major peanut growing regions
     Weibull n/odel:     PRBL = 1 -  exp[ -(W126/B) ** C]
     Weibull parameter:   B = 96.8
     Weibull parameter:   C = 1.890
     Based on estimates of 1990  maximum 3  month 12 hour W126 ozone exposure
Note:   NCLAN exposure-response  functions used in projections of  yield loss.
This map shows the  geographic range for this species.   It does not indicate
that this species will be found at  every -point within  its range.
  Percent Loss
^—Bi o -  1.9
      2.0 - 3.9
      4.0 - 5.9
      6.0 - 9.9
       > 10.0
                                                                                                                                 W
                                                                                                                                 I
                                                                                                                                 CT»

-------
                                               Estimated   Yield   Loss
                                                       Barley  -  1990
      Shadedfareas represent major  barley growing  regions
      Weibull!'model:     PRBL =  1 - expl -(w126/B)  ** C]
      Weibull parameter:   B = 6998.5
      Weibull parameter:   C = 1.388

      Based  on estimates of 1990 maximum 3 month 12 hour W126 ozone exposure

Note:   NOLAN exposure-response functions used in projections  of yield  loss.
This  map shows the  geographic  range ,for this species.   It does not indicate
that  this species will be found at every point within its range.
                                                                                                                                    W
                                                                                                                                    I
//dolo2/ondvh/bio Ioss3/!olol/ moo.ami
                              06/05/95

-------
                                             Estimated  Yield  Loss
                                                     Corn  -  1990

     Shaded areas represent major corn growing regions
     Weibull model:     PRBL = 1 - exp[ -(W126/B) ** C]
     Weibull parameter:   B = 97.9
     Weibull parameter:   C = 2.966
     Based on. estimates ol 1990 maximum 3 month 12  hour W126 ozone exposure
                                                                               M
Note:  NCLAN  exposure-response functions used in projections of yield loss.
This map shows the geographic range  for  this species.  It does not indicate
                                                                                                                                w
                                                                                                                                00

-------
                                            Estimated   Yield  Loss
                                                  Sorghum  -  1990

     Shadefl areas  represent major sorghum growing regions
     Welbflll model:     PRBL - 1  - expl -(W126/B) ** C]
     Weibull parameter:   B = 20S.3
     Weibull parameter:   C = 1.957                                                   ffl
     Based on estimates of 1990 maxmum 3 month 12 hour W126 ozone exposure                 "
Note:  NCLAN exposure-response  functions used in projections of yield  loss.
This map shows the geographic range for this species.   It does not indicate
that this  species will be  found at every point within  its range.
                                                                                                                              M
                                                                                                                               I
                                                                                                                              vo
            }/mini / moo oml

-------
                                         Estimated  Seedling   Biomass  Loss
                                                       Black  Cherry  -  1990
                                                                                                                                                  w
                                                                                                                                                   I
      Shaded areas represent black cherry growing regions
      Weibull model:      PRBL = 1 . exp[ .(W126/B) ** C]
      Weibull parameter:   B =  38.92
      Weibull parameter:   C =  0.9921
      Based on estimates of 1990 maximum 3 month 12 hour W126 ozone exposure

      NOTE: This map is (or seedlings, not mature trees of the same species.
      This map shows the geographic  range of this species.  It does not indicate
      that this species will be found at every point within its  range.
Percent Loss
     o - 4.9
    5.0 -  9.9.
    10.0  - 19.9
    19.0  - 29.9-
     > 30.0
//dgta2/ondvh/tii!> !

-------
                                        Estimated  Seedling   Biomass   Loss
                                                     Tulip Poplar -  1990
                                                                                    Percent Loss
                                                                                      EH3 0-4.9
                                                                                         5.0 -  9.9
Shaded areas represent tulip  poplar growing  regions
Weibull model:    PRBL = ,  . exp[  .  30.0 —^
                                                                                                                                          W
                                                                                                                                           I
//dotoZ/ondvh/bio losjt/iolot/ moo.ami

-------
                                        Estimated  Seedling   Biomass  Loss
                                                      White  Pine  -  1990
     Shaded areas represent white  pine growing regions
     Weibull model:     PRBL = i  .  exp[ -(w128/B) ** C]
     Weibull parameter:  B =  63.23
     Weibull parameter:  C =  1.6582
     Based on estimates  of 1990 maximum 3 month 12 hour W126 ozone  exposure

     NOTE: This  map is for seedlings, not mature trees of the same species.
     This  map shows the geographic  range of this  species.   It does not indicate
     that this species will be found at every point within its  range.
//dglc2/ondvh/bio Ioss4/mlol/ moo.oml
                               04/23/96

-------
                                          Estimated  Seedling   Biomass   Loss
                                                       Aspen  (wild)  -  1990
      Shaded areas represent aspen growing regions
      Weibull model:      PRBL = ,  . exp[ .(W126/B) "  Cl
      Weibull parameter:   B = 109.81
      Weibull parameter:   C = 1.2198
      Based on estimates ol  1090 maximum 3 month 12 hour W128 ozone exposure

      NOTE: This  map is for seedlings, not mature trees of the same species.
      This map shows the geographic range of this species.  It does not indicate
      that this species will be found at every point within its  range.
//dfltrj2/ondvh/bio Ioss4/zolot/ moo, cm I
                                 04/23/96

-------
                                        Estimated  Seedling   Biomass  Loss
                                                    Aspen  (clone)  -  1990
                                                                                    Percent Loss
                                                                                     C3 0-1.9
                                                                                        2.0 - 3.9
                                                                                        4.0 - 5.9
                                                                                        6.0 - 9.9
                                                                                         > 10.0
Shaded areas represent aspen growing regions
Weibull model:     PRBL = 1  . exp[ .(W126/B)  ** C]
Weibull parameter:  B = 49.56
Weibull parameter:  C =  1.4967
Based on  estimates of 1990  maximum 3 month 12 hour W126 ozone exposure

NOTE: This map  is (or seedlings,  not mature trees of the same species.
This  map  shows  the geographic range of this species.  It does not indicate
that this species  will be  found at  every point within its range.
//dolo2/ondvh/bio loss*/lolol/ moo.ami
                               04/23/9S

-------
                                       Estimated  Seedling  Biomass  Loss
                                                      Sugar  Maple  -  1990
     Shaded areas represent sugar maple  growing regions
     Weibull model:     PRBL = 1 .  exp[ .(W126/B)  ** C]
     Welbull parameter:  B = 36.35
     Weibull parameter:  C = 5.7785
     Based on estimates  of 1990 maximum 3 month 12 hour  W126 ozone  exposure

     NOTE: This  map is for seedlings, not mature trees of the same  species.
     This  map shows the  geographic range of this species.  It does not  indicate
     that this species will  be found at every point within its range.
                                                         6.0 - 9.9'
                                                          > 10.0
                                                                                                                                         W
                                                                                                                                          I
                                                                                                                                         H
                                                                                                                                         (Jl
//dotn2/on
-------
                                       Estimated  Seedling   Biomass   Loss
                                                   Ponderosa  Pine -  1990
                                                 our W126 ozone exposure
Shaded areas represent ponderosa pine growing r
Weibull model:    PRBL  =  , . exp[ .(w128/B) *
Weibull parameter: B = 159.63
Weibull parameter: C = 1.1190
Based  on estimates of 1990 maximum  3  month

NOTE: This map is for  seedlings, not mature tree
This map shows the  geographic range  of this  species!   It  does not  indicate
that this species will  be found at every point within  its  range.
Percent  Loss
     0-1.9
 133 2.0  - 3.9
    4.0  - 5.9
    6.0  - 9.9
     > 10.0
                                                                                                                                        W
                                                                                                                                        I
/dcta2/ondvh/bio Ioss4/zolot/ moo.ami
                                23/96

-------
                                        Estimated Seedling  Biomass Loss
                                                   Red  Alder  -   1990

     Shaded areas represent red alder growing regions
     Weibijtl model:      PRBL = 1  - exp[ -(W126/B) ** C]
     Weibull parameter:    B = 179.06
     Weibull parameter:    C = 1.2377
     Based on  estimates of 1990 maximum  3 month  12 hour W126  ozone exposure
Note:   This map is for seedlings,  not mature  trees of the  same species.
This map shows the geographic range of this species.  It does not  indicate
that this species  will be found at, every point within its  range.	
                                                                                                                                 M
                                                                                                                                  I
                                                                                                                                 H
                                                                                                                                 -J
//dolo2/ondvh/liio Ioss3/ii>lol/ moo.oml
                              06/06/95

-------
                                        Estimated  Seedling   Biomass   Loss
                                                     Douglas  Fir  -  1990
     Shaded areas represent douglas fir growing  regions
     Weibull model:     PRBL = 1 . exp[ .(Wi28/B) **  C]
     Weibull parameter:  B = 106.83
     Weibull parameter:  C = 5.9631
     Based on estimates  of 1990 maximum  3  month 12 hour W126 ozone exposure

     NOTE: This  map is for seedlings, not mature trees  of the same  species.
     This map shows the geographic range  of this  species.   It does not  indicate
     that this species will be found at every point within its  range.
                                                                                                                                          w
                                                                                                                                          I
                                                                                                                                          l->
                                                                                                                                          m
//data2/ondvh/bio lossl/zolot/ moD.oml

-------
                                       Estimated  Seedling  Biomass  Loss
                                                     Virginia Pine  -  1990
                                                                	I        _. +-<
                                                                        	   .1?
     Shaded areas represent  Virginia pine growing  regions
     Weibull model:     PRBL = 1 .  exp[ .(W126/B)  ** C]
     Weibull parameter:  B = 1714.64
     Weibull parameter:  C = 1.0
     Based on estimates  of 1990 maximum 3 month 12 hour W126 ozone exposure

     NOTE: This  map is for  seedlings, not  mature trees of the same species.
     This  map shows the geographic range of this species.   It does not indicate
     that this species will be found at every point within  its  range.
                                                                                                                                         M
//ddoZ/indvh/oio lgsi4/lDlol/ moo.ami
                               04/23/96

-------
                                 Estimated  Seedling  Biomass   Loss
                                                Red Maple  -  1990
Shaded areas represent red maple growing regions
Weibull model:     PRBL  = 1  . expl .(W126/B) ** C]
Weibull parameter:  B = 318.12
Weibull parameter:  C = 1.3756
Based on estimates  of 1990 maximum  3 month 12 hour W126 ozone exposure

NOTE: This  map is for seedlings,  not mature trees of the same species.
This map shows the geographic range  of this  species.   It does not indicate
that this species will be found at  every point within its  range.
                                                                                                                                a
                                                                                                                                 i
                                                                                                                                to
                                                                                                                                o

-------
                         APPENDIX F
Selected Ambient Ozone Air Quality Distributions for NCLAN,
Rural  (Class  I)  and Urban Sites  in  Terms  of  Three  Different
                      Exposure Indices

-------
                                            F-l
% Total
                  R85CO a
                    aa
                   SUM06
  <.055 .065  .085  105 .125  .145  165  185
     .055 .075  095  .115 .135  .155 .175  .195
             Concentration Levels
% Total
35
30
25
20
15
10
 5
 0
                 R85CO
                    aa
                  AOT06
                                                                             I  i  i _  i
  <.055 .065  .085  105  .125 .145 .165 185
     .055 .075  .095 .115  .135 .155  .175 .195
             Concentration levels
                           <.055 .065 .085 .105  .125 .145  .165  185
                              .055 .075 .095  .115 .135 .155  .175 .195
                                     Concentration levels
     •R85CO: Taken from  the  1985 NCLAN  cotton  (McNair)  study  in
      Raleigh,  NC.

-------
                                               F-2
% Total
                   C82CO
                     aa
                   SUM06
  < .055  .065  .085  .105  .125  .145  .165 .185
     .055  .075  .095  .115  .135  .155  .175  .195
              Concentration Levels
                C82CO
                  aa
                 AOT06
<.055 .065 .085  .105  .125 .145  .165  .185
   .055  .075 .095 .115  .135  .155  .175 .195
           Concentration Levels
                             V. Total
                             60
                             50
                             40
                             30
                             20
                             10
                              0
                                                C82CO
                                                  aa
                                                 W126
                               <.055 .065  .085  .105  .125  .145 .165  .185
                                  .055  .075  .095  .115  .135  .155  .175 .195
                                           Concentration Levels
      bC82CO:  Taken  from the  1982 NCLAN cotton  (Shafer)  study  in
        California

-------
                                                F-3
                 A85SO c
                   aa
                  SUM06
< .055 .065  .085  .105  .125  .145 .165  .165
   .055  .075  .095  .115  .135  .155  .175  .195
            Concentration Levels
                                                    % Total
                                                    50
                                                    40

                                                    30

                                                    20

                                                    10
                 A85SO
                   aa
                 AOT06
<.055 .065  .085  .105  .125  .145  .165  .185
   .055  .075  .095 .115  .135  .155  .175 .195
            Concentration Levels



% Total
en .
A85SO
aa
W126

                           < .055 .065  .085  .105  .125  .145  165  .185
                               .055  .075  .095  .115 .135  .155  .175  .195
                                       Concentration Levels
    CA85SO:  Taken from  the 1985  NCLAN  soybean  (Corsoy)  study  in
       Argonne,  Illinois.

-------
                                       F-4
       Madison, Co./91d
              SUM06
      t
<0055  065  085  105  125  145   165  185
  0055 075  095  115  135  155  175  185
                            Madison, Co./91
                                                            AOT06
                    <0055  065   065  105  125  145   165   185
                      0055 075  095  115  135  155  175  195
                               Madison, Co./91
                                      W126
                       10
lilt
                        O055 065  .085  105   125  145  165  165
                          0055  .075  095  115  135  155  175 ' 195
  ••Madison County,  1991:  Big Meadows  Site,  Shenandoah National
   Park,  Virginia.   Elevation 1073m.

-------
                                     F-5
         Warren Co./91
               SUM06
so
40
30
10
 "0055  065   065  105  125  145  165   185
   0055 075  095  115  135  155  175  195
                             Warren Co./91e
                                                            AOT06
                    60
                    so
                    40
                    30
                    4
                        ill,.
                      O055  065  085   105  125  145  165   165
                        0055  075  095  115  135  155  175  196
                               Warren Co. /91
                                     W126
                      60-
                       t
                      40 -

                      30 j-

                      20 h
                       i
                      10 (-
1
II,.
                       <0055  065   085  105  125  145   165  165
                         0055 075  085  115  135  155  175 ' 195
 'Warren County, 1991: Dickey Ridge Site,  Shenandoah National
   Park, Virginia.   Elevation  610m.

-------
                                         F-6
        Albuquerque, NM
                 SUM06
100





 eo





 60





 40





 20





 0
   <0 055 0055 0065 0075 0085 0095 0105 0115 0125 0135
                               Albuquerque, NM
                                                                    AOT06
                       100





                        80





                        80





                        40





                        20





                        0
                                                      O0550055 0065 0075 0085 0085 0105 0115 0125 0135
                                  Albuquerque, NM
                                          W126
                         100






                          80






                          60






                          40





                          20





                          0
_•••
                            •=00550055 0065 0075 0065 0065 010S 0115 0125 0135

-------
100





 80





 80





 40





 20





 0
              Atlanta, GA
                   SUM06
                                               F-7
   O0550055 0065 0075 0085 0095 0105 0115 0125 0135
100





 80





 80





 40





 20





 0
              Atlanta, GA
                                                                           AOT06
                                                           O0550055 0065 0075 0085 0095 0105 0115 0125 013£
                                         Atlanta, GA
                                               W126
                              O0550055 0065 0075 0065 0095 0105 0115 0125 0135

-------
                                          F-8
100
80
60
40
      Chicago - Cook Co., IL
                 SUM06
   <0 055 0056 0065 0075 0085 0095 0105 0115 0125 0135
100

80

60

40

20

 0
     Chicago - Cook Co., IL
                                                                    AOT06
                                                      O0550055 0065 0075 0065 0095 0105 0115 0125 0135
                            100 r
                            4
                            60
                            40
                            20
                             0
                                 Chicago - Cook Co., IL
                                             W126
                               00550055 0.065 0075 0085 0095 0105 0115 0125 0135

-------
                APPENDIX G

Letters of Closure on the Criteria Document
 and Staff Paper from the Chairman of the
  Clean Air Scientific Advisory Committee
      to the Administrator of the EPA

-------
                UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                               WASHINGTON, D.C. 20460
                                  November 28, 1995

EPA-SAB-CASAC-LTR-96-001

Honorable Carol M. Browner
Administrator                                                       OFFICE OF THE ADMNKTRATOR
U.S. Environmental Protection Agency                                    SCENCE ADVISORY BOARD
401 M Street, SW
Washington, DC  20460

                    RE     CASAC Closure on the Air Quality Criteria for Ozone and Related
                           Photochemical Oxidants

Dear Ms  Browner

      A Panel of the Clean Air Scientific Advisory Committee (CASAC) of EPA's Science
Advisory Board (SAB) met on  July 20 and 21, 1994 and March 21.and 22, 1995, to review drafts
of the document entitled Air Quality Criteria for Ozone and Related Photochemical Oxidants  At
these meetings and in subsequent written comments, the Committee made extensive
recommendations for strengthening the document. In August 1995,  revisions to the Criteria
Document were mailed to the  CASAC members for review. On September 19, 1995, the Panel
met to complete this review The resulting comments by the Committee members note with
satisfaction the improvements made in the scientific quality and completeness of the Criteria
Document  The changes are consistent with CASAC's recommendations.

      At the September 1995 meeting the Panel came to closure on the Criteria Document It
was the consensus of the Panel members that the Criteria Document provides an adequate
review of the available scientific data and relevant studies of ozone and related photochemical
oxidants  The document is quite comprehensive and will provide an  adequate scientific basis
for regulatory decisions on ozone  and related photochemical oxidants based on available
information  At the meeting and subsequently in writing, Panel members provided the Agency
with additional comments for consideration  Most of these comments were directed at Chapter
5 the ecological effects chapter   Although the Panel would like to have these comments
considered for incorporation in the Criteria Document, the Panel did not feel that it was
necessary to review another revised version

      On behalf of the  Panel. I would like to thank EPA staff for their efforts in preparing the
Criteria Document on the accelerated schedule  Our comments on the OAQPS Staff Paper will
follow shortly
                                             Sincerely,
                                       Dr. George T. Wolff, Chair
                                Clean Air Scientific Advisory Committee
                                                                   Recycled/Recyclable

-------
                           SCIENCE ADVISORY BOARD
                CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE
                              OZONE PANEL ROSTER

Chair
Dr. George T. Wolff, General Motors Environmental & Energy Staff, Detroit, Ml

Members
Dr. Stephen M. Ayres. Office of International Health Programs, Virginia Commonwealth University. Medical
       College of Virginia, Richmond, VA

Dr. Jay S. Jacobson. Boyce Thompson Institute, Cornell University, Ithaca, NY

Dr. Joe L. Mauderly. Inhalation Toxicology Research Institute, Lovelace Biomedical & Environmental
       Research Institute. Albuquerque NM

Dr. Paulette Middleton. Science & Policy Associates, Inc..Boulder, CO (did not participate)

Dr. James H. Price, Jr.. Texas Natural  Resource Conservation Commission, Austin. TX

Consultants to CASAC
Dr. Stephen D. Colome, Integrated Environmental Services. University Tower. Irvine CA

Dr. A. Mynck Freeman, Department of Economics. Bowdoin College,, Brunswick, ME

Dr. Allan Legge. Biosphere Solutions,Calgary, Alberta, CANADA

Dr. Morton Lippmann, Institute of Environmental Medicine, New York University Medical Center, Tuxedo, NY

Dr. William Manning. Department of Plant Pathology, University of Massachusetts. Amherst. MA

Dr. Roger O. McClellan. Chemical Industry  Institute of Toxicology, Research Triangle Park, NC

Dr. D. Warner North. Decision Focus, Inc., Mountain View, CA

Dr. Frank E.  Speizer. Harvard Medical School, Channmg Lab, Boston, MA

Dr. George Taylor, Biological Services Center, Desert Research Institute, University of Nevada, Reno, NV

Dr. Mark J. Utell, Pulmonary Disease Unit, University of Rochester Medical Center, Rochester, NY

Science Advisory Board Staff
Mr. A. Robert Flaak, Designated Federal Official, U. S. Environmental Protection Agency, Science Advisory
        Board (1400F). 401 M Street. SW, Washington, DC 20460

Ms. Connie Valentine, Staff Secretary, U S Environmental Protection Agency. Science Advisory  Board
        (1400F) 401  M Street, SW, Washington. DC  20460

-------
               UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                           WASHINGTON, D.C. 20460
                                    November 30, 1995

EPA-SAB-CASAC-LTR-96-002
                                                              SCENCE ADVISORY BOARD
Honorable Carol M. Browner
Administrator
U.S  Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460

                  RE.   CASAC Closure on the Primary Standard Portion of the Staff
                        Paper for Ozone

Dear Ms  Browner

      A Panel of the Clean Air Scientific Advisory Committee (CASAC) of EPA's
Science Advisory Board (SAB)  met on March 22,  1995, to review a draft of the primary
standard part of the document entitled Review of  National Ambient Air Quality
Standards for Ozone Assessment of Scientific and Technical Information OAQPS Staff
Paper  At that time, a draft of the secondary standard portion of the document was not
completed  At the March meeting, the Panel made extensive recommendations for
strengthening the document. In August 1995, a revised Staff Paper, which included a
first draft of the secondary standard portion was sent to CASAC panel members for
review. On September 19 and  20, 1995, the Panel met to compete this review. The
Pane! members' comments reflect their satisfaction with the improvements made in the
scientific quality and completeness of the primary standard portion of the Staff Paper
The changes made in that portion of the document are consistent with CASAC's
recommendations  However, the Panel Members provideo additional comments to your
staff at the meeting and subsequently in writing.  Although the Panel would like to have
these comments considered for incorporation in the  Staff Paper, the Panel did not feel
that it was necessary to review another revised version and came to closure on the
primary standard portion   It was the consensus of the Panel that although our
understanding of the health effects of ozone is far from complete, the document
provides an adequate scientific basis for making regulatory decisions concerning a
primary ozone standard

      The Panel could not come to closure, however, on the secondary standard
portion of the Staff Paper which was a first draft.  To facilitate further development of
                                                             flecycted/Recyciabi*
                                                             Pnnt«o with Soy/Canoa i<* on pao*r nag
                                                             oomara •> Maa £0% wcyctad Hbar

-------
this part of the Staff Paper, the Panel members have provided detailed comments to
your staff. The Panel felt that the suggested revisions were extensive enough to
warrant a review of the next draft.

       I would like to summarize for you the Panel's recommendations concerning the
primary standard. It was the consensus of the Panel that EPA's selection of ozone as
the surrogate for controlling photochemical oxidants is correct. It was also the
consensus of the Panel that an 8-hour standard was more appropriate for a human
health-based standard than a 1 -hour standard. The Panel was in unanimous
agreement that the present 1-hour standard be eliminated and replaced with an 8-hour
standard.

       The Panel felt that the weight of the health effects evidence indicates that there is
no threshold concentration for the onset of biological responses due to exposure to
ozone above background concentrations.  Based on information now available, it
appears  that ozone may elicit a continuum of biological responses down to background
concentrations This means that the paradigm of selecting a standard at the lowest-
observable-effects-level and then providing an "adequate margin of safety" is no longer
possible.  It further means that EPA's risK assessments must play a central role in
identifying an appropriate level

       To conduct the risk assessments, the Agency had to identify the population at
risk and the physiological responses of concern, develop a model to estimate the
exposure of this  population to ozone, and develop a model to estimate the probability of
an  adverse physiological response to the exposure. The Panel agrees with EPA that
the selection of "outdoor children" and  "outdoor workers," particularly those with
preexisting respiratory disease are  the appropriate populations with the highest risks
After considerable debate, it was the consensus of the Panel that the Agency's criteria
for the determination of an adverse physiological response was reasonable
Nevertheless, there was considerable concern that the criteria for grading physiological
and clinical responses to ozone was confusing if not misleading  The Panel concurs
with the Agency  that the models selected to estimate exposure and risk are appropriate
models  However, because of the myriad of assumptions that are made to estimate
population exposure and risk, large-uncertainties exist in these estimates.

       The results of two of the risk analyses are presented in Tables VI-1 and VI-2 in
the Staff Paper and are reproduced in the attached tables.  The ranges of the risk
estimates across nine cities for outdoor children are presented in Table VI-1   Because
of the large number of stochastic variables used in the exposure models, the exposure
estimates vary from run to run.  However, the ranges are not reflective of all of the

-------
uncertainties associated with the numerous assumptions that were made to develop the
estimates

      The single estimates presented in Table VI-2 do not reflect any of the
uncertainties associated with these estimates. (Table VI-2 contains only the estimated
hospital admissions due to asthma which account for over 85% of the estimated total
hospital admissions due to ozone exposure).  These uncertainties need to be explicitly
articulated in order to put the estimates in proper perspective. Nevertheless, based on
the results presented in these and other similar tables presented in the Staff Paper, the
Panel concluded that there is no "bright line" which distinguishes any of the proposed
standards (either the  level or the number of allowable exceedences) as being
significantly more protective  of public health.  For example, the differences in the
percent of outdoor children (Table VI-1) responding between the present standard and
the most stringent proposal (8H1EX at 0.07 ppm) are small and their ranges overlap for
all health endpoints  In Table VI-2, the estimates in row 1, which appeared in the draft
Staff Paper, suggest considerable differences between the several options.  However,
when ozone-aggravated asthma admissions are compared to total asthma admissions
(rows 5 and 6), the differences between the various options are small. Consequently,
the selection of a specific level and number of allowable exceedences is a policy
judgment  Although it was the consensus of the Panel that the ranges of concentrations
and allowable exceedences proposed by the Agency were appropriate, a number of
Panel members expressed "personal" preferences for the level and number of allowable
exceedences  Of the ten pane!  members who expressed their opinions, all ten favored
multiple allowable exceedences, three favored a level of 0.08 ppm, one favored the mid
to upper range (0.08 - 0.09 ppm),  three favored the upper range (0.09 ppm), one
favored a 0 009 - 0 10 ppm range with health  advisories issued when the 8-hour ozone
concentration was forecasted to exceed 0 007 ppm, and two just endorsed the range
presented by the Agency as  appropriate and stated that the selection should be a policy
decision  The members who favored  the lower numbers expressed concern over  the
evidence for chronic deep lung inflammation from the controlled human and animal
exposure studies and the observations of pain on deep inspiration in some subjects

      Because there is no apparent threshold for responses  and no "bright line" in the
risk assessment, a number of panel members recommended that an expanded air
pollution warning system be initiated so that sensitive individuals can  take appropriate
"exposure avoidance' behavior  Since many areas of the country already have an
infrastructure in place to designate 'ozone action days" when voluntary emission
reduction measures are put in place, this idea may  be fairly easy to implement.

      It was also the  consensus of the Panel  that the form of the 8-hr standard be more
robust than the present  1-hour standard  The present standard is based on an extreme

-------
value statistic which is significantly dependent on stochastic processes such as extreme
meteorological conditions.  The result is that areas which are near attainment will
randomly flip in and out of compliance.  A more robust, concentration-based form will
minimize the "flip-flops," and provide some insulation from the impacts of extreme
meteorological events. The Panel also endorses the staff recommendation for creating
a-"too close to call" category.

      Since the last ozone NAAQS review, the scientific community has made great
strides in their understanding of the health effects of ozone exposure because of
ongoing research programs.  Panel members were very  impressed with how much
more we understand now as compared  to the prior round. Nevertheless,  there are still
many gaps in our knowledge  and large  uncertainties in many of the assessments.  For
example, there is little information available on the frequency of human activity patterns
involving outdoor physical exercise. Little is also known about the possible chronic
health impacts of ozone exposure over  a period of many years. In addition, there is no
clear understanding of the significance  of the inflammatory response inferred from the
bronchoiavage data.  Panel members stated, however, that the scientific community is
now in a position to frame the questions that need to be  better resolved so the
uncertainties can be reduced  before the next ozone review in 5 years. For this reason,
it is important that research efforts on the health and ecological effects of ozone not be
reduced because we have come to closure on this review

      CASAC would appreciate being kept informed of progress on establishing a
revised or new ozone standard, and plans for research on ozone effects.  Please do not
hesitate to contact me if CASAC can be of further assistance in this matter.  We look
forward to receiving the revisions of the secondary standard portion of the Staff Paper

                                          Sincerely.
                                    Dr. George T. Wolff,
                              Clean Air Scientific Advisory Committee

-------
ExttrnaJ Rtvltw Drqft
August 1995
154
                          Do Not Quoit or Ctlt
                    TABLE VM. RANGE OF MEDIAN PERCENT OF OUTDOOR CHILDREN
                      RESPONDING ACROSS NINE U.S. URBAN AREAS UPON ATTAINING
                                 ALTERNATIVE AIR QUALITY STANDARDS.
Health Endpolnts
FEV, decrement J>.
15%'
FEV, decrement £_
20%'
Moderate or
Severe Pain on
Deep Inspiration*
Moderate .or
Severe Cough*
Range of Median Risk Estimates Associated With Just Attaining Alternative Standards
(percent of outdoor children responding)
Alternative 1-Hour
NAAQS
11I1EX
0.12
ppm
4.6-13.7
1.1-5.9
_ 	 a,. .
0.5-1.6 •
1.4-3.7
1II1EX
0.10
ppm
2.6-9.4
0.4-3.5
0.1-0.4
0.9-2.6
Alternative 8-Hour Dally Maximum Standards
1 Expected Exceedance Standards
8II1EX
0.10
ppm
6.9-
16.1
2.4-7.3
1.0-2.6
1.9-3.9
811 1 EX
0.09
ppm
4.9-11.9
1.5-4.8
0.6-1.7
1.5-3.3
8111 EX
0.08
ppm
3.3-8.3
0.9-2.8
0.4-1.7
1.0-2.5
8111 EX
0.07
ppm
1.7-5.1
0.2-1.2
0.2-0.7
0.7-1.9
5 Expected Exceedance
Standards
8II5EX
0.09
ppm
5.2-14.3
1.6-6.3
0.2-0.8
1.5-3.3
8II5EX
0.08
ppm
3.3-10.3
0.8-4.1
0.4-1.7
1.1-2.0
    'Risks associated with 8-hour exposures under moderate exertion (equivalent ventilation rate j>. 15 I mtn' m') based on
exposure-response relationships derived from Folinsbee et al. (1988), Horstman et al. (1989), and McDonnell et at. (1991).
    .         t                                                                 \
     Risks associated with 1-hr exposures under moderate exertion (equivalent ventilation rate j>. 16 and  30 I min'1 m') based on
              	i.*urta.it»« derived from McDonnell et al. (1983).

-------
                                              Table VI-2 (revised)
      ESTIMATED HOSPITAL ADMISSIONS FOR ASTHMATICS IN THE NEW YORK CITY AREA

Excess Admissions*
% A from present std
Excess + background"
% A from present std
All asthma admissions'
% A from present std
1H1EX
0.12
210
0%
890
0%
28,295
0%
1H1EX
0.10
130
-38%
810
-9%
28,215
-0 3%
8H1EX
0.10
240
+ 14%
920
+ 3%
28,325
+0 1%
8H1EX
0.09
180
-14%
860
-3%
28,265
-0.1%
8H1EX
0.08
110
-48%
790
-11%
28,195
-0 4%
8H1EX
0.07
60
-71%
740
-17%
28,145
-0.5%
8H5EX
0.09
180
-14%
860
-3%
28,265
-0 1%
8H5EX
0.08
120
-42%
800
-10%
28,205
-0 3%
AS IS
=385"
+83%
1065'
-t-20%
28,470'
+0.6%
a - excess asthma admissions attributed to ozone levels exceeding a background concentration of 0 04 ppm; from Table VI-2, page 155 in the
      August 1995 OAQPS Draft Staff Paper
b - asthma admissions included in (a) plus those due to background ozone concentrations; admissions due to background = 1065* - 385" = 680
c - asthma admissions due to all causes = 28,470' - 385d - Excess Admissions from row 1
d estimated from Figure V-15, page 125  in the August 1995 OAQPS Draft Staff Paper
e from page 127, line 13 in the August 1995 OAQPS Draft Staff Paper
f total admissions from asthma = total asthmatics (365,000 - from page 126, line 24) x hospitalization rate (78/1000 asthmatics - from page 126,
      line 29)

-------
     f*T        UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
       \                   WASHINGTON. D.C. 20460
                                April 4, 1996

EPA-SAB-CASAC-lTR-96-006

Honorable Carol M. Browner
Administrator
U S  Environmental Protection Agency
401 M  Street SW
Washington, DC 20460
                  RE:   Closure by the Clean Air Scientific Advisory Committee
                       (CASAC) on the Secondary Standard Portion of the Staff
                       Paper for Ozone
Dear Ms  Browner
      A Panel of the Clean Air Scientific Advisory Committee (CASAC) of EPA's
Science Advisory Board (SAB) met on March 22, 1995. to review a draft of the primary
standard portion of the document entitled Review of National Ambient Air Quality
Standards for Ozone Assessment of Scientific and Technical Information - OAQPS Staff
Paper  At  that time, a draft of the secondary standard portion of the document was not
completed  In August. 1995 a revised Staff Paper, which included a first draft of the
secondary standard portion was  sent to the CASAC panel members for review  On
September 19 and 20 1995. the Panel met to complete this review. The Pane!
members' comments reflect their satisfaction with the improvements made in the
scientific Quality ana completeness of the primary standard portion of the Staff Paper
and reacned closure on that part (see CASAC Letter Report: EPA-SAB-CASAC-LTR-
96-002  November 30 1995).  However, the Panel could  not come to closure on the
seconaary standard portion of the Staff  Paper which was  a first draft. To facilitate
further development of this part of the Staff Paper, the Panel members provided
detailed comments to your staff  The Panel felt that the suggestea revisions were
extensive enougn to warrant a review of the next draft

      On March 21   1 996  a subset of the Panel, consisting of all four of the Panel
members with expertise in ozone effects on vegetation plus three additional CASAC
members met in Research Triangle Park, NC to review a second draft of the secondary
portion of the Staff Pacer   In addition, a Panel member with expertise in economics
                                                                      7S*

-------
reviewed the Staff Paper and provided written comments.  Overall, the comments from
the Panel members reflected their satisfaction that the Staff Paper was much improved;
however, the verbal and written comments provided to your staff indicated that
important, additional modifications are still required. Nevertheless, it was the
consensus of the Panel that an additional review of the document by the Panel was not
necessary. Consequently, the majority of the Panel agreed to come to closure on the
Staff Paper assuming that the Agency would incorporate the Panel's latest comments.
It was the opinion of six of the seven members of the Panel who were present that the
Staff Paper will provide an appropriate scientific basis for making regulatory decisions
concerning a secondary ozone standard once the additional changes are incorporated.
The additional modifications are summarized below.

      It should be pointed out that the Panel members all agreed that damage is
occurring to vegetation and natural resources at concentrations below the present
1-hour national ambient air quality standard (NAAQS) of 0.12 ppm.  The vegetation
effects experts were in agreement that plants appear to  be more sensitive to ozone than
humans.  Further, it was agreed that a secondary NAAQS, more stringent than the
present primary standard, was necessary to protect vegetation from ozone.  However,
agreement on the level and form of such a standard is still elusive for a number of
reasons

      The first issue is the level of uncertainty associated with the crop loss risk
assessment presented in Tables Vll-5a-d through VII-7 of the Staff Paper. While some
of the sources of uncertainty are addressed earlier in  the Staff Paper, other sources of
uncertainty are not addressed at all.  The estimates in these Tables should only be
presented as rough estimates for a number of reasons.  First, the dose-response
functions are based upon open-top chamber studies which have the advantage of
providing the least amount of environmental  modification of any outdoor chamber but.
nevertheless, they still alter ambient  microclimate conditions which will introduce
uncertainty  In these studies, plant response to ozone has been optimized under
conditions which do not reflect the real-life ambient field 'conditions.  Two of the plant
experts said that the open-top chamber experiments by  their very design and execution
produced results that overestimated  the effects of ozone on plant yield  The other  two
experts agreed that the open-top chambers do alter the  environment in the chamber
with respect to ambient field conditions but did not agree with there being a positive
bias  Research has not yet provided methods that clearly are better than open-top
chambers for establishing ozone dose-response relationships for a wide variety of
crops  Second, the estimated exposures are based on a non-peer-reviewed, empirical
model which has not been subjected to any performance evaluation.  In addition,

-------
 insufficient details are given either in the Staff Paper or the unpublished Agency report
 for anyone to perform an evaluation. Third, the estimated exposures are then
 extrapolated to hypothetical scenarios where various secondary NAAQS are attained.
 Details of this extrapolation procedure are also insufficient to judge the appropriateness
 of the procedure.  Fourth, the exposure estimates are then extrapolated to the entire
 coterminous U.S. using a Geographic Information System (CIS) which is based on an
 unpublished, non-peer-reviewed, internal EPA memorandum that contains insufficient
 details to adequately evaluate the GIS. The exposure estimates and the dose-response
 function estimates are then input into the economic models which introduce additional
 uncertainties. Furthermore, the losses are computed from an assumed 12-hr.
 background ozone concentration of 0.025  ppm which is too low and will over-inflate  the
 crop loss estimates. A more reasonable 12-hr, daylight, summertime background is
 more likely closer to the 8-hr, background  of 0.03-0.05 ppm. As a result, the Panel felt
 that the absolute values of the numbers in Tables VII-5a-VII-7 are highly uncertain
 estimates of crop losses and are a result of a propagation of uncertainties.  They are
 rough estimates, and this should be explicitly stated in this discussion. The Panel
 believes, however, that these Tables can be of some use in identifying rough relative
 incremental benefits associated with a given NAAQS as long as it is recognized that
 small differences in benefits may have no significance because of these uncertainties.

       A related issue is the estimated yield losses and seedling biomass losses
 displayed on the maps in Appendix E of the Staff Paper.  Since these are also based on
 the results of open-top chamber experiments as well as the results of the GIS
 technology approach, the uncertainties are large. The concern here is that the maps
 will be used out of context and the caveats ignored. The limitations and uncertainties of
 the data need to be clearly stated in the legend of each map

       The SUM06 standard reflects a cnange in thinking over the current 1-hour
 standard with respect to  how plants resoond to ambient ozone exposure. This
 proposed form of the standard implicitly recognizes  that vegetation response to ambient
 ozone is cumulative.  However, there is disagreement over whether this is the best form
 for a cumulative standard and what the level of the standard should be to protect
 vegetation from damage by ozone. One of the Panel's ecology experts thinks the form
 and the range of between 25 to 38 ppm-hours proposed by the Agency is appropriate
 A second expert thinks the form proposed by the Agency is appropriate and biologically
 based, but feels that a level  of 20 ppm-hours is necessary to adequately protect natural
 resources. The other two experts are uncomfortable with a SUM06 form because they
feel it lacks a biological basis. One member stated that he feels very uncomfortable
with SUM06 and would not want to defend  it because he feels there is too much

-------
uncertainty associated with its derivation. The fourth expert is concerned that a SUM06
form is unnecessarily complicated, and the level proposed by the Agency would not
eliminate ozone damage.  Instead, he proposes that the 1-hour average ozone should
not exceed 0.05 ppm for more than one hour between the hours of 0700-1500.  In his
written comments, the Panel's economist noted that the welfare benefits of a secondary
standard depend on the decision regarding the primary standard. For example, he
points out that if the primary standard remains at 0.12 ppm for 1-hour, or is changed to
an 8-hour standard of 0.09 ppm with one allowable exceedence, Table Vll-5a suggests
potentially significant incremental benefits associated with a secondary standard based
on SUM06. He further states that if the primary standard is set at 0.07 or 0.08 ppm with
one exceedence, there is little to be gained by establishing a separate secondary
standard.

      Although the three remaining CASAC members were neither biologists or
economists, they offered their opinion on the secondary standard proposals. Two  think
the form proposed by the Agency is appropriate. One thinks that the level proposed by
the Agency is appropriate, while the other feels that the Administrator's discretion
should be broaaer than the range presented in the Staff Paper. One of these members
pointed out, however, that the Staff Paper does not make  it clear enough that the
SUM06 standard as proposed is a practical choice being made as to the level of effects
that will be tolerated and not a level that  will prevent effects from occurring.  The third is
uncomfortable with SUM06 and based on the estimates in Tables VII-5a-VII-7,
recommends an 8-hour standard at the same level as the  new primary standard. The
three members also concurred that given the crudeness of the risk assessment
estimates, policy decisions cannot be based firmly on science.

      A number of the Panelists offered  their insights as to why there are such
divergent opinions on the recommended  form and level of the standard.  The main
issues  are the lack of sufficient rural ozone data, ana the lack of relevant plant exposure
studies There are serious deficiencies in terms of the distribution of monitoring sites,
particularly in rural areas that prevent us from accurately assessing exposure once
ozone damage is observed.  The Panel is in agreement that plants are being damaged
by ozone and that the current secondary standard is not sufficiently protective, but there
remain important limitations to our understanding of the extent of the response of
vegetation to ozone under field conditions.  Five years from now, if we do not have the
results of research coupling ozone air quality and plant biology under conditions more
representative of ambient field conditions, to avoid the shortcomings  of the open-top
chamber experiments, then we will continue to be hampered by our inability to come  to
consensus on the levels of air quality that are protective of vegetation and ecosystems

-------
 at the most reasonable cost. In addition, a number of Panelists expressed the
 importance of knowing the consequences of decisions concerning National Ambient Air
 Quality Standards. Once a decision is made to change the standard or to maintain the
 status quo, we must be able to determine, by appropriate monitoring and research, what
 the consequences will be in terms of ambient air quality and effects on vegetation and
-ecosystems.

       In summary, a majority of the Panel has come to closure on the secondary part
 of the ozone Staff Paper despite the desire of the Panel for additional significant
 revisions. These revisions have been communicated to your staff by this letter and in
 written comments by individual Panel members.  The Panel trusts that your staff will
 address these concerns.

      CASAC would appreciate being kept informed of progress  on establishing a
 revised or new ozone standard, and plans for research on ozone  effects.  Please do not
 hesitate to contact me if CASAC can be of further assistance in this matter.  We look
 forward to seeing the final version of the secondary standard portion of the Staff Paper.

                                    Sincerely,
                                    Dr. Georcje T. Wolff, Cha;r
                                    Clean Air Scientific Advisory Committee

-------
                       SCIENCE ADVISORY BOARD
              CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE
            OZONE PANEL ROSTER - MARCH 21, 1996 MEETING
                       Secondary Standard Review

-Chair
 Dr. George T. Wolff, General Motors Environmental & Energy Staff, Detroit, Ml

 Members
 Dr. Stephen M. Ayres,  Office of International Health Programs, Virginia
      Commonwealth University, Medical College of Virginia, Richmond, VA

 Dr. Philip K. Hopke, Department of Chemistry, Clark University, Potsdam, NY

 Dr. Jay S. Jacobson, Boyce Thompson Institute, Cornell University, Ithaca, NY

 Dr. Joe L. Mauderly, Inhalation Toxicology Research Institute, Lovelace Biomedical &
      Environmental Research Institute, Albuquerque, NM

 Dr. James H. Price, Jr., Research & Technology Section, Texas Natural Resource
      Conservation Commission, Austin, TX

 Consultants to CASAC
 Dr. A. Myrick Freeman, Professor, Department of Economics, Bowdoin College,
      Brunswick, ME

 Dr. Allan Legge, Biosphere Solutions, Calgary, Alberta,  CANADA

 Dr. William Manning, Department of Plant Pathology, University of Massachusetts
      Amherst, MA

 Dr. George Taylor  Biological Services Center, Desert Research Institute University of
      Nevada. Reno, NV

 Science Advisory Board Staff
 Mr. A. Robert Flaak. Designated Federal Official,  U. S. Environmental Protection
      Agency, Science Advisory Board (1^00F), Washington, DC 20460

 Mrs. Dorothy Clark, Staff Secretary, U. S. Environmental Protection Agency. Science
      Advisory Board (1400F), Washington, DC  20460

-------
                                 NOTICE

      This report has been written as part of the activities of the Science Advisory
 Board, a public advisory group providing extramural scientific information and
 advice to the Administrator and other officials of the Environmental Protection
"Agency. The Board is structured to provide balanced, expert assessment of
 scientific matters related to problems facing the Agency.  This report has not
 been reviewed for approval by the Agency and, hence, the contents of this report
 do not necessarily represent the views and policies of the Environmental
 Protection Agency, nor of other agencies in the Executive Branch of the Federal
 government, nor does mention of trade names or commercial products constitute
 a recommendation for use.

-------
                      DISTRIBUTION LIST
Administrator
Deputy Administrator
Assistant Administrators
Director, Office of Air Quality Planning and Standards, OAR
Director, Office of Science Policy, ORD
EPA Regional Administrators
EPA Laboratory Directors
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
Library of Congress
National Technical Information Service

-------
TECHNICAL REPORT DATA
(f lease read Instructions on the reverse before completing)
1 REPORT NO 2
EPA-452/R-96-007
4 TITLE AND SUBTITLE
Review of the National Ambient Air Quality Standards
for Ozone-Assessment of Scientific and Technical
Information: OAQPS Staff Paper
7 AUTHOR(S)
McKee, D.J.; Atwell, V.V. ; Richmond, H.M. ; Freas, W.P.;
Rodriguez, R.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, North Carolina 27711
12 SPONSORING AGENCY NAME AND ADDRESS
3 RECIPIENT'S ACCESSION NO.
5 REPORT DATE
June 1996
6 PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT t*Q.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
 16 ABSTRACT

   This staff paper evaluates and interprets the updated scientific and technical
   information that EPA staff believes is most relevant to the review cf primary and,
   secondary national ambient air quality standards for ozone.  This assessment is
   intended to bridge the gap between the scientific review in the 1996 criteria
   document and the judgments required of the Administrator in setting ambient air
   quality standards for ozone.   The major recommendations presented in the staff pap«r
   include:  (1)  Ozone should remain as the surrogate for controlling ambient
   concentrations of photochemical oxidants; (2) The one-hour primary standard should
   be replaced by an 8-hour standard; (3) The range of consideration for the level of
   the primary standard should be 0.07 to 0.09 ppm; (4) Consideration should be given
   to the current expected exceedance form,  ranging from 1- to 5-expected exceedances,
   averaged over 3 years,  as well as to a concentration-based form, ranging from the
   second to the fifth highest 8-hour daily maximum concentration,  averaged over 3
   years; (5)  Consideration should be given to defining the primary standard in terms
   of a range of air quality values; (6)  If the Administrator determines that
   additional protection is needed beyond that provided by the alternative primary
   standards recommended,  or that no revisions to the primary are warranted, and/or
   that establishing a seasonal  form for  the secondary standard is justified,
   consideration should be given to a new secondary standard in the form of a 3-month,
   12-hour, SUM06 exposure index set at a level within the range of approximately 38 to
   25 ppm-hours.
"7 KEY WORDS A\-D DOCUMENT ANALYSIS
J DESCRIPTORS
Ozone
Photochemical Oxidants
Air Pollution
Health Effects
b IDENTIFIERS/OPEN1 ENDED TERMS
Air Quality Standards
v CCSATi I icId'Group

I  Welfare Effects
I  Exposure Assessment
  Risk Assessment  Economic Benefits
  Release to Public
8Q
                                                          (-'•/ ;-• .•<.-•*.i,c ^.

-------