EPA-450/2-77-021b
February 1978
PROCEDURES
FOR QUANTIFYING
RELATIONSHIPS BETWEEN
PHOTOCHEMICAL OXIDANTS
AND PRECURSORS:
SUPPORTING
DOCUMENTATION
U.S. ENVIRONMENTAL PROTECTION AGENCY
Monitoring and Data Analysis Division
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-W2-77-021b
PROCEDURES FOR QUANTIFYING RELATIONSHIPS BEIWEEN
PHOTOCHEMICAL OXIDAifTS AND PRECURSORS:
SUPPORTING DOOJNTATIQN
February 1978
Monitoring and Data Analysis Division
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from
the Library Services Office (MD-35), U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina 27711; or, for a fee,
from the National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161.
This document has been reviewed by the Office of Air Quality Planning
and Standards, U.S. Environmental Protection Agency, and approved for
publication. Subject to clarification, the contents reflect current
Agency thinking.
-------
Tab] e of Contents
Page
List of Figures iv
List of Tables v
Executive Summary vii
1.0 Introduction 1-1
2.0 Natural Background Concentrations of Ozone and Precursors 2-1
and Their Resulting Impact on Urban Ozone Levels
2.1 Introduction 2-1
2.1.1 Purpose 2-1
2.1.2 Background 2-1
2.2 Natural Ozone in Rural Areas 2-3
2.2.1 Stratospheric Source 2-4
2.2.1.1 Effect of Slow Transfer/Diffusion Process 2-6
2.2.1.2 Effect of Stratospheric Intrusions 2-6
2.2.2 Natural Sources of Precursors 2-8
2.2.3 Estimated Impact of Natural (Vegetative) Precursor 2-9
Sources
2.3 Estimated Impact of the Transport of Natural Background 2-12
Concentrations into Urban Areas
2.4 Summary and Conclusions 2-13
2.5 References 2-17
3.0 Estimating NMHC/NO Ratios for Use in the Empirical Kinetic 3-1
Modeling Approach CEKMA)
3.1 Introduction 3-1
3.1.1 Purpose 3-1
3.1.2 Background 3-1
3.2 NMHC/NOV Ratios for Use in EKMA 3-4
/\
-------
Table of Contents (cont'd)
Page
3.2.1 Spatial Variation in NMHC/NO Ratios 3-6
X
3.2.2 Uncertainty in Individual NMHC Readings 3-9
3.2.3 Diurnal and Spatial Variation in NMHC and NOx 3-11
Concentrations
3.2.4 NMHC/NO Ratios on High Ozone Days 3-14
A
3.2.5 Default NMHC/NO Ratios 3-16
X
3.3 Summary and Recommendations 3-21
3.4 References 3-25
4.0 Variation of Ozone and Precursor Levels in and Near 4-1
Urban Areas
4.1 Introduction 4-1
4.1.1 Purpose 4-1
4.1.2 Background 4-1
4.2 Ozone Concentrations Measured Upwind of Urban Areas 4-2
4.3 Urban Precursor Variations 4-4
4.4 Summary 4-8
4.5 References 4-12
5.0 The Impact of Inter-Urban Pollutant Transport on 5-1
Oxidant Concentrations
5.1 Introduction 5-1
5.2 The Simulation Model 5-2
5.3 Conditions Simulated 5-2
5.4 The Impact of Transport of Various Pollutants 5-6
on Ozone Concentrations
5.5 The Impact of Ozone Transport under Various Conditions 5-17
5.6 Anticipated Trends in the Impact of Ozone Transport 5-28
11
-------
Table of Contents (cont'd)
Page
5.7 Conclusions 5-32
5.8 References 5-35
6.0 Acknowledgements 6-1
m
-------
List of Figures
Page
2.1 Scavenging - mixing - dilution of natural 0, component 2-14
in vicinity of an urban area.
3.1 Sensitivity of maximum afternoon ozone concentrations 3-3
to morning precursor levels measured upwind.
3.2 RAPS monitoring network and surrounding urban areas. 3-5
3.3 Histograms of 6-9 a.m. NMHC/NO ratios at selected RAPS sites. 3-10
A
3.4 Average diurnal variation in NMHC 3-12
(St. Louis RAPS, June-October 1976).
3.5 Average diurnal variation in NO 3-13
(St. Louis RAPS, June-October 1§76).
3.6 Sensitivity of maximum afternoon ozone concentrations to 3-20
precursor concentrations, given default NMHC/NO ratios.
A
4.1 Ozone monitoring sites around the Philadelphia-Camden 4-5
metropolitan area.
4.2 Variation of 6-9 a.m. average NMHC with distance from 4-9
Site 101 (June-October 1976).
4.3 Variation of 6-9 a.m. average NO with distance from 4-10
Site 101 (June-October 1976). x
5.1 Role of ozone aloft on maximum surface ozone levels: 5-8
a conceptual view.
5.2 Primary set of simulated conditions. 5-20
5.3 Additional conditions simulated. 5-21
-------
Lj_st_ o_ __
Page
2.1 Typical Estimates of Natural Contribution to Ambient 2-7
Ozone Levels During the Photochemical Oxidant Season.
3.1 Ambient Data from Urban and Suburban Sites Used to 3-7
Estimate Default Ratios.
3.2 Example of Sensitivity of Suburban Ambient Data to Wind 3-8
Direction -- Portland, Maine.
3.3 Comparison of NMHC/NO. Ratios for Selected Sites 3-15
in St. Louis, 1976. x
3.4 Comparison of NMHC/NO Ratios on High Ozone Days 3-17
for Selected Cities. x
3.5 Comparison of Concentrations of NMHC and NO on High 3-18
Ozone Days with Seasonal Mean Concentration^ at RAPS Sites.
4.1 Upwind Ozone Concentrations for Selected Cities on High 4-3
Ozone Days During 1975.
4.2 Ozone Concentrations Measured Upwind of St. Louis on 4-6
Nine High Ozone Days in 1976.
4.3 Significance of Transported Precursor Concentrations. 4-7
5.1 Base Set of Conditions Assumed in Kinetics Model Simulations. 5-3
5.2 Conditions Assumed in Simulating the Transport of Various 5-11
Pollutants.
5.3A The Impact of Worst Case Transport of Various Pollutants 5-15
on Peak Ozone Concentrations.
5.38 The Impact of Typical Case Transport of Various Pollutants 5-16
on Peak Ozone Concentrations.
5.4 The Impact of Ozone Transport in Different Atmospheric Layers. 5-18
5.5 Impact of Ozone Transported Aloft. 5-23
5.6 Impact of Ozone Transported Aloft Under Different Emission 5-26
Conditions.
5.7 Impact of Ozone Transported Aloft Under Different Insolation 5-27
Assumptions.
-------
List of Tables (cont'd)
Page
5.8 The Impact of Ozone Transport Under Varying Conditions -- 5-29
Summary of Results.
5.9 The Impact of Ozone Transport as Air Quality Approaches 5-31
the NAAQS.
VI
-------
Executive Summary
The purpose of this report is to provide more extensive documentation
for many of the assertions made in EPA 450/2-77-021a, Uses, Limitations
and Technical Basis of Procedures for Quantifying Relationships Between
Photochemical Oxidants and Precursors. Several of the assertions are
applicable only to the Empirical Kinetic Modeling Approach (EKMA) de-
scribed at length in that document. Others, such as the prevailing
background concentration of ozone and the greater importance of trans-
ported ozone as opposed to transported precursors, have general signifi-
cance in a number of different procedures for quantifying ozone-precursor
relationships. Specifically, eight issues are addressed:
1. What levels of ambient ozone and precursors are attributable
to natural sources?
2. How significant are natural sources of ozone and precursors
likely to be in contributing to maximum ozone concentrations
observed within and downwind of urban areas?
3. What are typical ratios of non-methane hydrocarbon (NMHC) to
oxides of nitrogen (NO ) in urban and suburban areas?
X
4. How do ambient precursor concentrations vary, both spatially
and diurnally, in and near urban areas?
5. Is there any apparent difference in precursor levels and
NMHC/NO ratios on days experiencing high ozone concentrations
as opposed to other days?
6. Can existing ambient data be used to draw inferences about
ozone gradients upwind and downwind of major cities?
7. What is the relative importance of transported ozone, NMHC,
N0x, NOp/NO ratio and aldehydes in contributing to maximum
ozone concentrations observed in the vicinity of urban areas?
8. How is the impact of transported ozone likely to be affected
by such factors as changes in locally generated precursor
levels, prevailing NMHC/NO ratios, varying diurnal emission
patterns, level of transported ozone aloft, prevailing atmos-
pheric dilution rates, and sunlight intensity?
Vll
-------
To accomplish its purpose, the report is divided into five sections.
Section 1 discusses the purpose of the report and presents the format of
the sections that follow.
Section 2 discusses natural background concentations of ozone and
precursors and their impact on urban ozone levels. It is concluded that
natural ozone in the lower troposphere primarily results from slow
interchange between the stratosphere and troposphere. During the summer
months, natural ozone is typically about 0,04 ppm in rural areas. As a
result of scavenging by locally generated pollutants, natural sources of
ozone probably do not result in more than a 0.02 ppm increase in peak
ozone levels observed in urban areas.
Ambient levels of non-methane hydrocarbons measured in areas not
dominated by urban plumes are quite low, in the order of 0.1 ppmC. Only
a fraction of the measured NMHC in rural areas can be traced to natural
sources of hydrocarbons. It should be pointed out, however, that the
role played by reaction products of naturally emitted hydrocarbons in
the subsequent formation of ozone is still not certain. Nevertheless,
presently available information suggests that naturally emitted hydro-
carbons are not significant factors in the formation of urban ozone.
Ambient levels of NO in rural areas are frequently below the detectable
X
limits of commercially available instrumentation. Hence, naturally
emitted NO is not believed to be an important contributor to the urban
/\
ozone problem.
Section 3 discusses procedures for estimating NMHC/NO ratios for
)\
use with EKMA. Temporal and spatial variations in precursor concentrations
vm
-------
are also presented. Both NMHC and NO exhibit pronounced diurnal patterns
X
in urban areas, with high, or near peak, concentrations occurring in the
early morning (about 6-9 a.m.). Ambient precursor concentrations are
lower in suburban areas, and are lower still in rural areas. Diurnal
patterns in ambient precursors become less pronounced with increasing
distance from a city. Measured 6-9 a.m. NMHC/NO ratios appear to be
A
about 9.5:1 in urban areas. Suburban ratios appear to be close to this
value on days when the suburban location is downwind from the city.
Close agreement among sites is generally observed for 6-9 a.m. NMHC/NO
A
ratios on days with high oxidant.
Differences in ozone and precursor concentrations measured upwind
and downwind of urban areas are described in Section 4. Using available
data, ozone gradients upwind and downwind of a city can only be determined
for a limited number of cities. The problem is primarily one of not
having suitable upwind data. Such measurements as do exist suggest that
as a result of the transport of ozone over long distances (up to several
hundred kilometers), ground level concentrations upwind from a city may
be in the order of 0.05 to 0.13 ppm on days with high ozone. Since only
limited data were available for the analysis, however, higher values
cannot be ruled out.
In Section 5, the kinetics model underlying EKMA is used to estimate
which facet of long-range transport is likely to exert the most significant
impact on maximum ozone concentrations in urban areas. Simulations are
also performed to determine how the impact of transported ozone on
maximum urban ozone concentrations is affected by changes in local
-------
precursors and meteorology. The results of these simulations are somewhat
qualitative, since the model used assumes instantaneous, homogeneous
vertical mixing of pollutants beneath the atmospheric mixing depth. If
the actual time required for thorough vertical mixing is as great as or
greater than the time required for chemical reactions to occur, some
uncertainty is introduced.
The results of the simulations suggest that ozone transported aloft
overnight (i.e., above the morning mixing height but below the afternoon
mixing height) is generally the most significant facet of long-range
transport. The impact of transported aldehydes were found to be insignifi-
cant. The impact of 0.1 ppmC of NMHC (believed typical of levels found
in rural areas) on peak ozone concentrations was found to be quite low,
in the order of 0.01 to 0.02 ppm ozone. If a monitoring site is located
within an urban plume (e.g., NMHC % 0.3 ppmC and NO ^ 0.04 ppm), under
A
some circumstances the impact of transported NMHC and NO on peak ozone
X
concentrations may approach or violate the NAAQS for oxidant.
For the range of conditions examined, the simulations suggest that
ozone transported aloft overnight may be from "20 to 70% additive." For
example, if 0.10 ppm ozone were transported aloft and were "50% additive,"
its impact on maximum ground level ozone concentration in the urban area
would be 0.05 ppm. Modeling results imply that the impact of ozone
transported aloft increases: a) as local precursor levels are decreased,
b) as the NMHC/NO ratio decreases, c) as the concentration of transported
A
ozone increases, and d) as the atmospheric dilution rate increases.
These results imply that, as local emissions decrease in future years,
-------
the relative impact of a given transported ozone concentration vis a vis
local emissions may increase. However, as control programs are imple-
mented upwind, the transported ozone concentration can be expected to be
reduced. The result is likely to be a reduced impact of transport in
future years.
-------
1.0 Introduction
In a companion report, Uses, Limitations and Technical Basis of Procedures
for Quantifying Relationships Between Photochemical Oxidants and Precursors,
EPA-450/2-77-021a, several methods for relating ambient oxidant (measured
as ozone) to non-methane hydrocarbons and oxides of nitrogen have been
described. That document discusses Eulerian and Lagrangian photochemical
dispersion models, an Empirical Kinetic Modeling Approach, linear rollback
and statistical approaches. Procedures for integrating long-range transport
considerations into these approaches are also described. Because the EKMA
approach is less well known and documented than many of the other methods,
it is described in greater detail. In the companion report, the basis for
a number of statements concerning the propriety of several EKMA assumptions
was not fully documented. In addition, a number of assumptions concerning
the behavior and prevalence of ozone and precursors are common to the other
approaches as well as to EKMA. To be more specific, the following questions
are implied in the previous document:
1. What is the basis for the assertions concerning natural background
concentrations of ozone and precursors?
2. What levels of NMHC, NO and ozone appear to be transported over
long distances? x
3. How do urban NMHC and NO concentrations vary spatially and
temporally? Do these variations support assumptions made in
EKMA?
4. Which facet(s) of transport exerts significant impacts on the
urban ozone problem, and what is the basis for the conclusions
drawn about this issue?
5. To what extent is the impact of transported ozone likely to vary
under differing levels of local emissions and meteorological
conditions?
1-1
-------
The purposes of this report are: 1) to document more extensively the
basis for some of the assumptions which were made in the previously referenced
companion document, and 2) to answer as fully as possible, given the present
state of knowledge, a number of questions which may have arisen from a
reading of the companion document. The report is divided into this and
four additional sections. Section 2 discusses natural background levels of
ozone and precursors. The results of field studies, laboratory investigations
and kinetics model simulations are used to estimate the likely significance
of natural background concentrations in urban areas. Section 3 discusses
procedures for estimating NMHC/NO ratios for use with EKMA. Observed
A
temporal and spatial distributions of NMHC and NO concentrations are also
A
presented, with particular emphasis on data collected during the St. Louis
Regional Air Pollution Study (RAPS). Section 4 discusses differences in
ozone and precursor concentrations observed upwind and downwind of several
large cities. Section 5 describes the results of mathematical simulations
which have been obtained using the chemical kinetics model underlying EKMA.
These simulations are used to estimate the impact of transported ozone,
NMHC, NO , prevailing N09/N0 ratio and aldehydes on maximum ozone concen-
A L.- A
trations observed in urban areas. The degree to which the impact of trans-
ported ozone may be affected by changes in levels of locally generated
precursors, NMHC/NO ratios, diurnal emission patterns, atmospheric dilution
A
and sunlight intensity is also reported. These latter observations are
used to suggest the importance which transported ozone may assume in future
years when the National Ambient Air Quality Standard (NAAQS) for photo-
chemical oxidants is approached in the vicinity of urban areas.
1-2
-------
As with any photochemical model, it is difficult to verify the EKMA model
because of the rarity of suitable observed data. However, model verification
is essential to assure that the approach provides reasonable results.
Various EPA efforts are planned to verify EKMA. These efforts are being
directed along three lines of inquiry. These include: (1) comparisons of Los
Angeles ambient trend data with predictions using EKMA isopleths, (2) combining
of the kinetics model underlying EKMA with observed meteorological and emissions
data and comparison of the estimates with ambient air quality collected in the
RAPS monitoring network, and (3) comparisons of EKMA predictions with those
obtained from more complex Eulerian and Lagrangian models. Results of these
efforts will be reported as they become available.
Presently, only one comparison has been completed, the results of which
are described in a recent EPA report.* This effort compared the standard
isopleth version of EKMA for the basin and fixed sites with historical Los
Angeles ambient data. Considering the potential errors in estimating oxidant
and precursor changes for this analysis, the conclusion is that the basinwide
isopleth predictions "are not inconsistent with historical trends." However,
comparisons made at some specific sites for fixed irradiation times have
indicated that the standard isopleths tend to underestimate the oxidant level
improvements in Los Angeles. The recommended procedure is to apply EKMA
basinwide. Therefore, the verification results for the basinwide application
are most pertinent.
* J. Trijonis and D. Hunsaker, Verification of the Isopleth Method
for Relating Photochemical Oxidant to Precursors, EPA-600/3-78-019,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, February 1978.
1-3
-------
These results represent an initial review of the EKMA approach. Other
verification efforts are needed and are underway. More sophisticated compari-
sons of EKMA are being planned with the RAPS St. Louis data and in other urban
areas where a large data base exists. As previously mentioned, the results of
these verification studies will be reported as they become available.
1-4
-------
2.0 Natural Background Concentrations of Ozone and Precursors and
Their Resulting Impact on Urban Ozone Levels
2.1 Introduction
2.1.1 Purpose
The purpose of this section is to explain the basis for recommending
natural ozone (CL) and naturally emitted ozone precursor concentrations
that should generally be assumed to be transported into urban environments.
These values may be used in the application of various procedures described
in the companion document for quantifying oxidant-precursor relationships.
This section will describe the general circumstances for which the recom-
mended values are applicable and will identify possible exceptions. The
focus is on estimates of natural background values transported into urban
areas, since modeling approaches for oxidant control plan development will
largely be confined to urban areas and their environs, where oxidant problems
will tend to be most severe. Transported anthropogenic ozone is considered
separately (in Section 4), because, unlike natural ozone, it is likely to
be more area-specific.
2.1.2 Background
2
The original Appendix J method for estimating oxidant control require-
ments assumed zero background for CL and non-methane hydrocarbons. The
method did not consider oxides of nitrogen. When the method was conceived,
it was generally assumed that morning precursor emissions in the urban
center were associated with high afternoon oxidant concentrations in the
same central area. At that time, few oxidant or CL monitors existed beyond
the urban centers to suggest that transport could cause maximum concentrations
at significant distances downwind. The few oxidant measurements that were
2-1
-------
available for remote areas indicated a very low rural background (^ 0.02 ppm).
For the urban center, it was assumed that, during the night and early
morning, any ambient 0^ remaining from the preceding day would be completely
scavenged principally by reactions with nitric oxide (NO). This assumption
seemed to be borne out by existing surface measurements. Little evidence
existed of significant influence from 03 layers aloft or long-range horizontal
transport from other areas (urban or rural). However, under the stagnant
conditions generally assumed to accompany peak 0- concentrations, transport
considerations would be minimal. Therefore, on the basis of the information
then available and the assumptions deemed appropriate at the time, the zero
03 background value for urban areas was believed to be reasonable. Data on
NMHC were almost nonexistent outside urban centers but were surmised to be
low. Hence, a zero background was also assumed for NMHC.
More recent analyses of field data indicate that a zero background
^j r c ~i n Q
assumption for 03 is inappropriate. ' ' ' ' ' These investigations show
the following:
1. In many parts of the country during the oxidant season, rural
03 levels often violate the ambient photochemical oxidant standard. Such
rural 0-, levels reflect 0^ of anthropogenic origin as well as a natural
component. Measurements Tn remote areas indicate that the natural component
is greater than previously believed, although well below the NAAQS level,
except under infrequent circumstances unlikely to coincide with the season
of significant photochemical oxidant formation and buildup from anthropogenic
sources.
2. Maximum ozone concentrations resulting from precursor emissions
in medium/large urban areas usually occur several kilometers downwind of
urban centers (typically 15-30 kilometers). Even during light and variable
wind conditions, maximum concentrations are not normally found in the urban
centers, presumably because of the strong scavenging effect of NO emissions
throughout the day.
3. During wind transport conditions, rural ozone, consisting of
natural and anthropogenic components, is transported into urban areas and
contributes to the maximum 0., levels normally found downwind of the urban
2-2
-------
centers in the afternoon. The contribution of the rural 0- may not be en-
tirely additive due to the scavenging effect, especially of NO upon 03
transported into the urban area.
4. In a common air mass, the ozone concentration of the air
found aloft over urban environments in the early morning tends to resemble
that found near the surface in surrounding rural environments during the
preceding afternoon hours. This apparently results from well developed
vertical mixing during the afternoon, distributing pollutants rather uni-
formly over a relatively deep layer (typically 1-3 km). This uniform
vertical distribution of pollutants is transported into urban areas during
the night, but stability stratification separates or decouples the layer
aloft from that near the surface. The surface layer undergoes rapid chemical
scavenging upon entry into the urban area, while the layer aloft remains
virtually unaffected. (The above description reflects a phenomenon observed
with typical wind transport conditions for urban environments not directly
affected by plumes from other nearby urban areas.)
5. In most urban areas, maximum CL concentrations occur on warm
days during which the mixing height increases severalfold from morning to
afternoon. For instance, three-to fivefold increases in mixing height are
common in most inland locations in the eastern U.S. Hence, progressive
mixing with the air aloft should occur during downwind transport of the
urban plume accompanied by a contribution of the 0, mixed from aloft. The
significance of the (L entrained from aloft is a function of the initial 0.,
concentration aloft, the depletion effects of dilution during the mixing
process and scavenging (principally by NO) during the morning hours.
Presently, there is still a very limited amount of precursor data
available for determining levels of NO and NMHC advected into cities.
A
Those data that are available are discussed in this Section and in Section 4.
These data indicate that, while the incoming precursor values may not be
zero, they are considerably less than concentrations resulting from pre-
cursors emitted in the urban area. In the case of NMHC, organic species
data that have been collected indicate that only a small fraction of incoming
NMHC is of natural (vegetative) origin.
2.2 Natural Ozone in Rural Areas
Since anthropogenic 0, predominates in urban areas, the natural 0,
component is difficult, if not impossible, to estimate from urban measure-
ments. Therefore, ambient measurements in rural areas are essential to
2-3
-------
develop estimates of natural CU that may be transported into urban areas.
Natural ambient levels of 0-,, as may be reflected by measurements in more
remote rural areas, are believed to result principally from two sources:
(1) through interchange of air between the stratosphere and the troposphere,
and (2) the synthesis of naturally emitted organic precursors (e.g., from
vegetation). Other possible sources that have been mentioned are methane
oxidation, lightning and surface seepage of hydrocarbons, but there is no
7811
physical evidence of appreciable ozone production from such sources. ' '
Sinks for 0, occur as a consequence of contact with surface features and of
reactions with certain gases or aerosols. Surface features are considered
more important as sink mechanisms in rural areas, while in urban areas,
reactions with NO appear to have a greater scavenging effecc.
A description of the principal natural 03 sources and impacts in rural
areas is presented in the following sub-sections.
2.2.1 Stratospheric Source
A well known large reservoir of 0., exists in the stratosphere. Stratos-
pheric 0~ may reach concentrations of 10 ppm or more. This layer is usually
12
found at altitudes of 20-50 km. Formation of ozone within the stratosphere
occurs by reactions of ultraviolet light and oxygen. A usually abrupt
increase in atmospheric stability at the boundary of the stratosphere and
troposphere, termed the tropopause, normally prevents large quantities of
ozone from descending into the troposphere. However, by various mechanisms
which are not very well understood, there is some interchange between the
layers. Much of the 03 descending into the upper troposphere eventually
reaches the lower troposphere (or biosphere) as a result of slow diffusion
2-4
-------
and turbulent mixing processes. Dilution during downward transfer consider-
ably reduces the high concentrations sometimes found in the vicinity of the
tropopause.
Aside from the usual slow downward transfer of CU from the stratosphere,
there is superimposed a meridional, seasonal circulation. During winter in
polar regions, where the tropopause slopes toward lower elevations, relatively
large quantities of CL are transferred downward through the tropopause.
The relatively low altitude of the tropopause and frequent stormy conditions
in the polar winter make it easier for CL to penetrate downward through the
tropopause as a result of turbulent mixing. This 0~ gradually drifts
southward with the weather systems, reaching an early spring maximum at
mid-latitudes. Later, during the oxidant season, this meridional (L
12
transport decreases, reaching a seasonal minimum during the fall.
During early spring months, the maximum hourly values at remote sites may
approach the NAAQS. However, during the oxidant season, the stratospheric
contribution to ambient CL levels is normally well below the level of the
NAAQS.
Occasionally, intense tropospheric weather phenomena (e.g., thunder-
storms, jet streams, fronts) may cause turbulent, rapid "intrusions" into
I O
the lower troposphere of unusually large quantities of ozone. Ozone in
these intrusions may reach the surface in only a slightly diluted form.
However, such intrusions are sporadic, rare events. Also, the meteoro-
logical conditions that accompany these acute intrusions are not generally
conducive to ozone formation and buildup from anthropogenic sources.
There are presently no well accepted physical models for stratospheric-
2-5
-------
tropospheric ozone exchange that can account for the tropospheric ozone
burden. However, most experts tend to agree that the stratosphere is the
origin of the major portion of the natural ambient CL.
2.2.1.1 Effect of Slow Transfer/Diffusion Process
Numerous studies of (L measured by various means have been used to
estimate the stratospheric CL contribution to natural 03 background, parti-
cularly during the photochemical oxidant season. Techniques used to make
such estimates include 0^ measurements in remote areas not likely to be
influenced by vegetation (e.g., over oceans), ' analysis of ozonesonde
Q
balloon measurements made during the 1960's, and use of radioactive fallout
data to assess the degree of interchange between the stratosphere and the
o
earth's surface. In addition, some investigators have measured vertical
profiles of 0, in clean air above a low subsidence inversion and have
g
extrapolated those profiles to the surface. All of these measurements
indicate that typical or usual (L levels are attributed to slow transfer/
diffusion processes, rather than intrusions. Results from these studies,
as well as measurements made in remote land locations, are summarized in
Table 2.1. A range for stratospheric contribution from 0.02 to 0.05 ppm is
indicated as typical except for strong surface scavenging accompanying
nocturnally formed, low-level stable layers. The estimates generally apply
to the photochemical oxidant season.
2.2.1.2 Effect of Stratospheric Intrusions
It has been demonstrated by numerous studies, including those reviewed
8 13
by Reiter and by Davis, et a!., that intrusions of stratospheric aif
introducing unusually large concentrations of 0., into the troposphere are
2-6
-------
I
I
H
S
I H
(0
. O
I
I
tr
H
^ ^
E
a.
a.
z
o
i i
I
CQ
14
fv'
U-
h-
O
<_>
LU
O
a:
^3
o
co
a
LU
1
^gr*
» i
i fi
w J
LU
r""
(O
4->
O
1 '
U
r
s-
O)
(~
J
Q.
in
O
r
CQ
O
f
S-
cu
^~
lA II
CL
to
0
4->
(O
j_
4->
CO
CO
2:
o
1 I
1
<
o
o
_l
co
o:
1 O
-<
Z h-
i i co
ct: LU
a. >
z.
i »
**£.-
O
ir Lf^ ir> ^ 'o
C O O O O
lilt 1
* * * *
* * * *
CM CM CO i i
O O O O O
»
LO
O
1
CO «3~
O O
v'l
-M
>- to
(O
SO 4->
O to
«> +J tO
-U CU CO
)-> to t- O O
2: . ai o 2:
03 H-
QJ ^?~_ * ** r^
<_> -3
tO « CO 4J -P C3
«f- C S Q.
CU O rD
4-> 4-5 Jii if- 4->
r- j* q- o i C
jC r if- 3 O CU
3 LU O CQ 30
f CM OO «3- r CM
3- ^D
r CU
. CU 1 r-
r- 4-> IO O>
, 4-5 W-^
4-> CO 4-> CU T- CU
QJ T- S- 4->
E in ( 3
"OS- 4J
S +-> CU ^ J= T-
LU CD > O O 4->
co E !- H- s- in
co T- E D; to E
rs J^ ID LU cu i i
H to Q. in
CO IO D- QJ
eC 3 i i Ct:
cx a: ^-^
«
o
i
«* co
O O
v|
E
O CU
r- TJ
to E
S- 0
CU to
> CU
E E
I-H o
N
O
^
(O
CJ CO
^3
co
to
CU
> o
o vo
-Q CT>
*Y* ^_
"^ r
r^
*
i
fO
4->
a;
««
i
<
f^ CO
r^ ctC
LU LU
s: t
ra i
_J LU
CQ c;
_i
IS)
s:
.
o
o
o
vo
V
to
4->
E
CU
E
CU
3
to
to
jjj
"-
QJ
+J^ v
fO >s
4J 4J
co -i
to
O S-
o cu
10 >
O E
o
o
1C*
o
1
*
*
1
O
,_
0
V
LO
0
1
CM
O
CO
^3
co
^^
LU
CO
O
0
w
01
to QJ
3 -a
o o
s- to
O -<~
cn a.
r- CU
>
c
-E O
4-> <-
» i \
r (-*
S 3
p
4-> i
3 O
JQ a.
CU
a -a
0) E
S- E
CD to
£
JQ
O M-
O
E
CU U
a; r-
-0 4->
in
CO i
> i.
tO
i
CU T3
> E
O O
-Q O
tO
cn
tO E
CU X
> -r-
CU E
r
o
"io si
E CD
0 J=
r- Q.
to to
03 O
U E
O 4->
O tO
*
CU
o
ra
<+-
3
to
s~
to
CU
E
j^
O
4-»
to
cn
E
r
cn
E
CU
>
n}
u
to
CU
r
-P
4->
^:
cn
E
+->
o
CU
<^
CU
>,
io
CU
cn
E
s-
cu
o
_J
K
*
2-7
-------
caused by penetration of the tropopause by strong jet streams, deep frontal
systems and thunderstorm cell tops. These weather phenomena are most
prominent during the late winter and early spring months. Ozone measure-
ments at rural sites that seem to be associated with intrusions have exceeded
0.15 ppm for periods of from a few minutes to an hour or more. However,
these have been rarely observed, and the spatial extent of impact at the
surface is believed to be very limited, a few kilometers at most.
For urban oxidant control strategy planning, intrusion events can
usually be disregarded. Occasionally, peak observed ozone levels in the
data base may be identified with an intrusion. Such data (like that of
other sporadic, natural events that may cause natural pollutant contri-
bution to be temporarily high, i.e., dust storms and forest fires for
particulate matter), should not be used in strategy development.
2.2.2 Natural Sources of Precursors
Nationwide estimates of total natural NMHC and other organic emissions,
mainly from vegetation, are rough approximations. Nevertheless, the avail-
able estimates are consistent in indicating that global natural emissions
14 15
are several times greater than manmade emissions. ' This does not
necessarily imply that the emission densities or ambient concentrations of
natural hydrocarbons will be high. Because natural emissions are spread
over large areas, emission densities are low. In the vicinity of populated
areas, emission densities of manmade sources may be considerably greater
than those of natural sources. For example, an estimation for St. Louis
indicates vegetative emissions may account for only about five percent of
the total hydrocarbon emission burden from all sources within the metropolitan
area.16
2-8
-------
Vegetation emits some hydrocarbons, such as isoprene and various terpenes,
which are known to be highly reactive. However, in rural areas, which tend to
have extremely low, often underectable, NO -;0'icentrations, these natural
A
hydrocarbons may have more of a tendency to serve as a sink for 03 by reacting
with it to form aerosols. Recent smog chamber experiments tend to substantiate
17 ??
this point. One question that remains largely unresolved is how natural
organic emissions may react in and near urban areas (or other areas with sub-
stantial NO ). Some smog chamber experiments suggest that, in the presence of
.A
NO , reactive hydrocarbon compounds emitted from vegetation (i.e., terpenes and
/\
TO
isoprene) are capable of forming 0,- However, atmospheric concentrations of
these natural compounds in urban areas appear to be very low and are frequently
undetectable. A reasonable hypothesis is that these reactive hydrocarbons
have already reacted with 03 to form aerosols, or have been removed by other
means, and hence are not present in urban areas to form additional ozone.
Some natural NO can result from biological action in soils, but N09
A C-
levels in ambient air in remote areas are usually below the detectable limit
(0.005 ppm) of commercially available ambient measurement techniques. Thus,
the natural NO contribution to ozone production is largely unknown, but is
J\
believed to be insignificant. Hourly levels of N02 reported for some eastern
U.S. rural areas have infrequently ranged up to about 0.05 ppm, but such high
rural levels may reflect the occasional effect of transport from urban sources.
2.2.3 Estimated Impact of Natural (Vegetative) Precursor Sources
A number of approaches have been employed to evaluate the impact of
vegetative organic emissions on ambient 0., levels. Basically, these have
involved (1) direct measurements of emission rates, (2) smog chamber experi-
ments, and (3) ambient air studies.
2-9
-------
Natural organic emissions from vegetation over the entire land area of the
U.S. are estimated on an absolute weight basis to be greater than from anthro-
pogenic sources. But in and near urban areas, emission densities of organic
emissions from vegetation will be much lower than from manmade sources.
Comparative emission densities alone would indicate a low potential for natural
organics to form high concentrations of 0,.
Limited smog chamber data have indicated that terpenes and some other
organic species are photochemically reactive. However, field studies generally
conclude that very low concentrations of these organics are found in the
atmosphere, even in rural areas. In smog chamber experiments, both very low
and high terpene/NO ratios tend to inhibit CL formation. Smog chamber work by
A O
Washington State University and EPA groups shows that, even at optimum terpene/NO
ratios of about 20:1, 03 formation was very slight relative to other classes of
hydrocarbons. When typical rural terpene/NO ratios were tested in the chambers,
X
very little 0, was formed, with a great deal of the carbon present involved in
aerosol production. At optimum terpene/NO ratios and 100 ppbC, terpene con-
/\
centrations, smog chamber simulations indicate that the resulting 0^ concen-
18
trations would not exceed 5 ppb.
Too little information is available from smog chamber studies to determine
the effect of mixing natural hydrocarbons with urban air. Evidently, natural
hydrocarbons both enhance ozone formation through photochemical reactions as
well as scavenge available 0~. It is not known at present which effect may
predominate in an urban atmosphere. Of course, in urban areas, vegetation is
much less abundant than in rural areas. Also, as indicated previously, it is
unlikely that natural hydrocarbons from rural areas can survive long enough to
be an important influence on urban ozone formation.
2-10
-------
Ambient air studies have shown that vegetative-type hydrocarbons can be
detected in or near forested areas. However, the atmospheric burden of these
natural hydrocarbons is generally very low (from 10 to 50 ppbC), even in close
proximity to densely wooded areas. Also, they generally constitute a small
portion of the measured NMHC. The mix of hydrocarbons found at a rural site in
Ohio was a dilute urban mixture. Data analyzed for rural sites in Florida,
North Carolina, Montana, Iowa, Louisiana and Pennsylvania show early morning
17 23
levels of non-methane organics typically in the range of 50-200 ppbC. '
These levels are well below those observed in urbanized areas, which were
typically 5-12 times higher. In the extreme case, Los Angeles ambient levels
fi 17 ?"3
were about 31 times those observed in rural areas. ' '
Because of their relatively low densities of emission and their extremely
low concentrations in ambient air, both in absolute terms and in relation to
the manmade organic burden, as well as their low ozone forming potential in
areas with little NO (i.e., rural areas), natural organics are not likely to
x
1Q
be significant sources of natural 03 levels. Simple modeling estimates
based on measured emission rates indicate that the resulting 0, concentrations
will be extremely low, because of vertical dilution effects and because of
reactions with background 0., and hydroxyl radicals. These results suggest that
terpenes and other natural hydrocarbons may serve principally as a sink for 07
<3
in rural areas.
Ambient ozone measurements taken in remote land and ocean areas (Table 2.1)
also lead to the conclusion that vegetation is a negligible source of 0~
0
formation. In this table, ozone concentrations measured in remote land areas
(where urban tracer levels such as fluorocarbons indicate no detectable urban
2-11
-------
influence) can be compared with those from remote ocean areas (where vegetative
hydrocarbons are nonexistent). Only slight differences are evident. The
contribution of natural organic emissions to ambient ozone concentrations is
estimated to be less than 0.01 ppm. This leads to the general conclusion that
the most significant natural source of background 0., is the stratosphere.
2.3 Estimated Impact of the Transport of Natural Background Concentrations
into Urban Areas
From the information presented in Table 2.1 and other summaries of natural
11 1Q 21
source impact, '' the consensus is that the natural ozone contribution in
nonurban areas, in most cases, will range between 0.02 and 0.05 or 0.06 ppm
during the season and the part of the day that maximum ozone concentrations
resulting from manmade sources will tend to occur. The most likely value is
estimated to be 0.04 ppm.
An estimate can be made of the impact of the natural 0^ component in the
urban plume at the location and time of maximum 0., concentration. This esti-
mate is based on the following set of assumptions:
1. The initial surface concentration of ozone in the urban plume
during the 6-9 a.m. period of maximum precursor emissions is near zero.
2. Concentrations of natural Oo very similar to those found in
surrounding nonurban areas will exist a few hundred meters above the surface in
the urban area during the early morning hours. This layer is separated by an
inversion layer from freshly emitted NO and other surface scavengers.
3. As downwind transport occurs and surface heating proceeds during
the day, the mixing depth will increase, so that considerable mixing with the
air aloft will occur. The total increase in the mixing height will vary in
extent from a fraction of the initial height in a few locations to severalfold
in most locations.
4. Scavenging by NO during the early stages of plume transport is
significant. However, downwind from the urban areas in the afternoon, it is
likely to be of less importance because of lower NO emissions.
5. Relative to vertical mixing, lateral mixing with background air
2-12
-------
at the urban plume center!ine is considered negligible over the distance
traveled to the location of the maximum 03 concentration.
A scenario depicting the above assumptions is shown in Figure 2.1.
Simulations of this scenario utilizing the kinetics model underlying EKMA
(discussed in Section 5} indicate that, for most U.S. cities, the impact of the
background component is typically about 40-50 percent of the rural value when
added incrementally to the urban plume, assuming severalfold daily changes in
mixing height typical of most U.S. locations. However, in a few urban situ-
ations where the mixing height increases only slightly during the day (i.e.,
Los Angeles), the impact will be even less. Taking the more typical conditions
above, where the impact of the natural component will be about 50 percent of
the transported concentration, and taking the most probable value of 0.04 ppm
for natural background, yields a suggested value of 0.02 ppm for the impact of
natural background on maximum 03 concentrations observed in urban areas. As
indicated in the previous discussion, a local condition may justify using a
slightly different estimate.
2.4 Summary and Conclusions
Natural background levels of ambient ozone (03) appear to result mainly
from stratospheric-tropospheric exchange processes and slow transfer into the
lower troposphere. Concentrations of ambient 03 attributed to stratospheric
origin are estimated to be generally well below the NAAQS level for photo-
chemical oxidant, particularly during the oxidant season. However, on rare
occasions, more abrupt intrusions from the stratosphere can cause localized
elevated 0., levels which may violate the NAAQS. In rural areas, scavenging by
surface features provides the primary sink for natural 03- Natural organics
(i.e., terpenes) and/or their aerosol products may also scavenge 03. In urban
2-13
-------
CO
O
cc
«t
a
2
=3
O
en
14-'
O
H
-S
U
^
-S
O
1
O
I
H1
.s
S
2-U
-------
areas, 0, reaction with nitric oxide (NO) emissions provides the dominant
O
scavenging mechanism, although N02 produced by this scavenging reaction may at
a later time regenerate 03 through the photolytic reaction cycle.
In urban areas, early morning 0, concentrations near the surface tend to
approach zero. This results from the usual nighttime stratification of the
lower atmosphere which allows efficient scavenging of the available 03 by NO
and surface features within a confined surface-based layer. However, a reser-
voir of 0, remains aloft and insulated from these scavengers. This 0., reservoir
consists of natural 0, and 03 previously synthesized from anthropogenic pre-
cursors. These two components are usually considered as "transported CL"
During subsequent midday hours, the 0^ in the reservoir aloft is mixed downward
into the surface-based layer. This mixing will affect the concentration of
photochemical pollutants in the surface layer and will cause an increase in the
afternoon maximum 0, concentration observed in the vicinity of the urban area.
The impact of transported pollutants is discussed more fully in Section 5.
The consensus of existing experimental data is that, on those afternoons
conducive to high photochemical activity, the most likely range of natural 0,,
O
in rural areas is 0.02 to 0.05 or 0.06 ppm, with 0.04 ppm the most likely value
to assume. Nearly all of this natural Oo background is estimated to be of
stratospheric origin. Synthesis of natural precursors, mainly of vegetative
origin, is estimated to contribute a negligible increment.
The kinetics model which underlies the Empirical Kinetic Modeling Approach
described in Reference 1 was applied to simulate the effect of the transport of
natural background concentrations of ozone on urban ozone concentrations in
the presence of mixing, dilution and photochemistry. Results indicate the
2-15
-------
impact of transported natural cr?one in nr^ar j-eas is about 0.01-0.03 porn, with
0.02 ppm being the typical value that may be assumed at the location of maximum
afternoon 0~ concentration. Local pecularities in meteorological conditions,
levels of locally generated precursors and other factors may, in a few circum-
stances, justify slight deviations from the estimated 0.02 ppm value.
The sparse data available on precursor levels in rural areas that may be
transported into urban areas indicate that these concentration levels are low.
Moreover, only a small fraction of these low levels, at least of NMHC, have
been traced to natural sources. Typically, rural levels of non-methane hydro-
carbons (NMHC) appear to range from 50 to 200 ppbC. Slightly higher values may
be found in the periphery of urban areas (discussed in Section 4). Nitrogen
dioxide concentrations in rural areas are often near the lower threshold for
commercially available instruments. However, infrequent hourly concentrations
up to about 0.05 ppm may occur in some rural areas, particularly within a
reasonable range for transport from urban areas (discussed in Section 4). The
precursor levels typically found in rural areas are considerably less than
those typically found in urban centers. Simulations with the kinetics model in
EKMA (shown in Section 5) suggest that typical rural precursor levels will
generally not contribute significantly to the synthesis of high ozone levels if
transported into an urban area.
-------
2.5 References
1. uses, Limitations and Technical Basis of Procedures for Quantifying
ReTatTonThips Between PhotocFemical Oxidants~and Precursors.
EPA-450/2-77-021a, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, November 1977.
2. "Requirements for Preparation, Adoption, and Submittal of Implementation
Plans." m^e 40, Code of Federal Regulations: Part 51. 36 FR 15486
(August 147T97TT~~
3. Air Quality Criteria for Photochemical Oxidants. U.S. Public Health
Service, Washington, D.C., March 1970.
4. E. Robinson and R. A. Rasmussen, "Identification of Natural and
Anthropogenic Rural Ozone for Control Purposes." Proceedings,
Ozone/Oxidants Interactions With the Total Environment Specialty
Conference, Air Pollution Control Association, March 1976.
5. Investigation of Ozone and Ozone Precursor Concentrations at Nonurban
Locations in the Eastern United States, EPA-450/3-74-034, U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina, May 1974.
6. Investigation of Rural LeveU as Related to Urban Hydrocarbon Control
Strategies. EPA-450/3-75-036, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, March 1975.
7. Formation and Transport of Oxidants Along Gulf Coast and Northern U.S.
EPA-450/3-76-036, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, August 1976.
8. E. R. Reiter, "The Role of Stratospheric Import on Tropospheric Ozone
Concentrations." International Conference on Photochemical 0x1dant
Pollution and Its Control, Proceedings: Volume I, EPA-600/3-77-001a,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, January 1972. /
9. D. L. Blumenthal, W. H. White, R. L. Peace and T. B. Smith, Determination
of the Feasibility of Long-Range Transport of Ozone and Ozone Precursors.
EPA-450/3-74-061, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, January 1974.
10. G. C. Holzworth, Mixing Heights, Hind Speeds, and Potential for Urban
Air Pollution Throughout the Contiguous United States. AP-101, U.S.
Environmental Protection Agency, Research Triangle Park, North Carolina,
January 1972.
11. H. B. Singh, F. L. Ludwig and W. B. Johnson, Ozone in Clean Remote
Atmospheres: Concentrations and Variabilities. Final Report prepared
for Coordinating Research Council, Stanford Research Institute, Menlo
Park, California, March 1977.
2-17
-------
12. H. U. Dutsch, "Photochemistry of At»io*pher ic ^oin?." ^dwncps in
Geophysics. 15, 218-32? (1971).
13. D. R. Davis and R. E, Jensen, "Low-Level Ozone and Weather Systems."
EPA-600/3-77-001 a, ibid.
14. E. Robinson and R. C. Robbins, Sources, Abundance and Fate of Gaseous
Atmospneric Pol 1utants. Stanford Research Institute, Menlo Park,
California, February 1958. Supplement, April 1971.
15. P. Zimmerman and R. A. Rasmussen, "Measuring Organic Emissions from
Vegetation." Progress Report No. 4, EPA Contract 68-02-2071, Research
Triangle Park, North Carolina, November 1976.
16. Personal communication, Thomas Lahre to E. J. Lillis. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, March 1977.
17. W. A. Lonneman, "Ozone and Hydrocarbon Measurements in Recent Oxidant
Transport Studies." EPA-600/3-77-001a, ibid.
18. H. H. Westberg, The Issue of Natural Organic Emissions. EPA-600/3-77-116,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, December 1977.
19. Personal communication, J. J. Bufalini and W. A. Lonneman to E. J. Lillis.
Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, March 1977.
20. R. M. Angus and E. L. Martinez, "Rural Oxidant and Oxidant Transport."
Proceedings of Conference on State-pf-the-Art of Assessing Transportation
Related AiV QualTty Impacts, SpeciaT Report 167, National Research Council,
Washington, D.C., October 1976.
21. C. W. Spicer, "Experimental Evidence of Long Distance Pollutant Transport."
Presented at Annual Meeting, American Institute of Chemical Engineers,
Houston, Texas, March 1977.
22. B. W. Gay, Jr., and R. R. Arnts. "Chemistry of Naturally Emitted
Hydrocarbons." International Conference on Photochemical Oxidant
Pollution and It? Control. Proceedings: Volume II. EPA-600/3-77-001b,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, January 1977.
23. W. A. Lonneman, R. J. Seila and J. J. Bufalini, "Ambient Air
Hydrocarbon Concentrations in Florida." Unpublished manuscript.
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina.
-------
3.0 Estimating NMHC/NO Ratios For Use in the Empirical
Kinetic Modeling Approach (EKMA)
3.1 Introduction
3.1.1 Purpose
Recommendations concerning the use of non-methane hydrocarbon to oxides of
nitrogen ratios in the Empirical Kinetic Modeling Approach (EKMA) have been
made in Uses, Limitations and Technical Basis of Procedures for Quantifying
Relationships Between Photochemical Oxidants and Precursors. The purpose of
Section 3 is to present the results of analyses of existing NMHC and NO data
A
used to develop recommendations on the appropriate NMHC/NO ratio for use in
A
EKMA (e.g., calculation procedures such as appropriate averaging time, site
selection, number of sites to be examined, etc.). This section also provides a
review of temporal and spatial distributions of NMHC and NO in St. Louis and
A
ambient data for estimating default NMHC/NO ratios. The following three
A.
issues are addressed in this Section:
1. What are typical non-methane hydrocarbon to oxides
of nitrogen ratios in urban and suburban areas?
2. How are ambient NMHC and NO concentration data used to esti-
A
mate NMHC/NO ratios for use in EKMA?
X
3. How can a city without ambient NMHC and NO data apply EKMA?
A
3.1.2 Background
EKMA uses a set of ozone isopleths which depicts maximum afternoon con-
centrations of ozone downwind from a city as a function of initial concentra-
tions of NMHC and NO . The isopleths reflect specific assumptions about the
A
temporal distribution of emissions, meteorological conditions, reactivity of
the precursor mix and concentrations of ozone transported from upwind areas.
3-1
-------
There are two forms of EKMA, the city-specific isopleths described in Section 5
and in Reference I, and the Standard Isopleths shown in Figure 3.1. Use of
either set of isopleths for estimating the degree of reduction in ambient
precursor levels needed to attain a specified ozone level is similar. That is,
they both require the ozone design value (the highest second highest hourly
concentration observed in or near the city during the base period designated
for control strategy development purposes) and the prevailing NMHC/NO ratio.
X
Reference 1 contains a detailed discussion on monitoring requirements for
determining the ozone design value. Additional guidance on determining design
values is provided in References 7 and 8. Therefore, this Section focuses on
the second input requirement, the NMHC/NO ratio. As reported in the companion
A
document, the appropriate ratio for use with EKMA is the 6-9 a.m. (LOT)
NMHC/NOV ratio observed in the urban core* on the same day as the ozone design
A
value. Concentrations measured at this time in the urban core best reflect
initial conditions prior to the onset of photochemical reactions. Thus, this
ratio is considered to be consistent with the results of smog chamber experi-
ments which describe the sensitivity of afternoon maximum ozone levels to
initial concentrations of NMHC and NO . Further, concentrations of NMHC and
A
NO are typically higher during these hours. Hence, the reliability of the
A
monitoring data is greater.
The recommended procedure for estimating NMHC/NO ratios for use with EKMA
A
was developed using ambient data on prevailing precursor levels in and near
urban areas. Some sources of ambient monitoring data on NMHC and NO levels
A
are SAROAD, the St. LOUIS Regions"" Air Pollution Study (RAPS) and special
* Urban core is defined as the central city. In large cities this may encompass
areas within 3-4 kr,i of the center of the central business district.
LOT is local daylight time.
3-2
-------
"4H
4H r
00
O)
cu
s-
o
to
S-
3
O
, Q.
r- QJ
4-> S-
i- rj
to c/i
C ro
Ol OJ
oo E
CO
LU
Di
iudd 'X
3-3
-------
purpose monitoring programs (e.g., EPA sponsored summer oxidant studies).
The analyses of the spatial and temporal characteristics of NMHC and NO
X
rely primarily on data collected as part of RAPS. At the time the data sup-
porting EKMA were compiled, preliminary data on hourly concentrations of
pollutants (ozone, NO and NMHC) were available for May through October, 1976.
J\
The RAPS monitoring sites, arrayed approximately on four concentric rings
around Site 101 at distances of about 4, 9, 20 and 40 kilometers (km), are
shown in Figure 3.2, along with the surrounding metropolitan area and major
highways. As this figure illustrates, the design of the network is such that
urban-rural gradients can be estimated for each pollutant. Also, the use of
continuous monitoring equipment permits the examination of diurnal patterns in
pollutant concentrations.
3.2 NMHC/NO Ratios For Use in EKMA
/\
The 6-9 a.m. (LOT) NMHC/NO ratio is required to use either the standard
X
or the city-specific isopleths. Because of the importance of the ratio in
EKMA, questions concerning estimation procedures, instrument reliability and
temporal and spatial properties must be addressed. The four key issues listed
below are addressed in the sections which follow:
1) How do NMHC/NO ratios vary in and near urban areas?
/\
2) What effect do problems associated with measuring NMHC using
Flame lonization Detector (FID) instruments have on the estimate of the ratio?
3) How do ratios calculated on high oxidant days compare with
ratios observed on other days?
4) How can a specific city apply EKMA if ambient data on NMHC and
NO are not available?
/\
3-4
-------
RAPS CtNTRAI.
RAMS STATIONS
».«, // t-/.«W!S» J
ir
xftEy / v^~^^iii
^^N^^tes^t^J;
^^-S.-.TTS^;.^
SljJllSw^^
|gMisWilll*u
^^g??p-;«^?:^#,'o,^
^:Wxi-;y:x'.->.»lj?
-------
3.2.1 Spatial Variation in NMHC/NO Ratios
A
As noted in Reference 1, the appropriate ratio for use with EKMA is the
6-9 a.m. NMHC/NO ratio observed within the urban core of the city. Since an
A
estimate of the NMHC/NO ratio characteristic of the city under study is
A
required, it is desirable to examine estimates of NMHC/NO ratios from existing
A
monitoring sites in and near urban areas. A better understanding of the
spatial variability of the ratio should aid in the siting of ambient monitors
and in the interpretation of existing ratio data.
Table 3.1 provides ratio data from the fourteen urban and suburban sites
used to estimate default ratios described in Section 3.2.5. These data indi-
cate that suburban sites experience a wider range of ratios than do center city
sites. Since suburban sites are typically not encompassed by a fairly uniform
emissions field, as center city sites are, this wider range probably reflects
the greater dependence of the ratio on wind direction. Most of the suburban
sites in Table 3.1 are in the predominately downwind direction from manmade
sources of precursors. This may account for the relatively small variation
among seasonal median ratios at most of these urban and suburban sites. However,
high ratios are more likely on the relatively infrequent occasions when the
suburban sites are upwind from the city. Table 3.2 provides a good example of
the dependence of NMHC/NO ratios on wind direction at the Portland, Maine,
A
o
site. The median 6-9 a.m. NMHC/NO ratio for five days when the monitor is in
A
the Portland urban plume is 11:1, whereas, on other days the median ratio is
31:1. Because of the sensitivity of the NMHC/NO ratio to wind direction at
A
suburban sites, the most suitable sites for determining NMHC/NO ratios for use
A
in EKMA are those located in the urban core.
3-6
-------
D
D D
CO
Cfl
C
-P
a
-p
*^
JS
Q>
Q
0
o-j
t_ o
^
X Vt
§ cy o
^il ""pS
£ c=<^ _
~c°£^
a.1 -
^ f-.^ J 5;
c
- a- ^
-tj ^
~- 5
Oi-i/1 td C"O3'
O O QJ ^ ""
C? -) (J i^ T? T.
*n F 3 »xt c c "D
O o IT; O r- (i» ft*
>^ ro i-i- (_ J fO -~ «- ^. v u
a. E <- ro c IT ---a
O---C *-> *J CWO
Q. 3 d iy; L^J X. »-«*.' a
cc en m
c ^r ^r
-- «-
Ln CNJ O
i*-" ro in
en r
m co in'
f~ CTV Pv.
_ r-
un ro in
co CT> rv,
un r-. 10
."O vo ur>
C 0 r-
m cr» ^r
10 p-
r-- <\j
«3- CSJ
o
CO CM
in <-n
o o
ro ro
-x. ^^
^Z. -x.
CO rs.
in ro
2 c
at O
CTt *>
C C
< *t
(/) C
o as
_1 00
' O O
i. CO
*" ° 1* "
-.' O C
C 3 *.
ti E "~ - "a
_
^
^
a
o
^
CTi
^
O
CO
in
o
ro
en
i
--».
o
(D
X
V
>
ai
c
o
<
- i
O re
3 0 ~"'
a3 O *4-
300
m LU
^J ^ - UJ
01
"'
I
O
CO
^~
in
CO
10
ro
CM
in
1
S
vO
CO
ct
o
TJ
t/T
3
N
<
C, ^ O
ir n c_j
O tii. i
n?
% ^ 'e
o
w-i O CM
Cw
«»
r-
cc
ro
m
o
r-
ro
rv
^O
^.
O
i£>
m
^j-
PN
^,
CM
-v.
PN.
1
S
o
CM
X
O
c
0
4_>
c
HO
O
u
*.»
I/) ^
i*_ O. 03
000
t^ a
"7g~
I- O O
(U -i- *-> -
^3 E >>ro
3 « *J
x cn o rsj
r-
0
00
r-
cb
ro
CM
CO
^
,-
cn
CM
C^
5*
o
en
, in
^j-
r«*
CM
00
CO
r*.
«T
g
c
o
.*->
>)
a
CXI C C'
*-x. -O O
"" !ri a **- x
«j i- ca O Hi
«- O H
LU
Jt UJ 0 2T -
ifl uJ «o (U
a D- - i-
C '- f ^3
« 6 UJ ro -D
_
O
*J
T
o%
CJ
-er
o
«3
r~
o
o
CO
°
VO
^r
U)
Pv
**^
0
ro
^^
Ch
i
*v.
PS,
0*
CM
«-^.
j£
k.
«O
CL.
O -*-1
o ^
« <0
Q. O
(/>
«t
UJ'
in
3t
Q
Z. 0.
Cx. O -»
- O C O
>- U 0 0
c *
UJ <- O O
to o * «*
(
vO
l/>
^_
p*.
^J-
*
_
CO
CO
r-
m
en
en
C\J
o
^>
CM
in
rv
*v
CM
^
i
00
s
in
^
^
e>
c
2
£
(If
P-. *J V
13 O
*» U
0 O E
. ^i
^ 33
r>~
CO
L>1
CM
_
in
CO
r-
*o
P-.
J-v
o
CM
to
to
r^
-^
rn
0
I
to
o
in
A
1
_j
*j
00
c
o
*-t
f
<_>
I
U"
m
o
ro
^_
CM
cn
^
CM
P->.
VO
8
2 (
i
tf> '
^ -i
1
s \
m
«
, »
u
"
1 1
O)
c
!c
IA
3-7
-------
TABLE 3.2 Example of Sensitivity of Suburban Ambient Data to
Hind Direction Portland, Me.
Date
8/7/74
8/19/74
8/21/74
8/22/74
8/23/74
Median
Values
Days When Monitor
is Downwind From
Portland CBD
- NMHC/NC) "
J\
7:1
17:1
6:1
11:1
14:1
11:1
Date
7/30/74 ,
7/31/74
8/2/74
8/6/74
8/11/74
8/14/74
8/24/74
Other Days
NMHC/NO
^
36:1
160:1
23:1
31:1
200:1
23:1
26:1
31:1
3-8
-------
3.2.2 Uncertainty in Individual NMHC Readings
Within the past several years, there has been an increasing awareness of
the problems associated with measuring ambient NMHC levels. Because NMHC is
not measured directly in FID instruments but is computed from the difference
between total hydrocarbons (THC) and methane (CH.), errors in the individual
values for each of these pollutants may be compounded when NMHC concentrations
are calculated. This may even result in negative values for estimates of the
NMHC concentration. The many operational and mechanical problems associated
3
with FID instruments are discussed in detail elsewhere.
Studies of instrument response characteristics indicate that measurement
of NMHC by available commercial FID instruments yeilds unreliable individual
4
measurements. Inaccuracy in individual hourly values may be as high as ;+
several hundred percent in the range of concentrations below 0.5 ppmC. At
higher levels, the error is considerably less, but may still be +_ 25% at up to
2.0 ppmC. Because of the large potential for error that may be present in the
data, use of individual NMHC measurements at a single site for one day is not
recommended. Rather, if reasonable agreement is found among ratios at more
than one site on the day in which the ozone design value is observed (e.g., all
ratios within 50 percent of the mean ratio for all sites), the average of the
individual ratios should be used as the estimate of the ratio for use with
EKMA.
Figure 3.3 contrasts histograms of 6-9 a.m. NMHC/NO ratios calculated at
X
RAPS Sites 101, 105, 106 and 107 with a histogram of the average ratio for
these four sites. Data used in these calculations are from the period June
through October, 1976. The number of observations varies from site to site
because of incomplete data. For the composite of the four sites, the average
ratio for a day was only computed when data for at least two of the four sites
3-9
-------
i.
11
Site 101
18 22
NMHC/NO ratios
X
NMHC/NO ratios
Site 106
NMHC/NO ratios
X
12 16 20
NMHC/NOx ratios
Average NMHC/NOX ratios
FIGURE 3 3 Histograms of 6-9 a.m. NMHC/NO ratios at selected RAPS sites
3-10
-------
were available. As this Figure illustrates, the averaging process tends to
filter out the more extreme, and suspect, ratio values. Hence, this procedure
should provide a more representative value for the prevailing ratio. For the
situation when the estimated ratios among sites on a single day differ sub-
stantially, the procedure described in Section 3.2.4 is recommended.
3.2.3 Diurnal and Spatial Variation in NMHC and NO Concentrations
X
Average diurnal patterns in NMHC and NO concentrations at the five RAPS
.A
sites which lie approximately on a ray emanating to the north from Site 101 to
Site 122 are shown in Figure 3.4 and Figure 3.5. Both NMHC and NO concen-
X
trations exhibit pronounced peaks during the early morning hours, then decrease
rapidly to a minimum in the early afternoon before again rising during late
afternoon and evening. Spatial differences among the sites are also readily
apparent for both pollutants. As one proceeds outward from center city, both
the absolute concentration levels and the amplitude of the diurnal curves drop
off rapidly. This is particularly evident for NO at distant sites, where only
X
a slight diurnal pattern is perceptible and the concentrations are, on the
average, only slightly greater than the minimum detectable level of the moni-
toring instrument.
As noted in the previous section, the reliability of individual NMHC
measurements taken with an FID instrument increases as the concentration level
increases. Thus, the temporal and spatial properties of NMHC and NO con-
X
centrations suggest that the most reliable estimates of NMHC/NO ratios are
J\
those obtained at center city sites during the 6-9 a.m. period.
3-11
-------
0.7703 + 00
0000
0500 0700 1200
HOUR, LSI. *
2300
FIGURE 3.4 Average diurnal variation in NMHC (St. Louis RAPS, June-Oct. 1976
The hour associated with RAPS pollutant data is the beginning hour in
local standard time(LST),e.g. the hours denoted by 5-7 LST correspond
to the 6-9 a.m. LDT period, inclusive.
3-12
-------
X
O
0.9600-01
0.8708-01
\-x /
0.7816-01 ) \_/
0.6924 - 01
0.6032 - 01
0.5140 - 01
0.4248 - 01
0.3356 - 01
0.2464 - 01
0.1572-01
0.6800 - 02
TTT
/ \
\
\
A
o
0000
0500 0700
1200
*
2300
HOUR, LSI
FIGURE 3.5 Average diurnal variation in NOx (St. Louis RAPS, June-Oct. 1976)
The hour associated with RAPS pollutant data is the beginning hour in
local standard time(LST),e.q. the hours denoted by 5-7 LST correspond
to the 6-9 a.m. LOT period, inclusive.
3-13
-------
3,2.4 NMHC/NO Ratios on High Ozone Days
A
Application of EKMA requires an estimate of the prevailing 5-9 a.m.
MMHC/NCX ratio in the urban core on the same day as l.hf.- ozone concentration
'.,sed for control strategy development. Because of the concern about the
accuracy of individual NMHC concentrations measured with FID instruments, an
alternative procedure for estimating the ratio is suggr-sted. Unless readings
from more than one NMHC monitoring instrument are available on the same day as
the ozone design value, and these readings are comparable, it is recommended
that the median of the 6-9 a.m. NMHC/NO ratios observed on the five highest
X
ozone days (with accompanying NMHC and NO data) be jsf;d as an estimate of the
/\
prevailing NMHC/NO ratio for the city. Following the recommendations from
A
Section 3.2.2, when data from more than one monitoring site are available, the
6-9 a.m. ratios from each of these sites should be averaged on each day. The
estimate of the ratio for use with EKMA is the median of these five average
ratios.
Table 3.3 presents NMHC/NO ratios at seven RAPS sites (101 to 107) for
A
three time periods: (1) the entire oxidant season (June-October), (2) the day
with the second highest ozone concentration and (3) five high ozone days.* The
comparison of interest is between the NMHC/NO ratio on the day with the second
A
highest ozone concentration and the median ratio of the five high ozone days
(hereafter referred to as the median ratio). Good agreement is found between
The high ozone days were selected following application of quality control
r /-
procedures described by Hunt et__a]., to the preliminary RAPS hourly data.
Additional checks were also made for spatial continuity across the monitor-
ing network.
3-14
-------
oo
^
O
oo
t/1
a;
-o
cu
o
CD
O)
oo
i-
o
-M
ro
o;
X
4-
o
o
CO
S-
(O
c.
E
o
CJ
oo
CO
CQ
f S.
r^»
V -
S
^
LO
*. *
S s,
*d-
CO
0
f S.
CVJ
c
OO
i '
ro
CD
j
ro
0
00
ro
ro
O
to
ro
Ol
00
O
n-
u
o
i
o
z:
cr>
LC
Ld *
2 u"
00 ra
ro
~O T3
E ro
O O>
X
-o
C CO
CM O
1.
o
4-
cn
l
X.
^5
Z.
O
I
X
0
o
~1~
^-^
en
f
§U3
^ Pj
en
u>
= H,
> Dj
"Z.
f- t 1 CM >q «3< CO cn
r (
1 CxJ ^ LO 'd'
1 ... 1 .
O"t to to to i
r-~
en LO ^d" to oo i o
CO LO LO CO f- CO CTi
r CM co o o en co
o
o o o o o o o
^^ ^ cc to L^O ^J ^^
to LO co to r^ to to
*....
i CM CO «rf- LO IO l~^
o o o o o o o
OJ CU CL) OJ CD QJ (1J
+-> 4-> +-> 4-> -)-> 4-> 4->
OO OO OO OO OO CO OO
CM CO
-------
these two measures of the NMHC/NO. ratio at the five sites where the comparison
A
can be made. The largest difference is found at Site 102 with values of 9.1
for the day with the second maximum ozone concentration versjs 7.1 for the
median ratio of the five high days.
Similar comparisons are presented in Table 3.4 for the sites listed in
Table 3.1. In order to conduct these comparisons, given the sparseness of
precursor data at some of these sites, the NMHC/NO ratio on the highest ozone
X
day with accompanying NMHC and NO data is used. Again, good agreement is
X
found at the majority of these sites between the ratio observed on the high
ozone day(s) and the median ratio, i.e., columns (3) and (4) of Table 3.4. A
review of NMHC data for these high ozone days revealed that NMHC levels tended
to be higher on these days. Table 3.5 confirms this characteristic of pre-
cursor concentrations at RAPS Sites 101 to 107. Thus, greater confidence can
be placed in the ratio estimated on these high ozone days.
3.2.5 Default NMHC/NO Ratios
A
In order to apply EKMA to a specific city, both the ozone design value and
the prevailing NMHC/NO ratio must be known. Frequently, however, ambient
A
measurements of NMHC and NO may not be readily available for the city under
A
review. To accomodate such a situation, default NMHC/NO ratios have been
X
estimated for use with EKMA. These ratios are based on data from 14 sites in
cities with varying emission patterns and meteorological conditions. Of 25
locations reviewed, these sites are considered most appropriate for estimating
default ratios, because they meet the following criteria:
1) The sampling period is at least two weeks in duration, in order
3-16
-------
TABLE 3.4 Comparison of NMHC/NO Ratios on High Ozone Days for Selected Cities*
A
(1)
Site
(2)
Median
NMHC/NOX
Ratio
(All Data)
(3)
Median NMHC/NOX
Ratio on Five
Highest Oo Days
With NMHC, NOV Data
A
(4) 1
(NMHC/NOX) Ratio
on day with i
Highest 0 j
and NMHC, NOV Data|
X t
Austin
Corpus Christi
El Paso (1)
El Paso (A.P.)
Houston (Mae Dr.)
Los Angeles
San Antonio
Aldine, TX
Azusa, CA
Canton, OH
Dayton, OH
Groton, CT
St. Louis***
10.1:1
18.5:1
9.3:1
10.0:1
7.5:1
9.8:1
11.2:1
11.9:1
8.5:1
7.3:1
12.9:1
9.5:1
7.6:1
10.0:1
21.3:1
9.8:1
14.0:1
7.5:1
10.2:1
7.8:1
10.0:1
8.0:1
9.5:1
11.8:1
8.8:1
8.5:1
9.1:1
21.3:1, 47.3:1 **
9.8:1
14.0:1
12.5:1
8.3:1, 10.9:1 **
6.1:1
10.0:1
7.3:1, 7.9:1 **
5.4:1
8.0:1
7.2:1, n.Srl **
7.5:1
* All NMHC, NO and ratio data refer to 6 a.m. - 9 a.m. LOT values.
x
**
Two days with identical 03 concentrations.
*** Composite of Sites 101 - 107.
3-17
-------
I
*
>>
ro
O
OO
O
r^ .c
- cn
r-
LO
ON
|
vO
X
o
z:
!
«3" UD LO ^£j i i UD
ro r^ o LO
i- C
4J
QJ
O C
C fO
o a)
o ^
M- t
O (8
C
E O
O CO
CO (O
r- a>
s- co
LO
CO
03
ro
c o*\
» O !
LO CO ^O
ro X
CU O
co z:
-X
CO
ro
Q
OO
O ON
'-- 1
^-^ cn c_>
'£c 2:
LO
(O
^-. Q T
oo oo vo
-0 0
31
x s:
fO ir^^
2:
ro O
^ C 1
oo o vo
-co O
ro _L.
oj s:
co z:
c:
o
(->
^-^ d) ro
- -i- O
CO _J
i
^O *f OO «* OO CO
^o o^ r*^ r*^ co f^
o o o o o o
*»
LO
r- r- r- ^ ,_
ro r^- ^3- LO
c^ o^v o^ Ci cyi ^f~
2T * !?"
, p__ P
r>» cri co "^o LO CM
^.o LO oo ^X5 r*"-- ^o
r C\l OO «?3~ LO ^O
0 O O 0 O 0
"^" '
r"^ |
o
a
1
oo
r-
1
1
*-O
CSJ
. '
OJ
LO
UD
4
l>^
o
.
u
&,
S
w
g
-H
-P
s
§ .
Q
2:
cn
c
ro
Q-
E
O
o
o
ro
a>
(O
QJ
ro
CO
ro
"O
QJ
O
o
CO
QJ
'cn
i
.rr
Oi
T-T8
-------
to diminish the likelihood of a biased sample due to a single set of prevailing
synoptic conditions.
2) At least 10 days of valid measurements for each parameter are
available, so that extreme values can be eliminated when estimating 90th and
10th percentile NMHC/NO ratios.
A
3) The site is located in an urban or suburban area, and if in a
suburban area, the site is not dominated by point sources of industrial organics.
4) The sampling has been conducted since 1974, in order to minimize
the likelihood that ratios have been appreciably altered due to emission
control programs.
Summary statistics for the 14 sites meeting the above criteria were
presented previously in Table 3.1.
In the absence of ambient NMHC and NO data, the default ratio recommended
)\
for use with EKMA is a value of 9.5:1. This value is the median of the typical
ratios observed in these cities on high ozone days, i.e., it is the median of
the ratios listed in column (7) of Table 3.1, rounded to the nearest 0.5. Some
measure of the range of possible values for the default ratio also seems desir-
able. To obtain a conservative measure, the 10th percentile and 90th percentile
ratios observed at these sites are used. Estimates are based on the median of
the ratios listed in columns (8) and (9) of Table 3.1. The medians of the 90th
and 10th percentile ratios are, respectively, 15.9 and 5.9:1. Rounding these
values gives default ratios of 16:1 and 6:1. Thus, in the absence of ambient
measurements of the ratio, Figure 3.6 may be used to relate the sensitivity of
maximum afternoon ozone levels to changes in upwind precursor levels.
3-19
-------
(O C-J 00
r- T- «=>
O
e
ex
ex
X
s
2
iudd
-------
3.3 Summary and Recommendations
A review of ambient data on ozone, NMHC and NOY concentrations from a
A
number of cities has been conducted. Particular emphasis has been placed on
data collected as part of the RAPS program in St. Louis during 1976. Signifi-
cant findings from the study are:
1. NMHC and NO concentrations have very pronounced diurnal
A
patterns at typical urban monitoring sites. Peak
or. near peak concentrations occur during the period
6-9 a.m. LOT.
2. The more distant the site is from a large urban area,
the less pronounced is the diurnal variation and the
lower are the concentrations.
3. Because of problems associated with measuring NMHC levels
using an FID instrument, there is considerable uncertainty
in individual NMHC readings. Studies have shown, however,
that relative errors are less at higher concentration
values.
4. There appears to be good agreement between the 6-9 a.m.
NMHC/NQ ratio observed on the second highest ozone day
A
and ratios observed on other high ozone days.
5. A typical ratio from a cross section of urbanized areas
appears to be about 9.5:1.
These findings have several implications for how the ratio used in EKMA should
be estimated. As was previously noted, the use of the 6-9 a.m. NMHC/NO is
A
considered to be consistent with the results from smog chamber experiments on
which EKMA is based. Since both NMHC and NO levels are higher during this
A
3-21
-------
period, greater confidence can be placed in the estimate of this ratio.
However, because of the uncertainties in Individual NMHC readings, use in EKMA
of NMHC/NO ratios calculated at a single site for a single day is not recom-
A
mended. Considering FID instrument reliability and model sensitivity, the
suggested procedures for calculating NMHC/NO ratios for use in EKMA are:
A
1. If precursor measurements from more than one urban site are avail-
able for the same day as the ozone design value, and the individual
ratios at each site do not differ by more than 50 percent from the
average ratio, then the EKMA ratio (ER) j[s_ the average of the indi-
vidual 6-9 a.m. NMHC/NO ratios. A review of instrument operating
A
characteristics and data quality should be conducted prior to cal-
culating the average ratio and performing the comparisons.
ER = V (R,) - R
1=1
where
R. = 6-9 a.m. (NMHC/NOJ ratio at Site i
I A
R = the average ratio
n = the number of sites
Example
Given: The NMHC/NOv ratios for St. Louis in Table 3.3,
A
column (7).
Find: The NMHC/NOv ratio for use in EKMA
A
Solution:
Ratio data are available for five of the seven central
3-22
-------
sites. First calculate the average ratio,
R = 9.1 + 6.2 + 6.4 + 6.5 + 11.4
5
R" = 7.9
Note that all the ratios are within +_ 50% of R", i.e.,
all the ratios are between 4.0 and 11.9. Then, the EKMA
ratio is given by,
ER = R = 7.9
2. If only one site measuring both NMHC and NO is available, or
}\
if the values from different sites are not comparable (i.e., some
ratios differ by more than 50% from the mean ratio), the following
procedure is recommended. The ratio for use with EKMA is calculated
as the median of the ratios observed on the five highest ozone days
(with accompanying NMHC and NO data).
A
W
K) '
ER = median / Rn. \ ,for a single site
or
ER = median ( R^ } ,for multiple sites
where
j = a high day, and j = 1,2,...,5
Example
Given: Assume that only RAPS Site 101 is available for St.
Louis. The NMHC/NO ratios on the five days with
A
high ozone and accompanying precursor data are 8.8,
8.6, 15.5, 9.7 and 14.3.
3-23
-------
Find: The NMHC/NOv ratio for use in EKMA.
A
Solution:
Since only one site is available, the EKMA ratio is given by,
ER = median <
ER = 9.7
8.8, 8.6, 15.5, 9.7, 14.3
3. If ambient NMHC and NO data are not available, the use
/\
of the default ratio is recommended. The default ratio for
use in EKMA is 9.5:1. As a measure of the uncertainty in
estimating precursor controls introduced by the use of the
default ratio, estimates using ratios of 16:1 and 6:1 can be
made as well.
3-24
-------
3.4 References
1. Uses, Limitations and Technical Basis of Procedures for Quantifying
Relationships Between_ Photochemical Oxidants and f.recursors,
EPA-450/2-77-021a, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, November 1977.
2. D. J. Londergan and L. G. Polgar, Measurement Program for
Ambient 0,, NO and NMHC at Portland, MaineSummer 1974,
Unpublished Report, U.S. Environmental Protection Agency,
Research Triangle Park. North Carolina, November 1975.
3. Survey of Users of the EPA Reference Method for Measurement of
Non-Methane Hydrocarbons in Ambient Air, EPA-650/4-75-008,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, December 1974.
4. Hydroca rbo_n_ Measunement. Discrepancies Among Various Anatysers Using
Flame-Ionization Detectors. EPA-600/4-75-010, U.S. Environmental
Protection Agency, Research Triangle Park. North Carolina, September 1975.
5« W. F. Hunt, el a'i., "The Application of Quality Control Procedures
to the Ambient Air Pollution Problem in the U.S.A.," presented at
the 20th Annual European Organization for Quality Control Conference
in Copenhagen, Denmark. June 1976.
6. T. C. Curran, et_ a]_., "Quality Control for Hourly Air Pollution
Data," presented at the 31st Annual Technical Conference of the
American Society for Quality Control, Philadelphia, Pennsylvania, May 1977.
7. Interim Guidance on Air Quality Models, OAQPS 1.2-080, U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina, October 1977.
8. Guidelines for the Interpretation of Air Quality Standards, OAQPS 1.2-008,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, August 1974.
3-25
-------
4.0 Variatjon_0f Ozone and Precursor Levels In and Near Urban Areas.
4.1 Introduction
4.1,1 Purpose
The purpose of Section 4 is to provide estimates of ambient ozone and
precursor levels typically found upwind and downwind of urbanized areas.
Particular emphasis is placed on concentrations observed upwind of urban areas
on those days experiencing high oxidant levels.
4.1.2 Background
In Reference 1 and Section 5 of this report, it is concluded that un-
scavenged ozone remaining aloft at night is the facet of transport from upwind
cities likely to exert the greatest impact on maximum ozone concentrations in
urban areas, with a lesser impact also arising from transported organics. The
principal difficulty in measuring transported ozone and in interpreting the
results is having to rely primarily on surface measurements to estimate ozone
levels aloft. In the absence of locally appropriate data on the breakup of
nocturnal inversions, surface measurements of ozone between 1100 and 1300 LOT
are suggested for use in estimating ozone which is advected aloft. A dis-
cussion of the problems associated with using surface measurements to estimate
ozone levels aloft is contained in Section 3.2 of Reference 1.
Ozone data from monitoring networks around cities reporting to SAROAD were
examined in an effort to determine typical concentrations of transported ozone
and precursors on days experiencing high concentrations of ozone. Meteoro-
logical data from the nearest National Weather Service (NWS) reporting station
were used to identify the predominant wind flow on high ozone days, and thus,
the appropriate upwind monitoring site. Problems were frequently encountered
4-1
-------
when trying to make such estimates, resulting primarily from the absence of
suitable upwind monitoring sites.
Ambient precursor data upwind and downwind of urban areas are sparse. A
survey of data on ambient precursor levels in rural areas was presented in
Section 2. Rural non-mathane hydrocarbon levels as high as 200 ppbC were
found. This section focuses on precursor concentrations in the vicinity of
urban areas. The data base found most suitable for this analysis of precursor
gradients is the St. Louis RAPS for 1975. The RAPS ground sampling network is
shown in Figure 3.2. All sites provided continuous measurements of ambient
ozone, non-methane hydrocarbon and oxides of nitrogen concentrations. Measures
for other selected pollutants and meteorological variables were also provided.
4.2 Ozone Concentrations Measured Upwind of Urban Areas
For a large number of cities, suitable upwind sites were not available in
SAROAD. Major problems included: (1) no monitoring site located upwind and/or
outside the influence of an urban plume, (2) the wind field was poorly defined
(i.e., an upwind site could not be determined) and (3) sites were subject to
localized influences, e.g., nearby NO sources. These problems are understand-
A
able, because most networks are not designed to measure levels of ozone entering
a city but rather maximum ozone concentrations which typically occur along the
predominant downwind direction from the city.
Upwind ozone measurements on high ozone days for Philadelphia, Washington,
Pittsburgh and Cleveland are presented in Table 4.1. As the values in the
table indicate, average ozone concentrations measured upwind of these cities
during the 1100-1300 LOT period on high ozone days are typically above the
NAAQS of 0.08 ppm. Similar findings are reported by Cleveland, et al. in an
4-2
-------
TABLE 4.1 Upwind Ozone Concentrations for Selected
Cities on High Ozone Days During 1975
(ppm)
City
Philadelphia, PA
OWN = Bristol, PA
UP = Nbrristown, PA
Washington, D.C.
OWN = Suitland, MD
UP = Fairfax, VA
Pittsburgh, PA
OWN = Charleroi, PA
UP = Baden
Cleveland, OH
DWN = Painesville, OH
UP = Medina, OH
Downwind Site
Wind*
Max 0
PPM
.33
.29
.22
.18
.17
.26
.22
.18
.18
.23
.20
.19
,19
,15
.23
.18
.16
.16
Direction
DEG(lC's)
23-23-22
22-23-23
22-26-22
30-24-22
23-23-24
28-22-16
29-14-13
35-28-22
23-23-22
06-06-34
31-27-27
08-32-06
30-28-32
10-02-06
21-23-36
22-18-24
27-32-32
27-29-30
AVG.SPD
MPH
7.5
6.6
7.9
7.9
7.1
4.2
2.7
4.5
10.1
3.0
10.1
3.2
5.5
4.6
6.0
6.8
6.8
8.9
Upwind Site
AVG. 03
11-1300 LDT
PPM
.09
ND
.09
.10
.10
.13
.13
.09
.12
ND
.09
.07
.05
,06
.09
.09
.12
.12
ND No Data
* Wind direction is expressed in tens of degrees, at three hourly periods;
1100, 1400 ancq 1790 IDT.
-------
extensive study of ground level ozone data from the Northeast.
3
When calculating upwind ozone concentration'.;, some discretion can be used
in selecting an appropriate upwind monitoring site. Strictly speaking, the
site does not have to be directly upwind of the city. As long as the site is
not directly influenced by the urban plume from the cUy, it may be used as an
upwind site. For example, in Figure 4,1, which depicts ozone monitoring sites
in and around Philadelphia, the site located in Morristown can be taken as an
d
upwind site on days with northeasterly flow,'
Trie spatial distribution of the RAPS monitoring sites enables upwind/
downwind ozone concentrations to be determined under most wind flow patterns
(see Figure 3.2), Table 4.2 lists upwind ozone concentrations averaged between
1100-1300 LOT for nine high ozone days in St. Louis during 1976. These upwind
concentrations range from a low of 0.06 ppm to high of 0.12 ppm, with a median
value of 0.07 ppm.
4.3 Urban Precursor Variations
The data used to define the relative magnitude of precursor concentrations
across St. Louis are shown in Table 4.3. The table shows upwind, core and
downwind concentrations of NMHC and NO for nine representative high ozone days
/\
in 1976.* These days were selected on the basis of the ozone concentration
level as well as on a relatively consistent wind flow pattern across the
RAPS monitoring network. In all cases studied, the peak precursor concentrations
* Core sites are those sites within 4 km of Site 101, i.e, Sites 101 through
107.
4-4
-------
-------
TABU £1,2 nznriE aTicenwnms TOJREP iwin OF sr, LOUIS
fT! NINE HIGH OZOfiE DAYS If! 1976
(ppm)
DATE
*
10/1
6/8
7/13
6/7
8/25
9/17
10/2*
8/26
8/27
MAX SITE
^5
MAX (HR, - oITE)
i ^.H \_Li'/rJLiJ:^/
,22 (16-#U5)
, ZZ vJ.-5~~fr.LJ.Hy
,20 (15-#122)
,19 (13-0115)
,19 (12-*112>
, J_y \ J_u~"rrJ_ J.-3 /
,18 (13-^115)
,18 (11-#114)
UPWIND SITE
ru
1100-1300 AVG
,07
,12
,09
,10
,10
,OE
,06
,07
,07
i
i i
SITE
(# 124)
(// 124)
(# 125)
(# 125)
(# 124)
(// 122)
(# 124)
(# 125)
(# 124)
* Staonation period
4-6
-------
ro
ft
I-M
g
fa
§
P
1
=B=
(U
H
CO
*
e
H
3 1
-H n
-P Q
rd
1
u
u
C =M-
m Q}
H
P W
5
rd *
rr\ n\
U i \U
1 M
VD g
£>
"S
1
^*
Q)
-P
H
CO
*
"S
'>
o
CN
00
rH
rH
LO
rH
O
O
1 1
O
CN
in
rH
H
in
o
o
0 0
00 CN
H CN
rH rH
OO O
0 H
0 0
<3* in
O 0
rH rH
CTi CO
O CTi
O CN
r- o
CN CN
in co"
H rH
d r~l
pH ^3"*
o o
0 0
V
o
"*
CN
CN
rH
rH
0
O
rH
O
rH
CO
CN
0
O
^jT
CN
rH
rH
0
O
r-- r-
CN CN
in m
rH H
rH rH
O "3"
CN O
O O
in CN
0 O
rH rH
VD CN
rH rH
CN O
0
^
CN
rH
rH
O
2 O
V
o o in in o
CN T CN CN "*
CO <* rH rH CN
rH CN CN CN CN
rH rH rH i 1 rH
CN rH VD 00 CO
o o in o co
o o o o o
rH rH rH t^ ^
o o o o o
rH rH i 1 rH rH
r- in in ro co
m o r-- rH CN
0 0 rH 0 rH
in o
CN rr
rH CN
CN CN
rH rH
^f CN
o o
0 0* § § §
O
=*
CN
CN
rH
CN
0
0
rH
O
rH
rH
rH
0
0
^r
CN
rH
rH
0
O
O
^
CN
CN
rH
t^
ro
0
o
rH
0
CO
O
O
CN
00
rH
rH
1 I
O
o
\f
0 O
^ CN
CN O
CN CN
rH rH
ro i
h)
O
ro
H1
I-D
4-7
CN r^ co
rj< rH r 1 rH
-P -P -P -P
g, g, & g,
<<;<;<
&T
VD
-P
w g
Q |
^ ""V*
y Q
i ^
X )«
*
-------
were observed over center city. The center city concentration levels were
typically about three to thirty times greater than upwind-downwind values.
Figures 4.2 and 4.3 present average 6-9 a.m. NMHC and NO concentrations with
/\
distance from center city, along a north-south axis, during the oxidant season
(June-October). The use of seasonal average concentrations should provide a
robust estimate of the precursor gradients across the St. Louis urban area.
Based on a review of the RAPS data, reasonable estimates for precursor
transport into an urban area appear to be a concentration range for NMHC of
0.05-0.30 ppmC and for NO of < 0.01-0.05 ppm. The lower limit of the con-
/\
centration range for each precursor is assumed to represent precursor levels
typical of long range transport into an isolated city. The upper limit of the
concentration range for each precursor is assumed to represent precursor
levels typical of transport into a city that is about 40 kilometers downwind of
a medium/large city.
4.4 Summary
Ozone gradient data (upwind versus downwind) are difficult to obtain
because of limited data from appropriate upwind sites. Such data as have been
examined indicate a sizable increase in ozone concentrations downwind from
major cities during days with meteorological conditions conducive to ozone
formation. Surface measurements upwind from major cities not clearly dominated
by the urban plume from another city farther upwind suggest long range trans-
port values for ozone ranging from 0.05-0.13 ppm. Such values are reasonably
consistent with concentrations observed aloft overnight in balloon experi-
5
ments.
The sparse data available on precursor levels entering urban areas from
4-8
-------
w
«£1
(£51
A-l
!
§
u
(0
o
n
CM
4-9
I
-------
CN
(N
4-10
-------
rural regions indicate that these levels are normally low. Information in
Section 2 and this Section suggest that non-methane hydrocarbon levels from
about 0.05 to 0.30 ppmC, and nitrogen oxide levels from < 0.01 to 0.05 ppm are
typical ranges for precursor transport into an urban area. The upper limit of
these ranges may be suggestive of some impact of upwind urban areas or re-
circulation from the same urban area. On high oxidant days, the morning
(6-9 a.m.) NMHC and NO values in urban centers are about ten or more times
s\
greater than in adjacent rural areas. Simulations with the kinetics model in
EKMA (Section 5) suggest that these estimates of precursor background levels
will generally not contribute significantly to the synthesis of high ozone
levels in the vicinity of an urban area.
4-11
-------
4.5 References
^ Uggs, Limitations and Technical Basis of Procedure.; For
Quantifying Relationships Between P ho toe hern i :al_ 5x"idanl:3__and_
Precursors, EPA-450/2-77-021a} U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, November 1977,
2. Investigation of Rural Levels as Related to Urban Hydrocarbon
Control Strategies. EPA-45Q/3-75-036, U.S. EPA/ Research
Triangle Park, North Carolina, March 1975.
3, VI, S, Cleveland, et_al_., The Analysis_..p_f_Sround Level Ozone
D_ata From New Jersey, New York, Connecticut, and Massachusetts:
transport From the New York City' Metrppol 1 tan Area,' Bell
Laboratories, Murray Hill, New Jersey, 1976,
4. 'fi. S. Cleveland and B. Kleiner, "The Transport of Photochemical
Air Pollution From the Camden-PhHadelphia Urban Complex,"
Environmental Science & Technology, Vol. 9, No. 9, September 1975.
5. E, Robinson, F. Vukovich and D. H. Pack, "International Conference
on Oxidants 1976 Analysis of Evidence and Viewpoint. Part V. The
Issue of Oxidant Transport." U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, November 1977.
4-12
-------
5.0 The Impact of Inter-Urban Pollutant Transport on Oxidant Concentrations
5.] Introduction
The production of ozone from precursor emissions in an urban area can be
significantly affected by pollutants generated in areas upwind and transported
into the urban area. To assess the impact of these transported pollutants,
several simulations were conducted using the kinetics model underlying the
Empirical Kinetic Modeling Approach. This Section discusses these simulations
and their significance in terms of ozone-precursor relationships.
The simulations conducted in this study attempted to provide answers to
two major questions. The first question is "What aspect of the transport of
pollutants has the most significant effect on the peak ozone concentration
found in the vicinity of an urban area?" Several pollutants that affect
oxidant production, including ozone, non-methane hydrocarbons, oxides of
nitrogen and aldehydes, may be transported into an urban area. A number of
kinetics model simulations were performed to determine which transported pol-
lutants have the greatest impact on peak ozone concentrations. These simu-
lations also included evaluation of the relative importance of ozone and
precursor transport in a layer near the ground versus in a layer aloft.
The second major question is "What is the impact of transported ozone
under varying conditions?" A number of model simulations were conducted to
determine how the impact of transported ozone is affected by variations in
dilution rate, temporal emission distribution and sunlight intensity. Also
investigated was how the impact of transported ozone varies with the trans-
ported ozone concentration itself, the level of local emissions and the pre-
vailing NMHC/NO ratio. Although most of these simulations were intended to
5-1
-------
assess the impact of pollutant transport for a current range of situations,
some additional simulations were also conducted for situations anticipated when
the National Ambient Air Quality Standard for oxidants is approached.
5.2 The Simulation Model
All simulations discussed here employed the kinetics model used in gener-
ating the standard and city-specific isopleths presented in Uses, Limitations
and Technical Basis of Procedures for Quantifying Relationships Between
Photochemical Oxidants and Precursors. This kinetics model is discussed
2 3
further by Dodge and by Whitten and Hogo.
An important assumption inherent in this kinetics model is that pollutant
concentrations are uniformly mixed throughout a column of air extending through
the entire mixed layer. As mixing increases to greater heights, the model
assumes that entrained air is instantaneously mixed throughout the enlarged
mixed layer. Similarly, the model also assumes instantaneous upward mixing of
emissions. These assumptions could significantly affect the timing, and
possibly the quantitative impact, of the interaction of transported pollutants
and local emissions. Thus, this study is intended only to compare relative
impacts of various pollutant transport conditions rather than to assess the
absolute impact of any single set of pollutant transport conditions.
5.3 Conditions Simulated
Table 5.1 shows the basic set of conditions simulated in this study.
Reference 1 provides a more extensive discussion of the nature of these vari-
ables. As shown in Table 5.1, these conditions differ from those used in
deriving the standard set of isopleths presented in Reference 1. While the
standard isopleths are based on probable worst-case meteorological conditions,
5-2
-------
TABLE 5.1
Base Set of Conditions Assumed in Kinetics Model Simulations
Data Required
Conditions Assumed
in Standard
Isopleths
Conditions Assumed
in this study
Parameters
Affecting light
Intensity
34°N, 118°W
June 21
40°N, 75°W,
Sept. 15
Dilution rate
3%/hour
13%/hour and 29%/hour
Relative
Emission rates
Initial Relation-
ship of NOp and
NOX
All emissions
occur by 8 a.m. LOT
Time
by 8 a.m.
8 - 9 a.m.
9 -10 a.m.
10 -11 a.m.
11 a.m. -12 noon
after 12 noon
Relative
Emission Rates
100%
28.5%
23.3%
15.0%
6.3%
no emissions
**
NC0 = 25% x NO
f X
NO = 75% x NO
N02 = (15% x NOX) +(03)t
NO = NOX - N02
Emission rates are normalized relative to emissions prior to P a.m. LOT.
**
(0-)
3yt represents the ozone concentration transported into the city
in the surface layersee text.
5-3
-------
the simulations in this study are intended to represent a typical range of
conditions found throughout much of the United States during days experiencing
high ozone concentrations.
The first item in Table 5.1 concerns parameters that affect the diurnal
distribution cf light intensity. They are used to determine photolytic rate
constants in the kinetics model. At the time this study was conducted, it was
only possible to usa photolytic rate constant data for 40°N, 75°W on September 15
and for 34°N, 118°W on June 21. Since the highest ozona concentrations often
occur in the late summer or early fall, the first of these two sets of data was
used in these simulations. Nevertheless, these two sets of diurnal light
intensity distributions are not radically different. Specifically, the June
insolation has approximately a 17 percent higher peak insolation and a 10
percent higher integrated average light intensity between 8 a.m. and 5 p.m.
local daylight time.
The second item in Table 5.1, dilution rate, is used to simulate the
entrainment of relatively clean air from aloft as the mixing height increases,
a process which results in dilution of concentrations within the mixed layer.
In these simulations, the air above the mixed layer often contains significant
concentrations of pollutants. However, the kinetics model first calculates
dilution rate as if the entrained air were clean and then accounts separately
for entrainment of pollutants.
Dilution rate, expressed here as percent per hour, is defined as the
percentage increase in the mixing height that occurs over a period of an hour.
Table 5.1 shows the two dilution rates used in this study. The mixing height
was assumed to rise for six hours and to remain constant thereafter. Hence,
5-4
-------
the total increases in mixing height for the 13%/hr and the 29%/hr dilution
rates were approximately a twofold increase and approximately a fivefold in-
crease, respectively.
The third item shown on Table 5.1 is relative emission rates. These are
used to simulate the rates of emissions into the air parcel for each hour of
the simulation. These rates are typically based on a hypothetical air parcel
trajectory and an assumed spatial and temporal distribution of emissions
encountered along this trajectory. The rates are then normalized relative to
the emissions present at the beginning of the simulation (8 a.m. LOT). The
mixing height was assumed in this study to begin increasing at 9 a.m., and so
the calculation of rates of concentration addition required no adjustment of
relative emission rates prior to this time.
In calculating the basic set of relative emission rates used in this
study, the air parcel was assumed to remain at the center of the city until 9
a.m. At that time, the air parcel was assumed to begin moving downwind,
reaching the edge of the city three hours later. The spatial distribution of
emissions was assumed to be Gaussian in the direction of the wind, with the
2
edge of the city having e (about 0.14) of the emission density at the center
of the city. The temporal distribution of emissions was obtained from a U.S.
Department of Transportation study of traffic distribution in several cities.
The resulting set of relative emission rates as used in this study is shown in
Table 5.1.
The derivation of relative emission rates in this study involved some
simplifying assumptions, particularly those concerning the emission densities
of areas traversed by the air parcel. Unfortunately, it is often quite difficult
5-5
-------
to obtain more accurate Information on ?,ir parcel trajectories and on spatial
distributions of emissions, When such information is available, it is often
suitable to use a more sophisticated model. Thus, these assumptions should
reflect the accuracy appropriate for use in a kinetics model. In any case,
this study also supplies information regarding the effect of several alter-
native assumptions in calculating relative emission rates.
The fourth variable shown in Table 5.1 is initial NO and N02 concentra-
tions as a function of the initial NO concentration. Most NO, emissions are
.A t\
largely NO. However, the fraction of N0? is increased by photochemically
induced organic reactions and by reaction with ozone. As indicated in Table 5.1,
the standard isopleths used a fixed initial N09/N0 ratio, based on relatively
£- A
low transported pollutant concentrations. On the other hand, the initial
mixes of NO and NO? in these simulations were based on a ratio of N0n/N0
L~ L~ f\
attributable to organic reactions and an adjustment according to the trans-
ported ozone concentration. In some simulations with high transported ozone
concentrations, the initial NO was entirely NO^, and the simulations began
with an "excess" ozone concentration.
5.4 The Impact of Transport of Various Pollutants on Ozone Concentrations
Photochemical production of ozone in an urban area can be affected by
transport of various pollutants from upwind areas. In particular, it is
important to consider the transport of ozone, precursors (i.e., hydrocarbons
and NO ) and aldehydes.
A
A substantial number of studies have indicated that ozone can be trans-
ported over long distances between cities. It is necessary to consider this
ozone in terms of two atmospheric layers: (1) the surface layer (i.e., the
5-6
-------
morning mixed layer), and (2) the layer aloft, particularly the layer that is
above the mixed layer in the morning but within the mixed layer in the after-
noon. Ozone transported in both atmospheric layers can react with NO according
to the reaction NO + 03 -> NO- + 0-. However, ozone transported in these two
layers is added to the air parcel by different means, and it affects the
concentrations at different times. The ozone transported in the surface layer
is present in the air parcel prior to the simulation period and is simulated
by adjusting the initial concentrations. The kinetics model assumes that both
the transported ozone concentration and other pollutant concentrations are
horizontally homogeneous in the vicinity of the primary air parcel. Thus,
horizontal interchange with other air parcels is assumed not to affect the
concentration in the primary air parcel. Ozone transported aloft, on the
other hand, affects concentrations in the primary air parcel as the mixing
height increases and the air parcel entrains this ozone from aloft. Thus,
ozone transported aloft affects the air parcel concentrations in the late
morning and in the afternoon. Unlike the ozone transported in the surface
layer, the rate of addition of ozone aloft changes if the rate of the increase
in the mixing height changes. A typical scenario is pictured conceptually in
Figure 5-1.
Like the transport of ozone, the transport of precursors may also be
considered in terms of two layers, i.e., transport in the surface layer and
transport aloft. The pollutant concentration aloft is likely to be different
from that in the surface layer, due both to differences in the meteorological
history of the two layers of air and to incomplete vertical mixing of ground
level emissions.
5-7
-------
MID-AFTERNOON
iiivic uuuu-
usuu \i
HIGH 03 ALOFT (UNSCAVENGEO)
MIXING
DEPTH "X
9
uu (i IM
**
*r
**
**
S
'" \
-'f/ ^\\
s' 1 (* 03MIXED \ \
I"*-' / | IN SURFACE | 1
J'' I I I LAYER FROM / /
r-~-i i V V ALOFT / /
NMHC.NO, , - j \ X / /
03-0ppm { NMHC/NOX j , j \ S
E UF
.___
V1AX. C
MAGINARY
COLUMN
OF AIR
'3l
1
(WITH CONTINUOUS y7 / / / / / / 7 / / / ' /
NS) />'///
AND NOX EMISSIONS)
FIGURE 5.1 Role of ozone aloft on maximum surface ozone levels: a conceptual view
5-8
-------
Unfortunately, it is difficult to make reasonable assumptions about the
typical relationship between precursor concentrations at the surface and
aloft. This study tested two alternative assumptions: (1) precursor con-
centrations aloft are equal to the concentrations in the surface layer, and
(2) insignificant concentrations exist aloft. The true situation can prob-
ably be expected to fall between these two assumptions, but it is difficult to
assess which assumption is more appropriate.
Another class of pollutants which may be transported in significant
quantities is aldehydes. Aldehydes are both products and reactants in the
photochemical production of ozone. As a result, in an air mass in which the
photochemical pollutants have had ample time to react (an "aged air mass"),
aldehyde concentrations may be a somewhat higher proportion of organic species
than their proportion in emissions.
To compare the impact of the transport of these various pollutants, two
approaches were used. The first approach simulated approximately the maximum
pollutant concentrations that might be transported into an urban area. One
would seldom expect all of these concentrations to be transported in the same
air mass at the same time. However, this approach was not intended for esti-
mating the maximum total impact of pollutant transport. Rather, it was in-
tended to estimate and to compare the maximum impacts of the transport of each
pollutant separately, considering the worst case for each pollutant. The
second approach was based on the selection of more typical transported pol-
lutant concentrations. This approach was intended to suggest the typical
relative importance of the transport of various pollutants.
5-9
-------
The first part of Table 5.2 shows the transported pollutant concentra-
tions used in the comparison of worst cases. As indicated in the table, it
was assumed that the approximate upper limit to the ozone concentration trans-
ported in the surface layer is less than the approximate upper limit to ozone
concentrations transported aloft. This was based on the presumption that
ground level scavenging could significantly affect early morning ozone con-
centrations in the shallow surface layer but does not affect ozone concentra-
tions aloft. It was assumed that ozone concentrations transported aloft
seldom exceed 200 ppb and that ozone concentrations transported in the surface
layer will seldom exceed 120 ppb.
The transported precursor concentrations used in the comparison of worst
cases represent the concentrations found within 40 kilometers downwind of St.
Louis. These data are discussed in Section 4. Very few cities are within
40 kilometers of a city the size of St. Louis, and so very few cities should
experience transported precursor concentrations as high as those indicated on
Table 5.2.
As discussed previously, it is difficult to hypothesize the typical
relationship between precursor concentrations transported aloft and the con-
centrations transported in the surface layer. However, it may be possible to
make some generalizations about the relationship between the worst case pre-
cursor concentration transported aloft and the worst case precursor concen-
tration transported in the surface layer. The worst case precursor concen-
trations transported into the city in the two layers probably occur under two
different types of situations. The existence of high precursor concentrations
transported in the layer aloft is probably dependent upon the occurrence of
5-10
-------
TABLE 5.2
Conditions Assumed in Simulating the Transport of Various Pollutants
Comparison of Worst Cases
Pollutant
Ozone
Precursors
Aldehydes
Transported Concentration
Aloft: 200 ppb
Surface Layer: 120 ppb
.26 ppmC (90% butane,
10% propylene)
NMHC -
NO = .04 ppm (100% NO,)
X L,
Significance tested by doubling
initial concentrations
Pollutant
Ozone
Precursors
Aldehydes
Comparison of Typical Cases
Transported Concentration
120 ppb (in both layers)
NMHC = .10 ppmC (100% butane)
N0v =0.0 ppm
X
Negligible
k
Two alternative assumptions are used in simulating precursor transport:
1. Concentrations transported aloft equal concentrations transported
in the surface layer,
2. Concentrations transported aloft are negligible.
5-11
-------
substantial emissions on a previous day, when mixing extended above the current
surface layer. In contrast^ the highest precursor concentrations transported
in the surface layer probably occur when substantial emissions enter the layer
during the evening or at night, when the mixed layer is shallow. Therefore,
the highest precursor concentrations transported in the surface layer are
likely to on substantially greater than the nignest concentrations transported
aloft, due to the substantially lower mixing height as v;ell as the shorter
time during which dilution can occur, As a result, the actual worst case
impact of precursor transport is likely to be betfer represented by the as-
sumption of negligible concentrations aloft than by the assumption that the
concentration aloft equals the concentration in the surface layer.
Transported precursors differ significantly from urban emission in terms
of the composition of the pollutant mix. The kinetics model simulations of
smog chamber results suggest that the best representation of automotive
2
exhaust with a butane-propylene mix uses 75% butane by ppmC. However, a
transported hydrocarbon mix is lower in reactivity, due to prior reaction of
the more reactive species. Therefore, in these simulations, the 0.26 ppmC of
transported hydrocarbons was arbitrarily chosen to be 90% butane and 10%
propylene. The entire 0.04 ppm of NO was assumed to be N09.
X C-
Quantification of transported aldehyde concentrations is difficult,
especially due to an absence of reliable measurements. Therefore, the worst
case transport of aldehydes was simulated simply by doubling the 8 a.m. alde-
hyde concentrations. This may be considered as much an expression of the
uncertainty in aldehyde measurements as of the potential impact of aldehyde
transport. It seems unlikely that aldehyde transport would have such a large
effect on urban aldehyde concentrations. Nevertheless, given the dearth of
5-12
-------
reliable evidence on aldehyde concentrations, this should represent a reason-
able upper boundary of aldehyde transport for this study.
The typical case of transported pollutant concentrations is presented in
the second half of Table 5.2. These concentrations were chosen to reflect a
scenario in which an upwind city is located a considerable distance from the
city being modeled. As discussed by a number of authors, high ozone concen-
trations can be transported over long distances, but precursor concentrations
5
tend to decrease rapidly with distance. For this scenario, the transported
ozone concentration was taken as 120 ppb. For simplicity, the ozone concen-
tration transported in the surface layer was taken as equal to the concen-
tration transported aloft. The precursor concentrations were closer to the
background concentrations discussed in Section 2.
Simulations of precursor transport were conducted both for the assumption
that concentrations transported aloft equal those transported in the surface
layer and for the assumption that precursor concentrations transported aloft
are negligible. Little evidence is available to indicate which assumption
more closely approximates the typical case. In this case, the precursors will
often be emitted into the transported air mass on some previous day, providing
an opportunity for substantial mixing into the "layer aloft". Nevertheless,
uncertainties about the mixing process and about the subsequent dispersion and
transport in the layer aloft and in the surface layer make it difficult to
estimate the relationship between precursor concentrations in the two layers.
As indicated in Table 5.2, the 0.10 ppmC of transported hydrocarbons was
assumed to consist entirely of butane, reflecting substantial reduction of the
reactivity of the hydrocarbon mix due to reaction of the more reactive species.
5-13
-------
Transported concentrations of both NO and aldehydes were assumed to be neg-
A
ligible.
Tables 5.3A and 5.38 show the impacts on peak ozone concentrations of
worst case and typical case pollutant transport, respectively. The results
shown on these tables suggest that the transport of ozone is more significant
than the transport of other pollutants. The transport of concentrations of
precursors for worst case conditions can also have a significant effect,
particularly if concentrations transported aloft are as high as those trans-
ported in the surface layer. However, as discussed previously, the highest
precursor concentrations transported aloft are likely to be substantially less
than the highest concentrations transported in the surface layer. Table 5.3A
suggests that if negligible concentrations aloft are assumed, the worst case
impact of precursor transport is considerably less than the worst case impact
of ozone transport. Table 5.3B suggests that the typical impact of precursor
transport is much less than the typical impact of ozone transport. These two
tables suggest that the impact of worst case precursor transport is comparable
to the impact of a moderate level of ozone transport. Precursor transport
thus appears to be occasionally significant but generally much less signifi-
cant than ozone transport.
The results in Table 5.3A also suggest that increases in the initial
aldehyde concentration have a negligible impact on peak ozone concentrations.
In fact, an increase in aldehydes in one case caused a decrease in the peak
ozone concentration. This may be due to an increase in the depletion of NO
/\
by the forming of such compounds as HMO-.
Additional simulations were conducted to assess the relative significance
5-14
-------
LO
UJ
_J
CQ
taC
to
c:
o
4_)
to
1 ^
C
O)
o
c
0
OJ
c:
o
N
O
to
J3
rd
3
^
r.
o
Q-
to
Z3
O
fU
-a
to
4J *
3
.a
r
Sw
4-3 4-3
4_> ^_
a_
to
ta o
a> 4->
S-
a
c
S-
o
00
3
(J
0)
s_
Q-
S-
o
to
S-
^
o
2
Q-
a>
c
0
IM
O
a; s-
"O O
>, a.
au c
-a to
i S-
a: t
_Q
CL
CL
o <^>
f^
q
0
r
4-J CD *=C
i- 0
O 'O "O
a-1*- c
to S- fO
c~ 3
(O 00 S-
S- OJ
(- c >,
i- 03
1
-Q
a.
CL
CO 0
*>d~ ^"
4-> 01
s- u
o ,
C 3 «
ro OO I
S-
f c
r
J3
CL
CL
=s- to
4J
^
o
4J ^^5
S- O
O rOTD
O. H-c
c/) i~ rO
C 3
ro OO S-
S_ (D
J r~ ;>.
r- ro
_o
CL
CL
r-.. oo
10 un
"^
,
i
CO
^d~
^J-
oo co|
cu
en
ro O)
CQ Q- O
Q
CL
CL
00 CNJ
OJ CO
00 00
e
CL
to
C
To 4J
i- tO
4-> S-
"r~ -4^
c: c
i i
-------
to
c:
o
s_
0
to
i-
Z3
O
OJ
2
ro
jg
CD
D_
+-1 CD 0
S_ CJ r
o ro eC
O.M-
00 S- "O
ro oo ro
1 c: s-
- CD
ro
_Q
Q.
CL
CTl
CO CM
LT) CM
.a
CL
CL
UD
co
.
to
LU
1
J
CQ
c:
(U
u
c
o
CD
^
O
tsl
O
_^
ro
CD
Q_
o
en
4->
c~
ro
-t->
13
i
'o
D_
00
O
t-
ro
CD to
c: c:
o ro
N S-
-o t
^V^ 1 ^
ro c
^
c. i
"~ "o
CD CL.
to
ro o
CD +->
O
V 1
S- -M CD
o s- u
to o ro s_
S- O-M- QJ
rs oo s~ ^
U C 3 ro
cu ro co i
S- i-
cx 1 e
-
-Q
CL
CL
U3 CO CM
^j
0
+->. 1^
f U^ ^^-1-
O -r-t
CXo_ _.
CD OO r Z«
^ ij S- ro
o ro t
o£^ oj
^- ro'
_j
o
CL
Q.
CO CO LO
*± CO CM
-Q
Q.
CL
CTl
I--.
CO
COl
O.
Q.
0)
cn
ro
QJ
«=c
o
CL
to
C
ro
CD
to
ro
ro
u
CD
CD C
tO O
ro M
o o
CD -^
to ro
ro CD
CD Cu
s_
o
O-
to
£^
ro
S-
4->
O
c:
o
Q.
Q.
CX3
CM
CO
cvl
CO
CO
,
CM
CO
CO
0
CM
oo
o
CM
to x
c o
o -z:\
oo
o
O CM
o o
E
CL
Q.
CM
O
O
O
o
10
o
LO
«3- CM CM
CD CD CZ>
C3 O CZ>
CD
ro ro
- i-
u
ro
CL
E
CD
O
O Ol
CL
CL
CM
L£D r^v CTl
O O O
^" UO CM
CM CM CO
O O O
X
\ o
O -r-
:r +J
s: ro
o
5-16
-------
of ozone transported aloft and ozone transported in the surface layer. Specif-
ically, simulations were conducted to compare the impact of 120 ppb of ozone
transported aloft and the same concentration transported in the surface layer.
Table 5.4 shows that the impact of ozone transported aloft is much greater
than the impact of ozone transported in the surface layer. If consideration
is given to ground level ozone scavenging, the comparison suggests even more
strongly that ozone transported aloft is more significant than ozone trans-
ported in the surface layer.
5.5 The Impact of Ozone Transport Under Various Conditions
The results discussed in the previous section suggest that the best
variable for characterizing a given transport situation is the ozone con-
centration transported aloft. This section presents the results of numerous
simulations for various ozone concentrations transported aloft in a variety of
situations.
The impact of ozone transported aloft is influenced by several factors
that can be used to characterize a given day in a given city. These factors
include the NMHC/NOV ratio, the peak ozone concentration in the urban area,
A
dilution rate, the relative emission rates,* and the diurnal pattern of light
intensity. Table 5.1 and the discussion in Section 5.3 present the conditions
used in simulating pollutant transport. Using this basic set of conditions,
simulations were performed for various combinations of four concentrations of
ozone aloft, the two dilution rates, and the nine pairs of initial NMHC and
* Emission rates are normalized relative to emissions prior to 8 a.m. LOT.
5-17
-------
O)
>1
1X3
S-
CD
£:
a.
CO
O
I
ID
Q
CD
t-~
f>
rO
4_J
^
r"i
i
S_
4-> +->
=£ 0
Q.
,
CL ra
1
O a>
CM C CD
r O O
N ro
0 4-
S-
(^
o
0.
Q.
r~
I
Q.
Q.
CM CNJ CX3
CXI
LT)
CM
^ O)
01 o
Q,
o
i- O
ai
rO
0)
O
C
Q.
Q. QJ 4->
C 4-
O O O
CM Nl r
r- O
-------
NO concentrations. These are shown in Figure b.2. Additional simulations
/\
were performed for the four concentrations of ozone aloft and four selected
NMHC/NO pairs with: (1) two alternative sets of relative emission rates, and
X
(2) an alternative diurnal light distribution, as shown in Figure 5.3. As
indicated therein, all these additional simulations were based on a dilution
rate of 13%/hr.
Although this discussion focuses on the ozone concentration transported
aloft, other transported pollutant concentrations were also varied in these
simulations. The ozone concentration transported into the city in the surface
layer in the early morning was in most cases assumed equal to the ozone con-
centrations transported aloft. The only exception to this was for the case of
200 ppb of ozone aloft, in which case the ozone concentration transported at
ground level was assumed to be 120 ppb. The initial aldehyde concentrations
were left at their "normal" mix ([formaldehyde] = .02 [NMHC] and
[acetaldehyde] = .03 [NMHC] ) when the ozone concentration aloft was low
(0-, = 0 ppb or 40 ppb) and twice as high when the ozone concentration was high
(0., = 120 ppb or 200 ppb). As discussed in the previous section, this adjust-
ment is greater than might normally be expected, but this should not noticeably
affect the estimated impact of pollutant transport.
These simulations did not incorporate the transport of NMHC or NO . The
X
transport of these precursors can be assumed to cause an increase in the peak
ozone concentration. However, the relative variations of the impact of
pollutant transport under various conditions are assumed not to be affected by
the presence or absence of transported precursors.
Table 5.5 presents the results of the simulations shown in Figure 5.2 for
5-19
-------
co
E
cr
o
CJ)
CD
CO
*
*
E
CD
(J
E
O
CJ
4-> 4-
CO O
S- r-
-!-> CC
E
CD "O
CJ CD
E 4->
O S-
O O
Q-
CD CO
E E
O CO
N S-
.
UD
i
i
O
r"
r~^-
CD
UD
^
«^)~
|
0
^
1
CM
O
CM
^
CO
0
O
co"
cr>
o
i\
0
0
r^
0
f
0
to
o
C-J X
H °
CM
0
0
CM
CO
O
LT)
CM
O
.
O
^
CM
0
CD
,
CD
^>
CM
0
CD CJ
oo rn
i_
co
E
il
0- E
0
CN '^ OJ
^ 2.SL
K
X
o
S- i-
OO CD
r CM
JD JQ J2 -Q
CL CL Q. Q.
D. CL CL CL
O
o
CM
O
O
CM
4-* CD
M c~
O +->
r
ro C
0
CO
O CD
i- CO
4-> CO
s_
4-> CD
£3 £_
0) CD
<-> IS
E
O CO
CJ E
0
CD -r-
f J_J
O CO
Nl i
0 =3
E
4- T-
O to
CO 1
O S:
P
03
E CO
r- E
JD 0
E "1
O -4->
O ro
s_
t 4->
i E
CO CD
(J
i- E
O O
4- O
"d i
O) CO
!->
CJ 4->
13 *i
"0 E
E -r-
o
(J -o
E
CD
r
4-*
ro
i
CD
S-
O)
c~
43
"O
E
CO
E
O
1
4-3
rs
r~i
'i
S-
40
to
1
-o
4-J
c~
CD
r
i
CO
E
rs
i
^.
.^
c_3
E
Q.
Q.
E
r
'
^y
,
CO
i
[ ^
,
E
i
CD
>
.,
40
O
O)
CL
to
£
CD
^r
4->
C71
E
i
CO
E
O
4_j
CO
>
r-
to
CO
CO
E
CD
to
CD
d
CL
CD
s^.
S-
1
CO
CL
_E
CJ
CO
^ V
E
CL
O-
E
i
oo -a -o
CD
-O
CO
S_
CO
CO
E -O
O CD
r- 4J
4-> CO
"O ZS
E E
o <-
O 00
5-20
-------
__
a
OJ
-r-> a.
t O
3 Nl
E 0
1
(/}
<_ )
O
4~»
5
c
o
( ) !_)
r CD
rC 4-> t/1
d co -a
o -i- c
r -X CO -r
4-> -x <~ r-i
r- CO QJ
o QJ a.
T3 4->
e^ ro
rv
r r
->-> c
,o ° c
L0 T- 0
CO !
, CO 4-J
9j -.- (0
CD LU L§)
[~ ' ro
"-1- QJ 4->
> CO
)
4-J
ro
QJ
a;
a» QJ
CO (/)
ro (13
CD (_>
, -
-C Ci t
r- (- QJ 1
fO (T3 f~
CO r CcL -I
c; QJ --
o a: i/o
r t; (/
4-- oo O Q,
rC 3 -r- t
1 O 1/1 T-
3 -r- CO 1
E S- --
r- ro E
00 > LiJ
K
*X i
'VI
C i CNJ
, 0
- oo-
4" UJ |CM
r^ o")
CL'
O ' CJ
o ~ ~
0
r CM
'3 O
r- O ^
i CNJ CNJ
i r O
I
i
T
^ j
O "O
r CU -O o n _ci
4-J 4-J CL CL o. CL
r3 i- O. CL. CL CX
'_ O 4^
4-> CL<4- CD C CD CD
c: co o «rf- CM CD
C d i (CNJ
!_> (D c£
C S-
0 I
j
o^ VjD ^~ t^")
0 I 1 1
O CTi CO ^J-
i CNJ i
o-3 CO CO CD O O CC
O CO ^3 ^ LT) LD CC
i OJ CM CJ CM CNJ CN
r
s< en ro o ro
0 .... I
o co co in co
i CM CNJ r
I
Zl*=1^^51^2
e^ e^£ «r^ etC "--H r> c
OO CTl CD i CM c O
i i i
O
4-> 1 1 1 1 1
Q- CO O1 O i CM r-
' ) f f-^ r
x
-K
t/)
O
r
(a
4->
e
O)
o
o
CJ
(
ro
r-
4-J
I t
r_
O "C
r- QJ
40 4-J
1} '-^ JS_ + ^
~ > O L+~
O -M Cl. O
NJ C CO '
o CJ e cu
C_J ro
£J S-
0 t
1
1
III CO
c:
O
1 1 *
r
"U
O
^
4_)
i
5 co co r-^ co
3 i en «cr cu
J CO CO UT *->
c
h 4
4-J
C'
cr
r
i
i i i
4-J CO
C C
CJ O
SM *r !
QJ 4->
4- -r-
4- -o
i C"
"O O
U
-C
- s: s: 2: £ -M
- Q_ Q_ Q_ CO
00 C
J CO «3- Lf) CO)
O 4-J
'r C
III 4-J (-
03
- CNJ CO ^J- . 4->
3 -C
E C7
1 T
CO r
1
. ICNJ
r |0
O O
CO CM
r^ t^o
r- 0
CD
CNJ
0
o \.
CM CNJ
0
J
i
]
1
oa -Q -Q -Q
CL CL C_ Q.
d L-L CL CL
0 0 O 0
-i- CM O
r CM
1
UO
r-~ r
CM
Q. C
QJ Z3
CO "-3
*> f
'^ S
O
o CO
i_n i
f-^^ r
V «
zr" ii:
o o
o «a-
^)- co
O '-^
*4 C E
0 0-
>) CL
r ~c^
C QJ CI
O 4-J -1-
0
QJ X
4-J QJ O
US- «-
13 QJ ~^~
"CJ -d ^~~"
cz 2 c_j
o E
04-) CL
CL CL
QJ QJ
S_ O C
CU X T-
2 CL)
00 . CJ
C LT> E o.
O CU -SI
i i ro -^"
^ ) Q
*X3 (T3 OO i~~
^ 0 T-
E C 4-> 4->
OO S £1
c: o -r-
QJ -^
> S- QJ
c -r- a. >
2 C7> !-
O 00 4->
_c: QJ c u
CO i- O CU
QJ <- Q.
co 2 co co
C oo QJ
r- C i
4-> O QJ OJ
ro 'i -d
C 4-> O 4-J
r~ -I- 4->
ja ~o en
Ed QJ C
cj U -,- oo
4-> =5
r QJ (T3
r CO i d
(o -i at o
2 s- T-
i- S- 4->
O QJ "O (O
4- -d QJ i
4-> N 3
-a O .- E
QJ r T-
4-> ra CO
CJ E
-d-d 0
t: ^. d co
O o^ 4-J
o co QJ c:
r S- QJ
QJ ra oo
S- II QJ
QJ co S-
2 QJ QJ Q-
4-> 4-> QJ
CO rO QJ ra S-
C S- S- S-
O 13 i-
i c; cji d *>~
ro T- ^ -r- CL
4-J 00
13 3 co co -c:
e , -r- .r- 0
CO ~O 4-J LiJ LlJ
# -X *
-X *
5-21
-------
the various combinations of ozone concentration transported aloft, dilution
rate and initial concentrations. These results may be examined to consider
both the average impact of transported ozone and the variability of this
impact as a function of various parameters. As might be expected, the most
significant parameter is the transported ozone concentration itself. As shown
in Table 5.5, for ozone concentrations transported aloft of 40 ppb, 120 ppb
and 200 ppb, the average increases in peak ozone concentration constitute 40
percent, 48 percent and 45 percent of the respective ozone concentrations
aloft. These results suggest that the impact of ozone transport tends to
increase in proportion to increases in the transported ozone concentration.
The impact of ozone transport also seems to increase roughly in proportion to
increases in the transported ozone concentration for any of the various combin-
ations of peak ozone level, NMHC/NO ratio and dilution rate shown on Table 5.5.
X
In addition to looking at average impacts, it is also illuminating to
compare the impact of transported ozone on peak ozone concentrations to the
transported ozone concentration for individual simulations. The most extreme
of such comparisons is an impact of 21 ppb vs. the transport of 120 ppb (18
percent) and an impact of 85 ppb versus the transport of 120 ppb (71 percent),
although more extreme values may occasionally occur. It is clear that the
entrainment of ozone transported aloft does not cause a simple addition to
ozone concentrations in the air parcel. (Table 5.3 suggests that the trans-
port of precursors also does not cause any simple addition, either to pre-
cursor concentrations or to ozone concentrations). On the other hand, trans-
ported ozone clearly has at least some effect on the chemical processes that
produce ozone.
5-22
-------
-O
s-
&e"
-Q CTl
Q. CM
CL
CL
O
r
^
a
o s- .a
0 -C
CVJ \
CL
Q.
Lf)
CO
10
CO
in
CO .
r- O
o
CO
CM
LO
CTl
CT)
CM
LO
CM
CM
O
CT>
oo
ID
coi
-13
CL
O_
r LO
CTl ^J-
cu
O
it-
es
U
fO
CL
E
in
a
O)
+J
B3
+-> C
03 O
O
C
O
O
CO
O)
cr
ro
CU
o
<
co
O
O) -1-J
CO Q-
o
00 CM i
CM CO CM
OO CO CO
CO CO CO
o o o
CM CM i
CO
O
O
O
CM
o
(13
r-
ro
d
(JD
IO O
5-23
I
=£
*
-------
Table 5.5 may also be used to examine the impact of transported ozone as
a function of the peak ozone level, NMHC/NO ratio and dilution rate. The
X
first of these variables, peak ozone concentration, represents the maximum
ozone levels with no transport of pollutants into the city. This may be
considered an indicator of the quantity and concentration of locally emitted
precursors. For any given NMHC/NO ratio and dilution rate, the impact of
/\.
pollutant transport increases as the concentration of locally emitted pre-
cursors decreases. This is an expected result. As a second result, for each
peak ozone level and dilution rate, a decrease in the NMHC/NO ratio can cause
/\
a substantial increase in the impact of transport. This suggests the impor-
tance of the chemical interaction of the transported ozone and the precursor
emissions. If an air parcel with a low NMHC/NO ratio is assumed unable to
A
produce a high N09/N0 ratio, the role of transported ozone may be to foster
L- X
the high NO-/NO ratio needed to produce high ozone concentrations. As a
U- A
third result, an increase in the hourly dilution rate can be seen to cause an
increase in the impact of ozone transport. This is an expected result, since
a greater hourly dilution, i.e., a more rapid increase in the mixing height,
means more rapid entrainment of the ozone aloft.
Simulations were also conducted to investigate variations of the impact
of ozone transport under differing conditions of relative emission rates and
light intensity. Relative emission rates seem to be most critically affected
by the assumptions regarding location of the air parcel relative to the city.
Among the assumptions made in the base set of simulations are that the air
parcel being modeled stayed in the center of the city until 9 a.m. LOT and
then left the city three hours later. Relative emission rates were also
derived using two other assumptions: (1) the air parcel stagnated at the
5-24
-------
center of the city for the entire day, and (2) the air parcel moved steadily
across the city, reaching the center of the city at 7:30 a.m. LOT and leaving
the city three hours later. Figure 5.3 shows the relative emission rates
derived using the base set of assumptions and the rates derived using these
two alternate assumptions. This figure shows that the use of an assumption of
stagnation conditions results in a greater fraction of emissions entering the
air parcel later in the day. These simulations should represent a typical
range of temporal patterns of relative emission rates.
The peak ozone concentrations in the absence of ozone transport and with
the impact of ozone transport under these several assumed emissions conditions,
are shown in Table 5.6. The peak ozone concentration without ozone transport
obviously increases and decreases as the total amount of precursors injected
into the system increases and decreases. In this regard, it is interesting to
note that the peak ozone concentration is distinctly non-linear with the total
precursor emissions. Instead, a better hypothesis is that the peak ozone
concentration is mostly affected by emissions occurring prior to midmorning.
It is clear that the impact of ozone transport is noticeably reduced if
emissions into the air parcel occur later in the day, and it is almost always
slightly increased if emissions occur earlier. Thus, it appears that if any
change occurs that results in a greater fraction of the emissions into the air
parcel occurring later, e.g., a slowing of the wind or shifting the emissions
out away from the center of the city or to later in the day, the impact of
ozone transport will generally decrease.
The results of the simulations using alternative diurnal distributions of
light intensity are shown in Table 5.7. As stated earlier, the alternative
5-25
-------
t/1
c
o
(->
1
T3
C
O
o
c.
o
^
t/>
to
r~
E
LU
+J
C
CU
S-
0)
M-
f
S-
0)
"T3
C
ID
P
*4
o
o
ro
Q.
E
^
in
w
§
M
c
o
1
4_>
to
s-
-|~>
c
CU
c
o
o
CU
c
o
N
0
V*
s-
0
Q.
U1
C
CO
i.
h-
CU
JT
^J
>,
-a
o
CU
M
Zl
-l->
3
O
_c
4J
r~
*^T
C'
o
XJ
a.
Q.
o
o
CM
1
I «3 r r OOOCOO
LOCO OOVO COCMCTiCTi
OO
O
X)
a.
o.
o
CM
^~
OOO OOOO VOOOOCM
COi Ln«3" LOCMr~.lO
CO
o
.a
a.
Q.
o
^"
r V£) tO CO *Q" r^* 00 tO
r r r " p r r
4J
1-
o
coto
0
^
CU
D-
c
(0
S-
1
r~~
(^
CL
D-
oun <-r^ ocMr-»CM
r*^. oo co CM o o 0% cr»
CO CO i r CO CO
"H
EJ
gj
^
X
o
"^~
^^^
^ ^
t ^
^
Q.
a.
x»^
CJ
z
s:
z
to
c
o
-i->
i
-a
c
o
01
c
o
r
to
1/1
E
LU
i <* CM i <* CM
r-00 r 0 0
CM CM
OOO 0 0 0
C^ *^x "^-^ "***"* CD "^""^ "^x, ^^^
**^ **o ^ CNJ ^x. \£> "^ c\j
o*j r^- c\j c*') <.\j r^- c\j oo
r r- 00 i 0 O
+J
Z5
o c:
"D-,' """
^ C ^>> -i '
O o r c:
s- T- s- a>
JZ -M fO 4-»
I rO LJ to
c: -r-
l/l CT. tO U)
C O CU
r tO ' d- ' ^K
t/) v (/) - (ft
(SI (/) "D
r- >, -r; >, C
fci n3 £ t3 >"
LU a LU a 2
He
tr> i
r**^« i
CM r
LT> CM
OO 00
f-~
00 r
CM CM
OO CO
r
r~«
CM
O
0 ^,
"v^ ^£>
oj r^
r r
OJ
00
fd
t >
to
to
CO
CO
0
r
^-
o
o
^^
«sj*
CM
0
-a
c
1
^~
t: "^
CD ra
4-}
t^ CT>
r*
t/V ^~
t- CU
CU 4->
D- M-
' (0
in
CO
0^
LO
LO
f
CM
O
r
CM
O
*
0
"^
CM
CO
0
01
c.
5-26
-------
CO
c
o
i
4-^
Q.
H
3
CO
CO
^"i*
C
O
-11
03
r*
o
CO
c
1 1
+J
c:
S-
4^
C
u
c
o
0
0)
c
o
N
0
XX
(O
cu
Q.
cu
-C
4->
c
1
cu
CO
03
CU
s_
(_)
C"
i i
l[J
O
4->
s_
o
Q.
CO
C
c
1
CU
J=
^>
-O
O)
CO
^J
«3
t >
cp
o
_Q
Q.
Q.
O
0
CM
Or CD O ID^rlDLD
^~-«^"OOCO h^ ^- g> CO
CO
0
.0
Q.
Q.
O
CM
CDU3COVO CMi U3CT1
*CMVOLD IDCMVOLO
"~
co
O
JS
Q.
a
o
«3- r»- CT> «* OOCOCOtD
i i r r r
Q
s^
cu
a
c
Z3
4->
tf_
O
O
-C
P 4-^
r- S-
31 O
Q.
COW
O C
n5
-* S-
T3 H~*
cu
a.
CM vo r^ cvi
tjO "3" r r^
CO CO r~~ r~~
. s
pr
f*i
a.
o
^~
^ ^
C_5
E
O.
Q.
1^^ *
0
ac
^
r «J- CM
i O O
CM
o o o
o ^, -^ ^.
^>. VO <* CM
CM r»« CM oo
9 «
i i O O
OS 00
CM CM O
OO CO i
^ ^j-
O
CM
0 O
O ^ -^.
\ VD «*
CM r-~ CM.
i r O
CM
O
CM
O
O
^^
CM
CO
o
O
r~
ID
C
O
o
00 i
CM
« c
Z 3
0 --3
«*
CO
cu
J3
r-
(O
~ZL D-
cu o a)
co O C^)
03 «sf
OQ
5-27
-------
insolation distribution has a peak Tight intensity roughly 17 percent higher
and a" integrated average light intensity roughly 10 percent higher than the
base case insolation. Table 5.7 shows that, in these cases, the peak ozone
concentration increases with increased light intensity. However, the increase
is generally less than 10 percent and is not readily quantified in terms of
the increase in light intensity.
It is interesting to observe from Table 5.7 that, if the light intensity
is increased, the impact of ozone transport is generally slightly less.
Although there is insufficient evidence to support any single explanation of
this effect, this again suggests the importance of the interaction of trans-
ported ozone with the photochemical reactions occurring in the primary air
parcel. However, variations within the normal range of insolation during the
part of the year when high ozone concentrations generally occur would have
little effect on the impact of ozone transport.
Table 5.8 summarizes the results discussed in this section. This table
lists the variables discussed previously, in approximate order of the sensi-
tivity of the impact of ozone transport within the typical range of values of
these variables. For more quantitative results, it is necessary to refer to
Tables 5.5, 5.6, and 5.7.
5.6 Anticipated Trends in the Impact of Ozone Transport
The findings discussed in the previous section are the result of simul-
ations that consider a typical range of current conditions. That is, these
simulations represent conditions in which the peak ozone concentration in the
urban area is still above the NAAQS, and the transported ozone concentration
aloft is in some cases significantly above natural background levels. This
5-28
-------
TABLE 5.8
The Impact of Ozone Transport Under Varying ConditionsSummary of Results
Variable
Change 1n
Variable
Change 1n
Impact of
Ozone Transport
Transported Ozone
Dilution Rate
Peak Ozone Concentration
NMHC/NO Ratio
rt
Emissions Into the
A1r Parcel
Light Intensity
Increase
Increase
Decrease
Decrease
A Higher Proportion
Occurring Earlier
(e.g., more per-
sistent winds)
Decrease
Increase
Increase
Increase
Increase
Increase
Slight Increase
5-29
-------
section is intended to consitly* the :Tpa:t ^ ozone transport under conditions
that are likely to exist whan L,ie HAAQS :r, approached.
In making projections for Jays in which high ozone concentrations occur in
the future, it is generally assumed that the future and current meteorological
conditions for those days will be sinilar. However, the future impact of
pollutant transport may be significantly affected by reductions in local
emissions and, in many cases, by reductions in the transported pollutant
concentrations resulting from control programs in upwind areas. The reductions
in emissions will generally involve a greater reduction in hydrocarbon emis-
sions than in NO emissions. Therefore, future MMHC/NO ratios will probably
X A
be lower than current ratios. In those cases where transported pollutant con-
centrations are currently in excess of natural background concentrations, it is
reasonable to assume that, as upwind emission control programs take effect,
transported pollutant concentrations will be reduced to levels approaching
natural background concentrations.
In order to assess the future impact of ozone transport, simulations were
conducted with two different NMHC/NO / ratios that might be expected to occur in
A
the future, the two dilution rates used previously, and appropriate initial
concentrations such that, the peak ozone concentration in the presence of ozone
transport was about 0.08 ppm. These simulations all used the base case condi-
tions of this study for insolation, relative emission rates and initial N09/N0
L, A
relationship (i.e., those conditions shown in Table 5.1).
The 8 a.m. NMHC and NO concentrations that were simulated and the impact
A
of transport under these conditions are shown in Table 5.9. In each case, the
impact of ozone transport indicates the difference between the peak ozone
5-30
-------
TABLE 5.9
The Impact of Ozone Transport as Air Quality Approaches the NAAQS
Conditions Simulated
This table shows the initial
MMHC (in ppmC)/NP (in ppm) used
in each simulation. All simulations
were based on relative emission rates,
insolation, and the NO^/NO relation-
ships given in Table 5.1 All simulations
were conducted with and without the
transport of 0.04 ppm 0,. and all
simulations in the presence of ozone
transport produced a peak ozone
concentration of about 0.08 ppm.
The Increase in Peak Ozone Concentration Due to Consideration of Ozone Transport
NMHC/NO^
2.5:1
4:1
Impact of transport of 40 ppb 0,
Dilution rate = 13%/hr. Dilution rate = 29°//hr.
31 ppb 28 ppb
21 ppb 25 ppb
5-31
-------
concentration in the presence of transport of 40 ppb ozona and the peak ozone
concentration in the absence of transport. The information in this table is
in many ways merely an extension and a confirmation of the trends suggested by
the simulations of current conditions. The future impact of ozone transport
will in many cases be lessened simply because the transported ozone concen-
tration is less. However, if a comparison is made of the current and future
impacts of the transport of 40 ppb ozone, the future impact would be expected
to be greater than the current impact, due both to the generally reduced
emission levels of the city and to the reduced NMHC/NO ratio. In addition,
/\
Table 5.9 provides further evidence that the impact of the transport of a
given ozone concentration generally increases as the NMHC/NO. ratio decreases,
X
although the evidence regarding variations in the impact of ozone transport
with various dilution rates is somewhat ambiguous.
5.7 Conclusions
This discussion presents the results of a study using the kinetics model
underlying EKMA to estimate the impact of pollutant transport on the peak
ozone concentrations occurring in an urban area. The study was divided into
two parts: (1) analyzing the impact of the transport of various pollutants in
the surface layer and aloft, and (2) analyzing the impact of transported ozone
under various conditions. In the first part of this study, the transport of
ozone into a city was found to have a greater effect on the city's peak ozone
concentration than did the transport of other pollutants. The dominant cause
of the impact of ozone transport was found to be the ozone transported in a
layer aloft. The second part of this study assessed the changes in the impact
of ozone transport occurring with varying peak ozone levels, dilution rates,
NMHC/NO ratios, relative emission rates and diurnal light intensity distributions.
J\
5-32
-------
Most of these simulations were directed toward estimating the impact of ozone
transport under a range of current conditions. These findings were used,
together with some additional simulations, to estimate the probable impact of
ozone transport as the NAAQS for oxidants is approached.
As discussed previously, one of the assumptions inherent in a kinetics
model is that instantaneous vertical mixing occurs. The mixing of ground
level emissions and pollutants entrained from aloft were therefore simulated
as occurring more rapidly than is actually the case. This may be compared
with the treatment of vertical mixing found in most photochemical dispersion
models, whereby a column of air is divided into several layers and a finite
diffusivity and vertical flow are used to account for pollutant transfer
between layers. This study was based on the assumption that ambient vertical
mixing within the surface layer is rapid enough that a kinetics model can be
used to compare the impact of transport of various pollutants and to estimate
the sensitivity of the impact of pollutant transport for various conditions.
A clear need exists to assess further some of the findings of this study by
using a photochemical dispersion model.
Nevertheless, in spite of the limitations of this study, the results
support several conclusions. The first set of conclusions results from the
comparison of the relative impacts of various transported pollutants. Unless
another city is located immediately upwind, transport of ozone, particularly
ozone transported aloft, will almost certainly have a greater impact on an
urban area's peak ozone concentration than the transport of any other pol-
lutant. Thus, it is generally more important to obtain accurate measurements
of transported ozone than of other transported pollutants. However, the
5-33
-------
chemical interactions of this transported ozone with the urban emissions are
such that the impact is generally much less than if the transported ozone were
simply added like an inert pollutant. The second set of conclusions results
from the analysis of the variations of the impact of ozone transport under
various conditions. This study also suggests that for a given transported
ozone concentration, the absolute impact of ozone transport can be expected to
increase as the local emissions of precursors decrease and as the NMHC/NO
A
ratio decreases. However, this increase in the impact of ozone transport may
well be overwhelmed by the effect of reductions in the transported ozone
concentration, as emission control programs in upwind areas take effect. Due
to the assumptions used in the model and in the derivation of the input data,
it may be inappropriate to use the quantitative results in this study to esti-
mate the impact of transport in individual cities. However, if specific data
are available for an individual city, the techniques described herein might be
used to estimate the impact of transport on peak ozone concentrations. In any
case, the results of this study clearly demonstrate that ozone transported
into a city in a layer aloft can, in a wide variety of situations, have a
significant effect on that city's peak ozone concentration.
5-34
-------
5.8 References
1. Uses, .Limitations^ and Techrn'caj Basis__of F'rocedurgsJ'or^ Quantifying
Relationships Between Photochemical Oxidants and Precursors,
EPA-450/2-77-021a, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, November 1977.
2. M. C. Dodge, "Combined Use of Modeling Techniques and Smog Chamber
Data to Derive Ozone-Precursor Relationships," International
Conference on Photochemical Qxidant Pollution and Its Control,
Proceedings: Volume II, EPA-600/3-77-001b, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, January 1977,
p. 881.
3. G. Z. Whitten and H. Hogo, User's Manual for a Kinetics Model and
Ozone Isopleth Plotting Package, Draft Final Report, Systems
Applications, Inc., San Rafael, California, November 1977.
4. L. H. Titlemore, et a1., An Analysis of Urban Area Travel by
Time of Day, FH-11-7519, Federal Highway Administration, Washington,
D. C., January 1972.
5. B. Dimitriades, ed., International Conference on Photochemical
Qxidant Pollution andIts Control,Proceedings: Volume 1,
EPA-600/3-77-001a, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, January 1977, pp. 227-338.
5-35
-------
6.0 ACKNOWLEDGEMENTS
This document was prepared by the following members of the staff of the
Monitoring and Data Analysis Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency: Warren P. Freas, John E.
Summerhays, Emerico L. Martinez, Edwin L. Meyer, Jr., Norman C. Possiel and
Donald H. Sennett. Whitmel M. Joyner provided editorial assistance, and Mrs.
Carole Mask prepared the manuscript. In addition, numerous studies conducted
in-house or under contract to the Environmental Sciences Research Laboratory,
U.S. Environmental Protection Agency, were of great assistance in preparing
this document.
6-1
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse bcfo/e completing}
1. REPORT NO. 2.
EPA-450/2-77-021b
4. TITLE AND SUBTITLE
Procedures for Quantifying Relationships Between
Photochemical Oxidants and Precursors: Supporting
Documentation
7. AUTHOR(S)
W.P. Freas, E.L. Martinez, E.L. Meyer, N.C. Possiel,
D.H. Sennett and J.E. Surrrnerhays
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Monitoring and Data Analysis Division
Research Triangle Park, N.C. 27711
12. SPONSORING AGENCY NAME AND ADDRESS
3 RECIPIENT'S ACCESSION NO.
5. REPORT DATE
February 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NC
10. PROGRAM ELEMENT NO.
2AA635
11. CONTRACT/GRANT NO
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report is a companion document to EPA-450/2-77-021a, Uses, Limitations
and Technical Basis of Procedures for Quantifying Relationships Between Photochemical
Oxidants and Precursors. Particular emphasis is placed on providing additional
backqround documentation on the development of the Eitpirical Kinetic Modeling
Approach (EKMA). Specifically, the report describes: (1) procedures for estimating
the prevailing 6-9 a.m. NMHC/k)x ratio for a city, (2) the use of default NMHC/NOx
ratios, (3) procedures for estimating transported ozone concentrations and (4) the
results of kinetic model simulations for a range of meteorological conditions and
precursor levels. Topics which have general significance in a number of different
procedures for quantifying ozone-precursor relationships include the prevailing
background concentration of ozone and the greater importance of transported ozone
as opposed to transported precursors.
17.
a.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution EKMA Isopleths
Ozone
Hydrocarbons
Nitrogen Oxides
Meteorology
Transport
Kinetic Models
18.
DISTRIBUTION STATEMENT
Release unlimited
b. IDENTIFIERS/OPEN ENDED TERMS
19 SECURITY CLASS (This Report)
Unclassified
20 SECURITY CLASS (This page j
Unclassified
c. COS AT I Field/Group
21. NO. OF PAGES
109
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
PREVIOUS EDITION IS OBSOLETE
------- |