EPA-450/3-74-034-a
FEBRUARY 1975
INVESTIGATION
OF OZONE AND OZONE
PRECURSOR CONCENTRATIONS
i
AT NONURBAN LOCATIONS
IN THE
EASTERN
UNITED STATES,
PHASE II,
METEOROLOGICAL ANALYSES
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-74-034-a
INVESTIGATION
OF OZONE AND OZONE
PRECURSOR CONCENTRATIONS
AT NONURBAN LOCATIONS
IN THE
EASTERN
UNITED STATES,
PHASE II,
METEOROLOGICAL ANALYSES
by
W. D. Bach, Jr.
Research Triangle Institute
Research Triangle Park, N. C. 27709
Contract No. 68-02-1077
EPA Project Officer: E. L. Martinez
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, N. C. 27711
February 1975
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are available
free of charge to Federal employees, current contractors and grantees,
and nonprofit organizations - as supplies permit - from the Air Pollution
Technical Information Center, Environmental Protection Agency, Research
Triangle Park, North Carolina 27711; or, for a fee, from the National
Technical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22161.
This report was furnished to the Environmental Protection Agency
by Research Triangle Institute, Research Triangle Park, N. C. , in
fulfillment of Contract No. 68-02-1077. The contents of this report
are reproduced herein as received from Research Triangle Institute.
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 as an
endorsement by the Environmental Protection Agency.
Publication No. EPA-450/3-74-034-a
11
-------
ACKNOWLEDGEMENTS
The data analyses reported herein are the results of strong
support given by the government employees and by the staff of Research
Triangle Institute. Specifically, thanks are due Mr. John Clarke,
Division of Meteorology for supplying the basic trajectory analysis
program. The National Climatic Center provided upper air data, in a
given format as soon as it was available and also supplied surface
weather observations data. Mr. Doug McMann of the Research Triangle
Institute gave unselfish extra measure in the development and execution
of the numerous computations.
iii
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS iii
LIST OF FIGURES vii
LIST OF TABLES viii
INVESTIGATION SUMMARY ix
1.0 INTRODUCTION 1
2.0 OBJECTIVES 5
3.0 DATA SOURCES 7
3.1 Ozone Data 7
3.2 Meteorological Data 7
3.2.1 Upper Air Data 7
3.2.2 Surface Weather Observations 7
4.0 OZONE CONCENTRATIONS AND SYNOPTIC WEATHER 11
4.1 Ozone Averaging Periods 11
4.2 Frontal Analyses 11
4.3 Characteristics of Pressure Systems 19
4.4 Ozone Concentrations and Thunderstorms 19
4.5 Ozone Concentrations and Sky Cover 19
4.6 Time Section Analyses 21
4.6.1 Potential Temperature 21
4.6.2 Dewpoint Depression 21
4.6.3 Significant Temperature Layers 21
4.6.4 Computations and Presentation of Results 22
4.7 Data Interpretation 22
5.0 OZONE CONCENTRATIONS AND THE MIXED LAYER 41
5.1 Mixing Layer Parameters 41
5.2 Regression Analyses 42
5.3 Results 43
6.0 OZONE CONCENTRATIONS AND AIR PARCEL TRAJECTORIES 51
6.1 Computation of Trajectories 51
6.2 Selection of Trajectories 53
6.3 Trajectories with High and Low Ozone Concentrations 53
-------
TABLE OF CONTENTS (cont.)
Page
6.3.1 Kane 53
6.3.2 McHenry 59
6.3.3 Coshocton 59
6.3.4 Lewisburg 60
6.4 Summary 60
7.0 CLIMATIC INFLUENCES 61
7.1 Temperature 61
7.2 Average Sky Cover, Sunrise to Sunset 61
7.3 Relative Dispersion 61
7.4 Stagnation Periods 63
8.0 SELECTED STUDIES 65
8.1 Daily Maximum Temperature and Maximum Hourly Ozone 65
8.2 Concurrent Trajectories 65
8.3 Case Study 68
9.0 FINDINGS AND CONCLUSIONS 75
9.1 Fronts, Ozone and Potential Temperature Time Sections 75
9.2 Air Parcel Trajectories 75
9.3 Mixed Layer Properties 76
9.4 Maximum Temperature and Maximum Ozone Concentration 76
REFERENCES 77
APPENDIX A - BASIC DISCUSSION OF POTENTIAL TEMPERATURE ANALYSIS 79
TECHNIQUES
APPENDIX B - AIR PARCEL TRAJECTORIES 83
vi
-------
LIST OF FIGURES
Figure Page
1 Locations of ozone monitoring stations and rawinsonde
stations
2 Mean diurnal ozone concentrations 12
3 Tracks of centers of high pressure areas 20
4 Time section of potential temperature, time series of
ozone concentrations and frontal passages at indi-
cated locations during the study period 23
5 Sample results of regression model 45
6 Average annual hydrocarbon emission density 52
7 Trajectories associated with high and low ozone
concentrations 55
8 Maximum hourly ozone as a function of maximum daily
temperature 66
vii
-------
LIST OF TABLES
Table Page
1 Twelve-Hour Average Ozone Concentrations 13
2 Summary of Frontal Characteristics 17
3 Ranking by Significance of Mixed Layer Variables in
Regression Model of Ozone Concentrations 46
4 Linear Correlation Coefficients, R, of Each Mixed
Layer Variable with the Concurrent Twelve-Hour Ozone
Concentration 48
5 Ozone Concentrations for Indicated Percentile
Rankings at Each Monitoring Location 54
6 Comparison of Climatological Relative Dispersion
Values with those Occurring During the Study 62
7 Characteristics of Air Parcels Arriving at a Moni-
toring Station After Passing Near Other Monitoring
Stations 69
viii
-------
INVESTIGATION OF OZONE AND OZONE PRECURSOR CONCENTRATIONS
AT NONURBAN LOCATIONS IN THE EASTERN UNITED STATES
PHASE II
INVESTIGATION SUMMARY
Background
From July through the end of October 1973, the Research Triangle
Institute operated nonurban sampling locations at McHenry, Maryland;
Kane, Pennsylvania; Coshocton, Ohio; and Lewisburg, West Virginia. At
these locations, ozone, oxides of nitrogen, and hydrocarbons were moni-
tored continuously to determine, in part, the areal extent of high ozone
concentrations in nonurban areas of the northeastern United States', to
provide a data base of nonurban ozone and ozone precursor concentration
measurements for future detailed analysis; and to determine the inter-
relationship between ozone concentrations and its precursors at the selected
locations. Subsequent to that field investigation, this program examined
various meteorological conditions that accompanied these ozone concentrations,
high and low, attempting to relate the aerological and air quality measure-
ments at or near these four monitoring locations. The primary emphasis
of the meteorological portion of the study is at the synoptic scale—the
scale of motions that produce the day-to-day weather changes.
Analysis Techniques
Three primary methods of analysis were adopted:
1) The time-altitude sections of potential temperature and signi-
ficant temperature layers were developed showing some of the dynamical
processes of the lower atmosphere where the ozone is generated and trans-
ported. The relationship of inversions and/or fronts and atmospheric
structure to the twelve-hour average ozone concentraticn changes were
examined at each of the four locations using those sections.
2) Trajectories, approximating the path of air parcels during the
48-hours prior to their arrival, were computed for each location at each
twelve-hour interval during the field study. These trajectories were
examined for their association with high or low concentrations and the
ix
-------
implications as to the existence of preferred "source" regions (urban or
nonurban) as indicated by the spatial distribution of hydrocarbon emission
density.
3) The statistical relationships of the twelve-hour ozone concen-
trations to the properties of the mixing layer (usually confined within
the lowest 2 km) of the atmosphere at rawinsonde locations near the nonurban
sites were investigated. Correlation coefficients and the significance of
the meteorological parameters upon the ozone concentrations were determined.
Interpretation and Results
In earlier studies, data had indicated that ozone concentrations tended
to rise steadily prior to frontal passage and decrease rapidly, apparently
in association with the frontal passage. Thus, either the high ozone
concentrations were associated in some way with the frontal passage, or
they were associated with the characteristics of the air mass before and
after the frontal passage. There are several ways to investigate the
relationships of ozone with frontal passages in the 1973 data. At the same
time, properties of the air masses before and after the frontal systems and
other properties of the lower atmosphere where turbulence and transport
of ozone are taking place can be investigated. Some studies have also
suggested that high ozone concentrations at the ground or perhaps even at
nighttime might be associated with inversion layers aloft, which effectively
limit the vertical dispersion of pollutants.
Time-altitude sections were plotted with potential temperature (an
indicator of the state of the atmosphere), the date and time of frontal
passages, and the twelve-hour average ozone concentrations on the same time
axis for each of the four nonurban locations. Although changes in ozone
concentration were occasionally related to the frontal passages, only the
major disturbances of potential temperature shown in the cross-sections
were associated with the changes of the ozone concentrations. In many
instances of frontal passages, the potential temperature sections suggested
that hardly any changes of air masses took place. No major trends of
ozone concentrations developed before or after these frontal passages.
An association of high and low ozone concentrations before and after some
frontal passages were evident principally during early July and in October
x
-------
at Kane. The effects of those fronts could not be documented at the other
locations, since Kane was the only operative monitoring location at those
times.
The time sections clearly show the atmospheric changes—slowly descending,
stabilizing motion, and strong warming aloft—through the lower 5 km of
the atmosphere accompanying the onset of a stagnant weather system along with
a major increase of ozone concentration during day and night. During the
latter part of an eleven-day episode in late August and early September
at McHenry, those stagnant conditions remained, but descending motion gave
way to much less stable conditions. The ozone concentration decreased at
a slow rate, although deep vertical mixing was possible. Similar weather
conditions existed over the study area, but each locality showed somewhat
different trends of the ozone concentrations. The subsiding motion was too
weak and short lived to have transported ozone downward from the stratosphere
to account for the high ozone concentrations.
On the basis of these analyses, it became apparent that ozone concen-
trations change when there are major disturbances in the atmospheric
structure—extending above 2 km—but local factors, perhaps terrain, also
influence the daily trends of ozone.
A most elementary concept of high ozone concentrations might be to
consider that metropolitan and industrial areas are the most likely
anthropogenic sources of ozone precursor materials. Downwind of those
cities, precursors are mixed together and form ozone in sunlight. From
simple dispersion models, the ozone concentration would be expected to be
directly proportional to the emissions of precursor material and inversely
proportional to the wind speed, i.e., the highest ozone concentrations
should be associated with air that moves slowly over major source areas.
To test that overall concept, air parcel trajectories, indicative of the
path that air near the ground has taken for the past forty-eight hours, were
computed for each of the nonurban locations from upper air data. If high
ozone concentrations at the nonurban locations were related to specific
anthropogenic source regions upwind of the station, then the trajectories
with the high ozone concentrations should pass over those specific sources.
By the same reasoning, low ozone concentrations might be associated with
xi
-------
trajectories that have a small residence time (higher speed) over areas of
low emissions. On both counts, the trajectory analysis fails to show
consistent association of high ozone and high emission densities, or low
ozone and low emission density. Specific source regions for the occurrences
of high or low ozone concentrations at any of the four nonurban ozone
monitoring locations could not be identified. These data suggest that the
high ozone concentrations were associated with slower-moving air, especially
the prolonged stagnation period in late August and early September, and that:
low ozone concentrations were associated with faster moving air. Hence,
one concludes that the trajectory that the air takes en route to each of the
four ozone monitoring locations, influences but does not determine the extent
of the ozone concentration.
The trajectories relate to the horizontal transport of ozone or ozone
precursors; the potential temperature sections relate to the synoptic
motion producing changes in the vertical structure of the atmosphere. The
mixed layer—the layer of the air near the ground influenced by the daily
radiational heating and cooling—is important to the turbulent dispersion
of ozone or ozone precursors. The mixed layer varies from day to night and
as the weather systems move through. The response of the layer to these
influences is given by the variability of its properties, e.g., mixing depth,
ventilation, relative dispersion, mean wind velocity, and mean water vapor
content. The relationship of those properties to the comparable term
ozone concentration were examined.
The linear correlation coefficients of the ozone concentrations
(measured at a nonurban site) and the mixed layer parameters (measured at
a nearby urban rawinsonde location) were computed. In only two instances
did the magnitude of correlation exceed 0.30 with a variable other than the
time of day, the mixing depth, or the water vapor mixing ratio. The
correlation of water vapor mixing ratio with ozone was the only one to have
the same sign (positive) in all cases.
The best combined linear regression of all mixed layer properties with
the ozone showed that the water vapor mixing ratio and mixing depth were
usually the most significant terms of the regression. These results
suggest that high ozone concentrations are associated with high water
xii
-------
vapor content—an unexpected finding. The physical significance of the
statistical finding is not yet understood.
Another study showed that a high linear correlation exists between
the maximum hourly ozone concentration for the day at the nonurban locations
and the daily maximum temperature at a nearby location.
xiii
-------
INVESTIGATION OF OZONE AND OZONE PRECURSOR CONCENTRATIONS
AT NONURBAN LOCATIONS IN THE EASTERN UNITED STATES
PHASE II
1.0 INTRODUCTION
Until recent years, ozone, as an air pollutant, was considered as an
urban problem resulting from photochemical reaction of emitted reactive
hydrocarbons and oxides of nitrogen. Dilution with nonurban air during
transport and destructive mechanisms were thought to reduce the concen-
trations in nonurban areas.
In the course of a study of injury to Christmas trees, Environmen-
tal Protection Agency (EPA) investigators were surprised to find oxidant
concentrations at sites in western Maryland and eastern West Virginia
frequently exceeding the National Ambient Air Quality Standard (NAAQS)
_3
for ozone (160 yg m ) during the period May 29 through September 28,
1970. Corroborative measurements made with a chemiluminescent ozone
2
meter showed that virtually all of the oxidant at one site was ozone.
Of particular interest was the fact that the high ozone concentrations
persisted into the dark hours; i.e., ozone concentration did not exhibit
the typical diurnal pattern in which it decreases to near zero at night.
Nitrogen oxide concentrations during this period were reported to be
2
near background levels.
Ozonesonde measurements have shown that ozone-rich layers occur
3
within elevated but low-level inversion layers. The high ozone
concentrations observed aloft at Point Mugu, California, can be attrib-
uted to precursors originating in the Los Angeles area. The transport
of photochemical smog from Fresno to the Mineral King Valley of California
4
has been suggested.
Miller and Ahrens showed locally high ozone concentrations below
the West Coast inversion could be associated with breaks or folds in the
inversion over San Francisco Bay. Further inland, where and when the
inversion is dissipated by surface heating, ozone concentrations increased
-------
rapidly at the ground and increased aloft during the day. At night, ozone
aloft, insulated from the ground by the reestablished inversion, showed
small decreases of concentration. High ozone concentrations were found
in unstable air; and low ozone concentrations in stable air.
Vukovich and others showed diurnal changes of ozone concentration at
ground level and aloft are strongly influenced by the onset of radiative
inversions and hence, air stability. These studies implied only slight
nocturnal decreases in ozone concentrations above the inversion layer.
Ozone concentrations measured at Indio, California 140 miles from
Los Angeles, exceeded the air quality standards for oxidants more frequently
than at any other sampling site in California. The high concentrations
occur during day and night hours. The evidence indicates that transport
to Indio from the Los Angeles Basin is a contributing factor to the
observed concentrations.
Analysis of episodes of ozone concentrations exceeding the NAAQS near
Miami, Florida showed that long-range transport of the ozone or ozone
Q
precursors may have been responsible.
Q
RTl's 1972 study of ozone concentrations at McHenry, Maryland
identified increases of ozone during day and night when high pressure
systems stagnated in the northeastern United States. These episodes ended
rather abruptly as frontal systems moved through the area, replacing the
stagnant air with air having a different history. Approximate air
trajectories at the 850-mb pressure surface indicated that episodes of
high ozone concentrations at McHenry were associated with air which had
previously moved over the urban-industrial areas of the lower Great Lakes
and the mid-Atlantic States.
RTl's 1973 field study and analyses (Phase I of this contract)
sought to establish a data base of nonurban ozone and ozone precursor
concentrations at four locations in the northeastern United States; to
develop statistical summaries of the frequency of occurrence of ozone,
nitrogen dioxide, and non-methane hydrocarbons at those locations; and to
determine interrelationships between ozone concentrations and ozone
precursor concentrations at the locations. During this study, an EPA
aircraft equipped with an ozone meter made two flights over the study
-------
area. Ozone concentration measurements were taken enroute to each
station, where a low altitude fly-by of the ground stations was made.
Profiles of ozone concentrations during descent and ascent and a near-
ground check of the two ozone measurements were obtained. The validity
of measurements at each ground station was checked as part of a data
quality assurance program.
The statistical summaries show that the hourly average ozone
concentrations repeatedly exceeded the NAAQS for ozone. The frequency
of occurrence of hourly concentrations in excess of the ozone standard
at McHenry was 37 percent in 1973, in comparison to 11 percent found in
1972. Lag correlation studies between stations gave correlation coef-
ficients of 0.468 to 0.678. These data reinforced the conclusion that
the occurrence of high ozone concentrations at nonurban locations is
widespread, affecting a large area in the eastern United States.
Analysis of the airborne measurements indicated that high ozone
concentrations were prevalent over much of the study area; that hori-
zontal and vertical gradients of ozone concentrations existed aloft; and
that the ozone concentrations tend to persist.
Previous studies have investigated only a few of the possible
meteorological influences upon the ozone concentrations that are ob-
served in these nonurban areas over the northeastern United States.
This research expands upon those studies. It investigates the meteoro-
logical conditions present during the study period, tries to relate them
to the observed ozone concentrations at the nonurban locations, and tries
to find the influence of possible sources of ozone or ozone precursors.
The areal coverage and magnitude of the problem over the indus-
trialized east is not directly analogous of the problem of photochemical
smog in the Los Angeles Basin. Indeed, it represents a different
problem with a complex interaction of various source areas, atmospheric
dispersion in different terrain conditions, and finally, is subject to
transient weather systems.
-------
2.0 OBJECTIVES
In this study, the various meteorological conditions are quanti-
fied and related to the average ozone concentrations at each of the four
monitoring sites. The conditions are documented through observations
made at National Weather Service Offices. From these data, the meteoro-
logical variables that have a strong or weak influence upon the measured
ozone concentrations are determined. The documentation is concerned
with horizontal and vertical distribution of the meteorological variables.
Some of the synoptic weather characteristics investigated are: time
sections of the vertical structure of the lower troposphere; the charac-
teristics of the mixing layer which affect the dispersion and transport
of pollutants; and the characteristics of fronts and pressure systems.
Another objective is to develop and analyze air parcel trajectories
arriving at each of the sampling sites and relate them to the occurrences
of high or low ozone concentration at the nonurban locations. Investiga-
tion of low ozone concentrations is as important as that of high ozone
cases if geographic or meteorological differences are to be examined.
The magnitude of ozone concentrations observed during 1973 were
different from those observed in 1972, suggesting that climatic vari-
ables, e.g., temperature, precipitation, sunshine, or pressure-system
movement, during 1973 were substantially different than in 1972. An
examination of these differences was another research objective.
Finally, in the process of the investigation, certain avenues of
research appeared promising in explaining the high or low ozone concen-
tration. Some of these were pursued in an effort to give the research as
broad a scope of possible.
-------
3.0 DATA SOURCES
3.1 Ozone Data
The ozone data for this study were taken during the field monitoring
program conducted by RTI in the summer and early fall of 1973 at four
ozone monitoring locations at Kane, Pennsylvania, McHenry, Maryland,
Lewisburg, West Virginia and Coshocton, Ohio (Figure 1). The gathering,
reduction, and tabulation of these data have been documented and reported.
The basic unit of the ozone data provided for analysis was the one-hour
average concentration.
3.2 Meteorological Data
3.2.1 Upper Air Data
Upper air data for each of twenty-two rawinsonde sounding
stations within the northeastern United States were obtained from the
National Climatic Center for the period July 1 through October 31, 1973.
Two soundings are made daily at 0000 and 1200 Greenwich Mean Time (GMT).
For each processed sounding the reported data include the altitude,
pressure, temperature, relative humidity, wind speed and direction at
50-mb vertical increments and at any intermediate significant levels
within the troposphere.
The location of most of the rawinsonde stations are also shown in
Figure 1.
Four rawinsonde stations—Pittsburgh, Pennsylvania (PIT); Buffalo,
New York (BUF); Huntington, West Virginia (HTS); and Dayton, Ohio (DAY)—
were chosen as upper air stations "representative" of the McHenry, Kane,
Lewisburg, and Coshocton ozone monitoring stations, respectively, prin-
cipally because they were the closest upper air stations.
3.2.2 Surface Weather Observations
Surface synoptic meteorological data were derived from
surface pressure analyses for three-hour increments during the period
July 1 through September 30. These analyses show the movement and
extent of high and low pressure systems, frontal systems, cloud amounts,
and precipitation systems. Additional surface weather information was
obtained from the Local Climatological Data Summaries (LCD) for numerous
locations throughout the study area shown in Figure 1. Additional surface
-------
QJ
T3
C
O
CO
C
•H
0)
CO
tti
0) •
(3 >.
O T3
•H 3
w w
cd en
4-1
CO 0)
c
•H C
t-l -rH
O
4-1 T3
•H (1)
C w
so
O •— '
N
O CO
ti
M-l O
O -H
O
•H
O
O
i-J
£>0
-------
pressure information was obtained from the Climatological Data National
Summary. This publication gives analyses of the tracks of cyclones and
anticyclones across the United States during each month.
-------
4.0 OZONE CONCENTRATIONS AND SYNOPTIC WEATHER
4.1 Ozone Averaging Periods
Twelve-hour average ozone concentrations were computed for each of
the four monitoring stations. The averaging period conforms to the minimum
interval between rawinsonde releases. The rawinsonde data are the only
routine measurements of the state of the atmosphere above the ground within
the six hours prior until six hours following that observation. They
define the minimum spatial and time resolution of synoptic scale motion
aloft. Ozone concentrations averaged over the same period lose detail
on the intradiurnal structure, but show longer term trends of ozone con-
centrations. The changes which occur in the longer term averages are more
likely in response to the slowly changing synoptic meteorological conditions,
depicted by the changes in the upper air observations. In earlier studies,
the long-term trends of the ozone concentration with the synoptic meteoro-
logical conditions, were well depicted using this technique.
The averaging periods began at 0600 GMT (0200 EDT) and at 1800 GMT
(1400 EDT) and continued for the following twelve hours. The former period
corresponds to a twelve-hour interval centered about 1200 GMT whereas the
latter interval is centered about 0000 GMT. These central times coincide
with the times of rawinsonde observations. During the latter time period,
the daily maximum ozone concentrations are usually observed. The 1200 GMT
average value is usually lower, being associated with the nighttime and
early morning hours, as indicated in Figure 2. These averaging intervals
give a larger difference between twelve-hour average ozone concentrations
than would have been found using daylight (0800-2000 EDT) and dark (2000-
0800 EDT) averaging intervals.
Table 1 gives these twelve-hour averages for each station by months
from July through October. At least six one-hour averages in one period
were required to compute a twelve-hour average.
4.2 Frontal Analyses
The date and hour of each frontal passage at the four ground observing
stations was determined from the surface pressure analyses. The type,
intensity, characteristics of each front, as coded on the analyses, were
recorded. The speed and direction of frontal movement was computed from
the maps. This information is presented in Table 2. In all of the cases,
11
-------
« /•'- //• \-4 9
X M<'Henry , Md .
O Kane, Pa.
Coshocton, Ohio
• Lewishurg, W. Va
OUOn 0200 0^00 0600 0800 1000 1200 1400 1600 1800 2000 2200 0000
Time of Day (EOT)
Figure 2. Mean diurnal ozone concentrations at McHenry,
Maryland; Kane, Pennsylvania; Coshocton, Ohio;
and Lewisburg, West Virginia, from June 26 to
September 30, 1973.10
12
-------
§
W
O
§
o
W
§
W
W
Pi
o
PC
W
W
•9
H
GMT ObOO TO 1745 GMT
TON LEWISQURC MC HENRY KANE COSHQCTON LEHI3BURG
oa
a
t—
UJ
oz
0 «C
co *:
•*
>-
ee
z
IU
:r
^ ff-e-
o* o t>
O- O- O
*• c>
o» o»
k o> t>
C- O-ff
i
e. €>«•
O^ O^O AJ 0"
i
o-o- a- co to
o- o o- o- > o- o- o- o o- 9-oo>co o> o- c- o- o t> o o-o-o-aa;
o- o o
,'°,t*' -
t*\ ~* c
*
'f- (M
.CO —
» — • o
fM »/> 5
r r*. »*
O- (M f\J -• -•
— in o- o» o-
l
' !
-C IT IT 0- 0-
^^ »C CQ t^ ^
(^ -O O^ r» o
£?£££ £££££ ?£«?££
a-o*-c
. «~ *
00*0
^ o* c
1 * *
t-.o- o-
•O--O- <
i i
i
^ o^'ff1
i i
• o-c- o- • 0*0 o- o
l— nit* sr tr>
o o o o o
0- 0- (
o- o* e
1
* o* &•
-ru co
<>• <> O-e> «>•
1
O- O" O- AJ —
-.00 -0 0 1/1
O CO O Kl K>
i '
j
»•••••»•••
o- o- o- o- y-
o- o- o-o- o-
O t> f> O »
\** -O f
t
AI tr e- »* •<>
o- ^ r» IT o
f~to a\-ata
o- o- o- o- o-
o- o- o t> o-
!
r ••- «- «r «r «
O O- O l/> -O
O- O- AI ft O*
LP — ^; o- r~-
?,£,£?£
j
— 4 o^ LA o «> e> »•
, !
AJCO^v4«4 ITIK*!/^
p>«>aror>> co f- -o
•4 ,
1
'. 1
omo-oo
O^ O1 o O K> rOf^5r^i«i
i
t
i
7 •'. *7 «7 •! • T «~ *7
*O O* *4 ^ F** , CO f^ «N
tf> eo ru meo f>con
•4 «^ ^ «-t *-
i • i
f j
U"t f-* -« ^0 O o r-o *•
!
1
1
•^ «T
i h-O>
l/\ 1*1
> 0 C
_ -
lf^.a
•
i
k
'
*^S?» SS^S3
o-<>o-oi/» o-a-c
o-tro-OAj O--OJ
• 4 a
l O 0
> r» fi
K
k
11
i
iMAtp^^rin -or-cooo
AIAIAIAIAJ AJAlAfAftf
«
O
CO
•o
CO
-fl
•0
9
I •
a-
i •
•^
c>
. •
a-
AJ
• •
co
CO
«••
cd
13
§
•H
U
•H
v+-t
co
C
•H
I
ex
13
-------
(9
K
m
CO
5EO
O >-
u
in a
— c
u
o
Ul
oz
>-
or
«•* -C -A
o- o m
sr -o in
•O A| Al
r- r* -o
O K> O
m OAJ
V t\l ffl
Kl Al IT
2JC-
'-« co f
r »r •
!
1
i
r*>-0
r*> -O
•O t~
r- AI
*4
O Al
co o
-.-O
;5?
J — O
«,
o ar ro
i>- ^- in
in AI fA.
=J — . O
AJAJCJ-
-O W »A
in *r ^
£££
0-O'C
1
i
1
-0 CO
->0
1
^ t'
-O 9
o sr
•O 00
o- in
-«r-
•a KI
>*> tr
i
o> o- o-
O- 0 0-
o- t> o-
in AI sr
— O Al
•« CO t*
(T CO O
Al .O H>
in M »n,
o o- in
A: f» U"
Al AJ —
o- o o
1
i
i
COO
O- Al
tn o^
i
i
r«. -<
O- t>.
1
eo !T
H> r-
H> 5T
— a
o- oc
o c
ST n
• ««.
»
i
•
•o in sr
AJ — f-
«-• sr \n
«MO- O
•o •« — i
^- «v o-
•o-o o
"O Al M
££5
IT 0 <
: * •;
S
.0 «r
?w
w «
in o
o«n
j
r-Ai
or-*
O l*i
•to «o
in KI
1
1
>C U
r «.~
r
>
sr co o
o-o-o-
-•in o
«• AIf«
in o* AJ
0 — O-
a- «r co
AJ W»l Ai
o- A; c
CO K» —
,^^»
A|
M
-O
-o
•o
00
HI
C
Q
r-
0-
a-
r-
oo
v4
•«
^
^
<>• W^ -HP*
0- ^5 =T
^ C Al
O —
!•»• O -O
— Kl Al
Al AJ >O
O O O
in si in
r-in in
in oc a
•O'— «
r «T «
!
» to
0 0
-*o~
XT r-
^ -
--0
m CO
Al fs
t^ O
9 N
5T IS
1
I
f
i
9 &- in co co r*^ f^ ^^ 47 Ai ic-oor-eo
Bineo I
,oO • ;
0 •«•» I
O. a '
^ •-
LJ '
i~H <^ ^^ co ^5 *^ ^> f* o -»^ ^^ -A o* ' ^r ^ o* in in
iH •»« '-o— «>ooco m^ieoi^co A**^-o^in
^ —-•-.-• —A,— AJAI ^(V— — —
H i 1
^» ' ^5 ^« ?T f^ -jy ^T PA Al ^T -^ ^0 ^^ ^* *& **^
QT *•••***••• » •* •••'••* •*•*•*•*•••
7 co =j -^ ox co wftmfio %o in oo *• i/^
yj ^v^f^O*^ O -O -O «-• OO »•• U"* t/"» -"•»••
u : ;
a: • ' , i :
ikj i«« AJ *A sy>m ' -^P*- co o- o i *^ AT P^ sr in
4 CO COCO CO CO COCO COCO CO CO CO CO CO CO
o
1
in o -c
: < «i 4
in «> AJ
K> r- !•-
•*"* "
ro -O -O
^> o-
-.»« 0
t>-~ A,
ro Ai -O
o «o —
Al— A
.I--0
* *f *•
r* >o r*
A*I/» q
^^«
CO 00 OC
-C S3
C «.~ 4
O" 3T
•rf -H
A] AJ
r** t*
o* -o
5T »H
m a
sj *o
cO 00
t-
i
•
>
i
.
^o *y»o*
• «r «» •
•O O* AJ
J"» O =3"
-*•*-*
r-.
eo 7 f—
— t tn pfl
OlAf-
o «-4 aC
Al *-«
irr^r-
» *• *•
O^ €O IT
O O -
i** A; *
(V AJ ^
coco a
^*-^
• •* •
ino>
^f O*"
— 0-
1
i
1
— o
r- e»
— II
| i
i 1
i
i
0*0
NO cO
CO O
**
j
>* •• *
t GO OQ
i
|
• O'«* «a 9 CO K
°" ** **^ ""
^5? *4 A* *fi
oo fO eo ao 1/1
U>cO ^T O «O
»•* Al f^ Al *"
r* A» -o co A
»* *** in ^IT *•
^^ *^ O* OO A
— AJ— — A
• — sr or>o
* •• ^ •* •»
^- Al Kl J^ C
v» AJ Al l*> A
Al Al Al Al f*
CO CO CO CO CC
1
1
k
r
i
1
i
i
&
r •
0*
CO
—
•O
'JZ
—
cO
^
0^
**
o-
» •
-0
AJ
>-
CO
14
-------
o
or
i r- -o eo r- in iiri«»-«
! i !
o co co r- in •«• in o =»• oo
to
CO
X
Ul
_J
f-Z
3:0
13 >-
-o co -< -< o o o ru ru f» o-ooroo . KI •• •« ru
co co r*
CO (M
i
1 ;
! ' '
o =r a o -a
i '
| i j
| «
in f\ ^- sr o
o o o o
in
IA
i
ru
*o KI ru in ~«
^) M O -O O
v4
j
| !
^ JJ o o r-
o o o o co
M r~ a- in sr
1 ;
=o ru f> r- o
m o ^ o in ru N
sr x — • f r*< IT co
t- co ,— — ---«
0° i
Ui
oz
O < i CO O O IT O
o o to r- co O
i*> in o -« o
*-« ««-» V4 *« O
} '
!
•>• '
a: IT. o co -< o
uj =r o in o o
T" CO ~* *Q O O
"1 1
i
i
1
! '
O '
{£ f.. K-,,. IJ «J
co in -< co »-t o
co co in "i o in
3
U , i
-J 1
*—N
t> !
- r- sr rO o O
(V »•••**-.»•
z ru co in o
eo IT -O IM —
C cc =T r* -•
(M CO — «P~ CO
"1 oo * f\J fl
o
n- e r^. r\j o o
10 ^ fUfA CO
! « i
i l ;
in -o rjin rvi
in in r— ^-im
CO ^ —• -« O
'
M O -O «\l O
* f •" * «« •
in a- f> ^> in
r- in -o r*. ?•-
*
i ' i
.or-- » «
OOOOO
f~ o o -< o
in -o in o o
o
i
l
! '
m •** o -^ o
K< o o in o
O -O CO -» O
•* o
Kl !•». fw -^ O*
eo CJ in ^ O
f- o h- <\j o
-- -»
; |
! 1
i» «r>) ^i~-
«»••»•«•«•
t~- O —.-«•»
w in K» in o
IM AI »« sr in
OOOOO
O O O- IAO
o o o in in
o o o
i i .
i ; i
o o -o -«in
o o in r- -»
o o .0 -a in
oo
!
! !
or- =x fi co
f- O K CO ^
fl IT CO ^- O
; i ;
\f\ *c rj ^c r«- o in o
IT o vn r^ K>
~o r- co in r~
! • '
o -o m o -<
—i KI -o o in
in -< M =T •<
:
I < !
i ' i
]
.-ir-- =r-o eo
oin r- Ai •*
o r« fM -• -*
i ;
1 ' !
— CO o rO in
cr -« -o =r in
t». -<;r o K»
i ! !
=r r* -*co AI
fO IP *« O O*
cr c> i/> =j o*
1
i
r- inr- AJO
f m' «r •* •* «
<\> ~o >-• in i/»
! ' i
i .- /v K> cr i/i
OOOOO
ro o r— -o in
o co >o 10 r*»
1
«o co (V =r «J
CC -O (M -O IO
co eo o «r irt
t i
r* O -O fl a
in oo -< o fl
in >^if m v
»«*•»•- •*
IT =r r- r~o
1
1
!
1 :
i
^•r- «o o-r
, a* o o* a* o
i
o
-------
u
oc i
O >r\J
in i
U
o
t-
UJ
02
IO >O ^ ^ *O *? ^ ^ ^ J"* ' ST »O -^ ^*
1
i
I
|
1 ,
i
i
o< -~c-~~r^-O f\j «Mir eo K» ;srir-«o
-O
0
•O
; f- in ^ in w »o, AI to. a1 AJ. wrumsro
cr
!
1
1
1 !
• IT IT ^-«r «c
K15TOAJO1 C1 O- O* <> C-
v4 0 0*0*^^- <^>
1 i
: I
1
i
! t
' 1
i '
1 |
1 !
i
o o =r IM -« o fM o IT r^. o^ oj GO K< ?^-
-^CO h^t l/» CT i^O'OAjAl (M =J -O CO CT
; |
i !
' i
i
t
t • ;
.
— CO t~- O C- O- OO-W'O'O1' O- O- O* O- O- OO-O-O-O t> O» O- O O* oo-O-OO"
o
u
oc.
i
j
i
1
1
i
1
1 1
J
< 1
j
i
;
• O^cCI^AJ^ i9O>v*O l/> 9' (X <4! Ai
CT '• •" <" •• »r • :•;•.'• •. • :•*•-<•<
CO ' f^ CD f** f\l O* >O CO ?*• CO m A* *^ v4 CC
) O
T
r.
,
« j
1
!
1
!
. -C ?**• O «-*'O
1
i i
i ;
1 *
K- CC O O =T A; IT ^0 ' — l/»
• • •* «r •: «T * 'r «7 «e •' •* • r •*. «r «T «r 4
» IA ^ ^o AJ ••o > *r c* o* ^" ^ *^ •** ^f ^ ^>
(/} ; ^ p^ ^K«MO COO*^^CO lO'^Ov^CQ •^9'^«p«Ot> AjrVi^fO. «^ EO O «-• O^ IT
••i -< -< ,
UJ
H-Z
Z O
u
mo
^ zr
in co
•0 O
.u
o
UJ
!
!
*M -O IA «O -^
=»i-r-
CO AJ
•«•<•<
in -• K
*^ *4 ^ ^i O^ — * *^ , *—*^,
^ Al
o- rv, ^ ro -« c> <> r- ^ rvj o o PJ o
t/> ^> ^ r*> ^ ; eo r- o CD rvj ir» o ru h-
•4 "^4
|
j
|
1
w« «-t
i
i j
' !
•« «^ v*
«•
K
in - K> LI *r
miAo^o* C'd-o-o-O1
h
r*>f-^^>rotn in ^> ir» to o o* o o o OH
> - CO -^> O -O«£i ' -O F** *^ 1/1 O* ' O- ^ O* O^ O
; ~+ ~+ Q*> ,<^1O m o^ *o ^*^ co t> o^ ^ ^ o^ m^-o-oco '-^c^o^^^ iin AJ in ?n in AJ f\i ^ v*» in
to ^ -otno-roo LO=T-IMC> cr -o -o d
r H
1 =7tf>00£70 CT ^> 0 0 OD «T -*Kt ^0 ^
&•
o>
_«-•
^J
(M
«>
0
o>
A;
T •
•0
o-
0T
o-
o-
o»
V
r^-oo-o-o- o-c-o-o-o- o- o- a-o- o
— in — o- cr-
oo-r~o-e>
. «• o>e>
o- o- o- o- o-
o o- o- o o-
o- o- o-o- o-
o o- o o- o
o» o- o- ox e>
O" O- O- O- O
o o o o o
O* CJ- O- O- O-
o- o- o- o- o
o o- o o- o-
o- a- o- o- o-
O-C-O-O-O"
O O- O- O- O-
' t>
O
; «>
u
>-
•<
o
-Or-oooo
lit
f\i M w ru rv
o o o o o
-^ >^ eti o- o
ru rufu IM M
O <5 O O O
16
-------
Table 2. SUMMARY OF FRONTAL CHARACTERISTICS
n*TE
LOCATION D]R SP
TYPE INTENSITY
JUL
JUL
JUL
JUL
JUL
JUL
JUL
JUL
JUL
JUL
JL/L
JUL
JUL
JUL
JUL
JUL
-UL
JUL
JUL
JUL
JuL
JUL
jUL
JUL
-HA
JUL
JUL
JUL
JuL
JUL
JUL
JUL
JUL
JUL
JJUL
j ' I |
J ^ L
Jol
AUG
AUH
AJG
AUG
lrJG
AUG
AU-,
AU&
*UG
HUG
AUG
AUG
«UG
i
3
3
4
4
4
5
5.
5
5
9
9
11
11
10
1 C
10
11
IP
14
15
13
?1
21
?1
22
25
27
?;
27
27
2&
2S
29
29
2 9
29
1
1
2
2
11
11
12
12
13
14
14
14
14
if
15*
2C
2
It
It
?
6
11
12
2C
If.
10
11
3
1<
1 ''
1
1
19
16
20
'£
2
19
22
7
6
9
18
20
14
lo
t
7
h
It
21
2
3
^
7
19
19
13
3
1?
13
17
COSt-oCTON
KANE
MC ht^Y
LtMSnUr-o
C(ji>HUCTON
LEWlSrURQ
CDSnGcTCN
KA'-.E.
MC HE'<-: 5
'.Xl4
> W22
c 2
M 3
? 13
k- 1
N 5
t 1
".W 4
' WKJ
"W 5
r W 9
f 5
"E 3
EW23
l W 6
' WIG
swio
' «10
•;W13
Nw 3
».w 5
\Wi4
r w 6
Nw 3
f> 3
k 4
"• i
W 8
•,W 7
f.W 4
v 13
f 22
S 1
l, 0
S 1
^ 1
S 3
.3
.4
,B
.0
.6
.3
.A
.9
.2
.6
.6
.6
. 4
.C
.5
.0
.0
.4
1
.9
.6
. 6
, J
, 5
.3
.2
.6
.9
.9
.4
.4
.0
.2
.6 '
,1
2
.5
.9
.9
.9
.8
.6
.4
.4
.2
.9
.9
,4
.4
,0
JOLiJ
COLD
COLD
COLD
WA^M
WAPM
COLD
• COLD
COLJ
JOLU
COLD
CuLD
COLU
COLL
STAT
S;AT
STAT
WARM
STAT
CCLD
COLD
COLD
COLJ
COLD
COLD
STAT
WARM
COLD
COLJ
COLD
COLD
COLD
COLD
COLD
COLD
T f n
>-OL 3
COLD
COLD
COLD
CULU
CULD
COLD
COLD
COLD
COLD
STAT
STAT
STAT
STAT
STAT
WEAK
WEAK
WEAK
WEAK
WEAK
WEAK
WE'K
WtAK
WEAK
WEAK
WEAK
WEAK
WtAK
WfciK
WEAK
W£4K
WEAK
WEAK
WEAK
WEAK
WEAK
WEAK
WtAK
WcAK
WtAK
WEAK
WEAK
WEAK
WEAK
WEAK
WfcAK
WEAK
WEAK
WEAK
WEAK
W [- A K
WEAK
WE-AK
WEAK
WEAK
WbAK
WEAK
WEAK
WEAK
WEAK
WtAK
WtAK
WEAK
WEAK
WEAK
N
N
N
N
DC:'
DEI
N
»J
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
DEI
N
N
N
r,n t
DE1
DEC
N
N
N
N
N
N
N
N
N
N
N
N
N
u
C
c
c
;R
;R
C
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
:R
c
c
c
- p
:R
c
c
c
c
c
c
c
c
c
c
c
c
c
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSF
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
OIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
WITH w A v L s
DIFFUSE
WITH W»VfS
WITH
DIFFUSE
WITH WAVES
WITH
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSt
DIFFUSE
WITH
DIFFUSE
DlFFuSc
DIFFUSE
DIFF USE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSt
DIFFUSE
DIFFUSE
DIFFUSE
17
-------
Table 2
?»TI
»UG
AUG
AUG
*UG
AuG
«UG
AUG
AuG
AUG
iUG
4UG
AuG
AUG
<=£»
EgP
:EP
'.ia
CE"
st»
Sfc»
Sfc»
CfcP
Sfc»
Cj..
r- L- P
SE°
5E"
~fcP
SEP
SEP
SfcP
SfcP
SEP
sfcp
Sfc»
SEP
S£P
ccP
s£P
SEP
SEP
?E"
S£p
SEP
SEP
SfcP
e
1C
19
19
2
12
4
lS,,UR(i
CCSHOcTO.M
KANE
COShOcTON
KANF
COShOcTON
MC HE^RY
Lr-k'IiHURG
COShOcTOM
KANE
CCShGcTON
KANE
MC rE'irty
LE-ISf-UPG
COSHCCTON
CObHUCTCIx
CCSHOCTOU
CObnOCTON
DIK SP
K 14.
t< 7.
NK ».
NK 8,
w 10 .
NWi3.
\W 8,
c 9.
SK13.
5W t.
v. 1.
w 14.
''W20.
f W13.
N W 1 2 .
k'K 4.
NW14.
FW 9.
M"13.
»'W11 .
t.wll .
N 4.
N 7.
11 9.
*,« 9.
NW 5.
N 3.
SW 5.
pw 2.
- 4-
NW11.
NW 5.
NW 9.
* w 10
SW30.
*» 19.
\w ».
^ 3.
* 13.
N 7.
NW 8.
f 2,
». 6.
'. 10.
f.E 5.
? 3.
N 3.
S 6.
N 5.
8
9
2
8
4
9
3
3
4
9
6
8
4
2
0
G
4
3
0
8
a
4
4
7
5
8
5
6
3
9
1
3
7
.2
1
4
0
7
9
6
8
5
0
9
6
2
2
5
1
TYPE
COLD
COLD
COLD
COLD
COLD
COLU
COLD
CULD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
COLD
cai u
COlD
COLU
CuLU
COLU
COLD
COLD
WARM
WARM
OCCL
COLD
COLD
COLD
STAT
WARM
WARM
COLD
STAT
COLD
COLD
COLU
STAT
COLU
COLD
COLD
STAT
STAT
STAT
STAT
INTENSITY
WE*K
WEAK
WEAK
WEAK
WEAK
WEAK
WEAK
WcAK
WEAK
WEAK
WEAK
WEAK
WEAK
WEAK
WEAK
WtAK
WEAK
WEAK
WEAK
WEAK
WtAK
WfcAK
WEAK
MOD
MOD
WEAK
WEAK
WEAK
WEAK
WEAK
WtAK
WEAK
WEAK
MOD
WEAK
WEAK
WEAK
KtAK
WEAK
WEAK
WEAK
WfcAK
WEAK
WtAK
WEAK
WEAK
WEAK
WEAK
WEAK
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
c
C
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
c
. c
c
c
c
c
c
c
c
DECR
N
N
N
N
c
c
c
c
DEC«
N
N
N
N
N
N
N
N
N
N
c
c
c
c
c
c
c
c
c
c
CHARACTER
DIFFUSE
DIFFUSE
DIFFUSE
WITH WAVES
WITH WAVFS
WITH WAVfcS
WITH WAVFS
WITH WAVES
WITH ,(AVES
WITH WAVFS
WITH WAVES
WlTH WAVES
WITH WAVES
DIFFUSE
DIFFUSE
DIFFUSE
DIFFUSE
QUASI-STATIONARY
DIFFUSE
WITH WAVFS
WITH WAVFS
DjFr USE
DIFFUSE
WITH WAVES
WITH WAVFS
WITH WAVES
WITH WAVES
DIFFUSE
DIFFUSE '
ACTIVITY UNCHANGED
DIFFUSE
WITH WAVRS
WlT^ WAVES
DIFFUSE
DIFFUSE
DIFFUSE
WITH WAVFS
WITH WAVES
DIFFUSE
DIFFUSE
DIFFUSE
29 22 KANE
15.3
COLD WEAK
N C
18
-------
the frontal systems were classified as weak, with generally decreasing
or unchanged intensity.
4.3 Characteristics of Pressure Systems
Pressure systems can be characterized in several ways: The intensity
of the pressure system, i.e., the magnitude of the departure of the central
pressure from standard conditions; the pressure gradient, i.e., the rate
of change of pressure with distance; the areal coverage; and finally, the
rate of movement of the pressure system. An analysis of the areal extent
of several high pressure systems during August and September was attempted.
The area enclosed by the 1020-mb isobar and the magnitude of the central
pressure were recorded at twelve-hour intervals. The ratio of the maximum
pressure less 1020-mb to the area enclosed by the 1020-mb isobar served
as a measure of the pressure system intensity. The areal coverage varied
greatly from analysis to analysis. The ratio showed no relationship to
the ozone concentrations at McHenry. Attempts to recognize a relationship
between the intensity of central pressure or the areal coverage with ozone
concentration appeared fruitless and further research was abandoned.
The movement of high pressure systems during the four-month period
are presented in Figure 3. Fronts occasionally crossed the study area.
The low pressure centers associated with fronts passed generally west and
north of the study area leaving only the high pressure systems to penetrate
onto the study area.
4.4 Ozone Concentrations and Thunderstorms
Twelve-hour average ozone concentrations at McHenry were examined for
those days when thunder was reported at Pittsburgh and compared with the
previous and following day's concentration. (Pittsburgh was the nearest
National Weather Service office reporting thunder in its LCD) . Usually,
concentrations decreased following the thunder. At the time scale investi-
gated, no particular relationship of the two quantities was found.
4,5 Ozone Concentrations and Sky Cover
Several measures of cloudiness—daytime cloud cover, percent of
possible sunshine, and the hours of sunshine—were plotted for July through
September from the Pittsburgh LCD. In comparing those plots to the
twelve-hour average ozone concentration at McHenry, no consistent common
trends of the data sets were noted. This result does not suggest that
19
-------
July
August
September
October
Circle md.cates pOMtion of center at 7.00 am E S T Figure above circle indicates date, figure below, pressure to nearest millibar.
Dots indicate intervening 6-hourly positions Squares indicate position of stationary center for period shown Dashed line in track
indicates reformation at new position Only those centers which could be identified for 24 hours or more are included.
Figure 3. Tracks of centers of high pressure systems at sea level,
by months, 1973. (Source: Climatological Data, National
Summary, 24, 7-10, Environmental Data Service, Asheville, N.C.)
20
-------
sunlight, especially in the ultraviolet portion, has no effect on ozone
concentrations. The reported variables are not good indicators of the
light intensity, wavelength or amount.
4.6 Time-Section Analyses
4.6.1 Potential Temperature
A common technique used to analyze the vertical structure
of the atmosphere as a function of time is the vertical time section. It
is advantageous in such studies to examine conservative properties in order
to diagnose the ongoing processes. The potential temperature, 0, given
by the equation ,
C
where T is the temperature (K) ,
P is the pressure (mb) ,
R is gas constant for dry air, and
C is the specific heat at constant pressure for dry air (R/C = 2/7).
In adiabatic processes, 9 is conserved. Its distribution indicates the warm
J\ Q
or cold conditions. When the vertical gradient of 6, — , is large, the air
96
is quite stable. As — decreases, the air tends toward instability. The
o z
time section of Q show these changes. Since 9 is conserved in adiabatic
processes, downward movement of a 9 isopleth from one observation to another
indicates descending motion, and/or warm air advection. A basic discussion
of the use of 9 as an analysis variable is given in Appendix A.
4.6.2 Dewpoint Depression
A measure, although not conservative, of the amount of
moisture within the air, is the difference between the existing temperature
and the dewpoint temperature. When the depression is small (<_ 3°C) , the
air is nearly saturated. When the separation is large (> 15°C) , the air is
quite dry. The altitudes of these moist and dry conditions were computed
for each rawinsonde sounding.
4.6.3 Significant Temperature Layers
A third characteristic in time-section analyses is the
presence or absence of temperature inversions; i.e., layers in which air
temperature increases with altitude. Inversions are layers of stable air,
resisting transfer of momentum and mass across their boundaries. Aloft,
they often mark a boundary between types of air. Those inversions which
21
-------
persist, indicate that the contrast across them is maintained. Their
presence also indicates the potential for trapping ground-based emissions
and increasing concentrations. Lowering the altitude of the inversion,
decreases the dispersion volume, leading to high pollutant concentrations.
The longer the inversion persists, the greater the opportunity for
degrading the local air quality. Therefore, the altitude, motion or per-
sistence of the inversion layer, are documented and examined in relation
to ozone concentration. Surface-based inversions often develop during the
night due to radiative heat loss at the ground. These are quickly dissi-
pated by solar heating on the next day and are not considered in analyses
which are presented.
Superadiabatic layers, i.e., those layers where the temperature
decreases at a rate greater than 9.8°C km"-'- (the adiabatic lapse rate) are
ideal layers for transferring momentum and mass from one altitude to another.
These layers are common near the ground on a hot summer afternoon, but are
infrequently found aloft and seldom persist for any length of time because
of their unstable characteristics. The reader is referred to the discussions
of Appendix A.
4.6.4 Computations and Presentation of Results
The vertical time-sections of potential temperature, 6,
dewpoint depression and significant temperature layers were computed for
Pittsburgh, Buffalo, Huntington, and Dayton for each sounding from July 1
to October 31. The analyses were drawn from computer printouts of data at
100-meter increments in the vertical after each sounding was analyzed inde-
pendently of all the others. The analyses of the time-sections of 6 and
of temperature inversions that could be identified in two consecutive
rawinsonde observations are presented in Figure 4. Cross-sections of the
dewpoint depression were plotted in a similar manner, but are not given
here. The time and type of frontal passage at each ozone monitoring location
and a plot of the twelve-hour average ozone concentrations are shown on the
upper portion of Figure 4 so all of these data can be examined concurrently.
4.7 Data Interpretation
The plots, by month, of the ozone concentration at each of the four
observing stations indicate that Lewisburg and Coshocton normally exhibit
a fairly strong diurnal trend, i.e., high concentrations with the 0000 GMT
22
-------
fr
C
>>
1-1
3
•a
M
3
J2
W
4-1
4-1
•H
PM
4
4
t
4
4
4 _
4 ~
tii
4
4
4
4
4
4
iz:
4
111
4
4
tn
n
Li
o
o
0
0
01
O
0
CNI
o
o
( ra/§rt)
O
QJ
•H
H
auozo
0)
o
QJ 0) rH
4J 4J td
(I) 4->
iH a c
n) -H o
•H T3 J-i
w c m
C -H
a) M-I
4-1 OJ O
o j:
CX 4J 0)
o
•H w --a
4J r-l C
M CO cd
Q) >
> M •>
OJ 0)
-a 4J g
a c -H
(T! -H 4J
O 3 Q)
•H O 4J
4-1 X fl)
Cfl I P
)-i 01
4J >
C rH •
0) 0) ^3
U § 4J
C 4J C
O Q
o 4-1 B
n)
a) C
C C 0)
o o >
N -rl -H
O 4J 60
3
O) XI 0)
00 -H X!
a t-i 4-1
QJ
t-l
00
23
-------
•a
M
I
O
a
CO
3
60
•H
PM
o
O
-------
S-i
c
cu
so
M
QJ
•i
0)
4J
PL.
01
•a
n
3
J2
to
•H
PM
o
o
o-
o
o
ro
o
o
o
o
-------
00
C
0)
33
O
S
OJ
XI
O
3
,0
03
t
1
t
1
»
(
t
4
4
«
f
t
<
f
»
1
t
t
t
4
t
t
t
t
4
t
t
(
4
4
t
4
4
4
4
4
4
4
4
4
»
4
4
4
*
4
O
O
-tf
4 •
4
4
4
4
4
4
4
4
4
4
4
_ 4 ._
4
4
4 __
t
4
4
4
4
*
4
4
4
4
t
4
4
•4
4
4
4
4
4
4
4
4
4
4-
4
4
4
4
*
4
4 "
4
4
t
4
4
4
t
4
4
1
O
O
ro
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4 __
"4
4
4
4
4
4
4
4
4
4
t
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
0
O
CN
1
T
1
T
1
1
1
I
4
4
4
4
4
4
4
4
4 — "
4
4
4
4
4
4
4
4
4
4 —
4
4
4
4
>
4
4
4
4
4 —
4
4
_ 4
4
4
:4
4
4
4
4 -, - —
4
4
4
4
4
4
4
1
L
0 C
O
iH
-t-
CD
00
•H
(£m/3rt)
SUOZQ
26
-------
>-,
H
3
O
o
o
o
m
o
o
CM
O
o
rt
M-I
M-<
pq
(1)
6
s
O
^s
-------
•u
en
O
iH
tfl
O
o
si-
O
o
f")
o
o
r-l
O
o
01
0 /t
(U
4-1
g
si-
0)
3
60
•H
( m/3rl)
E
28
-------
.3
cu
4-1
tx
-------
0)
c
(-1
0)
rd
O
4-1
O
O
PQ
o
o
o
o
O
o
CM
o
o
0)
•H
H
13
3
c
o
y
0)
M
9UOZQ
(nq)
30
-------
bO
t-i
3
,0
tfl
•H
s*.
H
3
(3
o
4-J
00
e
•H
4J
O
O
O
o
O
o
CN
O
o
tu
•H
13
-------
00
M
3
,£>
05
4J
CO
3
o
4-J
60
C
•H
4J
C
4
}
4
4
4
t _
4
4
4
4
4
4
4
4
4 ' ....
4
4
t
4
+ u_
4 *™ "
1
4
4
t
4
4
4
4
4
4
+ u —
4
4
4
4
~1. 'I-
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
J
4
4
4
i __ .
4
I
4
4
4
4
4
4
4 _
4
4
t
4
={__ ..
4
4
4 4
4
4
4
" " 4"
4
4
4
4
f
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
._. ^
4
4
4
4
t
. .. f .. .
™i™^^*'t
4
4
O
O
o
o
o
o
CN
O
o
0 /T
0)
§
a
a
3
CxO
•H
Ul/8rt)
9UOZQ
32
-------
60
^
£>
cn
•H
o
o
o
o
ro
o
o
CNI
o
o
a)
i
0)
(U
Cfl
a
o
4-1
bO
c
•H
4-1
§
Cti
o /or
et
p
0
ZT
o
2t
0
^t
o
et
0
£T
o
ei
o
£T
o
et
0 /BT
0 /LI
/tx
/Zt
/ ^ t
o /OT
ZT/6
0 /6
0 /8
^t/i
0 /i
0/4
ZT/S
o /s
a /v
ET/C
o /r
0 (Z
tl/l
o /t
I
o
•^s
-------
00
h
3
w
•H
M
Q)
O
O
C
O
4J
00
•H
4-1
C
3
4
4
4
4
4
4
4
4
4
4
4
4_
4
4
4
4
4
4_
4
4
4
4
4
4
4
4
4
4
4
4
4
4
t
4
t
4
V"
4
4
4
4
4
4
4
4
^ .
4
4
4
4
4
4
4
4
*
4
4
4
4
4
4
4
f
4
.4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
1
4
4
4
4
4
4
i
4
4
4
4
I
4
4
4
1
4
4
4
4
4
1
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
- 4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
o
o
o
o
O
o
CM
O
o
( m/Srl)
o
SUOZQ
(unQ
34
-------
c
o
a
o
J
Cfl
o
o
a
o
tfl
O
•o
9
4-J
g
O
-*
0)
3
CO
•H
( ui/grt)
C/
auozo
35
-------
B
o
•U
o
o
JS
to
o
U
•U
W
3
00
c
o
tfl
o
(1)
B
*a
SI
a
CJ
-*
3.
•H
( ra/8rt)
36
-------
01
B
13
4-1
§
O
-------
t
0)
"O
0)
O
c
O
u
0)
M
3)
( ui/grt)
auozg
38
-------
average and low concentrations with the 1200 GMT average. Kane exhibits the
diurnal trend with smaller range and with less regularity, while McHenry shows
a diurnal trend which is easily overridden by some other contributory factor.
Coshocton, on the open rolling terrain of east central Ohio, at an
altitude of 354 m MSL, and Lewisburg, in the Greenbrier Valley (705 m MSL),
are located where strong nocturnal inversions can be expected, although for
different reasons. Such inversions effectively limit the vertical dispersion
of ozone and permit its destruction by ground contact or surface-based
emissions of destructive agents.
The McHenry monitoring location (altitude 884 m MSL) is higher than the
surrounding terrain. The depth and intensity of the nocturnal inversion may
be substantially less than would be observed at the other locations. Airflow
over the rougher terrain could have more opportunity to weaken the inversions
which occur, permitting the air at the ground at McHenry to become mixed with
air from aloft, and to maintain higher ozone concentrations during the night.
Kane, although at the highest point within its locale (630 m MSL), is
at a lower altitude than McHenry and is in smoother terrain. Nocturnal
inversions may be deeper and the airflow smoother, so that ozone from aloft
is not brought down as frequently as may occur at McHenry, and thus, increas-
ing the diurnal range.
The time and type of frontal passage at each of the stations are indi-
cated for July through September. At Kane, the ozone increases prior to
cold frontal passages on July 10, 14 and 21, and decreases with or just after
the front. Those fronts show sharp temperature contrasts to above 3 km.
At Coshocton, the limited evidence shows a similar relationship of
ozone changes with fronts in July. The Lewisburg and McHenry records are
incomplete in July. The strong association of ozone and fronts found at
Kane is suspected to have occurred at McHenry, because of the intensity and
depth of the fronts passing Pittsburgh.
In August and September, at all of the locations the ozone increases
and decreases are weakly associated with frontal passages. The fronts are
also weak and do not extend above 2 km.
The major stagnant high pressure situation occurred in conjunction with
a high ozone episode in late August and early September. During this time, high
ozone concentrations (> 200 yg/m ) were observed at all the stations. At
39
-------
•3
Coshocton, a maximum value of 348 yg/m at 0000 GMT on August 28, culminated
o
a 36-hour period of ozone concentrations exceeding 275 yg/m . For eleven
o
days, the twelve-hour average ozone concentrations exceeded 160 yg/m at
McHenry (see also Table 1).
Apparently, the subsidence aloft, associated with the stagnating anti-
cyclone, contributes to the potential for high ozone concentrations, near
the ground. Duration and intradiurnal changes are more of a local phenomena.
The descending motion and increasing ozone concentrations near the
ground might be associated with a stratospheric ozone source. However, the
evidence seems to the contrary. The slope of the 310°K 8 surface at
Pittsburgh indicates a descent rate of 2.2 km in 84 hours or 0.73 cm s~^-,
which is too small for effective downward advection of ozone. Furthermore,
the descending air tends to form a stable layer or an inversion inhibiting
the vertical exchange of ozone.
After August 29, the inversion aloft dissipates, leaving the atmosphere
below 5 km relatively unstable. Deep vertical mixing is quite possible. The
stability is less at Pittsburgh than elsewhere, which may help explain the
persistent ozone episode. The slow decrease of the ozone concentration is
accompanied by the slow increase in the stability (as judged by the vertical
separation of the 305 and 315°K surface).
In some instances, the ozone concentration increases when the 9 near
3 km are decreasing in altitude (Kane-Buffalo in July). In some instances,
these two variables increase or decrease together (McHenry-Pittsburgh in early
August) or are lagged slightly (Coshocton-Dayton in August).
The Lewisburg-Huntington data pair maintains a character quite different
from that of the other stations. The ozone concentration is strongly diurnal,
but shows only occasional periods in excess of the standard, all occurring with
the 0000 GMT averaging time. The occurrences of high and low ozone concen-
trations appear unrelated to the fronts, pressure systems, the vertical struc-
ture of the atmosphere, or to findings at other station pairs. One cause
may well be the separation of the stations from one another. The cross-section
data suggests that Huntington was in air more characteristic of the western
slopes of the Appalachian Mountains whereas the trajectory data (Section 6)
indicate Lewisburg was in air more characteristic of the eastern side. In
retrospect, Greensboro, North Carolina might have been more representative of
the airflow characteristic into Lewisburg.
40
-------
5.0 OZONE CONCENTRATIONS AND THE MIXED LAYER
Transport and dispersion of atmospheric pollutants depend upon the
properties of the mixed layer of the atmosphere. Friction and convective
processes disperse the pollutant. The wind, changing with altitude, trans-
ports the material. Without prior knowledge of other similar efforts, the
influence of the layer, as shown by its properties, on the ground level ozone
concentration was examined. These properties were deduced from the rawinsonde
ascents. The same pairings of ozone monitoring locations and rawinsonde
locations are used.
This part of the study explores, from a statistical viewpoint, the
influence of the boundary layer properties upon the average ozone concentra-
tions, and their relative importance. If a parameter shows strong or
consistent statistical influence or correlation with ozone, a physical reason
may be indicated which, in turn, may help "explain" high ozone concentrations
in nonurban locations.
There are necessarily some dangers in extrapolating mixed layer
properties from Buffalo to Kane, or any of the other three data pairs. Mixing
depth in central Pennsylvania is not the same as near the shore of Lake Erie.
Lacking anymore detailed information, it is necessary to assume that the
properties of the mixed layer at the rawinsonde station and the nonurban
locations are the same.
5.1 Mixing Layer Parameters
The research sought to relate the observed ozone concentrations to the
properties of the atmospheric boundary layer; i.e., the mixing depth of the
air, mean water vapor mixing ratio, vector mean wind speed through the mixing
layer, and the parameter x/Qj and the ventilation index for each of the
soundings. The mixing depth, Z . , as defined by Holzworth , is the thickness
in ix
of the air between the ground and the intersection of the adiabatic temper-
ature with the sounding temperature. In the morning (1200 GMT), the adiabatic
temperature is given by the potential temperature of the overnight minimum
temperature plus 5°C. In the afternoon, it is given by the potential
temperature corresponding to the maximum afternoon temperature. The mixing
depth indicates the depth to which mixing due to daytime heating will occur.
When the mixing depth is large, the pollutants should have ample opportunity
to disperse upward. Usually, these conditions are also associated with super-
41
-------
adiabatic or slightly unstable layers of the afternoon. A small mixing
depth indicates poor vertical dispersion of pollutants, thereby tending to
increase their concentration near the ground.
The mean water vapor mixing ratio of the air through the mixed layer,
R . , gives the indication of the amount of water vapor in the air. During
mix
well-mixed conditions, the mixing ratio should be nearly uniform, since it
is a conserved quantity.
Through the depth of the mixed layer, the wind speed, S, and direction,
a, may change with altitude. To approximately the mean transport by the
wind, the vector average wind speed and direction for the mixed layer was
computed from the average west (u) , and south (v) wind components for the
layer. Although the trajectory plots showed direction and speed of air
movement prior to arrival, u and v or S and a show the dependence of the
ozone concentration upon the existing wind velocity.
Average value, q, of a scalar quantity, q, for the mixed layer was
computed by the formula
J
i=2
where J was the number of data points required to define the mixing layer.
The ventilation index, V, is defined by the mean wind speed through
the mixed layer times the depth of the mixed layer. It measures the flux
of air into an area.
Partly to compare the summer of 1973 data with other climatic data,
and partly to include a regression analyses, the quantity x/Q, as defined
by Holzworth for a city of cross-section 100 km was computed for each
sounding. x/Q represents the relative dispersion of pollutants downwind
of a major city and is a function of the mixing depth and the mean wind speed,
5.2 Regression Analyses
The research tried to ascertain the properties of the mixed layer
having the greatest influence on the twelve-hour average ozone concentration.
Initially, one might assume that, since high ozone concentrations occurred
during the late afternoon and early evening when mixing depth and wind
42
-------
speed are greatest, and that since low ozone concentrations occurred
during the morning when the mixing depth and wind speeds are small, that
the mixing depth or the ventilation index, or both, might be good
12
predictors of the ozone concentration as was found in New York. However,
that assumption should be tested against the data. Other variables or
combination of variables may prove better indicators.
Selection of the boundary layer parameters which most strongly are
associated with the high or low ozone concentration was accomplished
using a regression analysis of the ozone concentration as a linear
function of mixed layer parameters. In Model I, the parameters are
Z . , R . , S, a, x/Q> V and in some cases the hour (0000 or 1200 GMT).
mix mix
In Model II, the mean wind components, u and v, were used instead of a and
S. All the other variables remained the same. The regression procedure
2
chosen was the "Maximum R Improvement". R is the linear correlation
coefficent of the regression. The procedure is described in the
1 -3
"User's Guide to the Statistical Analysis System" in the following
manner:
2
"Maximum R Improvement. This technique was developed by
James H. Goodnight; he considers it superior to the stepwise
technique and almost as good as calculating regressions on all
possible subsets of the independent variables. Unlike the three
techniques above, this technique does not settle on a single model.
Instead, it looks for the 'best' one-variable model, the 'best'
two-variable models, and so forth. It finds first the one-variable
2
model producing the highest R statistic. Then another variable,
2
the one which would yield the greatest increase in R , is added.
Once this two-variable model is obtained, each of the variables in
the model is compared to each variable not in the model. For each
comparison, the procedure determines if removing the variable in
the model and replacing it with the presently excluded variable
2
would increase R . After all the possible comparisons have been
2
made, the switch which produces the largest increase in R is made.
Comparisons are made again, and the process continues until the
2
procedure finds that no switch could increase R . The two-variable
model thus settled on is considered the 'best' two-variable model
43
-------
the technique can find. The technique then adds a third variable
to the model, according to the criteria used in adding the second
variable. The comparing and switching process is repeated, the
'best' three-variable model is discovered, and so forth."
The procedure also chooses the "best" N-variable model for which
the terms are significant at the 0.1000 level. Figure 5 shows typical
output of the procedure.
5.3 Results
Table 3 is a summary of the number of terms of the "best" model,
its correlation coefficient to the ozone concentration, and the relative
importance or the ranking of the variables making up that "best" model.
The model for Lewisburg ozone concentrations uses the Huntington mixed
layer data, etc. For Kane and McHenry ozone stations, the data were
further stratified by the observational period so that only 0000 GMT
data are used in the analysis or only 1200 GMT data are used on the analy-
sis. The regressions included only those date/times when all of the model
variables were available for a measuring location.
The mean water vapor mixing ratio of the mixed layer, R . , is the
most significant variable of either model for three of the data sets.
Stated alternatively, of the variables tested, R . is the best single
mix
predictor of the twelve-hour ozone concentrations. This was a most
surprising result. Furthermore, the correlation coefficient of R . and
^ & mix
ozone is positive (Table 4), meaning that ozone increases when R . increases.
If the high ozone concentrations near the ground were a result of vertical
transport or diffusion from the stratosphere, then the same processes would
have brought water vapor down also. The stratosphere is quite dry
( - 2g kg"-*- maximum) and water vapor mixing ratio is conserved in descending
motion. Stratospheric ozone would, therefore, be associated with a small
mixing ratio—contrary to the observations. Thus, the stratosphere does
not appear as a feasible principal source of ozone during the episodes of
high mixing ratio (and high ozone concentration)."
The mixing depth also appears as a variable which should be considered.
There is inconsistency in its importance between the data pairs (e.g.,
ranking 1 and 2 for McHenry and unranked at Kane).
44
-------
• • -a
tj r-
in
• •
I "
o.
UI O- —O -O
< -O o o- «>
>. -
•-f
a.
o
z
<
3
a
N
O
•>-
Or
3
a
u
X
VARIABLE OZONE
N-
z
UI
o
•z.
UI
a.
hi
a
K
o
u.
u
K
rs
o
UI
u
0
•or
a.
O J
if- M
en o
« a
u x
or
o z
u «-
a:
CO
U UI
eo _j
,.-. m
•* <
a. ••-•
ui or
i- «t
en >
h-
Z
UI
>
f-
Z or
UJ t-<
> o
QC o: or
•«•«•<
ca m CD
CE or or
X X X X
M fc* »4 ••*
X * X 3F
M IM N isl
AT THE OTlOOO SIGNIFICANCE UEVEU
b*
Z
<
U
ta«
•u,
n
Z
U
^4
eo
o
u<
X
UI
UI
o
z
u.
UI
ca
_l
_1
<
UI
^
<
X
_l
kJ
a
a
X
UI
>
a
03
^
_l
UI
O
o
X
UI
>
n
o
<
UI
X
fc-
or
ul
H-
»4
U.
U.
O
er>
u
^«
M
i-«
w>
«r
h>
ei
o
z
-<
*
crj
^*
2
UJ
O
^^
ItV
U.
w
o
u
o
iM
CA
.' RECRES:
UI
"_l
o
-t
VI
C
or
u.
A
0
a
K
a.
ui
3
_i
<
»
U.
UI
or
«
n
a
CO
z
•<
UI
X
e«»
UI
fy
<
3
O
Ol
a
X
3
m
u.
a
•«i
in
w
• •
o
•4
O
o
^
•
o
o
«•
CO
o
1 •
TO
O
9
•O
•e
o
•<•
0
r*
•
r-
Hl
in
0-
*4
•O
in
4
ni
9
O
CO
r >
•o
in
•4
CO
*-
sr
p—
r»
o>
•o
•e
CO
r-
*
'O
in
•^
10 in
9 O
in —
pg o-
in p»
r>. ••
in -a
m to
r « r <
sr in
9 o-
o -*
Kl -"
0- »~
*4
ni -O 9
< K»r» ro
cc a-4) ni -o
< • * » •
a. o in o -
^ 9 «4 ^* c>
> eo f- »«ro
* • • •
u. o HI in ro
•*t «T -O-O
-4fU l>- »O
_l - e>- -o
fr- r-'O f>- K1
Z -O«0 -O-O
ui f
ine> o- 10
co m t-o>
iniMo- -•
-oin «o-
c- eo o-o
o 9 oin
o\i
r-intow
r-< ^ r-- to
O-O- -.9
•a co 9 co
f 4r-*-«rt
in 9 >M-<
*
rO -a -« -O «M
IM tn-oivo
o *« o- r- in
r- to 9 co:—
o O O -O O-
o- in-o ~9
oo in in o o
r-.oioio.—
^ment technique choosing the besi
*«w
>
o
M
t
•H
f\J
P4
Q
3
.§
X
«
R
m •
O rH
(U
CO T3
W 0
^6
mple res
gr ess ion
cd o)
CA I-l
± y ru -1 10
uir-»-u»-''-z I
OC •* -A ** ^- ^ <•!
•« eo <•« co >e> z -•-> 0)
= .?iC-o«or J }-i
o o- «o r~ in -< < • 3
co H> _ ^ ^ eo >• J- 60
• .»- «o «M in ui a .H
a ,- .*? .*. .* o a z fl
A o o o < a ci
: i 1111 I -iigs I Ssiss
wo •- .«, -^ 9 •£ 5 S or u .S S iM-cca-o « x N-or > o
00
z
45
-------
Table 3. RANKING, BY SIGNIFICANCE, OF MIXED LAYER VARIABLES,
IN AN N-TERM LINEAR REGRESSION MODEL OF OZONE
CONCENTRATIONS
MODEL I
"Best" Model
Variables/Ranking
03 Site
Lewisburg
Coshocton
Kane
McHenry
Kane (00)
Kane (12)
McHenry (00)
McHenry (12)
Terms
4
4
4
4
1
2
4
2
Corr.
Coeff.
0.597
0.573
0.619
0.676
0.509
0.610
0.676
0.742
Hr
2
1
1*
NI
NI
NI
NI
NI
Wind
Z . R . Speed Dir x/Q V
mix mix A
134
42 3
1* 4 3
1* 1* 3 4
1
1 2
1* 1* 4 3
2 1
Significant at 0.0001
NI - Not Included in the Model.
46
-------
Table 3 (continued).
MODEL II
"Best" Model
Variables/Ranking
03 Site
Lewisburg
Coshocton
Kane
McHenry
Kane (00)
Kane (12)
McHenry (00)
McHenry (12)
Terms
4
6
4
4
3
2
4
3
Corr.
Coeff.
.613
.592
.624
.667
.572
.610
.678
.770
Hr
2
1
1*
NI
NI
NI
NI
NI
mix
1
4
1*
3
1
2
R . u v
mix
4 3
2 5
1* 3
1* 4
1 2
1
1 4
1 3
X/Q V
3 6
4
3
2
3
47
-------
a
i
§
|
13
H
IS
9
pg
H
PS
W
9
Q
^
a
s
w
p4
o
„
pj
H"
H
0
M
£
§
*g
0 H
M H
3 H
i i
O S3
CJ O
U
<. W
H §
M c«a
rJ O
*
flj
rH
ft
cO
H
S
w
o*
""*«••.
X
CO
(1)
rH >
•2
cfl
•H
t>
2
X
•d
s
PS
X
•H
Q
N
PS
cu
•H
CO
O
O
O
CM
0
O
o
CO
o
o
CM
•
O
1
m
o
o
i
o
rH
O
1
rH
0
•
O
CO
LO
•
0
CO
"7
o
1
00
3
CO
•H
1
CM
rH
O
O
o
o
-}-
CO
CM
O
0
rH
•
O
1
0
CM
O
rH
rH
0
m
CM
•
0
rH
l»^-
•
O
rH
•^
0
1
c
o
4-1
O
o
CO
5
00
0
o
H
rH
O
CO
0
0
CM
rH
•
0
1
CO
rH
O
CTi
rH
O
^
*3"
•
o
oo
o
•
o
CM
CO
d
1
CU
1
in
CM
d
00
H
6
i
00
0
o
m
o
•
o
t
CM
O
O
1
o
rH
O
0\
CO
,
0
oo
CO
•
o
u
£
g
"3
s
0
*
o
1
rH
rH
o
f^
o
O
1
CM
O
*
O
1
oo
CO
o
m
CM
CD
rH
•
O
CO
rH
•
O
1
CJ
S3
/•"s
O
O
0)
G
A
S
O
VO
O
(D
1
O
rH
O
1
rH
O
•
o
1
o
o
o
1
o\
0
d
o\
in
o
CO
rH
•
o
1
CJ
S3
^^
CM
rH
0)
VO
rH
d
CO
CM
d
i
00
o
o
1
00
o
*
o
1
o\
o
o
1
^
0
d
i
CO
co
t
o
CM
-a-
*
o
u
S3
0
O
i-i
g
m
CM
o
00
rH
d
i
rH
CM
O
1
CO
rH
,
O
^J.
rH
d
f^
CM
d
cy.
vo
d
^
CM
^
O
1
O
S3
CM
rH
£
g
1
T3
0)
•^
^"^
O
S3
1
U
5?
48
-------
The lack of a decided directional (a or u or v) preference (Table 4)
at any of the locations implies that the direction of the air flow during
the averaging period had little relationship to the amount of ozone measured.
Thus, the recent (six to twelve hour) location of an air parcel does not
seem to determine the ozone content. Wind speed, S, shows only a weak and
inconsistent relationship to the ozone concentration. The mixing depth
also ranked high in relationship to the ozone concentrations. Wind speed,
direction, relative dispersion of the atmosphere, and the ventilation at the
rawinsonde locations are poorly correlated to the amount of ozone measured
at the associated monitoring location.
The rate of synthesis or destruction of ozone in the presence of water
vapor alone may be written
d°
R
,, ; • -
03 dt 3' mix mix v '
where 0 is the ozone concentration
R . is the water vapor concentration, and
mix
k(CL, R . ) is the rate coefficient for the reaction.
3 mix
A regression analysis of the local percentage change of ozone with time as
a function of R . should give a straight line of slope k if the synthesis
or destruction of ozone depends solely upon the amount of water vapor present.
To test the hypothesis that the rate of change of ozone is independent
2
to the water vapor content, a maximum R analysis was run except that the
percentage change of ozone at McHenry from one twelve-hour period to the next
was used instead of the ozone concentration. That analysis picked Z . and
the wind direction, a, as the only two variables of the "best" model. The
correlation coefficient of the model was only 0.49. The correlation
coefficient of the percentage change of ozone with water vapor mixing ratio
was - 0.02, indicating that the two factors are not linearly related. It
must be remembered, however, that the twelve-hour ozone data and the mean
water vapor mixing ratio of the mixed layer were used in the analysis. At a
much smaller time period, such as five minutes, the production of ozone may
be related to the amount of water vapor present.
At this time, it is not known whether the correlation of ozone concentra-
tion to R . is only statistical or whether there is a physical basis for it.
mix
49
-------
Dr. Lyman Ripperton of RTI suggests that the presence of water vapor
influences the production of hydroxyl (OH) radicals in the ozone production
14
process . These reaction rates are such that the water vapor may be a
contributing factor to the high ozone. A definitive answer awaits further
experimentation and evidence.
50
-------
6.0 OZONE CONCENTRATIONS AND AIR PARCEL TRAJECTORIES
9
Previous research suggested that the high ozone concentrations
might be associated with air which had passed near one or more urban-
industrial areas, where ozone and/or its precursor might be abundant.
The average annual hydrocarbon emission density shown in Figure 6 is
used as a guide to the source of anthropogenic ozone precursors. Major
urban-industrial areas are associated with the larger emission densities.
Trajectories, showing the path of an air parcel for 48 hours prior to its
arrival at one of the four ozone monitoring locations, were computed for
each 0000 GMT and 1200 GMT between July 3 and October 31. The plotted
trajectories are presented in Appendix B.
6.1 Computation of Trajectories
A computer program to compute the air parcel trajectories was made
available to RTI by Mr. John Clarke, Division of Meteorology, EPA. This
program was modified to meet the specific needs of this research. The
2
program interpolates in time (linearly) and in space (1/R weighted, using
the three nearest observations) for the value of the u and v wind components
at a given location. The corresponding 2-hour latitude and longitude
displacements of the parcel are computed, and the "new" position, two
hours previous to the "old" position is found. Wind components are
interpolated for the new position and time.
Winds for the trajectory analysis were taken from the 22 rawinsonde
stations at 900 mb, approximately 1 km above mean sea level, for each
12-hourly observation during the four-month period. The 900-mb level is
usually within the mixed layer during the afternoon and very near the
top of surface-based inversions during the nightime hours. It was not
known a priori whether the inversions would be above or below this altitude.
The trajectories computed in this manner are only approximations to
the actual path taken by the air parcels because of errors in the assump-
tions, techniques, and measurements. Daytime mixing and terrain charac-
teristics cause the air to rise and fall relative to the ground, so
that one air parcel does not stay at 900 mb throughout the travel time.
The variable terrain of the study area also means that the 900 mb level
is not a fixed distance above the ground. The analysis technique tends
51
-------
CQ
4J
M
O
•r)
CO
C
o
•p)
CO
w
C
o
u
o
J-l
efl
1
0)
00
cd
VO
s,
52
-------
to smooth data by interpolation and loses contrast across fronts or shear
lines, which have important impact upon the initial wind directions and
hence, all of the resulting trajectories. Errors of the reported winds
may be + 1 m s and + 2 degrees in direction. In light and variable
winds, the errors may be much larger. The plotted trajectories should
be thought of as the "best estimate" of the actual path taken, but with
increasing uncertainty as the time and/or distance from the origin increases.
6.2 Selection of Trajectories
Since "high" or "low" ozone concentrations may have a different
meaning for each monitoring location, the highest (upper decile) and
lowest (lowest decile) ozone concentrations were determined for each
monitoring location, (see Table 5). Trajectories associated with ozone
concentrations in these deciles were plotted on a common map for each
location and are presented in Figure 7.
6.3 Trajectories with High and Low Ozone Concentrations
6.3.1 Kane
High ozone concentrations at Kane are associated with
trajectories from the west and west southwest, passing over central Ohio
and funneled down toward Kane. Only three of the trajectories come from
other directions. The positions twenty-four hours prior to arrival,
extend westward to nearly the Indiana-Illinois border, northward to the
very southern part of lower Michigan and remain almost exclusively
north of the Ohio River Valley. Thus, in almost every instance, a high
ozone concentration at Kane is associated with the trajectory over
central or northern Ohio during the immediately previous twenty-four
hours.
The trajectories associated with low concentrations at Kane are
predominantly from the northwest, principally across the southern portion
of Michigan and from Wisconsin. Several cross central Ohio. Thus, with
several exceptions, it seems that the air trajectory may strongly
influence the ozone concentration at Kane. Parcel speeds prior to
arrival in the low ozone cases are quite variable, some moving fast,
others slowly. In general, the low ozone concentrations are associated
with higher wind speeds than are the high ozone concentrations.
53
-------
Table 5. OZONE CONCENTRATIONS (ygnT3) FOR INDICATED
PERCENTILE RANKINGS AT EACH MONITORING LOCATION.
Percentile
STATIONS
McHenry
Kane
Lewisburg
Coshocton
All Stations
10
86
64
58
47
60
30
112
92
76
70
85
50
140
121
93
93
110
70
168
150
112
123
139
90
218
190
150
165
182
54
-------
0_> 190 ygm
-3
Kane
0~< 64 )jgm
-3
Figure 7. Trajectories associated with high and low at the indicated
location. Squares (o) and triangles (A) indicate the
twelve-hour positions prior to arrival.
55
-------
0_ > 218 ugm
-3
McHenry
0 < 86
-3
Figure 7 (continued)
56
-------
0_ > 165 ygm
-3
Coshocton
03 < 47 ygm'
-3
Figure 7 (continued).
57
-------
> 150
-3
Lewisburg
D_ < 47 ygm
-3
Figure 7 (concluded).
58
-------
6.3.2 McHenry
High ozone concentrations arriving at McHenry appear to come
from two general areas. One area is typical of northwest flow over indus-
trial southern Michigan and northeastern Ohio where hydrocarbon emissions
are large. The air seems to move fairly rapidly over this area before
arriving at McHenry. The second area is to the southwest, with the air
moving across West Virginia into McHenry at slow speeds on the trailing
side of a high pressure system. Over the past 48 hours, these trajectories
have been over the smallest hydrocarbon emission density area of the north-
eastern quarter of the country. With three exceptions, trajectories
associated with low ozone concentrations at McHenry are indicative of slow
moving air, principally flowing from the northwest into McHenry across
northeastern Ohio and southwestern and western Pennsylvania. That situation
is a stark contrast to the faster-moving trajectories which arrived at Kane
from the northwest. The three exceptions are air parcels which arrived
from the southwest rapidly passing over West Virginia, where slow-moving
air had shown higher ozone concentrations. The slower moving trajectories
moved at an average speed of 6 m s or less for the 48 hours prior to
arrival. Many of these trajectories passed near the Pittsburgh,
Pennsylvania area.
6.3.3 Coshocton
High ozone concentrations arrived at Coshocton from all
directions but principally from the west. These westerly trajectories
indicate speeds of 12 to 15 m s~ , traversing the distance from central
Illinois to Coshocton in a period of about 24 hours and are associated with
the ozone episode during late August. Clustered about Coshocton are many
trajectories that loop back upon themselves, indicating that stagnant air
conditions are associated with the other high ozone concentrations.
Most of the low ozone concentrations at Coshocton are associated with
rapidly moving air parcels arriving from the north to northwest and from
the southwest. Some air parcels cross the same areas associated with high
ozone concentrations. Thus, these data suggest that the speed of parcels
is correlated to the extreme ozone concentrations observed at Coshocton;
i.e., high wind speed—low ozone concentration, low wind speed—high ozone
concentration.
59
-------
6.3.4 Lewisburg
High ozone concentrations at Lewisburg are associated
with trajectories coming from all directions and with apparently equal
frequency. Each trajectory shows a very definite anticyclonic curvature
indicating that the high concentration was associated with a slow-moving
anticyclone probably passing to the north of the Lewisburg area.
The low ozone concentrations at Lewisburg are associated with
trajectories which arrived from the southwest through west and north-
west. Most of the trajectories indicated fairly rapid air movement
across industrialized areas, but moving more slowly approaching Lewis-
burg. This pattern, as was the case with the high ozone concentrations,
is unlike patterns observed at the other three stations. Thus, Lewis-
burg appears in a different circulation pattern associated with the high
and low ozone concentrations. The result is not surprising since two of
the stations are located to the west of the Appalachian Mountains and
the third is along the ridge line of the Appalachians and Lewisburg is
on the eastward side of the ridge in anticyclonic circulation.
6.4 Summary
The trajectory analyses do not implicate one particular area
of the northeastern United States as a source of high ozone or ozone
precursors concentrations at the four sampling locations. There
are common regions for some stations, such as northeastern Ohio; but air
coming from the Lake Erie-Lower Michigan regions appears favorable for
low ozone at Kane, but high ozone at McHenry.
Slow-moving air parcels tend to be associated with higher ozone
concentrations and vice versa. However, the characteristics of the
mixed layer wind speed and direction were not dominant variables in the
regression analysis. The trajectories associated with high or low ozone
do not indicate a preference having a maritime origin or any other indi-
cation that they might be associated with a high or low water vapor
content.
60
-------
7.0 CLIMATIC INFLUENCES
The influence of changes of climatic variables upon ozone concen-
trations is not clearly understood. Nevertheless, investigations of two
variables measured at Elkins, West Virginia, for comparison with McHenry
data show the following:
7.1 Temperature
Temperatures in July and August 1973 were 1 to 2°F warmer during
1973 than in 1972. The 1973 temperatures very nearly equal to the 30-
year mean. September 1972 was 2°F warmer than in 1973. Although the
results of Section 8.1, show the higher daily maximum temperatures are
associated with higher ozone concentration the small temperature differ-
ences seem insufficient to explain the differences in ozone concentrations
from 1972 to 1973.
7.2 Average Sky Cover, Sunrise to Sunset
The average sky cover, from sunrise to sunset is the only indication
of solar radiation amounts reported in the Elkins LCD. In July, August
and October of 1972, monthly average sky cover exceeded the 1973 monthly
average by at least one-tenth sky cover, whereas the September data
differed by only 0.1 tenth. The 1973 data exceeded the long-term mean
values in July and August, 1973 and are within 0.1 tenth of the mean the
other two months.
These indications suggest that more solar radiation was available
in 1972, perhaps meaning a greater potential for ozone generation.
However, the reported sky cover does not discriminate between high,
possibly thin, cirrus clouds or low-level stratus. An increase in the
former probably does not substantially reduce the incoming radiation
needed for ozone generation, whereas the latter may reduce the ultra-
violet light.
7.3 Relative Dispersion
Holzworth's climatology of the dispersion characteristics of the
mixed layer (x/£) were compared with computations of y^/Q at the four
rawinsonde stations. The computed values for the Dayton and Pittsburgh
soundings are compared with Holzworth's summer and autumn seasonal
statistics in Table 6. At Dayton, the afternoon soundings indicate
good dispersion with small variations from the climatological values.
61
-------
Table 6. MEDIAN, UPPER QUARTILE AND UPPER DECILE VALUES OF x/Q
(s nT1) VALUES OBSERVED IN 1973 AND CLIMATOLOGICAL VALUES
Season
Time
Summer*
Morning
Autumn*
Morning
Summer
Afternoon
Autumn
Afternoon
Dayton
Median
62
65
70
96
16
17
16
17
Upper
Quartile
174
165
173
215
19
22
21
23
Upper
Decile
420
260
394
410
25
30
28
30
Pittsburgh
Median
96
44
61
80
15
16
16
18
Upper
Quartile
458
122
380
230
18
22
21
25
Upper
Decile
740
206
820
488
21
SO
26
35
*Summer - June, July, August - Climatological Data (after Holzworth)
July, August - 1973 Data (italics)
*Autumn - September, October, November - Climatological Data
September, October - 1973 Data (italics)
62
-------
The summer morning values of 1973 showed somewhat better dispersion (lower
values of x/Q), while the climatological dispersion is slightly better
during the autumn.
Dispersion at Pittsburgh was substantially better than climatology
suggests during the summer mornings of 1973. The autumn morning values,
unlike those at Dayton showed better dispersion than Holzworth reported.
Although the afternoon x/Q at Pittsburgh showed poorer-than-normal
dispersion, the differences do not seem important compared to the differ-
ences of the morning values.
These results suggest slightly better overall dispersion during the
study period during the morning than might be expected from climatology.
Since x/Q does not exert much influence on the nonurban ozone concentrations,
according to the regression analysis, the better dispersion has little
relevance to explaining the high ozone concentration in 1973.
The basis for assessing an impact of climatic differences has not been
established because 1) there are insufficient ozone concentration measure-
ment to establish a climatology for one, and 2) the impact of weather
elements upon that climatology is not known. The reasons for differences
in ozone concentrations at McHenry in 1972 and 1973 remain unresolved.
7.4 Stagnation Periods
Despite the generally favorable dispersion conditions, for the season,
the prolonged stagnation period of late August into September surely
contributed to the increased frequency of exceedance of the NAAQS for ozone
over that frequency in the data taken in 1972. That event was unprecedented
in any summer of the past nine years according to data published in the
Climatological Data, National Summary. During the 1972 measurement period
at McHenry, Maryland, high pressure systems did not stagnate but moved
slowly through the study region at a nearly constant rate.
63
-------
8.0 SELECTED STUDIES
8.1 Maximum Temperature and Maximum Ozone
Jeffries'16 experimental results show that when the daily maximum
temperature remains below about 60°F (15.4°C) the generation of ozone
o
concentration above 160 pg m did not occur. The correlation of the
maximum hourly ozone concentration for a given day (midnight to mid-
night) as a function of the maximum daily temperature was investigated
for each of the four ozone monitoring stations. Since temperature was
not an observed quantity at each location throughout the period, daily
maximum temperature data were extracted from the Local Climatological
Data Summary for Beckley, West Virginia, Elkins, West Virginia, Youngs-
town, Ohio, and Columbus, Ohio. These temperatures were correlated with
the Lewisburg, McHenry, Kane and Coshocton ozone concentrations, respec-
tively.
The scatter diagrams and linear regression line for each tempera-
ture ozone station pairs are given in Figure 8. In all cases, an in-
crease of maximum hourly average ozone occurs with increasing tempera-
ture, but in none of the instances was an ozone concentration greater
_3
than 160 pg m found with temperatures less than 16.5°C. Of the four
cases, the best correlation (0.710) of the data occurred with the
McHenry-Elkins pair of data, whereas the poorest correlation of 0.526
occurred with the Lewisburg-Beckley data set. The greatest slope of the
-3 -1
regression line also occurs with the McHenry-Elkins data (11.95 yg m K )
—3 —1
and the smallest slope with the Lewisburg-Beckley data (5.02 pg m K ).
The correlation coefficient for the Kane-Youngstown data pair was 0.664
and for the Coshocton-Columbus data pair was 0.707.
In summary, these data confirm Jeffries' experimental findings in
the free air. However, the trend for higher ozone concentrations with
greater maximum temperatures does not establish that the ozone concentra-
tion is a function of the maximum temperature since the maximum temperature
is a function of other meteorological variables, including the sky
conditions, synoptic weather features, and the past history of the air.
8.2 Concurrent Trajectories
Trajectories that arrived at a particular station may pass very
near or over one ozone monitoring station enroute. Such instances
65
-------
400
E
00
g
*j
o
o
X
Ul
o
u
300
200
100
5 10 15 20 25
Temperature at Columbus (°C)
30
35
400
300
200
to
0)
c
o
o 100
A"
A
A * A
10 15 20 25
Temperature at Beckley (°C)
30
35
Figure 8. Scatter diagrams and least square linear fit of daily
maximum hourly ozone and daily maximum temperature at
indicated locations. (A-l data point; B-2 data points,
etc., fall within the 0.5°C temperature and 10 yg/m3
ozone concentration increment).
66
-------
400r
E
60
(J
c
ID
O
o
N
O
300 -
200 -
100 .
10 15 20 25
Temperature at Elkiiis (°C)
30
35
e
00
400
300
200
100
5 10 15 20 25
Temperature at Youngstown (°C)
30
35
Figure 8 (continued).
67
-------
afford an opportunity to examine the changes in ozone concentration as
the parcel moves from one station to another. Because of the uncertainty
in the precise time the parcel passed over a given location, if indeed
it did, a period approximately two hours before to two hours after it
passed one location is compared to the four-hour average concentrations
centered on the time of arrival. Travel times were estimated from the
plotted trajectory. No specific criteria were established to define
"passing near an indicated ozone monitoring station." The percentage change
in concentration with time as a parcel moved along a path is an indication
of net synthesis or destruction of the ozone. The data obtained for this
analysis are presented in Table 7.
A consistent, characteristic change from one station or another or
among the various stations as a function of time or as a function of ozone
concentration was not found. Changes of ozone concentration in a parcel
enroute from one position to another, in given length of time, are situation
dependent. More factors must be considered than just the ozone change,
especially the relevance of the computed change of ozone.
8.3 Case Study
The prolonged stagnation period with persistent high ozone concentra-
tion, August 25 to September 5, was the most significant event of the field
measurement program. Although a thrust of the research was to investigate
for consistent meteorological ozone relationships in all situations, this
event deserves some comment.
According to the data of Figure 3, the surface high pressure center
moved across Kentucky: then between McHenry and Coshocton on the mornine
of August 29th: became stationary near Kane on the 30th: and reached its
maximum sea level pressure 1031 mb on the 31st. By the next day, the
center drifted almost over McHenry, and remained there for five more days,
slowly decreasing in central pressure.
Figure 4 shows warming at each of the four rawinsonde stations. The
warming period begins about the 25th and persists until the cold front of
September 5, 6 and 7 traverses the area (Table 2). The warm core high is
further indicator of stagnant conditions.
The trajectories show that there was little air movement in the 48
hours prior to August 25. For the next three days, the air flow becomes
68
-------
Table la. CHARACTERISTICS OF AIR PARCELS ARRIVING AT KANE AFTER
PASSING NEAR THE INDICATED OZONE MONITORING LOCATIONS.
From Coshocton From Kane
Arrival Travel Time d(lnC)/dt Arrival Travel Time d(lnC)/dt
Time (hr) (hr"1) Time (hr) (hr"1)
7-27-12 6 .0574 8-14-12 18 _ .0179
8-2-00 18 .0380 8-19-12 44 .0062
8-7-12 12 -.0443 9-13-12 12 .0138
8-8-00 12 .1730
8-31-12 6 -.1007
9-1-00 10 .0614
9-1-12 14 -.0382
9-27-12 12 -.0419
From McHenry From Lewisburg
Arrival Travel Time d(lnC)/dt Arrival Travel Time d(lnC)/dt
Time (hr) (hr~l) Time (hr) (hr"1)
8-19-00 36 .0091
9-26-12 16 -.0346 NONE
69
-------
Table 7b. CHARACTERISTICS OF AIR PARCELS ARRIVING AT MCHENRY AFTER
PASSING NEAR THE INDICATED OZONE MONITORING LOCATIONS.
Arrival
Time
8-3-00
8-14-12
8-16-00
8-30-12
9-3-12
9-21-00
9-21-12
9-24-12
From Coshocton
Travel Time
(hr)
14
38
20
36
40
10
20
14
d(lnC)/dt
(hr-1)
.0170
-.0008
.0150
.0136
-.0051
.0558
-.0270
.0309
Arrival
Time
8-21-12
9-13-00
9-13-12
9-15-12
9-30-12
From Kane
Travel Time
(hr)
20
18
30
30
16
d(lnC)/dt
(hr-1)
.0241
.1095
-.0124
.0209
Arrival
Time
8-14-12
9-14-00
9-20-00
9-21-12
From McHenry
travel Time
(hr)
18
38
12
12
d(lnC)/dt
(hr-1)
-.0058
.0258
.0745
-.0589
Arrival
Time
9-1-00
9-27-12
From Lewisburg
Travel Time
(hr)
18
14
d(lnC)/dt
(hr-1)
.1204
.0190
70
-------
Table 7c. CHARACTERISTICS OF AIR PARCELS ARRIVING AT LEWISBURG AFTER
PASSING NEAR THE INDICATED OZONE MONITORING LOCATIONS.
Arrival
Time
8-5-12
8-6-00
8-16-12
8-21-12
8-29-12
9-11-12
From Coshocton
Travel Time
(hr)
38
48
26
18
22
48
d(lnC)/dt
(hr-1)
-.0110
-.0034
.0102
.0809
.0241
.0016
Arrival
Time
7-18-00
8-24-00
8-24-12
From Kane
Travel Time
(hr)
46
36
48
d(lnC)/dt
.0188
.1370
.0205
Arrival
Time
8-17-00
8-17-12
From McHenry
Travel Time
(hr)
20
36
d(lnC)/dt
-1
From Lewisburg
Arrival Travel Time d(lnC)/dt
.0288
-.0016
Time
8-19-00
9-2-12
9-17-12
(hr)
36
12
18
(hr-)
.0287
-.0534
-.0374
71
-------
Table 7d. CHARACTERISTICS OF AIR PARCELS ARRIVING AT COSHOCTON AFTER
PASSING NEAR THE INDICATED OZONE MONITORING LOCATIONS.
From Coshocton From Kane
Arrival Travel Time d(lnC)/dt Arrival Travel Time d(lnC)/dt
Time (hr) (hr~l) Time (hr) (hr"1)
8-18-12 48 .0448 8-17-12 38 -.0172
From McHenry From Lewisburg
Arrival Travel Time d(lnC)/dt Arrival Travel Time d(lnC)/dt
Time (hr) (hr'1) Time (hr) (hr~l)
NONE NONE
72
-------
progressively more rapid, beginning in the southwest and turning to a more
westerly flow by the 28th. The rapid rise in ozone concentration at
Coshocton is first apparent at 0000 GMT on August 27 when the wind flow has
been from the west. The ozone continued to rise while Coshocton was in a
5 to 6 m s~ (10 to 12 kt) westerly flow, until August 28, 1200 GMT when
the wind speed decreases to 3 m s or less. At Coshocton, the ozone
_3
concentrations decreased to near 160 ug m as the high center moved east.
The trajectories clearly show the movement of the high pressure system and
the return to a well defined anticyclonic flow over the entire area.
For the period following August 31, the trajectories show a slow
southerly air flow from similar geographic areas at all locations. On
September 1, the air arriving at McHenry and at Kane has previously passed
over Lewisburg and over Coshocton, respectively. Table 7, a) and b) ,
suggests that the air arriving at 0000 GMT showed an increase of concentration
in transit (during the daytime) and a concentration decrease occurred during
nighttime transit (arrival at 1200 GMT). These instances suggest that the
diurnal trends of ozone production and destruction in the local environment
may be a more predominant factor in the ozone changes that occur during
transport from one location to another.
The cross-sectional analysis shows a decided warming of the high
pressure center aloft and apparent subsidence of the air beginning about
August 25. Although subsidence normally produces an inversion layer aloft,
the subsidence inversion is not a persistent feature of the significant
temperature analyses. The inversion appears at all four rawinsonde locations
in late August, as the high pressure system is established, but it is
conspicuously absent at all of the four locations in September. Twelve-hour
ozone concentrations increased at Coshocton and McHenry during the time of
subsidence, whereas the concentrations at Kane and Lewisburg were not
substantially increased. After the subsidence ceased, the average concen-
_3
trations at Coshocton began to decrease rapidly, remaining above 160 yg m
When the subsidence above 3 km at Pittsburgh stopped, the ozone
concentration at McHenry continued to increase, before slowly decreasing.
Once the subsidence ended at all altitudes and locations (about August 29),
the lower 5 km of the atmosphere were only slightly stable (the isopleths
73
-------
of 6 are widely spaced), a condition conducive to vertical mixing and
dispersion; nevertheless, the ozone concentration remained quite high,
_3
day and night, at McHenry and near 160 yg m at the other locations.
During this episode, daily maximum temperatures at most locations in
the study area were at or near their highest values for the summer. Through
most of the daytime hours, skies were partly cloudy to clear. Usually,
scattered high (cirriform) clouds persisted and scattered, fair weather
cumulus clouds appeared during the day, but seldom did they cover more than
four-tenths of the sky. This type of sky condition permits sustained solar
radiation with few localized interruptions over the day producing high air
temperatures. Also, the abundant sunshine should be conducive to ozone
production.
During this episode of high ozone concentrations and high temperatures
the mean water vapor mixing ratio of the mixed layer, R . , also reached and
i mix
maintained its highest value. It exceeded 15.0 g kg , at Pittsburgh,
Dayton and Buffalo only during this episode. These magnitudes represent a
departure of at least 1.6 standard deviations from the mean value of 9.8 g kg
for those three locations. Moisture content decreased from about 13 to
about 7 g kg in 24 hours as the front passed. Temperatures and ozone con-
centrations also decreased with the frontal passage.
There were 24 consecutive twelve-hour periods of high ozone at McHenry
during late August and early September, a number that is three more than
occurred during the remaining sampling times. The weather conditions,
local and synoptic, that prevailed were unquestionably favorable for high
ozone occurrences at McHenry. Those conditions are summarized as being
characteristic of stagnant, warm core high pressure systems. Since the
system stagnated for eleven days, almost centered at McHenry, the effects
were most apparent there. Other locations were affected, to a lesser
degree, i.e., presumably because the conditions were not optimal for
continued high ozone concentrations.
74
-------
9.0 FINDINGS AND CONCLUSIONS
Investigations of meteorological conditions which accompanied ozone
concentrations during July through October at four nonurban locations
were conducted. Linear relationships of those conditions with the longer
term ozone concentrations were examined. Several avenues of approach
were explored—trajectory analysis, pressure system intensities, cross-
section analyses—which had not been used previously. Some of these
avenues were fruitful and others were fruitless. Analyses of the data
collected were conducted under four principal classifications. The
results of those analyses are followed by the primary conclusions.
9.1 Fronts, Ozone and Potential Temperature Time Sections
0 Large changes of ozone concentrations may be associated with
disturbances in the lower atmosphere which extend above 2 km, but which
are not necessarily associated with fronts;
o Relationships among fronts, ground level ozone concentrations,
and the potential temperature time-sections were inconsistent between ozone
monitoring locations, suggesting that local effects and situations are
equally important to determining concentrations;
o The location of the ozone monitoring station with respect to local
terrain and altitude may have a significant effect upon the daily range of
ozone concentrations;
o Downward transport and diffusion of ozone from the stratosphere
does not appear to account for occurrences of high ozone concentrations.
Therefore - The major anomalies -In the vertical structure of
the lower troposphere are -Important to the occurrences
of high ozone concentrations.
9.2 Air Parcel Trajectories
o High ozone concentrations are usually associated with slower moving
air, whereas the low ozone concentrations are usually associated with faster
air movement;
o No single region of the northeastern United States is implicated,
by trajectory analysis, as the source of the observed high ozone
concentration;
0 High and low ozone concentrations may be associated with trajectories
passing over nonurban areas as well as urban areas;
75
-------
o Low ozone concentrations may be associated with trajectories passing
over urbanized areas.
Therefore - The path which the air takes enroute to the four ozone
monitoring locations influences, but does not determine,
the measured ozone concentration.
9.3 Mixed Layer Properties
° Of the variables tested, the mean water vapor mixing ratio has the
greatest statistical relation to the twelve-hour average ozone concentrations.
o The poo correlation of the percentage rate of change of ozone concen-
tration at McHenry with the water vapor mixing ratio suggests that the
generation of ozone is not dependent upon the amount of water vapor present.
o The positive correlation of ozone concentration and the water vapor
mixing ratio precludes the descent of ozone from the stratosphere as an
explanation of the high ozone concentration at the ground.
o The ozone concentration has little relationship to the currently
observed wind vector.
Therefore - Measurements of the moisture content of the air should
be made with all ozone studies, until or unless experi-
mental evidence directs otherwise.
9.4 Maximum Temperature and Maximum Ozone Concentration
0 Ozone concentrations in excess of NAAQS did not occur on days when
the maximum temperature was less than 16.5°C.
o Ozone concentrations tend to increase as the daily maximum temperature
increases.
76
-------
REFERENCES
1. Environmental Protection Agency. Mount Storm, West Virginia-Gorman
Maryland, and Luke, MaryIand-Keyser, West Virginia, Air Pollution
Abatement Activity. Research Triangle Park, N. C. Publication
No. APTD-0656, April, 1971.
2. Richter, H. G. Special Ozone and Oxidant Measurements in Vicinity of
Mount Storm, West Virginia. Research Triangle Institute, Research
Triangle Park, N.C. 27709, October 1970.
3. Lea, D. A. Vertical Ozone Distribution in the Lower Troposphere Near
an Urban Pollution Center. J. Appl. Meteor., 7^:252-267, 1968.
4. Miller, P. R., M. H. McCutchan and H. P. Milligan. Oxidant Air
Pollution in the Central Valley, Sierra Nevada Foothills and
Mineral King Valley of California. Atmos. Environ., 6^:623-633, 1972.
5. Miller, A. and D. Ahrens. Ozone Within and Below the West Coast
Temperature Inversion. Tellus, 22:329-339, 1970.
6. Vukovich, F. M. Some Observations of the Variations of Ozone Concen-
trations at Night in the North Carolina Piedmont Boundary Layer.
J. Geophys. Res., ^8:4458-4462, 1973.
7. Maga, J., personal communication to L. A. Ripperton.
8. Baljet, P. J. Local Air Pollution Episode, Metropolitan Bade County
Pollution Control, Miami, Florida, June 1962.
9. Research Triangle Institute. Investigation of High Ozone Concentra-
tions in the Vicinity of Garrett County Maryland and Preston
County, West Virginia. Research Triangle Park, N.C. 27709,
Report No. EPA-R4-73-019, January 1973.
10. Research Triangle Institute. Investigation of Ozone and Ozone
Precursor Concentrations at Nonurban Locations in the Eastern
United States. Research Triangle Park, N.C. 27709, Report No.
EPA-450/3-74-034, May 1974.
11. Holzworth, G. Mixing Heights, Wind Speeds and Potential for Urban Air
Pollution Throughout the Contiguous United States. Environmental
Protection Agency, Research Triangle Park, N.C. 27709, Office of
Air Programs Publication No. AP-101, January 1972.
12. Stasiuk, W. N., Jr. and P. E. Coffey. Rural and Urban Oxidant
Relationships in New York State. J. Air Poll. Control Assoc.,
24:564-568, 1974.
77
-------
13. Service, J. A User's Guide to the Statistical Analysis System.
Student Supply Stores, North Carolina State University, Raleigh,
N.C., August 1972.
14. Ripperton, L. A., personal communication.
15. Davis, C. M., Preparation of Maps of Hydrocarbon Emissions from Point
and Area Sources. Research Triangle Institute, Research Triangle
Park, N. C., Final Report, Contract No. 68-02-1096, Task 6, July 1974.
16. Jeffries, H., personal communication to L. A. Ripperton.
78
-------
APPENDIX A
BASIC DISCUSSION OF POTENTIAL TEMPERATURE
ANALYSIS TECHNIQUES
Potential temperature is an ideal quantity to examine the static and
dynamic processes of the lower troposphere. Designated by 6, it equals the
temperature that a parcel of air would have if the parcel were brought to a
pressure of 1000 mb without exchanging heat energy with the environment. The
First Law of Thermodynamics shows that as air ascends or descends without
exchanging heat (an adiabatic process), 6 is unchanged (conserved). Thus, 0
is a measure of the thermodynamic energy potential of dry air. Within an
air parcel, changes of 6 occur through energy input or extraction. Nocturnal
radiational cooling, daytime solar heating, and condensation/evaporation of
moisture are the primary ways of changing 6 in a moving parcel of air. In the
absence of phase changes of water, the primary effect of diabatic heating
(changing 6), is confined to the near-ground layers (below 1 to 2 km) of the
atmosphere where the diurnal radiative heating and cooling cycle occurs.
Vertical Distribution and Stability
Normally, 9 increases with increasing altitude. The vertical gradient
ri ft ?}T
of 9, -—, is related to the vertical temperature gradient, —, by
dz dz
lii = I (.§!+r)
0 3z T ^az '
where
T = 9.8°K knf1
is the adiabatic temperature gradient. When the pressure is approximately
r\ A J\rp J\rp
1000 mb (~100m MSL) , T - 6 and — ^ — + F. Thus, unless — is less than
dz dZ dZ
-T, 9 will increase with altitude.
When an air parcel is lifted adiabatically from one altitude to another,
the air at the higher altitude will usually have a higher 6. The lifted
parcel must acquire heat energy in order to be in equilibrium with the new
environment, or it will return to its original energy level, i.e., the air
parcel is stable (not buoyant) with respect to vertical displacement when
r\ A £\ A ^T1
— > 0. Using the same scenario for the condition, -— = 0 (T— = -F near the
dZ dZ dZ
ground), reveals that the vertical displacement is neutral—neither stable
^ A <\m
or unstable—and, when — < 0 (— < -F near the ground), vertical displace-
dZ dZ
ments are buoyant in an unstable, superadiabatic state.
79
-------
8T
Inversion layers where — > 0 are very stable with respect to vertical
uZ
motion because 8 increases very rapidly with altitude. The inversion
layers may be only a few tens of meters thick, or a few hundred meters
thick. The potential temperature changes across the inversion, not its
thickness, measures the intensity of the inversion. A very strong inver-
sion may occur within a short distance, and a weaker inversion over a greater
depth. The presence of an inversion can best be seen when the analysis
interval is small enough (- 1.0 K or 0.5 K) to show details.
When air aloft descends (subsidence occurs) condensation does not
occur and radiative heat gains or losses are quite small, resulting in
an adiabatic process. A subsiding air parcel that has a 6 value at an
initial time and altitude will be found at a lower altitude, with the same
6 at a later time. During that time period, 9 will have increased at the
initial and the lower altitude. Subsidence stabilizes the air, often
forming inversion layers.
Time-Altitude Sections
In a time-altitude section of potential temperature, the viewer is
thought of as standing in one place, watching the air in a vertical column
of fixed height move past. The distribution of isopleths of 0 with altitude
and time show the changes which take place in the air.
If there were no transitory systems, the air above the heat-cooling
layer would remain stratified with little change in time or altitude in
the vertical distribution of atmospheric variables, including 6. Then
isopleths of 9 in the time-altitude section would be horizontal. When
events occur to change the distribution of 9, the structure of the atmos-
phere is changed because atmospheric energy is redistributed and dynamic
processes are at work changing other properties, perhaps including
atmospheric ozone.
For a given altitude, above surface heating influences, an increase
of 9 with time indicates a warming of the air, resulting from the hori-
zontal advection of warm air from upstream and/or the descent of air from
a higher altitude. Isopleths of 9 slope toward lower altitudes as time
increases. Conversely, a decrease of 9 might be associated with cold
air advection and/or uplifting of air. In either of those cases, isopleths
slope toward higher altitudes (upward) as time increases.
80
-------
Major cooling of the air would probably be associated with cold fronts
passing the sounding station. The fronts are followed by gradual warming.
The greater the contrast of 6 across the front, aloft and at the ground,
the greater the intensity of the front and the greater the contrasts of
the air masses. The stronger fronts exhibit frontal characteristics
(have greater changes of 9 with time) to greater depth of the atmosphere.
So, 9 time sections reveal frontal intensity and depth.
Major warming could arise from passing warm fronts, but they are
weak and rare in the study area during the summer. Large-scale subsidence
is the remaining mechanism for a major warming of the air, i.e., increasing
9 at a given altitude. If the warming occurs over a wide area, it could
only be in conjunction with a high pressure system. The individual time
sections of 9 from rawinsonde locations in the affected area should show
9 isopleths decreasing in altitude as time increased. The slope of the
isopleth is indicative of the magnitude of the subsidence rate.
Other basic discussions of potential temperature and cross section
analyses are found in general Meteorology by H. R. Byers, or Principles of
Meteorological Analysis, by W. J. Saucier.
81
-------
APPENDIX B
AIR PARCEL TRAJECTORIES
AT
900 mb
July 4 to October 31, 1973
for Air Arriving at
McHenry, Maryland
Kane, Pennsylvania
Coshocton, Ohio
Lewisburg, West Virginia
Air parcel positions at twelve-hour intervals prior to arrival are
indicated by a triangle (A) for those air parcels arriving at 0000 GMT and
by a square (a) for those arriving at 1200 GMT. Month and day are indicated
in the lower left corner of each panel.
83
-------
84
-------
85
-------
86
-------
87
-------
88
-------
89
-------
90
-------
91
-------
92
-------
93
-------
-------
95
-------
96
-------
97
-------
98
-------
99
-------
100
-------
101
-------
102
-------
103
-------
hfi
104
-------
105
-------
106
-------
107
-------
108
-------
109
-------
110
-------
Ill
-------
112
-------
113
-------
hrf
114
-------
115
-------
116
-------
117
-------
118
-------
34
119
-------
120
-------
121
-------
122
-------
123
-------
124
-------
125
-------
126
-------
127
-------
128
-------
129
-------
130
-------
131
-------
/IT
132
-------
133
-------
134
-------
135
-------
136
-------
137
-------
138
-------
139
-------
140
-------
141
-------
142
-------
143
-------
144
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-450/3-74-Q34-A
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE ANDSUBTITLE
Investigation of Ozone and Ozone
Precursor Concentrations at Nonurban Locations in the
Eastern United States, Phase II, Meteorological Analysi
5. REPORT DATE
February 1975
i 6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W. D. Bach, Jr.
8. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
Research Triangle Park, N. C.
10. PROGRAM ELEMENT NO.
1 HA326
27709
11. CONTRACT/GRANT NO.
68-02-1077
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Office of Air and Waste Management
Research Triangle Park, N. C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The meteorological conditions occurring during measurements of ozone at four
nonurban locations in Ohio, Pennsylvania, Maryland and West Virginia are examined
for their influence upon the ozone concentrations at the synoptic time and space
scale. Air parcel trajectories at the 900 mb level for the forty-eight hours prior
to arrival at each location were examined for the possible influence of transport
across urban-industrial areas on the measured non-urban ozone. Time-altitude
sections of potential temperature and stability from the ground to 5 km at four
nearby rawinsonde locations showed that major anomalies in the vertical structure
of the lower troposphere are important to the occurrences of high ozone concen-
trations. A regression analysis of twelve-hour average ozone concentrations as a
function of properties of the mixed layer showed that the taean mixing ratio of the
layer in the most significant variable of the eight shown. The mixing depth is
the next most significant.
2
Average hourly ozone concentrations did not exceed 160 yg/m when the maximum
daily temperature was less than 16.5°C.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Non-urban
Ozone
Mixed layer
Trajectories
Temperature
Cross-sections
Potential temperature
Mixing ratio
Water Vapor
13. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (ThisReport)
21. NO. OF PAGES
20. SECURITY CLASS (This page)
22. PRICE
EPA Form 2220-1 (9-73)
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NpTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. &20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverse)
------- |