United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Research Triangle Park, NC 27711
EPA/60Q/8-84/Q2QbF
August 1 986
Research and Development
Air Quality
Criteria for
Ozone and Other
Photochemical
Oxidants
Volume II of V
-------
-------
EPA/600/8-84/020bF
August 1986
Air Quality Criteria
for Ozone and Other
Photochemical Oxidants
Volume It of V
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, IM.G. 27711
-------
DISCLAIMER
This document has been reviewed ih accordance with U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade'
names or commercial products does not constitute endorsement or recommendation
for use. " ' L
IT
-------
ABSTRACT
Scientific information is presented and evaluated relative to the, health
and welfare effects associated with exposure to ozone and other phbtqqhemical
oxidants. Although it is not intended as a complete and detailed'literature
review, the document covers pertinent literature through early 1986. ;
Data on health and welfare effects are emphasized, but additional infor-
mation is provided for understanding the nature of the oxidant pollution pro-
blem and for evaluating the reliability of effects data as well as their
relevance to potential exposures to ozone and other oxidants at concentrations
occurring in ambient air. Information is provided on the following exposure-
related topics: nature, source, measurement, and concentrations of precursors
to ozone and other photochemical oxidants; the formation of ozone and other
photochemical oxidants and their transport once.formed; the properties, chem-
istry, and measurement of ozone and other photochemical oxidants; and the
concentrations of ozone and other photochemical oxidants that are typically
found in ambient air.
The specific areas addressed by chapters on health and welfare effects
are the toxicological appraisal of effects of ozone and other oxidants; effects
observed in controlled human exposures; effects observed in field and epidemi-
ological studies; effects on vegetation seen in field and controlled exposures;
effects on natural and agroecosystems; and effects on nonbiological materials
observed in field and chamber studies.
m
-------
AIR QUALITY CRITERIA FOR OZONE
AND OTHER PHOTOCHEMICAL OXIDANTS
Page
VOLUME I
Chapter 1. Summary and Conclusions 1-1
VOLUME II
Chapter 2. Introduction 2-1
Chapter 3. Properties, Chemistry, and Transport of Ozone and
Other Photochemical Oxidants and Their Precursors 3-1
Chapter 4. Sampling and Measurement of Ozone and Other
Photochemical Oxidants and Their Precursors . ..... 4-1
Chapter 5. Concentrations of Ozone and Other Photochemical
Oxidants in Ambient Air 5-1
VOLUME III
Chapter 6. Effects of Ozone and Other Photochemical Oxidants
on Vegetation 6-1
Chapter 7. Effects of Ozone on Natural Ecosystems and Their
Components 7-1
Chapter 8. Effects of Ozone and Other Photochemical Oxidants
on Nonbiological Materials 8-1
VOLUME IV
Chapter 9. Toxicological Effects of Ozone and Other
Photochemical Oxidants 9-1
VOLUME V
Chapter 10. Controlled Human Studies of the Effects of Ozone
and Other Photochemical Oxidants 10-1
Chapter 11. Field and Epidemic!ogical Studies of the Effects
of Ozone and Other Photochemical Oxidants 11-1
Chapter 12. Evaluation of Health Effects Data for Ozone and
Other Photochemical Oxidants 12-1
iv
-------
TABLE OF CONTENTS
LIST OF TABLES xi
LIST OF FIGURES xiv
LIST OF ABBREVIATIONS AND SYMBOLS xvii
AUTHORS, CONTRIBUTORS, AND REVIEWERS xxii
2. INTRODUCTION , 2-1
2.1 PURPOSE AND LEGISLATIVE BASIS OF THIS DOCUMENT 2-1
2.2 THE OXIDANT PROBLEM 2-2
2.3 SCOPE AND ORGANIZATION OF THIS DOCUMENT 2-4
2.4 REFERENCES 2-7
3. PROPERTIES, CHEMISTRY, AND TRANSPORT OF OZONE AND OTHER
PHOTOCHEMICAL OXIDANTS AND THEIR PRECURSORS : 3-1
3.1 INTRODUCTION 3-1
3.2 DESCRIPTIONS AND PROPERTIES OF OXIDANTS AND THEIR
PRECURSORS 3-2
3.2.1 Ozone and Other Photochemical Oxidants 3-2
3.2.1.1 Ozone 3-2
3.2.1.2 Peroxyacetyl Nitrate 3-3
3.2.1.3 Hydrogen Peroxide 3-4
3.2.1.4 Formic Acid 3-7
3.2.2 Organic Precursors 3-7
3.2.2.1 Hydrocarbons 3-8
3.2.2.2 Aldehydes 3-10
3.2.2.3 Other Organic Compounds 3-11
3.2.2.4 Volatility and Reactivity ". 3-11
3.2.3 Nitrogen Oxides 3-14
3.3 ATMOSPHERIC CHEMICAL PROCESSES: FORMATION AND TRANSFORMATION
OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS 3-15
3.3.1 Inorganic Reactions 3-16
3.3.1.1 Formation of Ozone: The NO-N02-03 Cycle .... 3-16
3.3.1.2 Formation of Radical Intermediates 3-17
3.3.1.3 Termination Reactions 3-21
3.3.1.4 Reactions Involving Nitrous Acid 3-22
3.3.1.5 Reactions Involving Nitric Acid and
Dinitrogen Pentoxide 3-23
3.3.2 Organic Reactions 3-24
3.3.2.1 Reactions with Hydroxyl Radicals 3-25
3.3.2.2 Reactions with Ozone 3-34
3.3.2.3 Reactions with Nitrate Radicals 3-41
3.3.3 Atmospheric Lifetimes of Organic Compounds 3-44
3.3.4 Atmospheric Reactions of Peroxyacetyl Nitrate 3-45
3.3.5 Role of Ozone in Aerosol Formation 3-47
3.3.5.1 Formation of Sulfate Aerosol 3-47
3.3.5.2 Formation of Nitrate Aerosol 3-48
3.3.5.3 Formation of Organic Aerosols 3-49
-------
TABLE.OF CONTENTS
(continued)
Page
3.3.6 Role of Ozone and Other Photochemical Oxidants in the
Acidification of Rain 3-50
3.3.6.1 Reactions of Ozone in Aqueous Droplets 3-50
3.3.6.2 Reactions of Hydrogen Peroxide in Aqueous
Droplets 3-51
3.3.6.3 Reactions of Formic Acid in Aqueous
Droplets 3-54
3.4 METEOROLOGICAL AND CLIMATOLQGICAL PROCESSES 3-54
3.4.1 Atmospheric Mixing 3-54
3.4.2 Wind Speed and Mixing *• 3-61
3.4.3 Effects of Sunlight and Temperature 3-66
3.4.4 Transport of Ozone and Other Oxidants and Their
Precursors 3-68
3.4.5 Surface Scavenging in Relation to Transport ...... 3-74
3.4.6 Stratospheric-Tropospheric Ozone Exchange 3-75
3.4.7 Stratospheric Ozone at Ground Level 3-80
3.4.8 Background Ozone from Photochemical Reactions 3-84
3.5 PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS 3-88
3.5.1 Sources and Emissions 3-88
3.5.1.1 Manmade Sources and Emissions 3-88
3.5.1.2 Natural Sources and Emissions 3-98
3.5.2 Representative Concentrations of Ozone Precursors
in Ambient Air 3-104
3.5.2.1 Concentrations of Nonmethane Organic
Compounds in Ambient Air 3-104
3.5.2.2 Concentrations of Nitrogen Oxides in
Ambient Air 3-108
3.6 SOURCE-RECEPTOR (OXIDANT-PRECURSOR) MODELS 3-110
3.6.1 Definitions, Descriptions, and Use 3-112
3.6.1.1 Statistical Models 3-112
3.6.1.2 Trajectory Models 3-113
3.6.1.3 Fixed-Grid Models 3-119
3.6.1.4 Box Model s 3-121
3.6.2 Validation and Sensitivity Analyses for Dynamic
Models 3-121
3.7 SUMMARY 3-124
3.7.1 Descriptions and Properties of Ozone and Other
Photochemical Oxidants 3-124
3.7.2 Nature of Precursors to Ozone and Other Photochemical
Oxidants 3-125
3.7.3 Atmospheric Reactions of Ozone and Other Oxidants
Including Their Role in Aerosol Formation 3-127
3.7.3.1 Formation and Transformation of Ozone
and Other Photochemical Oxidants 3-127
3.7.3.2 Atmospheric Chemical Processes Involving
Ozone .' ... 3-128
VI
-------
TABLE OF CONTENTS
(continued)
3.7.3.3 Atmospheric Reactions of PAN, H202, and
HCOOH 3-129
3.7.4 Meteorological and Climatological Processes 3-130
3.7.4.1 Atmospheric Mixing 3-131
3.7.4.2 Wind Speed and Direction 3-132
3.7.4.3 Effects of Sunlight and Temperature 3-133
3.7.4.4 Transport of Ozone and Other Oxidants
and Their Precursors 3-133
3.7.4.5 Stratospheric-Tropospheric Ozone Exchange ... 3-135
3.7.4.6 Stratospheric Ozone at Ground Level 3-135
3.7.4.7 Background Ozone from Photochemical
Reactions 3-136
3.7.5 Sources, Emissions, and Concentrations of Precursors
to Ozone and Other Photochemical Oxidants 3-137
3.7.5.1 Sources and Emissions of Precursors 3-137
3.7.5.2 Representative Concentrations in Ambient
Air 3-138
3.7.6 Source-Receptor (Oxidant-Precursor) Models 3-140
3.7.6.1 Trajectory Models 3-140
3.7.6.2 Fixed-Grid Models 3-141
3.7.6.3 Box Models 3-142
3.7.6.4 Validation and Sensitivity Analyses for
Dynamic Models 3-142
3.8 REFERENCES 3-142
4. SAMPLING AND MEASUREMENT OF OZONE AND OTHER PHOTOCHEMICAL
OXIDANTS AND THEIR PRECURSORS 4-1
4.1 INTRODUCTION 4-1
4.2 QUALITY ASSURANCE AND OTHER SAMPLING FACTORS IN
MONITORING FOR OZONE 4-2
4.2.1 Quality Assurance in Ambient Air Monitoring for
Ozone 4-2
4.2.2 Sampling Factors in Ambient Air Monitoring for
Ozone 4-3
4.2.2.1 Sampling Strategies and Air Monitoring
Needs : 4-4
4.2.2.2 Air Monitoring Site Selection 4-4
4.2.3 Measurement Methods for Total Oxidants and Ozone 4-6
4.2.3.1 Total Oxidants 4-6
4.2.3.2 Ozone 4-7
4.2.4 Generation and Calibration Methods for Ozone 4-13
4.2.4.1 Generation 4-13
4.2.4.2 Calibration. 4-14
4.2.5 Relationship between Methods for Total Oxidants
and Ozone 4-21
4.2.5.1 Predicted Relationship 4-22
Vll
-------
TABLE OF CONTENTS
(continued)
Page
4.2.5.2 Empirical Relationship Determined from
Simul taneous Measurements — 4-24
4.2.6 Methods for Sampling and Analysis of Peroxyacetyl
Nitrate and Its Homologues 4-33
4.2.6.1 Introduction 4-33
4.2.6.2 Analytical Methods for PAN 4-34
4.2.6.3 Generation and Calibration of PAN 4-39
4.2.6.4 Methods of Analysis of Higher Homologues ..... 4-42
4.2.7 Methods for Sampling and Analysis of Hydrogen
Peroxide 4-43
4.2.7.1 Introduction 4-43
4.2.7.2 Sampling 4-44
4.2.7.3 Measurement , 4-45
4.2.7.4 Generation and Calibration Methods 4-48
4.3 SAMPLING, MEASUREMENT, AND CALIBRATION METHODS FOR
PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS ..... 4-49
4.3.1 Nonmethane Organic Compounds 4-50
4.3.1.1 Nonmethane Hydrocarbons 4-50
4.3.1.2 Aldehydes , 4-61
4.3.1.3 Other Oxygenated Organic Species 4-66
4.3.2 Nitrogen Oxides .......... 4-66
4.3.2.1 Measurement Methods for N02 and NO 4-67
4.3.2.2 Sampling Requirements 4-70
4.3.2.3 Cal ibration 4-70
4.4 SUMMARY 4-71
4.4.1 Sampling and Measurement of Ozone and Other
Photochemical Oxidants 4-71
4.4,1.1 Quality Assurance and Sampling 4-72
4.4.1.2 Measurement Methods for Total Oxidants
and Ozone 4-72
4.4.1.3 Calibration Methods 4-74
4.4.1..4 Relationships of Total Oxidants and Ozone
Measurements 4-77
4.4.1.5 Methods for Sampling and Analysis of Peroxy-
acetyl Nitrate and Its Homologues 4-78
4.4.1.6 Methods for Sampling and Analysis of
Hydrogen Peroxide 4-80
4.4.2 Measurement of Precursors to Ozone and Other
Photochemical Oxidants 4-82
4.4.2.1 Nonmethane Organic Compounds 4-82
4.4.2.2 Nitrogen Oxides 4-84
4.5 REFERENCES 4-86
5. CONCENTRATIONS OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
IN AMBIENT AIR 5-1
5.1 INTRODUCTION 5-1
5.2 TRENDS IN NATIONWIDE OZONE CONCENTRATIONS 5-3
vi ii
-------
TABLE OF CONTENTS
(continued)
Page
5,3 OVERVIEW OF OZONE CONCENTRATIONS IN URBAN AREAS 5-7
5.4 OVERVIEW OF OZONE CONCENTRATIONS IN NONURBAN AREAS 5-16
5.4.1 National Air Pollution Background Network (NAPBN) 5-16
5.4.2 Sulfate Regional Experiment Sites (SURE) , 5-25
5.5 VARIATIONS IN OZONE CONCENTRATIONS: DATA FROM SELECTED
AND SITES 5-25
5.5.1 Temporal Variations in Ozone Concentrations 5-28
5.5.1.1 Diurnal Variations in Ozone Concentrations ... 5>-28
5.5.1.2 Seasonal Variations in Ozone Concentrations .. 5-42
5.5.1.3 Weekday-Weekend Variations in Ozone
Concentrations 5-48
5.5.2 Spatial Variations in Ozone Concentrations 5-49
5.5.2,1 Urban Versus Nonurban Variations 5-49
5.5.2.2 Intracity Variations 5-50
5.5.2.3 Indoor-Outdoor Concentration Ratios 5-55
5.5.2.4 Altitudinal and Latitudinal Variations 5-60
5.5.2.5 Vertical Gradients at Ground Level 5-69
5.6 CONCENTRATIONS OF PEROXYACETYL NITRATE AND PEROXYPROPIONYL
NITRATE IN AMBIENT AIR 5-71
5.6.1 Introduction 5-71
5.6.2 Historical Data 5-72
5.6.3 Ambient Air Concentrations of PAN and Its
Homo!ogues i n Urban Areas 5-* 74
5.6.4 Ambient Air Concentrations of PAN and Its
Homol ogues i n Nonurban Areas 5^80
5.6.5 Temporal Variations in Ambient Air
Concentrations of Peroxyacetyl Nitrate 5-83
5.6.5.1 Diurnal Patterns 5-83
5.6.5.2 Seasonal Patterns 5-89
5.6.6 Spatial Variations in Ambient Air Concentrations
of Peroxyacetyl Nitrate 5-89
5.6.6.1 Urban-Rural Gradients and Transport of PAN ... 5-89
5.6.6.2 Intracity Variations 5-93
5.6.6.3 Indoor-Outdoor Ratios of PAN Concentrations .. 5-93
5.7 CONCENTRATIONS OF OTHER PHOTOCHEMICAL OXIDANTS IN
AMBIENT AIR 5-93
5.8 SUMMARY 5-95
5.8.1 Ozone Concentrations in Urban Areas ..... ...... 5-98
5.8.2 Trends in Nationwide Ozone Concentrations ...,.., 5-101
5.8.3 Ozone Concentrations i n Nonurban Areas 5" 103
5.8.4 Diurnal and Seasonal Patterns in Ozone
Concentrati ons S-'IOS
5.8.5 Spatial Patterns in Ozone Concentrations 5-^107
5.8.5.1 Urban-Nonurban Differences in Ozone
Concentrati ons 5-107
5.8.5.2 Geographic, Vertical, and Altitudinal
Variations in Ozone Concentrations 5-109
ix
-------
TABLE OF CONTENTS
(continued)
Page
5.8.5.3 Other Spatial Variations in Ozone
Concentrations 5-112
5.8.6 Concentrations and Patterns of Other Photochemical
Oxidants 5-112
5.8.6.1 Concentrations 5-112
5.8.6.2 Patterns.. 5-114
5.8.7 Relationship Between Ozone and Other Photochemical
Oxi dants 5-114
5.9 REFERENCES 5-116
-------
LIST OF TABLES
Table Page
3-1 Physical properties of ozone , 3-3
3-2 Physical properties of peroxyacetyl nitrate 3-5
3-3 Infrared absorptivities of peroxyacetyl nitrate 3-6
3-4 Physical properties of hydrogen peroxide 3-6
3-5 Physical and chemical properties of nitric oxide and
nitrogen dioxide 3-15
3-6 Calculated lifetimes of selected organic compounds resulting
from atmospheric loss by reaction with 03 and OH and N03
radicals ". 3-44
3-7 Published episodes of transport of stratospheric ozone to
ground level ,.... 3-82
3-8 Emissions of VOC by decade, 1940 through 1980 3-90
3-9 Emissions of NO by decade, 1940 through 1980 3-93
3-10 Yearly quantities of motor vehicle fuels sold in the
United States for highway use, 1980 through 1983 3-97
3-11 Summary of NO emissions from mobile sources 3-98
3-12 Area-wide biogenic emission fluxes 3-101
3-13 Global estimates of nitrogen transformation 3-103
3-14 Nonmethane hydrocarbon concentrations measured between 6:00
and 9:00 a.m. in various United States cities 3-106
3-15 Nonmethane hydrocarbon concentrations measured in nonurban
atmospheres 3-107
3-16 Average 6:00 to 9:00 a.m. NO concentrations and HC/NO
ratios in urban areas 3-109
4-1 Performance specifications for automated methods of ozone
analysis 4-9
4-2 List of designated reference and equivalent methods of
ozone analysis — 4-10
4-3 Factors for intercomparison of data calibrated by UV
photometry versus KI colorimetry , 4-16
4-4 Response of NBKI reagent and Mast meter to various
oxidants 4-23
4-5 Comparison of corrected instrument readings to colorimetric
oxidant readings during atmospheric sampling 4-27
4-6 Summary of parameters used in determination of PAN by
GC-ECD 4-35
4-7 Infrared absorptivities of peroxyacetyl nitrate
(Base 10) 4-38
4-8 Measurement methods for hydrogen peroxide 4-46
4-9 Percentage difference from known concentrations of
nonmethane hydrocarbons obtained by sixteen users 4-52
4-10 Problems associated with gathering NMOC data with
automated analyzers and recommendations for reducing
these effects 4-53
4-11 Summary of advantages and disadvantages of primary
collection media for NMOC analysis 4-58
XI
-------
LIST OF TABLES
(continued)
Table
4-12 GC/continuous NMOC analyzer comparisons, least-squares
regressions 4-62
4-13 Summary of ozone monitoring techniques ... 4-73
4-14 Ozone calibration techniques 4-76
5-1 Second-highest 1-hr ozone concentrations reported for
Standard Metropolitan Statistical Areas having
populations > 0.5 million, 1981 through 1983 5-13
5-2 Annual ozone summary statistics for three sites of the
National Air Pollution Background Network 5-18
5-3 Concentrations of ozone during 6-day period of high
values at NAPBN site in Mark Twain National Forest,
Missouri, 1979 5-20
5-4 Percent!* le distributions of ozone concentrations at
sites of National Air Pollution Background Network,
aggregated by quarter across several years 5-22
5-5 Summary of ozone concentrations measured at Sulfate
Regional Experiment (SURE) nonurban stations, August
through December 1977 5-27
5-6 Number of consecutive-day exposures or respites when ,the
daily 1-hr maximum ozone concentration was >_ 0.06 ppm,
in four cities (April through September, 1979 through 1981) 5-38
5-7 Number of consecutive-day exposures or respites when the
daily 1-hr maximum ozone concentration was >_ 0.12 ppm, in
four cities (April through September, 1979 through 1981) 5-39
5-8 Number of consecutive-day exposures or respites when the
daily 1-hr maximum ozone concentration was > 0.18 ppm, in
four cities (April through September, 1979 through 1981) 5-40
5-9 Number of consecutive-day exposures or respites when the
daily 1-hr maximum ozone concentration was >_ 0.24 ppm, in
four cities (April through September, 1979 through 1981) 5-41
5-10 Ozone concentrations at sites in and around New Haven,
Connecticut, 1976 , 5-52
5-11 Quarterly maximum 1-hour ozone values at sites in and
around New Haven, Connecticut, 1976 ............................ 5-52
5-12 Peak ozone concentrations at eight sites in New York
City and adjacent Nassau County, 1980 5-53
5-13 Summary of reported indoor-outdoor ozone ratios 5-57
5-14 Comparison of ozone concentrations at three different
elevations, High Point Mountain, NJ, and at Bayonne, NJ,
July 1975 5-68
5-15 Means and standard errors of ozone concentrations measured
over 4 years at two sampling heights at three stations in
the rural, upper-midwestern United States 5-70
5-16 Summary of concentrations of peroxyacetyl nitrate in
ambient air in urban areas of the United States 5-75
XII
-------
LIST OF TABLES
(continued)
Tab! e Paige
5-17 Relationship of ozone and peroxyacetyl nitrate at urban
and suburban sites in the United States 5-79
5-18 Ambient air measurements of peroxypropionyl nitrate (PPN)
concentrations by electron capture gas chromatography
at urban sites in the United States 5-81
5-19 Concentrations of peroxyacetyl and peroxypropionyl nitrates
in Los Angeles, Oakland, and Phoenix, 1979 5-82
5-20 Concentrations in ambient air of peroxyacetyl and
peroxypropionyl nitrates and ozone at nonurban remote sites
in the United States 5-84
5-21 PAN and ozone concentrations in ambient air, New
Brunswick, N.J., for September 25, 1978, to May 10, 1980 5-91
5-22 Intracity variations in peak ozone and PAN concentrations
in Houston, October 26 and 27, 1977 5-94
5-23 Concentrations of hydrogen peroxide in ambient air at
urban and nonurban sites 5-97
5-24 Second-highest ozone concentrations among daily maximum
1-hr values in 1983 in Standard Metropolitan Statistical
Areas with populations >^ 1 million, given by census divisions
and regions 5-99
xm
-------
LIST OF FIGURES
F1gure . Page
3-1 Experimental time-concentration profiles for propene, NO,
N02s OB, HCHO, and PAH for an irradiated NO -propene-air
mixture '..... 3-18 .
3-2 Reaction scheme for OH radical-initiated oxidation of propene
in the presence of NO 3-28
3-3 Rate of aqueous-phase oxidation of S(IV) by 03 (30 ppb) and
H202 (1 ppb), as a function of solution pH 3-52
3-4 Isopleths (m x 102) of mean summer morning mixing heights, AGL .. 3-59
3-5 Isopleths (m x 1Q2) of mean summer afternoon mixing
heights, AGL 3-59
3-6 Percentage of summer 2315 GMT (6:15 p.m. EST, 3:15 p.m. PST)
soundings with an elevated inversion base between 1 and
500 m above ground 1 eve! 3-60
3-7 Mean resultant surface wind pattern for the United States
for July.j Direction and length of arrows indicate monthly
resultant wind 3-62
3-8 Percentage of summer 1115 GMT (6:15 a.m. EST, 3:15 a.m. PST)
soundings with an inversion base at the surface and wind
speeds at the surface £2.5 m/sec "...'.. 3-64
3-9 Isopleths (m/sec) of mean summer wind speed averaged
through the morning mixing layer ..... 3-65
3-10 Isopleths (m/sec) of mean summer wind speed averaged
through the afternoon mixing layer ............. 3-65
--3-11 Schematic cross section, looking downwind along the jet
stream, of a tropopause folding event as modeled by
Daniel sen 3-76
3-12 Measured vertical cross sections of (A) 03, (B) dewpoint, and
(C) the 500 mb chart and the flight track for October 5, 1978 ... 3-78
3-13 Hypothesized models of the process that mixes tropopause
folding events into the troposphere 3-79
3-14 National trend in estimated emissions of volatile organic
compounds, 1970 through 1983 3-91
3-15 Comparative trends in highway vehicle emissions of nitrogen
oxides (NO ) and volatile organic compounds (VOC) versus
vehicle mites traveled, 1970-1983 3-92
3-16 National trend,in estimated emissions of nitrogen oxides,
1970 through 1983 . 3-94
3-17 Schematics of the. three types, of dynamic models 3-114
- 3-18 Example of EKMA diagram for high-oxidant urban area 3-116
4-1 Comparison of ozone and total oxidant concentrations in the
Pasadena area, August 1955 4-26
v 4-2 Comparison of ozone and total oxidant concentrations in the
Los Angeles area, August 1955 4-26
4U3 Measurements of ozone and total oxidants in Los Angeles,
September 4 through September 30, 1971 4-29
4-4 Measurements of ozone and total oxidants in St. Louis,
October 14 through December 21, 1971 ;. .... 4-30
xiv
-------
LIST OF FIGURES
(continued) .
Figure Page
4-5 Measurements of ozone and total oxidants, Houston Ship
Channel, August 11, 1973 4-32
5-1 National trend in composite average of the second-highest
value among daily maximum 1-hour concentrations at
selected groups of sites, 1975 through 1983 ... : 5-4
'5-2 Comparison of the 1979-1980, 1981-1982, and 1983
composite average of the second-highest daily maximum
1-hour ozone concentrations across EPA regions 5r8
5-3 Average daylight (6:00 a.m. to 8:00 p.m.) concentrations
of ozone in the second and third quarters (April through
September), 1981 5-9
5-4 Average daylight (6:00 a.m. to 8:00 p.m.) concentrations
of ozone in the first and fourth quarters '(January through
March and October through December), 1981 .. 5-10
5-5 Distributions of the three highest 1-hour ozone
concentrations at'valid sites (906 station-years)
aggregated for 3 years (1979, 1980, and 1981) and
the highest ozone concentrations at NAPBN sites aggre-
gated for those years (24 station-years) ......... ... 5-12
5-6 Locations of the eight national forest (NF) stations
constituting the National Air Pollution Background Network
(NAPBN) ...... .'...;.,....... ,.-. .• 5-17
5-7 Trajectory analysis plots for the NAPBN site at Mark Twain
National Forest, .MQ, July 21, 1979 5-21
5-8 (A) Second-highest value among maximum 1-hr ozone
concentrations at five NAPBN monitoring stations, 1979
through 1983. (B) Composite averages of the second-highest
value among daily maximum 1-hr ozone concentrations at
five NAPBN stations, 1979 through 1983 .......:....; 5-24
5-9 Location, of Sulfate Regional Experiment (SURE) monitoring
stations .'..... ..-..' 5-26
5-10 Diurnal pattern of 1-hr ozone concentrations on July 13,
1979, Philadelphia, PA 5-29
5-11 Diurnal patterns of 1-hour ozone concentrations, September 20
and 21, 1980, Detroit, MI 5-30
5-12 Diurnal and 1-month composite diurnal variations in ozone
concentrations, Washington, DC, July 1981 ..., 5-31
5-13 Diurnal and 1-month composite diurnal variations in ozone
concentrations, St: Louis County, MO, September 1981 5-31
5-14 Diurnal and 1-month composite diurnal variations in ozone
concentrations, Alton, IL, October 1981 (fourth quarter) ... . 5-32
5-15 Composite diurnal patterns by quarter of ozone
concentrations, Alton, IL", 1981 .. ....' ...... ... 5-33
5-16 Three-day sequence of hourly ozone concentrations at
Montague, MA, SURE station'showing locally generated
midday peaks and transported late, peaks •; 5-36
xv
-------
LIST OF FIGURES
(continued)
Figure
5-17 Quarterly composite diurnal patterns of ozone concentrations
at selected sites representing potential for exposure of
major crops, 1981 , 5-43
5-18 Composite diurnal ozone pattern at a rural NCLAN site in
Argonne, IL, August 6 through September 30, 1980 5-45
5-19 Daily 7-hour and 24-hour average ozone concentrations at
a rural NCLAN site in Argonne, IL, 1980 5-45
5-20 Seasonal variations in ozone concentrations as indicated
by monthly averages and the 1-hour maximum in each month
at selected sites, 1981 5-46
5-21 New York State air monitoring sites for Northeast Corridor
Monitoring Program (NECRMP) . . 5-54
5-22 Altitudinal sequence of monitoring sites in the San Bernardino
Mountai ns 5-64
5-23 Relationship between elevation and diurnal patterns of total
oxidant concentrations, temperature, and vapor pressure at
four sites (A-D) in the San Bernardino Mountains, CA, July-
August 1969 5-66
5-24 Comparison of monthly daylight average and maximum PAN
concentrations at Riverside, CA, for 1967-1968 and 1980 5-78
5-25 Variation of mean 1-hour oxidant and PAN concentrations,
by hour of day, in downtown Los Angeles, 1965 5-86
5-26 Variation of mean 1-hour oxidant and PAN concentrations, by
hour of day, Air Pollution Research Center, Riverside CA,
September 1966 5-87
5-27 Diurnal profiles of ozone and PAN at Claremont, CA,
October 12 and 13, 1978, 2 days of a multi-day smog episode 5-88
5-28 Monthly variation of oxidant (Mast meter, continuous 24-hr)
concentrations and PAN (GC-ECD, sequential, 6:00 a.m. to
4:00-5:00 p.m.) concentrations, Air Pollution Research
Center, Riverside, CA, June 1966-June 1967 5-90
5-29 Average daily profile by month (July 7-October 10) for
PAN and ozone in New Brunswick, NJ, 1979 5-92
5-30 Diurnal profile of HCOOH, along with other oxidants and
smog constituents, on October 12 and 13, 1978, at
Claremont, CA 5-96
5-31 National trend in composite average of the second highest
value among daily maximum 1-hour ozone concentrations at
selected groups of sites, 1975 through 1983 5-102
5-32 Distributions of the three highest 1-hour ozone
concentrations at valid sites (906 station-years) aggregated
for 3 years (1979, 1980, and 1981) and the highest ozone
concentrations at NAPBN sites aggregated for those years
(24 station-years) 5-104
xvi
-------
LIST OF ABBREVIATIONS AND SYMBOLS
~ approximately
A wavelength
APR air:fuel ratio
APHA American Public Health Association
aq aqueous
AGL above ground level
ASL above sea level
atm atmosphere
avg average
b,p. boiling point
bz benzene
C carbon
°C degrees Celsius
CA chromotropic acid
CAMP Continuous Air Monitoring Program
CARB California Air Resources Board
cc cubic centimeter
CH4 methane
CO carbon monoxide
C02 carbon dioxide
cm centimeter
concn concentration
DBH tree diameter at breast height
DNPH 2,4-dinitrophenylhydrazine
DOT Department of Transportation
EA normal electrode potential
ECD electron-capture detector
EKMA Empirical Kinetic Modeling Approach
EPA U.S. Environmental Protection Agency
EST eastern standard time
FID flame ionization detector
FRM Federal Reference Method
ft foot
XVI1
-------
LIST OF ABBREVIATIONS AND SYMBOLS
(continued)
FTIR
9
g/mi
GC
GMT
GPT
hr
hv
HC
HCN
HCOOH
HFET
Hg
H202
H02
MONO
HON02
HPLC
HPPA
HRP
H20
H2S04
in
IR
k
KI
km
L
LAAPCD
LCV
In
LSI
Fourier-transform infrared
gram(s)
grams per mile
gas chromatography
Greenwich mean time
gas-phase titration
hour(s)
photon
hydrocarbons
hydrogen cyanide
formic acid
Highway Fuel Economy Driving Schedule
mercury
hydrogen peroxide
hydroperoxy
nitrous acid
nitric acid
high-pressure liquid chromatography; also,
high-performance liquid chromatography
3-(p_-hydroxyphenyl)propionic acid
horseradish peroxidase
water
sulfuric acid
inch(es)
infrared
constant
potassium iodide
kilometer
liter(s)
Los Angeles Air Pollution Control District
leuco crystal violet
natural logarithm (base e)
local standard time
xv 1.11
-------
LIST OF ABBREVIATIONS AND SYMBOLS
(continued)
M
tn
mb
MBTH
rag
mg/m3
MGE
min
ml
mm
mM
MMC
m.p.
mph
MS
MSL
MTBE
NA
NAAQS
NADB
NAMS
NAPBN
NAS
NBS
NECRMP
NEDS
NEROS
NH3
NH4N03
NF
nm
NMHC
NMOC
molar
meter(s)
mi llibar(s)
3-methyl-2-benzothiazolinone hydrazone
milligram(s)
milligrams per cubic meter
modified graphite electrode
minute(s)
milliliter(s)
millimeter(s)
millimolar
mean meridional circulation
melting point
miles per hour
mass spectrometry
mean sea level
methyl tertiary butyl ether
not available
National Ambient Air Quality Standard
National Aerometric Data Bank
National Aerometric Monitoring Stations
National Air Pollution Background Network
National Academy of Sciences
National Bureau of Standards
Northeast Corridor Regional Modeling Project
National Emissions Data System
Northeast Regional Oxidant Study
ammonia
ammonium nitrate
National Forest
nanometer(s)
nonmethane hydrocarbons
nonmethane organic compounds
xix
-------
LIST OF ABBREVIATIONS AND SYMBOLS
(continued)
NO
NOX
N02
N03
N20
NR
NYCC
02
03
PAN
PBzN
PNA
PPN
ppb
ppm
ppt
PSD
psig
PST
PUFA
RAPS
RTI
S.D.
SAROAD
SBR
SCAB
sec
SLAMS
SMSA
SRM
SSET
STA
STP
nitric oxide
nitrogen oxides
nitrogen dioxide
nitrogen trioxide
nitrous oxide
natural rubber
New York City Driving Schedule
oxygen
ozone
peroxyacetyl nitrate
peroxybenzoyl nitrate
peroxynitric acid
peroxypropionyl nitrate
parts per billion
parts per million
parts per trillion
Prevention of Significant Deterioration
pounds per square inch gauge
Pacific standard time
polyunsaturated fatty acids
Regional Air Pollution Study
Research Triangle Institute
standard deviation
Storage and Retrieval of Aerometric Data
styrene-butadiene rubber
South Coast Air Basin
second(s)
State and Local Air Monitoring Stations
Standard Metropolitan Statistical Area
Standard Reference Material
small-scale eddy transport
seasonal tropopause adjustment
standard temperature and pressure
xx
-------
LIST OF ABBREVIATIONS AND SYMBOLS
(continued)
SURE
TEL
Tenax GC
TF
Tg/yr
THC
TML
TNMHC
TWC
Mg/m3
MM
U
UHAC
U.S.
UV
V
v/v
VHAC
VOC
vol %
w/w
WCOT
XAD-2
XO
Sulfate Regional Experiment Sites
tetraethyl lead
adsorbent used in NMOC analysis
tropopause-folding events
teragrams per year
total hydrocarbon
tetramethyl lead
total nonmethane hydrocarbons
three-way catalyst
microgram(s) per cubic meter
micromolar
uranium
uranium hydroxamic acid chelates
United States
ultraviolet
vanadium
volume-volume
vanadium hydroxamic acid chelates
volatile organic compounds
volume percent
weight-weight
wall-coated open tubular (column)
absorbent used in NMOC analysis
xylenol orange
year(s)
xxi
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Chapter 3:
Pri nci pal Authors
*Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Mr. Elmer Robinson
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI 96720
*Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Halvor Westberg
Director, Laboratory for Atmospheric Research, and
Professor, Civil and Environmental Engineering
Washington State University
Pullman, WA 99164-2730
*Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
Contributing Authors
*Dr. A. Paul Altshuller
Atmospheric Sciences Research Laboratory
MD-59
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Basil Dimitriades
Atmospheric Sciences Research Laboratory
MD-59
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxn
-------
Chapter 3 Contributing Authors.(cont'd.)
*Dr. Marcia C. Dodge
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. James M. Kawecki
TRC Environmental Consultants, Inc. .
2001 Wisconsin Avenue, N.W.
Suite 261
Washington, DC 20007
*Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
The following people reviewed Chapter 3 at the request of EPA:
Dr. Joseph J. Bufalini
Atmospheric Sciences Research Laboratory
MD-84 •
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Bruce Gay
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Eric Ginsburg
Office of Air Quality Planning and Standards
Control Programs Development Division '
MD-15
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxi 11
-------
Chapter 3 Reviewers (cont'd.)
Mr. Gerald Gipson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Robert Hall
Industrial Environmental Research Laboratory
MD-65
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Michael R. Kuhlman
BatteH e, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. William A. Lonneman
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Chuck Mann
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. E. L. Martinez
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Edwin L. Meyer
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Johnnie Pearson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxiv
-------
Chapter 3 Reviewers (cont'd.)
Mr. Kenneth L. Schere
Atmospheric Sciences Research Laboratory
MD-80
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Stanley Sleva
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Dlvslon
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Chester W. Spicer
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. Bruce Tichenor
Industrial Environmental Research Laboratory
MD-54
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
^Authors also reviewed portions of this chapter.
xxv
-------
Chapter 4: Measurement of Ozone and Other Photochemical Oxidants and Their
Precursors
Principal Authors
*Dr. Jimmie A. Hodgeson
U.S. Environmental Protection Agency
Environmental Monitoring Systems Laboratory
26 W. St. Clair
Cincinnati, OH 45268)
*Mr. Michael W. Holdren •
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. M. Rene Surgi
Department of Chemistry
Louisiana State University
Baton Rouge, LA 70803
ContributingAuthors
Dr. Sandor Freedman
Piedmont Technical Services
Hillsborough, NC 27278
*Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC -27711
The following peoplereviewed Chapter 4 at the requestof EPA:
Dr. A. Paul Altshuller
Atmospheric Sciences Research Laboratory
MD-59
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Joseph J. Bufalini
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxvi
-------
Chapter 4 Reviewers (cont'd):
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Bruce Gay
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. William A. Lonneman
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Kenneth Rehme
Environmental Monitoring Systems Laboratory
MD-77
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Harold G. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Stanley Sleva
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Chester W. Spicer
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
*Authors also reviewed portions of this chapter.
xxv i i
-------
Chapter 5; Concentrations of Ozone and Other Photochemical Oxidants in Ambient
Air
Principal Authors
*Mr. Thomas B. McMullen
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
*Mr. Elmer Robinson
Director, Mauna Loa Observatory
National Oceanic and Atmospheric Administration (NOAA/CMCC)
Hilo, HI 96720
*Ms. Beverly E. Tilton
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Contributing Author
Dr. Sandor Freedman
Piedmont Technical Services
Hillsborough, NC 27278
The following people reviewed Chapter 5 at the request of EPA:
Mr. Gerald Akland
Environmental Monitoring Systems Laboratory
MD-56
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Joseph J. Bufalini
Environmental Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Gary Evans
Environmental Monitoring Systems Laboratory
MD-56
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxvi
-------
Chapter 5 Reviewers (cont'd):
Dr. Donald Fox
Associate Professor
Department of Environmental Science and Engineering
School of Public Health, 201H
University of North Carolina
Chapel Hill, NC 27514
Mr. Gerald Gipson
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Jimmie A. Hodgeson
Professor, Department of Chemistry
407 Choppin Hall
Louisiana State University
Baton Rouge, LA 70803
Mr. Michael W. Holdren
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Michael R. Kuhlman
Battelle, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Mr. William A. Lonneman
Atmospheric Sciences Research Laboratory
MD-84
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Thomas McCurdy
Office of Air Quality Planning and Standards
Strategies and Air Standards Division
MD-12
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Edwin L. Meyer
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
xxix
-------
Chapter 5 Reviewers (cont'd):
Dr. Harold 6. Richter
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
MD-14
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Chester W. Spicer
BatteHe, Columbus Laboratories
505 King Avenue
Columbus, OH 43201
Dr. Arthur M. Winer
Assistant Director
Statewide Air Pollution Research Center
University of California
Riverside, CA 92521
^Authors also reviewed portions of this chapter.
xxx
-------
SCIENCE ADVISORY BOARD
CLEAN AIR SCIENTIFIC ADVISORY COMMITTEE
The substance of this document was reviewed by the Clean Air Scientific
Advisory Committee of the Science Advisory Board Jn public sessions.
SUBCOMMITTEE ON OZONE
Chairman
Dr. Morton Lippmann
Professor
Department of Environmental Medicine
New York University Medical Center
Tuxedo, New York 10987
Members
Dr. Mary 0. Amdur
Senior Research Scientist
Energy Laboratory
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
Dr. Eileen G. Brennan
Professor
Department of Plant Pathology
Martin Hall, Room 213, Liptnan Drive
Cook College-NJAES
Rutgers University
New Brunswick, New Jersey 08903
Dr. Edward D. Crandall
Professor of Medicine
School of Medicine
Cornell University
New York, New York 10021
Dr. James D. Crapo
Associate Professor of Medicine
Chief, Division of Allergy, Critical
Care and Respiratory Medicine
Duke University Medical Center
Durham, North Carolina 27710
Dr. Robert Frank
Professor of Environmental Health
Sciences
Johns Hopkins School of Hygiene
and Public Health
615 N. Wolfe Street
Baltimore, Maryland 21205
Professor A. Myrick Freeman II
Department of Economics
Bowdoin College
Brunswick, Maine 04011
Dr. Ronald J. Hall
Senior Research Scientist and Leader
Aquatic and Terrestrial Ecosystems
Section
Ontario Ministry of the Environment
Dorset Research Center
Dorset, Ontario
Canada POA1EO
Dr. Jay S. Jacobson
Plant Physiologist
Boyce Thompson Institute
Tower Road
Ithaca, New York 14853
xxxi
-------
Dr. Warren B. Johnson
Director, Atmospheric Science Center
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
Dr. Jane Q. Koenig
Research Associate Professor
Department of Environmental Health /
University of Washington
Seattle, Washington 98195
Dr. Paul Kotin
Adjunct Professor of Pathology
University of Colorado Medical School
4505 S. Yosemite, #339
Denver, Colorado 80237
Dr. Timothy Larson
Associate Professor
Environmental Engineering and
Science Program
Department of Civil Engineering
University of Washington
Seattle, Washington 98195
Professor M. Granger Morgan
Head, Department of Engineering
and Public Policy
Carnegie-Mellon University
Pittsburgh, Pennsylvania 15253
Dr. D. Warner North
Principal
Decision Focus Inc., Los Altos
Office Center, Suite 200
4984 El Garni no Real
Los Altos, California 94022
Dr. Robert D. Rowe
Vice President, Environmental and
Resource Economics
Energy and Resources Consultants, Inc.
207 Canyon Boulevard
Boulder, Colorado 80302
Dr. George Taylor
Environmental Sciences Division
P.O. Box X
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37831
Dr. Michael Treshow
Professor
Department of Biology
University of Utah
Salt Lake City, Utah 84112
Dr. Mark J. Utell
Co-Director, Pulmonary Disease Unit
Associate Professor of Medicine and
Toxicology in Radiation Biology
and Biophysics
University of Rochester Medical
Center
Rochester, New York 14642
Dr. James H. Ware
Associate Professor
Harvard School of Public Health
Department of Biostatisties
677 Huntington Avenue
Boston, Massachusetts 02115
Dr. Jerry Wesolowski
Air and Industrial Hygiene Laboratory
California Department of Health
2151 Berkeley Way
Berkeley, California 94704
Dr. James L. Whittenberger
Director, University of California
Southern Occupational Health Center
Professor and Chair, Department of
Community and Environmental Medicine
California College of Medicine
University of California - Irvine
19772 MacArthur Boulevard
Irvine, California 92717
Dr. George T. Wolff
Senior Staff Research Scientist
General Motors Research Labs
Environmental Science Department
Warren, Michigan 48090
xxxi i
-------
PROJECT TEAM FOR DEVELOPMENT
OF
Air Quality Criteria for Ozone and Other Photochemical Oxldants
Ms. Beverly E. Tilton, Project Manager
and Coordinator for Chapters 1 through 5, Volumes I and II
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. Norman E. Chi Ids
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. J.H.B. Garner
Coordinator for Chapters 7 and 8, Volume III
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr; Thomas B. McMullen
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Mr. James A. Raub
Coordinator for Chapters 9 through 12, Volumes IV and V
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. David T. Tingey
Coordinator for Chapter 6, Volume III
Environmental Research Laboratory
U.S. Environmental Protection Agency
200 S.W. 35th Street
Corvallis, OR 97330
XXXI11
-------
-------
2. INTRODUCTION
2.1 PURPOSE AND LEGISLATIVE BASIS OF THIS DOCUMENT
The Clean Air Act specifies that the Administrator of the United States
Environmental Protection Agency (EPA) issue, and revise on a periodic basis,
air quality criteria for certain air pollutants. Air quality criteria may be
defined as qualitative and quantitative information that describes the effects
of a pollutant on public health and welfare in terms of the respective exposures
that elicited them. According to section 108 of the Clean Air Act, as amended
in 1977, criteria shall
...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 such pollutant in the
ambient air, in varying quantities.
(U.S.Code, 1982)
Air quality criteria provide the Agency with a scientific basis for deciding
whether regulations controlling given pollutants are necessary and for deriving
such ambient air quality standards as may be needed.
Among the air pollutants designated by the Administrator as criteria
pollutants are those known as photochemical oxidants. This document is a
revision of Air Quality Criteria for Ozone and Other Photochemical Oxidants
(U.S. Environmental Protection Agency, 1978a). Its purpose is to review and
evaluate the scientific literature on ozone and related oxidants and to document
their effects on public health and welfare.
The term "photochemical oxidants" has historically been defined as those
atmospheric pollutants that are products of photochemical reactions and that
are capable of oxidizing neutral iodide ions (U.S. Environmental Protection
Agency, 1978a). Research has established that photochemical oxidants in
ambient air consist mainly of ozone, peroxyacetyl nitrate, and nitrogen dioxide,
and of considerably lesser amounts of other peroxyacyl nitrates, hydrogen
peroxide, alkyl hydroperoxides, nitric and nitrous acids, and formic acid.
Other oxidants suspected to occur in ambient air but only in trace amounts
include peracids and ozonides.
2-1
-------
Although it is by definition a photochemical oxidant, nitrogen dioxide is
not included among the oxidants discussed in this document. The formation of
nitrogen dioxide clearly precedes the formation of ozone and other related
oxidants in the ambient air. Thus, nitrogen dioxide is the dominant oxidant
early in the day, while ozone and other related oxidants predominate from late
morning or midday through much of the afternoon. Nitrogen dioxide is known to
exert deleterious effects on human health and welfare. This fact, coupled
with temporal and spatial variations in concentrations that differ from those
of ozone and related oxidants, underlies the listing of nitrogen dioxide under
section 108(a)(l) of the Clean Air Act as a criteria pollutant separate from
«
ozone and other photochemical oxidants. The second criteria document prepared
by EPA on the oxides of nitrogen was completed in 1982 (thS. Environmental
Protection Agency, 1982a). That document discussed nitric and nitrous oxides,
nitrogen dioxide, nitric and nitrous acids, and nitrosamines. As used in this
document, the term "photochemical oxidants" refers to ozone, the peroxyacyl
nitrates, hydrogen peroxide, and formic acid. The oxides of nitrogen are dis-
cussed, but only in the context of their role as precursors to ozone and
related oxidants.
2.2 THE OXIDANT PROBLEM
Ozone (Og), a reactive allotrope of oxygen (02), occurs as a natural
component of the atmosphere. It is found in its highest concentrations in the
stratosphere, where it is formed through cyclic reactions resulting from the
photolysis of oxygen into atomic oxygen and the subsequent reaction of atomic
oxygen with other oxygen molecules.
Incursions of stratospherically produced ozone into the lower troposphere
occur through meteorological and atmospheric exchange phenomena, resulting in
a global background of ozone. To this global background of ozone of stratospheric
origin are added ozone formed in the free troposphere and the contributions of
ozone produced in the ambient air from photochemical reactions involving
manmade emissions and natural products (e.g., natural emissions of volatile
organic compounds). Manmade emissions of nitrogen oxides and volatile organic
compounds are the chief contributors to the ozone burden found in the ambient
air of urban areas. The presence of ozone in ambient air is the net result of
various formation, stratospheric-tropospheric exchange, transport, and de-
struction processes.
2-2
-------
Other photochemical oxidants occur in ambient air, but the nature and
source of their global backgrounds are not well established. The additional
photochemical oxidants of concern in this document, namely hydrogen peroxide,
peroxyacetyl nitrate and its homologues, and formic acid, have, except for
the latter, been detected in remote environments thought to be free of manmade
influences; and all have been detected in the ambient air of urban areas.
The toxicity of ozone is well known. It is a strong oxidizing agent that
is highly reactive with a wide spectrum of chemical moieties. Since ozone is
a gas, studies of its health-related toxicity have centered largely on its
capacity for affecting pulmonary function and the morphology of the respiratory
tract, which is now well documented. In addition, its effects on extrapulmo-
nary tissues and systems are also of concern and are the subject of some of
the research discussed in this document. Studies of toxic effects of ozone on
vegetation are also well documented and have focused on foliar injury and
reduction in growth and yield. The toxicities of the peroxyacyl nitrates, of
hydrogen peroxide, and of formic acid are less well documented than the toxicity
of ozone, having been the focus of considerably less research because the
levels at which these oxidants occur in the ambient air, even in urban areas,
appear to warrant much less concern.
Ozone, but not the other oxidants mentioned above, is regulated under
provisions and procedures spelled out in the Clean Air Act. Its concentra-
tions in ambient air are controlled through the promulgation and attainment of
primary and secondary national ambient air quality standards (NAAQS). As
described in the Clean Air Act, criteria pollutants are those atmospheric
pollutants that are ubiquitous and are emitted into the air from numerous and
diverse sources. While widespread, ozone and the other photochemical oxidants
found in ambient air are not emitted into the air as primary pollutants.
Rather, they are formed as secondary pollutants in the atmosphere from ubiqui-
tous primary organic and inorganic precursors that are emitted by a multi-
plicity of sources. Consequently, photochemical oxidant pollution in this
country is the result of a combination of many factors, such as local meteoro-
logical conditions and the concentrations, composition, and patterns of occur-
rence of the primary pollutants that give rise to the oxidants.
2-3
-------
2.3 SCOPE AND ORGANIZATION OF THIS DOCUMENT
The atmosphere does not easily lend itself to the partitioning required
for documentation. Nevertheless, certain boundaries are logical for purposes
of discussion as well as for purposes of regulatory decisions. Ozone and its
organic precursors are known* to give rise to secondary organic aerosols.
Likewise, ozone and hydrogen peroxide both appear to participate in the atmos-
pheric oxidation of nitrogen dioxide (N02) and sulfur dioxide (S02) to those
inorganic aerosols leading to visibility degradation in the atmosphere and to
acidic deposition. The contributions of ozone and hydrogen peroxide to the
oxidation of NO, and SQ2 cannot be quantified at present, but are known to be
minor compared to the oxidation of these compounds by hydroxyl radicals. In
addition, ozone, the principal photochemical oxidant in ambient air, has no
direct effects on visibility since, unlike N02 and S02, it does not absorb
energy in the visible region of the spectrum. Thus, this document includes
brief discussions, in Chapter 3, of the atmospheric chemistry of ozone and
hydrogen peroxide relative to the formation of inorganic nitrogen and sulfur
aerosols but does not include information on the actual effects associated
with visibility degradation or acidic deposition. Since N02 and S02 are the
immediate, direct precursors to the aerosol species involved, visibility
degradation and acidic deposition are discussed in the respective air quality
criteria documents on oxides of nitrogen and on particulate matter and sulfur
oxides (U.S. Environmental Protection Agency, 1982a,b).
This document has been divided into five volumes for ease of review,
printing, and distribution. Volume I consists of the summary and conclusions
for the entire document. Volume II contains the introduction to the document
(Chapter 2) and all chapters dealing with the formation, transport, and fate
of photochemical oxidants (Chapter 3); the measurement of oxidants and their
precursors (Chapter 4); and the concentrations of oxidants in ambient air
(Chapter 5). Volume III contains the documentation of the effects of photo-
chemical oxidants on vegetation, ecosystems, and nonbiological materials
(Chapters 6, 7, and 8, respectively). Volume IV reviews the extensive body of
data available on the toxicological effects of ozone and other oxidants in
experimental animals and on in vitro effects on human cells and body fluids
(Chapter 9). In Volume V, effects observed in human controlled exposures
(Chapter 10) and in field and epidemiological studies (Chapter 11) are pre-
sented. In addition, th'at volume contains an evaluation of the health effects
data of probable consequence for regulatory purposes (Chapter 12).
2-4
-------
Neither control techniques nor control strategies for the abatement of
photochemical oxidants are discussed in this document, although some of the
topics included are relevant to abatement strategies. Technology for control-
ling the emissions of nitrogen oxides and of volatile organic compounds is
discussed in documents issued by the Office of Air Quality Planning and Stand-
ards (OAQPS) of the U.S. Environmental Protection Agency (e.g., U.S. Environmen-
tal Protection Agency, 1978b, 1983). Likewise, issues germane to the scientific
basis for control strategies, but not pertinent to the development of cri-
teria, are addressed in numerous documents issued by OAQPS.
In addition, certain issues of direct relevance to standard-setting are
not explicitly addressed in this document, but are addressed instead in docu-
mentation prepared by OAQPS as part of its regulatory analyses. Such analyses
include: (1) discussion of what constitutes an "adverse effect," that is, the
effect or effects the standard is intended to protect against; (2) assessment
of risk; and (3) discussion of factors to be considered in providing an ade-
quate margin of safety. While scientific data contribute significantly to
decisions regarding these three issues, their resolution cannot be achieved
solely on the basis of experimentally acquired information. Final decisions
on items (1) and (3) are made by the Administrator.
A fourth issue directly pertinent to standard-setting is identification
of the population at risk, which is basically a selection by the Agency of the
population to be protected by the promulgation of a given standard. This
issue is addressed only partially in this document. For example, information
is presented in Chapter 12 on factors, such as pre-existing disease, that
biologically may predispose individuals and subpopulations to adverse effects
from exposures to ozone. The identification of a population at risk, however,
requires information above and beyond data on biological predisposition, such
as information on levels of exposure, activity patterns, and personal habits.
Such information is included in a staff paper developed by OAQPS. Thus, the
identification of the population at risk relative to standard-setting is the
purview of OAQPS and is not addressed in this document. For information on the
standard-setting process, see Padgett and Richmond (1983) and McKee et al.
(1985).
This document consists of the review and evaluation of relevant literature
on ozone and other photochemical oxidants through early 1986. The material
selected for review and comment in the text generally comes from the more recent
2-5
-------
literature, with emphasis on studies conducted at or near pollutant concentra-
tions found in ambient air. Older literature that was cited in the previous
criteria document for ozone and other photochemical oxidants (U.S. Environmental
Protection Agency, 1978) has often been summarized and presented briefly. An
attempt has been made, however, to discuss at greater length in the text older
studies (1) judged significant because of their usefulness in deriving the
1979 standards; (2) open to reinterpretation because of newer data; or
(3) potentially useful in deriving subsequent standards. The newer informa-
tion on oxidants now available may in some instances make possible a better
understanding of the earlier studies, such that a more detailed and comprehen-
sive picture of health effects is emerging on several issues. An attempt has
been made to discuss key literature in the text and present it in tables as
well. Reports of lesser importance to the purposes of this document may
appear in tables only.
Generally, .only published material that has undergone scientific peer
review is included. In the interest of admitting new and important information,
however, some material not published in the open literature but meeting other
standards of scientific reporting may be ihcluded. Emphasis has been placed
on studies in which exposure concentrations were <1 ppm. On this basis,
studies in which the lowest concentration employed exceeded this level have
been included only if they contain unique data, such as documentation of a
previously unreported effect or of mechanisms of effects; or if they were
multiple-concentration studies designed to provide information on concentration-
response relationships. Application of a concentration cutoff ,of 1 ppm to
health effects studies eliminates discussion of studies on mortality and
sublethal effects. In the areas of mutagenesis, teratogenesis, and reproduc-
tive effects, however, results of studies conducted at much higher than ambient
levels have been included because of the potential importance of these long-
term effects to public health and welfare.
In selecting studies for consideration, each paper or other publication
was reviewed in detail. Technical considerations for inclusion of a specific
study on health or welfare effects, for example, included, but were not
restricted to, an analysis of the exposure method; specificity or appropriate-
ness of the analytical method used to monitor the oxidant concentration;
information on oxidant monitoring practices such as location, calibration, and
sampling time; and the appropriateness of the technique used to measure the
2-6
-------
effect. In addition, for health effects studies technical considerations
included the characteristics of the subjects studied and the techniques used
for obtaining or selecting the study cohorts. Interpretation of the results
included consideration of the following factors: the end results of the
statistical analysis; the degree to which the results are plausible in the
context of other extant data; the appropriateness of the hypothesis developed;
and the agreement between the hypothesis and the results reported. Unless
otherwise stated, all results cited in the text for health and vegetation
studies are statistically significant at p £0.05.
The general policy of EPA is to express concentrations of air pollutants
3
in metric units, e.g., in micrograms per cubic meter (M9/m ); as well as in
the more widely used units, parts per million (ppm) or parts per billion
(ppb), which are neither metric nor English units. That policy has been
followed in those chapters in which most of the data have been obtained from
laboratory studies done at room temperature (e.g., Chapters 9 and 10). Data
reported in ppm for studies conducted outdoors, such as field and open-top
chamber vegetation studies, ambient air monitoring, and research on atmospheric
chemistry, have not been converted. Conversion of reported ppm and ppb units
is highly questionable in these cases because it assumes standard or uniform
temperatures and pressures. For data in the health chapters, the conversion
3 3
units used are 1 ppm ozone = 1962 pg/m , and 1 ppm PAN = 4945 Mfl/m ; at
1 atmosphere pressure and 25°C.
2.4 REFERENCES
McKee, D.; Johnson, P.; Richmond, H.; Jones, M.; McCurdy, T.; Walton, T.
(1985) Work plan for the ozone National Ambient Air Quality Standards.
Presentation to the Clean Air Scientific Advisory Committee, March;
Research Triangle Park, NC: U.S." Environmental Protection Agency, Office
of Air Quality Planning and Standards.
Padgett, J.; Richmond, H. (1983) The process of establishing and revising
national ambient air quality standards. J. Air Pollut. Control Assoc.
33:14.
U.S. Code. (1982) Clean Air Act, §108, air quality criteria and control
techniques. U.S.C. 42: §7408.
2-7
-------
U.S. Environmental Protection Agency. (1978a) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Criteria and Assessment Office;
EPA report no. EPA-60Q/8-78-004. Available from: NTIS, Springfield, VA;
PB80-124753.
U.S. Environmental Protection Agency. (1978b) Control techniques for volatile
organic emissions from stationary sources. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Air Quality, Planning and
Standards; EPA report no. EPA-450/2-78^022. Available from: NTIS,
Springfield, VA; PB-284804/2.
U.S. Environmental Protection Agency. (1982a) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Environmental Criteria and Assessment Office; EPA report no.
EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-163337.
U.S. Environmental Protection Agency. (1982b) Air quality criteria for parti-
culate matter and sulfur oxides. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Criteria and Assessment Office;
EPA report no. EPA-60Q/8-82-Q29. .Available from: NTIS, Springfield, VA;
PB84-156777.
U.S. Environmental Protection Agency. (1983) Control technology for nitrogen
oxides emissions from stationary sources. Revised second edition.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office
of Air Quality, Planning and Standards; EPA report no. EPA-450/3-83-002.
Available from: NTIS, Springfield, VA; PB84-118330/REB.
2-8
-------
3. PROPERTIES, CHEMISTRY, AND TRANSPORT OF OZONE AND .OTHER
PHOTOCHEMICAL OXIDANTS AND THEIR PRECURSORS
3.1 INTRODUCTION
Ozone and other oxidants found in ambient air, such as the peroxyacyl
nitrates and hydrogen peroxide, are formed as the result of atmospheric
physical and chemical processes involving two classes of precursor
pollutants, volatile nonmethane organic compounds (NMOC) and nitrogen
oxides (NO ). The formation of ozone and other oxidants from these
f\
precursors is a complex, nonlinear function of many factors, including the
intensity and spectral distribution of sunlight; atmospheric mixing and
related meteorological conditions; the concentrations of the precursors in
ambient air and, within reasonable concentration ranges, the ratio between
NMOC and NO (NMOC/NO ); and the reactivity of the organic precursors.
X X
This chapter describes the physical and chemical properties of
ozone and other photochemical oxidants (Section 3.2). It also character-
izes the nature of the precursors in terms of their physical and chemical
properties, their sources and emissions into the atmosphere, and their
concentrations in ambient air (Section 3.5). In addition, a brief description
is provided (Section 3.3) of the complex atmospheric chemical processes by
which ozone and other photochemical oxidants are formed from their precursors.
A brief discussion is also included of the relationship of ozone and other
oxidants to atmospheric phenomena that result from the formation of secondary
organic and inorganic aerosols.
In addition to the information on the chemistry of oxidants and their
precursors, the chapter includes a discussion of meteorological processes
(Section 3.4) that contribute to the formation of ozone and other oxidants
and that govern their transport and dispersion once formed. Finally, an
overview is given (Section 3.6) of models of source-receptor relationships
between precursor emissions and ozone formation in the atmosphere, which
either implicitly or explicitly include the relevant emissions, atmospheric
chemistry, and meteorological processes.
3-1
-------
3.2 DESCRIPTIONS AND PROPERTIES OF OXIDANTS AND THEIR PRECURSORS
3.2.1 Ozone and Other Photochemical 0x1dants
3.2.1,1 Ozone. Ozone (03) is a triangularly shaped molecule consisting
of three oxygen atoms arranged in four basic resonance structures:
* fi * *€" • fi • • fi« *fc~ »fi • * fi • *^~ * n * *
* u« -^. *u* *u* _^. «u» «u« ->. • u» i
(I) ~~ (II)"~ ""(III) "" (IV)
The first and fourth structures, which predominate, are characterized by
the presence of a terminal oxygen atom having only six electrons. The
resonance forms depicted above have no unshared electrons. As the result
of the presence of only six electrons on one of the oxygen atoms in ozone,
the chemical reactions of ozone are electrophilic; that is, ozone removes
electrons from or shares electrons with other molecules or ions. By
definition, then, ozone is an oxidant since the term "oxidant."
characterizes an ion, atom, or molecule that is capable of removing one or
more electrons from another ion, atom, or molecule, a process called
"oxidation." A "reducing agent" adds one or more electrons to another
ion, atom or molecule, a process called "reduction." Oxidation and
reduction reactions occur in pairs and the coupled reactions are known as
"redox reactions." In redox reactions, the oxidizing agent is reduced and
the reducing agent is oxidized. The two components of such redox
reactions are known as "redox pairs." The significance of redox reactions
involving ozone is discussed in Chapters 6 and 9. The capability of
a chemical species for oxidizing or reducing is termed "redox potential"
(positive or negative standard potential) and is expressed in volts.
Ozone is, in fact, a strong oxidant having a standard potential of +2.07
volts in aqueous systems (Weast, 1977).
The physical properties of ozone are given in Table 3-1 (U. S.
Department of Health, Education, and Welfare, 1970, modified).
3-2
-------
TABLE 3-1. PHYSICAL PROPERTIES OF OZONE
Physical state
Chemical formula
Molecular weight
Melting point
Boiling point
Specific gravity relative to air
Vapor density
At 0°C, 760 mm Hg
At 25°C, 760 mm Hg
Solubility at 0°C
(Indicated volume of ozone at
0°C, 760 mm Hg)
Henry's Law constant,
37°C and pH = 7
Conversion factors
At 0°C9 760 mm Hg
At 25°C, 760 mm Hg
Colorless gas; blue violet liquid
%
48.0
-192.7 ± 0.2°C
-111.9 ± 0.3°C
1.658
2.14 g/liter
1.96 g/liter
0.494 ml/100 ml water
8666 atm/mole fraction3
o
1 ppm = 2141 (j,g/m
1 g/nr = 4.670 x 10"4
1 ppm = 1962
3 = 5.097 x 10~4 ppm
Calculated by formula of Roth and Sullivan (1981).
Source: U. S. Department of Health, Education, and Welfare (1970), modified,
3.2.1.2 Peroxyacetyl Nitrate. Peroxyacetyl nitrate (PAN) has been
observed as a constituent of photochemical smog in many localities, though
its concentrations and its ratio to ozone differ as a function of time at
a given location as well as from place to place (Chapter 5). Peroxyacetyl
nitrate, which has the formula CH3C(0)02N02, exists in a temperature
dependent equilibrium with its decomposition products, N02 and acetyl-
peroxy radicals. It can persist for substantial periods of time in the
atmosphere, depending upon temperature and the N02/NO ratio (Cox and
Roffey, 1977).
3-3
-------
The chief property of interest regarding PAN is its oxidizing
ability. A second property of PAN of interest is its thermal instability.
In the laboratory, this thermal instability necessitates that precautions
be taken in synthesizing, handling and storing PAN, since improper
handling and storage have resulted in explosions (Stephens et a!.,
1969). The ready thermal decomposition of PAN results in a notable
temperature dependence in the rate of PAN decomposition in ambient air.
Partly because of the thermal instability of PAN, its properties have
not been as well characterized as those of 03 or H202. Recent work on the
physical properties of PAN, however, has confirmed data reported earlier,
and results of the earlier and more recent work are shown in Tables 3-2
and 3-3 (Stephens, 1969; U. S. Dept. of Health, Education, and Welfare, 1970;
Kacmarek et al., 1978; Bruckmann and Willner, 1983; Holdren et a!., 1984).
The infrared (1R) spectrum of PAN is important since most researchers
rely on it for establishing concentrations of PAN for calibration. Bruck-
mann and Willner (1983) reported the IR spectrum of pure PAN and the Raman
spectrum of liquid PAN at -40°C in an argon matrix. Their work confirmed
effects that correlate with the ultraviolet (UV) spectrum published
earlier by Stephens (1969); that is, PAN was shown to be stable at x>300
nm but was efficiently photolyzed at x<300 nm (Bruckmann and Willner,
1983). Actinic radiation falling upon the surface of the earth has
wavelengths >295 nm, and it is light at wavelengths between ~295 and ~430
nm which is involved in photochemical air pollution formation.
3.2.1.3 Hydrogen Peroxide. Hydrogen peroxide (H202) is an oxidant that
occurs in ambient air as a component of photochemical smog. It is
believed to be formed through the recombination of two hydroperoxy
radicals (H02) in the presence of a third, energy-absorbing molecule
(section 3.3.1.3). In aqueous media, H202 is an inorganic actd that has a
dissociation constant of 2.4 x 10~12 and a pK of 11.62 (at 25°C) (Weast,
1977). Hydrogen peroxide has a standard potential of +1.776 in the redox
pair, H202/H20. The physical properties of H202 are given in Table 3-4.
Additional properties should be noted here that are of interest
relative to whether effects of H202 in biological receptors are of
significance. First, H202, though classed as a reasonably strong oxidant
on the basis of its standard potential for the redox system H202/H20, has
3-4
-------
TABLE 3-2. PHYSICAL PROPERTIES OF PEROXYACETYL NITRATE
Physical state, @25°C
Chemical formula
Molecular weight
Boiling point, °C
Triple point, °C
Vapor pressure,
@room temperature
Vapor pressure curve
Hydrolysis
In alkaline solution
In acidic solution
@22°C, pH 5.6
<325°C, pH 5.6
Henry's Law constant,
-------
TABLE 3-3. INFRARED ABSORPTIVITIES OF PEROXYACETYL NITRATE
(RELATED TO 295°K and 973 mb) (ppm"1 m"1 x 104) (base 10)
Frequency
Reference
Bruekmann and Wi liner (1983a)
Stephens (1964, 1969b)
1842
12.4
10.0
1741
32. 6C
23. 6d
1302
13.6
11.2
, cm'1
1162.5
15.8
14.3
791.5
13.4
10.1
aAt 4 mbar; no diluent; resolution 1.2 cm""1.
bAt 7 mbar in N2 diluent at 973 mb total pressure and 295 K; grating
instrument.
CQ branch resolved (1.5 torr).
Q branch not resolved (1.4 mbar).
TABLE 3-4. PHYSICAL PROPERTIES OF HYDROGEN PEROXIDE
Physical state, @25°C
Chemical formula
Molecular weight
Melting point, °C
Boiling point, °C, @760 mm Hg
Density, @25°C, 760 mm Hg
Vapor pressure, §16.3°C
Conversion factors
§0°C, 760 mm Hg
@25°C, 760 mm Hg
Colorless liquid
H£02
34.01
-0.41
150.2
1.4422
~1 mm Hg
1 ppm = 152U
-4
= 6.594 x 10""* ppm
1 ppm = 1390
3 = 7.195 x lO'4 ppm
Source: Weast (1977).
3-6
-------
been reported to be a positive interference in measurements of total
oxidants made by the Mast meter but to give a very slow response (slow
color development) in the NBKI method for total oxidants (Chapter 4).
This difference should be borne in mind when effects attributed to
oxidants, as opposed to ozone, are evaluated. Second, #2®2 occurs
normally as a substrate in biological systems and is involved in several
redox pairs of biological importance (see, for example, West et al.,
1966). It should also be noted that enzymes are present, at least in
mammalian systems, that catalyze the breakdown of H20£.
3.2.1,4 FormicAcid. Formic acid is a stable product formed in photo-
chemical air pollution from, for example, the reaction of HO? radicals
. - , . C, „
with HCHO and from the reactions of the Criegee biradical CHpOO with water
vapor (Atkinson and Lloyd, 1984). It has been detected in polluted
ambient atmospheres by longpath infrared spectroscopy on the basis of its
characteristic Q-branch absorption at 1105 cm (Hanst et al., 1975;
Tuazon et al., 1978a, 1980, 1981a).
Formic acid has the structure of both an acid and an aldehyde and
hence it differs in chemical behavior from other carboxylic acids in which
the carboxyl group is linked to a hydrocarbon residue rather than to a
lone hydrogen atom. In concentrated form, HCOOH is a pungent-smelling,
highly corrosive liquid with a boiling point of 100.5°C.
3.2.2 Organic Precursors
This section briefly describes and defines those hydrocarbons and
other volatile organic compounds commonly found in the ambient air of the
United States and provides relevant information about their chemical and
physical properties.
The term "hydrocarbon" has been used since the initial investigations
of tropospheric photochemistry to represent those compounds of carbon and
hydrogen that exist as gases in the ambient air and that participate along
with oxides of nitrogen in reactions that form ozone and other photochemi-
cal oxidants. As knowledge of atmospheric chemistry has increased, some
carbon compounds containing elements such as oxygen and the halogens have
also been shown to be important in photochemical air pollution. Thus, the
term "volatile organic compounds" (VOC) has come to be used to describe
3-7
-------
stable organic compounds that exist as gases under normal atmospheric
conditions, most of which can participate in the formation of photo-
chemical oxidants. Recognition that methane (CH^) is virtually unreactive
in the photochemical formation of ozone and other oxidants has given rise
to the more accurate term, "nonmethane organic compounds" (NMOC), for
describing those gas-phase organic compounds in ambient air that serve as
precursors to ozone and other photochemical oxidants. While these three
terms may sometimes appear to be used interchangeably in this chapter, the
terminology used reflects that reported in the specific literature cited
in this chapter, though in some instances differentiations may have been
made for purposes of discussion.
As discussed in Chapter 4, methods for measuring total gas-phase
hydrocarbons are not specific for hydrocarbons but may also detect other
gas-phase organic compounds, though they will not measure them quantita-
tively. Where methods are used that permit speciation of the compounds
measured, organic compounds other than hydrocarbons can be and usually are
excluded from the summation of individual species used to arrive at a
total nonmethane hydrocarbon (TNMHC) concentration. Where researchers
have used methods that do not permit speciation, an indefinite and vari-
able fraction of the reported TNMHC concentration may, in fact, be the
result of the presence of nonhydrocarbon organics and such concentration
data are more properly reported as total nonmethane organic compounds
(NMOC).
The discussion that follows is aimed at presenting basic facts on
nomenclature and those characteristics of photochemically reactive
volatile organic compounds that are relevant to the information given
subsequently in this and later chapters.
3.2.2.1 Hydrocarbons. Hydrocarbons are compounds consisting of hydrogen
and carbon only. For a given homologous series the volatilities of hydro-
carbons are related generally to the number of carbon atoms in each
molecule, as well as to temperature. Hydrocarbons with a carbon number of
one to four are gaseous at ordinary temperatures, while those with a
carbon number of five or more are liquid or solid in pure state. Liquid
mixtures of hydrocarbons such as gasoline may include some compounds in
pure form that are gases, as well as those that are liquids. Likewise,
3-8
-------
gas-phase mixtures In ambient air will usually include compounds that are
liquid or solid in their pure form. Hydrocarbons with a carbon number of
about eight or less are abundant in ambient air, but those with a carbon
number greater than about 12 have generally not been reported in the gas
phase at significant concentrations, probably because of the inability of
analytical techniques to detect these high molecular weight organics.
A saturated hydrocarbon has each of its carbon atoms bonded to four
other atoms; whereas an unsaturated hydrocarbon has two or more carbon
atoms bonded to fewer than four other atoms.
Alkanes. Alkanes, also known as paraffins, are saturated hydro-
carbons having the general formula CnH2n+2* The first compound in the
series is methane, CH^» which, because of its low reactivity, does not
contribute significantly to photochemical air pollution in urban atmo-
spheres. Alkanes as a class are the least reactive of the photochemically
important hydrocarbons (U» S, Environmental Protection Agency, 1978a,b).
Alkanes may be straight- or branched-chain compounds, and comprise the
open-chain (acyclic) hydrocarbons known as aliphatic hydrocarbons.
Alkenes. Alkenes, also known as olefins, have at least one
unsaturated bond. The number of hydrogen atoms in the general formula of
alkenes is decreased by two with respect to the alkanes for each double
bond between carbon atoms; the general formula for alkenes with one double
bond, for example, is CnH2n. The first compound in the alkene class is
ethene, also known as ethylene; the second is propene, also known as
propylene. For alkenes containing more than three carbon atoms the
position of the double bond is specified by a numerical prefix (e.g., 1-
butene). Compounds with carbon numbers three or higher can have two
double bonds between the carbon atoms and are called dienes. The complete
name of a diene is formed by including a prefix with numbers that indicate
the location of the double bonds. Like alkanes, alkenes are aliphatic
hydrocarbons and may exist as straight or branched chains. As a class,
alkenes are among the most reactive hydrocarbons in photochemical systems
(see Section 3.2.2.4).
Terpenes. Terpenes are a naturally occurring subgroup of alkenes,
many of which have the formula C1QH16. Among the terpenes identified in
ambient air a- and p-pinene have been most frequently studied. Both a-
3-9
-------
and p-pinene contain six-membered rings, as do several other terpenes; but
at least one commonly occurring member of this group, myrcene, is an
acyclic or open-chain compound. Isoprene, also a naturally occurring
alkene, is a hemiterpene having the formula CgHg.
Alkynes. Alkynes are open-chain hydrocarbons that contain one or
more triple bonds. Acetylene, C2H2, is the simplest member of the class,
which as a whole is often referred to as the acetylenes. The general
formula for the acetylenes is CnH2n_2, and for each additional triple bond
in the molecule four hydrogen atoms must be removed from the general
formula. Acetylene is commonly present in ambient air, is thought to be
emitted largely from mobile sources, and has often been taken to be an
indicator of auto exhaust emissions, since it is relatively unreactive in
ambient air and persists in the atmosphere longer than most other exhaust
components.
Aromatics. Aromatic hydrocarbons include various compounds having
atoms arranged in six-membered carbon rings with only one additional atom
(of hydrogen or carbon) attached to each atom in the ring. Benzene is the
simplest compound in the series, having no side chains but only six carbon
atoms and six hydrogen atoms, linked by three conjugated double bonds.
Compounds containing the aromatic ring and elements other than carbon
and hydrogen are included with aromatic hydrocarbons in the general class-
ification "aromatics." The double bonds in aromatics are not nearly as
chemically active as those in alkenes because of an effect called
"resonance stabilization." As a class, aromatics exhibit a wide range of
photochemical reactivity, with benzene having a low photochemical reac-
tivity and 1,3,5-trimethylbenzene, for example, showing high reactivity.
3.2.2.2 Aldehydes. Aldehydes probably constitute the single most
abundant group of volatile organic compounds other than hydrocarbons in
ambient air. They are photochemically. important compounds because they
photolyze to form free radicals that will react with oxygen in ambient air
to form alkylperoxy or hydroperoxy radicals (National Research Council,
1977a) (see below).
Aldehydes are characterized by the presence of the formyl functional
group (CHO). A carbonyl group having a carbon-oxygen double bond, C=0, is
part of the formyl group. The carbonyl group is not unique to aldehydes,
3-10
-------
since it is found also in ketones and carboxylic acids; but it forms the
basis for one of the analytical methods used for measuring aldehydes in
ambient air (Chapter 4).
3.2.2.3 Other Organi cCompounds. Other organic compounds found in
ambient air are known to be photochemically reactive in the formation of
ozone and other photochemical oxidants. These other organic compounds do
not occur in ambient air collectively, much less singly, at concentrations
that approach the total concentrations of nonmethane hydrocarbons. Some
of them are suspected of having potentially adverse health effects,
however, and are therefore under scrutiny by the U. S. Environmental
Protection Agency. These compounds are mentioned here only because they
are photochemically reactive, can serve as precursors to oxidants, and
because they contribute a small but indeterminate fraction of the total
NMOC concentrations reported when continuous hydrocarbon analyzers
(Chapter 4} are used to determine ambient levels of volatile organic
compounds.
Many of the volatile organics in ambient air that are not hydro-
carbons are organic halides, in which one or more hydrogen atoms of a
hydrocarbon have been replaced by a halogen atom such as chlorine,
fluorine, or iodine. An enormous number of relatively simple organic
halides are possible, since a halogen atom can be attached to an organic
compound in many different positions.
3.2.2.4 Volatility and Reactivity. The physical and chemical properties
of nonmethane organic compounds that are most pertinent to their role as
precursors to ozone and other oxidants are those properties that govern
their emission into the atmosphere (volatility) and their lifetime in the
atmosphere, the latter being determined by photochemical reactions
(reactivity) and other removal processes (e.g., gas-to-particle conversion
and dry deposition).
To be significant in atmospheric reactions, an organic.compound must
have a sufficiently high volatility. Based upon a review of the available
o
literature, Singh et al. (1984) have chosen a vapor pressure of 10 atm
as the criteria for deciding whether an organic compound should be de-
scribed as volatile. Those compounds with vapor pressures less than 10
atm were considered by Singh et al. (1984) to occur predominately in
3-11
-------
the condensed phase and therefore not to participate in atmospheric
reactions. Clearly, a rigid cutoff for vapor pressures will not
necessarily be applicable to all organic compounds.
The photochemical reactivity of subclasses and individual species of
hydrocarbons and of other volatile organic compounds is relevant to mech-
anistic studies in atmospheric chemistry, to modeling, and to other
oxidant-control-related research; but it is not pertinent to the deriva-
tion of criteria. A major discussion of these properties therefore lies
outside the scope of this document, but a brief discussion of the concept
of hydrocarbon reactivity and its application is presented here.
Differences in reactivities among volatile organic compounds have
been the focus of considerable attention and research for nearly three
decades. In early research on photochemical air pollution, many different
definitions or criteria were used to evaluate the reactivity of organic
compounds. Examples of such criteria include: rate of NO-to-NOg conver-
sion, maximum ozone concentration formed, initial rate of disappearance of
the organic compound, eye irritation, damage to vegetation, and aerosol
formation.
Historically, reactivity classifications have been based on environ-
mental chamber measurements of these criteria, as observed in the photo-
oxidation of hydrocarbon-oxides of nitrogen mixtures under conditions
approximating those of polluted ambient atmospheres. Many of the reac-
tivity data that had been accumulated through 1969 for each of these
manifestations (except plant damage) were critically reviewed by
Altshuller and Bufalini in 1971. They noted general agreement in reac-
tivity trends from studies employing different reactivity criteria, but
they also cited a number of significant discrepancies in the specific
assignments of reactivity to individual compounds and even to whole
classes of compounds.
In recent years assessments of the reactivity of volatile organic
compounds have focused almost exclusively on the ability of an organic
compound to produce ozone and other photochemical oxidants. This focus
arises from interest in regulating most stringently the emissions of
organic compounds having the highest potential for forming ozone and other
photochemical oxidants.
3-12
-------
The 1978 criteria document for ozone and other photochemical oxidants
summarized reactivity data acquired from the mid-1960s to the mid-1970s
(U. S. Environmental Protection Agency, 1978a). Reference to tables of
reactivity schemes given in the 1978 document shows the relatively higher
reactivities of internally double-bonded alkenes, of aliphatic aldehydes
and other carbonyl compounds (such as branched alkylketones and
unsaturated ketones), of dienes, of 1-alkenes, of partially halogenated
alkenes, and of alkylbenzenes (primary and secondary monoalkylbenzenes,
and di-, tri- and tetraalkylbenzenes). Other compounds also have
relatively high reactivity but are not expected to be as abundant in
ambient air as the compounds cited above.
For more information the reader is referred to the 1978 criteria
document (U. S. Environmental Protection Agency, 1978a) and the references
therein [e.g., Dimitriades (1974) and Pitts et al. (1977)] for further
information on reactivities of specific compounds. The reader is also
referred to a more recent and comprehensive assessment of the present
literature on the reactivity and volatility of 118 organic chemicals by
Singh et al. (1984) in which a three-tiered classification scheme was
developed based on the potential for involvement of a given chemical in
photochemical air pollution formation.
Since a key reaction of volatile organic compounds in ambient air,
regardless of their class, is their oxidation via attack by hydroxyl
radicals (Atkinson et al., 1979, 1982c; Atkinson, 1985), the basis of
several proposed reactivity classifications is the rate of reaction
between an organic compound and the OH radical. This reaction is the
first step in a chain reaction that is propagated by various organic
peroxy radicals. As discussed in section 3.3.2.1, reaction with the OH
radical is thought to be the predominant loss process for most organ!cs in
the troposphere (Atkinson et al., 1979; Atkinson, 1985). On this basis, a
five-class reactivity scale was proposed by Pitts and coworkers (Darnall
et al., 1976; Pitts et al., 1977) based on the rate of reaction of more
than 100 VOC with OH radicals. In this scale, each class spanned an order
of magnitude in reactivity relative to methane, with Class I corresponding
to an atmospheric half-life of greater than 10 days and Class V a half-
life of less than 2.5 hours. The scale has the advantage that any
3-13
-------
compound whose OH rate constant has been measured can be placed in a
precise position in the scale. It has a number of limitations (Pitts et
al.» 1977), however, since it makes the implicit assumption that OH
radical reaction is the sole loss process for an organic, and that the
subsequent atmospheric chemistry is identical for all organic compounds.
Both these limitations and the advantages of the OH radical reactivity
scale are discussed in detail elsewhere (Pitts et a!., 1977, 1985).
3.2.3 Nitrpgen Oxides
The physical and chemical properties of the nitrogen oxides that
serve as precursors in the formation of ozone and other photochemical
oxidants have been documented in a recent air quality criteria document
(U. S. Environmental Protection Agency, 1982a). The most pertinent
properties are briefly summarized here. The role of nitrogen oxides in -
the formation of oxidants in the troposphere is discussed in Section 3,3
and in the document cited above.
The three most abundant oxides of nitrogen in ambient air are nitric
oxide (NO), nitrogen dioxide (N02), and nitrous oxide (N20). The latter,
a product of soil microbiology, is not known to participate in photo-
chemical reactions in the troposphere. The two important oxides of
nitrogen relative to photochemical processes in the troposphere are NO and
N0«, which are abundant in ambient air and participate in cyclic reactions
leading to the production of ozone and other oxidants, as described later
in this chapter.
The basic reactions of importance are (1) the photolysis of NQ2 (K <
430 nm); (2) subsequent formation of ozone from the reaction of atomic
o
oxygen [0('5P)] produced from the photolysis of N02 with $2 (^n tne
presence of a third, energy-absorbing molecule); and (3) the subsequent
regeneration of N02 by the reaction of NO with Oj. Coupled with these
basic reactions are reactions between NO and free radicals in the
atmosphere (hydroperoxy, alkylperoxy and acylperoxy) that oxidize NO to
N02, disturbing the N0-N02 equilibrium that would otherwise exist, and
leading, then, to the buildup of Og. These reactions, and further
information on the source of the free radicals, are given below. Basic
physical and chemical properties of NO and N02 are given in Table 3-5.
3-14
-------
TABLE .3-5. PHYSICAL AND CHEMICAL PROPERTIES
OF NITRIC OXIDE AND NITROGEN DIOXIDE
Property
Other
properties
of note:
NO
Odor
Taste
Color
Absorption
X, nma
None
-
None
<230
Pungent
"
Reddish-brown
Broad range,
both >400
and <400
Uneven number of
valence electrons
Corrosive, strong oxidant.
Photolyzes at x <430 nm.
Low partial pressure in ambient air.
Uneven number of valence electrons.
Forms dimers
aVisible light x >40Q nm; ultraviolet x <400 nm. Solar UV radiation in
the troposphere extends from about x290 nm to about x4UO nm.
Source: Derived from National Research Council (1977b) and U. S. Environ-
mental Protection Agency (1982a).
3.3 .ATMOSPHERIC CHEMICAL PROCESSES: FORMATION AND TRANSFORMATION OF
OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
The photochemistry of the polluted atmosphere is exceedingly
complex. Even if one considers only a single hydrocarbon pollutant, with
typical concentrations of nitrogen oxides, carbon monoxide, water vapor,
and other trace components of air, several hundred chemical reactions are
involved in a realistic assessment of the chemical evolution of such a
system. The, actual urban atmosphere contains not just one but hundreds of
different hydrocarbons, each with its own reactivity and oxidation
products.
(National Research Council, 1977a)
Despite the complexities of the chemistry of polluted atmospheres, it
is sufficient to understand certain basic processes involved in the formation
of photochemical oxidants from precursor compounds in the presence of sun-
light. The concentrations of ozone and other oxidants found in urban areas
and in downwind and rural receptor regions are the net result of at least
three general processes: (1) the initial emission, dispersion, and transport
of precursors; (2) the photochemical reactions that occur in the atmosphere
3-15
-------
as the dispersion and transport take place; and (3) the scavenging processes
along the trajectory that reduce the concentrations both of precursors and
the resulting oxidants.
Ozone (03) is formed in ambient air through the addition of an atom of
oxygen (0) to a molecule of oxygen (Op). The breakdown by sunlight (photoly-
sis) of nitrogen dioxide (N02) into nitric oxide (NO) and atomic oxygen provide
the atoms of oxygen involved. The NO formed in this reaction then reacts with
the 03 produced from the reaction between atomic and molecular oxygen. In
these cyclic reactions, no net increase in 0, occurs, with the result that an
equilibrium is set up among 03, N02, and NO. Any reactions that produce N02
without destroying 0- will upset this equilibrium, however, and will result in
a net increase in Oo. In ambient air, the oxidation of photochemically reactive
hydrocarbons and other nonmethane organic compounds (NMOC) provides a source of
reactive species (radicals) that convert NO to N02 without destroying 03, thus
upsetting the equilibrium. The reactions of these radicals with NO also consti-
tute a cyclic process. Since terminating reactions occur between N02 and these
radicals, as well, which remove both N02 and the radicals from the photochemical
reaction system, the cycles described above would gradually end, even in the
presence of sunlight, unless fresh NO emissions were injected into the atmos-
X
phere. The complexity of these cyclic, coupled reactions is such that ozone
concentrations in ambient air are a nonlinear function of the NMOC and NO
/\
concentrations and, within realistic ranges of precursor concentrations, of the
NMOC:NOV ratio.
f\
In the following sections, the processes just described in a simplistic and
summary manner are presented in more detail. For a complete and thorough dis-
cussion of the many complex reactions thought to take place in polluted atmos-
pheres, the primary literature should be consulted (e.g., Demerjian et al., 1974;
Finlayson and Pitts, 1976; Logan et al., 1981; Whitten, 1983; Atkinson and Lloyd,
1984; and Atkinson, 1985).
3.3.1 Inorganic Reactions
3.3.1.1 Formation of Ozone: The NO-NOo-0^ Cycle. Many aspects of the
inorganic reaction systems in the atmosphere are now well understood. The
photodissociation of N02 by near-ultraviolet solar radiation is a critical
process:
N02 + hv (295^\<430 nm) ^ NO + 0(3P) (3-1)
3-16
-------
The subsequent reaction of the resulting 0( P) atom with molecular oxygen
produces an ozone molecule:
0(3P) + 02 + M > 03 + M (3-2)
where M is a "third body" molecule (e.g., N2) which can carry away excess
energy of reaction. In the absence of any competing reactions, the rapid
reaction of NO with Og completes this reaction cycle, regenerating ah N02
molecule:
NO + 03 ->• N02 + 02 (3-3)
As a result of the above three reactions, an equilibrium or steady-state
condition is established among NO, N02 and 0^, and the concentration of Og
in the atmosphere is governed by the expression,
[NO ]
(3-4)
where K = ^3-1/^3-3 which depends on the sunlight intensity. Typically, K
in the lower troposphere is less than or equal to 0.025 ppm.
Because reaction (3-3) is rapid, ozone concentrations in urban
atmospheres cannot rise until most of the NO has been converted to f^.
This accounts in part for the fact that 03 levels may be lower on average
in city centers where high NO emissions occur, but higher in downwind
suburban areas to which the resulting N02 is transported and then photo-
dissociated, leading to 03 formation. The characteristic behavior of
irradiated NMOC-NOX systems in producing Og and other photochemical
oxidants is shown in Figure 3-1, which depicts data obtained from an
environmental chamber irradiation of a propene-NQ-N02 mixture (Pitts et
al., 1979).
3.3.1.2 Formation of RadicalIntermediates. Reactions (3-1) through
(3-3), however, cannot by themselves explain the buildup of ozone, since
for each molecule of NO oxidized to N02 in reaction (3-3) a molecule of
ozone is also destroyed. An alternate pathway of conversion of NO to N02
that does not destroy 03 is needed to explain the high ozone levels
observed in urban environments. Such an alternate pathway is available
through the oxidation of reactive organic compounds. In the atmosphere,
these compounds can be oxidized by ozone (03) and/or hydroxyl radicals (OH)
3-17
-------
0.6
60
120
180
240
300
360 420
ELAPSED TIME, minutes
Figure 3-1. Experimental time-concentration profiles for propene, NO, NO2,
O3, HCHO, and PAN for an irradiated NOx-propene-air mixture.
Source: Pitts et al. (1979).
3-18
-------
3.3.1.2.1 Hydroxyl and Hydroperoxyl Radicals. There are at least three
significant formation routes leading to the production of hydroxyl
radicals in the atmosphere. A pathway for OH radical formation, which
becomes important in the afternoon as ozone concentrations rise, is the
photolysis of 03:
03 + hv (\<319 nm) * 0(1D) + 02(1Ag) (3-5)
The electronically excited O^D) atoms may be quenched to ground state
O(^p) atoms or may react with water vapor to yield OH radicals with an
approximately 20 percent efficiency at 298 K and 50 percent relative
humidity:
O^D) + H20 * 2,OH " (3-6)
Nitrous acid, which has been shown to accumulate to concentrations of
~1 to 8 parts-per-billion (ppb) during the night in the Los Angeles basin
(Platt et al., 1980a; Harris et al.f 1982; Pitts et al., 1984c) will
photolyze at sunrise, producing a "pulse" of OH radicals (Harris et al.,
1982):
HOMO + hv (X<400 nm) •* OH + NO (3-7)
This photolytic reaction represents a major sink for HONO during daylight
hours. (Other aspects of the formation and atmospheric chemistry of this
important species are discussed in Section 3.3.1.4.)
A third significant source of OH radicals is the photolysis of HCHO-.
HCHO + hv (X<370 nm)
H + HCO (3-8a)
CO (3-8b)
Formaldehyde is both a primary (e.g., from motor vehicle exhausts) and
secondary pollutant that may occur in significant concentrations in the
morning hours as well as in the afternoon (Tuazon et al., 1978a, 1981a;
Grosjean, 1982).
3-19
-------
The H atoms formed in reaction (3-8a) or from reactions such as
(3-9):
OH + CO -> C02 + H (3-9)
can react with oxygen to produce hydroperoxyl radicals:
M
H + 02 -> H02 + M (3-10)
These can then react with NO to form hydroxyl radicals:
H02 + NO * OH + N02 (3-11)
Reaction (3-11) then completes a chain reaction involving reactions (3-9),
(3-10) and (3-11) and is a major pathway for the oxidation of nitric oxide
in ambient air.
The formyl radical in reaction (3-8a) may also serve as a precursor
to H02 radicals and hence OH radical formation:
HCO + 02 -> H02 + CO (3-12)
Other sources of formyl radicals include the photolysis of higher
aldehydes:
RCHO + hv U<350 nm) ->• R + HCO (3-13)
In addition to reactions (3-11) and (3-12), H02 radicals are produced
by H-atom abstraction from alkoxy radicals as discussed below:
RCH20 + 02 + RCHO + H02 (3-14)
These H02 radicals will also oxidize NO to N02 via reaction (3-11).
Based upon environmental chamber data, computer modeling studies and
measured ambient concentrations of unreactive organics such as fluoro-
chlorocarbons, concentrations of OH and H02 radicals in polluted
atmospheres are believed to be in the ranges 5 x 10 to 5 x 10^ and 10^ to
109 radicals cm~3, respectively.
3-20
-------
3.3.1.2.2 Mitrate Radicals. Ozone can react with N02 to produce the
nitrate radical and an oxygen molecule:
03 + N02 •» N03 + 02; (3-15)
however, because of its large photolytic cross section (Graham and
Johnston, 1978; Magnotta and Johnston, 1980) the N03 radical photolyzes
rapidly in sunlight:
N03 + hv •»• N02 + 0(3P) (3-16a)
-»• NO + 02 (3-165)
Direct spectroscopic measurements (Platt et al., 1980b, 1984; Noxoh. et al.,
1980; Pitts et al., 1984c) have confirmed that N03 radical concentrations only
rise above part-per-trillion (ppt) levels after sunset. The atmospheric
reactions of this important radical intermediate are discussed in Sections
3.3.1.5 and 3.3.2.3, including the fact that at night the N03 radical will
participate in a rapid equilibrium between N02» N03 and dinitrogen
pentoxide (N20g).
3.3.1.3 Termination Reactions. Although the photochemical reactions
described above require sunlight, the presence of sunlight does not mean
that the reactions continue indefinitely. Terminating reactions gradually
remove NO and N02 from the reaction mixtures such that the cycles would
slowly come to an end unless fresh NO emissions were injected into the
atmosphere. Specifically, the inorganic chemistry system includes termin-
ation reactions for OH and H02 radicals with NO and N02 to form nitrogen
acids such as nitrous acid (HOMO), nitric acid (HN03) and peroxynitric
acid (H02N02):
M
OH + NO •» MONO (3-17)
M
OH + N02 + HN03 (3-18)
M
H02 + N02 •* H02N02 (3-19)
3-21
-------
Pernitric acid, however, thermally back-dissociates rapidly U]y2 ~10 sec
at 298 K) and HONO photolyzes, so that under typical atmospheric
conditions OH + N02 is the major sink of NOX.
Nitrous acid and nitric acid have now both been reliably measured in
ambient air by longpath spectroscopic techniques and, in the case of
nitric acid by other techniques as well. Peroxynitric acid has not yet
been observed in the atmosphere, although it has been predicted to be
present at fractional ppb concentrations and has been extensively studied
in laboratory systems (Graham et al., .1977; Hanst and Gay, 1977; Howard,
1977; Niki et al., 1977; Graham1 et al., 1978).
In the presence of NO, radical-radical reactions are generally not of
major importance in the atmosphere, because concentrations of the radicals
are low. In the absence of NO, however (for example at night), the
reactions of peroxy radicals with H02 and with other peroxy and acylperoxy
radicals, and the self-combination of H02 radicals, can become
important. For example, the reaction of H02:
H02 + H02 -> H202 + 02 (3-20)
could be an important route to the oxidation of S02 to sulfate in solution
(see section 3.3.4).
Clearly, termination of the chain reactions can lead to the formation
of other oxidants as well as relatively stable organic nitrates in the
atmosphere. In addition to HONO, HN03, H02N02 and H202, these oxidants
include other peroxyacyl nitrates, organic hydroperoxides and organic
peracids which have been observed either in polluted atmospheres or in
irradiated laboratory mixtures (National Research Council, 1977a). These
compounds almost always occur in low concentrations in ambient air, but
they may play a significant or even critical role in atmospheric chemistry
(Pitts et al., 1983) as shown in the following sections.
3.3.1.4 Reactions Involving Nitrous Acid. An alternative pathway for
reaction (3-19) is the formation of HONO:
H0 + N0 + HONO + 0 (3-21)
3-22
-------
This reaction pathway has been shown, however (Graham et al . , 1977, 1978;
Howard, 1977), to be negligible compared to reaction (3-19).
The equilibrium between NO, NCL, and hLO results in still another
potential source of HONO:
NO + N0 + H0 -> 2 HONO (3-22)
2
2 HONO ^ NO + NQ2 + H20 (3-23)
Reactions (3-22) and (3-23) may proceed both homogeneously and hetero-
geneously, but they appear to be too slow to be of atmospheric importance
at part-per-million concentrations of NO .
X,
Similarly, the reaction of N02 with water may proceed in the gas
phase or on surfaces:
2 N02 + H20 -> HONO + HN03 . (3-24)
Recent work (Sakamaki et al., 1983; Pitts et al., 1984a) has shown that
HONO is produced in environmental chambers from the reaction of NQ2 with
water vapor, almost certainly via heterogeneous processes (although HNOg
was not observed in these studies). This process may be a minor source
of HONO in the atmosphere and in the exhaust plumes from combustion sources
(Pitts et al. , 1984b).
3.3.1.5 Reactions Involving Nitric Acid and Dinitrogen Pentoxide.
Another equilibrium reaction of importance is that involving NQ2, NO*,, and
N205: ;
M ••...--.
N02 + N03 $ N205 (3-25a,b)
The equilibrium constant for this system has been measured by several ,
groups (Graham and Johnston, 1978; Kircher et al., 1984; Tuazon et al.,
1984; Perner et al., 1985) and appears to be 3 (±0.5) x ID"11 cm molecule'1
at 298 K, which corresponds to a time to reach equilibrium of about one
minute. The equilibrium constant remains somewhat uncertain, however.
Dinitrogen pentoxide is a potentially important precursor to HNO,
(and hence acid deposition) through its reaction with water either in the
gas phase (Tuazon et al., 1983) or on surfaces (Heikes and Thompson, 1983):
3-23
-------
N2°5 + H2° "*" 2 HN03 (3-26)
Thus, for N0£ and NQg radical concentrations representative of receptor
sites downwind from major urban areas such as Los Angeles (Pitts et al.,
1983, 1984c; Platt et al.s 1984), and using an N00 + NO, $ N,0C
-11 3 -1
equilibrium constant of 3 x 10 cm molecule , an HNOQ formation rate
-1
of several ppb hr is obtained at ^-50 percent relative humidity [assuming
-?1 3 -1 -1
the upper limit rate constant of 1.3 x 10 cm molecule sec for
reaction (3-26)].
This estimated nighttime formation rate of HNCL via reaction (3-26)
-1
can be compared to a calculated daytime formation rate of *1 ppb hr from
reaction (3-18) for ^20 ppb of N02 and 1 x 106 molecule cm"3 of OH radicals.
Reaction (3-26) could potentially be an important loss process for NO and a
J\
significant nighttime pathway for HNQg formation in urban atmospheres.
3.3,2 Organic Reactions
It is now well recognized that all organic compounds emitted into the
atmosphere may be degraded by one or more of the following four pathways:
reaction with hydroxyl radicals, reaction with ozone, reaction with
nitrate radicals, or photolysis. Indeed, knowledge of the rates and
mechanisms of these processes has advanced to the point that the process that
will predominate for a given compound can be predicted with a reasonable
degree of certainty.
This progress notwithstanding, there remain substantial differences
in the degree to which the detailed atmospheric chemistry is understood
for the principal classes of hydrocarbons found in polluted atmospheres:
alkanes (paraffins), alkenes (olefins), and aromatics. Thus, the photo-
oxidation reactions of the smaller alkanes and the simple alkenes, such
as ethene, propene, and trans-2-butene, are fairly well understood. There
is much less certainty, however, about the detailed reactions undergone by
the higher alkanes, the higher alkenes, and the aromatics subsequent to
their initial reactions with OH radicals, ozone, or N03 radicals.
The following sections briefly summarize the basic features of the
four reaction pathways identified above for organic compounds emitted into
the atmosphere.
3-24
-------
3.3.2.1 Reactjons with Hydroxyl Radlcals. The following sections treat
separately the mechanisms of reaction of OH radicals with the major
classes of organic compounds including alkanes, alkenes, aromatics and
oxygenated compounds, as well as nitrogen- and sulfur-containing
compounds. The treatment here of relevant reactions is necessarily an
overview. For a comprehensive and current detailed description of the
kinetics and mechanisms of the atmospheric reactions of OH radicals with
organic compounds, the reader is referred to a review by Atkinson (1985).
3.3.2.1.1 Alkanes. It is now well established that the only significant
atmospheric chemical loss process for the alkanes is reaction with OH
radicals. These reactions proceed by hydrogen abstraction (Atkinson,
1985) to produce alky! radicals (R)» which then add 02 to form alkyl
peroxy radicals (R02):
OH + RH t- R* + H20 (3-27)
M
R' + 02 + R02' (3-28)
In polluted atmospheres R02 radicals rapidly oxidize NO to N02»
forming alkoxy radicals (RO); or add N02 to form alkyl peroxynitrates:
RU2* + NO f RO* + N02 (3-29)
R0£* + N02 + R02N02 (3-30a,b)
The latter, however, are not expected to be present in ambient air at
significant concentrations because of their short (<1 sec at 298 K)
lifetimes with respect to thermal decomposition (reaction 3-30b).
Alkoxy radicals may also undergo hydrogen abstraction by molecular
oxygen to form aldehydes or ketones (Baldwin et al., 1977):
R'RCHO* + 02 * R'OR + HQ2* (3-31)
or they may decompose to form oxygenates (Baldwin et al., 1977; Batt,
1979):
R'RCHO"
RCHO + R" (3-32)
R'CHO + R* (3-33)
3-25
-------
In both of these reaction sequences, however, H02 radicals are formed, and
hence OH radicals are regenerated. The carbonyl compounds thus formed may
subsequently react with OH radicals or may photodecompose (see Section
3.3.2.1.4).
A further important reaction pathway for acyl radicals is the
addition of 02, followed by reaction with N02 to form peroxyacyl nitrates:
0 0
RC* + 02 •* RCOO* (3-34)
0 0
RCOO' + N02 -» RCOON02 (3-35)
The simplest member of this class of compounds, peroxyacetyl nitrate, has
been measured in polluted atmospheres throughout the world (see Chapter 5).
The reactions described above suggest the importance of simple (C^) alkanes, namely alkoxy
radical isomerization (Carter et al., 1976; Baldwin et al., 1977; Hendry
et al., 1978; Batt, 1979; Batt and Robinson, 1979; Carter et al., 1979)
and alky! nitrate formation from the reaction of R02* with NO (Atkinson et
al., 1982a, 1983, 1984f):
M
R02* + NO •* RON02 (3-36)
Discussion of these processes is beyond the scope of this chapter and the
reader is referred to the original literature and to appropriate reviews
(e.g., Atkinson, 1985) for summaries of these aspects of the atmospheric
chemistry of longer-chain alkanes.
3.3.2.1.2 Alkenes. In polluted atmospheres, unsaturated hydrocarbons
react primarily with OH radicals and with Og (Herron and Huie, 1977, 1978;
Dodge and Arnts, 1979; Akimoto et al., 1980; Kan et al., 1981; Niki et
al., 1981; Atkinson et al., 1982c; Whitten, 1983; Atkinson, 1985). For
most alkenes studied to date, the reaction with OH radicals proceeds
3-26
-------
almost entirely by addition to the double bond. In the case of propene,
for example, addition of the OH radical to the double bond is expected to
be followed by 02 addition, with the oxidation of NO to N02;by the
resulting peroxy radical to form an alkoxy radical (Atkinson et al., 1985)
OH* + CH3CH=CH2 •* CH3CHCH2OH (3-37)
00*
I
CH3CHCH2OH + 02 •* CH3CHCH2OH (3-38)
00' 0 *
CH3CHCH2OH + NO + CH3CHCH2OH + N02 (3-39)
Decomposition of the alkoxy radical and subsequent reactions lead to
the formation of acetaldehyde and formaldehyde, both of which can be
detected in polluted atmospheres:
CH3CHCH2OH + CH3CHO + "CH^H (3-40)
HCHO + H0 . • (3-41)
OH + N02
The overall reaction resulting from reactions (3-37) to (3-41) is:
OH + CH3CH=CH2 + 202 + 2 NO + CH3CHO + HCHO + 2 N02 + OH (3-42)
1
0,
hv, 02
Thus, the reaction of OH radicals with alkenes increases the rate of NO-
to-N02 conversion and hence increases the yield of ozone. The specific
reaction sequence for the OH radical-initiated oxidation of propene
(Atkinson, 1985) is shown in Figure 3-2.
3-27
-------
OH
(-65%)
(-35%)
CH-CHCH OH CH,C1
J £-i -J
NO —
\
°2
-»* N02 NO -
^
N0n
f
_CH CHO_ + CH OH
1'
HO + HCHO
£. ~::
OH
CH-CHOH + HCHO
l°2
f
CH^CHO + HO^
Figure 3-2. Reaction scheme for OH radical-initiated oxidation of propene
in the presence of NO.
Source: Atkinson (198i).
3-28
-------
The reaction schemes presented for propane illustrate the major role
that organic compounds (not only alkenes but also alkanes and aromatics)
play in producing photochemical air pollution, namely acceleration of the
conversion of NO to NQ2 and the resulting formation of ozone.
Al k e n e s Emi 11 e d F rom Ve get at\onf ^ special class of alkenes
receiving considerable attention over the past decade are those
unsubstituted compounds emitted from vegetation. Examples include iso-
prene and monoterpenes such as a- and p-pinene, d-limonene, and myrcene.
Much research and discussion have been devoted to assessments of the
potential for such compounds, which are emitted from vegetation in large
quantities, to contribute to photochemical air pollution (Coffey, 1977;
Westberg, 1977; Arnts and Gay, 1979; Tingey and Burns, 1980; Bufalini and
Arnts, 1981; Dimitriades, 1981; Altshuller, 1983), but a detailed treat-
ment of this topic is beyond the scope of this document. Presented here,
and in other appropriate parts of section 3.3, are relevant aspects of
current knowledge of the atmospheric chemistry of organic compounds known
to be emitted from vegetation.
Although reliable rate constants for the reaction of these naturally
emitted organic compounds with OH radicals are now available (Winer et
al., 1976; Kleindienst et a!., 1982; Atkinson, 1985), with the exception
of isoprene their detailed atmospheric chemistry is still not well
characterized. Summarized briefly here are the OH radical-initiated
photooxidation reactions for isoprene and a-pinene.
Isoprene. Based on the relative atmospheric concentrations of OH
radicals and ozone, and their rate constants for reaction with isoprene,
the dominant atmospheric reaction pathway for isoprene is expected to be
OH radical addition to the olefinic double bonds (Lloyd et al., 1983):
CH0
I 3
—> HOCH2-C-CH=CH2 (3-43a)
CH
OH + CH2=C-CH=
I 3
H=CH2
1-
CH0
I 3
CH2C-CH=CH2 (3-43b)
OH
CH,
3
CH2=C-CHCH2OH (3-43c)
-> CH2=C-CH-CH2 (3-43d)
3-29 OH
-------
The hydroxyalkyl radicals formed in reactions (3-43a)-(3-43d) react
rapidly with 02 to form peroxy radicals, which can then rapidly oxidize NO
to N02 (Atkinson and Lloyd, 1984). The resulting hydroxyalkoxy radicals
decompose to form methyl vinyl ketone and methacrolein. The subsequent
atmospheric reactions of these products are described in detail by Lloyd
et al. (1983) and by Killus and Whitten (1984).
a-Pinene. Based on the rate constant for its reaction with OH
radicals (Atkinson et al., 1979; Kleindienst et al., 1982), a-pinene is
expected to react exclusively by OH radical addition at the least
substituted carbon atoms, with the resulting radical rapidly adding 02
(Lloyd et al., 1983):
°2
OH + > -» > (3-44)
A mechanism for the subsequent reaction pathways of this hydroxyperoxy
radical has been reported (Lloyd et al., 1983), but it was of necessity
largely parameterized because of the lack of data on the reaction products
resulting from the photooxidation of a-pinene under atmospheric
conditions. Even less information is available for other monoterpenes and
any detailed consideration of their atmospheric chemistry would be highly ,
speculative.
For further descriptions of the OH radical-initiated photooxidations
of isoprene and monoterpene, the reader should consult the,,primary
literature (Lloyd et al., 1983; Killus and Whitten, 1984).
3.3.2.1.3 Aromatics. The aromatic fraction in gasolines has increased in
recent years, partly as a result of the reduction of lead in gasoline.
Given that there are also substantial emissions of aromatic compounds
(e.g., benzene, toluene, xylenes, etc.) from a wide range of industrial
processes, the importance of aromatics in the hydrocarbon distribution in
ambient atmospheres has grown.
Reactions with OH radicals constitute the sole atmospheric loss
process for aromatic compounds. The available kinetic and mechanistic
3-30
-------
data concerning these reactions have been reviewed critically (Atkinson,
1985). It is clear from these data that two reaction pathways are
possible. The first of these is OH radical addition to the aromatic ring
to form an initially energy-rich QH-aromatic adduct which can either
decompose back to the reactants or be collisionally stabilized. This is
illustrated for toluene, an abundant aromatic constituent in urban
atmospheres:
OH
H,
(plus other isomers) (3-45)
For alkyl-substituted benzenes, the second reaction pathway involves
H-atom abstraction from the substituent group:
OH
(3-46)
This latter is a minor (<1Q$) process at room temperature (Atkinson,
1985). The reaction pathways subsequent to the H-atom abstraction
reaction pathway (3-46) are reasonably well understood (Atkinson and
Lloyd, 1984). Thus, under atmospheric conditions the benzyl radical is
expected to react via the following sequence of reactions:
CH200'
(3-47)
3-31
-------
+ NO
M
+ NO,
(3-48a)
(3-48b)
with reaction (3-48b) occurring approximately 10 percent of the time at
atmospheric pressure and room temperature (Hoshino et al., 1978).
The CgHgCHgO" radical then reacts with 02 to yield benzaldehyde and an H02
radical:
+ HO,
(3-49)
Analogous reaction pathways are expected to be applicable to the other
aromatic hydrocarbons, after H-atom abstraction from the substituent alkyl
groups (Atkinson and Lloyd, 1984).
The rates of reaction of OH radicals with aromatics are now well
characterized (Atkinson, 1985), as is the relative importance of the
alternative pathways (3-45) vs (3-46), at least for selected aromatics.
In the case of toluene, for example, OH radical addition is expected to
occur ~80 percent of the time at the ortho position (Kenley et al., 1981).
The fate, however, of the addition adducts formed from OH radical-
aromatic reactions remains unclear (Atkinson, 1985), although several
mechanisms have been proposed in the case of toluene (Atkinson et al.,
1980; Leone et al., 1985). Although substantial progress has been made in
understanding reaction mechanisms for toluene and certain other aromatic
hydrocarbons (Atkinson and Lloyd, 1984; Leone et al., 1985), much
additional research is needed before a complete understanding of the
complex N0x-photooxidation chemistry of aromatic compounds is obtained.
3-32
-------
3.3.2.1.4 Aldehydes. Aldehydes are consumed in the atmosphere both by
photolysis and attack by OH radicals. Photolysis of acetaldehyde, for
examples leads to methyl and formyl radicals that then react as discussed
above:
CHgCHO + hv ->• CH3 + HCO (3-50)
Attack by OH radicals forms acetyl radicals which can successively add Q£
and N02 to form PAN:
+ OH * CHjCO + HgO (3-51)
CHgCO + 02 * CHgCOO* —>• CHgCOONOg ' (3-52)
3.3.2.1.5 Nitrogen-Containing Compounds. Only limited information is
available for the reactions of OH radicals with nitrogen-containing
compounds and their subsequent reactions under atmospheric conditions.
The OH radical reactions with the aliphatic amines are rapid, with room
1 1 *3 -1
temperature rate constants being in the range (2-6) x 10 cm molecule"1
sec"-'- (Atkinson et al.» 1985). For the methyl-substituted amines, the
trend of the room temperature rate constants suggests that these reactions
proceed via abstraction from the C-H bonds and, where possible, the N-H
bonds. Product studies of several irradiated amine-air systems have been
reported in which plausible reaction,pathways following OH radical attack
have been proposed (Pitts et a!., 1978; Tuazon et al., 1978b; Lindley et
al., 1979).
To date, kinetic data are available only for OH radical reactions
with hydrazine and methylhydrazine (Atkinson, 1985), and only limited
product data are available for these reactions (Tuazon et al., 1981b,
1982). Reactions of nitrites with OH radicals are expected to proceed via
H-atom abstraction from the C-H bonds but, since no product data are
presently available, no reliable assessment of the initial reaction
pathway can be made.
No product or direct mechanistic data are available for organic
nitrates. Howevers the reactions of OH radicals with at least the smaller
3-33
-------
alky! nitrates (for which isomerization of the alkoxy radicals cannot
occur) will probably ultimately yield N0£ together with the corresponding
aldehydes (Atkinson et al., 1982d). These reactions may be of importance
in long-range transport and acid deposition, since alky! nitrates are
formed in significant yields from the atmospheric photooxidation of
certain alkanes (Atkinson and Lloyd, 1984).
3.3.2.1.6 Sulfur-Containing Compounds. The literature concerning reac-
tions of sulfur-containing compounds with OH radicals has been reviewed
recently by Atkinson (1985). Reactions of thiols with OH radicals must
proceed via either H-atom abstraction from the weak S-H bonds or, more
likely, by the formation of an QH-thiol adduct. The reaction of OH
radicals with the sulfides, RSR» can proceed via either H-atom abstraction
from the C-H bonds or OH radical addition to the sulfur atom. Product
data for the reaction of OH radicals with dimethyl sulfide under
atmospheric conditions have been obtained from numerous studies (Srosjean
and Lewis, 1982; Hatakeyama et al., 1982; Hatakeyama and Akimoto, 1983;
Niki et al., 1983; Grosjean, 1984); the major stable products are HCHO,
S02 and CHgSOgH, together with CHgSNO as an intermediate product.
Only for dimethyl disulfide have kinetic (Cox and Sheppard, 1980;
Wine et al., 1981) and product (Hatakeyama and Akimoto, 1983) data been
reported. On the basis of these data, it appears that the initial
reaction proceeds via OH radical addition to form an adduct, followed by
rapid decomposition of this adduct to CH3S and CH3SQH radicals.
Subsequent reactions of these CHgSOH and CH3S radicals then lead to the
observed products: S02, HCHO, and CH3S03H.
3.3.2.2 Reactions with Ozone. The atmospheric reactions of ozone are
complex and result in products and processes that have significant
environmental implications, including effects on biological systems,
visibility, and materials. Ozone, for example, is highly reactive towards
certain classes of organic compounds (e.g., alkenes) and certain of those
reactions lead to the formation of secondary organic aerosols. Ozone may
also play a role in the oxidation of S0£ to H^SO*, both indirectly in the
gas phase (via formation of OH radicals and Criegee biradicals) and
directly in aqueous droplets.
In the following sections, the atmospheric reactions of ozone with
organic compounds are summarized in some detail, including the mechanisms
3-34
-------
of certain of these reactions. Emphasis is placed, whenever possible, on
those reactions that lead to products or processes suspected or known to
have effects on biological or other important receptors.
In discussing the reactions of ozone with organic compounds in the
troposphere, it is important to recognize that organics undergo competing
reactions with OH radicals during daytime hours (Atkinson and Lloyd, 1984;
Atkinson, 1985) and, in certain cases, they can photolyze or react with
N03 radicals at night (Japar and Niki, 1975; Carter et al., 1981a;
Atkinson et al., 1984a,b,c,d; Winer et al., 1984). All organics except
the perhaloalkanes exhibit room temperature OH radical rate constants "of
>5 x 10"15 cm3 molecule"1 sec"1 (Atkinson, 1985). Since the ratio of 03
to OH radical concentrations in the unpolluted troposphere during daylight
hours is believed to be of the order of 106 (Singh et al., 1978; Crutzen,
1982), only for those organics whose 03 reaction rate constants are
pi O 11
greater than ~10 cm molecule sec can consumption by 03 be
considered atmospherically; important. These ozone reactions of interest
are summarized below.
3.3.2.2.1 Alkenes. Ozone reacts rapidly with the acyclic mono-, di- and
trialkenes and cyclic mono-, di-, and tri-alkenes. The rate constants for
these reactions range from ~1Q"18 to ~10"^ cm molecule" sec" (Atkinson
and Carter, 1984), corresponding to atmospheric lifetimes ranging from a
few minutes (for the more reactive cyclic alkenes such as the
monoterpenes) to several days. In polluted atmospheres, especially in the
afternoons during photochemical oxidant episodes, the principal
consumption of the more reactive alkenes will therefore occur via reaction
with 03, rather than with OH radicals.
It is now reasonably well established that the initial step in the
Q-j-alkene reaction involves the formation of a "molozonide" that rapidly
decomposes (Harding and Soddard, 1978; Herron et al., 1982) to a carbonyl
compound and a biradical (which is also initially energy rich):
3-35
-------
OO
.0-0
V V 3
>-<
V R4
i>
(3-53)
0-0'
X
Ci-C
where C 3 denotes an energy-rich species.
Based on an analysis of reported product and mechanistic studies of
the reactions of 03 with ethene (Herron and Huie, 1977; Su et al., 1980;
Kan et al., 1981; Niki et al.» 1981) and propene (Herron and Huie, 1978;
Dodge and Arnts, 1979), and the much less extensive studies of the higher
alkenes (Martinez et al.» 1981), Atkinson and Lloyd (1984) have suggested
that these initially energy-rich biradicals react under atmospheric
conditions as shown below:
M
CH200
CH
CH
/'
0*.
[HCOOH]
,/
X
H + HC02
(40%)
C02 -I- H2
CO + H90
(12%)
(42%) (3-54)
and
20,
(6%)
3-36
-------
[CH3CHOO]q
CHjCHOO
CH
(40%)
0 CHg + CO + OH (19%)
[CH^COH]*^
^ CH3 + C02 + H (24%)
CH4 + C02
(3-55)
(5%)
(12%)
where CH^OO and CHoCHOO denote thermalized biradicals. These thermal lied
biradicals have been shown (Calvert et'al., 1978; Herron et a!., 1982;
Atkinson and Lloyd, 1984) to undergo bimolecular reactions with aldehydes,
S0o> CO, and HoO, and it is believed that they will also react with HQ and
W$ (Calvert et al., 1978; Herron et al., 1982; Atkinson and Lloyd, 1984):
RCHOO + HQ + RCHO + HQ,
RCHOO + HQ2 > RCHO +
(3-56)
(3-57)
RCHOO + S02—> RCHO
RCHOO + H20 > RCOOH +
RCHOO + CO > products
(3-58)
(3-59)
(3-60)
RCHOO + R'CHO
(3-61)
Under atmospheric conditions, the reactions with HQt N02> or HgO are
expected to be the dominant loss processes of these thermalized
biradicals, with the precise major reaction pathway depending on the
relative concentrations of NO, N0 or Q (Atkinson and Lloyd, 1984).
3-37
-------
Hence, Og-alkene reactions in the atmosphere can lead ultimately to the
formation of aldehydes and acids, as well as to the conversion of SO^ to
H«S(L» although the latter is probably a minor process in the overall
oxidation of SO- during long-range transport (Fin!ayson-Pitts and Pitts,
1982).
Isoprene and monoterpenes are alkenes emitted from vegetation. The
role of these compounds in photochemical air pollution has been a subject
of discussion, and sometimes controversy, for more than a decade. The
case of ozone reactions with compounds such as isoprene and the monoterpenes
is particularly interesting since these reactions can represent a sink for
ozone as well as for the hydrocarbons themselves. This adds complexity
to an overall assessment of the role of hydrocarbons emitted from
vegetation since, depending upon the specific atmospheric conditions,
they may be both sources and sinks for ozone (Dimitriades, 1981; Altshuller,
1983).
Again, as in the case of their OH radical reactions, the detailed
reaction sequences following reaction of 0, with iosprene and the
monoterpenes are not well understood although substantial kinetic data are
available (Atkinson and Carter, 1984); thus only a brief summary of
available information is presented here.
Isoprene. The reaction of Q3 with isoprene (Kamens et a!,, 1982;
Lloyd et a!., 1983; Kill us and Whitten, 1984) leads to molozonides that
are presumed to decompose subsequently into stable products and radical
intermediates analogous to those produced in other ozone-alkene reactions
described earlier in this section. The major products, methlvinylketone
and methacrolein, undergo further reactions with OH radicals and ozone
(Lloyd et al., 1983; Killus and Whitten, 1984).
The reaction of ozone with isoprene can also lead to aerosol
formation (Kamens et al., 1982) and this is discussed in Section 3.3.4.
Honoterpenes. The detailed pathway for the reaction of 0, with the
monoterpenes under atmospheric conditions is unknown. Lloyd et al. (1983)
have proposed a reaction sequence for crpinene involving addition of 0, to
the double bond to form a molozonide, with subsequent ring opening. They
note, however, that this reaction sequence is entirely speculative, and
3-38
-------
many additional kinetic and mechanistic data will be required to elucidate
the detailed reactions of 03 with the monoterpenes.
9"\
3.3.2.2.2 ATkanes and A1 kynes. Given reported rate constants of 10~" to
10"^ cm3 molecule"1 sec"1 (Atkinson and Carter, 1984), there appears to
be no convincing evidence in the literature for an elementary reaction
between 03 and the alkanes. Similarly, although there is presently
substantial uncertainty concerning the rate constants for the reactions of
ozone with the simple alkynes (e.g., acetylene, propyne and 1-butyne),
most of the available room temperature data for these 50^ alkynes indicate
03 rate constants in the range of ~10"20 to i(T19 cm3 molecule'1 sec"1
(Atkinson and Carter, 1984). Thus, those reactions will also be
unimportant in the atmosphere since the corresponding OH radical rate
1Q _i p -a
constants are, for example, ~8 x 10 for acetylene and ~7 x 10 cm
molecule" sec"1 for propyne and 1-butyne (Atkinson, 1985).
3.3.2.2.3 Aromatics. As in the case of the alkanes, the aromatic
hydrocarbons react only very slowly with 03 (Atkinson and Carter, 1984)
and these reactions are not expected to be important in the atmosphere^
Although the cresols are significantly more reactive than the aromatic
hydrocarbons (Atkinson and Carter, 1984), under atmospheric conditions
their reactions with 03 are minor compared to their reactions with OH
radicals (Atkinson et a!., 1978) or N03 radicals (Carter et a!., 1981a).
3.3.2.2.4 Oxygen-Containing Organlcs. For those oxygen-containing
compounds that contain no unsaturated carbon-carbon bonds (e.g.,
formaldehyde, acetaldehyde, glyoxal, and methylglyoxal) the reactions with
ozone are very slow, and, by analogy, this is expected to be the case for
all ethers, alcohols, aldehydes, and ketones containing no unsaturated
carbon-carbon bonds. For the carbonyls and ethers (other than ketene)
that contain unsaturated carbon-carbon bonds, however, much faster
reactions are observed (Atkinson and Carter, 1984).
Few data are available, however, concerning the mechanisms of the
reactions of 03 with such oxygen-containing organics, the only published
information being that of Kamens et al. (1982). From a study of the
reactions of 03 with methacrolein and methylvinylketone, methylglyoxal was
observed as a product, along with other minor products (Kamens et al.,
1982), as anticipated from the reaction schemes:
3-39
-------
CH2=CHCOCH3+
HCHO + [CH,CQGHOQ]*
CH2-CHCOCH3
(3-62)
ChLCOCHO + [CH-OOT
•5 , £
(3-63)
and
HCHO
CHO
CHO
I •
CH3COO
'CHO
\
(3-64)
CH3COCHO + [CH200
Kamens et al. (1982) discuss the possible subsequent reactions.
3^3,2.2.5 N11rogen-Containing Organics. Studies of the kinetics of the
reactions of 03 with a variety of nitrites, nitriles, nitramines, nitroso-
amines, amines, and hydrazines (Atkinson and Carter, 1984) indicate that
only for the hydrazines are reactions with 03 sufficiently rapid to be of
atmospheric importance (Carter et al., 1981b; Tuazon et al., 1982).
3.3.2.2.6 Sul f ii r-Contai ni ng Organi cs. Based upon the kinetic data
available for dimethyl sulfide, thiirane (C2H^S) and thiophene, it appears
at the present time that the rates of reaction of Og with sulfur-
containing organics can be considered to be unimportant under atmospheric
conditions (Atkinson and Carter, 1984).
3.3.2.2.7 Organometallies. Rate constants have been reported only for
tetramethyl- and tetraethyl-lead (Harrison and Laxen, 1978), and no
mechanistic or product data are available for these reactions.
3.3.2.2.8 Radical Species. Because of the low concentration of 03 and
of both alky! and most alkoxy radicals in the atmosphere, and because
these radicals react at significant rates with 02 (which is present at a
concentration >105 higher than 03 in ambient atmospheres), the reactions
of ozone with such radical species can be considered to be of negligible
importance in the atmosphere. Of course, the reaction of ozone with the
hydroxyl radical must be considered, as discussed earlier.
3-40
-------
3.3.2.3 Reactions with Nitrate Radicals. The pioneering work of Nikl and
coworkers (Morris and Niki, 1974; Japan and Niki, 1975) showed that
gaseous N03 radicals react with alkenes, with the rate constants
increasing markedly with the degree of substitution on the double bond.
Carter et al. (1981a) showed that the hydroxy-substituted aromatics
(phenols and the cresols) also react rapidly with N03 radicals.
More recently, Atkinson and coworkers have investigated the kinetics
of the reactions of N03 radicals with a wide range of organics at room
temperature (Atkinson et al,, 1984a-e) and from these and the earlier
studies, information concerning the mechanisms of these reactions has been
forthcoming.
In the remainder of this section, current understanding of the
mechanisms of reaction of NQ3 radicals with the various classes of
organics is briefly summarized.
3.3.2.3.1 Alkanes. The relevant kinetic data (Atkinson et al . , 1984e) in-
dicate that these reactions proceed via H-atom abstraction from the C-H
bonds, almost certainly predominantly from secondary or tertiary C-H bonds:
. . . N03 + RH * HN03 + R' (3-65)
Hence, these reactions lead directly to HN03 formation. The measured room
17 "\
temperature rate constants for the alkenes range between 3.6 x 10 cm
molecule"1 sec'1 for jrj-butane to 2.2 x 1Q"16 cm3 molecule""1 sec"1 for 2,3-
di methyl butane.
3.3.2.3.2 Alkenes. The reactions of N03 radicals with the alkenes have
been shown from both kinetic (Japar and Niki, 1975; Atkinson et al.,
1984a) and product (Bandow et al., 1980) studies to proceed via initial
't '
addition of the N03 radical to the olefinic double bond:
ON00
I 2
+ 2CH3CH=CH2 * CH3CHCH2 + CH3CHCH2ON02, (3-66)
with addition at the terminal carbon expected to dominate (Atkinson and
Lloyd, 1984). Possible reaction sequences beyond this initial reaction
' I
have been discussed (Bandow et al., 1980; Atkinson and Lloyd, 1984) but
are highly uncertain at the present time (Atkinson and Lloyd, 1984).
3-41
-------
Thermally unstable nitro-peroxynitrates such as CH3CH{ON02)CH2OON02» and
stable dinitrates such as CH3CH(ON02)CH2ON02 have been reported as
products in NC^-NC^-propene-air systems (Bandow et al., 1980).
Monoterpenes. While no mechanistic information is available
concerning the reactions of N03 radicals with the monoterpenes, N03
radical reaction rate constants have recently been reported for a
substantial number of these compounds (Atkinson et al., 1984c, 1985).
Based in part on these kinetic data, Winer et al. (1984) have proposed
that reaction with N03 radicals at night may be an important reaction
pathway for certain naturally occurring organics such as the monoterpenes
and dimethyl sulfide. Conversely, these kinetic data also show that
reactions with the more reactive alkenes, including isoprene and certain
of the monoterpenes, as well as dimethyl sulfide and the hydroxy-
substituted aromatics, can be important loss processes for N03 radicals at
night (Winer et al., 1984). The importance of N03 radical reactions in
determining the atmospheric lifetimes of the monoterpenes is discussed in
section 3.3.3.
3.3.2.3.3 Aldehydes. Based upon the product data of Morris and Niki
(1974), i.e., the observed formation of HN03 from the reaction of N03
radicals with CH3CHO, it is expected that these reactions proceed via H-
atom abstraction from the relatively weak H-CO bonds:
N03 + RCHO * RCO + HN03 (3-67)
Thus the reaction of N03 radicals with acetaldehyde could be a nighttime
source of peroxyacetyl nitrate (PAN):
NO, + CH^CHO -» HNO.+ CH-CO (3-68)
O O *5 O
CH3CO + 0£ * CH3CO* (3-69)
?
CH-CO " + NO, * CHoCOON09 (3-70)
O O £ «3 £
Reaction of N03 radicals with the higher aldehydes will lead, by analogous
reaction schemes, to the higher peroxyacyl nitrates, RC03N02. Reaction
3-42
-------
with formaldehyde, however, will lead to H02 radical formation, since HCO
reacts rapidly with 02 (Atkinson and Lloyd, 1984):
HCO + 0 -> H0
CO
(3-71)
3.3.2.3.4 Aromatics. As discussed by Atkinson et al. (1984d), the
reactions of NOo radicals with the monocyclic aromatic hydrocarbons and
the hydroxy-substituted aromatics appear, based upon kinetic evidence, to
proceed via H-atom abstraction from the C-H or 0-H bonds on the .«
substituent groups. This conclusion is based upon the observation that
for the xylenes and the eresols the meta-isomer reacts more slowly (by a
factor of ~2) than the ortho- and para-isomers. This is in contrast to
the addition reactions of 0( P) atoms and OH radicals, in which the meta-
isomer is the most reactive (Atkinson, 1985). Furthermore, ^-nitrophenol
has been tentatively identified as a product of NoOg-NOg-phenol-air
reaction mixtures, presumably formed by the reaction sequence:
HN0
(3-72)
followed by (Nik1 et al., 1979):
OH OH
,NQ0
- NO,
(3-73)
Thus, these reactions can also be a direct source of nitric acid as well
as forming low-volatility organic nitro compounds.
In summary, it is now clear that reaction with NQ3 radicals at night
is a major atmospheric reaction pathway for many organic pollutants. It
must therefore be considered, along with the reactions of OH radicals and
03 and photolysis, as one of the dominant loss processes for organics in
the atmosphere.
3-43
-------
3.3.3 _Atmosjgher1 c LIfetImes of Organi c Compounds ;
Table 3-6 compares the atmospheric lifetimes (i.e., the time to reach
1/e of the initial concentration) for selected organic compounds, arising
from manmade and natural sources, as the result of reaction with 03 over a
24-hour period, with OH radicals during the day, and with N03 radicals at
night. It can be seen that under the atmospheric conditions assumed reac-
tions with Og are important for the higher alkenes, including the monoter-
penes during daylight when N03 radical concentrations are low, and for the
hydrazines. 'For the other organics for which kinetic data are available,
including alkanes and aromatics, reactions with 63 are generally of negli-
gible or minor importance in determining their atmospheric lifetimes.
TABLE 3-6. CALCULATED LIFETIMES OF SELECTED ORGANIC COMPOUNDS RESULTING
FROM ATMOSPHERIC LOSS BY REACTION WITH Do AND OH AND NO, RADICALS
o , %*
Organic lifetimes3*
Organic compound
Alkenes from manmade sources
Ethene
Propene
trans-2-Butene
2-Methyl-2-butene
2»3-Di methyl -2-butene
Naturally emitted alkenes
Isoprene
a-Pinene
p-Pinene
A -Carene
d-Limonene
24 hr
2.7 days
11 hr
35 min
17 min
6 min
10 hr
1.4 hr
5.5 hr
1.0 hr
11 min
OH,
daytime ,
16 hr
5.6 hr
2.0 hr
1.6 hr
1.3 hr
1.4 hr
2.3 hr
1.8 hr
1.7 hr
1.0 hr
N03,
nighttime
79 days
1.1 days
33 min
1.3 min
0.2 min
22 min
2 min
5 min
1.2 min
0.9 min
aT1me to reach 1/e of the initial concentration.
Assuming 100 ppb of 03 (24 hr average), 2 x 106 molecule cm"3 (0.08 ppt)
of OH radicals during daylight hours, and 100 ppt of N03 radicals during
nighttime hours, and at room temperature.
3-44
-------
3.3.4- Atmospheric Reactions of Peroxyacetyl Nitrate ,
With the recognition in recent years that PAN is a ubiquitous
nitrogenous species in the troposphere (Singh and Hanst, 1981; Aikin et
al., 1983; Penkett, 1983; Singh and Salas, 1983; Spicer et al., 1983) and
in the lower stratosphere (Aikin et al., 1983), there has been renewed
.focus on the atmospheric role of this organic compound.
Smog chamber studies have shown that, once formed, PAN can be
relatively stable under atmospheric thermal conditions (Pitts et al.,
1979; Akimoto et al,., 1980). Since PAN is in equilibrium, however, with
acetylperoxy radicals and N02:
8 8
CH3COON02 j CH3COO + N02 (3-74)
any process that removes either acetyl peroxy radicals or N02 will lead to
the decomposition ,of PAN. One such process is the reaction of NO with
CH3C(0)02 radicals. Because PAN has been shown to persist through the
night in urban atmospheres (Tuazon et al., 1980, 1981a) the reaction of
PAN with NO during the morning traffic peak can lead to the formation of
OH radicals via the foil owing,mechanism (Cox and Roffey, 1977; Garter et
al., 1981c): - . .
0 0
CH-COO* + NO -> CH,CO" + N09 (3-75)
*3 _ *3 c*. '• •' .
o
CH3C-0* + CH3* + C02 (3-76)
M
CH3* + °2 * CH3DO (3-77)
CH300* + NO -» CH30* + N02 (3-78)
CH^O- + 09 * HO, + HCHO (3-79)
*5 <•., <— ' , -
H02 + NO * OH + N0£ (3-80)
3-45
-------
M
OH + NO -> HONO (3-81)
M
OH + N0£ -> HN03 (3-82)
Thus, the reaction with NO of PAN carried over from previous air
pollution episodes will lead to enhanced smog formation on subsequent
days. This enhancement in reactivity results both from the fact that
these reactions form radicals that initiate the transformations occurring
in photochemical smog and from the fact that these reactions convert NO to
N02, which allows earlier formation of 03 and higher levels to be
attained. It should be noted that this enhancement will result even if
all of the PAN reacts with NO emitted at nighttime, since the NO
conversion does not require sunlight and since at least some of the
radicals formed will be "stored" as nitrous acid, to be released when
photolysis begins at sunrise (Harris et al., 1982).
These results could have important implications regarding multiday
photochemical pollution episodes in which significant buildup of PAN is
observed. Under such conditions, the carry-over of PAN may be a
significant factor in promoting ozone formation on subsequent days and
may, in part, contribute to the progressively higher 03 levels often
observed during such episodes (Tuazon et al., 1981a).
A second important role of PAN is its ability to contribute to the
long-range transport of NOX. In the absence of significant levels of NO
(i.e., in the cleaner troposphere) and in regions of lower temperature and
in the upper troposphere, when the thermal decomposition of PAN becomes
unimportant, the atmospheric lifetime of PAN will be determined by its
reaction with OH radicals. This reaction is sufficiently slow (Wellington
et al., 1984) that PAN will probably be long-lived and hence serve as a
reservoir for odd nitrogen in a manner analogous to HN03 (Aikin et al.,
1983). Also analogous to the case for HN03, dry deposition of PAN may be
a significant loss process in cleaner atmospheres.
3-46
-------
3*3.5 Role of Ozone In AerosolFormation
In addition to having direct effects on human health and on vegeta-
tion, ecosystems, and nonbiological materials, ozone can contribute
indirectly to visibility degradation and to acidic deposition through its
participation in the formation of both organic and inorganic aerosols
(National Research Council, 1977a; U. S. Environmental Protection Agency,
1982b).
3.3.5.1 Formation of SulfateAerosol. It is well established that the
source of the vast majority of manmade sulfate aerosol in the atmosphere
is the oxidation of sulfur dioxide (S02), ultimately to sulfuric acid
(H2SQ,|). The correlations between elevated levels of ozone and of sul fate
aerosol in ambient air have been noted by several investigators in field
studies concerned with visibility reduction by aerosols. Wilson (1978)
and Gillani et al, (1981) have pointed out that atmospheric mixing
intensity and the background 03 concentration are the two most important
factors in determining SOo oxidation at relative humidities lower than 75
percent. It is also clear, however, that the rate of reaction of Do with
S02 is far too slow to account for observed formation rates of sulfate
aerosol (U. S. Environmental Protection Agency, 1982b).
Of the many possible gas-phase reactions of S0£, only a few appear to
have any significance in the production of sulfate aerosol and the
reaction of OH radicals with SQ2 appears to be the dominant pathway for
the oxidation of S02 (Calvert and Stockwell, 1983, 1984; Calvert and Mohnen,
1983). A recent analysis by Stockwell and Calvert (1983) shows that the
formation of HOS02 radicals from the reaction of OH radicals with S02,
followed by reaction with 02, is the reaction mechanism for the formation
of S0q:
s>
OH + S02 (+ M) -> HOS02 (+ M); (3-83)
HOS02 + 0£ * H02 + S03. (3-84)
From the reaction of H02 radicals with NO, OH radicals are regenerated and
the cycle begins again as discussed earlier in Section 3.3.1.2,1.
3-47
-------
The importance of the reaction of OH radicals with S02 in the
atmosphere is supported by observations of power plant plumes, in which no
aerosol is formed at night when the OH radical concentration in ambient
air is negligible; and none is formed during the day before the plume is
well mixed with ambient air (the ambient air contains much higher
concentrations of OH radicals and 03 than the plume) (Davis et al., 1979;
Blumenthal et al., 1981).
Though it does not react directly with S02 at an appreciable rate, by
virtue of its role in OH radical production 03 plays an important indirect
role in the transformation of S02 to sulfate aerosol via the homogeneous
oxidation of S02 in both clean and polluted atmospheric systems. Ozone
plays a further role in the oxidation of sulfur in aqueous droplets as
discussed later in this section.
3.3.5.2 Formation of Nitrate Aerosol. Despite limited relevant data,
the possible contribution of nitrate aerosol to visibility reduction
should not be neglected and the role of 03 in the formation of this
aerosol species is briefly considered here.
The principal manmade nitrogen emissions in this case are NO and
N02. Nitric oxide is relatively insoluble in aqueous systems (Section
3.2) and does not react with water in any significant manner. Thus, NO
must be converted to a more highly oxidized form, for example N02, in
order to participate in the formation of particulate nitrate.
The oxidation of NO to N02 can occur through thermal oxidation at
high concentrations of NO such as those in and very near the stacks of
power plants (U. S. Environmental Protection Agency, 1982a). This
generates only a small portion of the N02 formed in the atmosphere,
however. As previously discussed (Section 3.3.1), the most important
reactions leading to formation of N02 in ambient air are the reaction of
NO with 03 and the oxidation of NO to N02 by hydroperoxyl radicals and
other peroxy radicals (reactions 3-3 and 3-11, respectively). Thus, if
N02 is a precursor of nitrate aerosol, 03 plays a significant direct role
in its formation by oxidizing NO, and an indirect role by leading to
formation of OH radicals (Section 3.3.1).
As discussed earlier, N02 can be converted in the gas phase to nitric
acid (HN03) vapor by reaction with OH radicals during the day or by
3-48
-------
reaction with Q3 to form NO, radicals, which at night are in equilibrium with
NgOg. As shown in Section 3.3.1.5, homogeneous or heterogeneous hydrolysis, or
both, of NgOg is an important nighttime pathway to nitric acid formation. Once
it has been produced in the gas phase, HNQ3 is sufficiently volatile to remain
in the atmosphere as a vapor. The available laboratory and ambient air data
indicate, however, that HN03 vapor reacts with ammonia in a reversible reaction
to form NH4NQ3 (Doyle et al., 1979; Stelson et al., 1979; Appel et al., 1980),
which, because of its low vapor pressure, will form nitrate aerosol particles:
HN03 + NH3 -» NH4N03 (3-85)
If acidic sulfate is present, however, it will react with NH,N03 to form HN03
again. Consequently, reaction (3-85) is not a major sink for nitric acid in
areas with high sulfate loading, such as the eastern United States. Evidence
also indicates that HN03 vapor will .react with NaCl aerosol in the following
way: ' . " -
HN03 + N.aCl •* NaN03 + HC1 (3-86)
This second reaction (equation 3-86) may account for the fact that much of the
observed particulate nitrate in Los Angeles is. found in the coarse mode (Farber
et al., 1982). Obviously, the importance of this mechanism for nitrate aerosol
formation is determined by the availability of sea salt particles.
3.3.5.3 Formation of Organic Aerosols. Sulfate and nitrate aerosols are
present at significant levels in the atmosphere in the form of just a few
compounds. In contrast, secondary organic aerosols are composed of a large
number of species, but there is no clear consensus concerning which ones
contribute most to the mass concentration. For all the species that are found
in the secondary organic aerosol,. however, the fundamental formation mechanism
is the same. The vapor-phase precursor undergoes some reaction that results
in formation of a product having an equilibrium vapor pressure sufficiently
low that condensation, nucleation, or both are possible at the gaseous concen-
tration achieved. From the available data, it seems clear that the more highly
oxygenated, larger-carbon-number species generally are those precursors likely
to form secondary aerosols in the atmosphere.
3-49
-------
For an earlier but thorough review of the formation of secondary
aerosol, the reader is referred to the 1977 monograph on ozone and other
photochemical oxidants prepared by the National Research Council
(1977a). This monograph reviews the reactions of manmade volatile organic
compounds that produce aerosol. Biogenic as well as manmade volatile
organic compounds, however, can participate in aerosol formation
(Altshuller and Bufalini, 1971; Arnts and Gay, 1979). Direct experimental
evidence of aerosol formation, along with product analysis, is only avail-
able, though, for a limited number of natural compounds, including iso-
prene (Kamens et al., 1982) and a-pinene and p-pinene (Schwartz, 1974;
National Research Council, 1977a; Hull, 1981), mainly because the
analysis and characterization of these kinds of products at ambient con-
centrations is extremely difficult. Hull (1981) has conducted experiments
with a- and p-pinene at high concentrations in a small tube reactor.
Analysis of the products showed, on a weight basis, that almost all of the
reacted a-pinene carbon was found in the condensed materials extracted
from the walls. Although the products identified from these experi-
ments were either in the condensed phase or on the walls, Hull suggested
that at the a-pinene levels found in ambient air these products have a
high enough vapor pressure to exist both in the gas phase and in aerosols
(Hull, 1981). In a recent review of the role of biogenic volatile organic
compounds, Altshuller (1983) has discussed at length the contribution of
these compounds to ambient air aerosols.
3.3.6 Role of Ozone and Other Photochemical Oxidants in the Acidification
of Rain
Two recent criteria documents prepared by the U. S. Environmental
Protection Agency (1982a; 1982b) contain thorough discussions of the
contributions of ozone and of hydrogen peroxide, also an oxidant, to the
oxidation of S02 and the roles of S02 as a precursor to acidic deposition..
3.3.6.1 Reactions of Ozone in Aqueous Droplets. While the thermal oxidation
of SCL by ozone in the gas phase appears to be too slow to be important in
acid deposition phenomena, the role of ozone in oxidizing SC^ dissolved in
water droplets (e.g., cloud, fog or rain) may be of considerable
significance. At 25°C ozone has a Henry's law constant of 10"^ mol L~^
(Kirk-Othmer, 1981). Given ambient concentrations ranging from 30
3-50
-------
to ~300 ppb, Oj would be expected to have concentrations in aqueous
droplets in the atmosphere of approximately 3-30 x 10"10 mol L"1. The
rate of reaction between 03 and S02, when both are dissolved in aqueous
droplets, has been shown in laboratory studies to be relatively fast
(Penkett et al., 1979; Kunen et al., 1983; Brock and Durham, 1984; Hoffman
and Jacob, 1984; Martin, 1984; Schwartz, 1984), but the rate of this
reaction is pH dependent and decreases as the acidity of the solution
increases.
Figure 3-3 shows data reported by Schwartz (1984) for the rate of the
aqueous-phase oxidation of S(IV) by 30 ppb of Og (and also by 1 ppb of
I^C^) as a function of solution pH, The aqueous-phase oxidation rate, R,
per part-per-billion S02 partial pressure decreases with decreasing pH by
roughly a factor of 20 per pH unit. This pH dependence reflects the
solubility of S(IV), as well as a slight pH dependence of the second-order
rate constant for the oxidation of S(IV) by 03 (Erickson et al,, 1977;
Larson et al., 1978; Penkett et al., 1979). Schwartz (1984) concluded,
from consideration of these data and uptake times for S0g» that the
oxidation of SOg by Og cannot produce solution pH values below ~4.5.
Schwartz (1984) has also, however, interpreted the field data of Hegg and
Hobbs (1981) for sulfate production rates at the inflow and outflow
regions of lenticular clouds as being consistent with the aqueous phase
oxidations of S(IV) by 03.
An additional aspect of the role of Oj in the chemistry of aqueous
droplets concerns its photolysis to yield OH radicals in solution (Graedel
and Weschler, 1981; Chameides and Davis, 1982):
(03)aq + hv -» 0(1D)aq + 02 aq (3-87)
OtXq + (H2°)aq * 2(°H)aq (3-88)
and its reactions with aqueous OH" ions and #2®2 to yield aqueous H02
radicals (Chameides and Davis, 1982). The OH radicals formed by this
aqueous process can result in the oxidation of S(IV).
3.3,6.2 Reactions of Hydrogen Peroxide in Aqueous Droplets. Although
hydrogen peroxide formed in the gas phase from the reactions of
3-51
-------
10"
10"'
o
CO
Q.
a.
"L
£
to
o.
QC
10'8
H2O2, 1 ppb
1000
100
10 ^
o
X
a
0,1
0.01
Figure 3-3. Rate of aqueous-phase oxidation
of S(IV) by O3 (30 ppb) and H2O2 (1 ppb), as a
function of solution pH. Gas-aqueous equilibria
are assumed for all reagents. R/pSO2 rep-
resents aqueous reaction rate per ppb of gas-
phase SO2; p/L represents rate of reaction
referred to gas-phase SO2 partial pressure per
cm3—nr3 liquid water volume fraction.
Source: Schwartz (1984).
3-52
-------
hydroperoxyl radicals plays a role in HOX chemistry in the troposphere,
and especially in the stratosphere (Crutzen and Fishman, 1977; Cox and
Burrows, 1979), the major importance of hydrogen peroxide arises from its
high solubility in water. The latter ensures that a large fraction of
gaseous H2Q2 will be taken up in aqueous droplets. Over the past decade,
evidence has accumulated that H202 .dissolved in cloud, fog, and rainwater
may play an important, and, in acidic droplets (i.e., pH <5), even a
dominant role in the oxidation of S02 to H2SO/j. (Hoffman and Edwards, 1975;
Penkett et a!., 1979; Dasgupta 1980a,b; Graedel and Weschler, 1981; Martin
and Damschen, 1981; Chameides and Davis, 1982; Calvert and Stockwel1,
1983, 1984; Brock and Durham, 1984; Hoffman and Jacob, 1984; Schwartz,
1984). Discussion of several proposed mechanisms for previous rate
studies of the oxidation of S(IV) by H202 are beyond the scope of this
document, but have recently been reviewed by several authors (see for
example, Calvert and Stockwell, 1983, 1984). Hydrogen peroxide may also
play a role in the oxidation of NQ2 dissolved in aqueous droplets,
although relevant data are limited (Halfpenny and Robinson, 1952a,b; Anbar
and Taube, 1954; Gertler et al., 1984) and additional research is
required. In addition to the direct oxidation of S02 and N02 dissolved in
aqueous droplets, the photolysis of H202 to produce aqueous OH radicals:
(H202)aq + hv •+• 2(OH)aq (3-89)
can lead to oxidation rates of S(IV) that can be competitive with calcu-
lated oxidation rates of S(IV) by (H202) and (03)a(. (Chameides and
Davis, 1982).
It should be emphasized, however, that substantial uncertainties
remain concerning the quantitative role of H202 in the acidification of
aqueous particles and droplets (Richards et al., 1983). This is further
complicated by the lack of reliable measurements of gas-phase H202
concentrations in the atmosphere (see Chapter 4). Moreover, it has also
been suggested recently that H202 may be formed in situ in aqueous
droplets as the result of absorption of OH and H02 radicals and other
precursors into solution from the gas phase (Graedel and Weschler, 1981;
Chameides and Davis, 1982; Heikes et al,, 1982).
3-53
-------
3.3.6.3 Reactions of Formic Acid in Aqueous Droplets. As a gas-phase species,
formic acid (HCQQH) cannot strictly be defined as a photochemical oxidant. It
can be scavenged rapidly into water droplets, however, and can potentially
function therefore as an oxidant in cloud water and rain water. It can also
be differentiated from other acids in that it is formed readily from the
reactions of the Criegee intermediates discussed earlier and of hydroperoxyl
radicals with formaldehyde (Calvert and Stockwell, 1983). The formation of
other acids may be orders of magnitude slower as the result of both the
apparently lower rates of reaction of H02 radicals with the higher aldehydes
and the much lower atmospheric concentrations of the higher aldehydes (Grosjean,
1982). Thus, formic acid is an example of a compound that is a non-oxidant or
weak oxidant in the gas phase but that is transformed upon incorporation in
aqueous solutions into an effective oxidizer of S(IV).
Formic acid (as well as acetic acid) has been identified among the acidic
components of rain (Galloway et al., 1982). Although much uncertainty remains
concerning their quantitative roles, HCOOH and the higher organic acids potentially
play a minor but still significant role in the acidification of rain.
3.4 METEOROLOGICAL AND CLIMATOLOGICAL PROCESSES
As discussed in Section 3.3, ozone and oxidants are formed by the action
of sunlight on the precursors, HQy and hydrocarbons. The accumulation of the
products to form an appreciable concentration is also dependent, however, on
the prevailing meteorology in the vicinity of the precursor emissions. To
understand the details of the effects of meteorology on air quality requires a
thorough knowledge of meteorology and climatology, but an appreciation of the
general factors important in the formation of elevated concentrations of
oxidants is relatively easy to acquire. Following is a brief presentation of
some features of atmospheric mixing and transport that will provide a basic
understanding of the meteorological factors that affect the concentrations of
ozone and other oxidants in urban and rural areas.
3.4.1 Atmospheric Mixing
The concentration of an air pollutant depends significantly on the degree
of mixing that occurs between the time a pollutant or its precursors are
emitted and the arrival of the pollutant at the receptor. Since, to a first
3-54
-------
approximation, the diurnal cycle of weekday urban emission patterns for ozone
and oxidant precursor pollutants is generally uniform, it is reasonable to
ascribe a significant proportion of the large day-to-day changes in pollutant
concentrations to changes in meteorological mixing processes. The rate at
which atmospheric mixing processes occur and the extent of the final dilution
of the pollutants depends on the amount of turbulent mixing and on wind speed
and wind direction. Moreover, the transport of pollutants and precursors from
a source region to a distant receptor is also dependent on wind speed and wind
direction.
The degree of turbulent mixing can be characterized by atmospheric sta-
bility. In an atmospheric layer with relatively low turbulence, pollutants do
not spread as rapidly as they do in an unstable layer. Also, because a stable
layer has a relatively low rate of mixing, pollutants in a lower layer will
not mix through it to higher altitudes. The stable layer can act as a trap
for air pollutants lying beneath it. Hence, an elevated inversion is ofte.n
referred to as a "trapping" inversion. Also, if pollutants are emitted into a
stable layer aloft, such as from a stack, the lack of turbulence will keep the
effluents from reaching the ground while the inversion persists.
In air pollution considerations, a stable atmospheric layer or situation
is usually equated with a temperature inversion, which is a layer of the
atmosphere in which the temperature increases with increasing altitude, because
inversions are common and also represent the most stable atmospheric conditions.
The lowest part of an inversion layer is called the base and is defined as the
altitude at which the temperature begins to increase. The top of the inversion
is the point at which the temperature begins to decrease with increasing
altitude. The distance between the base and top of the inversion layer is the
"depth" or "thickness" of the inversion. Inversion layers may begin at the
ground surface (i.e., the altitude of the base is zero), or the entire inver-
sion layer may be above the surface. The former is known as a "surface inver-
sion" and the latter as an "elevated inversion." The two types are usually
caused by different sets of weather conditions, but it is not unusual for both
types of inversions to be present at a given location at the same time. In
the United States, surface inversions are characteristic of nighttime and
early morning hours except when heavy cloud cover or windy and stormy conditions
prevai1.
Surface and elevated inversion layers are both important in determining
pollutant concentration patterns since, as noted above, mixing and dilution
3-55
-------
processes proceed at a relatively slow rate in such layers. Thus, if pollu-
tants are emitted into an inversion layer, relatively high concentrations can
persist for a considerable period of time or over a considerable distance of
wind travel from the source. For example, a surface inversion in the morning
could cause automotive exhaust pollutants released at the surface during the
morning rush hours to persist with minimum dilution near the ground surface
for an extended period of time, probably for 1 or 2 hours after sunrise, until
solar radiation heats the ground and causes the inversion to disappear or
"break" (Hosier, 1961; Slade, 1968). High concentrations may occur at the
ground even when an elevated inversion is present and the layers below the
inversion are unstable and are undergoing good mixing. Such a persistent
elevated inversion layer is a major meteorological factor that contributes to
high pollutant concentrations and photochemical smog situations along the
southern California coast (Holzworth, 1964; Hosier, 1961; Robinson, 1952).
The vertical mixing profile through the lower layers of the atmosphere
follows a typical and predictable cycle on a generally clear day. In such a
situation a surface inversion would be expected to form during the early
morning and to persist until surface heating becomes significant, probably 2
or 3 hours after sunrise. Pollutants initially trapped in the surface inversion
may cause relatively high, local concentrations, but these concentrations will
decrease rapidly when the surface inversion is broken by surface heating,
usually about midmorning. The surface inversion will begin to form again
during the early evening hours and pollutants from near-surface sources such
as automobiles or home fireplace chimneys will experience progressively less
dilution as the surface inversion develops.
Elevated inversions, when the base is above the ground, are also common
occurrences (Hosier, 1961; Holzworth, 1964). Since these conditions, however,
are identified with specific synoptic conditions, they are much less frequent
than the nighttime radiation inversion. Because it may persist throughout the
day and thus restrict vertical mixing, an elevated inversion is nevertheless a
very significant air pollution feature. Smog-plagued southern California is
adversely affected by persistent elevated inversions (Robinson, 1952). When
compared to a source near the surface and the effects of a radiation (surface)
inversion, the pollutant dispersion pattern is quite different for an elevated
source plume trapped in a layer near the base of an elevated inversion. This
plume will not be in contact with the ground surface in the early morning
3-56
-------
hours because there is no mixing downward through the surface radiation inver-
sion. Thus, the elevated plume will not affect surface pollutant concentra-
tions until the mixing processes become strong enough to reach the altitude of
the plume. At that time, the plume may be mixed downward quite rapidly in a
process called "fumigation." During fumigation, surface ozone concentrations
will increase if the morning ozone concentration is higher aloft than at,the
ground, and if insufficient scavenging by NO occurs at ground-level. After
this initial mixing, surface concentrations will decrease as the usual daytime
mixing processes continue to develop. If the daytime mixing becomes strong
enough to break the upper inversion, the pollutants may be mixed through an
increasingly deep layer of the atmosphere. When surface heating decreases in
the late afternoon and early evening, both the surface and elevated inversions
will form again. The surface inversion will again prevent pollutants from
elevated sources from reaching the ground and surface scavenging processes
will gradually reduce the concentrations of pollutants trapped during the
formation of the surface inversion.
Geography can have a significant impact on dispersion of pollutants
(e.g., along the coast of an ocean or one of the Great Lakes). Near the coast
or shore, the temperatures of land and water masses can be different, as can
the temperature of the air above such land and water masses. When the water
is warmer than the land, there is a,tendency toward reduction in the .frequency
of surface inversion conditions inland over a relatively narrow coastal strip
(Hosier, 1961). This in turn tends to increase pollutant dispersion in such
areas. Such conditions may occur frequently on the Gulf Coast and near the
Great Lakes in the winter. The opposite condition also occurs if,the water is
cooler than the land, as in summer or fall. Cool air near the water surface
will tend to increase the stability of the boundary layer in the coastal zone,
and thus decrease the mixing processes that act on pollutant emissions. These
conditions occur frequently, along the New England coast (Hosier, 1961).
Similarly, pollutants from, the Chicago area have been observed repeatedly to
be influenced by a stable boundary layer over Lake Michigan (Lyons and Olsson,
1972). This has been observed especially in summer and fall when the lake
surface is most likely to be cooler than the air that is carried over it from
the adjacent land.
Since the diurnal mixing conditions are such an important part of the
meteorological parameters for understanding pollutant mixing and diffusion, it
3-57
-------
is useful to have some knowledge of the mixing cycles that prevail over the
United States, Figures 3-4 and 3-5 show the average summer morning and after-
noon mixing heights (AGL» above ground level) as calculated on the basis of
upper air temperature data and an estimated midmorning urban temperature.
Since Holzworth (1972) attempted to include the influence of an urban heat
island in this estimated temperature, the morning results in Figure 3-4 are
probably most applicable to larger urban areas. Rural or nonurban areas would
be expected to have lower mixing heights.
Summer conditions are useful to consider because of the prevalence of
high photochemical oxidant concentrations during this season. As shown in
Figure 3-4, morning mixing heights are estimated to be greater than 300 meters
except for the central part of the Great Basin, where a 200-meter isopleth
includes parts of Oregon, Idaho, Utah, Arizona, and most of Nevada. By mid-
afternoon (Figure 3-5), the estimated mixing height at the time of maximum
temperature has increased markedly, and only a few coastal areas have an
average afternoon maximum mixing height of less than 1000 meters. In contrast
to the morning data, the central Great Basin area becomes the area of greatest
mixing in the afternoon. This would be expected since this is a hot, arid,
desert region, and the driving force generating the surface mixing layer is
the solar heating of the ground surface.
The magnitude of the afternoon mixing height is generally an indication
of the potential for recurring urban air pollution problems. If a trapping,
elevated inversion does not rise high enough in the afternoon to release the
generated pollutants that are trapped, an accumulating episode is likely.
From the average summer afternoon data shown in Figure 3-5, where the lowest
average mixing height is 600 meters and almost all of the area has a value
greater than 1500 meters, it would appear that urban air pollution should not
be severe. On the average this is probably correct; however, there are several
departures from the average, which result in relatively low mixing heights and
adverse dispersion over many areas of the United States on a recurring basis.
Figure 3-6 (Holzworth and Fisher, 1979) shows the frequency of occurrence
of elevated inversions in summer having a base between I and 500 meters (1600
feet) at the time of the afternoon upper air temperature measurement, 6:10
p.m. EST or 3:15 p.m. PST. The California coastal conditions, in which low
inversions occur with a frequency of nearly 90 percent, are clearly evident.
The northeastern coastal area from New Jersey north to Maine, where cool ocean
3-58
-------
11
Figure 3-4. Isopleths (m x 10^} of mean summer morning mixing heights, AGL.
Source: Holzworth (1972).
18
Figure 3-5. Isopleths (m x 10^) of mean summer afternoon mixing heights, AGL.
Source: Holzworth (1972).
3-59
-------
30
Figure 3-6. Percentage of summer 2315 GMT (6:15 p.m. EST, 3:15 p.m. PST)
soundings with an elevated inversion base between 1 and 500 m above ground level.
Source: Adapted from Holzworth and Fisher (1979).
water prevails, also has a relatively high percentage, above 20 percent,
compared to most of the rest of the country. Stations bordering one of the
Great Lakes—Green Bay, Sault St. Marie, and Buffalo—reflect a stabilizing
lake effect with percentages above 5 percent. Except along the Pacific Coast,
these ocean and lake coastal situations are probably limited to relatively
narrow coastal zones (Hosier, 1961). Examples are evident in Figure 3-6, in
which it may be noted that inversion frequencies of 21 to 28 percent occur in
coastal New England compared to only 2 percent at Albany in upstate New York.
A similar situation is evident in a comparison of the 3 percent inversion
frequency at Washington, D.C., with the 16 percent frequency on the Delaware
coast. A non-coastal region having summer afternoon low-level elevated inver-
sions more than 5 percent of the time is the Southeast, where an area from
Louisiana and Arkansas to the Atlantic coast shows frequency values between 5
and 10 percent. Other seasons differ in details, but the general patterns are
similar.
3-60
-------
This means that, for most of the United States, low-level stable layers
that persist through the afternoon hours are rare events, occurring on less
than 1 day in 20. Thus, air pollution situations in areas such as Kansas or
Iowa, although they can be the result of transport, may commonly be related to
the periods when the expected morning surface inversion persists later in the
morning than usual and when winds are not strong enough to carry pollutants
rapidly away from the local area. Along both the Pacific Coast and the
Northeast Coast, low-level afternoon inversions are frequent enough to be a
significant contributor to local and regional air pollution episodes.
3.4.2 Wind Speed and Mixing
Another major meteorological factor in the urban pollutant dispersion
problem is low-level or surface-layer wind. As would be expected, strong
winds across a source area will dilute pollutant concentrations even though
there is a strong, low-level inversion base. San Francisco is one example of
such a location where strong winds frequently provide good ventilation in
spite of a low inversion. Conversely, light and variable or calm wind condi-
tions over an area can lead to excessive pollutant accumulations even though
the afternoon mixing depth is quite large. Thus, it is necessary to include
wind direction and wind speed frequencies in any evaluation of air pollution
potential for a given area. It must also be recognized that both elevated
inversion conditions and surface wind patterns are governed to a major degree
by the synoptic, or large-scale, weather patterns. Both wind and inversion
factors tend to favor pollutant buildup when a deep, slow-moving high-pressure
system dominates the weather across an area (Korshover, 1967; Korshover,
1975).
Figure 3-7 shows the wind climatology across the United States in the
month of July by depicting the monthly resultant vector wind at major weather
stations (U.S. Department of Commerce, 1968), Note that the flow across the
West Coast is generally directed inland, from west to east. This contributes
to a typical situation for major California cities: significant urban pollu-
tant plumes are found east of the urban core source areas while the immediate
coastline or beach areas are relatively pollutant-free. In the Northeast
States, the average wind flow is from southwest to northeast more or less
parallel to the coastline. As a result, pollutant plumes from the major urban
areas along this coast are frequently additive along the trajectory of the
3-61
-------
CT>
SAN FRANCISCO
*^w:e—53"V --"-^"j|
wS N> ^/TA
ROI. j
t f\
t 1 \
'
/ M I
MILWAUKEE GRAND RAPIDS
ALBUQUEHOUE AMARILLO OKLAHOMA CITY MEMPHIS
LITTLE ROCK
rrii -r £
/T [ATLANTA*""*
. -, ^n^uno SHREVEPORT JACKSON muniuumcni y I
^™°v V \ /r i n 7
\ \ \ » f .. MOBILE—TALLAHASSEE-JljACKSl
\ N AUSTIN \ LAKE CHARLES-TjO^^ JC Wj'AUK,
V /*>w 1 GALVESTONg iPTT ^^ X. \\
X/ A \ «T~^^^V > GULFPORT A J
\ \ J^L I ^> TAMPA V
A \ rSr IA \
NOTE: BASED ON
HOURLY OBSERVATIONS
1951-1960.
SCALE IN mph
Figure 3-7. Mean resultant surface wind pattern for the United States for July. Direction and
length of arrows indicate monthly resultant wind.
Source: U.S. Dept. of Commerce (1968).
-------
wind. Polluted air moving toward the coast from major inland urban sources
may also be a factor in this Northeast region. Along the Gulf Coast, the
average winds form southerly, onshore flow. Under some weather situations,
however, there is often an offshore flow in one area (e.g., Texas) and an
onshore flow in an adjacent area. Thus, because of this recirculation, the
onshore Gulf air masses are not always pollutant-free (Price, 1976; Wolff et
al., 1981). Before the situation was examined carefully, the recirculating
pollutants were sometimes confused with natural background concentrations.
Wind climatology provides an average wind flow pattern, but it does not
provide a complete assessment of the influences of the wind on air pollution
dispersion. Wind speed and, in particular, the frequency of weak winds are an
important aspect to be considered. Figure 3-8, adapted from Holzworth and
Fisher (1979), shows the frequency with which early morning (6:15 a.m. EST or
3:15 a.m. PST) surface inversions occur with calm or weak surface winds; that
is, wind speeds equal to or less than 2.5 m/sec or 6 mi/hr. There is consi-
derable variation between stations because terrain and geography (e.g., coastal
locations) influence both the wind flow and inversion frequency. It is clear,
however, that over large areas of the United States, especially in heavily
industrialized inland areas east of the Mississippi River, calm amd stable
summer mornings are a frequent occurrence: 50 percent or more in many areas.
This means that there will be frequent incidents of morning pollutant accumu-
lation; but afternoon heating, as shown by Figure 3-5, will usually mix the
pollutant accumulations through a deep mixing layer and disperse them. Figure
3-9 (Holzworth and Fisher, 1979) shows the average wind speed through the
depth of the summer morning mixing layer. Note that the area east of the Rocky
Mountains, except for the Appalachians, can on the average, expect winds of 4
m/sec (about 10 mi/hr) or higher through the morning mixing layer. This
probably would provide acceptable midmorning dilution of accumulated pollu-
tants. In summer afternoons, as shown in Figure 3-10, the average wind speed
within the mixing layer increases in all areas and may even double over some
of the western mountain states. It should be noted, however, that since winds
normally increase with altitude above the ground, much of the increase in the
average afternoon mixing layer wind is probably the result of the considerable
increase in the depth of the mixed layer, as shown by the differences between
Figures 3-4 and 3-5.
3-63
-------
16
41
Figure 3-8. Percentage of summer 1115 GMT (6:15 a.m. EST, 3:15 a.m. PST) soundings
with an inversion base at the surface and wind speeds at the surface ^2.5 m/sec.
Source: Adapted from Holzworth and Fisher (1979).
3-64
-------
Figure 3-9. Isopleths (m/sec) of mean summer wind .speed averaged
through the morning mixing layer.
Source; Holzworth and Fisher (1979),
Figure 3-10. Isopleths (m/sec) of mean summer wind speed averaged
through the afternoon mixing layer.
Source; Holzworth and Fisher (1979),
3-65
-------
In summary, atmospheric mixing parameters of stability and wind in the
pollutant transport layers can exert controlling effects on 0, and oxidant
concentrations. The effects include the amount of dilution occurring in the
source area, as well as along the trajectory followed by an urban or source-
area plume. Regions having steady prevailing winds, such that a given air
parcel can pass over a number of significant source areas, can develop signifi-
cant levels of pollutants even in the absence of weather patterns that lead to
the stagnation type of air pollution episodes. The Northeast states are
highly susceptible to pollutant plume transport effects, although some notable
stagnation episodes have also affected this area (e.g., Lynn et al., 1964).
Along the Pacific Coast, especially along the coast of California, coastal
winds and a persistent low inversion layer contribute to major pollutant
buildups in urban source areas and downwind along the urban plume trajectory
(Robinson, 1952; Neiburger et al., 1961). In the southern Appalachians, the
weather favors longer-term air pollution episodes (Korshover, 1967; Korshover,
1975). Generally, low pollution potential results from the conditions that
occur in the Great Plains area and south to the Texas-Louisiana Gulf Coast;
and between the Mississippi River and the crest of the Rocky Mountains.
It should be clear even from this brief discussion that there are funda-
mental differences in regional meteorological conditions that cause the air
pollution potential applicable generally to California to be more severe than
in other parts of the United States. When adverse meteorology is coupled with
high population and source concentrations, it is quite evident why California
areas have severe photochemical air pollution problems. This very significant
difference in the magnitude of the photochemical air pollution problem in
California regions compared to non-California locations can serve as a basis
for separating air pollution statistics into two sets for evaluation, namely,
California and non-California groups.
3.4.3 Effects of Sunlight and Temperature
The significance of sunlight in photochemistry is related to its intensity
and its spectral distribution, both of which have direct effects upon the
specific chemical reaction steps that initiate and sustain oxidant formation.
Sunlight intensity varies with season and geographical latitude but the latter
effect is strong only during the winter months. During the summer, the maximum
light intensity is fairly constant throughout the contiguous United States and
only the duration of the solar day varies to a small degree with latitude.
3-66
-------
The effects of light intensity on individual photolytic reaction steps
and on the overall process of oxidant formation have been studied in the
laboratory (Peterson, 1976; Demerjian et a!., 1980). All of the early studies,
however, employed constant light intensities, in contrast to the diurnally
varying intensities that occur in the ambient atmosphere. More recently, the
diurnal variation of light intensity has been recognized and studied as a
factor in photochemical oxidant formation (Jeffries et al., 1975; Jeffries
et al., 1976). Such studies showed that the effect of this factor varies with
initial reactant concentrations. Most important was the observation that
similar NMOC/NO systems showed different oxidant potential depending on
/\
whether studies of these were conducted using constant or diurnal light. This
has led to incorporation of the effects of diurnal or variable light into
photochemical models (Tilden and Seinfeld, 1982).
While the effect of sunlight intensity is direct and has been amply
demonstrated (Leighton, 1961; Winer et al., 1979), the effect of wavelength
distribution on the overall oxidant formation process is subtle. Experimental
studies have shown the photolysis of aldehydes to be strongly dependent on
radiation wavelength in the near UV region (Leighton, 1961); and there is some
indication (Bass et al., 1980) that the photolysis rates for aldehydes may be
temperature-dependent. Since aldehydes are major products in the atmospheric
photooxidation of NMOC/NO mixtures, it is inferred that the radiation wave-
s\
length should have an effect on the overall photooxidation process. This
inference was directly verified, at least for the propylene/NO and rrbutane/
x\ """"""'
NO chemical systems, in smog chamber studies (Jaffee et al., 1974; Winer
X
et al., 1979). In the ambient atmosphere, some variation in the wavelength
distribution of sunlight does occur as a result of variations in time of day,
stratospheric Oo, ambient aerosol (Stair, 1961), and cloud cover.
It has been observed that days on which significant ozone-oxidant con-
centrations occur are usually days with warm, above-normal temperatures (Bach,
1975). The possibility that photochemical reactions show some temperature
dependence has been raised by smog chamber studies (e.g., Carter et al., 1979;
Countess et al., 1981), but is usually thought to be the production of chamber
artifacts concomitant with increases in temperature. In ambient air, the
observed correlation between temperature and oxidant concentrations can be
explained, at least in part, as a synoptic meteorological correlation rather
than as a temperature-photochemical rate constant effect, in that periods of
3-67
-------
clear skies and warm temperatures are periods of high air pollution potential,
as discussed above. Because of the close correlation between above-normal
temperatures and high photochemical air pollution potential, a maximum daily
temperature forecasting procedure is often useful as a substitute for a more
elaborate and specific program for forecasting possible photochemical air
pollution. The correlation between temperature and, thus, synoptic weather
conditions and photochemical air pollution intensity has been observed in a
number of areas. Evaluation of photochemical air pollution in Los Angeles as
early as 1948 showed a correlation with temperature. Recent studies of 0-
patterns in St. Louis, Missouri, have also shown a correspondence between
•daily maximum CL concentration and temperature (Shreffler and Evans, 1982).
Wolff and Lioy (1978) have shown high correlations between ozone concentration
and both the temperature for the current day and the temperature for the
previous day.
3.4.4 Transport of Ozone and Other Qxidants and Their Precursors
The 1978 air quality criteria document for ozone and photochemical oxidants
made a convincing case for the fact that ozone and other photochemical oxidants
are transported from urban source areas, other than those in California, to
downwind regions in concentrations of 0.1 ppm or greater (U.S. Environmental
Protection Agency, 1978a, and references therein). The 1972 study by Research
Triangle Institute at McHenry, Maryland, was an early examination of rural
oxidant in the eastern United States (Ripperton et al., 1977). Bach (1975)
examined the meteorology of these observed conditions and showed details of
the influence of transient high pressure systems. The transport of large
urban plumes in the northeastern states, especially from New York City into
Connecticut, was the subject of an EPA field study in 1975 (Westberg et al.,
1976; Wolff et al., 1977d; Westberg et al., 1978a). Transport of ozone on a
regional basis in the northeast was also described by Cleveland et al. (1976a,
1976b), while Wolff et al. (1977a) presented details of several east coast
urban oxidant plumes. The Northeast Regional Oxidant Study (NEROS) carried
out by EPA in 1979 and 1980 in the corridor from Washington, D.C., to Boston,
was designed specifically to support urban plume model development (Clarke et
al., 1982). Plume models have been based on other urban plume investigations,
as well (e.g., Wolff et al., 1977c), and reactive plume modeling for other
urban areas has progressed in recent years (e.g., U.S. Environmental Protection
3-68
-------
Agency, 1981a). These several early research programs and analyses served to
answer a number of perplexing problems that had been identified previously in
studies of ozone and other photochemical oxidants in nonurban areas (e.g.,
Ripperton et al., 1977; Bach, 1975). These questions included reasons for the
occurrence of high ozone or total oxidant concentrations in areas remote from
identifiable sources.
Studies of the transport of ozone and other photochemical oxidants (0~-0 )
*5 X
are classified into three regimes, depending upon transport distance (U.S.
Environmental Protection Agency, 1978a). In the first, urban-scale transport,
the occurrence of transport of photochemical pollutants can be detected in
most large urban areas if there is sufficient Q--Q monitoring information.
*3 X
It has been identified as a significant, characteristic feature of the 0~-Q
•C* X
problem in the Los Angeles basin (Tiao et al., 1975), as well as in San Fran-
cisco, New York, Houston, Phoenix, and St. Louis (Altshuller, 1975; Coffey and
Stasiuk, 1975; Shreffler and Evans, 1982; Wolff et al., 1977a). As noted
above, the recognition and assessment of the probable magnitude of urban
oxidant problems in locations other than in California has been a major research
topic since the mid-1970s (e.g., Bach, 1975; Cleveland et al.,,1976a, 1976b;
Westberg et al., 1976; Wolff et al., 1977a,b,c; and others).
Urban-scale transport patterns result from one or more of a combination
of factors. First is the simple advection of the photochemically reacting air
mass and -the development of maximum 0~-0 after 1 or 2 hours of downwind
O A
travel. Maximum concentrations may be displaced up to 20 or so miles from the
center of the major source area. It has been noted (U.S. Environmental Protec-
tion Agency, 1978a) that pollutant concentrations in air parcels in the central
core area of major source areas may not be the most conducive for 0~-0 forma-
*3 X
tion because of the tendency toward occurrence there of more effective scavenging,
especially scavenging related to NO and its reactions.
The distance of.the peak 0~-0 concentrations from the urban core area is
O X
dependent on the local wind pattern and is, in general, inversely related to
the peak 0~-0 concentration. Stronger winds will carry the air parcels
, «3 X
farther during the reaction period, increasingly diluting pollutant concentra-
tions along the trajectory. Weak winds and very restricted mixing heights
will tend to cause higher Q»-0 concentrations closer to the central source
*3 X
area. The diurnal wind cycle will also be an important factor, since in some
situations calm conditions may prevail until late in the morning but in others
3-69
-------
a steady wind may be present throughout the emission and reaction process,
Wolff and Lioy (1978) were able to model urban ozone concentrations on the
basis of meteorological observations, especially temperature.
The second, or mesoscale, kind of transport of CL-0 is in many respects
*5 X
an extension of the urban-scale transport and is characterized by urban plume
development. A report by Bell (1960) described November 1959 OVO incidents
O JTX
in northern coastal San Diego County, California. It showed conclusively that
these were caused by the 0,,-Q and precursors formed and emitted, respectively,
3 X
the previous day in the Los Angeles basin. The transport in these situations
was over the coastal Pacific Ocean, and the Q.j-0 arrived at the San Diego
*j X
receptor site as a contaminated sea breeze after overnight travel (Bell,
1960). The studies of Cleveland et al. (1976a, 1976b) are early documentation
of a similar scale of transport in the New York-Connecticut area. The results
of extensive aircraft measurements and modeling assessments of studies in the
Washington, D.C., to Boston corridor have also been reported by Wolff and his
colleagues (Wolff et al., 1977a,c; Wolff and Lioy, 1978).
In the 1978 0--0 criteria document, more than 30 references were cited
*3 X
relating to urban plume observations and investigations. Since that document
was published, the results of the 1975 New England oxidant study have been
published in detail (e.g., U.S. Environmental Protection Agency, 1977; Westberg
et al., 1978a), and results of a more comprehensive 2-year field program
carried out along the Washington, D.C.-Boston corridor in 1979 and 1980 have
appeared in the literature (Clark and Clarke, 1982; Clarke et al., 1982;
Vaughan et al,, 1982). A major field program supported by local industries
has been conducted in Houston, Texas, although the D.,-0 downwind plume phases
O f\
of that study were not as extensive as in NEROS. Chicago, Detroit, and adjacent
shoreline areas of Lake Michigan have also been the subject of a number of
ground-level and airborne studies over distances of 70 to 300 kilometers
(Lyons and Olsson, 1972; Sexton and Westberg, 1980; Westberg et al., 1981;
Kelly et al., 1986). As described above, Oo~0 plumes from major urban areas
O )\
can extend about 100 to 200 miles, with widths of tens of miles (Sexton,
1982), frequently up to half the length of the plume. Other field studies
conclusively demonstrating mesoscale transport over New England have been
reported (Spicer et al., 1979; Clarke et al., 1982; Cleveland et al., 1976a,b;
Rubino et al., 1976; Westberg et al., 1976; Westberg et al., 1978a; Wolff et
al., 1977a). Although urban plumes are frequently thought of as a problem
3-70
-------
related only to large source areas such as New York and other major metropolitan
areas, measurements in plumes from smaller urban areas have shown that these
sources cannot be ignored (Sticksel et al., 1979; Sexton, 1982; Spicer et al.,
1982; Wolff et al., 1977c).
In the third kind of pollutant transport, synoptic-scale, the transport
of 0~-0 and precursors is characterized by the general and widespread elevated
«5 y\
concentrations of pollutants that can occur on an air-mass scale under certain
favorable weather patterns. These weather situations are generally slow-moving,
well-developed high-pressure, or anticyclonic systems. This type of deep
high-pressure area was considered by Korshover (1967, 1975) as a prerequisite
for stagnating air pollution episodes.
A major criterion of these synoptic systems is the reinforcement of the
surface high-pressure area by a warm high-pressure circulation in the upper
air. The surface weather is then frequently characterized by weak winds,
stable surface layers, and high pollution potential over regional or air-mass-
sized areas. This is the generalized meteorological model pattern that is
involved in synoptic-scale pollutant transport (Korshover, 1967; Korshover,
1975). As with all generalized models, there have been specific oxidant
episodes that departed from this model. In particular, weak winds are not
always a prerequisite; and relatively strong winds have been observed to be
associated with some oxidant transport episodes (Mueller and Hidy, 1983; Wolff
and Lioy, 1980), when such winds did not also produce rapid plume dilutipn as
is usually expected.
While surface highs that are reinforced aloft by a warm high can lead to
air pollution episodes, there are other high-pressure systems that usually
cause few or no widespread air pollution problems. These are the strong
surface highs found behind rapidly moving storm systems, in which the surface
high underlies a cold low-pressure system aloft. These systems characteristi-
cally have good vertical mixing (instability), brisk winds, and low air
pollution potential.
Bach (1975) assessed in detail the relationship between elevated ozone
levels and synoptic systems. Other investigators have since described specific
instances of large-scale ozone transport and associated meteorological condi-
tions (e.g., Wolff and Lioy, 1980; Wolff et al., 1980; Wolff et al.,,1982).
For example, Wolff et al. (1982) have described synoptic meteorological systems
and the occurrence of haze and elevated ozone levels in an area extending from
3-71
-------
the Midwest to the Gulf Coast. In a separate study, Wolff and Lioy (1980)
examined the spatial and temporal distributions of ozone during three photo-
chemical smog episodes in July 1977. A stagnating high-pressure system formed
over the Gulf of Mexico and two high-pressure systems originating in Canada
were described as the respective meteorological systems associated with the
three episodes. In all three cases, elevated ozone concentrations (~120 to
130 ppb) were found to extend, in a virtual "ozone river", from the Gulf Coast
to New England (with 328 ppb ozone measured in Connecticut). The high-pressure
system originating in the Gulf area affected the entire southeastern United
States and extended from western Texas, northeastward through Illinois, and
east to the Atlantic Ocean. Elevated ozone levels affected anywhere from a
few hundred square miles to a thousand square miles during the 1-week period.
The importance of synoptic-scale or air-mass pollutant situations has
been recognized for many years, probably much longer than the importance of
major plumes has been apparent. The Donora, Pennsylvania, smog episode in
1948 (Schrenk et al., 1949), while not a photochemical smog situation, involved
the occurrence over a wide area of a regional air mass having relatively high
pollution levels simultaneous with the occurrence of a stagnating warm high-
pressure area over the Ohio Valley and the northern Appalachian area, Donora
was an especially adversely affected pocket within this larger system; in that
case, however, ozone was probably not one of the important pollutants.
The synoptic-scale high-pressure air pollution system is not charac-
terized by well-defined urban plumes. Rather, a warm, slow-moving or stagnant
anti-cyclone provides a synoptic-scale weather system that, because of weak
winds and limited vertical mixing, favors the accumulation of relatively high
concentrations of air pollutants. On a climatological basis, these systems
are most common in the summer and fall months over most of the United States,
as shown by the work of Korshover (1967; 1975). In many cases, an anti-cyclone
will stagnate or recurve and intensify over the Midwest or East as circulation
patterns in the upper air change and become more supportive of the surface
anti-cyclonic pattern (Schrenk et al., 1949; Lynn et al., 1964). The typical
paths followed by the air masses forming these slower-moving anticyclones in
the summer and fall months, as described by Bach (1975) and Wight et al.
(1978), is southeastward from Canada into the upper Mississippi Valley, eastward
across the Ohio Valley and then across the East Coast, either toward the
northeast into New England, east into the central Atlantic States, or southeast
3-72
-------
into the southeastern states. The actual stagnation of a pressure system for
one or more days, as occurred in the Donora case, is, of course, the most
severe example of a slow-moving system.
Along the West Coast, air pollution problems are also the result of
persistent high-pressure system influences. In this case, however, the high
is the persistent subtropical anti-cyclone of the eastern Pacific rather than
the series of transitory anti-cyclonic systems characteristic of the area east
of the Rocky Mountains. The persistent or semipermanent subtropical anti-
cyclone in the Pacific is linked to the large-scale general circulation of the
atmosphere rather than to moving wave systems (Neiburger et a!., 1961). The
effect is much the same, except that the area of limited mixing and more
adverse air pollutant effects is found on the eastern edge of the subtropical
anti-cyclone rather than on the trailing western edge as in the transitory
systems.
It is worthwhile to point out that the typical pattern of ozone concentra-
tions in the slow-moving midwest anticyclone, in contrast to the conditions in
California, shows the lowest values in the eastern portion of the system,
behind the advancing cold front. The central area of the anticyclonic system,
where dispersion is usually at a minimum, shows a broad expanse of elevated
ozone concentrations; but the highest ozone values in the typical system are
usually found in the western parts of the system, the so-called "back side,"
where dispersion conditions, although better than in the central portion of
the system, combine with increased residence time and longer exposure to
emission sources to cause the maximum ozone concentrations in a given high-
pressure system. Investigators describing this ozone pattern over the midwestern
and eastern parts of the United States include Bach (1975), Vukovich et al.
(1977), Wolff et al. (1977b), Wight et al. (1978), and Westberg et al. (1978a,
1981).
The identification and understanding of photochemical Q~-Q and precursor
•5 X
transport by weather systems has provided a significant advance in comprehending
photochemical air pollution and the potential extent of its effects. Considerable
progress has been made in the development of long-range photochemical modeling
techniques so that the likely impact of synoptic systems can be anticipated.
Such tools are very much in the research stage because the local impact of
0,-Q results from a complex interaction of distant and local precursor, sources,
«5 X
3-73
-------
urban plumes, mixing processes, atmospheric chemical reactions, and general
meteorology.
3.4.5 Surface Scavenging in Relation to Transport
Major scavenging processes for CU in the atmospheric boundary layer are
adsorption and subsequent destruction at the ground surface (i.e., dry deposi-
tion), and reactions with boundary layer pollutants, especially NO and
alkenes. In dry deposition, eddy diffusion moves air parcels downward through
the turbulent boundary layer to the laminar sub-layer where individual molecules,
such as ozone, can move by Brownian motion to the underlying surface. There
reactive molecules, such as ozone, can be removed by reactions at the surface.
Chemical reactions between ozone and NO and reactive hydrocarbons are described
in other sections of this document. When they occur in the near-surface
layers, these reactions can play an important role in the boundary layer
scavenging process. Dry deposition and boundary layer chemical reactions
result in a vertical concentration gradient, with the lowest concentrations
occurring at the surface of the ground.
Because of this surface-scavenging process, ozone will persist in an
atmospheric parcel in the absence of ozone-forming reactions only if the
parcel is dispersed such that contacts with the ground surface or surface
pollutant sources are minimized. It is likely that only in those air parcels
moving above the surface layer will ozone escape the surface reactions and
persist long enough to undergo long-distance transport or persist overnight.
It should be noted here that ozone transported aloft, judging from the limited
data available, may be 20 to 70 percent additive in urban areas, as determined
by simulations using the EKMA chemical kinetics model and several different
rates of vertical mixing (U.S. Environmental Protection Agency, 1977b). Thus,
if 0.1 ppm ozone were to be transported aloft and was 40 percent additive at
ground level, the contribution of transport to the peak ozone concentration
downwind would be 0.04 ppm.
Aircraft observations have documented frequently the occurrence of
relatively high ozone concentrations above lower-concentration surface layers
(e.g., Westberg et a!., 1976). This is a clear indication that ozone is
essentially preserved in layers above the surface and can be transported over
relatively long distances even when continual replenishment through precursor
3-74
-------
reactions is not a factor, such as at night. Boundary-layer scavenging pro-
cesses are also responsible for the fact that ozone concentrations in urban
areas rapidly fall to zero during the night.
3.4.6 Stratospheric-Trppospheric Ozone Exchange
The fact that 0, is formed in the stratosphere, mixed downward, and
incorporated into the troposphere, where it forms a more or less uniformly
mixed background concentration, has been known in various degrees of detail
for many years (Junge, 1963).
It is widely accepted that the long-term average tropospheric background
concentration at the surface ranges from about 30 ppb to 50 ppb (Fabian and
Pruchniewicz, 1977; Oltmans, 1981); and that it results primarily, though not
exclusively, from the transfer of stratospheric _ozone into the upper troposphere,
followed by subsequent dispersion throughout the troposphere (e.g., Singh et
a!., 1980; Kelly et al., 1982). Ozone residence time in the troposphere has
been estimated at 1 to 2 months by Junge (1963); and at 1 to 2 months for
spring and summer but at 2.5 to 3.5 months for fall and winter, respectively,
by Singh et al. (1980).
The mechanisms by which stratospheric air is mixed into the troposphere
have been examined by a number of authors. Danielsen conducted extensive
analyses of major synoptic weather events that injected stratospheric air into
the troposphere (Danielsen, 1968; Danielsen and Mohnen, 1977; Danielsen,
1980). Reiter has been especially active in describing the atmospheric
mechanisms by which stratospheric air injection takes place and in relating
these processes to the global circulation of the atmosphere (Reiter, 1963;
Reiter and Mahlman, 1965; Reiter, 1975). As a result of such research, exchange
between the stratosphere and troposphere in the middle latitudes has been
determined to occur to a major extent in events called "tropopause folds." In
a tropopause fold (TF), the jet stream in the upper troposphere plays a major
role in directing stratospheric air and high ozone concentrations into the
troposphere. Figure 3-11 is a schematic presentation of the intrusion process
as described by Danielsen (1968). The subsidence occurs along the poleward
side of the polar jet stream in the area where the jet is associated with a
cold front at ground level. The result is downward transport in the cold air
behind the cold front.
3-75
-------
Figure 3-11. Schematic cross section, looking downwind along the jet stream,
of a tropopause folding event as modeled by Danielsen (1968).
Source: Johnson and Viezee (1981).
3-76
-------
Since 1978, a considerable amount of research on TF and ozone injection
has been done, especially by SRI-International (Johnson and Viezee, 1981;
Ludwig et al,, 1977; Singh et a!., 1980; and Viezee et a!., 1979). Figure 3-12
from Johnson and Viezee (1981) shows one example of the probing by SRI of a TF
event In the midwestern United States. Concentrations of ozone in excess of
90 ppb were found as low as 13,000 feet (ca. 2,5 miles or 3.9 kilometers), as
shown in the upper part of Figure 3-12. These authors found that ozone intrusion
was lower during this fall study (October 5, 1978) than in a number of other
spring TF events. The dew point measurements in the second part of the figure
confirm the stratospheric injection. The weather pattern accompanying this TF
is shown at the bottom of Figure 3-12 by a 500 millibar (about 6 kilometer)
chart; the surface cold front is also indicated. Note that the intrusion was
detected well behind the cold front and appears to have assumed a layered
formation in the altitude range of 8,000 to 12,000 feet (2.4 to 3.6 kilometers).
From their analysis of measurement flights in a number of TF situations,
Johnson and Viezee (1981) concluded that the ozone-rich intrusions studied
sloped downward toward the south. In terms of dimensions, the average crosswind
width (north to south) at an altitude of 5.5 kilometers (ca. 18,000 feet or
3.4 miles) for six spring intrusions averaged 226 kilometers (364 miles), and
for four fall TF systems, 129 kilometers (208 miles). Ozone concentrations at
5.5 kilometers (ca. 18,000 feet or 3.4 miles) averaged 108 ppb in the spring
systems and 83 ppb in the fall systems. Previously it had been assumed that
only a few fairly intense systems would produce a TF event and trans-tropopause
mixing. From their data, however, Johnson and Viezee (1981) drew the very
important conclusion that all low-pressure trough systems, such as that
illustrated in Figure 3-12, may induce a TF event and cause the trans-tropopause
movement of ozone-rich air into the troposphere.
On the basis of their field studies and the earlier models and work of
Danielsen (1968), Johnson and Viezee (1981) proposed a set of model mechanisms
or types of TF injection, which are illustrated in Figure 3-13 and described
in the following general manner:
1- Type 1. The intrusion is broken up and dispersed by mixing and
diffusion in the middle or free troposphere.
2. Type 2. The intrusion persists down to the planetary boundary
layer or the top of the mixed layer, where the lower
part of the intrusion may be incorporated into the
mixed layer and may subsequently reach the ground.
3-77
-------
28
24
20
16
12
_j R
(/) O
5
* 4
1 0
a
3
3 '8.g'|iǤ8l3 4
50
I I
1 I
(A) CROSS-SECTION OF OZONE, ppb
I I I I I 1 I
CNG
I I I
J I
40-
, MEM
28
24
20
16
12
8
I
I I I I I I
-30
(BI CROSS-SECTION OF DEWPOINT TEMPERATURE, °C
CNG
I I I
J I
I
I
I I I L
MEM
I I I
as
7,3
6.1
4.9
3.7
2.4
1.2
0
8.5
7.3
6.1
4.9
3.7
2.4
1.2
CMI
-160
(-296)
-120
I-222)
-80
(-148)
-40
(-74)
0
(0)
+40
(74)
+80
(148)
+120
(222)
+ 160
(296)
DISTANCE FROM CNG, nautical mi and (km)
****** FLIGHT TRACK .
CONTOURS (DYNAMIC METERS)
V V f V SURFACE WEATHER FRONT
1C) 500-mb CHART AND FLIGHT TRACK
Figure 3-12. Measured vertical cross-sections of (A) ozone, (B)
dew-point, and (C) the 500-mb chart and the flight track for
October 5,1978. CMI = Champaign, IL; CNG = Cunningham, KS;
MEM = Memphis, TN.
Source: Johnson and Viezee (1981).
Cfl
5
j<
ul
a
3
3-78
-------
Q
*Z
3
O
DC
O
LATJRAL £
MIXING O
CO
I-
2
UJ
TYPE1
MIXING LAYER
TYPE 2
7///////////////////W/fl7/////W//7/////////////W
WMm/m//W//Mwm/w//^^^^
Figure 3-13. Hypothesized models of the process that mixes tropopause
folding events into the troposphere.
Source: Johnson and Viezee (1981).
3-79
-------
3- Type3. The intrusion occurs close behind the cold front, where
the air parcels are caught by the downdrafts behind the
cold front; and is brought to the ground by direct
circulations associated with the front.
4- Type 4. The ozone-rich parcels are incorporated into convective
cells and brought to the ground in association with
rain-showers and thunderstorm downdrafts; similar to
Type 3.
Johnson and Viezee (1981) summarized the possible impacts of these four
types of TF events by noting that Types 1 and 2 should produce "relatively
moderate effects" at the ground in comparison to those to be expected from
Types 3 and 4. The latter two could cause "substantial" effects in terms of
high surface ozone concentrations. The action described by Types 3 and 4 is
supported by meteorological theory (Bjerknes, 1951) and by observations of
surface ozone such as those made by Daniel sen and Mohnen (1977), Lamb (1977),
and Davis and Jensen (1976).
3.4.7 Stratospheric Ozone at Ground Level
After a detailed review of background tropospheric ozone, Viezee and
Singh (1982) concluded, as also concluded by other investigators (e.g., Kelly
et al., 1982), that the stratosphere is a major but not the sole source of
background ozone in the unpolluted troposphere. This stratospheric ozone is
brought to the surface mixed layer by vertical mixing processes that have been
known for many years (Junge, 1963). In the northern hemisphere, at midlatitudes
between 30°N and 50°N, annual average background surface ozone concentrations
generally range between 30 and 50 ppb, but in the tropics, lower concentrations,
15 to 20 ppb, prevail (e.g., Fabian and Pruchniewicz, 1977; Oltmans, 1981).
The stratospheric ozone reservoir has a strong seasonal variation, with a
maximum in the spring and a minimum in fall and winter months, especially at
middle latitudes. This seasonal cycle is reflected at ground-level background
observation stations, where the average spring background ozone at some stations
may be as high as 80 ppb and the average fall values range between 20 and 40 ppb
(e.g., Singh et al., 1977; Mohnen, 1977; U.S. Environmental Protection Agency,
1978). In the troposphere, concentrations generally increase gradually to the
tropopause, but the seasonal pattern is the same (Viezee and Singh, 1982).
Using data acquired in their studies of TF events, researchers at SRI-
International examined the frequency with which stratospheric intrusions occur.
3-80
-------
According to Singh et al. (1980), an intrusion can be expected to be present
over the United States on about 90 percent of the days. This frequency
diminishes somewhat in summer, However (Singh et al., 1980), which is the
season when most ground-level smog episodes occur.
Viezee and Singh (1982) concluded that relatively high ozone concentra-
tions can occur for short periods of time, minutes to a few hours, over local
areas as a result of stratospheric intrusions. They were able to document
from published literature ten situations of probable intrusion of stratospheric
ozone. These instances are shown in Table 3-7, reproduced from Viezee et al.
(1983). The concentrations reported in Table 3-7 were measured at ground-level
stations. Note that all • of the short-term situations in which peak concen-
trations exceeded 80 ppb occurred in fall, winter, or spring months and not in
the photochemically active summer season. Of the three summer instances that
were reported, two at Whiteface Mountain, New York, and one at Pierre, South
Dakota, the highest reported concentration was 56 ppb for a 1-hr average. The
lower incidence in summer of reported ground-level impact by stratospheric
ozone may be attributable in part to the reduced frequency of intrusions in
summer, as reported by Singh et.aT. (1980). In addition, however, the potential
for ground-level impact by stratospheric ozone in summer is lessened because
of the stability provided by the upper-level, warm anticyclone present in the
weather systems characteristic of summertime photochemical smog episodes.
There have been a number of attempts to quantify the proportion of the
surface ozone attributable to stratospheric sources. The most common method
used is based on the assumption that Be is a unique tracer for air parcels of
stratospheric origin. Both ozone and Be are measured and the proportion of
the surface ozone that might be of stratospheric origin is calculated by using
a derived ratio between ozone and Be. Some of the investigators using this
technique include Ferman and Monson (1978), Wolff et al. (1979), Husain et al.
(1977), Dutkiewicz and Husain (1979), Singh et al. (1980), and Johnson and
Viezee (1981), among others.
Calculated correlations between surface ozone and Be show that their
relationship is variable. The results of Kelly et al. (1982), from studies of
Be, 0Q, and air mass classification in South Dakota, showed that continental
7
tropical (cT) air masses frequently seemed to have higher Be and 0^ values on
the western side of the traveling cT anticyclone. A similar relationship was
not prominent in the polar air masses studied, however, and maritime tropical
3-81
-------
TABLE 3-7. PUBLISHED EPISODES OF TRANSPORT OF STRATOSPHERIC OZONE TO GROUND LEVEL
Case
no.
1
2
3
4
5
c*>
CO
rog
4
10
6
7
8
Date
3 March 1964
26 February 1971
19 November 1972
6 March 1974
8,9 January 1975
11, 12 July 1975
19 March 1977
24, 25, 28 June and
1 July 1977
4 March 1978
July 1978
15 March 1978
Ground-level 03
Geographic location concentration, ppb
Quincy, Florida (near Tallahassee)
Observatory Hohenpeissenberg
(1000 m MSL), SW of Munich, Germany
Santa iosa, California
Harwell, Oxon, United Kingdom
Zugspitze Mountain, near Garaisch-
Partenkirchen, Germany (3000 in MSL)
Whiteface Mountain, New York
(1150 m MSL)
Sibton, Suffolk, United Kingdom
Whiteface Mountain, New York
Denver, Colorado
Pierre, South Dakota
Kisatchie National Forest, Louisiana
100 to 300
415
250
200 to 230
110 to 115
160 to 193
< 37
100 to 110
< 47
82
< 56
< 46
100 to 105
Duration of
observed event
3 hr
10 rain
50 rain
1 hr
2 hr
4 hr
24-hr average
2 hr
24- hr average
1 hr
1 hr
24-hr average
2 hr
Length of data
record examined
July 1963 through
July 1973
December 1970 through
May 1971
November 1972
4 to 5 yr discontinuous
August 1973 through
February 1976
July 1975
4 to 5 yr discontinuous
June and July 1977
1975 to 1978
July through September
1978
Spring 1978
Source
Davis and Jensen (1976)
Atmannspacher and
Hartmannsgruber (1973)
Lamb (1977)
Derwent et al. (1978)
Singh et al. (1980)
Husain et al. (1977)
Derwent et al. (1978)
Dutkiewicz and Husain
(1979)
Haagenson et al. (1981)
Kelly et al. (1982)
Viezee et al. (1982)
Source: Viezee et al. (1983).
-------
air masses did not reach the South Dakota site. Kelly et al. (1982) found a
linear correlation, r, equal to 0,65, for 1978 summer measurements of Be and
ozone in South Dakota. Ferman and Monson (1978) reported r = 0.60 at McKee,
Kentucky, for 27 daily samples taken during August and September 1976. Johnson
and Viezee (1981) reported an r value of about 0.50 for Be and ozone data
from Dodge City, Kansas, during April and May 1978. A much lower correlation,
r = 0.15, was reported by Husain et al. (1977) for July and August 1975 measure-
ments at Whiteface Mountain, New York. These data imply that the variability
of the surface ozone concentration that can be explained by Be varies from
about 40 percent to less than 5 percent.
Singh et al. (1980) and Viezee and Singh (1982) have pointed out a number
of problems with this technique in their detailed analyses of the application
of Be measurements to the quantification of the amount of stratospheric ozone
in surface air. Some of the problems encountered when applying Be/0,, ratios
over short sampling periods (in contrast, for example, to seasonal averages)
include the following:
1. Because 7Be is an aerosol and 03 is a gas, they have fundamentally
different atmospheric scavenging mechanisms and thus respond differ-
ently to tropospheric meteorological conditions.
2. Although the stratosphere is its dominant source, on the average,
7Be is synthesized in both the troposphere and the stratosphere.
The 7Be source cannot be assumed to be equivalent to the 03 source
without meteorological verification.
3. 7Be sampling data are primarily averaged over 24 hours, which does
not give sufficient time resolution for disaggregating short-period,
rapidly moving stratospheric intrusion events from longer-term
processes.
4. The lower stratosphere is a highly variable and poorly characterized
region for which no 7Be/03 ratio has been firmly established.
As a result of these and other factors, Singh et al. (1980) concluded that
"the experimental technique involving a Be/0- ratio to estimate the daily
O
stratospheric component of ground-level 0~ is unverified and considered to be
inadequate for air quality applications" (Singh et al., 1980, p. 1009). The
investigators at SRI-International have suggested, however, that Be may be
used, under the appropriate meteorological conditions, as a qualitative tracer
for air masses of stratospheric origin (Johnson and Viezee, 1981; Viezee et
al., 1979). ,
3-83
-------
Other methods have been used to estimate the quantitative contribution of
stratospheric ozone to ozone concentrations measured at ground-level. These
include: (1) aircraft observations of TF events (Johnson and Viezee, 1981),
coupled with calculations of the downward flux of ozone in the troposphere
90
(Viezee et a!., 1983); (2) use of mean Sr surface measurements, from radio-
active fallout studies, to calculate a mean stratospheric ozone contribution
(Reiter, 1978); and (3) examination of data, especially multiyear data, on
surface ozone concentrations at remote sites (Viezee and Singh, 1982; Viezee
et a!., 1983). Using surface measurement data and their own data from aircraft
measurements and calculated ozone fluxes, Viezee et al. (1983) concluded that
direct ground-level impacts by stratospheric ozone may be infrequent, occurring
<1 percent of the time; that such ground-level events are short-lived and
episodic; and that they are most likely to be associated with ozone concentra-
tions <0.1 ppm. Viezee et al. (1983) recommended further study, however, on
the possible contribution at ground-level of stratospheric ozone.
On a qualitative basis, as mentioned earlier, there is no doubt that
stratospheric ozone is present in the atmospheric surface layers, and the
meteorological mechanisms that bring this about have been described by a
number of investigators, including Wolff et al. (1979), Johnson and Viezee
(1981), and others. Most investigators cite the basic meteorological analyses
of Danielsen (1968) as a basis for their exchange model.
The downward transfer of air parcels.and ozone from the stratosphere into
the troposphere has been described above. There is, of course, a compensating
transfer of tropospheric air upward .into the lower stratosphere. Reiter
(1975) has examined various mechanisms that contribute to this transfer. Air
parcels moving out of the troposphere will carry with them the background
concentrations of ozone that they had in the troposphere, and, as the air
parcels mix in the stratosphere, these ozone molecules will become part of the
stratospheric background ozone. Since the ozone concentrations are very much
lower in the troposphere compared to the stratosphere, however, this exchange
of tropospheric and stratospheric air parcels will not result in a net upward
transport of ozone and is not considered to be: a factor.in air pollution
situations. ,
3.4.8 Background Ozone from Photochemical Reactions
The apportionment of the background ozone concentrations measured at
remote and other nonurban locations to stratospheric versus tropospheric
3-84
-------
processes and sources Is the subject of continuing discussion and research.
In addition, the apportionment of ozone produced photochemically in the
troposphere to natural versus manmade sources of precursors, and, in turn,
their respective contributions to background levels of ozone, also remain a
focus of discussion and research. The preceding discussion has emphasized the
thesis that the major source of tropospheric background ozone is the
stratosphere. It must be recognized, however, that some investigators argue
for a much larger role for the formation of significant amounts of ozone
within the troposphere (e.g., Fishman et a!., 1980; Fishman and Seiler, 1983;
Fishman and Carney, 1984; Fishman et a!., 1985; Fishman and Crutzen, 1978;
Chameides and Walker, 1976). The question of the potential relative impacts
of the stratospheric source of ozone and tropospheric photochemical sources
has been evaluated critically by Singh et al. (1978, 1980).
Singh et al. (1980) concluded that background ozone in the troposphere is
"principally of stratospheric origin," and supported this position with a
number of arguments, including the following. First, the NO concentrations
X
in the free troposphere appear to be very low, probably 0.01 to 0.1 ppb; at
NO concentrations <0.03 ppb a tropospheric ozone reservoir cannot be gener-
X
ated, using photochemical simulation models. Second, Singh et al. (1980)
pointed out that the tropospheric concentration pattern of ozone, with its
spring maximum and fall minimum, is out of phase and inconsistent with a
photochemical source, which would be at a maximum during the summertime peak
of solar radiation flux. Also, at low latitudes, where the solar flux is
relatively constant, the ozone seasonal cycle is quite pronounced and is
consistent with the seasonal cycle of stratospheric-tropospheric exchange
processes. In addition, Singh et al. (1980) noted that the vertical gradient
of ozone in the troposphere is contrary to a dominant photochemical source for
background tropospheric ozone. ;
Other evidence, however, especially more recent modeling studies, indicates
that a substantial part of the 03 measured during the warmer months of the
year over the United States and Western Europe is of photochemical origin
(Altshuller, 1986). Attempts to obtain agreement with observed 03 concentrations
with a general circulation model including only a stratospheric source of 03
have been unsuccessful, especially over continents (Levy et al., 1985). From
their tropospheric modeling studies, Fishman and eoworkers have predicted the
photochemical formation of near-surface, summer ozone at concentrations of:
3-85
-------
(1) 40 ppb, using surface NO emissions corresponding to a concentration of
X
~0.25 ppb (Fishman and Seiler, 1983; Fishman and Carney, 1984); (2) 15 ppb,
assuming zero NO emissions (Fishman and Seiler, 1983); and (3) 80 ppb, using
X
NO emissions corresponding to a concentration of about 5 ppb and ignoring
J\
contributions from individual urban or industrial plumes (Fishman et al.,
1985). Using another model, Dignon and Hameed (1985) have predicted the
photochemical production of ozone at midlatitudes of the Northern Hemisphere
at annual average concentrations of: (1) <60 ppb, using a 1980 NO emissions
/v X
inventory; or (2) ~25 ppb, assuming zero manmade NO emissions. On the basis
s\
of his review of these and other studies, Altshuller (1986) has estimated that
average stratospheric contributions to near-surface ozone concentrations range
from 10 to 15 ppb.
As these data and the preceding discussion in Section 3.4.7 indicate, the
portion of background ozone concentrations attributable to stratospheric
sources versus tropospheric photochemistry remains uncertain, especially in
the absence of a quantitative technique for determining whether ozone is of
stratospheric origin.
Investigations on the contributions of photochemistry to background ozone
have focused on (1) the role of transport from urban into nonurban areas,
which was discussed in Section 3.4 (see also Chapter 5); and (2) the respective
roles of biogenic VOC and of NO , from all sources, in the photochemical
s\
formation of ozone in nonurban areas.
Altshuller (1983) has evaluated the specific question of whether naturally
emitted VOC (i.e., from biogenic VOC sources) could be a significant source of
background ozone. Since biogenic emissions are released into the atmospheric
boundary layer, this potential source of ozone is expected to affect only
ground-level or boundary-layer concentration patterns, in contrast to the
whole troposphere as described in the preceding discussion. While a matter of
controversy in recent years, the data on the role of biogenic VOC in boundary-
layer photochemistry now appear more conclusive. Inventories of biogenic VOC
emissions indicate that they are of the same order of magnitude as manmade VOC
emissions, as described in Section 3.5.1 (below). On the other hand, concen-
tration data indicate that biogenic VOC occur, even in nonurban areas, at low
levels in ambient air relative to VOC from manmade sources (Section 3.5.2
below). In an extensive review of the literature, Altshuller (1983) noted
that the concentrations of biogenic hydrocarbons are very low, constituting
3-86
-------
even at rural sites probably 10 percent or less of the total nonmethane hydro-
carbons. He concluded that the contribution of biogenic hydrocarbons to ozone
formation, as a result, does not appear to be significant under most atmospheric
conditions.
The role of NO in the photochemical formation of ozone in nonurban areas
X.
has been the subject of a number of modeling and experimental studies, as well
as a recent review (Altshuller, 1986). As summarized in Section 3.5 (below),
and as reviewed by Altshuller (1986), NO concentrations in nonurban areas of
/\
the United States-are appreciably lower than those in urban areas, sometimes
by an order of magnitude. Likewise, concentrations of NO in the western
}\
United States are usually lower by about a factor of ten than those in the
northeastern states (Section 3.5.2.2.2). At nonurban locations inside of
populated areas, however, the concentrations of NO are much higher than those
f\:
measured in clean air within the continental and maritime boundary layer and
in the free troposphere (Altshuller, 1986).
Global background concentrations of NO were previously thought to be
/s.
lower than more recent measurements show them to be. For example, the analysis
cited earlier in this section (Singh et a!., 1980), on the contributions of
photochemistry versus stratospheric intrusions, assumed NO concentrations in
/x
the free troposphere of 0.01 to 0.1 ppb and an NO reservoir of <0.03 ppb; and
f\
the global models described in Kelly et al. (1984) assume an NO background of
0.1 to 0.2 ppb. In contrast, the mean NO concentrations tabulated by Altshuller
f\
(1986) for remote surface sites, while <1 ppb and often <0.5 ppb, are still in
excess of the ranges described above. At nonurban sites east of the Rocky
Mountains in the United States, mean NO concentrations range from 1 to 10 ppb
(Altshuller, 1986).
Martinez and Singh (1979) analyzed the role of NO in nonurban ozone
formation using theoretical approaches and aerometric data from the Sulfate
Regional Experiment. They concluded that the impact of NO on nonurban ozone
is a function of geographical location since ozone production was NO -limited
at some sites they examined but not at others. Kelly et al. (1984) examined
the role of photochemistry in nonurban ozone at three very different rural
sites (in South Dakota, Louisiana, and Virginia, respectively) by ambient air
analyses, captive air irradiations, and photochemical modeling. Ambient air
.analyses indicated the formation of about 6 ppb ozone per ppb of NO . Irradia-
/v
tion and modeling results indicated similar, but slightly higher, ozone forma-
tion per ppb of initial NO . They concluded that photochemistry in these
3-87
-------
rural areas was not NOX~"limited but depended, as well, on hydrocarbon concen-
trations. In a study that included modeling and a comparison of the modeled
predictions with observed ozone concentrations, Liu et al. (1984) arrived at a
similar conclusion; i.e., that nonurban ozone formation in the areas they
studied was not exclusively controlled by the equilibrium between NO and
• x^
ozone (Equation 3-4, Section 3.3.1) but by the hydrocarbon reactions, as well.
Consistent with the results of the Kelly et al. (1984) study, Parrish et al.
(1986) developed empirical relationships between ozone and NOV concentrations
: - "
(June to September) at a high-altitude location (Niwot Ridge, Colorado) and
found the photochemical formation of about 17 ppb ozone for each ppb of NO .
s\
The studies and analyses cited in this and the preceding section indicate
that research is still needed to permit the apportionment of background ozone
to stratospheric and respective tropospheric sources. From his review of data
pertinent to the role of NO in boundary-layer photochemistry, Altshuller
/\
(1986) concluded that photochemically generated ozone should equal or exceed
ozone transported down from the stratosphere to relatively remote locations at
lower elevations. At more polluted rural locations, photochemically generated
ozone from manmade emissions are predicted to constitute most of the ozone
measured during the warmer months of the year. Thus, the evidence supporting
the stratosphere as being the major (but not exclusive) source for background
tropospheric ozone is relatively strong but still not conclusive. Research in
these topic areas is continuing, and the number of uncertainties and necessary
assumptions needed for the respective conclusions cited above should be reduced
as new data become available.
3.5 PRECURSORS TO OZONE AND OTHER PHOTOCHEMICAL OXIDANTS
3.5.1 Sources and Emissions
3.5.1.1 Manmade Sources and Emissions. This section presents information on
the manmade sources and emissions of precursors to ozone and other photochemical
oxidants. Estimates of annual emissions are presented by source category for
volatile organic compounds and nitrogen oxides. In addition, information is
presented on emission rates and the composition of emissions from principal
stationary sources and from mobile sources. The annual estimates are based on
emission inventories prepared according to procedures developed and published
by EPA (U.S. Environmental Protection Agency, 1980a,b; 1981a,b,c,d,e,f).
3-88
-------
3.5.1.1.1 Trends inemissionsof volatile organic compounds. Emission data
on volatile organic compounds (VOC) include data on hydrocarbons as well-as
other organic compounds found in ambient air. Because of their negligible
reactivity toward the photochemical production of ozone, methane, ethane,
methylene chloride, and several halogenated methanes and ethanes are also
excluded from emission inventories (U.S. Environmental Protection Agency,
1980b). Estimates of total emissions of volatile organic compounds (VOCs)
provide a gross measure of compounds .available for photochemical production of
ozone and other photochemical oxidants. Emissions of VOCs are reported here
as the collective mass of reactive VOC. (See the footnote to Table 3-8 for
the calculation of emissions-based NMOC/NO ratios.)
/v
Retrospective estimates of total manmade VOC emissions in the United
States, based on records of economic activity (e.g., fuel usage, industrial
production) have been prepared by-decade, beginning with 1940 (U.S., Environ-
mental Protection Agency, 1986). From a level of 18.5 Tg/yr in 1940, VOC
emissions increased about 14 percent each decade through 19704 then began to
decline (Table 3-8).
Figure 3-14 shows national trends in yearly emissions of manmade VOC by
general source category for the period 1970 through 1983 (U.S. Environmental
Protection Agency, 1984a). Emissions in 1983 of manmade VOCs in the. United
States have been estimated at 19.9 Tg/yr (U. S. Environmental Protection
Agency, 1984a); total manmade VOC emissions nationwide were 26 percent Jower
in 1983 than in 1970. The observed decrease is attributable largely, to a
decrease of 30 percent in estimated VOC emissions from highway vehicles during
this period. This decrease occurred in spite of a 42 percent increase in
vehicle miles traveled. Trends in manmade VOC emissions versus vehicle miles
traveled, urban only and total, for 1970 through 1983,.are shown in Figure
3-15 (U.S. Environmental Protection Agency, 1984a; Motor Vehicle Manufacturers
Association, 1984). The main sources nationwide of manmade VOC are industrial
processes, which emit a wide variety of VOCs such as chemical solvents; and
transportation, which includes the emission of VOCs in gasoline vapor as well
as in gasoline combustion products. .
3.5.1.1.2 Trendsin emissions of nitrogen oxides. Emissions of nitrogen
oxides (NO ) are reported here as the sum. of NO and NO,,, all expressed as
^\ c.
equivalent N09. Retrospective estimates of total manmade NO emissions in the
c. X,
3-89
-------
TABLE 3-8. EMISSIONS OF VOC BY DECADE, 1940 THROUGH 1980
(Tg/yr)a
Source category
Transportation
Stationary fuel combustion
Industrial processes
Solid waste
Miscellaneous
Total
1940
5.2
4.7
3.2
0.9
4.5
18.5
1950
7.9
3.1
5.2
1.0
3.6
20.8
1960
11.1
1.9
6.1
1.4
3.1
23.6
1970
12.3
1.0
8.7
1.8
3.3
27.1
1980
8.2
2.1
8.9
0.6
2.9
22.7
Source: U.S. Environmental Protection Agency (1986).
aTo calculate emissions-based NMOC/NO molar ratios, as used in control-
related data, a multistep procedure ^s required (Novak, 1986) as described
by the equation:
4
Total NMOC (Tg/yr) x SUM[Pct. x no../mw-]/[Total NO (Tg/yr)/46]
i=l 111 x
where Pet = percent of a compound class present in the NMOC emissions;
no. = average carbon number for a compound class;
mw = average molecular weight for a compound class;
i represents the four compound classes of interest; and
NO emissions are expressed as N02 (mw = 46).
/\
Values for terms in the equation vary in the published literature. Example
values, where no. and mw are derived from emissions inventories (Novak, 1986)
and pet from ambient air concentration data (Richter, 1986), are: paraffins
= [63.5 x 4.56/95]; olefins = [15.5 x 3.57/43]; aromatics = [15.5 x 7.56/97];
and aldehydes = [2.0 x 1.5/37].
United States have been prepared by decade, beginning with 1940 (U.S. Environ-
mental Protection Agency, 1986). From a level of 6.8 Tg/yr in 1940, NO
r\
•emissions increased almost 40 percent each decade through 1970 (Table 3-9).
The increase from 1970 to 1980 was 13 percent.
Total NO emissions in the United States in 1983 (19.4 Tg/yr) were some
/\
17 percent above the 1970 level, but appear to have declined slightly from
about 1980 through 1983 (Figure 3-16) (U.S. Environmental Protection Agency,
1984a). The net increase over the period 1970 through 1983 may be attributed
primarily to two causes: (1) increased fuel combustion in stationary sources
such as power plants; and (2) increased fuel combustion in highway motor
vehicles, as the result of the increase (42 percent) in vehicle miles driven
over the 14 years in question. Emissions associated with industrial processes
3-90
-------
30
25
20
05
O
15
ui
>
=>
O
10
i i r
TRANSPORTATION
INDUSTRIAL PROCESSES, STATIONARY SOURCES
<' SOLID WASTE
NON-INDUSTRIAL SOLVENTS
it [MISCELLANEOUS
1971 1972 1373 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
YEAR
Figure 3-14. National trend in estimated emissions of volatile organic compounds, 1970 through 1983.
Source: U.S. Environmental Protection Agency (1984a).
3-91
-------
160
150
140
130
2120
Q.
GO"
HI
D
< 110
100
o
oc
u.
LU
1 go
o
80
70
60
50
1 T
.VEH. MILES (URBAN)
•VEH. MILES (TOTAL)
•NOX
voc
\
s
I I
J I
J I
1971
1973
1975
1977
YEAR
1979
1981
1983
Figure 3-15. Comparative trends in highway vehicle emissions of nitrogen oxides (NOx) and
volatile organic compounds (VOC) versus vehicle miles traveled, 1970-1983.
Source: Motor Vehicle Manufacturers Association (1984); U.S. Environmental Protection
Agency (1984a).
3-92
-------
TABLE 3-9. EMISSIONS OF NO BY DECADE, 1940 THROUGH 1980
(Tff/yr)
Source category
Transportation
Stationary fuel combustion
Industrial processes
Solid waste
Miscellaneous
Total
1940
2.2
3.4
0.2
0.1
0.9
6.8
1950
3.5
4.7
0.3
0.2
0.6
9.3
I960
4.9
6.7
0.5
0.3
0.4
12.8
1970
7.6
9.1
0.7
0.4
0.3
18.1
1980
9.2
10.2
0.7
0.1
0.2
20.4
Source: U.S. Environmental Protection Agency (1986).
remained relatively constant, but solid waste and miscellaneous emissions
decreased slightly.
The national trends shown do not reflect the considerable local and
regional differences that exist in the relative amounts of NO emitted in the
s\
major source categories. For example, motor vehicle emissions in Los Angeles
County, California, increased sixfold from 1940 to 1970 (Los Angeles County,
1971), compared to a threefold national increase. Figure 3-15 compares the
relative trends in mobile source NO emissions versus vehicle miles traveled,
f\
urban only and total, from 1970 through 1983 (U.S. Environmental Protection
Agency, 1984a; Motor Vehicle Manufacturers Association, 1984).
3.5.1.1.3 Sources and emissions of volatile organic compounds. The source
category contributing the largest percentage of VOC emissions in 1983, 37.7
percent, is Industrial Processes (U.S. Environmental Protection Agency, 1984a).
The category consists almost entirely of point sources. The composition of
these emissions varies widely, depending on the process or product and the use
of emission reduction equipment and operating practices.
The second largest VOC source category, Transportation, accounting for
36.2 percent of the annual total in 1983, is discussed below.
The third largest VOC source category, Miscellaneous, accounts for 12.6
percent of the annual total, over half of which consists of the subcategory,
Miscellaneous Organic Solvents (U.S. Environmental Protection Agency, 1984a).
These emissions generally qualify as area-source emissions. Some of these
solvents are widely used in domestic products such as furniture polish, shoe
polish, shaving soap, perfumes, cosmetics, shampoo, hair spray, hand lotion,
3-93
-------
30
25
20
I
W-
O
55
15
UJ
I
i
=3
O
10
TRANSPORTATION
FUEL COMBUSTION (STATIONARY)
INDUSTRIAL PROCESSES, SOLID WASTE. MISCELLANEOUS I
I I I i i I I I
1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
YEAR
Figure 3-16. National trend in estimated emissions of nitrogen oxides, 1970 through 1983.
Source: U.S. Environmental Protection Agency (1984a).
3-94
-------
rubbing alcohol, and nail polish remover. The predominant compounds emitted
are isopropyl alcohol and ethyl alcohol (Bucon et a!., 1978).
Emissions of volatile organic compounds from the production and marketing
of gasolines and motor oils are classed in the Industrial Processes category
of VOC emissions. Emissions of VOC from these petroleum products after their
sale to vehicle owners are included in the Transportation category.
A significant portion of VOC emissions from gasoline-fueled mobile sources
arises from evaporation, but the most conspicuous mobile source emissions are
the combustion products. In a study by Black et al. (1980), evaporative
emissions were found to constitute one-third to one-half of total hydrocarbon
emissions from all of the vehicles tested (n = 60). Based upon actual
surveillance tests of in-use vehicles, Fisher (1980) found emission rates of
2.53 g/mi in 1968 and 0.15 g/mi in 1980 for composite evaporative and crankcase
emissions from gasoline-fueled cars.
Exhaust emissions from gasoline-fueled vehicles typically contain fuel
components and low-molecular-weight hydrocarbons that are not present in the
fuel. Typically, exhaust from a catalyst-equipped automobile contains about
62 percent alkanes, 17 percent aromatics, 18 percent alkenes, and 3 percent
acetylene. This may be compared with the corresponding typical values for
automobiles without converters: 40, 24, 26, and 11 percent, respectively.
Methane levels generally range from about 10 to 30 percent (Black and Bradow,
1975; Black, 1977). Exhaust gases from gasoline-fueled vehicles also contain
organic compounds such as aldehydes, ketones, ethers, esters, acids, and
phenols, amounting to as much as one-tenth of the total VOC content.
Factors other than gasoline composition influence the composition of
exhaust from internal combustion engines. These include driving patterns, the
specific configuration of emission control devices, ambient temperature and
humidity, and, of course, individual automobile parameters such as tuning,
make, and model year. Fuel additives can also influence emissions. For
example, in one study, tetraethyl lead increased hydrocarbon emissions by
about 5 percent but did not change the type of emissions (Leihkanen and Beckman,
1971).
Evaporative emissions from diesel vehicles are negligible because of low
fuel volatility (Linnell and Scott, 1962; McKee et al., 1962). In studies of
exhaust emissions from diesel automobiles (Black and High, 1979; Gibbs et al.,
1983), 15 to 40 percent of the hydrocarbons emitted were found to be associated
3-95
-------
with particles by the time the exhaust stream exited the tailpipe. Gibbs et
al. (1983) reported THC emissions from 19 in-use diesel automobiles, repre-
senting 1977 to 1979 model years, that were tested periodically over a 28-month
period. Emissions of THC at the end of the period ranged from 0.17 to 0.88
g/mi for individual vehicles and averaged 0.65 g/mi. It should be noted here
that the population of diesel-powered passenger cars is not growing as rapidly
as expected. Sales of diesel-powered cars peaked at 6 percent for 1981 models
and dropped to less than 3 percent by 1983 (Automotive News, 1982a,b,; 1983a,b;
Plegue, 1983).
A brief summary of both exhaust and evaporative emissions characteristic
of a variety of engines, fuel types, and control devices has been presented by
filton and Bruce (1981). Their summary indicates, for example, that exhaust
emissions of aldehydes and ethylene increase in cars fueled with ethanol-
gasoline blends (gasohol) as opposed to gasoline. Exhaust emissions of total
hydrocarbons (THC) decrease but evaporative emissions increase, for a net
increase in emissions from gasohol-fueled cars. In diesels, evaporative THC
emissions are virtually nonexistent, and net exhaust emissions of THC are
lower. The percentage of carbonyl compounds (aldehydes, ketones) in the
exhaust of diesels is higher than from gasoline combustion, but the net
photochemical reactivity per gram of emissions is lower nevertheless (Tilton
and Bruce, 1981).
In view of changes in emissions with fuel type, it is of interest to
examine fuel usage in the United States in recent years. Yearly sales of
vehicle fuels fluctuate in consequence of several factors, including retail
fuel price, general economic conditions, and the age and fuel efficiency of
the composite vehicle population. In addition to the year-to-year influence
of these factors on the sales of the principal vehicle fuels, gasoline and
diesel fuel, a change in fuel composition is emerging with the introduction in
the United States of gasolines containing a percentage of ethanol. As shown
in Table 3-10, sales of gasoline containing up to 10 percent ethanol (gasohol)
were first reported in 1981 and had increased almost sixfold 2 years later
(NPN: National Petroleum News, 1981-1985 summary issues).
3.5.1.1.4 Sources and emissions of nitrogen oxides. Fuel combustion is the
dominant source of NO emissions nationally. Stationary sources contributed
/\
50 percent and mobile sources contributed 45.4 percent in 1983 (U.S. Environ-
mental Protection Agency, 1984a). In contrast to their contributions to VOC
3-96
-------
TABLE 3-10. YEARLY QUANTITIES OF-MOTOR VEHICLE FUELS SOLD IN
THE UNITED STATES FOR HIGHWAY USE, 1980 THROUGH 1983
Gallons x 109
Fuel
Gasoline
Gasohol
Diesel
1980
106.45
->-
13.60
1981
104.14
0.72
14.30
1982
100.70
2.16
14.67
1983
101,55
4.29
15.87
Source: NPN: National Petroleum News (1981-1985 summary issues).
emissions, industrial process sources contributed only about 3 percent to the
national total of manmade NO emissions in 1983 (U.S. Environmental Protection
r\
Agency, 1984a).
Ratios of NO/NO in emissions vary depending upon the source. Nitric
"
oxide (NO) is the dominant oxide of nitrogen emitted by most sources; NOp
generally comprises less than 10 percent of the total NO emissions. More than
A
30 to 50 percent, however, of the total NO emissions from certain diesel
/\
(Braddock and Bradow, 1976; Springer and Stahman, 1977 a,b) and jet turbine
engines (Souza and Daley, 1978) can be NO,, under specific load conditions.
Likewise, tail gas from nitric acid plants, if uncontrolled, may contain about
50 percent NO, (Gerstle and Peterson, 1966). Variations in NO/NO ratios by
£. X
source type can be significant in local situations, as, for example, in the
immediate vicinity of a high-volume roadway carrying a significant number of
diesel-powered vehicles.
Emissions of NO from the principal categories of stationary combustion
-------
in the summer than in the spring (U.S. Department of Energy, 1978). Greater
degrees of variation and different seasonal patterns have been reported for
stationary sources in different regions of the country (California Board of
Sanitation, 1966).
Emissions of NO from mobile sources, gasoline- and diesel-fueled vehicles,
* if\
are affected by a number of variables such as speed, load, and air:fuel ratio
(APR), as reported by Billiard and Wheeler (1979). Seasonal variations in NO
J\
emissions from mobile sources will occur in relation to temperature (about a
35 percent decrease in emissions per vehicle mile with an ambient temperature
increase from 20 to 90°F) (Ashby et a!., 1974), and number of vehicle miles
traveled (about 18 percent higher in summer than in winter, nationwide) (Federal
Highway Administration, 1978). Emissions of NO also vary with vehicle miles
^\
traveled in urban versus rural areas and among states in different regions of
the country (Federal Highway Administration, 1978). Diurnal variations in NO
-A.
emissions associated with motor vehicle traffic are especially important
because of their potential impact on ambient air quality. Table 3-11 summa-
rizes data on NO emissions from mobile sources.
TABLE 3-11. SUMMARY OF NOX EMISSIONS FROM MOBILE SOURCES
Vehicle type
Gasoline-fueled
passenger cars
Emissions, g/mi
0.41 to 1184
Comments
In-use; equipped
with three-way
Reference
Smith and
Black, 1980
Diesel passenger
cars
Heavy-duty trucks
Gasoline (leaded)
Diesel
0.84 to 3.15
12.29 to 15.58
17.47 to 42.40
catalysts; four
test cycles used
In-use, 1977 to 1979
model years; tested
at ca. 35,000 mi
One test cycle
One test cycle;
several diesel fuels
Gibbs et al.,
1983
Dietzman et al.,
1980, 1981
3.5.1.2 Natural Sources and Emissions
3.5.1.2.1 Natural sources and emissions of volatile organic compounds. This
section presents a brief overview of the nature and quantity of hydrocarbon
emissions from biogenic sources. For detailed information, the reader is
3-98
-------
referred to several excellent review articles (Altshuller, 1983; Bufalini and
Arnts, 1981; DimitHades, 1981).
To date, isoprene and the monoterpenes are the only biogenic hydrocarbons
identified as emissions from vegetation (e.g., Sanadze and Dolidze, 1962;
Rasmussen, 1964, 1970; Evans et al., 1982). Other volatile organic emissions
from vegetation have been documented, but consist of oxygenated organic
compounds, e.g., camphor and 1,8-cineole (see, e.g., Graedel, 1978). Isoprene
and the monoterpenes are of interest to atmospheric scientists largely because
they are volatile enough to be released under normal environmental conditions,
because they are abundant relative to other biogenic VOCs, and because they
have been shown to be potential ozone precursors. The commonly identified
monoterpenes are a-pinene, p-pinene, camphene, A3-carene, limonene, myrcene,
and p-phellandrene. As a general rule, coniferous trees emit primarily
monoterpenes, and deciduous trees emit isoprene.
Biogenic emission rates. Biogenic emission rates have been determined
almost exclusively by techniques that involve enclosing the entire plant or a
portion of it, such as a branch of a tree, in a bag or chamber constructed of
light, transparent material. Isoprene and monoterpene emission rates given
here [as ug (g dry biomass) hr ] were measured by the enclosure method.
Because biogenic emission rates are species- and temperature-dependent, the
species tested and the enclosure temperature are also given.
Isoprene emission rates reported by Evans et al. (1982) for various
species of trees range from 3 pg g hr for spruce (Picea sitchensis, 28°C)
to 233 ug g"1 hr"1 for willow (Salix babylonica, 30°C). Reported emission
—I —1
rates for oak range from 9 |jg g hr (Quercus virginiana, 30°C; Zimmerman,
1979) to 49 ug g"1 hr"1 (Quercus agrifolia, 30°C; Winer, 1983). The reader is
referred to the studies by Zimmerman (1979), Evans et al. (1982), and Winer
(1983) for additional rate data and experimental details.
Monoterpene emission rates are probably best-documented for pine. Rates
obtained by Winer (1983) ranged from 0.6 ug g hr (Pinus radiata and Pinus
halepensis, 30°C) to 2 jjg g hr (Pinus canariensus, 30°C). For Pinus taeda,
Arnts et al. (1978) reported a rate of 4 ug g hr (at 30°C) and Knoppel et
al. (1982) reported a rate of 1 ug g hr (at 30°C). Among the higher monoter-
_1 —i
pene emission rates are those for Pinus clausa, 11 ug g hr at 30°C
(Zimmerman, 1979) and for Douglas fir (Pseudotsuga taxifolia), 15 ug g hr
at 30°C (Knoppel et al., 1982). See also the studies by Rasmussen (1972) and
by Evans et al. (1982) for additional monoterpene emission rate data.
3-99
-------
Besides temperature, biogenic emission rates are affected by other environ-
mental factors. Rasmussen (1972) reported that emission rates varied with
species, plant maturity, resin gland integrity, and leaf temperature. Dement
et al. (1975) found that the emission rate of monoterpenes from Sal via mellifera
(California Black Sage) is dependent on the vapor pressures of the terpenes,
the humidity, and the amount of oil present on the surface of the leaf; but
found that the emission rate is not directly dependent on the photosynthetic
activity or on the stomatal openings of the plant.
Tingey and his coworkers, in extensive studies on factors effecting
isoprene emission rates in live-oak seedlings (Tingey et al., 1981), found
light intensity to be a chief determinant of emissions, which decreased to
near zero levels in the dark. In contrast to isoprene, monoterpene emission
rates did not appear to be influenced by light intensity but were affected by
temperature (Tingey et al., 1980); a log-linear increase in emission rates of
monoterpenes with temperature was observed in studies of slash pine.
The validity of emission rate data obtained by the bag enclosure technique
has been widely discussed because of uncertainties associated with (1) isolating
the vegetation in an artificial environment; (2) possible damage to isolated
vegetation; (3) representativeness of emission rates measured from just one
branch; and (4) relationship of emission rates to ambient air concentrations
of biogenic hydrocarbons.
Attempts to validate the bag enclosure method have focused on comparing
enclosure emission estimates with those obtained by alternate procedures, such
as micrometeorological gradient procedures (e.g., Lamb et al., 1983; Knoerr
and Howry, 1981). In a study of Pennsylvania hardwood forest, the gradient
-2 -1
profile procedure gave a flux of 8,000 [jg m hr , while the enclosure tech-
-2 -1
nique yielded 7,300 |jg m hr (Lamb et al., 1983). Good agreement has been
reported, also, for orpinene emission fluxes measured by a micrometeorological
procedure and by the enclosure method (Knoerr and Mowry, 1981).
Although the micrometeorological approach yields mass fluxes similar to
the enclosure method, it, too, has certain limitations. For example, the
measurement of small vertical gradients above a forest canopy and the applica-
tion of surface layer theory to non-ideal sites can lead to erroneous results.
Such difficulties can largely be avoided by simulating the forest emissions
with an inert tracer release (such as SFg) and measuring ambient concentrations
of the tracer and biogenic gases along downwind sample lines (e.g., Arnts and
3-100
-------
Meeks, 1981). Isoprene fluxes obtained using the tracer procedure in a central
Washington oak grove compared well with flux estimates determined simultane-
ously with the enclosure technique (Allwine et a!., 1983).
Biogenic emission Inventories. Development of a biogenic emission inven-
tory requires knowledge of (1) emission rates for individual species; (2) the
vegetation coverage of an area, by species; (3) leaf biomass per tree (derived
through allometric equations); and (4) a biomass factor for the forested area
(derived from (2) and (3).
Table 3-12 contains a listing of area-wide biogenic emission fluxes that
have been reported for the United States and portions thereof. Accuracy of
the estimates reported in Table 3-12 depends upon the size of the area for
which the inventory has been prepared. Many of the problems and uncertainties
encountered in preparing inventories have been discussed in detail in the
literature (Altshuller, 1983; Zimmerman, 1979, 1980; Wells, 1981; Box, 1981;
Dimitriades, 1981).
TABLE 3-12. AREA-WIDE BIOGENIC EMISSION FLUXES
Location
Emission
flux,
jq m~2hr"1 Comment
Reference
South Coast Air Basin, <780
California
Lake Tahoe, California 1950
Lake Tahoe, California 2438
San Francisco Bay Area, 1388
California
San Francisco Bay Area, 2265
California
San Francisco Bay Area, 777
California
Tampa/St. Petersburg, 2540
Florida
Southeastern Virginia 8890
Pennsylvania 1660
Houston, Texas 1170
United States 1712
United States 1099
United States 884
Entire basin
Forested area
of basin
Daytime
Nighttime
Forested area only
Winer (1983)
JSA, Inc. (1978)
JSA, Inc. (1978)
Sandberg et al.
(1978)
Hunsaker (1981)
Hunsaker (1981)
Zimmerman (1979)
Salop et al. (1983)
Flyckt et al. (1980)
Zimmerman (1980)
Marchesant et al.
(1970)
Zimmerman (1977)
Zimmerman (1978)
3-101
-------
Considering all the variables that affect biogenic emissions and their
inventories, it is somewhat surprising that the area-wide emission fluxes
listed in Table 3-13 show no more variation than they do. With the exception
of the Southeastern Virginia area, which is a forested region with high biomass
coverage, most of the values in Table 3-12 differ by less than a factor of
three.
3.5.1.2.2 Natural sources and emissions of nitrogen oxides. Natural emissions
of nitrogen oxides (NO ) originate from the oxidation of nitrogen gas by
J\
electrical discharge in the atmosphere, from the ammonifi cation of organic
nitrogen during biological decomposition, and from the oxidation of organic
nitrogen during forest fires. Nitrogen fixation and electrical discharge are
normal processes of the nitrogen cycle that convert inert nitrogen gas to
biologically useful nitrate or ammonia. For a discussion of the nitrogen
cycle relative to NO emissions, the reader is referred to Air Quality Criteria
-------
TABLE 3-13. GLOBAL ESTIMATES OF NITROGEN TRANSFORMATION
(Tg N/yr)
Range of estimates
References'
Biological fixation
(N2 •* NH4 )
54 to 270
al. Delwiche (1970).
2. Burns and Hardy (1975).
3. Soderlund and Svensson (1976).
4. Robinson and Robbins (1975).
5. Liu et al. (1977).
6. Crutzen and Ehhalt (1977).
7. Noxon (1976).
8. Sze and Rice (1976).
9. Council for Agricultural Science and Technology (1976).
10. Chameides et al. (1977).
1-5
Electrochemical fixation
lightning (N2 -> NO )
atmospheric (N2 -* N02)
Biological denitrification
(N03" -*• N2)
(N03~ -> N20)
combined
Industrial denitrification
(Organic -N •» NO )
(Other -»• N0x) x
Atmospheric denitrification
(NH3 -» NOX)
Natural NO emissions
from lana and sea
NHs emissions to atmosphere
from land and sea
10 to 40
14 to 20
96 to 190
20 to 340
83 to 270
14 to 19
30 to 36
3 to 30
40 to 210
110 to 850
1
2, 6, 7, 8, 10
2, 4
2, 3
2-4
5, 8, 9 -. '
2-4
2, 3, 5
2, 3
3, 4
2-4
3-103
-------
are virtually nonexistent. In addition, scaling of emissions from such sources
as bacterial nitrification and denitrification for use in preparing area-wide
emission inventories is not possible. Thus, the emissions reported in this
section should be taken as very gross approximations that serve to identify
natural sources of NO and to present the estimated relative magnitude of
X
emissions from such sources.
3.5.2 Representative Concentrations of Ozone Precursorsin Ambient Air
As discussed earlier in this chapter (Section 3.3), nonmethane organic
compounds (NMOC) and the oxides of nitrogen (NO ) in the presence of sunlight
X
react to form ozone and other photochemical oxidants. The reaction sequence is
complex, so that dependable precursor-oxidant relationships are difficult to
establish. Factors such as absolute NMOC and NO concentrations, relative
X
NMOC and NOV concentrations (NMOC/NOV ratios), NMOC reactivity, and NOV compos!-
X X X
tion are known to affect the photochemical reactions that produce ozone and
other oxidants in ambient atmospheres. Concentration-based NMOC/NO ratios
X
are used in some precursor-oxidant models. Ratios of NMOC/NO that are
X
concentration-based are calculated directly, using measured concentrations of
NO expressed as N0? (ppm or ppb) and measured concentrations of hydrocarbons
f\ **»
expressed as carbon (ppm or ppb C). The latter is easily obtained from gas-
chromatographic measurements, since the chromatograph yields a known response
per carbon atom. This section provides summaries of NMOC and NO concentra-
A
tions recorded at various urban and nonurban locations in the United States.
In addition, HC/NO (or NMOC/NO ) ratios are given for some urban areas.
X- X
3.5.2.1 Concentrations of Nonmethane OrganicCompoundsin AmbientAir. The
NMOC data in this section are segregated into (1) nonmethane hydrocarbons
(NMHC) and (2) oxygenated hydrocarbons. The concentrations reported here for
NMOC were obtained by gas chromatographic methods for the identification and
quantification of individual NMOC species (see Chapter 4). A fairly substantial
data base exists for characterizing urban nonmethane hydrocarbon concentrations.
Measurements of nonurban hydrocarbon levels, as well as both nonurban and
urban oxygenated hydrocarbons, are much more limited. Among oxygenated hydro-
carbons, aldehydes have received the most attention. Insufficient information
exists for establishing ambient air concentrations of other classes of oxygenated
hydrocarbons such as alcohols, ketones, acids, and ethers.
3-104
-------
3.5.2.1.1 Urban nonmethane hydrocarbon concentrations. Most of the data on
ambient air concentrations of nonmethane hydrocarbons (NMHC) have been obtained
during the 6:00 to 9:00 a.m. time period. Since urban hydrocarbon emissions
peak during that period of the day and atmospheric dispersion is limited,
these concentrations generally reflect maximum diurnal levels. Table 3-14
lists the mean and range of NMHC concentrations recorded in a number of urban
areas throughout the United States. For most urban areas included in the
table, a mean NMHC value between 400 and 900 ppb C was observed, though mean
concentrations in some cities (e.g., Houston, Las Vegas, and Los Angeles) are
in excess of 1000 ppb C. The data in Table 3-14 are not meant to serve as a
comparison of NMHC levels in various cities but rather are shown to indicate
the mean and range of concentrations that have been reported. Comparisons are
invalid because of major differences in sample numbers, site classifications,
and seasonal sampling periods. In many cases, the range of values reported in
Table 3-14 might not reflect the true maximum and minimum concentrations that
occur in a particular urban area. Most of the hydrocarbon sampling programs
were of short duration (~1 month) and in some cases were not operated on a
daily basis. For example, the relatively high mean values reported in Las
Vegas are undoubtedly the result of the fact that ambient air samples were
only analyzed for hydrocarbons on days when conditions were appropriate for
oxidant formation. It is probably safe to assume, however, that NMHC levels
during the 6:00 to 9:00 a.m. time period in major urban areas will usually
exceed 50 ppb C but seldom surpass 10,000 ppb C.
Species in the C2-C-,g molecular-weight range dominate the hydrocarbon
composition of urban atmospheres, with the alkanes generally constituting 50
to 60 percent of the hydrocarbon burden, aromatics 20 to 30 percent, and
alkenes and acetylene making up the remaining 5 to 15 percent (Sexton and
Westberg, 1984). The alkane fraction is usually dominated by species in the
C.-Cfi molecular-weight range. Predominant aromatics include benzene, toluene,
ethyl benzene, and the xylenes. The most abundant alkenes are ethylene and
propene. The studies cited in Table 3-14 provide information on the individual
species of NMHC found in urban atmospheres.
3.5.2.1.2 Nonurban nonmethane hydrocarbon concentrations. Nonurban nonmethane
hydrocarbon concentrations are generally one to two orders of magnitude lower
than those measured in urban areas (Ferman, 1981; Sexton and Westberg, 1984).
Concentrations of individual species seldom exceed 10 ppb C. Total hydrocarbon
3-105
-------
TABLE 3-14. NONMETHANE HYDROCARBON CONCENTRATIONS
MEASURED BETWEEN 6:00 and 9:00 a.m. IN VARIOUS UNITED STATES CITIES
City
(Date)
Atlanta (1981)
Baltimore (1980)
Boston (1980)
Cincinnati (1981)
Detroit (1981) •
Houston (1976)
Houston (1978)
Las Vegas (1980)
Las Vegas (1983)
Los Angeles (1968)
Los Angeles (1982)
Milwaukee (1981)
Newark (1980)
New York (1969)
Philadelphia (1979)
St. Louis (1973)
Tulsa
Washington, DC (1980)
Mean
NMHC
concn. ,
ppb C
491
659
569
840
330
1414
1679
2506
2762
3388
2920
324
732
830
669
817
426
671
Range
113 to 1677
51 to 2798
83 to 4750
260 to 1870
60 to 1720
356 to 16,350
400 to 4500
689 to 4515
1835 to 4590
N.A.a
390 to 6430
24 to 3116
89 to 6946
N.A.
305 to 1710
N.A.
103 to 3684
210 to 2953
Reference
Westberg and Lamb (1983)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Holdren et al. (1982)
Kelly et al. (1986)
Sexton and Westberg (1984)
Lonneman (1979)
Nay lor et al. (1981)
Naylor et al. (1984)
Lonneman (1977)
Grosjean and Fung (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Lonneman (1977)
Sexton and Westberg (1984)
Lonneman (1977)
Eaton et al. (1979)
Sexton and Westberg (1984)
Data are not available.
concentrations range up to ~15Q ppb C, but usually fall in the range of about
5 to 100 ppb C. Alkanes comprise the bulk of species present, with C«-Cr
compounds most abundant. Ethylene and propene are occasionally reported at
concentrations of 1 ppb C or less, and toluene is usually present at ~1 ppb C.
Table 3-15 provides a summary of the range of hydrocarbon concentrations
measured at various nonurban locations in the United States. Samples were
carefully selected at most of the sites in order to guarantee their nonurban
character. At the coastal and near-coastal sites, only those samples collected
upwind of manmade sources (onshore advection) were included. The nonmethane
3-106
-------
TABLE 3-15. NONMETHANE HYDROCARBON CONCENTRATIONS
MEASURED IN NONURBAN ATMOSPHERES
Location
Belfast, ME
Benicia, CA
Miami, FL
Glascow, IL
Janesville, WI
Houston, TX
Robinson, IL
Smoky Mtns. , TN
Northern Idaho
Virginia
Atlanta (urban)
Whiteface Mtn. ,
NY
Elkton, MO
Eastern TX
North Carolina
Colorado
Species
analyzed
C2 - Cg
^2 ~ £5
C2 ~ C5
£-2 ~ £3.0
^2 ~ CIG
£•2 ~ ClO
^2 " CIQ
^2 ~ CIQ
Terpenes
Isoprene
Isoprene
Terpenes
Isoprene
ot-pinene
a-pinene
Terpenes
Concentration
range, ppb C
10
7
2
60
9
2
13
38
0.1
4
0
6
0
0.1
0.6
0
to 22
to 14
to 23
to 150
to 24
to 24
to 113
to 149
to 18
to 150
to 8
to 84
to 28
to 8
to 13
to 8
Reference
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Chatfield and Rasmussen (1977)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Sexton and Westberg (1984)
Cronn (1982)
Holdren et al. (1979)
Ferman (1981)
Westberg and Lamb (1983)
Whitby and Coffey (1977)
Rasmussen et al. (1976)
Seila (1981)
Seila (1981)
Roberts et al. (1983)
hydrocarbon concentrations reported at coastal sites (Belfast, Benicia, Miami,
and Houston) are definitely lower than those measured at most of the inland
sites. It should be pointed out, however, that the numbers of samples measured
for each of the nonurban locations listed in Table 3-15 is small. This,
coupled with the fact that only a limited range of hydrocarbons were monitored
i
in some cases, makes intersite comparisons tenuous at best.
Ambient air concentrations of naturally emitted hydrocarbons (e.g.,
isoprene, a-pinene, p-pinene, A-carene, and limonene) are generally reported
only in nonurban hydrocarbon sampling programs. Because they are present at
very low concentrations, natural hydrocarbons are extremely difficult to
identify unequivocally when they mix with manmade emissions in an urban area.
The one exception is isoprene, which has been reported in both urban and
3-107
-------
nonurban sampling programs. Monoterpene (ClnH.,g) concentrations in ambient
air seldom exceed 20 ppb C. Average concentrations of orpinene, the most
commonly reported monoterpene, are usually below 10 ppb C. During the summer
months, isoprene concentrations as high as 150 ppb C have been measured (Ferman,
1981), but maximum concentrations in the 30 to 40 ppb C range are more common.
Ambient concentrations of the naturally emitted hydrocarbons are site-dependent,
with the highest concentrations observed in or immediately adjacent to forested
areas. Concentrations vary with season, as well, because natural hydrocarbon
emission fluxes are directly related to the amount of biomass present and
increase with temperature. In a recent review article, Altshuller (1983) has
provided a more detailed discussion of natural hydrocarbons and their effect
on air quality.
3.5.2.2 Concentrations of Nitrogen Oxides in Ambient Air. Ambient air levels
of nitrogen oxides have been monitored throughout the United States for a
number of years. Since nitrogen dioxide (NOp) is the only oxide of nitrogen
for which an NAAQS has been promulgated, it has received the greatest attention.
The emphasis here is on NO measurements that can be related to the diurnal
Jr\,
photochemical processes that produce ozone.
3.5.2.2.1 Urban NO,, concentrations. Concentrations of NO , like hydrocarbon
"" -•^===-:T-;I-- ^'J"——"--'- J---1:::::::::: ^^, ^
concentrations, tend to peak in urban areas during the early morning period
when atmospheric dispersion is limited and automobile traffic is dense. Most
of the NO is emitted as nitric oxide (NO), but the NO is converted rapidly to
/%
N0« by ozone and peroxy radicals produced in atmospheric photochemical reac-
tions. Since ozone levels and photochemical activity vary diurnally and from
day to day, the relative concentrations of NO and N0? can fluctuate signifi-
cantly. Generally, urban NO concentrations peak during the 6:00 to 9:00 a.m.
period, followed by a rapid decrease caused by the photochemical conversion of
NO and NOg and increased atmosphere mixing. Nitric oxide levels remain low
during the daytime and then usually build up again through the nighttime
hours. Nitrogen dioxide concentrations typically increase during the mid-
morning hours and then abate as the afternoon progresses. Levels of NOg
increase again following the late afternoon rush-hour period, often continuing
to increase during the nighttime.
The average NO concentration in urban areas of the United States is
s\ •
about 70 ppb, with NO and N02 contributing about equally (Logan, 1983).
Monitoring data for 1975 through 1980 showed that peak 1-hr NOp concentrations
3-108
-------
equalled or exceeded 400 ppb in Los Angeles "and several other California
locations, as well as at sites in Kentucky (Ashland) and Michigan (Port Huron).
Cities with one peak hourly concentration exceeding 270 ppb during those years
include Phoenix; St. Louis; New York City; Springfield (Illinois); Cincinnati;
Saginaw and Southfield (Michigan); and more than a dozen sites in California.
Reported hourly concentrations in excess of 140 ppb were quite common nationwide
during the years between 1975 and 1980 (U.S. Environmental Protection Agency,
1982a).
Urban NO concentrations during the 6:00 to 9:00 a.m. period are of
f\
primary importance in terms of oxidant production. Average NO levels recorded
f\
in several urban areas during this morning period are listed in Table 3-16,
which shows mean 6:00 to 9:00 a.m. NO concentrations in the range of about 50
^\
to 150 ppb. Concurrent 6:00 to 9:00 a.m. hydrocarbon samples were also
obtained in the studies reported in Table 3-16, and the hydrocarbon-NQ ratios
f\
in each of these urban areas are included.
TABLE 3-16, AVERAGE 6:00 to 9:00 a.m. NO CONCENTRATIONS AND
HC/NO RATIOS IN URBAN AREAS
City
Atlanta
Baltimore
Boston
Houston
Detroit
Linden, NJ
Los Angeles
Milwaukee
St. Louis
Tulsa
Washington, DC
Average NO ,
ppb x
57
85
63
125
67
59
147
66
77
46
94
Average
HC/NOV
f\
9
10
10
13
5
16
10
5
8
13
14
References
Westberg and Lamb (1983)
Richter (1983)
Ri enter (1983)
Westberg et al. (1978b)
Kelly et al. (1986)
Richter (1983)
U.S. Environmental
Protection Agency (1978a)
Westberg and Lamb (1983)
U. S. Environmental Protection
Agency (1978a)
Eaton et al. (1979)
Richter (1983)
3-109
-------
Hydrocarbon concentrations (ppb C) exceeded the NO levels by a factor of
*\
5 to 16 during the 6:00 to 9:00 a.m. period. Smog chamber experiments indicate
that significant quantities of ozone can be produced when HC/NO ratios are in
/\
this range. Indeed, ozone production has been observed in the vicinity of
most of the cities referenced in Table 3-16.
3.5.2.2.2 Nonurban NO concentrations. Concentrations of NO in "clean"
jrST" /\
remote environments are usually below 0.5 ppb (Logan, 1983). For example,
median concentrations measured on Niwot Ridge in Colorado are about 0.3 ppb in
the summer and 0.24 ppb in winter. In exceptionally clean air, NO concentra-
}\
tions as low as 0.015 have been recorded (Bellinger et al., 1982). Slightly
higher NO concentrations have been reported at other remote locations in the
s\
western United States and Canada. Kelly et al. (1982) deduced a mean NO
A
concentration of about 1 ppb from measurements in South Dakota. At the South
Dakota site, nitric oxide generally contributed less than 20 percent of the
total NO . Measurements of NO during the 1970s at rural locations in Montana
S\ f\
(Decker et al., 1978) and Saskatchewan (McElroy and Kerr, 1977) yielded average
concentrations similar to those recorded in South Dakota.
At a rural site in Louisiana, Kelly et al. (1984) found mean concentrations
of ~1 ppb NO and ~4 ppb NOg. The same investigators observed mean concentrations
of ~0.7 ppb NO and ~1.6 ppb N02 at a rural site in Virginia (Kelly et al.,
1984).
In the northeastern United States, nonurban NO concentrations appear to
exceed those in the west by about a factor of ten. A median NO concentration
jf\
of 6.6 ppb was derived from data collected at nine rural sites utilized in the
Sulfate Regional Experiment (SURE) program (Mueller and Hidy, 1983). Median
concentrations at the individual stations, which extended eastward from the
Ohio River Valley to the Atlantic Coast, varied from 2 to 11 ppb. Measurements
at nonurban sites in Pennsylvania and Louisiana, during the summer of 1975
showed mean hourly NO concentrations of 4.7 and 4.1 ppb, respectively (Decker
/\ " .
et al., 1978). Nitric oxide composed approximately 40 percent of the total
NO at these latter two nonurban sites.
.A
3.6 SOURCE-RECEPTOR (OXIDANT-PRECURSOR) MODELS
In order to apply knowledge of the atmospheric chemistry of ozone and
other photochemical oxidants and their precursors during their dispersion and
3-110
-------
transport, models describing these phenomena have been developed in a variety of
forms over the past 15 years. Most of these models relate the rates of precursor
emissions from mobile and stationary sources, or precursor atmospheric concen-
trations, to the resulting ambient concentrations of secondary pollutants that
impact receptors at downwind sites. For this reason they have been described
as source-receptor or oxidant-precursor models.
A wide variety of source-receptor models exists, ranging in complexity
from empirical relationships based on air monitoring or smog chamber data to
complex computer-based grid or trajectory airshed models that may contain
detailed emission inventories, sophisticated dispersion and transport sub-
models, and lengthy chemical reaction mechanisms. Moving-box models represent
an approach of intermediate complexity.
All presently available source-receptor models require a degree of simpli-
fying assumptions to deal with practical limitations imposed by existing
computer capabilities, time and cost constraints, or lack of knowledge concern-
ing inputs such as boundary conditions, emissions, wind fields, or detailed
reaction mechanisms. The reliability and applicability of any particular
model therefore depend upon its specific limitations, data requirements, and
degree of validation against experimental data from ambient air measurements
or environmental chamber runs.
A detailed discussion of the range of available source-receptor models,
and their validation and applications, is beyond the scope of this document.
Instead, brief conceptual descriptions are provided of the major classes of
source-receptor models. It is important to recognize that such models are
continually undergoing evolution, revision, and refinement, particularly as
knowledge of atmospheric chemistry grows and, for example, as more sophisti-
cated approaches to dealing with boundary conditions become available. Even
under the best circumstances, however, the present generation of source-
receptor models should be viewed as being most useful for investigating the
relative effects on air quality of particular emission sources or emission
control strategies, rather than for predicting absolute concentrations of
secondary pollutants resulting from specific precursor emissions.
3-111
-------
3.6.1 Definitions, Descriptions, and Use
Current air-quality, or source-receptor, models can be classified as
either statistical or computational/dynamic. Statistical models are generally
based on an analysis of historical air quality data. An example of a statistical
model is the linear rollback concept.
Computational, or dynamic, models attempt to describe mathematically the
atmospheric chemical and physical processes Influencing air pollution formation
and impacts. Examples of computational models include trajectory and grid
airshed models. The basis of these models is the solution of the atmospheric
diffusion equation (Bird et al., 1960; Liu and Seinfeld, 1975).
Two phenomenologically different approaches have been employed in dynamic
models with respect to the coordinate systems chosen. A coordinate system
fixed with respect to the earth is termed Eulerian, while in Lagrangian models
the reference frame moves with the air parcel whose behavior is being simulated.
In the following section these and other models are briefly described.
3.6.1.1 Statistical Models. Two widely used models of this kind are linear
rollback and the Appendix J approach, both of which were employed by EPA prior
to the advent of more sophisticated dynamic modeling approaches. The concept
of linear rollback is based on the assumption that ambient concentrations of
air pollutants are directly proportional to emissions; and that a given reduc-
tion in emissions will result in a proportional decrease in the maximum ambient
concentrations of that pollutant. In principle, linear rollback models should
be applied only to inert primary pollutants and their original use was for
unreactive pollutants such as carbon monoxide (Larsen, 1969). Such models
have been applied, however, in modified form to secondary pollutants such as
ozone (Barth, 1970).
A prominent example of a statistical model was the Appendix J relationship
developed by EPA to relate maximum 1-hour average ozone concentrations in
several United States cities to 6:00-to-9:00 a.m. average nonmethane hydrocarbon
concentrations in those cities (F.R., 1971). This relationship was used to
calculate the amount of NMHC control needed to achieve the Federal standard
for photochemical oxidants.
Two important limitations of past statistical methods were their failure
to take into account the transport of primary and secondary pollutants from
source areas to downwind receptor sites, and lack of recognition of the role
of oxides of nitrogen in the formation of ozone and other photochemical oxidants.
3-112
-------
These and other weaknesses led EPA to abandon the use of statistical models in
state implementation plans (F.R., 1981).
3.6.1.2 Trajecto ry Mode1s. Figure 3~17(a) contains a schematic representation
of the trajectory model approach, in which a hypothetical air parcel moves
through the area of interest along a path calculated from wind trajectories.
Thus, a moving-coordinate system (Lagrangian) describes pollutant transport
under the influence of local meteorological conditions. Emissions are injected
into the air parcel and undergo vertical mixing and chemical transformations.
The data requirements for trajectory models include: (1) initial concen-
trations of all relevant pollutants arid species; (2) rates of emissions of
NMOC and NO precursors into the parcel along its trajectory; (3) meteorological
s\
characteristics such as wind speed and direction; and (4) solar ultraviolet
radiation. Various trajectory models exhibit a range of sophistication and
complexity with regard to such elements as chemical mechanisms (Atkinson
et al., 1982c), emission inventories (Braverman and Layland, 1982), treatment
of vertical mixing (Whitten and Hogo, 1978; Drivas, 1977; Meyers et al., 1979;
Lloyd et al., 1979; Lurmann et al., 1979), and trajectory determination (U.S.
Environmental Protection Agency, 1980c; Whitten and Hogo, 1978). Basic limi-
tations of trajectory models include the amount and density of data required
for precise calculations of emissions input, chemical transformations, and
dilution; neglect of horizontal wind shear; neglect of cell volume changes
resulting from convergence and divergence of the wind field; and uncertainties
in boundary conditions, including conditions aloft (Liu and Seinfeld, 1975).
Conversely, moving-cell models provide a dynamic description of atmospheric
source-receptor relationships that is simpler and less expensive to derive
than that obtained from fixed-cell models.
The simplest form of trajectory model is the empirical kinetic modeling
approach (EKMA). This modeling approach was developed from earlier efforts
(Dimitriades, 1972) to use smog chamber data to develop graphical relationships
between morning NMOC and NO levels and afternoon ozone maximum concentrations.
J\
Dodge (1977a,b) presented an approach in which smog chamber data (Dimitriades,
1970; 1972) were used to test and validate a photochemical kinetics model
(Durbin et al., 1975). Dimitriades (1977) used the resulting 03 isopleths to
define a method for obtaining the relative degree of precursor emissions
control needed to achieve a given percentage reduction in ozone. In the EKMA
approach, which has been extensively utilized (U.S. Environmental Protection
3-113
-------
SUNLIGHT GIVEN AS
FUNCTION OF TIME
TIME-DEPENDENT MIXING AND
REACTION COMPUTED FOR AIR
PARCEL UP TO MIXING HEIGHT, h
SPACE/TIME TRACK THROUGH SOURCE
GRID; DERIVED FROM WIND DATA
POLLUTANT INFLUXES AT ANY
ELEVATION (INCLUDING THE GROUND)
IMPOSED BY EMISSION SOURCE FUNCTIONS
(a) TRAJECTORY MODEL
TRANSPORT
TOP OF MODE LING
1 , REGION
«• 1—»• CHEMISTRY-ELEVATED EMISSIONS t—| fc
. TRANSPORT * TRANSPORT TOP °£
•^^^^ » TRANSPORT I LAYER
CHEMISTRY-ELEVATED EMISSIONS "—1 "
5 TRANSPORT 1 TRANSPORT
• TRANSPORT 1
3 * 1"~* CHEMISTRY-ELEVATED EMISSIONS ••—I ••
S TRANSPORT »_ TRANSPORT
j TRANSpoRT I
CHEMISTHY-ELEVATEO EMISSIONS ••—I ••
TRANSPORT , TRANSPORT
t
SURFACE T REMOVAL
1 TO 10 km
GROUND SURFACE
GRID SPECIFICATION
GRID CHEMISTRY AND TRANSPORT
(b) GRID MODEL
RISING MIXED
HEIGHT ENTRAINMENTOF
POLLUTANTS ALOFT
WIND DIRECTION
ADVECTIVE
INFLOW
ADVECTIVE
OUTFLOW
(c) BOX MODEL
Figure 3-17. Schematics of the three types of dynamic models.
3-114
-------
Agency, 1978a,b; 1980c), the Ozone Isopleth Plotting Package (OZIPP) (Whitten
and Hogo, 1978) is used to generate ozone isopleths at various levels of
sophistication corresponding to "standard" EKMA, "city-specific" EKMA, or the
"simplified trajectory" model (F.R., 1979). These models are designated as
Levels IV, III, and II, respectively, with Level IV being the least sophis-
ticated and Level II the most sophisticated. Substantial documentation and
guidance concerning the use of OZIPP and a more flexible modified version of
the Program (OZIPM) are available (U.S. Environmental Protection Agency,
1981a, 1984b,c).
The sensitivity of this method to variables such as the input hydrocarbon
composition and the choice of chemical kinetics mechanisms has been reported
(Carter et a!., 1982; Jeffries et a!., 1981; Shafer and Seinfeld, 1985) and
further refinements in the EKMA approach to accomodate these factors have been
made. For example, site-specific versions of EKMA allow the user to select
particular dilution rates, emissions, and solar intensity applicable to the
city or airshed of interest. Another version allows the user to employ an
alternative mechanism for making the EKMA calculations (U.S. Environmental
Protection Agency, 1984b).
Because it is often recommended by EPA for use in determining needed
precursor reductions and is in widespread use, EKMA is discussed here in
detail. An example of an EKMA diagram is presented in Figure 3-18, which
shows ozone isopleths for sites downwind of an urban source area in which
morning precursor emissions are high. The isopleths in this diagram depict
downwind, peak 1-hour ozone concentrations as an explicit function of initial
(i.e., morning) concentrations of nonmethane hydrocarbons (NMHC) and nitrogen
oxides (NO ); and as indirect functions of (1) NMHC and NO emissions occurring
A P\
later in the day; (2) specified meteorological conditions; (3) reactivity of
the precursor mix; and (4) concentrations of ozone and precursors transported
from upwind areas (U.S. Environmental Protection Agency, 1977). The relation-
ships between ozone and its precursors that are depicted in Figure 3-18 are
based on empirical data and the application of a chemical kinetics model
(Dodge, 1977a; Whitten and Hogo, 1977) that has been adjusted by comparing
model predictions against smog-chamber data obtained by irradiating automobile
exhaust (Dimitriades, 1972). Alternatively, EKMA diagrams can be constructed
using more recent chemical mechanisms that have been tested against smog
chamber data (Gipson, 1984; Whitten and Gery, 1986).
3-115
-------
0.28
0.2
0.4 0.6 0.8 1.0 1.2 1.4 1.6
NONMETHANE HYDROCARBON CONCENTRATION, ppm
1.8
0.28
0,24
0.20
- 0.16
- 0.12
- 0.08
0.04
2.0
Figure 3-18 Example of EKMA diagram for high-oxidant urban area.
3-116
-------
For the EKMA diagram given here as Figure 3-18, several general, inter-
related features are of interest. First, the isopleth lines in the lower
right quadrant are more or less parallel to the abscissa (the NMHC concen-
tration). Second, the isopleth lines in the upper left quadrant are slanted
with respect to the ordinate (the NO concentration). Because of the shape of
J\
the isopleth lines, varying the NMOC or NO concentration will have different
X
effects on ozone at different NMHC/NQ ratios.
yv • . , .
These two features of the diagram are related to the underlying photo-
chemistry (see, e.g., Whitten, 1983). For extremely low NO concentrations
A
(i.e., high NMHC/NO ratios), where the lines parallel the abscissa, the
J\.
formation of 0, is (1) insensitive to changes in NMHC concentrations, and (2)
NO -limited; that is, changes in NO concentrations cause co-directional
X X
changes in peak Og concentrations. As described early in Section 3.3 and in
Section 3.3.1, the atmospheric oxidation of hydrocarbons produces an abundance
of peroxy radicals (RQp*), more than enough to oxidize NO to NQp rapidly !and
completely. In addition, the NO present, before being removed from the
A.
cyclic reactions via termination reactions with various radicals, completes a
RO
number of NO ^—» NQp — »» NO (+ GO) cycles, thus producing a number
£m . O
of Oq molecules. Small-to-moderate changes in NMHC concentrations will there-
fore have little impact, since there will still be an abundance of R02* radicals
in the atmosphere. On the other hand, changing the already low concentration
of NOX does not have an appreciable impact on the ROp radicals, but it dres
change co-directionally the number of 0, molecules produced, since the photolysis
•3 , • . ~!
of NOp and the oxidation of NO to NO are essential for 0^ production.
For moderately higher concentrations of NO (lower NMHC/NO ratios),
X X
there is no longer a large abundance of ROp- relative to the NO present; and
changes in NMHC, therefore, have a co-directional effect on RQ2* and, hence,on
Og. Thus, for moderate NMHC/NO ratios, the effects on 03 formation of varying
the concentrations of NMHC or NO are similar in direction.
X
Finally, for much higher NOV concentrations (i.e., very low NMHC/NO
X . X
ratios), the dominant effects are the relative depletion of radicals, through
the reactions of N02 with these radicals (Equations 3-18 and 3-35, Section
3.3); with a resulting decrease in the rate of reaction of peroxy radicals
with NO (Equation 3-29, Section 3.3). The consequence of these effects is
that the conversion of NO to NOp is very slow. At such low NMHC/NOX ratios,
therefore, increasing the NMHC concentration enhances Qg formation but
increasing NO concentrations inhibits it.
3-117
-------
As examination of Figure 3-18 reveals, for an NMHC concentration of 0.6
ppm C, for example, increasing NO leads to increased 0~ until NMHC/NO ratios
X «3 X
of about 5:1 to 6:1 are reached; further NO increases, leading to lower
X
NMHC/NO ratios, inhibit 0- formation. Thus, in this example, there is a
X O
"critical" ratio (approximately 5,6:1) at which the NO effect on 0« changes
X * O
direction. Besides this "critical" ratio, an "equal control" NMHC/NO ratio
X
also exists, above which the reduction of NO is more beneficial in terms of
X
0,, reduction than an equal percentage reduction in NMHC. This ratio, for the
isopleths shown in Figure 3-18, is roughly 8:1 to 9:1 for low levels of control,
and as high as 20:1 for the levels of control needed to reduce 03 to 0.12 ppm.
Thus, for this particular case (Figure 3-18), the chemical mechanism modeling
evidence suggests that (1) NO control will increase the peak downwind 0.,
X O
concentration at NMHC/NO ratios of 5.6:1 or lower; (2) both NO control and
X X
NMHC control will be beneficial at somewhat higher ratios, with control of
NMHC being more effective; and (3) for ratios above 20:1, NO control is
X
relatively more effective in reducing 03<
The calculation of precursor controls necessary to reduce 03 to 0,12 ppm,
from the isopleths given in Figure 3-18, shows that NO control, although at
X
first beneficial, is ultimately detrimental because it makes the reduction of
03 to 0.12 ppm more difficult; This can be demonstrated through use of the
EKMA isopleths in Figure 3-18 as follows. For a high-oxidant atmosphere with,
for example, a peak 0,, concentration of 0.30 ppm and an NMHC/NO ratio of
O X
12:1, a 74 percent control of NMHC (1.84 ppm C to 0.48 ppm C) will be needed
in order to reduce downwind, 1-hour peak 03 to 0.12 ppm through the unilateral
control of NMHC (line AB in Figure 3-18). If, however, NO is first controlled
• /\ *•"
by, for example, 29 percent (from 0.152 ppm to 0.108 ppm, line AA1), this will
cause a 15 percent reduction in 03 (line AA1) but it will also increase the
NMHC control requirement (to reduce 03 to 0.12 ppm) from 74 percent (line AB)
to 81 percent (line A'B1). Since it is not ordinarily feasible to reduce 0,
to 0.12 ppm in a high-oxidant area through the unilateral control of NO (it
X
would take, in this case, an almost 85 percent control of NO , line AC), it
X
follows that 29 percent control (or any control) of NO , would ultimately be
X
detrimental for the situation represented by these isopleths. There may be
situations, however, in which the control of NO will not increase the NMHC
J\
control required to achieve a given percentage reduction in ozone. It must
be emphasized that the preceding discussion of the implications of controlling
VOC and NO is not applicable to all situations. These diagrams may differ
»
3-118
-------
when different chemical mechanisms or other model input data are used. Further-
more, conclusions drawn from Figure 3-18 may differ if a different starting
point on the diagram is used.
As noted early in Section 3.3, the effects on 0- formation of controlling
NO emissions are a matter of continuing discussion and research. The inhibition
-------
models (Roth et al., 1976). In addition to atmospheric data, various emissions
sources must be considered in each model. At the most sophisticated level,
emission models and inventories are constructed to provide estimates, by
category, of vehicular, aircraft, power plant, refinery, and distributed
source emission rates, including their temporal and seasonal variations (Roth
et al., 1976; Braverman and Layland, 1982; F.R., 1979).
Shortcomings in grid models stem from the theoretical and computational
complexities that are necessary in this type of simulation. Inaccuracies in
grid-model predictions arise from: (1) theoretical deficiencies in the mathe-
matical representation of atmospheric processes; (2) numerical inaccuracy in
the solution of the atmospheric diffusion equation (Liu and Seinfeld, 1975);
and (3) inadequate input data resulting from incomplete data bases. An
incomplete understanding of advection and turbulent diffusion, necessitates
the use of estimates or parameter!'zations to provide appropriate values
(Seinfeld and Wilson, 1977). Atmospheric chemical kinetics descriptions are
continually updated as new information is obtained, but uncertainties associ-
ated with these mechanisms may be propagated during solution. In addition,
sparse and often unrepresentative data are utilized to derive continuous
fields (wind fields, turbulence, and mixing depths) over the region (Seinfeld
and Wilson, 1977), a problem that is common to all dynamic models. In general,
wind and turbulence data are rarely collected aloft; surface data are much
more abundant but still vary widely in terms of number, frequency, and quality
of measurements (Roth et al., 1976). This implies that critical values,
especially aloft, must often be estimated to provide initial, boundary, and
operating conditions (Seinfeld and Wilson, 1977). Finally, uncertainty in
grid-model solutions also arises from emission inventories that are poorly
resolved, either spatially, temporally, or with respect to hydrocarbon
reactivity specifications (Braverman and Layland, 1982).
The structure and complexity of grid models also account for their utility.
The increased temporal resolution afforded by grid models can provide minute-
by-minute concentration estimates in each cell, and there can be as much
spatial resolution as the data will allow. Since it accounts for specific
atmospheric processes within the system, the model allows explicit insertion
of new information (e.g., on meteorological or chemical processes) into the
structure of the model. In addition, the impact of individual precursor
sources may be analyzed with this type of model (Association of Bay Area
Governments, 1979a,b).
3-120
-------
3.6.1.4 Box Models. Box models (Hanna, 1973; Demerjian and Schere, 1979;
Derwent and HOv, 1980) are the simplest of dynamic models. They treat the
atmosphere as a single cell, bounded by the mixing layer, with an area on the
order of 100 square miles [see Figure.3-15(c)]. The chemistry within the box
is affected by: (1) instantaneously mixed regional emissions, (2), dilution
from lifting of the inversion, (3) ventilation and transport resulting from a
characteristic wind field, and (4) entrainment of species from aloft. Because
the only consideration of spatial resolution occurs when the modeling boundaries
are chosen, data requirements are minimal. Results can only be interpreted
temporally, however, for a mass average of a species, and results can be
strongly affected by uncertainties in boundary conditions.
3.6.2 Validation andSensitivity Analyses for Dynamic Models
Dynamic models are mathematical representations of atmospheric processes.
They are based on many assumptions, however, and can only be considered approxi-
mations of real processes. Therefore, it is important to .investigate the
extent to which model predictions disagree with actual measurements, Deviations
occur for two basic reasons: (1) a completely valid mathematical description
of natural systems does not presently exist; and (2) input data and data for
comparison with predictions are often unresolved and imprecise (Seinfeld and
Wilson, 1977). It is therefore difficult to determine numerically the overall
accuracy of model calculations. Rather, attempts are made to validate model
predictions by comparing them with real observations and operating parameters
are often varied to determine the sensitivity of the model (Gelinas and Vajk,
1979). In addition, the extent of agreement between the results from two
simulations can be tested. "In this way, completely different models may be
compared, or an internal component, such as the chemical kinetics mechanism,
may be substituted and a model re-run to .ascertain the effects of such substi-
tutions. .
It should be noted that the validity of all dynamic .models depends, in
part, on the quality of the chemical kinetics mechanisms used to define the
Og-HC-NO relationship. These mechanisms have the advantage of being cause-
and-effect descriptions derived from actual experimental data. The data are
subject, however, to the effects of smog chamber artifacts (Carter et al.,
1982), which may or may not occur in the atmosphere. Also, there remain sub-
stantial uncertainties in the detailed chemistry of certain classes of organics
3-121
-------
such as the aromatics (Section 3.3), In addition, only recently have data
become available so that mechanisms can be tested against data bases in which
the hydrocarbon composition has been systematically varied or in which dynamic
dilution and injection of new reactants has occurred (Jeffries et a!,, 1981).
Therefore, dynamic models using existing chemical mechanisms may not accurately
describe all of the conditions that apply in the atmosphere (Jeffries et al.,
1981).
Evaluations of complex dynamic models have been of two forms: numerical
sensitivity analyses and simulation performance studies using ambient air
data. Sensitivity analyses have considered the effects of varying meteorolog-
ical factors, initial and boundary air quality data and emissions inputs,
model structure and computational factors, and reactions within the chemical
kinetics mechanism (U.S. Environmental Protection Agency, 1981g; Liu and
Seinfeld, 1975; Gelinas and Vajk, 1979; Til den and Seinfeld, 1982; Dunker,
1980, 1981; Falls et a!., 1979).
A comparative study of the Photochemical Box Model (PBM) of Demerjian and
Schere (1979), the Lagrangian Photochemical Model (LPM) developed by Lurmann
et al. (1979), and the Urban Airshed Model (UAM) (Ames et a!., 1978) was
performed in which the models were compared in "off-the-shelf" use (Schere and
Shreffler, 1982a,b). That is, no effort was made to adjust the model predic-
tions, although great care was taken in preparation of data and in model
execution. Based upon application of these models to St. Louis air quality
data, a fourth model, LIRAQ, was shown to be unsatisfactory and was not eval-
uated further (Schere and Shreffler, 1982a,b).
The remaining three models yielded adequate source-receptor information,
provided knowledgeable interpretation of the output was applied. For instance,
the PBM averaged 23 percent overprediction of ozone concentration over all
test days, but in the 5 stagnation days for which the maximum ozone observed
occurred within the PBM domain the average overprediction was only 8 percent.
The LPM showed the largest variance in the ozone concentration residuals,
possibly because the input data were not precise enough to fulfill the temporal
and spatial demands of the model. As in previous studies (Whitten and Hogo,
1981; Reynolds et al., 1982; Cole et al., 1982), the UAM predicted ozone
maximum concentrations with little bias (about 4 percent overprediction), but
had difficulty placing the "ozone cloud" at the correct time and place.
Again, this suggests uncertainty in specifying the wind field data. The user
3-122
-------
of any of these models must have a strong scientific background and must
exercise extreme care in implementing air quality simulations (Schere and
Shreffler, 1982a,b).
Because of their wide use, Levels II, III, and IV of EKMA have been the
subject of many evaluation efforts. These include parameter sensitivity
studies, comparison between the various levels of the approach, and studies
comparing EKMA with other dynamic models (Meyer et al., 1981).
Jeffries and coworkers (1981) evaluated the performance of EKMA Levels II
and III using 10 days of data from the 1976 St. Louis Regional Air Pollution
Study. To evaluate the effects of chemistry and meteorology inputs, four
different chemical kinetics mechanisms and three methods of calculating mixing-
height profiles were employed. The choice of mechanism, trajectory, and
mixing-height profile proved to have a large effect on the prediction of
absolute ozone levels. No one mechanism or mixing-height profile was superior,
however, at producing the "best fit" over all days. When the EKMA procedure
was used in a relative sense to estimate needed reductions in NMHC, results
obtained were not consistent. Improvements needed in the simple trajectory
model (Level III) were identified as: (1) improved chemical kinetics descrip-
tions, (2) smoother and more defined trajectories, (3) better treatment of
point sources, and (4) improved mixing-height profiles.
In another sensitivity study, standard and city-specific versions of
EKMA were used to simulate 100 pre-selected test days (Maxwell and Martinez,
1982). A statistical analysis was performed to determine how accurately these
models, using three different chemical mechanisms, could predict absolute
ozone levels. As above (Jeffries et al., 1981), none of the models was a
consistently good predictor of ozone and ozone levels were usually overpredicted
by more than 20 percent.
Finally, all three levels of EKMA were compared on a limited number of
test days with respect to: (1) level of EKMA, (2) chemical kinetics mechanisms,
and (3) isopleth diagram entry parameters (Hayes and Hogo, 1982). Again it
was found that substituting chemical mechanisms produced significant differences
in the shape of the isopleth curves. Also, and partly because of, this, EKMA
predictions were found to be quite sensitive to low NMHC/NO precursor ratios.
Jf\.
Precursor-ozone relationships derived for various levels of EKMA were not
particularly different, nor did they appear to be sensitive to the choice of
trajectory (for Level II analysis).
3-123
-------
Studies comparing EKMA performance with that of more complex dynamic
models have also been carried out (Whitten and Hogo, 1981; U.S. Environmental
Protection Agency, 1981a), Input to the EKMA model was often generated by the
comparison model so that specific features could be compared. In all studies,
the most sensitive difference, in terms of absolute prediction of ozone levels,
was found to be the choice of chemical mechanism. It was shown, however, that
application of the EKMA produced precursor-ozone descriptions similar to those
from the more complex models when the same chemical mechanism was used in each
of the models. Finally, it should be noted that the non-linear relationships
between 03 and its precursors means that good model performance in replicating
base case conditions does not ensure accurate emission control calculations.
To address this problem, emission control estimates obtained with EKMA have
been compared with trends and emission reduction estimates obtained with grid
models (Meyer et a!., 1981). Further information on the comparative performance
of EKMA is found in DimitHades and Dodge (1983).
Selection of a modeling approach for determining ozone concentrations
that is acceptable and appropriate for given circumstances necessitates making
many interrelated decisions. All models considered should be able, of course,
to simulate the physical and chemical processes known or suspected to be
important. Potential users must then weigh the advantages of greater credibi-
lity and capability against the disadvantages of greater cost, time, and
personnel requirements (Association of Bay Area Governments, 1979a). In
addition to the technical aspects of potential modeling approaches, specific
selection constraints also include: (1) extent of data requirements; (2) costs
of data collection, model implementation, and operation; (3) computer con-
straints; (4) personnel requirements; and (5) schedule constraints. These
criteria are not independent of one another, and time spent defining a selection
plan can result in substantial benefits throughout the modeling exercise.
3.7
3.7.1 Descriptions and Properties of Ozone and Other Photochemical Oxidants
Ozone (Q~) and other photochemical oxidants occurring at low concentra-
tions in ambient air, such as peroxyacetyl nitrate (PAN), hydrogen peroxide
(H202)5 and formic acid (HCOOH), are characterized chiefly by their ability to
remove electrons from, or to share electrons with, other molecules or ions (i.e.,
3-124
-------
oxidation). The capability of a chemical species for oxidizing or reducing
other chemical species is termed "redox potential" (positive or negative stan-
dard potential) and is expressed in volts. A reactive allotrope of oxygen
that is only about one-tenth as soluble as oxygen in water, ozone has a stan-
dard potential of +2.07 volts in aqueous systems for the redox pair, Q-XH-O
(Weast, 1977). Hydrogen peroxide, which is highly soluble in water and other
polar solvents, has a standard potential of +1.776 in the redox pair, 1-LOp/hLO
(Weast, 1977). No standard potential for peroxyacetyl nitrate in neutral or
buffered aqueous systems, such as those that occur in biological systems,
appears in the literature. In acidic solution (pH 5 to 6), PAN hydrolyzes
fairly rapidly (Lee et al.s 1983; Holdren et a!., 1984); in alkaline solution
it decomposes with the production of nitrite ion and molecular oxygen (Stephens,
1967; Nicksic et al., 1967). An important property of PAN, especially in the
laboratory, is its thermal instability. Its explosiveness dictates its synthesis
for experimental and calibration purposes by experienced personnel only.
Formic acid is formed as a stable product in photochemical air pollution.
It has the structure of both an acid and an aldehyde and in concentrated form
is a pungent-smelling, highly corrosive liquid.
The toxic effects of oxidants are attributable to their oxidizing abili-
ty. Their oxidizing properties also form the basis of several measurement
techniques for Q-, and PAN. The calibration of ozone and PAN measurements,
however, is achieved via their spectra in the ultraviolet and infrared regions,
respectively. All three pollutants of most concern in this document (0~, PAN,
and H-O-) must be generated i_n situ for the calibration of measurement tech-
niques. For ozone and H~Q?, generation of calibration gases is reasonably
straightforward.
3.7.2 Nature of Precursors to Ozone and Other Photochemical Oxidants
Photochemical oxidants are products of atmospheric reactions involving
volatile organic compounds (VOC) and oxides of nitrogen (NO ), as well as
s\
hydroxyl (OH) and other radicals, oxygen, and sunlight (see, e.g., Demerjian
et al., 1974; National Research Council, 1977a; U.S. Environmental Protection
Agency, 1978a; Atkinson, 1985). The oxidants are largely secondary pollutants
formed in the atmosphere from their precursors by processes that are a complex,
nonlinear function of precursor emissions and meteorological factors.
3-125
-------
The properties of organic compounds that are most relevant to their role
as precursors to ozone and other oxidants are their volatility, which governs
their emissions into the atmosphere; and their chemical reactivity, which
determines their lifetime in the atmosphere. Although vapor-phase hydrocar-
bons (compounds of carbon and hydrogen only) are the predominant organic
compounds in ambient air that serve as precursors to photochemical oxidants,
other volatile organic compounds are also photochemically reactive in those
atmospheric processes that give rise to oxidants. In particular, halogenated
organics (e.g., haloalkenes) that participate in photochemical reactions are
present in ambient air, although at lower concentrations than the hydrocarbons.
They are oxidized through the same initial step involved in the oxidation of
the hydrocarbons; that is, attack by hydroxyl radicals. Alkenes, haloalkenes,
and aliphatic aldehydes are, as classes, among the most reactive organic
compounds found in ambient air (e.g., Altshuller and Bufalini, 1971; Darnall
et al., 1976; Pitts et al., 1977; U.S. Environmental Protection Agency, 1978a,
and references therein). Alkenes and haloalkenes are unique among VOC in
ambient air in that they are susceptible both to attack by OH radicals (OH)
and by ozone (Niki et al., 1983). Methane, halomethanes, and certain haloe-
thenes are of negligible reactivity in ambient air and have been classed as
unreactive by the U.S. Environmental Protection Agency (1980a,b). Since methane
is considered only negligibly reactive in ambient air, the volatile organic
compounds of importance as oxidant precursors are usually referred to as
nonmethane hydrocarbons (NMHC) or, more properly, as nonmethane organic
compounds (NMOC).
The oxides of nitrogen that are important as precursors to ozone and
other photochemical oxidants are nitrogen dioxide (N0?) and nitric oxide (NO).
Nitrogen dioxide is itself an oxidant that produces deleterious effects, which
are the subject of a separate criteria document (U.S. Environmental Protection
Agency, 1982a). Nitrogen dioxide is an important precursor (1) because its
photolysis in ambient air leads to the formation of oxygen atoms that combine
with molecular oxygen to form ozone; and (2) because it reacts with acetyl-
peroxy radicals to form peroxyacetyl nitrate, a phytotoxicant and a lachryma-
tor. Although ubiquitous, nitrous oxide (N,,0) is unimportant in the production
of oxidants in ambient air because it is virtually inert in the troposphere.
3-126
-------
3.7.3 ' Atmospheric Reactions of Ozone and Other Oxidants Including Their Role
in Aerosol Formation
The chemistry of the polluted atmosphere is exceedingly complex, but an
understanding of the basic phenomena is not difficult to acquire. Three
processes occur: the emission of precursors to ozone from predominantly
manmade sources; photochemical reactions that take .place during the disper-
sion and transport of these precursors; and scavenging processes that reduce
the concentrations of both 03 and precursors along the trajectory.
The specific atmospheric reactions of ozone and of other photochemical
oxidants such as peroxyacetyl nitrate and hydrogen peroxide are becoming
increasingly well-character!zed. The reactions of these species result in
products and processes that may have significant environmental and health- and
weIfare-related implications, including effects on biological systems, nonbio-
logical materials, and such phenomena as visibility degradation and acidifica-
tion of cloud and rain water.
3.7.3.1 Formation and Transformationof Ozone and Other Photochemical Oxi-
dants. In the troposphere, ozone is formed through the dissociation of N0? by
sunlight to yield an oxygen atom, which then reacts with molecular oxygen (0_)
to produce an 0- molecule. If it is present, NO can react rapidly with 0, to
form NOp and an 0^ molecule. In the absence of competing reactions, a steady-
state or equilibrium concentration of 0- is soon established between GO, NOp,
and NO (National Research Council, 1977a). The injection of organic compounds
into the atmosphere upsets the equilibrium and allows the ozone to accumulate
at higher than steady-state concentrations. The length of the induction
period before the accumulation of 03 begins depends heavily on the initial
NO/NO,, and NMOC/NO ratios (National Research Council, 1977a).
i— /\
The major role played by organic compounds in smog reactions is attribut-
able to the hydroxyl radical (OH), since it reacts with essentially all organic
compounds (e.g., Atkinson, 1985; Herron and Huie, 1977, 1978; Dodge and Arnts,
1979; Niki et a!., 1981). Aldehydes, which are constituents of automobile
exhaust as well as decomposition products of most atmospheric photochemical
reactions involving hydrocarbons, and nitrous acid (MONO), are important
sources of OH radicals, as is 0~ itself. Other free radicals, such as hydro-
and alkylperoxy radicals and the nitrate (NO,) radical play important roles in
photochemical air pollution.
3-127
-------
The presence of organic compounds, oxides of nitrogen, and sunlight does
not mean that the photochemical reactions will continue indefinitely. Termi-
nation reactions gradually remove NCL from the reaction-mixtures, such that
the photochemical cycles slowly come to an end unless fresh NO and N0? emis-
sions are injected into the atmosphere. Compounds containing nitrogen, such
as PAN, nitric acid (HNO-), and peroxynitric acid (HNO»), as well as organic
and inorganic nitrates, are formed in these termination reactions.
Recent studies on the photooxidation of organic compounds under simulated
atmospheric conditions have been reasonably successful. The rate constants
for the reaction of OH radicals with a large number of organic compounds have
been measured (e.g., Atkinson et a!., 1979; Atkinson et al., 1985). The
mechanisms of the reactions of paraffinic compounds are fairly well under-
stood, as are those of olefinic compounds, at least for the smaller compounds.
Photooxidation reactions of the aromatic compounds, however, are poorly under-
stood.
In the presence of NO , natural hydrocarbons (i.e., those organic com-
s\
pounds emitted from vegetation) can also undergo photooxidation reactions to
yield 0,, although most naturally emitted hydrocarbons are olefins and are
scavengers as well as producers of 0- (e.g., Lloyd et a!., 1983; Atkinson
( . w
et al.» 1979; Kamens et al., 1982; Killus and Whitten, 1984; Atkinson and
Carter, 1984).
3.7.3.2 Atmospheric Chemical Processes InvolvingOzone. Ozone can react with
organic compounds in the boundary layer of the troposphere (Atkinson and
Carter, 1984). It is important to recognize, however, that organics undergo
competing reactions with OH radicals in the daytime (Atkinson et al., 1979;
Atkinson, 1985) and, in certain cases, with N03 radicals during the night
(Japar and Niki, 1975; Carter et al., 1981a; Atkinson et al., 1984a,b,c,d,e;
Winer et al., 1984), as well as photolysis, in the case of aldehydes and other
oxygenated organics. Only for organics whose ozone reaction rate constants
-21 3 ~1 -1
are greater than ~10 cm molecule sec can consumption by ozone be
considered to be atmospherically important (Atkinson and Carter, 1984).
Ozone reacts rapidly with the acyclic mono-, di-, and tri-alkenes and
— 1 O
with cyclic alkenes. The rate constants for these reactions range from ~10
to ~10 cm molecule" sec" (Atkinson and Carter, 1984), corresponding to
atmospheric lifetimes ranging from a few minutes for the more reactive cyclic
alkenes, such as the monoterpenes, to several days. In polluted atmospheres,
3-128
-------
a significant portion of the consumption Of the more reactive alkenes will
occur via reaction with ozone rather than with OH radicals, especially in the
afternoons during photochemical oxidant episodes. Reactions between ozone and
alkenes can result in aerosol formation (National Research Council, 1977a;
Schuetzle and Rasmussen, 1978), with alkenes of higher carbon numbers the
chief contributors.
Because of their respective rate constants, neither alkanes (Atkinson and
Carter, 1984) nor alkynes (Atkinson and Aschmann, 1984) are expected to react
with ozone in the atmosphere, since competing reactions with OH radicals have
higher rate constants (Atkinson et a!., 1979'; Atkinson, 1985).
The aromatics react with ozone, but quite slowly (Atkinson and Carter,
1984), such that their reactions with ozone are expected to be unimportant in
the atmosphere. Cresols are more reactive toward ozone than the aromatic
hydrocarbons (Atkinson and Carter, 1984), but their reactions with OH radicals
(Atkinson, 1985) or NQ3 radicals (Carter et a!., 1981a; Atkinson et a!.,
1984d) predominate.
For oxygen-containing organic compounds, especially those without carbon-
carbon double bonds, reactions with ozone are slow. For carbonyls and ethers
(other than ketene) that contain unsaturated carbon-carbon bonds, however,
much faster reactions are observed (Atkinson and Carter, 1984).
Certain reactions of ozone other than its reactions with organic com-
pounds are important in the atmosphere. Ozone reacts rapidly with NO to form
N02, and subsequently with NO,, to produce the nitrate (N03) radical and an
oxygen molecule. Photolysis of ozone can be a significant pathway for the
formation of OH radicals, particularly in polluted atmospheres when ozone
concentrations are at their peak.
Ozone may play a role in the oxidation of S02 to H^SO,, both indirectly
in the gas phase (via formation of OH radicals and Criegee biradicals) and
directly in aqueous droplets. -
3.7.3.3 Atmospheric Reactions of PAN, H00^,' and HCOQH. Because PAN is in
ULU"'" ^ \ " """'" """"" ~""r""vrr L ' ™«™ '«>« •'•"•"•" £"-"jT -"••- iuu.uuui-.ui
equilibrium with acetyl peroxy radicals and NOg, any process that leads to the
removal of either of these species will lead to the decomposition of PAN. One
such process is the reaction of NO with acetyl peroxy radicals. This can
lead, however, to the formation of OH radicals. Thus, PAN remaining overnight
from an episode on the previous day can react with NO emitted from morning
traffic to produce OH radicals (Cox and Roffey, 1977; Carter et al., 1981c)
3-129
-------
that will enhance smog formation on that day (e.g., Tuazon et al,, 1981a). In
the absence of significant NO concentrations, and in regions of moderate to
lower temperatures, PAN will persist in the atmosphere (Wellington et a!.,
1984; Aikin et a!., 1983) and contribute to the long-range transport of NO .
r\
Although hydrogen peroxide formed in the gas phase from the reactions of
hydroperoxyl radicals plays a role in HO chemistry in the troposphere, and
j\
especially in the stratosphere (Crutzen and Fishman, 1977; Cox and Burrows,
1979), its major importance arises from its high solubility in water. The
latter ensures that a large fraction of gaseous H-Op will be taken up in
aqueous droplets. Over the past decade, evidence has accumulated that tiJ^o
dissolved in cloud, fog, and rainwater may play an important, and, in acidic
droplets (i.e., pH <5), even a dominant role in the oxidation of SQ2 to HLSO,
(e.g., Hoffman and Edwards, 1975; Martin and Damschen, 1981; Chameides and
Davis, 1982; Calvert and Stockwell, 1983, 1984; Schwartz, 1984). Hydrogen
peroxide may also play a role in the oxidation of N0? dissolved in aqueous
droplets, although relevant data are limited and additional research is required
(see, e.g., Gertler et a!., 1984). Substantial uncertainties remain concerning
the quantitative role of HpOp in acidification of aqueous particles and droplets
(Richards et a!., 1983).
Because it can be scavenged rapidly into water droplets, formic acid can
t
potentially function as an oxidant in cloud water and rain water. Thus, HCOOH
is an example of a compound that is a non-oxidant or weak oxidant in the gas
phase but that is transformed upon incorporation in aqueous solutions into an
effective oxidizer of S(IV). Although much uncertainty remains concerning the
quantitative role of HCOOH and the higher organic acids, they potentially play
a minor but still significant role in the acidification of rain.
3.7.4 Meteorological and Climatological Processes
Meteorological and Climatological processes are important in determining
the extent to which precursors to ozone and other photochemical oxidants can
accumulate, and thereby the concentrations of ozone and other oxidants that
can result. The meteorological factors most important in the formation and
transport of ozone and other photochemical oxidants in the lower troposphere
are: (1) degree of atmospheric stability; (2) wind speed and direction;
(3) intensity and wavelength of sunlight; and (4) synoptic weather conditions.
These factors are in turn dependent upon or interrelated with geographic,
seasonal, and other Climatological factors.
3-130
-------
Incursions of ozone from the stratosphere are an additional source of the
ozone found in the lower troposphere. The physical and meteorological mechanisms
by which ozone is brought into the troposphere from the stratosphere are
important in determining the resulting ground-level concentrations, ground-level
locations impacted, and the seasonality of incursions of stratospheric ozone.
3.7.4.1 Atmospheric Mixing. The concentration of a pollutant in ambient air
depends significantly on the degree of atmospheric mixing that occurs from the
time the pollutant or its precursors are emitted and the arrival of the pollu-
tant at the receptor. The rate at which atmospheric mixing proceeds and the
extent of the final dilution depends on the amount of turbulent mixing that
occurs and on wind speed and direction. Atmospheric stability is one of the
chief determinants of turbulent mixing since pollutants do not spread rapidly
within stable layers nor do they mix upward through stable layers to higher
altitudes.
Temperature inversions, in which the temperature increases with increasing
altitude, represent the most stable atmospheric conditions. Surface inver-
sions (base at ground level) and elevated inversions (the entire layer is
above the surface) are both common (Hosier, 1961; Holzworth, 1964) and both
can occur simultaneously at the same location. Surface inversions show a
diurnal pattern, forming at night in the absence of solar radiation but break-
ing up by about mid-morning as the result of surface heating by the sun (Hosier,
1961; Slade, 1968). Elevated inversions can persist throughout the day and
pollutants can be trapped between the ground surface and the base of the
inversion. The persistence of elevated inversions is a major meteorological
factor contributing to high pollutant concentrations and photochemical smog
conditions along the California coast (Hosier, 1961; Holzworth, 1964; Robinson,
1952). In coastal areas generally, such as the New England coast (Hosier,
1961) and along the Great Lakes (Lyons and Olsson, 1972), increased atmospheric
stability (and diminished mixing) occurs in summer and fall as the result of
the temperature differential between the water and the land mass.
The depth of the layer in which turbulent mixing can occur (i.e., the
"mixing height") shows geographical dependence. Summer morning mixing heights
are usually >300 m in the United States except for the Great Basin (part of
Oregon, Idaho, Utah, Arizona, and most of Nevada), where the mixing height is
~200 m (Holzworth, 1972). By mid-morning, mixing heights increase markedly
such that only a few coastal areas have mixing heights <1000 m.
3-131
-------
Summer afternoon mixing heights are generally an indication of the poten-
tial for recurring photochemical oxidant problems. Photochemical smog problems
in the United States are somewhat unexpected since the lowest afternoon mixing
height is ~600 m (Holzworth, 1972). Elevated inversions having bases <500 m
(i.e.j low-level inversions) occur in the United States, however, with the
following frequencies: 90 percent on the California coast; >20 percent on the
Atlantic coast (New Jersey to Maine); >5 percent along the Great Lakes; and 5
to 10 percent from Louisiana to Arkansas and eastward to about Atlanta, Georgia.
For most areas of the United States, though, the persistence through the
afternoon of low-level stable layers is a rare event, occurring on <1 day in
20 (Holzworth and Fisher, 1979).
3.7.4.2 Wind Speed and Direction. For areas in which mixing heights are not
restrictive, wind speed and, in some cases, wind direction are major determi-
nants of pollution potential. Since strong winds dilute precursors to ozone
and other photochemical oxidants, a location may have good ventilation despite
the occurrence of persistent inversions (e.g., San Francisco). Conversely,
light winds can result in high oxidant levels even if the mixing layer is
deep.
The frequency of weak winds, then, is important in oxidant formation. In
industrialized, inland areas east of the Mississippi River, surface inversions
in the morning coupled with wind speeds £2.5 m/sec (£6 mi/hr) occur with a
frequency >50 percent (Holzworth and Fisher, 1979). These surface inversions
break up by afternoon, however, permitting dispersion.
The effects of wind speed and direction include the amount of dilution
occurring in the source areas, as well as along the trajectory followed by an
urban or source-area plume. Regions having steady prevailing winds, such that
a given air parcel can pass over a number of significant source areas, can
develop significant levels of pollutants even in the absence of weather patterns
that lead to the stagnation type of air pollution episodes. The Northeast
states are highly susceptible to pollutant plume transport effects, although
some notable stagnation episodes have also affected this area (e.g., Lynn
et a!., 1964). Along the Pacific Coast, especially along the coast of
California, coastal winds and a persistent low inversion layer contribute to
major pollutant buildups in urban source areas and downwind along the urban
plume trajectory (Robinson, 1952; Neiburger et a!., 1961).
3-132
-------
3.7.4.3 Effects of Sunlight and Temperature. The effects of sunlight on
photochemical oxidant formation, aside from the role of solar radiation in
meteorological processes, are related to its intensity and its spectral dis-
tribution. Intensity varies diurnally, seasonally, and with latitude, but the
effect of latitude is strong only in the winter. Experimental studies have
verified the effects on oxidant formation of light intensity (Peterson, 1976;
Demerjian et al., 1980) and its diurnal variations (Jeffries et al., 1975;
1976), as well as on the overall photooxidation process (Jaffee et al., 1974;
Winer et al., 1979).
A correlation between high oxidant concentrations and warm, above-normal
temperatures has been demonstrated generally (Bach, 1975; Wolff and Lioy,
1978) and- for specific locations, e.g., St. Louis (Shreffler and Evans, 1982).
Coincident meteorology appears to be the cause of the observed correlation.
Certain synoptic weather conditions are favorable both for the occurrence of
higher temperatures and for the formation of ozone and other oxidants, so that
temperature is often used to forecast the potential for high oxidant concen-
trations (e.g., Wolff and Lioy, 1978; Shreffler and Evans, 1982). Data from
smog chamber studies show an effect of temperature on ozone formation (e.g.,
Carter et al., 1979b; Countess et al., 1981), but the effect is thought to
result from the volatilization and reaction of chamber wall contaminants as
the temperature is increased.
3.7.4.4 Transport of OzoneandOther Oxidants and Their Precursors. -The
levels of ozone and other oxidants that will occur at a given receptor site
downwind of a precursor source area depend upon many interrelated factors,
which include but are not restricted to: (1) the concentrations of respective
precursors leaving the source area; (2) induction time; (3) turbulent mixing;
(4) wind speed and wind direction; (5) scavenging during transport; (6) in-
jection of new emissions from source areas in the trajectory of the air mass;
and (7) local and synoptic weather conditions.
Ozone and other photochemical oxidants can be transported hundreds of
miles from the place of origin of their precursors, as documented by the
numerous studies on transport phenomena that were described in the 1978 cri-
teria document for ozone and other photochemical oxidants (U.S. Environmental
Protection Agency, 1978a). In that document, transport phenomena were classi-
fied into three categories, depending upon transport distance: urban-scale,
3-133
-------
mesoscale, and synoptic-scale. In urban-scale transport, maximum concentra-
tions of Qg are produced about 20 miles or so (and about 2 to 3 hours) down-
wind from the major pollutant source areas. In mesoscale transport, 0~ has
been observed up to 200 miles downwind from the sources of its precursors.
Synoptic-scale transport is associated with large-scale, high-pressure air
masses that may extend over and persist for many hundreds of miles.
Urban-scale transport has been identified as a significant, characteris-
tic feature of the oxidant problem in the Los Angeles Basin (Tiao et al.,
1975), as well as in San Franciso, New York, Houston, Phoenix, and St. Louis
(e.g., Altshuller, 1975; Coffey and Stasiuk, 1975; Shreffler and Evans, 1982;
Wolff et al., 1977a). Simple advection of a photochemically reactive air
mass, local wind patterns, and diurnal wind cycles appear to be the main
factors involved in urban-scale transport.
Mesoscale transport is in many respects an extension of urban-scale
transport and is characterized by the development of urban plumes. Bell
documented cases in 1959 in which precursors from the Los Angeles Basin and
the resultant oxidant plume were transported over the coastal Pacific Ocean,
producing elevated oxidant concentrations in San Diego County the next day
(Bell, 1960). Similar scales of transport have been reported by Cleveland
et al. (1976a,b) for the New York-Connecticut area; by Wolff and coworkers and
others (Wolff et al., 1977a,c; Wolff and Lioy, 1978; Clark and Clarke, 1982;
Clarke et al., 1982; Vaughan et al., 1982) for the Washington, DC-Boston
corridor; and by Westberg and coworkers for the Chicago-Great Lakes area
(Sexton and Westberg, 1980; Westberg et al., 1981). These and other studies
have demonstrated that ozone-oxidant plumes from major urban areas can extend
downwind about 100 to 200 miles and can have widths of tens of miles (Sexton,
1982), frequently up to half the length of the plume.
Synoptic-scale transport is characterized by the general and widespread
occurrence of elevated oxidants and precursors on a regional or air-mass scale
as the result of certain favorable weather conditions, notably, slow-moving,
well-developed high-pressure, or anti-cyclonic, systems characterized by weak
winds and limited vertical mixing (Korshover, 1967; 1975), The size of the
region that can be affected has been described by Wolff and coworkers, who
reported the occurrence of haze and elevated ozone levels in an area extending
from the Midwest to the Gulf Coast (Wolff et al., 1982) and the occurrence of
elevated ozone concentrations extending in a virtual "ozone river" from the
3-134
-------
Gulf Coast to New England that affected anywhere from a few hundred square
miles to a thousand square miles during a 1-week period in July 1977 (Wolff
and Lioy, 1980).
3.7.4,5 Stratospheri'c-Tropospheric Ozone Exchange. The fact that ozone is
formed in the stratosphere, mixed downward, and incorporated into the tropo-
sphere, where it forms a more or less uniformly mixed background concentra-
tion, has been known in various degrees of detail for many years (Junge,
1963). It is widely accepted that the long-term average tropospheric back-
ground concentration of about 30 ppb to 50 ppb results primarily, though not
exclusively, from the transfer of stratospheric ozone into the upper tropo-
sphere, followed by subsequent dispersion throughout the troposphere (e.g.,
Kelly et a!., 1982).
The exchange of ozone between the stratosphere and the troposphere in the
middle latitudes occurs to a major extent in events called "tropopause folds"
(TF) (Reiter, 1963; Reiter and Mahlman, 1965; Danielsen, 1968; Reiter, 1975;
Danielsen and Mohnen, 1977; Danielsen, 1980), in which the polar jet stream
plays a major role. From recent studies, Johnson and Viezee (1981) proposed
four types or mechanisms of TF injection and concluded that two of these, both
of which are consistent with theory, could cause substantial effects in terms
of high ozone concentrations at ground level. They concluded, in addition,
that all low-pressure trough systems may induce TF events and cause the trans-
tropopause movement of ozone-rich air into the troposphere (Johnson and Viezee,
1981).
3.7.4.6 Stratospheric Ozone atGround Level. From a detailed review of
studies on background tropospheric ozone, Viezee and Singh (1982) concluded
that the stratosphere is a major but not the sole source of background ozone
in the unpolluted troposphere, a conclusion reached by other investigators as
well (e.g., Kelly et al., 1982). The stratospheric ozone reservoir shows a
strong seasonal cycle that is reflected at ground-level. At some stations
that monitor background ozone levels, average spring background levels may be
as high as 80 ppb, with average fall levels ranging from 20 to 40 ppb (e.g.,
Singh et al., 1977; Mohnen, 1977; U.S. Environmental Protection Agency, 1978a).
Viezee and Singh (1982) and Viezee et al. (1983) concluded that relatively
high ozone concentrations can occur for short periods of time (minutes, to a
few hours) over local areas as a result of stratospheric intrusions.
3-135
-------
A number of investigators have attempted to quantify the amount of the
surface ozone that can be attributed to stratospheric sources. The method
most commonly used is based on the assumption that beryl!ium-7 ( Be) is a
unique tracer for air parcels of stratospheric origin. Calculated correlations
between surface ozone and Be show, however, that their relationship is highly
variable (e.g., Kelly et al., 1982; Ferman and Monson, 1978; Johnson and
Viezee, 1981; Husain et al., 1977). Singh et al. (1980) and Viezee and Singh
(1982) have pointed out problems with using this technique to quantify the
contribution of stratospheric ozone to surface ozone. Singh et al. (1980)
concluded that "the experimental technique involving a Be/0- ratio to esti-
mate the daily stratospheric component of ground-level 0, is unverified and
considered to be inadequate for air quality applications" (p. 1009). This
group of investigators have suggested, however, that Be may be used, under
appropriate meteorological conditions, as a qualitative tracer for air masses
of stratospheric origin (Johnson and Viezee, 1981; Viezee et al., 1979).
Other methods used to attempt to quantify the stratospheric component of
surface ozone include aircraft observations of TF events coupled with calcula-
tions of downward ozone flux, and examination of surface ozone data records.
From such data, Viezee et al. (1983) concluded that direct ground-level contri-
butions from stratospheric ozone are infrequent (<1 percent of the time),
short-lived, and associated with ozone concentrations £0.1 ppm.
Notwithstanding difficulties with quantifying its contribution to surface
ozone, however, stratospheric ozone is clearly present in atmospheric surface
layers, and the meteorological mechanisms responsible have been described by a
number of investigators (e.g., Danielsen, 1968; Wolff et al.s 1979; Johnson
and Viezee, 1981).
3.7.4.7 Background Ozone from Photochemical Reactions. Whereas stratospheric
ozone is thought by many investigators to be the dominant contributor to
background levels of ozone, as discussed above, other investigators have
concluded that as much as two-thirds of the annual average background concen-
trations may result from photochemical reactions. For example, Altshuller
(1986), in a recent review article, has concluded that photochemically generated
ozone should equal or exceed the stratospheric contribution at lower-elevation
remote locations; and that photochemically generated ozone from manmade emissions
probably constitutes most of the ozone measured at more polluted rural locations
during the warmest months of the year. His conclusions were based, in part,
3-136
-------
on an analysis of global circulation (e.g., Levy et al., 1985) and photochemical
modeling .studies (e.g., Fishman and Seller, 1983; Fishman and Carney, 1984;
Fishman et al., 1985; Dignon and Hameed, 1985). In these modeling studies,
the photochemical contribution to background ozone levels was estimated to
range from ~15 ppb (long-term) to ~80 ppb (summertime), depending on the level
of NO emissions assumed.
/v
Studies on the role of NO in nonurban ozone photochemistry have shown
•r\
that ozone formation at many of the locations is not NO -limited, but depends
P\
on VOC reactions, as well (e.g., Martinez and Singh, 1979; Kelly et al., 1984;
Liu et al., 1984). Background NO concentrations at most remote, clean loca-
y\
tions range from <0.05 ppb upward. Mean concentrations of NO at nonurban
/\
locations in the United States east of the Rocky Mountains range from ~1 ppb
to 10 ppb (Altshuller, 1986; see also Sections 3.5 and 3.7.5). These background
concentrations of NO are higher than previously thought (see, e.g., Singh et
)\
al., 1980; Kelly et al., 1984, regarding global models and assumed reservoirs
of NOX).
The contributions of biogenic VOC to background ozone, although a matter
of controversy in recent years, appear not to be significant under most atmos-
pheric conditions, since ambient air concentrations of biogenic VOC are quite
low, even at rural sites (Altshuller, 1983).
Thus, photochemistry and stratospheric intrusions are both regarded as
contributing to background ozone concentrations, but the apportionment of
background to respective sources remains a matter of investigation.
3.7.5 Sources, Emissions,and Concentrations of Precursors toOzone and Other
Photochemical Oxidants
As noted earlier, photochemical production of ozone depends both on the
presence of precursors, volatile organic compounds (VOCs) and nitrogen oxides
(NO ), emitted by manmade and by natural sources; and on suitable conditions
of sunlight, temperature, and other meteorological factors. Because of the
intervening requirement for meteorological conditions conducive to the photo-
chemical generation of ozone, emission inventories are not as direct predic-
tors of ambient concentrations of secondary pollutants such as ozone and other
oxidants as they are for primary pollutants.
3.7.5.1 Sources and Emissions of Precursors. Emissions of manmade VOCs
(excluding several relatively unreactive compounds such as methane) in the
3-137
-------
United States have been estimated at 19.9 Tg/yr for 1983 (U.S. Environmental
Protection Agency, 1984a). Retrospective estimates show that manmade VOC
emissions rose from about 18.5 Tg/yr in 1940 to about 27.1 Tg/yr in 1970 (U.S.
Environmental Protection Agency, 1986). An examination of trends in manmade
VOC emissions for 1970 through 1983 shows that the annual emission rate for
manmade VOCs decreased some 26 percent during this period. The main sources
nationwide are industrial processes, which emit a wide variety of VOCs, such
as chemical solvents; and transportation, which includes the emission of VOCs
in gasoline vapor as well as in gasoline combustion products. Estimates of
biogenic emissions of organic compounds in the United States are highly
inferential but data suggest that the yearly rate is the same order of magni-
tude as manmade emissions. Most of the biogenic emissions actually occur
during the growing season, however, and the kinds of compounds emitted are
different from those arising from manmade sources.
Emissions of manmade NO in the United States were estimated at 19.4 Tg/yr
/\
for 1983. Retrospective estimates show that manmade NO emissions rose from
f\
about 6.8 Tg/yr in 1940 to about 18.1 Tg/yr' in 1970 (U.S. Environmental Protec-
tion Agency, 1986). Annual emissions of manmade NO were some 12 percent
jfX
higher in 1983 than in 1970, but the rate leveled off in the late 1970s and
exhibited a small decline from about 1980 through 1982 (U.S. Environmental
Protection Agency, 1984a). The increase over the period 1970 through 1983 had
two main causes: (1) increased fuel combustion in stationary sources such as
power plants; and (2) increased fuel combustion in highway motor vehicles, as
the result of the Increase in vehicle miles driven. Total vehicle miles
driven increased by 42 percent over the 14 years in question.
Estimated biogenic NO emissions are based on uncertain extrapolations
yx
from very limited studies, but appear to be about an order of magnitude less
than manmade NO emissions.
/\
3.7.5.2 Representative Concentrations in Ambient Air.
3.7.5.2.1 Hydrocarbons in urban areas. Most of the available ambient air
data on the concentrations of nonmethane hydrocarbons (NMHC) in urban areas
have been obtained during the 6:00 to 9:00 a.m. period. Since hydrocarbon
emissions are at their peak during that period of the day, and since atmospheric
dispersion is limited that early in the morning, NMHC concentrations measured
then generally reflect maximum diurnal levels. Representative data for urban
areas show mean NMHC concentrations between 0.4 and 0.9 ppm.'
3-138
-------
The hydrocarbon composition of urban atmospheres is dominated by species
in the C2 to C-,Q molecular-weight range. The paraffinic hydrocarbons (alkanes)
are most prominent, followed by aromatics and alkenes. Based on speciation
data obtained in a number of urban areas, alkanes generally constitute 50 to
60 percent of the hydrocarbon burden in ambient air, aromatics 20 to 30 percent,
with alkenes and acetylene making up the remaining 5 to 15 percent (Sexton and
Westberg, 1984).
3.7,5.2.2 Hydrocarbons in nonurbanareas. Rural nonmethane hydrocarbon
concentrations are usually one to two orders of magnitude lower than those
measured in urban areas (Ferman, 1981; Sexton and Westberg, 1984). In samples
from sites carefully selected to guarantee their rural character, total NMHC
concentrations ranged from 0.006 to 0.150 ppm C (e.g., Cronn, 1982; Seila,
1981; Holdren et al., 1979). Concentrations of individual species seldom
exceeded 0.010 ppm C. The bulk of species present in rural areas are alkanes;
ethane, propane, ri-butane, iso-pentane, and ri-pentane are most abundant.
Ethylene and propene are sometimes present at <0,001 ppm C, and toluene is
usually present at ~0.001 ppm C. Monoterpene concentrations are usually
£0,020 ppm C. During the summer months, isoprene concentrations as high as
0.150 ppm C have been measured (Ferman, 1981). The maximum concentrations of
isoprene usually encountered, however, are in the range of 0.030 to 0.040
ppm C.
3.7.5.2.3 Nitrogen oxides in urban areas. Concentrations of NO . like hydro-
f\
carbon concentrations, tend to peak in urban areas during the early morning,
when atmospheric dispersion is limited and automobile traffic is dense. Most
NO is emitted as nitric oxide (NO), but the NO is rapidly converted to N0?,
x, *~
initially by thermal oxidation and subsequently by ozone and peroxy radicals
produced in atmospheric photochemical reactions. The relative concentrations
of NO versus N0? fluctuate day-to-day, depending on diurnal and day-to-day
fluctuations in ozone levels and photochemical activity.
Urban NO concentrations during the 6:00 to 9:00 a.m. period in 10 cities
/\
ranged from 0.05 to 0.15 ppm in studies done in the last 5 to 7 years (e.g.,
Westberg and Lamb, 1983; Richter, 1983; Eaton et al., 1979), although concen-
trations two to three times higher occur in cities such as Los Angeles. ,
Concurrent NMHC measurements for these 10 cities showed that NMHC/NO ratios
J\
ranged from 5 to 16.
3-139
-------
3.7.5.2.4 Nitrogen oxides In nonurban areas. Concentrations of NO in clean
f\
remote environments are usually <0.5 ppb (Logan, 1983). In exceptionally
clean air, NO concentrations as low as 0.015 ppb have been recorded (Bellinger
f\
et a!,, 1982). Concentrations of NO at nonurban sites1.;in the northeastern
s f\.
United States appear to be higher than NO concentrations in the west by a
A,
factor of ten (Mueller and Hidy, 1983). From the limited amount of data
available, NO concentrations in unpopulated nonurban areas in the west average
*\
<1 ppb; but in nonurban northeastern areas average NO can exceed 10 ppb.
^~ " Px
3.7.6 Source-Receptor (Oxidant-Precursor) Models
In order to apply knowledge of the atmospheric chemistry of precursors,
and of ozone and other photochemical oxidants, during their dispersion and
transport, models describing these phenomena have been developed in a variety
of forms over the past 15 years. Most of these models relate the rates of
precursor emissions from mobile and stationary sources, or precursor atmos-
pheric concentrations, to the resulting ambient concentrations of secondary
pollutants that impact receptors at .downwind sites. For this reason they have
been described as source-receptor models.
Current air quality, or source-receptor, models can be classified as
either statistical or computational-dynamic. Statistical models are
generally based on a statistical analysis of historical air quality data,
and are not explicitly concerned with atmospheric chemistry or meteorology.
An example of empirical models is the linear rollback concept.
Computational, or dynamic, models attempt to describe mathematically the
atmospheric chemical and physical processes influencing air pollution forma-
tion and impacts. Examples of computational models include trajectory and
fixed-grid airshed models. Two phenomenologically different approaches have
been employed in dynamic models with respect to the coordinate systems chosen.
A coordinate system fixed with respect to the earth is termed Eulerian, while
in Lagrangian models the reference frame moves with the air parcel whose
behavior is being simulated.
3.7.6.1 Trajectory Models. In trajectory models, a moving-coordinate system
describes pollutant transport as influenced by local meteorological conditions.
Trajectory models provide dynamic descriptions of atmospheric source-receptor
relationships that are simpler and less expensive to derive than those obtained
from fixed-cell models.
3-140
-------
The simplest form of trajectory model is the empirical kinetic modeling
approach (EKMA). This approach was developed from earlier efforts
(Dimitriades, 1972) to use smog chamber data to develop graphical relationships
between- morning NMOC and NO levels and afternoon maximum concentrations of
/\ . • i
ozone. In applying EKMA, the Ozone Isopleth Plotting Package (OZIPP) (Whitten
and Hogo, 1978) is used to generate ozone isopleths at various levels of
sophistication corresponding to "standard" EKMA, "city-specific" EKMA, or the
simplified trajectory model (F.R., 1979). The EKMA isopleths generated are used
to determine the relative degree of control of precursor emissions needed to
achieve a given percentage reduction in ozone.
The use of EKMA in ozone abatement programs is relatively widespread. It
is therefore worth noting the general control implications of EKMA isopleths.
For areas with high levels of morning precursor emissions and meteorology
conducive to oxidant formation, such as Los Angeles, for example, EKMA isopleths
predict that (1) at high NMQC/NQ concentration ratios, reductions in NO will
y\ J\
decrease ozone formation; (2) at moderate NMOC/NO ratios, reductions in NMOC
s\.
and NO will decrease ozone formation; and (3) at very low NMOC/NO ratios,
/\ y\
increases in NO will inhibit ozone formation. These predictions cannot be
/\
assumed to apply to all urban areas, or even to all high-oxidant urban areas,
since the shape of the EKMA isopleths is a function of numerous factors, many
of which are location-specific. For discussions of the specific assumptions
employed in EKMA and the underlying chemistry and meteorology, the primary
literature should be consulted (e.g., Dimitriades, 1970, 1972, 1977a,b; Dodge,
1977a,b; Whitten and Hogo, 1977; U.S. Environmental Protection Agency, 1977,
1978a; Whitten, 1983). Likewise, the primary literature should be consulted
for additional data and discussions on the respective effects on ozone forma-
tion of controlling NMHC and NO (e.g., Liu and Grisinger, 1981; Chock et al.,
1981; Kelly, 1985; Kelly et al., 1986; Glasson and Tuesday, 1970; Dimitriades,
1970, 1972, 1977a,b).
3.7.6.2 Fixed-Grid Models. Fixed-grid models, also called regional airshed
models, are based on two- or three-dimensional arrays of grid cells and are
the most sophisticated source-receptor models presently available. Such
models are computationally complex and require the most extensive set of input
data; but they also provide the most realistic treatment of the various processes
involved in photochemical air pollution formation.
3-141
-------
3.7.6.3 Box Models. Box models (Hanna, 1973; Demerjian and Schere, 1979;
Derwent and Hov, 1980) are the simplest of dynamic models. They treat the
atmosphere as a single cell, bounded by the mixing layer, having an area on
the order of 100 square miles.
3.7.6.4 Validation and Sensitivity Analyses for Dynamic Models. All present-
ly available source-receptor models require a degree of simplifying assump-
tions to deal with practical limitations imposed by existing computer capabil-
ities, time and cost constraints, or lack of knowledge concerning inputs such
as boundary conditions, emissions, or detailed reaction mechanisms. The
reliability and applicability of any particular model therefore depends upon
its specific limitations, data requirements, and degree of validation against
experimental data from ambient air measurements or environmental chamber runs.
Reliability and applicability also depend on the quality of the chemical
kinetics mechanisms used to define the 0,-HC-NO relationship.
O X
Attempts are made to validate model predictions by comparing them with
real observations; and operating parameters are often varied to determine the
sensitivity of the model to respective parameter changes (Gelinas and Vajk,
1979). In addition, the extent of agreement between the results from two
simulations can be tested. In this way, completely different models may be
compared, or an internal component, such as the chemical kinetics mechanisms,
may be substituted and the model run again to ascertain the effect of such
substitutions.
3.8 REFERENCES
Aikin, A. C.; Herman, J. R.; Maier, E. J. R.; McQuillan, C. J. (1983) Influence
of peroxyacetyl nitrate (PAN) on odd nitrogen in the troposphere and lower
stratosphere. Planet. Space Sci. 31: 1075-1082.
Akimoto, H.; Bandow, H.; Sakamaki, F.; Inoue, G.; Hoshino, M.; Okuda, M. (1980)
Photooxidation of the propylene-NQ -air system: studies by long-path Fourier
transform infrared spectrometry. tnviron. Sci. Techno!. 14: 172-179.
Allwine, G.; Lamb, B.; Westberg, H. (1983) Application of atmospheric tracer
techniques for determining biogenic hydrocarbon fluxes from an oak tracer
In: Proceedings of forest environmental measurements conference; October;
Oak Ridge, TN.
3-142
-------
Altshuller, A. P. (1975) Evaluation of oxidant results at CAMP sites in the
United States. J. Air Pollut. Control Assoc. 25: 19-24.
Altshuller, A. P. (1983) Review: natural volatile organic substances and their
effect on air quality in the United States. Atmos. Environ. 17: 2131-2165.
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. Atmos. Environment. 20: 245-268.
Altshuller, A. P.; Bufalini, J. J. (1971) Photochemical aspects of air pollu-
tion: a review. Environ. Sci. Techno!. 5: 39-64.
Altshuller, A. P.; McPherson, S. P. (1963) Spectrpphotometric analysis of alde-
hydes in the Los Angeles atmosphere. J. Air Pollut. Control Assoc.
13: 109-111.
Ames, J.; Myers, T. C.; Reid, L. E.; Whitney, D. C.; Golding, S. H.; Hayes,
S. R.; Reynolds, S. D. (1978) The user's manual for the SAI airshed model.
San Raphael, CA: Systems Applications, Inc. No. EM78-89.
Anbar, M.; Taube, H. (1954) Interaction of nitrous acid with hydrogen peroxide
and with water. J. Am. Chem. Soc. 76: 6243-6247.
Anderson, G. E.; Hayes, S. R.; Hillyer, M, J.; Killus, J. P.; Mundkur, P. V.
(1977) ,Air quality in the Denver metropolitan region: .1974-2000. Denver,
CO: Region VIII, U.S. Environmental Protection Agency. EPA-908/1-77-002.
Andrews, A. W.; Thibault, L. H.; Lijinsky, W. (1978) The relationship between
mutagenicity and carcinogenicity of some nitrosamines. Mutat. Res. 51:
319-326.
Appel, B. R. ; Kothny, E. L.; Hoffer, E. M.; Wesolowski, J. J. (1980) Sulfate
and nitrate data from the California Aerosol Characterization Experiment
(ACHEX). In: Hidy, G.M.; Mueller, P. R.; Grosjean, D.; Appel, B. R.;
Wesolowski, J. J., eds. The character and origins of smog aerosols: a
digest of results from the California Aerosol Characterization Experiment
(ACHEX). New York, NY: John Wiley and Sons; pp. 315-335. (Advances in en-
vironmental science and technology: v. 9).
Arnts, R. R.; Gay, B. W., Jr. (1979) Photochemistry of some naturally emitted
hydrocarbons. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Research and Development; EPA report no. EPA-600/3-79-081.
Available from: NTIS, Springfield, VA; PB80-131980.
Arnts, R. R.; Meeks, S, A. (1981) Biogenic hydrocarbons: contribution to the
ambient air in selected areas. Atmos. Environ. 15: 1643-1651.
Arnts, R. R.; Seila, R. L.; Kuntz, R. L.; Mowry, F. L.; Knoerr, K. R.; Dudgeon,
A. C. (1978) Measurement of alpha-pinene fluxes from a loblolly pine forest.
In: Proceedings of the 4th joint conference on sensing of environmental
pollutants, American Chemical Society. Washington, DC: American Chemical
Society.
3-143
-------
Arnts, R. R.; Gay, B. W.; Bufalini, J. J. (1981) Photochemical oxidant poten-
tial of the biogenic hydrocarbons. In: Bufalini, J. J.; Arnts, R. R.,
eds. Atmospheric biogenic hydrocarbons, v. II. Ann Arbor, MI: Ann Arbor
Science Publishers; p. 117.
Ashby, H. A.; Stahman, R. C.; Eccleston, B. H.; Hum, R. W. (1974) Vehicle
emissions—summer to winter. In: Society of Automotive Engineers meeting;
October; Toronto, Canada. Warrendale, PA: Society of Automotive Engineers;
SAE technical paper 741053.
Association of Bay Area Governments. (1979a) Application of photochemical
models, Vol. I: The use of photochemical models in urban ozone studies.
Research Triangle Park, NC: U. S. Environmental Protection Agency; EPA
report no. EPA-450/4-79-025.
Association of Bay Area Governments. (1979b) Application of photochemical
models, Vol. Ill: Recent sensitivity tests and other applications of the
LIRAQ model. U. S. Environmental Protection Agency, Research Triangle Park,
NC; EPA report no. EPA-450/4-79-027.
Atkinson, R. (1985) Kinetics and mechanisms of the gas phase reactions of
the hydroxyl radical with organic compounds under atmospheric conditions.
Chem. Rev.: in press.
Atkinson, R.; Aschmann, S. M. (1984) Rate constants for the reactions of 03
and OH radicals with a series of alkynes. Int. J. Chem. Kinet. 16: 259-268.
Atkinson, R.; Carter, W. P. L. (1984) Kinetics and mechanisms of the reactions
of ozone with organic compounds in the gas phase. Chem. Rev. 84: 437-470.
Atkinson, R.; Lloyd, A. C. (1981) Smog chemistry and urban airshed modeling.
In: Rodgers, M. A. J.; Powers, E. L., eds. Oxygen and oxy-radicals in
chemistry and biology. New York; Academic Press, pp. 559-592.
Atkinson, R.; Lloyd, A. C. (1984) Evaluation of kinetic and mechanistic data
for modeling of photochemical smog. J. Phys. Chem. Ref. Data 13: 315-444.
Atkinson, R.; Darnall, K. R.; Pitts, J. N., Jr. (1978) Rate constants for the
reactions of OH radicals and ozone with cresols at 300 ± 1 K. J. Phys.
Chem. 82: 2759-2761.
Atkinson, R.; Darnall, K. R.; Lloyd, A. C. ; Winer, A. M.; Pitts, J. N., Jr.
(1979) Kinetics and mechanisms of the reactions of the hydroxyl radical
with organic compounds in the gas phase. In: Pitts, J. N., Jr.; Hammond,
G. S.; Gollnick, K.; Grosjean, D., eds. Advances in photochemistry, v.
11. New York, NY: John Wiley and Sons. pp. 375-488.
Atkinson, R.; Carter, W. P. L.; Darnall, K. R.; Winer, A. M.; Pitts, J. N.,
Jr. (1980) A smog chamber and modeling study of the gas phase NO -air
photooxidation of toluene and the cresols. Int. J. Chem. Kinet. 12: 779-836.
Atkinson, R.; Aschmann, S. M.; Winer, A. M.; Pitts, J. N., Jr. (1981) Rate
constants for the gas-phase reactions of 03 with a series of carbonyls at
296 K. Int. J. Chem. Kinet. 13: 1133-1142.
3-144
-------
Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M. ; Pitts, J. H. ,
Jr. (1982a) Alky! nitrate formation from the NO -air photooxidations of
C2-C8 n-alkanes. J. Phys. Chem. 86: 4563-4569.
Atkinson, R.; Aschmann, S. M. ; Fitz, D. R.; Winer, A, M.; Pitts, J. N. , Jr.
(1982b) Rate constants for the gas phase reactions of 03 with selected
organics at 296 K. Int. J. Chem. Kinet. 14: 13-18.
Atkinson, R.; Lloyd, A. C.; Winges, L. (1982c) An updated chemical mechanism
for hydrocarbon/NO /SO photooxidation suitable for inclusion in atmospheric
simulation models. Atmos. Environ. 16: 1341-1355.
Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N.,
Jr. (1982d) Kinetics of the gas-phase reactions of OH radicals with alkyl
nitrates at 299 ± 2 K. Int. J. Chem. Kinet. 14: 919-926.
Atkinson, R.; Carter, W. P. L.; Winer, A. M. (1983) Effects of temperature
and pressure on alkyl nitrate yields in the NO photooxidations of
n-pentane and n-heptane. J. Phys. Chem. 87: 2012-2018.
Atkinson, R.; Plum, C. N.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.
(1984a) Rate constants for the gas-phase reactions of N03 radicals with a
series of organics in air at 298 ± 1 K. J. Phys. Chem. 88: 1210-1215.
Atkinson, R.; Pitts, J. N., Jr.; Aschmann, S. M. (1984b) Tropospheric re-
actions of dimethyl sulfide with N03 and OH radicals. J. Phys. Chem. 88:
1584-1587.
Atkinson, R,; Aschmann, S. M.; Winer, A. M.; Pitts, J. N., Jr. (1984c)
Kinetics of the gas-phase reactions of N03 radicals with a series of
dialkenes, cycloalkenes, and monoterpenes at 295 ± 1 K. Environ. Sci.
Technol. 18: 370-375.
Atkinson, R.; Carter, W. P. L.; Plum, C. N.; Winer, A. M.; Pitts, J. N., Jr.
(1984d) Kinetics of the gas-phase reactions of N03 radicals with a series
of aromatics at 296 ± 2 K. Int. J. Chem. Kinet. 16: 887-898.
Atkinson, R.; Plum, C. N. ; Carter, W. P. L. ; Winer, A. M.; Pitts, J. N., Jr.
(1984e) Kinetics of the gas phase reactions of N03 radicals with a series
of alkanes at 296 ± 2 K. J. Phys. Chem. 88: 2361-2364.
Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N.,
Jr. (1984f) Formation of alkyl nitrates from the reaction of branched and
cyclic alky peroxy radicals with NO. Int. J. Chem. Kinet. 16: 1085-1101.
Atkinson, R.; Aschmann, S. M.; Winer, A. M.; Pitts, J. N., Jr. (1985) Kinetics
and atmospheric implications of the gas-phase reactions of N03 radicals
with a series of monoterpenes and related organics at 294 ± 2 K. Environ.
Sci. Technol. 19: 159-163. .
Attmannspacher, W.; Hartmannsgruber, R. (1973) On extremely high values of
ozone near the ground. Pure Appl. Geophys. 106-108: 1091-1096.
3-145
-------
Automotive News. (1982a) Sales of diesel-powered cars in the U.S. In: 1982
market data book issue. April 28: 74.
Automotive News. (1982b) Automotive news analysis of new-car registrations.
In: 1982 market data book issue. April 28: 8.
Automotive News. (1983a) Sales of new diesel-powered cars in the U.S. In:
1983 market data book issue. April 27: 70.
Automotive News. (1983b) Domestic car sales. 1982 market data book issue.
Bach, W. D., Jr. (1975) Investigation of ozone and ozone precursor concentra-
tions in nonurban locations in the eastern United States; phase II. meteo-
rological analyses. Research Triangle Park, NC: U.S. Environmental Protec-
tion Agency, Office of Air Quality Planning and Standards; EPA report no.
EPA-450/3-74-034a. Available from NTIS, Springfield, VA; PB-246899.
Baldwin, A. C.; Barker, J. R.; Golden, D. M.; Hendry, D. G. (1977) Photo-
chemical smog. Rate parameter estimates and computer simulations. J.
Phys. Chem. 81: 2483-2492.
Bandow, H.; Okuda, M.; Akimoto, H. (1980) Mechanism of the gas-phase reactions
of C3H6 and N03 radicals. J. Phys. Chem. 84: 3604-3608.
Barth, D. S. (1970) Federal motor vehicle emission goals for CO, HC and NO
based on desired air quality levels. J. Air Pollut. Contr. Assoc. 20:
519-523.
Bartok, W.; Crawford, A. R.; Piegari, J. (1971) Systematic field study of NO
emission control methods for utility boilers. Research Triangle Park, N(T:
U.S. Environmental Protection Agency; APTD-1163. Available from: NTIS,
Springfield, VA; PB-210739.
Bass, A. M.; Glasgow, L. C.; Miller, C.; Jesson, J. P.; Filkin, D. L. (1980)
Temperature dependent absorption cross sections for formaldehyde (CHgO):
the effect of formaldehyde on stratospheric chlorine chemistry. Planet.
Space Sci. 28: 675-679.
Batt, L. (1979) Gas phase decomposition of alkoxy radicals. Int. J. Chem.
Kinet. 11: 977-993.
Bird, R. B.; Stewart, W. E.; Lightfoot, E. N. (1960) Concentration distri-
bution in turbulent flow. In: Transport phenomena. New York: John Wiley
and Sons, pp. 626-635.
Bell, G. B. (1960) Meteorological conditions during oxidant episodes in
coastal San Diego County in October and November, 1959. Sacramento, CA:
Department of Public Health.
Bjerknes, J.. (1951) Extratropical cyclones. In: Compendium of meteorology.
Boston, MA: American Meteorological Society; pp. 557-598.
3-146
-------
Black, F. M. (1977) The impact of emissions control technology on passenger
car hydrocarbon emisson rates and patterns. In: Dimitriades, B., ed.
International conference on photochemical oxidant pollution and its
control; September 1976; > Raleigh, NC. Proceedings: v. II. Research
Triangle Park, NC: U.S. Environmental Protection Agency; pp. 1053-1068;
EPA report no. EPA-600/3-77-QQlb. Available from: NTIS, Springfield,
VA; PB-264233.
Black, F. M.; Bradow, R. L. (1975) Patterns of hydrocarbon emissions from 1975
production cars. Presented at: Society of Automotive Engineers' national
fuels and lubricants meeting; June; Houston, TX. Warrendale, PA: Society
of Automotive Engineers; SAE paper no. 750681.
Black, F. M.; High, L. E. (1979) Diesel hydrocarbon emissions, particulate and
gas phase. Presented at: Society of Automotive Engineers' symposium on
diesel particulate emissions measurement characterization; February; Detroit,
MI. Warrendale, PA: Society of Automotive Engineers.
Black, F. M.; High, L. E.; Lang, J. M. (1980) Composition of automobile evapo-
rative and tailpipe hydrocarbon emissions. J. Air Pollut. Control Assoc.
30: 1216-1221.
Bluraenthal, D. L.; Richards, L. W.; Macias, E. S.; Bergstrom, R. W.; Wilson,
W. E.; Bhardwaja, P. S. (1981) Effects of a coal-fired power plant and
other sources on southwestern visibility (interim summary of EPA's project
VISTTA). Atmos. Environ. 15: 1955-1969.
Bellinger, M. J.; Parrish, D. D.; Hahn, C.; Albritton, D. L.; Fehsenfeld, F.
C. (1982) NO measurements in clean continental air. In: Proceedings of
2nd symposium on composition of the nonurban troposphere; May; Williams-
burg, VA. (see also, Bellinger, M., et al. (1984) NO measurements in
clean continental air and analysis of the contributing meteorology. J.
Geophys. Res. 89: 9623-9631).
Box, E. 0. (1981) Foliar biomass: data base of the International Biological
Program and other sources. In: Bufalini, J. J.; Arnts, R. R., eds. Atmos-
pheric biogenic hydrocarbons, v. 1. Ann Arbor, MI: Ann Arbor Science;
pp. 107-119.
Braddock, J. N.; Bradow, R. L. (1976) Emission patterns of diesel-powered pas-
senger cars. Presented at meeting of Society of Automotive Engineers,
Houston, TX, June 3-5, 1975. SAE Paper No. 750682.
Braverman, T. N.; Layland, D. E. (1982) Application of urban emission inven-
tories to photochemical modeling. Proceedings of Air Pollution Control
Association Specialty Conference on Emission Inventories and Air Quality
Management, Kansas City, MO, April.
Breeding, R. J.; Lodge,; J. P., Jr.; Pate, J. B.; Sheesley, D. C.; Klonis, H.
B.; Fogle, B.; Anderson, J. A.; Englert, T. R.; Haagenson, P. L.; McBeth,
R. B.; Morris, A. L.; Pogue, R.; Wartburg, A. F. (1973) Background trace
gas concentrations in the central United States. JGR J. Geophys. Res. 78:
7057-7064.
3-147
-------
Brock, J. R.; Durham, J. L. (1984) Aqueous aerosol as a reactant i the atmos-
phere. In: Calvert, J. G., ed. S02, NO and N02 oxidation mechanisms: atmos-
pheric considerations, Boston, MA: Butterworth Publishers; pp. 209-250.
(Teasley, J. I., ed. Acid precipitation series, v. 3).
Bruckmann, P. W.; Winner, H. (1983) Infrared spectroscopic study of peroxy-
acetyl nitrate (PAN) and its decomposition products. Environ. Sci, Techno!.
17: 352-357.
Bucon, H. W.; Macko, J. F.; Taback, H. J. (1978) Volatile organic compound (VOC)
species data manual. Prepared by KVB Engineering, Inc., Tustin, CA.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA report
no. EPA-450/3-78-119. Available from: NTIS, Springfield, VA; PB80-113822.
Bufalini, J. J.; Arnts, R. R., eds. (1981) Atmospheric biogenic hydrocarbons,
Vols. 1 and 2. Ann Arbor, MI; Ann Arbor Science,
Bufalini, J. J.; Lonneman, W. A., eds. (1977) Proceedings of symposium on 1975
northeast oxidant transport study. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency. Report no. EPA-600/3-77-017.
Burns, R. C.; Hardy, R. W. F. (1975) Nitrogen fixation in bacteria and higher
plants. New York: Springer-Verlag.
California Bureau of Air Sanitation. (1966) The oxides of nitrogen in air
pollution. Sacramento, CA: California Department of Public Health.
Calvert, J. G.; Mohnen, V. (1983) The chemistry of acid formation. In: Acid
deposition: atmospheric processes in eastern North America. Washington,
DC: National Academy Press.
Calvert, J. G.; Stockwell, W. R. (1983) Acid generation in the troposphere by
gas-phase chemistry. Environ. Sci. Technol. 17: 428A-443A.
Calvert, J. G,; Stockwell, W. R. (1984) The mechanism and rates of the gas-phase
oxidation of sulfur dioxide and nitrogen oxides in the atmosphere. In:
Calvert, J. G., ed., Acid precipitation: S02, NO and N02 oxidation mechan-
isms: atmospheric considerations. Boston, MA: Butterworth Publishers.
(Teasley, J. I., ed., Acid precipitation series v. 3)
Calvert, J. G.; Su, F.;, Bottenheim, J. W.; Strausz, 0. P. (1978) Mechanism of
the homogeneous oxidation of sulfur dioxide in the troposphere. Atmos.
Environ. 12: 197-226.
Carter, W. P. L; Carnal 1, K. R.; Lloyd, A. C.; Winer, A. M.; Pitts, J. N.,
Jr. (1976) Evidence for alkoxyl radical isomerization in photooxidation
of C4~C6 alkanes under simulated atmospheric conditions. Chem. Phys, Lett.
42: 22-27. ,
Carter, W. P. L.; Lloyd, A. C.; Sprung, J. L.; Pitts, J. N., Jr. (1979) Computer
modeling of smog chamber data: progress in validation of detailed mecha-
nism for the photooxidation of propene and _n-butane in photochemical smog.
Int. J. Chem. Kinet. 11: 45-101.
3-148
-------
Carter, W. P. L.; Winer, A. M.; Darnall, K. R.;
chamber studies of temperature effects in
Sci, Techno!. 13: 1094-1100.
Pitts, J. N.,
photochemical
Jr. (1979b) Smog
smog. Environ.
Carter, W. P. L.; Winer, A. M.; Pitts,
sink for phenol and the cresols:
Environ. Sci. Techno!. 15:. 829-831.
J. N., Jr. (1981a) Major atmospheric
reaction with the nitrate radical.
Carter, W. P. L.; Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Gas
phase reactions of N,N-dimethylhydrazine with ozone and NO in simulated '
atmospheres. In: Scanlan, R, A.; Tannenbaum, S. R., eds. tr-nitroso com-
pounds. Washington, DC: American Chemical Society; pp. 118-131. (American
Chemical Society symposium series no. 174).
Carter, W. P. L.; Winer, A. M.;
nitrate on the initiation
15: 831-834.
Pitts, J. N., Jr. (1981c) Effect of peroxyacetyl
of photochemical smog. Environ. Sci. Techno!.
Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr. (1982) Effects of kinetic
mechanisms and hydrocarbon composition on oxidant-precursor relationships
predicted by the EKMA isopleth technique. Atmos. Environ. 16: 113-120.
Cato, G. A.; Buening, H. J.; DeVivo, C. C.; Morton, B. G.; Robinson, J. M. (1974)
Field testing: Application of combustion modifications to control pollutant
emissions from industrial boilers—phase I. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-650/2-74-078a.
Available from: NTIS, Springfield, VA; PB-238920.
Cato, G. A.; Muzio, L. J.; Shore, D. E. (1976) Field testing: Application of
combustion modifications to control pollutant emissions from industrial
boilers—phase II. Research Triangle Park, NC: U.S. Environmental Protec-
tion Agency; EPA report no. 600/2-76-086a.
ChameideSj W. L.; Davis, D. D. (1982) The free radical chemistry of cloud drop-
lets and its impact upon the composition of rain. JGR J. Geophys. Res.
87: 4863-4877.
Chameides, W. L.; Walker, J. C. G. (1976) A photochemical theory of tropospheric
ozone. J. Geophys. Res. 78: 8751-8760.
Chameides, W. L:; Stedman, D. H.; Dickerson, R. R.; Rusch, D. W.; Cicerone,
R. J. (1977) NO production in lightning. J. Atmos. Sci. 34: 143-149.
"
Chameides, W.L.; Tan, A. (1981). The two-demensional diagnostic model for
tropospheric OH: An uncertainty analysis. J. Geophys. Res. 86: 5209-5223.
Chatfield, R. B.; Rasmussen, R, A. (1977) Hydrocarbons, f1uorocarbons, and
oxidants observed at a rural site near St. Louis. In: Proceedings from
national AIChE meeting; March; Houston, TX.
Chock, D. P.; Dunker, A. M.; Kumar, S.; Sloane, C. S. (1981) Effect of NO
emission rates on smog formation in the California South Coast Air Basin.
Environ. Sci. Techno!. 15: 933-939.
3-149
-------
Clark, T. L.; Clarke, J. F. (1982) Boundary layer transport of NO and 03 from
Baltimore, Maryland—a case study. Presented at: 75th annual meeting of
the Air Pollution Control Association; June; New Orleans, LA. Pittsburgh,
PA: Air Pollution Control Association; paper no. 82-24.3.
Clarke, J. F.; Ching, J. K. S.; Brown, R. M.; Westberg, J.; White, J. H. (1982)
Regional transport of ozone. Presented at: Third conference on air pollu-
tion meteorology; January. Boston, MA: American Meteorological Society;
paper no. 1.1.
Cleveland, W. S.; Guarino, R.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976a)
The analysis of the ozone problem in the northeast United States. In:
Specialty conference on ozone/oxidants—interactions with the total environ-
ment. Pittsburgh, PA: Air Pollution Control Association; pp. 109-120.
Cleveland, W. S.; Kleiner, B.; McRae, J. E.; Warner, J. L. (1976b) Photochemical
air pollution: transport from New York City area into Connecticut and
Massachusetts. Science (Washington, DC) 191: 179-181.
Coffey, P. E. (1977) The issue of natural organic emissions. In: International
conference on oxidants, 1976—analysis of evidence and viewpoints: part IV.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA report
no. EPA-600/3-77-116. Available from: NTIS, Springfield, VA; PB-277462.
Coffey, P. E.; Stasiuk, W. N. (1975) Evidence of atmospheric transport of ozone
into urban areas. Environ. Sci. Techno!. 9: 59-62.
Coffey, P. E,; Stasiuk, W. N.; Mohnen, V. (1977) Ozone in rural and urban areas
of New York State. In: Dimitriades, B., ed. International conference on
photochemical oxidant pollution and its control; September 1976; Raleigh,
NC. Proceedings: volume I. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Research and Development; pp. 89-96; EPA report
no. EPA-600/3-77-001a. Available from: NTIS, Springfield, VA; PB-264232.
Cole, H. S.; Newberry, C. F.; Cox, W.; Moss, G. K.; Layland, D. (1982) Appli-
cation of the airshed model for ozone control in St. Louis. Proceedings
of the 75th Annual Meeting of the Air Pollution Control Association. New
Orleans, LA, June 20-25.
Council for Agricultural Science and Technology. (1976) Effect of increased
nitrogen fixation on stratospheric ozone. CAST report no. 53.
Countess, R. J.; Wolff, G. T.; Whitbeck, M. R. (1981) The effect of temperature
on ozone formation in the propene/nitrogen dioxide/air system. J. Environ.
Sci. Health. A16: 1-8.
Cox, R. A.; Burrows, J. P. (1979) Kinetics and mechanism of the disproportion-
ation of H02 in the gas phase. J. Phys. Chem. 83: 2560-2568.
Cox, R. A.; Penkett, S. A. (1972) Aerosol formation from sulfur dioxide in
the presence of ozone and olefinic hydrocarbons. J. Chem. Soc. Faraday
Trans. I 68: 1735-1753.
3-150
-------
Cox, R. A.; Roffey, M. J. (1977) Thermal decomposition of peroxyacetyl nitrate
in the presence of nitric oxide. Environ. Sci. Technol. 11: 900-906.
Cox, R.. A.; Sheppard, D. (1980) Reactions of OH radicals with gaseous sulfur-
compounds. Nature 284: 330-331.
Cox, R. A.; Derwent, R. G. ; Kearsey, S. V.; Batt, L.; Patrick, K. G. (1980)
Photolysis of methyl nitrite: kinetics of the reaction of the methoxy
radical with 02. J. Photochem. 13: 149-163.
Crawford, A. R. ; Manny, E. H. ; Bartok, W. (1974) Field testing: application of
combustion modifications to control NO emissions from utility boilers.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA report
no. EPA-650/2-74-066. Available from: NTIS, Springfield, VA; PB-237344.
Crawford, A. R. ; et al. (1978) Control of utility boiler and gas turbine pollu-
tant emissions by combustion modification-phase I. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-600/7-78-036a.
Available from NTIS PB, Springfield, VA; 281-078. Cited in Hall and Bowen
(1982).
Cronn, D. R. (1982) Monitoring of ambient halocarbons and natural hydrocarbons
in the Smoky Mountains. In: Proceedings from 2nd symposium on composi-
tion of the nonurban troposphere; May; Williamsburg, VA.
Crutzen, P. J. (1982) The global distribution of hydroxyl. In: Goldberg, E.
D., ed. Atmospheric chemistry. New York, NY: Springer-Verlag; pp. 313-328.
Crutzen, P. J.; Ehhalt, D. H. (1977) Effects of nitrogen fertilizers and
combustion in the stratospheric ozone layer. Ambio. 6: 112-117.
Crutzen, P. J.; Fishman, J. (1977) Average concentrations of OH in the tropos-
phere, and the budgets of CH4, CO, H2, and CH3CC13. Geophys. Res. Lett.
4: 321-324.
Daniel sen, E. F. (1968) Stratospheric-tropospheric exchange based on radioac-
tivity, ozone and potential vorticity. J. Atmos. Sci. 25: 502-518.
Daniel sen, E. F. (1980) Stratospheric source for unexpectedly large values of
ozone measured over the Pacific Ocean during Gametag, August 1977. JGR
J. Geophys. Res. 85: 401-412.
Danielsen, E. F. ; Mohnen, V. A. (1977) Project duststorm report: ozone trans-
port, jjn situ measurements, and meteorological analyses of tropopause
folding. JGR J. Geophys. Res. 82: 5867-5877.
Darnall, K. R. ; Lloyd, A. C. ; Winer, A. M. ; Pitts, J. N., Jr. (1976) Reac-
tivity scale for atmospheric hydrocarbons based on reaction with hydroxyl
radical. Environ. Sci. Technol. 10: 692-696.
Darnall, K. R.; Atkinson, R.; Pitts, J. N., Jr. (1979) Observation of biacetyl
from the reaction of OH radicals with o-xylene. Evidence for ring clea-
vage. J. Phys. Chem. 83: 1943-1946.
3-151
-------
Dasgupta, P. K. (1980a) Discussion on "The importance of atmospheric ozone and
hydrogen peroxide in oxidizing sulphur dioxide in cloud and rainwater [on
Penkett et a!., 1979]." Atmos. Environ. 14: 272-275.
Dasgupta, P. K. (1980b) The importance of atmospheric ozone and hydrogen per-
oxide in oxidizing sulfur dioxide in cloud and rainwater—further discussion
Con Penkett'et al.5 1979]. Atmos. Environ. 14: 620-621.
Davis, D. D.; Heaps, W.; Philen, D.; McGee, T. (1979) Boundary layer measure-
ments of the OH radical in the vicinity of an isolated power plant plume:
S02 and N02 chemical conversion times. Atmos. Environ. 13: 1197-1203.
Davis, D. R.; Jensen, R. E. (1976) Low level ozone and weather systems. In:
Specialty conference on ozone/oxidants—interactions with the total environ-
ment; March; Dallas, TX. Pittsburgh, PA: Air Pollution Control Association;
pp. 242-251. -
Decker, C. E.; Ripperton, L. A.; Worth, J. J. B. ; Vukovich, F. M.; Bach, W.
D.; Tommerdahl, J. B.; Smith, F.; Wagoner, D. E. (1978) Study of the forma-
tion and transport of ambient oxidants in the western Gulf Coast and north-
central and northeast regions of the United States. Research Triangle
Park, NC: Research Triangle Institute.
Delwiche, C. C. (1970) The nitrogen cycle. Sci. Amer. 223: 137-146.
Dement, W. A.; Tyson, B. J.; Mooney, H. A. (1975) Mechanism of monoterpene
volatilization in Sal via mellifera. Phytochemistry 14: 2555-2557.
Demerjian, K. L.; Schere, K. L. (1979) Applications of a photochemical box
model for 03 air quality in Houston, Texas, Proceedings of Air Pollution
Control Association Specialty Conference on Ozone/Oxidants: Interactions
with the Total Environment, pp. 329-352.
Demerjian, K. L.; Kerr, J. A.; Calvert, J. G. (1974) The mechanism of photo-
chemical smog formation. In: Pitts, J. N., Jr.; Metcalf, R. L., eds.
Advances in environmental science and technology: v. 4. New York, NY:
John Wiley and Sons, Inc.; pp. 1-262.
Demerjian, K. L.; Schere, K. L.; Peterson, J. T. (1980) Theoretical estimates
of actinic (spherically integrated) flux and photolytic rate constants of
atmospheric species in the lower troposphere. In: Pitts, J. N., Jr.;
Metcalf, R. L.; Grosjean, D., eds. Advances in environmental science and
technology: volume 10. New York, NY: John Wiley and Sons; pp. 369-459.
Derwent, R. G.; Hov, 0. (1980) Computer modeling studies of the impact of
vehicle exhaust emission controls on photochemical air pollution forma-
tion in the United Kingdom. Environ. Sci. Techno!. 14: 1360-1366.
Derwent, R. G.; Eggleton, A. E. J.; Williams, M. L.; Bell, C. A. (1978) Elevated
ozone levels from natural sources. Atmos. Environ. 12: 2173-2177.
Dietzman, H. E.; Parness, H. A.; Bradow, R. L. (1980) Emissions from trucks by
chassis version 1983 transient procedures. Presented at Society of Auto-
motive Engineer's Meeting. SAE Paper No. 801371.
3-152
-------
Dietzman, H. E.; Parness, M. A.; Bradow, R. L. (1981) Emissions from gasoline
and diesel delivery trucks by chassis transient cycle. ASME Energy Sources
Technology Conference and Exhibition. January; Houston, TX.
Dignon, J.; Hameed, S. (1985) A model investigation of the impact of increases
in anthropogenic NO emissions between 1967 and 1980 on tropospheric ozone.
J. Atmos. Chem. 3: 491-506.
Dimitriades, B, (1970) On the function of hydrocarbon and nitrogen oxides in
photochemical smog formation. Washington, DG: U. S. Department of Commerce,
Bureau of Mines. Report of Investigations, RI 7433.
Dimitriadesj B. (1972) Effects of hydrocarbon and nitrogen oxides on photo-
chemical smog formation. Environ. Set. Tech, 6: 253-260.
Dimitriades, B. (1974) Proceedings of the solvent reactivity conference. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-650/3-74-010. Available from: NTIS, Springfield, VA; PB-238296.
Dimitriades, B. (1977a) An alternative to the Appendix-J method for calculating
oxidant- and N02-related control requirements.. In: International Confer-
ence on Photochemical Oxidant Pollution and Its Control. Proceedings, Vol.
II. Research Triangle Park, NC: U. S. Environmental Protection Agency.
EPA-600/3-77-001b. .-.-....
Dimitriades, B. (1977b) Oxidant control strategies. Part I. Urban oxidant-
control strategy derived from existing smog chamber data. Environ. Sci.
Technol. 11: 80-88.
Dimitriades, B. (1981) The role of natural organics in photochemical air pol-
lution: issues and research needs. J. Air Pollut. Control Assoc. 31:
229-235. ..-•-.-
Dimitriades, B.; Dodge, M. C. s eds, (1983) Proceedings of the Empirical Kinetic
Modeling Approach (EKMA) validation workshop. Research Triangle Park, NC:
U.S. Environmental Protection Agency. Report no. EPA-600/9-83-014.
Dodge, M. C. (1977a) Combined use of modeling techniques and smog chamber data
to derive ozone-precursor relationships. In: International Conference on
Photochemical Oxidant Pollution and Its Control. Proceedings, Vol. II.
Research Triangle Park, NC: U. S> Enviornmental Protection Agency.
EPA-600/3-77-001b.
Dodge, M. C. (1977b) Effect of selected parameters on predictions of a photo-
chemical model. Research Triangle Park, NC: U. S. Environmental Protection
Agency. EPA 600/3-77-048.
Dodge, M. C.; Arnts, R. R. (1979) A new mechanism for the reaction of ozone
with olefins. Int. J. Chem.. Kinet. 11: 399-410.
Doyle, G. J.; Tuazon, E. C.; Graham, R, A.; Mischke, T. M.; Winer, A. M.; Pitts,
J. N., Jr. (1979) Simultaneous concentrations of ammonia and nitric acid
in a polluted atmosphere and their equilibrium relationship to particulate
ammonium nitrate. Environ. Sci. Technol. 13: 1416-1419.
3-153
-------
Drives, P. J. (1977) TRACE (Trajectory atmospheric chemistry and emissions)
user's guide. Santa Monica, CA: Pacific Environmental Services, Inc. PES
Public. No. TP-016.
Duce, R. A.; Mohnen, V. A.; Zimmerman, P. R.; Grosjean, D.; Cautreels, W.;
Chatfield, R.; Jaenicke, R.; Ogren, J. A.; Pellizzari, E, D.; Wallace, G.
T. (1983) Organic material in the global troposphere. Rev. Geophys. Space
Phys. 21: 921-952.
Dunker, A. M. (1980) The response of an atmospheric reaction-transport model
to changes in input function. Atmos. Environ. 14: 671-679.
Dunker, A. M. (1981) Efficient calculation of sensitivity coefficients for
complex atmospheric models. Atmos. Environ. 15: 1155-1161.
Durbin, P. A.; Hecht, T. A.; Whitten, G. I. (1975) Mathematical modeling of
simulated photochemical smog. Research Triangle Park, NC: U. S. Environ-
mental Protection Agency; EPA report no. EPA-650/4-75-026.
Dutkiewicz, V. A.; Husain, L. (1979) Determination of stratospheric ozone at
ground level using 'Be/Ozone ratios. Geophys. Res. Lett. 6: 171-174.
Eaton, W. C.; Decker, C. E.; TommerdahlI, J. B.; Dimmock, F. E. (1979) Study of
the nature of ozone, oxides of nitrogen, and nonmethane hydrocarbons in
Tulsa, Oklahoma; v. 1: project description and data summaries. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards; EPA report no. EPA-450/4-79-008a. Avail-
able from: NTIS, Springfield, VA; PB-300481.
Erickson, R. E.; Yates, L. M.; Clark, R. L.; McEwen, D. (1977) The reaction of
sulfur dioxide with ozone in water and its possible atmospheric signifi-
cance. Atmos. Environ. 11: 813-817.
Evans, R. C.; Tingey, D. T.; Gumpertz, M. L.; Burns, W. F. (1982) Estimates of
isoprene and monoterpene emission rates in plants. Bot. Gaz. (Chicago)
143: 304-310.
F.R. (1971, August 14) 36: 15486-15506.
F.R. (1979, November 14) 44: 65667-65670.
F.R. (1981, January 22) 46: 7193-7194.
Fabian, P.; Pruchniewicz, P. G. (1977) Meridional distribution of ozone in the
troposphere and its seasonal variations. J, Geophys. Res. 82: 2063-2073.
Falls, A. H.; McRae, G, J.; Seinfeld, J. H. (1979) Sensitivity and uncertainty
of reaction mechanisms for photochemical air pollution. Inter. J. Chem.
Kinet. 11: 1137-1162.
Farber, R. J.; Huang, A. A.; Bregman, L. D.; Mahoney, R. L.; Eatough, D. J.;
Hansen, L. D.; Blumenthal, D. L.; Keiffer, W. S. ; Allard, D. W. (1982)
The third dimension in the Los Angeles basin. Sci. Total Environ.
23: 345-360.
3-154
-------
Federal Highway Administration (1978) Traffic volume trends, tables 5A, 5B,
and 9A. November-December 1978, computer printout; available from
Washington, DC: Federal Highway Administration.
Ferman, M. Z. A. (1981) Rural nonmethane hydrocarbon concentrations and composi-
tion. In: Bufalini, J. J.; Arnts, R. R., eds. Atmospheric biogenic hydro-
carbons. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.
Ferman, M. Z. A.; Monson, P. R. (1978) Comparison of rural and urban air quality.
Paper presented at 71st annual meeting of the Air Pollution Control Asso-
ciation; Houston, June.
Finlayson, B. J.; Pitts, J. N., Jr. (1976) Photochemistry of the polluted
troposphere. Science 192: 111-119.
Finlayson-Pitts, B. J.; Pitts, J. N., Jr. (1982) An assessment of the atmospheric
chemistry of oxides of sulfur and nitrogen: acid deposition and ozone.
Final Report to Office of Technology Assessment, Congress of the United
States. Available from NTIS, Springfield, VA.
Finlayson, B. J.; Pitts, J. N., Jr.; Atkinson, R. (1976) Low pressure gas phase
ozone-olefin reactions: chemiluminescence, kinetics, and mechanisms. J.
Am. Chem. Soc. 96: 5356-5367.
Fisher, T. M. (1980) The evolution of automotive emission controls. Presented
at: 73rd annual meeting of the Air Pollution Control Association; June;
Montreal, Canada. Pittsburgh, PA: Air Pollution Control Association; paper
no. 80-1.1.
Fishman, J. (1984) Ozone in the troposphere. In: Stratospheric ozone. Whitten,
R. C.; Prasad, S., eds. New York; Van Nostrand Reinhold.
\
Fishman, J.; Carney, T. A. (1984) A one-dimensional photochemical model of the
troposphere with planetary boundary-layer parameterization. J. Atmos. Chem.
1: 351-376.
Fishman, J.; Crutzen, P. J. (1978) The origin of ozone in the troposphere.
Nature (London) 274: 855-858.
Fishman, J.; Seiler, W. (1983) Correlative nature of ozone and carbon monoxide
in the troposphere: Implications for the tropospheric ozone budget. J.
Geophys. Res. 88: 3662-3670.
Fishman, J.; Solomon, S.; Crutzen, P. J. (1980) Observational and theoretical
evidence in support of a significant jji situ photochemical source of tropo-
spheric ozone. Tell us 31: 432-446.
Fishman, J. ; Vukovich, F. M. ; Browell, E. V. (1985) The photochemistry of
synoptic-scale ozone synthesis: Implications for the global tropospheric
ozone budget. J. Atmos. Chem. 3:299-320.
3-155
-------
Flyckt, D. L.; Westberg, H. H.; Holdren, M. W. (1980) Natural organic emissions
and their impact on air quality. Presented at: 73rd annual meeting of
the Air Pollution Control Association; June; Montreal, Quebec. Pittsburgh,
PA: Air Pollution Control Association; Paper 80-69.2.
Galloway, J. N.; Likens, G. E.; Keene, W. C.; Miller, J. M. (1982) The composition
of precipitation in remote areas of the world. JGR J. Geophys. Res.
82: 8771-8786.
Gelinas, R. J.; Vajk, J. P. (1979) Systematic sensitivity analysis of air
quality simulation models. Research Triangle Park, NC: U.S. Environmental
Protection Agency. EPA-6QQ/4-49-035.
Gerstle, R. W.; Peterson, R. F. (1966) Atmospheric emissions from nitric acid
manufacturing processes. Cincinnati, OH: U.S. Department of Health,
Education, and Welfare, National Center for Air Pollution Control; Public
Health Service publication no. 999-AP-27. Available from: NTIS, Spring-
field, VA; PB-190239.
Gertler, A. W.; Miller, D. F.; Lamb, D.; Katz, U. (1984) Studies of sulfur dioxide
and nitrogen dioxide reactions in haze and cloud. In: Durham, J. L., ed.
Chemistry of particles, fogs and rain. Boston, MA: Butterworth Publishers;
pp. 131-160. (Teasley, J. I., ed. Acid precipitation series, v. 2).
Gibbs, R. E.; Hyde, J. D.; Whitby, R. A.; Choudhury, D. R. (1983) Characterization
of emissions and fuel economy of in-use diesel automobiles. Research Triangle
Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-83-087. Available from: NTIS, Springfield, VA; PB83-262071.
Gillani, N. V.; Kohli, S.; Wilson, W. E. (1981) Gas-to-particle conversion of
sulfur in power plant plumes—I. Parameterization of the conversion rate
for dry, moderately polluted ambient conditions. Atmos. Environ. 15:
2293-2313.
Gipson, G. L. (1984) Guidelines for using carbon-bond mechanisms in city-
specific EKMA. Research Triangle Park, NC: U.S. Environmental Protection
Agency. Report no. EPA-450/4-84-005.
Glasson, W. A.; Tuesday, C. S. (1970) Inhibition of atmospheric photooxidation
of hydrocarbons by nitric oxide. Environ. Sci. Techno!. 4: 37-44.
Graedel, T. E. (1978) Chemical compounds in the atmosphere. New York, NY:
Academic Press.
Graedel, T. E.; Farrow, L. A.; Weber, T. A. (1976) Kinetic studies of the
photochemistry of the urban troposphere. Atmos. Environ. 10: 1095-1116.
Graedel, T. E.; Weschler, C. J. (1981) Chemistry within aqueous atmospheric
aerosols and raindrops. Rev. Geophys. Space Phys. 19: 505-539.
Graham, R. A.; Winer, A. M.; Pitts, J. N., Jr. (1977) Temperature dependence
of the unimolecular decomposition of pernitric acid and its atmospheric
implications. Chem. Phys. Lett. 51: 215-220.
3-156
-------
Graham, R. A.; Johnston, H. S. (1978) The photochemistry of N03 and the kine-
tics of the N205-03 system. J. Chem. Phys. 82: 254-268.
Graham, R. A.; Winer, A. M.; Pitts, J. N., Jr. (1978) Pressure and temperature
dependence of the unimolecular decomposition of H02N02. J. Chem. Phys.
68: 4505-4510.
Grosjean, D. (1982) Formaldehyde and other carbonyls in Los Angeles ambient
air. Environ. Sci. Techno!. 16: 254-262.
Grosjean, D. (1984) Photooxidation of methyl sulfide, ethyl sulfide, and
methanethiol. Environ. Sci. Techno!. 18: 460-468.
Grosjean, D.; Lewis, R. (1982) Atmospheric photooxidation of methyl sulfide.
Geophys. Res. Lett. 9: 1203-1206.
Grosjean, D.; Fung, K. (1984) Hydrocarbons and carbonyls in Los Angeles Air.
J. Air Pollut. Control Assoc. 34: 537-543.
Haagenson, P. L.; Shapiro, M. A.; Middleton, P. (1981) A case study relating
high ground level ozone to enhanced photochemistry and isentropic trans-
port from the stratosphere. JGR J. Geophys. Res. 86: 5231-5237.
Halfpenny, E.; Robinson, P. L. (1952a) Pernitrous acid: the reaction between
hydrogen peroxide and nitrous acid, and the properties of an intermediate
product. J. Chem. Soc. 928-938.
Halfpenny, E.; Robinson, P. L. (1952b) The nitration and hydroxylation of
aromatic compounds by pernitrous acid. J. Chem. Soc. 939-946.
Hanna, S. R. (1973) A simple dispersion model for the analysis of chemically.
reactive pollutants. Atmos. Environ. 7: 803-817.
Hanst, P. L. .(1970) Infrared spectroscopy and infrared lasers in air pollution
research and monitoring. Appl. Spectrosc. 24: 161-174.
Hanst, P. L. ; Gay, B. W. (1977) Photochemical reactions among formaldehyde,
chlorine, nitrogen dioxide in air. Environ. Sci. Techno!. 11: 1105-1109.
Hanst, P. L. ; Wilson, W. E.; Patterson, R. K.; Gay, B. W., Jr.; Chaney, L. W.
(1975) A spectroscopic study of California smog. Research Triangle Park,
NC: U.S. Environmental Protection Agency, National Environmental Research
Center; EPA report no. EPA-650/4-75-006. Available from: NTIS, Spring-
field, VA; PB-241022.
Harding, L. B.; Goddard, W. A., III. (1978) Mechanisms of gas-phase and
liquid-phase ozonolysis. J. Am. Chem. Soc. 100: 7180-7188.
Harris, G. W.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr.; Platt, U.;
Perner, D. (1982) Observations of nitrous acid in the Los Angeles atmos-
phere and implications for predictions of ozone-precursor relationships.
Environ. Sci. Techno!. 16: 414-419.
3-157
-------
Harrison, R. M.; Laxen, D. P. H. (1978) Sink processes for tetralkyllead com-
pounds in the atmosphere. Environ. Sci. Techno!. 12: 1384-1392.
Hatakeyama, S.; Okuda, M.; Akimoto, H. (1982) Formation of sulfur dioxide
and methanesulfonic acid in the photooxidation of dimethyl sulfide in the
air. Geophys. Res. Lett. 9: 583-586.
Hatakeyama, S.; Akimoto, H. (1983) Reactions of OH radicals with mathanethiol,
dimethyl sulfide, and dimethyl disulfide in air. J. Phys. Chem. 87:
2387-2395.
Hayes, S. R.; Hogo, H. (1982) An uncertainty analysis of EKMA-derived control
requirements for ozone abatement in the South Coast Air Basin. Presented
at 75th annual meeting of the Air Pollution Control Association, New
Orleans, LA, June 20-25.
Hegg, D. A.; Hobbs, P. V. (1981) Cloud water chemistry and the production of
sulfates in clouds. Atmos. Environ. 15: 1597-1604.
Heikes, B. G.; Lazrus, A. L.; Kok, G. L.; Kunen, S. M.; Gandrud, B. W.;
Gitlin, S. N.; Sperry, P. D. (1982) Evidence for aqueous phase hydrogen
peroxide synthesis in the troposphere. JGR J. Geophys. Res. 87:
3045-3051.
Heikes, B. G.; Thompson, A. M. (1983) Effects of heterogeneous processes on
N03, HONO, and HN03 chemistry in the troposphere. J. Geophys. Res. 88:
10883-10895.
Hendry, D. G.; Baldwin, A. C.; Barker, J. R.; Golden, D. M. (1978) Computer
modeling of simulated photochemical smog. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Research and Development;
EPA report no. EPA-600/3-78-059. Available from: NTIS, Springfield, VA:
PB-284025.
Herron, J. T.; Huie, R. E. (1977) Stopped-flow studies of the mechanism of
ozone-alkene reactions in the gas phase. Ethylene. J. Am. Chem. Soc. 99:
5430-5435.
Herron, J. T.; Huie, R. E. (1978) Stopped flow studies of mechanisms of
ozone-alkene reactions in gas phase propene + isobutene. Int. J. Chem.
Kinet. 10: 1019-1040.
Herron, 0. T.; Martinez, R. I.; Huie, R. E. (1982) Kinetics and energetics of
the Criegee intermediate in the gas phase. I. The Criegee intermediate in
ozone-alkene reactions. Int. J. Chem. Kinet. 14: 201-224.
Hill, A. C. (1971) Vegetation: a sink for atmospheric pollutants. J. Air
Pollut. Control Assoc. 21: 341-346.
Hilliard, J. C.; Wheeler, R. W. (1979) Nitrogen dioxide in engine exhaust.
Presented at passenger car meeting; June; Dearborn, MI. Warrendale, PA:
Society of Automotive Engineers; SAE paper no. 790691.
3-158
-------
Hoffman, M. R.; Edwards, J. 0. (1975) Kinetics of the oxidation of sulfite by
hydrogen peroxide in acidic solution. J. Phys. Chem. 79; 2096-2098.
Hoffman, M. R.; Jacob, D. J. (1984) Kinetics and mechanisms of the catalytic
oxidation of dissolved sulfur dioxide in aqueous solution: an application
to nighttime fog water chemistry. In: Calvert, J. G., ed. S02, NO and N02
oxidation mechanisms: atmospheric considerations. Boston, MA: Butterworth
Publishers; pp. 101-172. (Teasley, J. I., ed. Acid precipitation series,
v. 3).
Holdren, M. W.; Westberg, H. H.; Zimmerman, P. R. (1979) Analysis of monoter-
pene hydrocarbons in rural atmospheres. JGR J. Geophys. Res. 84: 5083-5088.
Holdren, M. W.; Spicer, C. W.; Hales, J. M. (1984) Peroxyacetyl nitrate solu-
bility and decomposition rate in acidic water. Atmos. Environ. 18:
1171-1173.
Holdren, M.;'Spicer, C.; Sticksel, P.; Nepsund, K.; Ward, G.; Smith, R. (1982)
Implementation and analysis of hydrocarbon grab samples from Cleveland
and Cincinnati 1981. Ozone monitoring study. Research Triangle Park, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-905/4-82-001.
Available from: NTIS, Springfield, VA; PB83-197673.
Holzworth, G. C. (1964) Estimates of mean maximum mixing depths in the contig-
uous United States. Mon. Weather Rev. 92: 235-242.
Holzworth, G. C. (1972) Mixing heights, wind speeds, and potential for urban
air pollution throughout the contiguous United States. Research Triangle
Park, NC: U.S. Environmental Protection Agency; publication no. AP-101.
Available from: NTIS, Springfield, VA; PB-207103.
Holzworth, G. C.; Fisher, R. W. (1979) Climatological summaries of the lower
few kilometers of rawinsonde observations. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Environmental Sciences Research
Laboratory; EPA report no. EPA-600/4-79-026.
Hoshino, M.; Akimoto, H.; Okuda, M. (1978) Photochemical oxidation of benzene,
toluene, and ethyl benzene initiated by OH radicals in the gas phase.
Bull. Chein. Soc. Japan 51: 718-724.
Hosier, C. R. (1961) Low-level inversion frequency in the contiguous United
States. Mon. Weather Rev. 89: 319-339.
Howard, C. J. (1977) Kinetics of the reaction of H02 with N02. J. Chem. Phys.
67: 5258-5263.
Hull, L. A. (1981) Terpene ozonolysis products. In: Bufalim", J. J.; Arnts, R.
R., eds. Atmospheric biogenic hydrocarbons: v. II. Ann Arbor, MI: Ann
Arbor Science Publishers; p. 161.
Hunsaker, D. (1981) Selection of biogenic hydrocarbon emission factors for
land cover classes found in the San Francisco Bay Area. California
Association of Bay Area Governments; air quality technical memo 31.
3-159
-------
Hunter, S. C., et al. (1978) Evaluation of combustion modifications for emis-
sions reduction in industrial processes. ASME Paper No. 78-WA/APC-8.
Hunter, S. C.; Carter, W. A,; McElroy, M. W.; Cherry, S. S.; Buening, H. J.
(1977). Application of combustion modifications to industrial combustion
equipment. Research Triangle Park, NC: U. S. Environmental Protection
Agency; EPA report no. EPA-60Q/7-79~15a.
Husain, L.; Coffey, P. E.; Meyers, R. E.; Cederwall, R. T. (1977) Ozone
transport from stratosphere to troposphere. Geophys. Res. Lett. 4:
363-365.
Jaffee, R. J.; Smith, F. C., Jr.; West, K. W. (1974) Study of factors affecting
reactions in environmental chambers: final report on phase III. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
and Development; EPA report no. EPA-650/3-74-004b. Available from: NTIS,
Springfield, VA PB-256252.
Japar, S. M.; Niki, H. (1975) Gas-phase reactions of the nitrate radical with
olefins. J. Phys. Chem. 79: 1629-1632.
Jeffries, H. E,; Johnson, G. T. (1981) Predicting ozone frequency distri-
butions. In: Proceedings of EKMA Model Validation Workshop, December
15-17, Research Triangle Park, NC.
Jeffries, H.; Fox, D.; Kamens, R. (1975) Outdoor smog chamber studies. Effect
of hydrocarbon reduction on nitrogen dioxide. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-650/3-75-011.
Available from: NTIS, Springfield, VA; PB-245829.
Jeffries, H.; Fox, D.; Kamens, R. (1976) Outdoor smog chamber studies: light
effects relative to indoor chambers. Environ. Sci. Technol. 10:. 1006-1011.
Jeffries, H. E.; Sexton, K. G.; Salmi, C. N. (1981) Effects of chemistry and
meteorology on ozone calculations using simple trajectory models and the
EKMA procedure. Research Triangle Park, NC: U. S. Environmental Protection
Agency; EPA report no. EPA-450/4-81-034.
Jeffries, H. E.; Sexton, .K. G.; Kamens, R. M.; Holkman, M. S. (1985) Outdoor
smog chamber experiments to test photochemical models. Research Triangle
Park, NC: U. S. Environmental Protection Agency.
Johnson, W. B.; Viezee, W. (1981) Stratospheric ozone in the lower troposphere—
I. Presentation and interpretation of aircraft measurements. Atmos. Environ.
15: 1309-1323.
Joshi, S. B. (1979) Houston field study (1978): formaldehyde and total alde-
hydes monitoring program. EPA contract no. 68-02-2566. Research Triangle
Park, NC: Northrop Services, Inc.-Environmental Sciences; technical report
no. ESC-TR-79-22.
Joshi, S. B.; Adger, R. A.; Barnes, D. A.; Kaiserman, M. (1981) Aldehyde sampling
field study (1980). Research Triangle Park, NC: Northrop Services, Inc.-
Environmental Sciences; technical report no. TR-81-02.
3-160
-------
JSA, Inc. (1978) Draft environmental impact statement for the South Tahoe public
utilities district, waste water facilities planning program. Jones and
Stokes Associates, Inc.
Junge, C. E. (1963) Air chemistry and radioactivity. New York, NY: Academic
Press, (van Mieghen, J.; Hales, A. L., eds. International geophysics series
v. 4). . ,
Kacmarek, A. J. ; Solomon, I. J.; Lustig, M. (1978) Preparation and properties
of peroxyacetyl nitrate. J. Inorg. Nucl. Chem. 40: 574-576.
Kamens, R. M. ; Gery, M. W. ; Jeffries, H. E.; Jackson, M.; Cole, E. I. (1982)
Ozone-isoprene reactions: product formation and aerosol potential. Int.
J. Chem. Kinet. 14: 955-975.
Kan, C. S. ; Su, F.; Calvert, J. G.; Shaw, J. H. (1981) Mechanism of the ozone-
ethene reaction in dilute N2/02 mixtures. J. Phys. Chem. 85: 2359-2363.
Kelly, N. A. (1985) Ozone/precursor relationships in the Detroit metropolitan
area derived from captive-air irradiations and an empirical photochemical
model. J. Air Pollut. Control Assoc. 35: 27-34.
Kelly, N. A.; Wolff, G. T. ; Ferman, M. Z. A. (1982) Background pollutant measure-
ments in air masses affecting the eastern half of the United States--!. Air
masses arriving from the northwest. Atmos. Environ. 16: 1077-1088.
Kelly, N. A.; Wolff, G. T.; Ferman, M. A. (1984) Sources and sinks of ozone in
rural areas. Atmos. Environ. 18: 1251-1266.
Kelly, N. A.; Ferman, M. A.; Wolff, G. T. (1986) The chemical and meteorological
conditions associated with high and low ozone concentrations in south-
eastern Michigan and nearby areas of Ontario. J. Air Pollut. Control Assoc.
36: 150-158.
Kenley, R. A.; Davenport, J. E. ; Hendry, D. G. (1981) Gas-phase hydroxyl
radical reactions. Products and pathways for the reaction of OH with aro-
matic hydrocarbons. J. Phys. Chem. 85: 2740-2746.
Killus, J. P.; Whitten, G. Z. (1982) A mechanism describing the photochemical
oxidation of toluene in smog. Atmos. Environ. 16: 1973-1988.
Killus, J. P.; Whitten, G. Z. (1984) Isoprene: a photochemical kinetic
mechanism. Environ. Sci. Techno!. 18: 142-148.
Kim, C. M. (1973) Influence of vegetation type in the intensity of ammonia
and nitrogen dioxide liberation from soil. Biochem. 5: 163-166.
Kircher, C. C.; Margitan, J. J.; Sander, S. P. (1984) Pressure and temperature
dependence of the reaction N02 + N03 + M --> N205 + M. J. Phys. Chem. 88:
4370-4375.
Kirk^-Othmer encyclopedia of chemical technology. (1981) 3rd ed. New York, NY:
John Wiley & Sons, Inc., v. 16: 701.
3-161
-------
Knoerr, K. R.; Mowry, F. W. (1981) Energy balance/Bowen ratio technique for
estimating hydrocarbon fluxes. In: Bufalini, J. J.; Arnts, R. R., eds.
Atmospheric biogenic hydrocarbons, v. 1. Ann Arbor, MI: Ann Arbor Science.
Knoppel, H.; Versino, B.; Peil, A.; Schauenburg, H.; Vissers, H. (1982) Quanti-
tative determination of terpenes emitted by conifers. In: Versino, D.;
Ottj H., eds. Physicochemical behavior of atmospheric pollutants. Boston,
MA: D. Reidel Publishing Company; pp. 89-98.
Korshover, J. (1967) Climatology of stagnating anticyclones east of the Rocky
Mountains, 1936-1965. Cincinnati, Ohio: U.S. Department of Health,
Education and Welfare, National Center for Air Pollution Control; Public
Health Service publication no. 999-AP-34. Available from: NTIS, Springfield,
VA; PB-174709.
Korshover, J. (1975) Climatology of stagnating anticyclones east of the Rocky
Mountains, 1936-1975. Washington, DC: U.S. Department of Commerce, National
Oceanic and Atmospheric Administration; NOAA technical memorandum ERL AR-55.
Available from: NTIS, Springfield, VA; PB-257368.
Kunen, S. M.; Lazrus, A. L. ; Kok, G. L.; Heikes, B. G. (1983) Aqueous oxida-
tion of S02 by hydrogen peroxides. JGR J,. Geophys. Res. 88: 3671-3674.
Lamb, B. K.; Westberg, H. H.; Quarles, T.; Flyckt, D. L. (1983) Natural hydro-
carbon emission rate measurements from selected forest sites. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-84-001. Available from: NTIS, Springfield, VA; PB84-124981.
Lamb, R. G. (1977) A case study of stratospheric ozone affecting ground-level
oxidant concentrations. J. Appl. Meteorol. 16: 780-794.
Larsen, R. I. (1969) A new mathematical model of air pollutant concentration
averaging time and frequency. J. Air Pollut. Control Assoc. 19: 24-30.
Larson, T. V.; Horike, N. R.; Harrison, H. (1978) Oxidation of sulfur dioxide
by oxygen and ozone in aqueous solution: a kinetic study with significance
to atmospheric rate processes. Atmos. Environ. 12: 1597-1611.
Lee, Y.-N.; Senum, G. I.; Gaffney, J. S. (1983) Peroxyacetyl nitrate (PAN)
stability, solubility, and reactivity—implications for tropospheric nitrogen
cycles and precipitation chemistry. Presented at CACGP Symposium, Oxford,
England; September.
Leighton, P. A. (1961) Photochemistry of air pollution. New York, NY: Academic
Press, Inc.
Leihkanen, H. E.; Beckman, E. W. (1971) The effect of leaded and unleaded gaso-
lines on exhaust emissions as influenced by combustion chamber deposits.
Presented at Society of Automotive Engineers' national combined fuels and
lubricants, powerplant, and truck meetings; October; St. Louis, MO.
Warrendale, PA: Society of Automotive Engineers; SAE paper no. 71083.
3-162
-------
Leone, J. A.; Seinfeld, J. H. (1984) Updated chemical mechanism for atmos-
pheric photooxidation of toluene. Int. J, Chem. Kinet. 16: 159-193.
Leone, J. A.; Flagan, R. C.; Grosjean, D.; Seinfeld, J. H. (1985) An outdoor
smog chamber and modeling study of toluene-NO photooxidation. Int. J.
Chem. Kinet. 17: 177-216. x
Levy, H. II; Mahlman, J. D.; Moxim, W. J. (1985) Tropospheric ozone: The role
of transport. J. Geophys. Res. 90: 3753-3772.
Lindley, C. R. C.; Calvert, J. G.; Shaw, J. H. (1979) Rate studies of the
reactions of the (CH3)2N radical with 02, NO, and N02. Chem. Phys. Lett.
67: 57-62.
Linnell, R. H.; Scott, W. E. (1962) Diesel exhaust composition and odor
studies. J. Air Pollut. Control Assoc. 12: 510-515, 545.
Liu, C. S.; Grisinger, J. E. (1981) Review of SAI Airshed Model Sensitivity Study
Conducted for the South Coast Air Basin by California Air Resources Board.
AQMP Technical Paper No. 5. South Coast Air Quality Management District;
El Monte, CA.
Liu, M.-K.; Seinfeld, J. H. (1975) On the validity of grid and trajectory
models of urban air pollution. Atmos. Environ. 9: 555-574.
Liu, S. C.; Cicerone, R. J.; Donahue, T. M.; Chameides, W. L. (1977) Sources
and sinks of atmospheric N20 and the possible ozone reduction due to indus-
trial fixed nitrogen fertilizers. Tell us. 29: 251-263.
Liu, M.-K.; Morris, R. E.; Killus, J. P. (1984) Development of a regional oxidant
model and application to the northeastern United States. Atmos. Environ.
18: 1145-1161.
Lloyd, A. C. (1979) Tropospheric chemistry of aldehydes. In: Herron, J. T.;
Huie, R. E.; Hodgeson, J. A., eds. Chemical kinetic data needs for modeling
of the lower troposphere: proceedings of a workshop; May 1978; Reston,
VA. Gaithersburg, MD: U.S. Department of Commerce, National Measurement
Laboratory; Bureau of Standards special publication 557.
Lloyd, A. C.; Lurmann, F. W.; Godden, D. A.; Hutchins, J. F.; Eschenroeder,
A. Q. ; Nordsieck, R. A. (1979) Development of the ELSTAR photochemical
air quality simulation model and its evaluation relative to the LARPP data
base. ERT Doc. No. P-5287-500, Environmental Research and Technology,
Inc., Westlake Village, CA.
Lloyd, A. C.; Atkinson, R.; Lurmann, F. W.; Nitta, B. (1983) Modeling
potential ozone impacts from natural hydrocarbons. I. Development and
testing of a chemical mechanism for the NO -air photooxidations of isoprene
and orpinene under ambient conditions. Amos. Environ. 17: 1931-1950.
Logan, J. A. (1983) Nitrogen oxides in the troposphere: global and regional
budgets. JGR J. Geophys. Res. 88: 10785-10807.
3-163
-------
Logan, J.A.; Prather, M.J.; Wofsy, S.C.; McElroy, M.B. (1981) Tropospheric
chemistry: A global perspective. J. Geophys. Res. 86: 7210-7254.
Lonneman, W. A. (1977) Ozone and hydrocarbon measurements in recent oxidant
transport studies. In: Dimitriades, B,, ed. International conference on
photochemical oxidant pollution and its control; September 1976; Raleigh,
NC. Proceedings: volume I. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Sciences Research Laboratory; pp. 211-223;
EPA report no. EPA-600/3-77-001a. Available from: NTIS, Springfield, VA;
PB-264232.
Lonneman, W. A. (1979) Ambient hydrocarbon measurements in Houston. In: Ozone/
oxidants: interactions with the total environment II: proceedings of a
specialty conference; October; Houston, TX. Pittsburgh, PA: Air Pollu-
tion Control Association; pp. 278-287.
Los Angeles County, California. (1971) Profile of air pollution control. Los
Angeles, CA: Los Angeles Air Pollution Control District.
Ludwig^F. L.; Reiter, E.; Shelar, E.; Johnson, W. B. (1977) The relation of
oxidant levels to precursor emissions and meteorological features. Volume
I: analysis and findings. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Air Quality and Standards; EPA final report
no. EPA^450/3-77-022a. Available from: Springfield, VA; PB-275001.
Lurmann, F.; Godden, D.; Lloyd, A. C.; Nordsieck, R. A. (1979) A Lagrangian
photochemical air quality simulation model. Adaptation to the St. Louis-RAPS
Data Base. Vol. I. Model Formulation. Vol. II. User's Manual. Research
Triangle Park, NC: U. S. Environmental Protection Agency; EPA report no.
EPA-600/8-79-015a,b.
Lynn, D. A.; Steigerwald, B. J.; Ludwig, J, H. (1964) The November-December
1962 air pollution episode in the eastern United States. Cincinnati, OH:
U.S. Department of Health, Education and Welfare, Division of Air Pollu-
tion; Public Health Service publication no. 999-AP-77. Available from:
NTIS, Springfield, VA; PB-168878.
Lyons, W. A.; Olsson, L. E. (1972) Mesoscale air pollution transport in the
Chicago lake breeze. J. Air Pollut. Control Assoc. 22: 876-881.
MacCracken, M. C.; Wuebbles, D. J.; Walton, J. J.; Duewer, W. H.; Grant, K. E.
(1978) The Livermore regional air quality model: I. Concept and development.
J. Appl. Meteorol. 17: 254-272.
Magnotta, F.; Johnston, H. S. (1980) Photodssociation quantum yields for
the N03 free radical. Geophys. Res. Lett. 7: 769-772.
Makdo, M. W.; Troe, J. (1982) Analysis of unimolecular reaction of N205 + M
j N02 + N03 + M. Int. J. Chem. Kinet. 14: 399-416.
Marakov, B. N. (1969) [Liberation of nitrogen dioxide from soil]. Pochovove-
denie. 1: 49-58.
3-164
-------
Marchesani, V. J.; Towers, T.; Wohlers, H. C. (1970) Minor sources of air
pollutant emissions. J. Air Pollut. Control Assoc. 20: 19-22.
Martin, L. R. (1984) Kinetic studies of sulfite oxidation in aqueous solution.
In: Calvert, J. G., ed. S02s NO and NOg oxidation mechanisms: atmospheric
considerations. Boston, MA: Butterworth Publishers; pp. 63-100. (Teasley,
J. I., ed., Acid precipitation series, v. 3).
Martin, L. R.; Damschen, D. E. (1981) Aqueous oxidation of sulfur dioxide by
hydrogen peroxide at low pH. Atmos. Environ. 15: 1615-1621.
Martinez, J. R.; Singh, H. B. (1979) Survey of the role of NO in nonurban
ozone formation. SRI International, Menlo Park, CA. Final report to
U.S. Environmental Protection Agency, SRI Project 6780-8. Research Triangle
Park, NC: U.S. Environmental Protection Agency.
Martinez, R. I.; Herron, J. T.; Huie, R. E. (1981) The mechanism of ozone-
alkene reactions in the gas phase. A mass spectrometric study of the
reactions of eight linear and branched-chain alkenes. J. Am, Chem. Soc.
103: 3807-3820.
Maxwell, C.; Martinez, J. R. (1982) Performance evaluation of three chemical
models used with the Empirical Kinetic Modeling Approach (EKMA). Presented
at: 75th annual meeting of the Air Pollution Control Association. New
Orleans, LA, June 20-25.
McElroy, C. T.; Kerr, J. B. (1977) A limb scan measurement of atmospheric nitro-
gen dioxide. Presented at: fall meeting of the American Geophysical Union;
San Francisco, CA.
McKee, H. C.; Clark, J. M.; Wheeler, R. J. (1962) Measurement of diesel engine
emissions. J. Air Pollut. Control Assoc. 12: 516-521, 546.
Meyer, E. L.; Gipson, G. L.; Dimitriades, B. (1981) Evaluation of the empirical
kinetics modeling approach. Paper 81-6.4, Annual Meeting of the Air
Pollution Control Association, Philadelphia, Pennsylvania.
Meyers, T. C., et al. (1979) User's Guide to the SAI Trajectory Model. SAI
Report No. EP-79-3, Systems Application, Inc., San Rafael, CA.
Mohnen, V. A. (1977) Review and analysis. In: International conference on
oxidants,, 1976—Analysis of evidence and viewpoints. Part III. The issue
of stratospheric ozone intrusion. Research Triangle Park, NC: U.S.
Environmental Protection Agency. Report no. EPA-600/3-77-115.
Morris, E. D., Jr., Niki, H. (1974) Reaction of the nitrate radical with
acetaldehyde and propylene. J. Phys. Chem. 78: 1337-1338.
Motor Vehicle Manufacturers Association. (1984) MVMA Motor Vehicle Facts and
Figures (annual summary issues). Detroit, MI: Motor Vehicle Manufacturers
Association of the United States, Inc.
Mudd, J. B. (1966) Reaction of peroxyacetyl nitrate with glutathione. J. Biol.
Chem. 241: 4077-4080.
3-165
-------
Mueller, P. K.; Hidy, G. M. (1983) The sulfate regional experiment (SURE):
report of findings. Palo Alto, CA: Electric Power Research Institute;
report no. EA-1901.
National Research Council. (1977a) Ozone and other photochemical oxidants.
Washington, DC: National Academy of Sciences.
National Research Council. (1977b) Nitrogen oxides. Washington, DC: National
Academy of Sciences.
Naylor, M. H.; Lee, D. C.; Smith, M. J. (1981) Southeast Las Vegas valley air
quality measurement project 1980-81. Final report. Las Vegas, Nevada:
Clark County Health District.
Neiburger, M.; Johnson, D. S.; Chen, C. -W. (1961) Studies of the structure of
the atmosphere over the eastern Pacific Ocean in the summer: I. the inver-
sion over the eastern north Pacific Ocean. Berkeley, CA: University of
California Press; publications in meteorology, v. 1, no. 1.
Nicksic, S. W.; Harkins, J.; Mueller, P. K. (1967) Some analyses for PAN and
studies of its structure. Atmos. Environ. 1: 11-18.
Niki, H.; Daby, E. E.; Weinstock, B. (1972) Mechanisms of smog reactions. Adv.
• Chem. Ser. 113: 16-57.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1977) Fourier
transform IR spectroscopic observation of pernitric acid formed via HOO +
N02 -*• HOON02. Chem. Phys. Lett. 45: 564-566.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1978) Mechanism for
hydroxyl radical initiated oxidation of olefin-nitric oxide mixtures in
parts per mi 11 on concentrations. J. Phys. Chem. 82: 135-137.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1979) Fourier
transform infrared (FTIR) studies of gaseous and particulate nitrogenous
compounds. In: Grosjean, D., ed., Nitrogenous Air Pollutants. Ann Arbor,
MI: Ann Arbor Science Publishers, Inc.; pp. 1-16.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1981) An FT-IR study
of a transitory product in the gas-phase ozone-ethylene reaction. J. Phys.
Chem. 85: 1024-1027.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1983) An FT-IR
study of the mechanism for the reaction HO+ CH3SCH3. Int. J. Chem. Kinet.
15: 647-654.
Novak, J. (1986) Atmospheric Sciences Research Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, NC. Personal communication to
B. E. Tilton, Environmental Criteria and Assessment Office, U.S. Environ-
mental Protection Agency, Research Triangle Park, NC.
Noxon, J. F. (1976) Atmospheric nitrogen fixation by lightning. Geophys.
Res. Lett. 3: 463-465.
3-166
-------
Noxon, J, F. ; Norton, R. B.; Marovich," E. (1980) N03 in the troposphere.
Geophys. Res. Lett. 7: 125-128.
NPN: National Petroleum News (annual fact books). Des Plaines, IL: National
Petroleum News.
Oltmans, S. J. (1981) Surface measurements in clean air. J. Geophys. Res.
86: 1174-1180.
Parrish, D. D.; Fahey, D. W.; Williams, E. J.; Liu, S. C.; Trainer, M.; Murphy,
P. C.; Albritton, D. L.; Fehsenfeld, F. C. (1986) Background ozone and
anthropogenic ozone enhancement at Niwot Ridge, Colorado. J. Atmos. Chem.
4: 63-80.
PEDCo Environmental, Inc. (1978) Assessment of gasoline toxicity. Prepared for
U.S. Environmental Protection Agency under contract no. 68-02-2515.
Research Triangle Park, NC: Office of Air Quality Planning and Standards.
Penkett, S. A. (1983) Atmospheric chemistry: PAN in the natural atmosphere.
Nature (London) 302: 293-294.
Penkett, S. A.; Brice, K. A. (1986) The spring maximum in photooxidants in the
Northern Hemisphere troposphere. Nature 319: 655-657.
Penkett, S. A.; Jones, B. M. R.; Brice, K. A.; Eggleton, A. E. J. (1979) The
importance of atmospheric ozone and hydrogen peroxide in oxidising sulphur
dioxide i.n cloud and rainwater. Atmos. Environ. 13: 123-137.
Perner, D.; Schmeltekopf, A.; Winkler, R. H.; Johnston, H. S.; Calvert, J. G.;
Cantrell, C. A.; Stockwell, W. R. (1985) A laboratory and field study 'of
the equilibrium N205 5 N03 + N02. J. Geophys. Res. 90: 3807-3812.
Peterson, J. T. (1976) Calculated actinic fluxes (290-700 nm) for air pollution
photochemistry applications. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Sciences Research Laboratory; EPA
report no. EPA-6QO/4-76-025. Available from: NTIS, Springfield, VA;
PB-240125.
Pitts, J. N. , Jr.; Winer, A. M.; Darnall, K. R.; Lloyd, A. C.; Doyle,, G. J.
(1977) Hydrocarbon reactivity and the role of hydrocarbons, oxides of
nitrogen, and aged smog in the production of photochemical oxidants. In:
Dimitriades, B., ed. International conference on photochemical oxidant
pollution and its control; September 1976; Raleigh, NC. Proceedings: v.
II. Research Triangle Park, NC: U.S. Environmental Protection Agency;
EPA-600/3-77-001b.
Pitts, J. N. , Jr.; Grosjean, D. ; Van Cauwenberghe, K.; Schmid, J. P.; Fitz,
D. R. (1978) Photooxidation of aliphatic amines under simulated atmospheric
conditions: formation of nitrosamines, nitramines, amides, and photochemical
oxidant. Environ. Sci. Techno!. 12: 946-953.
3-167
-------
Pitts, J. N., Jr.; Darnall, K. R.; Carter, W. P. L.; Winer, A. M.; Atkinson,
R. (1979) Mechanisms of photochemical reactions in urban air. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-79-110. Available from: NTIS, Springfield, VA; PB80-132384.
Pitts, J. N., Jr.; Winer, A. M.; Harris, G. W.; Carter, W. P. L.; Tuazon, E.
C. (1983) Trace nitrogenous species in urban atmospheres. EHP Environ.
Health Perspect. 52: 153-157.
Pitts, J. N., Jr.; Sanhueza, E.; Atkinson, R.; Carter, W. P. L.; Winer, A.
M.; Harris, G. W.; Plum, C. N. (1984a) An investigation of the dark forma-
tion of nitrous acid in environmental chambers. Int. J. Chem. Kinet. 16:
919-939.
Pitts, J. N., Jr.; Biermann, H. W.; Winer, A. M.; Tuazon, E. C. (1984b)
Spectroscopic identification and measurement of gaseous nitrous acid in
dilute auto exhaust. Atmos. Environ. 18: 847-854.
Pitts, J. N., Jr.; Biermann, H. W.; Atkinson, R.; Winer, A. M. (1984c)
Atmospheric implications of simultaneous nighttime measurements of NOs
radicals and HONO. Geophys. Res. Lett. 11: 557-560.
Pitts, J. N., Jr.; Winer, A. M.; Aschmann, S. M.; Carter, W. P. L.; Atkinson,
R. (1985) Experimental protocol for determining hydroxyl radical reaction
rate constants. Estimation of atmospheric reactivity. Research Triangle
Park: U.S. Environmental Protection Agency; EPA report no. EPA-600/3-85-000.
Platt, U.; Perner, D.; Harris, G. W.; Winer, A. M.; Pitts, J. N., Jr. (1980a)
Observations of nitrous acid in an urban atmosphere by differential optical
absorption. Nature 285: 312-314.
Platt, U.; Perner, D.; Winer, A. M.; Harris, G. W.; Pitts, J. N., Jr. (1980b)
Detection of NOa in the polluted troposphere by differential optical ab-
sorption. Geophys. Res. Lett. 7: 89-92.
Platt, U.; Winer, A. M.; Biermann, H. W.; Atkinson, R.; Pitts, J. N., Jr.
(1984) Measurement of nitrate radical concentrations in continental air.
Environ. Sci. Techno!. 18: 365-369.
Plegue, J. (1983) Diesel-car sales plunge 48 pet. Automotive News (May 2): 1,
70.
Price, J. H. (1976) A study of factors associated with high urban ozone
concentrations in Texas. In: Specialty conference on ozone/oxidants—
interactions with the total environment. Pittsburgh, PA: Air Pollution
Control Association; pp. 282-292.
Prior, E. J.; Schiess, J. R.; McDougal, D. S. (1981) Approach to forecasting
daily maximum ozone levels in St. Louis. Environ. Sci. Technol. 15: 430-436.
Rao, T. K.; Young, J. A.; Lijinsky, W.; Epler, J. L. (1979) Mutagenicity of
aliphatic nitrosamines in Salmonella typhimurium. Mutat. Res. 66: 1-7.
3-168
-------
Rasmussen, R. A. (1964) Terpenes: their analysis and fate in the atmosphere.
St. Louis, MO: Washington University. Ph.D. thesis.
Rasmussen, R. A. (1970) Isoprene: identified as a forest-type emission to the
atmosphere. Environ. Sci. Technol. 4: 667-671.
Rasmussen, R. A. (1972) What do the hydrocarbons from trees contribute to air
pollution? J. Air Pollut. Control Assoc. 22: 537-543.
Rasmussen, R. A.; Chatfield, R. B.; Holdren, M. W.; Robinson, E. (1976) Hydro-
carbon levels in a midwest open-forested area. Technical report to Coor-
dinating Research Council, Inc., Atlanta, GA.
Reiter, E. R. (1963) A case study of radioactive fallout. J. Appl. Meteorol.
2: 691-705.
Reiter, E. R. (1975) Stratospheric-tropospheric exchange processes. Rev. Geophys.
Space Phys. 13: 459-473.
Reiter, E. R. (1978) Atmospheric transport processes - 4. Radioactive tracers.
Available from: National Technical Information Service, U.S. Department of
Commerce, Springfield, VA. Report no. TID-27114. (see also, Reiter, E. R.
(1977) The transport of radioactive debris and ozone from the stratosphere
to the ground. Arch. Met. Geophys. Biokl. 26(ser A): 323-340).
Reiter, E. R.; Mahlman, J. D. (1965) Heavy radioactive fallout over the southern
United States, November 1962. JGR J. Geophys. Res. 70: 4501-4519.
, j ,
Reynolds, S. D. et al. (1979) Photochemical modeling of transportation control
strategies. v.I. Model development, performance evaluation, and strategy
assessment. Prepared for the Federal Highway Administration, Office of
Research, Washington, DC.
Reynolds, S. D.; Hogo, J.; Oliver, W. R.; Reid, L, E. (1982) Application of
the SAI Airshed Model to the Tulsa metropolitan area. SAI No. 82004. Systems
Applications, Inc., San Rafael, CA.
Richards, L. W.; Anderson, J. A.; Blumenthal, D. L.; McDonald, J. A.; Kok, G.
L.; Lazrus, A. L. (1983) Hydrogen peroxide and sulfur (IV) in Los Angeles
cloud water. Atmos. Environ. 17: 911-914.
Richter, H. G. (1983) Analysis of organic compound data gathered during 1980
in northeast corridor cities. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards;
EPA report no. EPA-450/4-83-017. Available from: NTIS, Springfield, VA;
PB84-116052.
Richter, H. G. (1986) Monitoring and Data Analysis Division, Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC. Personal communication to B. E. Tilton,
Environmental Criteria and Assessment Office, U.S. Environmental-
Protection Agency, Research Triangle Park, NC.
3-169
-------
Ripperton, L. A.; Worth, J. J. B.; Vukovich, F. M.; Decker, C. E. (1977) Research
Triangle Institute studies of high ozone concentrations in nonurban areas.
In: Proceedings of the International Conference on Photochemical Oxidant
Pollution and Its Control, v.I. Dimitriades, B. , ed. Research Triangle
Park, NC: U.S. Environmental Protection Agency. EPA 600/3-77-OOla.
pp. 413-424.
Roberts, J. M. ; Fehsenfeld, F. C.; Albritton, D. L.; Sievers, R. E. (1983)
Measurement of monoterpene hydrocarbons at Niwot Ridge, Colorado. JGR J.
Geophys. Res. 88: 10,667-10,678.
Robinson, E. (1952) Some air pollution aspects of the Los Angeles temperature
inversion. Bull. Am. Meteorol. Soc. 33: 247-250.
Robinson, E. ; Robbins, R. C. (1975) Gaseous atmospheric pollutants from urban
and natural sources. In: Singer, S. F., ed. The changing global environment.
Boston: D. Reidel Publ. Co.
Roth, P. M.; Reynolds, S. D.; Anderson, G. E.; Pollack, R. I.; Yooke, M.
A.; and Killus, J. P. (1976) An evaluation of methodologies for assessing
the impact of oxidant control strategies. EF76-112R. Systems Applications
Inc., San Rafael, CA.
Roth, J. A.; Sullivan, D. E. (1981) Solubility of ozone in water. Ind. Eng.
Chem. Fundam. 20: 137-140.
Rubino, R. A.; Bruckman, L.; Magyar, J. (1976) Ozone transport. J. Air Pollut.
Control Assoc. 26: 972-975.
Salop, J.; Wakelyn, N. T.; Levy, G. F.; Middleton, E. M.; Gervin, J. C. (1983)
The application of forest classification from Landsat data as a basis for
natural hydrocarbon emission estimation and photochemical oxidant model
simulations in Southeastern Virginia. J. Air Pollut. Control Assoc. 33:
17-22.
Sakamaki, F.; Hatakeyama, S.; Akimoto, H.(1983) Formation of nitrous acid
and nitric oxide in the heterogeneous dark reaction of nitrogen dioxide
and water vapor in a smog chamber. Int. J. Chem. Kinet. 15: 1013-1029.
Sanadze, G. A.; Dolidze, G. M. (1962) C5H8 (isoprene) type in volatile emission
from leaves of plants [transl.]. Soobshch. Akad. Nauk Gruz. SSR 57: 1222.
Sandberg, J. S.; Basso, M. J.; Okin, B. A. (1978) Winter rain and summer ozone:
a predictive relationship. Science (Washington, DC) 200: 1051-1054.
Schere, K. L.; Shreffler, J. H. (1982a) Evaluation of four urban-scale photo-
chemical air quality simulation models. Draft. Office of Research and
Development, Research Triangle Park, NC: U.S. Environmental Protection
Agency.
Schere, K. L.; Shreffler, J. H. (1982b) Final evaluation of urban-scale
photochemical air quality simulation models. Draft. Office of Research
.and Development, U.S. Environmental Protection Agency, Research Triangle
Park, NC.
3-170
-------
Scheutzle, D, ; Rasmussen, R. A. (1978) The molecular composition of secondary
aerosol particles formed from terpenes. J. Air Pollut. Control Assoc.
28: 236-240.
Schrenk, H. H.; Heimann, H.; Clayton, G. D.; Gafafer, W. M.; Wexler, H. (1949)
Air pollution in Donora, PA: epidemiology of the unusual smog episode of
October 1948: preliminary report. Washington, DC: Public Health Service;
Public Health Service bulletin no. 306.
Schwartz, W. (1974) Chemical characterization of model aerosols. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-650/3-74-011. Available from: NTIS, Springfield, VA; PB-238557. '
Schwartz, S. E. (1984) Gas-aqueous reactions of sulfur and nitrogen oxides in
liquid-water clouds. In: Calvert, J. G., ed., S02, NO and N02 Oxidation
Mechamisms: Atmospheric Considerations, Chapter 4. Ann Arbor Science
Publishers, Acid Precipitation Series, v. 3. Boston, MA: Butterworth
Publishers, pp. 173-208.
Scott, W.E.; Stephens, E.R.; Hanst, P.L. ; Doerr, R.C. (1957) Further develop-
ments in the chemistry of the atmosphere. Proc. Amer. Petrol. Inst. 37
(III): 171-183.
Seila, R. L. (1981) Nonurban hydrocarbon concentrations in ambient air north
of Houston, TX. In: Bufalini, J. J.; Arnts, R. R., eds. Atmospheric
biogenic hydrocarbons. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.
Seinfeld, J. H. ; Wilson, K. R. (1977) The issue of air quality simulation
model utility. In: International Conference on Oxidants, 1976 - Analysis
of Evidence and Viewpoints, Part VI. Research Triangle Park, NC: U.S.
Environmental Protection Agency. EPA-600/3-77-118.
Sexton, K. (1982) Evidence of an additive effect for small city plumes. Pre-
sented at: the 75th annual meeting of the Air Pollution Control Associa-
tion; June; New Orleans, LA. Pittsburgh, PA: Air Pollution Control Asso-
ciation; paper no. 82-31.4. (see also, Sexton, K. (1983) Evidence for an
additive effect for ozone plumes from small cities. Environ. Sci. Technol.
17: 402-407.)
Sexton, K.; Westberg, H. (1980) Elevated ozone concentrations measured downwind
of the Chicago-Gary urban complex. J. Air Pollut. Control Assoc. 30: 911-
914.
Sexton, K.; Westberg, H. (1984) Nonmethane hydrocarbon composition of urban
and rural atmospheres. Atmos. Environ.: in press.
Shafer, T. B.; Seinfeld, J. H. (1985) Evaluation of chemical reaction
mechanisms for photochemical smog. Part III. Sensitivity of the empirical
kinetic modeling approach (EKMA) to chemical mechanism and input parameters.
Research Triangle Park, NC: U.S. Environmental Protection Agency.
Shreffler, J. H.; Evans, R. B. (1982) The surface ozone record from the regional
air pollution study, 1975-1976. Atmos. Environ. 16: 1311-1321.
3-171
-------
Singh, H. B.; Hanst, P. L. (1981) Reroxyacetyl nitrate (PAN) in the unpolluted
atmosphere: an important reservoir for nitrogen oxides. Geophys. Res. Lett.
8: 941-944.
Singh, H. B.; Salas, L. J. (1983) Peroxyacetyl nitrate in the free troposphere.
Nature (London) 302: 326-328.
Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1977) Ozone in clean remote
atmospheres: Concentrations and variabilities. SRI Project 5661. Menlo
Park, CA: Stanford Research Institute (SRI).
Singh, H, B.; Ludwig, F. L.; Johnson, W. B. (1978) Tropospheric ozone: concen-
trations and variabilities in clean remote atmospheres. Atmos. Environ.
12: 2185-2196.
Singh, H. B.; Viezee, W.; Johnson, W, B.; Ludwig, F. L. (1980) The impact of
stratospheric ozone on tropospheric air quality. J. Air Pollut. Control
Assoc. 30: 1009-1017.
Singh, H. B.; Salas, L. J.; Stiles, R. E. (1982) Distribution of selected
gaseous organic mutagens and suspect carcinogens in ambient air. Environ.
Sei. Techno!. 16: 872-879.
Singh, H. B.; Jaber, H. M.; Davenport, J. E. (1984) Reactivity/volatility
classification of selected organic chemicals: Existing data. (Prepared
by SRI for U. S. EPA). Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA-600/3-84-082.
Slade, D. H., ed. (1968) Meteorology and atomic energy 1968. Washington, DC:
U.S. Atomic Energy Commission. Available from: NTIS, Springfield, VA;
TID-24190.
Smith, L. R.; Black, F. H. (1980) Characterization of exhaust emissions from
passenger cars equipped with three-way catalyst control systems. Presented
at Passenger Car Meeting, Society of Automotive Engineers, Dearborn, MI;
June 9-13. SAE paper no. 800822.
Soderlund, R.; Svensson, B. H. (1976) The global nitrogen cycle. In: Soderlund,
R.; Svensson, B. H.., eds; Nitrogen, phosphorus, and sulfur—global cycles.
SCOPE report 7. Ecol. Bull (Stockholm) 22: 23-73.
Souza, A. F.; Daley, P. S. (1978) U.S. Air Force turbine engine emission survey;
v. Ill: engine model summaries. Tyndall Air Force Base', FL: Civil and
Environmental Engineering Development Office, Air Force Systems Command;
CEEDO-TR-78-34.
Spicer, C. W.; Holdren, M. W.; Keigley, 6. W. (1983) The ubiquity of peroxy-
acetyl nitrate in the continental boundary layer. Atmos. Environ. 17:
1055-1058.
Spicer, C. W.; Joseph, D. W.; Sticksel, P. R.; Ward, G. F. (1979) Ozone sources
and transport in the northeastern United States. Environ. Sci. Techno!.
13: 975-985.
3-172
-------
Spicer, C. W.; Joseph, D. W.; Sticksel, P. R. (1982) An investigation of the
ozone plume from a small city. J. Air Pollut. Control Assoc. 32: 278-281.
Springer, K. J.; Stahman, R. C. (1977a) Diesel car emissions—emphasis on parti-
culate and sulfate. Presented at: the international automotive engineering
congress and exposition; February-March; Detroit, MI. Warrendale, PA: Society
of Automotive Engineers; SAE paper no, 770254.
Springer, K. J.; Stahman, R. C. (1977b) Unregulated emissions from diesels used
in trucks and buses. Presented at: International automotive engineering
congress and exposition; February-March; Detroit, MI. Warrendale,, PA:
Society of Automotive Engineers; SAE paper no. 770258.
Stair, R. (1961) The spectral radiant energy from the sun through varying
degrees of smog at Los Angeles. In: Proceedings of the third national air
pollution symposium; 1955; Pasadena, CA. pp. 48-54.
Stelson, A. W.; Friedlander, S. K.; Seinfeld, J. H. (1979) A note on the equi-
librium relationship between ammonia and nitric acid and particulate
ammonium nitrate. Atmos. Environ. 13: 369-371.
Stephens, E. R. (1964) Absorptivities for infrared determination of peroxyacyl
nitrates. Anal. Chem. 36: 928r929.
Stephens, E. R. (1967) The formation of molecular oxygen by alkaline hydrolysis
of peroxyacetyl nitrate. Atmos. Environ. 1: 19-20.
Stephens, E. R. (1969) The formation, reactions, and properties of peroxyacyl
nitrates (PANs) in photochemical air pollution. In: Pitts, J. N., Jr.;
Metcalf, R. L., eds. Advances in environmental sciences v. 1. New York,
NY: Wiley-Interscience; pp. 119-146.
Stephens, E. R.; Burleson, F. R.; Holtzclaw, K. M. (1969) A damaging explosion
of peroxyacetyl nitrate. J. Air Pollut. Control Assoc. 19: 261.
Sticksel, P. R.;. Spicer, C. W.; Westberg, H. (1979) Ozone plumes from small
cities and airborne observations of ozone in a moving high pressure system.
Presented at: the 72nd annual meeting of the Air Pollution Control Associa-
tion; June; Cincinnati, OH. Pittsburgh, PA: Air Pollution Control Associa-
tion; paper no. 79-58.6
Stockwell, W. R.; Calvert, J. G. (1983) The mechanism of the H0-S02 reaction.
Atmos. Environ. 17: 2231-2235.
Su, F.; Calvert, J. G.; Shaw, J. H. (1980) An FT-IR spectroscopic study of the
ozone-ethene reaction mechanism in 02-rich mixtures. J. Phys. Chem.
84: 239-246.
Sze, N. D.; Rice, H. (1976) Nitrogen cycle factors contributing to N20
production from fertilizers. Geophys. Res. Lett. 3: 343-346.
Tiao, G. C.; Box, G. E. P.; Hamming, W. J. (1975) Analysis of Los Angeles photo-
chemical smog data: a statistical overview. J. Air Pollut. Control Assoc.
25: 260-268.
3-173
-------
Tilden, J. W.; Seinfeld, J, H. (1982) Sensitivity analysis of a mathematical
model for photochemical air pollution. Atmos. Environ. 16: 1357-1364.
Tilton, B, E.; Bruce, R. M. (1981) Review of criteria for vapor-phase hydro-
carbons. Research Triangle Park, NC: U.S. Environmental Protection
Agency: EPA report no. EPA-600/8-81-022. Available from: NTIS, Springfield,
VA; PB82-136516.
Tingey, D. T.; Burns, W. F. (1980) Hydrocarbons emissions from vegetation. In:
Proceedings, symposium on effects of air pollutants on Mediterranean and
temperate forest ecosystems, Riverside, CA, June 22-27.
Tingey, D. T.; Manning, M.; Grothaus, L. C.; Burns, W. F. (1980) Influence of
light and temperature on monoterpene emission rates from slash pine. Plant
Physio1, 65: 797-801.
Tingey, D. T.; Evans, R.; Gumpertz, M. (1981) Effects of environmental conditions
on isoprene emission from live oak. Planta 152: 565-570.
Tuazon, E. C.; Graham, R. A.; Winer, A. M.; Easton, R. R.; Pitts, J. N., Jr.
Hanst, P. L. (1978a) A kilometer pathlength Fourier-transform infrared
system for the study of trace pollutants in ambient and synthetic atmos-
pheres. Atmos. Environ. 12: 865-875.
Tuazon, E. C.; Winer, A. M.; Grahami R. A.; Schmid, J, P.; Pitts, J. N., Jr.
(1978b) Fourier transform infrared detection of nitramines in irradiated
amine-NQ systems. Environ. Sci. Techno!. 12: 954-958.
f\
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1980) Atmospheric
measurements of trace pollutants by kilometer-pathlength FT-IR spectroscopy.
In: Pitts, J. N., Jr.; Metcalf, R. L.; Grosjean, D., eds. Advances in
environmental science and technology: v. 10. New York, NY: John Wiley and
Sons; pp. 259-360.
Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981a) Trace pollutant concen-
trations in a multiday smog episode in the California South Coast Air
Basin by long-path!ength Fourier-transform infrared spectroscopy. Environ.
Sci. Techno!. 15: 1232-1237.
Tuazon, E. C.; Carter, W. P. L.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Reac-
tions of hydrazines with ozone under simulated atmospheric conditions.
Environ. Sci. Techno!. 15: 823-828.
Tuazon, E. C.; Carter, W. P. L.; Brown, R. V.; Atkinson, R.; Winer, A. M.; Pitts,
J. N., Jr. (1982) Atmospheric reaction mechanisms of amine fuels. U.S.
Air Force final report, ESL-TR-82-17.
Tuizon, E. C.; Atkinson, R.; Plum, C. N.; Winter, A. M.; Pitts, J. N., Jr.
(1983) The reaction of gas phase N205 with water vapor. Geophys. Res. Lett.
10: 953-956.
Tuazon, E. C.; Sanhueza, E.; Atkinson, R.; Carter, W. P. L.; Winer, A. M.;
Pitts, J. N., Jr. (1984) Direct determination of the equilibrium constant
at 298 K for the N02 + N03 -* N205 reactions. J. Phys. Chem. 88: 3095-3098.
3-174
-------
U.S. Department of Commerce. (1968) Climatic atlas of the United States. Asheville,
NC: National Climatic Center. [Reprinted (1977) by the National Oceanic
and Atmospheric Administration.]
U.S. Department of Energy. (1978) Energy data reports. Washington, DC: U.S.
Department of Energy; publication no. EIA-Q049/1.
U.S. Department of Health, Education, and Welfare (1970) Air quality criteria
for photochemical oxidants. Washington, DC: National Air Pollution Control
Administration; NAPCA report no. AP-63. Available from: NTIS, Springfield,
VA; PB82-242231.
U.S. Environmental Protection Agency. (1977) Uses, limitations and technical
basis of procedures for quantifying relationships between photochemical
oxidants and precursors. Research Triangle Park, NC: Office of Air Quality
Planning and Standards Report no. EPA-450/2-77-021a.
U.S. Environmental Protection Agency. (1978a) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Environmental Criteria and Assessment
Office; pp. 116-135; EPA report no. EPA-6QO/8-78-004. Available from:
NTIS, Springfield, VA; PB80-124753.
U.S. Environmental Protection Agency. (1978b) Procedures for quantifying
relationships between photochemical oxidants and precursors: supporting
documentation. Research Triangle Park, NC: Office of Air Quality Planning
and Standards; EPA report no. EPA-450/2-77-021b.
U.S. Environmental Protection Agency. (1980a) Procedures for the preparation
of emission inventories for volatile organic compounds: v. 1. 2nd ed.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-450/2-77-Q28. Available from: NTIS, Springfield, VA;
PB-275292.
U.S. Environmental Protection Agency. (1980b) Final emission inventory require-
ments for 1982 ozone state implementation plans. Research Triangle Park,
NC: U.S. Environmental Protection Agency; EPA report no. EPA-450/4-80-016.
Available from: NTIS, Springfield, VA; PB81-060434.
U. S. Environmental Protection Agency. (1980c) Guideline for use of city-specific
EKMA in preparing ozone SIPs. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-450/4-80-027.
U.S. Environmental Protection Agency. (1981a) Comparative application of the
EKMA in the Los Angeles area. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Office of Planning and Standards; EPA report
no. EPA-450/4-81-030d. Available from: NTIS, Springfield, VA; PB82-234204.
U.S. Environmental Protection Agency. (1981b) Procedures for emission inven-
tory preparation, v. I: emission inventory fundamentals. Research Triangle
Park, NC: Office of Air Quality Planning and Standards; EPA report no.
EPA-450/4-81-026a. Available from: NTIS, Springfield, VA; PB82-240102.
3-175
-------
U.S. Environmental Protection Agency. (1981c) Procedures for emission inven-
tory preparation, v. II: point sources. Research Triangle Park, NC: Office
of Air Quality Planning and Standards; EPA report no. EPA-450/ 4-81-026b.
Available from: NTIS, Springfield, VA; PB82-240110.
U.S. Environmental Protection Agency. (198ld) Procedures for emission inven-
tory preparation, v. Ill: area sources. Research Triangle Park, NC: Office
of Air Quality Planning and Standards; EPA report no. EPA-450/ 4-81-026c.
Available from: NTIS, Springfield, VA; PB82-240128.
U.S. Environmental Protection Agency. (1981e) Procedures for emission inven-
tory preparation, v. IV: mobile sources. Research Triangle Park, NC:
Office of Air Quality Planning and Standards; EPA report no.
EPA-450/4-81-026d. Available from: NTIS, Springfield, VA; PB82-240136.
U.S. Environmental Protection Agency. (1981f) Procedures for emission inven-
tory preparation, v. V: bibliography. Research Triangle Park, NC: Office
of Air Quality Planning and Standards; EPA report no. EPA-450/ 4-81-026e.
Available from: NTIS, Springfield, VA; PB82-240144.
U.S. Environmental Protection Agency. (1981g) The sensitivity of complex
photochemical model estimates to detail in input information. Research
Triangle Park, NC: U. S. Environmental Protection Agency, EPA report no.
EPA-450/4-81-031a.
U.S. Environmental Protection Agency. (1982a) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-
131011.
U.S. Environmental Protection Agency. (1982b) Air quality criteria for particu-
late matter and sulfur oxides. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA-600/8-82-029. Available from: NTIS, Springfield,
VA; PB84-156777.
U.S. Environmental Protection Agency (1984a). National air pollutant emission
estimates, 1940-1983. Research Triangle Park, NC: U.S. Environmental
Protection Agency. EPA-450/4-84-028.
U.S. Environmental Protection Agency (1984b) Guideline for using the carbon
bond mechanism in city-specific EKMA. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA report no. EPA-450/4-84-005.
U.S. Environmental Protection Agency (1984c) User's manual for OZIPM2: Ozone
isopleth plotting with optional mechanisms/version 2. Research Triangle
Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-450/4-84-024.
U.S. Environmental Protection Agency (1986) National air pollutant emission
estimates, 1940-1984. Research Triangle Park, NC: U.S. Environmental
Protection Agency. Report no. EPA-450/4-85-014.
3-176
-------
Vaughan, W. M.; Chan, M.; Cantrell, B.; Pooler, F. (1982) A study of persistent
elevated pollution episodes in the Northeastern United States. Bull. Am.
Meteorol. Soc. 63: 258-266.
Viezee, W.; Singh, H. B. (1982) Contribution of stratospheric ozone to ground
level ozone concentrations—a scientific review of existing evidence.
Final report (revised draft). EPA grant CR 809330010. Menlo Park, CA:
SRI International.
Viezee, W.; Johnson, W. B.; Singh, H. B. (1979) Airborne measurements of strato-
spheric ozone intrusions into the troposphere over the United States: final
report. Atlanta, GA: Coordinating Research Council; SRI project 6690.
Viezee, W.; Singh, H. B.; Shigeishi, H. (1982) The impact of stratospheric ozone
on tropospheric air quality—implications from an analysis of existing
field data: final report. Atlanta, GA: Coordinating Research Council.
Available from: NTIS, Springfield, VA; PB82-256231.
Viezee, W.; Johnson, W. B.; Singh, H. B. (1983) Stratospheric ozone in the lower
troposphere-II. Assessment of downward flux and ground-level impact. Atmos.
Environ. 17: 1979-1993.
Vukovieh, F. (1977) Review and analysis. In: International conference on oxi-
dants, 1976~-Analysis of evidence and viewpoints. Part V. The issue of
oxidant transport. Research Triangle Park, NC. U.S. Environmental Protec-
tion Agency; EPA report no. EPA-600/3-77-117. Available from: NTIS,
Springfield, VA; PB-278750.
Wellington, T. J.; Atkinson, R.; Winer, A. M. (1984) Rate constants for the
gas phase reaction of OH radicals with peroxyacetyl nitrate (PAN) at 273
and 296 K. Geophys. Res. Lett, 11: 861-864.
Weast, R. C., ed. (1977) CRC handbook of chemistry and physics. 57th ed. Boca
Raton, FL: CRC Press; p. D-194.
*
Wells, C. G. (1981). Methods of estimating foliar biomass. In: Bufalini, J.
J.; Arnts, R. R., eds. Atmospheric biogenic hydrocarbons, v. 1. Ann Arbor,
MI: Ann Arbor Science; pp. 107-114.
West, E. S.; Todd, W. R.; Mason, H. S.; Van Bruggen, J. T. (1966) Textbook of
Biochemistry, 4th Ed. New York, NY; The Macmillan Co.
Westberg, H. (1977) Analysis of evidence and viewpoints. Part IV. The issue
of natural organic emissions. International Conference on Oxidants, 1976.
Research Triangle Park, NC: U.S. Environmental Protection Agency. EPA report
no. EPA 600/3-77-116.
Westberg, H.; Lamb, B. (1983) Ozone production and transport in the Atlanta,
GAj region. Final report; EPA. grant no. CR 809221. Research Triangle
Park, NC: U.S. Environmental Protection Agency, -Atmospheric Sciences
Research Laboratory.
3-177
-------
Westberg, H.; Allwine, K. J.; Elias, D. (1976) Vertical ozone distribution above
several urban and adjacent rural areas across the United States. In:
Specialty conference on ozone/oxidants—interactions with the total
environment. Pittsburgh, PA: Air Pollution Control Association; pp. 84-95.
Westberg, H.; Allwine, K. J.; Robinson, E. (1978a) The transport of oxidant
beyond urban areas: light hydrocarbon and oxidant data, New England Study,
1975. Research Triangle Park, NC: U.S. Environmental Protection Agency,
Office of Research and Development; EPA report no. EPA-6007 3-78-006.
Available from: NTIS, Springfield, VA; PB-277043.
Westberg, H.; Allwine, K.; Robinson, E. (1978b) Measurement of light hydrocarbons
and studies of oxidant transport—Houston study 1976. Research Triangle
Park, NC: U.S. Environmental Protection Agency, Environmental Sciences
Research Laboratory; EPA report no. EPA-600/3-78-062. Available from: NTIS,
Springfield, VA; PB-285891.
Westberg, H.; Sexton, K.; Roberts, E. (1981) Transport of pollutants along the
western shore of Lake Michigan. J. Air Pollut. Control Assoc. 31: 385-388.
Whitby, R. A.; Coffey, P. E. (1977) Measurement of terpenes and other organics
in an Adirondack mountain pine forest. JGR J. Geophys. Res. 82: 5928-5934.
Whitten, G. Z. (1983) The chemistry of smog formation: a review of current
knowlege. Environ. Int. 9: 447-463.
Whitten, G. Z.; Gery, M. W.(1986) Development of CBM-X mechanisms for urban
and regional AQSM's. Prepared under EPA contract no. 68-02-3738 (M. C.
Dodge, Project Officer) by Systems Applications, Inc., San Raphael, CA.
Research Triangle Park, NC: U.S. Environmental Protection Agency.
Whitten, G. Z,; Hogo, H. (1977) Mathematical modeling of simulated photochemical
smog. Research Triangle Park, NC: U.S. Environmental Protection Agency.
Report no. EPA-600/3-77-011.
Whitten, G. Z.; Hogo, H. (1978) User's manual' for kinetics model and ozone
isopleth plotting package. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/8-78-014a.
Whitten, G. Z.; Hogo, H. (1981) Application of the empirical kinetic modeling
approach to urban areas, v. II: Tulsa. Research Triangle Park, NC: U.S.
Environmental Protection Agency. EPA report no. EPA-450/4-81-005b.
Wight, G. D.; Wolff, G. T.; Lloy, P. J.; Meyers, R. E.; Cederwall, R. T. (1978)
Formation and transport of ozone in the northeast quadrant of the United
States. In: Morris, A. L.; Barras, R. C.; eds. Air quality meteorology
and atmospheric ozone. Philadelphia, PA: American Society for Testing and
Materials; ASTM STP 653.
Wilson, W. E. (1978) Sulfates in the atmosphere: a progress report on project
MISTT. Atmos. Environ. 12: 537-548.
3-178
-------
Wine, P. H.; Kreutter, N. M.; Gump, C. A.; Ravishankara, A. R. (1981) Kinetics
of OH reactions with the atmospheric sulfur compounds H2S, CH3SH, CH3SCH3
and CH3SSCH3. J. Phys. Chem. 85: 2660-2665,
Winer, A. M. (1983) Investigation of the role of natural hydrocarbons in photo-
chemical smog formation in California: final report. Contract no. AO-056-32.
Sacramento, CA: California Air Resources Board.
Winer, A. M.; Lloyd, A. C.; Darnall, K. R.; Pitts, J. N., Jr. (1976) Relative
rate constants for the reaction of the hydroxyl radical with selected
ketone, chloroethenes and monoterpene hydrocarbons. J. Phys. Chem. 80:
1635-1639.
Winer, A. M.; Breuer, G. M.; Carter, W. P. L.; Darnall, K. R.; Pitts, J. N.,
Jr. (1979) Effects of ultraviolet spectral distribution on the photo-
chemistry of simulated polluted atmospheres. Atmos. Environ. 13: 989-998.
Winer, A. M.; Atkinson, R.; Pitts, J. N., Jr. (1984) Gaseous nitrate radical:
possible nighttime atmospheric sink for biogenic organic compounds.
Science (Washington, DC) 224: 156-159.
Wolff, G. T.; Lioy, P. J. (1978) An empirical model for forecasting maximum
daily ozone levels in the northeastern U.S. J. Air Pollut. Control Assoc.
28: 1034-1038.
Wolff, G. T.; Lioy, P. J. (1980) Development of an ozone river associated with
synoptic-scale episodes in the eastern United States. Environ. Sci. Techno!.
14: 1257-1260.
Wolff, G. T.; Lioy, P. J.; Meyers, R. E.; Cederwall, R. T.; Wight, G. D.; Pasceri
R; E.; Taylor, R. S. (1977a) Anatomy of two ozone transport episodes in
the Washington, D.C., to Boston, Mass., corridor. Environ. Sci. Techno!.
11: 506-510.
Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Meyers, R. E.; Cederwall, R. T. (1977b)
An investigation of long-range transport of ozone across the midwestern
and eastern United States. Atmos. Environ. 11: 797-802.
Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Pasceri, R. E. (1977c) Aerial inves-
tigation of the ozone plume phenomenon. J. Air Pollut. Control Assoc.
27: 460-463.
Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Pasceri, R. E. (1977d) Aerial investi-
gation of photochemical oxidants over the northeast. In: Bufalini, J. J.;
Lonneman, W. A., eds. Proceedings of symposium on 1975 northeast oxidant
transport study. Research Triangle Park, NC: U.S. Environmental Protection
Agency. Report no: EPA-600/3-77-017.
Wolff, G. T. ; Ferman, M. A.; Monson, P. R. (1979) The distribution of
beryl!ium-7 within high-pressure systems in the eastern United States.
Geophys. Res. Lett. 6: 637-639. (See also "Correction," Geophys. Res.
Lett. 6: 816, 1979.)
3-179
-------
Wolff, G. T.; Lioy, P. J.; Wight, G. D. (1980) Transport of ozone associated
with an air mass. J. Environ. Sci. Health. A15: 183-199.
Wolff, G. T.; Kelly, N. A.; Ferman, M. A. (1981) On the sources of summertime
haze in the eastern United States. Science (Washington, DC) 211:703-706.
Wolff, G. T.; Kelly, N. A.; Ferman, M. A. (1982) Source regions of summertime
ozone and haze episodes in the eastern United States. Water, Air, Soil
Pollut. 18: 65-81.
Zika, R.; Saltzman, E.; Chameides, W. L.; Davis, D. D. (1982) H202 levels in
rainwater collected in South Florida and the Bahama Islands. JGR J. Geophys.
Res: 87: 5015-5017.
Zimmerman, P. R. (1977) Procedures for conducting hydrocarbon emission inven-
tories of biogenic sources and some results of:recent investigations.
In: Emission inventory/factor workshop, v. 2. Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Air Quality Planning
and Standards; EPA report. Available from: NTIS, Springfield, VA;
PB-289123.
Zimmerman, P. R. (1978) Testing of Hydrocarbon Emissions from Vegetation and
Development of a Methodology for Estimating Emission Rates from Foliage.
Final Report, Draft. Prepared by Washington State University for U.S.
Environmental Protection Agency under•Contract No. DU-77-C063. U.S.
Environmental Protection Agency, Research Triangle Park, NC.
Zimmerman, P. R. (1979) Determination of emission rates of hydrocarbons from
indigenous species of vegetation in the Tampa/St. Petersburg, Florida,
area: final report; appendix C. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-904/9-77-028. Available from:
NTIS, Springfield, VA; PB-297057.
Zimmerman, P. R. (1980) Natural sources of ozone in Houston: natural organics.
In: Proceedings of specialty conference on ozone/oxidants—interactions
with the total environment; October 1979; Houston, TX. Pittsburgh, PA:
Air Pollution Control Association; pp. 299-310.
3-180
-------
4. SAMPLING AND MEASUREMENT OF OZONE AND OTHER
PHOTOCHEMICAL OXIDANTS AND THEIR PRECURSORS
4.1 INTRODUCTION
Detailed information is presented in this chapter on methods for sampling
and measuring ozone, "total oxidants," hydrogen peroxide, and peroxyacetyl
nitrate and its higher homologues. Because of their utility in atmospheric
research and in the application of oxidant abatement programs, methods for
sampling and measuring the organic and inorganic precursors to oxidants are
described as well. The information presented here should prove helpful to
state and local air pollution agencies and to researchers investigating health
and welfare effects. The chief reason for presenting such information, however,
is to provide relevant information: (1) for assessing the accuracy of aerometric
data on these pollutants; and (2) for determining the impact of respective
measurement and calibration methods on existing data on the health and welfare
effects of ozonej total oxidants, and individual other oxidants. Primary
emphasis is placed in this chapter on techniques considered satisfactory for
routine monitoring, on the effects of changes in calibration procedures for
ozone measurements, on the relationship between ozone and "total oxidant"
measurements, and on the accuracy and reliability of methods for measuring
oxidants not routinely monitored in ambient air.
Since the publication of the 1978 criteria document on ozone and other
photochemical oxidants (U.S. Environmental Protection Agency, 1978a), a new
procedure for calibrating ozone measurements has been promulgated by EPA as
the Federal Reference Method for calibration. In addition, EPA has continued
efforts to institute and codify a formal nationwide program of quality assurance
for the routine monitoring of pollutants in ambient air. Some examples of
these procedures are documented in this chapter as they apply to actual opera-
tion of the analytical instrumentation. Detailed descriptions of analytical
procedures, quality assurance procedures, and reporting requirements are
contained in Quality Assurance Handbook for Air Pollution Measurement Systems
(U.S. Environmental Protection Agency, 1977b). Pertinent rules and regulations
are contained in the Federal Register (1979a,b,c,d,e).
4-1
-------
Brief summaries are provided below of requirements pertaining to quality
assurance and sampling for ozone monitoring.
4.2 QUALITY ASSURANCE AND OTHER SAMPLING FACTORS IN MONITORING FOR
4.2.1 Quality Assurance in Ambient Air Monitoring for Ozone
Quality assurance as defined by EPA rules and regulations consists of two
distinct functions. One is the assessment of the quality of monitoring data
by estimating their precision and accuracy. The other is the control and
possible improvement of the quality of the ambient air data by implementation
of quality control policies, procedures, and corrective actions.
Each quality control program, developed by the individual States and
approved by the EPA Regional Administrator, must include operational procedures
for each of the following activities:
1. Selection of methods, analyzers, or samplers (prescribed refer-
ence and equivalent methods for ambient air monitoring are
described elsewhere in this chapter);
2. Installation of equipment;
3. Calibration (test concentrations for ozone must be obtained by
means of the ultraviolet (UV) photometric calibration procedure
described elsewhere in this chapter or by means of a certified
ozone transfer standard);
4. Zero/span checks and adjustments of automated analyzers;
5. Control checks and their frequency;
6. Control limits for zero, span, and other control checks, and
respective corrective actions when such limits are surpassed;
7. Calibration and zero/span checks for multiple range analyzers;
8. Preventive and remedial maintenance;
9. Quality control procedures for air pollution episode monitoring;
10. Recording and validation of data;
11. Documentation of quality control information.
A one-point precision check must be carried out at least every 2 weeks on
each automated analyzer used for ozone, using a precision test gas of known
concentration. Each calendar quarter, at least 25 percent of the analyzers
4-2
-------
used by the State and Local Air Monitoring Stations (SLAMS) for monitoring
ozone must be formally audited by an independent operator by challenging with
at least one audit gas of known concentration in each of the four concentration
ranges. Similar requirements are set forth for monitoring networks designed
to assess Prevention of Significant Deterioration (PSD) requirements.
In addition to requirements and recommendations associated with the
selection, installation, and maintenance of monitoring equipment, the above-cited
Federal Register publications discuss certain design criteria for monitoring
networks (SLAMS and the National Aerometric Monitoring Stations, NAMS). Included
are requirements on siting of monitors in order to obtain ozone concentrations
that are representative of regions of varying dimensions. For example, a
"middle scale" monitor would represent conditions close to sources of NO such
/v
that local ozone scavenging effects might be of significance. A "neighborhood
scale" monitor, on the other hand, would be located somewhere in a reasonably
homogeneous urban subregion having dimensions of a few kilometers. Other
"scales" applicable to siting of ozone monitors include urban scale, which
would be used to estimate concentrations characteristic of an area having
dimensions between several and 50 kilometers or to measure high concentrations
downwind of an area with high precursor emissions; and regional scale, used to
typify concentrations over portions of a major metropolitan complex up to
dimensions of hundreds of kilometers. For ozone SLAMS stations, applicable
scales are middle, neighborhood, urban, and regional. Requirements for NAMS
stations for ozone are neighborhood and urban scale. Two ozone NAMS stations
are expected to be sufficient for each urban area: one for specific transport
conditions leading to high ozone; and the other for monitoring peak concentra-
tions relative to population exposure.
4.2.2 SamplingFactors in Ambient Air Monitoring for Ozone
Sampling factors may have a crucial effect on the quality and utility of
measurements both in ambient air and in controlled laboratory situations.
Sampling techniques and strategies must preserve the integrity of a represen-
tative fraction of ambient air and must be consistent with the specific purpose
of the measurement. In this section, the significance of some sampling factors
will be discussed briefly. For more detailed discussions of this subject, the
reader is referred to Ott (1977) and to reports prepared for EPA by the National
Research Council (U.S. Environmental Protection Agency, 1977a; National Research
Council, 1977).
4-3
-------
4.2.2.1 Sampling Strategies and Air Monitoring Needs. Air monitoring data
relevant to assessing ambient Og or oxidant levels are collected for a variety
of specific needs, including:
1. Data to be used in trend analysis as indicators of the state of
attainment of ambient air quality standards.
2. Data to be used in development of Og control strategies and
evaluation of their effectiveness.
3. Data to be used in the development and validation of air quality
simulation models capable of application to the Qg problem.
4. Data to be used in investigation of causes of the ozone problem
both in general and in specific localities.
5. Data to be used in special research studies such as the effects
of ambient air pollution on human health and welfare.
Each specific purpose or need requires special considerations with regard
to air sampling strategy. For example, 5 or more years of On data might be
required for the adequate assessment of trends that resulted from the applica-
tion of a particular control strategy rather than trends that resulted from
chance local meteorological conditions. In contrast, the validation of an air
quality simulation model might require only a few carefully chosen days of
very detailed measurements of 0,, hydrocarbons, and NO , as well as detailed
meteorological data and time-varying emissions along the trajectory of the air
parcel in question.
4.2.2.2 Air Monitoring Site Selection. Ozone in the lower troposphere is a
product of photochemical reactions that involve sunlight, hydrocarbons, and
oxides of nitrogen. In typical urban atmospheres, ozone precursors react to
produce ozone at such a rate that the 0, reaches its daily peak level in the
middle of the day at locations downwind from the source-intensive center-city
area. Thus, if peak Og concentrations are to be measured, monitoring stations
should, in general, be located downwind from city centers. This downwind
distance may be on the order of 15 to 30 kilometers (9 to 19 miles), depending
on predominant wind patterns in the area (U.S. Environmental Protection Agency,
1977a). This distance may be highly area-specific, however. For example,
ozone maxima in the Los Angeles plume have been observed as far downwind as 50
to 70 km.
4-4
-------
Once a station is located, additional sampling considerations, arise
because of the chemical reactivity and instability of the (k molecule. Ozone
reacts extremely rapidly with NO and with some hydrocarbon compounds, including
most of those emitted by vegetation. Also, 0, decomposes readily on contact
with the surface of many materials. Consideration of these effects led to the
development of specific criteria for locating an Qg monitoring station (U.S.
Environmental Protection Agency, 1977a; National Research Council, 1977).
Briefly, the inlet of the sampling probe of the ozone analyzer should be
positioned 3 to 15 meters (10 to 49 feet) above ground, at least 4 meters
(13 feet) from large trees, and 120 meters (349 feet) from heavy automobile
traffic. Sampling probes should be designed so as to minimize 0, destruction
by surface reaction or by reaction with NO.
Another consideration that has significance for the selection of sites
for air monitoring stations is the fact that ambient .monitoring data, as
routinely obtained, have limited validity as absolute measures of air quality.
This limitation arises from the fact that, at ground level, the ambient atmos-
phere is inhomogeneous as a result of a continuous influx of fresh emissions,
incomplete mixing, and destruction of 0^ by fresh and unreacted emissions and
destruction on surfaces. In view of such inhomogeneity, monitoring data from
a fixed network provide measures of air quality at a discrete number of loca-
tions but may not detect temporal and spatial variations in ozone concentra-
tions of a localized nature. This problem can be alleviated by use of a
greater density of monitoring stations or by use of a validated air quality
model. Such models are capable of helping quantify the emission, dispersion,
and chemical reaction processes. Their outputs can provide data on the dis-
tribution of air quality concentrations between widely spaced ambient monitors.
The emphasis in this section has been on a brief discussion of sampling
strategies. The word sampling is also widely considered to mean those tech-
niques that are required to obtain a parcel of air that is representative of
the polluted atmosphere, and to maintain its integrity until a measurement of
concentration has been carried, out. Considerations relating to this, meaning
of sampling are discussed as appropriate in the following sections on measure-
ment techniques.
.4-5
-------
4.2.3 Measurement Methods forTotal Oxidants and Ozone
4.2.3.1 Total Oxidants. Although ozone was first unambiguously identified in
polluted atmospheres by infrared spectroscopy (Stephens et a!., 1956b; Scott
et al.» 1957), the earliest procedures for routinely monitoring 0, and other
oxidizing species in the atmosphere were based on iodometry. lodometric
techniques are inherently non-specific in that a variety of oxidizing species
in addition to Og may be positive interferences, whereas reducing agents are
negative interferences. Thus, the name "total oxidants" was coined because
the technique responded to 0, and other oxidants such as peroxides, peroxyacetyl
nitrate (PAN), and nitrogen dioxide (NO*). Total oxidants are then actually
defined by the particular iodometric procedure used, since the response to the
various oxidizing species present will depend on the details of the procedure.
This will be more evident when interferences are discussed below. The use of
the word "total" is in itself a misnomer. The measurement does not reflect a
sum of the oxidizing species present because the various oxidants present in
the atmosphere react to produce iodine at different stoichiometries and differ-
ent rates. In spite of these difficulties, the measurement of total oxidants
was a useful method for characterization of the atmosphere because of its
correlation with the principal oxidant, 0^; and, consequently, there is a
large oxidant data base available. For these reasons, the two principal
methods used for monitoring total oxidants are discussed briefly below.
The bulk of the total oxidants data base was obtained by the use of two
types of continuous monitoring instruments. In both types, an air sample is
continuously scrubbed by an aqueous reagent containing potassium iodide (KI).
Ozone and other oxidants produce iodine (tri-iodide ion) according to the
reaction:
03 + 31" + H20 -»• I3" + 02 + 20H (4-1)
In colorimetric oxidant instruments, the iodine is measured photometrically by
ultraviolet absorption (Littman and Benoliel, 1953). In the other common type
analyzer the iodine produced is measured by electrochemical means (Brewer and
Mil ford, 1960; Mast and Saunders, 1962). Many other chemical techniques for
oxidants have been proposed and in some cases applied, but for these reference
is made to the original literature (Hodgeson, 1972; Katz, 1976).
4-6
-------
The interferences for both colorimetric and amperometric 0^ analyzers are
other oxidizing and reducing species in the atmosphere. The major oxidant in
ambient air by far is 0^ (Chapter 5); and the other oxidants present, except
NOp, are considered part of the total oxidants measured rather than inter-
ferences. The only significant reducing interference known is SOp.
The magnitude of the NC^ interference is variable (Tokiwa et a!., 1972;
Intersociety Committee, 1970). For the Brewer amperometric cell, the inter-
ference from NOp is only 6 percent of an equivalent concentration of 03 (Tokiwa
et al.s 1972). For colorimetric oxidant analyzers, NO, interference equivalents
vary from 20 to 32 percent depending on 03 concentration (Tokiwa et al.,
1972), A "corrected" oxidant measurement is obtained by simultaneous measure-
ment of NOp and correction of the corresponding total oxidant measurement.
The interference from SOp is quantitative for both colorimetric and elec-
trochemical oxidant measurements, with one mole of SQp consuming one mole of
tri-iodide ion. If the SOp concentration is less than that of total oxidant
and SOp is simultaneously measured, the total oxidant may also be corrected
for SOp. This was the procedure previously applied in the older aerometric
data for California. For many areas of the East Coast and Midwest, such a
correction was not possible and preferential SOp scrubbers were used (Saltzman
and Wartburg, 1965; Mueller et al., 1973). These scrubbers could be effective
in the hands of skilled operators but their use was not without problems.
Among these problems were partial oxidation of NO to NOp and of HpS to SOp,
and partial removal of 03 when the scrubber was wet or contaminated (Hodgeson,
1972).
4.2.3.2 Ozone
4.2.3.2.1 Gas-phase chemi1uminescence. Many of the 0, oxidation reactions
are sufficiently energetic that they produce electronically excited products,
intermediates, or reactants, which in turn may chemiluminesce (Zocher and
Kautsky, 1923; Bowman and Alexander, 1966). Although well known for many
years, such reactions were not applied to chemical analysis until the 1960s.
In 1965, Nederbragt reported a detector that employed chemiluminescence from
the reaction of 0-, with ethylene for measurement of 0, in the vicinity of
large accelerators (Nederbragt et al., 1965; Warren and Babcock, 1970).
Applications to atmospheric analysis were a natural consequence (Stevens and
Hodgeson, 1973). The reference method for 0, originally promulgated by EPA
for compliance monitoring was the Og-ethylene chemiluminescence method (F.R.,
4-7
-------
1971). Appendix D of 40 CFR5 Part 50, describes the principle of the method,
including a method of calibration (F.R., 1971; C.F.R., 1971). Since then, the
measurement principle has remained the same but calibration procedures have
undergone extensive revision as discussed below (Section 4.2.4). It is also
noteworthy that the reference method is specific for 03, whereas the data used
for establishing the standard were based on measurement of total oxidants.
This issue is addressed in Section 4.2.5.
A flow of sample air (1 to 5 liter/min) containing 03 and a small flow of
pure ethylene are mixed at atmospheric pressure in a small reaction chamber
closely coupled to the photocathode of a photomultiplier tube. The reaction
between Qg and ethylene produces a small fraction of electronically excited
formaldehyde. Chemiluminescence from this excited state results in a broad
emission band centered at 430 nm (Finlayson et al., 1974). The emission
intensity that is monitored is a linear function of Q3 concentration from
0.001 to greater than 1 ppm. The relation between intensity and concentration,
i.e., instrument calibration, must be determined for each instrument with
standard concentrations of 0- in air. The minimum detection limit and the
response time are functions of detector design. Detection limits of 0.005 ppm
and response times of less than 30 seconds are readily attained, however, with
modest design features. For example, cooling the photomultiplier improves the
sensitivity but is not normally required. There are no known interferences
among the common atmospheric pollutants. There have been reports of a positive
interference when 0, is measured in the presence of water vapor, i.e., a
signal enhancement of 3 to 12 percent in high humidity as opposed to measure-
ment of the same concentration of 03 in dry air (California Air Resources
Board, 1976). Where this may be a real problem, it can be minimized by per-
forming calibrations with humidified air. Finally, in order to obtain accept-
able measurement precision and constant span, analyzers must contain means for
maintaining constant air and ethylene flow rates.
Ambient air monitoring reference and equivalent methods have been pub-
lished by EPA (F.R., 1975a). This regulation prescribes methods of testing and
performance specifications that commercial analyzers must meet in order to be
designated as a reference method or as an equivalent method. An analyzer may
be designated as a reference method if it is based on the same principle as
the reference chemiluminescence method and meets performance specifications.
An automated equivalent method must meet the prescribed performance specifi-
cations and show a consistent relationship with a reference method. These
4-8
-------
specifications for 03 analyzers are listed in Table 4-1. Commercial analyzers
that have been designated as reference or equivalent methods are listed in
Table 4-2. Information concerning the applications supporting the designation
of analyzers as reference or equivalent methods may be obtained by writing the
U.S. Environmental Protection Agency, Environmental Monitoring Systems
Laboratory, Research Triangle Park, NC 27711.
4.2.3.2.2 Gas-sol id chemi1uminescence. The first chemiluminescence technique
for (k was developed by Regener for stratospheric measurements (Regener, 1960)
and later for measurements in the troposphere (Regener, 1964). The reaction
of On with Rhodamine-B adsorbed on activated silica gel produces chemilumines-
cence in the red region of the spectrum characteristic of the fluorescence
spectrum of Rhodamine-B. The intensity is a linear function of CL concentration,
the minimum detection limit can be lower than 0.001 ppm, and no atmospheric
TABLE 4-1. PERFORMANCE SPECIFICATIONS FOR AUTOMATED
OF OZONE ANALYSIS
Performance parameter
Range
Noise
Lower detectable limit
Interference equivalent
Each interference
Total interference
Units
ppm
ppm
ppm
ppm
ppm
Specification
0 to 0.
0.005
0.01
±0.02
0.06
5
Zero drift, 12 and 24 hour
Span drift, 24 hour
ppm
±0.02
20% of upper range limit
80% of upper range limit
Lag time
Rise time
Fall time
Precision
20% of upper range limit
80% of upper range limit
percent
percent
minutes
minutes
minutes
ppm
ppm
±20.0
±5.0
20
15
15
0.01
0.01
Source: F.R. (1975a); C.F.R. (1975).
4-9
-------
TABLE 4-2. LIST OF DESIGNATED REFERENCE AND EQUIVALENT METHODS
OF OZONE ANALYSIS
Identification and source
Beckman Model 950A Ozone
Fed.
Vol.
42
Register notice
Page Date
28571 6/03/77
Designa-
tion
(E=equiv.
Ref . R=ref . )
(F.R. , 1977a) R
Analyzer
Beckman Instruments
2500 Harbor Boulevard
Fullerton, CA 92634
Bendix Model 8002 Ozone 41
Analyzer 45
The Bendix Corporation
Post Office Drawer 831
Lewisburg, WV 24901
Columbia Scientific Industries 44
Model 2000 Ozone Meter
11950 Jollyvilie Road
Austin, TX 78759
5145
18474
2/04/76
3/21/80
(F.R., 1976a)
(F.R., 1980a)
10429 2/20/79 (F.R.,1979a)
Dasibi Model 10Q8-AH Ozone
Analyzer
Dasibi Model 1003-AH
1003-PC or 1003-RS Ozone
Analyzers
Dasibi Environmental Corp.
616 E. Colorado Street
Glendale, CA 91205
MEC Model 1100-1 Ozone Meter, 41
MEC Model 1100-2 Ozone Meter, 42
or MEC Model 1100-3 Ozone Meter
Columbia Scientific Industries
11950 Jollyville Road
P.O. Box 9908
Austin, TX 78766
Meloy Model OA 325-2R Ozone 40
Analyzer
Meloy Model OA 350-2R Ozone 40
Analyzer
Columbia Scientific Industries
11950 Jollyville Road
Austin, TX 78759
48 10126 3/10/83 (F.R., 1983) E
42 28571 6/03/77 (F.R., 1977a) E
46647 10/22/76 (F.R., 1976b)
30235 6/13/77 (F.R., 1977b)
54856 11/26/75 (F.R., 1975b)
54856 11/26/75 (F.R., 1975b)
R
R
4-10
-------
TABLE 4-2 (continued).
LIST OF DESIGNATED REFERENCE AND EQUIVALENT METHODS
OF OZONE ANALYSIS
Identification and source
Fed.
Vol.
Register notice
Page Date
Designa-
tion
(E=equiv.
Ref . R=ref . )
Monitor Labs Model 8810 46
Photometric Ozone Analyzer
Monitor Labs, Incorporated
10180 Scripps Ranch Boulevard
San Diego, CA 92131
Monitor Labs Model 8410E 41
Ozone Analyzer
Monitor Labs, Incorporated
10180 Scripps Ranch
San Diego, CA 92121
PCI Ozone Corporation Model 47
LC-12 Ozone Analyzer
PCI Ozone Corporation
One Fairfield Crescent
West Caldwell, NJ 07006
Philips PW9771 03 Analyzer 42
Philips Electronic Instruments, 42
Incorporated
85 McKee Drive
Mahwah, NJ 07430
Thermo Electron Model 49
UV Photometric Ambient 03 45
Analyzer
Thermo Electron Corporation
Environmental Instruments Division
108 South Street
Hopkinton, MA 01748
52224 10/26/81 (F.R., 1981)
53684 12/08/76 (F.R., 1976d)
13572 03/31/82 (F.R., 1982) E
38931 08/01/77 (F.R,, 1977c)
57156 11/01/77 (F.R., 1977d)
57168 08/27/80 (F.R., 1980b)
interferences have been observed (Hodgeson et al., 1970). The technique is,
in fact, more sensitive than the gas-phase Nederbragt method and does not
require critical control of flow rate. It had the disadvantage in the original
analyzer built, however, that frequent and periodic internal calibration
cycles were required to compensate for changes and decaying sensitivity of the
surface of the detector (Regener, 1964; Hodgeson et al., 1970).
Improvement was made in the stability of the surface response in a modi-
fication added by Bersis and Vassiliou (1966), in which gallic acid is also
4-11
-------
adsorbed on the surface in excess. The Q3 apparently reacts with and consumes
the gallic acid rather than Rhodamine-B. An energy transfer step to Rhodamine-B
subsequent to the initial reaction results in the same chemiluminescence from
the dye compound, which is now no longer consumed. A commercial analyzer,
Phillips Model PW9771, is based on this principle and has been designated as
an equivalent method under EPA regulations.
4.2.3.2.3 Ultraviolet photometry. Ozone has a moderately strong absorption
band in the ultraviolet (UV), with a maximum very near the mercury 254 nm
emission line. This band is essentially a continuum near 250 nm. The molar
absorption coefficient at the mercury line has been measured by several inves-
—i ~i
tigators with good agreement and has an accepted value of 134 ± 2 M cm
(base 10) at 0°C and 1 atm (Hampson et a!., 1973). Ultraviolet absorption has
long been used as a method of measuring gas-phase 0, in fundamental chemical
and physical studies. Some of the first atmospheric 0, measurements were, in
fact, made by UV photometry; e.g., the Kruger Photometer. These early instru-
ments and the problems with their use are described more completely in the
first criteria document for photochemical oxidants (National Air Pollution
Control Administration, 1970). The major problem with the older photometric
instruments was the large imprecision involved in measuring the very small ab-
sorbance values obtained.
This problem of adequate sensitivity with moderate pathlengths has been
overcome by modern digital techniques for measuring small absorbancies. The
first instrument of this new generation of photometers was marketed by Dasibi
of Glendale, California, in the early 1970s. The details of this instrument
have been described by Bowman and Horak (1972). Other commercial instruments
have since been marketed and, along with the Dasibi, have been designated as
equivalent methods by EPA (Table 4-2). All of these instruments operate
effectively as double-beam digital photometers. A transmission signal is
averaged over a finite period of time with 0, present and is compared to a
similar transmission signal obtained through an otherwise identical reference
air stream from which the 0~ has been preferentially scrubbed; e.g., using a
manganese dioxide scrubber. The electronic comparison of the two signals can
be converted directly into a digital display of-Og concentration.
The UV photometric technique has the advantages, like gas-solid chemilu-
minescence, that a reagent gas flow is not required and that sample air flow
control is not critical. In addition, the measurement is in principle an
4-12
-------
absolute one, in that the concentration can be computed directly from the
measured absorbance since the absorption coefficient and the pathlength are
known. This capability is used extensively for the purpose of 03 calibration
as discussed in Section 5.5.5. Commercial UV photometers for Oo can serve a
dual function as a standard for 0^ calibration and as a means for measuring
ambient ozone concentrations. In practice, UV photometric analyzers that are
used for monitoring Og concentrations in the atmosphere are calibrated with
standard Og samples in order to compensate for possible 03 losses in the
sampling and inlet systems. A UV photometric analyzer has the potential
disadvantage that any molecular species that absorbs at 254 nm (e.g., SO,,,
benzene, mercury vapor) and that may also be removed along with 03 during the
reference cycle can interfere. Documentation of such interference during
atmospheric monitoring is lacking at present.
4.2.4 Generation and Calibration Methods for Ozone
Unlike the other criteria pollutants, 03 is a thermally unstable species
that must be generated in situ during the calibration of analyzers used for
atmospheric monitoring. This creates special requirements not encountered
with other pollutants and thus this section deals with means for generating
dynamic air streams containing stable 0^ concentrations and chemical and
physical means for absolute measurement of these concentrations.
4.2.4.1 Generation. Ozonized samples of air can be produced by a number of
means, including photolysis (Brewer and Mil ford, 1960), electrical discharge
(Toyama and Kobayashi, 1966), and radiochemical methods (Steinberg and Dietz,
1969). Electrical discharges are useful for producing high concentrations of
Oo in air for other applications; e.g., 0, chemistry. Radiochemical methods
would be ideal except for their cost and required safety features. By far the
most common method, however, for generating low concentrations of 0, in air in-
volves the photolysis of molecular oxygen.
02 + hY(X<200 nm) +20 (4-2)
0 + 02 + M (M = N2 or 02) •* 03 + M (4-3)
One of the most common photolytic generators uses a mercury vapor 6- or 8-inch
PenRay photolysis lamp positioned parallel to a quartz tube through which air
flows at a controlled rate. The 0, concentration is simply varied by means
4-13
-------
of an adjustable and calibrated mechanical sleeve placed over the lamp envelope
(Hodgeson et al., 1972) or by varying the voltage or current supplied to the
lamp.
4.2.4.2 Calibration
4.2.4.2.1 KI procedures: original EPA reference method. The output of
photolytic CL generators can provide air samples containing stable CL concen-
trations over a considerable period of time with careful control of flow rate,
lamp voltage, temperature, and pressure. It is necessary to calibrate these
generators periodically with an absolute reference method. Prior to 1975,
there were as many as seven different calibration methods for CL employed to
varying extents in this country (National Research Council, 1977). In an
attempt to standardize the methodology, EPA published a reference calibration
procedure with the reference method in 1971 when the oxidant (as CL) standards
were promulgated (F.R., 1971). This method was the 1 percent neutral buffered
potassium iodide (NBKI) procedure, a technique that had been used by EPA and
other agencies for some time.
During the early 1970s, it became evident that there were serious defi-
ciencies with the NBKI reference method. Several problems with the NBKI pro-
cedure, summarized by a joint EPA-NBS workshop in 1974 (Clements, 1975), in-
cluded the gradual continued release of iodine after sampling, variable results
obtained with different types of impingers, reagent impurities, and a positive
bias when compared to other 0, measurement methods. An interagency collabora-
tive study was undertaken to intercompare iodometric methods used by the Los
Angeles Air Pollution Control District (LAAPCD), the California Air Resources
Board (CARB), and EPA, using UV photometric 0, measurements as the reference.
The results of this study (DeMore et al., 1976) demonstrated the positive bias
of the NBKI methods. Concurrent with and after these earlier reports, a large
number of individual studies ensued. The history of these studies has been
reviewed in the previous criteria document (U.S. Environmental Protection
Agency, 1978a) and by Burton et al. (1976). The major conclusions from these
studies are presented below.
1. Results obtained by NBKI procedures are higher than those obtained
by UV photometry or gas-phase titration by 5 to 25 percent,
depending on details of the procedure.
4-14
-------
2. When Og is measured in the presence of humidified air, NBK!
results tend to be even higher by another 5 to 10 percent (e.g.,
California Air Resources Board, 1976). The reason for this
apparent moisture effect is not known.
3. In general, NBKI techniques are subject to large imprecision
because of procedural variation.
The EPA then evaluated four alternative calibration procedures (Rehme et a!,,
1981) and selected UV photometry as the reference procedure because of its
superior accuracy and precision and its simplicity of use.
Although NBKI methods are no longer used in this country for the purpose
of calibration, there is a considerable data base available on health and wel-
fare effects, as well as atmospheric chemistry and monitoring, that is based
on these methods as standards. Therefore, it is important to consider how
these data may be evaluated and compared to newer effects and aerometric data
based on the new UV calibration standard. Since a systematic bias is known to
exist between calibrations by KI methods and UV photometric methods, it should
be possible, in principle, to apply correction factors to convert from a KI
reference to a UV photometric reference. There are several problems inherent
in attempting such corrections, however. A fairly wide range of variations has
been reported in the literature on the comparison of KI and UV photometric
measurements. As discussed previously, the presence of moisture in the cali-
bration air increases the magnitude "of the bias. Fortunately, both the CARB
and the LAAPCD procedures called for the consistent use of humidified air,
whereas the EPA reference method prescribed the use of dry air. In addition,
the elapsed time between sample collection arid color measurement will also
affect the magnitude of the bias because of the slow liberation of iodine
after sampling (Clements, 1975; Beard et al., 1977; Hodgeson, 1976). Other
unknown experimental factors may also influence the bias, e.g., impinger
design (Clements, 1975; Beard et al., 1977).
An assessment has been made of the previous KI versus UV intercomparisons,
and recommendations are given in Table 4-3 for correction factors to apply to
calibration data for conversion from UV to a KI reference or vice versa. It
should be emphasized that these factors could validly be applied to correct
for a calibration bias only and can not be applied for comparison of data
where other effects are present, e.g., the comparison of oxidants versus O
4-15
-------
TABLE 4-3. FACTORS FOR INTERCOMPARISON OF DATA CALIBRATED BY
UV PHOTOMETRY VERSUS KI COLORIMETRY
Calibration method Ratio, KI/UV
EPA, 1% NBKI 1.12 ±0.05
CARB, 2% NBKI 1.20 ± 0.05
LAAPCD, 1% UKI 0.96a
Correction for this method not recommended; only one intercomparison has been -
reported.
data where the effects of oxidizing or reducing interferences must be consid-
ered. In this assessment, consideration was given only to those studies in
which the KI procedure was compared directly to UV photometry. The recommended
value for data based on the CARB method assumes the use of humidified air.
The value recommended for the EPA method assumes that dry air was used and
that color measurement was made immediately after sample collection.
The uncertainties assigned reflect the fact that a range of values has
been reported for the ratios in previous studies. Finally, whenever any
attempt is made to convert from one data base to another, these uncertainties
must be added by conventional error propagation techniques to the uncertainty
inherent with the original measurement.
4.2.4.2.2 Ultravi olet photometry method. A major reason UV photometry was
designated as the calibration procedure was the excellent precision of the
photometric measurement. In the collaborative study by Rehme et al. (1981),
measurement with ten individual UV photometers gave only a 3.4 percent varia-
bility when compared to a reference measurement system. Other significant
factors in the selection of UV photometry were the inherent simplicity of UV
photometric measurements and the ready availability of commercial instruments
that can also serve well as transfer standards between laboratory photometers
and field Og analyzers. (See National Research Council, 1977, and McElroy,
1979, for a discussion of transfer standards.)
It was also presumed that UV photometry gives more accurate results,
since the accuracy is determined primarily by the 03 absorption coefficient,
which is well known (Hampson et al., 1973; DeMore and Patapoff, 1976). Although
4-16
-------
there is little doubt that the accuracy of 0^ measurements has been signifi-
cantly improved by conversion to the UV basis, some question still exists
regarding the absolute relation between Og measurements by UV photometry and
Og measurements by GPT measurements based on either an NBS standard reference
material (SRM) nitric oxide (NO) gas cylinder or an N02 SRM permeation tube.
These intercomparisons have been made by several investigators over the past
10 years and have been summarized by Burton et al. (1976) and Paur and McElroy
(1979). The agreement between GPT and UV measurements was generally close to
unity, although in some cases GO measurements by GPT have shown a small positive
bias with respect to UV measurements (e.g., Rehme et al., 1981; DeMore et al.,
1976). DeMore and Patapoff (1976) reported a ratio of unity between simulta-
neous measurements of (K by GPT and UV with a 5 percent uncertainty on the
ratio of these measurements. In a recent detailed study conducted at the
National Bureau of Standards (NBS) (Fried and Hodgeson, 1982), 0-j measurements
made with an NBS standard photometer (Bass et al., 1977) were compared to GPT
measurements of 0,, that were standardized against both NO cylinders (NBS SRM)
and NQp permeation tubes (NBS SRM). Since the measurement of flow rates is a
critical GPT variable and has been considered as a major source of error in
GPT measurements (DeMore and Patapoff, 1976), NBS facilities were used for
making absolute flow measurements by both gravimetric and volumetric means.
The results of this study were that values of 0~ measured by GPT based on NOg
or NO SRMs agreed to within less than 1 percent, but that values of 0, measured
by UV were lower by 3 percent. For a consideration of possible error sources,
reference is made to the original article (Fried and Hodgeson, 1982). In
summary, the UV photometric 03 standard agrees quite closely with the NO and
N02 measurement standards by GPT, as it should in principle. The resolution
of any small biases that remain seems an appropriate matter for consideration
by EPA and NBS.
The measurement principle for the absolute measurement of Og by UV photom-
etry is the same as that used by instruments for monitoring atmospheric 0, as
described in Section 4.2.3.2.3 (Bowman and Horak, 1972; DeMore et al., 1976;
Bass et al., 1977). Ozone is measured in a dynamic flow system by measuring
the transmission, I/Io, of ozonized clean air in an absorption cell of path-
length H. When the concentration is to be expressed in units of ppm, meas-
urement of temperature and pressure is also required. The 0, concentration
may then be calculated directly from the Beer-Lambert equation:
4-17
-------
rn-1 = xo (4-4)
LU3Jppm «£ x 273 ^
where or = 0- absorption coefficient at 254 nm, 1 atm, and 0°C,
-1 -1
= 308 ± 4 atm cm (log base e),
and
T = temperature, °K;
P = pressure, torr.
Laboratory photometers used for primary Og calibrations have pathlengths of 1
to 5 meters and sophisticated digital electronic means for measuring small
absorbancies (Bass et a!., 1977; Bowman and Horak, 1972).
The conversion of a commercial 0~ photometric monitor to a photometer for
use as a transfer standard for calibration has been described by Paur and
McElroy (1979). Definitions are in order here. A primary standard UV photometer
is one that meets the requirements and specifications given in the 1979 revision
of the Oo measurement and calibration procedures (F.R. , 1979e). A transfer
standard as used by EPA is a device or a method that can be calibrated against
a primary photometer and transferred to another location for calibration of 03
analyzers. Commercial Q~ photometers have served well in this regard, but
other devices have been used as well; e.g., calibrated generators and GPT
apparatus. Guidelines on transfer standards for 0, have been published by EPA
(McElroy, 1979), and reference has already been made to the NAS discussion on
transfer standards (National Research Council, 1977).
The use of UV photometry is unique in air pollution measurements in that
it is based on a physical measurement principle rather than a chemical stan-
dard. It is then worthwhile to trace how the measurement chain works from a
primary standard to field measurements. The primary standard is referenced to
the accepted 0, absorption coefficient. Transfer standards are then calibrated
with primary photometers maintained at EPA, NBS, and elsewhere. The use of
commercial photometers in this regard has been described by several investiga-
tors (DeMore et al., 1976; Hodgeson et al., 1977). These and other kinds of
transfer standards are then used to calibrate 03 analyzers used for field
measurements.
4-18
-------
4.2.4.2.3 Other procedures. Although UV photometry has been specified as the
reference calibration procedure, other procedures are available that can give
equivalent results. These include a modified KI method, which was allowed as an
interim alternative method for the calibration of Og monitors when the UV
method was designated in 1979. Other KI methods that have been used success-
fully in Europe are also briefly discussed here. Finally, the GPT method is
reviewed since it has been used extensively in this country and was discussed
above with regard to the cross-check of method accuracies.
A major problem with NBKI techniques is the slow release of iodine and
continued color development after sampling. Flamm (1977) evaluated the rate
of this iodine production and found that it was the same as the rate with
which hydrogen peroxide (H^O^) releases iodine from the same solution. Based
on this observation and a consideration of other possible species that might
be responsible, Flamm concluded that certain buffer anions, including phosphate,
catalyze the formation of hLCU and yield stoichiometries for iodine production
greater than 1. Measurements made with a 1 percent KI reagent containing 0.1
M boric acid (BAKI), pH=5, did not exhibit this phenomenon, nor did the original
EPA method when phosphate was omitted from the reagent (Hodgeson, 1976).
The BAKI method was evaluated as one of four alternative techniques ,in
the collaborative study conducted by EPA (Rehme et a!,, 1981). No significant
bias was observed between BAKI and the reference technique based on UV photom-
etry. An analysis, however, of BAKI measurements by ten volunteers revealed
a large system-dependent variability, and thus the BAKI technique was not
recommended as an independent calibration method. It is noteworthy that the
system variability attributable to calibration was reduced somewhat if each
operator assumed a molar absorption coefficient for iodine (as I0) of 2.56 x
-1 -1
104 M cm at 352 nm rather than independently measuring the absorption
coefficient with standard I2 solutions as these procedures usually prescribe.
Measurement systems based on the BAKI procedure may still be certified as
transfer standards provided the guidelines for certification given in the EPA
technical assistance document for such standards are followed (McElroy, 1979).
Methods based on iodometry have been used in Europe for some time for the
calibration of 03 analyzers. Bergshoeff (1970) described a method for use in
the Netherlands, in which thiosulfate is added to the KI reagent (KIT method)
along with 0.1 M phosphate buffer. The iodine released is immediately reduced
by the thiosulfate and the amount of iodine consumed is determined by back-
titration of the thiosulfate. This method has the advantage that problems
4-19
-------
associated with iodine instability in solution are eliminated. In the Federal
Republic of Germany, the standard is based on a 2 percent KI reagent with 2
percent KBr (KIBr Method) and a low concentration (0.02 to 0.03 M) of phosphate
buffer (Van de Wiel et al.5 1979). These techniques have been compared to UV
and GPT measurement procedures by Van de Wiel, et al. (1979). Measurements
made with the KIBr method were in essential agreement with measurements by UV
or GPT, while measurements by the KIT method were too high by 15 to 25 percent,
depending on the relative humidity of the samples. Modifications have since
been made in the KIT method by the addition of KBr and reduction of the phos-
phate concentration.
The gas-phase titration (GPT) method employs the moderately rapid bimolec-
ular reaction between Og and NO to produce NG« (Rehme et al., 1974):
NO + 03 -» N02 + 02 (4-5)
This approach was, in fact, one of the early methods used to measure the
absorption coefficient of 0^ (Clyne and Coxon, 1968) and yielded excellent
agreement with other absolute techniques (DeMore and Patapoff, 1976). When NO
is present in excess, no side reactions occur and the stoichiometry is as
given above. This method has the distinct advantage that it gives an absolute
relation among three common pollutants. A measurement of the quantity of NO
or Og consumed or NOg produced provides a simultaneous measurement of the
other two species and the GPT procedure has been used in all three modes.
This calibration technique is often used in the calibration of chemilumines-
cence NO (NO + N09) analyzers. In order to obtain accurate concentration
f\ Lm
measurements in the procedure as normally employed, accurate flow measurements
are required; and this is the principal complexity and difficulty with this
procedure (DeMore and Patapoff, 1976). Stedman et al. (1976) has employed an
appropriate NO detector to make flow ratio measurements and thus avoid the
requirement for absolute flow measurements. Because of unexplained biases
between GPT measurement systems and the UV reference in the EPA collaborative
study, the GPT method was not recommended as an independent calibration tech-
nique (Rehme et al., 1981). It is still allowed, however, as a transfer
standard in accordance with the EPA guidelines for these standards (McElroy,
1979).
4-20
-------
4.2.5 Relationship between Methods for Total Oxldants and Ozone
When the ambient air quality standards for criteria pollutants were
originally established, a numerical standard was set for photochemical oxidants
as defined by measurements based on iodometric techniques. Much of the health-
related and welfare-related evidence used as the basis for the standards was
obtained using the total oxidant instrumentation discussed above. The reference
method specified in 1971, however, was the chemiluminescence measurement
of Og. Instrumental methods for the specific measurement of atmospheric 0,
became commercially available in 1970. These had several practical advantages
over total oxidant Kl-based instruments. These advantages were greater sensi-
tivity, precision, specificity—no interferences from ambient SOp and N02--and
improved reliability in routine monitoring. Second, the data available showed
that 03 was the major contributor to total oxidant measurements, that 03 was
the major contributor to observed health and welfare effects, and that 0,
could probably serve, then, as the best surrogate for measurements of total
oxidants and for controlling effects of oxidants in ambient air (see reviews
in Burton et al., 1976; U.S. Environmental Protection Agency, 1978a). :
Notwithstanding the promulgation of standards for ozone rather than
photochemical oxidants by EPA in 1979, an examination of the temporal .and
quantitative relationships between total oxidant and Q*3 data remains of con-
siderable interest, largely because earlier data and many newer data on health
and welfare effects were obtained by means of total oxidant methods. Aside
from the relative paucity of data on simultaneous measurements, there are two
distinct problems in making such comparisons. The first is the difficulty in
estimating the contributions to the total oxidant measurements from other
oxidizing species such as N02 and from reducing species such as SQp. 'The
presence of such species could cause the total oxidant measurements to be
either higher or lower than 0., concentrations. The second difficulty is in
estimating the bias created between past and present data as a result of the
change from the NBKI to the UV photometry calibration procedures. Fortunately,
these two problems can be treated separately and the latter problem vanishes
for comparison of simultaneous 0^ and oxidant data obtained using the same
calibration procedure. . . •
In the sections below, the relationship that should exist between total
oxidant and On is considered from an evaluation of the response of NBKI measure-
ments to other oxidizing and reducing species. The predicted relationship is
4-21
-------
then compared to data obtained in simultaneous field measurements of total
oxidant and Og.
4.2.5.1 Predicted Relationship. The predicted total oxidant measurements can
be expressed as the sum of the contributions from oxidizing and reducing
species that release or consume iodine in NBKI reagent:
[Total Ox] = a[03] + 2. b. [Ox]. + c[N02] (4-6)
- d[S02] - I. e. [Red].
In this equation, [Ox], and [Red], represent the concentrations of other
oxidizing and reducing species in the atmosphere. The atmospheric concentra-
tions of other reducing species, such as H2S, are normally quite low compared
to S02 concentrations (Stevens et al., 1972b, and references therein) and
these species will not be considered further here. If the concentrations
above are true atmospheric concentrations, the constants a, b, c, and d repre-
sent the efficiencies with which the various species release or consume iodine.
For example, the value of the constant a for an oxidant instrument calibrated
by the CARB 2 percent NBKI method would be approximately 1.2 (Section 4.2.4.2).
Since the instruments are calibrated with ozonized air, the factor a repre-
sents the bias of the calibration method used. If the 03 concentration is
overestimated because of calibration bias, then so are the contributions of
the other species by the same factor; i.e., the constants b, c, and d are all
higher than their true values by the same constant, a. Therefore, it should
in principle be possible to correct total oxidant data for calibration bias by
dividing both sides of the equation above by a.
[Total Ox]1 = [Total Ox]/a (4-7)
= [03] + I. b1. [Ox]. + c'[N02] - d'[S02]
Other atmospheric oxidants that have been identified and that may contrib-
ute to the total oxidant reading are hydrogen peroxide (H202), small organic
peroxides (e.g., methyl and ethyl hydroperoxide), peracetic acid, peroxyacyl
nitrates (Cohen et al., 1967), and pernitric acid (Niki et al., 1976). An
estimation of the contribution of these species to the total oxidant measure-
ment is quite difficult because individual b^'s have not been measured and
there are few data available on atmospheric concentrations of individual
species. The magnitude of the efficiency term will depend not only on the
4-22
-------
stoichiometry of the oxidation reaction, but also on the rate. A summary of
the effects of various oxidants on NBKI reagent and the Mast oxidant meter is
given in Table 4-4 (Cohen et a!., 1967; Purcell and Cohen, 1967; Burton et
al., 1976).
TABLE 4-4. RESPONSE OF NBKI REAGENT AND MAST METER TO VARIOUS OXIDANTS
Ozone
Peracetic acid
Hydrogen peroxide
Acetyl peroxide
Ethyl hydroperoxide
n- Butyl hydroperoxide
tert-Butyl hydroperoxide
Nitrogen oxides (NO )
f\
Peroxyacetyl nitrate (PAN)
Peroxypropionyl nitrate (PPN)
NBKI
Aa
A
B
B
B
B
B
D (10% as N02)
D
D
Mast
E
NAb
D
N
NA
NA
NA
D (10%)
N
D
aA = immediate color development; B = slow color development; 0 = positive
interference; E = good response; and N = no response (or negligible).
NA = Data not available.
Source: Cohen et al. (1967).
By contrast, reaction efficiencies for N02 and S02 are relatively well
known. Tokiwa et al. (1972) observed reaction efficiencies of 6 percent for
the Mast oxidant meter, 22 percent for a 10 percent KI colorimetric analyzer,
and variable (20 to 32 percent) for a 20 percent KI colorimetric analyzer. It
is well documented that SO^ is a quantitative negative interference with a 100
percent efficiency for reducing the oxidant reading by an amount equivalent to
the SO, concentration present (Cherniack and Bryan, 1965; Saltzman and Wartburg,
1965).
4-23
-------
Returning to the analytical expression for total oxidant,; an oxidant
value corrected for NQ2 and SOg interferences can be expressed as below, assum-
ing that no other significant reducing interferences are present:
[Total Ox]corr = [Total Ox] - c[N02] + [SOg] (4-8)
" = [03] + z. b. [Ox].
Thus, a total oxidant measurement for which legitimate corrections or compen-
sations for NOp and S02 have been made should always be higher than a
simultaneous 0, measurement by an amount that is a function of the type and
concentrations of other oxidants present. The only major qualifications to this
prediction are that both types of measurements must be sampling the same air
mass and be calibrated with respect to the same reference; that no other
significant reducing interferences are present; and that 03 losses within
sample inlet systems are insignificant. On the other hand, total oxidant data
uncorrected for SOp and NQ2 interferences may be higher or lower than corre-
sponding 03 data, depending on the concentrations of these pollutants. Because
of the potential presence of these interferences, it is quite difficult or
impractical to compare oxidant and 03 measurements during evening and early
morning hours, when 03 concentrations are quite low. Therefore, in the com-
parison of total oxidant and 03 simultaneous field measurements below, emphasis
is placed on comparison of peak hourly averages.
4.2.5.2 Empirical Relationship Determined from Simultaneous Measurements.
Several precautions should be taken in performing simultaneous measurements.
Both kinds of instruments must be calibrated frequently with the same ozonized
air stream that has been analyzed by a common reference method. In a simul-
taneous comparison, daily calibrations should be made with an 03 generator and
the generator output should be analyzed weekly. Both instruments should
sample the same air parcel. Routine maintenance should be frequent to ensure
constant gas and reagent flow rates, clean sample inlet systems, etc. Finally,
in any meaningful comparison of 03 and oxidant data, simultaneous measurements
of NOg and S02 should be made. If chromium trioxide scrubbers are used to
remove SOp in the inlet to the oxidant instrument, these must be frequently
tested to ensure that 03 is not also removed during continued use, particu-
larly under very humid conditions. These scrubbers may cause some additional
bias by oxidation of NO to NOo.
4-24
-------
The earliest comparative study reported was by Renzetti and Romanovsky
(1956). This study compared a phenolphthalein total oxidant monitor, a KI
continuous oxidant monitor, a rubber-cracking apparatus, and an open-path
ultraviolet spectrometer, which monitored the UV absorption at characteristic
Og absorption wavelengths. The only meaningful measurements for consideration
here are the KI oxidant and UV Og measurements, since these are similar to
measurement methods used later. The UV G~ spectrometer differed in a number
of respects from the 0, photometers in use today. Measurements were made
across an open optical path of 325 ft of the transmission at three wavelengths,
X-, = 265 nm, Xg = 313 nm, and A.., = 280 nm. Intensity ratios at the three
wavelengths were used to minimize the effects of other UV absorbers and of
particulate scattering. Some non-CU absorption may still have been present,
and, if so, the measured values would be higher than the true 0, values. The
published absorption coefficients at these three wavelengths were used to com-
pute 0, concentrations (Vigroux, 1952). Measurements were made over a 4-month
period. Figures 4-1 and 4-2 are illustrative of the data obtained for a
monthly average and a single day, respectively. Peaks of total oxidant and of
03 occurred at the same time, but the 0, maximum was usually less than the
total oxidant maximum. The UV 03 data were usually higher in the wings at low
Oo concentrations. Renzetti and Romanovsky attributed the higher total oxidant
reading to the presence of "other oxidants" and estimated concentrations of
other oxidants of 0.1 to 0.4 ppm, depending on 0- concentration. Since the
total oxidant instrument was calibrated by an NBKI method, this estimate is
almost certainly too high and a large portion of the difference between oxidant
and 0- may have been a result of the 20 to 25 percent positive calibration
bias. Since interferences may be present in the UV measurement and simultane-
ous measurements of N02 and SOp were not available at the time, no attempt is
made to make any more quantitative assessment of this study.
A later study (Cherniack and Bryan, 1965) compared a 10 percent colorimet-
ric KI oxidant instrument, a Mast oxidant meter (Brewer and Mil ford, 1960), a
galvanic-cell oxidant instrument (Hersch and Deuringer, 1963), and a UV 0^
photometer (Bryan and Romanovsky, 1956). This latter instrument was similar
in principle to present-day photometers. The precautions noted above were
taken. All the instruments were calibrated with respect to the 2 percent UKI
calibration procedure used by the LAAPCD. Simultaneous SOp and NO^ measure-
ments were made, but no corrections were made because the concentrations were
reported to be quite small during the period of comparison. Atmospheric sampling
4-25
-------
I
%
o
60
SO
40
30
QC
111 nn
Q ZO
z
o
0 10
I I I I I I I III
OXIDANTSBYKI
O3 BY UV
I I I I I I I I I I I I
I I I I I I I I I I I I I I I I I I ! I I I I
12 1 234
5 67 8 9 10 11 12 1 23 4 5 678 9 10 11 12
a.m. AUGUST 1955 p.m. P.S.T.
Figure 4-1. Comparison of ozone and total oxidant concentrations in the Pasadena area, August 1955.
Source: Renzetti and Romanovsky (1956).
60
50
40
30
20
10
I I I I I I I I I I I I I I f I I I I I I I I
OXIDANTS BY Ki
O
Ul
O
Z
o
O
0
12 1 2 3 45678 9 10 11 12 1 2 3456 78 9 10 11 12
a.m. AUGUST 27, 1955 p.m. P.S.T.
Figure 4-2. Comparison of ozone and total oxidant concentrations in the Los Angeles area,
August 1955.
Source; Renzetti and Romanovsky (1956).
4-26
-------
was conducted over an unspecified period of time, and the data were referenced
to the colorimetric oxidant measurements. The linear regression analysis of
the data over the concentration range 0 to 0.6 ppm gave the relationships
shown in Table 4-5 after correction for calibration factors. Thus, the data
show a much better absolute agreement and correlation between CL measurements
and colorimetric total oxidant than between electrochemical total oxidant and
colorimetric total oxidant. In addition, these data do not indicate any
significant contribution by "other oxidants" to the total oxidant measurement.
TABLE 4-5. COMPARISON OF CORRECTED INSTRUMENT READINGS TO COLORIMETRIC
OXIDANT READINGS DURING ATMOSPHERIC SAMPLING
Instrument
Mast meter
Galvanic cell
Ozone photometer
m
0.896
0.776
0.980
b
-0.013
+0.004
-0.005
r
0.868
0.867
0.982
(y = Instrument reading
= mx + b; x = colorimetric oxidant measurement; m = slope; b = intercept;
r = correlation coefficient)
Source: Cherniack and Bryan (1965).
During the 1970s, several studies were conducted on the intercomparison
of On and total oxidant instrumentation by the Research Triangle Institute
(RTI) of North Carolina (Ballard et al., 1971a,b; Stevens et a!., 1972a,b).
Measurements were made for 0^ by chemiluminescence and for total oxidant by a
colorimetric KI analyzer and a Mast meter. Calibrations were carried out
Mast meter. Calibrations were carried out frequently with an 0., generator
calibrated by the 1 percent NBKI method. Simultaneous N02 and S02 measurements
were also made and the oxidant data reported were corrected for these inter-
ferences. Clark et al. (1974) intercompared a commercial UV photometer, three
different commercial gas-phase chemiluminescence analyzers, and a gas-solid
chemiluminescence analyzer by monitoring in a rural environment. The instru-
ments were all calibrated by a common reference procedure and hourly-averaged
field measurements were collected over a 1-month period in August 1972. Davis
and Jensen (1976) reported intercomparisons of Mast meter total oxidant measure-
ments and chemiluminescence 0- measurements. The instruments were not calibrated
by the same procedure, however, nor were any corrections attempted for SOg and
4-27
-------
NOp interferences. Okita and Inugami (1971) reported an intercomparison of KI
total oxidant measurements with chemiluminescence 03 measurements in the urban
atmosphere of Musashino, Japan. An intercomparison of total oxidant by KI and
CL by chemiluminescence in irradiated auto exhaust was reported by Carroll et
al. (1972). In another extensive field study conducted at an air monitoring
station near the Houston ship channel, Severs and coworkers (Severs, 1975;
Neal et al., 1976) examined the relationship between ozone and total oxidants
for this area by making simultaneous measurements with a gas-phase chemilumi-
nescence Oo monitor and a Beckman colorimetric total oxidant analyzer. Primary
calibrations of the instruments were performed periodically using the EPA 1
percent NBKI method. No corrections were attempted for NOp or SOp interfer-
ences, but during the latter part of this study a chromium trioxide scrubber
was placed in the inlet of the total oxidant analyzer.
All of these 1970 studies were reviewed in the previous criteria document
(U.S. Environmental Protection Agency, 1978a). Only the major conclusions are
repeated here. In general, the averaged data showed fairly good qualitative
and quantitative agreement between the diurnal variations of total oxidants
and Og. The usual trend was a slightly higher value for the total oxidants
measurement at the maximum, a not unexpected result in view of the discussion
above. Comparisons of monthly-averaged data taken from studies in Los Angeles
and St. Louis are shown in Figures 4-3 and 4-4 (Stevens et al., 1972a,b).
The total oxidant data shown in Figure 4-4 are uncorrected and show distinct
morning and evening peaks resulting from N02 interference (see Chapter 3 for
diurnal patterns of N02). Examination of data taken from individual days
shows considerably more variation among the methods, with total oxidant measure-
ments both higher and lower than 03 measurements. Intercomparisons of only UV
photometric and chemiluminescence 03 analyzers have not shown these large
variations (Clark et al., 1974; Wendt, 1975). In all probability, these
variations result from the large imprecision and interferences in total oxidant
measurement.
Two of the studies described above reported consistently lower total
oxidant measurements. In one of these (Davis and Jensen, 1976), the reference
KI method was used for calibrating the chemiluminescence analyzer while a
factory calibration was used for the Mast meter. As pointed out above, other
studies have found low oxidant readings for the Mast meter as compared to
colorimetric analyzers (Cherniack and Bryan, 1965; Tokiwa et al., 1972; Stevens
et al., 1972a,b). The use of the factory calibration would cause the
4-28
-------
.10
E
Q.
D.
2 .06
1 J
OZONE-CHEM
TOTAL OX-MAST
TOTAL OX-TECH
Figure 4-3. Measurements of ozone and total oxidants in Los Angeles,
September 4 through September 30,1971.
Source: Stevens et al. (1972b).
-------
*>,
CO
o
E
a
a
h
Z
O
cc
I-
UJ
U
Z
o
o
0.12
0.10
,0,08
0.06
,0.04
COLORIMETRIC
COULOMETRIC
CHEMiLUMINESCENCE
0.02
r
I I I I I I
2 4 6 8 10 12 14 16 18 20 22 24
TIME, hours
Figure 4-4. Measurements of ozone and total oxidants in St. Louis,
October 14 through December 21,1971.
Source: Stevens et ai. (1972a).
-------
Mast readings to be even lower because of a calibration bias (Cherniack and
Bryan, 1965; Tokiwa et al., 1972). The results reported by Severs and coworkers
are more difficult to evaluate. Chemiluminescence CL values generally higher,
and sometimes considerably higher, than total oxidant measurements were re-
ported, although the measurements were referenced to the same calibration
procedure. Correlations were reported both with and without a chromium trioxide
scrubber in the oxidant inlet. These results are inconsistent with the known
responses of the instruments and the results of other investigators. The data
reported for one day of high 03, but abnormally low oxidants, are shown in
Figure 4-5. It is highly improbable that the problem is with the
chemiluminescence 0- measurement, since this response is typical of a normal
0- diurnal variation and no other species are known to interfere. It is more
probable that some other species of pollutant in the highly industrialized area
of the Houston ship channel repressed the response of the total oxidant analyzer,
which thus does not respond to Q3, much less to any other oxidant.
The most recent comparison in the literature involved simultaneous 03 and
total oxidant measurements in the Los Angeles basin by the California Air
Resources Board (1978) in the years 1974, 1976, and 1978. The maximum hourly
data pairs were correlated (Chock et al., 1982) and y-ielded the following
regression equation for 1978 data, in which a large number (927) of data pairs
were available:
Oxidant (ppm) = 0.870 03 + 0.005 (4-9)
(correlation coefficient = 0.92)
Thus, when the 03 levels were relatively high, they were actually slightly
higher than total oxidant. The total oxidant data were uncorrected for NO^
and S02 interferences.
In summary, specific 03 measurements agree fairly well with total oxidant
corrected for NO- and SO, interferences and Q~ is the dominant contributor to
total oxidant. Indeed, it is difficult to discern the presence of other
oxidants in most total oxidant data. There can, however, be major temporal
discrepancies between 03 and oxidant data, which are primarily a result of
oxidizing and reducing interferences with KI measurements. As a result of
these interferences, on any given day the total oxidant data may be higher
4-31
-------
UJ
to
0.150
a
a
Z
c 0.100
IU
o
z-
o
o
O 0.050
cc
o
m
o
— O3 BY CHEMILUMINESCENCE
,— OXIDANTS 1Y BECKMAN ACRALYZAR (Ki)
10 15
TIME, hours
20
Figure 4-5. Measurements of ozone and total oxidants, Houston Ship
Channsl, August 11,1973.
Source: Severs (1975).
-------
than or lower than simultaneous 03 data. The quantitative relationship be-
tween oxidant and (k data, such as that used by Chock et al. (1982), is proba-
bly quite location-dependent. From a methodologic standpoint, the measurement
of Og is a more reliable indicator than total oxidant measurements of oxidant
air quality, and such difficulties and controversy as may be involved in the
intercomparison of 0., and oxidant measurements are eliminated if the air
quality standard is defined in terms of CL.
4.2.6 Methods for Sampling and Analysis of Peroxyacetyl Nitrate and Its
HomoTbgues
4.2.6.1Introduction. Since the discovery of "compound X" (Stephens et al.,
1956a,b), later unambiguously identified as peroxyacetyl nitrate (PAN) (Stephens
et al., 1961), much effort has been directed toward its atmospheric measurement.
Peroxyacetyl nitrate and its homologues are products of photochemical reactions
involving hydrocarbons and oxides of nitrogen (NO ) in the atmosphere (Stephens
ar\
et al., 1961). The significance of atmospheric PAN is twofold. It is a
potent lachrymator and phytotoxicant in the ppb concentration range (Stephens
et al., 1961; Heuss and Glasson, 1968). Because of the reversible thermal
equilibrium (Hendry and Kenley, 1977),
CH30(02)N02 < CH3CO(02)- + N0£, (4-10)
which is sensitive to the NO^/NO ratio, PAN may serve as an important reservoir
for peroxy radicals and N02 (Singh and Salas, 1983a,b) and may play a signifi-
cant role in both the atmospheric nitrogen cycle and in tropospheric ozone
formation (Spicer et al., 1983).
Only two analytical techniques have been used to obtain significant data
on ambient PAN concentrations. These are gas chromatography with electron
capture detection (GC-ECD) and long-path Fourier-transform infrared (FTIR)
spectrometry. Atmospheric data on PAN concentrations have been obtained
predominantly by GC-ECD because of its relative simplicity and superior sensi-
tivity. These analytical techniques are described in Section 4.2.6.2 along
with attendant methods of sampling. Peroxyacetyl nitrate is somewhat analogous
to 0, in that it is a thermodynaraically unstable oxidant and PAN standards
must be generated and analyzed by some absolute technique for the purpose of
calibrating the GC-ECD. Generation and calibration techniques are discussed
separately in Section 4.2.6.3. Finally, the analysis of PAN homologues is
discussed briefly in Section 4.2.6.4.
4-33
-------
4.2.6.2 Analytical Methods for PAN. By far the most widely used technique
for the quantitative determination of ppb concentrations of PAN is GC-ECD
(Darley et al., 1963; Stephens, 1969). With-Carbowax or SE30 as a stationary
phase, PAN can be separated from components such as air, water, and other
atmospheric compounds, as well as ethyl nitrate, methyl nitrate, and other
contaminants that are present in PAN synthetic mixtures. Electron-capture
detection (using a nickel-63 source and a pulsed-current detector or a tritium
source and a direct-current detector) provides sensitivity for PAN in the ppb
and sub-ppb ranges. A typical column for the separation of PAN would be 3 to
5 feet in length and 1/8 inch in diameter (i.d.), and would be run isothermally
at 25° to 60°C. Under these conditions, a peak assignable to PAN appears
after 2 to 3 minutes. Table 4-6 shows parameters used by several investigators
to determine trace levels of PAN by GC-ECD.
Sample injection into the GC is accomplished by means of a gas-sampling
valve with a gas-sampling loop of a few milliliters volume (Stephens and
Price, 1973). Sample injection may be performed manually or automatically.
Typically, manual air samples are collected in 50 to 200 ml ungreased glass
syringes, and purged through the gas-sampling valve. Samples collected from
the atmosphere should be analyzed as soon as possible because PAN undergoes
thermal decomposition in the gas phase and at the surface of containers.
Automatic sample collection and injection may be accomplished by using a small
pump to pull ambient air continuously through the sampling loop of an automatic
sampling valve, which periodically injects the sample onto the column (Stephens
and Price, 1973). Recently, Singh and Sal as (1983b) have used cryogenic trap-
ping of PAN, with liquid argon, from relatively large air samples for the
purpose of measuring PAN concentrations in the sub-ppb range.
Most of the atmospheric PAN measurements have been made in polluted urban
environments, where maximum concentrations of 5 to 50 ppb and average concen-
trations of a few ppb may occur (Stephens, 1969; Lonneman et al., 1976). For
the purpose of such measurements, chromatographic detection limits of 0.1 to
1 ppb are sufficient. The recent work of Singh and Sal as (1983a,b) on the
measurement of PAN in the free (unpolluted) troposphere is illustrative of
current capabilities for measuring low concentrations. A 50- to 200-ml volume
of air was collected by preconcentration into an unpacked 0.15 cm o.d. stainless
tube of 1.24 ml volume, at liquid argon temperature, prior to analysis (see
Table 4-6). For measurements in humid environments, the air sample was passed
4-34
-------
TABLE 4-6. OF USED IN DETERMINATION OF PAN BY GC-ECD
Reference
Heuss and Glasson,
1968
Grosjean, 1983
Darl ey et al . ,
1963
Stephens and Price,
1973
-P>
i
CO
en
Lonneman et al. ,
1976
Holdren and Spicer,
1984
Peake and Sandhu,
1982
Singh and Sal as,
1983b
Grosjean et al . ,
1984
Nielson et al. ,
1982
Column dimensions
and materials
4 ft x 1/8 in
Glass
6 ft x 1/8 in
Teflon
3 ft x 1/9 in
Glass
1.5 ft x 1/8 in
Teflon
3 to 4 ft x 1/8 in
Glass
5 ft x 1/8 in
Teflon
3.3 ft x 1/8 in
Glass
1.2 ft x 1/8 in
Teflon
1.7 ft x 1/8 in
Teflon
3.9 ft x 1/12 in
Glass
Stationary phase
SE30
(3.8%)
10% Carbowax 400
5% Carbowax 400
51 Carbowax E 400
10% Carbowax 600
Carbowax 600
5% Carbowax 600
10% Carbowax 600
10% Carbowax 400
5% Carbowax 400
Solid support
80/100 Mesh
Diatoporte S ,
60/80 Mesh
Chrontosorb P
100/200 Mesh
Chromosorb W
Chromosorb G
80/100 Mesh
treated with
dimethyl
dichlorosilane
Gas Chrom Z
60/80 Mesh
Gas Chrom I
Chroinosorb W
80/100 Mesh
Supelcoport
60/80 Mesh
Chromosorb G
Chromosorb
W - AW - DCMS
Column
temperature ,
°C
25
30
35
25
25
35
33
33
30
25
Flow rate,
(ml/min)
NAa
40
25
60
70
70
50
30
40
40
Carrier
gas
NA
N2
Na
N2
95% Ar
5% CH4
90% Ar
10% CH4
N2
95% Ar
5% CH4
N2
N2
Elution
time,
rnin
1 NA
. NA
2.17
1.75
2.7
3.00
NA
NA
5.0
6.0
Concentration
range
ppb range
ppb range
3 to 5 ppb
37 ppb
0.1 to 100 ppb
ppb range
0.2 to 20 ppb
0.02 to 0.10 ppb
2 to 400 ppb
11 ppb
aNot available.
-------
through a Nation drier (Foulger and Simmonds, 1979) to reduce the humidity
prior to preconcentration. A minimum detection limit of 0.01 ppb was obtained.
Free tropospheric concentrations in the 0.01 to 0.1 ppb range were always
observed and indicated that PAN is a natural constituent of the atmosphere and
may constitute a significant fraction of the reactive nitrogen.
There are conflicting reports in the literature on the effects of variable
relative humidity on PAN measurements by GC-ECD. In 1973, Stephens and Price
stated that in preparing PAN calibration samples "the diluent (gas) should be
of normal humidity so that the chromatogram will be a realistic one." Subse-
quently, Holdren and Rasmussen (1976) observed a reduced response to PAN
calibration samples when the relative humidity of the sample was 30 percent or
lower and a tenfold decrease in PAN response when the relative humidity ap-
proached 0 percent. This effect was attributed to an interaction between
sample and GC column. Nieboer and van Ham (1976) reported that "the elution
gas stream was previously humidified . . . because it appeared that the height
of the PAN peak depends on the relative humidity of ambient air if dry elution
gas was used." In contrast to these studies, Lonneman (1977) observed no
effect on peak height in PAN calibration samples in which the relative humidity
varied from 10 to 50 percent.
In 1978, Watanabe and Stephens reported on a reexamination of the moisture
anomaly and investigated the effects of humidity on PAN storage flasks, columns,
and detectors. A consistent PAN loss to the walls of dry acid-washed glass
storage flasks was observed, but the PAN could be recovered by the addition of
moisture to the flasks. A tritium direct-current detector showed no humidity
effect except for a small (5 to 10 percent) decrease in peak height in a few
cases at very low humidities (2 to 3 percent). With a different GC instrument
employing a Ni-63 detector, erratic responses were observed at low humidities,
with responses reduced 30 to 95 percent from that obtained at 53 percent
relative humidity. No conclusion was drawn on whether this difference reflec-
ted a humidity effect on the detector, the column, or the sampling value.
Finally, the moisture anomaly did not appear to depend on column history or
loading even after a bake-out treatment.
In the most recent study (Grosjean et al., 1984), the humidity effect on
PAN calibration samples prepared by the dynamic and static irradiation of
CHoCHO-Clp-NOp mixtures was examined. A small decrease (<3 percent) was
observed in PAN peak height when the dry air stream was passed either over an
impinger containing water or directed through the water impinger. These
4-36
-------
results are in contradiction to all of the above, in which the response to
humidified PAN samples is either greater than or the same as dry PAN samples.
It is noteworthy that the chromatographic systems employed by different
investigators often employ different materials (e.g., glass, Teflon, stainless
steel), column loadings, and detectors. The resolution of all the differences
noted above in regard to a' suspected humidity effect might require considerable
effort. For the present, if a moisture effect is suspected in a PAN analysis,
the bulk of this evidence suggests that humidification of PAN calibration
samples (to a range approximating the humidity of the samples being analyzed)
would be advisable.
Conventional long-path infrared spectroscopy and Fourier-transform infrared
spectroscopy (FTIR) have been used to detect and measure atmospheric. PAN.
Sensitivity is enhanced by the use of FTIR over conventional long-path infrared
spectroscopy. Accurate knowledge of the absorptivities of many IR bands ,
assignable to PAN makes possible the quantitative analysis of PAN without the
use of calibration standards. The most frequently used IR bands have been
assigned and the absorptivities shown in Table 4-7 have been reported. Only
the key bands are shown, but all 27 fundamentals are infrared active and
Bruckmann and Wiliner (1983) have assigned most of them. The assignment by
Adamson and Guenthard (1980) of the bands at 1435, 1300, and 990 cm'1 to an
impurity, CHgQNQp, is apparently incorrect. Bruckmann and Willner (1983) ob-
served these same bands in a 99 percent pure PAN sample.
The initial discovery of PAN in simulated photochemical smog was accom-
plished by long-^path infrared absorption spectrometry (Stephens et al.,;1956b).
Some recent simultaneous measurements of PAN and other atmospheric pollutants
such as 03, HN03, HCOOH, and HCHO have been made by long-path FTIR spectrom-
etry during smog episodes in the Los Angeles Basin. Tuazon et al. (1978) have
described an FTIR system operable at pathlengths up to 2 km for ambient measure-
ments of PAN and their trace constituents. This system employed an eight-mirror
multiple reflection cell with a 22.5-m base path. The spectral windows avail-
able at pathlengths of 1 km were 760-1300, 2000-2230, and 2390-3000 cm"1.
Thus, PAN could be detected by the bands at 793 and 1162 cm"1. The 793 cm"1
»-"l
band is characteristic of peroxynitrates, while the 1162 cm band is reportedly
caused by PAN only (Stephens, 1969; Hanst et al., 1982). Tuazon et al. (1981a,b)
reported ambient measurements with this system during a smog episode in
Claremont, CA, in 1978. Maximum PAN concentrations ranged from 6 to 37 ppb
over a 5-day episode; the report presented diurnal patterns for PAN and several
019QQ/A ' 4-37
-------
TABLE 4-7. INFRARED ABSORPTIVITIES OF PEROXYACETYL NITRATE (BASE 10)
CO
CO
Vapor ppm""1™"1 x 104
Mode
v(c=0)
vas(N02)
vs(N02)
v(c-o)
v(o-o)
6(N02)
Frequency,
cm"1
1842
1741
1302
1162.5
930
791.5
0.1 m, no diluent
(Bruckmann and
Willner, 1983)a
12.4
32. 6d
13.6
15.8
NA
13.4
0.1 m path;
HZ at ~1 atrn h
(Stephens, 1964)°
10.0
23.6
11.2
14.3
1.8
10.1
120 m path;
air at ~1 atm h
(Stephens, 1964)°
NAC
NA
11.4
13.9
NA
10.3
Solution
pm""1 i
Frequency,
cm"1
1830
1728
1294
1153
NA
787,
in n- octane,
(Holdren and
Spicer, 1984)
0.00041
0.00115
0.00041
0. 00042
NA
0.00044
TTIR spectra 1-1.2
resolution.
Prism spectra -10 cnr* resolutioni
NA = Data not available.
Resolved Q branch.
-------
other pollutants for the 2 most severe days. The detection limit given for
PAN at a 900-m pathlength was 3 ppb.
Hanst et al. (1982) modified the FTIR system used by Tuazon et al. (1978)
by changing it from an eight-mirror to a three-mirror cell configuration and
by considerably reducing the cell volume. Measurements were made over a
1260-m optical path folded along a 23-m base path at 0.25 cm resolution.
Measurements were reported for PAN and a variety of other pollutants for a
2-day smog episode at California"State University, Los Angeles, in 1980, The
maximum PAN concentration observed was 15 ppb for this period of only moderate
smog intensity. An upper limit of 1 ppb of peroxybenzoyl nitrate (PbzN) was
placed based on observations in the vicinity of the PBzN band at 990 cm .
The reports by Tuazon et al. (1978) and Hanst et al. (1982) both refer to
earlier FTIR ambient air studies.
Sampling may constitute one of the major problems in the analysis of
trace reactive species, such as PAN, by long-path FTIR spectrometry. The
folded-path White cells have a large internal volume (15 m for Tuazon et al«,
1978; 3 m for Hanst et al., 1982). The large internal surface area may serve
to promote the decomposition or irreversible adsorption of reactive trace
species. To minimize these effects, both Tuazon et al. and Hanst et al.
employed high-speed blowers to pull ambient air through the cells at high
velocities. For interior cell linings, Hanst et al. employed 0.5 mm polyvinyl
chloride sheeting and Tuazon et al. used Plexiglas and FEP Teflon.
Pitts et al. (1973) proposed a chemiluminescence technique for continuous
monitoring of ambient concentrations of PAN. The reactions of both PAN and 0,
with triethyl amine in the gas phase produce chemiluminescence. The spectra
reported overlap somewhat with a h x of 520 nm for the 0, reaction and \_ax
of 650 nm for the PAN reaction. Pitts proposed a technique that included the
measurement of the emission intensity in the two regions by the use of optical
cut-off filters. Thus, the PAN concentration can be determined, provided the
absolute 0, concentration is simultaneously measured. Concentrations of 6 ppb
PAN were detected and a lower limit of detection of 1 ppb was estimated. No
interfering emissions were observed from methyl nitrate, ethyl nitrate, ethyl
nitrite, or N02- No further work has been reported on the development of this
technique, and there have been no atmospheric applications.
4.2.6.3 Generation and Calibration of PAN. Because of the thermal instability
of dilute PAN samples and the explosive nature of liquid PAN, calibration
samples are not commercially available; each laboratory involved in making
4-39
-------
such measurements must prepare its own standards. The PAN samples are prepared
by various means at concentrations in the ppm range and these must be analyzed
by some absolute technique. The analyzed samples must then be diluted to
obtain gas-phase samples in the low ppb range for direct calibration of GC
instruments. Thus, this section includes descriptions of various means of PAN
generation, methods of analysis, and the procedures for sample handling and
storage where applicable.
The earlier methods used for the preparation of PAN have been summarized
by Stephens (1969). These included (1) the photolysis of mixtures of nitrogen
oxides with organic compounds in air or oxygen (Stephens et a!., 1956b;
Stephens et a!., 1961); (2) the photolysis of alkyl nitrite vapor in oxygen
(Stephens et a!., 1965); (3) the dark reaction of aldehyde vapor with nitrogen
pentoxide (Tuesday, 1961); and (4) the nitration of peracetic acid. Of these
methods, the photolysis of alkyl nitrites was favored and used extensively by
Stephens and other investigators. As described by Stephens et al. (1965), the
liquefied crude mixture obtained at the outlet of the photolysis chamber is
purified by preparation-scale GC. [CAUTION: Both the liquid crude mixture
and the purified PAN samples are violently explosive and should be handled
behind explosion shields using plastic full-face protection, gloves, and a
leather coat at all times. These PAN samples should be handled in the frozen
or gaseous state whenever possible.] The pure PAN is usually diluted to about
1000 ppm in cylinders pressurized with nitrogen to approximately 100 psig.
When refrigerated at <15°C, PAN losses are less than 5 percent per month
(Stephens et al., 1965). Lonneman et al. (1976) used the photolysis products
without purification for the calibration of GC instruments in the field and
discussed the use of fedlar bags for the preparation and transport of cali-
bration samples.
Gay et al. (1976) have used the photolysis of CU: aldehyde: NOp mixtures
in air or oxygen for the preparation of PAN and a number of its homologues at
high yields:
d2 + hv > 2 Cl (4-11)
00
d + RC-H » R-C' + HC1 (4-12)
4-40
-------
00
RC- + 02 + M > RC-02' + M (4-13)
0 0
RC-02* + N02 > RC-02-N02 (4-14)
This procedure has been utilized in a portable PAN generator that can be used
for the calibration of GC-ECD instruments in the field (Grosjean, 1983;
Grosjean et al., 1984). The output of this generator is a dynamic flow of PAN
in air at a concentration of about 2 to 450 ppb. Dilute concentrations of
reactant gases for the photolysis chamber are obtained by passing a controlled
flow of air over C12, N02, and acetaldehyde permeation tubes.
The other technique for PAN preparation in current use involves the
nitration of peracetic acid. In the 1969 review (Stephens, 1969), this approach
was considered not useful for synthesis. Several investigators, however, have
recently reported on a condensed-phase synthesis of PAN with peracetic acid
that produces high yields of a pure product free of other alkyl nitrates
(Hendry and Kenley, 1977; Kravetz et al., 1980; Nielsen et al., 1982; Holdren
and Spicer, 1984). Most of these procedures call for the addition of peracetic
acid (40 percent in acetic acid) to a hydrocarbon solvent (pentane, heptane,
octane) maintained at -80°C in a dry-ice acetone bath, followed by acidification
with sulfuric acid. Nitric acid is formed in situ with stirring by the slow
addition of sodium nitrate. After the nitration is complete, the hydrocarbon
fraction, containing PAN concentrations of 2 to 4 mg/ml (Nielsen et al.,
1982), can be stored at -20°C for periods longer than a year (Holdren and
Spicer, 1984). After analysis, the PAN-hydrocarbon solutions can be used
directly for calibration by the evaporation of measured microliter volumes of
solution into Tedlar bags containing known volumes of clean air.
The most direct method for absolute analysis is by infrared absorption
using absorptivities given in Table 4-7. This is the technique used by Stephens
(1969; analysis of PAN in N2 cylinders), Lonneman et al. (1976; analysis of
gas-phase products from photolysis of ethyl nitrite); and Holdren and Spicer
(1984; analysis of PAN in octane solutions). Whereas long, folded-path cells
and FTIR spectrometry are required for the analysis of atmospheric PAN, conven-
tional IR instruments and 10-cm gas cells can analyze gas standards with
concentrations greater than 35 ppm (Stephens, 1969) and Holdren and Spicer
4-41
-------
(1984) used 50-|jm liquid microcells for the analysis of PAN in octane solutions.
Another candidate technique for absolute PAN analysis was gas-phase coulometry
using a tandem electron-capture detector (Lovelock et al., 1971), Singh and
Salas (1983b) have shown, however, that this technique is unsuitable for abso-
lute PAN analysis because a significant fraction of the PAN is destroyed prior
to coulometric detection.
The alkaline hydrolysis of PAN to acetate ion and nitrite ion in quanti-
tative yield (Nicksic et al., 1967) provides a means independent of infrared
for the quantitative analysis of PAN. Molecular oxygen is also produced in
quantitative yield by the reaction (Stephens, 1967):
00
CH3COON02 + 20H" -* CHgCO" + 0£ + N02" + HgO (4-15)
The colorimetric determination of nitrite ion with Saltzman reagent was first
used to measure PAN quantitatively (Nicksic et al., 1967; Stephens, 1969;
Kravetz et al., 1980). Nielsen et al. (1982) analyzed the hydrolyzed products
of pure PAN samples by ion chromatography for nitrite and nitrate and found
that 4 percent of the nitrite had been oxidized to nitrate. Some gas-phase
PAN calibration samples (e.g., photolysis of Cl«: acetaldehyde: N02) contain
impurities such as N02 that will yield nitrite and nitrate in aqueous solution.
Thus, Grosjean (1983) and Grosjean et al. (1984) performed ion chromatographic
analysis of the acetate ion to determine the PAN output of a portable generator.
An alternate calibration procedure has been proposed based on the thermal
decomposition of PAN in the presence of excess NO, and measurement by chemilumi-
nescence of the NO consumed (Lonneman et al., 1982). The acetylperoxy radical,
CH3C(0)02, and its decomposition products rapidly oxidize NO to NO,,. In the
presence of a small amount of benzaldehyde, which is used to scavenge the
hydroxyl radical and control the stoichiometry, simulation models predict that
5 molecules of NO will be removed per PAN molecule present. By the use of NO
and PAN standard mixtures and the chemiluminescent measurement of the NO
consumed, the experimental value was determined to be ANQ/APAN = 4.7 ± 0.2.
This measurement could be performed in field stations where chemiluminescent
NO analyzers are usually available.
4.2.6.4 Methods of Analysis of Higher Homologues. The GC-ECD analyzer is
likewise used for the higher homologues of PAN (Darley et al., 1963; Stephens,
1969; Heuss and Glasson, 1968). The higher homologues elute with longer
4-42
-------
retention times. The first observation of peroxypropionyl nitrate (PPN) in
heavily polluted air was by Darley et al. (1963), who also measured peroxybu-
tyryl (PBN) nitrate in synthetic mixtures by GC-ECD. The concentrations of
the higher homologues in ambient air are usually below the detection limits of
the GC-ECD technique. Heuss and Glasson (1968) measured peroxybenzoyl nitrate
(PBzN) in irradiated auto exhaust samples by GC-ECD and reported that this
homologue was 100 times more potent than PAN as a lachrymator. The direct
analysis of PBzN by GC-ECD is reported to be complicated by interferences
(Appel, 1973). Therefore, an analytical technique was developed in which the
PBzN was quantitatively hydrolyzed to methyl benzoate (MeOBz), followed by GC
analysis for MeOBz using a flame ionization detector (Appel, 1973). An upper
limit of 0.07 ppb was placed on the concentration of PBzN in the San Francisco
bay area. The analysis for the higher homologues of PAN in the atmosphere by
FTIR spectrometry is not feasible because of inadequate sensitivity, although
Hanst et al. (1982) placed an upper limit for PBzN in smoggy Los Angeles air
of 1 ppb based on absorption in the 990 cm region.
The higher homologues of PAN may be prepared for use in calibration in
the same manner as PAN by the use of a compound containing the parent alkyl
group. Thus, PPN and PBzN have been prepared by the photolysis of alkyl
nitrates in oxygen (Stephens, 1969) and parent aldehydes plus chlorine and N02
(Gay et al., 1976). The study of Gay et al. (1976) confirmed that the first
member of the series, peroxyformyl nitrate [HC(0)02N02], is too unstable to be
observed. There have been few reports of the absolute analyses for the higher
homologues. Infrared absorption analysis of purified samples should be the
preferred technique. Infrared absorptivities of homologues have been reported
by Stephens (1969) and Gay et al. (1976).
4.2.7 Methods for Sampling and Analysis of Hydrogen Peroxide
4.2.7.1 Introduction. Hydrogen peroxide (HpOp) is mechanistically significant
in photochemical smog as a chain terminator and as an index of the hydroperoxyl
radical (H02) concentration (Bufalini and Brubaker, 1969; Demerjian et al.,
1974). The major reaction leading to the formation of H202 is the recom-
bination of the hydroperoxyl radical (Graedel et al., 1976):
H02 + M -> H202 + 02 •+' M (4-16)
4-43
-------
Recent studies have implicated atmospheric H000 in the aqueous-phase oxidation
_n £. C.
of SOy to SO* and in the acidification of rain (Penkett et al., 1979;
Dasgupta, 1980a,b; Martin and Damschen, 1981; Overton and Durham, 1982).
One of the major problems in assessing the role of atmospheric HgO^ has
been a lack of adequate measurement methodology. Earlier measurements for
atmospheric hLCL were by the titanium colorimetric method (Gay and Bufalini,
1972a,b; Bufalini et al., 1972) and by a chemiluminescence technique (Kok et
al., 1978a,b). The reported HptL concentrations (0.01 to 0.18 ppm) are now
believed to be far too high, primarily as a result of artifact H202 formation
from reactions of absorbed 0, (Zika and Saltzman, 1982; Heikes et al., 19,82;
Heikes, 1984). Furthermore, maximum tropospheric HJ3,, concentrations pre-
dicted by modeling calculations (Chameides and Tan, 1981; Logan et al., 1981)
and observed in recent field studies (Das et al., 1983) are on the order of I
ppb. In a recent study, a chemiluminescence technique was employed with an
argon sample purge to remove 0~ interference, and a maximum H?0? concentration
of 1.2 ppb was observed in a polluted urban environment (Das et al., 1983).
Promising techniques that have been used or proposed for aqueous- and gas-phase
HpOg are discussed below, as well as methods for sampling, generation, and
standardization of H202 samples.
4.2.7.2 Sampling. Almost all of the methods used for the measurement of
atmospheric H^O^ have used aqueous traps for sampling. Midget impingers (Gay
and Bufalini, 1972a,b; Kok et al., 1978b; Das et al., 1983), continuous extrac-
tors (Kok et al., 1978a), and gas washing traps (Zika and Saltzman, 1982) have
been used. Aqueous traps have been found to be highly efficient in removing
trace concentrations of H202 from gas streams (Zika and Saltzman, 1982).
Atmospheric 03, however, which is also absorbed at concentrations much higher
than HgOp, reacts in the bulk aqueous phase and at surfaces to produce H^Op
and thus is a serious interference (Zika and Saltzman, 1982; Heikes et al.,
1982; Heikes, 1984). Details of the aqueous chemistry of QS can also be found
in other sources (Hoigne and Bader, 1976; Kilpatrick et al., 1955; Taube and
Bray, 1940). Because the rate of FLO^ production is relatively slow (Heikes,
1984), the removal of absorbed 03 by purging immediately after sample collec-
tion may remove or substantially reduce this interference. This was the
approach employed by Zika and Saltzman (1982) and by Das et al, (1983).
Another problem identified with aqueous sampling is that other atmospheric
species (in particular, S02) may interfere with the generation of H202 in
4-44
-------
aqueous traps and also react with collected Hg02 to reduce the apparent H^Op
concentration measured (Heikes et al., 1982).
4.2.7.3 Measurement. A number of methods for measuring low levels of H^CU
have been reported, including the following:
1. Titanium colorimetric methods.
2. Chemiluminescence methods.
3. Enzyme-catalyzed methods.
4. Laser diode infrared method.
5. Fourier-transform infrared method (Niki et al., 1980;
Hanst et al., 1975, 1982).
6. Electrochemical methods (Pisarevskii et al.,, 1980).
7. HpOp-olefi.n reaction (Hauser and Kolar, 1968).
8. Mixed-ligand complexes (Csanyi, 1981; Meloan et al., 1961).
9. lodometry.
Of these techniques, only the chemiluminescence and enzyme-catalyzed methods
are summarized below. A summary of methods reported in the literature is
given in Table 4-8. Although titanium colorimetric methods have been applied
in atmospheric measurements, these techniques are now thought to have inade-
quate sensitivity for the actual atmospheric concentrations that are present.
In addition, these techniques have questionable specificity. The same comments
apply to methods 6 through 8 and only the primary references are given above
for those methods. At the present state-of-art, the inevitable presence of
water vapor absorption limits the use of Fourier-transform infrared methods to
concentrations above about 0.040 ppm (Hanst et al., 1982). The use of a
tunable diode infrared laser source should eliminate the problem associated
with nearby water bands, and this method is currently under investigation for
atmospheric measurements (unpublished work in progress, Schiff, 1985). lodomet-
ric techniques are useful only for calibration of 1^2 standards and will be
discussed in that section.
4.2.7.3.1 Chemi1uminescence. Hydrogen peroxide in the atmosphere may be
detected at low concentrations by the chemi luminescence obtained from (^(ID-
catalyzed oxidation of luminol (5-amino-2,3-dihydro-l,4-phthalazinedione) by
H«00 (Armstrong and Humphreys, 1965; Kok et al., 1978a,b). The reagent is a
-5
solution containing NaOH (pH = 12.8) and luminol, 10 M Cu(II). The products
of the reaction with HpOg are 3-amino-phthalic acid, a nitrogen molecule, and
4-45
-------
TABLE 4-8. FOR PfROXIDE
,, Method
Titanium
colorimetry
Chewi luminescence
Enzyme-cata 1 yzed
Enzyme-catalyzed
Enzyme-catalyzed
-P>
i Fourier-transform
5J infrared absorption
Electrochemical
H202-o1efin
reactions
Mixed-ligand
complex reagents
Reagent(s)
(1) Titanium Sulfate
-8-Quinolinol
(2) Titaniui Tetrachloride
Luminol, Cu(II)
basic solution
Scopoletin, horseradish-
peroxidase (HRP)
Leuco crystal violet,
HRP
3- (p-hydroxyphenyl )
propionic acid
None
Aqueous solutions
l,2-di-(4-pyridyl)
ethyl ene
Vanadium and
uranium hydroxamic
acid chelates
Limits of
detection
(1) 1.6 x 10-6 H
(2) ca 10-° H
0.001 to 1 ppa
1.5 x 10-11 M
10-s H
10-8 to 10-4 M6
0,040 ppm (est.)
5 x 10-6 to 1 M
10-6 to 5 x 10-4 M
10-6 M
Interferences
Positive
Alkyl hydro-
peroxides
PANd
NA
NA
NA
NAf
NA
OB
NA
Negative
S02C
S02
NA
NA
NA
None
NA
NA
NA
Applications Primary reference
Atmospheric (1) Gay and Bufalini (1972a,b)
(2) Pilz and Johann (1974);
Kok et al. (1978a)
Atmospheric, Armstrong and Humphreys (1965);
rainwater Kok et al. (1978a,b)
Atmospheric, Andreae (1955); Perschke
rainwater and Broda (1961); Zika and
Saltzman (1982)
Mottola et al. (1970)
Zaitsu and Okhura (1980)
Atmospheric3 Hanst et al. (1982)
Pisarevskii et al. (1980)
Hauser and Kolar (1968)
Csanyi (1981);
Heloan et al. (1961)
Except where noted, detection limits are in moles/1iter(M) in aqueous solution.
QS is an interference in all these procedures using aqueous sampling. See Text. NA = not available.
cThe S02 interference is reported to be small at high S02 concentrations (Gay and Bufalini, 1972b). Studies of potential positive and
negative interferences are incomplete for these methods.
The reported PAN interference is small (Kok et al., 1978b).
eThe lower limit could presumably be reduced by the use of larger samples.
With sufficient resolution, there should be no interferences. IR absorption by atmospheric water vapor is the major analytical
limitation.
^Oz bands have not been observed in any long-path FTIR studies. The estimated,lower limit of detection in these studies is
approximately 0.04 ppm.
-------
a photon of light at 450 nm. The detection limit for atmospheric samples was
given as 0.001 ppm, and the linear dynamic range is 0.001 to 1 ppm. This
technique as initially employed suffered the interferences from 0^ and SC^
discussed above for aqueous traps. Das et al. (1982) employed a static version
of the method of Kok et al. (1978a) to measure H?0? concentrations in the 0.01
to 1 ppb range. In addition, samples were purged with argon immediately after
collection to eliminate, reportedly, the 0^ interference.
Recently, a modified chemiluminescence method has been reported that used
hemin, a blood component, as a catalyst for the lumino!-based hLOp oxidation
(Yoshizumi et al., 1984). This method was applied to the measurement of Op
in rainwater. There was no interference for SOp but a significant positive 0,
interference was reported.
4.2.7.3.2 Enzyme-Catalyzed Methods (Peroxidase). This general method involves
three components: a substrate that is oxidizable; the enzyme, horseradish
peroxidase (HRP); and hLOg. The production or decay of the fluorescence in-
tensity of the substrate or reaction product is measured as it is oxidized by
HpOg, catalyzed by HRP. Some of the more widely used chromogenic substrates
have been scopoletin (6-methoxy-7-hydroxy-l,2-benzopyrone) (Andreae, 1955;
Perschke and Broda, 1961); 3-(p_-hydroxyphenyl)propionic acid (HPPA) (Zaitsu
and Okhura, 1980); and leuco crystal violet (LCV) (Mottola et al., 1970).
In the scopoletin method, the reagent solution is mixed with a second
solution containing the 1^2, the concentration of which must not be less than
0.33 nor more than 0.84 times the concentration of scopoletin (Perschke and
Broda, 1961). The disappearance of scopoletin fluorescence is monitored and
the fluorescence intensity can be used to obtain the concentration of H^O^
from a calibration curve. The most significant advantages of this method of
analysis are the specificity for H^Op, the sensitivity, and the stoichiometry
of the scopoletinrH^Op reaction (1:1 mole per mole as long as scopoletin is
present in a 20-fold excess over HRP). The chief disadvantage of the scopoletin
method is that the concentration of H^Oo must be within a narrow range in
order to obtain an accurately measurable decrease in fluorescence. This
limits the usefulness of the technique in determining unknown HgOg concentra-
tions over several orders of magnitude (Armstrong and Humphreys, 1965; Andreae,
1955). Detection limits for this technique are quite low and are in the range
of 1.5 x 10"11 M (Perschke and Broda, 1961) to 2 x 10"10 M (Andreae, 1955).
This technique has been applied to measurements of H202 in rainwater by Zika
and Saltzman (1982).
4-47
-------
With the leuco crystal violet (LCV) substrate, intensely colored crystal
violet is formed from the reaction of HpOp with LCV, catalyzed by HRP (Mottola
et al., 1970). The absorbance is measured at 596 nm, where the absorption
coefficient of crystal violet is 10 M cm , a very high value and an inherent
advantage of this method. The concentration of HpOp is a linear function of
the concentration of crystal violet produced. The detection limit reported is
_Q
about 10 M HpOp for an absorbance of 0.005 in a 5-cm cell.
The HRP catalyzes the oxidation of a wide variety of hydrogen-donating
substrates by HpOg. Zaitsu and Okhura (1980) have tested a number of 4-hydroxy-
phenyl compounds and found that 3-(p_-hydroxyphenyl) propionic acid (HPPA) pro-
vided the most sensitive and rapid means for determining H202. When HPPA
reagent solution is mixed with HRP solution and a test solution containing
H202, a product is formed that fluoresces at 404 nm following excitation at
320 nm. The intensity of this fluorescence is monitored as a function of H^O,,
_in £• t-
concentration. The detection limit was reported to be 10 mole H000 with a
-8
linear range extending to 10 mole HpOo when a test solution of only 0.1 ml
volume was used. Presumably the molar sensitivity could be improved by the
use of larger sample volumes. No interference studies were reported. More
recently, the acetic acid homologue of HPPA has been employed (Kunen et al.,
1983; Dasgupta and Hwang, 1985). The fluorescence intensity of the product
dimer is directly proportional to H/,0,, concentration.
The enzymatic methods appear to be the most promising aqueous, colori-
metric methods for H202 and have considerably greater sensitivity than other
colorimetric methods. Studies of potential atmospheric interferences, however,
have not been reported for any of these three substrates.
4.2.7.4 Generation and Calibration Methods
4.2.7.4.1 Generation. Standard samples containing trace concentrations of
H202 are required for testing and calibration of various measurement methodol-
ogies. As with DO, such standards are not available and are usually prepared
at the time of use. A number of techniques have been employed for generating
aqueous standards, but convenient methods for the generation of gas-phase
standards are noticeably lacking.
Techniques for the generation of high concentrations of H202 have been
discussed by Shanley (1951). Commercial solutions of 30 percent aqueous H202
are readily available. Trace levels of H009 in water may be generated by the
fiO
irradiation of water with CO ^-radiation (Hochanadel, 1952; Armstrong and
Humphreys, 1965) and by enzymatic means (Andreae, 1955). By far the most
4-48
-------
convenient method for generating aqueous standards containing micromolar
concentrations of HpOp is simply the serial dilution of commercial-grade.30
percent Og (Fisher Analytical Reagent). Samples prepared in this manner
must be standardized and the method usually employed is the iodometric tech-
nique discussed below. Stock standard solutions of H^00 as low in concentra-
_4 ^ £
tion as 10 M have been found to be stable for many weeks if kept in the dark
(Armstrong and Humphreys, 1965),
Techniques for the convenient generation of gas-phase standards are not
available. With the increased interest in atmospheric H«Q,,, there is,an
obvious need for an hLO^ generator comparable to the photolytic 0^ generator
discussed in Section 4.2.4.1. A technique that has often been used for gener-
ating ppm concentration levels of H^O^ in air has been described by Cohen and
Purcell (1967). Micro!iter quantities of 30 percent \\J^^ solution are injected
into a metered stream of air that flows into a Teflon bag. The concentration
of HgQp in the bag is then determined by the iodometric method discussed
below.
4.2.7.4.2 Calibration. By far the most common method for standardizing low
concentrations of H^O* is based on iodometry (Allen et al., 1952; Hochanadel,
1952; Cohen et al., 1967). Hydrogen peroxide liberates iodine from an iodide
solution quite slowly, but in the presence of a molybdate catalyst the reaction
is quite rapid. The iodine liberated may be determined by titration with
standard thiosulfate at higher concentrations or by photometric measurement of
the tri-iodide ion at low concentrations. The molar absorption coefficient of
the tri-iodide ion at 350 nm has been determined to be 2.44 x 10 by measuring
mmfc _jfl
10 to 10 M H202 solutions prepared from 0.2 M stock H202 solution standard-
ized against primary grade arsenious oxide (Armstrong and Humphreys, 1965).
The stoichiometry is apparently 1 mole of iodine released per mole of HpOo-
Definitive studies of the stoichiometry, however, have not been performed to
the same extent as those of the stoichiometry for the iodometric determination
of03.
4.3 SAMPLING, MEASUREMENT, AND CALIBRATION METHODS FOR PRECURSORS TO OZONE
AND OTHER PHOTOCHEMICAL OXIDANTS
During the last decade, a number of advances have been made in the method-
ology for determining nonmethane organic compounds and oxides of nitrogen. An
overview of these advances is discussed in the following sections. In the case
of measurement methods for nonmethane organic compounds (NMOCs), early ozone
4-49
-------
control models did not require speciation of the complex mixture of organics
in ambient air. As a result, non-speciation methods were employed for the
purpose of providing a data base for early photochemical modeling studies. As
the air quality models grew more sophisticated, however, the need arose for
more specific information concerning the organic composition of the atmosphere.
Consequently, methodology was developed to provide for detailed speciation of
NMOCs. In addition to improving the data base for photochemical modeling, the
NMOC speciation techniques have also been utilized to characterize various
sources of pollution (e.g.3 mobile versus stationary).
The development of methodology for oxides of nitrogen has likewise ad-
vanced since the original EPA Federal Reference Method for measurement of
nitrogen dioxide, the Jacobs-Hochheiser technique, was withdrawn in 1973. A
number of methods for nitric oxide and nitrogen dioxide have been proposed and
evaluated since then and are described here.
4.3.1 Nonmethane Organic Compounds
Numerous sampling, analytical, and calibration methods have been employed
to determine vapor-phase NMOCs in ambient air. Some of the analytical methods
utilize detection techniques that are highly selective and sensitive to specific
functional groups or atoms of a compound (e.g., formyl group of aldehydes,
halogen), while others respond in a more universal manner; that is, to the
number of carbon atoms present in the organic molecule. In this overview of
the most pertinent measurement methods, NMOC have been arranged in three major
classifications: nonmethane hydrocarbons, aldehydes, and other oxygenated
compounds. Sampling, analytical, and calibration procedures are discussed for
each classification. Reference is also made to those analytical methods
utilized in more than one of the above classifications.
4.3.1.1 Nonmethane Hydrocarbons. Nonmethane hydrocarbons (NMHC) constitute
the major portion of NMOC in ambient air (Chapter 3). Traditionally, NMHC
have been measured by methods that employ a flame ionization detector (FID) as
the sensing element. This detector was originally developed for gas chromato-
graphy and employs a sensitive electrometer that measures a change in ion
intensity resulting from the combustion of air containing organic compounds.
Ion formation is essentially proportional to the number of carbon atoms present
in the organic molecule (Sevcik, 1975). Thus, aliphatic, aromatic, alkenic,
and acetylenic compounds all respond similarly to give relative responses of
1.00 ± 0.10 for each carbon atom present in the molecule (e.g., 1 ppm hexane =
4-50
-------
6 ppm C; 1 ppm benzene = 6 ppm C; 1 ppm propane = 3 ppm C). Carbon atoms
bound to oxygen, nitrogen, or halogens give reduced relative responses (Dietz,
1967). Consequently, the FID, which is primarily used as a hydrocarbon measur-
ing method, should more correctly be viewed as an organic carbon analyzer.
In the following sections, discussion focuses on the various methods
utilizing this detector to measure total nonmethane organics. Methods in
which no compound speciation is obtained are covered first. Methods for
determining individual organic compounds are then discussed.
4.3.1.1.1 Non-speciation methods. The EPA reference method for nonmethane
organic compounds, which was promulgated in 1971, involves the gas chromato-
graphic separation of methane (CH,) from the remaining organics in an ,air
sample (F.R., 1971). A second sample is injected directly to the detector
without methane separation. Subtraction of the first value from the second
produces a nonmethane organic concentration.
A number of studies of commercial analyzers employing the Federal Refer-
ence Method have been reported (Reckner, 1974; McElroy and Thompson, 1975;
Harrison et al., 1977; Sexton et a!., 1982). In one of the first studies,
analyses of known synthetic mixtures of hydrocarbons were conducted by 16
users of the reference method (Reckner, 1974). The nonmethane concentrations
tested in this study were 0.23 and 2.90 ppm C. The results shown in Table 4-9
indicate the percentage error from the two known concentrations. At the 0.23
ppm level, the majority of the measurements were in error by amounts greater
than 50 percent. At 2.90 ppm, most of the measurements were in error by only
20 percent or less.
In general, all of the above studies indicated an overall poor performance
of the commercial instruments when either calibration or ambient mixtures
containing NMOC concentrations less than 1 ppm C were used. The major problems
associated with using these NMOC instruments have been reported in a recent
technical assistance document (U.S. Environmental Protection Agency, 1981).
The technical assistance document also suggests ways to reduce the effects of
existing problems. Table 4-10 summarizes the major problem areas and lists
recommendations for reducing these effects.
Other approaches to the measurement of nonmethane organics have also been
investigated. One such method, developed in 1973, utilizes the fact that CH*
requires more heat for combustion than other organics (Poll and Zinn, 1973).
One portion of the air sample passes through a catalyst bed where all hydro-
carbons except CH^ are combusted. This sample stream then enters an FID where
4-51
-------
TABLE 4-9. PERCENTAGE DIFFERENCE FROM KNOWN CONCENTRATIONS
OF NONMETHANE HYDROCARBONS OBTAINED BY SIXTEEN USERS
Known % difference from given concentration
ppm
0.23
2.90
>100
6
2
50 to 100
4
0
20 to 50
3
3
10 to 20
2
2
0 to 10
1
9
Source: Reckner (1974).
the CH* concentration alone is recorded. The remaining portion of the sample
passes directly to a second FID for a total organic carbon measurement. The
simultaneous processing of both signals yields an NMOC value. Although it
provides a continuous measurement of NMOC levels, this method is also subject
to many of the same shortcomings attributed to the EPA reference method.
Recently, a prototype instrument that measures NMOCs by optical absorption
has been developed (Manos et a!., 1982). The unit oxidizes NMOCs to carbon
dioxide (CO*) and uses a non-dispersive infrared absorption technique to
measure the organic burden indirectly. Ascarite serves to remove C02 initially
present in air and a hopcalite catalyst selectively oxidizes organics other
than methane to COg and H20. Since carbon monoxide (CO) will also oxidize to
COg during this process, a dual-channel system is utilized to correct for the
contribution from ambient CO concentrations. This unit performed well during
a brief laboratory evaluation using calibration standards; however, more
extensive laboratory and field tests are needed before the unit can be con-
sidered suitable for NMOC measurement.
Other methods under development and evaluation include oxidation-reduction
schemes in which nonmethane organics are chromatographically separated from
methane and non-organic species and then oxidized to C02, reduced to CH«, and
detected by FID (U.S. Environmental Protection Agency, 1978b). When organic
carbon concentrations are greater than 100 ppb, the reduction step can be
eliminated and a non-dispersive infrared analyzer can be used to detect the C02
formed during the oxidation step (Salo et a!., 1975).
A method for measuring NMOC directly has been reported by Cox et al.
(1982). This approach involves the cryogenic preconcentration of nonmethane
organic compounds and the measurement of the revolatilized NMOCs using flame
4-52
-------
TABLE 4-10. PROBLEMS ASSOCIATED WITH GATHERING NMOC DATA
WITH AUTOMATED ANALYZERS AND RECOMMENDATIONS FOR REDUCING THESE EFFECTS
Problem Areas:
1. Contaminants may be present in compressed-gas cylinders containing
calibration gases.
2. Compressed-gas cylinders of calibration gases sometimes contain the
standard in a nitrogen or argon background. When no oxygen is blended
with these gases, FID sensitivity is altered.
3. The assay of calibration gases contained in compressed-gas cylinders (as
received from the supplier) is sometimes incorrect.
4. There are wide differences in the per-carbon response to different NMOC
species.
5. FID analyzers require hydrogen, which presents a potential operational
hazard.
6. The NMOC concentration is obtained by subtraction of two relatively large
and nearly equal numbers (TOC-CH«=NMOC) and thus is subject to large,5 rela-
tive errors.
7. NMOC analyzers may exhibit excessive zero and span drift during unattended
operation.
8. The complex design of some NMOC analyzers creates unique problems that are
generally not experienced in other pollutant analyzers. Meticulous set-up,
calibration, and operation procedures (which are analyzer-specific) are
difficult to understand and follow.
i i
Recommendations: . ,
1. Calibration gases should be checked to determine the concentration of
contaminants. i
2. Calibration concentrations should be obtained by dynamic dilution of a
pollutant standard with zero-grade air containing oxygen. The dilution
ratio should be sufficiently high (vLOO:l) to ensure that the calibration
sample contains 20.9% ± 0,3% oxygen.
3. All calibration standards contained in compressed-gas cylinders should be
traceable to Standard Reference Materials from the National Bureau of
Standards.
4. The NMOC response should be calibrated to a propane standard.
5. The operator should use documented procedures for hydrogen safety.
6. All channels should be properly calibrated.
7. The FIDs should be operated in accordance with instructions supplied by
the manufacturer and this document.
8. The training of qualified operators should be augmented with a Technical
Assistance Document, which provides detailed calibration and operation
procedures for NMOC analyzers.
Source: U.S. Environmental Protection Agency (1981).
4-53
-------
ionization detection. The procedure involves the following steps. A fixed
volume of sample is drawn through a trap cooled to liquid argon temperature
(liquid Hy cannot be used since it will also condense methane and oxygen). At
this temperature all NMOCs are condensed into the trap. After the residual
CH* and oxygen are cleared from the trap by the helium carrier gas, the trap
temperature is raised to revolatilize the NMOC. Oxygen removal reduces the
variation in response to different organic compounds.
Jayanty et al. (1982) used a similar system to study the responses of a
variety of aliphatic and aromatic compounds. Good linearity was observed with
the cryogenic trapping procedure (500 ml of air) over a concentration range of
50 to 5,000 ppb C. Humidity did not generally interfere with the analysis.
Sample precisions of ±5 percent for single- and multiple-component gas standards
and ±10 percent for ambient samples were consistently achieved. Responses for
aromatic compounds, however, were less than expected. The researchers recom-
mended additional testing and instrument refinement in order to resolve this
problem.
4.3.1.1.2 Speciation methods. The primary separation technique utilized for
NMOC speciation is gas chromatography (GC). Coupled with flame ionization
detection, this analytical method permits the separation and identification of
many of the organic species present in ambient air.
Compound separation is accomplished by means of both packed and capillary
GC columns. If high resolution is not required and large sample volumes are
to be injected, packed columns are employed. The traditional packed column
may contain either (1) a solid polymeric adsorbent (gas-solid chromatography)
or (2) an inert support, coated with a liquid (gas-liquid chromatography).
Packed columns containing an adsorbent substrate are required to separate
C^-Co compounds. The second type of column can be a support-coated or wall-
coated open tubular capillary column. The latter column has been widely used
for environmental analysis because of its superior resolution and broader
applicability. The wall-coated capillary column consists of a liquid station-
ary phase coated or bonded (cross-linked) to the specially treated glass or
fused-silica tubing. Fused-silica tubing is most commonly used because of its .
physical durability and flexibility. When a complex mixture is introduced
into a GC column, the carrier gas (mobile phase) moves the sample through the
packed or coated capillary column (stationary phase). The chromatographic
process occurs as a result of repeated sorption-desorption of the sample
4-54
-------
components (solute) as they move along the stationary phase. Separation
results from the different affinities that the solute components have for the
stationary phase.
The GC-FID technique has been used by numerous researchers to obtain
ambient NMOC data. Singh (1980) drew on the cumulative experience of these
researchers to prepare a guidance document for state and local air pollution
agencies interested in obtaining speciation data. In general, most researchers
have employed two gas chromatographic units to carry out analyses of NMOC
species in ambient air. The more volatile organic compounds (Cp through Cg)
are generally measured on one unit using packed-column technology, while the
other GC separates the less volatile organics using a capillary column. In
typical chromatograms of urban air, all major peaks are identified and, on a
mass basis, represent from 65 to 90 percent of the measurable nonmethane
organic burden. Identification of GC peaks is based upon matching retention
times of unknown compounds with those of standard mixtures. The use of dedi-
cated computer systems facilitates this task, but close scrutiny of the data
is still necessary to correct periodic misidentification of unknown compounds
resulting from variations in retention time. Subsequent verification of the
individual species is normally accomplished with gas chromatographie/mass
spectrometric (GC/MS) techniques. Compound-specific detection systems,, such
as electron capture, flame photometry, and spectroscopic techniques, have also
been employed for peak identifications. A discussion of these systems, however,
is beyond the scope of this report. Several documents covering these detection
systems are available (Lamb et al., 1980; Riggin, 1983).
Because the organic components of the ambient atmosphere are present at
ppb levels or lower, some means of sample preconcentration is necessary to
provide sufficient material for the GC-FID system. The two primary techniques
utilized for this purpose are the use of solid adsorbents and cryogenic collec-
tion. The more commonly used sorbent materials are divided into three catego-
ries: (1) organic polymeric adsorbents, (2) inorganic adsorbents, and (3)
carbon adsorbents. Primary organic polymeric adsorbents used for NMOC analyses
include the materials Tenax GC and XAD-2. These materials have a low retention
of water vapor and, hence, large volumes of air can be collected. These
materials do not, however, efficiently capture highly volatile compounds such
as C« to Cg hydrocarbons, nor certain polar compounds such as methanol and
acetone. Primary inorganic adsorbents are silica gel, alumina, and molecular
sieves. These materials are more polar than the organic polymeric adsorbents
4-55
-------
and are thus more efficient for the collection of the more volatile and polar
compounds. Unfortunately, water is also efficiently collected, which in many
instances leads to rapid deactivation of the adsorbent. Carbon adsorbents are
less polar than the inorganic adsorbents and, as a result, water adsorption by
carbon adsorbents is a less significant problem. The carbon-based materials
also tend to exhibit much stronger adsorption properties than organic polymeric
adsorbents; thus, lighter-molecular-weight species are more easily retained.
These same adsorption effects result, however, in irreversible adsorption of
many compounds. Furthermore, the very high thermal desorption temperatures
required (350 to 400 °C) limit their use and also may lead to degradation of
labile compounds. The commonly available classes of carbon adsorbents include:
(1) various conventional activated carbons; (2) carbon molecular sieves (Sphere-
carb, Carbosphere, Carbosieve); and (3) carbonaceous polymeric adsorbents
(Ambersorb XE-340, XE-347, XE-348).
Although a number of researchers have employed solid adsorbents for the
characterization of selected organic species in air, only a few attempts have
been made to identify and quantitate the range of organic compounds from C~
and above. Westberg et al. (1980) evaluated several carbon and organic poly-
meric adsorbents and found that Tenax-GC exhibited good collection and recovery
efficiencies for >Cg organics; the remaining adsorbents tested (XAD-4, XE-340)
were found unacceptable for the lighter organic fraction. The XAD-4 retained
>C« organic gases, but it was impossible to completely desorb these species
without partially decomposing the XAD-4. Good collection and recovery efficien-
cies were provided by XE-340 only for organics of C* and above.
Ogle et al. (1982) used a combination of adsorbents in series and designed
an automated GC-FID system for analyzing C^ through C^g hydrocarbons. Tenax
GC was utilized for Cg and above, while Carbosieve S trapped C, through Cg
organics. Silica gel followed these adsorbents and effectively removed water
vapor while passing the C2 hydrocarbons onto a molecular-sieve 5A adsorbent.
The combined sorbents have been laboratory-tested with a SB-component hydro-
carbon mixture. Good collection and recovery efficiencies were obtained.
Preliminary field tests have also been successful, but a very limited data
base exists. Futhermore, the current chromatographic procedures utilize
packed-column technology. The addition of capillary columns to this system
would permit better resolution of the complex mixtures typically found in
ambient air.
4-56
-------
The preferred method for obtaining NMOC data is cryogenic preconcentra-
tion (Singh, 1980). Sample preconcentration is accomplished by directing air
through a packed trap immersed in either liquid oxygen (b.p. -183°C) or liquid
argon (b.p. -186°C). For the detection of about 1 ppb C of an individual
compound, a 250-cc air sample is normally processed. The collection trap is
generally filled with deactivated 60/80 mesh glass beads (Westberg et a!.,
1974), although coated chromatographic supports, have also been used (Lonneman
et al., 1974). Both of the above cryogens are sufficiently warm to allow air
to pass completely through the trap, yet cold enough to collect trace organies
efficiently. This collection procedure also condenses water vapor. An air
volume of 250 cc at 50 percent relative humidity and 25°C contains approxi-
mately 2.5 mg of water, which appears as ice in the collection trap. The
collected ice at times will plug the trap and stop the sample flow; furthermore,
water transferred to the capillary column during the thermal des.orption step
occasionally causes plugging and other deleterious column effects. To circum-
vent water condensation problems, McClenny et al. (1984) have employed a Nafion
tube drier to remove water vapor selectively during the sample collection
step. These researchers have constructed an automated cryogenic preconcentra-
tion sampling and analysis GC system using this drier and are currently conduc-
ting field evaluaton studies on their system.
In addition to direct sampling via preconcentration with sorbents and
cryogenic techniques, collection of whole air samples is frequently used to
obtain NMOC data. Rigid devices such as syringes, glass bulbs, or metal
containers and non-rigid devices such as Tedlar and Teflon plastic bags are
often utilized during sampling. The primary purpose of whole-air collection
is to store an air sample temporarily until subsequent laboratory analysis is
performed. The major problem with this approach is assuring the integrity of
the sample contents prior to analysis. Of concern is whether sample components
of interest are adsorbed or decomposed through interaction with the container
walls or reaction with co-collected gases such as ozone and nitrogen dioxide.
Sample condensation may also occur at elevated concentrations or when samples
are stored under high pressures (i.e., in metal containers). Contamination
from sampling containers is likewise a frequent occurrence (Lonneman et al.,
1981; Seila et al., 1976). Table 4-11 summarizes the advantages and disadvan-
tages of the primary collection media for NMOC analysis,,
4.3.1.1.3 Calibration. Calibration procedures for NMOC instrumentation
require the generation of dilute mixtures at concentrations expected to be
4-57
-------
TABLE 4-11. OF ADVANTAGES AMD DISADVANTAGES OF PRIMARY COLLECTION MEDIA FOR NMOC ANALYSIS
Sampling technique
Advantages
Disadvantages
1. Solid adsorbents
en
00
2. Cryogenic pre-
concentration
Many sorbents do not retain H20
vapor; thus, large volumes of air
can be processed.
Integrated samples from a period
of minutes to days are easily
obtained.
Sample cartridges are convenient
for field use.
o A wide range of organic material
can be collected,
o Artifact problems are avoided.
o Collected organics are immediately
available for analysis, without
solvent removal or use of high
thermal desorption temperatures.
o Collected species are stable; good
recovery efficiencies are obtained.
No single adsorbent can be used to
collect and recover organics of G£
and above.
Contamination and artifact peaks
plague many sorbent systems.
o Many adsorbents require high
(>300°C) thermal desorption tem-
peratures, which may lead to
degradation of labile compounds,
artifact peak formation, etc.
o Breakthrough volume and collection
and recovery efficiencies must be
determined for each compound of
interest.
o Volume of air collected is limited
by amount of moisture condensing.
o Liquid argon or oxygen is necessary
for preconcentration.
o Field collection apparatus is bulky
compared to adsorbent cartridges.
-------
TABLE 4-11 (continued). SUMMARY OF ADVANTAGES OF PRIMARY COLLECTION MEDIA FOR NMOC ANALYSIS
Sampling technique
Advantages
Disadvantages
•p*
en
3. Rigid containers
(Metal canisters)
4. Non-rigid containers
(Teflon and Tedlar
Bags)
o Can be treated to make them
chemically unreactive.
o Durable; easy to clean, transport,
and use.
o Can be pressurized; leakage and
permeation minimized.
o Excellent stability for many trace
species; long-term storage is
possible.
o Readily available.
o Convenient for collecting integrated
samples.
o Good short-term stability of trace
species.
o High initial cost.
o Limited collection volume.
Difficult to collect integrated
samples.
Sample may be lost to walls by
condensation.
o Subject to breakage (at seams)
during handling.
o Admits sunlight.
o Slow permeation of chemicals out
of and into plastic bags during
storage.
o Outgassing contamination from bag
material.
o Short storage life.
Source: Derived from Singh (1980); Jayanty et al. (1982); Sexton et al
(1976); Lonneman et al. (1981); Holdren et al. (1982).
(1982); National Research Council
-------
found in ambient air. Methods for generating such mixtures are classified as
static or dynamic systems.
Static systems are generally preferred for quantitating NMOCs, The most
commonly used static system is a compressed gas cylinder containing the appro-
priate concentration of the compound of interest. These cylinder gases may
also be diluted with hydrocarbon-free air to provide multi-point calibrations.
Calibration and hydrocarbon-free air cylinders are available commercially.
Additionally, some standard gases such as propane and benzene are available
from the National Bureau of Standards (NBS) as certified standard reference
materials (SRM). Commercial mixtures are generally referenced against these
NBS standards. In its recent technical assistance document for operating and
calibrating continuous NMOC analyzers, the U.S. Environmental Protection
Agency (1981) recommended propane-in-air standards for calibration. Some
commercially available propane cylinders have been found to contain other
hydrocarbons (Cox et a!., 1982), so that all calibration data should be refer-
enced to NBS standards. Because of the uniform carbon response of a GC-FIO
system (±10 percent) to hydrocarbons (Dietz, 1967), a common response factor
is assigned to both identified and unknown compounds obtained from the specia-
tion systems. If these compounds are oxygenated species, an underestimation
of the actual concentrations will be reported. Dynamic calibration systems
are employed when better accuracy is needed for these oxygenated hydrocarbon
species. Dynamic systems are normally employed to generate jji situ concentra-
tions of the individual compound of concern and include devices such as permea-
tion and diffusion tubes and syringe delivery systems.
4.3.1.1.4 Comparison of non-speciationversus speciation methods. Speciation
methods involving cryogenic preconcentration have been compared with non-speci-
ation NMOC analyzers in the following studies.
Jayanty et al. (1982) conducted a laboratory comparison between the pro-
totype non-speciation method described earlier (Section 4.3.1.1.1) and their
gas chromatographic separation method. Comparison of the two methods for 12
ambient air samples collected in stainless steel canisters showed agreement
within ±15 percent. Ambient air concentrations ranged from 100 to 1000 ppb C.
Lonneman (1979) compared total NMOC and speciation methods during field
studies in Houston in 1978. Samples were collected during 3-hour integrated
time periods (6 to 9 a.m., 1 to 4 p.m.) in Tedlar bags for subsequent analysis.
The correlation coefficients for 150 measurement pairs from five sites averaged
0.74. For data pairs of 500 ppb C and less, an average correlation coefficient
4-60
-------
(r) of 0.55 was calculated, with a low value of 0.12 at one site. Lonneman
attributed the low correlations to maintenance and calibration problems in the
continuous analyzers and concluded that the results from the, non-speciation
method are "at best marginal for use in photochemical model applications."
Holdren et al. (1982) made a similar comparison during a 2^month study at
urban sites in Cincinnati arid Cleveland, Ohio. They utilized a GC/cryogenic
trapping technique and compared their results with data from state-operated
NMOC analyzers (SAROAD data). Ambient air samples were collected in Teflon
bags (6 to 9 a.m. integrated collection) and were transferred immediately to
pretreated aluminum cylinders for shipment and analysis at the central labora-
tory. Concentrations of NMOC ranged from 200 to 2100 ppb C. Linear regres-
sion analyses of the non~speciated versus speciated data resulted in correla-
tion coefficients that ranged from 0.75 to 0.92 for the four urban sites
(total of 67 comparisons). Limiting the comparisons to concentrations of 500
ppb C and lower resulted in an average correlation coefficient of 0,10.
Richter (1983) compared continuous total NMOC with GC speciation results
obtained at seven fixed ground-level sites used in the Northeast Corridor
Regional Modeling Project (NECRMP). The NMOC data were obtained in real time,
while Teflon bags were used to collected integrated samples (6 to 9 a.m.) for
the GC/cryogenic analyses. Over 60 comparisons were available from each site.
Table 4-12 summarizes statistical information obtained from least-squares
analysis of the data (Richter, 1983). As the table indicates, only data from
the East Boston site exhibited a high correlation coefficient. This study
represents the most extensive effort made yet to compare the two NMOC measuring
methods. Emphasis was placed on correct instrument operation, calibration,
etc.; and only verified data were compared. Yet the above results indicate that
much more work is needed to resolve the differences between the two methods.
4.3.1.2 Aldehydes. Historically, the major problem in measuring concentrations
of aldehydes in ambient air has been to find an appropriate monitoring technique
that is sensitive to low concentrations and specific for the various homologues.
Early techniques for measuring formaldehyde, the most abundant aldehyde, were
subject to many interferences and lacked sensitivity at low ppb concentrations.
The more recently developed techniques can be utilized to accurately measure
the various types and amounts of aldehydes at ppb levels. This section de-
scribes those methods currently used for measuring aldehydes in ambient air.
4-61
-------
TABLE 4-12. GC/CONTINUOUS NMOC ANALYZER COMPARISONS,
LEAST-SQUARES REGRESSIONS
Standard ,
Location Slope Intercept, ppm C error r
West End Library, 0.552 -0.552 0.672 0.169
Washington, DC
Read Street, 0.113 -0.283 0.713 0.0077
Baltimore, MD
Essex, MD
Linden, NJ
Newark, NJ
East Boston, MA
Watertown, MA
0.835
0.531
0.987
1.108
0.750
-0.101
NAa
-0.277
+0.095
-0.568
0.599
0.865
0.574
0.327
0.574
0.354
0.141
0.467
0.887
0.475
aData not available.
Source: Richter (1983).
These include the chromotropic acid (CA) method for formaldehyde, the 3-methyl-
2-benzothiazolone hydrazone (MBTH) technique for total aldehydes, Fourier-trans-
form infrared (FTIR) spectroscopy, and the high-performance liquid ehromato-
graphy (HPLC) method employing 2,4-dinitrophenylhydrazine (DNPH) derivatization.
4.3.1.2.1 Chromotropic acid method. The chromotropic acid method (CA) involves
the collection of formaldehyde in a midget impinger containing an aqueous
mixture of chromotropic and sulfuric acids, followed by the spectrophotometric
measurement of absorbance of the resulting color (Altshuller and McPherson,
1963; U.S. Dept. of Health, Education and Welfare, 1965). A modification
described by Johnson et al. (1981) improved the accuracy and sensitivity of
the CA method by reducing the concentration of sulfuric acid and by eliminating
a heating cycle, relying solely on the heat of solution generated by sulfuric
acid (Altshuller et al., 1961; Olansky and Deming, 1976). Trapping formalde-
hyde in a 1 percent bisulfite solution before adding the CA solution increased
collection efficiency from 84 percent to 92 percent with no sulfite interfer-
ences.
4-62
-------
The CA method for measuring formaldehyde has been widely studied (Salas
and Singh, 1982; Grosjean and Kok, 1981; National Research Council, 1981;
Tuazon et al., 1980; Lloyd, 1979). Originally developed as a spot-test by
Eegriwe (1937), it was adopted to quantitate formaldehyde spectrophotometri-
cally (Bricker and Johnson, 1945; West and Sen, 1956) and was modified for
ambient air measurements in the early 1960s (Altshuller et al., 1961; Altshuller
and McPherson, 1963; U.S. Dept. of Health, Education, and Welfare, 1965).
While used widely today for both occupational and ambient air environments,
its specificity for formaldehyde, which accounts for approximately half of
total ambient air aldehydes (see Section 3.5), limits its usefulness for
characterizing aldehyde concentrations in ambient air.
The CA method has been reported to be sensitive to acrolein, acetaldehyde,
phenol, nitrogen dioxide, and nitrates (National Research Council, 1981; Krug
and Hirt, 1977; U.S. Dept. of Health, Education, and Welfare, 1973; Sleva,
1965; Altshuller et al., 1961). Recent work, however, indicates that neither
nitrates, nitrites, NO^, nor acetaldehyde at elevated ambient air levels
interfere with the CA analysis (Johnson et al., 1981; Grosjean and Kok, 1981).
Relevant data on other interfering agents were not found.
4.3.1.2.2 MBTH method. A spectrophotometric technique for total aldehydes
was developed in the early 1960s by Sawicki et al. (1961). Known as the MBTH
method, it involves the reaction of aldehyde with 3-methyl-2-benzothiazolone
hydrazone to form an azine that is oxidized by a ferric chloric-sulfamic acid
solution to form a blue cationic dye (Altshuller, 1983; U.S. Dept. of Health,
Education, and Welfare, 1965; Hauser and Cummins, 1964; Altshuller and McPherson,
1963; Altshuller and Leng, 1963; Altshuller et al., 1961).
The MBTH method has a reported sensitivity of 15 ppb for, primarily,
low-molecular-weight aldehydes (National Research Council, 1981). The method
is subject to interferences by N02 and gives an inconsistent response to
higher-molecular-weight aldehydes (Sawicki et al., 1961; Altshuller etal.,
1961). Nonetheless, the Intersociety Committee of the American Public Health
Association recommends the MBTH colorimetric method for determining total
aldehydes in air (Intersociety Committee, 1977a). Miksch and Anthon (1982)
devised a sampling and analysis scheme that permits a single MBTH sample to be
used for both formaldehyde and total aliphatic aldehyde determinations.
4.3.1.2.3 Fourier-transform infrared spectroscopy. Infrared absorption
spectroscopy has been used to identify and measure aldehydes in ambient air
(Hanst et al., 1975, 1982; Tuazon et al., 1978, 1980, 1981a). These studies
4-63
-------
employ Fourier-transform infrared (FTIR) spectrometers interfaced to multiple
reflection cells operating at total optical paths of up to 1 km. At such
pathlengths, a detection limit of a few ppb is achieved for formaldehyde. The
advantages of the long-path! ength FTIR method for ambient air aldehyde measure-
ments (i.e., good specificity and direct jhn situ analysis) are offset by the
relatively high cost and lack of portability of such systems.
4.3.1.2.4 High-performance 1 iquid chromatography (HPLC) 2,4-dim'trophenylhydra-
zlne (DNPH) method. This method takes advantage of the reaction of carbonyl
compounds with 2,4-dinitrophenylhydrazine to form a 2,4-dinitrophenylhydrazone:
RR'C=0 + NH2NHCgH3(N02)2 — »• H20 + RR'C=NNHCgH3(N02)2 (4-17)
Since DNPH is a weak nucleophile, the reaction is carried out in the presence
of acid in order to increase protonation of the carbonyl.
The HPLC-DNPH method is the preferred way of measuring aldehydes in
ambient air. Atmospheric sampling is usually conducted with micro-irnpingers
containing an organic solvent and aqueous, acidified DNPH reagent (Papa and
Turner, 1972; Katz, 1976; Smith and Drummond, 1979; Fung and Grosjean, 1981).
After sampling is completed, the hydrazone derivatives are extracted and the
extract is washed with deionized water to remove the remaining acid and unre-
acted DNPH reagent. The organic layer is then evaporated to dryness, subse-
quently dissolved in a small volume of solvent, and analyzed by reversed-phase
liquid chromatographic techniques employing an ultraviolet (UV) detection
system. Analysis by a flame ionization detection (FID) system has proved less
successful than UV because the derivatives are not really amenable to GC-FID
analysis.
An improved procedure has been reported that is much simpler than the
above aqueous impinger method (Lipari and Swarin, 1982; Kuntz et a!,, 1980;
Tanner and Meng, 1984). This scheme utilizes a midget impinger containing an
acetonitrile solution of DNPH and an acid catalyst. After sampling, an aliquot
of the original collection solution is directly injected into the liquid
chromatograph. This approach eliminates the extraction step and several
sample-handling procedures associated with the DNPH-aqueous solution; and
provides much better recovery efficiencies. Lipari and Swarin (1982) have
reported detection limits of 20, 10, 5, and 4 ppb for formaldehyde, acetalde-
hyde, acrolein, and benzaldehyde, respectively, in 20-liter air samples.
These researchers have also developed a newer technique employing a solid
4-64
-------
adsorbent cartridge containing Florisil coated with 2,4-dinitrophenyldrazine
(Lipari and Swarin, 1985). Collection efficiencies greater than 90 percent
were obtained at low ppb levels, A detection limit of 1 ppb of HCHO for a
100-liter air sample was achieved.
4.3.1.2.5 Calibrationof aldehyde measurements. Since they are reactive
compounds, it is extremely difficult to make stable calibration mixtures of
aldehydes in pressurized gas cylinders. Although gas-phase aldehyde standards
are available commercially, the vendors do not guarantee any level of accuracy.
Formaldehyde standards are generally prepared by one of several methods.
The first method utilizes dilute commercial formalin (37 percent formaldehyde,
w/w). Calibration is accomplished by the direct spiking into sampling impin-
gers of the diluted mixture or by evaporation into known test volumes, followed
by impinger collection. Formaldehyde can also be prepared by heating known
amounts of paraformaldehyde, passing the effluent gases through a methanpl-
liquid nitrogen slush trap to remove impurities, and collecting the remaining
formaldehyde. Paraformaldehyde permeation tubes have also been used (Tanner
and Meng, 1984).
For the higher-molecular-weight aldehydes, liquid solutions can be evap-
orated or pure aldehyde vapor can be generated in dynamic gas-flow systems
(permeation tubes, diffusion tubes, syringe delivery systems, etc.). These
test atmospheres are then passed through the appropriate aldehyde collection
system and analyzed. A comparison of these data, with the direct spiking of
liquid aldehydes into the particular collection system, provides a measure of
the overall collection efficiency.
4.3.1.2.6 Comparison of measurement methods. Several side-by-side comparisons
of the chromotropic acid method (CA) with other methods have been reported.
Grosjean and Kok (1981) compared the CA method (Johnson et al., 1981) with
HPLC-DNPH (Fung and Grosjean, 1981) and FTIR spectroscopy (Tuazon et al.,
1978). They found fairly close agreement between the CA and HPLC-DNPH methods,
but noted consistently higher results with FTIR. Corse (1981) sampled ambient
air with CA (U.S. Dept. of Health, Education, and Welfare, 1965), MB'TH (U.S.
Dept. of Health, Education, and Welfare, 1965), and HPLC-DNPH methods (Kuntz
et al., 1980). An examination of tabulated data from the Corse study shows a
consistent and considerable difference between CA and HPLC measurements. For
25 CA measurements, formaldehyde averaged 8.8 ppb; while HPLC measurements
from the same sampling train averaged 14.2 ppb. Overall, formaldehyde, levels
were approximately 60 percent higher with HPLC than with CA measurements.
4-65
-------
Because blanks were not utilized for the HPLC analyses, however, the HPLC data
are subject to uncertainty, since blank corrections can affect results substan-
tially (Altshuller, 1983a). During laboratory studies, Kuntz et al. (1980)
reported reasonable agreement (±7 percent} among the HPLC-DNPH, CA, FTIR, and
GC-FID methods when low ppb levels of formaldehyde, acetal dehyde , propional-
dehyde, hexanal , and benzal dehyde were generated.
4.3.1.3 Other Oxygenated Organic Species. As mentioned earlier, unsubsti-
tuted hydrocarbons constitute the major fraction of vapor-phase organic com-
pounds occurring in ambient air. Aldehydes as a class of volatile organics
appear second in abundance. With the exception of formic acid (Hanst et al.,
1982; Tuazon et al., 1981a, 1980, 1978), other oxygenated species are seldom
reported. The lack of data for oxygenated hydrocarbons is somewhat surprising
since significant quantities of these species are emitted directly into the
atmosphere by solvent-related industries (methanol, ethanol, acetone, etc.;
see Chapter 3). Along with manmade emissions, natural sources of oxygenated
hydrocarbons also contribute to this total. In addition to these direct
emissions, it is also expected that photochemical reactions of hydrocarbons
with oxides of nitrogen, ozone, and hydroxyl radicals will produce significant
quantities of oxygenated products.
Difficulties in sample collection and analysis may account for this lack
of data. The adsorptive nature of the surfaces that contact these oxygenated
species often complicates the process of compound quantisation. The approach
used for analysis of oxygenated and other polar organic compounds is to decrease
adsorption by deactivating the interior surfaces of analytical hardware. A
novel method has been reported, however, in which the reactive compounds of
interest are modified rather than the surfaces with which these compounds
interact (Osman et al., 1979; Westberg et al., 1980). In these studies, the
laboratory derivatization of vapor-phase alcohols and acids (silylation) was
investigated to evaluate the potential of such a procedure for stabilizing
these polar compounds prior to analysis. Results indicate that silylation
procedures greatly reduce adsorption of alcohols and acids and that, qualita-
tively, the silylated derivatives can be detected via the GC-FID system.
4.3.2 Nitrogen Oxides
In highly polluted urban air, the two most abundant oxides of nitrogen
compounds are nitric oxide (NO) and nitrogen dioxide (NO). Both compounds,
together designated as NO , participate in the photooxidation reactions in the
/s,
4-66
-------
atmosphere that lead to the formation of ozone (Chapter 3). Other minor
reactive oxides of nitrogen in ambient air include peroxyacyl nitrates, nitrogen
trioxide, dinitrogen pentoxide, and peroxynitric acid. In rural areas, and in
urban areas late in the day and at night, these species may constitute a
significant fraction of the total oxides of nitrogen in ambient air (Singh and
Hanst, 1981; Kill us and Whitten, 1985).
Analytical methods for NOp and NO are briefly described in this section.
Older methods, including the former Federal Reference Method (Jacobs-Hochheiser
method), are described in the criteria document on nitrogen oxides prepared by
the U.S. Environmental Protection Agency (1982).
4.3.2.1 Measurement Methods forNQg and NO. In 1976, the continuous chemi-
luminescence method was promulgated as the new Federal Reference Method
(F.R., 1976c). This method measures atmospheric concentrations of N02 indirectly
by first reducing or thermally decomposing the gas quantitatively to NO (with
a converter), reacting NO with 0.,, and measuring the light intensity from the
chemiluminescent reaction (Fontijn et a!., 1970). Two types of converters have
been employed for converting NOg to NO. In carbon or ferrous sulfate converters,
NO/, is chemically reduced to NO with the concurrent oxidation of converter
material. Thermal converters, such as stainless or molybdenum steel converters,
spontaneously decompose N02 to NO at elevated temperatures. The type and
severity of interfering species will depend upon the converter used and the
temperature of operation. The NO in the air stream is measured separately and
subtracted from the previous NO (NO plus NQ«) measurement to yield the NQ2
concentration. Typical commercial chemiluminescence instruments can detect
levels as low as 2.5 ng/m3 (0.002 ppm) (Katz, 1976).
While all oxides of nitrogen and organic nitrogen compounds are thermody-
namically unstable with respect to the formation of NO, the rate of the conver-
\
sion to NO is infinitesmally slow under normal conditions. In the presence of
reducing agents or at elevated temperatures, however, the conversion may
become quite rapid. A number of studies have shown for example, that some
species found in ambient air can undergo reduction by converters, resulting in
positive interferences. Winer et al. (1974), for example, found that peroxy-
acetyl nitrate (PAN) and various nitrogen compounds are reduced by the converter
to NO and that nitroethane and nitric acid are partially reduced in the system
when a carbon (reducing) converter or a molybdenum (thermal decomposition)
converter is used. Joshi and Bufalini (1978) reported positive interferences
from halocarbons when a heated carbon converter was used; they also suggested
4-67
-------
that stainless steel converters may be subject to interferences from chlorinated
hydrocarbons. Other evidence suggests that ammonia (NH3) may be converted to
NO in high-temperature thermal converters (U.S. Environmental Protection
Agency, 1982). Positive interferences resulting from the presence of PAN and
HNO« on afternoons of high oxidant concentrations can exceed 30 percent of the
NOg concentrations (Spicer et a!., 1979). Grosjean (1982) also reported that
positive interferences from nitric acid and PAN during NOg analysis by chemi-
luminescence can cause a 50 to 60 percent NO, overestimation during smog
conditions in Los Angeles. In less severe smog, the overestimation should not
be this high.
Other acceptable methods for measuring ambient NQ2 levels, including two
methods designated as equivalent methods, are the Lyshkow-modified Griess-
Saltzman method, the instrumental colorimetric Griess-Saltzman method, the
triethanolamine method, the sodium arsenite method, and the TGS-ANSA method
[TGS-ANSA = triethanolamine, guaiacol (crmethoxyphenol), sodium metabisulfite,
and 8-anilino-l-naphthalene sulfonic acid]. The sodium arsenite method and
the TGS-ANSA method were designated as equivalent methods in 1977. While some
of these methods measure the species of interest directly, others require
oxidation, reduction, or thermal decomposition of the sample, or separation
from interferences, before measurement. These methods have been described in
the 1982 criteria document on nitrogen oxides (U.S. Environmental Protection
Agency, 1982).
In addition to the wet chemical methods for measuring N02, other tech-
niques have been investigated. Maeda et al. (1980) have reported a new chemi-
luminescence method based on the reaction of NO, with luminol (5-amirio-2,3~
dihydro-l,4-phthalazine dione), with a detection limit of about 50 parts per
trillion (ppt) and linearity over a range of 0.5 ppb to 100 ppm. Work is
under way by Maeda and coworkers to remove the interferences of 0^ and SOp.
Wendel et al. (1983) have reported the development of a lumino!-based instrument
for the continuous measurement of N02 in ambient air. In the early instrument
developed by Maeda et al. (1980), response time was slow and the luminol solution
pool in the detector cell showed chemiluminescence long after NO^ was removed.
The Wendel et al. (1983) instrument has rapid response time and luminol contacts
the air sample via a filter paper strip rather than a pool of liquid. The
method has a detection limit of 30 parts per trillion and has been freed of 03
interference through modifications to the inlet system and the addition of
4-68
-------
Na/jSQo to the luminol solution. Other workers have endeavored to improve
chemiluminescence analyzers through physical- modifications (e.g., Ridley and
Hewlett, 1974; Schiff et al., 1979).
Lipari (1984) has developed a solid-sorbent method for measuring N0» in
ambient air. The NOj, is quantitatively collected by means of a commeridally
available air sampling cartridge containing Florisil (magnesium silicate)
coated with diphenylamine (DPA). The N02 reacts with DPA to form 2-nitro-4-
nitro-, and N-nitroso-DPA derivatives. The products are eluted from the
cartridge with methanol, acidified with a 1 N HC1 catalyst to ensure recovery
of the DPA derivatives, and analyzed by HPLC-UV. No interferences from NO, '
0,5 HNO,, S02 and humidity were found, but PAN was found to produce a 50%
positive interference. Storage stability tests indicated that cartridge
blanks are stable for at least 3 months and that samples can be transported,
stored, and analyzed for at least a period of 4 weeks after collection without
significant sample loss or degradation. The reported limit of detection is
0.1 ppb NO,, for a 2000-liter air sample, which corresponds to an 8-hr sampling
period at a flow rate of 4 liters/min.
Molecular correlation spectrometry, in which an absorption band of a
sample is compared with a corresponding band stored in the spectrometer, has
been applied in analysis of N02 (Williams and Kolitz, 1968). Instruments
processing the second derivative of sample transmissivity have also been used
(Hagar and Anderson, 1970), as have infrared lasers and infrared spectrometers
(Hanst, 1970; Hinkley and Kelley, 1971; and Kreuzer and Patel, 1971). Tucker
et al. (1973, 1975) reported on instruments based on the principle of laser-
induced fluorescence at optical frequencies. Fincher et al. (1977) described
detection of 1 ppb N0? with a technique based on fluorescence by a pulsed
xenon flashlamp. Long-pathlength differential optical absorption spectroscopy
has also been employed to monitor N0? in the troposphere (Platt et al., 1980,
1984).
The preferred approach for measuring NO is also the continuous chemilumi-
nescence method. Other methods for measuring NO directly include ferrous
sulfate absorption and spectrophotometric measurement of the resulting ion
(Norwitz, 1966), ultraviolet spectroscopy (Sweeny et al., 1964), infrared
spectroscopy (Lord et al., 1975), and ultraviolet fluorescence (Okabe and
Schwartz, 1975). Mass spectrometry and gas chromatography may also be employed.
4-69
-------
4.3.2.2 Sampling Requirements. When sampling for NO , long residence times
in sampling lines should be avoided. In the ambient air, the rate of photoly-
sis of N09 (forming NO and 0, and thus 0.,) is almost equal to the rate of the
£» o
reaction of the NO and 03 to form N0«, In sampling lines, photolysis stops
but NO continues to react with Q3, producing NOp. The magnitude of the dark
reaction of NO with 03 depends, of course, on the concentrations of NO and 03
in the sample being analyzed, as well as on the residence time of the sample
in the line. This dark reaction has practical consequences only under certain
conditions. In moderately polluted urban areas, steady-state concentrations
of NO are almost certainly too low at the period of maximal 03 to cause signi-
ficant errors in obtaining NOX or 03 measurements. Conversely, when 03 is at a
minimum, as in the early morning or possibly even in the late afternoon, NO
(and NOp) may be at maximal levels, and no significant errors would be intro-
duced. If the concentrations of NO and 03 are both low as in some rural
areas, or during those brief periods in polluted areas when NO and 0, diurnal
patterns cross, then measurement errors could be introduced. Such errors are
not likely to be significant, however.
Techniques for limiting sampling errors sampling to given levels of
tolerance are reviewed by Butcher and Ruff (1971). In general, only glass or
Teflon materials should be used in sampling trains. Among absorbents, granules
impregnated with triethanolamine are reported to be the best, converting only
2 to 4 percent of the incoming NOp to NO (Intersociety Committee, 1977b;
Huygen, 1970). The most frequently used oxidizer is chromic oxide on a fire-
brick granule support (Intersociety Committee, 1977b; Levaggi et al., 1974).
4.3.2.3 Calibration. Calibration procedures for NO measurements methods are
critical for obtaining accurate analyses. Measurement methods for NO and NOp
are calibrated (1) by sampling a gas stream of known concentration prepared
from standard reference materials (SRMs); and (2) in the case of chemilumines-
cence analyzers, by determining converter efficiency. The SRM for NO is a
cylinder of compressed NO in N2 (50 and 100 ppm). The initial accuracy and
the stability with time of this mixture were found to be quite good in a study
conducted at the National Bureau of Standards (Hughes, 1975). An accuracy of
1 percent of the labeled concentrations was obtained as determined from pressure
or gravimetric measurements. Concentration changes of less than 1 percent
were observed over a 7-month period. The SRM for NOp is the N0« permeation
tube (O'Keeffe and Ortman, 1966; Scaringelli et al., 1970). These'tubes are
4-70
-------
calibrated by determining the permeation rate by means of weight-Toss measure-
ments at a constant temperature; or, in some cases, by means of micromanometric
measurements. Both the NO and NOg SRMs are commercially available from sources
traceable to SRMs maintained at the National Bureau of Standards.
Procedures for preparing calibrated gas streams call for an accurately
measured dilution of the output of the NO and NOp SRMs. Likewise, the flow
from the NO-in^ cylinder must be accurately measured, as well as the diluent
air flow. The flow of diluent air over the NO, permeation tube, maintained at
constant temperature, must also be accurately measured. An alternative proce-
dure commonly employed for the calibration of NO, N09, and NO measurements is
f— s\
the use of gas-phase titration (Rehme et a!., 1974). A constant flow of 0., is
added to a diluted gas stream of NO at a higher concentration and the reaction
mixture is monitored with an NO monitor, e.g., chemiluminescence. Because of
the straightforward bimolecular reaction,
NO + 03 •* N02 + 02 " (4-18)
the NO consumed is equivalent to the 0- added and the N00 produced. Thus, if
-------
large existing data base that employed measurements for "total oxidants,"
non-specific iodometric techniques are discussed and compared to current
specific (U measurements.
4.4.1.1 Quality Assurance and Sampling. A quality assurance program is
employed by the U.S. Environmental Protection Agency for assessing the accuracy
and precision of monitoring data and for maintaining and improving the quality
of ambient air data. Procedures and operational details have been prescribed
in each of the following areas: selection of analytical methods and instrumen-
tation (i.e., reference and equivalent methods); method specifications for
gaseous standards; methods for primary and secondary (transfer standards)
calibration; instrumental zero and span check requirements, including frequency
of checks, multiple-point calibration procedures, and preventive and remedial
maintenance requirements; procedures for air pollution episode monitoring;
methods for recording and validating data; and information on documenting
quality control (U.S. Environmental Protection Agency, 1977b). Design criteria
for DO monitoring stations, to help ensure the quality of aerometric data,
have been established (U.S. Environmental Protection Agency, 1977a; National
Research Council, 1977).
4.4.1.2 Measurement Methods for Total Oxidantsand Ozone. Techniques for the
continuous monitoring of total oxidants and 0^ in ambient air are summarized
in Table 4-13. The earliest methods used for routinely monitoring oxidants in
the atmosphere were based on iodometry. When atmospheric oxidants are absorbed
in potassium iodide (KI) reagent, the iodine produced,
03 + 31" + H20 -»• I3~ + 02 + 20H" (4-18)
is measured by ultraviolet absorption in colorimetric instruments and by
amperometric means in electrochemical instruments. The term "total oxidants"
is of historical significance only and should not be construed to mean that
such measurements yield a sum of the concentrations of the oxidants in the
atmosphere. The various oxidants in the atmosphere react to yield iodine at
different rates and with different stoichiometries. Only ozone reacts immedi-
ately to give a quantitative yield of iodine. As discussed below, the total
oxidants measurement correlates fairly well with the specific measurement of
ozone, except during periods when significant nitrogen dioxide (NO^) and
sulfur dioxide (S02) interferences are present. The major problem with the
total oxidants measurement was the effect -of these interferences. Total
4-72
-------
TABLE 4-13. SUMMARY OF OZONE MONITORING TECHNIQUES
Principle
Continuous
colorimetric
Continuous
electrochemical
-p> Chemi luminescence
i
Chemi luminescence
Ultraviolet
photometry
Reagent
10(20)% KI
buffered at
pH = 6.8
2% KI
buffered at
pH = 6.8
Ethyl ene,
gas-phase
Rhodamine-B
None
Response
Total
oxidants
Total
oxidants
03-specific
03-specific
03-specific
Minimum
detection limit
0.010 ppm
0.010 ppm
0.005 ppm
0.001 ppm
0.005 ppm
Response ,
time, 90% FSa
3 to 5 minutes
1 minute
< 30 seconds
< 1 minute
30 seconds
Major
interferences
N02(+20%, 10%l
S02(-100%)
N02(+6%)
S02(-100%)
None"
None
Species that
absorb at 254
References
(I) Littman and Benoliel (1953)
Tokiwa et al. (1972)
Brewer and Mil ford (1960)
Mast and Saunders (1962)
Tokiwa et al. (1972)
Nederbragt et al. (1965)
Stevens and Hodgeson (1973)
Regener (1960, 1964)
Hodgeson et al. (1970)
Bowman and Horak (1972)
nm
aFS = full response.
A signal enhancement of 3 to 12% has been reported for measurement of 03 in humid versus dry air (California Air Resources Board, 1976).
cNo significant interferences have been reported in routine ambient air monitoring. If abnormally high concentrations of species that
absorb at 254 nm (e.g., aromatic hydrocarbons and mercury vapor) are present, some positive response may be expected. If high aerosol
concentrations are sampled, inlet filters must be used to avoid a positive response.
-------
oxidants instruments have now been replaced by specific ozone monitors in all
aerometric networks and in most research laboratories. Biases among and
between total oxidants and ozone methods are still important, however, for
evaluating existing data on health and welfare effects levels where concentra-
tions were measured by total oxidants methods.
The reference method promulgated by EPA for compliance monitoring for
ozone is the chemiluminescence technique based on the gas-phase ozone-ethylene
reaction (F.R., 1971). The technique is specific for ozone, the response is a
linear function of concentration, detection limits of 0.001 to 0.005 ppm are
readily obtained, and response times are 30 seconds or less. Prescribed
methods of testing and prescribed performance specifications that a commercial
analyzer must meet in order to be designated as a reference method or an
equivalent method have been published by EPA (F.R., 1975b). An analyzer may be
designated as a reference method if it is based on the same principle as the
reference method and meets performance specifications. An acceptable equiva-,
lent method must meet the prescribed performance specifications and also show
a consistent relation with the reference method.
The designated equivalent methods are based on either the gas-solid
chemiluminescence procedure or the ultraviolet absorption procedure (Table
4-13). The first designated equivalent method was based on ultraviolet absorp-
tion by ozone of the mercury 254 nm emission line. The measurement is in
principle an absolute one, in that the ozone concentration can be computed
from the measured transmission signal since the absorption coefficient and
pathlength are accurately known. In the gas-solid chemiluminescence analyzer,
the reaction between ozone and Rhodamine-B adsorbed on activated silica pro-
duces chemiluminescence, the intensity of which is directly proportional to
ozone concentration.
4.4.1.3 Calibration Methods. All the analyzers discussed above must be
calibrated periodically with ozonized air streams, in which the ozone concen-
tration has been determined by some absolute technique. This includes the
ultraviolet (UV) absorption analyzer, which, when used for continuous ambient
monitoring, may experience ozone losses in the inlet system because of contami-
nation.
A primary ozone calibration system consists of a clean air source, ozone
generator, sampling manifold, and means for measuring absolute ozone concentra-
tion. The ozone generator most often used is a photolytic source employing a
mercury lamp that irradiates a quartz tube through which clean air flows at a
4-74
-------
controlled rate (Hodgeson et al., 1972). Once the output of" the generator has
been calibrated by a primary reference method, it may be used to calibrate 0~
transfer standards, which are portable generators, instruments, or other
devices used to calibrate analyzers in the field. Reference calibration
procedures that have been used for total oxidants and ozone-specific analyzers
in the United States are summarized in Table 4-14.
The original reference calibration procedure promulgated by EPA was the
1 percent neutral buffered potassium iodide (NBKI) method (F.R., 1971). This
technique was employed in most of the United States, with the exception of
California. The California Air Resources Board (CARB) (1976) and the Los
Angeles Air Pollution Control District (LAAPCD) employed different versions of
iodometric techniques. Procedural details of the calibration methods are
summarized in Table 4-14. A number of studies conducted between 1974 and 1978
revealed several deficiencies with KI methods, the most notable of which were
poor precision or inter!aboratory comparability and a positive bias of NBKI
measurements relative to simultaneous absolute UV absorption measurements.
The positive bias observed is peculiar to the use of phosphate buffer in the
NBKI techniques. The bias was not observed in the unbuffered LAAPCD method
(which nevertheless suffered from poor precision), nor in the 1 percent EPA KI
method without phosphate buffer (Hodgeson et al., 1977), nor in a KI procedure
that used boric acid as buffer (Flamm, 1977). A summary of results of these
prior studies was presented in the previous criteria document (U.S. Environ-
mental Protection Agency, 1978a) and in a review by Burton et al. (1976).
Correction factors for converting' NBKI calibration data to a UV photometry
basis are given in Table 4-14 and discussed in Section 4.2.4.2.1,
Subsequently, EPA evaluated four alternative reference calibration proce-
dures and selected UV photometry -on the basis of superior accuracy and precision
and simplicity of use (Rehme et al., 1981). In 1979 regulations were amended
to specify UV photometry as the reference calibration procedure (F.R., 1979e).
Laboratory photometers used as reference systems for absolute 03 measurements
have been described by DeMore and Patapoff (1976) and Bass et al. (1977).
These laboratory photometers contain long path cells (1 to 5 m) and
employ sophisticated digital techniques for making effective double beam
measurements of small absorbancies at low ozone concentrations. A primary
standard UV photometer is one that meets the requirements and specifications
given in the revised ozone calibration procedures (F.R., 1979e). Since these
are currently available in only a few laboratories, EPA has allowed the use of
4-75
-------
TABLE 4-14. OZONE CALIBRATION TECHNIQUES
Method
1% NBKI
2% NBKIC
1% Unbuffered
KI
UV photometry
Gas-phase
titration (GPT)
1% BAKI
Reagent
1¥ KT
X»fc M ,
phosphate buffer
pH = 6.8
2% KI
phosphate buffer
pH = 6.8
1% KI
pH = 7
None
Nitric oxide
standard reference
gas
•w KT
-Lfc M,
boric acid buffer
pH = 5
Primary standard3
Reagent grade
arsenious oxide
Reagent grade
potassium biiodate
03 absorptivity at
Hg 254 nm emission
line
Nitric oxide SRM
(50 ppm in N2)
from NBS .
Standard KI039
solutions
Organization
and dates
EPA
1971-1976
CARB
until 1975
LAAPCD
until 1975
All
1979-present
EPA, States
1973-present
EPA
1975-1979
Bias,
Purpose Ws] -/[03]
Primary reference 1.12 ± 0.05
procedure -
Primary reference 1.20 ± 0.01
procedure
Primary reference 0.96
procedure
Primary reference
procedure
Alternative reference 1.030 ± 0.015f
procedure (1973-1979)
Transfer standard (1979-present)
Alternative reference 1.00 ± 0.05
procedure • .
In the case of the iodometric methods, the primary standard is the reagent used to prepare or standardize iodine solutions.
The uncertainty limits represent the range of values obtained in several independent studies.
cPre-humidified air used for the ozone source. .
Only one study available (DeMore et al., 1976). , -
eUV photometry used as reference method by CARB since 1975. This technique used as an interim, alternative reference procedure by
EPA from 1976 to 1979.
This is the value reported in the latest definitive study (Fried and Hodgeson, 1982). Previous-studies reported biases fpng,ing::Trom
0 to 10 percent (Burton et al., 1976; Paur and McElroy, 1979). , .'; ,: : '" ;,, :• f: ;t; ^ -k
procedure also recommended a standard I3 solution absorptivity to be used instead of the preparation of standard iodine solutions.
-------
transfer standards, which are devices or methods that can be calibrated against
a primary standard and transferred to another location for calibration of 0,
analyzers. Examples of transfer standards that have been used are commercial
DO photometers, calibrated generators, and gas-phase titration (GPT) apparatus.
Guidelines on transfer standards have been published by EPA (McElroy, 1979).
4.4.1.4; Relationships of Total Oxidants and OzoneMeasurements. The temporal
and quantitative relationship between simultaneous total oxidants and ozone
measurements has been examined in this chapter because of the existence of a
data base obtained by total oxidants measurements. Such a comparison is com-
plicated by the relative scarcity of simultaneous data, the presence of both
positive (NOg) and negative ($02) interferences in total oxidants measurements
of ambient air, and the change in the basis of calibration. In particular,
the presence of NOp and SOp interferences prevent the establishment of a
definite quantitative relationship between ozone and oxidants data. Neverthe-
less, some interesting conclusions can be drawn and are summarized below.
Studies concluded in the early to mid-1970s were reviewed.,in the previous
criteria document (U.S. Environmental Protection Agency, 1978a). Averaged data
showed fairly good qualitative and quantitative agreement between diurnal
variations of total oxidants and ozone. For example, uncorrected monthly
averaged data from Los Angeles and St. Louis showed distinct morning and
evening peaks resulting from NOp interference (Stevens et al., 1972a,b). The
most recent comparison in the literature involved simultaneous ozone and total
oxidant measurements in the Los Angeles basin by the California Air Resources
Board (1978) in 1974, 1976, and 1978. The maximum hourly data pairs were
correlated (Chock et al., 1982) and yielded the following regression equation
for 1978, in which a large number (927) of data pairs were available:
Oxidant (ppm) = 0.870 03 + 0.005
(Correlation coefficient = 0.92) (4-19)
The oxidant data were uncorrected for NOg and SO, interferences, and on individ-
ual days maximum oxidant averages were both higher than and lower than ozone
averages.
In summary, specific ozone measurements agree fairly well with total oxi-
dants corrected for N0? and S02 interferences, and in such corrected total
oxidants measurements ozone is the dominant contributor. Indeed, it is diffi-
cult to discern the presence of other oxidants in corrected .total oxidant
4-77
-------
data. Without corrections there can be major temporal discrepancies between
ozone and oxidants data, which are primarily a result of oxidizing jmd reducing
interferences with KI measurements. As a result of these interferences, on
any given day the total oxidant values may be higher than or lower than simul-
taneous ozone data. The measurement of ozone is a more reliable Indicator
than total oxidant measurements of oxidant air quality.
4.4.1.5 Methods for Sampling and Analysis of Peroxyacetyl Nitrate andIts
Homologues. Only two analytical techniques have been used to obtain signifi-
•cant data on ambient peroxyacetyl nitrate (PAN) concentrations. These are gas
chromatography with electron capture detection (GC-ECD) and long-path Fourier
transform infrared (FTIR) spectrometry. Atmospheric data on PAN concentrations
have been obtained predominantly by GC-ECD because of its relative simplicity
and superior sensitivity. These techniques have been described in this chapter
along with attendant methods of sampling, PAN generation, absolute analysis,
and calibration.
By far the most widely used technique for the quantitative determination
of ppb concentrations of PAN and its homologues is GC-ECD (Darley et al.,
1963; Stephens, 1969). With carbowax or SE30 as a stationary phase, PAN,
peroxypropionyl nitrate (PPN), peroxybenzoyl nitrate (PBzN), and other homo-
logues (e.g., peroxybutyryl nitrate) are readily separated from components such
as air, water, and other atmospheric compounds, as well as ethyl nitrate,
methyl nitrate, and other contaminants that are present in synthetic mixtures.
Electron-capture detection provides sensitivities in the ppb and sub-ppb
ranges. Typically, manual air samples are collected in 50- to 200-ml ungreased
glass syringes and purged through the gas-sampling valve. Continuous analyses
are performed by pumping ambient air through a gas sampling loop of an auto-
matic valve, which periodically injects the sample onto the column. Samples
collected from the atmosphere should be analyzed as soon as possible because
PAN and its homologues undergo thermal decomposition in the gas phase and at
the surface of containers. The recent work of Singh and Salas (1983a,b) on
the measurement of PAN in the free (unpolluted) troposphere (see Chapter 5) is
illustrative of current capabilities for measuring low concentrations. A
minimum detection limit of 0.010 ppb was obtained.
The literature contains conflicting reports on the effects of variable
relative humidity on PAN measurements by GC-ECD. If a moisture, effect is
4-78
-------
suspected in a PAN analysis, the bulk of this evidence suggests that humidiff-
cation of PAN calibration samples (to a range approximating the humidity of
the samples being analyzed) would be advisable.
Conventional long-path infrared spectroscopy and Fourier-transform in-
frared spectroscopy (FTIR) have been used to detect and measure atmospheric
PAN. Sensitivity is enhanced by the use of FTIR. The most frequently used IR
bands have been assigned and the absorptivities reported in the literature
(Stephens, 1964; Bruckmann and Willner, 1983; Holdren and Spicer, 1984) permit
the quantitative analysis of PAN without calibration standards. The absorptiv-
ity of the 990 cm band of PBzN, a higher homologue of PAN, has been reported
by Stephens (1969). Tuazon et al. (1978) describes an FTIR system operable at
v
pathlengths up to 2 km for ambient measurements of PAN and other trace constit-
uents. This system employed an eight-mirror multiple reflection cell with a
22.5-m base path. Tuazon et al. (1981b) reported maximum PAN concentrations
ranging from 6 to 37 ppb over a 5-day smog episode in Claremont, CA. Hanst et
al. (1982) made measurements with a 1260-m folded optical path system during a
2-day smog episode in Los Angeles in 1980. An upper limit of 1 ppb of PBzN
was placed, and the maximum PAN concentration observed was 15 ppb. The Jarge
internal surface area of the White cells may serve to promote the decomposition
or irreversible adsorption of reactive trace species such as PAN. High volume
sampling rates and inert internal surface materials are used to minimize these
effects.
Because of the thermal instability of dilute PAN samples and the explosive
nature of liquefied PAN, calibration samples are not commercially available
and must be prepared. Earlier methods used to synthesize PAN have been summa-
rized by Stephens (1969). The photolysis of alkyl nitrites in oxygen was the
most commonly used procedure and may still be used by some investigators. As
described by Stephens et al. (1965), the liquefied crude mixture obtained at
the outlet of the photolysis chamber is purified by preparative-scale GC.
[CAUTION: Both the liquid crude mixture and the purified PAN samples are
violently explosive and should be handled behind explosion shields using
plastic full-face protection, gloves, and a leather coat at all times.] The
pure PAN is usually diluted to about 1000 ppm in cylinders pressurized with
nitrogen to 100 psig and stored at reduced temperatures, <15°C.
Gay et al. (1976) have used the photolysis of chlorine: aldehyde: nitrogen
dioxide mixtures in air or oxygen for the preparation of PAN and a number of
its homologues at high yields. This procedure has been utilized in a portable
4-79
-------
PAN generator that can be used for the calibration of GC-ECD instruments in
the field (Grosjean, 1983; Grosjean et a!., 1984).
Several investigators have recently reported on a condensed-phase synthesis
of PAN by nitration of peracetic acid in a hydrocarbon solvent. High yields
are produced of a pure product free of other alkyl nitrates (Hendry ,and, Kenley,
1977; Kravetz et al., 1980; Nielsen et a!., 1982; Holdren and Spicer, 1984).
After the nitration is complete, the hydrocarbon fraction containing PAN
concentrations of 2 to 4 mg/ml can be stored at -20°C for periods longer than
a year (Holdren and Spicer, 1984).
The most direct method for absolute analysis of these PAN samples is by
infrared absorption, using the IR absorptivities mentioned earlier. Cooyen-
W
tional IR instruments and 10-cm gas cells can analyze gas standards of concen-
trations >35 ppm. Liquid microcells can be used for the analysis of PAN in
isooctane solutions. The alkaline hydrolysis of PAN to acetate ion and nitrite
ion in quantitative yield (Nicksic et al., 1967) provides a means independent
of infrared for the quantitative analysis of PAN. Following hydrolysis,
nitrite ion may be quantitatively analyzed by the Saltzman colorimetric proce-
dure (Stephens, 1969). The favored procedures now use ion chromatography to
analyze for nitrite (Nielsen et al., 1982) or acetate (Grosjean, 1983; Grosjean
et al., 1984) ions. Another calibration procedure has been proposed that is
based on the thermal decomposition of PAN in the presence of excess and measured
NO concentrations (Lonneman et al., 1982). The acetylperoxy radical, CHgC^Op,
and its decomposition products rapidly oxidize nitric oxide (NO) to NO^ with a
stoichiometry that has been experimentally determined.
4.4.1.6 Methods for Sampling and Analysis of Hydrogen Peroxide. Hydrogen
peroxide (H^Op) is significant in photochemical smog as a chain terminator; as
an index of the hydroperoxyl radical (H02) concentration (Bufalini and Brubaker,
1969; Demerjian et al., 1974); and as a reactant in the aqueous-phase oxidation
«. O
of SQy to SO^ and in the acidification of rain (Penkett et al., 1979; Dasgupta,
1980a,b; Martin and Damschen, 1981; Overton and Durham, 1982).
One of the major problems, however, in assessing the role of atmospheric
HgOp has been a lack of adequate measurement methodology. Earlier measurements
(Gay and Bufalini, 1972a,b; Bufalini et al., 1972; Kok et al., 1978a,b) reporting
Hp02 concentrations of 0.01 to 0.18 ppm are now believed to be far too high,
and to be the result of artifact H^Op formation from reactions of absorbed Og
(Zika and Saltzman, 1982; Heikes et al., 1982; Heikes, 1984). Maximum tropo-
spheric H^Oz concentrations predicted by modeling calculations (Chameides and
4-80
-------
Tan ,: 1981; Logan et al., 1981) and observed in recent field studies (Das et
al., 1983) are on the order of 1 ppb.
Almost all of the methods used for the measurement of atmospheric H202
have used aqueous traps for sampling. Atmospheric 0,, however, which is also
absorbed at concentrations much higher than H2Q2, reacts in the bulk aqueous
phase and at surfaces to produce H202 and thus is a serious interference (Zika
and Saltzman, 1982; Heikes et al,, 1982; Heikes, 1984). The removal of absorbed
On by purging immediately after sample collection may remove or substantially
reduce this interference (Zika and Saltzman, 1982; Das et al., 1983). Another
problem identified with aqueous sampling is that other atmospheric species (in
particular, SOp) may interfere with the generation of H^O/j in aqueous .traps
and also react with collected HpO^ to reduce the apparent concentration measured
(Heikes et al., 1982).
Of the techniques that have been used for the measurement of aqueous and
gas-phase HJ)^, only the chemi luminescence and enzyme-catalyzed methods are
summarized below. The other techniques are now believed to have inadequate
sensitivity and specificity for the atmospheric concentrations actually present,
In addition, the use of a tunable diode infrared laser source should eliminate
the problem associated with nearby water bands, and this method is currently
under investigation for atmospheric measurements (unpublished work in progress,
Schiff, 1985).
Hydrogen peroxide in the atmosphere may be detected at low concentrations
by the chemi luminescence obtained from copper(II)-catalyzed oxidaton of luminol
(5-amino-2,3-dihydro-l,4-phthalazinedtone) by H2Q2 (Armstrong and Humphreys,
1965; Kok et al., 1978a,b). This technique as initially employed suffered the
interferences from Og and SQ^ discussed above for aqueous traps. Das et al.
(1982) employed a static version of the method of Kok et al. (1978a) to measure
HpO/j concentrations in the 0.01 to 1 ppb range. In addition, samples were
purged with argon immediately after collection to eliminate, reportedly, the
Qg interference. Recently, a modified chemi luminescence method has been
reported which used hemin, a blood component, as a catalyst for the luminol-
based HgQ/? oxl'dat1on (Yoshizumi et al., 1984).
The most promising chemical approach employs the catalytic activity of
the enzyme horseradish peroxidase (HRP) on the oxidation of organic substrates
by H2Q2- The production or decay of the fluorescence intensity of the substrate
or reaction product is measured as it is oxidized by HOo* catalyzed by HRP.
4-81
-------
Some of the more widely used substrates have been scopoletin (6-methoxy~7-
hydroxyl,2-benzopyrone) (Andreae, 1955; Perschke and Broda, 1961); 3-(g-hydroxy-
phenyl)propionic acid (HPPA) (Zaltsu and Okhura, 1980); and leuco crystal
violet (LCV) (Mottola et al., 1970).
The decrease in the fluorescence intensity of scopoletin is measured as a
function of HgO, concentration. Detection limits have been reported to be
quite low (10 M), The chief disadvantage of this approach is that the
concentration of HpOg must be within a narrow range to obtain an accurately
measureable decrease in fluorescence. Oxidation of LCV produces intensely
5 -1
colored crystal violet, which has a molar absorption coefficient of 10 M
cm at the analytical wavelength, 596 nm. The detection limit reported was
~8
10 M in 5 cm cells. Two quite similar hydrogen donor substrates have been
used. Zaitsu and Okhura (1980) employed 3-(p_-hydroxyphenyl) propionic acid
and more recently the j>-hydroxyphenyl acetic acid homologue is being used
(Kunen et al., 1983; Dasgupta and Hwang, 1985). The measurement of the fluores-
cence intensity of the product dimer provides a quite sensitive means for the
assay of HgOg-
As with Q-, HgOp calibration standards are not commercially available and
are usually prepared at the time of use. The most convenient method for pre-
paring aqueous samples containing micromolar concentrations of t-LOo is simply
the serial dilution of commercial grade 30 percent HgOg (Fisher Analytical
Reagent). Techniques for the convenient generation of gas-phase standards are
not available. A technique often used for generating ppm concentrations of
HgOg in air involves the injection of microliter quantities of 30 percent Hy®?
solution into a metered stream of air that flows into a Teflon bag. Aqueous
and gas-phase samples are then standardized by conventional iodometric proce-
dures (Allen et al., 1952; Cohen et al., 1967).
4.4.2 Measurement of Precursors to Ozone and Other Photochemical Oxidants
4.4.2.1 Nontnethane Organic Compounds. Numerous analytical methods have been
employed to determine nonmethane organic compounds (NMOC) in ambient air.
Measurement methods for the organic species may be grouped according to three
major classifications: nonmethane hydrocarbons, aldehydes, and other oxygenated
compounds.
Nonmethane hydrocarbons have been determined primarily by methods that
employ a flame ionization detector (FID) as the sensing element. Early methods
for the measurement of total nonmethane hydrocarbons did not provide for
4-82
-------
speeiation of the complex mixture of organics in ambient air. These methods,
still in use for analysis of total nonmethane organic compounds, are essentially
organic carbon analyzers, since the response of the FID detector is essentially
proportional to the number of carbon atoms present in the organic molecule
(Sevcik, 1975). Carbon atoms bound, however, to oxygen, nitrogen, or halogens
give reduced relative responses (Dietz, 1967). The FID detector has been
utilized both as a stand-alone continuous detection system (non-speciation
analyzers have indicated an overall poor performance of the commercial instru-
ments when calibration or ambient mixtures containing nonmethane organic
compounds (NMOC) concentrations less than 1 ppm C were analyzed (e.g., Reckner,
1974; McElroy and Thompson, 1975; Sexton et a!,, 1982). The major problems
associated with the non-speciation analyzers have been summarized in a recent
technical assistance document published by the U.S. Environmental Protection
Agency (1981). The document also presents ways to reduce some of the existing
problems.
Because of the above deficiencies, other approaches to the measurement of
nonmethane hydrocarbons are currently under development. The use of gas
chromatography coupled to an FID system circumvents many of the problems
associated with continuous non-speciation analyzers. This method, however,
requires sample preconcentration because the organic components are present at
part-per-billion (ppb) levels or lower in ambient air. The two main preconcen-
tration techniques in present use are cryogenic collection and the use of
solid adsorbents (McClenny et a!., 1984; Jayanty et a!., 1982; Westberg.
et al., 1980; Ogle et al., 1982). The preferred preconcentration method for
obtaining speciated data is cryogenic collection. Speciation methods involving
cryogenic preconcentration have also been compared with continuous nonspeciation
analyzers (e.g., Richter, 1983). Results indicate poor correlation between
methods at ambient concentrations below 1 part-per-million carbon (ppm C).
Aldehydes, which are both primary and secondary pollutants in ambient
air, are detected by total NMOC and NMHC speciation methods but can not be
quantitatively determined by those methods. Primary measurement techniques
for aldehydes include the chromotropic acid (CA) method for formaldehyde
(Altshuller and McPherson, 1963; Johnson et al., 1981), the 3-methyl-2-benz-
othiazolene (MBTH) technique for total aldehydes (e.g., Sawicki et a1», 1961;
Hauser and Cummins, 1964), Fourier-transform infrared (FTIR) spectroscopy
4-83
-------
(e.g., Hanst et al.} 1982; Tuazon et a!., 1978, 1980, 1981a), and high-perfor-
mance liquid chromatography employing 2»4-dinitrophenyl~hydrazine derivatiza-
tion (HPLC-DNPH) for aldehyde speciation (e.g., Lipari and Swarin, 1982; Kuntz
et a!., 1980). The CA and MBTH methods utilize wet chemical procedures and
spectrophotometric detection. Interferences from other compounds have been
reported for both techniques. The FTIR method offers good specificity and
direct jn situ analysis of ambient air. These advantages are offset, however,
by the relatively high cost and lack of portability of the instrumentation.
On the other hand, the HPLC-DNPH method not only offers good specificity but
can also be easily transported to field sites. A few intercomparison studies
of the above methods have been conducted and considerable differences in
measured concentrations were found. The data base is still quite limited at
present, however, and further intercomparisons are needed.
Literature reports describing the vapor-phase organic composition of
ambient air indicate that the major fraction of material consists of unsubsti-
tuted hydrocarbons and aldehydes. With the exception of formic acid, other
oxygenated species are seldom reported. The lack of oxygenated hydrocarbon
data is somewhat surprising since significant quantities of these species are
emitted into the atmosphere by solvent-related industries and since at least
some oxygenated species appear to be emitted by vegetation. In addition to
direct emissions, it is also expected that photochemical reactions of hydro-
carbons with oxides of nitrogen, 0,, and hydroxyl radicals will produce signi-
ficant quantities of oxygenated species. Difficulties during sample collection
and analysis may account for the apparent lack of data. Attempts have been
made to decrease adsorption by deactivating the reactive surface or by modifying
the compound of interest (Osman et a!., 1979; Westberg et al., 1980). Additional
research efforts should focus on this area.
4.4.2.2 Nitrogen Oxides. Aside from the essentially unreactive nitrous oxide
(NgO), only two oxides of nitrogen occur in ambient air at appreciable concen-
trations: nitric oxide (NO) and nitrogen dioxide (N02). Both compounds,
together designated as NO , participate in the cyclic reactions in the atmosphere
r\
that lead to the formation of ozone. Other minor reactive oxides of nitrogen
in ambient air include peroxyacyl nitrates, nitrogen trioxide, dinitrogen
pentoxide, and peroxynitric acid.
The preferred means (Federal Reference Method) of measuring NO and NOp is
the chemiluminescence method (F.R., 1976c). The measurement principle is the
gas-phase chemiluminescent reaction of 0, and NO (Fontijn et al., 1970).
4-84
-------
While;NO is determined directly in this fashion, N02 is detected indirectly by
first reducing or thermally decomposing the gas quantitatively to NO with a
converter. The reaction of NO and 0, forms excited N02 molecules that release
light energy that is proportional to the NO concentration. Although the NO
chemiluminescence is interference-free, other nitrogen compounds do interfere
when ^directed through the N02 converter. The magnitude of these interferences
is dependent upon the type of converter used (Winer et al., 1974; Joshi and
Bufalini, 1978). The detection limit of commercial chemiluminescence instru-
ments for N02 measurement is 2.5 MO/m3 (0.002 ppm) (Katz, 1976).
Development of an instrument based on the chemiluminescent reaction of
N02 with luminol (5-amino-2, 3-dihydro-l, 4-phthalazine dione) has been reported
by Maeda et al. (1980). Wendel et al. (1983) have reported modifications of
this luminol-based method in which better response time and less interference
from Oo have been achieved.
Other acceptable methods for measuring ambient N02 levels, including two
methods designated as equivalent methods, are the Lyshkow-modified Griess-
Saltzman method, the instrumental colorimetric Griess-Saltzman method, the
triethanolamine method, the sodium arsenite method, and the TGS-ANSA method
[TGS-ANSA = triethanolamine, guaiacol (o-methoxyphenol), sodium metabisulfite,
and 8-anilino-l-naphthalene sulfonic acid]. The sodium arsenite method, and
the TGS-ANSA method were designated as equivalent methods in 1977. For descrip-
tions of these methods, the reader is referred to the EPA criteria document
for nitrogen oxides (U.S. Environmental Protection Agency, 1982). While some
of these methods measure the species of interest directly, others require
oxidation, reduction, or thermal decomposition of the sample, or separation
from interferences, before measurement. None of these other techniques,
however, is widely used to monitor air quality.
Careful adherence to specified calibration procedures is essential for
obtaining accurate NO measurements. The U.S. Environmental Protection Agency
/\ '
(1975) has issued a technical assistance document that describes in detail the
two acceptable calibration methods for NO : (1) the use of standard reference
materials (SRMs) and (2) gas-phase titration (GPT) of NO with 03. The SRM for
NO is a cylinder of compressed NO in N2; the mixture is both accurate and
stable (Hughes, 1975). The SRM for N02 is the N02 permeation tube (O'Keeffe
and Ortman, 1966; Scaringelli et al., 1970). The gas-phase titration, described
by Rehme et al. (1974), is based upon the bimolecular reaction of NO with 03
to form N02 plus 02. The U.S. Environmental Protection Agency (1975) recommends
4-85
-------
the combined use of GPT plus SRM procedures, using one technique to check the
other.
4.5 REFERENCES
Adamson, P.; Guenthard, Hs, H. (1980) Reinvestigation of the matrix i.r. spectrum
of peroxyacetylnitrate (PAN). Spectrochim. Acta Part A 36: 473-475.
Allen, A. p.; Hochanadel, C. J.; Ghormley, J. A.; Davis, T. W. (1952) Decom-
position of water and aqueous solutions under mixed fast neutron and
gamma-radiation. J. Phys. Chem. 56: 575-586.
Altshuller, A. P. (1983) Measurements of the products of atmospheric photo-
chemical reactions in laboratory studies and in ambient air—relationships
between ozone and other products. Atmos. Environ. 17: 2383-2427.
Altshuller, A, P.; Leng, L J. (1963) Application of the 3-methyl-2-benzothia-
zolone hydrazone method for atmospheric analysis of aliphatic aldehydes.
Anal. Chem. 35: 1541-1542.
Altshuller, A. P.; McPherson, S. P. (1963) Spectrophotometric analysis of
aldehydes in the Los Angeles atmosphere. J. Air Pollut. Control Assoc,
13: 109-111.
Altshuller, A. P.; Cohen, I. R.; Meyer, M. E.; Wartburg, A. F., Jr. (1961)
Analysis of aliphatic aldehydes in source effluents and in the atmosphere.
Anal. Chim. Acta 25: 101-117.
Andreae, W. A. (1955) A sensitive method for the estimation of hydrogen peroxide
in biological materials. Nature (London) 175: 859-860.
Appel, B. R. (1973) A new and more sensitive procedure for analysis of peroxy-
benzoyl nitrate. J. Air Pollut. Control Assoc. 23: 1042-1044.
Armstrong, W. A.; Humphreys, W. G. (1965) A L.E.T. independent dosimeter based
on the chemiluminescent determination of H202- Can. J. Chem. 43: 2576-2584,
Ballard, L. F.; Tommerdahl, J. B.; Decker, C. E.; Royal, T. M,; Nifong, D. R.
(1971a) Field evaluation of new air pollution monitoring systems: the Los
Angeles Study: interim report. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; report no. APTD-0775. Available from: NTIS,
Springfield, VA; PB-204444.
Ballard, L. F.; Tommerdahl, J. B.; Decker, C. E.; Royal, T. M.; Matus, L. K.
(1971b) Field evaluation of new air pollution monitoring systems: St.
Louis Study, phase 1. Contract no. CPA 70-101, U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC. Available from: NTIS, Springfield,
VA; PB-264232.
4-86
-------
Bass,: A. M.; Ledford, A. E., Jr.; Whittaker, J. K. (1977) Ultraviolet photometer
for ozone calibration. In: Dimitriades, B., ed. International conference
on photochemical oxidant pollution and its control; September 1976;
Raleigh, NC. Proceedings: v. I. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency; pp. 13-17; EPA report no. EPA-600/3-77-001a.
Available from: NTIS, Springfield, VA; PB-264232.
Beard, M. E.; Margeson, J. H.; Ellis, E. C. (1977) Evaluation of 1 percent
neutral buffered potassium iodide procedure for calibration of ozone
monitors. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-6Q074-77-QQ5, Available from: NTIS, Springfield,
VA; PB-267985.
Bergshoeff, G. (1970) Ozone determination in air. In: Pittsburgh conference:
analytical chemistry and applied spectroscopy; publication no. 342.
Delft, The Netherlands: IG-TNO.
Bersis, D.; Vassiliou, E. (1966) A chemiluminescence method for determining
ozone. Analyst 91: 499-505.
Bowman, L D,; Horak, R. F. (1972) A continuous ultraviolet absorption ozone
photometer. Anal. Instrum. 10: 103-108.
Bowman, R. L.; Alexander, N. (1966) Ozone-induced chemiluminescence of organic
compounds. Science (Washington, DC) 154: 1454-1456.
Brewer, A. W.; Milford, J. R. (1960) The Oxford-Kew ozone sonde. Proc. R.
Soc. London Ser. A 256: 470-495.
Bricker, C. E.; Johnson, H. R. (1945) Spectrophotometric method for deter^
mining formaldehyde. Ind. Eng. Chem. Anal. Ed. 17: 400-402.
Bruckmann, P. W.; Winner, H. (1983) Infrared spectroscopic study of peroxy-
acetyl nitrate (PAN) and its decomposition products. Environ. Sci.
Technol. 17: 352-357.
Bryan, R. J.; Romanovsky, J. C, (1956) Instrumentation for air pollution.
Instrum. Autom. 29: 2432-2438.
Bufalini, J. J.; Brubaker, K. L. (1969) Symposium on chemical reactions in
urban atmospheres; October. Warren, MI: General Motors Research Labora-
tories.
Bufalini, J. J.; Gay, B. W., Jr.; Brubaker, K. L. (1972) Hydrogen peroxide
formation from formaldehyde photooxidation and its presence in urban
atmospheres. Environ. Sci. Technol. 6: 816-821.
Burton, C. S.; Bekowies, P. J.; Pollack, R. I.; Connell, P. (1976) Oxidant/
ozone ambient measurement methods: an assessment and evaluation. San
Rafael, CA: Systems Applications, Inc.; EF76-111R.
Butcher, S. S.; Ruff, R. E. (1971) Effect of inlet residence time on analysis
of atmospheric nitrogen oxides and ozone. Anal. Chem. 43: 1890-1892.
4-87
-------
California Air Resources Board. (1976) A study of the effect of atmospheric
humidity on analytical oxidant measurement methods. Presented ;;at: 15th
conference [on] methods in air pollution studies; January; Long; Beach,
CA. Berkeley, CA; California Air Resources Board,, Air and Industrial
Hygiene Laboratory.
• ' ' ' •"!.•''
California Air Resources Board. (1978) California air quality data;; July-
August-September. 10(13): 2-3. Available from: Technical Service Div.,
California Air Resources Board, Sacramento, CA.
Carroll, H.; Dimitriades, B.; Ray, R. M. (1972) A chemiluminescence.detector
for ozone measurement. Washington, DC: U.S. Department of the Interior,
Bureau of Mines; report of investigations no. RI7650.
C.F.R. (1971) 40:§50. Reference methods. . ;
C.F.R. (1975) 40:§53. Specifications for automated methods.
Chameides, W.L.; Tan, A. (1981). The two-demensional diagnostic model for
tropospheric OH: An uncertainty analysis. J. Geophys. Res. 86: 5209-5223.
Cherniack, I.; Bryan, R.. J. (1965) A comparison study of various types of
ozone and oxidant detectors which are used for atmospheric air sampling.
J. Air Pollut. Control Assoc. 15: 351-354.
Chock, D. P.; Kumar, S.; Herrmann, R. W. (1982) An analysis of trends in
oxidant air quality in the South Coast Air Basin of California. Atmos.
Environ. 16: 2615-2624.
Clark, T. A.; Baumgardner, R. E.; Stevens, R. K.; Krost, K. J. (1974) Evalua-
tion of new ozone monitoring instruments by measuring in nonurban atmos-
pheres. In: Instrumentation for monitoring air quality: a symposium;
August 1973; Boulder, CO. . Philadelphia, PA: American Society for Testing
and Materials; pp. 101-111.
Clements, J. B. (1975) Workshop on ozone measurement by the potassium iodide
method summary. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA summary report no. EPA-650/4-75-007. Available from: NTIS,
Springfield, VA; PB-240939.
Clyne, M. A. A.; Coxon, J. A. (1968) Kinetic studies of oxy-halogen radical
systems. Proc. R. Soc. London Ser. A 303: 207-231.
Cohen, I. R.; Purcell, T. C. (1967) Spectrophotometric determination of hydro-
gen peroxide with 8-quinolinol. Anal. Chem. 39: 131-132.
Cohen, I. R.; Purcell, T. C.; Altshuller, A. P. (1967) Analysis of the oxidant
in photooxidation reactions. Environ. Sci. Techno!. 1: 247-252.
Corse, E. W. (1981) Aldehyde analyses for the 1981 Atlanta study. EPA contract
no. 68-02-2566. Research Triangle Park, NG: Northrop Services, Inc.-
Environmental Sciences.
4-88
-------
Cox, ;R±i D,; Balfour, W. D.; Langley, G. J. (1982) Quality control for ambient
"level hydrocarbon sampling and analyses. Presented at: 75th annual
meeting of the Air Pollution Control Association; June; New OrleanSj LA.
Pittsburgh, PA: Air Pollution Control Association.
Csanyi, L. J. (1981) Analytical applications of mixed-ligand complexes. III.
vanadium(V)-xylenol orange reagent mixture for the spectrophotometric
determination of traces of hydrogen peroxide. Microchem. J. 26: 10-16.
Darley, E. F.; Kettner, K. A.; Stephens, E. R. (1963) Analysis of peroxyacyl
nitrates by gas chromatography with electron capture detection. Anal.
"Chem. 35: 589-591.
Das, T.N.; Moorthy, P.N.;
determination of low
Chem. Soc. 59: 85-89.
Rao, K.N. (1982).
concentrations of
Chemiluminescent method for the
hydrogen peroxide. J. Indian
Das, T.N.; Moorthy, P.N.; Rao, K.N. (1983). Chemiluminescent measurement of
hydrogen peroxide in the Bombay area. Atmos. Environ. 17: 79-82.
Dasgupta, P. K. (1980a) Discussion on "The importance of atmospheric ozone and.
hydrogen peroxide in oxidizing sulphur dioxide in cloud and rainwater [on
Penkett et al., 1979]." Atmos. Environ. 14: 272-275.
Dasgupta, P. K. (1980b) The importance of atmospheric ozone and hydrogen
peroxide in oxidizing sulfur dioxide in cloud and rainwater—further
discussion [on Penkett et al., 1979], Atmos. Environ. 14: 620-621.
Dasgupta, P. K.; Hwang, H. (1985). Application of a nested loop system for
the flow injection analysis of .trace aqueous peroxide. Anal. Chem. In
press.
Davis, D. R.; Jensen, R. E. (1976) Low level ozone and weather systems. In:
Specialty conference on ozone/oxidants—interactions with the total
environment. Philadelphia, PA: Air Pollution Control Association; pp.
242-251.
Demerjian, K. L.; Kerr, J. A.; Calvert, J.
chemical smog formation. In: Pitts,
Advances in environmental science and
John Wiley and Sons, Inc.; pp. 1-262.
G. (1974) The mechanism of photo-
J. N., Jr.; Metcalf, R. L., eds.
technology: v. 4. New York, NY:
DeMore, W. B.; Patapoff, M. (1976) Comparison of ozone determinations by
ultraviolet photometry and gas-phase titration. Environ. Sci. Techno!.
10: 897-899.
DeMore, W. B.; Romanovsky, J. C.; Feldstein, M.; Hamming, W. J.; Mueller, P.
K. (1976) Interagency comparison of iodometric methods for ozone determi-
nation. In: Calibration in air monitoring: a symposium; August 1975; .
Boulder, CO. Philadelphia, PA: American Society for Testing and Materials;
pp. 131-143; ASTM special technical publication 598.
Dietz, W. A. (1967) Response factors for gas chromatographic analyses.
Chromatogr. 5: 68-71.
J. Gas
4-89
-------
Eegriwe, E. (1937) Reactions and reagents for the detection of organic com-
pounds. Z. Anal. Chem. 110: 22025.
Fincher, C. L; Tucker, A. W.; Birnbaum, M.; Paur, R. J.; McClenny, W. A.
(1977) Fluorescence ambient N02 monitor with flashlamp pumping. Appl
Optics 16: 1359-1365.
Finlayson, B. J.; Pitts, J. N., Jr.; Atkinson, R. (1974) Low pressure gas
phase ozone-olefin reactions: chemiluminescence, kinetics, and mechanisms.
J. Am. Chem. Soc. 96: 5356-5367.
Flamm, D. L. (1977) Analysis of ozone at low concentrations with boric acid buffered
KI. Environ. Sci. Techno!. 11: 978-983.
Fontijn, A.; Sabadell,"A. J.; Ronco, R. J. (1970) Homogeneous chemiluminescent
measurement of nitric oxide with ozone: implications for continuous
selective monitoring of gaseous air pollutants. Anal. Chem. 42: 575-579.
Foulger, B. E.; Simmonds, P. G. (1979) Drier for field use in the determination
of trace atmospheric gases. Anal. Chem. 51: 1089-1090.
F. R. (1971, April 30) 36: 8186-8201. National primary and secondary air quality
standards.
F.R. (1975a, February 18) 40: 7042-7070. Ambient air monitoring reference and
equivalent methods.
F.R. (1975b, February 18) 40: 54856. Ambient air monitoring reference and
equivalent methods: notice of reference method designations.
F.R. (1976a, February 4) 41: 5145-5146. Ambient air monitoring reference and
equivalent methods: reference method designation.
F.R. (1976b, October 22) 41: 46647-46648. Ambient air monitoring reference and
equivalent methods: reference method designation.
F.R. (1976c, December 1) 41: 52686-52692. National primary and secondary ambient
air quality standards: nitrogen dioxide measurement principle and calibra-
tion procedure.
F.R. (1976d, December 8) 41: 53684-53685. Ambient air monitoring reference and
equivalent methods: reference method designation.
F.R. (1977a, June 3) 42:28571-28572. Ambient air monitoring reference and
equivalent methods: reference and equivalent method designation.
F.R. (1977b, June 13) 42:30235. Ambient air monitoring reference and equivalent
methods: Amendment to reference methods for ozone.
F.R. (1977c, August 1) 42:38931-38932. Ambient air monitoring reference and
equivalent methods: equivalent method designation.
F.R. (1977d, November 1) 42: 57156. Ambient air monitoring reference and
equivalent methods: Philips PW9771 03 analyzer; correction.
4-90
-------
F.R- (19793, February 20) 44: 10429-10430. Ambient air monitoring reference
and equivalent method designations: Columbia Scientific Industries Model
2000 Ozone Meter Thermo Electron Model 14D/E Chemiluminescent NO-N02-N0-
Analyser.
F.R. (1979b, May 10) 44: 27558-27604. Air programs; ambient air quality monitor-
ing, data reporting, and surveillance provisions.
F.R. (1979c, November 9) 44: 65069-65071. Regulations for ambient air quality
monitoring and data reporting.
F.R. (1979d, December 14) 44: 72589-72592. Air programs; ambient, air quality
monitoring, data reporting, and surveillance provisions [errata].
F.R., (1979e, February 8) 44: 8221-8233. Calibration of ozone reference methods.
F.R. (1980s, March 21) 45: 18474. Ambient air monitoring reference and equiva-
lent methods; amendment to reference method for ozone.
F.R. (1980b, August 27) 45: 57168-57169. Ambient air monitoring reference and
equivalent methods; equivalent method designation.
F.R. (1981, October 26) 46: 52224. Ambient air monitoring reference and equiva-
lent methods; equivalent method designation.
F.R. (1982, March 31) 47: 13572-13573. Ambient air monitoring reference and
equivalent methods; equivalent method designation.
F.R. (1983, March 10) 48: 10126. Ambient air monitoring reference and equiva-
lent methods; equivalent method designation.
Fried, A.; Hodgeson, J. (1982) Laser photoacoustic detection of nitrogen dio-
xide in the gas-phase titration of nitric oxide with ozone. Anal. Chem.
54: 278-282.
Fung, K.; Grosjean, D. (1981) Determination of nanogram amounts of carbonyls
as 2,4-dinitrophenylhydrazones by high-performance liquid chromatography.
Anal. Chem. 53: 168-171.
Gay, B. W., Jr.; Bufalini, J. J. (1972a) Hydrogen peroxide in the urban atmos-
phere. In: Gould, R. F., ed., Photochemical smog and ozone reactions.
Washington, DC: American Chemical Society; pp. 225-263. (Advances in
chemistry series: 113.)
Gay, B. W., Jr.; Bufalini, J. J. (1972b) Hydrogen peroxide in the urban atmos-
phere. Environ. Lett. 3: 21-24.
Gay, B. W.; Noonan, R. C.; Bufalini, J. J.; Hanst, P. L. (1976) Photochemical
synthesis of peroxyacyl nitrates in gas phase via chlorine-aldehyde
reaction. Environ. Sci. Techno!. 10: 82-85.
Graedel, T. E.; Farrow, L. A.; Weber, T. A. (1976) Kinetic studies of the
photochemistry of the urban troposphere. Atmos. Environ. 10: 1095-1116.
4-91
-------
Grosjean, D. (1982) Distribution of nitrogenous pollutants during photochemi-
cal smog episodes. Presented at: 75th annual meeting of the Air Pollution
Control Association; June; New Orleans, LA. Pittsburgh, PA: Air Pollution
Control Association; paper no. 82-41.4.
Grosjean, D. (1983) Distribution of atmospheric nitrogenous pollutants at a
Los Angeles area smog receptor site. Environ. Sci. Technol. 17: 13-19.
Grosjean, D.; Kpk, G. L. (1981) Inter!aboratory comparison study of methods
for measuring formaldehyde and other aldehydes in ambient air. Westlake
Village, CA: Environmental Research and Technology, Inc. Available from:
NTIS, Springfield, VA; PB82-224486.
Grosjean, D.; Fung, K.; Collins, J.; Harrison, J.; Breitung, E. (1984) Portable
generator for on-site calibration of peroxyacetyl nitrate analyzers.
Anal. Chem. 56: 569-573.
Hager, R. N., Jr.; Anderson, R. C. (1970) Theory of the derivative spectro-
meter. J. Opt. Soc. Am. 60: 1444-1449.
Hampson, R. F., ed.; Braun, W.; Brown, R. L.; Garvin, D.; Herron, J. T.; Huie,
R. E.; Kurylo, M, J.; Laufer, A. H.; McKinley, J. D.; Qkabe, H.; Scheer,
M. D.; Tsang, W.; Stedman, D. H. (1973) Survey of photochemical and rate
data for twenty-eight reactions of interest in atmospheric chemistry. J.
Phys. Chem. Ref. Data 2: 267-311.
Hanst, P. L. (1970) Infrared spectroscopy and infrared lasers in air pollution
research and monitoring. Appl. Spectrosc. 24: 161-174.
Hanst, P. L.; Wilson, W. E.; Patterson, R. K.; Gay, B. W., Jr.; Chaney, L. W.
(1975) A spectroscopic study of California smog. Research Triangle Park,
NC: U.S. Environmental Protection Agency, National Environmental Research
Center; EPA report no. EPA-650/4-75-006. Available from: NTIS, Springfield,
VA; PB-241022.
Hanst, P. L.; Wong, N. W.; Bragin, J. (1982) A long-path infrared study of Los
Angeles smog. Atmos. Environ. 16: 969-981.
Harrison, J. W.; Timmons, M. L.; Denyszyn, R. B.; Decker, C. E. (1977) Evalua-
tion of the EPA Reference Method for measurement of non-methane hydrocar-
bons. Research Triangle Park, NC: U.S. Environmental Protection Agency;
EPA report no. EPA-600/4-77-033. Available from: NTIS, Springfield, VA;
PB-278296.
>/
Hauser, T. R.; Cummins, R. L. (1964) Increasing sensitivity of 3-methyl-2-benzo-
thiazolone hydrazone test for analysis of aliphatic aldehydes in air.
Anal. Chem. 36: 679-681.
Hauser, T. R.; Kolar, A. M. (1968) Spectrophotometric determination of hydrogen
peroxide in aqueous media with l,2-Di-(4-pyriclyl) ethylene. Anal. Chem.
40: 231-232.
Heikes, B.G. (1984) Aqueous H202 production from 03 in glass impingers.
Atmos. Environ. 18: 1433-1445.
4-92
-------
Heikes; B. G. ; Lazrus, A. L. ; Kok, G. L. ; Kunen, S. M. ; Gandrud, B. W. ; Gitlin,
-^ S. N. ; Sperry, P. D. (1982) Evidence for aqueous phase hydrogen peroxide
: "^synthesis in the troposphere. JGR J. Geophys. Res. 87: 3045-3051.
Hendry, D. G. ; Kenley, R. A. (1977) Generation of peroxy radicals from peroxy
'">m'trates (RQ2N02). Decomposition of peroxyacyl nitrates. Atmos. Environ.
• 99: 3198-3199.
Hersch, P.; Deuringer, R. (1963) Galvanic monitoring of ozone in air. Anal.
35: 897-899.
Heuss, J. M. ; Glasson, W. A. (1968) Hydrocarbon reactivity and eye irritation.
.Environ. Sci. Techno!. 2: 1109-1116.
Hinkley, E. D. ; Kelley, P. L. (1971) Detection of air pollutants with tunable
diode lasers. Science (Washington, DC) 171: 635-639.
Hochanadel , C. J. (1952) Effects of cobalt ^-radiation on water and aqueous
solutions. J. Phys. Chem. 56: 587-594.
Hodgeson j J. A. (1972) Review of analytical methods for atmospheric oxidants
measurements. Int. J. Environ. Anal. Chem. 2: 113-132.
Hodgeson, J. A. (1976) A survey of calibration techniques for atmospheric
ozone monitors. Washington, DC: National Bureau of Standards; publica-
tion no. NBSIR 76-1191. Available from: NTIS, Springfield, VA; PB-263534.
Hodgeson, J. A.; Krost, K. J. ; Q'Keeffe, A. E. ; Stevens, R. K. (1970) Chemi-
luminescent measurement of atmospheric ozone: response characteristics
and operating variables. Anal. Chem. 42: 1795-1802.
Hodgeson, J. A.; Stevens, R. K. ; Martin, B. E. (1972) A stable ozone source
applicable as a secondary standard for calibration of atmospheric monitors.
ISA Trans. 11: 161-167.
Hodgeson, J. A.; Hughes, E. E. ; Schmidt, W. P.; Bass, A. M. (1977) Methodology
for standardization of atmospheric ozone measurements. In: Dimitriades,
B. , ed. International conference on photochemical oxidant pollution and
its control; September 1976; Raleigh, NC. Proceedings: v. I. Research
Triangle Park, NC: U.S. Environmental Protection Agency; pp. 3-12; EPA
report no. EPA-600/3-77-001a. Available from: NTIS, Springfield, VA;
PB-264232.
Hoigne, J. ; Bader, H. (1976) The role of hydroxyl radical reactions in ozona-
tion processes in aqueous solutions. Water Res. 10: 377-386.
Holdren, M. W. ; Rasmussen, R. A. (1976) Moisture anomaly of peroxyacetyl
nitrate (PAN). Environ. Sci. Techno!. 10: 185-187.
Holdren, M. W. ; Spicer, C. W. (1984) Field compatible calibration procedure
for peroxyacetyl nitrate. Environ. Sci. Techno!. 18: 113-116.
4-93
-------
Holdren, M,; Spicer, C.; Sticksel, P.; Nepsund, K.; Ward, G.; Smith, R. (1982)
Implementation and analysis of hydrocarbon grab samples from Cleveland
and Cincinnati 1981. Ozone monitoring study. Research Triangle Rark, NC:
U.S. Environmental Protection Agency; EPA report no. EPA-905/4-82-001.
Available from: NTIS, Springfield, VA; PB83-197673. :
Hughes, E. E. (1975) Development of standard reference materials for afr quality
measurement. ISA Trans. 14: 281-291.
Huygen, I. C. (1970) Reaction of nitrogen dioxide with Griess type reagents.
Anal. Chem. 42: 407-409.
Intersociety Committee, American Public Health Association. (1970) Manual and
methods of air sampling and analysis: tentative method for continuous
monitoring of atmospheric oxidants with amperometric instruments. Health
Lab. Sci. 7: 1.
Intersociety Committee, American Public Health Association. (1977a) Tentative
method of analysis for formaldehyde content of the atmosphere (MBTH-colori-
metric method—application to other aldehydes). In: Katz, M., ed. Methods
of air sampling and analysis. 2nd ed. Washington DC: American Public
Health Association; pp. 308-313.
Intersociety Committee, American Public Health Association, (1977b) Tentative
method of analysis for nitrogen dioxide content of the atmosphere. In:
Katz, M., ed. Methods of air sampling and analysis. 2nd ed. Washington,
DC: American Public Health Association.
Jayanty, R. K. M.; Blackard, A.; McElroy, F. F.; McClenny, W. A. (1982)
Laboratory evaluation of nonmethane organic carbon determination in
ambient air by cryogenic preconcentration and flame ionization detection.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-6QO/4-82-019. Available from: NTIS, Springfield, VA;
PB82-224965.
Johnson, D. F.; Kok, G. L.; Sonner, R. J. (1981) Improved chromatographic acid
technique for the determination of formaldehyde. In: Grosjean, D.; Kok,
G. L., eds. Inter!aboratory comparison study of methods for measuring
formaldehyde and other aldehydes in ambient air: final report. Atlanta,
GA: Coordinating Research Council.
Joshi, S. B.; Bufalini, J. J. (1978) Halocarbon interferences in chemilumines-
cent measurements of NO . Environ. Sci. Techno!. 12: 597-599.
f\
Katz, M. (1976) Nitrogen compounds and oxidants. In: Stern, A. C., ed. Air
pollution; v. III. measurement, monitoring and surveillance of air pollu-
tion. 3rd ed. New York, NY: Academic Press; pp. 259-305.
Killus, J. P.; Whitten, G. Z. (1985) Behavior of trace NO species in
the nighttime urban atmosphere. J. Geophys. Res. 90; pp. 2430-2432.
Kilpatrick, M. L.; Herrich, C. C.; Kilpatrick, M. (1955) The decomposition of
ozone in aqueous solution. J. Am. Chem. Soc. 78: 1784-1789.
4-94
-------
Kok, G._L.; Partial!, K. R.; Winer, A. M.; Pitts, J, N., Jr.; Gay, B. W., Jr.
(1978a) Ambient air measurements of hydrogen peroxide in the California
South Coast Air Basin. Environ. Sci. Technol. 12: 1077-1080.
. •;•," I; at. .
Kok, G. L.; Holler, T. P.; Lopez, M. B.; Nachtrieb, H. A.; Yuan, M. (1978b)
Chemiluminescent method for determination of hydrogen peroxide in the
ambient atmosphere. Environ. Sci. Technol. 12: 1072-1076.
Kravetz, T. M.; Martin, S. W.; Mendenhall, G. D. (1980) Synthesis of peroxy-
acetyl and peroxyaroyl nitrates: complexion of peroxyacetyl nitrate with
benzene. Environ. Sci. Technol. 14: 1262-1264.
Kreuzer, L. B.; Patel, C. K. N. (1971) Nitric oxide air pollution: detection
by optoacoustic spectroscopy. Science (Washington, DC) 173: 45-47.
Krug, E. L. R.; Hirt, W. E. (1977) Interference of nitrate in the determination
of formaldehyde by the chromotropic acid method. Anal. Chem. 49: 1865-1867.
Kunen, S. M.; Lazrus, A. L.; Kok, G. L.; Heikes, B. G. (1983) Aqueous oxida-
tion of S02 by hydrogen peroxides. JGR J. Geophys. Res. 88: 3671-3674.
Kuntz, R.; Lonneman, W.; Namie, G.; Hull, L. A. (1980) Rapid determination of
aldehydes in air analyses. Anal. Lett. 13: 1409-1415.
Lamb, S. I.; Petrowski, C.; Kaplan, I. R. ; Simoneit, B. R. T. (1980) Organic
compounds in urban atmospheres: a review of distribution, collection and
analysis. J. Air Pollut. Control Assoc. 30: 1098-1115.
Levaggi, D. A.; Kothny, E. L.; Belsky, T.; de Vera, E. R.; Mueller, P. K.
(1974) Quantitative analysis of nitric oxide in presence of nitrogen
dioxide at atmospheric concentrations. Environ. Sci. Technol. 8: 348-350.
Lipari, F. (1984) New solid-sorbent method for ambient nitrogen dioxide monitor-
ing. Anal. Chem. 56: 1820-1826.
Lipari, F.; Swarin, S. (1982) Determination of formaldehyde and other aldehydes
in automobile exhaust with an improved 2,4-dinitrophenylhydrazine method.
J. Chromatogr. 247: 297-306.
Lipari, F.; Swarin, S. (1985) 2,4-Dinitrophenylhydrazine-coated Florisil
sampling cartridges for the determination of formaldehyde in air. Environ.
Sci. Technol. 19: 70-74.
Littman, F. E.; Benoliel, R. W. (1953) Continuous oxidant recorder. Anal.
Chem. 25: 1480-1483.
Lloyd, A. C. (1979) Tropospheric chemistry of aldehydes. In: Herron, J. T.;
Huie, R. E.; Hodgeson, J. A., eds. Chemical kinetic data needs for model-
ing of the lower troposphere: proceedings of a workshop; May 1978; Reston,
VA. Gaithersburg, MD: U.S. Department of Commerce, National Measurement
Laboratory; Bureau of Standards special publication 557.
Logan, J.A.; Prather, M.J.; Wofsy, S.C.; McElroy, M.B. (1981) Tropospheric
chemistry: A global perspective. J. Geophys. Res. 86: 7210-7254.
4-95
-------
Lonneman, W.A. (1977) PAN measurement in dry and humid atmospheres. Environ.
Sci. Techno!. 11: 194-196. ; ---,"
Lonneman, W. A. (1979) Ambient hydrocarbon measurements in Houston. In-: Ozone/
oxidants: interactions with the total environment II: proceedings of a
specialty conference; October; Houston, TX. Pittsburgh, PA: . J-iir Pollu-
tion Control Association; pp. 278-287.
Lonneman, W. A.; Kopczynski, S. L.; Parley, P. E.; Sutterfield, F. D, (1974)
Hydrocarbon composition of urban air pollution. Environ. Sci. Techno!.
8: 229-236.
Lonneman, W. A.; Bufalini, J. J.; Seila, R. L. (1976) PAN and oxidant measure-
ment in ambient atmospheres. Environ. Sci. Techno!. 10: 374-380.:
Lonneman, W. A.; Bufalini, J. J.; Kuntz, R. L.; Meeks, S. A. (1981) Contamina-
tion from fluorocarbon films. Environ. Sci. Techno!. 15; 99-103.
Lonneman, W. A.; Bufalini, J. J.; Namie, G. R. (1982) Calibration procedure
for PAN based on its thermal decomposition in the presence of nitric
oxide. Environ. Sci, Techno!. 16: 655-660.
Lord, H. C.; Egan, W. D.; Johnson, F. L.; Mclntosh, L. C. (1975) Measurement
of exhaust emissions in piston and diese! engines by dispersive spectrum.
J. Air Pollut. Control Assoc. 4: 136-139.
Lovelock, J. E.; Maggs, R. J.; Adlard, E. R. (1971) Gas-phase coulometry by
thermal electron attachment. Anal. Chem. 43: 1962-1965.
Maeda, Y.; Aoki, K.; Munemori, M. (1980) Chemiluminescence method for the
determination of nitrogen dioxide. Ana!. Chem. 52: 307-311.
Manos, D. M.; Volltrauer, H. N.; Allen, J.; Burch, D. E.; Chudy, J. S.; Pembrook,
J. D. (1982) Totally optical ambient NMHC monitor. EPA contract no. 68-02-
3292. Research Triangle Park, NC: U.S. Environmental Protection Agency.
Martin, L. R.; Damschen, D. E. (1981) Aqueous oxidation of sulfur dioxide by
hydrogen peroxide at low pH. Atmos. Environ. 15: 1615-1621.
Mast, G. M.; Saunders, H. E. (1962) Research and development of the instru-
mentation of ozone sensing. ISA Trans. 1: 325-328.
McClenny, W. A.; Pleil, J. D.; Holdren, M. W.; Smith, R. N. (1984) Automated
cryogenic preconcentration and gas chromatographic determination of
volatile organic compounds in air. Ana!, Chem. 56: 2947-2951.
McElroy, F. F. (1979) Transfer standards for the calibration of ambient air
monitoring analyzers for ozone: technical assistance document. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory; EPA report no. EPA-600/4-79-056.
Available from: NTIS, Springfield, VA; PB-146871.
McElroy, F. F.; Thompson, V. L. (1975) Hydrocarbon measurement discrepancies
among various analyzers using flame-ionization detectors. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory; EPA report no. EPA-60Q/4-75-Q1Q.
4-96
-------
Meloan, C. E.; Mauck, M.; Huffman, C. (1961) Spectrophotometric determination
of traces of hydrogen peroxide. Anal. Chem. 33: 104-106.
Miksch, P. R.; Anthon, D. W. (1982) A recommendation for combining the standard
analytical methods for the determinations of formaldehyde and total
aldehydes in air. Am. Ind. Hyg. Assoc. J. 43: 362-365.
Mottola, H. A.; Simpson, B. E.; Gorin, G. (1970) Absorptiometric determination
of hydrogen peroxide in submicrogram amounts with leuco crystal violet
and peroxidase as catalyst. Anal. Chem. 42: 410-411.
Mueller, P. K.; Tokiwa, Y.; deVera, E. R.; Wehrmeister, W. J.; Belsky, T.;
TWiss, S.; Inada, M. (1973) A guide for the evaluation of atmospheric
analyzers. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Quality Assurance and Environmental Monitoring Laboratory; EPA
report no. EPA-650/4-74-014. Available from: NTIS, Springfield, VA;
PB-233048.
National Air Pollution Control Administration. (1970) Air quality criteria for
photochemical oxidants. Washington, DC: U.S. Department of Health, Educa-
tion, and Welfare; publication no. AP-63. Available from: NTIS, Springfield,
VA; PB82-242231.
National Research Council. (1976) Vapor-phase organic pollutants: volatile
hydrocarbons and oxidation products. Washington, DC: National Academy
of Sciences.
National Research Council. (1977) Ozone and other photochemical oxidants.
Washington, DC: National Academy of Sciences; pp. 45-125 and chapters
4 and 6.
National Research Council. (1981) Formaldehyde and other aldehydes. Washington,
DC: National Academy of Sciences.
Neal, R.; Severs, R. K.; Wenzel, L.; MacKenzie, K. (1976) Simultaneous chemi-
luminescent ozone and KI oxidant measurements in Houston, Texas, 1975.
In: Speciality conference on ozone/oxidants—interaction with the total
environment. Philadelphia, PA: Air Pollution Control Association; pp.
180-188.
Nederbragt, G. W.; van der Horst, A.; van Duijn, J. (1965) Rapid ozone deter-
mination near an accelerator. Nature (London) 206: 87.
Nicksic, S. W.; Harkins, J.; Mueller, P. K. (1967) Some analyses for PAN and
studies of its structure. Atmos. Environ. 1: 11-18.
Nieboer, H.; van Ham, J. (1976) Peroxyacetyl nitrate (PAN) in relation to
ozone and some meteorological parameters at Delft in The Netherlands.
Atmos. Environ. 10: 115-120.
Nielsen, T.; Hansen, A. M.; Thomsen, E. L. (1982) A convenient method for pre-
paration of pure standards of peroxyacetyl nitrate for atmospheric analyses.
Atmos. Environ. 16: 2447-2450.
4-97
-------
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1976) Infrared
Fourier-transform spectroscoplc studies of atmospheric reactions. In:
12th informal conference on photochemistry, National Bureau of Standards;
June-July; Gaithersburg, MD: National Bureau of Standards; paper N2.
Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. (1980) An FTIR study
of the mechanism for the gas phase reaction between H02 radicals. Chem.
Phys. Lett. 73: 43-46.
Norwitz, 6. (1966) A colorimetric method for the determination of oxides of
nitrogen. Analyst 91: 553-558.
Ogle, L. D.; Hall, R. C.; Crow, W. L.; Jones, A. E.; Gise, J. P. (1982) Devel-
opment of preconcentration and chromatographic procedures for the contin-
uous and unattended monitoring of hydrocarbons in ambient air. Presented
at: 184th national meeting of the American Chemical Society; September;
Kansas City, MO. Austin, TX: Radian Corporation.
Okabe, H.; Schwarz, F. P. (1975) Nitric oxide pollution monitor. U.S. Patent No.
3,906,226 (Cl 250-304, G01J) Sept. 16, 1975.
O'Keeffe, A. E.; Ortman, G. C. (1966) Primary standards for trace gas analysis.
Anal. Chem. 38: 760-763.
Okita, T.; Inugami, M. (1971) Measurement of ozone by chemiluminescent method.
Koshu Eiseiin Kenkyu Hokoku 20: 207-211.
Olansky, A. S.; Deminp, S. N. (1976) Optimization and interpretation of absor-
bance response in the determination of formaldehyde with chromotropic
acid. Anal. Chim. Acta 83: 241-249.
Osman, M.; Hill, H. H.; Holdren, M. W.; Westberg, H. H. (1979) Vapor-phase
silylation of alcohols for air analysis. In: Zlatkis, A., ed. Advances
in chromatography, ehromatography symposium; Houston, TX.
Ott, W. (1977) Development of criteria for siting air monitoring stations. J.
Air Pollut. Control Assoc. 27: 543-547.
Overton, J. H., Jr.; Durham, J. L. (1982) Acidification of rain in the presence
of S02, H202, 03, and HN03. In: Keith, L. H., ed. Energy and environmental
chemistry, v. 2: acid rain. Ann Arbor, MI: Ann Arbor Science; pp. 245-262.
Papa, L. J.; Turner, L. P. (1972) Chromatographic determination of carbonyl
compounds as their 2, 4-dinitrophenylhydrazones. II. High pressure liquid
chromatography. J. Chromatogr. Sci. 10: 747-50.
Paur, R. J.; McElroy, F. F. (1979) Technical assistance document for the
calibration of ambient ozone monitors. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Environmental Monitoring and Support
Laboratory; EPA final report no. EPA-600/4-79-057. Available from: NTIS,
Springfield, VA; PB80-149552.
4-98
-------
Peake, E.; Sandhu, H. S. (1982) The formation of ozone and peroxyacetyl nitrate
(PAN) in the urban atmospheres of Alberta. Presented at: 65th annual
conference of the Chemical Institute of Canada; May-June; Toronto, ON,
Canada. Can. J. Chem. 61: 927-935.
Penkett, S. A.; Jones, B. M. R.; Brice, K. A.; Eggleton, A. E. J. (1979) The
• importance of atmospheric ozone and hydrogen peroxide in oxidising sulphur
dioxide in cloud and rainwater. Atmos. Environ. 13: 123-137.
Perschke, H.; Broda, E. (1961) Determination of very small amounts of hydrogen
peroxide. Nature (London) 190: 257-258.
Pilz, W.; Johann, I. (1974) Die Bestimmung kleinster Mengen von Wasserstoff-
peroxid in Luft. [Measurement of very small quantities of hydrogen pero-
xide in air.] Int. J. Environ. Anal. Chem. 3: 257-270.
Pisarevskii, A. M.; Polozova, I. P. Avramenko, L. V. (1980) An electrode for
potentiometric determination of hydrogen peroxide in alkaline media. J.
Anal. Chem. USSR 35: 1419-1425.
Pitts, J.N., Jr.; Fuhr, H.; Gaffney, J.S.; Peters J.W. (1973). Chemiluminescent
reactions of peroxyacetyl nitrate and ozone with triethyl amine. Possible
atmospheric monitor for peroxyacetyl nitrate. Environ. Sci. Tech. 7:
550-552.
Platt, U.; Perner, D.; Winer, A. M.; Harris, G. W.; Pitts, J. N., Jr. (1980)
Detection of N03 in the polluted troposphere by differential optical ab-
sorption. Geophys. Res. Lett. 7: 89-92.
Platt, U.; Winer, A. M.; Biermann, H. W.; Atkinson, R.; Pitts, J. N., Jr.
(1984) Measurement of nitrate .radical concentrations in continental air.
Environ. Sci. Techno!.: in press.
Poli, A. A.; Zinn, T. L. (1973) A new continuous monitor for methane and
non-methane total hydrocarbons. Anal. Instrum. 11: 135-142.
Purcell, T. C.; Cohen, I. R. (1967) Microdetermination of peroxides by kinetic
colorimetry. Environ. Sci. Technol. 1: 431-433.
Reckner, L. R. (1974) Survey of users of EPA Reference Method for measurement
of non-methane hydrocarbon in ambient air. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Research and Development;
EPA report no. EPA-650/4-75-008. Available from: NTIS, Springfield, VA;
PB-247515.
Regener, V. H. (1960) On a sensitive method for the recording of atmospheric
ozone. JGR J. Geophys. Res. 65: 3975-3977.
Regener, V. H. (1964) Measurement of atmospheric ozone with the chemilumine-
scent method. JGR J. Geophys. Res. 69: 3795-3800.
Rehme, K.A.; Martin, B.E.; Hodgeson, J.A. (1974) Tentative method for the
calibration of nitric oxide, nitrogen dioxide, and ozone analyzers by gas
phase titration. U.S. Environmental Protection Agency, Research Triangle
Park, N.C. EPA-R2-73-246.
4-99
-------
Rehme, K. A.; Puzak, J. C.; Beard, M. E.; Smith, C. F.; Paur, R. J... (1981)
Evaluation of ozone calibration procedures. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Environmental Monitoring Systems
Laboratory; EPA final report no. EPA-600/4-80-050. Available from: NTIS,
Springfield, VA; PB81-118911.
Renzetti, N. A.; Romanovsky, J. C. (1956) Comparative study of oxidants in the
Los Angeles atmosphere. J. Air Pollut. Control Assoc. 6: 154-159.
Richter, H. G. (1983) Analysis of organic compound data gathered during 1980
in northeast corridor cities. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards;
EPA report no. EPA-450/4-83-017. Available from: NTIS, Springfield, VA;
PB84-116052.
Ridley, B. A.; Hewlett, L. C. (1974) An instrument for nitric oxide measure-
ments in the stratosphere. Rev. Sci. Instrum. 45: 742-746.
Riggin, R. M. (1983) Technical assistance document for sampling and analysis
of toxic organic compounds in ambient air. Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards; EPA report no. EPA-600/4-83-027. Available from: NTIS,
Springfield, VA; PB83-239020.
Salas, L. J.; Singh, H. B. (1982) Measurements of formaldehyde and acetaldehyde
in the ambient air. Environ. Sci. Techno!.: submitted.
Salo, A. E.; Whitz, S.; MacPhee, R. D. (1975) Determination of solvent vapor
concentrations by total combustion analysis: a comparison of infrared
with flame ionization detectors. Presented at: the 68th annual meeting of
the Air Pollution Control Association; June; Boston, MA. Pittsburgh, PA:
Air Pollution Control Association; paper no. 75-33.2.
Saltzman, B. E.; Wartburg, A. F., Jr. (1965) Absorption tube for removal of
interfering sulfur dioxide in analysis of atmospheric oxidant. Anal.
Chem. 37: 779-782.
Sawicki, E.; Hauser, T. R.; Stanley, T. W.; Elbert, W. (1961). The 3-methyl-2-
benzothiazolone hydrazone test: sensitive new methods for the detection,
rapid estimation, and determination of aliphatic aldehydes. Anal. Chem.
33: 93-96.
Scaringelli, F. P.; Rosenberg, P. E.; Rehme, K. A. (1970) Comparison of per-
meation devices and nitrite ion as standards for the colorimetric deter-
mination of nitrogen dioxide. Environ. Sci. Technol. 4: 924-929.
Schiff, H. (1985). York University, Toronto, Canada. Personal communication:
work in progress.
Schiff, H. L; Pepper, D.; Ridley, B. A. (1979) Tropospheric NO measurements
up to 7 km. JGR J. Geophys. Res, 84: 7895-7897.
Scott, W.E.; Stephens, E.R.; Hanst, P.L.; Doerr, R.C. (1957). Further develop-
ments in the chemistry of the atmosphere. Proc. Amer. Petrol. Inst, 37
(III): 171-183.
4-100
-------
Seila, R. L.; Lonneman, W. A.; Meeks, S. A. (1976) Evaluation of polyv.inyl
fluoride as a container material for air pollution samples. J. Environ.
Sei. Health Part A 11: 121-130.
Sevcfk, J. (1975) Detectors in gas chromatography. Prague, Czechoslovakia:
Elsevier Scientific Publishing Company.
Severs, R. K. (1975) Simultaneous total oxidant and chemiluminescent ozone
measurements in ambient air. J. Air Pollut. Control Assoc. 25: 394-396.
Sexton, F. W. ; McElroy, F. F.; Michie, R. M., Jr.; Thompson, V. L. (1982) A
comparative evaluation of seven automated ambient non-methane organic
compound analyzers. Research Triangle .Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/4-82-046, Available from: NTIS,
Springfield, VA; PB82-230798.
Shanley. E. S. (1951) Hydrogen peroxide. J. Chem. Educ. 28: 260-266.
Singh, H. B. (1980). Guidance for the collection and use of ambient hydrocar-
bon species data in development of ozone control strategies. Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards; EPA final report no. EPA-450/4-80-008.
Available from: NTIS, Springfield, VA; PB80-202120.
Singh, H. B.; Hanst, P. L. (1981) Peroxyacetyl nitrate (PAN) in the unpolluted
atmosphere: an important reservoir for nitrogen oxides. Geqphys. Res.
Lett. 8: 941-944.
Singh, H. B.; Salas, L. J. (1983a) Peroxyacetyl nitrate in the free troposphere.
Nature (London) 302: 326-328.
Singh, H. B.; Salas, 1. J. (1983b) Methodology for the analysis of peroxyacetyl
nitrate (PAN) in the unpolluted atmosphere. Atmos. Environ. 17: 1507-1516.
Sleva, S. F. (1965) Determination of formaldehyde-chromotropiG acid method.
In: National Air Pollution Control Administration. Selected methods for
the measurement of air pollutants. Durham, NC: U.S. Department of Health,
Education, and Welfare, Public Health Service; publication no. 999-AP-ll.
Available from: NTIS, Springfield, VA; PB^-169677.
Smith, R. A.; Drummond, L (1979) Trace determination of carbonyl compounds in air
by gas chromatography of their 2, 4-dinitrophenylhydrazones. Analyst 104:
875-877.
Spicer, C. W.; Joseph, D. W.; Sticksel, R. R.; Ward, G. F. (1979) Ozone sources
and transport in the northeastern United States. Environ. Sci. Techno!.
13: 975-985.
Spicer, C. W.; Holdren, M. W,; Keigley, G. W. (1983) The ubiquity of peroxy-
acetyl nitrate in the continental boundary layer. Atmos. Environ. 17:
1055-1058.
4-101
-------
Stedman, D. H.; Kok, G.; Delumyea, R.; Alvord, H. H. (1976) Redundant calibra-
tion of nitric oxide, carbon monoxide, nitrogen dioxide, and o2one air
pollution monitors by chemical and gravimetric techniques. In: Calibration
in air monitoring, pp. 337-344. Philadelphia, PA: Amer. Soc. for Testing
and Materials. ASTM STP 598.
'i *; - P
Steinberg, M.; Dietz, R. N. (1969) Sr-90 ozone generator for sub-ppiri concentra-
tion range—instruction manual. Upton, Long Island, NY: Brookhaven National
Laboratory, Department of Applied Science.
Stephens, E. R. (1964) Absorptivities for infrared determination of peroxyacyl
nitrates. Anal. Chem. 36: 928-929.
Stephens, E. R. (1967) The formation of molecular oxygen by alkaline hydrolysis
of peroxyacetyl nitrate. Atmos. Environ. 1: 19-20.
Stephens, E. R. (1969) The formation, reactions, and properties of peroxyacyl
nitrates (PANs) in photochemical air pollution. In: Pitts, J. N.-, Jr.;
Metcalf, R. L., eds. Advances in environmental sciences v. 1. New York,
NY: Wiley-Interscience; pp. 119-146.
Stephens, E. R.; Price, M. A. (1973) Analysis of an important air pollutant:
peroxyacetyl nitrate. J. Chem. Educ. 50: 351-354.
Stephens, E. R.; Hanst, P. L.; Doerr, R. C.; Scott, W. E. (1956a) Reactions of
nitrogen dioxide and organic compounds in air. Ind. Eng. Chem. 48: 1498-1504.
Stephens, E. R.; Scott, W. E.; Hanst, P. L.; Doerr, R. C. (1956b) Recent
developments in the study of the organic chemistry of the atmosphere.
Presented at: 21st midyear meeting of the meeting of the American Petro-
leum Institute's Division of Refining; May; Montreal, PQ, Canada. Proc.
Am. Pet. Inst. 36 (III): 288-297.
Stephens, E. R.; Darley, E. F.; Taylor, 0. C.; Scott, W. E. (1961) Photochem-
ical reaction products in air pollution. Int. J. Air Water Pollut. 4:
79-100.
Stephens, E. R.; Burleson, F. R.; Cardiff, E. A. (1965) The production of pure
peroxyacyl nitrates. J. Air Pollut. Control Assoc. 15: 87-89.
Stevens, R. K.; Hodgeson, J. A. (1973) Applications of chemiluminescent reac-
tions to the measurement of air pollutants. Anal. Chem. 45: 443A-446A,
449A.
Stevens, R. K.; Clark, T. A.; Decker, C. E.; Ballard, L. F. (1972a) Field per-
formance of advanced monitors for oxides of nitrogen, ozone, sulfur
dioxide, carbon monoxide, methane and nonmethane hydrocarbons. Presented
at: 65th annual meeting of the Air Pollution Control Association; June;
Miami, FL. Pittsburgh, PA: Air Pollution Control Association.
Stevens, R. K.; Hodgeson, J. A.; Ballard, L. F.; Decker, C. E. (1972b) Ratio
of sulfur dioxide to total gaseous sulfur compounds and ozone to total
oxidants in the Los Angeles atmosphere - an instrument evaluation study.
In: Mamantov, G.; Shults, W. D., eds. Determination of air quality:
proceedings of the ACS symposium on determination of air quality; April
1971; Los Angeles, CA. New York, NY: Plenum Press; pp. 83-108.
4-102
-------
Sweeney, M. P.; Swartz, D. J.; Rost, G. A.; MacPhee, R.; Chad, J. (1964)
Continuous measurements of oxides of nitrogen in auto exhaust. J. Air
..Rollut. Control Assoc. 14: 249-254.
Tanner, R. L.; Meng, Z. (1984) Seasonal variations in ambient atmospheric
levels of formaldehyde and acetaldehyde. Environ. Sci. Techno!. 18:
723-726.
Taube, H.; Bray, W. C. (1940) Chain reactions in aqueous solutions containing
ozone, hydrogen peroxide and acid. J. Am. Chem. Soc. 62: 3357-3373.
Tokiwa, Y.; Twiss, S.; de Vera, E. R.; Mueller, P. K. (1972) Atmospheric ozone
determination by amperometry and colorimetry. In: Mamantov, G.; Shults,
W. D., eds. Determination of air quality: proceedings of the ACS symposium
on determination of air quality; April 1971; Los Angeles, CA. New York,
NY: Plenum Press; pp. 109-130.
Toyama, Y.; Kobayashi, J. (1966) Notes on an ozone generator and its calibra-
tor. Pap. Meteorol. Geophys. 17: 65-75.
Tuazon, E. C.; Graham, R. A.; Winer, A. M.; Easton, R. R.; Pitts, J. N., Jr.
Hanst, P. L. (1978) A kilometer pathlength Fourier-transform infrared
system for the study of trace pollutants in ambient and synthetic atmos-
pheres. Atmos. Environ. 12: 865-875.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1980) Atmos-
pheric measurements of trace pollutants by kilometer-pathlength FTIR
spectroscopy. Adv. Environ. Sci. Techno!. 10: 259-300.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1981a) Atmos-
pheric measurements of trace pollutants: long path Fourier-transform
infrared spectroscopy. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/3-81-026. Available from: NTIS,
Springfield, VA; PB81-179848.
Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Trace pollutant concen-
trations in a multiday smog episode in the California South Coast Air
Basin by long path length Fourier transform infrared spectroscopy.
Environ. Sci. Techno!: 15: 1232-1237.
Tucker, A. W.; Petersen, A. B.; Birnbaum, M. (1973) Fluorescence determination
of atmospheric NO and N02. Appl. Opt. 12: 2036-2038.
Tucker, A. W.; Birnbaum, M.; Fincher, C. L. (1975) Atmospheric N02 determi-
nation by 442-nm laser induced fluorescence. Appl. Opt. 14: 1418-1422.
U.S. Department of Health, Education, and Welfare. (1965) Selected methods for
the measurement of air pollutants. Durham, NC: National Air Pollution
Control Administration; publication no. PHS 999-AP-ll. Available from:
NTIS, Springfield, VA; PB-169677.
U.S. Department of Health, Education, and Welfare. (1973) Manual of analytical
methods, Vol. 1. Washington, DC: U.S. Government Printing Office. Cited
in: National Academy of Sciences. (1981) Formaldehyde and Other Aldehydes.
Washington, DC: National Academy Press.
4-103
-------
U.S. Environmental Protection Agency, (1975) Technical assistance document for
the chemiluminescence measurement of nitrogen dioxide. Researchfilriangle
Park, NC: U.S. Environmental Protection Agency; EPA report no. EPA-600/
4-75-003. Available from: NTISS Springfield, VA.
U.S.
• .-?•&
Environmental Protection Agency, (1976a) Ambient air monitoring references
and equivalent methods. Code of Federal Regulations, Title 40,;Part 53.
- '"•' *' '\
U.S. Environmental Protection Agency, (1976b) Nitrogen dioxide measurement
principle and calibration procedure. Notice of final rulemaking. Fed.
Regist. 41: 52686-52694, December 1.
U.S. Environmental Protection Agency. (1977a) Air monitoring strategy,for
state implementation plans. Research Triangle Park, NC: Office -of Air
Quality Planning and Standards, Standing Air Monitoring Group; EPA report
no. EPA-450/2-77-010. Available from: NTIS, Springfield, VA; PB-274014.
U.S. Environmental Protection Agency. (1977b) Quality assurance handbook for
air pollution measurement systems, volume II: ambient air specific methods.
Research Triangle Park, NC: U.S. Environmental Protection Agency; EPA
report no. EPA-600/4-77-027. Available from: NTIS, Springfield, VA;
PB-273518.
U.S. Environmental Protection Agency. (1978a) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Criteria and Assessment Office;
pp. 116-135; EPA report no. EPA-600/8-78-004. Available from: NTIS,
Springfield, VA; PB80-124753.
U.S. Environmental Protection Agency. (1978b) Measurement of volatile organic
compounds. Research Triangle Park, NC: U.S. Environmental Protection
Agency; EPA report no. EPA-450/2-78-041. Available from: NTIS, Springfield,
VA. - .
U.S. Environmental Protection Agency. (1981) Technical assistance document
for the calibration and operation of automated ambient non-methane organic
compound analyzers. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/4-81-015. Available from: NTIS,
Springfield, VA; PB82-147406.
U.S. Environmental Protection Agency. (1982) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: Environmental Criteria and
Assessment Office; EPA report no. EPA-600/8-82-026. Available from: NTIS,
Springfield, VA; PB83-131011.
Van de Wiel, H. J.; Reijnders, H. F. R.; Seifert, R.; Gies, H.; Lanting, R.
W.; Rudolf, W. (1979) International comparative programme on ozone refer-
ence methods. Atmos. Environ. 13: 555-557.
o
Vigroux, E. (1952) Spectrophotometry absorption of ozone at 18°C below 3130A.
Compt. Rend: 2529-2530.
Warren, G. J.; Babcock, G. (1970) Portable ethylene chemiluminescence ozone
monitor. Rev. SciVInstrum. 41: 280-282.
4-104
-------
Watanabe, L; Stephens, E, R. (1978) Reexamination of moisture anomaly in
analysis of peroxyacetyl nitrate. Environ. Sci. Techno!. 12: 222-223.
Wendel, G. J.; Stedman, D. H.; Cantre11, C. A. (1983) Lumino!-based nitrogen
dioxide detector. Anal. Chem. 55: 937-940.
Wendt;; (1975) Personal communication. Information regarding California air
quality data files. El Monte, CA: California Air Resources Board. i
West, P. W; Sen, B. Z. (1956) Spectrophotometric determination of traces of
formaldehyde. I. Anal. Chem. 153: 177-183.
Westberg, H. H.; Rasmussen, R. A.; Holdren, M. W, (1974) Gas chromatographic
analysis of ambient air for light hydrocarbons using a chemically bonded
stationary phase. Anal. Chem. 46: 1852-1854.
Westberg, H. H.; Holdren, M. W.; HiVI, H. H. (1980) Analytical methodology for
the identification and quantification of vapor phase organic pollutants.
New York, NY: Coordinating Research Council; CRC-APRAC project no. CAPA-
11-71.
Williams, D. T.; Kolitz, B. L. (1968) Molecular correlation spectrometry.
Appl, Opt. 7: 607-616.
Winer, A. M.; Peters, J. M.; Smith, J. P.; Pitts, J. N., Jr. (1974) Response
of commercial chemiluminescent NO-N02 analyzers to other nitrogen-containing
compounds. Environ. Sci. Technol. 8: 1118-1121.
Yoshizumi, K.; Aoki, K.; Nouchi, I.; Okita, T.; Kobayashi, T.; Kamakura, S.;
Tajima, M. (1984). Measurement of the concentration in rainwater and of
the Henry's Law constant of hydrogen peroxide. Atmos. Environ. 18:395-401.
Zaitsu, K.; Okhura, Y. (1980) New fluorogenic substrates for horseradish pero-
xidase: rapid and sensitive assays for hydrogen peroxide and peroxidase.
Anal. Biochem. 109: 109-113.
Zika, R. G.; Saltzman, E. S. (1982) Interaction of ozone and hydrogen peroxide
in water: implications for analysis of H202 in air. Geophys. Res. Lett.
9: 231-234.
Zocher, H.; Kautsky, H. (1923) Uber Lumineszenz bei chemischen Reaktionen.
[Luminescence during chemical reactions.] Naturwissenschaften 11: 194-199.
4-105
-------
-------
5. CONCENTRATIONS OF OZONE AND OTHER PHOTOCHEMICAL OXIDANTS IN AMBIENT AIR
5.1 INTRODUCTION
The data presented in this chapter on the concentrations of ozone and
other photochemical oxidants in ambient air are intended to support and com-
plement information presented in subsequent chapters on the effects of these
compounds. Thus, this chapter describes potential exposures of human popula-
tions, crops and other vegetation, ecosystems, and nonbiological materials in
general terms for the entire nation and in specific terms for selected areas
of the country. Since the health and welfare effects of ozone have been much
more thoroughly documented than those of other related oxidants, primary
emphasis in this chapter is placed on the concentrations of ozone found in
ambient air. Potential exposures are described by presenting data on peak and
average concentrations nationwide and on seasonal and diurnal patterns in
selected urban and nonurban areas. The recurrence on consecutive days of
selected levels of ozone has been examined for a few urban sites to aid in
understanding the significance of health effects documented in subsequent
chapters. Likewise, data have been included that portray representative urban
and rural concentrations by season and by time of day. Spatial variations in
ozone concentrations are briefly addressed since the effects on concentrations
of latitude, altitude, intracity variations, and indoor-outdoor gradients are
pertinent to the assessment of potential exposures of human populations, and,
except for the indoor-outdoor gradients, of crops and other vegetation and
ecosystems.
Ozone is the only photochemical oxidant other than nitrogen dioxide that
is routinely monitored and for which a comprehensive aerometric data base
exists. Data for peroxyacetyl nitrate (PAN) and its homologues and for hydro-
gen peroxide (H202) and formic acid (HCOOH) have all been obtained as part of
special research investigations. Consequently, no data on nationwide patterns
of occurrence are available for these non-ozone oxidants; nor are extensive
data available on the correlations of levels and patterns of these oxidants
with those of ozone. Data on these oxidants are considerably more abundant
now, however, than in 1978, when the previous criteria document for ozone and
other photochemical oxidants was published (U.S. Environmental Protection
Agency, 1978), Sections 5.6 and 5.7 summarize the available data on these
other oxidants.
5-1
-------
The concentrations of ozone and related photochemical oxidants observed
in ambient air are the net result, as shown in Chapter 3, of various combina-
tions of a variety of atmospheric processes, including:
1. Local photochemical production from oxides of nitrogen and
reactive volatile organic compounds.
2. Atmospheric mixing after sunrise, such that ozone-rich air from
above the nocturnal inversion layer is mixed to the surface,
resulting in a steady increase in ozone during the morning and
early afternoon. This increase, from ozone preserved aloft, is
independent of photochemistry, occurring even in the absence of
precursors and photochemical processes.
3. Transport of ozone produced photochemically but not locally.
4. Intrusion into the troposphere, even to ground level, of ozone-
rich air from the stratospheric reservoir.
5. Formation of ozone photochemically in the mid-troposphere, with
subsequent intrusion into the boundary layer.
6. Chemical scavenging in the atmosphere of ozone and other oxi-
dants; e.g., the reaction of ozone with nitric oxide (NO) or the
reaction of H202 with sulfur dioxide (S02).
7. Physical scavenging in the atmosphere of ozone and other oxi-
dants; e.g., the temperature-dependent decomposition of PAN,
the precipitation scavenging of H^O^, and the photolytic dis-
sociation of ozone.
8. Combined physical and chemical scavenging processes at the
surface of the earth; e.g., the deposition of ozone on reactive
biological or nonbiological surfaces, such as vegetation,
soils, or certain polymers.
These processes include, obviously, both manmade and natural processes
and driving mechanisms. The occurrence of high ozone concentrations is most
commonly associated with recognized meteorological conditions that involve
intense sunlight and elevated temperatures, and the variety of processes
involved contribute to strong diurnal cycles. Peak concentrations have been
observed to occur, however, at almost any time of day. Ozone may be trans-
ported after its formation for distances up to 1000 km or more (e.g., Hov et
a!., 1978; Wolff and Lioy, 1980). As a result, high concentrations of ozone
and related oxidants occur not only near large sources of precursors but also
in downwind nonurban areas, and usually later in the day at these downwind
receptor sites. Ozone, and apparently PAN, as well, can be transported at
5-2
-------
night above the surface pollutant and nocturnal inversion layer (Chapter 3),
but during daylight hours can be transported considerable distances at or near
ground level (Coffey and Stasiuk; 1975a,b).
The emphasis of this chapter is on the documentation of concentrations
rather than on explanations of possible causes of observed concentrations.
Probable causes and explanations are mentioned, however, where they are perti-
nent to the discussion.
Most of the data presented in this chapter to characterize both nationwide
and site-specific ozone concentrations in ambient air were obtained after
1978. Two factors influenced the use of post-1978 data. First, the current
Federal Reference Method for ozone, chemiluminescence, and the equivalent UV
method were almost universally employed by 1979. Second, EPA promulgated a UV
calibration method for ozone in 1979. Thus, these data form a relatively
homogeneous set for purposes of intercomparison. Because of the well-recog-
nized difficulties in converting from older data sets to the current reference
method, few pre-1979 aerometric data for ozone are presented in this chapter,
and then only to demonstrate spatial variations in concentrations that are
pertinent to exposure assessments.
5.2 TRENDS IN NATIONWIDE OZONE CONCENTRATIONS
Whether ozone concentrations in ambient air are static, rising, or declining
over time must be determined from statistical tests using comparable aerometric
data for a number of years. The national trend in ozone concentrations is
shown in Figure 5-1 for the 9-year period, 1975 through 1983 (U.S. Environmental
Protection Agency, 1984a). The concentrations depicted are the average
second-highest 1-hour concentrations for selected stations for each year.
In this context, the second-highest 1-hour value for each station is selected
from all daily maximum 1-hour values (n < 365) recorded per year at that station.
Subsequently, this statistic will be referred to as the second-highest daily
maximum 1-hour value or simply the second-highest 1-hour value. The 176
monitoring stations included in this analysis reported at least 50 percent of
the possible hourly values for the ozone season in at least 7 of these 9 years.
(The ozone season varies at the respective sites from 4 to 12 months, depending
upon local conditions. The sampling period is specified in State Implementation
Plans.)
5-3
-------
0.18
0.17
i
n.
a
cc
UJ
o
z
o
o
UJ
z
o
N
o
0.16
0.15
0.14
0.13
0.12
CA (27 Stations) _
O NAMS STATIONS (62)
T 35% CONFIDENCE
1. INTERVALS
O ALL STATIONS (176)
T 95% CONFIDENCE
i INTERVALS
A CALIFORNIA STATIONS (27)
i? ALL STATIONS EXCEPT
CALIFORNIA (149)
I i I I
1975 1976 1977 1978 1979 1980 1981 1982 1983
YEAR
Figure 5-1. National trend in composite average of the second highest
value among daily maximum 1-hour ozone concentrations at selected
groups of sites, 1975 through 1983.
Source: U.S. Environmental Protection Agency (1984a).
5-4
-------
Ambient air monitoring stations in the nation are generally operated by
state and local agencies, but among the stations is a small group of National
Air Monitoring Stations (NAMS) whose data are reported directly to EPA. The
trend line for the subset of the 62 NAMS ozone stations, also shown in Figure
5-1, tracks fairly closely the trend for all 176 stations. To permit a com-
parison of aerometric data for California, a high-oxidant area, with the
nationwide trends data, data for California and data for all states other than
California are also plotted separately in Figure 5-1 and are compared below.
For the entire 9-year period, 1975 through 1983, all subsets of monitoring
stations show a decline in the composite second-highest daily maximum 1-hour
ozone concentration. In 1979, a new, more accurate ozone calibration procedure
was promulgated by the U.S. Environmental Protection Agency (see Chapter 4).
Between 1979 and 1982, a small decline of 9 to 10 percent in nationwide ozone
concentrations occurred. From 1982 to 1983, however, concentrations increased
by about 10 percent in California, by about 12 percent nationwide, and by about
14 percent outside California. Recently published data for 1984 from a somewhat
smaller number of stations (163 sites), not depicted in Figure 5-1, show a
decrease in second-highest 1-hour concentrations nationwide, with ozone levels
in 1984 approximating those recorded in 1981, which are shown in Figure 5-1
(U. S. Environmental Protection Agency, 1986).
Note the influence of ozone data from Calfornia. The California data
heavily influence the trends data because of the number of California monitor-
ing sites represented in the data and because of the actual concentrations of
ozone in California.
Evaluation of national trends, as well as local or regional trends, in
reported concentrations of ozone in ambient air over the past 5 to 10 years is
complicated by a number of factors, including: (1) the change in calibration
procedure recommended by EPA in 1978 and promulgated in 1979 (see Chapter 4);
(2) the possible effects on aerometric data of quality assurance procedures
instituted by EPA in 1979; (3) the influence of diverse regional meteorological
conditions; and (4) changes in precursor emissions.
How much of the observed decline in ozone concentrations from 1975 through
1982 is attributable to the 1979 promulgation of the ultraviolet (UV) calibra-
tion method as the Federal Reference Method is uncertain. To determine that,
the monitoring practices at each of the 176 sites would have to be examined in
detail to find out the calibration methods used, the date when the UV method
5-5
-------
was adopted for this site, and whether states applied correction factors to
prior data (see, e.g., Walker, 1985; Hoggan, 1986; Walker, 1986). Such an
examination would be necessary because not all monitoring sites switched to
the use of the UV method simultaneously. For example, the state of California
(in EPA Region IX) had already been using the UV method before it was
promulgated in February 1979. In addition, other states, in other regions,
may have used the boric acid-potassium iodide (BAKI) method before, after, or
both before and after promulgation of the UV method, since the BAKI procedure
was allowed by EPA as an interim method for 18 months following the 1979 UV
•promulgation (see Chapter 4). Likewise, other states used gas-phase titration
prior' to 1979 but either BAKI or UV procedures following the UV promulgation.
In addition, random errors associated with individual operator practices
occur. Thus, the relationships among these three methods, even if monitoring
practices at individual sites were known, are complex and would preclude the
simple application of a single correction factor to these trend data (see
Chapter 4).
Hunt and Curran (1982) have noted that only EPA Region IX showed improve-
ment in ozone air quality between 1979 and 1981 but, in contrast to other EPA
regions, showed no improvement in that period versus the 1975 to 1979 period.
Since California, which dominates Region IX, changed its calibration method in
1975, the decrease in ozone concentrations seen in California from 1979 through
19S2 (Figure 5-1) cannot be attributed to the introduction of the UV calibra-
tion method. The shape of the trend line for California is quite similar to
that for the rest of the nation over the 4-year period (1980 through 1983)
following the promulgation of the UV method. From 1982 to 1983, an increase
in ozone concentrations occurred nationwide, with the percentage increase in
California roughly paralleling that for the rest of the nation. Increased
precursor emissions and meteorological conditions conducive to oxidant forma-
tion appear to be the most likely causes of the increase in 1983 in the com-
posite average of the second-highest daily maximum 1-hour ozone concentrations.
An examination of emissions data, however, indicates that NO emissions did
){
not change significantly from 1982 to 1983 and that VOC emissions rose only 3
percent in that period (U;S. Environmental Protection Agency, 1984b; also see
Chapter 3). Therefore, precursor emissions are not thought to account for all
of the 12 percent increase in 0- from 1982 to 1983 nationwide (U.S. Environ-
mental Protection Agency, 1984a).
5-6
-------
The influence of meteorological conditions on ozone concentrations has
been explored for 10 cities using an experimental index based on meteorological
parameters conducive to ozone formation (U.S. Environmental Protection Agency,
1984a). The index suggests that the potential for ozone formation was greater
in 1983 than in 1981 and 1982; however, the index also indicates that even
greater potential for ozone formation existed in 1980 than in 1983 and that
conditions in 1979 might be comparable to those in 1981 and 1982. Thus, this
simple index may indicate the direction but not the magnitude of the effects
of meteorological factors on ozone formation.
From a study of climate in the upper Midwest during the summer of 1983,
Wendland et al. (1984) have reported that temperatures, which correlate reason-
ably well with insolation, were generally higher in the summer of 1983 than
the 1950~through-1980 norm for the 12-state region; and that cooling-degre.e
days over about one-third of the region were 50 percent higher than normal.
While not quantitative or conclusive, these studies do suggest that meteorolo-
gical conditions in 1983 contributed to the increase in the second-highest
value among daily maximum 1-hour ozone concentrations.
Trends in the composite average second-highest daily maximum 1-hour ozone
values in the 10 EPA regions are shown in Figure 5-2 (U.S. Environmental
Protection Agency, 1984a). The use of data beginning with 1979 avoids some of
the potential effects of changes in calibration and quality assurance procedures
mentioned earlier. Nine of the 10 regions show a 7 to 15 percent decrease in
this statistic from the 1979-1980 period to the 1981-1982 period. The same
nine regions showed increases of 6 to 16 percent from the 1981-1982 period to
1983, demonstrating the pervasiveness of the trend. Only in the Pacific
Northwest, Region X, was there an opposite trend: +6 percent in 1981-1982 and
-9 percent in 1983.
5.3 OVERVIEW OF OZONE CONCENTRATIONS IN URBAN AREAS
An overview of nationwide urban ozone concentrations for 1981 is,provided
in Figures 5-3 and 5-4, which depict graphically the average daylight concentra-
tions. Figure 5-3 shows data for spring and summer months, the months that make
up the smog season in most though not all areas of the nation; and Figure 5-4
shows daylight concentrations during the fall and winter months. The daylight
period of 6:00 a.m. to 8:00 p.m. includes the hours of greatest human activity
5-7
-------
0,24-1
0.20
E
D.
O.
Z
O
Z
ui
o
z
o
u
UJ
Z
o
N
o
0.16-
0,12-
0.08-
0.04-
0.00
EPA REGION3
NO. OF SITES
1979-1980 COMPOSITE AVERAGE
1981 -1982 COMPOSITE AVERAGE
1983 COMPOSITE AVERAGE
II
19
III
26
IV
14
V
40
VI
15
VII VIII
8 7
IX
36
X
5
Figure 5-2. Comparison of the 1979-1980, 1981-1982, and 1983
composite average of the second highest daily maximum 1-hour ozone
concentrations across EPA Regions.
aWote: Regions are composed of the following states:
I CT, MA, ME, NH, RI.VT
II NJ, NY, PR, VI
ill DE, MD, PA, VA, WV
IV AL, FL, GA. KY, MS, NC, SC, TN
V IL, IN, Ml, MM, OH,WI
VI IA, KS, MO, NE
VII AR, LA, NM, OK, TX
VIII CO, MT, ND, SD. UT, WY
IX AZ, CA. HI. NV
X AK, ID, OR, WA
Source: U.S. Environmental Protection Agency (1984a).
5-8
-------
en
10
.14 - .16 PPM
.16 - .18 PPM
Figure 5-3. Average daylight (6:00 a.m. to 8:00 p.m.) concentrations of ozone in the second
and third quarters (April through September), 1981.
01 Apr 1983
Source: SAROAD (19S5d). Derived by G. Duggan, OAQPS.
-------
tn
i
.16 - .18 PPM
Figure 5-4. Average daylight (6:00 a.m. to 8:00 p.m.) concentrations of ozone in the first and
fourth quarters (January through March and October through December), 1981. 01 Apr 1983
Source: SAROAD (1985d). Derived by G. Duggan, OAQPS,
-------
outdoors; the hours when exposure of vegetation and ecosystems would be expected
to have the greatest consequences (stomata are open in daylight and photosyn-
thesis is taking place; see Chapter 6); and the hours of greatest local forma-
tion of ozone and other oxidants via photochemistry in the atmosphere (see
Chapter 3). The average concentrations during the spring and summer months
(second and third quarters of the year) are clustered mainly in the 0.04 to
0.06 ppm range. Averages for the winter and fall months (first and fourth
quarters) are clustered mainly in the 0.02 to 0.04 ppm range.
The stations used in Figures 5-3 and 5-4 reported at least 75 percent of
the possible 1-hour values per quarter. Some stations, however, monitor ozone
only during the months when the potential for photochemical ozone formation is
significant in those localities. Also, certain areas of the United States are
not monitored routinely for ozone because of the lack of emission sources or
transport events and thus the low potential for significant ozone or oxidant
concentrations. The Great Basin and the Great Plains, for example, are such
areas.
Figure 5-5 shows the nationwide frequency distributions of the first-,
second-, and third-highest 1-hour 0- concentrations at predominantly urban
stations aggregated for 1979, 1980, and 1981, as well as the highest 1-hour 03
concentration at sites of the National Air Pollution Background Network (NAPBN)
(see Section 5.4.1) aggregated .for the same 3 years. Only data collected by
the Federal Reference Method (chemiluminescence) or the equivalent UV method
(see Chapter 4) have been used in this analysis. A "valid site" is one report-
ing at least 75 percent of the 8760 possible 1-hour values in a year. There
were 282 such sites in 1979, 266 in 1980, and 358 in 1981 (U.S. Environmental
Protection Agency, 1980, 1981, 1982a). As shown by Figure 5-5, 50 percent of
the second-highest 1-hour values from non-NAPBN sites in this 3-year period
were 0.12 ppm or less and about 10 percent were equal to or greater than 0.20
ppm. At the NAPBN sites, the collective 3-year distribution (1979 through
1981) is such that about 60 percent of the values are less than 0.10 ppm and
fewer than 20 percent are higher than 0.12 ppm.
Table 5-1 lists the second-highest 1-hour 0- values reported for 1981
through 1983 for the 80 most populous Standard Metropolitan Statistical Areas
(SMSAs), grouped by population (see Table 5-1 for a definition of the "second-
highest" values). Collectively these SMSAs account for about 53 to 54"percent
of the 1981 United States population of 229.3 million. The significant obser-
vation to be drawn from this table of second-highest values is that the lowest
5-11
-------
99.999.8
99 98 " iS 90
70 60 SO 40 30 20
10
2 1 0.5 02 0.1 0.05 0.01
0.45
0.40
OJi
o.
a 030
»
O
oc
01
o
z
O
O
IU
Z
O
N
O
0.2S
0.20
0.15
0.10
0.05
I I i i I i I I I ! I I I I I
HIGHEST
— — 2nd-HIGHEST
3rd HIGHEST
....... HIGHEST, NAPBN SITES
I i I i i I i i _L i I i i i i I I II
I I
0.01 0.05 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9
STATIONS WITH PEAK 1-hour CONCENTRATIONS < SELECTED VALUE, percent
99.99
Figure 5-5. Distributions of the three highest 1 -hour ozone concentrations at valid sites
(906 station-years) aggregated for 3 years (1979,1980, and 1981) and the highest
ozone concentrations at NAPBN sites aggregated for those years (24 station-years).
Source: U.S. Environmental Protection Agency (1980, 1981, 1982).
-------
TABLE 5-1. SECOND-HIGHEST 1-hr OZONE CONCENTRATIONS*>b REPORTED FOR
STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION,
1981 THROUGH 1983
Ozone concentration, ppm
Standard Metropolitan Statistical Area 1981
1982
1983
Population >2 million
New York, NY - NJ 0.18
Los Angeles - Long Beach, CA 0.35
Chicago, IL 0.14
Philadelphia, PA - NJ 0.17
Detroit, MI 0.15
San Francisco - Oakland, CA 0.14
Washington, DC - MD - VA 0.15
Dallas - Fort Worth, TX 0.15
Houston, TX 0.23
Boston, MA 0.13
Nassau - Suffolk, NY 0.14
St. Louis, MO - IL 0.15
Pittsburgh, PA 0.16
Baltimore, MD 0.17
Minneapolis - St. Paul, MN - WI 0.10
Atlanta, GA 0.14
Summary statistics:
Minimum 1-hour value 0.10
Median 1-hour value 0.15
Maximum 1-hour value 0.35
Population 1 to < 2 million
Newark, NJ 0.14
Anaheim - Santa Ana - Garden Grove, CA 0.31
Cleveland, OH : 0.12
San Diego, CA 0.24
Miami, FL 0.14
Denver - Boulder, CO 0.13
Seattle - Everett, WA 0.12
Tampa - St. Petersburg, FL 0.12
Riverside - San Bernardino - Ontario, CA 0.34
Phoenix, AZ 0.16
Cincinnati, OH - KY - IN 0.13
Milwaukee, WI 0.17
Kansas City, MO - KS 0.12
San Jose, CA 0.14
Buffalo, NY 0.12C
Portland, OR - WA 0.15
New Orleans, LA ' 0.12
Indianapolis, IN 0.13
Columbus, OH . 0.11
0.17
0.32
0.12
0.18
0.16
0.14
0.15
0.17
0.21
0.16°
0.13
0.16
0.14
0.14
0.10
0.14
0.10
0.15
0.32
0.17
0.18C
0.13
0.21
0.14
0.14
0.09
0.12
0.32
0.12
0.13
0.13
0.10
0.14
0.11
0.12
0.17
0.12
0.13
0.19
0.37
0.17
0.10
0.17
0.17
0.17
0.16
0.28
0.18
0.17
0.18
0.14
0.19
0.13
0.17
0.10
0.17
0.37
0.25
0.28
0.15
0.20
0.12
0.14
0.10
0.14
0.34
0.16
0.15
0.18
0.13
0.16
0.12
0.12
0.12
0.14
0.12
5-13
-------
TABLE 5-1 (continued). SECOND-HIGHEST 1-hr OZONE CONCENTRATIONS3'b REPORTED
FOR STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION,
1981 THROUGH 1983
Standard Metropolitan Statistical Area
San Juan, PR
San Antonio, TX
Fort Lauderdale - Hollywood, FL
Sacramento, CA
Summary statistics:
Minimum 1-hour value
Median 1-hour value
Maximum 1-hour value
Population 0.5 to < 1 million
Rochester, NY
Salt Lake City - Ogden, UT
Providence - Harwick - Pawtucket, RI - MA
Memphis, TN - AR - MS
Louisville, KY - IN
Nashville - Davidson, TN
Birmingham, AL
Oklahoma City, OK
Dayton, OH
Greensboro - Winston-Salem - High Point, NC
Norfolk - Virginia Beach - Portsmouth, VA - NC
Albany - Schenectady - Troy, NY ,
Toledo, OH - MI
Honolulu, HI
Jacksonville, FL
Hartford, CT
Orlando, FL
Tulsa, OK
Akron, OH
Gary - Hammond - East Chicago, IN
Syracuse, NY
Northeast Pennsylvania
Charlotte - Gastonia, NC
Allentown - Bethlehem - Easton, PA - NJ
Richmond, VA
Grand Rapids, MI
New Brunswick - Perth Amboy - Sayreville, NJ
West Palm Beach - Boca Raton, FL
Omaha, NE - IA
Greenville - Spartanburg, SC
Jersey City, NJ
Austin, TX
Ozone
1981
0.07
0.12
0.11
0.17
0.07
0.13
0.34
0.12
0.16
0.15
0.14
0.14
0.13
0.16
0.11
0.13
0.11
0.12
0.12
0.13
0.05
0.11
0.15
0.10
0.15
0.24e
0.14
0.11
0.10
0.12
0.12
0.12
0.11
0.13
0.09
0.08
0.11
0.14
0.12
concentration,
1982
0.04C
0.14
0.09
0.16
0.04
0.13
0.32
0.11
0.14
0.15
0.13
0.14
0.11
0.15 '
O.llc
0.16
0.11
0.11
0.12
0.13
0.07
0.12
0.17r
0.10C
0.13
0.14
0.13
0.12
0.16
0.12
0.15
0.12
0.11
0.16
0.09
0.09
0.11
0.14
0.11
ppm
1983
NDd
0.12
0.10
0.15
0.10
0.14
0.34
0.12
0.14
0.15
0.15
0.16
0.12
0.15
0.11
0.13
0.12
0.13
0.12
0.13
0.06
0.14
0.19
0.11
0.13
0.13
0.17
0.08
0.13
0.15
0.14
0.14
0.14
0.19
0.09
0.09
0.11
0.16
0.12
5-14
-------
TABLE 5-1 (continued). SECOND-HIGHEST 1-hr OZONE CONCENTRATIONS9>b REPORTED
FOR STANDARD METROPOLITAN STATISTICAL AREAS HAVING POPULATIONS >0.5 MILLION,
1981 THROUGH 1983
Standard Metropolitan Statistical Area
Ozone concentration, ppm
1981
1982
1983
Youngstown ~ Warren, OH
Tucson, AZ
Raleigh - Durham, NC
Springfield - Chicopee - Holyoke, MA - CT
Oxnard - Simi Valley - Ventura, CA
Wilmington, DE - NJ - MD
Flint, MI
Fresno, CA
Long Branch - Asbury Park, NJ
0.13
0.10
0.12
0.16
0.20
0.12C
0.11
0.11
ND
0.11
0.12
0.09
0.16
0.22
0.16
0.11
0.16
ND
0.11
0.11
0.13
0.19
0.21
0.18
0.11
0.16
ND
Summary statistics:
Minimum 1-hour value
Median 1-hour value
Maximum 1-hour value
0.05
0.12
0.24
0.07
0.12
0.22
0.06
0.13
0.21
d
The maximum 1-hour value for each day forms the data set from which the highest
and "second-highest" concentrations are determined. Thus, the "second-highest"
1-hour concentration occurs on a different day from the highest and therefore
is not necessarily the absolute second-highest concentration for the monitoring
period.
These values permit a comparison of potential exposures and are not necessarily
equivalent to the "design value" used for control strategy development.
'Less than 50% of days in ozone season.
ND = no data.
^Questionable data; third-highest value for 1981 was 0.125 ppm.
Source: U.S. Environmental Protection Agency (1984a).
5-15
-------
median concentrations in 1981, 1982, and 1983, for SMSAs having populations of
0.5 to 1 million, are 0.12, 0.12, and 0,13 ppm, values that equal or exceed
the current national ambient air quality standard for ozone. With the possible
exception of the concentration reported for Akron, Ohio, in 1981, which appears
to be questionable (see footnote, Table 5-1), the highest of all the second-
highest 1-hour concentrations in these 80 SMSAs occurred in California in all
3 years.
5.4 OVERVIEW OF OZONE CONCENTRATIONS IN NONURBAN AREAS
As mentioned in the preceding section, very few ozone monitoring stations
are located in nonurban areas. Consequently, the aerometric data base for
nonurban areas is not comparable to that for urban areas. The nonurban data
presented in this section were obtained from two special-purpose monitoring
networks that were designed to measure ozone concentrations at sites specifi-
cally selected to represent a variety of pristine, rural, or suburban environ-
ments. These sites do not all represent areas totally unaffected by manmade
ozone or its precursors, as shown by the fact that some data records contain a
significant number of high values that are best explained as resulting from
the transport of ozone or its precursors from upwind urban areas. The data
given here are intended to show an overview of nonurban concentrations in
areas with relatively infrequent urban influences. Additional data on speci-
fic rural areas are presented in Sections 5.5.1 and 5.5.2.
5.4.1 National Air Pollution Background Network (NAPBN)
The NAPBN consists of eight stations located in eight National Forests
(NF) across the country (Figure 5-6). The first three stations began opera-
tion in 1976 (Green Mountain NF, Vermont; Kisatchie NF, Louisiana; and Custer
NF, Montana); the second three in 1978 (Chequamegon NF, Wisconsin; Mark Twain
NF, Missouri; and Croatan NF, North Carolina); and the last two in 1979 (Apache
NF, Arizona; and Ochoco NF, Oregon). Yearly summaries of ozone concentrations
through 1980 are shown in Table 5-2 for the three sites established first
(Evans et a!., 1983). The principal points of interest in these summary sta-
tistics are the range and the arithmetic mean of the ozone concentrations
measured at these National Forest sites. The arithmetic mean concentrations
for the three sites ranged from 0.027 ± 0.015 ppm at the Kisatchie NF site in
5-16
-------
CROATAN IMF
Figure 5-6, Locations of the eight national forest (IMF) stations
constituting the National Air Pollution Background Network
(NAPBN).
Source: Evans et al. (1983),
5-17
-------
TABLE 5-2. ANNUAL OZONE SUMMARY STATISTICS FOR THREE SITES OF THE
NATIONAL AIR POLLUTION BACKGROUND NETWORK
Site
Kisatchie NFS LA
en
i
co Custer NF, MT
Green Mt. NF, VT
Year
1976
1977
1978
1979
1980
1976
1977
1978
1979
1980
1976
. 1977
1978
1979
1980
No. 1-hr
meas.
3448
6793
5636
6993
4438
275
7603
7674
8488
7754
1058
6483
3671
6423
8574
% of
possible
1-hr meas.
39,4
77.5
64.3
79.8
50.7
3.1
86.8
87.6
96.9
88.5
12.1
74.0
41.9
73.3
97.9
Concn. , ppm
Concn.
M1n.
NDa
ND
ND
ND
ND
0.020
ND
ND
ND
ND
ND
ND
ND
ND
ND
, Ppm
Max.
0.125
0.135
0.125
0.100
0.105
0.060
0.080
0.075
0.070
0.070
0.060
0.145
0.105
0.105
0.115
Arith.
mean
0.032
0.033
0.034
0.027
0.028
0.039
0.040
0.030
0.032
0.037
0.029
0.038
0.029
0.032
0.032
Arith.
std. dev.
0.021
0.023
0.021
0.015
0.016
0.008
0.011
0.017
0.012
0.012
0.011
0.021
0.018
0.017
0.017
Concn. , ppm
Geom.
mean
0.024
0.025
0.027
0.023
0.023
0.038
0.039
0.023
0.029
0.035
0.026
0.031
0.024
0.027
0.027
Geom.
std. dev.
2.19
2.25
2.14
1.92
1.94
1.22
1.37
2.14
1.59 .
1.41
1.76
2.00
2.01
1.86
1.90
aND = not detectable.
Source: Evans et al. (1983)
-------
1979 to 0.040 ± 0.011 ppm at the Custer NF site in 1977. The arithmetic mean
concentration across all years and all three sites was 0.033 ppm. Fluctua-
tions in the observed concentrations from year-to-year and site-to-site are
demonstrated by the range of concentrations measured and by the size of the
standard deviations as well. The lowest concentrations seen were below the
limits of detection of the chemiluminescence monitor employed, but the highest
concentrations observed at the Kisatchie NF and Green Mt. NF sites were both
above the present ozone standard of 0.12 ppm. In 1979, only one excursion
over 0.12 ppm 03 was recorded at an NAPBN site. That excursion was recorded
at the Mark Twain NF site in Missouri (Evans et al., 1983; Lefohn, 1984). In
1980, seven 1-hour excursions over 0.12 ppm were reported for Croatan NF, North
Carolina (Lefohn, 1984). None of the NAPBN sites recorded an ozone concentra-
tion greater than 0.12 ppm in 1981 (Evans et al., 1983).
These summary statistics show somewhat higher mean concentrations, lower
maximum concentrations, and lower, standard deviations in data obtained at the
Custer NF site than at the other two, which may indicate that meteorological
conditions are less variable at that site or that the site is much less affec-
ted, if not altogether unaffected, by manmade ozone or its precursors.
During a 6-day period in 1979, the NAPBN site in the Mark Twain NF,
Missouri, showed ozone concentrations well in excess of typical values.
Table 5-3 shows the peak 1-hour value for each of the 6 days. A 1-hour value
of 0.125 ppm, the maximum observed at any NAPBN site in 1979, was measured at
that site on July 21, 1979. Evans et al. (1983) calculated the trajectories
of air masses reaching the site during the 6-day period of July 18 through
July 23, 1979. They ascribed the unusually high values, including the peak
value on the 21st, to pollutants picked up as the trajectory passed over urban
areas in the Ohio River Valley and the Great Lakes region. Figure 5-7 shows
the trajectories for the air parcels reaching the Mark Twain NF site at midnight
(0000), 8 a.m. (0800), noon (1200), and 6 p.m. (1800) on July 21, 1979. On
July 23, clouds and rain spread over the region and the airflow trajectories
shifted to the east and south, reducing both the quantities of transported
precursors and the potential for photochemical ozone generation.
More recent and more comprehensive data from the NAPBN sites are presented
in Table 5-4 and in Figures 5-8A and 5-8B. Table 5-4 presents the percentile
distributions of ozone concentrations at all eight of the sites, aggregated by
quarter across several years.
5-19
-------
TABLE 5-3. CONCENTRATIONS OF OZONE DURING 6-day PERIOD OF HIGH
VALUES AT NAPBN SITE IN MARK.TWAIN NATIONAL FOREST, MISSOURI, 1979
Date
July 18
July 19
July 20
July 21
July 22
July 23
1-hr maximum
03 concentration, ppm
0.080
0.100
0.115
0.125
0.120
0.050
Source: Evans et al. (1983)
As the data in Table 5-4 show, the arithmetic mean concentrations of
ozone, for the years of data averaged, are generally higher in the second
quarter of the year (April, May, June) than at other times at the NAPBN sites.
Although a few excursions of the 1-hour concentration above 0.12 ppm were
recorded at some of these sites, as discussed above (Evans et al., 1983;
Lefohn, 1984), the distribution given in Table 5-4 clearly indicates that 99
percent of the 1-hour ozone concentrations measured at these sites are well
below the present ozone standard. This is true even at those sites thought to
be influenced by transport of ozone (Green Mt. NF) or demonstrated to be
influenced by transport (Mark Twain NF, Evans et al., 1983). As shown in
Table 5-4, the highest 99th percent!le value was 0.093 ppm, reached at both
the Green Mt, NF, VT, site (second quarter) and the Mark Twain NF, MO, site
(third quarter). The maximum 1-hour ozone concentrations at these sites ranged
from 0.050 ppm at Custer NF, MT (in the fourth quarter) to 0.155 ppm at Mark
Twain NF, MO (in the third quarter). The second-highest 1-hour concentration
among maximum daily 1-hour values ranged from 0.050 ppm at Custer NF, MT
(fourth quarter) to 0.150 ppm at Mark Twain NF, MO (third quarter).
Five of the NAPBN monitoring stations (Apache NF, AZ; Mark Twain NF, MO;
Custer NF, MT; Croatan NF, NC; Ochoco NF, OR) reported sufficient data for
1979 through 1983 to support an examination of their second-highest 1-hour
5-20
-------
MINNEAPOLIS
1200 1
MILWAUKEE
DES MOINES
> KANSAS CITY
ST. LOUIS
MARK TWAIN N.F.
INCINNATI
^
LOUISVILLE
Figure 5-7. Trajectory analysis plots for the
NAPBN site at Mark Twain National Forest, MO,
July 21, 1979 (distance between bars represents
12hr).
Source: Evans et al. (1983).
5-21
-------
TABLE 5-4. PERCENTILE DISTRIBUTIONS OF OZONE CONCENTRATIONS
AT SITES OF NATIONAL AIR POLLUTION BACKGROUND NETWORK,
AGGREGATED BY QUARTER ACROSS SEVERAL YEARS
(ppb)
en
i
Site, yr» No. 1-hr
and qtr meas.
Apache, AZ
Ki satchi e ,
Mark Twain,
Custer, MT
Croatan, NC
(1980-1983)
1
2
3
4
LA (1977-1980, 1982)
1
2
3
4
MO (1979-1983)
1
2
3
4
(1979-1983)
1
2
3
4
(1978-1983)
1 • • ••••••
2
3
4
7587
7971
8407
8537
7093
6333
5462
6872
9484
10294
9155
8624
6675
8646
8751
9956
10640
11491
8389
12036
Percentile
10%
32
40
28
27
16
14
6
8
14
29
23
11
22
30
27
18
10
12
5
4
50%
39
50
40
35
32
37
22
22
32
46
42
27
31
40
39
28
26
36
23
18
80%
45
56
46
39
43
55
42
37
41
60
55
38
37
46
47
32
i
37
51
43
29
90%
49
59
51
40
53
65
53
48
47
68
65
46
40
50
50
35
44
60
53
37
95%
50
64
54
44
60
72
60
56
52
73
75
57
43
53
53
36
50
66
60
41
2nd
highest
W%
54
68
60
46
73
85
70
77
63
85
93
72
50
59
58
40
61
79
72
54
1-hr
65
90
85
55
120
115
110
90
80
115
150
85
65
80
85
50
85
110
• 80
85
Max
1-hr
75
90
90
55
125
135
110
105
85
115
155
100
65
80
85
50
95
150
95
85
Arith.
mean
39.3
49.4
39.5
34.2
32.4
38.7
25.8
25.7
31.0
47.4
43.2
28.6
30.9
40.3
38.6
26.9
27.4
36.1
27.0
19.8
-------
TABLE 5-4 (continued). PERCENTILE DISTRIBUTIONS OF OZONE CONCENTRATIONS
AT SITES OF NATIONAL AIR POLLUTION BACKGROUND NETWORK,
AGGREGATED BY QUARTER ACROSS SEVERAL YEARS
(ppb)
(Jl
ro
CO
Site, yr,
and qtr
Ochoco, OR
Green Mt. ,
Chequamegon
(1980-1983)
1
2
3
4
VT (1977-1981)
1
2
3
4
, WI (1979-1981)
1
2
3
4
No. 1-hr
meas.
7236
7861
8041
7467
7387
7752
8636
8712
5548
6085
4577
5909
Percent! le
10%
26
28
26
22
21
17
5
11
22
32
15
15
50%
35
38
38
31
31
43
26
25
36
45
30
25
80%
37
45
46
32
41
58
44
33
45
58
45
30
90%
40
46
49
36
45
69
56
37
51
68
53
32
95%
42
51
53
39
48
81
65
41
56
75
58
35
2nd
highest
99%
43
54
59
41
61
93
86
54
65
88
75
43
1-hr
55
60
75
55
120
140
110
75
80
110
90
55
Max
1-hr
55
75
80
60
135
145
115
85
80
115
95
60
Arith.
mean
33.0
37.1
37.8
29.2
32.8
42.2
29.2
24.8
36.1
48.3
33.0
23.7
Data are weighted by the number of 1-hr concentrations measured. Since data records were
inadequate for some sites for some years, the years of data presented differ from site
to site. The percentile distributions were derived from all 1-hr values for each quarter
and all years listed. The maximum 1-hr and second-highest 1-hr concentrations represent
the single value for each out of the entire data record for respective sites.
b •' '•"'"
These are the two highest values in the "daily maximum 1-hour" data set.
Source: Derived frojji Ivans (1985). ,,'.- ...: •_., .-.,.', =.-' ._'.. v ., :1 ;;', -^.
-------
a
a
%
Z
o
§
I
ui
o
z
o
u
ui
z
o
a.
a.
*,
O
§
1
ui
U
Z
o
u
UI
z
o
M
o
1979
1380
1981
PERIOD, years
1982
1983
Figure 5-8A. Second-highest value among maximum 1-hr ozone
concentrations at five IMAPBRI monitoring stations, 1979 through
1983.
Source; Derived from SAROAD (1985 b-f).
0.14
0.12
0,10
0,08
0.06
0.04
I I I
T 95% CONFIDENCE INTERVAL
T
I
I
-L
1979
1980
1981
1982
1983
PERIOD, years
Figure 5-8B. Composite averages of the second-highest value
among daily maximum 1 -hr ozone concentrations at five NAPBN
stations, 1979 through 1983.
Source: Derived from SAROAD (1985 b-f).
5-24
-------
ozone values for evidence of trends. As seen in Figure 5-8B, however, the
diversity of the NAPBN site characteristics, the consequent large variation in
recorded values, and the small sample size result in such a large confidence
interval around the average for each year that no trend can be determined.
The data presented in Figures 5-8A and 5-8B show the NAPBN second-highest
daily maximum 1-hour ozone concentrations to be about one-half those reported
for urban-suburban sites for the same period (Figure 5-1).
5.4.2 Sulfate Regional Experiment Sites (SURE)
As part of a comprehensive air monitoring project sponsored by the Electric
Power Research Institute (Martinez and Singh, 1979), ozone measurements were
made by the chemiluminescence method in the last 6 months of 1977 at nine
"nonurban" SURE sites in the eastern United States, shown in Figure 5-9. On
the basis of diurnal NO patterns that indicated the influence of traffic
)\
emissions, five of the sites were classed as "suburban"; the other four were
classed as "rural." The ozone data from these nine stations are summarized in
Table 5-5. Martinez and Singh (1979) noted that the four rural stations
occasionally recorded high values comparable with those in urban areas, but that
the incidence was low. They concluded that infrequent transport of ozone or
its precursors, or both, rather than local ozone generation, was the most prob-
able cause of these high values.
5.5 VARIATIONS IN OZONE CONCENTRATIONS: DATA FROM SELECTED URBAN AND NONURBAN
SITES
Variations of ozone concentrations by season and by time of day have
been long known and are well documented. First studied in smog chambers,
diurnal patterns have since been corroborated by field investigations, and
exceptions to such general patterns have been examined and documented. Like-
wise, field investigations have substantiated general seasonal patterns and
exceptions to them, and have also established a number of spatial variations
in concentration, such as those that occur with latitude or with altitude.
While it is difficult to discuss temporal and spatial variations separately,
this section is subdivided along those lines for convenience.
5-25
-------
MN
050 150250
I ' ' ' ' I
km
Figure 5-9. Location of Sulfate Regional Experiment (SURE)
monitoring stations.
Source: Martinez and Singh (1979).
5-26
-------
TABLE 5-5. SUMMARY OF OZONE CONCENTRATIONS MEASURED AT SULFATE REGIONAL EXPERIMENT
(SURE) NONURBAN STATIONS, AUGUST THROUGH DECEMBER 1977
Total no. of
measurements
Number of measurements
with concentrations:
>0.08 ppm
>0.10 ppm
>0.12 ppm
Mean
concn,. ppm
Mean of
daily
1-hour
maxima,
ppm
1-hour
maximum,
ppm
Rural sites
01
i
•xl
#1
#4
#6
#9
Montague, MA
Duncan Falls, OH
Giles Co., TN
Lewisburg, WV
3419
3441
3632
3459
60
52
63
23
33
2
5
3
21
0
0
0
0.
0.
0.
0.
021
029
026
035
0.
0.
0.
0.
044
049
052
054
0.153
0.107
0.117
0.106
Suburban sites
#2
#3
#5
#7
#8
Scranton, PA
Indian River, DE
Rockport, IN
Ft. Wayne, IN
Research Triangle
Park, NC
3410
3017
3462
3438
3495
0
29
29
0
80
0
0
0
0
10
0
0
0
0
0
0.
0.
0.
0.
0.
023
030
025
020
025
0.
0.
0.
0.
0.
035
049
046
039
050
0.077
0.099
0.099
0.080
0.118
Source: Martinez and Singh (1979).
-------
5.5.1 Temporal Variations In Ozone Concentrations
5.5.1.1 Diurnal Variations in OzoneConcentrations. By definition, diurnal
variations are those that occur during a 24-hour period. Diurnal patterns of
ozone may be expected to vary with location, depending on the balance among
the many factors affecting ozone formation, transport, and destruction, as de-
scribed in Chapter 3 and noted in Section 5.1. Although they vary with locality,
diurnal patterns for ozone typically show a rise in concentration from low or
near-zero levels to an early afternoon peak. The 1978 criteria document
ascribed the diurnal pattern of concentrations to three simultaneous processes:
(1) downward transport of ozone from layers aloft; (2) destruction of ozone
through contact with surfaces and through reaction with nitric oxide (NO) at
ground level; and (3) in situ photochemical production of ozone (U.S. Environ-
mental Protection Agency, 1978; Coffey et al., 1977; Mohnen, 1977; Reiter,
1977a). Figure 5-10 shows the diurnal pattern of ozone concentrations on
July 13, 1979, in Philadelphia, Pennsylvania. On this day a peak 1-hour average
concentration of 0.20 ppm, the highest for the month, was reached at 2:00 p.m.,
presumably as the result of meteorological factors, such as atmospheric mixing,
and local photochemical processes. The severe depression of concentrations to
below detection limits (less than 10 ppb) between 3:00 and 6:00 a.m. is usually
explained as resulting from the scavenging of ozone by local nitric oxide
emissions. In this regard, this station is typical of most urban locations.
Diurnal profiles of ozone concentrations can vary from day to day at a
specific site, however, because of changes in the various factors that influence
concentrations. Such day-to-day variations are clearly demonstrated in Figure
5-11 (SAROAD, 1985c), which shows diurnal variations in ozone concentrations on
2 consecutive days at the same monitoring site in Detroit, Michigan. Differ-
ences in timing and magnitude occur that are especially noticeable between
midnight and about 7:00 a.m. Transport is probably involved in these nighttime
variations. The afternoon peak concentrations, the actual maxima for the 2
days, differ in magnitude but not in timing.
Composite diurnal data, that is, concentrations for each hour of the day
averaged over multiple days or months, often differ markedly from the diurnal
cycle shown by concentrations for a specific day. In Figures 5-12 through
5-14 (SAROAD, 1985d), diurnal data for 2 consecutive days are compared with
composite diurnal data (1-month averages of hour-by-hour measurements) at
three different kinds of sites: center city (Washington, DC), rural but near
5-28
-------
O
§
LU
I
I I I I I I I I I
12 1 234567 89 10 lit 1 23456789
NOON
« -a.m. HOUR OF DAY
1011
"a.m."
•p.m.-
Figure 5-10. Diurnal pattern of 1 -hr ozone
concentrations on July 13, 1979, Philadelphia, PA.
Source: SAROAD (1985b).
5-29
-------
0.12
0.10
i%
Q.
*
z
o
= 0.08 —
cc
UJ
o
z
o
o
UJ
z
o
N
o
-I I I I I I I I
SEPTEMBER 20
- O SEPTEMBER 21
I I I I I I I i i I
0.06 . -
0.04 -
0.02
12 123456783 10 11 I 123456789 10
NOON
am.HOUR OF DAY P'm- *
11
o
o
its
CM
@l
m
E
X
ro
Z
O
Z
o
o
LLJ
Z
o
N
o
Figure 5-11. Diurnal patterns of 1 -hr ozone
concentrations, September 20 and 21, 1980,
Detroit, Ml.
Source: SAROAD (1985c).
5-30
-------
U. \£.
£
g 0.10
Z
o
P 0.08
oc
g 0.06
U
z
S 0.04
Ul
O
N 0,02
O
0
CENTER CITV-CO*MMERCIAL'1 1 1 1 I 1 I
—
—
-
•^
t
i
L
I
i
mmr-L
I ,
"
\ i
\;
0^
. i •
^
i
i
i
r
i
i
i
i
r /
7
/
J
. i
s,
%
^'
"**"**•,
1
JULY 8-9"
M —"~"— 1 -month
i —
i
\
1
1
I .
I
<
(
i /
\!
V
L
N —
• 1 i
-»' S
x>
^
r 1 ,
\ y
i
t
i
i
i
*
j
V
, 1 •
/*•"
1
t J
1 /
1
*
m «
k»**\e;
•«.*^
i i I
—
ww
—
—
— ^
^1
^ ^
T
• ' WM
. N-
• I 1 i lilt
24 4 8 12 16 20 24 4 8 12 16 20 24
I* a.m.-
— NOON n.m._ M-!
HOUR OF DAY
Figure 5-12. Diurnal and 1-month composite
diurnal variations in ozone concentrations,
Washington, DC, July 1981.
Source: SAROAD(1985d).
a.
2
O
5
c
z
Ul
U
z
O
U
Ul
1
O
U. !£.
0.10
0.08
0.06
0.04
0.02
0
I I i
1 1 1 1 M i 1 1 1 1 1 1
1 1 1 I 1
RURAL-NEAR URBAN SEPT. 29-30
— — — 1 -month —
— . . —
_ -
—
V
£
j
4
/
I
1
*\ '
7
•
>
| |
/
\
| .
f
S
. i
24
8 12 16 20 24
8 12 16 20 24
J-*—a.m.—NOON—p.m.—*•{•*—a.m.—NOON—p.m.—*J
HOUR OF DAY
Figure 5-13. Diurnal and 1 -month composite
diurnal variations in ozone concentrations, St.
Louis County, MO, September 1981.
Source: SAROAD (1985d).
5-31
-------
i 0.10
Q.
Z
2 0.08
| 0.06
O
z
0 0.04
UJ
Z
i 0.02
0
0
111 1
1 I i I 1 1 1 1 I I 1 I I I 1 1 i
SUBURBAN-RESIDENTIAL
OCT. 11-12 -
• 1-month
- -
— _
a
i
— i
f
_ ' *
~ ^?
CS^* A
>£7 ^
\ - f
\
"**^»
, J. , J. , , , •
* ^ -^ ""
T ' % T
* v>^ ~j^.
i ' i y^ "* i-\ •—
V| ^^r^f i*^"* i •**' 'w X ^-
'Jr ^I^L^' ^-^ ^.^
, 1 , 1 , 1 | .. ! , , ! | 1 |s
24 4 8 12 16 20 24 4 8 12 16 20 24
L»— a.m.— NOON— p.m.
a.m — NOON — p.m — 1~\
HOUR OF DAY
Figure 5-1 4. Diurnal and 1 -month composite
diurnal variations in ozone concentrations, Alton,
IL, October 1981 (fourth quarter).
Source: SAROAD (1985d).
an urban area (St. Louis County, MO), and suburban-residential but in a crop-
growing area (Alton, IL). Several obvious points of interest present themselves
in these graphs: (1) at some sites, at least, peaks can occur at virtually
any hour of the day or night but these peaks may not show up strongly in the
longer-terra average data; (2) some sites may be exposed to multiple peaks
during a 24-hour period; and (3) disparities, some of them large, can exist
between peaks (the diurnal data) and the 1-month mean (the composite diurnal
data) of hourly ozone concentrations. These figures are given simply as exam-
ples of the differences that can occur between daily and monthly mean concen-
tration patterns.
The effects of averaging are readily apparent when diurnal or short-term
composite diurnal ozone concentrations are compared with longer-term composite
diurnal ozone concentrations. When compared with Figures 5-12 through 5-14
(daily values and 1-month averages), Figure 5-15 (SAROAD, 1985d), based on
3-month averages, demonstrates rather graphically the effects of lengthening
the period of time over which Values are averaged. This figure shows a
composite diurnal pattern calculated on the basis of 3 months. While seasonal
5-32
-------
•i
1
o
i
fE
z
o
0
LU
2
8
V.IU
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
2
I I 1 I I I I 1 1 I i I i 1 i I 1
- ALTON, IL
—
—
•»
- .*—•. 3rd
/ - — N
— ^" x** 2nd Q ^
* ^
"• -^^
[-••-=.•?- < ....-*' 4th Q -.
1 i 1 1 1 i I 1 1 1 1 1 1 I i 1 1
42 4 6 8 10 12 14 16
I*. a.m. NOON
i I i I I 1
. —
—
—
•m*m
Q - -
mm*
\
\
Nk
ss—S
1 I I 1 1 1
18 20 22 2
HOUR OF DAY
Figure 5-15. Composite diurnal patterns by
quarter of ozone concentrations, Alton, IL, 1981.
Source: SAROAD (1985d).
differences are seen, and will be discussed later, the comparison of 3-month
and 1-month composite diurnal concentrations at the Alton, IL, site readily
demonstrates the smoothing out of peak concentrations as the averaging period
is lengthened. Thus, a fourth pertinent point of interest, related to the
third point cited above, emerges from the data presented above: that is,
increasing the averaging time obscures the magnitude and time of occurrence of
peak ozone concentrations. This is an obvious and familiar result in the
statistical treatment of monitoring data, but one that is highly pertinent to
the protection of human health and welfare from the effects of ozone.
The significance of the relationships of peak and mean concentrations
depends upon whether the health and welfare effects of exposure to ozone are
solely concentration-dependent, heavily concentration-dependent, or both
concentration- and time-dependent. Depending upon whether acute or chronic
exposures elicit the health and welfare effects of concern, careful attention
5-33
-------
may have to be paid to the relationship of short-term (e.g., 1-hour) versus
longer averaging times.
Quantitative analyses of the relationships among maximum or second-highest
1-hour concentrations and daylight, diurnal, monthly, seasonal, and yearly
average ozone concentrations lie outside the scope of this document. An
example of one study, however, serves to describe the kinds of analyses that
may be done to determine relationships among various exposure statistics or
"averaging times." Lefohn and Benedict (1985) have examined the relationship
between 7-hour and 12-hour (daylight) mean ozone concentrations averaged over
a season (7 months). They also examined the relationships between each of
these two exposure statistics and the 1-hour values during the same season
>0.10 ppm. For their analyses they evaluated data records from all sites in
the SAROAD network that met the following criteria: (1) existence of 4 years
of monitoring data (1978 through 1981); (2) >4300 1-hour concentrations re-
ported; and (3) the occurrence of exposures ^1^.2 ppm-hours, which was an
index of sites exposed repeatedly to ozone concentrations MK10 ppm over the
specified time period. Sites so selected were then subjected to additional
criteria that resulted in the identification of monitoring sites located in
areas where agriculturally important cash crops were grown.
On the basis of calculated integrated exposures (hours times 1-hour
concentrations >0.10 ppm), the monitoring sites thus identified were subdivided
by Lefohn and Benedict (1985) into two categories: (1) those with 80 to 200
occurrences of 1-hour values <0.10 ppm (i.e., 80 to 200 hours at >0.10 ppm
ozone), and (2) those with 200 to 1612 hours at >0.10 ppm 03. To describe the
quantitative relationships among the exposure statistics at sites in the two
categories, a Pearson linear correlation matrix was calculated. For sites
with 200 to 1612 occurrences of 1-hour concentrations >0.10 ppm, the correla-
tion was 0.88 between the number of occurrences and the 7-hour seasonal mean
and 0.93 between the number of occurrences and the 12-hour seasonal mean. The
correlation between the 7- and 12-hour seasonal means was 0.96. When correla-
tions were calculated from data obtained at sites with 80 to 200 occurrences
of 1-hour concentrations >.0.10 ppm, the correlations between occurrences and
the 7- and 12-hour seasonal means were 0.40 and 0.47, respectively. The
correlation for the two seasonal means at these sites was 0.94 (Lefohn and
Benedict, 1985).
5-34
-------
No attempt is made in this section to document the respective contribu-
tions of local formation of ozone versus transport of ozone; however, the
occurrence of multiple peak ozone concentrations within a 24-hour period is
usually construed as indicating the presence of ozone transported to the site
from elsewhere. Figure 5-16 illustrates the diurnal variations that can be
seen when transport occurs. Note the occurrence of dual peaks on each of 3
successive days at this site, part of the Sulfate Regional Experiment (SURE)
network.
A familiar measure of ozone air quality is the number or percentage of
days on which some specified concentration is equalled or exceeded. This
measure, however, does not shed light on one of the more important questions
regarding the effects of ozone on both people and plants; that is, the possible
significance of high concentrations lasting 1 hour or longer and then recurring
on 2 or more successive days.
In human controlled exposures, attenuation of responses to ozone has been
observed at about 0.20 to 0.50 ppm in exercising subjects receiving repeated,
consecutive-day exposures (see Chapter 10). That attenuation is lost after
exposures to those levels cease (see Chapter 10 for the time course of loss of
attenuation). It becomes of interest, therefore, to examine how many days in
a row the maximum 1-hour ozone concentration reaches or exceeds .specified
levels in communities in high-ozone areas, as well as in other parts of the
country.
The recurrence of high ozone concentrations on consecutive days was exam-
ined in data (SAROAD, 1985b-d) for one site in each of four cities: Pasadena
and Pomona, CA; Washington, DC; and Dallas, TX. The numbers of multiple-day
events were tallied by length of event (i.e., how many consecutive days) using
data for the daylight hours (6:00 a.m. to 8:00 p.m.) in the second and third
quarters of 1979 through 1981. These sites were chosen because they include
areas known to experience high ozone concentrations (California), and because
they represent different geographic regions of the country (west, southwest,
and east). Similar data could be compiled for any city for which sufficient
aerometric data exist. The choice of the 14-hour daylight period and of the
second and third quarters is consistent with known diurnal and seasonal patterns
of ozone concentrations and with typical human, crop, and ecosystem exposures.
In this discussion of the recurrence of respective specified ozone levels,
a day or series of days on which the daily 1-hour maximum reached or exceeded
5-35
-------
0.16
E 0.14
2
tt
H
g
a
0.12
0.10
0.08
0.06
0.04
Z
o
0 0.02
0
tn
i
oo
01
24 a.m. NOON p.m.
SATURDAY, 27 AUGUST
24 a.m. NOON p.m.
SUNDAY, 28 AUGUST
24 a.m. NOON p.m.
MONDAY, 29 AUGUST 1977
24
Figure 5-16. Three-day sequence of hourly ozone concentrations at Montague, MA, SURE
station showing locally generated midday peaks and transported late peaks.
Source: Martinez and Singh (1979).
-------
the specified level is called an "exposure"; the intervening day or days when
that level was not reached is called a "respite." Four ozone concentrations
were chosen: 0.06, 0.12, 0.18, and 0.24 ppm. The 1- to 7-day events (i.e.,
exposures or respites) are individually tabulated and events longer than
7 days are grouped together.
The summaries of "exposure" and "respite" events in Tables 5-6 through
5-9 show expected regional differences in the successive recurrence of these
respective 1-hour concentrations. At the Dallas station, for example, the
3-year data tally shows 11 exposure events when the daily 1-hour maximum ozone
value equalled or exceeded 0.06 ppm for more than 7 days in a row (Table 5-6).
At first glance, the Pasadena station, with 10 such exposures, appears similar,
but those 10 exposure events spanned 443 days; in Dallas the 11 exposures in-
volved only 168 days. At this lowest concentration (> 0.06 ppm), the Dallas
station recorded more short-duration (< 7 days) exposures (45) involving more
days (159) than the Pasadena station (14 exposures over 45 days), simply because
the daily 1-hour maximum statistic in Pasadena remained above 0.06 ppm for such
protracted periods. At concentrations >_ 0.12 ppm (Table 5-7), the lengthy
exposures at the Pasadena site resolved into numerous shorter exposures; whereas
in Dallas the exposures markedly dwindled in number and duration. At concen-
trations >_ 0.18 ppm (Table 5-8), short-term (< 7 days) exposures in Pasadena
were yet more numerous; Dallas had only two 1-day exposures. At > 0.24 ppm
(Table 5-9), the incidence rate of exposures at the Pasadena station finally
decreased.
Note that these tables (Tables 5-6 through 5-9) also present the single
and multiple-day periods when concentrations were lower than specified levels
("respites"); "event-days" include all "exposures" to and "respites" from
specified concentrations, such that "total event-days" for a given site equal
the total days monitored.
These tables give a more extensive indication of the long-term severity
of ozone air quality at some of the selected sites than is given by the use of
a 1-hour exposure statistic alone. The human responses to single acute expo-
sures to ozone are well documented (Chapter 10)5 but the full significance of
the attenuation of human responses to ozone over the course of a multiple-day
exposure, and any possible consequences of the repetition of such multiple-day
exposures within a smog season or over a number of years remain uncertain.
5-37
-------
en
to
Co
TABLE 5-6. OF CONSECUTIVE-DAY EXPOSURES OR RESPITES WHEN THE DAILY
1-hr MAXIMUM OZONE CONCENTRATION WAS > 0.06 ppm, IN FOUR CITIES
(APRIL THROUGH SEPTEMBER, 1979 THROUGH 1981)
Length of
event,
days
1
2
3
4
5
6
7
Subtotals
Events
(Event-days)
>7
(Event- days)
Total events
(Total event-
days)
Total days
monitored
Source: SAROAD
a
Pasadena
3
5
0
2
2
0
2
14
(45)
10
(443)
24
(488)
532
(1985b-d)
Exposures"
Pomona
6
2
4
3
1
0
1
17
(46)
16
(426)
33
(472) .
542
.
(>0.06 ppm)
Washington
21
10
6
2
2
0
2
43
(91)
5
(55)
48
(146)
442
Dallas
8
10
6
5
8
3
5
45
(159)
11
(168)
56
(327)
451
ii
Pasadena
9
8
2
2
1
0
0
22
(44)
0
(44)
22
(44)
532
Respites
Pomona
11
11
6
2
1
1
0
32
(70)
0
(0)
32
(70)
542
11 (<0.06 ppm)
Washington
17
16
0
3
3
3
1
43
(101)
8
(195)
51
(296)
442
Dallas
25
13
8
4
1
0
1
52
(103)
2
(2.1)
54
(124)
451
-------
CJI
TABLE 5-7. NUMBER OF CONSECUTIVE-DAY EXPOSURES OR RESPITES WHEN THE DAILY
1-hr MAXIMUM OZONE CONCENTRATION WAS > 0.12 ppm, IN FOUR CITIES
(APRIL THROUGH SEPTEMBER, 1979 1981)
Length of
event,
days
1
2
3
4
5
6
7
Subtotals
Events
(Event- days)
>7
(Event- days)
Total events
(Total event-
days )
Total days
monitored:
Source: SAROAD
11
Pasadena
11
10
8
4
7
2
_
42
(118)
14
(254)
56
(372)
532
(1985b-d)
Exposures"
Pomona
13
16
8
5
2
3
1
48
(124)
13
(211)
61
(335)
542
(>0.12 ppm)
Washington
3
1
0
0
0
0
0
4
(5)
0
(0)
4
(5)
442
Dallas
20
4
2
0
1
0
0
27
(39)
0
(0)
27
(39)
451
11
Pasadena
21
14
10
1
3
2
3
54
(131)
3
(29)
57
(160)
532
Respites
Pomona
16
18
9
3
6
3
2
57
(153)
5
(54)
62
(207)
542
" (<0.12 ppm)
Washington
0
0
0
0
0
0
0
0
0
7
(437)
7
(437)
442
Dallas
3
2
-
1
2
2
1
11
(40)
18
(372)
29
(412)
451
-------
TABLE 5-8. NUMBER OF CONSECUTIVE-DAY EXPOSURES OR RESPITES WHEN THE DAILY
1-hr MAXIMUM OZONE CONCENTRATION WAS > 0.18 ppm, IN FOUR CITIES
(APRIL THROUGH SEPTEMBER, 1979 THROUGH 1981)
Length of
event,
days
1
2
3
4
5
6
7
Subtotals
Events
(Event-days)
>7
(Event-days)
Total events
(Total event-
days)
Total days
monitored
"Exposures" (>0.18 ppm)
Pasadena
21
9
10
6
3
4
1
54
(139)
7
(90)
61
(229)
532
Pomona
24
12
9
4
4
4
2
59
(149)
2
(20)
61
(169)
542
Washington
0
0
0
0
0
0
0
0
(0)
0
(0)
0
(0)
442
Dallas
2
0
0
0
0
0
0
2
(2)
0
(0)
2
(2)
451
ii
Pasadena
15
8
14
3
7
2
3
52
(153)
11
(150)
63
(303)
532 .
Respites
Pomona
15
11
9
4
5
1
2
47
(125)
16
(248)
63
(373)
542
" (<0.18 ppm)
Washington
0
0
0
0
0
0
0
0
(0)
3
(442)
3
(442)
442
Dallas
0
0
0
0
0
0
0
0
(0)
5
(449)
5
(449)
451
Source: SAROAD (1985b-d).
-------
TABLE 5-9. NUMBER OF CONSECUTIVE-DAY EXPOSURES OR RESPITES WHEN THE DAILY
1-hr MAXIMUM OZONE CONCENTRATION WAS > 0.24 ppm, IN FOUR CITIES
(APRIL THROUGH SEPTEMBER, 1979 THROUGH 1981)
en
Length of
event,
"Exposures" (>0.24 ppm)
"Respites" (<0.24 ppm)
days
1
2
3
4
5
6
7
Subtotals
Events
(Event-days)
>7
(Event-days)
Total events
(Total event-
days)
Total days
monitored
Source: SAROAD
Pasadena
20
13
5
2
2
0
0
42
(79)
1
(9)
43
(88)
532
(1985b-d).
Pomona
21
10
0
2
1
0
0
34
(54)
1
(9)
35
(63)
542
Washington
0
0
0
0
0
0
0
0
(0)
0
(0)
0
(0)
442
Dallas
0
0
0
0
0
0
0
0
(0)
0
(0)
0
(0)
449
Pasadena
7
1
6
4
3
1
0
22
(64)
23
(380)
45
(444)
532
Pomona
8
4
5
2
2
1
1
23
(62)
14
(417)
37
(479)
542
Washington
0
0
0
0
0
0
0
0
(0)
3
(442)
3
(442)
442
Dallas
0
0
0
0
0
0
0
0
(0)
3
(449)
3
(449)
449
-------
5.5.1,2 Seasonal Variations In Ozone Concentrations. In addition to the
diurnal cycles and between-day variations discussed in the preceding section,
seasonal variations in ozone concentrations occur (for,the reasons discussed
in Chapter 3 and Section 5.1) and usually assume characteristic patterns.
In order to compile an assessment of potential ozone damage to the six
leading commercial crops in the United States (corn, soybeans, hay, wheat,
cotton, and tobacco), Lefohn (1982) surveyed 304 ozone monitoring stations and
identified 24 that (1) were located in counties producing significant quanti-
ties of one or more of these six crops in 1978; (2) reported at least 50
percent of possible hourly data in 1978; (3) reported an hourly maximum of at
least 0.1 ppm 0~; and (4) ranked high in cumulative ozone exposure for the
<3 - -• - . ... • .
period April to October 1978. Six of these sites represented counties with
high soybean, wheat, or hay production. Quarterly composite diurnal patterns
for six of these sites with reasonably complete (>75 percent) 1981 data are
shown in Figure 5-17 (SAROAD, 1985d). The average levels are apparently
comparable with the long-term averages at the NAPBN sites previously discussed
(Section 5.4.1). In addition, the diurnal patterns for these sites clearly
show the division of the afternoon ozone concentrations into two seasonal
patterns, the low "winter" levels in the first and fourth quarters and the
higher "summer" levels in the second and third quarters of the year.
Although averaging causes details to be obscured, the average diurnal
patterns in Figure 5-17 show that the time of occurrence of peaks differs
among sites, and, to an extent, between seasons. The seasonal differences in
time of day are especially noticeable for the fourth quarter. Among the sites
shown in Figure 5-17, ozone concentrations appear to peak at 2:00 to 2:30 p.m.
in Little Rock in the higher-concentration second and third quarters. At
Bakersfield in the second and third quarters, there is evidence, even from
these composite data, of two peaks, the first at about 1:00 p.m. and the second
at 5:00 to 6:00 p.m. At the Clark County, Ohio, site, the peak concentrations
in the second and third quarters center around about 5:00 p.m., but they do not
return to "baseline" until after midnight. The patterns at the Bakersfield and
Clark County sites appear to indicate transport into the areas. It is also
possible that single peaks that are shifted to mid- to late afternoon, as at
the Little Rock site, are the product of transport. Depending upon proximity
to urban centers and wind speed and direction, rural areas are typically
exposed to their peak concentrations later than urban areas, usually within
daylight hours but not always.
5-42
-------
Q.
a
<
cc
ui
O
z
o
o
o
o
a
a
O
P
K
Z
O
o
1
a
z
o
H
cc
o
o
IU
1
O
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
I I I I I I I I I I I I M I I i M I
N. LITTLE ROCK, AR
(A)
1st Q
.— — • 2ndQ
3r
-------
Composite diurnal variations in ozone concentrations at a rural site in
Argonne, Illinois, over a 7-week period of the third quarter of 1980 are shown
in Figure 5-18 (Kress and Miller, 1983). The actual day-to-day variations in
ozone concentration over the entire third quarter of 1980 at that site are
shown for comparison in Figure 5-19 (Kress and Miller, 1983). As part of the
National Crop Loss Assessment Network, the site at Argonne monitors ozone
concentrations over a 7-hour daytime period approximating the period of peak
photosynthesis in crops. The data in Figure 5-18 yield a 24-hour average
ozone concentration of 0.026 ppm (Kress and Miller, 1983). The 7-hour average
for the same 7-week period at the Argonne site is 0.042 ppm (Kress and Miller,
1983). The day-to-day variations in both the 7-hour and the 24-hour averages
(Figure 5-19) generally appear to be greater than the average difference
within a day for either the 7-hour or 24-hour periods (Figure 5-18). The
fluctuations in 1-hour values either within a day or from day to day, however,
would be larger than within-day or between-day variations in either the 7-hour
or the 24-hour average. The 7-hour average will be higher than the 24-hour
average because the former excludes the low nighttime concentrations. As
Figure 5-19 and its data illustrate, the selection of the appropriate averaging
time is critical for the accurate description of dose-response relationships
and for the protection of human, vegetation, and other receptors from the
effects of ozone.
It is worth noting in this context that air pollution exposure statistics
are of two basic functional types: descriptive and preventive. Descriptive
exposure statistics simply define the conditions of concentration and exposure
duration (averaging time) under which a specified effect has been observed or
detected. Preventive exposure statistics are used prescript!*vely with the
expectation of keeping a specified effect from happening. To be effective
bases for an air quality standard, preventive exposure statistics must be
(1) related to the true time-course of development of the specified effect(s),
and (2) based on observed or predictable distributions, or both, of exposure
conditions over the range of concentrations and durations producing the
effect(s) of concern.
In Figure 5-20 (A-H), seasonal variations in ozone concentrations in 1981
are depicted using 1-month averages and the single 1-hour maximum concen-
tration within the month for eight sites across the nation (SAROAD, 1985d).
The data from most of these sites exhibit the expected pattern of high ozone
5-44
-------
a
a
DC
Z
111
u
Z
o
u
111
Z
o
N
o
0.06
0.04
0.02
24
a.m.
NOON
p.m.
24
Figure 5-18. Composite diurnal ozone pattern at a rural
NCLAN site in Argonne, IL, August 6 through September
30,1980.
Source: Kress and Miller (1983).
0.15
a
a
2 o.io —
<
DC
U
o
u
LU
Z
o
N
o
0.05
JUL
AUG
SEP
OCT
MONTH OF YEAR
Figure 5-19. Daily 7-hour and 24-hour average
ozone concentrations at a rural NCLAN site in
Argonne, IL, 1980.
Source: Kress and Miller (1983).
5-45
-------
u.oa
0.30
1
u.
2 0.26
g
e 020
ty
U
Z 0.15
O
U
Ul
o a1°
8
0.05
0
0.35
0.30
n.
2 °'25
O
5
5 0.20
us
O
O
UJ
Z 0.10
N
O
0.05
0
_
MM
_
—
_
TUCSON, AZ
(AJ
B MAXIMUM ~
AVERAGE
I
MMVMH
1
11
_
1
••^•K
1
_
1
—
i
»*»***
•i'x
—
.ri
tTO
—
m
—
fff
JFMAMJJASOND
-.
—
—
_
_
—
POMONA. CA
(0 __,
T"
«_
55
1
—
fx
MM*
11
rrr
1
—
WT
P-:
—
S!
—
—
1
-
_
—
Wl-
—
— ~1
_
0.30
0.25
0.20
0.15
0.10
0.05
0
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0
LENNOX, CA
_ (B)
"~
_
—
-n
H
j
"""
~
•J-X
_
9
^
:•:•:•:
I
_
r?9
:JS
1
1 1 ~
—
1 W ~
FMAMJJASOND
DENVER, CO
_ (D) _
_
_
-..
"r-i
D
_
_
_
i
—
P
—
^
*!'
••H
((•(•I
—
1
— i
m_
J F M A M J J A S O N D
MONTH OF YEAR
J FMAMJJASOND
MONTH OF YEAR
Figure 5-20. (A-H). Seasonal variations in ozone concentrations as indicated
by monthly averages and the 1 -hour maximum in each month at selected sites,
1981.
5-46
-------
u.oa
0.30
D.
^ 0.25
O
2 0.20
Ul
i 0.15
O
o
Z 0.10
o
s
0.05
0
0.35
0.30
Q.
^ 0.25
g
< 0.20
y
Ul
Z 0.15
O
U
z o.io
1
0.05
0
WASHINGTON, DC
_
(E
)
—
B MAXIMUM —
AVERAGE
_
—
I
n
j
n
F
M
M
MORRIS
—
D
W.
^™
I
i
_
TO
•X**
—
PWafcWi LwJ
AMJJASOND
CO., NJ
_ (6) _
—
—
_
^^
W!
PJJJ3
"*""
m
—
—
__ —
ff™^^
—
ra
M
—
1
^™^^i
W:
I
— __
DO
0.30
0.25
0.20
0.15
0.10
0.05
0
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0
_
•"""
—
_
—
_
_
—
—
_
^^
TAMPA, FL
IF) _
—
_
-
D
rr>
I
-
m
m
IT-
—
1
m
—
xT
""""•
1!
—
M tea
JFMAMJJASOND
DALLAS, TX
(H) _
—
—
— . —_ —
—
—
~
_
I
m
55
—
n
1
1.
^
53
„ „. „„,„
-—
ii
sltei
JFMAMJJASOND
MONTH OF YEAR
J FMAMJJ ASOND
MONTH OF YEAR
Figure 5-20. (A-H) (continued). Seasonal variations in ozone concentrations as
indicated by monthly averages and the 1 -hour maximum in each month at
selected sites, 1981.
Source: SAROAD (1985d).
5-47
-------
levels in late spring or in summer and low levels in the winter. Data for
Pomona (Figure 5-20C) and Denver (Figure 5-20D) show summer maxima. Tampa
shows a late spring maximum but with concentrations in the fall (October)
approaching those of spring (June) (Figure 5-20F). Dallas data also tend to
be skewed toward higher spring concentrations; but note that November concen-
trations are also relatively high (Figure 5-20H). Averaging together data for
several years would give a smoother "characteristic" pattern but also would
obscure the fact that local, and even national, weather in a particular year
plays at least as big a role in the formation of ozone as the regular seasonal
changes in the elevation of the sun and the resulting variations in insolation.
Because of seasonal changes in storm tracks from year to year, the general
weather conditions in a given year may be more favorable for the formation of
ozone and other oxidants than during the prior or following year. Thus,
short-term concentration trends may not be indicative of real changes in air
quality.
5.5.1.3 Weekday-Weekend Variations in Ozone Concentrations. Atmospheric
ozone concentrations represent the combined effects of emission sources and
meteorological conditions. The various sections of this chapter have been
based on the assumption that ozone precursor sources operate on a generally
steady-state or at least on an average, repeatable diurnal cycle. For the
most partj urban source patterns of oxidant precursors do appear to be reason-
ably constant; however, in most urban areas changes occur in traffic and'
commercial emission patterns that are keyed to a weekday-weekend activity
cycle. The effects of these changes have been observed in corresponding
changes in ozone concentration patterns.
Debate in the 1960s over the role of nitric oxide (NO) in scavenging
ozone led to the examination of whether weekday-weekend differences in ozone
concentrations might occur in urban areas, on the assumption that NO emissions
would be lower on weekends when some NO -emitting sources virtually shut down.
y\
Altshuller (1975) reported the observation that no alerts had ever occurred in
the Los Angeles Basin on a Sunday. Likewise, Elkus and Wilson (1976), Levitt
and Chock (1976), and Graedel et al. (1977) reported the existence of a weekday-
weekend difference in ozone concentrations. While not reporting lower average
ozone concentrations on weekends for the Los Angeles Basin, Schuck et al.
(1966) noted a spatial shift of concentrations on summer weekends, away from
the central commercial and urban areas to suburban and coastal areas. They
5-48
-------
attributed the spatial shift in concentrations to changes in traffic patterns
resulting from a shift from job-related travel to recreational travel.
No newer reports of possible weekend-weekday variations are available in
the literature, but any exhaustive exposure assessment should take into consid-
eration the possibility that subtle differences in absolute concentrations or
their spatial and temporal patterns could have some effect on total exposure.
5.5.2 Spatial Variations in Ozone Concentrations
Ozone is commonly thought of as a relatively homogeneous, regional pollu-
tant, largely because of the nature of the sources of its precursors and the
processes that contribute to its formation as a secondary pollutant. Numerous
factors affect the occurrence of ozone in ambient air, however, so that dif-
ferences in the true exposures sustained by respective receptors (e.g., human
populations, crops, natural ecosystems) will vary within regions, airsheds, or
other defined areas. The data presented in this section demonstrate a few of
the known spatial variations in ozone concentrations that should be taken into
account when assessing actual or even potential exposures. The data presented
in this brief section are intended to be illustrative rather than exhaustive,
since numerous meteorological, topographical, and physicochemical factors will
influence the spatial and temporal distributions of ozone concentrations at
specific sites.
5.5.2.1 Urban versus Nonurban Variations. Data were presented in the 1978
criteria document demonstrating that peak concentrations of ozone in rural
areas are generally lower than those in urban areas, but that "dosages or
average concentrations in rural areas are comparable to or even higher than
those in urban areas" (U.S. Environmental Protection Agency, 1978). The
diurnal concentration data presented in the preceding section indicate that
peak ozone concentrations can occur later in the day in rural areas than in
urban, with the distance downwind from urban centers generally determining how
much later the peaks occur. The data presented in the preceding section for
Montague, Massachusetts, in Figure 5-16 (Martinez and Singh, 1979) exemplify
high late-afternoon secondary peak concentrations resulting from transport.
The NAPBN and other nonurban data presented in Section 5,4 illustrate the
urban-nonurban gradient in ozone concentrations that generally exists, and
support the statement quoted above from the 1978 document. While corroboration
of that statement would require the calculation of means or ppm-hours, the
5-49
-------
data given in Section 5.4.2 tend to support that conclusion. For example,
data presented in that section (Table 5-5) showed that rural sites of the SURE
network were exposed, on the average, to daily maximum 1-hour ozone concentra-
tions in excess of those to which the suburban sites were exposed (Martinez
and Singh, 1979). While one rural site recorded 21 occurrences of 1-hour
ozone concentrations >0.12 ppm during August through December 1977 (referred
to below as A-D), the suburban sites recorded no such occurrences. All four
of the rural sites in the SURE network recorded two or more occurrences (and
one recorded 33 such occurrences) of daily maximum 1-hour concentrations >0.10
ppm; only one of the five suburban sites recorded the occurrence (n = 10) of
that concentration over the same period (A-D). All exposure statistics exam-
ined by Martinez and Singh (1979) for the SURE concentration data were higher
at the rural sites, on the average, than at the suburban sites. The respective
statistics, weighted by the number of observations in the data record, were:
average of 24-hour mean concentrations (A-D), 0.028 ppm at rural versus 0.024
ppm at suburban sites; mean of daily maximum 1-hour (A-D), 0.050 ppm at rural
versus 0.044 ppm at suburban sites; and the 1-hour maximum for the whole
period (A-D), 0.121 ppm at rural sites versus 0.094 ppm at suburban sites.
In another example, data reported by Lefohn (1984) from sites classified
as rural in the SAROAD data bank showed 9 occurrences of daily maximum 1-hour
ozone concentrations >0.12 ppm at a rural Preble County, Ohio, site; and 19
such occurrences at a rural site in Camden County, New Jersey.
In addition to the occurrence of higher mean concentrations and occasion-
ally higher peak concentrations in nonurban areas than in urban, it is well
documented that ozone persists longer in nonurban than in urban areas (Coffey
et al., 1977; Wolff et a!., 1977; Isaksen et al., 1978). The absence of
chemical scavengers appears to be the chief reason.
5.5,2.2 Intracity Variations. Despite relative intraregional homogeneity,
evidence exists for intracity variations in concentrations that are pertinent
to potential exposures of human populations and to the assessment of actual
exposures sustained in epidemiologic studies. Two illustrative pieces of data
are presented in this section, one a case of relative homogeneity in a city
with a population under 500,000 (New Haven, Connecticut) and one a case of
relative inhomogeneity of concentrations in a city of greater than 9 million
population (New York City) (U.S. Department of Commerce, 1982).
5-50
-------
New Haven, Connecticut, was the site of an epidemiological study in 1976
by Zagraniski et al. (1979). Symptoms recorded in diaries kept by study
participants were correlated with ozone concentrations measured by the chemi-
"luminescence method at a downtown New Haven site characterized as Center
City-Residential. Table 5-10 shows several percentiles in the distribution of
hourly values for that site plus two other monitoring stations in the county
that were operating at the time, one in Derby, Connecticut, 9 miles west of
New Haven, and one in Hamden, Connecticut, 6 miles north of New Haven. The
Derby site also is characterized as Center City-Commercial, the Hamden site as
Rural-Agricultural. The general similarity of values among the three sites
appears to substantiate the New Haven data used in the epidemiological study
since there was probably a reasonable temporal correlation between these close
sites. Thus, wherever individuals might have traveled about the county, they
probably were exposed to similar concentrations of ambient ozone. This conclu-
sion is reinforced by the data in Table 5-11, showing the date and time of the
maximum hourly concentrations by quarter at these three sites. The significant
data are those for the second and third quarters when the potential for 0~
formation and for exposure was the greatest. Differences in peak concentrations
varied from 0.006 ppm in the fourth quarter to 0.055 ppm in the third quarter
among sites.
The source of much of the ozone found in the New Haven, Connecticut, area
is the greater New York City area (e.g., Wolff et al., 19/5; Cleveland et al.,
1976a,b) and an urban plume transported over the distance from New York City
to New Haven would tend to be relatively well-mixed and uniform, such that
intracity variations in New Haven would probably be minimal.
The highest or second-highest 1-hour maximum ozone concentration reported
from a given station during a given year frequently gives an indication of the
potential for repeated human exposure to high ozone levels. Nevertheless, a
one-to-one correspondence between peak levels and either the number of days or
the number of hours that a given level may be exceeded does not necessarily
exist. Data obtained in the metropolitan New York area in 1980 illustrate
this latter fact (Smith, 1981); and illustrate, as well, that intracity
gradients can exist that should be taken into account in exposure assessment.
The data given in Table 5-12 were obtained at the monitoring sites shown in
Figure 5-21. The second-highest 1-hour ozone readings at the Eisenhower Park
and Queens College stations have values only a few percentage points apart,
5-51
-------
TABLE 5-10. OZONE CONCENTRATIONS AT SITES IN AND AROUND
NEW HAVEN, CONNECTICUT, 1976
CCHEMILUMINESCENCE METHOD, HOURLY VALUES IN
of values < stated concentration
No.
Site (SAROAD No.) measurements
50%
90%
95%
99% Max concn.
New Haven, CT
(070700123F01)
Derby, CT
(070190123F01)
Hamden, CT
(070400001F01)
4119
5698
3853
0.021 0.035 0,091 0.162 0.274
0.023 0.038 0.071 0.095 0.290
0.030 0.045 0.075 0.098 0.240
Source: SAROAD (1985a).
TABLE 5-11. QUARTERLY MAXIMUM 1-HOUR OZONE VALUES AT SITES
IN AND AROUND NEW HAVEN, CONNECTICUT, 1976
(CHEMILUMINESCENCE METHOD, HOURLY VALUES IN ppm)
New Haven, CT
No. measurements
Max 1-hr, ppm
Hour of day
Date
Derby^ CT
No. measurements
Max 1-hr, ppm
Hour of day
Date
Hamden, CT
No. measurements
Max 1-hr, ppm
Hour of day
Date
1
10
0.045
11:00 a.m.
3/29
11
0.015
11:00 p.m.
3/31
56
0.050
Noon
3/29
Quarter of
2
1964
0.274
2:00 p.m.
6/24
2140
0.280
2:00 p.m.
6/24
2065
0.240
3:00 p.m.
6/24
Year
3
2079
0.235
2:00 p.m.
8/12
2187
0.290
2:00 p.m.
8/12
1446
0.240
1:00 p.m.
7/20
4
66
0.066
10:00 p.m.
10/3
1360
0.060
7:00 p.m.
12/20
286
0.065
3:00 p.m.
10/7
Source: SAROAD (1985a).
5-52
-------
TABLE 5-12. PEAK OZONE CONCENTRATIONS AT EIGHT SITES IN NEW YORK CITY
AND ADJACENT NASSAU COUNTY, 1980
en
Site
Susan Wagner H.S.
Mabel Dean H.S.
Woolsey Post Office
Mamaroneck
P.S, 321
Sheepshead Bay H.S.
Queens College
Eisenhower Park
Site
no.
1 •
2
3
4
5
6
7
8
No. 1-hr
averages
>0.12 ppm
20
19
37
0
24
44
51
7
Days
with 1-hr
averages
>0.12 ppm
8
10
6
0
9
12
15
2
Four highest 1-hr daily
values, ppm, and date
1st
0.174
(8/28)
0.155
(7/21)
0.188
(7/20)
0.092
(6/14)
0.148
(7/26)
0.184
(7/31)
0.174
(8/28)
0.175
(8/28)
2nd
0.152
(7/18)
0.154
(7/26)
0.163
(7/21)
0.080
(8/28)
0.146
(8/28)
0.173
(7/18)
0.164
(7/21)
0.158
(7/21)
3rd
0.140
(7/26)
0.144
(7/18)
0.151
(7/22)
0.076
(7/2)
0.145
(7/18)
0.165
(8/7)
0.163
(6/14)
0.119
(7/20)
4th
0.131
(9/1)
0.139
(8/28)
0.148
(8/28)
0.075
(7/26)
0.144
(7/9)
0.164
(7/14)
0.159
(8/24)
0.118
(8/24)
Sites monitored during the Northeast Corridor Monitoring Program (NECRMP); site
numbers assigned here are keyed to Figure 5-21.
Source: Smith (1981).
-------
SITES
1. Susan Wagner High School
2. Mabel Dean High School
3. Woolsey Post Office (Astoria)
4. Mamaroneck
5. Public School 321
6. Sheepshead Bay High School
7. Queens College
8. Eisenhower Park (Nassau Co.)
NEW JERSEY
Figure 5-21. New York State air monitoring sites for
Northeast Corridor Monitoring Program (NECRMP).
Source: Smith (1981).
5-54
-------
yet there were 51 hours of ozone concentrations exceeding 0.12 ppm and 15 days
when ozone levels exceeded 0.12 for at least 1 hour at the Queens College
Station; whereas corresponding values were recorded at Eisenhower Park for
only 7 hours during 2 days. At both stations, data for about 94 percent of
possible hours were recorded as valid.
The range of first-, second-, third-, and fourth-highest values, along
with frequencies of values >0.12 ppm, establishes an apparent concentration
gradient in the area from sites 6 and 5 to site 4. Exposures of human popula-
tions living and working in metropolitan New York City could differ appreciably
if the residences were located and all activities were centered in lower.
Brooklyn as opposed to the upper Bronx. Differences in peak concentrations at
the respective sites varied by date (6/14 to 8/28) and by level on the same
day (8/28), when ozone was 0.080 ppm at site 4 and 0.174 at site 7, a differ-
ence of 0.094 ppm (a factor of 2.18).
Intracity gradients in ozone concentrations have also been reported by
Kelly et al. (1986) for a 1981 study in Detroit, MI. Ozone concentrations
were measured for about 3 months at 16 sites in the metropolitan Detroit area
and in nearby Ontario, Canada. Values at 15 sites were correlated with those
at a site adjacent to the Detroit Science Center, about 3 km north of the
central business district 'in Detroit. In general, the correlation decreased
as distance from the Science Center site increased; and, in general, the
actual concentrations increased with distance from that site toward the north-
northeast. The highest ozone concentrations were recorded at sites about 10
to 70 km north-northeast of the urban core. At greater distances or in other
directions, ozone maxima decreased.
5.5.2.3 Indoor-Outdoor Concentration Ratios. Most people in the United
States spend a large proportion of their time indoors. Although knowledge of
actual exposures of populations to ozone is essential for optimal interpreta-
tion and use of the results of epidemiological studies, essentially all air
pollution monitoring is done on outdoor air. The modeling of actual exposures,
as opposed to potential exposures, therefore necessitates knowing general
activity patterns and at least approximate indoor/outdoor ratios (I/O) of
ozone.
For slowly reacting compounds such as carbon monoxide, long-term average
ratios of indoor to outdoor concentrations tend to be close to unity, in the
absence of indoor sources. 'Over short time periods, however, the I/O may be
5-55
-------
significantly different because of non-equilibrium factors (Yocom, 1982). The
situation for reactive pollutants such as ozone is much more complex, and
reported I/O values for ozone are highly variable. Unfortunately, the number
of experiments and kinds of structures examined to date provide only limited
data for use in modeling indoor exposures. Yocom (1982) has presented a
chronological summary of studies in which either ozone or photochemical oxidant
indoor-outdoor gradients in buildings and residences were measured. Studies
summarized by Yocom have been conducted over the period 1971 through the
present (one ongoing study) by five research organizations: University of
California, the California Institute of Technology, GEOMET, Inc., Lawrence
Berkeley Laboratory, and TRC Environmental Consultants. Structures examined
have included hospitals, schools, office buildings, single-dwelling homes,
"experimental" dwellings, apartments, and mobile homes. Private homes included
those with and without gas stoves and fireplaces, and those inhabited by
smokers versus nonsmokers. Areas of the country in which the buildings were
located ranged from Southern California to Boston, including such cities in
between as Denver, Chicago, Washington, Baltimore, Pittsburgh, as well as
other unspecified locations (Yocom, 1982). The indoor-outdoor ratios reported
from these studies are summarized in Table 5-13 and are discussed later.
Among newer reports of indoor/outdoor gradients in ozone concentrations,
also summarized in Table 5-13, are the studies of Stock et al. (1983) and
Contant et al. (1985), undertaken to provide exposure assessments for an
epidemiological study of asthmatics in Houston, Texas, in 1981 (see Holguin et
al., 1985, in Chapter 11). Stock et al. (1983) found I/O ratios of nearly
zero to 0.09 in one air-conditioned residence (maximum ozone concentration of
5 ppb) and an I/O ratio of 1.0 in a residence ventilated completely by outdoor
air. (Indoor concentrations were monitored via a sampling manifold connected
by Teflon lines to a chemiluminescence analyzer in a mobile van parked near
the respective houses.) Contant et al. (1985), in a continuation and extension
of the same study, used "personal monitors" (i.e., portable analyzers) to
measure indoor and outdoor air in the immediate environs of participants in
the asthma study. Mean, median, and maximal ozone concentrations, respectively,
were 10.8, 5, and 147 ppb indoors and 51.8, 42, and 250 ppb outdoors. The
respective I/O ratios were 0.21, 0.12, and 0.59.
Davies et al. (1984) found I/O ratios of 0.7 ± 0.1 at an art gallery in
rural Norwich, England, where the outside ozone concentrations ranged from 18
5-56
-------
TABLE 5-13. SUMMARY OF REPORTED INDOOR-OUTDOOR OZONE RATIOS
Structure
Indoor-outdoor
ratio (I/O)
Reference
Residence
(with evaporative cooler)
Office
(air-conditioned; 100% outside
air intake)
(air-conditioned; 70% outside
air intake)
Residence
Residence
Residence
(gas stoves)
(all electric)
Office
School room
Residence
Residences (1 each)
(air-conditioned)
(100% outside air; no
air conditioning)
Residences (12)
(ai r-condi tioned)
Art gallery
(three modes of ventilation
in each 24-hr period:
recirculation, mixture of
recirculated and outside
air, and 100% outside air)
0.60*
0.80 + 0.10
0.65 + 0.10
0.70
0.50 to 0.70
0.19
0.20
0.29
0.19
(maximum concn.)
0.10 to 0.25
0.00 to 0.09
1.0
0.21 (mean concn.)
0.12 (median concn.)
0.59 (maximum concn.)
0.70 + 0.10
(mean~concn.)
Thompson et al. (1973)
Sabersky et al. (1973)
Sabersky et .al. (1973)
Moschandreas et al. (1978)
Moschandreas et al. (1981)
Moschandreas et al. (1978)
Berk et al. (1980)
Berk et al. (1981)
Stock et al. (1983)
Contant et al. (1985)
Davies et al. (1984)
Measured as total oxidants.
5-57
-------
to 58 ppb in the 3-week study period. The interior of the modernistic gallery
had few reactive surfaces (two plate glass walls and two walls and ceiling of
painted aluminum, with minimal other walls and furnishings). In addition,
ventilation cycled between recirculation air, a mixture of outside and recireu-
lation air, and 100 percent outside air over each 24-hour period.
The results of all the studies cited above are shown in Table 5-13.
These results are highly variable, to say the least. The variability is not
surprising, however, considering the diversity of structures and locations
included in the studies. In Table 5-13, the highest I/O, 1.0, was reported
for a residence ventilated by outside air (Stock et al., 1983). The second-
highest I/O, 0,80, was determined by smog-season measurements made in a multi-
-story, air-conditioned building on the Pasadena campus of the California
Institute of Technology (Sabersky et al., 1973). Air exchange in this building
was at a rate of 10 changes per hour with 100 percent outside air (i.e., no
recirculation of inside air). For another Cal Tech building, in which there
was a mixture of 70 percent outside air and 30 percent recirculated inside air,
Sabersky et al. (1973) found an indoor-outdoor ozone ratio of 0.65. The
lowest indoor-outdoor ozone concentration ratios shown in Table 5-13 are those
reported by Stock et al. (1983) for one residence; the indoor concentration
recorded is barely above the limit of detection of the analyzer.
A relatively large number of factors can affect the difference in ozone
concentrations between the inside of a structure and the outside air. In
general, outside air infiltration or exchange rates, interior air circulation
rates, and interior surface composition (e.g., rugs, draperies, furniture,
walls) affect the balance between replenishment and decomposition of ozone
within buildings (Thompson et al., 1973; Sabersky et al., 1973; Berk et al.,
1980; Moschandreas et al., 1978). The rate at which exterior air enters a
building depends on local wind speed and direction, on how well-sealed the
building is, on how frequently doors and windows are opened, and on the operat-
ing characteristics and cycles of heating/air conditioning/ventilating systems.
A significant factor that increases infiltration is an increasing temperature
differential between warm interior air and cold outside air (Moschandreas et
al., 1978), although a high differential would be unusual for a photochemical
smog period. Moschandreas et al. (1978) reported exterior-interior exchange
'rates ranging from ten changes per hour in an office building to one change
every 5 hours in a residence. At the higher exchange rates, inducted ozone
5-58
-------
remains at a level indoors that is closer to the outdoor level. As the exchange
rate decreases, surface decomposition processes can result in progressively
lower equilibrium ozone concentrations. Other factors, such as relative
humidity, also affect decomposition. The half-life of ozone inside residences
has been estimated at 2 to 6 minutes (Moschandreas et al., 1978; Mueller
et al., 1973; Sabersky et al., 1973), while its half-life in an office environ-
ment has been estimated at 11 minutes (Mueller et al., 1973). These results
are indicative of the relatively rapid reaction rate that can be expected for
ozone in a building or room environment. The problem of I/O values in buildings
was the subject of a model development program by Shair and Heitner (1974) in
which they tried to account for ventilation and for losses by reactions and
surface scavenging. Considering the research results shown in Table 5-13 and
summarized by Yocom (1982), any estimates of indoor ozone exposures must be
considered as having a large degree of probable variability.
Automobiles and other vehicles constitute another indoor environment in
which people may spend appreciable amounts of time. As with buildings, the
mode of ventilation and cooling helps determine the inside concentrations.
Peterson and Sabersky (1975) reported ozone measurements made inside and
outside a car traveling a freeway in southern California. Concentrations
inside were higher when the air-conditioning was operating on "maximum" than
when it was operating on "normal," largely because at higher air turnover
rates ozone intake exceeded ozone decay. When windows were opened, an equili-
brium between inside and outside ozone concentrations was established in about
3 minutes, with inside concentrations then remaining significantly lower than
outside concentrations as the result of the exponential decay of ozone from
contact with interior surfaces. Inside ozone concentrations were one-third or
less of the outside concentrations (I/O < 0.33), on the average, under all
ventilation and cooling conditions. Ozone concentrations measured outside the
car were lower than those measured at nearby fixed sites, a result attributed
to scavenging of ozone by nitric oxide. In the exposure study of Content et
al. (1985) in Houston, I/O ratios from 49 measurements inside vehicles were '
0,44 for mean, 0.33 for median, and 0.56 for maximum concentrations measured.
Driving routes and ventilation and cooling modes were not described.
Concentrations of ozone inside vehicles obviously depend first on the
concentration outside and then on the various effects of cooling and ventilation
conditions and surface decay. Thus, the titration of 03 by NO on and near
5-59
-------
roadways is an important factor. The titration reaction is rapid but depends
on the amount of NO emitted, which in turn depends on the amount of traffic on
a given road or freeway. Rhodes and Holland (1981) found average ozone levels
on the nine-lane freeway between Los Angeles and San Diego, which carries
200,000 cars each day, to be one-ninth to one-sixth the levels found 30 meters
upwind.
At present there are no long-term monitoring data on indoor air pollutant
concentrations that are comparable to the concentration data available for
outdoor locations. Thus, for estimates of the exposure of building or vehicle
occupants to ozone and other photochemical oxidants, it is necessary to rely
on extrapolations of very limited I/O data such as those presented above and
in Table 5-13.
5.5.2.4 Altitudinal and Latitudinal Variations. Concentrations of ozone vary
with altitude and with latitude. These variations occur for a number of
reasons, including the nature of the interchange mechanisms involved in strato-
spheric- tropospheric exchange, the decay of stratospheric ozone as it traverses
the troposphere, and the known production of ozone in the apparently unpolluted
free troposphere at certain altitude ranges above mean sea level (MSL). In
addition, other meteorological, physical, and chemical factors contribute to
the concentration gradients found with changes in altitude and latitude at the
surface of the earth.
Among specific factors known to contribute to variations in ozone concen-
trations with altitude and latitude, and, to a much lesser extent, longitude,
are the following:
1. Incursions of stratospheric ozone;
2. Global circulation patterns, along with accompanying altitudinal and
geographic differences in atmospheric pressure, direction of prevail-
ing winds, wind speeds, temperature, humidity, and precipitation;
3. Intensity and spectral distribution of sunlight; and
4. Meteorological factors, such as mixing heights, frequency and persist-
ence of inversions, upslope flows in mountainous areas, and cloud
cover and precipitation patterns.
Most of these factors were discussed in Chapter 3, as well as in the 1978
criteria document (U.S. Environmental Protection Agency, 1978). A few are
briefly noted here to present an overview of the macroscale differences in
ozone concentrations that may be expected. Concentration gradients found with
increases in altitude at the earth's surface are presented in more detail
because of their relevance for specific kinds of exposure assessments.
5-60
-------
Latitudinal differences in ozone concentrations have been demonstrated.
These are in part attributable to the variations, temporal as well as geogra-
phic, in actinic irradiance at the earth's surface. Both the intensity and
the spectral distribution of sunlight have direct effects on the photochemical
reactions that initiate and sustain oxidant formation (Chapter 3 and U.S.
Environmental Protection Agency, 1978). The effect of latitude on ozone
formation is significant during the winter but not during the summer, since in
summer, for the contiguous United States, the maximum light intensity is
fairly constant and the duration of the solar day varies only slightly with
latitude. Light intensity varies somewhat with longitude in the summer, with
the highest intensities occurring in the western states.
Incursions of stratospheric ozone also contribute to variations in ozone
concentrations with latitude. These are expected to be strongest in the
mid-latitudes of the northern hemisphere because of the nature of the mechanisms
by which stratospheric-tropospheric exchange occurs.
The 1978 criteria document presented discussions of the effects of tropo-
pause-folding events (TF) and of the seasonal tropopause adjustment (STA) and
small-scale eddy transport (SSET) mechanisms on stratospheric-trospheric
exchange. As described previously in Chapter 3 and in the 1978 document, TF
events would be expected to produce sporadic increases in ground-level ozone
concentrations, resulting from strong incursions of stratospheric ozone, in
the southern and eastern United States (latitudes of < 37°N and longitudes of
< 90°) (U.S. Environmental Protection Agency, 1978). The STA-produced incur-
sions also occur in these latitudes. Both STA and TF produce incursions in
the winter or early spring.
Logan et al. (1981) summarized earlier measurements of ozone in the lower
troposphere. Hemispheric asymmetry is apparent in the data they presented,
with higher concentrations occurring in the northern hemisphere than in the
southern. Also apparent is a seasonal increase in lower-tropospheric ozone in
the summer at mid-latitudes in the northern hemisphere. The authors noted,
however, that ozone data for the lower troposphere are sparse, especially for
the southern hemisphere. Compared with ozone data from two east-coast sites
at mid-latitudes in the northern hemisphere, ozone data from one site in
Australia showed (1) little seasonal variation at <3 km (MSL), (2) about 40
percent lower concentrations at 2 km (MSL) in summer, and (3) similar concen-
trations in autumn and winter. Altitudinal profiles of ozone concentrations
5-61
-------
for the tropical northern hemisphere (Panama, 9°N, and Hawaii, 19.5°N) show a
considerably lower gradient with altitude than that seen at mid-latitudes
(Oltmans, 1981; cited in Logan et a!,,, 1981). Logan et al. (1981) noted the
difficulty, given the sparseness of the data, of separating hemispheric effects
from variations in concentrations attributable to other factors (e.g., oceanic,
coastal, inland meteorology).
In a more recent report, the same investigators presented other data
showing hemispheric differences in ozone concentrations and the seasonal
distributions of those concentrations (McElroy et al., 1985). Ozone concentra-
tions in the lower troposphere measured at Cape Grim, Tasmania (ca. 41°S),
showed strong similarities, including seasonal distributions, to those measured
at three different sites, at different longitudes, in Canada (latitudes of ca.
53°N to 59°N). Ozone concentrations from sites in West Germany, Switzerland,
and the United States showed even greater similarity, both in level and seasonal
distribution.
Although of interest and concern when estimating global ozone budgets,
variations in ozone concentrations with latitude have little practical signifi-
cance for assessing exposure within the contiguous United States. The effects
on ozone concentrations of latitude, as well as longitude, within the contiguous
states are minor.
A number of additional reports are available on the increase of tropospheric
ozone concentrations with altitude (e.g., Kroening and Ney, 1962; Galbally and
Roy, 1980). The data cited above showing latitudinal effects also show alti-
tudinal effects, particularly at mid-latitudes in both the northern and southern
hemispheres (Oltmans, 1981; cited in Logan et al., 1981). At mid-latitudes,
the increase in background ozone concentrations rises rather sharply (from a
spring-summer mixing ratio of about 40 to 60 ppb to one of about 100 ppb)
between altitudes of about 8 and 12 km (ca. 5 to 7.4 mi, or ca. 26,400 to
39,600 ft). Between the surface and about 2 km (ca. 1.2 mi, or ca. 6600 ft),
another relatively sharp increase in ozone concentrations is observed for the
mid-latitudes (from a spring-summer mixing ratio of about 35 ppb to one of
about 50 to 55 ppb) (Oltmans, 1981; cited in Logan et al., 1981). Other data
corroborate these findings. Seller and Fishman (1981) reported ozone measure-
ments taken on flights in remote tropospheric air during July and August 1974.
Their data also show that ozone concentrations increase with increasing altitude
and in general substantiate the accepted view that the lower atmosphere and
the surface of the earth act as ozone sinks.
5-62
-------
Increases in ozone concentrations at higher altitudes are of potential
concern relative to exposures sustained in-flight by airline passengers and
employees on high-altitude flights (see Chapter 11). Air-filtration systems
are employed on airplanes, however, that reduce ozone concentrations. Increases
in concentrations with altitude at lower altitudes (e.g., in the zero to 2 km
range) have potentially greater importance, since these gradients may affect
forest ecosystems in mountainous areas, (It must be borne in mind, however,
that atmospheric pressure decreases with altitude and thus, for a given concen-
tration, the mass of ozone per cubic meter of air also decreases with altitude.)
Contributors to increased ozone concentrations at the higher altitudes of
relevance for forest ecosystems include stratospheric ozone intrusions and ozone
transported aloft and conserved overnight in nocturnal inversions layers. The
latter, in particular, may be an important consideration for the accurate
assessment of the ozone exposures sustained by mountain forest ecosystems, as
discussed below (see Chapter 7 also).
While a number of reports contain data on ozone concentrations at high
altitudes (e.g., Coffey et a!., 1977; Reiter, 1977b; Singh et a!., 1977;
Evans, 1985; Lefohn and Jones, 1986), fewer reports are available that present
data for different elevations at the same locality. Studies by Berry (1964)
and Miller and Ahrens (1969) are among earlier reports documenting ozone
gradients with altitude in forested areas. Two more recent studies are dis-
cussed here because the investigators acquired and presented data on the
diurnal patterns of ozone at those sites, as well as on differences in ozone
concentrations at the respective altitudes. In addition, the two studies were
conducted in different parts of the country and presented data that demonstrate
differences in the meteorological influences on the concentrations found in
the respective studies.
The mixed-conifer forest ecosystem of the San Bernardino Mountains in
California is the most thoroughly studied forest ecosystem in the United
States (see Chapter 7). In conjunction with vegetation and forest ecosystem
studies conducted there in the 1960s and 1970s, a number of monitoring stations
were established at different elevations on the south-facing slopes of the San
Bernardino Mountains. The mountain range begins about 80 miles east of Los
Angeles and its center is almost due east of Pasadena. The first four monitor-
ing stations were established at Highland, City Creek, Mud Flat, and Rim
Forest, located as shown in Figure 5-22.
5-63
-------
2745
90
CD
EH
z
2135
1525
915
305
0
I
REDLANDS
I I I
I
CITY CREEK
—N-
HIGHLAND
I I I I I I
DEEP
CREEK
70
50
30
10
0
CD
0
8 12 16 20 24
DISTANCE, kilometers
28
32
36
Figure 5-22. Altitudinal sequence of monitoring sites in the San
Bernardino Mountains.
Source: Adapted from Miller et al. (1972).
5-64
-------
Total oxidant, temperature, and relative humidity (RH) were measured
continuously during 16 days in July and August at all but the Highland station
(Miller et a!., 1972). The vapor pressure gradient was calculated from RH and
temperature data. The time of the daily peak oxidant concentration was pro-
gressively later at stations of higher elevation, as shown in Figure 5-23.
Temperatures and vapor-pressure gradients were also progressively lower at
higher elevations at the time of the oxidant peak. The average duration of
3
oxidant concentrations exceeding 200 |jg/m (0.10 ppm) was 9, 13, 9, and
8 hr/day going from lower to higher stations for June, July, and August 1969.
The longer duration at City Creek (elevation 817 m; 2680 ft) probably coincided
with the zone where the inversion layer most often contacted the mountain slope.
3
The oxidant concentrations rarely decreased below 98 ug/m (0.05 ppm) at night
on the slope of the mountain crest, whereas they usually decayed to near zero
at the basin station (Highland, elevation 442 m; 1459 ft).
The vertical and horizontal distributions of oxidant air pollution in the
Los Angeles Basin have been described by several investigators and presented
in the 1978 criteria document (Blumenthal et a!., 1974; Edinger, 1973; Edinger
et al., 1972; Miller et al., 1972; as cited in U.S. Environmental Protection
Agency, 1978). In Los Angeles, a marine temperature inversion layer frequently
forms above the heavily urbanized metropolitan area and often extends inland
as far as 144 km (90 miles), depending on season and time of day. Surface
heating of air under the inversion increases with distance eastward in the
basin and often disrupts the inversion by midmorning at its eastern edge. The
marine temperature inversion layer encounters the mountain slopes, usually
below 1200 m (4000 ft). In the morning, the temperature inversion often
remains intact at this juncture, and air pollutants are confined beneath it.
The heated mountain slopes act to vent oxidant air pollutants over the crest
of the mountains and cause the injection of pollutants into the stable inversion
layer horizontally away from the slope. Oxidant concentrations within the
inversion are not uniform, but occur in multiple layers and strong vertical
gradients. In some cases, the inversion may serve as a reservoir for ozone,
which may arrive at downwind locations along the mountain slopes relatively
undiluted because of a lack of vertical mixing within the inversion layer and
a lack of contact with ozone-destroying material generated at the ground. The
important result of the trapping of oxidant in these layers is its prolonged
contact with high terrain at night.
5-65
-------
§
Q.
0.35
°'3°
0.25
S 0.20
§
J 0.15
O
»- 0.10
0.05
0
0.35
i 0.30
Q.
Q.
H* 0.25
Z
—, 0.20
0.15
0.10
0.05
0
HIGHLAND
(442m; 1459 ft)
95
90
85
g 80
Q.
Ul
«- 75
6 9 12 15
TIME, hours
18
21
95
90
". 85
ui
DC
1 8°
ui
a.
S 75
P
70
65
VPG
CITY CREEK
(817m; 2696 ft)
I
I
24
18
12
6 9 12 15
TIME, hours
18
21
E
UJ
111
CC.
«>
co
Ul
DC
Q.
6 .§
Q.
<
Figure 5-23. Relationship between elevation and diurnal patterns of
total oxidant concentrations, temperature, and vapor pressure at
four sites (A-D) in the San Bernardino Mountains, CA, July-August
1969.
Source: Miller et al. (1972).
5-66
-------
0.35
E 0.30
Q.
Q.
H* 0.25
g 0.20
O
< 0.15
O
H 0.10
0.05
0
LL
O
K
tt
Ml
O.
2
UJ
95
90
85
80
75
70
65
MUD FLAT
(1110m; 3663 ft)
24
18
12
_L
I
I
9 12 15
TIME, hours
18
21
en
X
E
E
UJ
UJ
cc
3
(0
(A
UJ
K
Q.
K
O
o.
a.
Q.
0.35
0.30
0.25
Q 0.20
X
O
-J 0.15
0.10
0.05
0
Ul
K
E
U4
Q.
5
95
90
85
80
75
70
65
1
I
RIM FOREST
(1725m; 5792 ft)
VPG
24
18
12
69 12 15
TIME, hours
18 21
S.
UJ
CC
o
01
cc
3
W5
W
UJ
DC
Q.
CC
2
Figure 5-23 (Cont'd). Relationship between elevation and diurnal
patterns of total oxidant concentrations, temperature, and vapor
pressure at four sites (A-D) in the San Bernardino Mountains, CA,
July-August 1969.
Source; Miller et al. (1972).
-5-67
-------
The studies in the Los Angeles Basin and the nearby San Bernardino Moun-
tains underscore the importance of understanding the interactions between
topography and local meteorology for specific areas. Consideration of those
factors is essential for accurate exposure assessments at higher altitudes,
especially in the absence of monitoring sites and adequate aerometric data.
Wolff et al. (1986) conducted a study of the effects of altitude on ozone
concentrations at three sites located at three separate elevations on High
Point Mountain in northwestern New Jersey. Ozone concentrations and temperature
were measured at ground level at the respective sites. In addition, temperature
and relative humidity were measured above the surface by balloon-borne sensors,
with wind direction and speed data obtained by means of a tracking theodolite.
Data for several days (July 1975) indicate that in mid-day, when atmos-
pheric mixing was good, vertical profiles were nearly constant, with concentra-
tions increasing only slightly with elevation. Likewise, the daily ozone
maxima were similar at different elevations. At night, however, ozone concen-
trations were nearly zero in the valley (the lowest-elevation site) and in-
creased with elevation. Comparison of the ozone dosages at the three sites
(number of hours > 0.08 ppm) showed that greater cumulative doses were sustained
at the higher elevations (Table 5-14). Comparable data from an urban area
(Bayonne, NJ) about 80 km southeast of High Point Mountain showed that the
cumulative doses were higher at all three of the mountain sites than in the
urban area (Table 5-14) (Wolff et al., 1986).
TABLE 5-14. COMPARISON OF OZONE CONCENTRATIONS AT THREE DIFFERENT
ELEVATIONS, HIGH POINT MOUNTAIN, NJ, AND
AT BAYONNE, NJ, JULY 1975
Locati on/el evati on
High Point, 500 m
'High Point, 300 m
High Point, 140 m
Bayonne, sea level
Source: Wolff et al.
1-hr
max,
ppb
66
61
59
69
(1986).
24- hr
mean,
ppb
49
38
26
33
1-hr
max,
ppb
130
110
120
119
24- hr
mean,
ppb
81
61
52
48
No. hr
>80 ppb
13
9
9
7
5-68
-------
The maximum depth of the nocturnal inversion at High Point Mountain was
about 900 m (ca. 2970 ft). Extrapolating from the gradient observed from the
ground sites, Wolff et al. (1986) estimated that mountain peaks exceeding 900
m (2970 ft) would sustain exposures to relatively constant diurnal concentra-
tions that would approximate the maximum ozone concentration found at the
valley site. The investigators concluded from their concentration and meteorol-
ogical data that elevated, mountainous sites in the eastern United States may
A'be expected to be exposed to higher ozone dosages than valley sites throughout
-1 the year.
5.5.2,5 Vertical Gradients at Ground Level. Just as macroscale variations in
ozone concentrations have been observed by measuring vertical and horizontal
profiles at various altitudes, so too have microscale vertical variations at
ground level been observed as a function of placement of sampling probes.
Data drawn from a recent study on rural ozone concentrations illustrate the
effects on concentration data of the placement of monitoring probes. These
effects are pertinent for vegetation exposures, in particular, although the
vertical gradient at ground level has obvious implications for human exposures
as well.
Pratt et al. (1983) studied concentrations of ozone and oxides of nitrogen
in the upper-midwestern part of the United States. Concentration data were
obtained over 4 years by means of monitors placed at two different sampling
heights (ca. 3 and 6, or 3 and 9 meters) at three air quality monitoring
sites: LaMoure County, North Dakota; Traverse County, Minnesota; and Wright
County, Minnesota. All stations were rural sites. The mean ozone concentra-
tions did not differ greatly among the sites, but in at least some instances
the mean differences between sampling heights were as large or larger than the
differences among the scattered sites. Table 5-15 presents the mean ozone
concentrations measured at two separate sampling heights (Pratt et al., 1983).
Annual average concentrations were 1 to 3 ppb lower at the 3.05-meter height
than at the 6.10 or 9.14-meter heights, reflecting the depletion of ozone near
the surface. As might be expected, the gradient was especially conspicuous at
night because of continued surface scavenging and a decrease in the rate of
transfer from layers aloft. The concentrations of ozone occurring at these
sites were near background in all years measured. In areas with higher ozone
concentrations, one would expect to see larger absolute gradients between
5-69
-------
TABLE 5-15. MEANS AND STANDARD ERRORS OF OZONE CONCENTRATIONS MEASURED OVER 4 YEARS
AT TWO SAMPLING HEIGHTS AT THREE STATIONS IN THE RURAL, UPPER-MIDWESTERN UNITED STATES
(ppb/v/v)
CJl
1
•-4
o
Site
LaMoure
County, ND
Traverse
County, MN
Wright
County, MN
Sampling
height, m
3.
9.
3.
9.
3.
6.
05
14
05
14
05
10
1977
22.97
(6
24.14
(6
24.44
(9,
26.29
(9,
± 0.20
,413)a
± 0.19
,410)
± 0.19
672)
± 0.19
810)
-
1978
35.30 ± 0.20
(15,218)
35.67 ± 0.19
(15,220)
35.2 ± 0.19
(22,675)
36.77 ± 0.19
(22,624)
34.64 ±0.23
(17,437)
35.61 ±0.23
(17,440)
1979
30.80 ± 0.16
(21,064)
32.25 ± 0.15
(20,470)
30.64 ± 0.17
(19,900)
32.39 ±0.16
(19,289)
28.71 ± 0.18
(17,771)
29.27 ±0.18
(17,775)
Years
1980
34.98 ± 0.25
(17,157)
37.53 ± 0.25
(17,157)
34.66 ± 0.23
(22,629)
37.60 ±0.23
(22,625)
J4.27 ± 0.26
(18,222)
35.16 ±0.27
(18,280)
31.
34.
28.
31.
31.
32.
1981
92 ± 0.26
(6,699)
23 ± 0.26
(6,364)
78 ± 0.23
(7,141)
08 ±0.22
(7,142)
60 ± 0.24
(7,764)
28 ± 0.24
(7,766)
1977-1981
32.26
(66,551)
33.82
65,597
32.09 .
(82,017). .
34.21
(81,550)
32.42
(61,254)
33.21
(61,261)
The numbers in parentheses refer to the number of hours of monitoring included in the reported values. Values are based on
all valid: data per site. For each sampling height at each site, values for three monitors separated by 76 m are included
in the calculations. Monitoring was conducted only during the second half of 1977 and only until 30 June in 1981.
Source: Pratt et al. (1983).
-------
monitors at different heights. In fact, the careful measurement of concentra-
tion gradients over di.stancesrof 1 to 10 meters above a surface is a recognized
method for estimating the scavenging potential of the surface.
In this context, it should be noted that the dry deposition of ozone on
various surfaces has been fairly extensively studied. Summaries on surface
scavenging and other natural removal processes appeared in the previous criteria
document for ozone and other photochemical oxidants (U.S. Environmental Protec-
tion Agency, 1978) and in the review by the National Research Council (National
Research Council, 1977). Newer review articles on dry deposition and surface
scavenging have been published by Wesely (1983) and Gal bally and Roy (1980),
among others.
5.6 CONCENTRATIONS OF PEROXYACETYL NITRATE AND PEROXYPROPIONYL NITRATE IN
AMBIENT AIR
5,6.1 Introduction ' .
As noted in the introduction to this chapter (Section 5.1), published
data on the concentrations in ambient air of photochemical oxidants other than
ozone are not comprehensive or abundant. Much more is known now, however,
about their atmospheric concentrations than was known when the 1978 criteria
document for ozone and other photochemical oxidants was published. Review of
the data that follow will show that peroxyacetyl nitrate (PAN) and peroxypro-
pionyl nitrate (PPN) are the most abundant of,the non-ozone oxidants in ambient
air in the United 'States other than the inorganic nitrogenous oxidants such as
nitrogen dioxide (NO,,), and possibly nitric acid (HNO~), in some areas. The
inorganic nitrogenous oxidants found in ambient air are reviewed in air quality
criteria documents on the nitrogen oxides and are not treated in this document
except for the review of the role of nitric oxide (NO) and N02 in atmospheric
photochemistry.
Available data show that PAN is toxic to vegetation and is potentially
toxic to humans. No data indicating potentially adverse effects of PPN are
available. It exists in trace quantities and apparently no research has ever
been undertaken to determine its potential toxicity. At least one study
(Heuss and Glasson, 1968) has reported that a higher homologue of the series,
peroxybenzoyl nitrate (PBzN), like PAN, is a lachrymator. No unambiguous
identification of PBzN in the ambient air of the United States has been made,
however. • ' ,- .•! • • ''. : . ,
5-71
-------
Given the information available on PAN, the concentrations of PAN that
are of most concern are those to which vegetation could potentially be exposed,
especially during daylight hours in agricultural areas; followed in importance
by those both indoors and outdoors, in both urban and nonurban areas, to which
human populations could potentially be exposed. Most of the available data on
concentrations of PAN and PPN in ambient air are from urban areas. The levels
to be found in nonurban areas will be highly dependent upon the transport of
PAN and PPN or their precursors from urban areas, since the concentrations of
the NO precursors to these compounds are considerably lower in nonurban than
4 ppb around 10:00 a.m. and one of ~10 ppb between 4:00 and 6:00 p.m.
Seasonal data from Riverside for June 1966 to June 1967 showed that PAN concen-
trations were highest in September 1966 and in March and June 1967.
Total oxidants (by Mast meter) in these Los Angeles sites reached a peak
concentration (the same composite diurnal data as above) of close to 140 ppb
in September 1965 at about the same hour of day as the PAN peak. In October,
total oxidants peaked at nearly 200 ppb, again coinciding in time with peak
PAN concentrations. In Riverside, the morning PAN peak preceded the oxidant
peak (~110 ppb around noon) by almost 2 hours but the afternoon PAN peak
trailed the afternoon oxidant peak (~160 ppb around 2:00 to 4:00 p.m.) by
about 2 hours. The ozone/PAN ratio was variable from month to month but was
generally lower from November through April than during the rest of the year,
5-72
-------
i.e., PAN was generally a greater percentage of the total oxidant during the
November through April period.
In the 1978 criteria document for ozone and other photochemical oxidants
(U.S. Environmental Protection Agency, 1978), additional PAN data for Los
Angeles as well as two other cities were presented. Lonneman et al. (1976)
measured PAN and total oxidants in Los Angeles in 1968. In 118 samples col-
lected for the period 10:00 a.m. to 4:00 p.m. during the study, the median PAN
concentration was 13 ppb and the average PAN concentration was 18 ppb. The
median total oxidant concentration (measured by UBKI) was 97 ppb and the
average was 117 ppb. Thus, the median oxidant/PAN ratio was 7.5 and the average
oxidant/PAN ratio was 6.5; the median PAN concentration was 13.4 percent of
the median total oxidant concentration, and the average PAN concentration was
15.4 percent of the average total oxidant concentration.
Lonneman et al. (1976) conducted similar studies in Hoboken, New Jerseyi
in 1970 and in St. Louis, Missouri, in 1973. Samples were measured during the
period 10:00 a.m. to 4:00 p.m. over the course of the study. In Hoboken, PAN
concentrations averaged 3.7 ppb. Ozone concentrations (measured by chemilumi-
nescence) averaged 90.5 ppb. In St. Louis, PAN averaged 6.4 ppb, ozone (meas-
ured by chemiluminescence) averaged 50.1 ppb, and total oxidants (by UBKI)
averaged 74.3 ppb.
From these 1966 through 1973 urban data, it is clear that PAN concentr^r
tions in urban areas are appreciably lower than those of ozone, even in the
winter when PAN is higher relative to ozone than in summer.
In addition to urban data, the 1978 criteria document (U.S. Environmental
Protection Agency, 1978) also included PAN concentration data from one nonurban
agricultural area, Wilmington, Ohio (Lonneman et al., 1976). The maximum PAN
concentration observed in 1500 samples taken during August 1974 was 4.,1 ppb.
The daily maximum PAN concentration rarely exceeded 3.0 ppb even though the
daily maximum ozone concentration frequently exceeded 80 ppb.
Conclusions reached in the 1978 document were (1) that PAN concentrations
are much lower in the ambient air of nonurban than of urban areas; and (2)
that ozone/PAN or total oxidant/PAN ratios vary with location, such ratios
being higher in nonurban areas than in urban; that is, PAN concentrations are
a lower percentage of total oxidant concentrations in nonurban areas (U.S.
Environmental Protection Agency, 1978). From data presented in the 1970
document, it may be concluded (1) that oxidant/PAN ratios vary with season;
5-73
-------
(2) that PAN concentrations are lower than ozone concentrations in urban
areas; and (3) that PAN and ozone concentrations exhibit similar but not
identical diurnal patterns. ,
The historical data presented above have been given in some .detail because
the information about PAN conveyed by those data remains valid. - Examination
of the more recent data presented in subsequent sections shows that newer data
extend and corroborate, for the most part, the findings of the older literature.
5.6.3 Ambient Air Concentrations of .PAN and Its Homologues in Urban Areas
Additional studies on concentrations of PAN and its homologues in both
urban and nonurban areas are now available. Newer data are presented in this
section in summary form, where possible; and the individual studies or examples
presented are merely a few of many, chosen to represent current knowledge
regarding ambient air PAN concentrations and their patterns.
The existing literature on the concentrations in ambient air of the per-
oxyacyl nitrates, PAN and its higher homologues5 has been compiled and examined
in two recent review articles (Temple and Taylor, 1983; Altshuller, 1983). In
the first, Temple and Taylor reviewed the concentrations of PAN in the ambient
air in Europe, Japan, and North America in the context of the phytotoxicity of
PAN. Altshuller, in the second, reviewed the published concentrations of PAN
and of PPN in ambient air, also within and outside the United States. In
addition, Altshuller analyzed the relationships to ozone of PAN and other
photochemical reaction products. The reader is referred to these reviews for
detailed information and for references therein. •
Table 5-16 presents a summary of PAN concentrations observed in the
ambient air of urban areas of the United States. Data in this table include
the results of studies cited in Section 516.2 above, along with the results of
newer studies. This table was derived from the reviews of Altshuller (1983) and
Temple and Taylor (1983), as well as from a few additional sources (Jorgen et
a!., 1978; Lewis et a!., 1983). The data are summarized in the table by
region of the United States and by date, with the newer studies .reported first
for each region. ,
Because of variations in diurnal patterns of PAN by location and season,
and because no national, uniform aerometric data base for PAN exists, few of
the data reported in Table 5-16 are really comparable. Thus, data in this
table lend themselves to general conclusions but not to the analysis of trends
5-74
-------
.3.
ID
TABLE 1-16. SUMMARY OF CONCENTRATIONS OF PEROXYACETYL NITRATE IN AMBIENT AIR IN URBAN AREAS OF THE UNITED STATES
PAN concentrations,
Site
West : . • •
Riverside, CA.
• W. Los Angeles, CA
Claremont, CA-
• Claremont, CA
Claremont, CA
en
i
-sj
01 Riverside, CA
Riverside, CA
Riverside, CA
Riverside, CA
West Covina, CA
West Covina, CA
Pasadena, CA
Riverside, CA
Downtown
Los Angeles, CA
Time
of yr
All year
June
Sept. -Oct.
Aug. -Sept.
Oct.
Apr. , May
July, Aug. ,
Oct.
All year
Oct.
Aug. , Sept.
Aug. , Sept.
July
Aug. -Apr.
Sept.-
Nov.
Yr
1980
1980
1980
1979
1978
1977
1977
1975 (May)-
1976 (Oct.)
1976
1977
1973
1973
1967-1968
1968
Hours
sampled
8 a.m. -8 p.m.
6:35 a.m. -1:35 p.m.
. 24 hr/day
Morning to late
evening
Late morning to
late evening
24 hr/day
.Late, morning to
evening
24 hr/day
Late morning to
early evening
23 hr/day
NA
7 a.m. -4:30 p.m.
24 hr/day
10 a.m. -4 p.m. .
No. days
sampled
365
2
.11
8
.5
10
10
533
3
24
NA
3
273
NA
Method9
GC-ECD
LP-FTIR '
• GC-ECD
GC-ECD
LP-FTIR
GC-ECD
LP-FTIR
"GC-ECD -
LP-FTIR
GC-ECD
GC-ECD
LP-FTIR
GC-ECD
GC-ECD
Avg.
NAb
1
'.13
4,
6
1.6
7
3.6
9
9
NA
30
NA
8
Max.
41.6
16
47 :
11
37
5.7
18
32
18
20
46
, 53
58
68
ppb
Monthly
mean
4.9.
NA
NA
NA
NA
NA
NA
NA
NA
NA
8.8
NA
4.6
NA
Original reference
Temple and Taylor (1983)
Hanst.et al. (1982)
Grosjean (1983)
Tuazon et al. (1981a)
Tuazon et al. (1981a;
1981b)
Singh et al. (1979)
Tuazon et al. (1980)
Pitts and Grosjean (1979)
Tuazon et al. ' (1978)
Spicer (1977)
Spicer (1977) .
Hanst et al. (1975)
Taylor (1969)
Lonneman et al. (1976)
-------
TABLE 5-16 (continued). SUMMARY OF CONCENTRATIONS OF PEROXYACETYL NITRATE IN AMBIENT AIR IN URBAN AREAS OF THE UNITED STATES
.in
i
cr>
Site
Salt Lake
City, UT
Los Angeles, CA
Downtown
Los Angeles, CA
Southwest
Houston, TX
Houston, TX
Midwest
Dayton, OH
(Huber Heights, OH)
St. Louis, MO
St. Louis, MO
East
New Brunswick, NJ
New Brunswick, NJe
Hoboken, NJ
Time
of yr
July-
Sept.
Sept.-
Oct.
July-
Oct.
Oct.
July
July,
Aug.
June-
Aug.
Aug.
All year
Sept.-
- Dec.
June, July
Yr
1966
1965
1960
1977
1976
1974
1973
1973
1978 (Sept.)
-1980 (May)
1978
1970
Hours No. days
sampled sampled
7 a.m. -3 p.m.
8 a.m.-l p.m. 35
9 a.m. -Noon 9
8 a.m. -8 p.m.
24 hr/day 22
24 hr/day 20
23 hr/day 26
8 -to 24 hr/day 12
24 hr/day 400
(9600 1-hr values)
8 a.m. -6 p.m.
10 a.m. -4 p.m.
PAN
Method3
GC-ECD
GC-ECD
IR
GC-ECD
. GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
GC-ECD
•GC-ECD
concentrations, ppb
Avg. Max.
54
31 214
~20 70
15.6
0.4 11.5
0.7 10
1.8 19
6.3 25
0.5 10.6
10.5
9.9
Monthly
mean Original reference
4.4 Tingey and Hill (1968)
38 (Sept.) Mayrsohn and Brooks
40 (Oct.) (1965)
Renzetti and
Bryan (1961)
Jorgen et al. (1978)
Westberg et al. (1978)
Spicer et al. (1976)
Spicer (1976)
. - Lonneman et al. (1976)
Lewis et al. (1983)
2.7 Brennan (1980)
3.7 Lonneman et al. (1976)
GC-ECD, gas chromatography with electron-capture detector; LP-FTIR, long-path Fourier-transform infrared spectroscopy; IR, infrared spectroscopy.
Not available.
cSubset of data reported by Lewis et al. (1983), cited in table above.
Source: Altshuller (1983); Temple and Taylor (1983); Lewis et al. (1983); Jorgen et al. (1978).
-------
over the past decade and a half or necessarily to analysis of between-city or
between-region similarities or differences. For example, concentrations of
PAN in Los Angeles appear to have been a great deal higher in 1965 (Mayrsohn
and Brooks, 1965) than in 1980 (Hanst et a!., 1982) for nearly the same time
of day, but the 1965 concentrations were measured in September and October and
the 1980 concentrations were measured in June. Other data from California
indicate that September and October are more likely to be part of the smog
season there than June, A comparison by Temple and Taylor (1983) of PAN
concentrations in Riverside in 1980 with those in 1967 and 1968 indicates
little difference. Again, however, the sampling periods were not identical
relative to averaging time or time of year. Tabulated data from Temple and
Taylor (1983) have been plotted in Figure 5-24.
In addition to compiling existing data on the concentrations of PAN in
ambient air, Altshuller (1983) also related PAN concentrations to ozone concen-
trations where data for both exist. It must be borne in mind for this review,
as well, that sampling periods (years, time of year, number of measurements)
are not identical and in many cases are not even similar between studies (see
Table 5-16). Neither are the averaging times over which samples were collected
and calculated within respective studies identical or even necessarily similar.
Nevertheless, the data of Altshuller constitute a comprehensive review and
examination of the relationships among respective photochemical oxidants in
urban areas of the United States. Concentrations of PAN as a percentage of
ozone concentrations are given in Table 5-17, but Table 5-16 should be consulted
for information on sampling periods and averaging times.
The existence of peroxybenzoyl nitrate (PbzN) in ambient air of urban
areas was postulated in the 1978 criteria document for ozone and other oxidants
(U.S. Environmental Protection Agency, 1978), but PBzN has not been clearly
identified in ambient air in the United States. Hanst et al. (1982) estimated
that as much as 2 ppb PBzN would be clearly detectable in FTIR measurements
but reported no clear absorption band for PBzN in their measurements during a
smoggy period in Los Angeles. They estimated an upper limit of 1 ppb PBzN
during their 1980 study (the maximum ozone concentration was 272 ppb and the
maximum PAN concentration was 16 ppb during that period).
The only homologue of PAN that has been unambiguously identified in
ambient air in the United States is peroxypropionyl nitrate (PPN). In his
5-77
-------
0.07
0.06
J 1967-1968
- I I 198°
I 0.05
a
g
fc 0.04
ui
S 0.03
o
o
0.02
0.01
AUG SEPT OCT MOV DEC JAN FEB MAR APR
Figure 5-24. Comparison of monthly daylight average and
maximum PAN concentrations at Riverside, CA, for 1967-
1968 and 1980.
Source: Derived from Temple and Taylor (1983).
5-78
-------
TABLE 5-17. RELATIONSHIP OF OZONE AND PEROXYACETYL NITRATE AT
URBAN AND SUBURBAN SITES IN THE UNITED STATES
Site/year
California
Downtown Los Angeles, 1960
Downtown Los Angeles, 1965
Downtown Los Angeles, 1968
West Los Angeles, 1980
Pasadena, 1973
West Covina, 1977
Claremont, 1978
Claremont, 1979
Riverside, 1967-1968
Riverside, 1975-1976
Riverside, 1976
Riverside, 1977
Riverside, 1977
Southwest
Houston, TX, 1976
Midwest
St. Louis, MO, 1973
St. Louis, MO, 1973
Dayton, OH, 1974
(Huber Hts, OH)
East
Hoboken, NJ, 1970
New Brunswick, NJ, 1978-1980
Avg.
8
NA
13
9
10
20
7
4
8
9
5
4
4
3
13
5
2
4
4
PAN/OS, %
At 03 peak
7
7
NAa
6
8
12
6
4
NA
5
4
4
NA
3
NA
5
1
NA
2
Reference
Renzetti and Bryan (1961)
Mayrsohn and Brooks (1965)
Lonneman et al . (1976)
Hanst et al . (1982)
Hanst et al . (1975)
Spicer (1977)
Tuazon et al. (1981a, 1981b)
Tuazon et al . (1981a)
Taylor (1969)
Pitts and Grosjean (1979)
Tuazon et al . (1978)
Tuazon et al. (1980)
Singh et al . (1979)
Westberg et al . (1978)
Lonneman et al . (1976)
Spicer (1977)
Spicer et al . (1976)
Lonneman et al . (1976)
Brennan (1980)
Not available.
Source: Adapted from Altshuller (1983).
5-79
-------
review of existing literature, Altshuller (1983) compiled data on the concen-
trations of PPN in ambient air in urban areas. In addition, He calculated
ratios of the concentrations of PPN and PAN., His data, expressed as
percentages [(PPN/PAN) x 100], are presented as Table 3-18 (Altshuller, 1983).
As Altshuller has pointed out, average PPN concentrations are 10 to 30
percent of the average PAN concentrations shown in Table 5-18 except for those
reported for San Jose (8 percent) and Oakland (42 percent). The maximum PPN
concentrations reported are highly variable, however, ranging from 0.13 ppb in
San Jose (August 1978) to 6.0 ppb in Riverside (month and year as well as
sampling period of day unknown). Thus, the PPN/PAN ratio at maximum concentra-
tions of PPN is highly variable, as well. Among more recent data, the maximum
PPN concentration was 5.0 ppb in St. Louis in August 1973. Note, however,
that the sampling period in St. Louis was 10:00 a.m. to 3:30 p.m. Depending
upon temperature, concentrations of precursors, and other factors, a true PPN
maximum may or may not have occurred by 3:30 p.m.
Table 5-19 presents PAN and PPN concentrations reported by Singh et al.
(1981) for three cities, Los Angeles, Oakland, and Phoenix; as well as PPN
concentrations as percentages of PAN concentrations (PPN/PAN x 100). Comparison
of the data from these three cities helps demonstrate the variability of PPN
and PAN concentrations with location.
5.6.4 Ambient Air Concentrations of PAN and Its Homologues in Nonurban Areas
Data on the concentrations of PAN and PPN in agricultural and other non-
urban areas of the United States are sparse. They include data from the study
done by Lonneman et al. (1976) in Wilmington, Ohio, in August 1974, and cited
earlier in Section 5.6.2. In that study, measurements made by GC-ECD from
10:00 a.m. to 4:00 p.m., local .time, showed a maximum concentration for the
study period of 4.1 ppb. The average daily maximum was 2.0 ppb. While the
4:00 p.m. cutoff used by Lonneman et al. (1976) could possibly have resulted
in missing some peak PAN concentrations, especially in transported air masses,
the data are within the range reported by Westberg et al. (1978) and by Spicer
and Sverdrup (1981) for 24-hour measurements at nonurban sites. At the Sheldon
Wildlife Reserve, Texas, Westberg et al. (1978) found a 24-hour average PAN
concentration of 0.64 ppb and a maximum concentration of 2.8 ppb for the study
period (October). Spicer and Sverdrup (1981) found a 24-hour average PAN
5-80
-------
TABLE 5-18. AMBIENT AIR MEASUREMENTS OF PEROXYPROPIONYL NITRATE (PPN) CONCENTRATIONS
BY ELECTRON CAPTURE GAS CHROMATOGRAPHY AT URBAN SITES IN THE UNITED STATES
CJl
Site
Los Angeles, CA
Riverside, CA
Riverside, CA
Riverside, CA
San Jose, CA
Oakland, CA
Phoenix, AZ
Denver, CO
Houston, TX
St. Louis, MO
St. Louis, MO -,-.
.Chicago, IL
Pittsburgh, PA
Staten Island, NY
Period of
measurement/
no. of days (n)
April 1979 (13)
NAa (1)
April -May
July 1980 (13)
August 1978 (7)
June-July 1979
(13)
April -May 1979
(14)
June 1980 (14)
May 1980 (12)
10 Aug. 1973
(1)
-May- June, 1980
(9)
April -May, 1981
(13)
April 1981 (11)
March- April
1981 (11)
Period
of day
24
NA
24
24
24
24
24
24
24
100
.(LT
24
24
24
24
hr
hr
hr
hr
hr
hr
hr
hr
g-1530
hr
hr
hr
hr
PPJf,
Avg.
0.7
NA
0.3
0.2
0.08
0.15
0.09
0.05
0.11
3.0
0.66
0.05
0.05
0.20
ppb
Max.
2.7
6
1.8
0.9
0.13
0.5
0.33
0.32
0.63
5.0
0.25
0.13
0.07
3.1
PPN/PAN
Avg.
15
NA
21
16
8
28
12
10
14
17
23
12
17
27
,*
Max.
16
12
32
16
10
42
9
3
25
20
28
8
10
80
Reference
Singh
Darley
Singh
Singh
Singh
Singh
Singh
Singh
Singh
et
et
et
et
et
et
et
et
et
Lonneman
Singh
Singh
Singh
Singh
et
et
et
et
al.
al
al.
al.
al.
al.
al.
al.
al.
et
al.
al.
al.
al.
(1981)
. (1963)
(1979)
(1981)
(1979)
(1981)
(1982)
(1982)
(1982)
al. (1976)
(1982)
(1982)
(1982)
(1982)
Not available.
DLocal time.
Source: Altshuller (1983).
-------
en
i
00
ro
TABLE 5-19. CONCENTRATIONS OF PEROKYACETYL AND PEROXYPROPIONYL NITRATES
IN LOS ANGELES, OAKLAND, AND PHOENIX, 1979a
(ppb)
Los Angeles
Value
Mean
Std. dev.
Maximum
Minimum
PAN
4.98
4.48
16.82
0.03
PPN
0.72
0.67
2.74
NDb
PPN/PAN, %
14
_
16
-
PAN
0.36
0.42
1.85
0.05
Oakland
PPN PPN/PAN, %
0. 159 42
0.12
0.50 27
NDb
PAN
0.78
0.77
3.72
NDb
Phoenix
PPN PPN/PAN, %
0.09 12
0.08
0.33 9
NDb
Measurements were made by gas-phase coulometry.
bNot detectable.
Source: Singh et al. (1981).
-------
concentration of 0.50 ppb and a maximum of 6.5 ppb for the study period (July
and August) at Van Hiseville, New Jersey, in the New Jersey pine barrens; and
Spicer et al. (1983) reported averages of 0.46 ppb at a nonurban site in
Indiana (Huntington Lake) and 0.74 at a nonurban site in east central Missouri
(42 km west of St. Louis).
In contrast with these values, 24-hour average concentrations of PAN at
one nonurban and four remote sites (see Table 5-20) were reported by Singh et al.
(1979) to range from 0.08 to 0.30 ppb, with maxima at those sites ranging from
0.22 to 0.83 ppb. The Texas, Ohio, and New Jersey sites sustained higher PAN
concentrations than the nonurban sites where Singh et al. (1979) measured PAN.
The higher concentrations may reflect the influence at those sites .of nearby
metropolitan areas.
Concentrations of PAN have recently been reported by Singh and Salas
(1983) for a Pacific marine site, Point Arena, California, at which earlier
measurements were also made and reported by Singh et al. (1979) (see Table
5-19). Data collected in August 1982 showed concentrations of PAN ranging
from 0.01 to 0.12 ppb during the 5-day study period. The average concentration
for the period was 0.032 ± 0.024 ppb. Winds were west-to-northwest 90 percent
of the time and northerly the rest of the time. Modeled trajectories confirmed
that air masses passing over the site were of a marine origin. The site is
thought to be free of manmade pollutants..
Data from two nonurban sites in Canada are of interest even though they
are outside the United States. Cherniak and Corkum (1981; cited in Temple and
Taylor, 1983) measured PAN at a nonurban site in Simcoe, Ontario, Canada, for
6 months. Measurements made by GC-ECD showed monthly means of <2 ppb and a
maximum concentration during the study of 5.6 ppb. At a remote site in the
Kananaskis Valley of Alberta, Canada, monthly mean concentrations were <1 ppb
for samples taken at half-hour intervals, 24 hr/day, for 110 days (Peake et al.,
1983). The site, located at the base of a mountain range and about 50 miles west
of Calgary, is thought to be free of manmade pollutants, including transported
pollutants.
5.6.5 Temporal Variations in Ambient Air Concentrationsof Peroxyacetyl Nitrate
5.6.5.1 Diurnal Patterns. Concentration data obtained in the 1960s were
briefly discussed in Section 5.6.2, where it was noted that the first criteria
5-83
-------
TABLE 5-20. CONCENTRATIONS IN AMBIENT AIR OF PEROXYACETYL AND PEROXYPROPIONYL NITRATES AMD OZONE
AT NONURBAN REMOTE SITES IN THE UNITED STATES
Site
Mill Valley, CA
Point Arena, CA
Badger Pass, CA
Reese River, NV
Jetmore, KA
Wilmington, OH
Van Hiseville, NJ
Reference
Singh et al. (1979)
Singh et al. (1979)
Singh et al. (1979)
Singh et al. (1979)
Singh et al. (1979)
Westberg et al.
(1978)
Lonneman et al.
(1976)
Spicer and Sverdrup
(1981)
Nature of site
Maritime
Clean-Maritime
Remote-high
altitude
Remote-high
altitude
Rural-
continental
Rural -
continental
Rural -
continental
Rural -
continental
Period of
measurement and
no. of days (n)
Jan. 1977 (12)
Aug. -Sept. 1978 (7)
Hay 1977 (10)
May 1977 (7)
June 1978 (7)
October 1978 (9)
August 1974 (9)
July-Aug. 1979 (31)
Period
of day
24 hr
24 hr
24 hr
24 hr
24 hr
24 hr
10:00
a.m.-
4:00
p,n.
24 hr
Average
concentrations
PAN PPN 03
0.30
0.08
0.13
0.11
0.25
0.64
NAb
0.50
0.04
N0a
0,05
0.04
NDa
NDa
NDa
NDa
38
NDa
46
39
31
47
NAb
36
Maximum
concentrations
PAN
0.83
0.28
0.22
0.26
0.52
2.8
4.1
6.5
PPN
0.11
NDa
0.09
0.09
NDa
NO3
NDa
NDa
03
0,55
NDa
54
56
53
148
107
161
Avg. PAN/
Avg. 03, %
0.8
NDa
0.3
0.3
0.8
1.4
NAb
*
1.4
Not determined.
Measured, but results not given in the reference.
Source: Altshuller (1983).
-------
document for photochemical oxidants (U.S. Department of Health, Education, and
Welfare, 1970) reported concentrations and patterns for PAN that remain valid
now. In that document, the general proximity in time of PAN and oxidant peaks
was shown in data from Los Angeles and Riverside, California. Maximum PAN
concentrations, although varying from location to location, generally occurred
in midday; i.e., late morning to mid-afternoon. Figures 5-25 and 5-26, taken
from the 1970 criteria document (U.S. Department of Health, Education, and
Welfare, 1970), graphically present the diurnal patterns of PAN in Los Angeles
in 1965 and in Riverside in 1966. The occurrence of the second PAN peak in
Riverside, which appears to trail a second total oxidant peak by an hour or
two, was ascribed to transport, as verified by the occurrence of maximum
oxidant concentrations at three receptor sites east of West Los Angeles (down-
town Los Angeles, Azusa, and Riverside), at times that corresponded, wind
speed factored in, with respective distances from West Los Angeles.
Examples drawn from recent data substantiate that the general diurnal
pattern (as it appears in composite diurnal data averaged over a week, a
month, or longer) remains the same as the pattern established by data obtained
in the mid-1960s.
Using FTIR spectroscopy, Tuazon and coworkers (Tuazon et al., 1978, 1980,
1981b) measured concentrations of PAN at Claremont and Riverside, California,
over a 5-year period. Concentrations of PAN ranged from about 5 to 40 ppb
over the course of the study. The diurnal profiles for PAN and ozone at
Claremont are shown for 2 days of a multiday smog episode in October 1978 in
Figure 5-27 (Tuazon et al., 1981b). Note the qualitative relationship of the
two pollutants, with the peak concentrations of the two occurring almost
simultaneously. The relationship between PAN and ozone concentrations and
behavior in the atmosphere is neither constant nor monotonic, however, as is
borne out by the slight differences in time of occurrence of their peak concen-
trations but especially by the persistence of somewhat elevated PAN concentra-
tions before return to "baseline" levels. It appears that PAN concentrations,
in this instance at least, closely parallel the nitric acid (HNO,) concentra-
tions, persisting after ozone concentrations have subsided. The percentage of
PAN relative to ozone differed slightly at the peak concentrations of the two
on the 2 days. On October 12, the peak PAN concentration was close to 6 percent
of the peak ozone concentration; on October 13, the peak PAN concentration was
nearly 8 percent of the peak ozone concentration.
5-85
-------
£
Q.
Q.
CO*
z
o
I
UJ
o
z
o
o
Q.
Q
Q
X
O
I
0.20
0.18
0.16
0,14
0.12
0.10
0.08
0.06
0.04
0.02
0
^H MMM ^^ 1
AVERAGES:
19 WEEKDAYS,
— I OCTOBER
- 116 WEEKDAYS,
SEPTEMBER
9 10
a.m—
11 12
1 2 3
—p.m.i. • t»-j
HOUR OF DAY, PST
Figure 5-25. Variation of mean 1 -hour oxidant
and PAN concentrations, by hour of day, in
downtown Los Angeles, 1965.
Source: U.S. Department of Health, Education,
and Welfare (1970).
5-86
-------
Q.
a.
g
S
01
u
I
I
o
X
o
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
1IIIT~T
OXIDANT
0.010
Q.
Q.
0.008 2
0.006
0.004
0.002
O
5
ec
01
o
z
O
u
6
8 10
12
8
•a.m.-
•p.m.
HOUR OF DAY. PST
Figure 5-26. Variation of mean 1-hour
oxidant and PAN concentrations, by hour of
day. Air Pollution Research Center,
Riverside, CA, September, 1966.
Source: U.S. Department of Health,
Education and Welfare (1970).
5-87
-------
en
00
00
a.
a
cc
z
Ul
u
z
o
u
Ul
z
O
N
O
0.50
1000 1400 1800
OCTOBER 12,1978
2200 0200 0600 1000
TIME OF DAY, PDT
1400 1800 2200
i—J
•OCTOBER 13,1978-
a
a
o
z
X
Figure 5-27. Diurnal profiles of ozone and PAN at Claremont, CA,
October 12 and 13,1978, 2 days of a multi-day smog episode.
Source: Tuazon et al. (1981b).
-------
5.6.5.2 Seasonal Patterns. Seasonal differences in PAN concentrations were
alluded to in Section 5.6.3 and mean and maximum PAN concentrations were pre-
sented by month for 2 years (1967-1968 and 1980) for Riverside, California, in
Figure 5-24. That seasonal differences exist in PAN concentrations was also
documented in the 1970 criteria document for photochemical oxidants (U.S.
Department of Health, Education, and Welfare, 1970). Total oxidants and PAN
were monitored for 13 months in Riverside, California. The oxidant concentra-
tions were obtained by continuous Mast meter measurements, 24 hr/day. Concen-
trations of PAN were measured in sequential samples analyzed by GC-ECD from
6:00 a.m. to about 4:00 or 5:00 p.m. The data are not strictly comparable,
since the shorter, daylight averaging time for PAN would be expected to result
in somewhat higher mean concentrations of PAN than would be obtained across a
24-hour averaging period. Nevertheless, the patterns given in Figure 5-28 are
of interest and demonstrate that peak PAN concentrations can constitute a
higher percentage of the peak ozone concentrations during winter months than
during the rest of the year. This observation is still valid (Spicer et al.,
1983) and has been attributed by Lewis et al. (1983) and by Holdren et al.
(1984) to greater PAN stability in the winter months because of cooler tempera-
tures (Cox and Roffey, 1977). The possibility, however, that the somewhat
greater NO emissions of the winter heating season also contribute to this
JTl
phenomenon should not be overlooked.
Data from one additional study complement the older data already presented
on diurnal and seasonal patterns (Section 5.4.2). Lewis et al. (1983) measured
PAN and ozone concentrations from September 25, 1978, to May 10, 1980, in New
Brunswick, New Jersey. Average (10-hr and 24-hr) and maximum concentrations
of both pollutants are given in Table 5-21 by month of the year (Lewis et al.,
1983). Note that the highest monthly mean concentrations, both the 10-hr and
24-hr means, occurred during the smog season (August and September) but that
the next highest occurred in October and February, respectively. Average
diurnal profiles were obtained during this same study and are shown, by month,
in Figure 5-29 (Lewis et al., 1983).
5.6.6 Spatial Variations in Ambient Air Concentrations of Peroxyacetyl Nitrate
5.6.6.1 Urban-Rural Gradients and Transport of PAN. As noted earlier, precur-
sors to PAN, especially N02, are lower in nonurban than in urban areas, such
that little local formation is expected in nonurban areas. Available data on
5-89
-------
III!)
MONTHLY MEANS OF DAILY MAXIMUM
1-hour AVERAGE CONCENTRATIONS
MONTHLY MEANS OF 1-hour AVERAGE
CONCENTRATIONS
-O -O— —
—.*-• ^ PAN
I L
JUN. JUL AUG. SEP, OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUN.
I MONTH I
U 1966— »-U 1967-
si
O
z
O
O
Figure 5-28. Monthly variation of oxidant (Mast meter, continuous
24-hr) concentrations and PAN (GC-ECD, sequential, 6:00 a.m. to
4:00-5:00 p.m.) concentrations. Air Pollution Research Center,
Riverside, CA, June 1966 - June 1967.
Source: U.S. Department of Health, Education, and Welfare (1970).
5-90
-------
TABLE 5-21. PAN AND OZONE CONCENTRATIONS IN AMBIENT AIR,
NEW BRUNSWICK, N.J., FOR SEPTEMBER 25, 1978, TO MAY 10, 1980
PAN concentration, ppb
Month
January
February
March
April
May -
June
July
August
September
October
November
December
24- hr
average
0.12
0.61
0.36
0.45
0.23
0.09
0.26
1.17
1.04
0.93
0.25
0.57
10-hr
average
0.19
0.69
0.41
0.57
0.28
0.17
' 0.44
1.63
1.41
1.08
0.31
0.62
Hourly
maximum
1.3
4
1.3
2.5
1
0.8
3.5
10.6
7.5
5.8
3.5
2.5
QS concentration, ppb
24- hr
average
11.5
17.2
23
28.5
31.4
NA
37,5
37.4
22.4
15.8
11.6
9.7
10-hr
average
15.5
23.2
29.1
37.3
40.9
NA
57.6
55.9
33.9
22.6
15.8
12.8
Hourly
maximum
34
40
58
80
78
NA
" 130
145
110 .
68.''
40
35
These results are lower than expected; however, there was no evidence of
instrument malfunction. ,
Source: Lewis et al. (1983). ;
PAN concentrations indicate clearly that they are lower in nonurban areas than
in urban (Section 5.5.3). It should be noted, however, that few data exist on
concentrations in agricultural areas and that the possibility that PAN is
transported is therefore important in assessing exposures of vegetation to
PAN. Lonneman et al. (1976) and Nieboer and Van Ham (1976), in studies cited
in the 1978 criteria document (U.S. Environmental Protection Agency, 1978),
reported the transport of PAN. The more recent study by Nielsen et al. (1981)
has confirmed that PAN can be present at relatively high concentrations in
photochemically polluted air after long-range transport. Variations in con-
centrations of PAN and other oxidants measured in Claremont, California
(Grosjean, 1983), are consistent with transport patterns. The recent work of
Singh and Salas (1983) has shown that significant nighttime PAN concentrations
can occur aloft, at least in a relatively clean environment. It is possible
that the transport of PAN occurs aloft, as with ozone, and that under favorable
conditions PAN can be transported long distances.
5-91
-------
E
O
Z
O
0.080
0.060
0.040
0.020
0.025
0.015
0.005
_ OZONE
_ PAN
-a.m.-
12
-p.m.
TIME OF DAY, hour
24
Figure 5-29. Average daily profile by month
(July 7 - October 10) for PAN and ozone in
New Brunswick, NJ, 1979. Numbers refer to
months of the year.
Source: Lewis et al. (1983).
5-92
-------
5.6.6,2 Intraclty Variations. Few data on PAN concentrations at different
sites in the same city are available. One study is available for Houston,
Texas (Jorgen et a!., 1978), in which PAN was measured on October 26 and 27,
1977, at three separate sites, two in Houston and one north of the city.
Comparison of the levels of ozone and PAN peaks among the three sites on those
2 days reveals significant differences (Table 5-22). On October 26, the
highest ozone concentrations (two peaks of 110 and 95 ppb at ~3:00 p.m. and
~7:00 p.m., respectively) were seen at site 3, in southeast Houston; while the
highest PAN concentration on that day was a secondary peak of 15 ppb at ~8:00
p.m. at site 2 (mid-city). On October 27, the highest ozone concentration,
180 ppb, occurred at site I, about 4 miles north-northeast of Houston, at
~1:00 p.m. That was accompanied at the same time by the highest PAN concentra-
tion for the day, 16 ppb.
Examination of Table 5-22 shows the lack of a consistent, quantitative
relationship between PAN and ozone as indicated by intracity differences of
more than twofold in ozone concentrations (afternoon peaks, sites 1 and 2,
October 27) and differences of about fivefold in PAN concentrations between
sites (afternoon peaks, sites 1 and 2, October 27).
5.6.6.3 Indoor-Outdoor Ratios of PAN Concentrations. No recent studies
appear in the literature on indoor concentrations of PAN or indoor-outdoor
ratios (I/O). In three school buildings in southern California, Thompson et
al. (1973) found I/O ratios (expressed here as percentages) of 89, 97, and 148
percent, respectively, in the absence of air conditioning. With air condi-
tioning, the I/O ratios were 75, 108, and 117 percent, respectively. Total
oxidants were nearly constant all day, remaining about 30 percent (in air
conditioning) of the average outside concentration. The higher I/O gradients
for PAN than for oxidants were attributed by the authors to the greater break-
down of ozone ("oxidants") through its reaction with surfaces. The cooler
temperatures indoors are the probable cause of the greater persistence of PAN
indoors. Like ozone, PAN also decays indoors, but over an extended period
(Thompson et al., 1973).
5.7 CONCENTRATIONS OF OTHER PHOTOCHEMICAL OXIDANTS IN AMBIENT AIR
Concentrations of nitrogen dioxide (N02) and related nitrogenous oxidants
are presented in a recent criteria document on oxides of nitrogen (U.S. Environ-
mental Protection Agency, 1982b) and are not given here. In addition, the
5-93
-------
TABLE 5-22. INTRACITY VARIATIONS IN PEAK OZONE AND PAN CONCENTRATIONS IN
HOUSTON, OCTOBER 26 AND 27, 1977
1
2
Site
(~4 mi NNE
of Houston)
(mid- city)
Date
10/26
10/27
10/26
o8,
ppb
75
180
50
Time
1:
2:
2:
00
00
00-
of peak
p.
P-
3:
m.
m.
00 p.m.
PAN,
ppb
2
16
2
Time
9:00
p.m.
2:00
2:00-
a
of peak
.m. - 5:00
(plateau)
P
6
.m.
:00 p.m.
(plateau)
3
(SE Houston)
10/26
10/27
10/26
10/26
10/27
10/27
50
70
110
95
90
55
8:
1:
3:
7:
1:
7:
00
P-
00-3:
00
00
00
30
P.
P-
P-
P.
m.
00 p.m.
m.
m.
m.
m.
10
• 3 •
8
12
4
4
7:00
P
12:30-
2:00
5:30
11:30
6:30
P
P
P
.m.
3:00 p.m.
.m.
.m.
a.m.
.m.
Source: Jorgen et al. (1978).
comprehensive review by Altshuller (1983) also documents available information
on nitrogenous oxidants such as nitric acid (HNO-) and peroxynitric acid, as
well as on formic acid (HCOOH) and hydrogen peroxide (H202). The reader is
referred to these reviews for information on these oxidants. The two non-
nitrogenous oxidants, formic acid and hydrogen peroxide, are appropriate
concerns for this document, however, and the limited information on concentra-
tions of these two pollutants is summarized below.
Studies on the toxicologic effects of formic acid, though limited in
number, have shown only negligible effects, even in animals exposed to levels
as high as 20 ppm (aerosol vapor) (see Chapter 9). These levels are three
orders of magnitude greater than the concentrations seen in polluted urban
atmospheres. For example, maximum concentrations of HCOOH observed by Tuazon
and coworkers, using FTIR (Tuazon et al., 1978; 1980; 1981b), in Claremont and
Riverside, California, were in the range of 5 to 20 ppb in a study covering
5 years. The ranges of concentrations of HCOOH measured by Tuazon et al. were
consistent with those found in a long-path FTIR study by Hanst et al. (1982).
The FTIR method offers a reliable assessment of the ambient air concentrations
of HCOOH and reported concentrations are believed to be accurate.
5-94
-------
Data on HCOOH concentrations for 2 days in October 1978 are shown in
Figure 5-30 (Tuazon et a!., 1981b). The diurnal pattern is similar to that of
related oxidants and of some of the other smog products.
The measurement of hydrogen peroxide (H-O^) in ambient air is fraught
with difficulties that remain unresolved. Ozone itself is known to be an
interference in virtually all of the past and current measurement methods for
H?0? (Chapter 4). In his review of non-ozone oxidants and other smog consti-
tuents and photochemical products, Altshuller (1983) examined data obtained in
the South Coast Air Basin of California in the late 1970s and in 1980 for
possible consistency in the interference of ozone in H?0p measurements.
Laboratory experiments (Heikes et al. , 1982) have indicated that 1 ppb H?0?
would be generated per 100 ppb ozone. The analysis by Altshuller shows that
this relationship does not hold in ambient air in the South Coast Air Basin
once H^Op levels exceed about 5 to 10 ppb; and Altshuller (1983) concluded
that variations in l-LOj, measurements there remain unexplained.
Because of measurement problems, the true levels of hLO^ in ambient air
are unknown, especially in polluted areas, where multiple interferences may
possibly occur. Attempts to detect hydrogen peroxide by means of FTIR spectro-
scopy have all been negative, even in polluted areas. The method can measure
HLOp in ambient air with specificity at H^Qp levels > 40 ppb, which is the
limit of detection for an FTIR instrument with a 1-km pathlength (see Chapter
4>-
Notwithstanding measurement difficulties, some ranges of H?0p concentra-
tions at urban and nonurban sites have been reported in the literature. These
are given in Table 5-23, along with the general type of measurement method
used to obtain the reported concentrations. It must be kept in mind, however,
that the reported concentrations, though they represent state-of-the-art
measurements, are thought not to be accurate.
5.8 SUMMARY
In the context of this document, the concentrations of ozone and other
photochemical oxidants found in ambient air are important for: (1) assessing
potential exposures of human and other receptors; (2) determining the range of
ambient air concentrations of ozone and other photochemical oxidants relative
5-95
-------
CJ1
I
ID
CTt
E
a
a
•>
O
I
ui
U
z
O
O
LU
Z
O
N
O
-OCTOBER 12,1978
1000 1400 1800 2200
TIME OF DAY, POT
0200 0600 1000
OCTOBER 13,1978
1400 1800 2200
1000 1400 1800 2200
-OCTOBER 12,1978
0200 0600 1000
TIME OF DAY, PDT
1400 1800 2200
— OCTOBER 13,1978-
Figure 5-30. Diurnal profile of HCOOH, along with other oxidants and smog
constituents, on October 12 and 13,1978, at Claremont, CA.
a
a.
Z
O
Z
-------
TABLE 5-23. CONCENTRATIONS11 OF HYDROGEN PEROXIDE IN AMBIENT AIR
AT URBAN AND NONURBAN SITES
(ppb)
Location
Hoboken, NJ
(urban)
Riverside, CA
(urban)
Riverside and
Claremont, CA
(urban)
Minneapolis, MN
(urban)
Boulder, CO
(urban)
Boulder, CO
(nonurban,
east of
Boulder)
Tucson, AZ
(nonurban,
54 km SE
of Tucson)
Tucson, AZ
(remote, near
Tucson)
Concns. , and
Date comments
1970 < 40
1970 <180 (during
smog episode
with 650 ppb
oxidants)
July-Aug. 100 max. (ozone
1977 also 100 and
increasing);
10 to 50 on
most days
NAb <6
NA <0.5
Feb. 1978 0.2 to 3
NA <7
NA ~1
Method
Titanium (IV)
sill f ate/8-qui no-
linol
Titanium (IV)
s u 1 fate/8- qui no-
linol
Luminol
chemi 1 umi nescence
Wet chemical
Wet chemical0
Luminol
chemi 1 umi nescence
Luminol
chemi 1 umi nesence
Luminol
chemi 1 umi nescence
Reference
Bufalini et
al. (1972)
Bufalini et
al. (1972)
Kok et al.
(1978)
He ikes et al.
(1982)
Hei kes et al .
(1982)
Kelly et al .
(1979)
Farmer and
Dawson (1982)
Farmer and
Dawson (1982)
aNote: Since methods used to obtain these data are all subject to positive inter-
ference by ozone, data presented here are not reliable. They are included as a
summary of reported concentrations.
Not available.
cSee Chapter 4 for method used by Heikes et al. (1982).
Source: Derived from data in Altshuller (1983).
5-97
-------
to demonstrated "effects levels" (Chapters 6-12); (3) determining indoor-
outdoor gradients for exposure analyses; (4) assessing whether the concentra-
tions of oxidants other than ozone, singly, collectively, or in combination
with ozone, are cause for concern; and (5) evaluating the adequacy of ozone as
a control surrogate for other photochemical oxidants, if concentrations of the
other oxidants are cause for concern given the effects and the "effects levels"
of those oxidants.
5.8.1 Ozone Concentrations in Urban Areas
In Table 5-24, 1983 ozone concentrations for Standard Metropolitan Stat-
istical Areas (SMSAs) having populations >_ 1 million are given by geographic
area, demarcated according to United States Census divisions and regions (U.S.
Department of Commerce, 1982). The second-highest concentrations among daily
maximum 1-hour values measured in 1983 in the 38 SMSAs having populations of
at least 1 million ranged from 0.10 ppm in the Ft. Lauderdale, Florida; Phila-
delphia, Pennsylvania; and Seattle, Washington, areas to 0.37 ppm in the Los
Angeles-Long Beach, California, area. The second-highest value among daily
maximum 1-hour ozone concentrations for 35 of the 38 SMSAs in Table 5-24
equaled or exceeded 0.12 ppm. The data clearly show, as well, that the
highest 1-hour ozone concentrations in the United States occurred in the
northeast (New England and Middle Atlantic states), in the Gulf Coast area
(West South Central states), and on the west coast (Pacific states). Second-
highest daily maximum 1-hour concentrations in 1983 in the SMSAs, within each
of these three areas averaged 0.16, 0.17, and 0.21 ppm, respectively. It
should be emphasized that these three areas of the United States are subject
to those meteorological and climatological factors that are conducive to local
oxidant formation, or transport, or both. It should also be emphasized that 9
of the 16 SMSAs in the country with populations :> 2 million are located in
these areas.
Emissions of manmade oxidant precursors are usually correlated with
population, especially emissions from area source categories such as transpor-
tation and residential heating (Chapter 3). Accordingly, when grouped by
population, the 80 largest SMSAs had the following median values for their
collective second-highest daily maximum 1-hour ozone concentrations in 1983:
populations > 2 million, 0.17 ppm QS; populations of 1 to 2 million, 0.14 ppm
03; and populations of 0.5 to 1 million, 0,13 ppm Og. As noted above, however,
5-98
-------
TABLE 5-24. SECOND-HIGHEST OZONE CONCENTRATIONS AMONG DAILY MAXIMUM 1-hr
VALUES IN 1983 IN STANDARD METROPOLITAN STATISTICAL AREAS -WITH POPULATIONS
> 1 MILLION, GIVEN BY CENSUS DIVISIONS AND REGIONS3
Division
and region
SMSA
population,
SMSA millions
Second-highest
1983 03
concn. , ppm
Northeast
New England
Boston, MA
Middle Atlantic Buffalo, NY
. Nassau-Suffolk, NY
Newark, NJ
New York, NY/NJ
Philadelphia, PA/NJ
Pittsburgh, PA
South
. . >2
1 to <2
>2
1 to <2
>2
>2
>2
0.18
0.12
0.17
0.25
0.19
0.10
0.14
South Atlantic
South
West South
Central
North Central
East North
Central
West North
Central
Atlanta, GA >2
Baltimore, MD >2 :•
Ft. Lauderdale-Hollywood, FL 1 to <2
Miami, FL 1 to <2
Tampa-St. Petersburg, FL 1 to <2
Washington, DC/MD/VA >2
Dallas-Ft. Worth, TX >2
Houston, TX >2
New Orleans, LA 1 to <2
San Antonio, TX , , 1 to <2
Chicago, IL >2
Detroit, MI >2
Cleveland, OH 1 to <2
Cincinnati, OH/KY/IN 1 to <2
Milwaukee, WI 1 to <2
Indianapolis, IN 1 to <2
Columbus, OH 1 to <2
St. Louis, MO/IL
Minneapolis-St. Paul
Kansas City, MO/KS
MN/WI
>2
. >2
1 to <2
0.17
0,19
0.10
0.12
0.14
0.17
0.16
0.28
0.12
0.12
0.17
0.17
0.15
0.15
0.18
0.14
0.12
0.18
0.13
0.13
5-99
-------
TABLE 5-24 (cont'd). SECOND-HIGHEST OZONE CONCENTRATIONS AMONG DAILY MAXIMUM
1-hr VALUES IN 1983 IN STANDARD METROPOLITAN STATISTICAL AREAS
WITH > 1 MILLION, GIVEN BY CENSUS DIVISIONS AND REGIONS9
Division
and region
West
Mountain
Pacific
SMSA
Denver-Boulder, CO
Phoenix, AZ
Los Angeles- Long Beach, CA
San Francisco-Oakland, CA
Anaheim-Santa Ana-
Garden Grove, CA
San Diego, CA
Seattle- Everett, WA
Riverside-San Bernardino-
Ontario, CA
San Jose, CA
Portland, OR/WA
Sacramento, CA
SMSA Second-highest
population, 1983 03
millions concn. , ppm
1 to <2
1 to <2
>2
>2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
1 to <2
0.14
0.16
0.37
0.17
0.28
0.20
0.10
0.34
0.16
0.12
0.15
Standard Metropolitan Statistical Areas and geographic divisions and regions
as defined by Statistical Abstract of the United States (U.S. Department of
Commerce, 1982).
Source: U.S. Environmental Protection Agency (1984a).
coincident meteorology favorable for oxidant formation undoubtedly contributes
to the apparent correlation between population and ozone levels.
Among all stations reporting valid ozone data (:> 75 percent of possible
hourly values per year) in 1979, 1980, and 1981 (collectively, 906 station-
years) in the United States, the median second-highest 1-hour ozone value was
0.12 ppm, and 5 percent of the stations reported second-highest 1-hour values
> 0.28 ppm.
A pattern of concern in assessing responses to ozone in human populations
and in vegetation is the occurrence of repeated or prolonged multiday periods
when the ozone concentrations in ambient air are in the range of those known
to elicit responses (see Chapters 10 and 12). In addition, the number of days
of respite between such multiple-day periods of high ozone is of possible
consequence. Data show that repeated, consecutive-day exposures to or respites
5-100
-------
from specified concentrations are location-specific. At a site in Dallas,
Texas, for example, daily maximum 1-hour concentrations were >^ 0.06 ppm for
2 to 7 days in a row 37 times in a 3-year period (1979 through 1981). A con-
centration of >0.18 ppm was recorded at that site on only 2 single days,
however, and no multiple-day recurrences of that concentration or greater were
recorded over the 3-year period. At a site in Pasadena, California, daily
maximum 1-hour concentrations >_0.18 ppm recurred on 2 to 7 consecutive days
33 times in that same 3-year period (1979 through 1981) and occurred, as well,
on 21 separate days. These and other data demonstrate the occurrence in some
urban areas of multiple-day potential exposures to relatively high concentra-
tions of ozone.
5.8.2. Trends in Nationwide Ozone Concentrations
Trends in ozone concentrations nationwide are important for estimating
potential exposures in the future of human populations and other receptors, as
well as for examining the effectiveness of abatement programs. The determina-
tion of nationwide trends requires the application of statistical tests to
comparable, representative, multiyear aerometric data. The derivation of
recent trends in ozone concentrations and the interpretation of those trends
is complicated by two potentially significant factors that have affected
aerometric data since 1979: (1) the promulgation by EPA in 1979 of a new
calibration procedure for ozone monitoring (see Chapter 4); and (2) the intro-
duction by EPA of a quality-assurance program that has resulted.in improved
data-quality audits. The effects of these factors on ozone concentration
measurements are superimposed on the effects on concentrations of any changes
in meteorology or in precursor emission rates that may have occurred over the
same time span.
The nationwide trends in ozone concentrations for a 9-year period, 1975
through 1983, are shown in Figure 5-31 (U. S. Environmental Protection Agency,
1984a). The data given are trends as gauged by the composite average of the
second-highest value among daily maximum 1-hour ozone concentrations. Data
from four subsets of monitoring stations, most of them urban stations, are
given: (1) California stations only; (2) all stations except those in
California; (3) all stations including those in California; and (4) all
National Air Monitoring Stations (NAMS), which report data directly to EPA.
Only stations reporting > 75 percent of possible hourly values in the respective
years are represented in the data.
5-101
-------
0.18
0.17
I
Q.
*
z
o
§
QC
Ul
o
z
o
o
UJ
Z
o
N
O
0.16
0.15
0.14
0.13
0.12
CA (27 Stations) —
f
""""? -9
0 NAMS STATIONS (62)
T 95% CONFIDENCE
1 INTERVALS
O ALL STATIONS (176)
J 95% CONFIDENCE
JL INTERVALS
A CALIFORNIA STATIONS (27)
V ALL STATIONS EXCEPT
CALIFORNIA (149)
I I I I
V
I
1975 1976 1977 1978 1979 1980 1981 1982 1983
YEAR
Figure 5-31. National trend in composite average of the second highest value
among daily maximum 1 -hour ozone concentrations at selected groups of
sites, 1975 through 1983.
Source: U.S. Environmental Protection Agency (1984a).
5-102
-------
For the entire 9-year period, 1975 through 1983, all subsets of monitoring
stations show a decline in the composite second-highest daily maximum 1-hour
ozone concentration. Between 1979, when the new, more accurate calibration
procedure was promulgated, and 1982, a small decline of 9 to 10 percent in
nationwide ozone concentrations occurred. From 1982 to 1983, however, concen-
trations increased by about 10 percent in California, by about 14 percent
outside California, and by about 12 percent nationwide (all states). Recently
published data for 1984 from a somewhat smaller number of sites (163) (U.S.
Environmental, Protection Agency, 1986) show a decrease in nationwide ozone
concentrations from 1983 to 1984, with 1984 levels approximating those recorded
in 1981. The portion of the apparent decline in ozone nationwide from 1975
through 1984 that is attributable to the calibration change of 1979 cannot be
determined by simply applying a correction factor to all data, since not all
monitoring stations began using the UV calibration procedure in the same year.
Figure 5-32 shows the nationwide frequency distributions of the first-,
second-, and third-highest 1-hour 0,, concentrations at predominantly urban
stations aggregated for 1979, 1980, and 1981, as well as the highest 1-hour 0.,
concentration at site of the National Air Pollution Background Network (NAPBN)
aggregated for the same 3 years. As shown by Figure 5-32, 50 percent of the
second-highest 1-hour values from non-NAPBN sites in this 3-year period were
0.12 ppm or less and about 10 percent were equal to or greater than 0.20 ppm.
At the NAPBN sites, the collective 3-year distribution (1979 through 1981) is
such that about 6 percent of the values are less than 0.10 ppm and fewer than
20 percent are higher than 0,12 ppm.
5.8.3. Ozone Concentrations in Nonurban Areas
Few nonurban areas have been routinely monitored for ozone concentrations.
Consequently, the aerometric data base for nonurban areas is considerably less
substantial than for urban areas. Data are available, however, from two
special-purpose networks, the National Air Pollution Background Network (NAPBN)
and the Sulfate Regional Experiment network (SURE). Data on maximum 1-hour
concentrations and arithmetic mean 1-hour concentrations reveal that maximum
1-hour concentrations at nonurban sites classified as rural (SURE study,
Martinez and Singh, 1979; NAPBN studies, Evans et a!., 1983) can sometimes
exceed the concentrations observed at sites classified as suburban (SURE
study, Martinez and Singh, 1979). For example, maximum 1-hour ozone concentra-
tions measured in 1980 at Kisatchie National Forest (NF), Louisiana; Custer
5-103
-------
in
i
^>
99.99
0.4S
0.40
0,35
0.30
99.9 99.8
99 38 96 90
80 70 60 60 40 30 20
10
1 0,5 0.2 0,1 0.05 0.01
< 0.25
ui
O
§ 0.20
O
ui
O 0.15
O
0.10
0.05
II I I I I I I I I I II
HIGHEST
2nd-HIGHEST
•• 3rd-HlGHEST
HIGHEST, NAPBN SITES
I I I I I I I
I I
0.01 0.06 0.1 0.2 0.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.8 99.9 99.99
STATIONS WITH PEAK 1-hour CONCENTRATIONS < SELECTED VALUE, percent
Figure 5-32, Distributions of the three highest 1 -hour ozone concentrations at valid sites (906
station-years) aggregated for 3 years (1979,1980, and 1981) and the highest ozone
concentrations at NAPBN sites aggregated for those years (24 station-years).
Source: U.S. Environmental Protection Agency (1980,1981,1982).
-------
NF, Montana; and Green Mt. NF, Vermont, were 0.105, 0.070, and 0.115 ppm,
respectively. Arithmetic mean 1-hour concentrations for 1980 were 0.028,
0.037, and 0.032 ppm at the respective sites. For four nonurban (rural) sites
in the SURE study, maximum 1-hour ozone concentrations were 0.106, 0.107,
0.117, and 0.153; and mean 1-hour concentrations ranged from 0.021 to 0.035
ppm. At the five nonurban (suburban) sites of the SURE study, maximum concen-
trations were 0.077, 0.099, 0.099, 0.080, and 0.118 ppm, respectively. The
mean 1-hour concentrations at those sites were 0.023, 0.030, 0.025, 0.020, and
0.025 ppm, respectively. ;
Ranges of concentrations and the maximum 1-hour concentrations at some of
the NAPBN and SURE sites show the probable influence of ozone transported from
urban areas. In one documented case, for example, a 1-hour peak ozone concen-
tration of 0.125 ppm at an NAPBN site in Mark Twain National Forest, Missouri,
was measured during passage of an air mass whose trajectory was calculated to
have included Detroit, Cincinnati, and Louisville in the preceding hours
(Evans et al., 1983).
The second-highest concentration among all the daily maximum 1-hour
concentrations measured at the NAPBN sites appear to be about one-half the
corresponding concentrations measured at urban sites in the same years.' No
trend in concentrations at these NAPBN sites is discernible in the data record
for 1979 through 1983.
These data corroborate the conclusion given in the 1978 criteria document
(U.S. Environmental Protection Agency, 1978) regarding urban-nonurban and
urban-suburban gradients; i.e., nonurban areas may sometimes sustain higher
peak ozone concentrations than those found in urban areas.
5.8.4. Diurnal and Seasonal Patterns in Ozone Concentrations
Since the photochemical reactions of precursors that result in ozone for-
mation are driven by sunlight, as well as by emissions, the patterns of ozone
occurrence in ambient air depend on daily and seasonal variations in sunlight
intensity. The typical diurnal pattern of ozone in ambient air has a minimum
ozone level around sunrise (near zero in most urban areas), increasing through
the morning to a peak concentration in early afternoon, and decreasing toward
minimal levels again in the evening. The 1978 criteria document ascribed.th§
daily ozone pattern to three simultaneous processes: (1) downward transport of
ozone from layers aloft; (2) destruction of ozone through contact with surfaces
and through reaction with nitric oxide (NO) at ground level; and (3) in situ
5-105
-------
photochemical production of ozone (U.S. Environmental Protection Agency, 1978;
Coffey et a!., 1977; Mohnen, 1977; Reiter, 1977a). Obviously, meteorology is
a controlling factor; if strong winds disperse the precursors or heavy clouds
intercept the sunlight, high ozone levels will not develop. Another important
variation on the basic diurnal pattern appears in some localities as a secondary
peak in addition to the early afternoon peak. This secondary peak may occur
any time from midafternoon to the middle of the night and is attributed to
ozone transported from upwind areas where high ozone levels have occurred
earlier in the day. Secondary peak concentrations can be higher than concen-
trations resulting from the photochemical reactions of locally emitted precursors
(Martinez and Singh, 1979). At one nonurban site in Massachusetts (August
1977), for example, primary peak concentrations of about 0.11, 0.14, and 0.14
occurred at noon, from noon to about 4:00 p.m., and at noon, respectively, on
3 successive days of high ozone levels (Martinez and Singh, 1979). Secondary
peaks at the same site for those same 3 days were about 0.150, 0.157, and
0.130 ppm at about 6:00 p.m., 8:00 p.m., and 8:00 p.m., respectively (Martinez
and Singh, 1979).
Because weather patterns, ambient temperatures, and the intensity and
wavelengths of sunlight all play important roles in oxidant formation, strong
seasonal as well as diurnal patterns exist. The highest ozone levels generally
occur in the spring and summer (second and third quarters), when sunlight
reaching the lower troposphere is most intense and stagnant meteorological
conditions augment the potential for ozone formation and accumulation. Average
summer afternoon levels can be from 150 to 250 percent of the average winter
afternoon levels. Minor variations in the smog season occur with location,
however. In addition, it is possible for the maximum and second-highest
1-hour ozone concentration to occur outside the two quarters of highest .average
ozone concentrations. Exceptions to seasonal patterns are potentially important
considerations with regard to the protection of crops from ozone damage,
especially since respective crops have different growing seasons in terms of
length, time of year, and areas of the country in which they are grown.
In addition to the seasonal meteorological conditions that obtain in the
lower troposphere, stratospheric-tropospheric exchange mechanisms exist that
produce relatively frequent but sporadic, short-term incursions into the
troposphere of stratospheric ozone (see Chapter 3). Such incursions show a
seasonal pattern, usually occurring in late winter or early spring.
5-106
-------
Percentile distributions, by season of the year, of concentration data
from all eight NAPBN sites show that the arithmetic mean 1-hour concentration
(averaged over a minimum of 3 years of data at each site, for 1977 through
1983) was higher in the second quarter of the year (April, May, June) at seven
of the eight stations; and was only negligibly lower than the third-quarter
value at the eighth station. The maximum 1-hour concentrations at respective
sites, aggregated over 3 to 6 years, depending on the data record for each
site, ranged from 0.050 ppm at Custer NF, MT (in the fourth quarter) to
0.155 ppm at Mark Twain NF, MO (in the third quarter). The second-highest
1-hour concentration among maximum daily 1-hour values ranged from 0.050 ppm
at Custer NF, MT (in the fourth quarter), to 0.150 ppm at Mark Twain NF, MO
(in the third quarter). The data also show that 99 percent of the 1-hour
concentrations measured were well below 0.12 ppm, even in the second quarter
of the year, when incursions of stratospheric ozone are expected to affect,
at least to some degree, the concentrations measured at these stations.
Excursions above 0.12 ppm were recorded in 1979 and 1980 at NAPBN sites; but
none were recorded in 1981 (Evans et a!., 1983; Lefohn, 1984).
Because of the diurnal patterns of ozone, averaging across longer-term
periods such as a month, a season, or longer masks the occurrence of peak.
concentrations (see, e.g., Lefohn and Benedict, 1985). This is an obvious and
familiar statistical phenomenon. It is pointed out, however, because it has
direct relevance to the protection of public health and welfare. Averaging
times must correspond to, or be related in a consistent manner to, the pattern
of exposure that elicits untoward responses. • . •
5.8.5 Spatial Patterns in Ozone Concentrations ,>
In addition to temporal variations, both macro- and microscale spatial
variations in ozone concentrations occur that have relevance ranging from
important to inconsequential for exposure assessments. Differences in concen-
trations or patterns of occurrence, or both, are known to exist, for example,
between urban and nonurban areas, between indoor and outdoor air, within large
metropolises, and between lower and higher elevations. The more important
variations are summarized below. t
5.8.5.1 Urban-Nonurban Differences in Ozone Concentrations. Ozone concentra-
tions differ between urban and rural, between urban and remote, and even
between rural and remote sites, as discussed in part in the preceding section
5-107
-------
on temporal variations. The variations with area and type of site are varia-
tions both in the timing and the magnitude of the peak concentrations, and, in
the case of transported ozone, are related to the temporal variations between
urban and nonurban areas discussed above. Data from urban, suburban, rural,
and remote sites (see, e.g., SAROAD, 1985a-f; Martinez and Singh, 1979; Lefohn,
1984; Evans, 1985; respectively) corroborate the conclusion drawn in the 1978
criteria document (U.S. Environmental Protection Agency, 1978) that ozone
concentrations can sometimes be higher in some suburban or even rural areas
downwind of urban plumes than in the urban areas themselves; and, furthermore,
that higher concentrations can persist longer in rural and remote areas,
largely because of the absence of nitric oxide (NO) for chemical scavenging.
In nonurban areas downwind of urban plumes, peak concentrations can
occur, as the result of transport, at virtually any hour of the day or night,
depending upon many factors, such as the strength of the emission source,
induction time, scavenging, and wind speed (travel time) and other meteorological
factors. The dependence of the timing of peak exposures upon these transport-
related factors is well-known and is illustrated here by two studies. Evans
et al. (1983) calculated multiday trajectories for air parcels reaching a
nonurban sites in the Mark Twain National Forest, Missouri, during an episode.
Four separate trajectories, all of which passed over the Ohio River Valley and
the Great Lakes region, impacted the forest site at different times in a
24-hour period (in which the maximum 1-hour concentration measured was 0.125
ppm). Subsequently, regional cloud cover and rains produced shifts in air
flow and also reduced the potential for ozone formation, alleviating the
impact at the site. Kelly et al. (1986) showed in the Detroit area that peak
ozone concentrations occurred at distances of 10 to 70 km (ca. 6 to 43 mi)
north-northeast of the urban center. Consequently, it would be possible for
peak ozone concentrations to occur in the late afternoon or early evening in
nonurban areas downwind of Detroit. Kelly et al. (1986) also found that
concentrations diminished again beyond 70 km (ca. 43 mi) downwind of the urban
center. Thus, as illustrated by these and similar data, beyond the distance
traversed in the time required for maximum ozone formation in an urban plume,
ozone concentrations will decrease (unless fresh emissions are injected into
the plume) as the rate of ozone formation decreases, the plume disperses,
surface deposition or other scavenging occurs, and meteorological conditions
intervene.
5-108
-------
It is not surprising, therefore, that in rural areas lying beyond the
point of maximum ozone formation, for a given set of conditions, peak concen-
trations are lower and average diurnal profiles are flatter than in urban and
near-urban areas (see, e.g., SAROAD, 1985b-f, for rural and remote sites). In
remote areas beyond the influence of urban plumes, average peak concentrations
will be still lower and average diurnal profiles still flatter (see e.g.,
Evans, 1985). Exceptions to these generalizations occur, of course, because
of the complex interactions of topography, meteorology, and photochemistry.
Such temporal and spatial differences between ozone concentrations in
urban versus nonurban areas are important considerations for accurately assessing
actual and potential exposures for human populations and for vegetation in
nonurban areas, especially since the aerometric data for nonurban areas are
far from abundant.
5.8.5.2 Geographic, Vertical, and Altitudinal Variations in Ozone Concentrations.
Although of interest and concern when estimating global ozone budgets, demon-
strated variations in ozone concentrations with latitude and the lesser variations
with longitude have little practical significance for assessing exposure
within the contiguous United States. The effects on ozone concentrations of
latitude and longitude within the contiguous states are minor, and are reflected
in the aerometric data bases. Of more importance, ozone concentrations are
known to increase with increasing height above the surface of the earth.
Conversely, they may be viewed as decreasing with proximity to the surface of
the earth, since the earth's surface acts as a sink for ozone (see, e.g.,
Seller and Fishman, 1981; Galbally and Roy, 1980; Oltmans, 1981, cited in
Logan et al.5 1981). The most pertinent vertical and altitudinal gradients in
ozone concentrations are: (1) increases in concentration with height above
the surface of the ground (regardless of altitude); (2) increases in concentra-
tion with altitude; and (3) variations in concentrations with elevation in
mountainous areas, attributable to transport and overnight conservation of
ozone aloft, nocturnal inversions, trapping inversions, upslope flows, and
other, often location-specific interactions between topography, meteorology,
*
and photochemistry.
5-109
-------
The importance of monitoring concentrations at the proper height.above
the surface of the ground has been known for a long time, and EPA guidance on
the placement of monitoring instruments, (see Chapter 4) is predicated on the
existence of a vertical gradient as ozone is depleted by reaction with ground-
level emissions of NO or by deposition on reactive surfaces such as vegetation.
Data illustrative of the near-surface gradient were reported by Pratt et al.
(1983), who measured ozone concentrations at two separate heights (3 and ,9 or
6 and 9 meters) above the ground at three rural,.vegetated sites. Although
the maximum mean difference between 3 and 9 meters was 3 ppb, this difference
was similar to the mean difference between sites at the same height. Given
the height of some vegetation canopies, especially forests, even such small
differences over a spread of 6 meters should probably be taken into considera-
tion when interpreting reported dose-response functions.
The gradual increase in ozone concentrations with altitude has been
documented by a number of workers (see e.g., Viezee et al., 1979; Seiler and
Fishman, 1981; Oltmans, 1981, as cited in Logan et al., 1981). There is a
general increase in concentration with altitude, but as described by Seiler
and Fishman (1981) and Oltmans (1981; cited in Logan et al., 1981), for example,
two relatively pronounced gradients, exist, one between the surface of the
earth and 2 km (ca. 1 mi) and one even more pronounced between 8 and 12 km
(ca. 5 and 7.5 mi).
Increases in concentrations with altitude could potentially be of some
consequence for passengers and airline personnel on high-altitude flights in
the absence of adequate ventilation-filtration systems (see Chapter 11).
Variations with height above the surface and with elevation, in mountainous
areas, however, should be taken into account to ensure the accurate assessment
of exposures and the accurate derivation of dose-response functions, especially
for forests and other vegetation.
Variations in ozone concentrations with elevation, not always consistent
or predictable, have been reported by researchers investigating the effects of
ozone on .the mixed-conifer forest ecosystem of the San Bernardino Mountains of
*
California. Measurements taken at four monitoring stations at four different
elevations showed that peak ozone concentrations occurred progressively later
in the day at progressively higher elevations (Miller et al., 1972). Ozone
concentrations >0.10 ppm occurred for average durations of 9, 13, 9, and
5-110
-------
8 hr/day at the four respective stations, going from lower to higher elevations.
The occurrence for 13 hr/day of concentrations >0.10, ppm at the station at
817 m (2860 ft) was probably the result of contact of that zone of the mountain-
side with the inversion layer (U.S. Environmental Protection Agency, 1978).
Nighttime concentrations rarely decreased below 0.05 ppm at the mountain crest;
whereas at the lowest elevation, the basin station at 442 m (1459 ft), the
nighttime concentration usually decayed to near zero. Trapping inversions
were major.contributors to the elevational gradients observed i,n this study,
which was conducted in the 1970s. Oxidant concentrations within the inversion
were found not to be uniform but to occur in multiple layers and strong vertical
gradients. The important result of the trapping of oxidants in the inversion
layers was the prolonged contact of high terrain with oxidants at night (U.S.
Environmental Protection Agency, 1978).
In a more recent report, Wolff et al. (1986) described measurements made
in July 1975 at three separate elevations at High Point Mountain in northeastern
New Jersey. The daily ozone maxima were similar at different elevations. At
night, however, ozone concentrations were nearly zero in the valley but increased
with elevation on the mountainside. Greater cumulative doses .(number of hours
at >0.08 ppm) were sustained at the higher elevations, 300 and 500 m, respec-
tively (ca. 990 and 1650 ft, respectively). Wolff et al. (1986) related this
phenomenon to the depth of the nocturnal inversion layer and the contact with
the mountainside of ozone conserved aloft at night. -.••.-
These concentration gradients with increased elevation are important for
accurately describing concentrations at which injury or damage to vegetation,
especially forests, may occur. Researchers investigating the effects of ozone
on forest ecosystems have seldom measured nighttime ozone concentrations
because the stomates of most species are thought to be closed at night, thus
preventing the internal flux of ozone that produces injury or damage (see
Chapter 6). If stomates remain even partially open at night, however, the
possible occurrence of nighttime peak concentrations of ozone, the occurrence
of multiple peaks in a 24-hour period, or the persistence of elevated concen-
trations that do not decay to near zero overnight should not be overlooked.
Furthermore, the lack of NO for nighttime scavenging in nonurban areas and the
•persistence of ozone overnight at higher elevations will result in the presence
of relatively higher concentrations in such areas at sunrise when the stomates
5-111
-------
open and photosynthesis begins. This possibility requires that exposure
assessments, in the absence of sufficient aerometric data for forests and
other vegetated areas, take such factors into consideration.
5.8.5.3 Other SpatialVariations in Ozone Concentrations. Other spatial varia-
tions are important for exposure assessments for human populations. Indoor-
outdoor gradients in ozone concentrations are known to occur even in buildings
or vehicles ventilated by fresh air rather than air conditioning (e.g., Sabersky
et al., 1973; Thompson et a!., 1973; Peterson and Sabersky, 1975). Ozone reacts
with surfaces inside buildings, so that decay may occur fairly rapidly, depending
upon the nature of interior surfaces and furnishings (e.g., Davies et al., 1984;
Contant et al., 1985). Ratios of indoor-to-outdoor (I/O) ozone concentrations
are quite variable, however, since cooling and ventilation systems, air infil-
tration or exchange rates, interior air circulation rates, and the composition
of interior surfaces all affect indoor ozone concentrations. Ratios (I/O,
expressed as percentages) in the literature thus vary from 100 percent in a
non-air-conditioned residence (Contant et al., 1985); to 80 ± 10 percent
(Sabersky et al., 1973) in an air-conditioned office building (but with 100 per-
cent outside air intake); to 10 to 25 percent in air-conditioned residences
(Berk et al., 1981); and to as low as near zero in air-conditioned residences
(Stock et al., 1983; Contant et al., 1985).
On a larger scale, within-city variations in ozone concentrations can
occur, even though ozone is a "regional" pollutant. Data show, for example,
relatively homogeneous ozone concentrations in New Haven, Connecticut (SAROAD,
1985a), a moderately large city that is downwind of a reasonably well-mixed
urban plume (Wolff et al., 1975; Cleveland et al.; 1976a,b). In a large
metropolis, however, appreciable gradients in ozone concentrations can exist
from one side of the city to the other, as demonstrated for New York City
(Smith, 1981), and for Detroit (Kelly et al., 1986). Such gradients should be
taken into consideration, where possible, in exposure assessments.
5.8.6 Concentrations and Patterns of Other Photochemical Oxidants
5.8.6.1 Concentrations. No aerometric data are routinely obtained by Federal,
state, or local air pollution agencies for any photochemical oxidants other
than nitrogen dioxide and ozone. The concentrations presented in this document
for non-ozone oxidants were all obtained in special field investigations. The
limitations in the number of locations and areas of the country represented in
5-112
-------
the Information presented simply reflect the relative paucity of data in the
published literature.
The four non-ozone photochemical oxidants for which at least minimal
concentration data are available are formic acid, peroxyacetyl nitrate (PAN),
peroxypropionyl nitrate (PPN), and hydrogen peroxide (H~0p). Peroxybenzoyl
nitrate has not been clearly identified in ambient air in the United States.
The highest concentrations of PAN reported in the older literature, 1960
through the present, were those found in the Los Angeles area: 70 ppb (1960),
214 ppb (1965); and 68 ppb (1968) (Renzetti and Bryan, 1961; Mayrsohn and
Brooks, 1965; Lonneman et al., 1976; respectively).
The highest concentrations of PAN measured and reported in urban areas in
the past 5 years were 42 ppb at Riverside, California, in 1980 (Temple and
Taylor, 1983) and 47 ppb at Claremont, California, also in 1980 (Grosjean
1981). These are clearly the maximum concentrations of PAN reported for
California and for the entire country in this period. Other maximum PAN
concentrations measured in the last decade in the Los Angeles Basin have been
in the range of 11 to 37 ppb. Average concentrations of PAN in the Los Angeles
Basin in the past 5 years have ranged from 4 to 13 ppb (Tuazon et a!,, 1981a;
Grosjean, 1983; respectively). The only published study covering urban PAN
concentrations outside California in the past 5 years is that of Lewis et al.
(1983) for New Brunswick, New Jersey, in which the average PAN concentration
was 0.5 ppb and the maximum was 11 ppb during September 1978 through May 1980.
Studies outside California from the early 1970s through 1978 showed average
PAN concentrations ranging from 0.4 ppb in Houston, Texas, in 1976 (Westberg
et al., 1978) to 6.3 ppb in St. Louis, Missouri, in 1973 (Lonneman et al.,
1976). Maximum PAN concentrations outside California for the same period
ranged from 10 ppb in Dayton, Ohio, in 1974 (Spicer et al., 1976) to 25 ppb in
St. Louis (Lonneman et al., 1976).
The highest PPN concentration reported in studies over the period 1963
through the present was 6 ppb in Riverside, California (Darley et al., 1963).
The next highest reported PPN concentration was 5 ppb at St. Louis, Missouri,
in 1973 (Lonneman et al., 1976). Among more recent data, maximum PPN concentra-
tions at respective sites ranged from 0.07 ppb in Pittsburgh, Pennsylvania, in
1981 (Singh et al., 1982) to 3.1 ppb at Staten Island, New York (Singh et al.,
1982). California concentrations fell within this range. Average PPN concentra-
tions at the respective sites for the more recent data ranged from 0.05 ppb at
Denver and Pittsburgh to 0.7 ppb at Los Angeles in 1979 (Singh et al., 1981).
5-113
-------
Altshuller (1983) has succinctly summarized the nonurban concentrations
of PAN and PPN by pointing out that they overlap the lower end of the range of
urban concentrations at sites outside California, At remote locations, PAN
and PPN concentrations are lower than even the lowest of the urban concentra-
tions by a factor of 3 to 4.
The concentrations of FLC^ reported in, the literature to date must be
regarded as inaccurate since ozone is now thought to be an interference in all
methods used to date except FTIR (Chapter 4). Measurements by FTIR, the most
specific and accurate method now available, have not demonstrated unambiguously
the presence of HpQ^ in ambient air, even in the high-oxidant atmosphere of the
Los Angeles area. (The limit of detection for a 1-km-pathlength FTIR system
is around 0.04 ppm.)
Recent data indicate the presence in urban atmospheres of only trace
amounts of formic acid: < 15 ppb, measured by FTIR (Tuazon et al., 1981b).
Estimates in the earlier literature (1950s) of 600 to 700 ppb of formic acid
in smoggy atmospheres were erroneous because of faulty measurement methodology
(Hanst et a!., 1982).
5.8.6.2 Patterns. The patterns of formic acid (HCOOH), PAN, PPN, and H£02
may be summarized fairly succinctly. Qualitatively, diurnal patterns are
similar to those of ozone, with peak concentrations of each of these occurring
in close proximity to the time of the ozone peak. The correspondence in time
of day is not exact, but is close. As demonstrated by the work of Tuazon
et al. (1981b), ozone concentrations return to baseline levels somewhat faster
than the concentrations of PAN, HCOOH, or H202 (PPN was not measured).
Seasonally, winter concentrations (third and fourth quarters) of PAN are
lower than summer'concentrations (second and third quarters). The percentage
of PAN concentrations (PAN/03 x 100) relative to ozone, however, is higher in
winter than in summer. Data are not readily available on the seasonal patterns
of the other non-ozone oxidants.
Indoor-outdoor data on PAN are limited to one report (Thompson et al.,
1973), which confirms the pattern to be expected from the known chemistry of
PAN; that is, it persists longer indoors than ozone. Data are lacking on
indoor-outdoor ratios for the other non-ozone oxidants.
5.8.7 Relationship Between Ozone and Other Photochemical Oxidants
The relationship between ozone concentrations and the concentrations of
PAN, PPN, HpOp, and HCOOH is important only if these non-ozone oxidants are
5-114
-------
shown to produce potentially adverse health or welfare effects, singly, in
combination with each other, or in various combinations with ozone at concentra-
tions correponding to those found in ambient air. If only ozone is shown to
produce adverse health or welfare effects in the concentration ranges of
concern, then only ozone must be controlled. If any or all of these other
four oxidants are shown to produce potentially adverse health or welfare
effects, at or near levels found in ambient air, then such oxidants will also
have to be controlled. Since ozone and all four of the other oxidants arise
from reactions among the same organic and inorganic precursors, an obvious
question is whether the control of ozone will also result in the control of
the other four oxidants.
Controlled-exposure studies on these non-ozone oxidants have employed
concentrations much higher than those found in ambient air (see Chapters 9 and
10). Because PAN may have contributed, however, to the eye irritation symptoms
reported in earlier epidemiologies! studiesj and because PAN is the most
abundant of these non-ozone oxidants, the relationship between ozone and PAN
concentrations in ambient air remains of interest.
•The patterns of PAN and ozone concentrations are not quantitatively
similar but do show qualitative similarities for most locations at which both
pollutants have been measured in the same study. That a quantitative, monotonic
relationship between ozone and PAN is lacking, however, is shown by the range
of PAN-to-ozone ratios, expressed as'percentages, between locations and at the
same location, as reported in the review of Altshuller (1983);
Certain other information bears out the lack of a monotonic relationship
between PAN and ozone. Not only are PAN-ozone relationships not consistent
between different urban areas, but they are not consistent in urban versus
nonurban areas, in summer versus winter, in indoor versus outdoor environments,
or even, as the data show, in location, timing, or magnitude of diurnal peak
concentrations within the same city. Data obtained in Houston by Jorgen
et aV. (1978), for example, show variations in peak concentrations of PAN and
in relationships to ozone concentrations of those peaks among three separate
monitoring sites. Temple and Taylor (1983) have shown that PAN concentrations
are a greater percentage of ozone concentrations in winter than in the.remainder
of the year in California. Lonneman et al. (1976) demonstrated that PAN,
absolutely and as a percentage of ozone, is considerably lower in nonurban
than in urban areas. Thompson et al. (1973), in what is apparently the only
5-115
-------
published report on indoor concentrations of PAN, showed that PAN persists
longer than ozone indoors. (This is to be expected from its enhanced stability
at cooler-than-ambient temperatures such as found in air-conditioned.buildings.)
Tuazon et al. (1981b) demonstrated that PAN persists in ambient air longer
than ozone, its persistence paralleling that of nitric acid, at least in the
locality studied (Claremont, CA). Reactivity data presented in the 1978
criteria document for ozone and other photochemical oxidants indicated that
all precursors that give rise to PAN also give rise to ozone. The data also
showed, however, that not all precursors giving rise to ozone also give rise
to PAN, and that not all that give rise to both are equally reactive toward
both, with some precursors preferentially giving rise, on the basis of units
of product per unit of reactant, to more of one product than the other (U.S.
Environmental Protection Agency, 1978).
In the review cited earlier, Altshuller (1983) examined the relationships
between ozone and a variety of other smog components, including PAN, PPN,
HpOpj HCOOH, aldehydes, aerosols, and nitric acid. He concluded that "the
ambient air measurements indicate that ozone may serve directionally, but
cannot be expected to serve quantitatively, as a surrogate for the other
products" (Altshuller, 1983). It must be emphasized that the issue Altshuller
examined was whether ozone could serve as an abatement surrogate for all
photochemical products, not just the subset of non-ozone oxidants of concern
in this document. Nevertheless, a review of the data presented indicates that
his conclusion is applicable to the non-ozone oxidants examined in this docu-
ment.
5.9 REFERENCES
Altshuller, A. P. (1975) Evaluation of oxidant results at CAMP sites in the
United States. J. Air Pollut. Control Assoc. 25: 19-24.
Altshuller, A. P. (1983) Measurements of the products of atmospheric photo-
chemical reactions in laboratory studies and in ambient air—relation-
ships between ozone and other products. Atmos. Environ. 17: 2383-2427.
Berk, J. V.; Young, R.; Hollowell, C. D.; et al. (1980) The effects of energy-
efficient ventilation rates on indoor air quality in an Ohio elementary
school. Berkeley, CA: University of California, Lawrence Berkeley
Laboratory; report no. EEB-Vent 80-9. Available from: NTIS, Springfield,
VA; LBL-10223.
5-116
-------
Berk, J. V,; Young, R. A.; Brown, S. R.; Hollowell, C. D. (1981) Impact of
energy-conserving retrofits on indoor air quality in residential housing.
Presented at: 74th annual meeting of the Air Pollution Control Assoc-
ciation; June; Philadelphia, PA. Pittsburgh, PA: Air Pollution Control
Association; paper no. 81-22.1
Berry, C. R. (1964) Differences in concentrations of surface oxidant between
valley and mountaintop conditions in the southern Appalachians. J. Air.
Pollut. Control Assoc. 14:238-239.
Blumenthal, D. L,; White, W. H.; Peace, R. L ; Smith T. B. (1974) Determination
of the feasibility of the long-range transport of ozone or ozone precursors.
EPA-450/3-74-061, U.S. Environmental Protection Agency, Research Triangle
Park, NC.
Brennan, E. (1980) PAN concentrations in ambient air in New Brunswick, NJ.
Final report. Prepared under grant 805827 for U.S. Environmental Protec-
tion Agency, Research Triangle Park, NC.
Bufalini, J. J.; Gay, B. W., Jr.; Brubaker, K. L. (1972) Hydrogen peroxide
formation from formaldehyde photooxidation and its presence in urban
atmospheres. Environ. Sci. Techno!. 6: 816-821.
Cherniak, E. A.; Corkum, R. (1981) Continuous monitoring of PAN in the polluted
troposphere. Brock University, St. Catherines, ON, Canada. Cited in Temple,
P. J.; Taylor, O.C. (1983).
Cleveland, W. S.; Graedel, T. E.; Kleiner, B. (1976a) Photochemical air pollu-
tion: transport from the New York City area into Connecticut and Massa-
chusetts. Science (Washington, DC) 191: 179-181.
Cleveland, W. S.; Guarino, R.; Kleiner, B.; McRae, J. E,; Warner, J. L. (1976b)
The analysis of the ozone problem in the northeast United States. In:
Specialty Conference on: Ozone/oxidants—interactions with the total
environment. Pittsburgh, PA: Air Pollution Control Association; pp.
109-120.
Coffey, P. E.; Stasiuk, W. N. (1975a) Evidence of atmospheric transport of
ozone into urban areas. Environ. Sci. Technol. 9: 59-62.
Coffey, P. E.; Stasiuk, W. N. (1975b) Urban ozone: its local and extraregional
components. Presented at the 79th annual national meeting, American
Institute of Chemical Engineers; March; Houston, TX. Paper no. 60C.
Coffey, P. E.; Stasiuk, W. N.; Mohnen, V. (1977) Ozone in rural and urban
areas of New York State. In: Dimitriades, B., ed. International confer-
ence on photochemical oxidant pollution and its control. Proceedings: v.
I. Research Triangle Park, NC: U.S. Environmental Protection Agency,
Office of Research and Development; pp. 89-96; EPA report no. EPA-600/
3-77-OOla. Available from: NTIS, Springfield, VA; PB-624232.
5-117
-------
Content, C. F.; Gehan, B. M,; Stock, T. H.; Holguin, A. H.; Buffler, P. A.
(1985) Estimation of Individual ozone exposures using microenvironment
measurements. In: APCA Specialty Conference: Scientific Basis for the
Ozone Standard, Houston, TX, November 28-December 1, 1984. Pittsburgh;
APCA. In Press.
Cox, R. A.; Roffey, M. J. (1977) Thermal decomposition of peroxyacetyl nitrate
in the presence of nitric oxide. Environ. Sci. Techno!. 11: 900-906.
Parley, E. F.; Kettner, K. A.; Stephens, E. R. (1963) Analysis of peroxyacyl
nitrates by gas chromatography with electron capture detection. Anal.
Chem. 35: 589-591.
Davies, T. D.; Ramer, B.; Kaspyzok, G.; Delany, A. C. (1984) Indoor/outdoor
ozone concentrations at a contemporary art gallery. J. Air Pollut. Control
Assoc. 31: 135-137.
Edinger, J, G. (1973) Vertical distribution of photochemical smog in Los Angeles
basin. Environ. Sci. Techno!. 7: 247-252.
Edinger, J. G.; McCutchan, M. H.; Miller, P. R.; Ryan, B. C.; Schroeder, M. J.;
Behar, J. V. (1972) Penetration and duration of oxidant air pollution in .
the south coast air basin of California. J. Air Pollut. Control Assoc.
22: 882-886.
Elkus, B.; Wilson, K. R. (1976) Photochemical oxidant: weekday-weekend differences.
Draft report. Cited in U. S. Environmental Protection Agency (1978).
Evans, G. F. (1985) The National Air Pollution Background Network: final
project report. Research Triangle Park, NC: U. S. Environmental Protection
Agency, Office of Research and Development; EPA report no. EPA-600/4-85-038.
Available from: NTIS, Springfield, VA; PB85-212413.
Evans, G.; Finkelstein, P.; Martin, B.; Possiel, N.; Graves, M. (1982) The
National Air Pollution Background Network, 1976-1980. Research Triangle
Park, NC: U.S. Environmental Protection Agency; EPA report no. EPA-600/
4-82-058. Available from: NTIS, Springfield, VA; PB83-100412.
Evans, G.; Finkelstein, P.; Martin, B.; Possiel, N.; Graves, M. (1983). Ozone
measurements from a network of remote sites. J. Air Pollut. Control
Assoc. 33: 291-296.
Farmer, J. C.; Dawson, G. A. (1982) Condensation sampling of soluble atmos-
pheric trace gases. JGR J. Geophys. Res. 87: 8931-8942.
Galbally, I. E.; Roy, C. R. (1980) Destruction of ozone at the earth's surface.
Quart. J. R. Met. Soc. 106: 599-620.
Graedel, T. E.; Farrow, L. A.; Weber, T. A. (1977) Photochemistry of the "Sunday
effect." Environ. Sci. Techno!. 11: 690-694.
Grosjean, D. (1981) Critical evaluation and comparison of measurement methods
for nitrogenous compounds in the atmosphere. Final report, A 706-05, for
Coordinating Research Council. Environmental Research and Technology.
Westlake Village, CA.
5-118
-------
Grosjean, D. (1983) Distribution of atmospheric nitrogenous pollutants at a
Los Angeles area smog receptor site. Environ. Sci. Techno!. 17: 13-19.
Hanst, P. L. ; Wilson, W. E.; Patterson, R. K.; Gay, B. W., Jr.; Chaney, L. W.;
Burton, C. S. (1975) A spectroscopic study of California smog. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-650/4-75-006. Available from: NTIS, Springfield, VA; PB-241022.
Hanst, P. L.; Wong, N. W.; Bragin, J. (1982) A long path infrared study of Los
Angeles smog. Atmos. Environ. 16: 969-981.
Heikes, B. G.; Lazrus, A. L.; Kok, G. L.; Kunen, S. M.; Gandrud, B. W.; Gitlin,
S. N. ; Sperry, P. D. (1982) Evidence of aqueous phase hydrogen peroxide
synthesis in the troposphere. JGR J. Geophys. Res. 87: 3045-3051.
Heuss, J. M.; Glasson, W. A. (1968) Hydrocarbon reactivity and eye irritation.
Environ. Sci. Techno!. 2: 1109-1116.
Hoggan, M. (1986) Comment on "Ten-year ozone trends in California and Texas."
J. Air Pollut. Control Assoc. 36: 599-600.
Holdren, M. W. ; Spicer, C. W.; Hales, J. M. (1984) Peroxyacetyl nitrate solu-
bility and decomposition rate in water. Atmos. Environ.: in press.
Holguin, A. H. ; Buffler, P. A.; Contant, C. F., Jr.; Stock, T. H.; Kotchmar,
D. J.; Hsi, B. P.; Jenkins, D. E.; Gehan, B. M.; Noel, L. M. ; Mei, M.
(1985) The effects of ozone on asthmatics in the Houston area. In: Proceed-
ings of APCA Speciality Conference: Evaluation of the Scientific Basis
for Ozone/Oxidants Standards. Pittsburgh, PA: Air Pollution Control
Assoc. Houston, TX, November 28-30, 1984.
Hov, 0.; Hesstvedt, E. ; Isaksen, I. S. A. (1978) Long-range transport of
tropospheric ozone. Nature (London) 1: 341-344.
Hunt, W. F., Jr.; Curran, T. C. (1982) National and regional trends in ambient
ozone measurements, 1975-1981. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards.
Isaksen, I. S. A. ; Hov, 0.; Hesstvedt, E. (1978) Ozone generation over rural
areas. Environ. Sci. Technol. 12: 1279-1284.
Jorgen, R. T.; Meyer, R. A.; Hughes, R. A. (1978) Routine peroxyacetyl nitrate
(PAN) monitoring applied to the Houston Area Oxidant Study. Presented at:
71st annual meeting of the Air Pollution Control Association; June;
Houston, TX. Pittsburgh, PA: Air Pollution Control Association; paper no.
78-50.1.
Kelly, T. J. ; Stedman, D. H.; Kok, G. L. (1979) Measurements of H202 and HN03
in rural air. Geophys. Res. Lett. 6: 375-378.
Kelly, N. A.; Ferman, M. A.; Wolff, G. T. (1986) The chemical and meteorological
conditions associated with high and low ozone concentrations in south-
eastern Michigan and nearby areas of Ontario. J. Air Pollut. Control Assoc.
36: 150-158.
5-119
-------
Kok, G. L.; Darnall, K. R.; Winer, A. M.; Pitts, J. N., Jr.; Gay, B. W.;
(1978) Ambient air measurements of hydrogen peroxide in the California
South Coast Air Basin. Environ. Sci. Techno!. 12: 1077-1080.
Kress, L. W.; Miller, J. E. (1983) Impact of ozone on soybean yield. J. Environ.
Qual. 12: 276-281.
Kroening, J. L.; Ney, E. P. (1962) Atmospheric ozone. J. Geophys. Res.
67: 1867-1875.
Lefohn, A. S. (1982) The identification and description of ozone monitoring
sites which represent high yield reduction potential for agricultural
crops. Prepared for U.S. Environmental Protection Agency, Corvallis, OR.
A.S.L. and Assoc. Helena, MT.
Lefohn, A. S. (1984) A comparison of ambient ozone exposures for selected
nonurban sites. Proceedings of 77th Annual Meeting of the Air Pollution
Control Association. Vol. 6. Paper no. 84-104.1. Pittsburgh, PA: Air
Pollution Control Assoc.
Lefohn, A. S.; Benedict,1 H. M. (1985) Exposure considerations associated with
characterizing ozone ambient air quality monitoring data. In: Proceedings
of the APCA Specialty Conference: Evaluation of the Scientific Basis for
Ozone/Oxidants Standards, Houston, TX, November 28-30, 1984. Pittsburgh,
PA: Air Pollution Control Assoc.
Lefohn, A. S.; Jones, C. K. (1986) The characterization of 03 and S02 air
quality data for assessing possible vegetation effects. J. Air Pollut.
Control Assoc. 36: (In Press). (Presented at annual meeting of Air Pollu-
tion Control Assoc., Minneapolis, MN, July 1986.)
Levitt, S. B.; Chock, D. P. (1976) Weekday-weekend pollutant studies on the Los
Angeles Basin. J. Air Pollut. Control Assoc. 26: 1091-1092.
Lewis, T. E.; Brennan, E.; Lonneman, W. A. (1983) PAN concentrations in ambient
air in New Jersey. J. Air Pollut. Control Assoc. 33: 885-887.
Logan, J. A.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B. (1981) Tropospheric
chemistry: a global perspective. JGR J. Geophys. Res. 86: 7210-7254.
Lonneman, W. A.; Bufalini, J. J.; Seila, R. L. (1976) PAN and oxidant measure-
ment in ambient atmospheres. Environ. Sci. Techno!. 10: 374-380.
Martinez, J. R.; Singh, H. B. (1979) Survey of the role of NOx in nonurban
ozone formation. Research Triangle Park, NC: prepared by SRI International
for U.S. Environmental Protection Agency. EPA report no. EPA-450/4-79-035.
Available from: NTIS, Springfield, VA; PB80-122815.
Mayrsohn, H.; Brooks, C. (1965) The analysis of PAN by electron capture gas
chromatography. Presented at: Western regional meeting of the American
Chemical Society; November; Los Angeles, CA. Los Angeles, CA: California
State Department of Public Health.
5-120
-------
McElroy, M. B.; Wofsy, S. C.; Logan, J. A.; Prather, M. J. (1985) Theoretical
studies of long-range transport of oxidants. Annual technical report to
Coordinating Research Council. Atlanta, GA: Coordinating Research Council,
CRC Report No. CAPA-22-83.
Miller, A.; Ahrens, C. P. (1969) Ozone within and below the West Coast tempera-
ture inversion. San Jose, CA; San Jose State College. Report No. 6.
Miller, P. R. ; McCutchan, M. H.; Ryan, B. C. (1972) Influence of climate and
topography on oxidant air pollution concentrations that damage conifer
forests in sourthern California. Mitt. Forst. Bundesversuchsanst. Wien
97: 585-607.
Mohnen, V. A. (1977) Review and analysis. In: International Conference on
Oxidants, 1976 — Analysis of evidence and viewpoints. Part III. The
issue of stratospheric ozone intrusion. U.S. Environmental Protection
Agency, Research Triangle Park, NC; Report no. EPA-600/3-77-115.
Moschandreas, D. J. ; Stark, J. W. C. ; McFadden, J. E.; Morse, S. S. (1978)
Indoor air pollution in the residential environment. Vol. I. Data collec-
tion, analysis, and interpretation. Research Triangle Park, NC: U.S.
Environmental Protection Agency; EPA report no. EPA-600/7-78-229a. Avail-
able from: NTIS, Springfield, VA; PB-290999.
Moschandreas, D. J.; Zabransky, J.; Petton, D. J. (1981) Comparison of indoor
and outdoor air quality. Palo Alto, CA: Geomet, Inc. Electric Power
Research Institute. Report no. EA-1733. Available from: NTIS, Springfield,
VA; EPRI-EA-1733.
Mueller, F. X.; Loeb, L.; Mapes, W. H. (1973) Decomposition rates of ozone in
living areas. Environ. Sci. Techno!. 7: 342-346.
National Research Council, (1977) Ozone and other photochemical oxidants.
Washington, DC; National Academy of Sciences.
Nieboer, H.; van Ham, J. (1976) Perpxyacetyl nitrate (PAN) in relation to
ozone and some meteorological parameters at Delft in the Netherlands.
Atmos. Environ. 10: 115-120.
Nielsen, T.; Samuelsson, U.; Grennfelt, P.; Thomsen, E. L. (1981) Peroxyacetyl
nitrate in long-range transported polluted air. Nature (London) 293:
553-555.
Oltmans, S. J. (1981) Surface ozone measurements in clean air. J. Geophys. Res.
86: 1174-1180. Cited in: Logan et al., 1981.
Peaks, E.; MacLean, M. A.; Sandhu, H. S. (1983) Surface ozone and peroxyacetyl
nitrate (PAN) observations at rural locations in Alberta, Canada. J. Air
Pollut. Control Assoc. 33:881-883.
Peterson, G. A.; Sabersky, R. H. (1975) Measurements of pollutants inside an
automobile. J. Air Pollut. Control Assoc. 25: 1028-1032.
5-121
-------
Pitts, J. N., Jr.; Grosjean, D. (1979) Detailed characterization of gaseous
and size-resolved participate pollutants at a south coast air basin smog
receptor site: levels and modes of formation of sulfate, nitrate and
organic particulates and their implications for control strategies.
Sacramento, CA: California Air Resources Board; report no. ARB-R-5-384-
79-100. Available from: NTIS, Springfield, VA; PB-301294.
Pratt, G. C.; Hendrickson, R. C.; Chevone, B. I.; Christopherson, D. A.;
O'Brien, M. V.; Krupa, S. V. (1983) Ozone and oxides of nitrogen in the
rural upper-midwestern U.S.A. Atmos. Environ. 17: 2013-2023.
Reiter, E. R. (1977a) Review and analysis. In: International Conference on
Oxidants 1976 -- Analysis of evidence and viewpoints. Part III. The issue
of stratospheric ozone intrusion. U.S. Environmental Protection Agency,
Research Triangle Park, NC; Report no. EPA-600/3-77-115.
Reiter, E. R. (1977b) The role of stratospheric import or tropospheric ozone
concentrations. In: International Conference on Photochemical Oxidant
Pollution and Its Control. Proceedings. V. I. Dimitriades, B., ed. U.S.
Environmental Protection Agency, Research Triangle Park, NC; Report no.
EPA-600/3-77-001a.
Renzetti, N. A.; Bryan, R. J. (1961) Atmospheric sampling for aldehydes and
eye irritation in Los Angeles smog - 1960. J. Air Pollut. Control Assoc.
11: 421-424.
Rhodes, C. E.; Holland, D. M. (1981) Variations of NO, N02, and 03 concentra-
tions downwind of a Los Angeles freeway. Atmos. Environ. 15: 243-250.
Sabersky, R. H.; Sinema, D. A.; Shair, F. H. (1973) Concentrations, decay
rates, and removal of ozone and their relation to establishing clean
indoor air. Environ. Sci. Techno!. 1: 347-353.
SAROAD, Storage and Retrieval of Aerometric Data [data base]. (1985a) Data
file for 1976. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
SAROAD, Storage and Retrieval of Aerometric Data [data base]. (1985b) Data
file for 1979. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
SAROAD, Storage and Retrieval of Aerometric Data [data base]. (1985c) Data
file for 1980. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
SAROAD, Storage and Retrieval of Aerometric Data [data base]. (1985d) Data
file for 1981. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
SAROAD, Storage and Retrieval of Aerometric Data [data base]. (1985e) Data
file for 1982. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
5-122
-------
SARGAD, Storage and Retrieval of Aerometric Data [data base]. (1985f) Data
file for 1983. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Disc; ASCII.
Schuck, E. A.; Pitts, J. N., Jr.; Wan, J. K. S. (1966) Relationships between
certain meteorological factors and photochemical smog. Int. J. Air Water
Pollut. 10: 689-711.
Seiler, W.; Fishman, J. (19810 The distribution of carbon monoxide and ozone
in the free troposphere. JGR J. Geophys. Res. 86: 7255-7265.
Shair, F. H.; Heitner, K. L. (1974) Theoretical model for relating indoor
pollutant concentrations to those outside. Environ. Set. Techno!.
8: 444-451. ... . <
Singh, H. B.; Hanst, P. W. (1981) Peroxyacetyl nitrate (PAN) in the unpolluted
atmosphere: an important reservoir for nitrogen oxides. Geophys. Res.
Lett. 8: 941-944.
Singh, H. B.;,Salas, L. J. (1983) Methodology for the analysis of peroxyacetyl
nitrate (PAN) in the unpolluted atmosphere. Atmos. Environ. 17: 1507-1516.
Singh, H. B.; Ludwig, F. L.; Johnson, W. B. (1977) Ozone in clean remote,
atmospheres. Concentrations and variabilities. SRI Project No. 5661. Menlo
Park, CA: SRI International. ,,
Singh, H. B.; Salas, L. J.; Shigeishi, H.; Smith, A, J.; Scribner, E.; Gavanagh,
L. A. (1979) Atmospheric distributions, sources and sinks of selected
halocarbons, hydrocarbons, SF6, and N20. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Environmental Sciences Research Laboratory;
EPA report no. EPA-600/3-79-107. Available from: NTIS, Springfield, VA;
PB80-142706.
Singh, H. B.; Salas, L. J. ; Smith, A. J..; Shigeishi, H. (1981) Measurements of
some potentially hazardous organic chemicals in urban environments.
Atmos. Environ. 15: 601-612.
Singh, H. B. , Salas, L. J.; Stiles, R.; Shigeishi, H. (1982) Measurements of
hazardous organic chemicals in the ambient atmosphere. Research Triangle
Park, NC: U.S. Environmental Protection Agency, Environmental Sciences
Research Laboratory. EPA report no. EPA-600/3-83-002. Available from:
NTIS, Springfield, VA; PB83-156935.
Smith, W. J. (1981) New York State air monitoring data report for the Northeast
Corridor Regional Modeling Project (NECRMP). Albany, NY: New York State
Department of Environment and Conservation.
Spicer, C. W. (1976) The fate of nitrogen oxides in the atmosphere. Research
Triangle Park, NC: U.S. Environmental Protection Agency; EPA report no.
EPA-600/3-76-030. Available from: NTIS, Springfield, VA; PB-248-260.
Spicer, C. W. (1977) The fate, of nitrogen oxides in the atmosphere. In: Pitts,
J. N., Jr.; Metcalf, R. L.; Lloyd, A. C.; eds. Advances in environmental
science and technology: v. 7. New York, NY: John Wiley and Sons; pp.
163-261.
5-123
-------
Spieer, C. W.; Sverdrup, G. W. (1981). Trace nitrogen chemistry during the
Philadelphia oxidant data enhancement study (1979). Report on Contract
no. 68-02-0338. U.S. Environmental Protection Agency, Research Triangle
Park, NC.
Spicer, C. W. ; Gemma, J. L; Joseph, D. W.; Sticksel, P. R. ; Ward, G. F. (1976)
The transport of oxidant beyond urban areas. Columbus, OH: Battelle
Columbus Laboratories; EPA report no. EPA-600/3-76-018a. Available from:
NTIS, Springfield, VA; PB-253736.
Spicer, C. W.; Holdren,
acetyl nitrate in
17: 1055-1058.
M. W.; Keigley,
the continental
G. W. (1983) The
boundary layer.
ubiquity of peroxy-
Atmos. Environ.
Stock, T. H.; Holguin, A. H.; Selwyn, B. J.; Hsi, B. P.; Contant, C. F. ;
Buffler, P. A.; Kotchmar, D. J. (1983). Exposure estimates for the Houston
area asthma and runners studies. In: Lee, S. D.; Mustafa, M. G.; Mehlman,
M. A. eds. Adv. Modern Environ. Toxicol. Vol. V. International Symposium
on the Biomedical Effects of Ozone and Related Photochemical Oxidants.
Pinehurst, NC; March 14-18, 1982. Princeton, NJ: Princeton Scientific
Publishers, Inc.
Taylor, 0. C. (1969) Importance of peroxyacetyl nitrate (PAN) as a phytotoxic
air pollutant. J. Air. Pollut. Control Assoc. 19: 347-351.
Temple, P. J.; Taylor, 0.
acetyl nitrate (PAN)
17: 1583-1587.
C. (1983) World-wide ambient measurements of peroxy-
and implications for plant injury. Atmos. Environ.
Thompson, C. R.; Hensel, E. G.; Kats, G. (1973) Outdoor-indoor levels of six
air pollutants. J. Air Pollut. Control Assoc. 23: 881-886.
Tingey, D. T.; Hill, A. C. (1968) The occurrence of photochemical phytotoxi-
cants in the Salt Lake Valley. Utah Acad. Sci. Proc. 44: 387-395.
Tuazon, E. C.; Graham, R. A.; Winer, A. M.; Easton, R. R.; Pitts, J. N., Jr.
(1978) A kilometer pathlength Fourier-transform infrared system for the
study of trace pollutants in ambient and synthetic atmospheres. Atmos.
Environ. 12: 865-875.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1980) Atmos-
pheric measurements of trace pollutants by kilometer-pathlength FT-IR
spectroscopy. In: Pitts, J. N.; Metcalf, R. L., eds. Advances in environ-
mental sciences and technology: v. 10. New York, NY: John Wiley and Sons;
pp. 254-300.
Tuazon, E. C.; Winer, A. M.; Graham, R. A.; Pitts, J. N., Jr. (1981a) Atmos-
pheric measurements of trace pollutants: long path Fourier transform
infrared spectroscopy. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Environmental Sciences Research Laboratory. Available
from: NTIS, Springfield, VA; PB81-179848.
5-124
-------
Tuazon, E. C.; Winer, A. M.; Pitts, J. N., Jr. (1981b) Trace pollutant concen-
trations in a multiday smog episode in the California South Coast Air
Basin by long path length Fourier transform infrared spectrometry. Environ.
Sci. Techno!. 15: 1232-1237.
U.S. Department of Commerce, Bureau of the Census. (1982) Statistical Abstract
of the United States, 1982-1983. 103rd Ed. Washington, DC.
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; PB82-242231.
U.S. Environmental Protection Agency. (1978) Air quality criteria for ozone
and other photochemical oxidants. Research Triangle Park, NC: U.S. Environ-
mental Protection Agency, Environmental Criteria and Assessment Office;
EPA report no. EPA-600/8-78-004. Available from: NTIS, Springfield, VA;
PB80-124 753.
U. S. Environmental Protection Agency. (1980) Air quality data—1979 annual
statistics. Research Triangle Park, NC: U. S. Environmental Protection
Agency, Office of Air Quality Planning and Standards; EPA report no.
EPA-450/4-80-014.
U. S. Environmental Protection Agency. (1981) Air quality data—1980 annual
statistics. Research Triangle Park, NC: U. S. Environmental Protection
Agency, Office of Air Quality Planning and Standards; EPA report no.
EPA-45Q/4-81-027.
U. S. Environmental Protection Agency. (1982a) Air quality data—1981 annual
statistics. Research Triangle Park, NC: U. S. Environmental Protection
Agency, Office of Air Quality Planning and Standards; EPA report no.
EPA-450/4-82-007.
U.S. Environmental Protection Agency. (1982b) Air quality criteria for oxides
of nitrogen. Research Triangle Park, NC: U.S. Environmental Protection
Agency, Environmental Criteria and Assessment Office; publication no.
EPA-600/8-82-026. Available from: NTIS, Springfield, VA; PB83-131011.
U.S. Environmental Protection Agency. (1984a) National air quality and emission
trends report, 1983. Research Triangle Park, NC: U.S. Environmental
Protection Agency. Office of Air Quality Planning and Standards. EPA
report no. EPA-450/4-84-029.
U.S. Environmental Protection Agency. (1984b) National air pollutant emission
estimates, 1940-1983. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards. EPA
report no. EPA-450/4-84-028.
U.S. Environmental Protection Agency. (1986) National air quality and emissions
trends report, 1984. Research Triangle Park, NC: Office of Air Quality
Planning and Standards. Report no. EPA-450/4-86-001,
5-125
-------
Viezee, W.; Johnson, W. B.; Singh, H. B. (1979) Airborne measurements of
stratospheric ozone intrusions into the troposphere over the United
States. Final report. SRI Proj. 6690, CRC-APRAC Proj. no. CAPA-15-76
(1-77), Coordinating Research Council, Atlanta, GA.
Walker, H. M. (1985) Ten-year ozone trends in California and Texas. J, Air
Pollut. Control Assoc. 35: 903-912.
Walker, H. M. (1986) Author's reply [to Comment on "Ten-year ozone trends in
California and Texas"]. J. Air Pollut. Control Assoc, 36: 600-602.
Wendland, W. M.; Bark, L. D.; Clark, D. R.; Curry, R. B.; Enz, J. W.; Hubbard,
K. G.; Jones, V.; Kuehnast, E. L.; Lytle, W,; Newman, J.; Nurnberger, F. V.;
Waite, P. (1984) A climatic review of summer 1983 in the upper Midwest.
Bull. Amer. Meteor. Soc. 65(10).
Wesely, M. (1983) Turbulent transport of ozone to surfaces common in the eastern
half of the United States. In: Trace atmospheric constituents: Properties,
transformations, and fates. Schwartz, S. E., ed. New York, NY: John Wiley
& Sons, Inc.
Westberg, H.; Allwine, K.; Robinson, E. (1978) Measurement of light hydrocar-
bons and oxidant transport: Houston area 1976. Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Research and Develop-
ment; EPA report no. EPA-600/3-78-662. Available from: NTIS, Springfield,
VA; PB-285891.
Wolff, G. T.; Lioy, P. J. (1980) Development of an ozone river associated with
synoptic scale episodes in the eastern United States. Environ. Sci.
Techno!. 14: 1257-1260.
Wolff, G. T.; Stasiuk, W. N.; Coffey, P. E.; Pasceri, R. E. (1975) Aerial
ozone measurements over New Jersey, New York, and Connecticut. Presented
at the 68th Annual Meeting of the Air Pollution Control Association,
Boston, MA: June. Pittsburgh, PA: Air Pollution Control Association.
Paper no. 75-58.6.
Wolff, G. T.; Lioy, P. T.; Wight, G. D.; Meyers, R. F.; Cederwall, R. T.
(1977a) An investigation of long-range transport of ozone across the
midwestern and eastern United States. In: Dimitriades, B., ed. Inter-
national conference on photochemical oxidant pollution and its control.
Proceedings: volume I. Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Research and Development; pp. 307-317; EPA
report no. EPA-600/3-77-001a. Available from: NTIS, Springfield, VA;
PB-264232.
Wolff, G. T.; Lioy, P. J.; Wight, G. D.; Pasceri, R. E. (1977b) Aerial
investigation of photochemical oxidants over the northeast. In: Bufalini,
J. J.; Lonneman, W. A., eds. Proceedings of symposium on 1975 northeast
oxidant transport study, Research Triangle Park, NC, January 20-21, 1976.
Research Triangle Park, NC: U.S. Environmental Protection Agency. Report
no. EPA-600/3-77-017.
5-126
-------
Wolff, G. T.; Kelly, N. A.; Ferman, M. A. (1981) On the sources of summertime
haze in the eastern United States. Science (Washington, DC) 211: 703-705.
Wolff, G. T.; Lioy, P. J.; Taylor, R. S. (1986) The diurnal variations of ozone
at different altitudes on a rural mountain in the eastern United States.
J. Air Pollut. Control Assoc. 36: (In Press)
Yocom, J. E. (1982) Indoor-outdoor air quality relationships: a critical
review. J. Air Pollut. Control Assoc. 32: 500-520.
Zagraniski, R. T.; Leaderer, B. P.; Stolwijk, J. A. J. (1979) Ambient sulfates,
photochemical oxidants, and acute adverse health effects: an epidemiolog-
ical study. Environ. Res. 19: 306-320.
5-127
U.S. GOVEMMENT PRIWOTG OFFICE! 1986- 6 % 6- 1 1 6/ i»0661
-------
-------
-------
Environmental Protection
Agency
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use, $300
Please make all necessary changes on tha abovt label,
detach or copy, and return to the address in the upper
left-hand corner.
if you do not wish lo receive these reports CHECK HERE o;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
EPA/600/8-84/020bF
------- |