United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/4-79-017
August 1979
Air
Assessment of Vertical
Distributions of
Photochemical Pollutants
and Meteorological Variables
in the Vicinity of Urban Areas
-------
EPA-450/4-79-017
ASSESSMENT OF VERTICAL DISTRIBUTIONS
OF
PHOTOCHEMICAL POLLUTANTS AND METEOROLOGICAL VARIABLES
IN THE VICINITY OF URBAN AREAS
by
F.L. Ludwig
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
SRI Project 6869
Contract No. 68-02-2662
Project Officer
E.L. Martinez
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR, NOISE AND RADIATION
OFFICE OF AIR QUALITY PLANNING AND STANDARDS
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
August 1979
.!9.RA.RY
-------
This report is issued by the U.S. 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 con-
tractors and grantees, and nonprofit organizations - in limited quantities - from the Library Services
Office (MD-35), Research Triangle Park, North Carolina 27711; or, for a fee, from the National
Technical Information Service, 5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by SRI International, 333
Ravenswood Avenue, Menlo Park, California 94025, in fulfillment of Contract No. 68-02-2662.
The contents of this report are reproduced herein as received from SRI. The opinions, findings and
conclusions expressed are those of the author and not necessarily those of the Environmental
Protection Agency. Mention of company or product names is not to be considered an endorsement
by the Agency.
Publication No. EPA-450/4-79-017
-------
ACKNOWLEDGMENTS
Numerous people helped in the preparation of this report. Kenneth
Nitz, Eugene Shelar, Rosemary Maughan, Frances Adams, and Kathleen
Chaiken assisted with data processing and analysis. Joyce Kealoha, L.H.
Wu, Jack Byrne, Shirley Bartels, Jane Hazlett, and Josette Louvigny
assisted in the preparation of the figures and text.
E. L. Martinez and Warren Freas of EPA had many valuable sugges-
tions during the course of the research. Messrs. Martinez and Freas,
along with Dave Mage of EPA, provided most of the data necessary for the
analyses described in this report. Mr- Dale Coventry of EPA also pro-
vided data and was extremely helpful with calculations of air
trajectories.
iii
-------
ABSTRACT
A system for classifying vertical profiles of ozone and other pho-
tochemical pollutants has been derived by analyzing 268 ozone profiles
obtained from published reports and EPA data files—93 from St. Louis,
40 from Los Angeles, 30 from Houston, 31 from Washington, B.C., 53 from
Toronto, and the remaining 21 from Indianapolis, Tampa, and Denver. The
profiles fall into six different categories that are related to the
relative importance of the following factors: destruction of ozone near
the surface, vertical mixing, and photochemical production. Vertical
mixing tends to produce profiles that are nearly uniform with height.
Strong mixing offsets the destruction at the surface. Unless effective
vertical mixing of ozone occurs, destruction at the surface will produce
profiles with steeply increasing concentrations with height in the
lowest layers. The report presents and discusses 108 representative
examples of the vertical profiles, most with corresponding temperature
data and many with aerosol (nephelometer), NOX, or hydrocarbon data.
Linear regression has been used to define the relation between ozone
concentrations in the mixing layer and those aloft, for the different
profile types. The locations relative to an urban area and the times
when each profile type is most apt to occur are discussed in terms of
the statistics of the data set and the physical processes involved.
iv
-------
CONTENTS
ACKNOWLEDGMENTS iii
ABSTRACT iv
LIST OF ILLUSTRATIONS vii
LIST OF TABLES xi
SUMMARY AND CONCLUSIONS xiii
I INTRODUCTION 1
A. Purposes of the Study 1
B. Sources of Data 1
1. Background 1
2. Primary Areas Selected for Study 2
C. General Approach 4
II PHYSICAL PROCESSES GOVERNING PHOTOCHEMICAL POLLUTANT
DISTRIBUTIONS IN THE VERTICAL AND CONSEQUENT PROFILES 7
A. Factors Governing the Vertical Distributions
of Ozone in the Vicinity of Urban Areas 7
1. Background 7
2. Formation, Destruction, Transport, and Mixing
of Ozone in the Vicinity of Cities 10
B. Types of Vertical Ozone Profiles 13
C. Other Considerations Concerning Ozone Profiles 18
III OBSERVED VERTICAL PROFILES 21
A. General 21
B. Primary Study Areas 21
1. St. Louis 21
2. Los Angeles 38
3. Houston 63
4. Washington, D.C 89
5. Toronto 103
6. Indianapolis 108
7. Denver 108
8. Tampa Bay, Florida 110
C. Vertical Cross Sections of Ozone Concentration
in the Northeastern United States 110
-------
D. Summary of Ozone Profile Features • 118
1. General 118
2. Characteristics of the Different Types
of Profile 119
3. Relationship Between Surface Ozone
Concentrations and Those in the Mixed Layer 122
E. Summary of Profile Features
for Other Photochemical Pollutants 139
IV VARIATIONS IN THE MIXING LAYER AND WINDS NEAR CITIES 143
A. Background • 143
B. The Mixing Layer 143
C. Winds 157
REFERENCES 161
vi
-------
ILLUSTRATIONS
1 Schematic Representation of Ozone Variations
at the Surface and in the Free Troposphere...
2 Idealized Ozone Variations at Remote Locations
3 Schematic Diagram of the Transport
of Precursors, Ozone 10
4 Six Types of Vertical Profile 14
5 Typical Diurnal Evolution of the Vertical
Ozone Profile 16
6 Monitoring and Helicopter Sounding Locations
in the St. Louis Area 26
7 St. Louis Ozone Concentrations on the Morning
of July 19, 1976 28
8 St. Louis Ozone Concentrations on the Morning
of July 23, 1976 30
9 St. Louis Ozone Concentrations Around Midday
on July 23, 1976 32
10 St. Louis Ozone Concentrations on the Morning
of July 29, 1976 33
11 St. Louis Ozone Concentrations on the Morning
of July 30, 1976 34
12 St. Louis Ozone Concentrations During Midmorning
on July 30, 1976 36
13 St. Louis Ozone Concentrations at 0830 to 0930 CST
on the Morning of August 4, 1977 37
14 St. Louis Ozone Concentrations at 0930 to 1030 CST
on the Morning of August 4, 1976 39
15 St. Louis Ozone Concentrations During the Morning
of August 12, 1976 40
16 Map of Southern California 43
17 Example of Ozone Layer Aloft, Probably Caused
by Upslope Flow 44
18 Pasadena Profiles During the Afternoon
of September 19, 1973 46
19 El Monte Profiles During the Afternoon (1336 PST)
of September 19, 1973 47
20 El Monte Profiles During the Afternoon (1413 PST)
of September 19, 1973 48
vii
-------
21 Compton Profiles During the Afternoon
of September 19, 1973 49
22 Compton Profiles During the Afternoon
of September 24, 1973 51
23 El Monte Profiles During the Afternoon
of October 5, 1973 52
24 Upland Profiles During the Afternoon
of October 5, 1973 53
25 Crlendale Profiles During the Afternoon
of October 25, 1973 54
26 Mira Loma Profiles During the Afternoon
of October 28, 1973 .•••• 55
27 Arcadia Profiles During the Afternoon
of October 29, 1973 56
28 Arcadia Profiles During the Afternoon
of October 30, 1973 58
29 San Marino Profiles During the Afternoon
of October 31, 1973 59
30 Profiles Over the Santa Fe Flood Control Basin
During the Afternoon of November 7, 1973 60
31 Arcadia Profiles During the Afternoon
of November 7, 1973 ._. 61
32 Cucamonga Profiles During the Afternoon
of November 7, 1973 62
33 Houston Ozone Concentrations During the Late Afternoon
of July 2, 1976 66
34 Houston Ozone Concentrations Near Midday
on July 4, 1976 67
35 Houston Ozone Concentrations During the Late Afternoon
of July 5, 1976 68
36 Houston Ozone Concentrations During the Morning
of July 7, 1976 70
37 Houston Ozone Concentrations on July 8, 1976 71
38 Houston Ozone Concentrations During the Afternoon
of July 10, 1976 72
39 Houston Ozone Concentrations During the Morning
of July 12, 1976 73
40 Houston Ozone Concentrations During the Afternoon
of July 12, 1976 74
41 Houston Ozone Concentrations During the Late Afternoon
of July 13, 1976 76
viii
-------
42 Houston Ozone Concentrations on July 14, 1976 77
43 Houston Ozone Concentrations on July 15, 1976 79
44 Houston Ozone Concentrations During the Morning
of July 18, 1976 80
45 Houston Ozone Concentrations on July 20, 1976 81
46 Houston Ozone Concentrations During the Morning
of July 21, 1976 82
47 Houston Ozone Concentrations on July 22, 1976 84
48 Houston Ozone Concentrations During the Morning
of July 23, 1976 85
49 Houston Ozone Concentrations on July 24, 1976 86
50 Houston Ozone Concentrations During the Morning
of July 25, 1976 87
51 Location 48 Hours Earlier of Air Arriving in Houston
on August 24-25, 1976 88
52 Washington, D.C. Ozone Profile, 1113 EST,
August 17, 1976 91
53 Washington, D.C. Ozone Profiles During the Morning
of August 18, 1976 93
54 Washington, D.C. Ozone Profiles During the
Late Afternoon of August 18, 1976 94
55 Trajectories of Air Arriving at Washington, D.C.
on August 18, 1976 at 0700 and 1900 EST 95
56 Washington, D.C. Ozone Profiles During the Morning
of August 19, 1966 96
57 Washington, D.C. Ozone Profiles During the Afternoon
of August 19, 1976 98
58 Washington, D.C. Ozone Profiles During the Morning
of August 20, 1976 99
59 Washington, D.C. Ozone Profiles During
the Late Afternoon of August 20, 1976 100
60 Trajectories of Air Arriving at Washington, D.C.
on August 2, 1900 EST; August 21, 1700 EST;
and August 23, 0700 and 1900 EST, 1976 101
61 Washington, D.C. Ozone Concentrations at About Midday
on August 21, 1976 102
62 Washington, D.C. Ozone Profiles, August 23, 1976 104
63 Area in Which Toronto Profiles Were Taken 107
64 Vertical Cross Sections over Connecticut 1545-1715,
August 21, 1976 112
ix
-------
65 Vertical Cross Section from Long Island ^
through Cape Cod, 1258-1520, August 12, 1975 ..............
66 Vertical Cross Sections of Ozone Concentration,
0815-1055, August 21, 1975 ................................
67 Vertical Cross Section of Ozone Concentration,
1324-1500 EST, August 21, 1975 ............................ '
68 Histograms of the Ratio of Surface Ozone Concentration
to the Average Concentration in the Mixing Layer .......... 125
69 Histograms of the Difference Between Average
Concentrations of Mixing-Layer Ozone and
Surface Ozone ............................................. 126
70 Histograms of Ratios and Differences Between
Concentrations of Ozone at the Surface and in
the Mixing Layer, for Different Times of Day .............. 128
71 Histograms of Ratios and Differences Between
Concentrations of Ozone Concentrations
for Different Atmospheric Stability Classes ............... 131
72 Scattergrams of Mixed-Layer and Surface Ozone
Concentrations for the Six Types of Vertical Profile ...... 135
73 Scattergrams of Average Mixing-Layer Ozone
Concentrations versus Surface Concentrations
for Profile Types Associated with a
Well-Mixed Surface Layer ........................... ....... _ L36
74 Scattergram of Average Mixing-Layer Ozone
Concentrations versus Surface Concentrations
for All Measurements Between 0900 and 1500 LST ............ 138
75 Gray-Scale Display of Lidar Data for August 3, 1976
from 0929 to 1145 Showing the Changes Produced
by Thermal Mixing ......................................... 146
76 Diurnal Changes in Mixing Depth in St. Louis,
August 9, 1976 ............................................ 147
77 Gray-Scale Display of Lidar Data for August 3, 1976
from 1958 to 2124 Showing Establishment of the
Mixing Depth at a Low Level ............................... 148
78 Location and Elevation of the Lidar Route Used
During the 1975 Metromex Study ............................ J5Q
79 Lidar Observed Maximum and Minimum Mixing Depths
as a Function of Time .....................................
80 Average Increase in Urban and Rural Influences
on the Depth of the Mixing Layer
Near St. Louis
. „
81 Typical Wind Profiles over Urban, Suburban
and Rural Areas .......................................
-------
TABLES
1 Summary of Ozone Profiles Studied from the
St. Louis Area 23
2 Relative Frequencies of Ozone Profile Categories
in the St. Louis Data 27
3 Summary of Ozone Profiles Studied from the
Los Angeles Basin 41
4 Summary of Ozone Profiles Studied from the
Houston Area 64
5 Summary of Ozone Profiles Studied from the
Washington, B.C., Area 90
6 Summary of Ozone Profiles Studied from the
Toronto Area 105
7 Summary of Ozone Profiles Studied from the
Indianapolis Area 109
8 Locations of Spiral Measurements Relative
to the Center of Indianapolis 108
9 Summary of Ozone Profiles Studied from the
Denver Area 109
10 Summary of Ozone Profiles Studied from the
Tampa Bay Area Ill
11 Frequency of Occurrence of Different Profile Types
at Different Times of Day 120
12 Frequency of Occurrency of Different Profile Types
at Different Locations Relative to the City 123
13 Average Differences and Ratios Between Ozone
Concentrations at the Surface and Aloft,
for Different Profile Types 127
14 Average Differences and Ratios Between Ozone
Concentrations at the Surface and Aloft,
for Different Times of Day 130
15 Averages and Standard Deviations for the Differences
and Ratios Between Ozone Concentrations at the
Surface and Aloft, for Different Stability Classes 132
16 Average Differences and Ratios Between Ozone
Concentrations at the Surface and Aloft,
for Different Locations Relative to the City 133
17 Ozone Profile Types Observed with Various
Nephelometer Profile Types 140
xi
-------
144
18 Mixing Depth at Different Times of Day
19 Variation of Wind Exponent and the Height
of Gradient Wind with Surface Characteristics—
After Davenport (1968)
xii
-------
SUMMARY AND CONCLUSIONS
A. Findings
1. General
The development of effective ozone control strategies requires at
least some basic knowledge of the three-dimensional distribution of pho-
tochemical pollutant concentrations and meteorological conditions.
Available information, however, is usually limited to a few observations
made near the surface. The objective of this study has been to develop
a characterization of the three-dimensional nature of the atmosphere
which might be inferred from the surface observations and perhaps, a few
measurements obtained aloft. Those special studies where vertical pro-
files of pollutant concentrations were measured around cities have
served as the starting point for the study.
The open literature and the U.S. Environmental Protection Agency
(EPA) have provided 268 vertical profiles of ozone from seven different
areas that have been studied in detail and classified. The data base
included 93 profiles from the St. Louis area, 40 from the Los Angeles
Basin, 30 from the Houston area, 31 from the Washington, D.C. area, and
smaller numbers from Toronto, Indianapolis, and Tampa. Most of the pro-
files are accompanied by measurements of the vertical distribution of
temperature; in many cases, the vertical distribution of one or more of
the following parameters is also available: dew point, aerosol back-
scatter (Bscat), oxides of nitrogen (NO), total hydrocarbons (THC), and
nonmethane hydrocarbons (NMHC).
Careful examination of the pollutant profiles in general, and the
ozone profiles in particular, showed that their shapes could be classi-
fied into six categories, shown in the figure on the next page (Figure 4
in the text). The shapes are determined by three different physical
processes: vertical mixing, surface generation or destruction, and pho-
tochemical production. Section III presents many of the vertical pro-
files from the cities of St. Louis, Los Angeles, Houston, and Washington
and discusses the classification of the profiles.
2. Types of Pollutant Profiles
a. Type .A Profile
The Type A profile is characterized by nearly uniform concentra-
tions with height. Such profiles are produced when the vertical mixing
is strong enough to offset destruction of the ozone at the surface and
when little or no ozone is being formed from precursors within the boun-
dary layer. Thus, one might expect this to be the most common profile
during the forenoon, when mixing is likely to be reasonably good but the
photochemical production of ozone is not as pronounced as during the
afternoon. The data confirm the expectations: about 45 percent of all
xiil
-------
profiles measured between 0900 and 1200 Local Standard Time (LSI) were
of Type A. About 55 percent of the measurements during this morning
time period upwind or to one side of the city fell in this category if
only profiles measured upwind or to one side of the city were
considered.
b . Type _B Profile
The Type B profile is similar to the Type A except that concentra-
tions in the lowest layers are much less than they are at higher alti-
tudes, because destruction processes near the surface dominate vertical
mixing. Ozone is destroyed much more rapidly near the surface than it
can be replaced by mixing from aloft. This situation can arise when the
mixing is inhibited by a stable layer at the surface or when the
destruction processes are very strong because of the presence of consid-
erable amounts of nitric oxide (NO). Nitric oxide will be present near
the surface in areas of strong emissions. A stable layer will frequently
form at the surface with nighttime cooling of the ground or when air
passes over a colder surface, such as a body of water in the summertime.
Few Type B profiles were observed, but this was probably because there
were very few measurements available from nighttime hours. None of the
13 profiles categorized as Type B were measured after 1030 LST during
the daytime. Nine of the 13 were measured between 2100 and 0900. One
might expect that with the onset of mixing during the morning, Type B
profiles would evolve into Type A profiles.
c. Type C Profile
A Type C profile is characterized by concentrations that are
greatest near the ground and tend to decrease with height to the top of
the mixing layer, remaining relatively constant above it. We believe
this type of profile is most likely to occur during periods when photo-
chemistry is proceeding rapidly and vertical mixing is strong enough to
offset surface destruction processes, but has not proceeded to the point
where ozone formed from precursors near ground level has been mixed
throughout the boundary layer. One would expect a Type C profile to be
most common in the afternoon and, as it turned out, more than three-
fourths of the 21 times C profiles were observed were between noon and
1800 LST. None was observed at night.
d. Type D Profile
With continued mixing, a profile of Type C should evolve into a
Type D profile. Type D displays nearly uniform concentrations throughout
the mixing layer and a rapid decrease with altitude at the mixing
layer's top. The processes leading to a Type D profile are the same as
those leading to a Type C profile, but the mixing is more complete. It
is sometimes difficult to distinguish between these two types of
xiv
-------
profiles. Type D profiles were second most common among the profiles
examined in this program; 68 of the 268 classified cases were of this
type. As with the Type C profiles, they were most frequent during the
afternoon hours; nearly 80 percent occurred between noon and 1800 LST.
e. Type _E Profile
The Type E profile bears much the same relationship to a Type D
profile as Type B does to Type A. Above the lowest layers, Type E and D
profiles are similar- The surface concentrations are much less with a
Type E, because these profiles are formed when destructive processes at
the surface more than offset vertical mixing that brings ozone down from
aloft. A Type E profile will form either when vertical mixing is inhib-
ited by the formation of a stable layer at the surface, or when surface
destruction processes are particularly pronounced. As with Type B, Type
E profiles are not common during the afternoon; about two-thirds of the
32 profiles classified as Type E were measured before noon. However,
this may be somewhat misleading because very few measurements were made
at night. If more measurements were available, one might expect to find
many Type E profiles during the early evening when the ozone formed by
afternoon photochemical processes would be destroyed in the stable layer
forming at the surface. The elevated layer formed in this way is likely
to be found farther downwind in suburban and rural areas later at night
and during the early morning hours.
f . Type _F Profile
If a Type E profile persists through the night, it may evolve into
a Type F profile during the following morning. As the mixing layer
develops and becomes deeper during the morning, the ozone trapped in the
layer aloft during the night will be eroded and mixed downward into the
mixing layer. If the mixing never reaches the height that it did on the
preceding day, then the Type F profile will last through the afternoon.
However, this should be less common and one would expect to find Type F
profiles most frequently during the morning hours. Of the 54 cases clas-
sified as Type F, nearly 80 percent were found between 0600 and noon
LST.
3. Mixing Depth Changes During the Day
It is obvious from the preceding section that the type of profile
present and the transitions from one type of profile to another depend
heavily on changes in atmospheric stability and the depth of the mixing
layer- It is therefore important to know something of the typical diur-
nal cycle of the mixing depths, especially for the clear, warm, light-
wind conditions that usually accompany high ozone concentrations (Meyer
et al., 1976; Ludwig et al., 1977).
xv
-------
On clear nights outside the city, a stable layer is likely to
develop at the surface and extend upward for a few hundred meters.
Within the city, the formation of this surface-based stable layer is
inhibited by the heat released by power consumption, air conditioning,
space heating, and the heat stored by the thermal mass of all the city's
buildings. There is some empirical evidence suggesting a minimum mixing
depth over a city. This minimum mixing depth is dependent on city size,
but in general, it will be of the order of 100 meters or so (Ludwig and
Dabberdt, 1973).
During the afternoon, the heating of the lower layers of the atmo-
sphere will cause mixing to reach its greatest heights. In general,
these heights will be from a few hundred to a few thousand meters; they
can usually be estimated from temperature profiles available from
radiosonde stations in the morning and afternoon. Usually the mixing
depth during the afternoon will be lower on days with high ozone concen-
trations. The transitions between daytime and nighttime mixing depths
take place rapidly, especially the evening transition. Around sunset, a
stable layer forms very rapidly near the surface outside the city. The
corresponding transitions to Type B or Type E ozone profiles (from Type
A or Type D) will probably occur somewhat later, because destruction of
the ozone already within the stable layer takes time.
The morning transition from nighttime to daytime mixing conditions
takes place more slowly, because it takes a finite period of time for
the heating to overcome the effects of the night's cooling of surface
layers. The data suggest that this transition can take anywhere from 1
or 2 hours to 5 or 6 hours. If there is a strong subsidence inversion
above the surface but at a relatively low altitude, surface heating will
cause the mixing to reach the altitude of that inversion fairly quickly,
and then it will rise only slowly thereafter. Subsidence inversions are
frequently a feature of high-pressure areas, which are known to be
favored areas for the development of high ozone concentrations when suf-
ficient precursor sources exist.
The preceding discussion applies primarily to simple continental
geographic locations, but in more complicated locations, such as those.
near land-water interfaces, the mixing layer may behave in a more com-
plicated fashion. For instance, if air moves from a warmed land surface
to a colder water surface, the effect will be similar to that which
occurs at night. The air will be cooled from below and a stable layer
will form near the surface. Within this stable layer, the ozone may be
destroyed and there will be a transition from a Type D to a Type E pro-
file or from a Type A to a Type B profile, even in the daytime, when one
might not expect such transitions to occur. Sometimes synoptic meteoro-
logical conditions will cause convective mixing at night that will pro-
duce a transition from a Type B to a Type A profile, even though this
transition is much more frequently associated with the onset of daytime
heating.
xv i
-------
4. Evolution of Profile Types
This section discusses the typical evolution of profiles during the
day as air moves over an urban area. It should be remembered that there
are other, atypical ways in which the same or different transitions
might occur. Imagine a column of air upwind of a city during the early
morning hours. This column of air has traveled for a considerable
period of time without having been exposed to major sources of pollu-
tion. Ozone concentrations are near "background" values (a few tens of
ppb) and distributed uniformly with height, except near the surface.
Because it has been a clear night, a shallow stable layer has formed
near the surface, and much of the ozone within that layer has been
destroyed by contact with the surface. Thus the vertical distribution
of ozone within our column is Type B. As the column moves toward the
city and the sun rises, the ground begins to heat. As the heating con-
tinues, the stable layer is gradually destroyed and the ozone aloft is
mixed down, so that the vertical distribution of ozone becomes nearly
uniform with height, a Type A profile.
As the air enters the city, perhaps around 0900 LST, it encounters
the NO emissions from the city, and considerable ozone in the lower
layers is destroyed. Mixing above the city may not be sufficiently
vigorous to replace the destroyed ozone, so the profile can revert to
Type B. As the column of air leaves the city, the photochemical
processes act on the hydrocarbons and oxides of nitrogen that have been
introduced into the lower layers of the atmosphere, and ozone is pro-
duced. The mixing of precursors and ozone has not gone on long enough
to distribute ozone throughout the mixing layer, so a Type C profile
results.
As the column of air travels farther downwind of the city, mixing
continues, and soon the ozone and precursors are distributed uniformly
within the mixing layer. Thus, the Type C profile evolves into a Type D
profile. By now it is late afternoon or early evening and the surface
has begun to cool. A stable layer forms near the surface and much of the
ozone within that stable layer is destroyed, leading to the Type E pro-
file (see, e.g., Harrison et al., 1978). It was noted earlier that the
Type E profile could also be formed if the column of air passed over a
cold surface, such as a body of water.
The Type E profile formed in our column of air will probably per-
sist through the night, unless weather patterns develop that cause the
atmosphere to be unstable and nighttime mixing to occur. If this should
happen, the profile might evolve back into a Type D profile, and the
concentrations at the surface would increase anomalously at night.
Ludwig and Shelar (1978) have observed some nighttime instances of high
ozone concentrations when this anomalous transition from Type E to Type
D seems to have occurred. A more usual situation is for the Type E pro-
file to evolve into a Type F during the next day's morning heating
period (see, e.g., Johnson and Singh, 1977). If the heating and mixing
proceed to the point where the mixing depth is equal to, or greater
xvii
-------
than, what it was on the preceding day, the profile will change back to
a Type D once again. As time goes on and more and more ozone is lost,
the Type D profile will revert to a Type A profile unless more pollut-
ants are introduced and more ozone is formed photochemically-
If, as will be the case under the worst of conditions, the air
remains in a region of emissions or passes successively over one emis-
sions area after another, the concentrations within the mixing layer
will continue to build up. The Type D profiles will become more and more
pronounced and the concentrations within the mixed layer will become
greater and greater, while those aloft will change very little.
B. Applications of Findings
1. General
There have been two goals for the research summarized here. The
first was to use the existing data to describe and improve our under-
standing of the processes that affect the vertical distribution of pho-
tochemical pollutants, particularly ozone, in the atmosphere. The second
goal was to take that general description and use it as a basis for
analyzing specific cases. That is, some quidelines are sought by which
conventional surface pollutant measurements and routinely available
meteorological observations might be used to infer the concentrations
aloft, especially during periods when the national ambient air quality
standards (NAAQS) were violated.
Knowing the distribution of ozone around a city is fundamental to
developing strategies for reducing pollution. Even the simplest model-
ing approaches—e.g., the empirical kinetic modeling approach (EKMA) —
require knowledge of the concentrations in the mixing layer entering the
city and those in the mixing layer downwind of the city (Office of Air
Quality Planning and Standards, 1977, 1978). Certain relationships are
assumed between precursor emissions within the city and the concentra-
tions within the mixing layer downwind. One can then postulate changes
in emissions and use the EKMA or some other model to estimate how the
downwind ozone concentrations would be changed. More complicated models
require similar kinds of information, but in greater detail. The follow-
ing sections provide some guidelines on how the results of this research
might be applied to specific situations.
2. Assessment of Air Quality Conditions from Limited Surface Data
a. "Background" Conditions
The data have shown that when a Type A or D profile is present,
surface measurements of ozone will do a good job of characterizing
mixing-layer concentrations. The data also indicate that the Type A or D
profiles can reasonably be expected to accompany well-established
xviii
-------
vertical mixing. The question then becomes one of determining when mix-
ing has proceeded far enough to have established a relatively uniform
vertical pollutant profile. Several routinely measured parameters are
available that can be used as guides. One of these is the ozone concen-
tration itself. During the morning, one might expect to find a rela-
tively rapid rise in ozone concentration accompanying the onset of mix-
ing; this rapid rise in the morning would be uncharacteristic of photo-
chemical production because of the hour of the day. Temperature might
also rise rather rapidly during the early morning because the effects of
solar heating would be confined to the shallow layer within the
surface-based nocturnal inversion, at least until that inversion was
erased by the heating. Finally, wind speed can be used as a guide to
determining when mixing has taken place in the morning. Momentum is con-
tinually being transferred downward to replace that lost at the surface
by friction. When this vertical transfer is inhibited, there are likely
to be near-calm winds, but when the mixing is reestablished the wind
speed should rise rather pronouncedly.
Routine data can be used to determine when measurements at surface
locations outside the city are representative of concentrations through
the mixing layer. If the location is upwind of the city, these measure-
ments will certainly be representative of the "background" conditions.
Even if the available measurements are downwind, they might still be
useful for estimating the "background" conditions if vertical mixing
occurs sufficiently early in the day that little photochemical activity
would have taken place, say before about 1000 LST.
The data have shown that after about noon, it can usually be safely
assumed that good vertical mixing has taken place. However, it cannot
always be assumed that the concentrations entering the city during the
afternoon are representative of the "background" conditions that pre-
vailed early in the day. Some of the data from this study suggest that
new photochemical ozone might form in "background" air, even when no new
precursors have been added. Presumably, this new ozone forms from resid-
ual precursors.
b. Concentrations Above and Downwind of the City
It is very important that any ozone-monitoring site within a city
be well removed from local sources of NO, but this does not seem to be
sufficient to ensure that the data from such a station can be used to
describe conditions throughout the mixing layer above the city. In gen-
eral, the stations of the Regional Air Pollution Study (RAPS) are well
located, but nevertheless give readings that were substantially less
than the concentrations in the mixing layer, even in cases where Type A
profiles were well established in the areas outside the city. However,
the St. Louis data do show that ozone concentrations above the surface
effects were fairly uniform throughout the region during the morning
hours. Thus, measurements made outside the city can be used to estimate
concentrations over the city—above the destructive effects of NO
xix
-------
emissions. During the afternoon, in lieu of better information, one
might assume that the upwind conditions prevailed to about the center of
the city; interpolation could then be used to estimate conditions from
the center of the city to a downwind point where data are available.
The interpolation between the center of the city and a downwind
location could be aided by the use of a simple Gaussian model. For
example, Turner's (1970) workbook shows how the depth to which emissions
will mix will increase with travel distance, for different classes of
atmospheric stability. The atmospheric stability classes can be deter-
mined from conventional meteorological measurements (Turner, 1964;
Ludwig and Dabberdt, 1976). These considerations could be used to esti-
mate the transition of the vertical profile from the Type C to the Type
D case.
3. Modeling
It is apparent from the discussion in the preceding section that
the distribution of ozone around a city can be estimated reasonably
well, if sufficient surface pollutant and meteorological data are avail-
able. Complex models require estimates of morning conditions throughout
the area as inputs. The model's outputs describe the three-dimensional
ozone distribution for later in the day. These outputs must be compared
with the estimates obtained from the available surface-ozone observa-
tions. It might well be that good mesoscale meteorological models could
be used to provide better estimates of the mixing and transport, and to
improve the estimates of pollutant distribution around the city. Obvi-
ously, it would be better if comprehensive airborne measurements were
available as inputs for the model and the required validation data, but
this would be prohibitively expensive for most routine applications.
Certainly, special programs such as the RAPS are warranted for model
development and validation, but it appears that reasonable estimates of
the necessary inputs can be obtained from more routine measurement pro-
grams, especially if the ozone monitoring sites are placed away from
local influences in areas where their data can be used to characterize
conditions through the mixed layer.
4. Selection ojf Monitoring Sites
The analysis of the data makes it clear that ozone monitoring sites
within a heavily populated area are not apt to be as useful as those
outside the area. It is not the purpose of this research to develop
comprehensive guidelines for siting monitoring stations, but it is worth
pointing out that three or four locations outside a city would provide
very valuable information for most practical applications. The locations
of those sites relative to the city would depend on wind directions and
numerous other considerations (see, for example, Ludwig and Shelar
1978). Sites outside the cities would provide estimates of the incoming
"background" conditions and the ozone formed from the precursors emitted
xx
-------
within the cities. The precursor concentrations should probably be
characterized by measurements within the urban area.
C. Remaining Shortcomings
As is always the case, the data base has some shortcomings, and the
analysis could not include everything that one might hope for. However,
the data base is much more comprehensive than was expected when the work
began, its most important shortcoming being the relative absence of
nighttime cases. With more nighttime cases available, the characteris-
tics of Type B and E profiles could be better defined.
Some kinds of data that are available were not analyzed during this
project. For example, vertical profiles of ozone in rural areas were not
studied, although at least some data are available from such areas. The
study was also limited to the lower parts of the atmosphere, leaving out
studies of stratospheric intrusions or other interchanges with the stra-
tosphere. Perhaps, at some future date, it would be worthwhile to
develop a more comprehensive characterization of the three-dimensional
distribution of ozone by including data from higher altitudes and from
less urban areas.
Finally, time, money, and contractual limitations have kept us from
exploring the possibility of using the temporal histories of pollutant
concentration and meteorological variables to determine objectively the
likely vertical profiles for an area. At present, the general subjec-
tive guidelines given earlier are available, but it might be worthwhile
to examine the data more carefully in the future to see if objective
techniques can be devised. Even without such objective techniques, the
analyses presented here provide the information required for estimating
conditions aloft from surface observations.
xxi
-------
I INTRODUCTION
A. Purposes of the Study
The meteorological and chemical processes involved in the produc-
tion of atmospheric oxidants are all three-dimensional; consequently,
the development of effective control strategies for ozone requires an
understanding of the three-dimensional distributions of photochemical
pollutants—principally oxides of nitrogen, nonmethane hydrocarbon, and,
most important, ozone. Meteorological conditions play an important role
in the production, destruction, and transport of photochemical pollut-
ants. Therefore, it is also important to know something of the three-
dimensional behavior of the atmosphere itself. Most routine measure-
ments of air pollution are confined to the surface, leaving a void in
our knowledge of the vertical distributions. The purpose of the
research described in this report has been to fill that void by making
use of the considerable amount of three-dimensional data (from aircraft
and sounding devices) that have been collected in recent years.
The analyses of the data have been directed toward finding ways of
estimating vertical distributions of the photochemical pollutants in the
vicinity of urban centers, when no special measurements have been made
at higher altitudes. Toward this end, we have sought to characterize the
vertical profiles and relate them to more commonly available kinds of
information, such as time of day, conventional meteorological observa-
tions, and location relative to the city itself. Ultimately, the results
are intended to be useful for formulating oxidant control strategies;
they will provide the basis for estimating the effects of the uncon-
trollable levels of oxidant and its precursors that are apt to enter a
city from the upwind direction, and also for estimating total impacts by
considering ozone concentrations at higher altitudes that can be
injected into the mixing layer when that layer deepens. Thus we are not
limited to consideration only of what is observed at the surface.
B. Sources of Data
1. Background
No measurements were made as part of this project. All the analyses
that follow are based exclusively on data that were collected either
routinely or during the course of special projects. The work began with
a survey to find data that might be appropriate to the purposes of this
project. It was originally planned that five cities be selected for
intensive study. However, during the course of the project, it was
decided that the extensive study of four cities and somewhat less
detailed analysis of data from other data sources would be preferable.
The following cities were considered: St. Louis, Missouri; Houston,
Texas; Washington, D.C.; Los Angeles, California; Toronto, Canada;
Indianapolis, Indiana; the urbanized area of New York, New Jersey, and
-------
southern New England; Tampa, Florida; Dayton/Columbus, Ohio; the San
Francisco Bay Area; Fresno and Bakersfield, California; and Tulsa,
Oklahoma.
The first four cities on the list were chosen for detailed
analysis. Although the choice of the cities was subjective, it was
based on fairly well-defined criteria. The criteria were
e Quality of the data; we have tried to avoid unreliable data that
could be misleading.
• Quantity of the data; in choosing among areas that were gen-
erally equal in most other respects, we opted for the area with
the larger data set.
• Form of the data; data in computer-compatible, or already
displayed graphically, were judged to be more desirable because
analysis was easier and less time had to be spent on tedious
data-reduction processes.
• The magnitude of photochemical air-quality problems; this was
judged to be an important factor because of the practical con-
siderations underlying the project.
• Geographical diversity; this was sought because the potential
for practical application of the results dictated that a wide
variety of climatological and geographical conditions be stud-
ied, so that the results would be as general as possible.
• Availability of data; some data sets were not considered because
they had been collected for proprietary reasons; other data sets
were not available in time for use.
The areas selected for study and the reasons for their selection are
discussed in the next section. The amounts and kinds of data that were
available for analysis will be discussed in more detail later.
2. Primary Areas Selected for Study
a. St. Louis
The Regional Air Pollution Study (RAPS) in St. Louis provided one
of the most comprehensive collections of air-quality data available any-
where. Furthermore, all the hourly surface data from 24 stations in the
St. Louis area were completely reduced and available on magnetic tape
for the year 1976. In many cases, these same stations also provided
meteorological information such as wind speed and direction, tempera-
ture, humidity, and so forth. Furthermore, special radiosonde and pilot
balloon measurements were available from four locations in the St. Louis
area for the months of July and August 1976. Laser radar (lidar) and
acoustic sounder records were also on hand to characterize the behavior
of the mixing layer for part of the period. Concentrations of ozone in
the area are known to exceed the National Ambient Air Quality Standards
-------
(NAAQS), which further supports the choice of St. Louis as a study area.
Finally, St. Louis is representative of midwestern cities and certainly
there are more data available from St. Louis than from any other
midwestern location.
b. Los Angeles
Los Angeles is the classic city for photochemical pollution and
probably deserves to be included in any list of cities to be studied for
that reason, if for no other- Its warm climate, its subsidence inver-
sions, and the California-type emissions all require consideration, if
the study is to be general. In 1973, the Coordinating Research Council
sponsored an extensive program of helicopter measurements of air quality
in the Los Angeles Basin. Several other agencies, including the U.S.
Environmental Protection Agency (EPA) and the California Air Resources
Board (GARB), also participated in the Los Angeles Reactive Pollutant
Program (LARPP). The computer-compatible data from this program (Parker
and Martinez, 1975) were available and Johnson and Singh (1977) had
identified and plotted most of the usable vertical profiles from this
data set, further increasing its usefulness. Other data collected by
Blumenthal et al. (1974) were also available.
c. Houston
Houston has some unique characteristics that make it a candidate
for study. It has a particularly high density of petrochemical indus-
tries with concomitant hydrocarbon emissions. It lies along the Gulf
coast, with warm, humid summers. Photochemical air pollution is known to
be a problem in the Houston area, so the area has the geographical loca-
tion and the air quality problems that make it worth including in the
list of cities to be studied.
In recent years, at least two fairly extensive aircraft measurement
programs have been undertaken in the Houston area. One of these (Shauck
and Alexander, 1978) was sponsored by the Houston Chamber of Commerce as
part of their Houston Area Oxidant Study (HAOS). Another extensive air-
craft monitoring project was conducted by Washington State University
(Westberg et al., 1977) under the sponsorship of the EPA. The data from
the EPA-sponsored study has proven to be more useful for two reasons.
While computer-compatible data were not available from either study,
graphical representations of the vertical ozone and temperature profiles
were available from the report of Westberg et al. (1977). Furthermore,
the vertical profiles from the Westberg study generally seemed to pro-
vide more information about upwind and downwind conditions than did the
HAOS study. Also, the vertical profiles generally began nearer the
ground and had better vertical resolution in the Westberg data set than
in the HAOS data. The HAOS project had assembled most of the relevant
surface meteorological and air-quality data (see Ludwig et al., 1978)
-------
necessary to supplement the Westberg data, and they were in computer-
rnnrna H Vil e> fnvm J
compatible form.
d. Washington. D_.£.
Once the West Coast, Midwest, and Gulf Coast areas were
represented, the one remaining area with major photochemical air-quality
problems was the east coast. Certainly the coastal corridor from
Philadelphia to Boston is a logical place to study, especially in view
of the fact that many aircraft measurements are available from the area
(e.g., Spicer, et al., 1976; Siple et al., 1976; Washington State
University, 1976; Wolff et al., 1975) and are in computer-compatible
form (Ruff et al., 1977). However, these data had already been exten-
sively analyzed, especially by Ludwig and Shelar (1977), so we decided
to direct our efforts to a different area, while still making use of the
analyses already available from southern New England.
The only other east coast area for which aircraft data are known to
be available is the region surrounding Washington, B.C. and Baltimore,
Maryland. Good aircraft data, in computer-compatible form, were col-
lected by the EPA (Fitzsimmons, et al., 1978) during the latter half of
August, 1976. Although the surface air quality and meteorological mea-
surements were not as extensive as could be desired, we nevertheless
chose Washington, B.C. because the analysis of the available data would
add another city with somewhat different characteristics to the existing
inventory of information.
e. Other Areas
As noted earlier, we decided to supplement detailed analyses of
data from the four cities named above with studies conducted in other
areas. In this regard, the aforementioned analyses of New England data
by Ludwig and Shelar (1977) have proven to be quite useful, as have
those of Lovelace, et al. (1975) from the Indianapolis area and the work
of Wiebe et al. (1975) in the Toronto area. Some obvious gaps remain in
the geograhical coverage, particularly the Pacific Northwest and the
Rocky Mountain areas. Photochemical pollution problems in the Northwest
are not generally as severe as in other parts of the country. However,
the mountain states and areas such as Denver should be considered for
more intensive analysis as more data become available.
C. General Approach
After the major data sources had been identified and the cities
selected for intensive study, we acquired the necessary data and
prepared them for analysis. In general, this report is devoted to the
results of the analysis from that point on. The conduct of the analysis
can be divided into four parts, as follows:
-------
• The classification of the observed vertical profiles of the pho-
tochemical pollutants, especially ozone, according to the shape
of those profiles.
• The consideration of those factors that govern the distribution
of ozone and other photochemical pollutants in the atmosphere.
• The determination of how the different types of vertical air
pollutant profiles relate to those factors that govern the dis-
tribution of ozone and other photochemical pollutants in the
atmosphere.
• The development of rules and methods for determining just what
kind of vertical profile might be expected, based on conven-
tional meteorological and surface-based air-quality information.
This report is organized along the lines just described. The next
section presents a discussion of those factors that govern the vertical
distribution of pollutants in the vicinity of urban areas, the kinds of
vertical profiles that have been observed, and how they are related to
those factors. Although the discussion proceeds from the physical
processes to the profile shapes that result from those processes, the
actual research took a somewhat different course. The major categories
of profile shape were recognized from the available data and from a
knowledge of the processes. The classification system that is described
here evolved from the simultaneous consideration of theory and observa-
tion. The third section of the report presents examples from the cities
selected for intensive study and from other areas, to illustrate the
different types of profiles and when they tend to occur in relationship
to the cities, times of day, and meteorological conditions. The report
includes a summary of the findings and a discussion of how those find-
ings can be applied to the real problem of estimating the three-
dimensional distribution of ozone and the other photochemical pollutants
in the vicinity of cities, when only surface measurements and conven-
tional meteorological data are available.
-------
II PHYSICAL PROCESSES GOVERNING PHOTOCHEMICAL POLLUTANT DISTRIBUTIONS
IN THE VERTICAL AND CONSEQUENT PROFILES
A. Factors Governing the Vertical Distributions
of Ozone in the Vicinity £f_ Urban Areas
1. Background
The processes governing the distribution of ozone in the atmosphere
are reasonably well understood qualitatively, if not quantitatively.
The sources of ozone in regions well removed from urban areas are sub-
ject to some controversy, especially with regard to their relative mag-
nitude. However, regardless of the source, it seems reasonably well
established (e.g., Singh et al., 1978) that the yearly mean concentra-
tion of background ozone in the troposphere at midlatitude is a few tens
of parts per billion.
It is worth examining how this natural ozone behaves, because it
provides a convenient introduction into the processes governing the
vertical distribution of ozone in the atmosphere. It is also quite easy
to mistake the natural ozone behavior for similar behavior observed for
ozone of anthropogenic origin. Singh, et al., (1977) have provided an
idealized depiction of the behavior of natural ozone in remote loca-
tions. Figure 1 is a schematic representation based on a figure from
their report. It shows that there is a large reservoir of ozone aloft
and that this reservoir is relatively unaffected by daily short-term
changes. The diurnal variation in ozone concentration above the mixed
layer is rather small, but, as shown on the right-hand side of the fig-
ure, there can be considerable variability in the concentration at the
surface during the day. This diurnal variation at the surface results
from the ozone that is within or below the nocturnal inversion being
destroyed and not replaced by ozone from aloft; the presence of the noc-
turnal inversion inhibits mixing. However, after the sun rises and the
ground is warmed, there will be convective mixing, which will bring
ozone from the reservoir aloft down to ground level, to replace that
which is destroyed. This process causes ground-level concentrations to
increase during the times of day when there is good mixing. As will be
discussed later, the sunshine that warms the ground and causes mixing
can also lead to the photochemical production of ozone, and to similar
diurnal changes in ozone concentration at the surface.
Singh, et al. (1979), have also shown schematically typical annual
variations in the natural tropospheric ozone burden. These are shown in
Figure 2. At remote sites that are unaffected by anthropogenic emis-
sions, the ozone concentrations tend to reach their maximum in the early
spring. In general, the natural ozone falls somewhere in the shaded
area marked A in the figure. Natural concentrations are at a minimum in
the late fall or early winter. It is believed that the autumnal decline
of ozone concentrations in these remote locations is caused by decreases
in the injection of stratospheric ozone into the troposphere.
-------
00
— 225
FREE TROPOSPHERE
O3 DIURNAL PROFILE
TOP OF THE AFTERNOON MIXED LAYER
FREE TROPOSPHERE
AFTERNOON
03
POLLUTION
AFTERNOON
°3 WITHOUT
POLLUTION
tso
a
a.
o"
OZONE (ppb)
5 10 15 20
LOCAL TIME (hrs)
FIGURE 1 SCHEMATIC REPRESENTATION OF OZONE VARIATIONS AT THE SURFACE AND IN THE FREE TROPOSPHERE
-------
120
100
FROM URBAN CENTERS
LOCAL OZONE
SYNTHESIS
^^&<
-------
2. Formation, Destruction, Transport, and Mixing
of_ Ozone in the Vicinity of^ Cities
Before proceeding, it should be noted that profiles very similar to
those shown in Figure 1 for unpolluted atmospheres can also be found in
polluted atmospheres, if mixing has been sufficient to cause the pollut-
ants to be distributed nearly uniformly in the vertical. Thus, we might
expect ozone profiles in air entering any isolated city to be shaped
somewhat like that shown in Figure 1 for unpolluted air masses.
Figure 3 is a schematic illustration of what one might expect to
observe in the plume of emissions from a city. The top part of the fig-
ure shows the plume of hydrocarbons and oxides of nitrogen emitted dur-
ing the morning rush hour- These emissions are reasonably well mixed
HOUR OF
THE DAY
0900
NO,
HC '
fen
NO
KO» ^C HC
NO HO NO
nr **C HC
NO Hf. NO,
NO HC
NO HC HC HC N0. °3
N0« °3 n NO,
HC NO o °3 HC "
HC HC °« * HC
3 HC
1300
HC
NO,
, NO, HC HC
NO
HC NO,
•^jj^o^o^!^.
1800
2300
FIGURE 3 SCHEMATIC DIAGRAM OF THE TRANSPORT OF PRECURSORS, OZONE
PRODUCTION, AND OZONE DESTRUCTION DOWNWIND OF A CITY
10
-------
and have traveled some distance downwind of the city. At greater
downwind distances, their densities are less because they have been
diluted by mixing and also because they were emitted during the evening,
when emission rates were much lower. There is some ozone scattered
among the oxides of nitrogen and the hydrocarbons; at this morning hour,
the ozone at the higher elevations is unlikely to have been produced by
photochemical activity and probably reflects the "background" concentra-
tions that have been advected into the city. The second panel of the
figure shows the situation at about midday. The plume is dense with
oxides of nitrogen and hydrocarbons for some distance downwind. Of con-
siderable significance is the fact that much ozone has been produced by
photochemical processes during the day. This ozone is distributed
nearly uniformly in the vertical because of the strong daytime mixing.
In general, the concentrations increase downwind of the city for some
distance, a result of the time needed by the photochemical processes to
produce the ozone. Eventually the production of ozone by the photochem-
ical processes is offset by dilution, so concentrations fall off at
greater distances from the city.
The third panel of Figure 3 illustrates the situation during late
afternoon or early evening, when the oxides of nitrogen and the hydro-
carbons have been uniformly mixed in the vertical. At this time, rela-
tively dense concentrations occur for considerable distances downwind of
the city. However, it should be noted that there are few ozone symbols
scattered among the emissions that occurred later in the day (and are
still close to the city), because photochemical processes are rather
ineffective for producing ozone in the late afternoon. Another impor-
tant characteristic of this diagram is the fact that ozone is confined
to the upper parts of the plume at the greater downwind distances. This
is caused by the formation of a stable layer near ground level in the
late afternoon or early evening; the stable layer prevents mixing of
ozone through the entire depth of the plume; the ozone that was in the
lowest layers is destroyed at the ground and is not replaced by mixing
from above.
The final panel of Figure 3 is an extension of the situation shown
in the preceding panel- Concentrations of hydrocarbons and oxides of
nitrogen are smaller near the city than they were earlier in the day
because the emission rates have declined during the night. Ozone is not
produced after sunset. Only the "background" ozone and that produced
earlier in the day and advected far from the city are present. This
accounts for the fact that very little ozone is found near the city.
Farther downwind, in that part of the plume released earlier in the day
when photochemical activity could take place, there is considerable
ozone, but it is confined to the upper parts of the plume. Ozone in the
lower parts of the plume has been destroyed at the surface.
The discussion to this point has shown that there are a limited
number of processes contributing to the development of vertical ozone
profiles. The relative importance of the processes varies with time of
day, with meteorological conditions such as atmospheric stability, and
11
-------
with distance from the city. The three most important processes are
destruction of ozone, production of ozone, and vertical mixing.
Destruction of ozone occurs mainly near the surface. If there is
no nitric oxide (NO) present, destruction at the surface will occur
fairly rapidly. The rapidity probably depends on the amount of surface
area available for contact with the air, or as suggested by Vukovich
(1973), the presence of natural agents that promote gas phase destruc-
tion of ozone. If nitric oxide emissions are present, as they generally
will be from traffic and high-temperature combustion sources within the
city, then the destruction of ozone will be very rapid. The presence of
NO in plumes from large stationary sources such as power plants can also
lead to the rapid destruction of ozone aloft, but in general the
destruction of ozone in the free atmosphere away from the surface
proceeds rather slowly.
The processes that operate to destroy ozone are active at all hours
of the day and night, but the processes that operate to produce ozone
from hydrocarbons (HC) and oxides of nitrogen (NOX) are important only
during the daytime. Another very important difference between the pro-
ductive and destructive processes for ozone is the rate at which they
occur. The important destructive processes occurring at the surface are
quite rapid, while the productive processes tend to be slow by compari-
son For this reason, the destructive effects associated with a city's NO
emissions will be observed over the city and in its immediate environs,
while the ozone production that will result from those same emissions
will take place some distance downwind. Ludwig and Shelar (1978) have
estimated that the maximum concentrations of locally produced ozone will
be found downwind of a large city at a distance that corresponds to a
travel time of five to seven hours from the upwind edge of the urban
area. Although the maximum ozone concentrations generally occur within a
few tens of kilometers of the city, the ozone produced by emissions from
a city can be detectable much farther downwind. Discernible ozone
plumes generated by emissions from East Coast cities have been observed
well over 100 km downwind of those cities (e.g., Cleveland, et al«,
1976; Zeller, et al., 1977; Spicer et al., 1977).
Vertical mixing is the third factor that influences the nature of
the vertical distributions of pollutants, especially ozone, in the
vicinity of cities. Mixing causes the pollutants to be transported ver-
tically to and from the sinks and sources. For the oxides of nitrogen
and the hydrocarbons, there is a source at the surface. For ozone, the
surface is a sink; the source of ozone tends to be rather diffuse and
coincides with the volume occupied by the oxides of nitrogen and the
hydrocarbons. Obviously, if mixing is weak, the effect of the source or
sink will not extend very far into the overlying atmosphere, and the
gradients in the immediate vicinity of that source or sink are apt to be
quite large. If the mixing is very vigorous, then the effect of the
source or sink will extend to much higher levels, and the gradients are
apt to be much more uniform.
12
-------
There is one other important process to be considered in analyzing
vertical profiles in the vicinity of cities—the horizontal transport
process. Transport, of course, determines where the air entering the
city has come from, and, indirectly, the pollutant burden that it is
carrying with it as it enters the city. It also determines where the
air goes after it leaves the city. The rate at which the air is
moving—i.e., the resultant wind speed—determines the volume of air
into which the emissions are mixed and hence their concentration. The
wind speed also determines how far the air travels downwind of the city
during the time the photochemical reactions producing ozone are opera-
tive. There are also relationships between wind speed and vertical mix-
ing; rapid air movement over a rough surface can induce mechanical tur-
bulence. High^wind speeds near the surface are also an indication that
momentum is being transferred downward by vertical mixing.
B. Types of Vertical Ozone Profiles
The processes discussed in the preceding section can interact in
different ways to produce a variety of vertical ozone profiles. The
relative strength and vertical extent of mixing, the relative importance
of the sources and sinks, and the time over which all the processes
operate will contribute to the development of the vertical pollutant
profile. It was anticipated that a taxonomy of vertical pollutant pro-
files might be developed and that it might be possible to relate the
different types of profile within that taxonomy to different kinds of
interrelationships among the processes that produce the profiles. With
this in mind, the vertical profiles from the four major study areas were
examined to see if a reasonable classification system could be
developed.
Figure 4 shows the classification system that evolved from this
effort. Six basic types of ozone profile were commonly observed. To
some extent, these categories have evolved from earlier work presented
by Johnson and Singh (1977). However, their work focused on profiles of
the type found in the last two categories of Figure 4. The classifica-
tion system discussed below is more general than their work. (It should
be noted that the height and concentration scales that are used in Fig-
ure 4 are for illustrative purposes only. The values used are typical,
but observed conditions can differ widely.) Types A and B are closely
related to the "background" conditions shown earlier in Figure 1. Type
A gives little evidence of any nearby anthropogenic effects. Such a
profile arises when mixing is very thorough. The small gradients near
the surface are an indication that either the mixing is vigorous or the
surface sink is weak. The uniformity of the profile in the upper level
is an indication that sufficient time has elapsed since any ozone form-
ing precursors were introduced so that the precursors and ozone have
been thoroughly mixed through a very deep layer.
13
-------
E
.*
HEIGHT -
j
2
1
0
C
3
2
1
0
A
-
I I/ 1 1
) 40 80 C
B
^k^Xl I I
C
lllllsffllllis:
r! MIXED LAYER*::::::::
) 40 80 120 C
1 D
1
llil'ml
XY MIXED LAYER •'••{•/••••:••'•••
mmMmmmkm \
i
i
i
i
i
—
40
I
1
1
::x:: MIXED:::
mmm
E
\
1
STABLE LAYER
80 120 160
""""• — -x
1
Y/*: Y, LAYER -X-X-
mmmmm
40 80
40
80 120
40
80 120 160
OZONE CONCENTRATION - ppb
FIGURE 4 SIX TYPES OF VERTICAL PROFILE
-------
The Type B profile in Figure 4 is related to the Type A profile in
one of two ways. Either the magnitude of the sink at the surface is
greater than for the Type A profile, or the vertical mixing, especially
in the lower layers, is weaker. Weaker vertical mixing will generally
result from formation of a stable layer near the surface. A stable
layer will form when the air is cooled from below. Typically, this will
occur by radiative cooling in the late afternoon and evening. It can
also occur when air is advected over a cooler surface such as water
(see, for example, Lyons and Cole, 1976).
The most common way in which a surface ozone sink can be enhanced
is by the introduction of NO. Thus, a vertical ozone profile of Type B
can evolve from a Type A when lower layers of the atmosphere become very
stable or when the air passes over a source of NO, such as a city, or by
some combination of these two causes.
The remaining four types of profiles show the effects of the intro-
duction of anthropogenic pollutants, and hence they might be considered
to be "urban" or "downwind" types. The major differences between the
conditions that would cause a Type C profile and those that would cause
a Type D are the elapsed time since the original emissions were released
and the vigor of the mixing during the intervening time. For a profile
of Type C, mixing has not proceeded to the point where the pollutants
are uniformly distributed through the mixing layer, as they are for Type
D. In both cases, the slight increase with height just above the sur-
face shows the effect of the surface sink for ozone. To a large extent,
the sink is offset by the vertical mixing, but not completely. If pro-
files were shown for pollutants that are not destroyed at the surface,
there would not be the gradient in the lowest layers. For primary pol-
lutants, emitted directly into the atmosphere, Type C or D profiles
could occur at almost any time of day, but for ozone they are most
likely to occur in the afternoon. This is because sunshine is required
to produce the ozone, and mixing (caused by warming of the surface) is
required to offset the destruction at the surface.
One can imagine a profile of Type D evolving to one of Type E in
exactly the same fashion as a profile of Type A evolves to Type B. In
one case, a stable layer develops near the surface, as shown in Figure
4, so that surface destruction or gas phase destruction by natural
agents (Vukovich 1973) are not offset by vertical mixing. The second
way that the types E or B profiles evolve is when the ozone-laden air
passes over an area where the surface destruction of ozone is very
strong, such as a city with heavy NO emissions. Thus, a Type E profile
could occur in at least four different situations:
• Downwind of a city at night, when a stable layer has formed by
cooling of the surface, especially if natural destructive agents
are present.
• In air that has passed over a cool surface, such as the sea,
after it has left the city.
15
-------
• In air that has left one city but is passing over another with
strong NO emission.
• In air that is recirculating over the city of origin or continu-
ing to pass over that city, if the city is very large.
Type F profiles can result from several different processes. The
most common requires several steps. Johnson and Singh (1977) have
described a sequence of events that leads to a profile of Type F. Fig-
ure 5, from their report, shows the situation in which air moves from an
urban area to a rural area. In the first panel, there is a Type D pro-
file that has resulted from photochemical reactions involving precursors
distributed throughout the mixing layer. At night, this profile evolves
into the Type E profile shown in the center panel, as the ozone in the
stable layer near the surface is destroyed. The final section of Figure
5 shows conditions on the following day; in this example, the mixing
depth is not as great as it was on the preceding day. Because the mix-
ing depth is less on the second day, only the lower levels of the ozone
layer aloft are eroded and mixed downward. This leaves intact the
greater ozone concentrations at the higher altitudes.
HEIGH1
DAY 1
V 1 y
| -Pr
1
x,
**~""\
\
03
/
NOXRHC /
T>ft /
1
i
,
MIX
DEF
ING
>TH
NIGHT
1
1
1 -^
1 ^
V
"""*v
— __j
))
I)
,
1
°3
""TMIXING
f DEPTH
DAY 2
! -$-
I^>C^^
1 <•
v
^M
LEFT-OVER N-
("OLD") 03^
)
*— — ^_^-.
^-~?
FAVORABLE
PRODUCTION
ZONE?
-fT~
MIXING
DEPTH
J \
12 18 00 06 12 18
|
20
LOCAL TIME
1
1
1 1
0 100 200 300 400 500
|
600
MILES
URBAN
RURAL
SA-6321-15
FIGURE 5 TYPICAL DIURNAL EVOLUTION OF THE VERTICAL OZONE PROFILE
AS AIR MOVES FROM URBAN TO RURAL AREAS
16
-------
Johnson and Singh (1977) point out that reduction in the depth of
the mixed layer from one day to the next can occur for a number of rea-
sons. For example:
• Reduced insolation (from increased cloudiness), causing less
pronounced convection.
• Increased surface moisture that reduces transfer of sensible
heat and also leads to weaker convection.
• Passage of the air from a warmer to a colder surface.
• Large scale subsidence.
Changes in the mixing depth are not unusual from day to day. If the
depth of the mixing layer decreases from one day to the next, it can
leave a relatively concentrated layer of ozone isolated aloft and result
in a profile of Type F. If the mixing depth increases again on subse-
quent days, that layer will tend to be mixed more uniformly and the
resulting profile will evolve toward Type A, completing the cycle.
As noted before, the Type F profile can result from processes other
than that just described. For example, Edinger (1973) suggested that
elevated layers of pollution are formed in the Los Angeles Basin when
polluted air near the ground moves up heated mountain slopes at the
northern part of the Los Angeles Basin, and then moves out from the
mountains horizontally within the inversion that commonly caps the
marine air in Los Angeles. The presence of polluted layers aloft in the
Los Angeles area has been observed frequently (e.g., Blumenthal et al.,
1974; Johnson and Ruff, 1975). This report does not give much attention
to this particular mechanism for the formation of profiles of Type F,
since it appears to have rather unique geographical requirements. It is
probably necessary to have both mountainous terrain, to deflect the air-
flow, and a very strong inversion aloft. Los Angeles is among the few
places that have the combination of a frequently occurring subsidence
inversion (in association with the Pacific anticyclone) and the
requisite terrain.
There is yet another way that profiles of Type F can be formed.
This is through the introduction of photochemical precursors into a
stable layer aloft. Blumenthal et al. (1974) have also observed this
phenomenon in the Los Angeles area. While this mechanism produces Type
F profiles for a primary pollutant, such as NO and HC, it results ini-
tially in a very different profile for ozone. The high concentrations
of NO in a plume will severely reduce the ozone concentrations within
the plume. This will produce a vertical profile that is a complement of
Type F profile. There will be a layer aloft with less ozone concentra-
tion. However, if there is sufficient sunshine, the increased concen-
trations of NO and HC will result in corrrespondingly increased concen-
trations of ozone after some time has elapsed. Although this mechanism
can occur in almost any geographical area, its consequences are likely
to be of rather limited extent. If the plume is injected into a layer
sufficiently stable that it will maintain its identity for a prolonged
17
-------
period of time, that same stability will tend to limit the horizontal
extent of the plume. For this reason, we have chosen not to deal with
this phenomenon in any great detail.
The system described above for classifying vertical pollutant pro-
files provides a basis for examining observations in the vicinity of
cities. Of course, the ultimate objective is to use this classification
scheme for extrapolating surface observations of ozone and other pollut-
ants to higher elevations. Profiles of Types A and D certainly should
allow this to be done. The question becomes, how does one recognize,
from ground based pollution measurements and conventional meteorological
observations, just when such profiles are likely to exist? The observed
profiles from several cities will be examined later in an attempt to
answer this question.
C. Other Considerations Concerning Ozone Profiles
The classification system for vertical profiles (discussed in the
preceding section) is only one of the important factors that need to be
considered. Other important factors include
• Time of day
• Depth of the mixing layer
• Atmospheric stability
• Temperature and cloud cover
• Winds
• Location relative to the city.
Time of day is important because it provides a first estimate of
mixing conditions and of the strength of photochemical processes that
might be producing ozone. Photochemical processes occur only in the
daytime. They require sunshine and have also been found to depend on
temperature; this is one of the reasons some consideration should be
given to temperature and cloud cover-
The mixing depth and the presence of stable layers will generally
determine at what altitudes the major transition points of the different
types of pollutant profile will occur- Obviously, mixing depth will
play an important role in any attempt to define vertical profiles of
pollutants. Atmospheric stability will determine the vigor of the mix-
ing process and influence the nature of the profile. Atmospheric sta-
bility is related to cloud cover, wind speed, and time of day, a fact
that contributes to the inclusion of those items on the list of impor-
tant factors.
18
-------
Winds, of course, determine where the pollutants are carried
(direction) and the volume of air into which they are mixed (speed)—
hence, how much they will be diluted. It is essential to know the wind
direction in order to define whether a location relative to the city is
upwind, downwind, or to one side. It should be evident from the discus-
sion in the preceding section that location relative to the city is an
important determinant of vertical ozone profile type.
19
-------
Ill OBSERVED VERTICAL PROFILES
A. General
An attempt has been made in the following sections to group the
examples from each of the primary study areas along the lines discussed
in the preceding sections. The first point that should be made is that
the classifications and the groupings are often based on subjective
judgments. Sometimes the vertical extent of the measurements was not
sufficient to determine the mixing depth or to adequately classify the
character of the vertical profile. Not all hours of the day are
represented. In spite of the shortcomings, there is an encouraging con-
sistency and orderliness to the results.
Each of the following sections includes a tabulation of the ver-
tical profiles considered. The majority of those profiles for the pri-
mary study areas are displayed in this report. An attempt has been made
to bring together as much information as possible in the graphical
displays. In general, the displays include the vertical profiles and
maps of the distribution of ozone at ground level. An indication of the
wind direction is included to define the location relative to the city
at which the sounding was made. The primary emphasis has been on ozone,
and that pollutant is always included in the displays. The Los Angeles,
St. Louis, and Washington data generally include nephelometer readings
(Bscat). Vertical profiles of temperature usually accompany the ozone
profiles. Frequently, dew point temperature is also available. Unfor-
tunately, reliable NOX and hydrocarbon measurements were not generally
available.
Finally, it should be noted that the formats of the graphical
presentations differ somewhat from one section to the next. There are
two reasons for this. First, the available data differed from city to
city and so the displays must reflect that fact. Second, when data were
not available in computer-compatible form, but had been graphed, we made
use of the preexisting graphical displays. The text calls attention to
differences in display and the reasons for those differences, so it
should cause the reader little trouble.
B. Primary Study Areas
1. St. Louis
a. Background
By far the most numerous data are available from the St. Louis
area. Several days suitable for analysis were selected from the larger
body of data. Mage et al. (1978) provided a list of days•on which spe-
cial helicopter missions had been flown during the Regional Air
21
-------
Pollution Study (RAPS) in St. Louis. This list was compared with a list
of days during 1976 when high ozone concentrations were observed in the
St. Louis area. The latter list was provided by the project officer,
Mr. E. L. Martinez. Several of the days when special helicopter mis-
sions had been flown and relatively high ozone concentrations had been
observed were selected for intensive study and the data were obtained.
Some days with very high ozone concentrations were not selected for
analysis, because those dates were not listed among the days when spe-
cial helicopter missions had been flown. However, some of those days
may have helicopter data of a more routine nature that was collected,
but not tabulated, by Mage et al. (1978). Table 1 summarizes the 94
vertical profiles of ozone concentration in the St. Louis area that were
analyzed during the course of this project. The times given in the table
are Central Standard Time (CST) and refer to the approximate time at
which the vertical profile was completed. The mixing depths are given in
kilometers above sea level; the approximate elevation in the St. Louis
area is 0.15 km. The atmospheric stability class, as shown in Table 1,
was derived from information given on the surface weather charts. The
classification system is the one suggested by Pasquill (1961) and
Gifford (1961); Turner's (1969) algorithm was used. The wind speeds and
directions shown in Table 1 were estimated from the surface weather
maps.
The vertical profiles were taken at a limited number of locations,
which are shown in Figure 6. In Table 1, the profiles taken in the
vicinity of Smartt Field are designated with location number 131. Wher-
ever possible, the vertical profiles were taken directly above one of
the stations in the RAPS network. In those cases, the profile is iden-
tified by the number of the station over which it was taken. At some
locations, it was not feasible to fly the helicopter at low levels
directly above a station so an alternative site was chosen nearby. In
those cases, the profile location is designated with a number that
corresponds to the nearest RAPS station followed by the letter (H). The
classification of the location of a profile relative to the city is
determined by the extent to which that profile might have been affected
by the city. If the profile was taken upwind of the city or to one side
of it, then the first column in Table 1 under the heading "Location
Relative to the City" is checked. The second column is checked when the
profile was taken in the heavily urbanized part of the city, and the
third column when the profile was clearly taken downwind of the city. In
general, an attempt has been made to take into account changes of wind
direction that might have caused air flow from the city to reach the
place where the profile was taken, even though the wind direction at the
time of measurement might indicate that that site was uninfluenced. In
some cases, it was impossible to determine whether a profile had been
influenced by recent emissions from the city; in those cases, the last
column was checked.
The ozone profiles were categorized according to the system already
discussed. It is important to remember that the classifications are sub-
jective and that there are inevitably some profiles whose classification
22
-------
Table 1
SUMMARY OF OZONE PROFILES STUDIED FROM THE ST. LOUIS AREA
Date
760719
760723
Time
(CST)
0621
0721
0724
0750
0801
0814
0821
0831
0839
0849
0850
0901
0911
0912
0926
0938
0938
0947
0959
1011
1034
1257
1315
1338
0442
0501
0707
0741
0842
1128
1320
Mixing
Depth
(km msl)
0.5
0.5
0.5
0.4
0.4
0.5
0.4
0.5
0.5
0.5
1.3
1.4
1.4
1.4
1.4
1.4
1.8
1.8
1.8
0.4
0.4
0.4
0.3
0.6
Pasquill
Gifford
Stability
Class*
D
D
D
D
D
D
D
D
D
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
D
D
D
C
D
C
C
Wind (Approximate)
Speed
(m/sec)
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
•' 5
5
5
2.5
2.5
2.5
3
5
7.5
7.5
Direction
(deg)
180
180
180
190
190
190
200
200
200
200
200
200
200
200
200
190
190
190
190
190
190
170
170
170
240
240
240
240
240
240
240
Location
of Ozone
Profile
(See Fig. 6)
131
124
131
102
103H
106H
124
105H
102
103H
102
106H
106H
105H
105H
131
102
10 3H
106H
105H
131
131
121
115
131
131
103H
131
131
131
131
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
X
X
X
X
X
X
X
X
Downwind
X
X
X
X
Unknown
X
Ozone
Profile
Type
E
E
E
.E
D
F
A
A
F
A
F
F
F
F
F
D
F
A
D
A
A
D
D
D
E
E
F
A
A
A
A
Approximate Ozone
Concentration (ppb)
At
Surface
55
65
55
40
90
40
80
70
60
75
60
50
50
80
80
95
75
85
75
80
100
110
120
100
5
10
20
60
70
75
80
Average in
Mixed Layer
75
90
70
60
100
65
85
85
70
85
70
70
70
90
90
100
70
95
85
95
100
105
120
95
30
35
45
60
70
70
70
ro
OJ
-------
Table 1 (Continued)
Date
760729
760730
760803
Time
(CST)
0619
0639
0648
0733
0800
0807
0821
0834
0844
0902
0915
0925
0952
1030
1139
0608
0637
0657
0707
0718
0733
0744
0752
0802
0813
0841
0839
1058
1115
1122
1125
Mixing
Depth
(km msl)
0.3
0.5
0.5
0.6
0.4
0.3
0.4
0.5
0.5
0.5
0.5
0.5
0.5
0.7
0.8
0.4
0.4
0.3
0.3
0.4
0.4
0.3
0.3
0.3
0.3
0.4
0.8
1.5
1.6
1.7
1.7
Pasquill
Gifford
Stability
Class*
D
D
D
C
C
C
C
C
C
C
C
C
C
B
B
D
C
B
B
B
B
B
B
B
B
B
A
A
A
A
A
Wind (Approximate)
Speed
(m/sec)
4
4
5
4
4
5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
2.5
2.5
0
2
2.5
2.5
2
2
2
2
2
2.5
3
2
1.5
2.5
2.5
2.5
Direction
(deg)
250
270
290
180
180
180
190
190
190'
190
190
240
240
280
260
_
80
80
80
70
70
60
60
60
80
120
90
90
90
90
90
Location
of Ozone
Profile
(See Fig. 6)
131
131
131
124
102
103H
106H
105H
102
131
103H
106H
131
131
131
131
123
122
103H
106H
105H
102
103H
106H
105H
131
131
131
102
10 3H
123
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
X
X
X
X
X
X
Downwind
X
X
Unknown
X
X
X
X
Ozone
Profile
Type
E
E
E
E
F
F
F
F
F
A
A
F
A
A
A
F
F
F
F
F
F
F
F
F
F
F
A
A
A
A
A
Approximate Ozone
Concentration (ppb)
At
Surface
0
5
5
60
30
35
30
40
30
45
60
80
90
100
90
20
10
15
20
10
25
20
20
15
20
25
60
60
80
75
80
Average in
Mixed Layer
20
15
20
75
30
35
45
60
50
55
65
55
80
80
75
30
35
30
35
15
30
25
25
15
30
30
60
65
70
75
80
KJ
-------
Table 1 (Concluded)
Date
760803
760804
760812
Time
(CST)
1138
1145
1151
1153
1159
1201
1210
1214
1227
1234
1240
1243
1313
1549
1556
1607
1619
1631
1645
1745
0743
0831
0856
0903
0924
0933
0941
1001
1027
1258
0629
0841
Mixing
Depth
(km msl)
1.6
1.6
1.7
1.7
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.5
1.5
1.5
1.6
1.7
1.8
1.8
0.4
0.5
0.6
0.6
0.6
0.6
0.8
0.9
1.0
1.6
0.5
0.5
Pasquill
Gifford
Stability
Class*
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
C
D
C
B
B
B
B
C
C
C
C
C
D
C
Wind (Approximate)
Speed
(m/sec)
3
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
2
2
2.5
2.5
3
3
2
3.5
2.5
2
2
3
3
3.5
4
3.5
3.5
4.5
4.5
Direction
(deg)
90
90
90
90
90
90
80
70
60
50
50
40
40
330
330
320
320
310
300
30
190
210
220
220
230
230
240
240
230
240
160
240
Location
of Ozone
Profile
(See Fig. 6)
105H
102
103H
102
103H
106H
106H
105H
102
103H
123
106H
131
102
103H
106H
105H
123
106H
131
131
124
102
10 3H
105H
102
103H
10511
131
131
131
131
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
X
X
X
X
X
X
X
X
X
X
X
X
Downwind
X
X
X
X
X
X
X
Unknown
X
Ozone
Profile
Type
C
A
A
A
A
A
A
A
A
A
A
A
A
D
D
E
D
E
A
F
A
F
C
D
D
D
D
D
D
B
A
Approximate Ozone
Concentration (ppb)
At
Surface
70
70
75
85
85
90
90
80
90
90
95
100
95
120
100
110
100
100
60
25
40
60
60
55
75
75
75
75
90
25
70
Average in
Mixed Layer
80
75
75
80
80
75
80
75
85
80
85
90
80
130
120
135
110
145
120
70
40
50
75
60
75
85
75
80
70
100
70
75
*A = extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
A RADIOSONDE SITES
Q SPECIAL HELICOPTER SPIRAL SITES
• RAMS AND HELICOPTER SPIRAL SITES
122
A'4"
• 115
125 «
• 109
A 143
10
124 •
FIGURE 6 MONITORING AND HELICOPTER SOUNDING LOCATIONS
IN THE ST. LOUIS AREA
is ambiguous. The last two columns of Table 1 give the estimated ozone
concentrations at the surface and aloft. The surface concentrations were
estimated from objective isopleth analyses of data collected at the RAPS
stations. These objective analyses involve some smoothing, which should
minimize localized, unrepresentative influences that might be reflected
in the data from a single station. The concentrations shown in the last
column of the table are approximate averages of the concentrations
through the mixed layer. Of course, concentrations above the mixed layer
might be considerably higher, as in the case of ozone profiles of Type
F, or considerably lower, as in the case of ozone profiles of Type D.
26
-------
The available sample of ozone profiles from St. Louis is heavily
biased toward those taken during the morning hours. Over three-quarters
ol the cases shown in Table 1 represent hours between 0600 and noon. As
a result, all the stability classes represented are neutral (Class D) or
one of the unstable class (A = extremely unstable, B = moderately
unstable, C = slightly unstable). More afternoon profiles are available
from some of the other cities, but there are few nighttime profiles
available anywhere. This seems to be one of the serious shortcomings of
the available data set. Table 2 summarizes the frequency with which dif-
ferent types of ozone profiles are represented in the St. Louis data
set. Obviously this is not necessarily the frequency with which these
categories occur in the St. Louis area in general, since the data set is
biased toward high ozone concentration, toward the morning hours, and
probably toward fair weather conditions conducive to flying. The table
shows that the ozone profile Types A and F were most frequently
observed, but appreciable numbers of Types D and E were also found. The
scarcity of Type B profiles probably reflects the scarcity of nighttime
observations.
Table 2
RELATIVE FREQUENCIES OF OZONE
PROFILE CATEGORIES IN THE
ST. LOUIS DATA
Profile
Type
A
B
C
D
E
F
Relative
Frequency
(percent)
38%
1
2
15
14
30
27
-------
b. Examples
This section gives examples of vertical profiles and the
corresponding distribution of ozone concentrations near the surface in
the St. Louis area. Figure 7 shows the conditions that prevailed on the
morning of July 19, 1976. Five vertical ozone and nephelometer profiles,
measured between 0814 and 0849 CST, are shown. Where more than one pro-
file was measured at nearly the same time and in the same location, dou-
ble lines are shown. The locations at which the various profiles were
measured are indicated on the map at the left center of the figure. This
map also shows isopleths of ozone concentrations determined from the
RAPS data for the hour in which the profiles were measured. The arrows
indicate wind directions measured at the RAPS stations; the lengths of
u 600
G
•'0
.1_ I 1000 ,-
TlME 8M
100 150 2oo
OZONEIcolidl-PPB BSCATIIarqe dashl-l/IOKM
250 300
100 150
200 250
~VE 82
100 150 200 250
FIGURE 7 ST. LOUIS OZONE CONCENTRATIONS ON THE MORNING OF JULY 19, 1976
28
-------
the arrows equal one hour's travel distance at the measured speed. The
location of the rivers and the most heavily urbanized part of St. Louis
are marked by dashed lines.
On the preceding day, St. Louis had been near the center of a large
high-pressure system. On this morning, the high-pressure system was
somewhat to the east, but St. Louis was still in the western part of the
anticyclone and heavily influenced by the same system. There is every
indication that considerable accumulation of ozone occurred within the
high-pressure system; concentrations in excess of the then existing
National Ambient Air Quality Standard (NAAQS) of 80 ppb were found
throughout the area outside the city itself. Within the city, the emis-
sions of NO reduced ozone concentrations to less than 60 ppb. The pro-
file taken south, and upwind, of the city at 0821 shows that concentra-
tions of 80 ppb were present through a considerable depth of the atmo-
sphere. The profiles taken to the side of the city at 0831 and 0839 also
show that concentrations of 80 ppb or more extended from the surface
upward for some height. All these profiles have been classified as Type
A because the high concentrations extended through the top of the mixed
layer; this is assuming that the upper boundary of the mixed layer was
accurately marked by the abrupt decline seen in nephelometer (aerosol
concentrations) readings. However, there is some evidence that the ozone
concentrations begin to drop at higher levels, and so there may be some
justification for classifying these as Type D profiles. The influence of
the city is clearly seen in the profiles taken at 0814 and 0839. The
upper parts of these profiles are very similar to those taken outside
the city, but the lower parts show concentrations that are much below
those outside the city. Both these profiles have been classified as Type
F. Ozone concentrations above the mixed layer reflect the effects of
earlier accumulations of precursors and photochemistry and thus are
higher than the concentrations of ozone within the mixed layer. Obvi-
ously, the emissions from St. Louis had not had time to produce ozone
concentrations as great as those arising from the preceding day, and in
fact have reduced those concentrations at the lower altitudes over the
city. The shape of the profiles are characteristic of Type F profiles.
Figure 8 shows conditions in the St. Louis area on the morning of
July 23, 1976. Three of the profiles were taken at Smartt Field,
northwest of St. Louis at different times during the morning. The fourth
profile was measured at 0707 CST, just downwind of the city; that pro-
file, along with the corresponding nephelometer and temperature profile,
is shown at the top right of Figure 8. There is evidence of ozone
preserved in the stable air shown in the temperature profile at about
400 m. The ozone profile gives evidence of considerable destruction of
ozone by the NO emissions from the city below that same height. These
factors have led to a Type F classification. The other three profiles
in Figure 8 show the evolution of ozone profiles during the period when
the surface stability was wiped out by the onset of solar heating. The
temperature profile measured at 0501 shows a strong, surface-based
stable layer. Correspondingly, the ozone profile shows a sharp gradient
connecting concentrations of 60 ppb or more at a height of less than
29
-------
50 100 150 200 J5 JO
«nM-H/IOKM Thohdl M) DfTtlMWl)
50 100 150 2O 25 30
100 ISO 2O 25 30
~ •
50 IOO '50 20 25 30
OZONEIsolidl—PPB BSCATIiarge (JasM—t/IOKM Tt«o!i01 «nd
FIGURE 8 ST. LOUIS OZONE CONCENTRATIONS ON THE MORNING OF JULY 23, 1976
-------
AOO m, with concentrations less than 20 ppb at the surface. By 0741 CST,
the temperature profile shows that temperature decreased with height up
to about 300 m. This evidence of mixing is supported by evidence in the
ozone profile, which no longer had the very low concentrations in the
lower layers. Finally, by 0842, the concentrations of ozone in the
lowest layers had increased still further. Some rather interesting
features are shown in the upper parts of the nephelometer and ozone pro-
files. For example, at about 600 m, the nephelometer profile shows a
sharp peak, while there is a minimum in the ozone profile. This sug-
gests the presence of a distinct plume confined by the stable layer evi-
dent in the temperature profile at about the same height. The plume
appears to contain increased amounts of particulate material and reduced
amounts of ozone, probably caused by the presence of NO in the same
plume.
Figure 9 shows two vertical profiles taken later in the day on July
23, 1976, again at Smartt Field. By this time, mixing had progressed to
the point where concentrations were virtually uniform with height
throughout the depth of the vertical profile. Since the measurements did
not go above the depth of the mixing layer, it is not possible to say if
the uniformity of concentration continued above that height. However,
the profiles were classified as Type A on the presumption that the con-
centrations of about 60 or 70 ppb would not be unreasonable, throughout
a fairly substantial depth of the atmosphere, in a high-pressure area
such as was found to the south and east of St. Louis on the morning of
this day. By the time represented by the two maps in Figure 9, it is
apparent that precursors from St. Louis were causing a build-up of con-
centrations at the surface in areas downwind of St. Louis.
Figure 10 shows conditions in the early morning hours of July 29,
1976. The two ozone profiles shown are classified as Type E, although
they might well have been classified as Type B. Inversion or isothermal
lapse rates are present in the lowest few hundred meters of the atmo-
sphere. As a consequence, ozone destruction at the surface reduced con-
centrations to near zero values in the lowest layer; there was insuffi-
cient mixing from above to offset the destruction. The reason for
categorizing these as Type E rather than Type B profiles is found in the
nephelometer profiles. The nephelometer profiles show evidence of a
strong pollution layer aloft. It has been presumed that high ozone con-
centrations were associated with this evidence of higher pollution,
rather than background levels, as would be the case with Type B pro-
files. Obviously, this is a subjective choice, and the interpretation
is complicated by the fact that there was a stationary front in the
immediate vicinity of St. Louis at this time, so the nephelometer read-
ings may well reflect something other than pollution, such as the pres-
ence of some condensation. The general flow pattern at the two times
shown in the figure indicates that there was a wind shift such as might
be associated with the passage of the front.
Figure 11 shows five profiles taken in the vicinity of St. Louis
between about 0730 and 0830 CST on the morning of July 30, 1976.
31
-------
S3
80 PPb 1200
80
80 ppb
80
100
100
100
120
1300
O 00 100 25 JO
OZONElKllKlt-PPe HWKll •>*
B5CATIIwg« d«sn>—I/IOKM
1200
"—•...„
s1000
t
r
f 600
I I
1320
I
\
JULY 2.V
O SO 100 20 25
020NE(soti
-------
"•"-'
£ «x»
I
1 m
5 200
o
I
0619
i\
i\
~\ )
s /
- : \
I
) •*•-''*'
1 r
X -^-^
A *" ^v
"
i . i . i . i
20 25 JOO SO IOO ISO 200 25
TIMril ml 02ONEI»ol*ll-<>f>8 BSCATlivgt tucn>-i/lOKM
DPTIOKhl — °C
0600
20 ppb
E «o
UJ 400
o
D
t
i 200
0
2
0
y f
i i
64
8
':) #-'"
J
1.1.1.
0 25 30 0 SO IOO IbO 2O
Tltoltdt «nd OZONCI»oiid>— PP6 6SC»TH*»g« 0*tM— I/IOKM
20 ppb
0700
\\ /;-—--.:::
v» //'»»• » *x-»
*»*n..»v% //-** NN*C»
/ /^ Vv\
X ,/• Xv.^ -
JUl1! :'). l')76
FIGURE 10 ST. LOUIS OZONE CONCENTRATIONS ON THE MORNING OF JULY 29, 1976
-------
50 100 150 200 250 30O IS 20 25 3O
200 250 30015 2O 25 SO
50 100 150 200 250 30015 20 25 JO
50 100 150 200 250 300 15 2O 25 JO
50 100 150 200 250 300 15 20 25 30
OZONEl.oliOl—PPB B5CATIIarg< omnl-l/IOKM TlioliO) ino DPTIdMXI —°C
FIGURE 11 ST. LOUIS OZONE CONCENTRATIONS ON THE MORNING OF JULY 30, 1976
-------
Generally, the air flow was light and from the south. At this time, St.
Louis was located on the cold-air side of a weak stationary front, and
in the southwest quadrant of a weak high-pressure area. The upper three
profiles on the right side of the figure were taken at 0800, 0807, and
0821 GST. They show clearly the influence of the city. The observed sur-
face concentrations of about 30 ppb are much less than the 60 to 110 ppb
that was observed above the same sites. There is even evidence of the
city's effects in the temperature profile. Apparently the surface rough-
ness and added heat from the city prevented the formation of an inver-
sion or isothermal layer at the surface, such as is evident in the pro-
file taken ouside the city at 0834 GST and from the upwind station at
0733 CST- Both those stations show some evidence of lower concentrations
at the surface, but in the case of the upwind station, the lower surface
concentration is probably the result of the surface-based stable layer
acting in combination with surface destruction in the manner described
before. The nephelometer profile taken at 0834 CST shows some evidence
of high aerosol concentrations at the surface, so the reduced ozone con-
centration might well have been caused by increased NO. This same pair
of profiles also shows some evidence of a plume aloft, with reduced
ozone and high nephelometer readings at around 400-500 m altitude.
Figure 12 shows conditions during the day of July 30, 1976. It is
somewhat difficult to classify the two profiles taken inside the city at
0844 and 0925 CST, because of their limited extent. However, on the
basis of profiles shown in Figure 11, it is presumed that ozone concen-
trations increased somewhat with height above the top of the profiles
that are shown in Figure 12, so the later profiles have been classified
as Type F. The profile taken at 0915 CST shows very uniform ozone con-
centration with height, with little evidence of influence from the NO
emissions of the city, although it does appear to be downwind of the
city, as evidenced by the high nephelometer reading in the lowest layer-
The profile taken above Smartt Field at 0952 CST shows evidence of being
reasonably constant, with height up to the top of the mixing layer at
about 500 m; the other profiles taken on the same day at other locations
suggest that the concentration is probably much the same above the mix-
ing layer as within it, so this profile has been classified as Type A.
Figure 13 shows five profiles taken between 0830 and 0930 CST on
the morning of August 4, 1976. All these profiles have rather limited
vertical extent and so their classification is difficult. St. Louis was
in the western part of a very large high-pressure area on this day, thus
the light south-southeasterly winds shown in the figure. The profile
taken over the city at 0856 CST was classified as a Type F, based on the
surface level concentration derived from the analysis of ground-level
ozone observations. It was not possible to classify the vertical profile
taken over the city at 0913 CST because of its limited vertical extent.
The upwind profile taken at 0831 CST was assumed to be a Type A, partly
on the basis of its upwind location and partly because of the analysis
of surface concentrations. The downwind profile taken at 0903 CST is
probably a Type C profile, because of the higher concentrations observed
at the surface. The remaining profile, taken at 0924 CST, was classified
35
-------
SO IOO ISO 200 250 300 10 15 2O 25
OZON£l«oliO>-l>« B5C«ril«n> OPTIOj.nl —
50 100 150 200 250 3OO 20 25 10
=----^^-_~ 0925 •. \
^ii^-~ \ \ \
00 150 200 250 IOO 2O ZS JO
BO IOO 150 200 250
OZONEIsolK))—PP6 6SCATIi*rq« dnM—I/IOKM
300 20 25 30
T(MIK»> *nd DPIIOMh) —
JULY 30.
FIGURE 12 ST. LOUIS OZONE CONCENTRATIONS DURING MIDMORNING ON JULY 30, 1976
-------
U)
0
A
i
R'
56 , J "
i '
1.1.
20 25 O
ir
0913
2O 24 O
20 n o
100 ISO
KX> ISO
\ "
I
I
1 >l.
a
! ? -!
i.t.
100 160
SO KX>
_1_
_1_
'00
M>-
onM—t/iony
AUCUST 4.
FIGURE 13 ST. LOUIS OZONE CONCENTRATIONS AT 0830 TO 0930 CST ON THE MORNING OF AUGUST 4, 1976
-------
as Type D because of the uniformity of its concentration within the
mixed layer. It was assumed that concentrations above the mixed layer
were somewhat lower, more nearly like those observed at the upwind sta-
tion. The depth of the mixed layer could not be determined from these
vertical profiles, but it was available from analysis of lidar (laser
radar) observations taken at site 141 (see Figure 6) and reported by
Endlich et al. (1978).
Figure 14 shows five more vertical profiles taken later in the day
on August 4, 1976. With the exception of the profile taken at Smartt
Field, it appears that the ozone concentrations are reasonably uniform
in the vertical from the surface upward, even within the city itself.
This is probably because vertical mixing had been fairly vigorous and
offset most of the NO or surface destruction. The profile taken at 1027
CST above Smartt Field shows reasonably uniform concentrations above the
surface, but the much higher concentrations shown at the surface are
likely incorrect and may reflect an artifact of the smoothing process
used in the data analysis (see Endlich and Mancuso, 1968; Mancuso and
Endlich, 1973). In classifying this profile as Type D, we have assumed
that the surface concentration is about 70 to 80 ppb.
Figure 15 shows two profiles measured on the morning of August 12,
1976. St. Louis was in the western part of a high-pressure area at this
time, but showers had been observed to the west at Kansas City and to
the north at Springfield, Illinois during the morning. That shower
activity may have been associated with the squall line shown over the
Great Lakes on the National Weather Service analysis. Such a trough or
squall line may have caused the wind shift shown in the figure. In any
event, it appears that there was substantial vertical mixing. That
strong vertical mixing would normally be associated with nearly uniform
vertical distribution of ozone. If the analysis of surface ozone con-
centrations is correct, then there was probably a stable layer near the
surface at the time of the earlier profile (0629 CST). The rather low
concentrations at the surface are characteristic of a Type B profile.
Above the lowest layers, concentrations were nearly uniform except for
the pronounced depression in ozone concentration at around 500 m. That
depression was probably caused by emissions from a power plant. By 0841
CST, the profile was nearly uniform with height, ranging from about 75
ppb at the surface down to about 55 ppb at low altitudes, but staying
generally between 70 and 80 ppb at higher altitudes. The 0900 CST
analysis indicates that ozone production had begun downwind of St. Louis
by that hour.
2. Los Angeles
a. Background
Table 3 summarizes the vertical profile data available from the Los
Angeles Basin. The profiles measured before September of 1973 are based
on the work of Blumenthal et al- (1974). The profiles for the later
38
-------
•00
•00
400
1OO
0
600
400
100
400
no
i r
1027
20 23 JOfiO
1 I
0962
\
0
; \
i
93
3
1
ISO
'III
20 25 3060 IOO 150
TlMMI OIOtttltoW)-PPe
OMM-VIOKM
1000
.
—°±J
AUCHISI 4.
FIGURE 14 ST. LOUIS OZONE CONCENTRATIONS AT 0930 TO 1030 CST ON THE MORNING OF AUGUST 4, 1976
-------
MOO
1100
K»0
MOO
1200
MOO
•00
too
400
0629
K) 15 10 K 30 0
Tlwwl vid DPTIdnM —°C
0841
*
1
IS ZO . 25 30 0
CM) mm DPTfdnn) —°C
BO 100 150
BSC*Tllv«« Onhl—I/IOKM
AlKJUSI \2. l')7(.
FIGURE 15 ST. LOUIS OZONE CONCENTRATIONS DURING THE MORNING OF AUGUST 12, 1976
-------
Table 3
SUMMARY OF OZONE PROFILES STUDIED FROM THE LOS ANGELES BASIN
Date
720921
721018
730719
730725
730726
730727
730816
730824
Time
(PST)
1220
1252
1525
1636
0920
1257
1638
1306
1327
1555
1607
1624
1705
0725
1538
2139
0022
0400
1217
1237
1600
1100
1221
Mixing
Depth
(km msl)
2.0
0.3
1.3
2.0
0.5
0.7
0.8
0.2
0.5
0.4
0.8
0.4
0.9
0.7
0.5
0.5
0.6
0.8
1.3
1.1
1.1
1.0
1.3
Pasquill
Gifford
Stability
Class*
D
D
Wind (Approximate)
Speed
(m/sec)
3
5
7
7
8
7
3.5
3.5
4
Direction
(deg)
290
190
190
250
250
280
230
230
250
Location
of Ozone
Profile
(See Fig. 16)
Redlands
Long Beach
Bracket!
Redlands
Shepherd
Shepherd
Rialto
Hawthorne
Shepherd
El Monte
Brackett
Corona
Brackett
Rialto
Riverside
Riverside
Riverside
Redlands
Rialto
Redlands
Upland
Rialto
Location Relative to the City
Upwind
or Side
Within
the City
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Downwind
X
Unknown
Ozone
Profile
Type
A
D
F
A
B
A
D
F
D
F
D
D
C
F
D
B
B
B
D
D
D
F
F
Approximate Ozone
Concentration (ppb)
At
Surface
70
170
140
140
0
5
300
130
390
180
320
210
320
10
140
0
0
0
230
240
320
160
180
Average in
Mixed Layer
70
160
120
140
20
10
300
150
400
210
350
210
250
50
140
50
60
0
240
230
340
160
190
-------
Table 3 (Concluded)
Date
730919
730924
731004
731005
731025
731028
731029
731030
731031
731107
Time
(PST)
1320
1336
1413
1452
1554
1521
1325
1551
1630
1415
1600
1412
1431
1231
1247
1319
1435
Mixing
Depth
(km msl)
0.6
0.5
0.8
1.2
1.0
1.0
0.6
0.8
0.8
0.6
1.4
0.7
0.6
0.8
0.8
0.9
0.9
Pasquill
Gifford
Stability
Class*
C
C
C
D
D
D
D
D
D
C
C
C
C
C
C
Wind (Approximate)
Speed
(m/sec)
4
4
4
6
4
5
6
7
5
3
3
4
3
3
4
Direction
(deg)
180
200
200
200
260
210
190
240
140
200
160
230
180
180
230
Location
of Ozone
Profile
(See Fig. 16)
Pasadena
El Monte
El Monte
Ontario
Compton
Compton
Glendale
El Monte
Upland
Glendale
Mira Loma
Arcadia
Arcadia
San Marino
Santa Fe Basin
Arcadia
Cucamonga
Location Relative to the City
Upwind
or Side
Within
the City
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Downwind
X
X
X
Unknown
Ozone
Profile
Type
D
D
D
D
F
A
E
F
D
D
D
E
D
D
E
E
E
Approximate Ozone
Concentration (ppb)
At
Surface
200
125
175
175
40
25
150
50
175
120
200
140
120
80
80
110
170
Average in
Mixed Layer
160
200
100
25
300
190
150
210
180
145
100
140
180
200
to
A = extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
dates come from the Los Angeles Reactive Pollutant Program (LARPP) and
were taken from a report by Johnson and Singh (1977). This latter group
was examined in somewhat more detail than the earlier profiles. For the
earlier profiles, only that information which could be deduced from
Blumenthal et al. (1974) is included in Table 3. Figure 16 shows key
locations in the Los Angeles Basin.
FIGURE 16 MAP OF SOUTHERN CALIFORNIA
Examination of Table 3 shows that most of the vertical profile
measurements made in the Los Angeles Basin were made during afternoon
hours. In fact, 35 of the 40 profiles were measured between noon and
1700 Pacific Standard Time (PST). Another bias in the data is caused by
the extremely large urbanized area within the Los Angeles Basin. Virtu-
ally all the profiles must be considered as having been measured above
the city. To a large extent, the Los Angeles profiles complement those
from St. Louis by providing more afternoon coverage and more illustra-
tions of the effects of urban emissions on the lower portions of an
ozone profile.
Los Angeles also has some unique characteristics that cannot be
generalized to other areas. For example, it was noted earlier that the
persistent summertime subsidence inversion, in combination with the
mountains surrounding the basin, sometimes produces ozone profiles of
Type F by a mechanism different than that prevalent in other parts of
the country. We discussed earlier Edinger's (1973) suggested mechanism
for the injection of ozone into stable layers aloft. The July 19, 1973,
43
-------
observation of Blumenthal et al - (1974) at Rialto (shown in Figure 17)
illustrates injection of ozone into the elevated inversion layer by
upslope flow along the heated mountain slopes. Some of the examples from
the LARPP study that are presented on later pages also seem to show the
effect.
1600
1400
1200
£ 1000
+*
| 800
UJ
O
K 600
_i
400
200
(
Vw ' 1 ' '
/-** ^ \ Temp
"\ ^•-— °3 1
- "C"; \
Ccjii^l. x^
c:~, *-•-•—. ^
i \/
* i
t i
^ •-y//'///^^////, GROUND ELEVATION ''^/////--/>///////r,
RIALTO, CALIFORNIA
19 JULY 1973
1738 PDT
till
] 01 0.2 0.3 0.4 0
03 — ppm
1 1 1 1
124681
LIGHT SCATTERING 0>scat) — lO^nr1
1 1 1 1
LAYER CAUSED BY
UPSLOPE FLOW
SURFACE MIXING
LAYER
5
0
-5
15 25
TEMPERATURE — °C
35
45
SA-5321-9
FIGURE 17 EXAMPLE OF OZONE LAYER ALOFT, PROBABLY CAUSED BY UPSLOPE FLOW
-------
b. Examples
Figures 18 through 22 show vertical ozone profiles that were mea-
sured on September 19, 1973, in various parts of the Los Angeles Basin.
Vertical profiles of nephelometer readings (Bscat), nonmethane hydrocar-
bons (NMHC), and oxides of nitrogen (NOX) are also shown. The tempera-
ture and dew point instruments were not functioning on this day. The
locations where the profiles were measured are marked on the maps by
circled stars. Surface wind observations and surface ozone concentra-
tions are also shown on the maps, as they were for the St. Louis data.
However, the plotting and analysis of the Los Angeles data were done by
hand, rather than with an objective, computerized system, as was used
for the St. Louis analyses.
The vertical profiles collected during the LARPP study were fre-
quently interrupted at various altitudes by short intervals when the
helicopter flew a rectangular pattern at constant altitude. All the data
collected during one of these rectangular segments were averaged and
plotted on the profile. The horizontal bars show the standard error of
those observations. The large standard errors resulted primarily from
the fact that the data collection system was subject to considerable
interference, and there were frequent "noise spikes" that caused scatter
in the data points. Usually it is quite easy to identify the errant
points visually and to ignore them in the interpretation of the
profiles .
Figure 18 shows the ozone profile measured at about 1307 to 1320
PST above Pasadena on September 19, 1973. The bottom of the profile
corresponds with the approximate top of the mixing layer, so that it is
not possible to classify the profile with absolute certainty. Profiles
collected elsewhere during the afternoon showed rapid decreases in pol-
lution concentrations with height at the approximate top of the mixing
layer. This suggests that the 200 ppb (20 pphm) surface concentrations
probably extended throughout the depth of the mixing layer. At about
1300 m there was a peak in ozone concentration, and also in the back-
scatter and NOX profiles which suggests upslope injection of pollutants
into the stable layer aloft.
Figure 19 shows the vertical profiles over El Monte somewhat later
on the same day. The ozone profile shown in Figure 19 is a good example
of the Type D profile and lends further support to classifying the pro-
file in Figure 18 as Type D. The profile in Figure 20 has also been
classified as Type D, although there is considerable scatter in the data
points. We assumed that there was little change in the characteristics
of the profile during the 1/2-hour that elapsed between the measurements
shown in Figure 19 and those shown in Figure 20.
Figure 21 shows the ozone profile measured over the Compton area
between about 1523 and 1554 PST. This profile has been classified as
Type F. The mechanism producing this profile is not entirely clear, but
it appears that there is considerable ozone aloft within the stable
45
-------
. NO. OB wore io ami o»-i*-n tiro
"•*
racxr — KD «-•
uw" w>. ». oe wooz 10 onti m-ir-n ra«t:
1320
I I
1320
io »
OZDNC - M**!*
0.4 o.e o.s
W>« — pen
- ^150 ppb--^
1300
JO 111 .
SI I'll MHI K I", l'n.1
FIGURE 18 PASADENA PROFILES DURING THE AFTERNOON OF SEPTEMBER 19, 1973
-------
m. OB swxn roi ont o»-»n TIME: IS»:«O-UMJO uuw» o*. MO. oewro mi O«TI: at-n-n nut,
*>
^J
noo
zooo
*
!~
' UJU
i i i
j • 1 • 1 ' 1 ' 1
1336
'"^ ""%
I.I.I
— —
-
r-....
l~~
0 5 a 19 zo 0 «> zo JO «0 50
HOT — «f «-• OttKC — KM
noo
1000
r
1 noo
1
c
-
-
-
1 .1 1 1
— r • • i • i • r
-
_
-
i .1.1.1.
-10 0 XI a » V! •"> 0 10 10 JO «0
OmwiT — «r IWITIWTVXI — "C
l«
g
B
5
soc
J \
-
- v
^-•-
*^'
'••-.
»<.
I.I.I f
133
5
' r .
• '•••'. ^ . . •
, -- — .T1'^ •
'•T* .• '
! . 1 . 1 . 1
O 2 4 e n ,rn n? p.« 08 OP '
WM: — <*„ ' am — ~-
= %•^••^; "^ - ' "'
=>l|MlllHlllllllllllllMI^
_ . ! MMlil'MHtltplfl
'-'~^' -'^;.'v^
C~*^, "^- •- "'
SI I'll \1»! R I1'. WM
FIGURE 19 EL MONTE PROFILES DURING THE AFTERNOON (1336 PST) OF SEPTEMBER 19, 1973
-------
oo
ooo
2000
moo
1000
eoo
o
2000
noo
«00
000
°0
290O
»»
«0
1000
n
iMtff OP. NO. 06 9H003 (C) DATE: 09-19-73 TIME: I39O*X>-MI3:OO
!")'['
'
"
4
Hi.
-'^^^^
— *. *^_, —
^T
'i.
I.I.I
» w is a
•SOW — 10-" «-
till
J
"
r
t
i(
I
•r
i.i.i i
OJ 0^ 0^ OJ 1.
NO — w»
-
\
-I,
*T
i . i . i . i
1413
3 C
1413
0 C
J
1413
uatrr OP. w. o» woos 10 o«re: om-rj me: HMKX>-«I):OO
i i i i
_
_..«
-
1— '
, ,-''',
1 . 1 . 1 . 1
K> ZO 30 4O 50
020M _ pgft,,,
!•••••••••
"
, ._.
*"*!*
' ' .'"••-^. ..
inir1
' ••*•
t-
i . i . i . i .
OJ 0.4 o.e o^ i.
NOX ^ BV"
1 • ' 1 ' 1 ' 1 '
-
-
-
1 . .1.1.1
1400
10 -to
10 20 90
mrcwnum — °c
SHPTHMBhR 19, 1973
FIGURE 20 EL MONTE PROFILES DURING THE AFTERNOON (1413 PST) OF SEPTEMBER 19, 1973
-------
: »SZ3:OO-<»3:JO
1554
OZtMC — •*•»
«0 90
1 • 1 • 1 ' T
•t .
~i • r
1554
0.4 OJt 0.0
SIIPTI MB! R I1). I97J
FIGURE 21 COMPTON PROFILES DURING THE AFTERNOON OF SEPTEMBER 19. 1973
-------
layer and that this ozone aloft is being undercut by clean marine air
flowing onshore with the sea breeze. The wind arrows (shown in the fig-
ure) give clear evidence of a well established sea breeze.
Figure 22 shows the conditions above Cotnpton on the afternoon of
September 24, 1973. Unlike the case shown for the same location in Fig-
ure 21, the low concentrations observed at the surface were found to
extend upward through the top of the marine layer (at about 1000 m as
shown by the temperature and dew point traces) and well into the stable
layer. The nephelometer readings show a marked decrease at the top of
the mixing layer, but this may simply reflect the change in humidity,
causing particles to dry out and become smaller, less effective light
scatterers.
Figure 23 shows the ozone profiles measured above El Monte on the
afternoon of October 5, 1973. Figure 24 shows conditions above Upland,
somewhat farther to the east, later during the same day. No data are
available below about 800 m above El Monte. Very dry conditions (indi-
cated by the negative dew point temperatures) suggest that all the data
were collected above the marine layer- The ozone concentrations within
the inversion layer, above the marine layer, were about 100 ppb (10
pphm) or greater. At the surface, the concentrations were about 50 ppb.
The high concentrations of ozone trapped within the inversion layer and
the lower concentrations below provide reason for classifying this pro-
file as Type F. The Upland temperature and dew point profiles shown in
Figure 24 confirm that the top of the mixed layer was below 800 m. It
has been assumed that the relatively high ozone concentrations shown in
the lower parts of the profiles extend to ground level; this is based on
an extrapolation of the isopleth analysis shown in the figure. Based on
this assumption, the profile was classified as Type D, but this is
admittedly uncertain.
Figure 25 shows conditions on October 25, 1973, above Glendale.
The demarcation between the relatively polluted marine layer below about
600 m and the cleaner, drier air above that height was quite pronounced.
The NOX and nephelometer (Bscat) readings dropped to near zero values
above the stable layer. Ozone concentrations also dropped abruptly at
the top of the mixed layer, in the manner of a Type D profile.
Figure 26 shows the conditions in the eastern parts of the Los
Angeles Basin on the afternoon of October 28, 1973. The mixing layer
appears to have been fairly deep, about 1300 m judging by the nephelome-
ter readings. Very high concentrations of ozone had built up within the
mixing layer as the air moved over the basin and beyond the most heavily
urbanized areas. Concentrations of 200 ppb or greater were found
throughout the mixing layer. This relatively uniform distribution of
ozone within the mixing layer, and the abrupt decrease at the top, is
characteristic of a Type D profile.
Figure 27 shows the conditions during the afternoon of the next
day, October 29, 1973, above Arcadia. Although there are no temperature
50
-------
UW*» Of. WO. OB
n°°i • —r
*. MO. OO WOOS (Cl D*T: O9 ?
: 1433.30-13*1:00
i
10
men — a* .-
'*T7
:.*
4i -
1521
1521
TO SO *O
OIONC — a*nm
., , . 1 . r
1521
0 J 0,« 0 0
^ ,'--.,_ -i'-'O 1500
~ iniitiinmim ,
si rn MMI K :•».
FIGURE 22 COMPTON PROFILES DURING THE AFTERNOON OF SEPTEMBER 24, 1973
-------
LWFF Of. WO. 19 3WOO2 IC1 0*TC: KJ-O9-T3 TIME: IS39;?0-I35I:OO UMf* Of. HO. 19 3MO02 1C) 0*TE: IO-OS-T3 TBIE: IS59:2O-'M':OQ
\
1 - . - 1
1551
1551
10 20 JO 40
i r
1551
ZO SO 40 OO
OZOW — wv»
^50"•
t-v '^'"""iimiiiiiiiiiiiii
~~1;-; 4 '-.'•"
90 40
20 mi.
oj 0.4 o.e
OCTOBI R 5. I973
FIGURE 23 EL MONTE PROFILES DURING THE AFTERNOON OF OCTOBER 5, 1973
-------
Ul
0». MO. H 9UOO2 (01
: i6Oe.-OO-i«SO;00
I • i • i • i
1630
\,
•
1 . 1
K>
— KT* wr*
20
— »C
1630
1 ---- ^. I.I.I
10 O OJ 0.4 0.8 o.fl
200K>2O30«O9O
oiONt — §•*«
m
1,00
1 woo
s
*„
0
1
.
>
-
,
••••••• 1 1 •
_.
'r.
?
I.I 1
1630
_
i
i
-
'%
?
J
I.I.I,
,.'^000
1600
30*0
(X'TOHl R 5. I''7.1
FIGURE 24 UPLAND PROFILES DURING THE AFTERNOON OF OCTOBER 5, 1973
-------
or. NO. » swos ici witi e-M-ra ™t: iM9.oo-»i9«o LABO-OP. «. gi jnota ici are: n-n-n TIME: m»j»-Mi»ao
Ui
o
aoo
1
1
r
i
i
i
3 9
.r'
.'•-. "v.
''^r
'«*
d
't.
"° *
1
-
i
'
-
*~*
K> 19 2
oca — wr* •-•
~
'"•^^ -
i.l.i.
n zo BO •
-
I.I.I.
1415
o o
1415
o
1415
f. ' "
• |
'^r '
*\ i
' ^~"
10 ZO 30 40 M
ozone — n>i>ii
i • ,i -t- • i • i
\ :
s,'-, ,
j
'it
' >*H
1 .1.1.1.
•<0 0 10 ZO SO «
, , , ,
•-'^^^
140°
iliyillllllllllllllllHIl1 ""A
,)^l-. "
'20 mi.
10 ° oj 0.4 0.8 o.a
OCTOBER 25. 1973
FIGURE 25 GLENDALE PROFILES DURING THE AFTERNOON OF OCTOBER 25, 1973
-------
o» NO. zr stxxa 10 ami -o-a-n no,
o>. no. 2T 3»oe2 ici tm ro-a-ra m.t:
Ul
Ul
noo
MOO
*
1.
C
' "OOO
!
* m
"
C900
1000
i™
i
i
8
P
i
SCO
Z900
ZOOO
!«.
f
' KWO
|
000
ft
• \ • 1 ' 1 '
_ -
__
*""£
'"*,
• "'>
£ v
' ' ~ '' **
1 . 1 . I
5 K) I* 3
»3C« — . «T* M-*
.,-,.,. j
-
J\M %
*v
', \
•-.1
— ''^\ —
1 .1.1.1.
-K> o n a so *
~
-4"
- \
y
>..
• 1 . 1 . 1 i
1600
0 C
1600
0
1600
• y • 1 • 1 1 '
—
1 ; . ;
••:'.-
1 * ' *Vm
_ *i -
»7^-«
—. ' '
~ '- ' 1
1 . 1 . 1 . 1
10 » SO «O 51
' 1 I *' 1 ' I ' I '
._
V.
V.
v\
?.
. V,
— C —
'*•
-10 0 10 » 3O «C
-
!'
^ _
}
fe
1 . 1 . 1 . 1 .
, 50 ppb
1600
0.2 0.4 0.6 00
OTTOBhK 28. I97J
FIGURE 26 MIRA LOMA PROFILES DURING THE AFTERNOON OF OCTOBER 28, 1973
-------
UMFf> OP. NO. 28 9MOOZ K] tHTE: 10-29-73 TIME: I33O:OO-I4I2:OO UMn> Of>. MO. 2B 31MO2 (Cl [MTC: IO-29-73 TIME: I33O:OO-M v~- r , - j
'^"C^r ^ " ^-. --1,.-x :• c;
^.-. '-~-C^-3 '" '^'-v-
- -. r ^'oo/i.-,vv ^T ^.^_j^jj
"°""Q/i5'"X "o'jjv"
100' —
"" V™S
r=.-r"\
.'0 mi.
10 0 0.2 0.« 0,8 0.8 1.0
MM — wm
OCTOBI R 2'). 1973
FIGURE 27 ARCADIA PROFILES DURING THE AFTERNOON OF OCTOBER 29, 1973
-------
or dew point measurements available to define the location of the inver-
sion, it appears from the NMHC, NOX and nephelometer (Bscat) profiles
that the top of the polluted layer was at about 600 m. Ozone concentra-
tions within the upper part of this layer were near 200 ppb, although at
the surface they were less than 150 ppb. The fact that the ozone con-
centrations do not decrease rapidly with height at the top of the mixing
layer may indicate the presence of ozone trapped in the stable layer
above this foothill location.
Figure 28 shows conditions above Arcadia on the afternoon of
October 30, 1973. Although the dew point profiles suggest that the top
of the marine layer might be at 600 m, all the other traces shown in the
figure indicate a deeper mixing layer, with the top at about 1000 m. It
appears that the mixing is vigorous enough to have offset most of the
effects of destruction near the surface. The surface concentrations are
around 120 ppb. Those at higher altitudes are about 140 to 150 ppb.
As is shown in Figure 29, ozone concentrations above San Marino
during the afternoon of October 31, 1973, were fairly uniform up to a
height of 1400 m where they dropped from about 100 ppb to 50 ppb. Con-
centrations at the surface were around 80 ppb. There is some evidence
in the NOX and NMHC profiles that the top of the polluted layer is at
about 900 m rather than 1400 m. The ozone profile gives some evidence
of this also, but there appears to have been an influx of higher ozone
concentrations aloft. The dew point profile shows a corresponding,
slightly moist, layer at about the same height as the increased ozone
concentrations. The lower portions of the ozone profiles shown in Fig-
ure 29 are typical of the Type D profile, while the upper portions are
like the Type F profile that arises from Edinger'S (1973) hypothesized
mechanism, which introduces ozone into a stable layer aloft.
Figures 30 through 32 show conditions over three different loca-
tions in the Los Angeles Basin during the early afternoon of November 7,
1973. In all three examples, the ozone profiles have been classified as
Type E, because the vertical mixing has not been sufficient to compen-
sate for the destruction at the surface. Although fairly vigorous mix-
ing is indicated, it appears that the destructive processes were suffi-
cient to reduce the ozone concentrations in the lower layers by 100 ppb
or more below the concentrations observed at higher altitudes. For
example, Figure 30 shows that while concentrations aloft were 180 ppb or
more, surface concentrations were about 80 ppb in this area. Figure 31
shows that concentrations above Arcadia were generally greater than 200
ppb within the mixing layer, but at the surface they were only about 100
ppb. For the example shown in Figure 32, surface concentrations were
around 150 ppb, while aloft over Cucamonga, the concentrations exceeded
250 ppb.
57
-------
CO
__ UWPF of. NO, 29 3MOO2 1C) 0*Tt: IO-3O-T3 TIME: t4»:08-l43t:OO uWPP OP. MO. 29 3MOC2 (C) O*Tt IO-3O-T3 TIME: r42O:C»-nJirOO
2000
*
t
" 1300
!
1 KXX>
|
S «0
0
1 ' F ' I ' 1
1431
_
-I -
i.t.i.
_
- -•:;(::
1 . 1 . 1 . 1
B5CHT — KT* «•'
raw
2000
£
f
* «00
fe
1 MOD
§
f-
5 ,00
2900
2000
1 1000
I
5«,
D
1
' ' ' J
'
1431
~
^
*l!>"
\
-ro 0 tO 20 SO 40
oewpowr — °c
i • i • i • i
J
1 .
'
;
5-
''-.
i ...
•10 0 10 20 50 4O
I'll
1431
T
4
i-i.i i
•f
"l"'
1 . 1 . 1 • . 1
1400
20 nil.
10 0 O.Z 0.4 0.6 O.B
OCTOBER 30. l')73
FIGURE 28 ARCADIA PROFILES DURING THE AFTERNOON OF OCTOBER 30, 1973
-------
utrr or. «o. 10 wou KI out a-si-n me-. l
uarr o». MX 30
o>Tt io-ai-rs mt:
1322
to >o
OtOf — IV—
« a x «o
«•>«•» — «c
1322
C,'
PL
-
^
f
1
'\-
^1
t
,'
•-•••• 1 [••
-
-
, , !
1322
-
-
,
1
-
"V
\
X
*
h
H
t . i . 1
90 «0
20 mi.
O.4 O.A O.S
OCTOHt R M. 1*)73
FIGURE 29 SAN MARINO PROFILES DURING THE AFTERNOON OF OCTOBER 31, 1973
-------
Of. NO. 55 SMOGS IC» DArt: H-07-T3 TIME: I2IS:OCMZ«7;00 UMPP OP. NO- 35 SM003 1CI D*Tt; H-OT-7B TIME: IZIfl:OO-l2«T:OQ
2900
2000
*
1,500
|
1 1000
X
£
* .00
0
290O
2000
*
1 1000
1
, WOO
*
l~
1
E
1 1000
S
L
o
o
1 • 1
—
-
v.
»
i
- "'t
i.i.i.
3 5 10 15
MOT — «r" "I'1
I ' I ' ,
'' ' )
' .'. "t
1 .1.1.1
•to o 10 zo jo 4
DCVPOtNT — »C
j
-
_
•
I
"?
-\
z « 6 a i
1
1247
o t
1247
0
1247
) 0
1 • 1 ' 1 ' 1
—
-
it: .• .
'.:;•:'•!
• • • • « t :
! >i • • •
i!;-
10 2O SO *O *
OZONC — POtan
1 ' '
•S
: I ;
1 .1.1.1.
-IO 0 (0 20 3O 4
Twfcwumc-oc
'
_
i-
\m ^
V""'1''
i4..', .;•
'M
i - i . i . i
0.2 0.4 0.8 O.B I.
1200
JO mi .
NOV1.MB1 R 7. 147.1
FIGURE 30 PROFILES OVER THE SANTA FE FLOOD CONTROL BASIN DURING THE AFTERNOON OF NOVEMBER 7, 1973
-------
JIW» or 10. n MJCJ 101 0>Tt ll.07.ri tmt.-
0" •*>• 38 3"OOS IOI B«t! ll-or-T3 Tint:
O\
I 000
3
1 - • - 1
1
1319
"I,
OUT — w* m"
0.2 0.« O.O 0.0
NO* — p*m
20 so «o so
BOO
1000
i
* iwo
i
1 noo
|
J
* 000
1
-
-
_
"
1
' ' ' ' J
-
-
\
^
f
•f
t ^
I.I.I.
1319
'
-
-
-
1
1 1 1
-
-
£
V
s
T
vl
I.I.I.
-IOO 10 10 SO 40 '*° ° W 2O SO *
nOO
J—
r
1 000
1
5
eoo
o
•••••••• J 1 •'•••'•'• 1
-
i ;
• \
— . 1 .I.I.I.
1319
p
1"
/..
• :'i:
' i
I.I.I.,
5)llMlllii|IIIIIIMMIIII "',
UliimiuiMiimmiilf
20 mi.
NOVI MB1-R 7 1473
FIGURE 31 ARCADIA PROFILES DURING THE AFTERNOON OF NOVEMBER 7, 1973
-------
KJ
2900
2000
* 1300
*
1UITUOC —
, § 1
D
25OO
20OO
s
1 m
e
\
' 1000
I
900
Q
UWl"r Of. NO. 39 SUOC2 (Cl DATE: 11-07-73 TIME: I4U;(XH«35:00
1 ' I ' 1
'
T ~
L
• -.- • •
_\ '
. ' * :
1455
map* OP. NO. 35 5X002 Id 0»TI: II OT-7J TIME: MI«:OO-WM:OO
• 1 ' 1 ' 1 ' I '
,..,.
• : i
" i!
l|
i . i . i . i .
9 K> 19 20 0 10 20 90 40 90
•9WT — ID"* m-* 0«)« — m«m
-
*^
'V
\1
— ,. —
\
-
1 .1.1.1
1455
1 • i • i • i • i
T ;
V
X
I
\
V
-
i .1.1.1.
-IO 0 H> 2O BO 40 -10 0 10 20 » 4fl
OnrOMT — «C TCMPERATU« — °C
2900
2000
L
I
' KXX)
§
5 «o
T —
"1
'••,
'l
"1
-
_ . 1 . 1 . 1 . 1 .
1455
-
••;
fc
i
M| J-
ii
-
i . i . i ' i
0 2 4 0 0 1C 0 0.2 0.4 0.0 0.8 1.
1400
100
JO mi .
NOVl-lMBt'R 1 1973
FIGURE 32 CUCAMONGA PROFILES DURING THE AFTERNOON OF NOVEMBER 7, 1973
-------
3. Houston
a. Background
Table 4 summarizes the 30 profiles examined from the Houston area.
All the vertical profile information and much of the meteorological
information were obtained from the report of Westberg et al. (1977).
The measurements of surface ozone concentrations were obtained from the
data archived as part of the Houston Area Oxidant Study (HAOS; see,
e.g., Ludwig, Martinez, and Nitz, 1978). Morning and afternoon hours
are reasonably well represented in this data set. About 57 per cent of
the cases are for the hours between 0700 and noon, and about 40 per cent
between noon and 1800. No profiles were measured directly over the city
itself, but the number of profiles measured upwind is approximately the
same as the number of profiles measured downwind of the city.
All the profiles examined were measured during the month of July,
1976. The aircraft data collected by Dr - M. E. Shauck of Baylor Univer-
sity (see Ludwig, Martinez, and Nitz, 1978) were also investigated, but
most of those profiles were collected at higher altitudes and had not
been graphed or digitized. Also, the tabulated summaries of Shauck*s
data had rather poor vertical resolution. For these reasons, it was
decided to concentrate the analysis on the data of Westberg et al.
(1977), although their data suffers from some of the same deficiencies.
In particular, the vertical resolution in the graphical presentations is
not as good as was available for the Los Angeles, St. Louis, and
Washington, D.C. data sets.
The examples presented in the next section use reproductions of
graphs given by Westberg et al. (1977). The reader should be aware that
those authors do not always use the same ozone and temperature scales in
their graphs. Each vertical profile from Westberg et al- (1977) is
accompanied by a map of the Houston area showing the distribution of
ozone concentrations observed at the surface. The isopleth analyses in
these diagrams are subjective interpretations of the observed values.
The winds at the 850-mb level (approximately 1500 m) were determined
from the National Weather Service analyses and are shown in the upper
right corners of the figures. The wind direction is indicated by an
arrow whose length is equal to one hour's travel at the observed speed.
Stability classes were determined from the solar radiation and wind data
of Westberg et al. (1977) using the algorithm given by Ludwig and Dab-
berdt (1976). One final note is in order concerning the figures in the
following section. The altitude shown on the graphs obtained from West-
berg et al. (1977) are given in thousands of feet above sea level (msl).
b. Examples
Figure 33 shows the ozone profile taken about 60 km northwest of
Houston during the late afternoon of July 2, 1976. At its highest levels
the ozone increased to about 60 ppb. However, up to about 2.4 km
63
-------
Table 4
SUMMARY OF OZONE PROFILES STUDIED FROM THE HOUSTON AREA
Date
760702
760704
760705
760707
760708
760708
760710
760712
760713
760714
760715
760718
760720
760721
Time
(CST)
1720
1110
1707
0715
0809
0923
0740
1410
1410
1439
0804
1005
1420
1610
1700
1105
1535
1035
1925
0810
0830
1640
0715
Mixing
Depth
(km msl)
1.2
0.5
0.6
0.6
1.0
1.1
0.6
0.6
0.6
0.6
1.1
1.2
1.2
0.9
0.5
0.6
0.6
1.1
0.3
1.0
Pasquill
Gifford
Stability
Class*
D
A
D
C
A
A
B
B
D
D
C
B
B
C
C
B
C
C
D
B
C
C
C
Wind (Approximate)
Speed
(m/sec)
4
3
5
4
4
2
2
2
7
7
4
5
4.5
4.5
6
2.5
3.5
4.5
2.5
2
4.5
4.5
4.5
Direction
(deg)
180
190
90
60
50
60
50
200
140
140
70
140
170
170
190
120
190
190
190
190
190
150
100
' Approximate
Location
of Ozone
Profile
60 km NW
40 km SW
60 km NW
40 km W
50 km E
50 km NW
30 km W
40 km W
30 km W
15 km SW
45 km W
20 km W
50 km W
75 km NW
30 km W
35 km W
60 km NW
40 km NW
45 km W
45 km W
35 km W
30 km W
30 km W
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
Downwind
X
X
X
X
X
X
X
X
X
X
Unknown
Ozone
Profile
Type
A
A
F
A
A
A
B
F
C
C
B
C
C
C
F
A
E
F
D
D
F
D
B
Approximate Ozone
Concentration (ppb)
At
Surface
30
30
30
25
35
35
15
80
70
130
10
80
160
170
25
30
80
20
20
20
25
70
30
Ave rage in
Mixed Layer
30
30
30
35
35
35
30
80
50
100
35
80
120
150
25
35
120
40
25
20
30
80
50
-------
Table 4 (Concluded)
Date
760722
760723
760724
760725
Time
(CSX)
0805
1215
0825
0930
1400
0750
0930
Mixing
Depth
(km msl)
1.8
2.7
1.2
0.6
1.7
0.6
0.3
Pasquill
Gifford
Stability
Class*
A
A
B
B
B
B
A
Wind (Approximate)
Speed
(m/sec)
5
2.5
0
5
3.5
0
1.5
Direction
(deg)
130
140
-
40
180
270
Approximate
Location
of Ozone
Profile
35 km W
25 km W
40 km W
35 km W
40 km SW
40 km W
45 km SW
Location Relative to the City
Upwind
or Side
X
X
X
Within
the City
Downwind
X
X
X
Unknown
X
Ozone
Profile
Type
F
A
F
F
D
B
F
Approximate Ozone
Concentration (ppb)
At
Surface
25
45
25
35
80
40
45
Average in
Mixed Layer
30
45
25
45
90
70
60
Ul
extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
10 JO SO 40 50 60
03 (ppb)
/ > v
/ \ \
' > >•
20 ppb
._
• JULY 2 1800
' 20 km
FIGURE 33 HOUSTON OZONE CONCENTRATIONS DURING THE LATE AFTERNOON
OF JULY 2, 1976
(8,000 ft), concentrations remain very nearly constant between 30 and 40
ppb, in spite of the fact that there is a stable layer with a base at
about 1.2 km (4,000 ft). Thus, the profile has been classified as Type
A. Although the 850-mb wind direction suggests that this sounding was
made generally downwind of the Houston area, the shape of the ozone pro-
file indicates little or no influence from city emissions, so this pro-
file was not put in the downwind category.
66
-------
Figure 34 shows another Type A ozone profile
July 4, 1976. The reasons for classifying this as
much the same as those used to classify the July
33. Ozone concentrations were nearly constant at
elevation of 2 km (6,500 ft), although there is a
a much lower altitude. This profile was measured
to one side of the Houston area, and was clearly
emissions.
, taken near midday on
a Type A profile are
2 case shown in Figure
about 30 ppb, up to an
strong stable layer at
slightly upwind, well
uninfluenced by Houston
o
t-4
CO
E 7
K
*-s
V\ 'M 62 £.6 70 74 70
TEMP
10 K> 30 «0 SO GO
03 (ppb)
40 ppb
JULY A 1200
H
20 km
FIGURE 34 HOUSTON OZONE CONCENTRATIONS NEAR MIDDAY OF JULY 4, 1976
67
-------
The profile shown in Figure 35, for the late afternoon of July 5,
1976, has been classified Type F. Ozone concentrations up to about 2 km
(6,500 ft) were nearly uniform at 30 to 40 ppb. This appears to be the
height of the afternoon mixing layer, as evidenced by the presence of a
nearly isothermal layer in the temperature profile. Above the top of
the afternoon mixing layer, the ozone concentrations nearly doubled, to
60 ppb. The profile appears to be little influenced by the Houston
emissions, probably from being off to one side, although generally in a
downwind direction.
^ 8
t—I
01
E 7
jj
*" 6
H 3
20 30 «0 50 CO
03 (ppb)
O
20 ppb
JULY 5 1800
20 km
FIGURE 35 HOUSTON OZONE CONCENTRATIONS DURING THE LATE AFTERNOON
OF JULY 5, 1976
68
-------
Figure 36 shows the vertical distributions of ozone at three dif-
ferent times during the morning of July 7, 1976. Although the 850-mb
wind direction was from the south to south-southeast during this morn-
ing, Table 4 shows that the winds reported by Westberg et al. (1977) at
the surface were from the northeast or east-northeast. This probably
accounts for the buildup of ozone concentrations to the west of the
city. The profile taken at 0715 CST about 40-km west of the city was
classified as a Type A profile, although the decrease in concentration
near the surface might have warranted a B classification. There was a
general, gradual increase in ozone concentration with height from about
35 ppb at around 0.3 km (1,000 ft) to about 55 ppb at about 3 km (10,000
ft). This gradual increase with height is fairly characteristic of the
troposphere (Singh, et al., 1977). The other two profiles shown in Fig-
ure 36 have also been classified as Type A. They have concentrations
between 30 and 40 ppb in the lowest layer, but there are no data above
about 1-km (3,300 ft). The earlier sounding that extended to higher
altitudes suggests that concentrations at higher altitudes were probably
consistent with a Type A ozone profile. None of the profiles shown in
Figure 36 appear to have been appreciably influenced by emissions from
Houston.
Figure 37 shows ozone concentrations during the morning and after-
noon of July 8, 1976. The morning profile was classified as Type B;
concentrations above about 0.3 km (1,000 ft) were generally uniform,
between about 30 and 40 ppb, but it appears that the concentration at
the surface was probably much less than that. During the afternoon of
July 8, 1976, ozone concentrations at the surface increased downwind of
the Houston and Galveston areas. Concentrations at the surface in the
downwind areas exceeded 80 ppb and were even higher aloft, reaching over
100 ppb in a layer between about 0.8 and 1.1 km (2,500 to 3,500 ft).
The profiles certainly have the shape of a Type F profile, but the ori-
gins of the ozone trapped aloft are unclear. According to Westberg et
al. (1977), surface winds during much of the night were calm, so that
the pollutants seen in this layer may be the result of precursors emit-
ted during the night and early morning. It is not possible to determine
with any certainty the degree to which this profile was influenced by
urban emissions, but on the basis of the 850 mb winds and the surface
winds near the time of the observation, the profile has been put in the
category of those uninfluenced by the urban emissions.
Figure 38 shows an example of a Type C profile. This profile was
measured downwind of the Houston area on the afternoon of July 10, 1976.
Concentrations at the surface were about 70 ppb. Between about 0.3 and
0.6 km altitude (1,000 to 2,000 ft), ozone concentrations dropped from
about 75 to 20 ppb. The 20-ppb concentration coincided with an abrupt
change to a more stable lapse rate. Above 0.6 km, ozone concentration
increased from about 20 to about 45 ppb at 2.6 km (8,500 ft). The lack
of vertical resolution in the data makes it difficult to determine
whether the ozone concentrations within the lowest, mixed layer were
uniform, but it appears that the sharp decrease in concentration
69
-------
C 20 30 «O W 60
(ppb)
20 ppb
JULY
-->
7 800
•••-•" s \
tC iV 70 72 7" 76 76
TEl/P
0809
30 40
03 (ppb)
40 ppb
JULY 7 0900
20 JO
66 72 76 80 61
TEMP
0923
10
03 (ppb)
- - , i i
', .""'' '-. / ' \
' ' \
'•-. \
60 ppb
JULY 7 1000
20 km
20 km
20 km
FIGURE 36 HOUSTON OZONE CONCENTRATIONS DURING THE MORNING OF JULY 7, 1976
-------
X
<^s
w
a '
3 .
x
i i i
52 5C 60 64 60
TEMP
10 20 30 40
55 61 67 73 79 83 9i
TEMP
(•'I
1410
i i i
25 JO fj 15 65 75 65 95 105
03 (ppb)
03 (ppb)
t
JULY 8 800
20 km
20 ki
FIGURE 37 HOUSTON OZONE CONCENTRATIONS ON JULY 8, 1976
-------
10 20 30 40 50 60 70 BO
03 (ppb)
JULY 10 1500
v;
20 km
FIGURE 38 HOUSTON OZONE CONCENTRATIONS
DURING THE AFTERNOON OF
JULY 10, 1976
72
-------
o
*-4
X
< 2
80
TEMP
1005
50 60 70 80 90 IOO 110 120
03 (ppb)
JULY 12 1100
S
<
10 JO «0
03 (ppb)
20 km
20 km
FIGURE 39 HOUSTON OZONE CONCENTRATIONS DURING THE MORNING OF JULY 12, 1976
-------
9
x-x e
f~4
0
B 7
o
^
X
I
H
40 60 BO ICO 120 140 100 ISO
03 (ppb)
o
i-H
X
3
<
8 92 9G 100 IO4
TEMP cr°l
1610
I1O ICO IUO 200
03 (ppb)
-' • -'.V
'14'°",Vn~' '•'"'•• •'' 'A
/ }20 100 X../ \\
» I I 80 »_\
JULY 12 1700
f-
20 km ' ' 20 km
FIGURE 40 HOUSTON OZONE CONCENTRATIONS DURING THE AFTERNOON OF JULY 12, 1976
-------
of the day. However, the vertical ozone profiles taken downwind of
Houston, especially during the late morning and during the afternoon,
suggest that ground-level ozone concentrations were much greater than
those observed in, and upwind of, Houston. The earliest profile was
classified as Type B; concentrations at the surface were very much lower
than those aloft. The concentrations aloft were generally between 30
and 45 ppb. Concentrations above 1.5 km (5,000 ft) seem to have
remained in the 40- to 50-ppb range, at least until afternoon. Figures
39 and 40 show that the concentrations near the surface continued to
increase during the day, at least until 1610 GST. However, it does not
appear that these high concentrations were uniformly mixed throughout
the mixing layer- It is not easy to define the top of the mixing layer
from the available data, but it appears to have been around 1.1 to 1.2
km (3,500 to 4,000 ft) after 1000 CST. Thus these profiles show a sur-
face concentration appreciably greater than that observed at the top of
the mixing layer; this is an important characteristic of Type C
profiles.
Figure 41 shows an ozone profile taken about 30 km west of Houston
during the late afternoon of July. 18, 1976. Ozone concentrations
observed at ground level generally ranged from about 7 to 13 ppb. Con-
centrations above about 0.8 km (2,500 ft) were greater, about 50 to 60
ppb. If the temperature profile showed evidence of the formation of a
stable layer near the surface, this profile would have been classified
as a Type E. However, the category F has been chosen because there
appears to be a reservoir of higher ozone concentrations above the mix-
ing layer, but the choice of this category is admittedly very
subjective.
Figure 42 shows ozone concentrations during the late morning and
the afternoon of July 14, 1976. The sounding made west of the city at
1105 CST shows that concentrations of around 30 ppb extended from near
the surface to at least 2.7 km (9,000 ft). Clearly, this is a Type A
profile that is generally free of urban influence. Surface concentra-
tions of ozone nearer the urban area reach as high as 100 ppb. The sur-
face winds reported by Westberg et al. (1977) suggest that the precur-
sors emitted in the Houston area moved generally toward the southwest
during the morning hours and then, about midday, they turned and moved
toward the north. This would have brought the morning precursors, and
the ozone formed from them, over the area where the profile was measured
at 1535 CST. It is evident that there is considerable ozone within the
mixed layer; above the mixed layer the concentrations remained at about
30 ppb, as during the morning hours. Although concentrations were about
150 ppb at an altitude of 0.8 km (2,500 ft), surface concentrations were
less than 90 ppb. The temperature profile does not indicate strong sta-
bility in the lowest layers, but destruction at the surface appears to
have been too rapid to be offset by downward mixing of the ozone aloft.
This profile was classified as Type E because there was ozone aloft
within the mixing layer and the downward mixing in the lowest layers was
not sufficient to offset surface destruction.
75
-------
SZ 56 60 64 68 73 76 BO
TEMP
I'FI
1700
20 JO 40 50 60
03 (ppb)
JULY 13 1800
20 km
FIGURE 41 HOUSTON OZONE CONCENTRATIONS
DURING THE LATE AFTERNOON
OF JULY 13, 1976
76
-------
e 7 -
H
M
a
<
50 S« 56 £2 66 70 7« 74
TEMP
CF]
1105
20 X) 40
03 (ppb)
T 1 1 1 1 T
62 C£ TO 7« Tfl 82
TtMP
rn
1535
o1—±.
30 SO 70 90 MO I30
03 (ppb)
40',
JULY M 1200
•o
\
JULY 14 1600
40
20
20 km
20 km
FIGURE 42 HOUSTON OZONE CONCENTRATIONS, JULY 14, 1976
-------
Figure 43 shows the ozone concentrations in the late morning and
early evening of July 15, 1976. The morning profile, taken at 1035 GST,
shows that there was a layer of higher ozone concentration associated
with a more stable portion of the temperature profile, at about 0.8 to
0.9 km (2,500 to 3,000 ft). Ozone concentrations observed at the sur-
face were quite low throughout the area, generally less than 30 ppb.
These concentrations are consistent with the lower portions of the ver-
tical profiles. The 1035 CST profile has been classified as a Type F
because of the ozone layer aloft. By early evening, 1925 CST, ground-
level ozone concentrations had fallen below 20 ppb everywhere in the
area. In most places close to the city, the concentrations were near
zero. The vertical profile about 45 km west of the city suggests that
ground-level concentrations were about 15 ppb. At higher altitudes, the
ozone concentrations were about 20 to 30 ppb and relatively uniform to
the top of the sounding. If we knew that these same concentrations con-
tinued upward above the height to which afternoon mixing had penetrated,
we could confidently classify this profile as Type A. However, it has
been classified as Type D because we do not know whether the same con-
centrations were found above the afternoon mixing layer.
Figure 44 shows morning ozone concentrations around Houston for
July 18, 1976. Although there appears to be an isothermal layer at the
surface, the ozone concentrations were no less there than at higher
altitudes. The concentration is uniform at 20 ppb from the surface up
to the elevated stable layer at about 1.1 to 1.4 km (3,500 to 4,500 ft).
Concentrations above this altitude increase slightly to about 30 ppb.
It is uncertain what the concentrations were above the profile, so it
was classified as Type D; it would probably be equally easy to justify a
Type A classification.
Figure 45 shows ozone concentrations during the morning and after-
noon of July 20, 1976. The 0830 CST profile is Type F. Mixing had not
proceeded to very high altitudes and so the concentrations of ozone
aloft are much greater than those at the surface. There is some evi-
dence that the mixing may have been more pronounced within the city, and
hence there are greater ozone concentrations at the surface, 40 to 60
ppb in the city versus about 25 ppb where the profile was measured.
Similarly, there may have been greater mixing over the warm water sur-
faces to the southeast of Houston, where a surface ozone concentration
of 60 ppb was observed. Figure 45 shows that by afternoon, considerable
ozone had been formed within the mixing layer. Ozone concentrations of
75 to 85 ppb were found from the surface to 1.1 km (3,500 ft). The 1640
CST profile seems to have been influenced by the emissions from the
city, but there are other areas more directly downwind where concentra-
tions at the surface exceeded the 100 ppb.
Figure 46 shows the ozone concentrations around Houston during the
morning of July 21, 1976. The profile taken west of the city at 0715 CST
shows that concentrations above 0.8 km (2,500 ft) were generally 50 to
65 ppb. No profile data are available below 0.8 km, but extrapolating
78
-------
72 74 7G 78 80 62 81
TEMP
I'FI
H 3
80 82 M 80 88 90 92
TEMP
t'f I
1925
20 30 10 50 CO
10 15 20 25 JO
03 (ppb)
03 (ppb)
JULY 15 IIOO
20
\ O
> I i
JULY 15 2000
20 km
20 km
FIGURE 43 HOUSTON OZONE CONCENTRATIONS ON JULY 15, 1976
-------
66 70 72 7« 76 76 80
TEMP
rn
0810
I5 20 25 30 35
03 (ppb)
JULY IB 900
20 km
FIGURE 44 HOUSTON OZONE CONCENTRATIONS
DURING THE MORNING OF
JULY 18, 1976
80
-------
oo
60 70
03 (ppb)
60
JULY 20 900
2i 30 3
10 00 60 70 Tj BO
03 (ppb)
100 ppb'
80
60
JULY 20 1700
20 km
20 km
FIGURE 45 HOUSTON OZONE CONCENTRATIONS ON JULY 20, 1976
-------
e*i
o
1
H
62 66 70 74 78
TEMP
IT)
0715
60 70
03 (ppb)
X O
JULY 21 800
20 km
FIGURE 46 HOUSTON OZONE CONCENTRATIONS
DURING THE MORNING OF
JULY 21, 1976
the analysis of the surface observations suggests that the concentration
near ground level was around 20 ppb. It is not unreasonable to expect
that the thermal stratification in the lower layers was fairly stable at
this early hour. Therefore, vertical mixing was probably inhibited and
the destruction of ozone at the surface was probably not offset by down-
ward mixing from aloft, making this a Type B profile.
82
-------
Figure 47 shows the ozone concentrations during the morning and at
about midday on July 22, 1976. During the morning (0805 CST), concen-
trations were 25 to 30 ppb from the surface up to about 1.7 km (5,500
ft). The winds reported earlier in the morning at ground level by West-
berg et al. (1977) were calm. Some ozone may have remained aloft in the
relatively stable layer above 1.7 km, making this a Type F profile. The
midday (1215 CST) profile is a Type A profile; concentrations were
between 40 and 50 ppb from the surface up to an altitude of at least 2.7
km (9,000 ft). The location where the profile was measured was uninflu-
enced by the emissions from the city; downwind of the city, concentra-
tions exceeded 160 ppb.
Figure 48 shows the ozone concentrations during the morning of July
23, 1976. The profile of this morning was very similar to that on the
preceding morning. Concentrations were around 25 ppb below the isother-
mal layer at 1.2 km (4,000 ft). They were more than 55 ppb at higher
altitudes. The profile has been classified as Type F because of the
greater ozone concentrations above the mixed layer. The generally calm
winds reported by Westberg et al. (1977) suggest that the air throughout
the lowest layers was relatively stagnant and may have been influenced
by the nearby urban area, but this is uncertain.
Figure 49 shows morning and afternoon concentrations on July 24,
1976. The morning profile, taken downwind of the city, shows relatively
low concentrations at the surface, around 35 ppb. There were higher
concentrations aloft, about 60 to 75 ppb. Trajectory calculations and
the very light surface-pressure gradients prevailing in the gulf coast
area at this time indicate that there was probably considerable "milling
around" of the air with an accompanying accumulation of urban precursor
emissions. It appears from Figure 49 that the greatest ozone concentra-
tions at the surface were found to the southeast of the Houston area at
this time. However, this may reflect enhanced mixing over the warmer
water surfaces during the morning. Such mixing would have brought ozone
down to the surface from aloft. Under light wind conditions, virtually
the entire area has to be considered as having been influenced by urban
precursors. During the afternoon, the ozone profile shows near uniform
concentrations of 85 to 105 ppb throughout the mixed layer. Above the
mixed layer, ozone concentrations were in the 70-ppb range.
Figure 50 shows the ozone concentrations on the morning of July 25,
1976. At 0750 CST, the temperature profile shows that the lowest layers
were reasonably stable. An analysis of surface ozone observations sug-
gests that the surface concentration was around 40 or 50 ppb in the
lowest layers. The profile shows ozone concentrations of about 60 to 75
ppb above the lowest layers, appreciably more than the concentrations at
ground level, making this a Type B profile. Somewhat later in the morn-
ing, at 0930 CST, there was evidence of mixing in the lower layers at
the site where the profile was taken, to the southwest of Houston. Above
the mixed surface layer was a stable layer with a base at 0.3 km (1,000
ft). Within the stable layer, ozone concentrations of 80 ppb were
observed; within the surface-based mixed layer, the concentrations rose
83
-------
H 3
25 30 J5 40
45
03 (ppb)
70 80 84 86 92
TEMP
CFI
0805
9
>-> 8
i-4
m
B 7
4J
•" 6
GO ,70 BO K> 00 HO
TEMP
CD
1215
40 44 48 52
03 (ppb)
oo
140
20 km
20 km
FIGURE 47 HOUSTON OZONE CONCENTRATIONS ON JULY 22, 1976
-------
6 7
i->
"" 6
O
K '
V_X
1 '
£ '
< 2
I
25 50 55 40 45 50 55
03 (ppb)
JULY 23 900
20 km
FIGURE 48 HOUSTON OZONE CONCENTRATIONS
DURING THE MORNING OF
JULY 23, 1976
85
-------
oo
o
X
w 4
g
H J
M
< 2
t
62 66 70 7« TB 82 86
TEUP
0930
55 65 75
03 (ppb)
JULY 24 1000
O
X
*~s
w
H
62 6G 70 7« 76 82 86 30
TEMP
IT I
1400
iO GO 70 80 90 100
03 (ppb)
100
JULY 24 1500
20 km 20 km
FIGURE 49 HOUSTON OZONE CONCENTRATIONS ON JULY 24, 1976
-------
00
K
^v>
1 •
K '
-------
from a value at the surface of about 50 ppb to the 80 ppb within the
stable layer. A profile like this, indicating higher concentrations of
ozone stored in a stable layer aloft is a Type F profile.
The presence of such an ozone layer aloft is generally an indica-
tion of pollutant effects from the preceding day. The light and vari-
able winds prevailing on this and the preceding day make it quite possi-
ble that the ozone concentrations measured aloft anywhere in the area
might have come from anthropogenic and other emissions during the
preceding day or so. Figure 51 shows the trajectories of the air arriv-
ing at Houston on August 24-25, 1976. The air had remained within 200
km of Houston for an extended period of time, so there was probably con-
siderable accumulation of pollutants, but the locations where the two
profiles were taken on the morning of July 25, 1976, seem to be reason-
ably free of very recent urban influence.
KEY
• 25 JULY, 1976 - 0600 CST
O 25 JULY, 1976 - 1800 CST
A 24 JULY, 1976 - 0600 CST
FIGURE 51 LOCATION 48 HOURS EARLIER OF AIR ARRIVING IN HOUSTON
ON JULY 24-25, 1976
88
-------
4- Washington. D.«.C«
a. Background
Table 5 summarizes the 31 profiles available from the Washington,
D.C. area. The data provide a good balance between morning and after-
noon cases. Twelve profiles are for the hours between 0600 and 1200
EST, and 19 cases come from the hours between 1200 and 1900 EST.
Unfortunately, the surface ozone observations available from the
area did not provide very wide coverage. Most of the available data
come from the urbanized parts of Washington and Baltimore. However,
flight restrictions prevented any profiles from being measured over the
central part of the cities themselves, so it was difficult to combine
surface and aircraft data. Flight restrictions also prevented the air-
craft from obtaining data at the lowest altitudes, generally below 500
m. The lack of surface and low-altitude measurements makes interpreta-
tion of some of the profiles somewhat uncertain. Counterbalancing these
shortcomings is the fact that many measurements were made during a
period of prolonged stagnation. Thus, these data from Washington, D.C.,
provide a good picture of the buildup of ozone within the mixing layer
during stagnant conditions.
The winds at 850 mb (approximately 1500 m, msl) were determined
from the National Weather Service analysis and are indicated by arrows
plotted on the maps accompanying the ozone profiles shown in the next
section. The length of the arrows represents one hour's travel dis-
tance, except where dashed arrows are shown. In those cases, the winds
were too light to be displayed in the same way; one hour^s travel time
would usually be less than the length of the arrow head.
The surface winds shown in Table 5 were estimated from the surface
weather map analyses of the National Weather Service. The estimates of
stability class shown in Table 5 were derived from wind and cloud cover
information contained in these same analyses, using Turnerxs (1964)
algorithm. The ozone profile types and the surface ozone concentrations
shown in Table 5 have been estimated largely on the basis of the upper
parts of the profiles measured, and by use of the kinds of reasoning
that went into the development of the profile classification system.
b. Examples
Figure 52 shows the ozone profile measured to the southeast of
Washington, D.C., during the late morning of August 17, 1976. The two
lines shown in the graph indicate that measurements were made during
both an ascending and a descending spiral, as was usually the case in
the Washington data set. The temperature and dew point profiles indi-
cate that there was very good mixing up to the base of a stable layer at
about 2500 m. The ozone concentrations were nearly uniform, about 40
ppb, throughout the mixed layer and above. In the lowest layers, there
89
-------
Table 5
SUMMARY OF OZONE PROFILES STUDIED FROM THE WASHINGTON, D.C., AREA
Date
760817
760818
760819
760820
760821
760823
Time
(EST)
1113
0921
0940
1008
1031
1648
1711
1750
1810
1843
0955
1019
1047
1109
1625
1645
1711
1743
0636
0703
1620
1644
1726
1745
1210
1232
1251
1318
1343
0640
1636
Mixing
Depth
(km msl)
1.8
0.8
1.6
1.5
1.6
2.0
1.8
1.8
2.1
2.0
0.8
0.7
1.0
1.8
1.6
1.8
1.8
2.0
0.5
0.5
1.5
1.3
1.4
1.3
0.9
1.0
1.2
0.9
0.8
0.4
0.9
Pasquill
Gifford
Stability
Class*
B
D
D
D
D
D
D
D
D
D
C
C
B
B
D
D
D
D
C
C
C
C
C
D
B
B
B
B
B
D
D
Wind (Approximate)
Speed
(m/sec)
2.5
7.5
7.5
7.5
7.5
5
5
5
5
5
5
5
5
5
5
5
5
5
2
2
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
2.5
5
Direction
(deg)
330
30
30
30
30
60
60
60
60
60
50
50
50
50
90
90
90
90
360
360
150
150
150
150
70
70
70
70
70
360
320
Location of
Ozone Profile
(As shown in
Figs. 52-62)
1
2
3
4
1
4
1
5
6
2
8
9
7
10
8
9
7
10
7
4
7
4
8
9
7
11
12
1
13
7
7
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
Downwind
X
X
X
X
X
X
X
X
X
X
X
X
Unknown
X
X
X
X
Ozone
Profile
Type
A
A
A
A
A
D
D
C
C
D
F
F
F
A
D
D
C
C
E
E
D
D
C
D
D
D
D
D
C
A
D
Approximate Ozone
Concentration (ppb)
At
Surface
40
30
40
50
100
90
90
90
90
30
30
30
40
90
80
100
110
10
10
110
90
100
90
110
120
90
100
190
60
70
Average in
Mixed Layer
40
40
40
40
40
90
100
70
75
100
35
40
40
50
70
70
80
90
40
50
110
90
85
85
100
110
100
100
125
60
70
vO
O
A = extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
3800
SOOO
2500
>
; 2000
° 1500
1000
900
O 50
ozoNO
BSCATdanj.
i©r
1113
100 O
10 15 20 25
AUGUST 17, 1976
FIGURE 52 WASHINGTON, D.C. OZONE PROFILE, 1113 EST, AUGUST 17, 1976
-------
are some indications that concentrations of light-scattering particles
near the surface were greater than aloft. No similar indications are
seen in the ozone profile, so it has been classified as a Type A pro-
file. The 850-mb winds indicate that this location was downwind of
Baltimore; the surface winds suggest it was downwind of Washington. In
either event, it was probably influenced by urban emissions and was
therefore classified as being downwind of a city.
Figure 53 shows the data collected during the morning of August 18,
1976. Spiral ascents and descents were made at four different locations
around the Washington-Baltimore area. All the profiles show ozone con-
centrations of about 40 ppb throughout their entire vertical extent.
The earliest data, for the sounding ending 0921 EST, indicate that a
stable layer was present near the surface, with its top below 1000 m.
The dew point trace suggests that the moisture is confined to this
layer, but that aerosols, as indicated by the nephelometer (Bscat) read-
ings, were mixed to levels above 1500 m. Similarly, at the other loca-
tions, the nephelometer profiles show the top of the layer with the
greatest aerosol concentrations to be at 1500 to 1600 m. In general, the
temperature profiles also indicate the same height for the top of the
mixed layer. Apparently the ozone that was present was mixed rather
thoroughly through a much greater depth of the atmosphere, and little or
no new ozone formed from precursors confined within the mixing layer-
Only the 0940 EST profile that was taken southwest of Baltimore seems
likely to have been influenced by precursor emissions from a nearby
urban area, but even it shows little or no sign of excess ozone forma-
tion within the mixed layer.
Figure 54 shows the measurements taken at five locations during the
late afternoon and early evening of August 18, 1976. There are many
important differences between the profiles shown in Figure 54 and those
shown for the same morning in Figure 53. The most important of these
differences is the buildup of ozone, presumably formed from precursors
within the mixed layer. In three of the five locations, the ozone con-
centrations are reasonably uniform throughout the depth of the mixed
layer. The two profiles taken farthest from the urban area show rela-
tively high concentrations at the lowest altitude; the concentrations
decreased with height through the mixed layer. These latter two pro-
files are of Type C; the other three are Type D. According to the sur-
face winds and the winds aloft, none of the profile locations should
have been influenced appreciably by urban emissions from nearby cities
during the afternoon of August 18, 1976, except perhaps for the profiles
taken at the southern edge of Washington at 1648 EST.
Figure 55 shows the trajectories of the air arriving at Washington,
B.C., on August 18, 1976, as calculated by Mr. Dale Coventry of the
Environmental Protection Agency using Heffter and Taylor's (1975) compu-
terized trajectory program. The air arriving in Washington, B.C., had
passed over most of the urbanized areas along the east coast, so it is
quite likely that a considerable amount of ozone precursors had been
injected into the atmosphere and that those precursors and the ozone
92
-------
FIGURE 53 WASHINGTON, D.C. OZONE PROFILES DURING THE MORNING OF AUGUST 18, 1976
-------
vD
.P-
5 10 15 20 25 0
TIsolid) »nd OPTtdash) —°C
50 100 150
BSCATMa'gs dashl—M^MD"*
FIGURE 54 WASHINGTON, D.C. OZONE PROFILES DURING THE LATE AFTERNOON OF AUGUST 18, 1976
-------
KEY
• 18 AUGUST, 1976 - 0700 EST
O 18 AUGUST, 1976 - 1900 EST
FIGURE 55 TRAJECTORIES OF AIR ARRIVING AT WASHINGTON, D.C.
ON AUGUST 18, 1976 AT 0700 AND 1900 EST
formed from them were reasonably uniformly mixed through the depth of
the mixing layer- There is no apparent explanation for the two southern-
most profiles, where the ozone is not uniformly mixed through the mixing
layer.
Figure 56 shows the ozone profiles measured around the Washington-
Baltimore area on the morning of August 19, 1976. The three more north-
erly ozone profiles, measured before 1100 EST, are all of Type F. In
each case, there was a layer between about 800 and 1200 m where concen-
trations exceeded 50 ppb. Below this layer, concentrations were around
30 ppb; it appears that mixing has not proceeded to the point where this
elevated layer has been eroded. The profile measured downwind of
Washington, D.C., at 1109 EST also shows some very slight signs of an
95
-------
FIGURE 56 WASHINGTON, D.C. OZONE PROFILES DUFIING THE MORNING OF AUGUST 19, 1976
-------
elevated layer, but concentrations were much more nearly uniform at 40
ppb.
Figure 57 shows the ozone profiles during the afternoon of August
19, 1976. The two upwind profiles taken to the northeast of Baltimore
are typical Type D profiles with nearly uniform concentrations of 70 to
80 ppb up to the top of the mixed layer, at about 1600 to 1800 m. The
upper parts of the profiles taken downwind of Baltimore (at 1711 EST)
and downwind of Washington (at 1743 EST) are very similar to the upper
parts of the upwind profiles. However, at the lower altitudes, downwind
concentrations were appreciably greater, being 100 ppb or more near the
surface. This appears to have been the effect of ozone produced from
urban precursors that had not mixed thoroughly throughout the mixing
layer. These profiles have both been classified as Type C.
Figure 58 shows ozone profiles measured south of Washington and
between Washington and Baltimore early on the morning of August 20,
1976. All the available surface observations of ozone concentrations in
Baltimore and Washington, B.C., as well as those from two stations to
the northwest of Baltimore, show that concentrations at ground level
were less than 10 ppb at this hour. This almost certainly reflects the
effects of a stable layer at the surface, which probably formed during
the night. It appears that the ozone concentrations increased with
height from near zero values at ground level to more than 70 ppb in the
air above the ground-based inversion. Above the base of the elevated
inversion that was present at about 1300 m, the ozone concentrations
were 30 to 40 ppb. Winds were very light throughout the layer below
1500 m; as shown in Figure 58, the morning 850-mb wind was very light
from the east and the surface wind was very light from the north (see
Table 5). Both profiles were taken downwind of urban areas, but at the
early hour at which they were taken, there had not been sufficient pho-
tochemical activity to produce any discernible effects in the lower
parts of the ozone profiles.
Figure 59 shows conditions in the Washington area during the after-
noon of August 20, 1976. A strong subsidence inversion was evident
throughout the area at around 1300 to 1400 m. Winds within the mixed
layer were very light and from the east-southeast. Figure 60 shows that
air arriving at Washington, D.C., during the early evening of August 20,
1976, had come from the east at a rather slow rate of speed. The dots
on the trajectory mark the air position at 12-hour intervals; during the
preceding 48 hours, the air had traveled only about 450 km. The gen-
erally uniform distribution of the 100-ppb ozone concentrations below
the subsidence inversion indicates that the mixing was fairly good. The
nature of the nephelometer and the ozone traces indicates that very lit-
tle material was mixed upward out of the layer below the inversion. A
restriction to upward diffusion and the slow air movement allowed enough
precursors to accumulate in the lowest layers to produce the 100-ppb
ozone concentrations observed.
97
-------
VO
CO
0
2500
2000
1500
1000
I 2SOO°
2000
1500
1000
500
1500
1000
500
W I
1625 I
20 25 30
-
1645 I.
TO 25 30 0 50 100
1711
I . I . I . I
?0 25 SO 0
1743
I .-I . I .. I
) —°C
FIGURE 57 WASHINGTON, D.C. OZONE PROFILES DURING THE AFTERNOON OF AUGUST 19, 1976
-------
3500
3000
2900
15 ?0 25 0
20 .'•5 0
AUC'-UST :0.
FIGURE 58 WASHINGTON, D.C. OZONE PROFILES DURING THE MORNING OF AUGUST 20, 1976
-------
o
o
5 20 25 30 O 50 IOO
15 JO 25 JO 0 50 100 150 200
10 '5 ?0 25 30 0 50 100 "50 200
"M
1745
10 15 20 25 30 o
AliCUST -0. I<>7(i
'•-•0 200
FIGURE 59 WASHINGTON, D.C. OZONE PROFILES DURING THE LATE AFTERNOON OF AUGUST 20, 1976
-------
KEY
• 20 AUGUST. 1976 - 1900 EST
O 21 AUGUST, 1976 - 1700 EST
V 23 AUGUST, 1976 - 0700 EST
A 23 AUGUST, 1976 - 1900 EST
FIGURE 60 TRAJECTORIES OF AIR ARRIVING AT WASHINGTON, D.C. ON AUGUST 20,
1900 EST; AUGUST 21, 1700 EST; AND AUGUST 23, 0700 AND 1900 EST, 1976
The light winds and near-stagnant conditions persisted in the
Washington area through the following day, August 21, 1976. Ozone con-
centrations remained high and were confined below the subsidence inver-
sion, as can be seen in Figure 61. This figure shows ozone concentra-
tions shortly after midday on August 21, 1976. Figure 60 shows that air
arriving in Washington, B.C., during the morning had come from the north
and had traveled only slightly more than 200 km during the preceding two
days. Most of the profiles show ozone concentrations to have been about
100 ppb throughout the depth of the mixing layer. Ozone concentrations
measured over the water to the east of Baltimore at 1343 EST were appre-
ciably higher, approaching 200 ppb in the lowest layers. Although the
101
-------
o
to
0 15 20 25 30 O
100 150 200 25O i
5 10 15 20 25 0 ^ 50 100
Tlsoi'dl and DPTtdash) —°c ^ OZONEIsotidl—PP8
dl-PP0 fl^CATdarqp jashl— M''»IO"B
5 10 15 20 25 30 O
TIsoi'dl and DPTIdash] — °c
AIKIIIST 2\. IP7h
FIGURE 61 WASHINGTON, D.C. OZONE CONCENTRATIONS AT ABOUT MIDDAY ON AUGUST 21, 1976
-------
wind measurements indicate otherwise, this profile gives evidence of the
recent addition of urban precursors.
Figure 62 shows two ozone profiles measured between Washington and
Baltimore during the morning and afternoon of August 23, 1976. The
850-mb winds were very light, from approximately north in the morning
and shifting to north northwest during the late afternoon. The trajec-
tories of the air arriving at Washington, D.C., on this day are shown in
Figure 60. These two profiles illustrate the evolution of a Type A pro-
file, as observed in the morning, into a Type D profile as the precur-
sors contained within the mixed layer are converted to ozone. The light
winds make it very difficult to determine the extent to which these pro-
files were influenced by urban emissions. However, it seems likely that
if there had been great influence, the ozone concentrations in the
afternoon would have exceeded the 70 ppb that was observed.
c. Other Areas
The preceding section presented many examples of vertical profiles
of ozone and other photochemical pollutants in the four primary study
areas. The intention in presenting so many individual profiles, with
the accompanying discussions of the factors considered when classifying
those profiles, was to show how the various physical processes had to be
considered in order to classify the data. The numerous examples were
also intended to show the kinds of uncertainties inevitably encountered
when a subjective classification scheme is applied to the real world.
With those objectives accomplished, it is sufficient to present tabu-
lated summaries of the vertical profiles as observed in other areas from
which data are available. The data discussed below have been combined
with those from the primary study areas in order to develop some sta-
tistical descriptions of the combined data sets. For example, the rela-
tive frequencies of the various kinds of profile have been determined
for different periods of the day and for different locations relative to
the city. The statistical summaries are discussed later.
5. Toronto
Table 6 summarizes the Toronto area ozone profiles that were clas-
sified. The classified profiles were obtained from the report of Wiebe,
Lusis and Anlauf (1975). All the data were collected near Toronto, over
Lake Ontario or near Niagara, during August of 1975. All 53 profiles
were measured during daylight hours, between about 0800 and 1700 EST.
The forenoon accounted for 20 of the profiles, the afternoon for the
remaining 33.
Figure 63 is a map of the area in which the Toronto ozone and tem-
perature profiles were measured. The important locations referred to in
Table 6 are shown on the map. The winds given in the table are gen-
erally estimated from those observed with pilot balloons in the lowest
103
-------
IO 15 20 29 30O SO 100 150 ZOO 250 300
9 10 15 20 25
Tttolcd) «r« OPT(dnM —°C
50 100 150 200 250
OZONEdolidt—PPB BSCATdarqs Jasht—M"'«IO'8
AUGUST 2.t, 1976
FIGURE 62 WASHINGTON, D.C. OZONE PROFILES, AUGUST 23, 1976
-------
Table 6
SUMMARY OF OZONE PROFILES STUDIED FROM THE TORONTO AREA
Date
750808
750809
750810
750811
750812
Time
(EST)
1206
1233
1245
1450
1512
1545
1130
1150
1210
1527
1545
1604
1125
1155
1435
1455
1035
1105
1316
1343
1353
0947
1001
1015
1103
1117
1130
1405
1430
Mixing
Depth
(km msl)
1.0
0.7
1.1
1.2
1.2
0.3
1.2
1.1
1.2
1.5
1.4
1.4
0.4
0.4
1.5
0.5
0.2
0.2
0.3
0.3
0.5
0.4
0.5
0.4
0.2
0.2
Pasquill
Gifford
Stability
Class*
A
A
A
A
A
B
B
B
B
B
B
B
B
B
C
C
D
D
D
D
D
A
A
A
A
A
A
B
B
Wind (Approximate)
Speed
(m/sec)
1
1
1
1
1
1
5
5
5
3
3
3
3.5
5
4
3
4
4
2
2
2
1.5
1.5
2.5
3.5
4
Direction
(deg)
240
240
240
240
240
240
200
200
200
190
190
190
180
180
160
150
210
210
170
170
170
170
170
160
170
160
Location
of Ozone
Profile
(See Fig. 63)
Toronto Island
Lake Ontario
Niagara
Toronto Island
Lake Ontario
Niagara
Toronto Island
Lake Ontario
Niagara
Toronto Island
Lake Ontario
Niagara
Toronto Island
Niagara
Toronto Island
Niagara
Toronto Island
Niagara
Toronto Island
Lake Ontario
Niagara
Toronto Island
Niagara
Lake Ontario
Toronto Island
Niagara
Lake Ontario
Toronto Island
Lake Ontario
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
K
X
X
X
X
X
X
Within
the City
Downwind
Unknown
X
X
X
X
X
X
Ozone
Profile
Type
F
A
C
D
D
D
D
D
D
D
D
D
C
D
D
A
E
A
C
D
D
E
E
E
E
E
E
D
F
Approximate Ozone
Concentration (ppb)
At
Surface
75
45
40
70
85
120
125
125
115
135
155
145
105
80
130
105
65
75
120
115
125
35
45
50
85
85
55
110
55
Average in
Mixed Layer
60
55
55
90
90
110
115
130
110
130
130
130
105
90
90
90
80
80
100
100
95
60
60
75
115
110
80
110
55
o
U1
-------
lable 6 (Concluded)
Date
760814
750819
750827
750828
Time
(EST)
0835
0907
0935
1005
1445
1509
1535
1200
1221
1245
1340
1345
1400
1433
1510
0853
0917
0945
1015
1340
1357
1427
1455
1512
Mixing
Depth
(km msl)
0.6
0.4
0.5
0.2
0.3
1.0
1.0
0.2
0.2
0.2
0.2
0.5
0.7
0.2
0.2
1.2
0.9
0.2
0.3
Pasquill
Gifford
Stability
Class*
B
B
B
B
B
B
B
C
C
C
C
A
A
B
B
B
A
A
A
B
B
B
B
B
Wind (Approximate)
Speed
(m/sec)
3
3
3
3
2.5
2.5
2
7
7
8
9
2
2
3
4
3
3
2.5
2
4
4
3.5
3
3
Direction
(deg)
270
270
270
300
150
150
150
310
310
300
290
270
270
290
310
170
170
160
150
180
180
180
170
.170
Location
of Ozone
Profile
(See Fig. 63)
Toronto Island
Niagara
Oakville
Ajax
Toronto Island
Niagara
Olcott
Toronto Island
Niagara
Olcott
40 km East
of Oshawa
Toronto Island
Niagara
Olcott
Ajax
Toronto Island
Niagara
Olcott
Ajax
Toronto Island
Niagara
Olcott
Ajax
Oakville
Location Relative to the City
Upwind
or Side
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Within
the City
Downwind
X
X
X
X
X
Unknown
Ozone
Profile
Type
A
A
A
A
A
A
C
A
A
A
A
A
A
C
A
A
F
A
A
E
C
C
E
D
Approximate Ozone
Concentration (cob)
At
Surface
25
30
30
40
60
60
85
50
55
60
60
40
65
90
60
30
40
45
45
70
90
90
50
110
Average in
Mixed Layer
25
30
30
40
50
60
60
50
55
55
60
40
50
50
50
40
40
40
45
60
65
65
90
80
*A = extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
• Holland Landing
• Orangeville
Stouffville
Woodbndge •
FIGURE 63 AREA IN WHICH TORONTO PROFILES WERE TAKEN
atmospheric layers above the Toronto area, or from National Weather
Service analyses. Stability estimates are based on data from the same
sources. It should be noted that wind observations from a single loca-
tion in a lakeside environment may not adequately describe the lake
breeze circulations. Therefore specifying whether a particular profile
was "upwind" or "downwind" was subject to considerable uncertainty.
Lusis et al- (1976) have analyzed many of the profiles tabulated in
Table 6. Their analysis suggests that some of the profiles classified
as having little or no recent influence from the city might contain
ozone and precursors from the preceding day that was recirculated by the
107
-------
lake breeze. For example, they suggested that such recirculation was
responsible for the profiles observed on the morning of August 12, 1975.
Their paper should be consulted for details.
6. Indianapolis
The profiles summarized in Table 7 were contained in the report of
Lovelace et al- (1975). The approximate locations, relative to the
center of Indianapolis, of the three areas where spiral measurements
were made are given in Table 8. More details can be obtained from
Lovelace et al. (1975).
Table 8
LOCATIONS OF SPIRAL MEASUREMENTS
RELATIVE TO THE CENTER OF
INDIANAPOLIS
Site
Location Relative
to City Center
Indianapolis
West Newton
Oaklandon
Over center of city
20 km southwest
25 km northeast
It is evident from Table 7 that the profiles were taken in pairs,
one over the city and one upwind of the city. Two pairs of observations
were obtained during morning hours and three pairs during the afternoon.
The wind and other meteorological data necessary for estimating stabil-
ity class were reported by Lovelace, et al. (1975). One important
shortcoming of these data is the relative sparsity of observations of
high ozone concentration.
7. Denver
Table 9 summarizes the characteristics of the six ozone profiles
measured November 21, 1973, in the vicinity of Denver, Colorado, by
Blumenthal et al. (1974). Those authors gave maps showing the location
108
-------
Table 7
SUMMARY OF OZONE PROFILES STUDIED FROM THE INDIANAPOLIS AREA
Date
740823
740825
740912
740917
Time
(CST)
1431
1454
0844
0910
1453
1518
1010
1036
1510
1531
Mixing
Depth
(km msl)
1.7
0.8
0.8
0.9
1.8
1.8
0.6
0.7
1.1
1.1
Pasquill
Gifford
Stability
Class*
B
B
B
B
D
D
B
B
B
B
Wind (Approximate)
Speed
(m/sec)
3.5
3.5
2.5
2.5
5
5
5
5
5
5
Direction
(deg)
270
270
20
20
220
220
250
250
250
250
Location
of Ozone
Profile
West Newton
Indianapolis
Oaklandon
Indianapolis
West Newton
Indianapolis
West Newton
Indianapolis
West Newton
Indianapolis
Location Relative to the City
Upwind
or Side
X
X
X
X
X
Within
the City
X
X
X
X
X
Downwind
Unknown
Ozone
Profile
Type
A
A
F
A
A
A
B
E
D
D
Approximate Ozone
Concentration (ppb)
At
Surface
75
70
15
25
20
20
0
0
25
30
Average in
Mixed Layer
90
90
40
20
15
30
10
30
20
30
extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
Table 9
SUMMARY OF OZONE PROFILES STUDIED FROM THE DENVER AREA
Date
731121
Time
(MST)
0822
0915
0925
1135
1234
1242
Mixing
Depth (km
msl above
ground)
0.3
0.2
0.2
0.3
0.1
0.2
Pasquill
Gifford
Stability
Class*
Wind (Approximate)
Speed
(m/sec)
Direction
(deg)
Location
of Ozone
Profile
(See text)
Site 3
Site 2
Site 1
Sice J
Sice 2
Site 1
Location Relative Co the City
Upwind
or Side
X
X
X
Within
the City
Downwind
X
X
X
Unknown
Ozone
Profile
Type
B
B
B
D
A
A
Approximate Ozone
Concentration (ppb)
At
Surface
5
10
10
70
50
50
Average in
Mixed Layer
40
40
30
75
50
50
extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
where the ozone profiles were measured. Sites 1 and 2 are to the
northeast of the Denver Civic Center at distances of about 12 and 18 km,
respectively. Site 3 is about 17 km to the northwest. The determination
of whether or not the sites were downwind of Denver was based on the
analyses presented by Blumenthal et al- (1974). Needless to say,
November is not a month conducive to photochemical activity, but these
data do extend the geographical and seasonal scope of the available pro-
file data.
8. Tampa Bay, Florida
Washington State University measured ozone concentrations at vari-
ous altitudes in the vicinity of Tampa Bay during May of 1976. Some of
the data collected were included in an EPA memorandum from Lonneman
(1977) to J. J. Bufalini. The nature of the profiles included with that
memorandum is summarized in Table 10. It is anticipated that more
detailed information about these profiles and more profiles from the
same study will be published at a later date. However, the limited sam-
ple in Table 10 was included for purposes of statistical analysis.
These data broaden the geographical scope of the total data base.
C. Vertical Cross Sections of Ozone Concentration
in the Northeastern United States
The examples of the vertical distribution of ozone given in the
preceding sections have been limited to isolated vertical profiles taken
at locations directly above or outside a city, either upwind, downwind,
or to one side. Ludwig and Shelar (1977) analyzed data collected by
four groups that participated in the Northeast Oxidant Study Environmen-
tal Monitoring and Support Laboratory, 1975; Siple et al. 1976; Spicer
et al. 1976; Washington State University, 1976; Wolff et al. (1975). The
large numbers of vertical profiles taken at different locations at
approximately the same times allowed Ludwig and Shelar to analyze the
data in the form of vertical cross sections that show the distribution
of ozone concentration in a plane perpendicular to the surface. The
vertical cross sections provide interesting displays of the urban plume
structure. The presence of elevated ozone layers is also seen easily in
this kind of display.
Vertical cross sections are similar to the ozone isopleth maps that
were used to display ground-level concentration patterns in the St.
Louis, Los Angeles, and Houston areas, but the plane in which the con-
centrations are mapped is vertical, as opposed to being the ground sur-
face itself. In the figures presented later in this section, the line
above which a cross section applies has been shown on a map of the
region. The basic data were obtained from aircraft measurements above
points that are marked on the maps. The vertical extent of the aircraft
measurements is indicated by vertical lines on the cross section.
Stable layers have been marked by stippled bars. Wind directions at
.110
-------
Table 10
SUMMARY OF OZONE PROFILES STUDIED FROM THE TAMPA BAY AREA
Dace
760513
760514
Time
(EST)
0823
1229
1423
2134
1049
Mixing
Depth
(km msl)
0.9
0.9
0.8
0.5
1.0
Pasquill
Gifford
Stability
Class*
Wind (Approximate)
Speed
(m/sec)
3
3.5
3.5
5
7
Direction
(deg)
160
200
200
190
150
Location
of Ozone
Profile
SW
SW
N
SW
SW
Location Relative to the City
Upwind
or Side
X
X
X
X
Within
the City
Downwind
X
Unknown
Ozone
Profile
Type
F
F
E
A
F
Approximate Ozone
Concentration (ppb)
At
Surface
10
25
25
20
25
Average in
Mixed Layer
20
25
40
25
35
A = extremely unstable; B = moderately unstable; C = slightly unstable; D = neutral
-------
850 mb (about 1500 m) are shown by stream lines. The times (EST) of the
aircraft measurements are shown on the cross sections. In many
instances, analyses of the ozone concentrations observed at ground level
have been used to extend the analysis of the vertical soundings down to
the surface.
Figure 64 shows two cross sections based on data collected during
the afternoon of August 10, 1975. The stream lines show the 850-mb
winds for 1900 EST. The cross sections are based on data collected
somewhat earlier, between 1545 and 1715 EST. The two cross sections
show that there were layers of high ozone concentration aloft. In the
.TTITim, L.I. BRIDGE LITCHHELD
ALTITUDE soum) poRT
(ft, MSL) 16QO 4Q 16JO EST
ALTITUDE
(ft. MSL)
8000
ATLANTIC L.I. GRCTON PUTNAM
^,1545 1615 IMS 1715 EST
NC \D
0, CONTOURS IN ppb
FIGURE 64 VERTICAL CROSS SECTIONS OVER CONNECTICUT 1545-1715, AUGUST 10, 1975
112
-------
western cross section, concentrations exceed 180 ppb at a height of
about 1300 ft (400 m) above the Connecticut coast. Farther downwind,
ozone concentrations were less. The eastern cross section shows concen-
trations of about 125 ppb. At the southern end of this same eastern
cross section, concentrations aloft exceeded 230 ppb. The southernmost
vertical sounding in the eastern cross section shows that there was a
stable layer at the surface, probably caused by the passage of air over
the colder Atlantic waters. Within the surface-based stable layer, con-
centrations fell below 140 ppb, while aloft they exceeded 230 ppb.
Another stable layer had a base at about 600 m (2000 ft); high ozone
concentrations were confined below the upper stable layer- It is
apparent that the profile at this southernmost point on the cross sec-
tion is a Type E. It is also apparent from these analyses that a Type E
profile can result from the urban "plume," especially when it travels
over a surface that causes the lower layers to become stable, or when
the air passes over an area where the surface emissions of NO deplete
the ozone in the lowest layer.
Figure 65 provides another example of the elevated urban plumes.
The measurements shown in this figure were made during the afternoon of
August 12, 1975. At the northeastern end of this cross section, concen-
trations were relatively uniform in the vertical, about 60 ppb. Farther
to the southwest along the cross section, the plume from Boston becomes
apparent, with concentrations greater than 160 ppb at about 500 m (1650
ft). At the lower altitudes over Cape Cod Bay and the Atlantic Ocean,
the concentrations were less than 80 ppb. This again appears to be a
Type E profile, formed when the lowest layer"of the atmosphere was sta-
bilized by passage over a cool water surface, and the ozone that was
removed at the surface could not be replaced by downward mixing of the
ozone aloft. Part of the plume from New York is evident at the
southwest end of the cross section shown in Figure 65. Concentrations
at about 300 m (1000 ft) exceed 240 ppb, while surface concentrations
are slightly less, around 200 ppb. This suggests that mixing was more
pronounced in the lowest layers in this area than above the cooler water
surfaces, and that this was a Type D profile.
Figure 66 shows two cross sections based on measurements during the
morning of August 21, 1975. This particular date provides an example of
the behavior of urban plumes when an area is under the influence of a
large anticyclone (high-pressure area) to the east. Although the two
cross sections shown in Figure 66 are reasonably close together in
space, the line that is more to the northwest is based on data that were
collected upwind of, or above, the major urban centers. Ozone concen-
trations over central New Jersey exceeded 80 ppb, but no values as high
as 100 ppb were found. In general, concentrations were quite uniform
with height but tended to decrease toward the northeast at all alti-
tudes. The uniformity in concentration with height indicates that this
was a well mixed body of air, representing "background conditions," and
that the vertical profiles were Type A.
113
-------
ATLANTIC OCEAN
ALTITUDE ^-—~—v
(ft, MSL) 1520
8000
CAPE COD BAY
1430
ATLANTIC OCEAN
1315 1258 EST
FIGURE 65 VERTICAL CROSS SECTION FROM LONG ISLAND THROUGH CAPE COD, 1258-1520, AUGUST 12, 1975
-------
ALTITUDE [LCTO*
(ft. KSI.) 0955
1000
KEW
TREHTOM BRUNSWICK
1045 0935 DOVER
1003
BRIDGEPORT
1055
BURLINGTON
0940 1ST
FIGURE 66 VERTICAL CROSS SECTIONS OF OZONE CONCENTRATION, 0815-1055, AUGUST 21, 1975
-------
The other cross section in Figure 66 has a very different appear-
ance. Ozone concentrations as great as 140 ppb were observed in a layer
between 1500 and 2000 ft (450 and 600 m), even though the measurements
were made in the morning before the time when one would expect maximum
photochemical production of ozone. Morning 850-mb wind patterns shown
in the figure suggest that the high concentrations observed over the
Atlantic might have arisen from emissions in the Newark-New York area.
The vertical uniformity of the low concentrations at the northeast end
of this cross section, where it is close to the other cross section,
gives much the same picture of "background conditions" as can be seen in
the other cross section. The high concentrations that were observed at
the southwest end of this cross section are more difficult to explain;
the only metropolitan area lying upwind is Wilmington, Delaware, but it
is also possible that these high concentrations are at the northern edge
of a plume from Baltimore. That explanation would require a more south-
erly component to the air motion than is seen in the 850-mb wind field.
However, surface weather maps show that ground-level winds were light
and from the south-southeast, so it is quite possible that net transport
at 2000 ft (600 m) might have been from the Baltimore area. Ludwig and
Shelar (1977) attribute the lack of evidence for a plume of high ozone
concentration downwind of Philadelphia to the lack of vertical profile
measurements of concentration in the appropriate places, suggesting that
had such measurements been made, they would have revealed evidence of
Philadelphia^ effects.
Figure 67, based on data collected during the midafternoon hours of
August 21, 1975, displays complex patterns aloft that are difficult to
interpret. Concentrations in excess of 200 ppb were observed at about
4000 ft (1200 m) over southern Connecticut. Concentrations in excess of
140 ppb were found over New Jersey, and near the eastern end of the
cross section. Part of the complexity in that data arises from the
rather large time span covered by the data, from 1325 to 1550 EST.
Presumably, the patterns would have been somewhat more organized had the
measurements been made at more nearly the same time. The most important
physical reason for the complexity is probably the light wind conditions
that prevailed. According to Wolff et al. (1975), the early morning
winds this day were light and from north to northeast. The surface winds
shifted to southwesterly during the afternoon. Wolff et al. (1976) have
analyzed this day extensively and concluded that the ozone and precursor
emissions from urban areas such as Philadelphia, Baltimore, and Washing-
ton had accumulated during a period of light winds that lasted from late
evening on August 20 until the early morning hours the following day.
Zeller (1976) believes that the plumes seen in Figure 67 may have come
from Philadelphia, New York, and Hartford, having traveled to the
southwest in the morning and then returned to the north later during the
day. In any event, there was a complex mixture of ozone and precursors
that came from several different urban areas. During the morning, the
mixture was relatively uniform and showed only the effects of the most
recent precursor emissions. As the day progressed, ozone was formed
from the residual precursors and from the more recent emissions, which
resulted in the very complex patterns seen in Figure 67. Figure 67
116
-------
ALTITUDE
(ft, MSL) 1450
8000
NEW
TRENTON BRUNSWICK
1540 1431
LONG IS.
SOUND BRIDGEPORT FISHERS IS.
1325 1550
1540
MARTHA'S
VINEYARD
1445 EST
21 AUGUST 1975 V
\f
FIGURE 67 VERTICAL CROSS SECTION OF OZONE CONCENTRATION, 1325-1550 EST, AUGUST 21, 1975
-------
shows that in a region with many closely spaced areas of precursor emis-
sion, it is not always possible to define vertical ozone profiles in
terms of "upwind" or "downwind". Generally, the characteristics of each
profile can still be fit into the classification scheme described here,
but much of the physical significance that goes with that classification
system is obscured.
D. Summary of Ozone Profile Features
1. General
Earlier sections of this report have described how the pollutant
profiles can be categorized according to their shape and how the shapes
relate to various physical processes operating in the atmosphere. Many
examples were presented and discussed. In this section, we present some
statistics to show when and where certain types of profiles are most
likely to be found. The relationships between surface ozone concentra-
tions and those aloft are also examined in terms of how they relate to
the different categories of profile. The effects of stability, time of
day, and location relative to the city on the surface/mixing-layer con-
centrations are also examined.
All comparisons are based on analysis of the complete data sample-
-that is, the information for all the profiles enumerated in Tables 1,
3, 4, 5, 6, 7, and 9. It should be borne in mind while interpreting the
results of the analysis that there are substantial differences in the
data sets for the different cities. The data sets for some cities
emphasize the morning hours, for others the afternoon; different seasons
are represented and the mixtures of upwind, downwind, and in-city obser-
vations change from one data set to another- Thus, the possibility
exists that results appearing to represent diurnal effects or the
effects of positions relative to the city might in fact have been pro-
duced by differences among the cities. However, the results obtained
are consistent enough with known physical processes that the conclusions
drawn are unlikely to have been substantially influenced by city-to-city
differences on the data sets.
One serious shortcoming remains in the available data. Nighttime,
stable atmospheric conditions are virtually unrepresented. This almost
certainly means that the occurrence of profiles of Type B and Type E is
more frequent than indicated by the data set. Similarly, the data set
contains more examples of profiles taken in areas relatively free of
recent urban influence than it does of profiles taken over the city
itself or in downwind areas. This, too, is likely to bias the estimates
of the relative frequency of occurrence of the different types of
profile.
It must be recognized that the statistics on ozone profile classes
that are presented in this section are not the product of a wholly
objective process. The statistics should be useful for defining the
118
-------
relative frequencies found for the different types of profile under
varying circumstances, but they should not be applied in an absolute
sense to other areas. The classification was done with the physical
factors that govern the shape of the profiles in mind. Thus, there is
apt to be a bias that makes the statistics more as would be expected
from physical principles than would be the case if the classifications
had been made in isolation from a knowledge of the governing factors.
2. Characteristics of the Different Types of Profile
a. General
Of the 268 cases classified, 80 were judged to be Type A profiles.
The next most commonly observed were Type D (68 cases) and Type F (54).
Only 13 Type B and 32 Type E profiles were in the sample. This is prob-
ably a reflection of the sampling bias that exists with regard to time
of day. Profiles were categorized as Type C in 21 instances. The Type C
profile should probably occur downwind of an urban area, but relatively
close to it, and it may well be that vertical profiles were seldom taken
at the appropriate locations.
With regard to time of day, the overall sample includes only 9
cases for the hours between 1800 and 0600 local time. For the hours
between 0600 and 0900, there were 58 cases, and 77 cases between 0900
and noon. During the afternoon, 74 cases were observed between noon and
1500, and 51 cases between 1500 and 1800.
By far the largest number of profiles were measured in locations
judged to be relatively free of recent urban influences; 133 profiles
were measured in such locations. Downwind cases numbered 51, while 67
were taken above the city. This leaves only 18 cases that could not be
classified.
b. Relative Frequency During Different Times of Day
Table 11 is a contingency table showing the frequency at which dif-
ferent combinations of profile type and times of day were observed. The
first number in each square is the number of cases having the combina-
tion of profile type and time of day represented by that square. The
second number shows the percentage of the total number of cases in that
row (time interval) that had the characteristics of that column (profile
type). The third figure in the square shows the percentage of the total
cases in that column (profile type) that occurred with the characteris-
tics represented by that row (time interval). Finally, the last number
in the square shows the percentage of the total cases that had the com-
bination of characteristics represented by the square.
For example, consider the second row and second column of Table 11.
In this instance, the sample contained 12 Type A profiles that occurred
119
-------
Table 11
FREQUENCY OF OCCURRENCE OF DIFFERENT PROFILE TYPES
AT DIFFERENT TIMES OF DAY
BEFORF 603
603-900
| 900-1200
1200-lbOO
en
UJ
UJ
p
1500-1800
eoo-2ioo
2100-2400
COUNT I
ROW PCT I
COL PCT I
TOT PCT I UNCERTAIN
I.I 0
3 I J
I 0
I 0
2. 1 0
I 0
I 0
I 0
3. I 0
I 0
I 0
1 0
4.1 0
I 0
I 0
I 0
IS. I 1
I 2.0
I 10J.O
I .4
6.1 0
I 0
I 0
I 0
7.1 0
I 0
1 O
I 0
- r-_ 1
I A
I 0
3
0
~\
12
20 .7
lb.0
4.b
35
45.5
43.8
13.0
25
33.8
31 .3
9.3
7
13.7
8.8
^ .6
0
0
0
0
1
130. 0
1 .2
.<»
IE
2
to. o
15.4
. 7
c
10.3
46.2
2.2
4 1
5.2
30. t: l
1 . 5
0 ]
0
0
0
0
0
o
0
0
0
0
0
1
50. 0
7.7
.4
PROFILE TYPE
1C
0
0
0
J
0
0
0
0
4
[ 5.2
19.0
1 1.5
I *
t 12.2
^2. 5
3.3
7
13.7
33.3
1
33.3
•4. e
.4
0
0
0 I
0
1
D IE
0 I 2
0 I 50. 0
0 I 6.3
J 1 .7
2 10
3.4 1 7.2
2.9 31 .3
.7 3.7
1 3 9
13.0 11 .7
14.7 I 2H.1
3.7 1 3.3
27 I B
36.5 I 10. d
39.7 I 25. D
10.0 I 3. 0
27 I 3
52.9 I 5.9
39.7 I 9.4
10.0 I 1.1
2 I 0
66.7 I 0
2. ? I 0
.71 0
0 I 0
I J I 0
0 I 0
[ 01 0
_ _ i _ _
IF
0
0
0
0
28
48.3
51 .9
I 0.4
15
19.5
27 .8
S.t>
5
6.8
9 .3
1.9
6
11 .8
1 1 .1
2.2
0
0
0
0
O
0
0
0
HOW
TOTAL
4
1 .b
58
COLUMN
TOTAL
1
,4
80
1 j
4. e
7.8
68
25.3
32
11.9
54
20. 1
21
77
23.6
74
27.5
51
19.0
3
1.1
2
.7
269
100.0
-------
between 0600 and 0900. This represented 20.7 percent of all the data%
collected between 0600 and 0900 and 15.0 percent of all the Type A pro-
files. Finally, 4.5 percent of all the observations were taken between
0600 and 0900 and were of Type A. The first column of the table shows
that the sample includes one case where the ozone profile was not clas-
sified as to type.
Examination of the columns in Table 11 shows that the times at
which the different profile types are observed differed considerably
from one to another- For example, 83.9 percent of the Type A profiles
were observed between 0900 and 1800 local time. By contrast, only 30.8
percent of the Type B profiles were observed during this time period,
and all of those were before noon. Well over 90 percent of the C and D
profiles were also observed during the daytime hours between 0900 and
1800. Type F profiles, which represent those cases with ozone layers
isolated aloft, were far more common during the forenoon, when nearly 80
percent of the cases occurred, than during the afternoon, when the ozone
isolated aloft was likely to be mixed more thoroughly through the mixing
layer. Type E profiles were distributed reasonably evenly through the
day; those that occurred during the late forenoon and afternoon are
probably the result of instances when the strong vertical gradients of
ozone concentration near the surface, associated with this type of pro-
file, were caused by the destruction of ozone by urban NO emissions.
The occurrences of Type E profiles at other hours of the day are more
likely to have been caused by a nocturnal, surface-based stable layer
that isolated ozone aloft, preventing surface destruction.
Table 11 shows the expected differences in diurnal distribution of
the different kinds of ozone profile. Application of the chi-square
test (Nie et al., 1975) to these data indicates that the differences in
the diurnal distributions are highly significant. The chances of
obtaining results like those shown in Table 11 are less than 0.01 per-
cent if there are no differences in the frequencies at which the profile
types occur at different times of day.
c. Relative Frequencies at Different Locations
Relative to the Urban Areas
Table 12 shows the relative frequency at which the different pro-
file types occurred in different locations relative to the city. The
key to the entries is the same as for Table 11. The entries in the
second column of the first row shows that there were 51 observations of
Type A profiles upwind or to one side of urban areas; this represented
38.3 percent of all observations made upwind or to one side of the city
and 63.8 percent of all Type A profile observations.
Table 12 shows the expected relationships. For example, Type A
profiles were observed over 35 percent of the time in locations rela-
tively uninfluenced by the city, but only around 20 percent of the time
in areas above or downwind of the city. Type D profiles were found in
121
-------
only 18 percent of the observations where the profile was believed to be
uninfluenced by the city, but in over 37 percent of those instances
downwind of the city. Similarly, the Type C profile represents a much
larger proportion of downwind profiles than of profiles uninfluenced by
the city.
The differences in the frequency at which the different profile
types occurred in different locations relative to the city are signifi-
cant, according to the chi-square test (Nie, 1975). The probability of
observing these differences would be less than 1 percent if there were
no differences from place to place.
3. Relationship Between Surface Ozone Concentrations
and Those in the Mixed Layer
a. General
One of the major purposes for analyzing the ozone profiles is to be
able to relate surface ozone measurements to those aloft. In this sec-
tion, surface ozone concentrations are compared with those aloft to see
how they vary with ozone profile type, time of day, atmospheric stabil-
ity, location relative to the city, and, finally, the concentration of
the ozone at the surface itself.
Some difficulties arise trying to define average ozone concentra-
tions "aloft." Some upper height limit must be established before it is
possible to estimate the average ozone concentration aloft. The top of
the mixing layer has physical significance and measurements usually were
taken at least to that height or high enough so that one could reason-
ably infer the concentrations that were likely between the top of the
measured profile and the top of the mixing layer. Nevertheless, the use
of the mixing depth to define the region in which average concentrations
should be estimated presents some problems, especially when the lowest
layers are stable or mixing is otherwise confined to a very shallow
layer. In such cases, much higher concentrations were often isolated
above the layer for which the average concentration was determined.
Thus, for profiles of Types E or F (Figure 4), the concentrations deemed
to be the average for the mixing layer may be substantially less than
the concentrations just above the top of the layer. This should be borne
in mind when interpreting the results, especially for Types B, E, and F.
The estimation of the depth of the mixed layer was to some extent
subjective. In most cases there was very little ambiguity. Frequently,
the top of the mixed layer would be marked by the bottom of an inversion
in the lapse rate, the top of the moist layer, and the top of a layer
polluted with aerosols (Bscat) and ozone. Above the layer, the air would
be warmer, drier, and cleaner. Examples of such unambiguous cases can be
seen in Figure 57. For those cases with surface cooling, at night or
during the late afternoon, the top of the mixed layer was judged to
coincide with the top of the inversion or the height at which the ozone
122
-------
Table 12
FREQUENCY OF OCCURRENCE OF DIFFERENT PROFILE TYPES
AT DIFFERENT LOCATIONS RELATIVE TO THE CITY
N>
LO
CJ
01
I
o
>
H
z
o
o
o
COUNT
ROW PCT
CUL FCT
TOT PCT
UPWIND OR
TO ONE SIDE
ABOVE
CITY
DOWNWIND
UNCERTAIN
—
UNCERTAIN
1
.«
100.0
« 4
0
0
0
0
0
0
0
J
0
0
0
0
A
51
30. 3
63 .8
13.0
14
2 J. 9
17.5
6.2
5
17. t
1 1 .2
3. 3
6
33.3
7.5
2.2
PROFILE TYPE
E 1C
5 t <>
j. o i t. e
38. b I 42.9
1.9 I J . 3
4 I 1
o. J I 1.5
30 .8 I 4 . 8
1.5 I .4
4 I 9
7.6 I 17. b
30.8 I 42.9
1.5 I 3.3
0 I 2
0 I 11.1
0 I 9.5
01 .7
I
ID
24
i e. j
35.3
H.9
13
28.4
27.9
7. 1
1 ci
37. 3
27. <>
7. 1
o
33.3
A.-i
2. 2
i
E
19
14.3
59.4
7. 1
9
13.4
2fl. 1
3.3
4
7.3
12. ti
1 .5
J
0
0
0
IF
t 24
I 1H .0
44.4
8.9
20
29.9
37.0
7.4
6
1 1 .8
11.1
2.2
4
22.2
7.4
1 .5
COLUMN
TOTAL
1
,4
29.7
1 3
4 .«
32
25.3
54
20. 1
HOW
TOTAL
1 J3
49.4
67
24.9
51
1 9.0
18
6.7
269
100.0
-------
concentrations cease to increase rapidly with height. The rationale for
this choice is that the effects of the surface, on cooling and ozone
destruction, have not penetrated to higher altitudes. Hence these
heights mark the uppermost points where the air has been influenced
appreciably by the surface through mixing. In many cases, the choice of
the depths of the mixing layer was quite ambiguous and was based solely
on the author's subjective judgment as to what mixing height would be
most consistent with the observed profiles of temperature, humidity (dew
point), aerosol concentration (Bscat), ozone concentration, and, where
available, lidar or sodar records.
b. Differences Among Profile Types
Figure 68 shows the frequency distribution of the ratio of surface
concentration to the average within the mixed layer, for each of the six
different profile types. The numerals at the end of each bar indicate
the number of instances when the ratio fell within the interval defined
below the bar. More than half the Type A profiles had ratios between
0.8 and 1; more than 80 percent were between 0.8 and 1.2. By contrast,
none of the eight Type B profiles had a ratio greater than 0.6. As is
to be expected, Type C profiles are the only ones with an appreciable
skew toward values of the ratio greater than 1. The frequency distribu-
tion of the ratio of surface to mixing-layer concentrations of ozone for
Type D profiles is quite similar to that for Type A profiles. This is
not surprising because the two types of profiles are quite similar in
shape within the mixed layer; their differences are at altitudes above
the mixed layer. This is also true of Type F profiles, and the histo-
grams in Figure 68 are consistent with this fact. Similarly, Type E
profiles resemble Type B profiles within the mixed layer, although they
differ substantially above it. The histogram of the ratios for the Type
E profiles is skewed toward smaller values, as is the histogram for the
Type B profile, although not so pronouncedly.
The type of display shown in Figure 68 can also be used with the
difference between the average mixed-layer concentration and the surface
concentration. Figure 69 shows the distribution of the differences
(mixed-layer concentration minus surface concentration). The three
types of profile characterized by relatively uniform distribution of
ozone within the mixed layer (A, D, F) all show the expected clustering
of difference values around zero. Types B and E show a preponderance of
positive differences, indicating that the average concentration within
the mixing layer was greater than at the surface. Finally, Type C pro-
files show a bias toward negative differences, indicating that surface
concentrations were higher than those aloft.
Table 13 presents some of the same information contained in Figures
68 and 69, but in a different way. This table gives the average ratio
and the average difference values for each of the six profile types. It
is obvious from Table 13 and from Figures 68 and 69 that the surface
124
-------
(a) TYPE A
(b) TYPE B
(c) TYPE C
.4 TO .6
^^^^^ 1
.6 TU .9 1
I 1
• B TO 1. UNDEH .2 1
1
1. TO 1.2 .2 TO. 4 1
I
1.2 TU 1.4 .4 TO .0 I
1
1
0 10 20 JO 4U SO C
FatOUENCY f
(d) TYPE D
N3 1
.b TO .8
.a TO i.
1 . TO 1.2
1 .2 TO 1.4
••» I 1 )
.6 TO .rt
.8 TU 1.
1. TU 1 ..:
1.2 TO 1.4
2 4 b H 100 2 4 C fl 10
BEQUENCr FOEQUKNCY
(e) TYPE E i. i ii (f) TYPE F
\ (JNO€i> .2
1
UNPEH .2 1 .2 TO .4
1
.2 TO .4 1 .4 TO .6
1
.4 TO .6 1
1
. 0 TO .8 1
1
.8 TU | . I
« ( 1 1 4»
1 . TO 1.2 1
1
1
I
.6 TO .a
.8 TO 1 .
( J 1
1. TO 1.2
1 .2 TO 1 .4
0 1 J 20 Jd 40 bO 0 4 8 12 It 20 0 13 20 JO 4 ,> •> J
FQFQUFMrv FOEQUENCY FREQUENCY
FIGURE 68 HISTOGRAMS OF THE RATIO OF SURFACE OZONE CONCENTRATION TO THE AVERAGE CONCENTRATION
IN THE MIXING LAYER
-------
(a) TYPE A
(bl TYPE B
(c) TYPE C
N3
1
•• 1 11
-40 TU -JJ
-<»0 TO 0
a TO 20
j 20 40 tjO no 100
FREOUENCV
(d) TYPE D
1
-60 TO -40
-40 TO -20
-20 TO 0
0 TO 20
20 TO 40
mmmmmmm^mmmmmm^ ) j>
0 1 1 ) 2 0
20 Tl) <|J
40 TU GO
LESS THAN -bO
-60 TO -40
-40 TO -20
-20 TO 0
mmmmmmmmmm ( zt
0 TO 2J
2 4 b d 100
Hd QUCNCY FHEQ
(e) TYPE E
— • ( i)
-40 TO -^0
•• ( 1 )
-20 TO 0
0 TC 20
20 TO 40
40 TO 60
•• ( 1 I
60 TO 30
mm { it
140 TG loO
9 10 20 JO 4O 50 0 4 8 1 ^ Ifc 20
PBEOUENCY FREQUENCY
f
2 4 e e 10
UENCY
(f) TYPE F
••• ( 2)
-40 TO -20
-20 TO 0
0 TO 20
mmmmmm t b)
20 TO 40
» ( I >
4 a m 60
) 10 20 JO 40 50
RECUENCY
FIGURE 69 HISTOGRAMS OF THE DIFFERENCE BETWEEN AVERAGE CONCENTRATIONS OF MIXING-LAYER OZONE
AND SURFACE OZONE
-------
Table 13
AVERAGE DIFFERENCES AND RATIOS BETWEEN OZONE
CONCENTRATIONS AT THE SURFACE AND ALOFT,
FOR DIFFERENT PROFILE TYPES
Profile
Type
A
B
C
D
E
F
No.
of
Cases
79
8
21
66
30
53
Average Difference,
Mixed Layer
Minus Surface Ozone
(ppb)
-0.19
27.50
-21.19
2.42
28.33
9.91
Average Ratio
(surface concentration/
mixed- layer concentration)
1.00
0.38
1.23
0.99
0.65
0.79
concentration is a good measure of the concentration within the mixed
layer when Type A or Type D profiles are present. This will be dis-
cussed in greater detail later.
c. Different Times of Day
As already shown, there are significant differences in the frequen-
cies at which the different profile types occur for different times of
day. Also, the preceding section has shown that the differences between
between surface ozone concentration and the average concentration within
the mixed layer, expected to be associated with each of the different
ozone profile types, do, in fact, occur. It is therefore reasonable to
expect diurnal changes in the ratios and differences between surface and
mixed-layer concentration. Figure 70 shows the frequency distributions
of the differences and ratios for four different time periods. The fig-
ure shows that, as expected, surface concentrations are invariably less
than those aloft before 0900. Throughout the day, the most common ratio
of surface to mixed-layer concentration lies between 0.8 and 1, but as
the day wears on, the relative frequency of surface concentrations
greater than those aloft increases. According to the chi-square test
(Nie et al., 1970), the diurnal
127
-------
1 UNDER .2
™^^"» (
1 ,2 TO .4
TIME: | •* T0 •"
0900 LST ! •" T0 '"
i .a TO i.
1
i
i
0 10
FREQUENCY
1
.2 TO .«
< "• I 21
.4 TO .6
.6 TO .H
TIME:
0900 .a to I.
1 . TO 1.2
•»• < Jl
1.2 TO 1.4
0 10
FREQUENCY
I
1 .4 TO .0
1
•^••••••••B (
I .0 TO -d
I
TIME: { •" T0 '•
1500 LST i '' T° 1<2
••••^••^•v (
I 1 .2 TC 1 .4
I
1
0 10
FREQUENCY
I
• ( 1 >
.2 TO .4
.0 TO .0
TIME:
1 50° 1. TO 1.2 '
1800 LST (
1.2 TO 1.4
FREQUENCY
71 ^^"^^^^ ( 101
-20 TO J
3 TO 20
23 TO 40
40 TO 60
20 JO 40 bO 0 10 20
FREQUENCY
•• ( 151 VM^^ ( J 1
-40 TO -20
-20 TO 0
0 TO 20
m^^^ 1 61
-iO TO 40
20 JO 40 50 0 10 20
FREQUENCY
1
• < 11
t
5 I
91 <
LESS THAN -(, J
^•B ( Jl
-00 TO -40
mmmmmf ( ol
-40 TO -JO
-20 TO 0
0 TU 2J
•^•^^ I 71
20 TO 40
^ < 11
40 TO 60
• ( 11
60 TO 80
• ( 11
140 TO 160
20 JO 40 500 10 20
FREQUENCY
•• 1 II
1 LESS THAN -60
1
111
71
-40 TO -20
-20 TO 0
0 TO 20
•••• ( 45
20 TO 40
• ( 11
40 TO 60
20 JO 40 bO 0 10 20
FREQUENCY
JO 40 bO
"
JO 40 50
JO 40 bO
221
JO 40 50
RATIO — surface/mixing layer
DIFFERENCE — average mixing layer
minus surface ozone concentration
FIGURE 70 HISTOGRAMS OF RATIOS AND DIFFERENCES BETWEEN CONCENTRATIONS
OF OZONE AT THE SURFACE AND IN THE MIXING LAYER,
FOR DIFFERENT TIMES OF DAY
128
-------
Table 14 summarizes the average difference and ratio between sur-
face and mixed-layer ozone concentrations for different periods of the
day. The trend toward greater uniformity in the vertical during the
afternoon is apparent. During the morning hours, from 0600 to 0900, the
surface ozone concentration is only slightly more than two-thirds the
value of the mixed layer, in the mean. By afternoon, virtual uniformity
is found, on average.
d. For Different Stability Classes
It should be apparent from much of the discussion to this point
that the vigor of atmospheric mixing plays a large part in determining
the nature of the ozone profile in the vertical. Thus we might expect
to find the most unstable atmospheric conditions associated with the
most nearly uniform vertical distributions of ozone. Figure 71 gives
strong indications that this is true. For those measurements taken dur-
ing extremely unstable conditions (A stability), the ratios of surface
to mixing-layer ozone concentrations cluster around 1 and the differ-
ences between mixed-layer and surface concentrations cluster around
zero. Stability increases toward Class D (neutral stability) and so
does the frequency of ratios much greater and much smaller than 1.
Large negative and larger positive differences also are more common
under the more nearly stable conditions. The differences among the dis-
tributions of the ratios (when those cases are included that could not
be classified) are significant at better than the 1-percent level. The
differences among the distributions of mixing-layer and surface ozone
concentrations could be due to chance, being significant only at about
the 12-percent level.
Table 15 shows the averages and the standard deviations for the
differences and ratio, stratified according to the stability that pre-
vailed when the measurements were made. The ratio of surface to mixed-
layer concentration varies with stability class, but the standard devia-
tion shows a monotonic increase from the most unstable (A) to neutral
(D) conditions. Similar, but much less pronounced, tendencies are evi-
dent in the difference tabulations.
e. For Different Locations Relative to the City
Although the ratios and differences discussed in preceding sections
might be expected to differ from one location to another (relative to a
city), it turns out that the differences are not very pronounced. For
the ratio of surface ozone concentration to concentration aloft, the
differences among the frequency distributions upwind, downwind, and over
the city are totally insignificant. The distributions of differences
between ozone concentration aloft and that at the surface have signifi-
cant changes from one location to another, but only at the 10-percent
level. Table 16 shows the average ratios and average differences, and
it can be seen that the surface ozone concentrations tend to be about 90
129
-------
Table 14
AVERAGE DIFFERENCES AND RATIOS BETWEEN OZONE CONCENTRATIONS
AT THE SURFACE AND ALOFT, FOR DIFFERENT TIMES OF DAY
Time of Day
(LST)
0600-0900
0900-1200
1200-1500
1500-1800
No.
of
Cases
57
73
72
49
Average Difference,
Mixed Layer
Minus Surface Ozone
(ppb)
13.6
4.7
1.9
-0.4
Average Ratio
(surface concentration/
mixed-layer concentration)
0.69
0.92
1.02
1.02
-------
STABILITY
CLASS
A
B
C
D
1
«• ( l )
1 .4 TO .6
1 6 TO 6
1
1 . « TO 1 .
1
1 1 . TO l . -t
\
1 1 2 TO l . •>
1
1 1 1 1 1 , .
o 4 e 12 IE
FREQUENCY
I
•• ( 1 )
1 .2 T8 .4
1
•^-^ ( 8)
1 . 4 TO . 6
1
1 .6 TO .B
1
1 . 8 TO 1 .
1
1 1 . TO 1 . 2
1
1 1 . 2 TO 1 . 1
1
1
1
0 10 20 30 40
FREQUENCY
1
•• ( 1 )
1 UNDER .2
1
1 .2 TO . «
1
1 .4 TO .6
1
1 .6 TO .e
1
1 . » TO 1 .
1
1 1 . TO 1 . 2
1
1
0 10 20 30 40
FREQUENCY
1
•» ( 1 )
1 UNDER .2
1
1 .2 TO .4
1
1 .4 TO .6
1
1 .6 TO .B
1
i . a TO i .
1 1 . TO 1 . Z
1
1 1 . 2 TO 1 . 4
I
1
1
I . .1 1 1 | ...
0 4 8 12 16
FREQUENCY
I
1 -20 TO 0
1
1 0 TO 20
1
1 20 TO 10
1
... I 1 .... . 1 .... 1
20 o d e
FREQUENCY
1
VI 1 )
1 LESS THAN -60
1
^ < 2>
1 -60 TO -40
[
1 -40 TO -20
1
1 -20 TO 0
1
1 0 TO 20
1
«^^» ( 4)
I 20 TO 40
I
1
1
. . I | 1 1
90 0 10 20
FREQUENCY
1
«« { 1 )
1 -60 TO -40
— ( 3)
1 -40 TO -20
1
1 -20 TO 0
I
1 0 TO 20
1
1 20 TO 40
1
1 40 TO 60
I
«• ( 1 )
1 60 TO 60
1
1
1
50 0 10 20
FREQUENCY
1
-40 TO -20
-20 TO 0
0 TO 20
20 TO 40
40 TO 60
20 0 4 a
FREQUENCY
. 1 1 . 1
12 16 20
. . 1 1 1
30 40 30
. . . | 1 1
30 40 50
12 16 20
RATIO — surface/mixing layer
DIFFERENCE — average mixing layer
minus surface ozone concentration
FIGURE 71 HISTOGRAMS OF RATIOS AND DIFFERENCES BETWEEN
SURFACE AND MIXING-LAYER OZONE CONCENTRATIONS
FOR DIFFERENT ATMOSPHERIC STABILITY CLASSES
131
-------
Table 15
AVERAGES AND STANDARD DEVIATIONS FOR THE DIFFERENCES AND RATIOS
BETWEEN OZONE CONCENTRATIONS AT THE SURFACE AND ALOFT, FOR
DIFFERENT STABILITY CLASSES
to
Stability
Class*
A
B
C
D
No.
of
Cases
27
81
68
50
Difference, Mixed
Layer Minus Surface
(Ppb)
Average
6.5
-1.1
10.1
3.5
Standard
Deviation
12.5
16.0
19.4
16.5
Ratio
(surface concentration/
mixed-layer concentration)
Average
0.92
0.99
0.86
0.90
Standard
Deviation
0.18
0.24
0.25
0.31
*A = extremely unstable; B = moderately unstable; C = slightly unstable;
and D = neutral.
-------
Table 16
AVERAGE DIFFERENCES AND RATIOS BETWEEN OZONE CONCENTRATIONS
AT THE SURFACE AND ALOFT, FOR DIFFERENT LOCATIONS
RELATIVE TO THE CITY
CO
co
Location
Relative
to City
Upwind or
to one side
Above city
Downwind of
city
No.
of
Cases
130
60
49
Average Difference,
Mixed Layer Minus
Surface Concentration
(ppb)
1.9
13.6
5.1
Average Ratio
(surface concentration/
mixed-layer concentration)
0.94
0.87
0.87
-------
percent of those in the mixed layer, regardless of location relative to
the city. However, on average, the surface ozone concentration in the
city is more than 13 ppb less than that averaged through the mixing
layer. Outside the city, upwind or downwind, the surface ozone concen-
trations much more nearly approximate those averaged through the mixing
layer. If the ratios remain about the same, but important changes occur
in the difference, this suggests that the absolute values must be moving
more or less in concert with the differences. This, in turn, suggests
that the average concentrations above the city are greater than those
outside it. For this sample, that is true, but it has little physical
significance, because it simply reflects the dominance of Los Angeles
observations in the subset of observations taken above a city. Very few
Los Angeles observations were classified as being other than above the
city.
f. Relations Between Surface Concentrations
and Those Aloft for Different Profile Types
Figure 72 presents scattergrams of surface ozone concentration
versus the average through the mixing layer, for each of the six dif-
ferent types of profile. The least-square fit linear regression lines
and the linear correlation coefficients are also shown. If more than
one case had the same combination of surface and mixing-layer concentra-
tions, then a numeral was plotted to denote the number of cases
represented by the point. As is to be expected, Figure 72 shows that
mixing-layer ozone concentrations are highly correlated with surface
ozone concentrations for those profiles classified as either Type A or
Type D. Also as expected, the slopes of the regression lines are near
unity, with small values for the intercept. The correlation and the
slope of the regression line should be as they are because the profiles
were classified "A" or "D" when the surface and mixing layer concentra-
tions were about the same. It is important to note that such conditions
were not at all unusual.
Although the Type C profiles have high correlations between concen-
trations of surface and mixing-layer ozone, the slope of the regression
line is not near 1. Type F profiles exhibit good correlation between
surface and mixing-layer concentrations. The slope of the regression
line is near unity but its intercept is fairly large. The remaining two
types of profile, Types B and E, do not show as high a correlation
between surface concentrations and those averaged through the mixing
layer as do the other categories, although the relationship in the case
of Type E profiles is still sufficiently strong (correlation coefficient
= 0.9) to be quite useful.
13-4
-------
SURFACE OZONE CONCENTRATIONS — ppb
I.) TYPE E PROFILE
(«) TYPE f PROFILE
FIGURE 72 SCATTERGRAMS OF MIXED-LAYER AND SURFACE OZONE CONCENTRATIONS
FOR THE SIX TYPES OF VERTICAL PROFILE
Figure 73(a) shows a scattergram of surface ozone concentration
versus that averaged through the mixing layer for the combined subsets
of data for profiles Type A and D. Obviously, if one could identify
those situations where the vertical profile of ozone fell in one of
these two categories, it would then be quite legitimate to infer that
the ozone concentration averaged through the mixing layer was essen-
tially the same as that observed at the surface. The regression line
135
-------
.D
Q 166.00
2 1^7.00
o
H
< 06.00
cn
o
o
u
in
2
O
N
O
t'5.0') ('j.Od lOb.OJ 1*5.00 l-!t.JO ^.Jt.JO ^O'J.OO Job.10 J«b.OO
V 1.015x - 0.33
CORRELATION = 0.98
(a) PROFILE TYPES A AND D
y = 0.981 x + 4.81
CORRELATION = 0.98
12?* 00 |t5. (JO £Cv»UO i
SURFACE OZONE CONCENTRATION — ppb
(b) PROFILE TYPES A, D, AND F
FIGURE 73 SCATTERGRAMS OF AVERAGE MIXING-LAYER OZONE CONCENTRATIONS
VERSUS SURFACE CONCENTRATIONS FOR PROFILE TYPES ASSOCIATED
WITH A WELL-MIXED SURFACE LAYER
136
-------
shown in the figure has a standard error estimate for the averaged
mixing-layer ozone concentration of only about 11.4 ppb. It is impor-
tant to note that the predictive abilities of the regression line are
not limited to those cases where the ozone concentrations were low. It
is obvious from Figure 73(a) that the high concentration values are also
well defined by the regression line.
Profile Types A and D are associated with an atmosphere that is
well mixed in the layers near the surface. Profile Type F is also asso-
ciated with strong mixing in the low layers, but with the difference
that the mixing has not penetrated vertically to altitudes as high as
those to which earlier pollution has been mixed. Figure 73 (b) shows
that the inclusion of Type F profiles in the data set causes very little
change in the regression line or the correlation. Thus, it is apparent
that surface ozone concentrations measured under Type F conditions are
representative of values through the mixing layer, but it should be
remembered that this representativeness is limited to the current
mixing-layer depth and that there will be higher ozone concentrations
farther aloft, which are likely to be mixed downward later.
The results shown in Figure 73 would be extremely useful if it were
possible to identify those times and places where one could confidently
expect to find profiles of Types A and D, or even Types A, D, and F.
Table 11 shows that the odds of encountering such profiles are improved
considerably when the search is limited to daylight hours after 0900
local time. Between 0900 and noon, 58.5 percent of the cases are Types A
or D and 78 percent are Types A, D, or F. Between noon and 1500, 73.3
percent are Types A or D, and 80.1 percent are observed to be A, D, or
F.
Types B and E occur when surface destruction is not offset by ver-
tical mixing. This can happen either when there is a strong stable
layer at the surface or when surface destruction is very strong. For
daytime hours after 0900, the presence of a strong stable layer will
generally be unlikely, particularly on days sunny enough to be photo-
chemically active. For the daytime hours, Types B and E will probably be
associated with the strong destruction of ozone by NO that occurs in the
city. Therefore, if only profiles taken outside the city were con-
sidered, we might expect the proportion of profile types B and E to be
reduced relative to other types during the late forenoon and early
afternoon hours, but this is not the case during the late forenoon, when
there is no change. During the early afternoon, the proportion of Type
E profiles (there are no Type B profiles during this time period) is
about 10.8 percent in the total sample and 6.1 percent among the pro-
files taken outside the city. The expected effect is present, but not
very pronounced.
Even if profiles measured above the city are excluded, this does
not appreciably alter the chances of finding a profile with uniform
ozone concentrations through the mixing layer. Figure 74 is a scatter-
gram relating mixing-layer ozone concentration to surface concentration
137
-------
* £• 2 •* • •
b* 2»
• • *Jti 42to
2*2 »2 2 2
• a 3 ••
2 2.122 •• • •
• 2
FIGURE 74 SCATTERGRAM OF AVERAGE MIXING-LAYER OZONE CONCENTRATIONS
VERSUS SURFACE CONCENTRATION FOR ALL MEASUREMENTS
BETWEEN 0900 AND 1500 LST
for all the observations made between 0900 and 1500 LST. Obviously, the
relationship is close enough to be quite useful. The regression line
shown in the figure estimates average mixing-layer concentrations with a
standard error of about 21 ppb.
Probably an astute analyst could use readily available meteorologi-
cal observations and radiosonde data to select cases with sufficient
mixing to ensure that the surface ozone concentration was representative
of that throughout the mixing layer- Some attempts were made to select
data points from among those shown in Figure 74, on the basis of atmo-
spheric stability class and location relative to the city. When only
those profiles measured during periods when the atmosphere was
moderately or extremely unstable are included, the standard error of the
regression estimate drops from about 21 ppb to about 13 ppb. However,
the correlation is not as good, being reduced to 0.87 from 0.91. Fur-
thermore, the relationship between surface and mixing-layer ozone con-
centrations in the data set for unstable atmospheric conditions is not
one of near-equality, as it is in Figure 74. The slope of the line in
the unstable data set is only 0.7 (as opposed to 0.96), and the inter-
cept is 21 ppb (as opposed to 7 ppb).
138
-------
E. Summary gf_ Profile Features for Other Photochemical Pollutants
It should be apparent from the preceding discussions that one of
the major factors in the determination of vertical profile shapes for
ozone is the destruction process. In general, destruction takes place
near the surface by reaction with NO, or with natural, gas-phase
destructive agents (Vukovich, 1973), or by contact with the surface
itself. Destruction is largely responsible for the formation of pro-
files of types B, E, and F. If there were no destruction near the sur-
face, the type A profile would not evolve to a type B, and the type D
would not evolve to type E. Thus, if a pollutant has no sink near the
surface it would be unlikely to find profile shapes like type B or E.
If instead of a sink at the surface, a pollutant has sources there,
then one might expect to find a much larger percentage of profile types
C and D. If the pollutants emitted at the surface were thoroughly dis-
tributed through the mixing layer, then the profile would be of type D.
If the mixing were not complete, then the profile would be type C. Most
of the major pollutant groups, e.g. NMHC and NOX are not destroyed at
the surface. In general, their sources are near the surface. Unfor-
tunately, only a few reliable profile measurements have been made for
these pollutants, so it is not possible to develop a comprehensive sta-
tistical description of observed profile types as was done for ozone.
Nephelometer measurements have been made more extensively than
those for NMHC and NO . Nephelometer measurements give an indication of
the concentration of submicron particles in the atmosphere. If it is
presumed that the aerosol source and behavior are similar to the pollu-
tants of interest, then the profiles of nephelometer readings can be
examined as surrogates for the other pollutants. Even when this is
done, the available data are still fewer than for ozone, partly because
fewer airborne nephelometer measurements were made, and partly because
ground based nephelometer measurements that would allow us to extrapo-
late the profiles to the ground are not available.
The useable profiles of nephelometer readings from St. Louis,
Washington, D.C., and Los Angeles have been examined; 116 profiles were
subjectively classified according to shape. None of the profiles fell
into either category B or E. This may be because the lowest part of
many profiles were not available, but it is probably due to the lack of
strong sinks near the surface.
Table 17 summarizes the joint frequencies of occurence of various
combinations of ozone and nephelometer profile types. The most obvious
thing about the table is that type D profiles are by far the most com-
monly occurring type. In essence, this shows that pollutants that are
emitted near the surface, but that have no appreciable sinks there, tend
to be uniformly mixed through the boundary layer. The differences
139
-------
Table 17
OZONE PROFILE TYPES OBSERVED
WITH VARIOUS NEPHELOMETER PROFILE TYPES
Nephelometer
P ret f i 1 p
Type
A
C
D
F
Ozone Profile Type
A
6
2
17
1
B
0
0
3
2
C
0
0
6
0
D
0
1
31
1
E
0
0
9
5
F
1
0
23
8
Total
7
3
89
17
between the shapes of the ozone profiles and the concurrent nephelometer
profiles are generally explainable in terms of the differences between
pollutants that have a source at the surface and those that have a sink.
For example, when the ozone profile is type A. nearly all the nephelom-
eter profiles were either the same type (A) or type D, indicating that
aerosol material had been released into the boundary layer although no
ozone had been produced recently within the boundary layer. When the
ozone profile was of type B, the nephelometer readings were of type D
because aerosol had been added, rather than removed at the surface. In
some cases, the type B ozone profiles were accompanied by type F
nephelometer profiles, which suggests that there may have been some
removal of aerosol near the surface, but not as much as for ozone.
All ozone profiles of type C were accompanied by type D nephelome-
ter profiles. This indicates that in those cases the mixing of aerosol
through the boundary layer was more complete than it had been for ozone
and its precursors. Nearly all the type D ozone profiles were accom-
panied by type D nephelometer profiles. In one case, the nephelometer
profile was of type C, indicating that the aerosols had not been as com-
pletely mixed as the ozone, but this was quite unusual. Similarly,
there was only one case where the nephelometer profile was type F when
the ozone profiles was type D.
As was noted before, the type E ozone profiles evolve from type D
when ozone is destroyed near the surface at a rate faster than it can be
mixed downward from aloft. Thus, for those pollutants that are not
destroyed near the surface, the profile should remain a type D. Table
17 shows that most of the type E ozone profiles were accompanied by type
D nephelometer profiles, as expected. In five instances, the nephelome-
ter profile was of type F which suggests that there may have been some
losses within the mixing layer, but not severe enough to produce a type
E shape profile. Finally, Table 17 shows that when ozone was found
140
-------
isolated in a layer aloft, in a type F profile, the vast majority of the
accompanying nephelometer profiles were of type D. Again, this is a
reflection of the lack of strong sinks for the aerosol.
The question remains, do the precursors, NMHC and NO , behave in
the same way that the nephelometer observations of aerosol concentration
behave? Eleven NMHC profiles were available for comparison with
nephelometer profiles from the Los Angeles basin. In all cases, the
nephelometer profile was of type D. In seven instances, the NMHC pro-
file was the same. Three of the remaining four were type C, indicating
that the NMHC had not been mixed as thoroughly in the boundary layer as
had the particulates. The one remaining case was type A, as was the
ozone profile at the time.
There are sixteen NOX profiles from the Los Angeles basin that can
be compared to nephelometer profiles; fourteen of these nephelometer
profiles are of type D. Two of the NOX profiles associated with type D
ozone profiles are type C and one is type A. There is another type D
profile of NOX associated with a type F nephelometer profile. Finally,
there was one instance when both the nephelometer and NO profiles were
of type F.
In summary, the nephelometer profiles seem to behave much like the
precursor profiles. Both categories of pollutants most commonly display
profiles that are type D, as one would expect of materials with sources
rather than sinks at the surface. Type D profiles indicate that the
precursors emitted at the surface are rather quickly mixed within the
boundary layer and achieve a reasonably uniform concentration profile.
Under such conditions, precursor measurements made at the ground should
be reasonably representative of concentrations existing aloft. It
appears that if conditions are such that the ozone profiles are of type
A or type D, then one can reasonably expect surface precursor observa-
tions to be representative of conditions through the boundary layer. It
should be understood, that the precursors referred to are NMHC and NOV.
The individual oxides of nitrogen, such as NO and N02, may be involved
in reactions with ozone. In such cases, the processes that reduce ozone
concentrations near the surface might also reduce the NO concentration
and increase the N02 concentrations in the lowest layers.
The above analysis is necessarily tentative and uncertain because
of the very few measurements that are available. Physical reasoning
suggests that the dust concentrations represented by the nephelometer
data should behave in a manner similar to other pollutants that are not
formed or removed rapidly. Nevertheless, more data from future field
studies will be needed before one can be certain that the nephelometer
data do serve the intended role as surrogates for the behavior of oxides
of nitrogen and hycrocarbons.
141
-------
IV VARIATIONS IN THE MIXING LAYER AND WINDS NEAR CITIES
A. Background
The four most important meteorological factors governing ozone con-
centrations in the vicinity of cities are probably transport by the
wind, availability of sunshine, mixing and dilution by turbulent diffu-
sion, and restrictions to the vertical extent of mixing. In general,
the transport winds can be inferred from surface wind measurement; pilot
balloon observations of winds aloft are also sometimes available; and
weather service analyses of conditions at the 850-mb level (approxi-
mately 1500 m) are also useful for determining transport, but the city
will change the wind during its passage and these changes should be con-
sidered. Available sunshine can usually be estimated from conventional
cloud cover observations and time of day- Similarly, atmospheric sta-
bility can be estimated from surface weather observations (e.g. Turner
1964; Ludwig and Dabberdt, 1976), but the atmosphere will be less stable
over the city than over its environs because of differences in surface
roughness and heat capacity (see, e.g., Ludwig and Dabberdt, 1976).
While most of the important meteorological factors governing the
distribution of chemical pollution can be inferred from the conventional
weather observations available throughout the day (often at hourly
intervals), the mixing depth is not so easily obtained. Its determina-
tion generally requires radiosonde observations, which are made only
twice daily and at rather widely spaced locations. Therefore, it is
important to have some understanding of how the mixing depth varies in
time and space, so that the rather limited radiosonde observations can
be usefully interpreted. This section discusses the spatial and tem-
poral changes in the mixing layer and winds.
B. The Mixing Layer
Table 18 shows how mixing depth was distributed during the dif-
ferent time intervals for the data sample considered in this report.
For the nighttime hours, the observed mixing depths were always below 1
km. During the day, greater mixing depths were more common. During the
night hours, two-thirds of the cases were less than 500 m. Between 0900
and 1200, only one-third of the cases were less than 500 m, and in the
afternoon the proportion of mixing depths less than 500 m was less than
one-sixth. Correspondingly, the numbers of cases of mixing depths in
the 1500- to 2000-m range increased during the day from a small fraction
before 0900 to about one-fourth during the afternoon. A chi-square test
(Nie et al., 1975) of the data shown in Table 18 indicates that the
differences in the frequency distribution from one time of day to
another are highly significant; the chance probability is less than 0.01
percent.
143
-------
Table 18
MIXING DEPTH AT DIFFERENT TIMES OF DAY
MIXING
DEPTH — km
LESS THAN
.5 TO 1
1 TO 1.5
1.5 TO 2
2 TO 2.5
2.5 TO J
COUNT
ROW PCT
COL PCT
TOT PCT
1.
.6
2.
3.
4.
5.
6.
:
1
—
flEFQRE t
00
1.
2
2.3
50.0
.8
2
a. a
50.0
.8
0
0
0
0
a
0
0
0
0
0
0
0
0
0
0
0
r
tOO-900
2.
40
4t> . G
71 .4
15. 8
1 1
15.3
19. fa
4.3
3
7.0
5.4
1 .2
2
4.1
3.6
.6
—
0
0
0
0
0
0
0
0
1
<;oo- i2cu
3.
26
,:9.9
35.6
10. 3
21
29.2
2-3.C
8.3
1 1
25. fc
15.1
4.3
15
30.6
20. 5
5.5
0
0
0
0
0
0
0
I 0
TIME PERIOD
1230-1 tO
0
•+.
1 1
1 2.6
16.4
4.3
26
36. 1
30.8
10.3
1 1
25.6
16.4
4.3
18
3fc.7
26.9
7. 1
0
0
3
0
1
100.0
1.5
.4
i e JG-I eo
0
5.
b
fc. 9
12.2
2.4
12
It. 7
24.5
1 . 7
1 8
41.9
36.7
7. 1
13
26.5
26. 'j
5. 1
0
0
0
0
0
0
0
0
i a jo-2 1 o
0
6.
0
0
0
0
0
0
0
J
0
0
0
0
1
2.0
bO. J
.4
1
100.0
50.0
.4
0
0
0
[ 0
I
2100-240
0
7 .1
2
2.3
100.0
.8
0
0
0
0
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0
0
COLUMN
TOTAL
4
1 .6
56
22. 1
73
28.9
67
26.5
49
19.4
2
,8
2
,8
ROM
TOTAL
87
34.4
72
28- b
43
17.0
49
19.4
1
. 4
1
,4
253
100.0
-------
There is some suggestion in Table 18 that the transitions between
the low nighttime mixing depths and the higher daytime values and back
again to the low nighttime values may occur over rather short periods of
time. Case studies verify this. Auer and Eaton (1976), Endlich et al-
(1978) and Uthe et al- (1978) have examined lidar observations taken
during the summer in St. Louis, and provide many examples of transition
from nighttime to daytime conditions. Figure 75, from the report of
Endlich et al. (1978), illustrates a typical case. The figure shows the
distribution of backscatter in the vertical, which is a measure of the
distribution of aerosol. Figure 75(b) shows the vertical gradients in
the backscatter, with the bright areas representing the largest negative
gradients. The increasing depth to which the aerosol was mixed later in
the day is apparent in the figure. The heavy line in Figure 75 (c) marks
the climb of the mixing depth during the morning. From about 1000 to
1115 CDT, the depth of the mixing layer rises from about 350 m to about
1400 m, approximately the height to which aerosol from preceding days
had been mixed, as is evident in the figure. Figure 76 is another exam-
ple of the temporal changes in mixing depth over St. Louis. The rises
in mixing depths seen in Figures 75 and 76 are rather typical of what
has been found in the analyses of St. Louis lidar data.
A major part of the increase from low nighttime mixing depths to
the higher daytime values comes over a period of 1 or 2 hours during the
late forenoon. The rise then usually continues more slowly until later
afternoon or early evening. Around sunset, as seen in Figure 76, there
is often a very rapid transition to the nighttime, shallow mixing-layer
conditions. The transition to nighttime conditions is almost discon-
tinuous, whereas the increase in mixing depth usually proceeds over a
period of a few hours as the ground heats during the morning and the
resulting convection pushes farther and farther upward. At night, the
ground cools and a stable layer forms very quickly in the lowest layers.
This stable layer damps out the convection so that a new, very shallow
mixing layer is established within the previously formed mixing layer.
Figures 75 and 77 (from Endlich et al., 1978) provide good examples of
this. There is evidence throughout Figure 77 that mixing sometime in
the past extended above 1800 m, but the more recent afternoon mixing had
carried pollutants only to about 1500 to 1600 m. By 2030-2045 CDT, vir-
tually all the pollutants emitted at ground level were confined within a
very shallow layer, no more than 300 m deep. The objective analyses of
Endlich et al. (1978) shown in Figure 77 (c) shift abruptly from the
higher mixing layer to the lower one during this period ot time.
The diurnal behavior of mixing depth discussed above was measured
in the city of St. Louis. Within a city, nighttime mixing tends to be
more vigorous and extends to greater heights than it does outside the
city, because of the heat generated and stored in urban structures
(Tyson et al., 1973). Uthe et al. (1978) operated a mobile lidar system
in St. Louis to study differences between mixing heights outside the
city and those over the city. The lidar was operated along the traffic
145
-------
3.0
2.4
S 1-2
r
0.6
3.0
2.4
o
< IB
I
O
1.2
0.6 —
I I I I I I I I I I I I
' I I I I I I I I ' ' |
0933 0953 1025 1045 1113
TIME, CDT
1133
FIGURE 75 GRAY-SCALE DISPLAY OF LIDAR DATA FOR AUGUST 3, 1976 FROM 0929
TO 1145 SHOWING THE CHANGES PRODUCED BY THERMAL MIXING.
(a) Backscatter data, (b) Vertical gradients, (c) Objectively identified negative
gradients. A heavy line indicates the largest negative gradient, which is the mixing
depth.
146
-------
1400
1200 —
0600
0800
1000 1200 1400
T1Mg - CST
1600
1800
2000
FIGURE 76 DIURNAL CHANGES IN MIXING DEPTH IN ST. LOUIS, AUGUST 9, 1976
147
-------
3.0
2.4
3
5 1.2
z
0.6
MC?< l*t TO OV
3.0
2.4
2 1.2
tu
0.6
I I I I I I
Me)
1 I r r
111 ij i i
i i i i
2004 2024 2048 2108
TIME. CDT
FIGURE 77 GRAY-SCALE DISPLAY OF LIDAR DATA FOR AUGUST 3, 1976
FROM 1958 TO 2124 SHOWING ESTABLISHMENT OF THE MIXING DEPTH
AT A LOW LEVEL
(a) Backscatter data, (b) Vertical gradients, (c) Objectively identified negative gradients.
148
-------
route shown by the heavy line between Site 2 and Site 4 in Figure 78.
Auer and Eaton (1976) analyzed the data as contours of mixing depth on a
time/distance mapping. Their analysis indicated that the rise in the
urban mixing depth precedes the rise in the rural mixing depth by thirty
to forty-five minutes. Earlier studies by Uthe and Russell (1973) indi-
cated that the morning rise of the rural mixing depth lagged behind the
rise in the city by one to two hours.
Uthe et al. (1978) have used the results of Auer and Eaton (1976)
to determine minimum and maximum mixing depths observed along the tran-
sect line shown in Figure 78. The minima and maxima were plotted as
functions of time for eight different days. The results are shown in
Figure 79. The numbers near the data points on the curves are the dis-
tance (km) from the Gateway Arch at which the minimum or maximum was
observed. Positive values are east of the Arch and negative values are
to the west. The Arch is just east of the central business district of
St. Louis and near the Mississippi River. The metropolitan area extends
east of the Mississippi River to include East St. Louis and other Illi-
nois suburbs.
Inspection of Figure 79 shows that nearly all the maximum mixing
depths were in urban areas, within about ten kilometers of the Arch.
The minimum mixing depths were, in almost all cases, in the more rural
locations well removed from the Arch. The difference between maximum
and minimum mixing depth, i.e. between rural and urban mixing depth,
tends to increase during the daytime; the greatest differences are
observed in the afternoon. The average difference between all maximum
(urban) and minimum (rural) mixing depths was computed to be 203 meters.
Because Uthe et al. (1978) used the highest and lowest values along the
lidar route, the average that they obtained probably overestimates the
difference between rural and urban conditions. Figure 80 shows Auer and
Eaton's (1976) analysis of the same data, using more typical urban and
rural values. Figure 80 shows that the mixing depths over the city are
greater than over the rural areas, both during the daytime and at night
and that this difference increases with time of day to an average of
about 160 or 170 meters. There are some indications that the mixing
depths later in the afternoon are more nearly constant than indicated by
the figure. An empirical formula exists (Ludwig, 1970) for estimating
the depth of the mixing layer over a city, using rural observations of
the lapse rate in the lowest layers of the atmosphere. It was derived
from an empirical relationship between the magnitude of the urban heat
149
-------
Mlaiaippi River
FIGURE 78 LOCATION AND ELEVATION OF THE LIDAR ROUTE USED
DURING THE 1975 METROMEX STUDY (Uthe et al., 1978)
150
-------
IK
16
14
!»
i
E 10
1
HEIGHT
00
MIXING
O)
4
2
0
1 1 1 1 1 1
28 JULY 1975 ^ _
•2 / . -9 -
7 -i -
./'/'
.//•'"
-2/ /"10
-«/ /1 3
2>X ^
- '• xx;^"13
•x+7*8
— +7 —
I I I I I I
1 I 1 1 1 1
_ 31 JULY H78
-
•8 ~
•10. /
y
.,„/ ^
*3/ x/iV2
•«+2./ /-13
- /^A"
./ /-
_ x'-13
-Vs""13
1 1 1 1 1 1
08
10 12
HOUR OF DAY - CDT
14
08
10 12
HOUR OF DAY -
CDT
FIGURE 79 LIDAR-OBSERVED MAXIMUM AND MINIMUM MIXING DEPTHS
AS A FUNCTION OF TIME (Uthe et al., 1978)
Positive numbers indicate distance (km) east and negative numbers
west of the Gateway Arch
14
-------
18
16
14
8,,
X
10
(3 O
iu
z
o
x 6
I
_ 7 AUGUST 1978
2 -
08
•8 x
A"
Xe
+6
1
1
1
10 12
HOUR OF DAY - CDT
14
8 AUGUST 1976
+ 10+* +•
*•/•
+13
1
1
1
08
10 12
HOUR OF DAY - CDT
14
FIGURE 79 LIDAR-OBSERVED MAXIMUM AND MINIMUM MIXING DEPTHS
AS A FUNCTION OF TIME (CONTINUED)
Positive numbers indicate distance (km) east and negative numbers
west of the Gateway Arch
-------
18
16
14
8
x 12
§
* 10
1
£
2 8
Ul
I
0
X 6
I
Ol
4
2
1 1 1 1 1 1
9 AUGUST 1975
-
"^
-7 -7 -7 .5 .4
- -a"?-'"""" ' ""
-9/'X
>2'7/
— * / ^ ""
• --* i"** < *&* « """*
/ «^*ii^ 4- 13+13
•»/ X+13+13
/ X+13
-8//;#3
/13
• 13
—
1 1 I 1 1 1
! 1 I 1 1 1
_ 10 AUGUST 197B
~ -
—
~ —
-3.
"~ -3 "^"' —
-2 /
^*
_ -a/ ^X;i3
-------
10
16
14
8
N 0
. x ueiaui _
HEIGHT
00
O
X 6
4
2
0
01
1 1 1 1 | 1
_ 11 AUGUST 1975 _
— —
~ +6 ~"
...
+8 '
*tV /+13
*-y' /;»'
. */'' X+13
— -*^' i'"*'14 ~
*13'M4+14
I I I 1 I I
J 10 12 14
0
1 ' ' ' ' .,?'•
12 AUGUST 1976 .j /' _
J
- —
I / *14
-2 / /" *14
/ /
- *}/-
•9 //••>*
+ 11.-.'x'^14"14
•14
— -
1 I I 1 I I
8 10 12 14
HOUR OF DAY - CDT
FIGURE 79 LIDAR-OBSERVED MAXIMUM AND MINIMUM MIXING DEPTHS
AS A FUNCTION OF TIME (CONCLUDED)
Positive numbers indicate distance (km) east and negative numbers
west of the Gateway Arch.
-------
— — — — URBAN
RURAL
07
08
09 10 11 12
LOCAL TIME—hours
SOURCE: AUER AND EATON, 1976
13
14
FIGURE 80 AVERAGE INCREASE IN URBAN AND RURAL INFLUENCES
ON THE DEPTH OF THE MIXING LAYER NEAR ST. LOUIS
155
-------
island and rural lapse rate, using Summers' (1965) simple heat island
model. The relation is
h=29.3^°'25(0.298dT/dp-0.0633)/(pdT/dp-0.287)
where
<| = population of the city
p = average pressure through the depth (in mb)
T = average temperature through the depth (in K)
h = depth of the mixing layer over the city (in m)
dT/dp = rate of temperature change with pressure (in °K/mb), in the
lowest layers outside the city
Ludwig and Dabberdt (1973) pointed out an interesting characteris-
tic of the above equation. The equation suggests that there is a
minimum positive value that hm-^n can assume, implying a minimum mixing
depth for any given city. For typical conditions, e.g. T = 280°K and p
= 950 mb, the equation becomes the following, when dT/dP approaches a
negatively infinite value:
0.25
hmin=2'6*
If one needs to estimate mixing depths in the vicinity of some
urban areas, it can be done using the nearest available radiosonde mea-
surements. The mixing depths outside the city can be estimated directly
from the temperature and humidity soundings. Holzworth (1967) has
described the conventional method for estimating afternoon mixing depth
from the afternoon maximum temperature and the morning sounding (in the
United States, the morning sounding will typically be the one taken nom-
inally at 1200 GMT). The same sounding can also be used to estimate
rural mixing depths at night. Urban mixing depths can be derived from
the equation given earlier for nighttime cases and by assuming them to
be 100 to 200 m greater than the rural values for daytime cases. The
transition from daytime to nighttime conditions can be presumed to take
place quite rapidly at around sundown. The transition from nighttime to
daytime conditions is more gradual and must be estimated. The curves
shown in Figures 75, 76 and 80 will serve as guides, if better informa-
tion is available.
The preceding discussion applies primarily to simple continental
locations. In more complicated locations, especially those involving a
water-land interface, the behavior of the mixing depth can be much more
complex. Lyons and Cole (1976) describe mixing near the shoreline of
Lake Michigan. Edinger (1958) modeled the behavior of mixing in the Los
Angeles basin as driven by the diurnal heating cycle and the onshore
flow. In these more complicated situations, generalizations of mixing-
156
-------
depth behavior are probably not possible and it will be necessary to
examine each such case separately.
C. Winds
Winds are driven by pressure gradients, frictional forces and the
Coriolis acceleration. One might expect air flow over the city to be
different from that in the environs inasmuch as the pressure gradients
around a city are altered by the thermal effects of the city, and the
frictional effects of the rough city are different from those of its
smoother surroundings.
There is some evidence of thermal effects on urban winds, but the
effects are generally weak. At night, when the city is appreciably
warmer than its surroundings (i.e., the so-called urban "heat island"
effect) warm air rising over the city would be expected to produce a
radially inward flow near the surface. Shreffler (1978) has found that
there are daytime effects as well; although the daytime heat island
intensities are generally less than those at night, the lower daytime
stabilities require smaller surface temperature increments to maintain
upward motion over the city. In any event, the thermally driven com-
ponent of the wind is generally limited to about one or two meters per
second and tends to be undetectable above the background wind speeds
produced by large-scale pressure fields.
Thermal effects also have an indirect effect on wind speeds over
cities. The nighttime destabilization of the air over the city results
in greater vertical momentum transport, so winds at the lower altitudes
will tend to be greater at night in the cities than in their rural sur-
roundings. However, the surface winds are not the only winds of impor-
tance to the air quality problem. The integrated wind speeds through
the lower layers of the atmosphere will tend to be less over the city
throughout the day. At night the stabilization of the lower rural atmo-
sphere isolates the air aloft from surface friction and its slowing
effects. In fact, the removal of the friction can result in accelera-
tion as the motions adjust. By contrast, over the city with its greater
vertical motions, momentum continues to be lost to the surface.
During the day, the frictional effects are somewhat more direct.
An exponential function is frequently used to approximate wind speed
changes with height :
where U is the wind speed at height z, while UQ is the speed at some
reference height ZQ . Munn (1970) reports that Davenport (1968) pub-
lished the average values for c( shown in Table 19 for different types of
surface. The table also shows Davenport's (1968) reported averages for
157
-------
Table 19
VARIATION OF WIND EXPONENT AND THE HEIGHT
OF GRADIENT WIND WITH SURFACE CHARACTERISTICS--
AFTER DAVENPORT (1968)
Surface Type
Flat, open terrain
Suburban
Build-up urban center
Exponent
0.16
0.28
0.40
Altitude of
Gradient Wind
(m)
270
390
420
the height of gradient wind, above which the winds become more nearly
constant with altitude. Figure 81 (from Singer and Smith, 1970) plots
much the same information graphically. There are slight differences in
the exponents and the top for the urban profile—Singer and Smith extend
it to 500 m before the wind becomes constant versus 420 m for Davenport.
In addition to changes in wind speed with height, there are also
apt to be changes in wind direction that are accentuated by the urban
roughness. The wind direction near the surface will usually be rotated
counterclockwise relative to directions aloft. Munn (1970) reports a
case over New York where the change of direction from the surface to the
gradient wind level amounts to 59°. According to Munn (1970), the
directional shear has been reported by Ariel and Kluichnikova (1960) to
vary with atmospheric stability from about 15° with strong instability
to 40° during inversion conditions.
One final comment is in order with regard to urban effects on air
flow. Over cities, the airflow tends to be more turbulent. Ludwig and
Dabberdt (1973) have estimated that a city's effects are about
equivalent to one Pasquill-Gifford Stability Class (Pasquill, 1961; Gif-
ford, 1961). Thus, the dispersion of precursor pollutants through the
mixing layer will proceed more rapidly over the city than over its sur-
roundings .
158
-------
600
500
400
HEIGHT
(m)
300
200
100
URBAN AREA
GRADIENT WIND
SUBURBS
LEVEL COUNTRY
GRADIENT WIND
0 5
WIND SPEED (m/sec)
FIGURE 81 TYPICAL WIND PROFILES OVER URBAN, SUBURBS AND RURAL AREAS
(Singer and Smith, 1970)
-------
REFERENCES
Auer, A. H., and F. D. Eaton, 1976: Final report, NSF Grant AEN-76-
00898; U. of Wyoming, Laramie, Wyoming; 40 pp.
Ariel, N. Z. and L. A. Kluchnikova, 1960: Wind Over a City, Glavnaia
GgoJLiz. Observator. 94, 29-32.
Bluraenthal, D. L., W. H. White, R. L. Pease, and T. B. Smith, 1974:
Determination of the feasibiliy of the long-range transport of
ozone or ozone precursors; EPA Report 450/3-74-061; 117 pp.
Cleveland, W. S., B. Kleiner, J. E. McRae, and J. L. Warner, 1976: Pho-
tochemical air pollution: Transport from the New York City area
into Connecticut and Massachusetts; Science, 191, 179-181.
Davenport, A. G., 1968: The dependence of wind loads on meteorological
parameters, Proc. Internat. Seminar on Wind Effects on Buildings
and Structures, Nat. Research Council of Canada, Ottawa.
Edinger, J. G., 1978: The meteorology of Los Angeles' polluted layer;
UCLA Department of Meteorology; 22 pp.
Edinger, J. G., 1973: Vertical distribution of smog in Los Angeles
basin; Envir. Sci. Tech., 7, 247-252.
Endlich, R. M., and R. L. Mancuso, 1968: Objective analysis of environ-
mental conditions associated with severe thunderstorms and torna-
does; Wea. Rev., 96, 342-350.
Endlich, R. M., F. L. Ludwig, and E. E. Uthe, 1978: An objective method
for determining the mixing depth from lidar records; final report,
EPA Contract 68-02-2418; SRI International, Menlo Park, California;
194 pp.
Environmental Monitoring and Support Laboratory, 1975: Meteorological
data for the northeast oxidant transport study; EPA, Las Vegas,
Nevada; Xeroxed draft.
Fitzsimmons, C. K., K. Zeller, and N. J. Pearson, 1978: Analysis of
aerometric data collected by aircraft during a stagnation period in
Washington, D.C., August 1976; presented at APCA meeting, Houston,
Texas; paper 78-10.7.
Gifford, F. A., 1961: Use of routine meteorological observations for
estimating atmospheric dispersion; Nucl. Saf., 2, 47-51.
Harrison, R« M., C. D. Holman, H. A. McCartney, and J. F. R. Mcllveen,
1978: Nocturnal depletion of photochemical ozone at a rural site;
Atmos. Envir., 12, 202-2026.
161
-------
Heffter, J. L., and A. D. Taylor, 1975: A regional-continental scale
transport, diffusion and deposition model, part I: Trajectory
model; NOAA Tech. Memo ERI ARL-50; pp. 1-16.
Holzworth, G. C., 1967: Mixing depths, wind speeds and air pollution
potential for selected locations in the United States; jJ. Appl-
Meterol., 6, 1039-1044.
Johnson, W. B., and R. E. Ruff, 1975: Observational systems and tech-
niques in air pollution meteorology; Proc. Workshop Meteor. Envir -
Assessment, Boston; Amer. Meteorol. Soc., Boston, Massachusetts.
Johnson, W. B., and H. B. Singh, 1977: The origin and significance of
ozone maxima aloft; interim progress report for Coordinating
Research Council Contract CRC-APRAC CAPA-12-72 (1-76); 65 pp.
Lonneman, W. A., 1977: Results of May 13-18 Florida bag samples; EPA
memorandum to J. J. Bufalini; February 14.
Lovelace, D. E., T. Kapsalis, R. C. Bourke, and P. P. Cook, 1975: Indi-
anapolis 1974 summer ozone study; Indianapolis Center for Advanced
Research, Indianapolis, Indiana; 126 pp.
Ludwig, F. L., 1970: Determination of mixing depths for use with synop-
tic model; Proc. Symp. Multiple-source Urban Diffusion Models; EPA
Pub. AP-86; pp. 5-33 to 5-38.
Ludwig, F. L., and W. F. Dabberdt, 1973: Effects of urbanization on tur-
bulent diffusion and mixing depth; Int. J_. Biometeol., 17, 1-11.
Ludwig, F. L., and W. F. Dabberdt, 1976: Comparison of two practical
atmospheric stability classification schemes in an urban applica-
tion; J_. Appl. Meteorol., 15, 117-1176.
Ludwig, F. L., J. R. Martinez, and K. C. Nitz, 1978: Data summaries for
the Houston area oxidant study: interim report for Houston Area
oxidant Study Contract DA-1; 261 pp.
Ludwig, F. L., E. Rieter, E. Shelar, and W. B. Johnson, 1977: The rela-
tion of oxidant levels to precursor emissions and meteorological
features; Vol 1: Analysis and findings, EPA report 450/3-77-09;
153 pp.
Ludwig, F. L., and E. Shelar, 1977: Ozone in the Northeastern United
States, EPA report 901/9-6-007; 276 pp.
Ludwig, F. L., and E. Shelar, 1978: Site selection for the monitoring of
photochemical air pollutants; EPA report 450/3-78-013; 130 pp.
162
-------
Lusis, M. A., K. G. Anlauf, Y. S. Chung, and H. A. Wiebe, 1976: Aircraft
03 measurements in the vicinity of Toronto, Canada; Air Poll. Cont .
Assoc. paper 76-6.1.
Lyons, W. A., and H. S. Cole, 1976: Photochemical oxidant transport:
Mesoscale lake breeze and synoptic-scale aspects; ^J. Appl -
Meteorol.. 15, 733-743.
Mage, D. T., et al., 1978: The RAPS helicopter air pollution measurement
program, St. Louis, Missouri, 1974-1976; EPA, Las Vegas, Nevada;
draft report.
Mancuso, R. L., and R. M. Endlich, 1973: Wind editing and analysis pro-
gram spherical grid (WEAP-1A); User's Manual; U.S. Army Contract
DAHC04-71-C-0013; SRI International, Menlo Park, California; 69 pp.
Meyer, E. L., W. P. Freas, J. E. Summerhays, and P. L. Youngblood, 1976:
The use of trajectory analysis for determining empirical relation-
ships among ambient ozone levels and meteorological and emissions
variables, pres. at Int. Conf. on Photochemical Oxidant Pollution
and Its Control, Raleigh, North Carolina, 12-17 September -
Munn, R. E., 1970: Airflow in Urban Areas, Urban Climates, World
Meteorol. Org., Tech. Note 108, 15-39.
Nie, N. H. et al., 1975: SPSS, Statistical Package for the Social Sci-
ences (second ed.); McGraw-Hill Book Company, New York; 675 pp.
Office of Air Quality Planning and Standards, 1977: Uses, limitations
and technical basis of procedures for quantifying relationships
between photochemical oxidants and precursors; EPA report 450/2-
77-021a.
Parker, R. 0., and J. R. Martinez, 1975: Los Angeles reactive pollutant
program (LARPP) data archiving and retrieval; NTIS No. 244 295;
130 pp.
Pasquill, F., 1961: The estimation of the dispersion of windborne
material; Meteorol. Mag., 90-49.
Ruff, R. E., L. S. Gasiorek, and H. Shigeishi, 1977: Master data file
for the summer 1975 northeast oxidant transport study; Users'
manual; EPA report No. 901/9-76-004; 46 pp.
Shaulk, M. E., and W. M. Alexander, 1978: Houston area oxidant study:
Airborne ozone study, volume II data; Baylor U., Waco, Texas;
359 pp.
Shreffler, J. H., 1978: Detection of Centripetal Heat-Island Circula-
tions from Tower Data in St. Louis, Boundary Layer Meteorology, 15,
229-242.
163
-------
Singer, I. A. and M. E. Smith, 1970: A Summary of the Recommended Guide
for the Prediction of the Dispersion of Air-borne Effluents (ASME),
Urban Climates, World Meteorol. Org. Tech. Note 108, 306-324.
Singh, H. B., F. L. Ludwig, and W. B. Johnson, 1977: Ozone in clean
remote atmospheres: Concentrations and variabilities; final report,
Coordinating Research Council contract CAPA-15-76; SRI Interna-
tional, Menlo Park, California; 147 pp.
Singh, H. B., F. L. Ludwig, and W. B. Johnson, 1978: Tropospheric
ozone: Concentrations and variabilities in clean remote atmo-
spheres, Atmos. Environ., 12, 2185-2196.
Siple, G. W., K. F. Zeller and T. M. Zeller, 1976: data for the
northeast oxidant transport study, EPA Off. of Res. and Dev.,
Environ. Monitor- and Support Lab., Las Vegas, Nevada.
Spicer, C. W., J. W. Gemma, and P. R. Sticksel, 1977: Ozone: Transport
considerations and predictive models for southern New England:
draft report, EPA contract 68-02-2241; Battelle, Columbus, Ohio.
Spicer, C. W. D. W. Joseph, and G. W. Ward, 1976: Final data report on
the transport of oxidants beyond urban areas; final report, EPA
Contract 68-02-2441; 380 pp.
Summers, P. W., 1965: An urban heat island model—its role in air pol-
lution problems, with application to Montreal; pres. at First
Canadian Conference on Micrometeorology, Toronto; 22 pp.
Turner, D. B., 1964: A diffusion model for an urban area; J^. Appl.
Meteorol., fB3, 83-91.
Turner, D. B., 1970: Workbook of Atmospheric Diffusion Estimates; EPA
Office of Air Programs Publications No. AP-26; 84 pp.
Tyson, P. D., M. Garstang, and G. D. Emmitt, 1973: The structure of heat
islands; Dept. Geograph. and Environ. Studies, U. Witwatersrand,
Johannesburg, Occasional paper No. 12; 71 pp.
Uthe, E. E., 1976: Preliminary analog lidar records from SRI METROMEX
study during the period 29 July - 14 August 1975. Informal report
submitted to U. of Wyoming, Stanford Research Institute, Menlo
Park, 59 pp.
Uthe, E. E. and P. B. Russell, 1973: METROMEX 1972 Radiation/Aerosol
Experiment, Final Rept. Project 1976, Stanford Research Institute,
Menlo Park, California.
Uthe, E. E., P. B. Russell, C. M. Bhumralkar and R. M. Endlich, 1978:
Contributions to METROMEX, Final Report NSF Grant ENV76-14835, SRI
International, Menlo Park, California, 91 pp.
164
-------
Vukovich, F. M., 1973: Some Observations of Ozone Concentrations at
Night in the North Carolina Piedmont Boundary Layer, J^. Geophys.
Res., 78, 4458-4462.
Washington State University, 1976: Measurement of light hydrocarbons and
studies of oxidant transport beyond urban areas; final report, EPA
Contract 602-2339; 317 pp.
Westberg, H., K. Allwine, and E. Robinson, 1977: Measurement of light
hydrocarbons and studies of oxidant beyond urban areas, Houston
study-1976; final report EPA contract 68-02-2298; 240 pp.
Wiebe, H. A., M. Lusis and K. G. Anlauf, 1975: The Toronto oxidants
study: Aircraft ozone survey of the metropolitan Toronto region;
Canadian Atmospheric Environment Service Report No. ARQA 29-75;
174 pp.
Wolff, G. T., et al., 1975: An aerial investigation of photochemical
oxidants over New Jersey, southwestern New York and Long Island,
western Connecticut, northern Deleware, southeastern Pennsylvania
and northeastern Maryland; Interstate Sanitation Commission, New
York; 118 pp.
Wolff, G. T., et al., 1976: Anatomy of two ozone transport episodes in
the Washington, B.C. to Boston, Massachusetts corridor; pres. at
10th Midatlantic states Amer- Chem. Soc. meeting, Philadelphia;
February, 23-26.
Zeller, K. F., 1976: Personal communication.
Zeller, K. F., R. B. Evans, C. K. Fitzsimmons, and G. W. Siple, 1977:
Mesoscale analysis of ozone measurements in the Boston environs;
J.. Geophys. Res.. 82, 5879-5888.
165
-------
REPORT NO
EPA-450/4-79-017
TECHNICAL REPORT DATA
(Please rcjd Instructions on the "tverse before completing}
TITLE AND SUBTITLE
Assessment of Vertical Distributions of Photochemical Pollutants
and Meteorological Variables in the Vicinity of Urban Areas
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 1979
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
F.L. Ludwig
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
SRI International
333 Ravenswood Avenue
Menlo Park, California 94025
10. PROGRAM ELEMENT NO.
2AA635
11. CONTRACT/GRANT NO.
68-02-2662
2. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air, Noise and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
5. SUPPLEMENTARY NOTES
16. ABSTRACT
A system for classifying vertical profiles of ozone and other photochemical pollutants has been derived by analyzing
268 ozone profiles obtained from published reports and EPA data files — 93 from St. Louis, 40 from Los Angeles,
30 from Houston, 31 from Washington, D.C., 53 from Toronto, and the remaining 21 from Indianapolis, Tampa,
and Denver. The profiles fall into six different categories that are related to the relative importance of the following
factors: destruction of zone near the surface, vertical mixing, and photochemical production. Vertical mixing tends
to produce profiles that are nearly uniform with height. Strong mixing offsets the destruction at the surface. Unless
effective vertical mixing of ozone occurs, destruction at the surface will produce profiles with steeply increasing
concentrations with height in the lowest layers. The report presents and discusses 108 representative examples of
the vertical profiles, most with corresponding temperature data and many with aerosol (nephelometer), NOX, or
hydrocarbon data. Linear regression has been used to define the relation between ozone concentrations in the
mixing layer and those aloft, for the different profile types. The locations relative to an urban area and the times
when each profile type is most apt to occur are discussed in terms of the statistics of the data set and the physical
processes involved.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Ozone
Oxidants
Photochemical
Nitrogen dioxide
Nitrogen oxide
Meteorological
Vertical Profiles
Temperature Profiles
Temperature Inversions
Hydrocarbons
Aerosols
Nephelometer
Mixing height
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
21 . NO. OF PAGES
20. SECURITY CLASS (This page)
22. PRICE
EPA
Form 2220-1 (Rev. 4-77) PREVIOUS EDI T.ON i s OBSOLE T E
------- |