United States
Environmental Protection
Agency
Office of Aii Quality
Planning and Standards
Research Triangle Paric, NC 27711
EPA-454/R-98-002 V
August 1998
Air
?, EPA
GUIDELINE ON OZONE
MONITORING SITE SELECTION
-------
EPA-454/R-98-002
Guideline On Ozone Monitoring Site Selection
United States Environmental Protection Agency
Office of Air Quality Planning and Standards
Emissions, Monitoring, and Analysis Division
Research Triangle Park, NC 27711
August 1998
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, II 60604-3590
-------
TABLE OF CONTENTS
Section
LIST OF FIGURES v
LIST OF TABLES vii
LIST OF ABBREVIATIONS AND ACRONYMS viii
EXECUTIVE SUMMARY ES-1
1. INTRODUCTION '.....' 1-1"
2. IMPORTANT CHARACTERISTICS OF OZONE FORMATION 2-1
2.1 OZONE FORMATION CHEMISTRY 2-1
2.1.1 Basic Principles 2-1
2.1.2 Atmospheric Sinks for Ozone 2-3
2.2 OZONE PRECURSOR EMISSIONS 2-5
2.2.1 Sources of Ozone Precursor Emissions 2-5
2.2.2 Spatial Distribution of Emissions 2-5
2.3 METEOROLOGICAL CONDITIONS CONDUCIVE TO OZONE
FORMATION 2-5
3. SPATIAL DISTRIBUTION OF OZONE CONCENTRATIONS 3-1
3.1 EXAMPLE 1-HOUR AND 8-HOUR OZONE PATTERNS 3-1
3.1.1 San Francisco Bay Area 3-1
3.1.2 Northeastern United States 3-11
3.2 SUMMARY AND CONCLUSIONS 3-21
4. MONITORING OBJECTIVES AND NETWORK DESIGN 4-1
4.1 INTRODUCTION 4-1
4.2 PRINCIPLES INVOLVING LOCALIZED VS. REGIONAL MONITORING... 4-2
4.3 MONITORING OBJECTIVES AND SITE SELECTION 4-3
4.3.1 Regulatory Compliance 4-6
4.3.2 Control Strategy Development and Assessment 4-6
4.3.3 Health Effect Studies 4-6
4.3.4 Vegetative Impacts Related to a Secondary Ozone Standard 4-7
4.4 OZONE SEASON 4-7
5. MACROSCALE CONSIDERATIONS IN MONITOR SITING 5-1
5.1 DEVELOPMENT OF A NEW NETWORK OR ENHANCEMENT OF AN
EXISTING NETWORK 5-1
5.1.1 Maximum Population Exposure Site(s) 5-1
5.1.2 Upwind and Downwind Site(s) 5-1
5.2 EXAMINATION AND ADJUSTMENT OF AN EXISTING OZONE MONITORING
NETWORK TO REFLECT 8-HOUR MONITORING
NEEDS 5-11
iii
-------
TABLE OF CONTENTS (Concluded)
Section Page
6. MICROSCALE CONSIDERATIONS FOR MONITOR SITING 6-1
6.1 GUIDANCE FOR MONITOR PLACEMENT '.'. 6-1
6.1.1 Vertical and Horizontal Probe Placement 6-1
6.1.2 Effects of Obstructions 6-1
6.1.3 Separation from Roadways -. 6-3
6.1.4 Spacing from trees and other vegetation 6-3
6.2USEOF SATURATION MONITORING 6-3
7. RECOMMENDATIONS 7-1
8. REFERENCES 8-1
APPENDIX A: BACKGROUND ON OZONE FORMATION A-l
APPENDIX B: SUPPLEMENTAL REFERENCE MATERIALS B-l
IV
-------
LIST OF FIGURES
Page
igure
2-1. Schematic diagram of photochemical air pollution from emission to deposition
(NAS, 1991) 2-2
2-2. Typical ozone isopleths used in the EPA's Empirical Kinetic Modeling Approach
(EKMA) 2-4
2-3. Spatial distribution of the anthropogenic VOC emissions (tons per year) in the OTR
during 1990 2-7
2-4. Spatial distribution of the anthropogenic NOX emissions (tons per year) in the OTR
during 1990 2-8
3-1. Selected cities in the San Francisco Bay Area and downwind regions 3-3
3-2. Selected monitoring sites that operated in the San Francisco Area and downwind
regions between 1992-1995 3-4
3-3a. Exceedances of the 0.12 ppm ozone threshold between 1992-1995 in the
San Francisco Bay Area and downwind regions 3-6
3-3b. Exceedances of the 0.12 ppm ozone threshold between 1992-1995 in the
San Francisco Bay Area and downwind regions 3-7
3-4a. Contour plot of the exceedances of the 0.08 ppm 8-hour threshold between
1992-1995 in the San Francisco Bay Area and downwind regions 3-8
3-4b. Exceedances of the 0.08 ppm 8-hour threshold between 1992-1995 in the
San Francisco Bay Area and downwind regions 3-9
3-5. Ozone concentration profiles for Bethel Island and Livermore for July 27,1995 3-10
3-6. Ozone concentration profile for Woodland, California (northeast of the
San Francisco Bay Area) on August 20,1995 3-12
3-7. Major cities in the Northeastern United States 3-13
3-8. Number of site exceedances of the 1 hour 0.12 ppm ozone threshold (greater than
or equal to 125 ppb) in the Northeast OTR during 1988 3-15
3-9. Number of site exceedances of the 1 hour 0.12 ppm ozone threshold (greater than
or equal to 125 ppb) in the Northeast OTR during 1995 3-16
-------
LIST OF FIGURES (Concluded)
Figure Page
3-10. Number of site exceedances of the 0.08 ppm 8-hour running average in the
Northeast OTR during 1988 3-17
3-11. Number of site exceedances of the 0.08 ppm 8-hour running average in the
Northeast OTR during 1995 - 3-18
3-12. Ratio of the concentration of the 0.08 ppm 8-hour running average standard and
the 1 hour 0.12 ppm ozone threshold for the ozone episode of July 20, 1991 3-19
3-13. Ratio of the concentration of the 0.08 ppm 8-hour running average standard and
the 1 hour 0.12 ppm ozone threshold for the ozone episode of July 15, 1995 3-20
5-1. Morning wind roses (7-10 a.m.) for Muskegon, Michigan from 1981-1990 for
summer days (June-September) for (a) days that exceeded the 0.12 ppm 1-hour
ozone threshold, and (b) all days (Chu, 1995) 5-2
5-2. Depiction of the 30, 35, and 40 degree latitude lines for the United States 5-5
5-3. Morning (7-10 a.m.) and .afternoon (1-4 p.m.) wind roses for Milwaukee,
Wisconsin (Chu, 1995) 5-7
5-4. Morning (7-10 a.m.) and afternoon (1-4 p.m.) wind roses for Atlanta, Georgia
(Chu, 1995) 5-8
5-5. Example network design for a metropolitan area 5-9
5-6. Morning wind roses (7-10 a.m.) for two stagnation dominated metropolitan areas:
Louisville, Kentucky and Huntington, West Virginia (Chu, 1995) 5-10
5-7. Example network design for a metropolitan area with no predominant wind
direction (Chu, 1995) 5-12
6-1. Schematic representation of the airflow around an obstacle such as a building
(EPA,1978) 6-2
6-2. Monitor locations for the Wasatch Front Ozone Saturation Study summer 1993
(Start, 1994) 6-6
VI
-------
LIST OF TABLES
Table Page
2-1. Summary of Anthropogenic VOC and NOX emissions in the United States during 1994
(EPA, 1994) 2-6
3-1. Monitoring sites in the San Francisco Bay Area and downwind regions 3-5
4-1. Ozone and ozone precursor monitoring site types with corresponding
monitoring objectives 4-5
5-1. Criteria for ozone conducive conditions for the eastern United States 5-4
6-1. Separation distance between ozone monitors and roadways 6-4
vn
-------
LIST OF ABBREVIATIONS AND ACRONYMS
Act - Clean Air Act
CBD - Central Business District
CFR - Code of Federal Regulations
EKMA - Empirical Kinetic Modeling Approach
EPA - United States Environmental Protection Agency •
FR - Federal Register
hv - Photochemical energy from ultraviolet radiation or a photon
k,, - Rate Constant
M - A molecule which absorbs a reaction's excess vibrational energy
NAAQS - National Ambient Air Quality Standard(s)
NAMS - National Air Monitoring Stations
NMHC - Non-Methane Hydrocarbon(s)
NO - Nitric Oxide
NOX - Oxides of Nitrogen
N0y - Total Reactive Oxides of Nitrogen
NO2 - Nitrogen Dioxide
O - Oxygen Atom
O2 - Oxygen Molecule
O3 - Ozone
PAMS - Photochemical Assessment Monitoring Stations
PWD - Prevailing Wind Direction
RH - Relative Humidity
SIP - State Implementation Plan
SLAMS - State and Local Air Monitoring Stations
T - Temperature
VOC - Volatile Organic Compound(s)
W - Wind Speed
Vlll
-------
EXECUTIVE SUMMARY
OVERVIEW
This report provides background information for State and local agencies responsible for
designing and implementing an ozone monitoring program. The purpose of this report is to
update a 1978 U.S. Environmental Protection Agency (EPA) repert entitled: Site Selection for
the Monitoring of Photochemical Air Pollutants (EPA, 1978). This updated report, Guideline
on Ozone Monitoring Site Selection, is motivated by the need to help air quality agencies and
others incorporate 8-hour monitoring into their ozone monitoring program. Since the 1970s,
ozone monitoring programs have been designed to address the 1-hour ozone National Ambient
Air Quality Standard (NAAQS); however, on July 18,1997, the U.S. EPA promulgated an 8-
hour ozone NAAQS. This report assists those agencies considering how to design, implement or
revise ozone monitoring networks, particularly in light of the need to collect 8-hour ozone
concentration data. Although this report does not specifically address special purpose monitors
(SPMs), these additional monitors deployed by air pollution control agencies can serve multiple
purposes including:
fulfilling the need to obtain information on where to locate permanent monitoring
stations,
• providing data to support pollutant formation and transport analyses,
• assessing air quality in a particular location, or
• determining where the monitored domain near urban areas should be expanded to ensure
measurement of a maximum 8-hour concentration.
Where SPM data are used as part of a demonstration of attainment or nonattainment or in
computing an ozone design value, the SPM must meet the requirements for SLAMS. Note that
the EPA is obligated to consider all publicly available, valid (i.e., collected in accordance with 40
CFR 58), and relevant data, including that collected by SPMs, in the NAAQS regulatory process.
More detail on SPMs may be found in 40 CFR 58.14.
Existing ozone monitoring networks should be reviewed annually in accordance with the
requirements of 40 CFR 58.20 (d); any changes to the networks to accommodate the 8-hour
ozone standard should be addressed at that time (see also the SLAMS/NAMS/PAMS Network
Review Guidance, EPA-454/R-98-003, for further detail on network review.) The EPA does not
expect major changes to the monitoring networks as a result of these reviews since the analysis in
this document indicates that, generally, the 1 -hour and 8-hour concentrations are well correlated
spatially.
The report covers basic issues essential to understanding monitor siting considerations,
including the following:
ES-1
-------
The physics and chemistry of ozone formation: The report includes a description of
ozone chemistry, the relationship between precursor and ozone concentrations, and
meteorological conditions that are most conducive to ozone formation.
Spatial distribution of ozone concentrations: Distributions of ozone concentrations
across metropolitan areas and ozone transport regions or broad geographical areas are
examined and related to ozone chemistry and meteorology. Also included is a
comparison of the spatial distribution of 1-hour and 8-hour ozone concentrations.
Monitoring objectives and network design: Purposes of ozone monitoring are described,
as well as the types and number of monitoring sites that are recommended.
Macroscale considerations for monitor placement: Techniques are recommended to
identify general locations for the required monitor types for a metropolitan area.
Microscale considerations for monitor placement: Techniques are recommended for
finding specific locations for ozone monitors.
Appendices include background information on ozone formation, and references to
supplemental materials that can assist those involved with siting efforts.
REVIEW OF KEY FINDINGS AND RECOMMENDATIONS
In this report, improved knowledge and understanding of ozone photochemistry is used to
reach a series of conclusions and recommendations regarding ozone monitoring networks. In
particular:
Ozone Formation (Section 2.)
Section 2. includes a description of the physics and chemistry of ozone formation (a
supplement to this section is included as Appendix A). The following key principles are
illustrated:
Ozone is a secondary pollutant formed in the atmosphere by reactions between oxides of
nitrogen (NOX) and volatile organic compounds (VOC).
The relationship between the precursor concentrations and ozone formation must be
considered when developing control strategies. Ozone formation in some geographic
areas may be limited by the supply of NOX, while in other areas ozone formation may be
limited by the supply of VOC.
Ozone formation is most conducive during warm, dry, and cloudless days with low wind
speeds; these conditions most often occur during high-pressure systems. Consequently,
ozone monitoring should take place during the warmer periods of the year.
ES-2
-------
Spatial Distribution of Ozone Concentrations (Section 3.)
Using examples from the San Francisco Bay Area and the northeastern United States,
Section 3. includes a discussion of the urban and regional distribution of ozone concentrations.
The following key principles are discussed:
Generally, the highest ozone concentrations in a metropolitan area are found downwind
of the urban fringe.
The locations of the peak concentrations of a 0.12 ppm 1-hour and a 0.08 ppm 8-hour
threshold are similar; however, downwind sites may be more prone to 8-hour
exceedances relative to 1-hour exceedances.
Subtle differences exist between the 1-hour and 8-hour forms. For example, downwind
sites which receive transported, dispersed ozone plumes and have low NO concentrations
available to titrate ozone, will have flatter diurnal concentration profiles that are more
prone to 0.08 ppm 8-hour exceedances, relative to 0.12 ppm 1-hour exceedances. Also,
the ratio of the 8-hour running average ozone concentration and the 1-hour daily
maximum ozone concentration generally increases from urban centers, suggesting that
rural sites may be more prone to 8-hour exceedances than to 1-hour exceedances.
Monitoring Objectives and Network Design (Section 4.)
This section introduces the basis for establishing a monitoring network. Key points
include:
Ozone monitoring objectives, which include determining regulatory compliance,
developing control strategies, and collecting data for health studies and air quality studies.
For both the 1-hour and 8-hour ozone standards, a minimum of two NAMS (National Air
Monitoring Stations) ozone monitors are required by 40 CFR 58, Appendix D, §3.4, for
urbanized areas with populations greater than 200,000. If a nonattainment area is
designated as serious, severe, or extreme for the 1-hour ozone standard, then up to five
monitors are required as a part of the Photochemical Assessment Monitoring Stations
(PAMS) program (40 CFR 58, Appendix D, §4.4).
Although 8-hour ozone monitoring seasons vary across the United States, in general, 8-
hour ozone monitoring seasons are slightly shorter than those established previously for
the 1-hour standard (based on a 6-year data review). See the "Guideline for Selecting
and Modifying the Ozone Monitoring Season Based on an 8-Hour Standard", EPA-
454/R-98-001, for further information on ozone monitoring seasons.
ES-3
-------
Macroscale Considerations in Monitor Siting (Section 5.)
This section focuses on identifying general locations within a metropolitan area to site
ozone monitors. Recommendations include:
Upwind monitors should be located along the trajectory of the morning prevailing wind
direction (PWD), and the downwind monitor should be located along the trajectory of the
afternoon PWD. Upwind monitoring will help establish background ozone and ozone
precursor concentrations (if PAMS) entering the geographic area, and downwind
monitors will help evaluate the peak ozone concentrations experienced within the
geographic area.
PWDs should be determined from wind rose plots of days where high concentrations of
ozone are measured, or days conducive to ozone formation. Specific meteorological
criteria based upon temperature, wind speed, and relative humidity can help identify days
conducive to ozone formation. These factors vary depending upon local conditions, and
the section describes how these criteria can be applied to different parts of the country.
For areas where ozone episodes are dominated by stagnation conditions and for which a
dominant PWD cannot be discerned, ozone monitors should be located along the major
axes of the emission sources within 10 miles of the urban fringe (i.e., no further than 10
miles beyond the outermost portion of the urban fringe).
For re-evaluating a network pursuant to the 8-hour running average concentration, the
potential for changes in the spatial distribution of exceedances needs to be evaluated.
Generally, 1-hour and 8-hour ozone concentrations are well correlated spatially.
However, since downwind sites are more prone to ozone exceedances under the 8-hour
NAAQS, the total geographic area monitored may need to be expanded to encompass the
expanded, downwind areas where these 8-hour ozone exceedances may occur.
Microscale Considerations in Monitor Siting (Section 6.)
This section provides guidance on determining specific monitor locations after the
general locations are determined (Appendix B provides references to supplemental materials
useful in developing a monitoring program). Key recommendations from this section include:
Probe inlets should continue to be placed 3 to 15 meters above ground level. (40 CFR 58,
Appendix E, §2.1)
Samplers should be separated from any obstruction such as a building by at least twice
the height that the obstruction protrudes above the sampling inlet. (40 CFR 58, Appendix
E, §2.3)
Monitors should be placed an adequate distance from roadways. The more traffic on a
ES-4
-------
road, the greater the recommended distance between the roadway and the monitor
location. For example, for roads with 10,000 or less vehicles per day, monitors should be
at least 10 meters away from the roadway; roads with 110,000 vehicles per day or more
should have monitors sited at least 250 meters away from the roadway. (40 CFR 58,
Appendix E, §2.5)
Inlet probes should be placed at least 20 meters from the drip line of trees to minimize
interferences. Large trees or stands of trees should be treated as large obstructions, e.g.,
as buildings, however, the probe must not be located less than 10 meters from the drip
line of such trees treated as obstructions. (40 CFR 58, Appendix E, §2.4)
"Saturation monitoring" can also be used to site monitors. Saturation monitoring
involves the deployment of numerous portable monitors on a short-term basis to quickly
develop a better understanding of how ozone concentrations vary across a geographic
area.
Summary (Section 7.)
The recommendations in Section 7. of this document provide real world examples to
illustrate how geographic areas should conduct their monitoring activities, answer common
questions, and provide practical "rules of thumb" for those engaged hi monitoring network
design. As a primer on ozone monitoring issues, this report touches on a number of key issues
and their implications for monitoring network design including: ozone formation and spatial
distribution; monitoring objectives, such as regulatory compliance and scientific research; big
picture geographic issues, such as the need for siting monitors upwind of an urban area; and
micro-scale issues, such as the need to avoid siting monitors too close to trees and buildings.
Readers with an interest in either developing a new ozone monitoring network, assessing the
efficacy of an existing network, or updating an existing network to monitor 8-hour ozone
concentrations should all find this report's contents to be of value. The recommendations are
organized in response to commonly asked questions such as:
How many monitors are required for a particular region?
How should monitors be sited to properly account for wind direction?
How should monitors be sited to avoid problems in the immediate vicinity of the monitor
(e.g., from roads, trees, buildings, etc.)?
Are existing 1-hour monitoring sites still useful for 8-hour monitoring* or should they be
moved? In particular, does an 8-hour standard mean that the geographic area monitored
needs to be enlarged, and if so, by how much?
How can monitor sites be evaluated to determine if they are located in appropriate
downwind locations to identify peak ozone concentrations?
Should 8-hour ozone monitoring be conducted during the same time of the year as 1 -hour
ozone monitoring?
ES-5
-------
How should monitor siting differ depending upon whether an area is an isolated urban
area, is part of an urban corridor which includes numerous metropolitan areas, or is an
urban area with multiple prevailing wind directions?
What PAMS requirements will be necessary in new ozone nonattainment areas for only
the 8-hour ozone NAAQS?
ES-6
-------
1. INTRODUCTION
This report provides background information for those responsible for designing and
implementing an ozone monitoring program. Its purpose is to update the 1978 U.S.
Environmental Protection Agency (EPA) report entitled: Site Selection for the Monitoring of
Photochemical Air Pollutants (EPA, 1978). This updated report, Guideline on Ozone
Monitoring Site Selection, is motivated by the need to help air quality agencies and others
incorporate 8-hour monitoring into their ozone monitoring program. Since the 1970s, ozone
monitoring programs have been designed to address the 1-hour ozone National Ambient Air
Quality Standard (NAAQS); however, on July 18,1998, the U.S. EPA promulgated an 8-hour
ozone NAAQS. This report assists those agencies considering how to design and implement
ozone monitoring in light of the need to collect 8-hour ozone concentration data. In addition,
knowledge of ozone photochemistry has improved in recent years and it is timely to revisit
monitoring network design issues in light of new scientific knowledge.
One of the first steps hi recognizing or diagnosing an air pollution problem is the
development of an adequate monitoring network to quantify pollution concentrations.
Specifically, there are five major objectives to air monitoring:
Determine the highest concentrations expected to occur hi the area covered by the
network.
Determine representative concentrations hi areas of high population density.
Determine the impact of specific sources or source categories on ambient pollution
concentrations.
Determine general background concentrations.
Determine the extent of air pollution transport into and out of an area.
Numerous decisions need to be made in designing a monitoring network, including:
How many monitors are needed?
What pollutants need to be measured?
Where should the monitors be placed?
During which part of the year is monitoring needed?
In the years since the EPA produced its 1978 guidance for designing monitoring networks
for photochemical pollutants (EPA, 1978), there have been significant advances in the
understanding of the chemistry of photochemical air pollution. Also, moving from a 1-hour daily
maximum ozone concentration-based NAAQS to one based on an 8-hour running average
concentration means different geographic areas may be affected. Such differences suggest that
changes to traditional monitoring networks should be considered.
The principal focus of this report is ozone, but a few remarks are made regarding NO
monitoring as it relates to ozone formation. EPA's 1978 document addressed monitoring
placement criteria for nonmethane hydrocarbons (NMHC), nitrogen dioxide (NO2), nitric oxide
(NO), and ozone (or oxidants, on which the standard was then based). There is no longer a
-------
NMHC standard, and there are presently (as of 1996) no areas in violation of the current NO2
NAAQS. Therefore, these pollutants are not addressed. It is worth noting that since 1991,
hydrocarbons that are considered precursors to ozone are now required to be measured hi
selected ozone nonattainment areas at sites called Photochemical Assessment Monitoring
Stations (PAMS). A separate document was prepared by the EPA to address the selection of
PAMS sites (EPA, 1994).
This report is organized into seven main sections:
1. Introduction (this discussion).
2. Important Characteristics of Ozone Formation: Provides the reader with a general
explanation of ozone formation chemistry, the importance of ozone precursor emissions, and
meteorological conditions conducive to ozone formation.
3. Spatial Distribution of Ozone Concentrations: Describes the spatial nature of ozone
concentrations and how these differ on a 1-hour and 8-hour basis. The section provides
example ozone concentration situations hi the western and northeastern United States.
4. Monitoring Objectives and Network Design: Discusses under what conditions to monitor in
neighborhoods versus urban or regional settings, monitoring for regulatory compliance versus
control strategy development, the number of required monitor sites, and the seasons during
which monitoring should occur.
5. Macroscale Considerations in Monitor Siting: Provides guidance for determining the
general location of monitors in a metropolitan area.
6. Microscale Considerations for Monitor Siting: Provides guidance for the specific placement
of monitors, once the general location has been identified.
7. Recommendations: Summarizes the key recommendations included in the report.
The report also includes references and two appendices: Appendix A, which contains
information on ozone formation, and Appendix B, which contains citations for additional
information resources of interest to those involved with monitor siting.
-------
2. IMPORTANT CHARACTERISTICS OF OZONE FORMATION
Ozone is not generally emitted by sources, but instead is formed in the atmosphere by a
series of complex chemical reactions between oxides of nitrogen and organic compounds. The
process of ozone formation is complex. However, a basic understanding of this chemistry is
necessary for predicting the spatial distribution of the ozone concentration within a metropolitan
or regional area, and for developing an ozone monitoring network. This section provides
background information on ozone formation and includes brief discussions on the following
(More detail on .the photochemical reactions contributing to ozone formation can be found in
Appendix A.):
Ozone formation chemistry in the troposphere
Sources and spatial distribution of ozone precursor emissions
Meteorological conditions conducive to ozone formation.
Figure 2-1 shows a general schematic diagram of ozone formation in the atmosphere.
This section describes these processes and their interactions. These principles are illustrated
further in Section 3., which describes the spatial and temporal distribution of ozone
concentrations.
2.1 OZONE FORMATION CHEMISTRY
2.1.1 Basic Principles
The basis for ozone formation is the photolysis of nitrogen dioxide (N02), by the
following reactions (Seinfeld, 1986):
NQ+hv—^->AD+0 (2-1)
O+Q+M—O^+M (2-2)
where hv represents photochemical energy from ultraviolet radiation (or a photon), k represents
a rate constant for the reaction of NO2 and hv and M represents N2 or O2 or another molecule
that absorbs the reaction's excess vibrational energy. Once formed, ozone is rapidly dissociated
by reaction with NO, as follows:
(2-3)
2-1
-------
Dry
transformation
RH + OH—»H,O + H«
«• + O, + M-»ROt + M
HO«| 4- NO NO, + RO
NO, + *v—»NO + O
O + Ot + M—»0t 4- M
Cloud
evaporation
i.
Air
concentration*
Wet
transformation
+ 0,
Figure 2-1. Schematic diagram of photochemical air pollution from
emission to deposition (NAS, 1991)
2-2
-------
The NO2 molecule is regenerated, and in the absence of other species a steady state is achieved
through reactions (2-1) through (2-3) in which the ozone concentration can be estimated by the
following relationship:
In the natural troposphere, these reactions normally result in a background ozone
concentration of 15 to 45 parts per billion (ppb) (Altshuller and Lefohn, 1996).
Emissions of volatile organic compounds (VOC) are key contributors to ozone
concentrations above normal background values. Most combustion and biogenic emissions of
oxides of nitrogen are emitted in the form of NO, although some of the NO in the combustion
gases are oxidized to NO2. In the atmosphere, VOC facilitate the oxidation of NO to NO2, thus
allowing for continued ozone production while reducing the destruction of ozone by NO.
VOC sources are both anthropogenic and natural in origin. Anthropogenic sources
include automobiles, chemical production, and other industrial activities. Many species of
vegetation including trees and plants naturally emit VOC. In the natural troposphere, there is a
sufficient amount of natural (or biogenic) VOC to oxidize some of the NO to NO2, which results
in the background ozone concentration discussed above. When additional VOC is added to the
atmosphere, a greater proportion of the NO is oxidized to NO2, resulting in greater ozone
formation. Additionally, anthropogenic sources of NO, including sources such as automobiles
and electric utility plants, result in greater levels of NO2 in the atmosphere, which is then
available for photolysis to NO and O (an oxygen atom) and, ultimately, NO and ozone.
Using the chemical principles elucidated above, EPA's Empirical Kinetic Modeling
Approach (EKMA) can be used to depict a simplistic but concise diagram for predicting the
effects of changes in VOC or NOX concentrations. Figure 2-2 shows a typical set of EKMA
ozone isopleths. The ozone isopleths depend on the VOC/NOX ratio. For low VOC/NOX ratios,
the system is VOC limited (i.e., reductions in VOC emissions will reduce the ozone
concentrations). Conversely, for high VOC/NOX ratios, the system is NOX limited (i.e.,
reductions in NOX emissions will reduce ozone concentrations). The NOx-limited region is
typical of locations downwind of urban and suburban areas, while the VOC-limited region is
typical of highly polluted urban areas.
2.1.2 Atmospheric Sinks for Ozone
Once in the atmosphere, ozone can be removed by wet and/or dry deposition. Ozone in
contact with the surface can be deposited or adsorbed onto vegetation such as trees and other
plants. This scavenging includes ozone uptake into vegetation which may also result in
detrimental effects to the plant or tree. Also, ozone can be scavenged by precipitation, and can
be transformed in the aqueous phase by reaction with other atmospheric constituents such as
hydrogen peroxide.
2-3
-------
0.28 r
0.24
0.20
0.16
§ 0.12
0.08
0.04
0
03 (ppm) = 0.08 0.16 0.24
0.40
-VOC
LIMITED,
VOC _Jl
NO. ~ 1
0.30
.12/0.20//0.28U.32
0 0.2 0.4 0.6 0.8 1-0 1.2 1.4 1.6 1.8 2.0
VOC (ppmC)
Figure 2-2. Typical ozone isopleths used in the EPA's
Empirical Kinetic Modeling Approach (EKMA). The NOX-
limited region is typical of locations downwind of urban
and suburban areas, whereas the VOC-limited region is
typical of highly polluted urban areas (NAS, 191).
2-4
-------
2.2 OZONE PRECURSOR EMISSIONS
2.2.1 Sources of Ozone Precursor Emissions
Table 2-1 summarizes the total anthropogenic emissions of VOC and NOX in the United
States for 1994. The dominant activity for producing NOX is combustion processes, including
industrial and electrical generation processes, and mobile sources such as automobiles. Mobile
sources also account for a large portion of VOC emissions. Other industries that process organic
materials, such as the chemical industry or others that use solvents, also account for a large
portion of the VOC emissions. Although not included in Table 2-1, the EPA also estimates the
annual biogenic emissions. In 1996, for example, estimated VOC biogenic emissions were 29
million short tons (EPA, 1997), which is higher than the anthropogenic emissions. Biogenic
VOC emissions include the highly reactive compound isoprene. However, it must be noted that a
large proportion of the biogenic VOC emissions are hi forested and vegetative areas, where these
emissions may not impact urban ozone formation. In contrast to biogenic VOC, which is large
relative to anthropogenic emissions, biogenic NOX emissions (2000 short tons) are much smaller.
2.2.2 Spatial Distribution of Emissions
Anthropogenic VOC and NOX emissions are generally concentrated around urban areas,
due to the nature of the emission sources (e.g., mobile sources [cars, trucks, buses], area sources
[including non-road mobile sources], and industry being located in the populated areas of the
country). The resulting spatial emission patterns reflect this urban-centered orientation. For
example, Figures 2-3 and 2-4 show the VOC and NOX emissions (both anthropogenic and
biogenic) for 1990 in the northeastern United States. The highest emission densities are near the
urban centers of Boston, New York City, Philadelphia, and Baltimore. The emission densities
drop rapidly expanding outwards from the urban centers. As will be discussed hi more detail
later, ozone formation is generally highest downwind of urban centers as a result of the
photochemical reaction of ozone precursor emissions in these areas.
2.3 METEOROLOGICAL CONDITIONS CONDUCIVE TO OZONE FORMATION
The chemistry of ozone formation is complex and dependent on numerous variables.
Similarly, there are a variety of meteorological variables that influence ozone formation.
Although changes in daily emissions can affect daily ozone concentrations, it is the daily
variation in meteorological parameters that best explain the daily variations of ozone
concentrations. The meteorological conditions which are conducive to ozone formation include:
Days -when solar radiation is high: Solar radiation is an important factor because of
ozone formation reactions (see Equation 2-1). Solar radiation is highest during cloudless,
summer days.
2-5
-------
Table 2-1. Summary of Anthropogenic VOC and NOX emissions in the United States during 1996
(EPA, 1997).
Source Type
*
Fuel Combustion, Electric
Utility
On-Road Vehicles
Fuel Combustion,
Industrial
Non-Road Sources
Fuel Combustion, Other
Miscellaneous
Other Industrial
Processes
Chemical & Allied
Product Manufacturing
Petroleum & Related
Industries
Waste Disposal &
Recycling
Metals Processing
Solvent Utilization
Storage & Transport
Total
MO^lmi&sioas
Emissions ,
f^oBsan<3sfeoittats)
6034
7171
3170
4610
1289
239
403
159
110
100
98
3
6
23^92
Percentage
• of Total
25.8.
30.7
13.6
19.7
5.5
1.0
1.7
0.7
0.5
0.4
0.4
0.01
.03
100%
VOC Emissions
, V JSmlsssiojis
^thousand short tons)
45
5502
208
2426
822
601
439
436
517
433
70
6273
1312
19,084
P«reestag
* of Total
0.2
28.8
1.1
12.7
4.3
3.1
2.3
2.3
2.7
2.3
0.4
32.9
6.9
100%
2-6
-------
L'Z
a,
£
31?
if
§*
11
»• iv
Is
1
Latitude (dedmal degrees)
if??
- r
-------
Highest NOx Emissions Are in the Urban Areas
to
00
1800
1800
1400
1200
1000
800
800'
b
too
90
25
10
0
-400
Longitude (decimal degrees)
Figure 2-4. Sj^ distribution of the anthropogenfc^ EmiuioiK data arc from the
1990 EPA Interim Inventory (EPA, 1992). Emissions are on acountywide basis and assigned to flic county centrokL
-------
Days with low wind speed: Low wind speeds lead to poorer dispersion and generally result in
increased concentration of ozone and ozone precursors within a smaller area. However, areas
affected by long-range transport can experience high ozone concentrations during periods of
moderate wind speeds.
Days when mixing heights are low: Strong temperature inversions can cause low mixing heights
to occur, which generally lead to increased concentrations of ozone and ozone precursors in a
shallow vertical layer.
Days with low relative humidity: Since days with high relative humidity or precipitation are likely
to be associated with cloudy skies or hazy conditions, which reduce temperatures and solar
radiation, and result hi decreased ozone concentrations, days with low relative humidity are more
conducive to ozone formation.
The presence of ozone-conducive meteorological conditions can be indicated by the synoptic-scale
meteorology (i.e., meteorological conditions over a broad regional area). In particular, ozone
concentrations generally accumulate during periods with slow-moving, high-pressure weather systems
(NAS, 1991) for the following reasons:
High-pressure systems are characterized by widespread sinking of air through most of the
troposphere, which warms the subsiding ah". This tends to make the troposphere more stable and
less conducive to mixing, which would disperse ozone and ozone precursor emissions.
Temperature normally decreases with altitude hi the troposphere, which allows warm air parcels to
rise from the surface and mix through the upper troposphere. However, the subsidence of air
creates a temperature inversion, which effectively results hi a "warm air lid" placed on top of
cooler air. The cooler air near the surface fails to mix with the warmer layer above, and there are
higher concentrations of ozone and ozone precursors near the surface.
The light winds associated with high-pressure systems allow for ozone and ozone precursors to
accumulate near the emission sources.
The clear skies associated with high-pressure systems result in increased solar radiation at the
surface, which is favorable for photochemical processes.
Ozone-conducive meteorological conditions normally occur during the summertime in most of the
United States, except in the southern or "sunbelt" areas where ozone is not restricted to the summer
season. Consequently, ozone monitoring is not necessary during the entire year in most areas.
Determining the precise monitoring requirement for the ozone season in each metropolitan area of the
United States is beyond the scope of this report. However, the periods during which ozone monitoring is
likely to be needed for regulatory compliance in each area of the country are outlined in Section 4, and in
the Guideline for Selecting and Modifying the Ozone Monitoring Season Based on an 8-Hour Ozone
Standard, EPA-454/R-98-001.
2-9
-------
3. SPATIAL DISTRIBUTION OF OZONE CONCENTRATIONS
When designing a monitoring network for measuring ozone, it is important to understand the
spatial distribution of ozone in the monitoring area. However, the complex process of ozone formation
can lead to complex distributions of ozone within a metropolitan area. Several ozone formation processes
influence this spatial distribution, including:
Proximity to fresh combustion emissions: NOX from fresh combustion emissions is normally in the
form of NO, which destroys ozone by chemical reaction (sometimes referred to as NO titration).
This results in low ozone concentrations relatively near combustion sources.
Distance from primary emission sources: The process of ozone formation in the atmosphere takes
several hours. The maximum concentration normally forms several miles downwind of the
primary emission sources because the conversion of precursors requires several hours.
Direction and wind speeds through the city: The direction and speed of winds through the
maximum emissions area will determine where the ozone and ozone precursors are transported.
Furthermore, 1-hour and 8-hour ozone concentration averaging times have different spatial
implications. For example, high 8-hour ozone concentrations can be observed over a broader geographic
area than high 1-hour ozone concentrations. To illustrate the differences between 1-hour and 8-hour
spatial concentration patterns, the next two sections provide examples using observed ozone data from the
San Francisco Bay Area (in California) and the northeastern United States. Much of the data and plots for
these examples are based on an analysis by Reiss et al., 1995. All of the data portrayed in this report as
representing exceedances of the 1-hour and 8-hour thresholds were based on the following "rules of
thumb": ozone concentrations equal to or greater than 0.085 ppm over an 8-hour period were considered
an exceedance of the 0.08 ppm 8-hour threshold; all concentrations equal to or greater than 0.125 ppm
over a 1-hour period were considered an exceedance of the 0.12 ppm 1-hour threshold. These conventions
comply with 40 CFR 50, Appendix I, §2.3, which stipulates the following:
The number of significant figures in the level of the standard dictates the rounding
convention for comparing the...ozone concentration with the level of the standard. The
third decimal place of the computed value is rounded, with values equal to or greater than
5 rounding up.
3.1 EXAMPLE 1-HOUR AND 8-HOUR OZONE PATTERNS
3.1.1 San Francisco Bay Area
The San Francisco Bay Area has a dense monitoring network encompassing upwind, source, and
receptor areas. Thus, it can be used to provide a good example of the spatial distribution of ozone within
and downwind of a major metropolitan area. The largest urban centers are the cities of San Francisco,
Oakland, and San Jose. Generally, the city of San Francisco, the areas south of the city (the San Jose
area), and the areas east of the San Francisco Bay are urbanized. Most of the ozone precursor emissions
are associated with the urban areas and the industrialized eastern shoreline of the San Francisco Bay and
-------
Sacramento River Delta. The areas north of San Francisco (Marin, Sonoma, and Napa counties) are
relatively rural and produce fewer ozone precursor emissions. Figure 3-1 depicts major cities in the San
Francisco Bay Area. The meteorology of the region is dominated by coastal westerly winds and water-to-
land airflows. Therefore, much of the daytime emissions in the urbanized portions of the San Francisco
Bay Area are transported to the east and south.
Selected monitoring sites that were active between 1992 and 1995 in the San Francisco Bay Area
and downwind areas are shown in Figure 3-2. Table 3-1 lists the sites (city and county) corresponding to
the numbers in Figure 3-2. Figures 3-3a and 3-3b, and 3-4a and 3-4b, show the number of exceedances
for 1991-1995 of a 0.12 ppm 1-hour threshold, and a 0.08 ppm 8-hour threshold. Figures 3-3a and 3-4a
show contours using the number of exceedances to mark each contour interval; Figures 3-3b and 3-4b
show the number of exceedances at specific monitors. The figures show information for the San
Francisco region and nearby downwind areas. For both the 1-hour and 8-hour thresholds, most of the
exceedances are in the eastern and southern portions. There are no exceedances in the city of San
Francisco, and there are very few exceedances in the counties north of the city. Three important factors
contribute to the clean air in San Francisco: (1) the upwind region is the Pacific Ocean, and thus the air
blowing into the San Francisco area is clean to begin with; (2) San Francisco's automobile and truck
traffic contributes NO emissions that can titrate ozone; and, (3) except for the area immediately south of
San Francisco, nearby areas produce relatively few emissions, so there are fewer emissions that contribute
to ozone formation in the immediate vicinity of the city (counties north of San Francisco are largely rural;
the area immediately to the east of the city is the San Francisco Bay). In contrast to San Francisco, San
Jose to the south and east experiences higher ozone concentrations. Key differences include: (1) winds
blowing into San Jose are already polluted with neighboring areas' emissions (for example, the southern
portion of the San Francisco Bay Area, near San Jose, provides transported pollutants from upwind
portions of the Bay Area including San Francisco and East Bay cities such as Oakland) and (2) San Jose is
bordered by heavily developed areas with significant emissions activity.
In general, Figures 3-3 (a and b) and 3-4 (a and b) show that the distribution of the 1-hour and 8-
hour exceedances are relatively similar. However, at several of the downwind sites (e.g., Bethel Island
and Livermore) there were considerably more 8-hour exceedances compared to 1-hour exceedances.
While these sites have some local emissions that may contribute to their ozone concentrations, ozone
transport from the San Francisco Bay Area is likely. This simple analysis shows that the area of influence
of an urban area such as San Francisco may be greater for the 8-hour standard compared to the 1-hour
standard. Therefore, States should consider expanding ozone monitoring networks for the 8-hour standard
to larger geographic areas during the annual network review required by 40 CFR 50.20 (d).
The effects of downwind transport can be graphically illustrated with the ozone data from the San
Francisco Bay Area. Downwind sites can be more prone to exceedances of a 0.08 ppm 8-hour threshold
due to transported pollutants that increase ozone concentrations in the early evening hours. This increased
exceedance count can be explained in part by the diurnal ozone concentration profile of the site. For
example, Figure 3-5 shows a comparison of the diurnal profile for Bethel Island (site #8) and Livermore
(site #1) for July 27, 1995. Livermore exceeded the 1-hour standard 10 times during 1992-1995, and
3-2
-------
U)
u>
M*-
I - 1 - 1 - 1
I - 1 - 1 - 1 _ I
-li*
Loagitufe (decimal degree*)
Figure 3-1. Selected cidet in the San Francisco Bay Area and downwind regions.
-------
u>
Figure 3-2. Selected monitoring sites that operated in the San Francisco Area and downwind regions ben
1992-1995.
-------
Table 3-1. Monitoring sites in the San Francisco Bay Area and downwind regions.
The number corresponds to the label used in Figure 3-2.
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
City
Livermore
Oakland
San Leandro
Fremont
Livermore
Concord
Richmond
Bethel Island
Pittsburgh
San Rafael
Point Reyes
Napa
San Francisco
Tracy
Redwood City
Gilroy
San Jose (4th Street)
Los Gates
Mountain View
San Jose (Carlos Street)
San Jose (Piedmont Rd.)
San Martin
Davenport
Santa Cruz (Airport Blvd.)
Scotts Valley (Vine Hill School Rd.)
Scotts Valley (Scotts Valley Dr.)
Santa Cruz (Bostwick Ln.)
Fairfield
Vallejo
Vacaville
Santa Rosa
Healdsburg
Sonoma
Davis
County
Alameda
Alameda
Alameda
Alameda
Alameda
Contra Costa
Contra Costa
Contra Costa
Contra Costa
Marin
Marin
Napa
San Francisco
San Joaquin
San Mateo
Santa Clara
Santa Clara
Santa Clara
Santa Clara
Santa Clara
Santa Clara
Santa Clara
Santa Cruz
Santa Cruz
Santa Cruz
Santa Cruz
Santa Cruz
Solano
Solano
Solano
Sonoma
Sonoma
Sonoma
Yolo
3-5
-------
Highest Number of 1-hour Exceedances Are Within the Eastern and Southern Portions of the San Francisco Bay Area
u>
ON
120
118
16
14
112
110
Is
I HfUtJfr11** ("*•"""•*'**l[f"*)
Figure 3-3a. Exceedances of the 0.12 ppm ozone threshold between 1992-1995 in the San Frandico Bay Area .and
downwind regions. The Kale (0-20) represent! the number of exceedances over the four-year period.
The plot displays concentration differences among Bay Area and downwind monitor locations; the plot
excludes several Sacramento sites around the northeast corner of the map.
-------
Highest Number of 1-hour Exceedances Are Within the Eastern and Southern Portions of the San Francisco Bay Area
Figure 3-3b. Exceedances of the 0.12 ppm ozone threshold between 1992-1995 in the San Francisco Bay Area and
downwind regions. The size of the tirdes are proportional to the Bomber of exceedances (which is shown
above die circle). The plot displays data for the San Francisco Bay Area and downwind monitor locations;
the plot excludes several Sacramento skes around die northeast corner of the map.
-------
Downwind Areas Experience More 8-hour Threshold Exceedances Than 1-hour Threshold Exceedances
UJ
oo
1 20
1 18
16
14
1 12
'o
-rfw -rtu -li* -ti.4
Longitude (decimal degree*)
•«* -«iji
Figure 3-4a. Contour plot of the exccedances of the 0.08 ppm 8-hour threshold between 1992-1995 in the San Francisco
Bay Area and downwind regions. The scale (0-20) represents the number of exccedances over the four year period.
For the purposes of displaying differences betweensites within the influence of emissions from the Sao Francisco
Area, several Sacramento sites around the nortbeas corner of the map are not shown.
-------
Downwind Areas Experience More 8-hour Threshold Exceedances Than 1-hour Threshold JSxceeaances
u>
-ixu -m« -m* -m*
Longitude (decimal degree*)
-m»
Figure 3-4b. Exceedances of the 0.08 ppm 8-hour threshold between 1992-1995 in the San Francisco Bay Area and
downwind regions. The size of the circles are proportional to the number of exceedances (which is sbown
above the circle). The plot displays data for the SanFrantisco Bay Area and downwind locations; the plot
excludes several Sacramento sites around the northeast comer of the map.
-------
Livermore exceeds both the 1-hour and 8-hour standards, while
Bethel Island only exceeds the 8-hour standard
Bethel
Island
Livermore
Hour of Day (PST)
Figure 3-5. OZOIK concentration profiles for Bethel Jilaud and Livermore for July 27,1995. Livermore •
exceed* both the 0.12 ppro l-Jwur and O.Q8 |)t>m 8-hour Qireshokto. Bethel Island, tt more
rural site than Liveimorc. experiences lower mid-day ozone concentrations. An early evening
rise in Bethel Island's ozone concentrations reflects transported pollutants from upwind areas.
Transport on (his day almost causes Bethel bland (o exceetf (be 1-hour threshold; transport
combined with locally generated ozone causes an 8-hour exeeedance.
-------
exceeded the 0.08 ppm 8-hour threshold 20 times. By comparison, Bethel Island only exceeded the 1-hour
standard one time during 1992-1995, but it exceeded the 0.08 ppm 8-hour threshold nine tunes. The
diurnal profiles for both sites are very similar in the morning hours until about 10:00 a.m. After 10:00
a.m., the concentration at the Livermore site continues to rise rapidly, while the concentration at Bethel
Island levels off. After about 2:00 p.m., the concentration at Livermore declines sharply, while the Bethel
Island concentration actually rises to its peak at 6:00 p.m. For the remainder of the evening, the
concentration at Bethel Island is higher than at Livermore. Livermore exceeded both the 1-hour and 8-
hour standards on this day, while Bethel Island only exceeded the 8-hour standard.
Both Livermore and Bethel Island receive transported ozone and ozone precursors from the
urbanized areas to the west. However, the difference is likely explained by the difference in urbanization
between Livermore and Bethel Island. Neither are very urban compared to San Francisco and San Jose,
but Livermore is a moderately sized community near a major Interstate, which connects with Oakland;
Bethel Island is a more rural site and not near any major roadways. Livermore likely has some ozone
buildup in the afternoon as the result of local emissions which add to the transported ozone and ozone
precursors, and the rush hour traffic which provides fresh NO to quickly titrate the ozone concentration
after the peak is reached in the late afternoon. Bethel Island likely receives a dispersed plume of ozone
late in the afternoon, adding to ozone formed locally. Unlike Livermore, there is no late afternoon source
of fresh NO to titrate the ozone; therefore, the ozone concentration lingers for a longer number of hours
than at Livermore.. The more elongated peak and the lingering of the ozone through the early evening at
Bethel Island make this site more prone to 8-hour exceedances than Livermore. Therefore, with an 8-
hour form of the standard, special attention should be given to rural sites receiving transported ozone
without fresh NO in the late afternoon to titrate the ozone.
Rural or suburban areas that do not exceed 1-hour ozone thresholds can exceed 8-hour ozone
thresholds. One example of this is the ozone concentrations monitored at Woodland, California.
Woodland is a rural site located northeast of the San Francisco Bay Area, not far from the Sacramento
metropolitan area. Figure 3-6 shows ozone concentrations at Woodland on August 20,1995. This site
did not exceed the 0.12 ppm 1-hour ozone threshold from 1991 through 1995, but exceeded the 0.08 ppm
8-hour threshold 10 times during this period. The plot shows the concentration was above 80 ppb for 8
hours, without ever going above 100 ppb. Thus, this site has a steady ozone concentration throughout the
afternoon, without reaching a high peak. This type of site is more likely to exceed a standard based on an
8-hour running average set at 0.08 ppm rather than the 1-hour peak concentration of 0.12 ppm.
3.1.2 Northeastern United States
The northeastern United States includes many major metropolitan areas and has been designated as
an ozone transport region (OTR). In this section, spatial ozone patterns are presented for the entire OTR,
including many metropolitan areas as well as rural areas which are influenced by ozone transport. The
OTR in the northeastern United States extends from Virginia to Maine and contains several large
metropolitan areas, including Washington, D.C., Baltimore, Philadelphia, New York City, and Boston.
Figure 3-7 presents a map of the OTR showing the location of all of the major cities. The close proximity
3-11
-------
Areas can exceed 8-hour thresholds without exceeding 1-hour
thresholds.
120
Woodland
Hour of Day (PST)
Figure 3-6. Ozone concentration profile for Woodland, California (northeast of the San Francisco Bay Area)
on August 20,1995. Under the conditions experienced on this day, Woodland exceeded the
0.08 ppm 8-hour threshold without exceeding the 0.12 ppm 1-hour threshold.
-------
Areas of NE U.S. within a MSA/CMSA
(Census Bureau statistics as of July 1, 1994.)
MSA/CMSAs with populations > 1,000,000 are labeled.
Buffalo-Niagara Falls MSA
Boston-Worcester-Lawrence CMSA
ice-FaU River-Warwick MSA
:-No. New Jersey-Long Island CMSA
Wilmington-Atlantic City CMSA
Itimore CMSA
Figure 3-7. Major cities in the Northeastern United States.
-------
of these cities along a southwest to northeast line and a generally southerly or southwesterly flow in the
summertime ozone season results hi a complex interaction of ozone and ozone precursors in the region.
Figures 3-8 and 3-9 depict the spatial pattern of exceedances of the 0.12 ppm 1-hour ozone
threshold for 1988 and 1995. These years were chosen for display because in the northeast and
nationwide, 1988 was the year with the highest number of ozone exceedances in the 1980s and 1990s, and
1995 is the most recent year (as of this writing) with available ozone data. Figures 3-10 and 3-11 show
contour plots of the number of exceedances of the 0.08 ppm 8-hour threshold for 1988 and 1995. For both
the 1-hour and 8-hour indices, there are a high number of exceedances hi the Baltimore- Washington D.C.
area, Philadelphia, and New York City. Areas such as Maine and western Pennsylvania also recorded a
significant number of exceedances. The exceedances hi Maine are likely the result of transport from the
metropolitan areas of Boston and New York. Note that there are significantly more exceedances of the 8-
hour standard than the 1 -hour standard hi the rural areas.
Husar (1996) conducted a similar analysis for the Ozone Transport Assessment Group (OTAG)
region, which includes 37 eastern states. In Husar (1996) the spatial pattern of exceedances of the 0.12
ppm 1-hour threshold and the 0.08 ppm 8-hour running average were compared from 1991-1995 (June
through August). Over the entire OTAG region, the mean 8-hour average was 86 percent of the 1-hour
daily maximum. Husar (1996) concluded that switching from a 0.12 ppm 1-hour NAAQS to a 0.08 ppm
8-hour NAAQS would yield an increase in exceedances over the industrial midwestern states from Illinois
to Pennsylvania and a decrease of exceedances around Houston and New York City. However, the overall
exceedance pattern of the 1-hour and 8-hour indices were very similar. Additionally, Husar (1996)
examined the correlation between the 1-hour and 8-hour concentrations among monitoring sites. The
correlation coefficient was extremely good (r2=0.96). Husar (1996) shows that there are only subtle
differences in the overall exceedance patterns for the 1-hour and 8-hour forms of the standard.
To further examine the subtle differences between the 1-hour and 8-hour forms, contour plots of
the ratio of the 8-hour maximum ozone concentration and the 1-hour maximum ozone concentration were
computed for several high ozone concentration days. The significance of the 8-hour/l-hour ratio is that
higher values tend to indicate that a site is more prone to 8-hour exceedances, and lower values tend to
indicate that a site is more prone to 1-hour exceedances. Figures 3-12 and 3-13 show plots of this ratio
for two regional ozone episodes in the northeastern United States (i.e., ozone exceedances were observed
throughout the OTR on June 15, 1988, and July 20,1991). Typically, the 8-hour/l-hour ratio is lower in
the center of metropolitan areas. The figures show low ratios in Washington D.C. and New York City
3-14
-------
Exceedances of the 1-hour Standard in 1988 Were Predominantly in the Urban Regions
45.0
44.0-
43.0-
42.0-
4UO-
40.0-
30 jO-
380
•80.0 -74.0 -78.0 -7T.O
-75.0 -74.0 -73.0 -72.0 -7J.O -70.0 4OO
122
I20
I1*
lie
114
112
r
-lo
•J
Figure 3-8. Number of site exceedaoces of the 1 hour 0.12 ppm ozone threshold (greater than or equal to 125 ppb) hi the Northeast OTR
during 1988. Monitoring sites are represented by the black dots.
-------
9I-£
Latitude (decimal degrees)
a
'!
M •»
S5
!?
f
8
5
5-
-------
Ll~£
Latitude (decimal degrees)
f
3
U)
o
a
o
•••>
r+
00.
t/3
i
I
9
I
I
I
I
s
-------
8l-£
Latitude (decimal degrees)
A
•n
t
3
UJ
Sz
o e
P
3 O
OP i-ti
jn g.
55
3
3
*!
II
g. g
B. OO
M
£
9
5'
f
z
&
oo
I
5
o.
I
s
en
1
O
t
I
3
I
CD
»»
O
-------
OJ
Co
Differences Between the 1-hour and 8-hour Concentrations are Greater at the Urban Sites (1991)
45.0
M.O
••1.0
40.0
Longitude (decimal degrees)
Figure 3-12. Ratio of the concentration of the 0.08 ppm 8-hour running average standard and the 1 hour 0.12 ppm ozone threshold
for the ozone episode of July 20, 1991. Higher values tend to indicate a site is more prone to 8-hour exceedances;
lower values indicate the site is more prone to 1-hour exceedances.
-------
03-e
Latitude (decimal degrees)
s{
1
i
n
§
t
(D
I
g
cn
I
-------
with the ratio increasing with distance from the city centers. This effect is likely a result of NO titration of
ozone during the morning and afternoon rush hours, which prevents sustained high concentrations of
ozone in the metropolitan areas. Downwind sites away from cities receive transported ozone and ozone
precursors from cities. There is less NO to titrate the ozone in the rural and suburban areas outside of the
cities. For downwind sites that are far from any cities, the transported ozone plume is likely to be spatially
dispersed, resulting in flatter diurnal profiles. Areas such as western Pennsylvania, upstate New York, and
Maine had high values of the ratio, illustrating this effect. These are areas which receive transported
ozone (likely from dispersed plumes) and have low NO concentrations available for titration. This effect
points to the increased importance of rural ozone monitoring outside the urban/suburban core for
measuring 8-hour concentrations as compared to 1-hour values.
3.2 SUMMARY AND CONCLUSIONS
The following general observations can be made from the analysis of the San Francisco and
northeastern United States ozone concentrations:
The locations of the exceedances of the 0.12 ppm 1-hour and the 0.08 ppm 8-hour average ozone
thresholds were generally similar, with some exceptions.
The locations of the maximum concentration during exceedances were generally similar for the 1 -
hour and 8-hour thresholds within the San Francisco Bay Area, however, far downwind sites were
more prone to 8-hour exceedances relative to 1-hour exceedances. Therefore, the geographic
scope of ozone monitoring for an urban area may need to be expanded for the 8-hour threshold.
Differences between the 1-hour and 8-hour ozone values among sites can partially be explained by
the ozone diurnal profile at the sites. For example, downwind sites which receive transported,
dispersed ozone plumes and have low NO concentrations available to titrate ozone will have flatter
diurnal concentration profiles and are more prone to 8-hour exceedances, relative to 1-hour
exceedances.
The ratio of the 0.08 ppm 8-hour running average to the 1-hour maximum concentration increased
with distance from the center of the major cities on episode days. In other words, as you move
away from the center of major cities, the outlying areas become more prone to exceeding the 0.08
ppm 8-hour threshold rather than experiencing peak 1-hour concentrations. Some of the highest
values of this ratio were observed in areas outside the urban/suburban core.
3-21
-------
3-22
-------
4. MONITORING OBJECTIVES AND NETWORK DESIGN
4.1 INTRODUCTION
EPA requires different monitoring efforts depending upon an area's nonattainment status. In
general, the more polluted and populous a region is, the more extensive the required monitoring. For
ozone, there are two monitoring network systems: (1) the State and Local Air Monitoring Stations
(SLAMS), and (2) the National Air Monitoring Stations (NAMS). The objectives of the SLAMS network
are (CFR, 1997):
Determine the highest concentrations expected to occur in the area covered by the network.
Determine representative concentrations in areas of high population density.
Determine the impact on ambient pollution levels of significant sources or source categories.
Determine general background concentration levels.
The SLAMS stations are used to determine compliance with the ozone NAAQS. The EPA does not
specify the number of required stations, except to prescribe a minimum number of stations . States and
localities are given flexibility in determining the size of their network based on data needs and available
resources.
The NAMS stations are selected from the SLAMS stations. Thus, the NAMS network is a subset
of the SLAMS network. Areas to be monitored must be selected based on urbanized population and
pollutant concentrations. Accordingly, there are two categories of NAMS monitors: category (a) stations
located in areas of expected maximum concentrations, and category (b) stations which combine poor air
quality with a high population density but are not necessarily located in an area of expected maximum
concentrations (sometimes referred to as the maximum exposure monitor). The NAMS monitors are
designed to provide data for national policy analyses and trend analyses, and for providing the public with
information about the air quality in major metropolitan areas. Therefore, it is intended that these stations
remain at the same location. The criterion for designating NAMS sites is any urban area with a population
over 200,000. A minimum of two NAMS ozone monitors are required in each of these urban areas; this
requirement remains unchanged as a result of the promulgation of the 8-hour ozone NAAQS in 1997.
In addition, serious, severe, and extreme ozone nonattainment areas (for the 1 -hour ozone
NAAQS) are required to implement enhanced ozone monitoring. This enhanced monitoring is called the
Photochemical Assessment Monitoring Stations, or PAMS program. PAMS requirements are extensive
and include requirements for the monitoring of an array of ozone precursors and meteorological
parameters. At this time, EPA has not extended the PAMS requirements to new ozone nonattainment
areas that violate only the 8-hour NAAQS. Appendix B includes PAMS references for readers seeking
PAMS guidance materials.
This section describes the issues that determine what type of monitoring network is appropriate for
4-1
-------
a given metropolitan area. Specific topics covered include:
Monitoring principles involving localized versus regional scale monitoring.
Monitoring objectives, such as the need to demonstrate compliance with federal air quality
standards, and how they relate to monitor placement.
Seasonal monitoring requirements.
4.2 PRINCIPLES INVOLVING LOCALIZED VS. REGIONAL MONITORING
As a secondary pollutant, ozone requires appreciable formation time; longer time periods allow
precursor emissions to distribute more uniformly across a region, and thus allow ozone concentrations to
develop more uniformly across subregions and even large-scale regions. This does not imply that ozone
concentrations do not vary across a metropolitan area, it simply means that the gradient in ozone
concentrations is not as great as some other pollutants that derive directly from emission sources. For
example, carbon monoxide can reach significantly high concentrations in localized "hot spots" such as
busy intersections. High ozone concentrations generally occur over larger areas.
In contrast to ozone concentrations, ozone precursor emissions can vary significantly in small
areas. Monitoring, depending upon whether it is done for ozone or emissions precursors, needs to reflect
an understanding of the spatial scale over which precursor emissions and ozone concentrations occur.
Given the importance of large-scale transport in the ozone formation process, meteorological conditions
are particularly important to the site selection process. For monitoring secondary pollutants, one must
identify areas generally downwind of the primary pollutant sources during periods conducive to formation
of the secondary pollutant. It is important to consider the winds, in combination with the length of time it
takes ozone to form, and the locations of the major sources of the reactants. These factors will be useful
for determining where in an urban region the maximum ozone concentration occurs. Since the mixing of
the ozone precursors will occur over a large volume of air, the monitoring of small-scale variability is
usually not necessary for ozone.
Given these principles, four types of spatial scales can be defined for ozone and ozone precursor
monitoring (40 CFR Part 58, Appendix D; CFR, 1997):
Middle-Scale: Measurements at this scale represent conditions close to the sources of VOC and
•NOX emissions such as roads where it would be expected that suppression of ozone concentrations
would occur. Trees may also have a strong scavenging effect on ozone concentrations and may
tend to suppress ozone concentrations in their immediate vicinity. Measurements at these stations
would represent conditions over relatively small portions of the urban area.
Neighborhood: Measurements in this category represent conditions throughout some reasonably
homogeneous urban subregion, with dimensions of a few kilometers. These types of stations are
useful for the assessment of health effects because they represent conditions in areas where people
4-2
-------
work and live. These sites, if also monitoring for ozone precursors such as VOC or NOX, are
useful for understanding sources of ozone precursors.
Urban: This measurement scale is useful for quantifying concentrations over very large portions
of a metropolitan area with dimensions of 50 or more kilometers. Such measurements can be used
for determining trends, and designing area-wide control strategies. The urban-scale sites can also
be used to measure high concentrations downwind of the area having the highest precursor
emissions.
Regional: This scale of measurement can be used to represent concentrations over large portions
of a metropolitan area and even larger rural areas with dimensions of as much as hundreds of
kilometers. Data from such sites can be useful for assessing the ozone that is transported into or
out of an urban area. Also, data from such stations can be useful for quantifying the amount of
ozone and ozone precursors that cannot be reduced by control strategies within that urban area.
4.3 MONITORING OBJECTIVES AND SITE SELECTION
The data from an ozone monitoring network serve a myriad of purposes. Air quality program
managers use the aerometric data to determine compliance with the ozone NAAQS. When an area
violates the NAAQS, the EPA designates the area nonattainment. The Clean Air Act requires that these
nonattainment areas submit plans with sufficient control measures adopted to achieve attainment with the
standards. Some metropolitan areas use ozone monitoring networks to warn residents when unhealthy
ozone concentrations are expected. A detailed understanding of the spatial and temporal distribution of
ozone is needed for researchers that develop models to simulate the atmosphere. These models can be
used by regulators in the development of control strategies or by researchers trying to better understand
atmospheric chemistry, meteorology, and the relationship between precursor emissions and ozone
formation. Yet another purpose of the ozone monitoring network is for human exposure assessment.
Researchers conducting health studies need to estimate the exposure of residents to ozone and other
pollutants in both high and low concentration areas. Exposure is generally best estimated by sites close to
where people live, work (particularly for outdoor workers), and play (e.g., school playgrounds). Ozone
monitoring networks also provide data useful for evaluating the exposure experienced by urban vegetation
or nearby agricultural/natural areas.
The EPA (CFR, 1997) lists the major objectives of ozone monitoring:
Determine the highest concentrations expected to occur in the area covered by the network.
Determine representative concentrations in areas of high population density.
Determine the impact on ambient pollution levels by significant sources or source categories.
Determine background levels.
This suggests four major types of monitoring sites: (1) maximum downwind concentration sites, (2)
maximum exposure sites, (3) maximum emissions sites, and (4) upwind sites. These site types are
summarized in Table 4-1. Additional monitoring objectives are:
4-3
-------
4-4
-------
Table 4-1. Ozone and ozone precursor monitoring site types with corresponding monitoring objectives.
Type of Site
Relevant Pollutants Monitoring Objective'
Spatial-Scale Notes
Maximum Downwind
Concentration
Ozone
Regulatory
Compliance
Urban to
Regional
This site is required as part of the NAMS network and is
designed to measure the maximum ozone concentration
due to an urban area.
Maximum Exposure
Maximum Emissions
Upwind
Exposure
Exposure Ozone
Research
Ozone
NOX, VOCb
Regulatory
Compliance
Control Strategy
Development
Ozone, NOX, VOCb Control Strategy
Development
Ozone
Data for Vegetation
Ozone, NOX, VOC
Data for Health Studies
Urban to
Studies
Research
Neighborhood to
Urban
Middle-Scale
Regional
Neighborhood to
Urban
Regional
Middle-Scale to
Regional
This site is required as part of the NAMS network and is
designed to measure the highest concentration in a
heavily populated area.
This site is designed to measure the concentration of NO
and VOC in proximity to a source. This data would be
used in modeling ozone formation.
This site is designed to measure the ozone and ozone
precursor concentrations entering an urban area from an
upwind source region.
This site provides additional (i.e., more than the maximum
exposure site required for NAMS) exposure data for
health studies.
This site is used to quantify the exposure of vegetation to
ozone to assess the deleteri6us impact oh the vegetation.
This site is established for a specific research purpose
independent of regulations. Often these sites will operate
only temporarily (i.e., during a single summer season).
a For all MSAs or CMSAs with more than 200,000 people, two ozone NAMS sites are required, the maximum downwind site and the maximum exposure site. If the MSA/CMSA is a
serious, severe, or extreme nonattainment area; up to five sites (PAMS) are required. (Note that PAMS sites are based on population - see 40 CFR Part 58, Appendix D, Table 2.)
b EPA requires PAMS VOC monitoring for serious, severe, and extreme ozone nonattainment areas. Up to two of the NAMS monitors may be part of the PAMS program in
these areas. Appendix B to this document includes references for additional information on PAMS monitoring.
Additional downwind, exposure, and emission sites can be installed as necessary, or as resources permit. Definitions of the spatial scale and type of monitoring site are found in 40 CFR Part
58, Appendix D, "Network design for SLAMS, NAMS, and PAMS."
4-5
-------
Determination of human exposure for health effect studies.
Data for air quality research studies.
Determination of vegetation exposure for vegetation effects studies.
These different monitoring objectives require different site selection criteria. The remainder of
this section describes how these different objectives affect site selection.
4.3.1 Regulatory Compliance
For regulatory compliance, the principle objective is to measure the ozone concentration in the
high population density areas and the maximum downwind concentration from the urban region. It is
important to be careful when selecting the high population sites because, particularly in dense urban areas,
the greatest concentration of people may be in an area with heavy automobile traffic, which may result in
low ozone concentration due to NO titration. Therefore, the high population sites must be chosen by
considering both population and the level of nearby fresh NOX emissions. Generally, the area with the
highest population density that might be reasonably exposed to a significant ozone concentration should
be chosen. (Section 5. discusses strategies for determining where the maximum downwind concentration
is located.) For areas needing more than two ozone monitoring sites, additional monitors can be placed in
other areas downwind of the urban region (other than the maximum concentration downwind area) that
may receive transported ozone on a particular day. Note that EPA has not promulgated any special
monitoring requirements for areas designated nonattainment for the 8-hour ozone standard.
4.3.2 Control Strategy Development and Assessment
Another important objective for monitoring is the development of control strategies. This requires
monitoring of both ozone and ozone precursors. For ozone, the sites for regulatory compliance will be
useful for this purpose, as well as additional downwind sites. However, an additional site type is critical
for control strategy development; monitors should be placed upwind of the urban area to determine the
concentration of ozone and ozone precursors entering the urban region. This will help to determine what
portion of an area's ozone problem is due to local emissions and what portion is due to transported ozone
and ozone precursors. It is also important to measure NOX and VOC concentrations for control strategy
development. In particular, NOX and VOC monitors should be placed near the major NOX and VOC
source regions, which are normally in the areas within the urbanized portion with the highest vehicle
traffic. Note that EPA has prepared separate guidance concerning VOC monitoring. EPA has, for
example, detailed enhanced monitoring requirements for serious, severe, and extreme ozone
nonattainment areas, and these requirements provide for the monitoring of over 50 individual VOC
species. Appendix B to this document includes references for those interested in learning more about
VOC monitoring.
4.3.3 Health Effect Studies
For health effect studies, the emphasis is on neighborhood-scale monitors which can be used to
quantify the exposure of the residents of a particular community. Typically, the more differentiation one
can make in the exposure of residents within an urban area, the more statistical power there is to detect
any health effects that may be associated with the exposure. Therefore, areas with good monitor coverage
are ideal for health effect studies.
4-6
-------
4.3.4 Vegetative Impacts Related to a Secondary Ozone Standard
Secondary standards relate to ecological and welfare effects, as opposed to health effects. Since
ozone adversely affects vegetative growth (e.g., commercially produced crops as well as natural forests),
additional regional-scale monitors for vegetation may be necessary. Information on monitoring network
design to address ecological effects is provided by Reiss et al., 1995. Generally, Class I areas and areas
with substantial acreage devoted to agriculture, are important areas for monitoring the effects for the
secondary standard.
4.4 OZONE SEASON
For the purposes of regulatory monitoring, an ozone season is defined for each state and dominion
in the United States. This ozone season is defined as the period for which exceedances of the standard are
expected. Monitoring for ozone is only required during this season. The EPA recently sponsored an
update to the existing ozone season definitions (EPA, 1998) to better refleftt the ozone season for the 0.08
ppm 8-hour threshold. The methodology used to determine the new ozone seasons includes points such as
(EPA, 1998):
1. 8-hour average ozone concentrations were calculated by the U.S. EPA for each day and station in
the United States and its territories using hourly data from the Aerometric Information Retrieval
System (AIRS) database; the EPA provided this data for the six-year period 1990-1995.
2. For determining the ozone season for 8-hour averaging times, exceedances of a conservative
0.080 ppm threshold for each station-month were computed.
3. These data were used to construct state-by-state histograms for the months in which the 8-hour
concentrations were greater than 0.080 ppm in the six-year period 1990-1995 using all available
data for each state.
4. Initial ozone seasons for each state were defined as starting with the first month with any daily 8-
hour concentration greater or equal to the 0.080 ppm threshold, and ending with the month when
the last exceedances of the 0.080 ppm threshold occurred. These seasons vary from state to state
and region to region, but as a whole are generally shorter than seasons based on a 1-hour standard.
With an 8-hour standard, adjustments to these initial ozone monitoring seasons are expected to better
reflect the need for similar ozone monitoring seasons within ozone transport areas; and to consider states
with large geographic areas without available ozone data. Ozone seasons for major metropolitan areas, as
well as a complete description of the methodology used to define these seasons, can be found in the
Guideline for Selecting and Modifying the Ozone Monitoring Season Based on an 8-Hour Ozone
Standard, EPA-454/R-98-001.(EPA 1998).
4-7
-------
5. MACROSCALE CONSIDERATIONS IN MONITOR SITING
5.1 DEVELOPMENT OF A NEW NETWORK OR ENHANCEMENT OF AN EXISTING
NETWORK
This section provides guidance on the siting of four major types of monitors that are needed in an
ozone monitoring network. As discussed hi the last section, these sites are:
• Maximum population exposure sites
• Upwind background sites
• Downwind maximum concentration sites
• Maximum emissions impact (maximum precursor concentration) sites
In addition, this section presents how these monitoring site types are typically deployed to measure
maximum ozone concentrations under two meteorological scenarios: (1) During conditions when
transport occurs from upwind sources or polluted regions (maximum ozone concentration is often
determined by the direction of winds from the emission sources, thus the maximum ozone concentration
often occurs along the path of the predominant wind direction); and (2) during stagnant conditions when
ozone forms primarily from local emission sources.
5.1.1 Maximum Population Exposure Site(s)
This site is fairly straightforward to locate. Census tract data or other population data may be used
to locate the maximum population area. However, care must be given not to locate this monitor in an area
which is too heavily influenced by local emission sources, given the titration of ozone by nitric oxide,
which is part of fresh industrial emissions or automobile exhaust. Therefore, this monitor should be
located in the area with the highest population that is expected to be exposed to a relatively high ozone
concentration. This area is likely to be located on the urban fringe hi an suburb slightly downwind of the
urban area.
5.1.2 Upwind Background and Downwind Maximum Concentration Site(s)
As discussed before, ozone forms in the atmosphere by reactions between VOC and NOX. These
reactions take some time to occur, thus the maximum ozone concentration usually occurs 4 to 6 hours
after maximum emissions, and under conditions of light winds, usually downwind of the urban region
(EPA, 1978; Chu, 1995). Therefore, the key factor for identifying the area where the maximum ozone
concentration is expected to occur is to determine the prevailing wind direction (PWD) from the urban
area. Also, the most significant amount of transported ozone and ozone precursors will come from the
area where the winds enter the city. As meteorological conditions influence the formation of ozone, it is
important to consider PWDs on days when the ozone standard is violated or, if there is no existing
network, on days when conditions are expected to be conducive to ozone formation.
This point is illustrated in Figure 5-1, which shows the morning wind roses for Muskegon,
Michigan from 1981 through 1990 for summer days (June-September) that exceeded the 0.12 ppm 1-hour
ozone threshold, as well as for all summer days. Wind roses should be interpreted as follows:
The arms of the roses point to where the wind was coming from.
.5-1
-------
Muskegon am windrose (ozone GE120 PPB)
10 20 30
Frequency!*/,)
40
Muskegon am windrose (summertime)
Wind speed (m/s)
<>l.5-3 I >3-6 I >6-IO I >IO
10 20 30
Frequency (•/•)
40
Figure 5-1. Morning wind roses (7-10 a.m.) for Muskegon, Michigan
from 1981-1990 for summer days (June-September) for (a) days that
exceeded the 0.12 ppm 1-hour ozone threshold, and (b) all days (Chu,
1995). The figure illustrates that prevailing winds on ozone exceedance
days can be distinct from average wind conditions.
5-2
-------
The length of the arm is proportional to the percentage of time that the wind was coming from that
direction.
The various pieces of the arm represent speed categories.
A program to construct wind roses is available from the EPA Technology Transfer Network (TTN), which
is operated by the EPA's Office of Air Quality Planning and Standards (OAQPS) in Research Triangle
Park, North Carolina. Within the TTN, the Support Center for Regulatory Air Models (SCRAM) provides
a program which constructs wind rose plots from user-provided meteorological data. Appendix B to this
document includes references directing the reader on how to access the TTN and other relevant EPA
information.
As shown in Figure 5-1, there is no distinctive morning PWD for Muskegon if all the summer days
are included. However, for days with peak ozone levels above the 0.12 ppm 1-hour threshold, the majority
of winds come from the southwest. Thus, for Muskegon, an upwind monitor should be placed to the
southwest and similarly, the downwind maximum ozone site should be located based on an afternoon
exceedance day wind rose. Therefore, when determining the appropriate downwind location for an ozone
monitor, it is important to concentrate on the meteorological conditions present during an exceedance day.
Given the generally good correlation between the 0.12 ppm 1-hour and the 0.08 ppm 8-hour thresholds, it
is expected that historical information describing prevailing wind direction during 1-hour exceedance
conditions will generally still be valid during 8-hour exceedances. Using existing data, agencies are
encouraged to conduct similar analyses.
For areas without existing monitoring data or areas which are not certain if their monitors are
located in the maximum downwind direction, Chu (1995) has developed a methodology for determining
from meteorological data which days were most conducive to ozone formation. The methodology is based
on the model developed by Cox and Chu (1993) for relating meteorological variables to the ozone
concentration. Cox and Chu (1993) have found that the maximum ozone concentration is best correlated
with the following meteorological variables: (1) daily maximum 1-hour temperature, (2) morning (7-10
a.m.) winds, (3) afternoon (1-4 p.m.) winds, and (4) midday (10 a.m.-4 p.m.) relative humidity (RH).
Using these relationships, Chu developed a set of meteorological conditions that are likely to occur during
a day that exceeds the NAAQS. These conditions are summarized in Table 5-1. Figure 5-2 depicts the
latitude delineations used in Table 5-1. The data used to develop these relationships were based on 31
eastern cites, but the relationships were found to be adequate for the entire United States. These criteria
were developed based on the 0.12 ppm 1-hour NAAQS. However, the general correlation between the 1-
hour and the 8-hour ozone thresholds suggests that the 1-hour relationships should be generally applicable
during 8-hour conditions. Nonetheless, it would be useful to re-evaluate and possibly refine the criteria
using the 8-hour threshold.
5-3
-------
Table 5-1. Criteria for ozone conducive conditions for the eastern United States.
1. T ^ 26.5°C (79.7°F) for cities north of 40°N
T * 29°C (84.2°F) for cities between 35°N and 40°N
T ^ 32°C (89.6°F) for cities south of 35°N
2. W^m <, 5 m/s for cities in transport regions (i.e., Midwest and Northeast)
Wam < 4 m/s for cities outside transport regions.
3. Wpm < 7.5 m/s for cities in transport regions
Wpm £ 6 m/s for cities outside the transport regions
Wpm <: 5 m/s for Gulf Coast cites and Florida
4. RH < 75% for coastal cities north of 40°N
RH < 65% for inland cities between 30 °N and 40 °N
RH ^ 70% for all cities south of 30°N.
Source: Chu, 1995.
Notes: T refers to temperature; W refers to wind speed; RH refers to relative humidity.
5-4
-------
Figure 5-2. Depiction of the 30, 35, and 40 degree latitude lines for the United States
5-5
-------
After determining the ozone conducive days over the analysis period, wind roses should be
calculated for the morning (7-10 a.m.) and afternoon (1-4 p.m.) for these days. The morning wind rose
will be used to determine the location of upwind monitors, and the afternoon wind rose will be used to
determine the location of downwind ozone monitors. Sometimes the morning and afternoon wind roses
can be different. However, over simple terrain areas, the morning and afternoon wind roses and
corresponding PWDs are usually similar except for some urban heat island effect which creates a
deflection immediately downwind of the city. Over complex terrain and over coastal areas, the two wind
roses can be different. This is illustrated hi Figures 5-3 and 5-4, which show the morning and afternoon
wind roses for Milwaukee, Wisconsin and Atlanta, Georgia. While the morning and afternoon wind roses
are similar in the flat terrain metropolitan area of Atlanta, the complex interactions between the synoptic
winds and the mesoscale lake breeze of Lake Michigan are evident in the difference between the morning
and afternoon wind roses in Milwaukee. Therefore, it is recommended that both the morning and
afternoon wind roses be examined.
As an example, Figure 5-5 shows a four-monitor PAMS/ozone network design for a metropolitan
area. The sites in this sample monitoring design are as follows: (1) background ozone concentration (i.e.,
upwind site), (2) maximum precursor emissions site, (3) daily maximum ozone concentration (i.e.,
downwind site), and (4) ozone impact on downwind areas. The morning PWD was used as a guide for
siting the background 'ozone concentration and maximum precursor emission sites, Sites #1 and #2. The
afternoon PWD was used to site the ozone monitors, Sites #3 and #4, downwind of the metropolitan area.
Chu (1995) recommends that Site #1 be located along the morning PWD upwind from the city limit near
the edge of the photochemical grid model domain, and Site #2 should be located along the morning PWD
near the downwind edge of the central business district (CBD). Site #3 should be located along the
afternoon PWD at the location where the daily maximum ozone concentrations are likely to occur. For
coastal cities, the site should be located as near as possible to the sea/lake breeze convergence zone.
For the 0.12 ppm 1-hour NAAQS, the maximum 1-hour ozone concentration is likely to occur in
the early afternoon about 4 to 6 hours after the precursors are emitted. Therefore, the morning and
afternoon wind roses can be used to find an approximate location of the maximum ozone concentration.
For the 0.08 ppm 8-hour threshold, the maximum concentration generally occurs roughly in the same
downwind area (or slightly further downwind) as the 1-hour daily maximum. Site #4 should be located
along the afternoon PWD near the downwind edge of the photochemical grid model domain.
5.1.2.1 Stagnation Conditions
Chu (1995) notes that the approach described above is not applicable for ozone episodes occurring
during extreme stagnation conditions (average winds < 1.5 m/sec) because, in some
cases, no PWD can be resolved. However, for most stagnation dominated areas, distinct PWDs are still
identifiable. As an example, morning wind roses are shown in Figure 5-6 for two stagnation dominated
metropolitan areas: Louisville, Kentucky and Huntington, West Virginia. While the winds in these cities
are much lighter than the winds in the other cities that were shown, the light but steady winds in these
cities appear to play an important role in the transport, mixing, and accumulation of pollutants.
5-6
-------
Milwaukee am windrose (ozone GE 120 PPB)
Wind fpMd(m/*)
10 20 30
Frequency <•/•)
40
Milwaukee pm windrose (ozone GE I2O PPB)
10 20 30
Fr«au«ncy(V«>
40
Figure 5-3. Morning (7-10 a.m.) and afternoon (1-4 p.m .) wind
roses for Milwaukee, Wisconsin (Chu, 1995). The figure illustrates
that in an area with complex terrain, wind patterns may shift during
the day; therefore, upwind and downwind sites will not necessarily
be located in a straight line through the region.
5-7
-------
Atlanto am windrose (ozone GE-120 PPB).
wino «pM0 (M/C)
10 20 30
4O
Atlanta am wind rote (ozone conducive)
10 20 30
Fr«qutney I'M
4O
Figure 5-4. Morning (7-10 a.m.) and afternoon (1-4 p.m.) wind roses for
Atlanta, Georgia (Chu, 1995). In contrast to the Milwaukee illustration,
the Atlanta example illustrates that hi an area with a relatively flat terrain,
prevailing wind conditions may remain steady throughout the day. In this
5-8
-------
Sample MSA/CMSA
Morning
Wind ro»«t
\Eminion* e«nt
\ /
Central busi new district
Figure 5-5. Example network design for a metropolitan area. Sites
include: (1) background ozone concentration sites (i.e., upwind sites) (2)
maximum precursor emission sites, (3) daily maximum ozone concentration
(i.e., downwind sites), and (4) ozone impact on downwind sites (Chu
5-9
-------
Louisville am windrose (ozone GE120 PPB)
Huntington am windrose (ozone GE 120 PPB)
10
M
10
Figure 5-6. Morning wind roses (7-10 a.m.) for two stagnation dominated metropolitan
areas: Louisville, Kentucky and Huntington; West Virginia (Chu, 1995).
This illustrates that even with light winds, an area may still have a prevailing
wind direction that affects the siting of upwind and downwind monitors.
5-10
-------
For stagnation dominated areas where a well defined PWD may not be discernible, Chu (1995)
recommends that the major axes of the precursor emission sources be substituted for PWDs. Figure 5-7
shows an example PAMS/ozone monitoring network using this strategy. As shown, the axes of the
emission sources extend across the industrial park and CBD. Chu (1995) recommends that monitors be
located within 10 miles of the urban fringe (i.e., no further than 10 miles beyond the outermost portion of
the urban fringe). The reasons for this choice are as follows: (1) completely calm conditions seldom last
more than an hour during the day because most calm days have light variable winds, and (2) ozone
concentrations are likely to be highest when the winds are along the axis of emissions because the
precursor concentrations are likely to be highest.1
5.2 EXAMINATION AND ADJUSTMENT OF AN EXISTING OZONE MONITORING
NETWORK TO REFLECT 8-HOUR MONITORING NEEDS
The need to monitor ozone during 8-hour running average periods raises the issue of whether
monitoring networks designed around the 1-hour standard are adequate to measure compliance with
the 8-hour standard. The analysis presented in Section 3 suggests that, generally, the 1-hour and 8-
hour concentrations are well correlated spatially, and that major changes in the monitoring network
are not needed. However, there are subtle factors that need to be considered to monitor 8-hour ozone
concentrations, including:
Downwind monitoring sites which receive dispersed, transported ozone plumes are more prone to
8-hour exceedances than 1-hour exceedances. This is particularly true for areas without afternoon
rush hours which provide fresh NO emissions to titrate the transported ozone.
The increased stringency of a threshold such as 0.08 ppm 8-hour (in comparison to 0.12 ppm 1 -
hour) may encourage monitoring further downwind of metropolitan areas than is currently done,
especially in situations where these areas contain a public welfare area of concern (e.g.,
agriculture, Class I Areas, etc.).
Several methods are available to analyze where to site 8-hour ozone monitors; these were described in
Section 3, and include the following:
Calculate, tabulate, and graphically display the number of exceedances for 1-hour and 8-hour
thresholds.
Compare locations of the 1-hour and 8-hour exceedances.
Prepare and compare diurnal ozone profiles.
Evaluate the potential for farther downwind sites to exceed the 8-hour standards.
Evaluate the potential for NO titration to influence ozone concentrations.
Examine the correlation between the 1-hour and 8-hour standards.
The terminology in Figure 5-7, #3/#l, indicates that a site can be used both as a maximum downwind
ozone site as well as an upwind site depending upon the time of day/wind direction.
5-11
-------
Sample MSA/CMSA
Morning
Emission* centroia
Central business district
Wind roses
Figure 5-7. Example network design for a metropolitan area with no predominant
wind direction (Chu, 1995)
5-12
-------
For all SLAMS monitoring networks, EPA requires that each agency conduct an annual review to
determine how well it is achieving its required air monitoring objectives, how well it is meeting data users
needs, and how it should be modified (e.g., through termination of existing stations, relocation of stations,
or establishment of new stations) to continue to meet its monitoring objectives and data needs. The main
purpose of the review is to improve the network to ensure that it provides adequate, representative, and
useful air quality data (EPA, 1998). Adjustment to ozone monitoring site locations to respond to the 8-
hour standard should be considered in this annual review. Annual network review criteria may be found
in the SLAMS/NAMS/PAMS Network Review Guidance, EPA-454/R-98-003.
5-13
-------
6. MICROSCALE CONSIDERATIONS FOR MONITOR SITING
6.1 GUIDANCE FOR MONITOR PLACEMENT
The previous section provided guidance on determining the general location in a metropolitan area
for siting photochemical pollutant monitors. Once the general location is selected, care must be given to
selecting a specific location. The site should be selected so that it is in an area of small concentration
gradients (i.e., the concentrations measured by the monitor are not affected by small changes in the
location of the station). This criterion will be met if the site is located such that no single source
contributes disproportionately to the concentrations that are measured, but rather that the measured
concentrations reflect the contributions of numerous sources in the area. This section provides criteria for
selecting station locations that are representative of the larger area around the station by avoiding large
influences of nearby sources and sinks. These recommendations reflect those found in Title 40, Code of
Federal Regulations (CFR), Part 58, Appendix E (Probe Siting Criteria for Ambient Air Quality
Monitoring).
6.1.1 Vertical and Horizontal Probe Placement
Human exposure is a principal concern associated with high ozone concentrations. The monitor's
ozone inlet probe should be placed at a height and location that best approximates where people are
usually located. However, due to complicating factors such as obstructions and adsorbing surfaces,
probes sometimes need to be more elevated. Considering these issues, probe inlets should be located 3 to
15 meters above ground level. The probe must also be located more than 1 meter vertically and
horizontally away from any supporting structure. See also 40 CFR 58, Appendix E, "Probe and
Monitoring Path Siting Criteria for Ambient Air Quality Monitoring" for further information on probe
placement.
6.1.2 Effects of Obstructions
The effects of obstructions such as buildings, trees, and nearby surfaces are important
considerations when siting monitors for ozone and nitric oxide because of the effect obstructions have on
airflow and pollutant mixing, and the destruction of these species upon contact with some surfaces.
Therefore, it is important that sampling be done at a location where the air has had as little contact as
possible with nearby surfaces. Figure 6-1 shows a schematic representation of airflow around a sharp-
edged building based on the work of Halitsky (1961) and Briggs (1973). This figure shows that air in the
"cavity zone" (the region beyond the building where the airflow is influenced by the obstruction of the
building) will make considerable contact with the building. Conversely, air outside the cavity zone will
have passed over the building with minimal contact. According to Briggs (1973), the cavity zone extends
to roughly 1.5 building heights downwind of the building. Using this as a guide, the minimal allowable .
distance separating samplers and obstructions should be at least twice the height of the obstruction
protrudes above the inlet. Figure 6-1 also illustrates why it is recommended that inlets along the side of a
building be avoided. There is airflow up the side of the building which has considerable contact with the
building and allows a substantial possibility for deposition of the pollutant on the surface.
6-1
-------
'///////////A/////////'/'//'
CAVITY ZONE
SA-34OO-1
Figure 6-1. Schematic representation of the airflow around an obstacle such as a
building [EPA, 1978, based on the work of Halitsky (1961) and Briggs (1973)].
6-2
-------
The inlet for sampling must extend above the roof of the building to avoid the complicated airflow
within the cavity zone. If the building which houses the instrument is small (i.e., less than 2 meters high),
then an extension of the inlet above the roof by a distance of about 1.5 meters should be sufficient (1
meter minimum). For taller buildings it may not be possible to avoid the cavity zone on top of the
building without using an inlet line that is so long that it will introduce pollutant losses of its own. If
sampling from the top of a tall building is unavoidable, then it is recommended that the inlet be placed
toward the upwind side of the building. Airflow must be unrestricted in a horizontal arc of at least 270
degrees around the inlet probe, and the predominant wind direction for the season of greatest pollutant
concentration potential must be included in the 270 degree arc. If the probe is located on the side of a
building, a 180-degree clearance is required.
6.1.3 Separation from Roadways
For oxides of nitrogen, most fresh combustion emissions are in the form of NO, which rapidly
reacts with ozone to form NO2 resulting in a decrease in ozone concentrations. EPA (1978) included a
kinetic model to predict the change in the ozone concentration for incremental additions of NOX. Using
this model, it was possible to estimate the proper distance from a roadway for placement of an ozone
monitor so that the ozone concentration is not depressed to a level where the concentration measurement
is not representative of the surrounding area. The EPA has updated this analysis, and the results are
shown in Table 6-1. This table provides the minimum separation distance between roadways and ozone
monitors based on the vehicle traffic of the road.
6.1.4 Spacing from trees and other vegetation
Ozone is known to deposit on trees and other vegetation, thereby reducing the ambient ozone
concentration in that area. Additionally, large trees or stands of trees can obstruct normal wind flow
patterns. To minimize the effects of individual trees on measured ozone concentrations, inlet probes
should be placed at least 20 meters from the "drip line" of trees (the area around the tree where water
dripping from the tree may fall). If the tree or stand of trees could be considered an obstruction, the probe
must be located at least 10 meters from the drip line and at least twice the height of the obstruction
protrudes above the inlet, whichever is furthest.
6.2 USE OF SATURATION MONITORING
Following the use of the sector analysis described in Section 5. for locating a suitable area for a
monitor, one approach for finding a specific location is to develop a short-term sampling study. In this
type of study, numerous portable monitors are employed during a period of high ozone concentration at
various candidate sites within the sectors identified by the meteorological analysis as having the potential
for maximum ozone concentrations. Analysis of data over the period of about a month could be used to
determine the best monitor location.
6-3
-------
Table 6-1. Separation distance between ozone monitors and
roadways (i.e., the minimum distance between the edge of the
nearest traffic lane and the monitor).
Roadway Average Daily
Traffic Vehicles Per Day
si 0,000
15,000
20,000
40,000
70,000
* 110,000
Minimum Separation
Distance Between Roadways
and Stations (meters)*
>10
20
30
50
100
>250
a Distances should be interpolated based on traffic flow.
Source: 40 CFR Part 58, Appendix E, "Probe siting criteria for ambient air quality monitoring."
6-4
-------
This technique was successfully applied during the Wasatch Front Ozone Saturation Study (Start et
al., 1994). The State of Utah was required to install two new monitoring sites in the Wasatch Front and
conducted this study to determine the optimal location of these monitoring sites. The study was also useful
for assessing whether the existing ozone monitoring sites were properly located. The study employed
Ogawa passive samplers (Koutrakis et al., 1993), which are small, inexpensive diffusion samplers (i.e., no
pump is required) that are useful for obtaining integrated average ozone concentrations over a sampling
period, usually on the order of 12 to 24 hours. The study made measurements at 62 passive sampler
locations, and also placed passive samplers at each of the current ozone monitoring sites. The locations of
the sites are shown in Figure 6-2. Sampling occurred over a 30-day period. However, sampling only
occurred on 7 of the 30 days which were expected to have elevated ozone concentrations. Each sampler
was exposed for 24 hours to give a one-day integrated sample. The passive samplers performed well, and
the study provided an inexpensive means of better understanding the spatial distribution of ozone in the
Wasatch Front.
6-5
-------
Figure 6-2. Monitor locations for the Wasatch Front Ozone Saturation
Study summer 1993 (Start, 1994)
6-6
-------
7. RECOMMENDATIONS
As a primer on ozone monitoring issues, this report has touched on a number of key issues and their
implications for monitoring network design including: ozone formation and spatial distribution;
monitoring objectives, such as regulatory compliance and scientific research; big picture geographic issues,
such as the need for siting monitors upwind of an urban area; and micro-scale issues, such as the need to
avoid siting monitors too close to trees and buildings. Readers with an interest in either developing a new
ozone monitoring network, assessing the efficacy of an existing network, or updating an existing network
to monitor 8-hour ozone concentrations should all find this report's contents to be of value. In
summarizing the report the focus has been to organize information hi a way that answers commonly asked
questions. As has been mentioned throughout the text, the reader is referred to the Appendices for more
detailed information concerning the science behind ozone formation, as well as the many additional
resources available to assist in designing and implementing monitoring networks. Each of the question and
answer discussions in this section concludes with a reference to the earlier report chapters that provides
additional relevant information.
How many monitors are required for a particular region?
The EPA has established minimum network regulatory requirements on how many NAMS and
PAMS monitors are required. The EPA guidelines serve as minimum requirements beyond which
additional monitoring through the SLAMS network provides better documentation of air quality in a
community. Adequately defining the ambient ozone problem in any given area, especially the larger urban
areas, generally necessitates the establishment of additional SLAMS monitors, in addition to the minimum
requirements.
EPA requirements vary depending upon the population and nonattainment status of a particular
area. For all urbanized areas with more than 200,000 people, two ozone NAMS sites are required, the
maximum downwind site and the maximum exposure site. In addition, EPA requires PAMS VOC, ozone,
nitrogen oxides, and meteorological parameter monitoring for serious, severe, and extreme ozone
nonattainment areas (for the 1-hour standard). Areas with less than 500,000 people are required to have at
least two PAMS sites, areas with 500,000 to 1,000,000 people are required to have at least three sites, areas
with 1,000,000 to 2,000,000 people need four sites, and areas with more than 2,000,000 people need at
least five PAMS sites (EPA, 1994). Appendix B to this document includes references for more PAMS
related information. These PAMS requirements have not been extended to include areas which are
nonattainment for the 8-hour standard only.
Scientifically, the number of ozone monitors needed is a function of the region's size and
topographic and meteorological complexity. At a minimum, two monitors can be sufficient for a relatively
small metropolitan area. EPA recommends that a region using only two monitors site these as (1) a
downwind site to measure peak ozone concentrations hi a populated area and (2) a site to measure
maximum population exposure. A relatively low-cost enhancement would be a monitor at a third location
designed to measure upwind, or background, ozone concentrations. Examples of when additional monitors
would be necessary include:
• Additional monitors will be needed if the region's high ozone concentration days are triggered by any
one of several different meteorological scenarios. When prevailing winds differ depending upon the
high-ozone meteorological scenario, additional monitors are needed to evaluate each scenario's upwind
7-1
-------
and downwind locations.
Multiple monitors will be needed if an area is quite large geographically, with multiple communities in
downwind locations. Multiple downwind monitors will be needed to adequately measure peak ozone
in the affected areas.
Additional monitors would also be necessary to measure concentrations in an urban core; to address
special studies focused on particular areas; to monitor areas with complex terrain; to monitor the effects of
specific sources; and to quality assure the existing monitoring network irom time to time by collocating
monitors near existing monitors, or by conducting saturation monitoring to ensure that the highest upwind
and downwind concentrations are being measured.
[Refer to Section 4 of the report for more detailed information.]
How should monitors be sited to properly account for wind direction?
Some straightforward meteorological evaluations can help determine how to site monitors to
properly account for prevailing winds. Wind roses need to be constructed for the metropolitan area to
identify what direction the wind originates from at given times of day, as well as the wind strength during
these time periods. Note that the wind strength is important to help determine how quickly pollutants will
travel downwind and be dispersed. Some general "rules of thumb" can assist in this evaluation:
Identify days when high ozone concentrations are known to occur. If these days are not yet known,
identify days when high ozone concentrations are likely to occur. Methodologies are available to help
with this analysis, which utilize wind, temperature, and relative humidity data (Cox, 1995).
Construct wind roses for morning (7:00-10:00 a.m.) and afternoon (1:00-4:00 p.m.) winds for high
ozone days. Morning wind direction can help determine where to site upwind monitors; morning and
afternoon wind directions can help determine where to site downwind monitors.
Note that it is important to distinguish between average summer days, and days when high ozone
concentrations are likely to occur. Prevailing average winds may be different than winds on high ozone
days. Also note it is possible that more than one meteorological regime is contributing to high ozone
formation.
[Refer to Section 5 of the report for more detailed information.]
How should monitors be sited to avoid problems in the immediate vicinity of the monitor (e.g., from
roads, trees, buildings)?
Large obstructions, such as tall buildings, and small obstructions, such as trees, can each
contribute to inaccurate pollutant measurements if monitors are sited too close to the obstruction. "Rules
of thumb" to consider include:
Generally, monitoring probe inlets should be located 3 to 15 meters above ground, and at least 1 meter
vertically and horizontally away from the supporting structure.
The downwind sides of large obstructions such as buildings or stands of trees include wind pockets
that are not representative of the surrounding air mass. Multiplying the height of the obstruction by 10
7-2
-------
yields a distance equal to the desired distance between the base of the obstruction and the monitor. If
space of that magnitude is not available, the minimal acceptable distance between such obstructions
and monitors is 2 times the height of the obstruction.
• Fresh NO emissions titrate ozone. Thus, ozone is quickly reduced in the vicinity of roadways with any
significant amount of traffic. Monitors should be placed anywhere from at least 10 to at least 250
meters from a roadway, depending upon the extent of traffic. Existing monitors can be quality assured
by analyzing their monitoring data. Frequent measured ozone values in the range of 0 to 30 ppb of
ozone may be an indicator that the site is too close to fresh NO emissions. Natural ozone background
concentrations are on the order of 40 ppb, prior to pollutant contributions that increase observed
concentrations. Measured concentrations hi the 0 to 30 ppb range are an indicator of ozone titration
and should trigger a review of the monitor site and its proximity to emission sources.
[Refer to Section 6 of the report for more detailed information.]
Are existing 1-hour monitors still useful for 8-hour monitoring, or should they be moved? In
particular, does an 8-hour standard mean that the geographic area monitored needs to be enlarged,
and if so, by how much?
The locations of existing 1-hour monitors are likely to remain appropriate for monitoring upwind
contributions and downwind peak concentrations. There will possibly be a need however, for additional
monitors to be located further downwind than those established for 1-hour monitoring. These will help
determine the maximum downwind distance where 8-hour exceedances occur.
There is a strong correlation between 0.12 ppm 1-hour and 0.08 ppm 8-hour ozone concentrations,
in terms of the general locations of exceedances. There is also a strong correlation between the locations
of peak 1-hour and 8-hour ozone concentrations. A key difference between 1-hour and 8-hour ozone
concentrations, however, is that downwind of a metropolitan area, 8-hour concentrations tend to exceed
the 0.08 threshold more often than 1-hour concentrations exceed the 0.12 ppm threshold. This means that
when monitoring 8-hour ozone concentrations, it is probable that the area to be monitored needs to be
increased in geographic size in comparison to the area monitored for 1-hour ozone concentrations.
A "rule of thumb" may be as follows: 1-hour peak ozone concentrations are observed
approximately 4 to 6 hours downwind of the location of the maximum emissions region (usually the urban
core). When monitoring for 8-hour ozone concentrations, however, ozone exceedances may be observed
over a broader geographic area than the typical 1-hour exceedance locations. An 8-hour monitoring
network would likely benefit by adding monitors to the existing 1-hour network. One suggestion might be
to place the additional monitors approximately 1 to 2 hours further downwind from where monitors are
normally sited to measure peak 1-hour concentrations. Consideration of the location for the monitor for
the potentially affected population must be balanced with the need to capture maximum downwind
concentrations outside populated areas.
[Refer to Sections 3 and 5 of the report for more detailed information.]
How can monitor sites be quality assured to determine if they are located in appropriate downwind
locations to identify peak ozone concentrations?
The only practical way to assure whether downwind monitors are observing peak ozone
concentrations is to locate additional monitors further upwind and downwind to see whether measured
7-3
-------
ozone concentrations are higher or lower than the existing monitor. Another possibility is to conduct
photochemical air quality computer modeling to predict the location of peak ozone concentrations, and to
compare the modeling results to the existing or planned network-(EPA has issued separate guidance with
detailed computer modeling guidelines). An important consideration is how long ago the existing
monitoring network was established. If the monitors were sited many years ago, and the area has since
undergone significant growth, peak ozone concentrations may be migrating further downwind and/or in
the direction of the growth. Additionally, special studies, such as those described in Section 6.2 of this
document, can be used as an adjunct to network/site evaluation.
[Refer to Section 3 of the report for more detailed information.]
Should 8-hour ozone monitoring be conducted during the same time of the year as 1-hour ozone
monitoring?
The EPA has established ozone seasons for the 1-hour standard which are published in the Code of
Federal Regulations (40 CFR 58, Appendix D). However, given differences in the spatial distribution of
1-hour and 8-hour exceedances, and because the 8-hour standard is stricter than the 1-hour standard,
adjustments to these seasons may be necessary. The EPA is recommending (EPA, 1998) initial 8-hour
ozone seasons for each state to be defined as the period of time beginning with the first month with any
daily 8-hour concentration greater to or equal to a 0.080 ppm threshold, and ending with the last month
with such exceedances during a 6-year period. Adjustments to these initial ozone seasons are expected to
better reflect the need for similar ozone monitoring seasons within ozone transport areas; and to consider
states with large geographic areas without available ozone data.
[Refer to Section 4 of the report for more detailed information.]
How should monitor siting differ depending upon whether an area is an isolated urban area, is part of
an urban corridor which includes numerous metropolitan areas, or is an urban area with multiple
prevailing wind directions?
Answering this question helps summarize a number of the issues presented in this report. The
information presented here draws upon the real world examples provided by the San Francisco Bay Area's
monitoring network. The San Francisco Bay Area offers a good example of a reasonably complex area
with monitors in a variety of locations that serve multiple needs. Among the San Francisco Bay Area's
numerous monitoring locations are several major metropolitan areas: San Francisco itself, San Jose to the
south and east of San Francisco, Oakland to the east, and a number of less populous areas to the north,
east, and south (Figure 3-1 illustrates the geographic relationship among Bay Area cities). In addition,
Sacramento is approximately 90 miles to the northeast of San Francisco, and the proximity between the
two regions helps to illustrate the value of monitoring sites located between neighbor regions. This
section concludes with five different monitoring scenario discussions that utilize the San Francisco Bay
Area as a real world example.
San Jose: an urban area influenced by pollutant transport. San Jose experiences pollutant transport
from its upwind neighbors in the Bay Area, and contributes pollutants to downwind sites. Prevailing
winds influencing San Jose's ozone concentrations are generally from the northwest. Therefore,
useful monitor locations include San Jose itself, upwind locations to the northwest, and downwind
locations to the southeast. Bay Area cities within this geographic sphere include Oakland, San
Leandro, Hay ward and Fremont to the northwest, and Gilroy to the southeast, all of which have
7-4
-------
monitors. San Jose is one of the areas where peak ozone concentrations are observed in the San
Francisco Bay Area. The Gilroy monitor to the southeast of San Jose both monitors the effects of
downwind transport from San Jose on the Gilroy area and serves to confirm that ozone
concentrations monitored in San Jose are higher and, therefore, representative of peak
concentrations. The monitors stretching to the northwest of San Jose provide a range of upwind
ozone concentration measurements that enable San Jose to track the transport of ozone and ozone
precursors into its metropolitan area.
San Francisco: an urban area influenced bv clean upwind air conditions. Metropolitan areas that are
generally located downwind from and near other metropolitan areas should expect to receive and be
influenced by pollutant transport unless one of two conditions occur: (1) extremely fast-rising
mixing heights occur which enable ozone concentrations to mix and become dispersed aloft or (2)
strong afternoon winds with clean air sweep through the region and remove locally generated and
transported pollutants. San Francisco is a coastal city with the Pacific ocean to its west. Afternoon
winds tend to be strong due to the sea breeze, and are relatively clean coming from the Pacific ocean.
In contrast to its neighbor city, San Jose, San Francisco is relatively free from the influence of
pollutant transport, and thus has little need for upwind monitoring.
Bethel Island: a downwind area that exceeds the 8-hour but not the 1 -hour ozone thresholds. The
San Francisco Bay Area's peak 1-hour and 8-hour ozone concentrations occur in the same general
areas: to the east and to the south of the Bay Area. However, there are areas in the far downwind
portions of the Bay Area that exceed a 0.08 ppm 8-hour ozone threshold without exceeding the 0.12
ppm 1-hour threshold. Bethel Island, east of the Bay Area, is an example site exceeding the 8-hour,
but not the 1-hour threshold. Bethel Island receives transported ozone and ozone precursors late hi
the day that contribute to higher late afternoon ozone concentrations. These peak concentrations are
on the order of 120 ppb ozone, but are not quite high enough to exceed the 1-hour 125 ppb threshold.
However, when added to the locally generated ozone formed earlier in the day, this transported
pollution creates an 8-hour concentration that exceeds the 8-hour threshold.
The greater San Francisco Bav Area: an area with several predominant wind directions contributing
to high ozone concentrations. The greater San Francisco Bay Area experiences high ozone
concentrations to the northeast, the east, and the southeast of downtown San Francisco. These
coincide with the varying meteorological regimes that contribute to ozone formation in the Bay Area,
as well as the region's topographic complexity, which includes two mountain ranges that run
northwest to southeast (to the south and east of the Bay Area); a river channel that heads to the
northeast; a southeastern basin surrounded by San Francisco Bay to the north; and the two mountain
ranges that sweep toward one another at the Bay Area's southern rim. To accommodate these
meteorological and topographic conditions, the Bay Area has established ozone monitors in
numerous locations. Several are located in a line extending through the region from the southwest to
the northeast, to the east, and to the southeast. There are also numerous background monitors located
in rural areas to the north. By arranging monitors in this manner, the Bay Area is able to measure
ozone concentrations along the predominant wind directions that occur during conditions favorable
to ozone formation. The large number of monitors enables the Bay Area to measure peak
concentrations with relative confidence, since the area also measures downwind concentrations
which are less than peak concentrations.
7-5
-------
San Francisco to Sacramento: metropolitan areas with neighboring monitoring networks. When
two metropolitan areas are located in close proximity to one another, their monitoring networks
provide an opportunity to measure pollutant concentrations relevant to each other's area. The San
Francisco and Sacramento, California areas provide an example of this situation. Under some
conditions, pollutants are transported to the northeast of San Francisco as far as the Sacramento
metropolitan area, some 90 miles away. The San Francisco to Sacramento corridor is an area of
continued growth, and several monitoring sites are located along the northeast path that connects
the two cities. Monitors located in between these metropolitan areas can serve a dual purpose: for
example, monitors to the southwest of Sacramento may have been sited to provide background or
upwind readings for the Sacramento area. Depending upon the meteorological conditions present,
these same monitors may also serve to provide downwind ozone concentrations for the San
Francisco Bay Area. Note that hi areas where pollutant transport is a significant ongoing event,
such as in the northeastern U.S. ozone transport region (OTR), monitors between major urban
areas may record more uniform pollutant concentrations in comparison to the San Francisco
example cited here. In those cases, the appropriate location of downwind monitors will be a
function of (1) downwind ozone concentrations and (2) proximity to neighbor metropolitan areas.
[Refer to Section 3 of the report for more information and maps of the areas discussed.]
What PAMS requirements will be necessary in new areas designated nonattainment only for the 8-
hour standard ?
PAMS requirements are extensive and include requirements for the monitoring of an array of ozone
precursors and meteorological parameters. At this time, EPA has not extended the PAMS requirements to
new ozone nonattainment areas which are nonattainment only for the 8-hour NAAQS (a new 8-hour ozone
nonattainment area in an area EPA designates nonattainment for the 8-hour NAAQS that was not
previously nonattainment for the 1-hour NAAQS). Appendix B to this document includes PAMS
references for readers seeking PAMS guidance materials.
7-6
-------
8. REFERENCES
Altshuller A.P. and Lefohn A.S. (1996) "Background ozone in the planetary boundary layer
over the United States". J. Air & Waste Manag. Assoc. 46, 134-141.
Briggs G.A. (1973) Diffusion Estimation for Small Emissions. Draft No. 79 in 1973
annual report of the Atmospheric Turbulence and Diffusion Laboratory, Oak Ridge,
TN., Report No. ATDL-106, 83-195.
Carter W.P.L. (1994) "Development of Ozone Reactivity Scales for Volatile Organic
Compounds". J. Air & Waste Manag. Assoc. 44, 881-899.
Chu S.-H. (1995) "Meteorological Considerations in Siting Photochemical Pollutant
Monitors". Atmos. Environ. 29,2905-2913.
Code of Federal Regulations (1997) Title 40, Part 58, Appendix D. "Network Design for
SLAMS, NAMS, and PAMS"., published by United States Government.
Cox W.M. and Chu S.H. (1993) "Meteorologically Adjusted Ozone Trends in Urban Areas:
a Probabilistic Approach". Atmos. Environ. 27B, 425-434.
Halitsky J. (1961) "Estimation of Stack Height Requirements to Limit Contamination of
Building Air Intakes". J. Amer. Ind Hyg. Assoc. 26,106-115.
Husar R.B. (1996) Pattern of 8-hour Daily Maximum Ozone and Comparison with the 1-
hour Standard. Report prepared by the Center for Air Pollution Impact and Trend
Analysis for the OTAG Air Analysis Workgroup, posted on the OTAG Air Analysis
Web Site, URL='http://capita.wustl.edu/otag/1.
Koutrakis P., Wolfson J.M., Bunyaviroch A., Froehlich S.E., Hirano K., and Mulik J.D.
(1993) "Measurement of Ambient Ozone Using a Nitrite-coated Filter". Anal.
Chem. 65,209-214.
National Research Council (1991) Rethinking the Ozone Problem in Urban and Regional
Air Pollution. National Academy of Sciences/National Research Council, National
Academy Press, Washington, DC.
Reiss R., Chinkin L.R., and Main H.H. (1995a) "Ozone NAAQS Review-ambient Air
Monitoring Support Target Strategy. Work assignment 1-95. Final report prepared
for U.S. Environmental Protection Agency, Research Triangle Park, NC by Sonoma
Technology, Inc., Santa Rosa, CA, EPA Contract No. 68D30020, STI-94510-1535-
FR, September.
8-1
-------
Reiss R., Chinkin L.R., Roberts P.T., Main H.H., and Eisinger D.S. (1995b) Investigation
of Monitoring Networks for an Alternative Ozone NAAQS. Work assignment 7-
95. Final report prepared for U.S. Environmental Protection Agency, Research
Triangle Park, NC by Sonoma Technology, Inc., Santa Rosa, CA, EPA Contract No.
68D30020, STI-94571-1553-FR, December.
Seinfeld J.H. (1986) Atmospheric Chemistry and Physics of Air Pollution. Wiley
Interscience, New York, NY, pp. 154-156.
Start G.E. (1994) Wasatch Front Ozone Saturation Study Summer 1993. Report prepared
for the State of Utah Department of Environmental Quality by U.S. Department of
Commerce, National Oceanic and Atmospheric Administration, Idaho Falls, ID.
U.S. Environmental Protection Agency ,(1978) Site Selection for the Monitoring of
Photochemical Air Pollutants. Office of Air Quality Planning and Standards,
Research Triangle Park, NC, EPA-450/3-78-013, April.
U.S. Environmental Protection Agency (1989) Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume IV - meteorological measurements.
Report prepared by the Office of Research and Development, Atmospheric Research
and Exposure Assessment Laboratory, U.S. Environmental Protection Agency,
Research Triangle Park, NC, EPA/600/4-90-003.
U.S. Environmental Protection Agency (1996) Clearinghouse for inventories and emission
factors. On U.S. Environmental Protection Agency electronic bulletin board.
U.S. Environmental Protection Agency (1998) Guideline for Selecting and Modifying the
Ozone Monitoring Season Based on an 8-Hour Standard, Office of Air Quality
Planning and Standards, Research Triangle Park, North Carolina, EPA-454/R-98-
001, June.
U.S. Environmental Protection Agency (1994) Photochemical Assessment Monitoring
Stations Implementation Manual, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina, EPA-454/B-93-051, March.
U.S. Environmental Protection Agency (1998) SLAMS/NAMS/PAMS Network Review
Guidance, Office of Air Quality Planning and Standards, Research Triangle Park,
North Carolina, EPA-454/R-98-003, March.
U.S. Environmental Protection Agency (1998) National Air Quality and Emissions Trends
Report, 1996, Office of Ah" Quality Planning and Standards, Research Triangle
Park, North Carolina, EPA-454/R-97-013, January.
8-2
-------
APPENDIX A
BACKGROUND ON OZONE FORMATION
-------
APPENDIX A: BACKGROUND ON OZONE FORMATION
OVERVIEW
Ozone is formed as a product of photochemical reactions involving several components,
the primary contributors being volatile organic compounds (VOC) and oxides of nitrogen (NOX).
This appendix includes a brief discussion of the atmospheric chemistry governing ozone
formation. The material in this discussion is based on Seinfeld (1986).
As discussed in the main body of the report, the photolysis of NO2 in sunlight produces
NO and O, and the free oxygen atom rapidly combines with O2 and forms ozone (O3):
O (A-l)
O + O2 + M—^-> 03 + M (A-2)
where hv represents photochemical energy from ultraviolet radiation (or a photon), k, and k2
represent rate constants for the reaction of NO2 and hv, and M represents N2, O2 or another
molecule that absorbs the reaction's excess vibrational energy. Once formed, ozone is rapidly
dissociated by reaction with NO, as follows:
Os + NO — ^-> NO2 + O2 (A-3)
The NO2 molecule is regenerated, and in the absence of other species a steady state is achieved
through reactions (A-l) through (A-3) in which the ozone concentration can be estimated by the
following relationship:
(A'4)
where k, and k3 are the-rate constants for reactions (A-l) and (A-3).
Ozone in the clean troposphere is governed almost solely by the relationship described by
reaction (A-4). In the more polluted troposphere, oxidation of VOC by free radicals have
significant effects on ozone formation by converting NO to NO2. The key to this oxidation
process is the OH radical (a radical is a highly reactive molecule) which is formed by ozone
photolysis and other reactions. Its reaction with many VOC leads to the formation of peroxyalkyl
radicals, as follows:
RH+OH-^ R- + H2O (A-5)
A-3
-------
Where RH is a general symbol for a hydrocarbon (which in this context is synonymous with a
VOC); OH • is a hydroxy radical (the dot following the OH symbolizes a free electron, which
defines a radical); and, R • is the RH molecule without the H (it is thus a free radical
hydrocarbon). In the presence of oxygen, the next step in this process is:
> ROz- • (A-6)
where ROz • is a free radical oxidized hydrocarbon (a peroxy radical).
Similarly, OH radicals react with aldehydes ( RCHO ) to form acyl ( RCO • ) and
acylperoxy [ RC(O)O2 • ] radicals.
The peroxy radicals react rapidly with NO to form NO2 and other free radicals, as follows:
RO2-+NO^> NOz + RO-^> RONOz (A-7)
RC(O)O2 • +NO -> NO2 + RC(O)O • (A-8)
The [ RC(O)O2 • ] radicals normally decompose to alkyl radicals which leads to the generation
of another peroxyalkyl radical. This peroxyalkyl radical can oxidize additional NO to NO2.
Typically, smaller alkoxy radicals react with O2 to form HO2 radicals and a carbonyl
compound, as follows:
(A-9)
where R' CHO is an aldehyde or carbonyl, R is a general term for a carbon chain, and R' is a
different carbon chain. Like other peroxy radicals, the hydroperoxy radical (HOz ) can oxidize
NO to NO2 as follows:
HO2 -+NO^> NOz + OH • (A- 10)
Note that the OH radical is regenerated in this reaction, which completes the cycle. Therefore,
this OH • radical is available to react with other VOC species and continue the chain of reactions.
A-4
-------
APPENDIX B
SUPPLEMENTAL REFERENCE MATERIALS
-------
APPENDIX B: SUPPLEMENTAL REFERENCE MATERIALS
On-line Resources
The U.S. EPA operates a world wide web site with a variety of useful information. The
Internet address is 'www.epa.gov'. The Office of Air Quality Planning and Standards
(OAQPS) has a sub-directory on the site at 'www.epa.gov/oar/oaqps'.
EPA Quality Assurance Guidelines
1. Quality Assurance Handbook for Air Pollution Measurement Systems, Volume I: A
Field Guide to Environmental Quality Assurance U.S. EPA, EPA-600/R-94-038a,
April 1994 (or later revision).
2. Quality Assurance Handbook for Air Pollution Measurement Systems, Volume II:
Ambient Air Specific Methods U.S. EPA, EPA-600/R-94-038b, April 1994 (or later
revision due.out Spring 98).
3. Quality Assurance Handbook for Air Pollution Measurement Systems, Volume IV:
Meteorological Measurements U.S. EPA, EPA/600/4-90/003, August 1989
These 3 quality assurance documents are available through the EPA's Center for
Environmental Research Information, 26 W. Martin Luther King Drive, Cincinnati, Ohio
45268, telephone (513) 569-7562, fax (513) 569-7566.
Photochemical Air Monitoring System (PAMS)
Photochemical Assessment Monitoring Stations Implementation Manual U.S. EPA, EPA-
454/B-93-051, March 1994
This reference document is available through the National Technical Information Service
(NTIS), 5285 Port Royal Road, Springfield, VA 22161, telephone (703) 487-4650, fax (703)
321-8547. The PAMS NTIS document number is PB 941 873 82.
Meteorological Data
National Climatic Data Center (NCDC), Asheville, North Carolina, (704) 271-4476. NCDC
stores a wide variety of meteorological data from the United States, which can be used for
analysis.
-------
EPA Technology Transfer Network (TTN) contains a program to construct wind rose plots.
The program can be found within the directory for the Support Center for Regulatory Air
Models (SCRAM). The TTN electronic bulletin board can be reached by modem at (919)
541-5742 of yia^Jnten^
The National Weather Service (NWS) operates regional clunate centers, which can provide
meteorological data. The regional centers include:
1.) High Plains Clunate Center
L.W. Chase Hall
University of Nebraska
Lincoln, NE 68583-0728
(402) 472-6706
2.) Midwestern Clunate Center
Illinois State Water Survey
2205 Griffith Drive
Champaign, IL 61820-7495
(217) 244-8226
3.) Northeast Regional Clunate Center
1 123 Bradfield Hall
Cornell University
Ithaca, NY 14853
(607) 255-1751
4.) Southeast Regional Climate Center
1201 Main Street
Suite 1100
Columbia, SC 29201
(803) 765-0849
5.) Southern Regional Clunate Center
245 Howe-Russell Complex
Louisiana State University
Baton Rouge, LA 70803
(504) 388-5021
6.) Western Regional Climate Center
Desert Research Institute
P.O. Box 60220
Reno, NV 89506-0220
(702) 677-3106
B-4
-------
Code of Federal Regulations
The following citations include specific monitoring requirements as outlined in the Code of
Federal Regulations.
40 CFR Part 58, Appendix D - Network Design for State and Local Air Monitoring Stations
(SLAMS), National Ah- Monitoring Stations (NAMS), and Photochemical Assessment
Monitoring Stations (PAMS).
40 CFR Part 58, Appendix E - Probe Siting Criteria for Ambient Air Quality.
B-5
------- |