\~ i '/< 6 r« /' "
December 1982
j .<•
f
." ;Sf.OpIES IN AIR QUALITY
iY: METEOROLOGY AT
STATE UNIVERSITY
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
3- $3 -
-------
STUDIES IN AIR QUALITY
METEOROLOGY AT
NORTH CAROLINA STATE UNIVERSITY
by
Gerald F. Watson
Allen J. Riordan
Walter J. Saucier
and
Ted L. Tsui
Department of Marine, Earth and Atmospheric Sciences
North Carolina State University
Raleigh, North Carolina 27650
805554
Project Officer
Lawrence E. Niemeyer
Meteorology and Assessment Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency,and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
ii
-------
ABSTRACT
This report summarizes eight studies in diverse areas of
air quality meteorology resulting from a cooperative research
program involving the graduate students and faculty of the
atmospheric sciences program of North Carolina State University
and the staff and facilities of the EPA Meteorology and
Assessment Division.
Meteorological analysis of the St. Louis RAPS data has
shown that: (1) The urban heat island strength during the day
amounts to about 0.5°C and varies little with wind speed, cloud
cover, or season. The nocturnal heat island, however, is highly
responsive to all three factors as well as to anthropogenic heat
sources. (2) Profiles of ozone above the urban nocturnal surface
inversion are highly variable and apparently related to details
of the wind structure. (3) Evaluation of the energy budget over
concrete, blacktop, and soil surfaces suggests improvements in
the parameterization of the surface heat flux in air quality
models.
Studies of atmospheric visibility and suspended
particulates reveal that: (1) Air transparency nation-wide has
declined significantly between 1955 and 1972. The annual cycle
in visibility is inversely related to that in relative humidity.
(2) The relative contributions of haze, smoke, and dust to
lowering visibilities in eastern Texas between 1949 and 1968 are
determined. (3) Most sulfate concentration variability at fixed
sites is due to the vagaries in long-range transport rather than
to local conditions of temperature, humidity, or insolation.
Studies of the boundary layer mesoscale wind structure over
the Appalachian Mountains indicate that: (1) A pronounced
low-level jet with significant diurnal variability can form even
during a period of air stagnation. (2) The standard 850 mb-level
winds can be used to estimate surface winds in complex terrain
given the valley-ridge orientation and time of day.
This report was submitted in fullfillraent of Contract No.
R805554-01 by North Carolina State University under the
sponsorship of the U.S. Environmental Protection Agency. This
report covers the period from October 19, 1977 to October 19,
1981 and work was completed as of October 31, 1981.
iii
-------
CONTENTS
ABSTRACT iii
FIGURES vi
TABLES tx
ACKNOWLEDGMENTS xi
1. Introduction 1
2. Conclusions 2
3. Recommendations 4
4. Climatological Aspects of the St. Louis
Urban Heat Island 5
5. A Study of the Vertical Distribution of Ozone
and the Variability of the Wind Field
Above a Nocturnal Radiation Inversion .... 18
6. Temporal Variation in the Surface Energy
Budget Components for Three Land Use
Patterns 31
7. Trends in Atmospheric Visibility Across the
United States from 1955 to 1972 42
8. A Visibility Study in the Eastern Half of Texas
from 1949 to 1968 50
9. Synoptic-Scale Variability in Atmospheric
Suspended Sulfate Concentrations 62
10. Diurnal Variation of Wind Profiles Across
Mountainous Terrain During an Air
Stagnation Period 73
11. A Mesoscale Analysis of Air Flow in Complex
Terrain 86
REFERENCES 100
-------
FIGURES
Number o Page
" 4.1
4.2
4.3
4.4
4.5
4.6
5.1
5.2
5.3
5.4
5.5
5.6
6.1
6.2
6.3
6.4
Heat island strength by cloud cover category. . .
Heat island strength by wind speed category . . .
Heat island strength by cloud cover category
Heat island strength by wind speed category
Rate of change of temperature. .seasonal
Anthropogenic heat output. .annual average ....
Map depicting RAPS facilities in St. Louis,
Height-time section of wind speed for day 216 . .
Height- time section of wind direction for
day 216
200 m layer-average trajectory for day 216. . . .
Ozone and temperature versus altitude, day 216. .
Duirnal variation of net radiation over blacktop,
Same as Fig. 6.1 but for ground heat storage.
Water content^O.SS
Same as Fig. 6.1 but for the sum of the sensible
and latent heat fluxes. Water content=0.35 . . .
Diurnal variation of the ratio G/Fn for blacktop
11
11
12
13
14
15
25
26
27
28
29
30
37
38
39
40
VI
-------
Number Page
6.5
7.1
7.2
7.3
7.4
8.1
8.2
8.3
9.1
9.2
9.3
9.4
9.5
9.6
10.1
1 0.2
Same as Fig. 6.4 but for soil
Trend determined from deviations in visibility
from the respective monthly mean (1955-1972) at
Fast Fourier transform of visibility time series
at Raleigh- Durham, NC
Bimodal annual oscillation of visibility and
relative humidity at National Airport (D.C.). . .
Bimodal annual oscillation of visibility and
relative humidity at Cleveland, OH
Precent frequency of the number of smoke days
from a particular resultant wind direction at
Houston, 1949-1968
Same as Fig. 8.1 except for haze days
Annual visibility frequencies ( %) and linear
trend lines in the four visibility ranges for
Houston, 1949-1968
National Air Surveillance Network stations whose
data are employed in the present research ....
Time series of one year's sulfate data (1972)
for Nashville, TN
Regionally representative seasonal means
( 1969-1974)
Patterns of synoptic weather (12 GMT) and sulfate
concentration
Sulfate concentration versus wind direction at
St. Louis, MO, January- December 1969
Patterns of synoptic weather (12 GMT) and sulfate
concentrations. A reconsideration of Fig. 9.4. .
Geographic variation in annual S02 emission
Area and stations analyzed
41
46
47
48
48
54
55
56
67
68
69
70
71
72
77
78
Vll
-------
Number Page
10.3 Representative sea level pressure analyses during
the air stagnation period 16-22 July 1957 .... 79
10.4 Average wind speed profile for westerly flow
at Washington, D.C 80
10.5 Average wind speed profile for westerly flow at
Akron, Ohio 80
10.6 Time variation of the deviation in westerly wind
from its daily mean at Washington, D.C 81
10.7 Time variation of the deviation in westerly
wind from its daily mean at Akron, Ohio 82
10.8 Cross section of 06 GMT easterly flow
( U component) across terrain 83
10.9 Cross section of 12 GMT westerly flow
(U component) across terrain. . 84
10.10 Hodographs of the wind variation at DCA in the
easterly and westerly flow 85
11.1 Variations in the wind field within a
symmetrical valley due to uneven heating of the
variably sloped terrain (Defant, 1951) 93
11.2 Night and day wind roses at a valley (Nash)
site 94
11.3 Terrain surrounding the Nash site 94
11.4 Night and day wind roses at a mountain (Hockey)
site 95
11.5 Terrain surrounding at the Hockey site 95
Vlll
-------
TABLES
Number Page
4.1 Heat island strength annual and seasonal
averages 16
4.2 Maximum contribution from anthropogenic heat
emissions. Predictions from Summers' (1965)
model 16
4.3 Duirnal temperature amplitudes (July 8-11,
1976) 17
4.4 Wind speed — annual average 17
5.1 Comparison of the character of the morning profile
and the estimated time of trajectory passage
over southern Lake Michigan 30
7.1 Stations used in this study and grouped according
to geographical region 49
8.1 Station information and data years used 57
8.2 Number of days with obstructions to vision and
mean daily visibility 57
8.3 Results of the correlation between annual mean
visibilities and the number of days with
obstructions, 1949-1968 58
8.4 The total number of days with smoke and the mean
population for all locations 58
8.5 Number of days with haze and average relative
humidity for each station 59
8.6 Correlation coefficients between population
growth and (a) the frequency with mean visibility
less than 7 miles and (b) visibilities in the
13 to 15 mile range 59
IX
-------
Number Page
8.7 Distance and direction from the Gulf of Mexico
and mean visibility for all locations ...... 60
8.8 Visibility trend analysis for season and locations
with a downward trend (1949-1968) 61
10.1 Station identifiers and elevations 78
11.1 Data collected at each of the eight fixed
stations 96
11.2 The frequency of surface winds coincident with
the upper level winds 97
11.3 Day to night variations in poor (<25%) and good
( >60%) agreement of surface and upper level
directions during the SW parallel and NW
perpendicular analysis periods 98
11.4 Percentage of time Tower and Hockey wind
directions were within the indicated degrees of
the 850 mb wind during the 00 and 12 GMT
radiosonde releases 99
-------
ACKNOWLEDGMENTS
The bulk of the work contained in this report is primarily
the achievement of graduate students in the atmospheric sciences
program of North Carolina State University, through the
completion of their Master's theses. These students, whose
names appear as principal authors of their respective
contributions to this report, have adequately vindicated the
original concept of "Graduate Research in Air Quality
Meteorology."
The support of personnel of the EPA Meteorology and
Assessment Division, through guidance of several of the graduate
students in their research efforts, has been invaluable. Mr.
George C. Holzworth and Dr. Jason K.S. Ching are deserving of
special recognition in this regard.
Appreciation is also expressed to Dr. Walter D. Bach of the
Research Triangle .Institute for his contribution.
XI
-------
-------
SECTION 1
INTRODUCTION
The proximity of the atmospheric sciences program at North
Carolina State University (NCSU) to the Meteorology and
Assessment Division and other EPA facilities has engendered over
the past several years professional and scientific cooperation
among personnel of the two organizations. NCSU graduate
students have benefitted through work at the EPA laboratory and
helpful guidance in thesis research by EPA staff. It was
proposed that this on-going cooperative research effort be
enhanced through a formal program which focused the joint effort
of NCSU students and faculty, and EPA staff, toward solving
specific problems of the EPA in air quality meteorology.
This report contains results of this cooperative effort in
the form of summaries of eight research theses on diverse
topics. The first three contributions (SECTION'S 4-6) are
concerned broadly with the meteorological analysis of the St.
Louis Regional Air Pollution Study (RAPS) data. The topics
include the climatology of the St. Louis heat island, the
problem of ozone variability in the urban boundary layer, and
the analysis of the urban surface-energy budget in an effort to
improve parameterization of the surface heat fluxes in air
quality models.
SECTION'S 7-9 examine the impact of suspended particulates
as manifest in the generally deteriorating transparency of the
air, both regionally and nationally. The last paper in this
series is concerned with the relationship of suspended sulfates
to the large-scale air circulation and to other meteorological
variables.
The final two reports (SECTION'S 10 and 11) tackle the
difficult problem of wind structure over the Appalachian
Mountains. The implications of this wind structure to pollutant
transport and dispersal in complex terrain is also considered.
The format of each of the following eight research
contributions consists of an 'Abstract', an 'Introduction1 which
presents the scope of the problem and relevant literature, and a
summary of major 'Results and Conclusions'. Tables and figures
follow the text of each paper, and 'References' for all papers
are given in a separate section. The abbreviated nature of
these papers has necessitated the omission of detail regarding
data sources, analysis methods, and explanation of results. The
full research theses are, however, available to the EPA.
-------
SECTION 2
CONCLUSIONS
Conclusions based on the eight research theses summarized
in this report and worthy of special note are here listed by
major topic.
Meteorological analysis of the St. Louis RAPS data:
( 1) The urban heat island strength during the day amounts
to about 0.5°C and varies little with wind speed, cloud
cover, or season. The nocturnal heat island, however,
is highly responsive to all three factors, as well as
to anthropogenic heat sources.
( 2) Variability in ozone profiles above the urban nocturnal
surface inversion is significant and appears to be
related to a similar variability in wind profiles. The
relevant wind variability occurs on space and time
scales unresolvable by the conventional radiosonde
network.
(3) Evaluation of the ratio of subsurface heat flux to net
radiation over concrete, blacktop, and soil surfaces
has shown that this ratio can be better parameterized
in air quality models by incorporating its dependence
on surface type and on day versus night conditions.
Atmospheric visibility and suspended particulates:
( 4) Visibility at 14 sites throughout the United States has
declined significantly between 1955 and 1972. The
annual cycle in atmospheric transparency is inversely
related to that in relative humidity.
( 5) The relative contribution of haze, smoke, and dust to
decreasing visibilities at several sites in eastern
Texas during the period 1949 to 1968 are identified.
The haze contribution is part of a regional-scale
problem, while the smoke contribution is associated
with local sources and the wind direction.
-------
(6) M9SC sulfate concentration variability at particular
sites appears explained by long-range transport as
dependent on the wind pattern, rather than by local
conditions of temperature, humidity, or insolation.
Boundary layer wind structure in complex terrain:
( 7) The unexpected occurrence of a nocturnal low-level jet
over the north-central Appalachian Mountains during a
period of air stagnation has significant implications
for regional pollutant transport.
(8) Winds at the 850 mb level can be useful in describing
near-surface winds in mountainous terrain depending on
valley-ridge orientation and time of day. This
relationship provides a simple parameterization of
winds in complex terrain for anticipating pollutant
dispersal.
-------
SECTION 3
RECOMMENDATIONS
Results of the work described in this report should, in
many cases, be considered preliminary. Specific recommendations
for further research work by the EPA include:
( 1) Use of the full St. Louis RAPS data-set (1974-77) to
strengthen the conclusions of the study of the urban
heat island based on 1976 data (SECTION 4). In
particular, the role of rural soil moisture and of
urban anthropogenic heat sources to the diurnal and
seasonal heat island strength need further
investigation.
( 2) Use of the full RAPS data-set to extend the two-month
study period of energy budget evaluations for various
types of surfaces (SECTION 6). The ultimate goal of
formulating an improved parameterization of the surface
heat flux in air quality models which properly accounts
for seasonal variability and the urban environment
seems achievable with an extension of this study.
( 3) Further regional assessments of trends in atmospheric
visibility, like that in eastern Texas (SECTION 8), is
one means of detecting the environmental impact of mass
migrations of the U.S. populous as, for example, the
apparent shift in population from the northeast to the
"Sun Belt" states.
( 4) Further study of suspended sulfates and their
relationship to source areas and meteorological
variables (SECTION 9). However, major revelations
would only seem possible with increased sampling
resolution over that provided by the present National
Air Surveillance Network, as well as improved means of
particulate collection and chemical analysis.
( 5) Additional studies of the mesoscale boundary layer wind
structure in complex terrain (SECTION'S 10 and 11) with
emphasis on consequent pollutant transport and
dispersal over such regions.
-------
SECTION 4
CLIMATOLOGICAL ASPECTS OF .THE
ST. LOUIS URBAN HEAT ISLAND
Joseph E. Steigerwald, G.F. Watson, and J.K.S. Ching
Abstract
The St. Louis urban heat island is studied using one year
of temperature data (1976). Fluctuation in the strength of the
urban heat island due to changes in cloud cover, wind speed, and
season are explored.
The heat island strength varies diurnally and seasonally,
and is largest on fall evenings and weakest on summer mornings.
It is also very strong under low wind speed and clear sky
conditions, with the wind speed exerting more control on the
nocturnal heat island strength than cloud cover. The daytime
heat island, the nocturnal heat island, and the transition
periods are examined in detail. The nocturnal heat island and
the transition periods vary greatly with changes in cloud cover,
wind speed, and season; the daytime heat island varies little
with changes in these factors.
Summers' (1965) heat island model was used to evaluate the
contribution of the anthropogenic heat output to the urban heat
island. The results were compared with results derived from an
analysis of the St. Louis Heat Emission Inventory. The emission
data comparison was inconclusive.
-------
INTRODUCTION
The term 'heat island' is used to describe the ground level
air temperature difference between urban and rural areas. The
urban area is normally warmer than the surrounding rural area.
A heat island forms because cities are composed of materials
whose thermal characteristics differ from the natural materials
that they replace. This factor combined with anthropogenic heat
emissions and the effect on wind speed and albedo caused by the
city's physical presence produce a change in the microclimate
encompassing the city. This microclimate is warmer and drier
than that in the country-side, except in winter when it is
warmer but more moist (Hage, 1977).
Many studies of the urban heat island have been made over
the years. Howard (1883) conducted one of the first of many
studies of the London, England heat island. Subsequently the
heat islands of Paris, France (Renou, 1862); Dallas/Fort Worth,
Texas (Ludwig, 1968); Cincinnati, Ohio (Clarke, 1969); Akron,
Ohio (Martin and Powell, 1977); and Montreal, Canada (Oke,
1977), just to mention a few, have been described.
The first heat island studies were usually comparisons of
two temperatures, one urban and one rural. Spacial features
were pointed out by Schmidt who used an automobile to traverse
the urban area to measure temperatures (Duckworth and Sandburg,
1954). Since then automobiles have been a major platform used
in many heat island studies to get temperature readings at a
number of points along a set route.
The use of a network of instruments at fixed sites is
fairly common in recent field investigations of heat islands.
The heat island studies of Austin, Texas (Wood, 1971) and St.
Louis, Missouri (Shreffler, 1979) both used such a network.
The altered climate found in the cities needs to be studied
because it affects the air flow in and around the city. Cities
emit undesirable pollutants that affect those who live in and
downwind of their source, and it is necessary to understand the
transport capacity of the urban induced circulation. This
circulation is, to a large degree, controlled by the magnitude
of the urban heat island.
The present study utilizes data from the St. Louis Regional
Air Pollution Study (RAPS) as a basis for better documenting and
-------
understanding the urban heat island and how it varies with"
season and various meteorological parameters. As the urban heat
island is better understood, the urban centers may be planned to
enhance or dilute the heat island effect depending on what is
most beneficial to the area. Once a reliable climatology for
the urban area has been established, it can be incorporated into
pollution distribution models to ensure that they model the
pollutants as realistically as possible. The present study
hopes to make a contribution to the climatology of the urban
heat island.
RESULTS AND CONCLUSIONS
The daytime urban-rural temperature difference, or heat
island strength (H.I.S.), varies little from day to day;
different winds speeds and cloud cover conditions have little
effect on it. The annual average daytime H.I.S. is 0.6°C,
varying only from 0.5°C in fall to 0.7°C in winter.
The daytime heat island is seasonally consistent because a
very deep mixed layer is present through the day. Any changes
or inputs that might affect the daytime H.I.S. are mixed
throughout this layer diluting their effect. Thermal mixing
accounts for this very deep mixed layer, often called the
'planetary boundary layer', which extends up to a kilometer or
more in depth.
The nocturnal heat island is more complex and not as steady
as the daytime heat island. The H.I.S. varies significantly
with wind speed, cloud cover, and season. It is strongest in
the fall (2.9°C) and weakest in the w.inter (1.6°C).
Wind speed exerts more influence on the urban heat island
than does the amount of cloud cover. Figures 4.1 and 4.2 show
the H.I.S. versus cloud cover and the wind speed, respectively.
In general, the clearer the skies or the lower the winds, the
more pronounced is the nocturnal heat island. These
meteorological conditions favor a greater and more rapid cooling
of the rural surface layer thus producing a larger urban-rural
temperature difference than under other conditions. Table 4.1
lists the average nocturnal H.I.S. for all three cloud cover and
wind speed groups. If cloud cover is constant, the H.I.S. drops
steadily as wind speed increases. If wind speed is constant,
the H.I.S. usually drops as cloud cover increases, but not in
all cases, and certainly not at a steady rate.
The nocturnal heat island strength exhibits seasonal trends
and these trends are modified by weather conditions. The
winter, spring, and summer nocturnal heat islands attain their
maximum values just after sunset, then the nocturnal H.I.S.
-------
declines until morning. Fall is different, though, in that the
maximum strength occurs just before sunrise. These trends are
most obvious under partly-cloudly skies and/or moderate winds as
suggested by Figures 4.3 and 4.4. High winds and overcast
conditions muddle these trends, and low winds and clear skies
yield a constant nocturnal H.I.S.
Graphs of the diurnal heating and/or cooling rates for the
urban and rural area are a useful tool. When compared, the
differences in these two curves show how the heat island grows,
matures, and decays. The seasonal curves of the rate of
temperature change (Figure 4.5) for the urban area are
remarkably consistent throughout the year. All of the urban
curves, except fall, have the same maximum and minimum rate of
temperature change (about 1.5°C/hour and 1.0°C/hour,
respectively). The rural rate of temperature change varies from
season to season. It is the way in which temperature in the
rural area changes, that governs the strength and timing of the
urban heat island (Oke, 1977). The rural rate is in turn
controlled by such factors as city size, degree of urban
industrialization, amount of vegetation, and geography of the
area (Oke and Maxwell, 1975).
The exact contribution to the urban heat island from
anthropogenic heat sources is difficult to assess. A
comprehensive listing of all of the contributors would be much
too complicated and costly. Even though this is true, heat
emission inventories are the only way to accurately estimate the
contribution of anthropogenic heat sources in the H.I.S.
One such heat emission inventory was conducted in St. Louis
as part of the RAPS/RAMS project. This study, conducted using
1976 data, took into account all point and area sources,
stationary and mobile. Specific source data were used for all
point sources; annual records were used for the area sources.
The annual curve for both rural and urban emissions are
shown in Figure 4.6. The data are confusing and difficult to
explain. The midnight maximum in emissions and the 0800 GST
minimum are contrary to the activity pattern of industry or the
population. In addition, none of the seasonal curves have any
features that can be identified as a traffic peak (a peak in the
heat emission curve caused by rush-hour traffic), even though 30
percent of the total' sources are mobile. A fall maximum at 1600
GST, a spring minimum at 1800 GST, and an unexpected minimum at
0500 GST in the summer also present problems in data
interpretation.
Details of the heat emission inventory data were reviewed.
The diurnal variations of anthropogenic heat output, which are
on the order of 5 percent, cannot be responsible for the diurnal
fluctuations on the order of 2°C in the annual average H.I.S.
-------
Clearly, the diurnal fluctuations in the H.I.S. curves are due
to some other factor or combination of factors.
The influence of anthropogenic heat output on the urban
heat island is overshadowed by the effect of solar radiation
during the day. However, anthropogenic heat output at night
becomes an important factor in the formation and strengthing of
the urban heat island (Bowling and Benson, 1978).
While individual diurnal fluctuations are negligible, the
anthropogenic heat output is the correct order of magnitude and
exhibits the expected seasonal variation. When substituted into
Summers' model (Summers, 1965)in order to predict the
contribution of anthropogenic heat sources to the H.I.S., the
heat emission data suggest a contribution of about 1.5°G to the
urban heat island (under a wind speed of 2m/s). The
contribution changes with the season becoming less than 1,4°C in
the summer. In percentages, these figures represent 110 percent
and 56 percent contributions (winter and summer, respectively)
of anthropogenic heat sources relative to the solar contribution
(see Table 4.2). Obivously, the anthropogenic heat output does
not contribute 110 percent to the winter urban heat island.
This result is a statistical anomaly caused by comparing the
maximum contribution made by anthropogenic heat as a percentage
of the average winter H.I.S.
The amount of moisture trapped within the soil exerts some
control on the urban heat island, but unfortunately the soil's
moisture content was not one of the parameters recorded during
the study. A qualitative assessment of the importance of this
variable can still be made using temperature measurements taken
during a period of drying. This can be accomplished by
recording and then analyzing the diurnal temperature cycle for
the urban and rural areas. Decreased moisture within the urban
area should have little effect on the diurnal temperature cycle,
since up to 50 percent of the urban fabric is composed of
materials that are impervious to moisture and thus greatly
reduce evapotranspiration (Landsberg, 1968).
Table 4.3 shows the amplitude of the urban and rural
diurnal temperature traces for July 8 and 11, 1976 as well as
the H.I.S. This four day period was relatively dry with no
precipitation and only scattered clouds. The data show no trend
towards an increasing diurnal amplitude in temperatures for the
rural area. In addition, the H.I.S. does not increase
significantly. These results contradict those of Lyall (1977)
for London and environs.
Table 4.4 compares seasonally and annually averaged rural
and urban wind speeds. The increased roughness length found in
the urban area slows the wind speeds by as much as 18 percent.
This maximum reduction occurs in the spring; the annual
-------
reduction averages to 11 percent. These results are in agreement
with those presented by Lowry (1972), who states that a 20 to 30
percent drop in the wind speed is observed in the urban area. At
these lower wind speeds, the air has a greater residence time in
the urban canyon enabling it to absorb more of the heat stored
by the buildings and other urban surfaces (called the 'canyon
effect'). This increases the H.I.S.
It appears possible to draw five basic conclusions from
this analysis of the climatology of the St. Louis urban heat
island.
1) The daytime H.I.S. remains constant (at about 0.6°C);
changes in season, wind speed, and cloud cover have little
effect on it.
2) The nocturnal heat island strength varies significantly with
season, amount of cloud cover, and changes in wind speed.
Of the latter two, increased wind speed has a greater effect
on the nocturnal heat island strength, causing a rapid
reduction.
3) The 1976 RAPS Heat Emission Inventory needs further
scrutinizing. The data, while not in a very useful form,
are of the correct order of magnitude and give some
indication of the importance of anthropogenic heat emissions
to the formation of the urban heat island. While diurnal
fluctuations are negligible, and the contribution during the
day seems insignificant, anthropogenic heat sources do
contribute to the nocturnal H.I.S. The exact proportion of
their contribution in St. Louis needs further study.
4) The effect of moisture on the urban heat island is unclear.
The results presented here disagree with the results
presented elsewhere. Enhanced cooling was not observed in
the rural areas surrounding the St. Louis metropolitan
area. However, the data set used was limited and the
results inconclusive.
5) The RAPS/RAMS data need and deserve more study. The
experiment continued from 1974 to 1977, amassing a huge
quantity of data, meteorological as well as ambient air
concentrations of certain pollutants. Very little of its
full potential has been realized. Analyses similar to those
attempted in this study should be undertaken with the full
data set. This should further refine the conclusions of
this study and make them of real value in understanding and
modeling the urban heat island.
10
-------
4.0
£ 3.0
(9
U*
OK
S 2.0
o
I
2 1.0
0.0
elMr
partly-cloudy
OVMClSt
M
12
it
i
24
Figure 4.1 Heat island strength by cloud cover category,
4.0 r
2.0
1.0
0.0
low winds
modtntt winds
hi|ti winds
I
M
12
TIME
II
1
24
Figure 4.2 Heat island strength by wind speed category,
II
-------
e!«r
^-pirtly-cloudy
ovtrust
eiur
> partly-cloudy
ov«rc«t
elMr
* p«rtly-«loudy
ovircist
elMf
""*"•"* partly-cloudy
ov»re«t
Figure
Heat island strength by cloud cover category and
season.
12
-------
mod«nt»
low
!-*••*_».•* modtntt
24
Figure 4.4 Heat island strength by wind speed category and
season.
13
-------
1.0 r
TtiC
Figure 4.5:
Rate of change of temperature • • seasonal averages
Solid line •• rural temperatures; dashed line ••
urban temperatures.
14
-------
0.57$
TIME
- Q.JII
- 8.914
- 0.3M
- O.J7I 2
- O.J72
- 0.941
- 0.9«4
FIGURE 4.6: Anthropogenic heac output •• annual average,
15
-------
Table 4.1 Heat Island Strength
Averages.
Annual and Seasonal
Annual
Winter
Spring
Summer
Fall
Day
0.60°C
0.75°C
0.70°C
0.45°C
0.45°C
Night
2.60°C
1.60°C
2.20°C
2.50°C
2.9QOC
Table 4.2 Maximum Contribution from Anthropogenic Heat
Emissions. Predictions From Summers' (1965)
Model.
Season
Max imum
contribution
Percentage of
average
Winter
Spr i ng
Summer
Fall
1.77°C
1.540C
1.390C
1.55°C
110%
70%
56%
53%
16
-------
Table 4.3 Diurnal Temperature Amplitudes. (July 8-11,
1976). (M= morning; E- evening).
Rural
M
July
July
July
8
9
10
11.
11.
13.
2
2
5
E
10.
10.
13.
5
6
0
Urban
M
10.0
8.2
12.5
E
7
9
11
.8
.2
.0
M
1
3
1
H.I.
.2
.0
.0
,s.
E
2.7
1.4
2.0
Average 12.0 11.4 10.2 9.3 1.7 2.0
Table 4.4 Wind Speed Annual Average,
Annual
Winter
Spring
Summer
Fall
Rural (m/s)
4.25
5.29
5.03
3.34
3.41
Urban (m/s)
3.77
4.47
4.13
3.22
3.33
Percent
reduction
11.0
15.5
17.9
3.6
2.4
17
-------
SECTION 5
A STUDY OF THE VERTICAL DISTRIBUTION OF OZONE
AND THE VARIABILITY OF THE WIND FIELD ABOVE
A NOCTURNAL RADIATION INVERSION
Donald R. Hood and A. J. Riordan
Abstract
The vertical distribution of ozone prior to the morning
breakdown of the radiation inversion is related to the changes
in the nocturnal wind field and the trajectory of the layer
containing the ozone. Helicopter and hourly pilot-balloon data
collected on five days during August 1976, in St. Louis,
Missouri, were analyzed.
The ozone was assumed to be uniformly mixed throughout the
boundary layer at sunset the previous evening. The study
indicates that the similarities and differences between the
vertical distribution of ozone assumed at sunset and observed
the following morning are related to the variability of the
nocturnal wind field.
Analysis of height-time cross sections of the nocturnal
wind field indicated that the winds exhibited both vertical and
temporal variability above an observation point. Most of the
temporal variability in the nocturnal wind field occurs between
scheduled radiosonde launch times and will not be completely
detected. Therefore, the trajectory computations will not
entirely represent the changes which occur. The study also
indicates that due to the vertical variability of the wind
field, more representative trajectories can be obtained by
computing the trajectories for several thin layers versus
computing a single trajectory for a thick layer.
18
-------
INTRODUCTION
The formation, destruction, and transport of ozone can be
investigated through examination of air pollution and
meteorological parameters measured in St. Louis, Missouri, as
part of the Regional Air Pollution Study (RAPS). RAPS was a
multi-year research program conducted by the Environmental
Protection Agency (EPA) in St. Louis, in which air quality and
meteorological parameters were measured both at the surface and
in the lower troposphere from 1974 through 1976. As part of the
RAPS, a network of air monitoring stations, called the Regional
Air Monitoring System (RAMS), was established to provide a large
body of ground level air monitoring data (Figure 5.1). Figure
5.1 also locates four upper air network sites (UAN) numbered
141, 142, 143, and 144, whose purpose was to provide upper air
wind data using pilot-balloons (PIBALs) and radiosondes. In
addition, during data-intensive periods, helicopter flights
originating at Smartt Field were used to obtain the vertical
distribution of ozone, total oxides of nitrogen, carbon
monoxide, sulfur dioxide, particulate ligh.t scattering,
temperature and dewpoint as a function of time and altitude.
Also during data intensive periods PIBALs were launched each
hour from all four UAN sites, with the exception of 0400, 1000,
1600, and 2200 GST, at which times radiosondes were launched
from all four UAN sites.
The case studies to follow will use data collected during
the summer 1976 data intensive period. The helicopter spirals
used in the case studies were observed over Smartt Field which
is located in a rural setting approximately 35 km north-north-
west of the center of St. Louis. The case studies will also use
PIBAL observations taken at UAN site 144 which is located
approximately 25 km to the east-southeast of Smartt Field.
The purpose of this study is to describe the early morning
vertical distribution of ozone (the ozone profile) above Smartt
Field and to relate the shape of the ozone profile to the
changes in the wind field that have occurred since sunset on the
previous evening and to the trajectory of the air containing the
ozone prior to its arrival over Smartt Field.
19
-------
RESULTS AND CONCLUSIONS
Five case studies detailing the changes in the wind field
above a nocturnal radiation inversion and the effect of these
changes on the shape of an early morning ozone profile are
presented.
The five days were chosen based on the following criteria:
( 1) Either a. surface-based radiation inversion or a low-
level elevated inversion were present at the time of the
helicopter spiral so that the profile was protected from
scavenging from below.
(2) No frontal system was present in the St. Louis area
from sunset the previous evening to the time of the helicopter
spiral.
(3) The helicopter spiral was high enough to detect the
subsidence inversion above Smartt Airfield. Thus, the changes
in the wind field between the top of the radiation inversion and
the base of the subsidence inversion can be related to the shape
of the ozone profile.
(4) The helicopter spiral occurred within 3 hours after
sunrise so that changes in the shape of the profile due to
photochemical activity will be at a minimum.
Using the above criteria the 2nd, 3rd, 4th, 7th, and 8th
of August 1976 were chosen from a total of eleven days with
morning helicopter spirals for which edited helicopter data were
available. Throughout the case studies these five days will be
referred to as day 215, 216, 217, 220, and 221, respectively.
The height-time sections for each of the five case-study
days have shown that the nocturnal wind field between the top of
the radiation inversion and the base of the subsidence inversion
is highly variable. The wind field not only exhibits
variability with time but also exhibits a large amount of
vertical variability above an observation site at a given time.
Figures 5.2 and 5.3 for day 216 serve as examples. In addition,
the height-time sections have shown that most of the variability
in wind speed and direction occurs between the scheduled launch
times (1800 CST and 0600 CST) of radiosondes routinely used to
obtain wind observations.
The vertical variability of wind direction is most evident
on days 216, 217, and 221. On days 215 and 220 the winds were
fairly uniform at the time of the helicopter spiral, but the
winds had exhibited changes in direction earlier in the night.
20
-------
The structure of the wind field seems closely related to
the shape of the ozone profile. On day 216, for example, a
well-developed jet persisted at 200-300 m above the ground for
the 8 hours previous to the ozone measurement. A sharp spike in
ozone concentration was then found just above the level of the
jet maximum. Concentrations decreased rapidly from 400 to 800 m
in a zone of over 90 degrees of wind shear which descended from
much higher elevations during the night as the low—level.jet
became concentrated in the lowest layers. Thus, the morning
ozone profile has probably been shaped by important structures
and variations in the nocturnal wind field. More ozone
measurements would certainly have aided in quantifying the
relationship.
In all five cases examined the trajectory computed for the
original layer (maximum top 2000 m) followed almost the same
path as the trajectory for the 900-1100 m AGL layer. Compare,
for example, Figures 5.4 and 5.5 for day 216. This agreement
might be expected since the 900-1100 m AGL layer is
approximately in the middle of the original layer (maximum top
2000 m AGL) depending on the top and base of the layer chosen by
the trajectory model. The maximum height of the layer base was
5000 m AGL.
The 200 m-layer trajectories outline the region which could
be considered the source region for the pollutants arriving over
Smartt Field. The trajectories for days 215 and 216 indicated
the least variance in the size of the source region for the
entire 60-hour trajectory while day 220 showed the greatest
variance between layers, and thus possibly the largest source
region.
The Heffter-Taylor Trajectory Model is intended for use as
a regional scale trajectory model (Heffter _e_t al. , 1975).
Therefore, it may not be entirely correct to compare the results
of the model to the 12-hour hand computed trajectory, since the
hand computed trajectory uses only the winds obtained over one
site. However, the comparison is important since it gives an
idea of the error that may be present in the first 12-hour
segment due to the variability of the nocturnal wind field. The
end point of the first segment is the starting point of the
second segment; thus, the errors would be cumulative with
time.
The morning ozone profiles for days 216 (Figure 5.6) and
221 have almost the same shape between the top of the radiation
inversion and the base of the subsidence inversion. The ozone
profiles for both days exhibit a sharp "spike" of ozone above
the top of the radiation inversion with uniform ozone
concentrations between the point where the ozone concentrations
decrease to a minimum (possibly natural background levels) and
the base of the subsidence inversion. On day 221 layers of
21
-------
ozone are present above the top of the subsidence inversion as
opposed to uniform ozone concentration above the subsidence
inversion on day 216.
The time of day when the trajectory passes over an urban-
industrial area is a primary determinant of the pollutant burden
that will be carried away from the area (Ludwig, 1979b; Hester
e_t _al. , 1977; White _e_t al. , 1977). If the trajectory passes
over an urban area at night or in the early morning hours (prior
to rush hour traffic), the pollutant burden of the trajectory
will not be increased as much as it will be if passage occurs
between mid-morning and sunset. This is especially true if a
radiation inversion is present in the morning when passage
occurs. The term "pollutant burden" refers not only to ozone
but also to the precursors from which ozone will form downwind
of the urban-industrial area.
Table 5.1 indicates that the area around southern Lake
Michigan is most likely to make a significant contribution to
the pollutant burden on days 215 and 220 and least likely to
make a significant contribution on days 216 and 221, while the
contribution to the pollutant burden on day 217 is uncertain.
The following conclusions can be drawn from analysis of
the trajectories and the height-time sections:
( 1) The height-time sections of wind speed and direction
have shown that the nocturnal wind fields are highly variable
with time and that significant changes can occur between
scheduled launch times of radiosondes routinely used to obtain
wind observations.
( 2) Since the radiosonde observations do not detect changes
in the nocturnal wind field which occur between scheduled launch
times, these changes cannot be reflected in the trajectory
computations.
( 3) The height-time sections of wind speed and direction
also indicate that the wind field exhibits a large amount of
vertical variability above an observation point at a given time.
The vertical variability of the wind field in a layer can be
represented, at least in part, by choosing smaller layers within
the original layer and computing trajectories for the smaller
layers. The need for better time-resolution mentioned in (2) is
vital here.
(4) Even with the possibility that vertical motion might be
taking place, the choice of several small layers as opposed to
one overall layer appears to give a better indication of the
source region from which the trajectory may have obtained its
pollutant burden.
22
-------
The shape of the morning ozone profile observed in the case
studies is highly variable and is sometimes radically different
from the morning ozone profile predicted by Ludwig (1979a). If
the ozone profile at sunset on the previous evening is assumed
to be a Type D profile, which is described by Ludwig as a
profile in which the precursors and the ozone that has formed
from them are mixed throughout the boundary layer, then the case
studies indicate that the shape of the ozone profile on the
following morning could be due to one of or a combination of the
following:
(1) The variability of the nocturnal wind field, including
the variability of the wind speed and direction, and including
both the variability with time and vertical variability above an
observation point.
( 2) The transport of air parcels containing different
pollutant burdens in different layers, with little mixing
occurring between the layers, over Smartt Field at the time of
the helicopter spiral.
( 3) The transport of air parcels containing different
pollutant burdens by the nocturnal jet.
(4) The proximity of the source region to major urban-
industrial areas and the time of day when the trajectory passes
the source region.
( 5) The formation of, and the change in height of,
nocturnal radiation inversions and the more persistent
subsidence inversions. These changes could also affect the
nocturnal wind field.
(6) The natural background levels present in the trajec-
tory air mass. This would determine the minimum levels of ozone
that could be observed if the air came from "clean" remote
areas.
The five case studies indicate that the most plausible
explanation for the shape of the ozone profile observed over
Smartt Field at the time of the helicopter spiral, is the
transport of air containing different pollutant burdens in
different layers. However, a major problem with the case studies
has been a lack of knowledge of the shape of the ozone profile
at sunset on the evening prior to the observation of the ozone
profile and the changes that occurred in the profile moving with
the trajectory prior to the arrival of the trajectory over
Smartt Field. Due to the variability of the nocturnal wind
field, it does not seem that this problem can be solved easily.
Therefore, the best approach to follow would be to observe
hourly profiles of ozone, NO, NOX, and temperature in
conjunction with hourly wind observations over a chosen point to
23
-------
determine the changes which are taking place in each of the
quantities and the possible relationship between the changes,
24
-------
3
O
CO
CO
2 z
< <
CO
O
CO
0)
O
CO
00
c
-U
O
CL
CU
•o
m
0)
oo
25
-------
1900*
two
100*
300
••*
'•'
\\
\\
\\
s V
* \
l
I I
\ l
> I
I
\ % •
X> '* /
V \ '
\\w
V %
»•» » * N
'» \
• V «v
.-3
..-z
* • ;
' ' S
'• H M 21 22 23 I* 01 02
ns CST
94 OS
Figure 5.2 Height-time section of wind speed (m s~^) for
day 216. Shaded area represents an inversion.
A "+" indicates a wind speed maximum. A"-"
indicates a wind speed minimum. Q indicates
ozone concentration in ppb.
26
-------
1900
I SOOf
sod
10
3W 340
\ \
330
20--
30—
« ~*
-v/\\%\ \ \ •>
• ^\\ X \ • ""
\ \V<\\ \
> \ \
' \ \
\
. \ \N ^
N » \\X
34O
330
\ V '
\ \ '1*^1
.J30
»
40s
1 VN\ '\ \ >J<'' .•''..-'-
-:---si>-\^-'->'-;,/-
---.v-x-:--\ \ / /
>• -.:v:\\ \\ -..- ^
t| It 20 21 22 23 M 01 93 03 04 OS 0« 07 0«
CST
219 218
Figure 5.3 Same as Figure 5.2 except for wind direction.
27
-------
Figure 5.4 Layer-average trajectory for day 216.
Maximum top of layer is 2000 m; base
of the layer is chosen by the trajectory
model. "0* indicates a radiosonde
station within 300 nm (approximately 550 km)
of trajectory. Each segment represents 12
hours.
28
-------
Figure 5.5
200 m layer-average trajectory for day
216. Trajectory A is for 200-400 m AGL
layer, Trajectory B is for 900-1100 m
AGL layer, and Trajectory C is for 1700-1900
m AGL layer. "•<*•" indicates the end point
of the 12-hour computed 200-400 m AGL
trajectory. Each segment represents 12 hours
29
------- An error occurred while trying to OCR this image.
-------
SECTION 6
TEMPORAL VARIATION IN THE SURFACE
ENERGY BUDGET COMPONENTS FOR THREE
LAND USE PATTERNS
Dennis C. Doll, G.F. Watson, and J.K.S. Ching
Abstract
Special St. Louis RAPS data were used to examine the
diurnal variation of the net radiation and heat flux components
of the surface energy budget for a two-month spring time period
for blacktop, concrete, and soil surfaces. Sensible and latent
heat fluxes were analyzed from measured and calculated net
radiation and ground heat fluxes for each surface.
Ground heat flux for concrete and blacktop was determined
for different water contents of the underlying soil layer. A
lower water content in the soil layer reduces the magnitude of
the total ground flux throughout the blacktop and concrete layer
and soil sublayer. An accurate determination of soil water
content appears necessary in any ground heat flux evaluation.
The sura of sensible and latent fluxes for the soil surface
was found to be nearly twice as large as that for the blacktop
layer and approximately three times larger than for the concrete
surface during the daytime period of the diurnal cycle. Latent
heating appears to be an important component of the surface
energy balance for the soil.
Hourly average net radiation and ground heat flux values
resulted in an improvement of a parameterization technique based
on the ground heat flux to net radiation ratio (G/Fn). Diurnal
variation in G/Fn for each surface is distinctly different. For
concrete and blacktop, G/Fn is constant at night but variable
during the day. For soil, G/Fn changes throughout the diurnal
period.
31
-------
INTRODUCTION
An essential part of determining the behavior of the
boundary layer lies in knowing the surface layer fluxes,
particularly that of heat, without which one cannot hope to
specify pollution transport or mixed layer- dynamics (Carlson and
Boland, 1978). These surface layer fluxes, particularly heating
patterns, are strongly tied to the character of the ground
surface temperature and its substrate. The direction of the
energy flow across the atmosphere-ground surface interface is
primarily determined by the difference between the ground
surface temperature and the air temperature above it
(Bhumralkar, 1975). Therefore, in studies of atmospheric
circulation it is necessary to know the ground surface
temperature accurately.
• The determination of ground surface temperature within a
general circulation or boundary layer model is usually
accomplished by solution of the surface energy balance equation
( Deardoff, 1978). It is important that the components of the
surface energy budget be rather carefully specified if an
accurate determination of ground surface temperature is desired.
If a model is being considered for a region where different
land-use patterns exist, for example urban and rural areas,
these same components must be evaluated for each surface type
considered in the modeling in order to eventually determine
representative surface temperatures. It is important,
therefore, that comparisons of the different surface energy
budget fluxes be assessed in order to determine how these fluxes
vary among the surfaces. These fluxes can be compared for
annual or diurnal characteristics depending on the time scale of
the modeling effort.
Data collected during the subsurface heat flux experiment
of St. Louis Regional Air Pollution Study (RAPS) are the basis
for the present evaluations. The field site was the St. Charles
County Airport (formerly Sraartt Field) about 35 km NW of
downtown St. Louis. The concrete runways and adjacent
soil/grassy area were to represent urban land use materials. A
portion of the runway was painted black to simulate asphalt.
Net radiometers were installed directly over each surface, while
thermistors embedded at several levels from the surface to 50 cm
depth provided measurements for determination of ground heat
flux. This particular RAPS experiment was run from January 1976
to May 1977. However, only the last two months (April-May) had
32
-------
concurrent soil moisture data, and the present study is
restricted to this period.
Soil moisture affects both the soil density and heat
capacity and is crucial to the accurate determination of the
ground heat flux. Soil moisture was directly determined for the
soil/grassy surface. However, for the soil sublayer beneath the
12 cm thick concrete runway it was necessary to assume values.
An assumed 35% water content by weight was the average of that
for neighboring soil, while a 60% value was appropriate to
saturation.
As mentioned by Deardorff (1978), the one component of the
surface energy balance that is particularly difficult to
evaluate in numerical models is the ground heat flux term.
Several authors have suggested parameterizing the ground heat
flux as a function of the more easily defined net radiation.
This parameterization involves defining a proportionality
constant for the ratio of ground heat flux (G) to net radiation
(Fn). Nickerson and Smiley (1976), for example, have suggested
using the constant 0.19 to represent G/Fn for daytime periods,
and G/Fn = 0.32 at night. Idso et al. (1975) have suggested a
range of values for G/Fn around 0.4"~d"epending on the type and
moisture content of the soil.
In this study, then, Fn and G are directly measured. The
surface energy budget equation Fn - G = H + LE also provides
informatin on the sum of the sensible (H) and latent (LE) heat
fluxes. The following results are the mean diurnal variation of
the various fluxes and ratio G/Fn over the two-month spring
period.
RESULTS AND CONCLUSIONS
Evaluation of the energy balance components for each
surface pattern considered in this analysis showed significant
differences between soil and concerte surfaces and soil and
blacktop surfaces. Variations noted between the concrete and
blacktop were primarily attributed to differences in albedo.
The diurnal variation of mean net radiation for all
surfaces (Figure 6.1) showed a very similar pattern with the
maximum occurring around 1300 Local Time (LT) near the time of
the maximum flux of solar shortwave radiation. Variations in
magnitude are primarily attritubed to differences in albedo
between the concrete, blacktop and soil surfaces. Nighttime
values of net radiation showed larger magnitudes for concrete
and blacktop as compared to the soil.
Figure 6.2 shows the magnitude of ground heat storage to be
33
-------
more Chan twice as large for concrete and blacktop as compared
to that for soil. The maximum ground heat flux occurs near the
time of maximum solar shortwave radiation for the concrete and
blacktop. However, for soil, the time of maximum ground heat
flux is shifted nearly two hours earlier to approximately 1000
LT. A large upward flux of heat occurs within concrete and
blacktop at night due to the large amount of heat stored during
the daylight hours. For soil, there is a much smaller upward
flux compared to the concrete and blacktop due to the much
smaller amount of heat stored during the daylight hours.
The diurnal variation of sensible heat flux (H) plus latent
heat flux (LE) is found to be much greater in amplitude for the
soil compared to the concrete and blacktop. The maximum daytime
value for soil is nearly twice as large as that of the blacktop
and approximately three times greater than that for concrete
(Figure 6.3). Since the magnitude of H plus LE can be
considered to be primarily due to H for the concrete and
blacktop, it appears that latent heating effects are a major
contributor to the large magnitude of H plus LE noted for the
soil.
Figure 6.3 also shows an important characteristic of the
diurnal curve of H plus LE for concrete and blacktop, namely,
the occurrence of positive values during the nighttime hours.
There is thus a continuation of a flux of heat from the surface
to the atmosphere. Nighttime values of H plus LE for the*soil
were found to be small in magnitude and only slightly positive.
The diurnal pattern of the quantity G/Fn for concrete and
blacktop in Figure 6.4 reveals that nighttime values are
positive and greater than 1.0, indicating that the nighttime
ground heat flux is of greater magnitude than the net radiative
flux. The nighttime temporal variation of G/Fn is seen to be
rather small. During the daylight hours, G/Fn varies little,
except near the time of the evening transition period around
1700 to 1800 LT when G/Fn becomes negative. This is due to the
change in the sign of ground heat flux that occurs prior to that
for net radiation.
The mean diurnal pattern of G/Fn for soil in Figure 6.5
illustrates a much different pattern than that for concrete and
blacktop. However, nighttime values are again found to be
positive and greater than 1.0, similar to that for concrete and
blacktop, although smaller in magnitude for the soil. Values of
G/Fn just after the sunrise transition period around 0600 LT
show values decreasing from approximately 1.0 to nearly 0.0 at
about 1500 LT when the sign change in ground heat flux occurs.
This is almost three hours earlier than the sign change in net
radiation.
In conclusion, it seems clear that soil moisture content
34
-------
has a very important effect in the heating rate of a soil
sublayer. Generally, a moist soil layer will have a greater
heat flux than a dry soil layer due to the larger thermal
conductivity for the moist layer.
A concrete material over soil transmits heat effectively to
the soil sublayer. Concrete generally has a higher thermal
conductivity and lower heat capacity than soil allowing a
greater heat flux. This greater heat flux for concrete allows a
larger storage of heat to occur during the day compared to soil.
This results in a greater release of heat during nighttime hours
which aids in maintaining the surface temperature of the
concrete and blacktop at warmer values than that of the soil.
Although the sensible heat flux (H) over the concrete and
blacktop may be larger than H over soil, the total of H plus the
latent heat flux (LE) is still larger for the soil than that of
concrete and blacktop. This suggests that the latent heat flux
is a major component of the energy budget for the soil
material.
The contrast of energy budgets for these adjoining
materials demonstrates the complexities involved in undertaking
energy budget analyses. In mesoscale modeling, it has been
demonstrated that land use variability greatly influences the
energy budget component magnitudes, and must be considered in
any type of mesoscale energy budget analysis.
The parameterization of G to Fn ratio is possible under
undisturbed weather conditions. A strong diurnal effect of G/Fn
is observed for all three surfaces although hour to hour
variations may be small. This suggests that a traditional use
of a single value for G/Fn as applied by some authors during the
day and night does not accurately describe the parameterization.
The results presented here show G to be greater than Fn at night
for all three surfaces, although those for soil are less than
those for the concrete and blacktop. Also, these results appear
larger than those reported in most other studies.
Some suggestions for further research include:
1) An attempt to expand this two-month study using data
obtained over all seasons so that seasonal variations
in the energy budget can be evaluated.
2) A repeat of this rural site experiment at a more
urbanized area. The magnitudes and characteristics of
the energy budget for various surface materials can be
contrasted to those found in this study.
35
-------
3) A test to see whether applying the results presented
here for the variation of G/Fn with respect to time of
day and land use would in fact provide for more accurate
flow predictions in boundary layer models.
36
-------
ca
I I
I I I I I I
.J *
LD
^& O ^& ^p ^^ ^^ ^S*
® Lft ® Lfi O LD ®
T cn n cu cxi •** ^*+
CN
il
,
(U
•U)
I I I I I I I FT
^9 ^^ ^&
U) 9
I •*-<
I
E
\
3
o: C
O
r^ O
33
3
afl
•r-l
IL,
37
-------
J
<:
TIT I I
III
v
Till
.">
cu
.9
OJ
O
.®
•U)
-------
as
o
on
0)O
^ II
n jj
CO CU
c w
-------
I
.in
cu
cu
o
OLf»OLftOljD«S>L/>Ln«S)Lf)OLr)«3>
• •*•»*•••••••••
•^•nfjfUfu^-i^-tooo^H^-iajojr'J
i i i i i i
"J/9
Q.
O
u
X,
o
O 0)
M-l C
•-I
Cc-l
En
•^T3
O 0)
JS
o en
•—i cQ
4JT3
(U -U
j- CU
•u u
o
H-4 C
O O
U
C
o-o
•^ C
U CO
CO C
C •-<
i-lr-1
3 O
•^ co
U
U
3
00
40
-------
1
9
1
<
1
>
1
<
1
>
1
<
1
>
1
X
1
O
1
<
]
4
X
*
X
K
X
1
>
X
»
t
1
<
X
1
>
X
I
<
'
^J
X
X
1
>
X
X
<
1
<
X
1
>
1
<
1
9
,
1
<
1
>
1
<
1
>
1
<
1
>
1
<
1
>
1
<
OI
-
-
m
cu
" i l-l
: o
u>
<«-« u
o
QZ u
O -Q
M *
vO
0
3
00
•t-4
CO
(0
«
CO
•
* •
°
0)
3
00
»
I I I I I I
"J/9
41
-------
SECTION 7
TRENDS IN ATMOSPHERIC VISIBILITY ACROSS
THE UNITED STATES FROM 1955 TO 1972
Jerrold S. Foster, T.L. Tsui, and G.C. Holzworth
Abstract
Regular daytime synoptic observations from 14 stations
across the United States but excluding those with relative
humidity greater than or equal to 90% and those with
precipitation or fog present, were used to determine the trend
in visibility from 1955 to 1972.
Comparison of statistics at the beginning and the end of
the period, and linear regression of the monthly mean
visibilities were used to discover relationships between
visibility and other meteorological variables such as wind
speed, wind direction, and relative humidity. Spectrum and
cross-spectrum analysis of the time series of visibility and
relative humidity were also performed.
The results of this study indicate: (1) a deterioration of
the visibility at all 14 stations, (2) a bimodal (six and twelve
month) oscillation in the annual cycle of visibility, and (3)
an inverse correlation of relative humidity with the annual
cycle of visibility.
42
-------
INTRODUCTION
Growing concern for the quality of our air, not only with
regard to health, but also for aircraft safety and for the
aesthetic beauty of our environment has motivated this study.
The concentration of various constituents of the atmosphere,
such as water droplets, dust, smoke and other pollution
particles, condensation nuclei, and the air molecules
themselves, determines the optical transparency. Therefore,
'visibility* is often used in evaluating the quality of our air,
especially in aviation operations. Consequently, it is of
interest to aircraft safety and air clarity to understand how
visibility changes and the relationship between visibility and
other meteorological variables.
However, because of the human factor involved in the
estimation of visibility, it is sometimes difficult to evaluate
the changes in visibility. That is why a very large data set, 18
years, is used in this study. Such a' large sample should
override most of the subjectivity inherent in the data. In
addition, subjectivity can be overcome by using some sort of
smoothing technique, such as averaging, in order to make each
member of the data set less dependent on the preceding one.
In meteorological usage, the term 'visibility' means both
the condition of being visible and the "visual range." The
latter term was introduced by Bennett in 1930 to indicate the
distance at which something can be seen, i.e., the clearness
with which objects stand out from their surroundings. Since
then the definition of visibility has undergone many revisions.
Throughout the period covered by this study "prevailing visi-
has been defined to be the greatest visibility, equaled or
exceeded throughout at least half the horizon circle, which need
not be continuous. One difficulty in the analysis was the use
of the "+" in reporting the visibility. This symbol was used
when the observer judged the prevailing visibility to be more
than seven miles and also more than twice the distance to the
most distant marker visible. The observer then encoded the
visibility as twice the distance to that marker, or seven miles,
whichever was greater. If the visibility was greater than the
coded value, then a "+" was added. This symbol was deleted from
the observations when they were archived thus making it
impossible to determine if, and by how much, the visibility
exceeded the recorded value. Therefore, any study of visibility
trends becomes more problematic, and may be somewhat biased.
43
-------
Complete airways observations from National Weather Service
reporting stations, archived on magnetic tape at the National
Climatic Center at Asheville, North Carolina were used in this
study. Observations from 1955 through 1972 for sites with
continuous records and which had not significantly changed
location were selected. Of these 14 stations only observations
between 0900 and 1600 local time for which precipitation, fog,
or relative humidity of 90% or more were not reported were
included.
Trend analysis methods by Holzworth and Maga (1960) and
standard time series and cross-spectrum analysis were used to
examine the changes of visibility and its relation to variations
in wind speed, wind direction, and relative humidity at each
reporting site.
Research in visibility as an indicator of air quality has
been going on most of this century. One technique described by
Holzworth (1959) and common to many studies is to determine the
percent frequency of occurrence of the visibility in a few (four
or five) categories within the range of the visibility at the
station in question. The percent frequency is calculated at the
beginning and the end of the period being studied. Holzworth
( 1959) used this technique and found that the visibility near
Sacramento, California had deteriorated from 1935 to 1956. The
author also related poor visibility to wind directions which
brought pollutants from urban areas. Holzworth (1962)
determined that the visibility at various cities was lower from
1955 to 1959 than from 1929 to 1938 because of an increase in
the frequency of visibility less than seven miles. This
decrease was found to relate to the wind speed and direction
relative to pollution sources. Miller et al. (1972), using
comparisons of statistics at the beginning and the end of the
period coupled with multiple regression showed that the summer
daytime visibility in Ohio, Kentucky, and Tennessee deteriorated
from 1962 to 1969, and that wind direction and relative humidity
were significantly related to visibility.
The southwest United States is an area of great interest in
visibility studies today. Because of the typically large
visibilities, over 100 miles, reported by stations in this area,
a decrease of only 10 or 15 per cent in the prevailing
visibility is considered a large change, and is of more concern
than a similar percentage decrease in the more populated eastern
United States. A recent study by Trijonis and Yuan (1977) showed
the visibility throughout the southwest United States to be
decreasing. The concern over the deterioration of the
visibility in this area, as well as over the remainder of the
United States, points to the need for continuing research of
visibility trends.
Because of these findings, the present visibility study was
44
-------
undertaken to verify whether deteriorating visibility was common
to many stations across the country. If this is the case then
what meteorological elements, or their combinations, are
responsible? Linear regression and time series analysis will
seek relationships between visibility and other meteorological
variables.
RESULTS AND CONCLUSIONS
Some of the difficulties inherent to an investigation of
visibility were brought out in this study. Problems were
lessened by using a large data base of 18 years. One particular
problem encountered by this study was the use of the "+" in
reporting visibility. Unfortunately the "+" symbol was not in
the archived data and the results biased by changes in the usage
of the "+" symbol cannot be estimated. If the "+" symbol did
impose an influence on the outcome of the trend analysis, the
trend deduced should have been discontinuous. This was shown in
the Austin, Texas visibility data. In general, the "+" symbol
has not seriously affected this study.
The first major finding was the downward trend in
visibility at all 14 stations listed in Table 7.1. This was
shown by the linear regression method (for example, Figure 7.1),
and more dramatically by the method developed by Holzworth and
Maga (1960). This result may be due to new urban and/or
industrial developments around the airports or to an increase in
the amount of aircraft traffic. The deterioration of visibility
from 1964 to 1970 may also be due to an increase in the sulfate
percentage of the total particulate matter in the air.
Another finding of this study was a bimodal annual
oscillation in the mean monthly visibility cycle. Spectral
analysis indicated that two major components of the total
variance occur with periods of approximately six and twelve
months. The spectrum for Raleigh-Durham, NC is displayed in
Figure 7.2. The positive correlation of the visibility time
series was high among stations within similar geographical
areas, indicating that geographical region is one major factor
in determining the periodicity of the visibility.
Further examination of this result determined that the
monthly mean relative humidity also exhibits a bimodal annual
oscillation, but 180 degrees out of phase with the visibility.
This relationship is illustrated for two stations in Figures 7.3
and 7.4. There seems little doubt that seasonal variability in
relative humidity is a major, if not the dominant, cause of the
bimodal oscillation in the annual visibility cycle. Other
factors such as wind speed, wind direction, the number of daily
observations used in this study, and station location may also
play an important role in the periodic variability of
visibility.
45
-------
• ft 6 A
IAN
— •—.
55
DEC
...4...
S'J
MONTH
• 4 4 4
oec
oec
MONTH
"• 7 2
Figure 7.1 Trend deterained from deviations in
visibility (miles) from the respective
monthly mean (1955-1972) at Birmingham,
AL. Note the general decrease in visibility
in the latter half of the period.
46
-------
•*— PERIOD (mo) s
Figure 7.2 Fast Fourier Transform of Visibility
Time Series of Raleigh-Durham, N.C.
47
-------
S44
52
48H
f'S
4-13
•12
^VISIBILITY -mi
IAN
JUN
MONTH
DEC
Figure 7.3 Bimodal annual oscillation of visibility
and relative humidity at National
Airport (D.C.).
DEC
Figure 7.4 Bimodal annual oscillation of visibility
and relative humidity at Cleveland, OH.
48
-------
Table 7.1 Stations used in this study and grouped according
to geographical region.
Distance and direction Elevation
Station of airport from city (ft)
1. Washington, D. C. (National) 3 mi. S 65
Raleigh-Durham, N. C. 10 mi. NW 441
Atlanta, Ga. 7 mi. SW 1034
Birmingham, Ala. 6 mi. NE 630
2. Houston, Tx. 7 mi. SE 662
Austin, Tx. 3 mi. E 621
3. Tucson, Az. - 2555
Phoenix, Az. 3 mi. E 1107
4. St. Louis, Mo. 13 mi. NW 564
Peoria, 111. 6 mi. SW 662
Milwaukee, Wis. 7 mi. S 663
Ft. Wayne, Ind. 5 mi. S. 828
Cleveland, Ohio 10 mi. SW 805
Youngs town, Ohio 8 mi. N 1186
49
-------
SECTION 8
A VISIBILITY STUDY IN THE EASTERN HALF OF
TEXAS FROM 1949 TO 1968
Alfonse J. Mazurowski, A.J. Riordan, and G.C. Holzworth
Abstract
Hourly weather observations for times 0900-1500 GST were
used to examine visibility trends, obstructions to vision, and
the effects of different parameters on visibility for eight
stations in the eastern half of Texas from 1949 to 1968.
Observations with any obstruction to vision, along with the
individual obstructions by haze, smoke, and dust, were examined
to determine their effect on visibility. Mean visibility
categorized by wind direction was employed to investigate the
influence of winds from industrial areas and cities on
visibility at airports. Trend analyses of mean visibility data,
both annual and seasonal, were performed using mean ridits and
shifts in frequency from one visibility range to another. The
effects on visibility of population growth and distance of
stations from the Gulf of Mexico were also analyzed.
The main results of this study are: ( 1) a deteriorating trend
in annual visibilities was apparent for seven of the eight
stations, (2) smoke was a major factor in the reduction of
visibility in Houston, (3) smoke appeared to be a result of
local transport of air pollution while haze and dust seem to be
associated with long-range transport processes, (4) daily
resultant winds from nearby industrial areas produced the lowest
mean visibilities, (5) mean visibility increased significantly
as the distance of stations from the Gulf of Mexico increased.
50
-------
INTRODUCTION
Visibility has been considered a valuable tool in
determining the quality of the air for many years. Since its
inception, the aviation industry has been concerned with
visibility, but recently individuals and groups have become
involved with investigating the apparent degradation of the air
quality. This involvement has resulted in an increased
awareness of the trend of declining visibility throughout the
United States. As a result, visibility studies have become more
numerous in the interest of aircraft safety and air quality
improvements.
Visibility studies have been conducted at various locations
in the United States. References cited in the previous section
of this report (SECTION 7) are also relevant to this study, and
these are not repeated. Some additional references include
Holzworth (1962) who examined the percent frequency of all
visibilities less than 7 miles at stations scattered throughout
the United States in the period 1929 to 1961. Results indicated
that visibility was improving in many of these urban areas.
Holzworth suggested that the possible reduction of the emission
of visibility-reducing materials to the atmosphere, such as the
smoke from coal combustion, may have been responsible for the
visibility increases. Faulkenberry and Craig (1976) showed that
a trend of declining visibility was apparent for Salem, Oregon
in January and August from 1950 to 1971. But at the same time,
no significant trends were indicated at Portland and Eugene,
Oregon.
The long range transport of air pollution has been studied
by following areas of reduced visibility during periods of
increased air stagnation. Hall _et _al. (1973) used surface
visibilities and an 850-mb wind trajectory analysis to show that
air pollution advanced as much as 1100 km from the central
midwest into the upper midwest and Great Plains in August 1970.
Visibility contour maps, daily weather maps, and long-range
trajectory analyses were employed by Husar et al. (1976) to
monitor the long range transport of pollutants~Tn June and July
1975 in the eastern half of the United States. They found that
haziness may be caused by long range transport from sources 1000
km or more away.
51
-------
This report examines visibility at eight stations in the
eastern half of Texas (Table 8.1). The area stretches from the
Gulf of Mexico to the Oklahoma border. The region is of
interest because it has had large population increases since the
1940's, and the effect of topography appears negligible.
The eight stations were selected based on continuity of
location, and data were examined for trends or discontinuities
in visibility relating to changes of the observation site. The
largest change was 5.6 km at Corpus Christi, but there was no
significant change in visibility associated with that
re-location. Data after 1968 were not included for any site
because after this some sites began reporting maximum
visibilities as 7+ rather than 15 miles (see Section 7).
Finally, only observations for 0900 to 1500 GST without
precipitation or fog were included. Prior to 1965 observations
were hourly, but from 1965 through 1968 they were archived only
for 3-hour intervals.
The influence of the Gulf of Mexico on the stations is
investigated as are the obstructions to vision that are
associated with air pollution problems. The existence of
visibility trends, the effect of various non-meteorological
parameters on visibility, and the influence of long range and
local transport of air pollutants on visibility also are
examined.
RESULTS AND CONCLUSIONS
The predominant resultant winds on an annual basis are from
the direction of the Gulf of Mexico for all eight stations. The
Gulf of Mexico, therefore, has a large influence on the climate
of the area. These resultant winds from the Gulf of Mexico are
also conducive to unstable conditions over land, giving large
mixing heights and decreased potential for air pollution
problems.
Houston has by far the greatest number of days with
obstructions to vision reported. The remaining stations have
considerably fewer such days (Table 8.2). Every station, except
Fort Worth, has significant negative correlation between mean
annual visibilities and the number of days per year with
obstructions to vision (Table 8.3).
Table 8.4 shows that smoke appears to be a major factor in
reducing the mean visibility in Houston. In Dallas and Houston,
smoke is reported a greater percentage of the time when
resultant winds are from the nearby industrial areas. Figure
8.1, and others, suggest that smoke is a local rather than a
regional problem.
52
-------
Days with haze occur most frequently in the South Central
and Gulf Coastal regions, but differences among all the stations
are small (Table 8.5). The uniformity of the number of days with
haze at all the stations suggests that haze is generally a
regional problem. Further evidence is provided by Figure 8.2 in
which the predominant wind direction associated with haze
occurrence at Houston is not from the industrial area.
Observations with dust are widespread over the eastern half
of Texas, with small differences in the number of days with dust
noted among the stations. Drought conditions in the 1950's were
responsible for the majority of the dust reported in the study
period, and a case study during this period showed that
large-scale transport processes contributed to the dust
observed.
The most important result of the present study is the
apparent trend of deteriorating visibility (Table 8.8) on an
annual basis at every station, except Brownsville. Seasonal
analysis shows that Houston is the only station with a trend of
declining visibility in all seasons. Two contrasting methods
are utilized for determining the visibility trends. Both give
similar results. The Holzworth and Maga (1960) method employs
the res-ultant shift between visibility ranges to determine
trends, while the method introduced by Faulkenberry and Craig
(1976) uses mean ridits. Ridit analysis (Bross, 1958) is
appropriate when the response variable has numerical values but
the measurement system is heavily dependent on the technical
skill of the scientist. Because visibility observations involve
a subjective decision by an observer, the results reported
depend on the skills of the observer.
Increases in city population, when correlated with the
number of days with mean visibility in the 13-15 mile range, are
found to decrease significantly the frequency of visibilities
in this range at all stations except Brownsville. Another
factor that influences visibility is the distance of a station
from the Gulf of Mexico. As this distance increases, the mean
visibilities for the study period increase significantly.
Tables 8.6 and 8.7 document, respectively, visibility dependence
on population trends and distance from the Gulf.
Because of the continued migration of people and industries
to the "Sun Belt" states, future visibility studies should be
considered in this portion of the country to monitor the effects
of this migration. Detailed case studies would be especially
beneficial in determining the transport processes of both haze
and smoke, and to determine if the smoke observed in Houston
might be affecting the visibility in nearby cities.
53
-------
Figure 8.1
of S2SLfjequency of the ratio of the number
Sfr!?MS t73 urOID a Partic^*r resultant wind
«»I J i the.tocal number of days with the
same resultant wind direction at Houston, 1949-
54
-------
CiTvTiMiTS
•% . .» .•.-.-.•
\ .....
..•Vtf-x'S
HOUSTON .XjivilXx'ixXv
XvXvXvXvX'l*
..;
.*!*•
/ it
ji ||
**x*
:::::::::;::>::^::::x::^::
;X%;Xv.';X;X;.';X;:;.'
:;:;:;:;:|:;:;:x'x;:;:;:;x|
S^Si^i^SS^:
•X;X;X;X;.';X;.';X;X
xx'.:
•"'X*
" ".".*
V**-v
.V;X\;.";X;'
:-:•:•:•:•:•
:;XX\>;
*""i"""»*"*r
|:| i N 0
.••:5:-::-:-:
•'i-'i&'x'
• ::•:•:•:•:•;•
•^ *t *x\
'u s T'R' A
A
.'
j
^
REA
if
•;-x '-XT
vX;X;
!*"*I*X*
1 &X+
:•:•:•:••.'
:.::x'x:
:|xj:v
:' :-x'S:
-.«.--.-.-.»•-- , • ,- x.:.;.: ;.:
..
% ;.
: .';
• %
.*.*.».
X"
• • •;
A
'.~.f.
.:x
:•:•
.:;.'l
"Vfalf
,
.* .".
• '"'X
; i .*
; »XVN^ ^*
T'. ..
- '•'
IBMMM
:.::'x: :
*i"!"^*.* *
.*.*.*.*
••<••:•'•'»'
j LA POSTg 7 ]
*^ ^ A ^^^ 1
GALVESTON
BAY
9 ! ?
Se«i*
Figure 8.2 Same as Figure 8.1 except for haze days
55
-------
n
a *;
« oe
at
z
2
o
x
<
00
OS
c
•»^ •
^ oo
v£>
T3 ON
CrH
0) I
tt ON
4J >
05
1-4
>
03 3
•^ O
> M-l
3 4J
c
5 c
-------
Table 8.1 Station information and data years used.
STATION
Houston Intl. Aprt.
Dallas Love Field
San Antonio Intl. Aprt.
Greater Southwest Intl. Aprt.
(Fort Worth)
Austin Municipal Aprt.
Corpus ChHstl Intl. Aprt.
Ulchita Falls Municipal Aprt.
RGV Intl. Aprt.
(Brownsville)
LOCATION
IDENTIFIER
HOU
DAL
SAT
GSM
AUS
CRP
SPS
BRO
DATA YEARS
USED
1949-1968
1949-1968
1949-1968
1954-1968
1949-1968
1949-1968
1949-1968
1949-1968
DIRECTION i
DISTANCE FROM CITY
16 km
9 km
13 km
27 km
4 km
3 km
11 km
5 km
SE
NW
N
ENE
ENE
w
N
E
ELEVATION
IS «
145 a
241 *
166 •
187 n
13 «
305 a
5 *
Table 8.2 Number of days with obstructions to vision and
mean daily visibility.
STATION
Houston
San Antonio
Dallas
Corpus ChrisM
Brownsville
Austin
Fort Worth
UlchUa Falls
• 1954-1968
« 1952-1968
NUMBER OF DAYS
WITH OBSTRUCTIONS
(1949-1963)
1743
496
484 (379)»
361
344
303
134-
166
MEAN DAILY VISIBILITY
(MILES)
(1949-1963)
11.2
13.2
13.3
12.9*»
12.3
12.9
13.9*
14.1
57
-------
Table 8.3 Results of the correlation between annual mean
visibilities and the number of days with
obstructions, 1949-1968.
STATION
Houston
0*11*s
S«n Antonio
Fort Worth *
Austin
Corpus Christl •*
Wichita Fills
8rcwnsvi11e
CORRELATION
COEFFICIENT
-.360
-.306
-.854
-.422
-.593
-.685
-.472
-.451
SIGNIFICANT AT
THE 5"! LEVEL
X
X
X
X
X
X
X
• 1954-1968
** 1952-1968
Table 8.4 The number of days with smoke and the mean
population for all locations.
STATION
HOU
DAL
CftP
SAT
AUS
GSU
SPS
8RO
NUMBER OF SMOKE DAYS
(1949-1968)
1344
243 (175)«
73
70
30
12*
6
3
MEAN POPULATION
(1949-1968)
379,018
628,937
154,759
540.638
180,491
354,349
39,914
44,335
1954-1963
58
-------
Table 8.5 Number of days with haze,
STATION
SAT
am
AUS
cap
HOO
OAL
ssu
SPS
NUMBER OF DAYS WITH
(1949-1968)
341
233
203
136
1S3
123 (100
116*
43
HAZE
*)
1954-1968
Table 8.6 Correlation coefficients between population
growth and (a) the frequency of days with
mean visibility less than 7 miles and (b)
visibilities in the 13 to 15 mile range for
all locations.
STATION
Houston
3*1 Us
San Antonio
Austin
Corpus Chrlstl
Fort Worth
Wichita Falls
Brownsville
CORRELATION
COEFFICIENT
(•) (b)
.399-
-.012
.331-
-.013
.230
-.392
-.031
-.207
-.739*
-.821*
-.501*
-.616*
-.320*
-.723*
-.564*
.023
POPULATION GROWTH
1949-1968
593,387
390 ,963
247.383
110.749
93,970
76,260*-
32,547
16.958
PERCENT
GROWTH
51",
48?
395
461
431
205—
33t
335
Statistically significant *t the 5^ level
1954-1963
59
-------
Table 8.7 Distance and direction from the Gulf of Mexico
and mean visibility for all locations.
DISTANCE (KH) AND MEAN VISIBILITY
STATION DIRECTION FSCM THE (NilM)
SUIF OF MEXICO 1949-1968
Corpus Christ! 23 MNH 12.9*
Brownsville 29 « 12.3
Houston 84 NU 11.2
San Antonio 22S FM 13.2
Austin 259 I'M 12.9
Dallas 450 ' NNU 13.3
Fort Worth 475 NM 13.9*
Vlchtta Falls 644 NW U.I
• 1952-1968
*• 1954-1963
60
-------
^
c
^-M>CO
O •«- cn
cn •!-
F- J3 O
I/)
it) -r-cn
cn > «•
ro en
z: a*.—
•si's
« t-
m s-
j= .— aj
•M a.
i- O
§•*-> o
.c
rsj co 4J
i— r—
O O>
a: aj c
O 3
•a "o "°
o -o
•»-> u ai
a; c s_
E OJ -M
H-
(rt O) 4J
2 §•**$
(O it]
C J= CO
it] U C
o
^ aj •^-
c cn+J
V 03 03
i. -I-1 U
*J C O
> u
•4-> i- -a
•r- CO
•r- <1J C
ul C O
=" .c S
fr^ (/)
CO
CO
Q£
Ul
Z
_J
i
ZB
O
"-•
1—
4^
t—
ty^
^o ^r ^o ^o tp
^r u^ C^ *••• v^^
Lf5 CSJ O> LO UO
CM CN4 CV4 C\J CM
1 1 t 1 1
^m ^ft ^* CO O
^- ^ O yD CSJ
• • • • •
CO i*^ vo cn m
CM i— »— ^- <—
1 1 1 1 1
en ? In cn r^ rC
• •••••
^^ cn f^ ^r ^r cn
CM i— CM r— CM cn
1 1 1 1 1 1
o —
•»- JZ S.
C *J J=
o ^ O
C *J O
a vt c 2 c u)
*J tS < ••- 3
U) ^— • 4>J 4«) CX
3 <— C S. Ul U
O Ki
4-> U)
"- C
g™ 2
o o
— i.
2 ca
— I
61
-------
SECTION 9
SYNOPTIC-SCALE VARIABILITY IN
ATMOSPHERIC SUSPENDED SULFATE CONCENTRATIONS'
Brian W. Galusha, G.F. Watson, and G.C. Holzworth
Abstract
The spatial variation in atmospheric suspended sulfate
concentrations is studied for evidence of meteorologically
linked sulfate transformation and transportation. Low frequency
( every 12-14 days) data from 41 National Air Surveillance
Network stations as well as higher frequency (every 2-3 days)
data available from two special studies are examined.
Variations in high frequency data with wind direction for St.
Louis, Missouri, are compared to local and regional sources of
precursor pollutants for two one-year periods; January-December
1969 and April 1975-March 1976. Spatial variations in low
frequency sulfate concentrations during the spring season
(March-May) for the six-year period 1969-1974 are compared with
synoptic weather and wind circulation patterns. These
comparisons indicate that regional and local sulfate transport
can largely account for large-scale sulfate variations. Sulfate
transformation due to humidity, temperature, and sunlight
intensity do not adequately explain observed variations, but may
be of secondary importance.
The existence of regionally high concentrations of sulfate
in the northeastern United States and of a general summertime
peak in sulfate values is confirmed.
62
-------
INTRODUCTION
Atmospheric suspended sulfates have been identified as a
significant pollution problem by the Environmental Protection
Agency (EPA). These sulfates have been related to adverse
health effects (EPA, 1974c), reduced visibilities (Wilson et
al., 1977), and the corrosion, discoloration, and deterioration
oT materials (EPA, 1975).
The term "sulfates" applies collectively to a large class
of sulfur commpounds. Any single physical property which
characterizes all is difficult to define. One of the major
distinctions is that some sulfates are hygroscopic and others
are not. Although some researchers have found it convenient or
necessary to assume homogeneity, generalizations about sulfates
are often invalid. Thus, results of studies conducted in one
section of the country may not be applicable to other regions.
Adding complexity to the problem is the fact that sulfates are
usually secondary pollutants transformed chemically from the
precursor agents such as sulfur dioxide (S02) and hydrogen
sulfide (H2S).
The present study is undertaken with the hope that a more
detailed meteorological investigation can provide a better
understanding of the mechanisms involved in sulfate transport
and transformation. In particular, this study proposes to
determine whether the sulfate concentrations display
well-defined patterns over areas resolvable by the normal
meteorological surface observation network. This implies areal
extents of a few hundred kilometers (synoptic-scale).
Patterns in sulfate distribution on this scale will be
determined from data collected by the National Air Surveillance
Network (NASN). There are about 100 more or less continuously
reporting stations in the NASN with sampling frequency about
once each 12 days. In this study only urban sampling sites in
close proximity to National Weather Service (NWS) stations and
in the eastern United States are employed. Further restrictions
include: (i) samples must be chemically analyzed by the same
(methylthymol blue) method; (ii) each station must have at least
five samples from each season during any year; iii) each station
must have four valid years of data out of the six-year period.
The 41 NASN stations satisfying these criteria are shown in
Figure 9.1 .
63
-------
Synoptic scale patterns in sulfate distribution will next
be presented and related to patterns of meteorological
variables.
RESULTS AND CONCLUSIONS
The presence of significant synoptic-scale variability in
sulfate concentrations in NASN data is illustrated by that at
Nashville, TN in Figure 9.2. Seasonal mean values at a number
of stations (Figure 9.3) show the summer maximum and generally
higher values in the northeast states.
This study has arrived at a number of conclusions. Some
conclusions must be considered tentative due to the marginally
acceptable sulfate data presently available. Future
researchers, utilizing data of better quality and quantity, may
be able to determine whether these conclusions remain valid.
When sulfate values are represented as deviations from
seasonal means and plotted on maps, synoptic-scale patterns
emerge as in Figure 9.4. This figure, and others analyzed for
the weather active spring season, show coherent patterns of +
and - deviations. However, a systematic relationship of these
patterns to those of weather variables is not obvious. For
example, the large positive values over the Carolinas and
Virginia on May 11, 1974 are associated with cloudy skies and
warm temperatures. On March 10, 1969 positive values in the
same area are related to mostly clear skies and near-freezing
temperatures. Large negative values over Nebraska and Kansas
are related to clear skies and warm temperatures on May 11, but
to cloudier skies and sub-freezing temperatues on March 10.
Although expected correlations with temperature, moisture, and
sunshine were not clearly evident, these are still thought to
exist. Apparently, such relationships are overshadowed by
sulfate transport.
The importance of sulfate transport was most readily
demonstrated at St. Louis, Missouri where sufficient data were
available and sources of sulfate or precursor agents are
nonuniformly distributed around the horizon (see Figure 9.5).
Similar results were apparent at other stations with relatively
high frequency observations. Recognition of the dominance of
the transport process led to a re-examination of previously
puzzling sulfate distributions in which no obvious correlation
with other meteorological variables was evident. Figure 9.6
represents a reconsideration of the patterns in Figure 9.4. If
streamlines are used to infer transport, the positive deviations
over the Carolinas on May 11 are evidently the result of
transport from the sulfate producing northeast states. On March
1 0 the mid-Atlantic states are experiencing higher than normal
sulfate concentrations as a result of winds passing over the
64
-------
Ohio Valley source region. The large negative deviations over
the mid-west mentioned in connection witn Figure 9.4 can now be
understood as arising in the northwest air flow coming from the
relatively sulfate-free north central states. Reinterpretation
of these distributions in terms of regional sulfate distribution
and wind circulation convincingly demonstrates the primary role
of long-range transport. This confirms and strengthens the
conclusions of Spirtas and Levin (1970).
The strength of the sulfate source/wind relationship can be
expected to depend on the general sulfate environment of a
station. Stations in a rather extensive area of more or less
uniform sulfate background might show less variablity with wind
direction than, say, St. Louis. Such a station might allow the
unmasking of relationships with other meteorological variables.
Some complications to any interpretation of sulfate
variability include such local effects as terrain and proximity
to water bodies. For example, Lyons (1977) showed that large
bodies of water can affect low-level atmospheric stability which
in turn produces unexpected pollution problems. Additional
problems can occur in analyzing sulfate pollution near the sea
since sea salt nuclei may contain sulfate.
Because large scale sulfate concentrations depend upon wind
direction, annual means, and interannual trends, inferences
about intrinsic sulfate variability based on such data (Frank,
1974) should be reevaluated. The present monitoring network
with 12 or so days between measurements may not be adequate to
remove through averaging preferential wind directions which in
turn may bias the derived annual mean. Furthermore, such means
may be expected to vary from year to year because of shifting
storm tracks and flow patterns associated with the planet's
general circulation.
The results of this study suggest topics for further
research. In some instances these new efforts may not be
practicable without an improved sulfate sampling network or
firmer knowledge of sulfate chemistry.
The direct measurement of sulfate concentration needs to be
performed as a separate sampling process, not merely as a
by-product of the total particulate evaluation. A separate
sampling method for sulfate would eliminate much of the
systematic error incurred by artifact production on the intake
filter in the presence of other pollutants. Furthermore, the
laboratory analysis would be simpler and more reliable if it
were not first necessary to separate sulfates from the mixture
of various substances. If and when this improvement becomes
available shorter sampling periods and more frequent
observations would be justified. Such data would help resolve
some of the weather/sulfate relationships suspected but not
65
-------
detected in this study.
Research into appropriate ways of filtering out the
influence of wind direction in annual means and variability is
needed if underlying variations due to source effluxes are to be
determined. This might be accomplished by taking into account
changes in annual wind roses and significant changes in regional
sulfate sources. These filtered data would aid the
Environmental Protection Agency in establishing meaningful
sulfate standards, as well as providing data that could be
subjected to analysis for relationships with other
meteorological variables.
66
-------
Figure 9.1
National Air Surveillance Network (NASN)
stations whose data are employed in the
present research.
67
-------
; i
X
u
•
» X
-------
abed
Figure 9.3
Regionally representative seasonal means
(1969-1974). Each bar in the graph
represents a mean seasonal value. Vertical
scales are in units of
69
-------
(A)
nation
aoacgra
•mutn
(3)
10 MASCH 69
figure 9.4 Patterns of synoptic weather (12GMT) and
sulfate concentration. Sulfate
concentrations are expressed as deviations
from the seasonal mean. Only deviations
greater than 1 ug/m^ are plotted.
70
-------
o
° (Sll) xn-i 4) CO
o 3-1-1 Q^ r;
P 03 C J-»
a M
5 > c •£ u
° O 3 CO
i ^ o a
* ' S o -a c
^ en AJ
W i— 1 0)
CO 03 O
u u o o
U dJ J= 0
C ,0 4J r-4
OJ S
O OJ Ui C
C O O -rH
O 0) M-l
U O W
I 03 C
CD >, — i
3 co > g
CO 1-5 co co
.
3%
2
3
00
71
-------
Figure 9.6 Patterns of synoptic weather (12 GMT) and
sulfate concentrations. Streamlines (arrows)
suggest sulfate transport. Synoptic stations
are the same as in Figure 9.4, but have been
omitted for figure clarity.
72
-------
SECTION 10
DIURNAL VARIATION OF WIND PROFILES ACROSS
MOUNTAINOUS TERRAIN DURING AN AIR STAGNATION PERIOD
Julius A. Jackson, W.J. Saucier, and W.D. Bach
Abstract
The diurnal variation of wind profiles across mountainous
terrain during an air stagnation period was evaluated for seven
days in the summer of 1957. The study was conducted across the
north-central Appalachian Mountains, an area of heavy pollution
concentration. The study was divided into easterly (16-18
July, 1957) and westerly (19-22 July, 1957) flow across the
mountains.
Examination over the seven days showed a diurnal variation
in boundary layer winds on the eastern side of the mountain
range with a maximum amplitude of about 3 to 4 m sec-1 at
1000-1500 m MSL in both the easterly and westerly flows. On the
western side of the mountain range, a diurnal variation with a
maximum amplitude of about 4 m sec-1 at 600-1100 m MSL occurred
in both flows.
This oscillation in the lower levels showed the presence of
a low-level jet, which was unexpected in that this study was
conducted during an air stagnation period. The low-level jet in
the easterly flow across the mountains reaches a maximum wind
speed at approximately 0600 GMT at about 300 m above ground
level. In the westerly flow, the low-level jet occurs at
approximately 1200 GMT at 600-800 m above the ground. This
low-level jet is due to an inertial-type oscillation driven by
the diurnal varition of the frictional forces aided by thermal
forcing.
73
-------
INTRODUCTION
Stagnating anticyclones are often associated with incidents
of heavy air pollution in urban areas. These anticyclones
usually linger over an area for a protracted period (four days
or more). During this period, surface wind speeds may be very
low, and visibility and vertical mixing are often restricted.
Thus, the circulation is often thought to be insufficient to
disperse the accumulated pollutants of the atmosphere. The
resulting accumulation may cause distressful and possibly
hazardous conditions for inhabitants of the area.
The stagnating anticyclones that produce the major air
pollution episodes are usually found in the eastern United
States (Korshover, 1975), and are most likely to occur in late
summer and autumn.
The northeast United States is very susceptible to air
pollution episodes due to the heavy concentration of industry
and population centers. Figure 10.1 shows the annual S0£
emission density in the northeast and Ohio Valley areas. These
pollution episodes not only affect the urban areas, but the
non-urban areas are affected due to dispersion and transport of
pollutants.
The purpose of this study is to investigate the effects of
the mountains on the wind profile in the boundary layer during
an air stagnation period, as well as the effects of boundary
layer winds on air pollution dispersion and transport in a
stagnation period over mountainous terrain. Various studies
have been completed on diurnal wind variation in the boundary
layer, but none (as could be found) during air stagnation
period.
Bonner and Paegle (1970) found that boundary layer winds
oscillate diurnally, reaching a maximum speed at night and a
minimum during the day at elevtions between 50 and 2000 m above
the ground. The wind variation is especially pronounced with
southerly flow over the central plains to the east of the Rocky
Mountains. The amplitude of this oscillation is 2 to 3 m sec-1
at levels of 0.5 and 2.0 km above the ground (Hering and Borden,
1962).
Boundary layer wind oscillations may arise from periodic
variation in the horizontal pressure gradient force as in
74
-------
mountain wind circulation, or they may be driven by day-to-night
variation in the frictional force. A number of theories have
been put forward to explain this diurnal wind oscillation which
is often associated with the low-level jet. Blackadar (1957),
argues that the jet results from a free inertial oscillation
superimposed on the geostrophic basic flow, initiated by the
rapid nocturnal breakdown of the frictional restraining force in
the boundary layer. As the eddy diffusivity decreases during
the onset of stable stratification in the evening hours, the
motion in the upper part of the boundary layer becomes uncoupled
from the frictional force and undergoes an inertial clockwise
oscillation about the gradient wind vector. For the latitude
of Washington, D.C., a maximum at lowest levels is most likely 9
hours after the oscillation begins. Attempts (Blackadar and
Buajitti, 1957) to simulate this behavior indicated that both
the mean value of the eddy viscosity and the amplitude of its
diurnal variation must decrease rapidly with height. Numerical
experiments with constant geostrophic wind and diurnally varying
eddy viscosity duplicate reasonably well the observed
oscillation over the central plains.
The present study attempts to describe the diurnal
variation of wind profiles across the north-central
Appalachian Mountains, and to find what effects the mountains
have on the boundary layer wind profile. The six stations used
in this study, which lie along an axis approximately
perpendicular to the mountains, are listed in Table 10.1 and
located in Figure 10.2.
RESULTS AND CONCLUSIONS
Wind profiles across the Appalachian Mountains and
available at six-hourly intervals showed the presence of a
low-level jet during two air stagnation periods. These periods
included July 19-22, 1957 when the flow was generally westerly,
and July 16-18, 1957 when easterly flow prevailed. A
representative 12 GMT surface analysis for each of the two
periods is shown in Figure 10.3. Both periods are dominated by
weak high pressure systems. Diurnal variability in profile
shape is significant, but typicially the wind maximum occurred
at 0.5 to 1 km above ground. Wind speeds decreased rapidly
above this level and then remained about constant to 3 km. Some
of these features can be noted in the average wind profiles for
two stations presented in Figures 10.4 and 10.5.
Diurnal oscillations about the daily mean in the zonal (U)
and (V) wind components, at the same stations are portrayed in
Figures 10.6 and 10.7. Amplitudes at jet level range from 1.5
to 3.0 m sec-1.
75
-------
The analysis of the flow over the terrain has shown that
the low-level jet is at different altitudes depending on the
wind direction across the mountains. The time of the day is
also a factor in the occurrence of the low-level jet. The
easterly maximum low-level wind flow occurs at 0600 GMT (about 6
hours after sunset), and in the westerly flow it occurs at 1200
GMT (about 2 hours after sunrise) (Figures 10.8 and 10.9).
Inertial oscillations above the terrain as revealed by
hodographs (for example, Figure 10.10) indicate that below the
mountain top, on the windward side, the mountains tend partially
to divert the component of the mean motion normal to the
mountains to that parallel to the mountains. It also aids in
explaining the wind maximum in the low-levels as an
inertial-type oscillation driven by the diurnal variation of the
frictional force aided by thermal effects.
Extensive analysis of the diurnal winds across the north
central Appalachian Mountains has shown that a low-level jet
exists across mountainous terrain during air stagnation periods.
This jet is similar to that over the Great Plains, but is
reduced in magnitude. Nevertheless, the low-level wind maximum
described in this study can be effective in transporting and
dispersing pollutants across the Appalachians.
76
-------
j
en
o
LJ
s
o
cn
c
o
O O
03
C
^
o <
•rH 3U
JT Cx3
flu
CO V4
U
-------
Table 10.1 Station identifiers and elevations
Station
Identifier
Elevations above
MSL(m)
Washington, D. C.
Harrisburg, Pa.
Pittsburgh, Pa.
Akron, Oh.
Toledo, Oh.
Flint, Mich.
DCA
HAR
PIT
AKN
TOL
FNT
20
107
373
377
211
233
Figure 10.2 Area and stations analyzed.
78
-------
c
o
•f—
•(-» T3
ta c
c »
c •
C 'T- ig
•i- S o;
s. s_
3 >> (O
^T >,
(/> , ra CO
i— a>
en
«•<-!.
S. r- 0)
3 >
U» O> O
VI C
,
a. 3 —
•M 3
i— 03 O
0) OJ
> i+- O
(U OJ
r- r-. — •
un
> 3
^ VO Wl
wl •— OJ
OJ 2
t. -O
a. o +-»
aj •»- .c
OS 1- CT>
OJ -r-
Q.I—
s_
3
o>
79
-------
WIND PROFILES
UJ
a
6 8 10
WIND SPEED,M/S
Figure 10.4 Average wind speed profile for westerly
flow at Washington, D.C.
tu
a
3.0
2.5
2.0
1.5
1.0
0.5
0
WIND PROFILES
* 6 8 10
WIND SPEED,M/S
Figure 10.5 Average wind speed profile for westerly
flow at Akron, Ohio.
80
-------
TJ COMPONENT
S
-X
VJ
H-
r
o
3.0
2.5
-0.7
-0.2
/^2 1.3 !\-0.
1.1
_. ^^^^ v ^^.r^1**^,' \
.3 \\l-3 \-0.2 \\1.6 /
0.
0.3 -0.3 -0.2
V COMPONENT
£
w
>—
r
o
Figure
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.8
o.3
0.2
-0.4
-1.1 ^b;8
00 06 12
TIME CGMT5
13 00
10.6 Time variation of Che deviation in
westerly wind from its daily mean at
Washington, D.C. Values are in meters
per second; times are in GMT.
81
-------
TJ COMPONENT
o
IU
x
0.1 -0.4 \ 2.7 { K-2.2; | -0
V COMPONENT
s
_«:
s^
H-
r
o
o.o
Figure 10.7 Time variation of the deviation in
westerly wind from its daily mean at
Akron, Ohio. Values are in meters per
second; times are in GMT.
82
-------
o
r—I
4H
^
r-4
U
(U
4J
CO •
«J C
o>M
O
O 03
>O 0>
O O
S-J
<4-l CJ
O «8
•^ C
•U O
o c
01 O
co a.
ca O
CO U
o
CO
O
(U
u
3
00
o
•
o
83
-------
1 1 I I I 1
o
I—I
14-1
>
I—I
-------
EASTERLY (0.5km)
V 095
4 m/s
WESTERLY(0.5 km)
00
V 235
4 m/s
Figure 10.10 Hodographs of the wind variation at
DCA in the easterly and westerly flow.
Heights are MSL; times are GMT.
85
-------
SECTION 11
A MESOSCALE ANALYSIS OF AIR FLOW
IN COMPLEX TERRAIN
Christofer Maxwell, W.J. Saucier, T. L. Tsui and G.C. Holzworth
Abstract
Analyses of wind data are presented which examine the
coupling of the surface and 850mb wind directions in a region of
complex terrain. The data were grouped based on periods when
the 850 mb wind directions were parallel or perpendicular to
the major valley axis. The diurnal and pressure gradient
influences were also examined.
The analyses show that the surface and 850mb wind
directions agree best when the 850mb winds are about parallel to
the valley axis. The agreements are best during daytime periods
when the 850mb wind speeds are strong. In addition, wind
measurement sites in valleys agreed better with the 850mb winds
than did sites located on ridges. When the 850mb winds were
perpendicular to the valley axis, the wind measurement sites on
ridges tended to agree better with the 850mb winds than did the
valley sites.
Further analyses show that 30m tower wind sensors can be
discriminately used to predict the direction of power plant
plume travel in complex terrain regions. The tower sensors are
unreliable during nighttime hours and are best during daytime
hours when the 850mb winds are parallel to the valley axis.
86
-------
INTRODUCTION
The structure of wind fields over flat terrain is well
understood. Both the statistical and turbulent quantities have
been examined and are fairly well expressed mathematically.
Recently, however, attention has been shifted toward learning
more about air flow in complex terrain. The reasons for the
shift are varied but strong influence comes from atmospheric
pollution control agencies as well as energy and defense
concerns.
Very few locations in the United States conform to the
terrain and source characteristics of the places where empirical
determinations were made of the atmospheric dispersion
parameters. The terrain in areas where there is an interest in
calculating the pollution potential will exhibit varying degrees
of complexity, either a step change in roughness, variations
in temperature fields, or actual terrain variations. Early
studies of air flow in irregular terrain were limited (Defant,
1951; Forchtgott, 1949; Gleeson, 1951), and even at present
relatively few investigations are being conducted to study the
flow characteristics in complex terrain. Consequently, there is
a strong need to learn more about the basic influences of
terrain on the winds.
A simplistic description of the diurnally varying
circulations resulting from the variably-sloped terrain within a
valley was presented by Defant (1951) (Figure 11.1). Using a
symmetrical valley with an axis running east to west, Defant
showed the diurnal changes in winds due only to heating by the
sun passing directly overhead and due to the subsequent
radiative cooling at night.
Defant's description explains the basic principle of the
diurnal pattern of slope and valley winds but considers an ideal
valley. Other researchers expand and add to Defant1s
description of mountain/valley winds. Gleeson (1951) developed
Defant1s description further by explaining the single cellular
pattern that would result when the sun warms one valley slope
and not the other. Instead of air sinking down the middle of
87
-------
the valley, the air tends to sink along the. slope that is not in
sunlight. Forchtgott (1949) and Musaelyan (1964) discussed
vortices produced in the lee of ridges due to winds blowing
perpendicular to the ridges. Forchtgott noted that vertical and
horizontal extent of the lee side vortices increased with
increasing wind speeds. More recently, Nappo (1977) presented
analyses of the three-dimensional wind fields over a large area
of eastern Tennessee mountains. In particular, he compared the
vertical wind field structure of complex terrain to uniform
terrain.
The analyses to be presented here examine the statistical
relationships between the surface and upper level winds within a
mountainous or complex terrain region. The terrain influence
will be measured by examining wind direction differences between
the surface and upper levels during periods when the upper level
winds are perpendicular and parallel to the valley or ridge
axis. In addition, the turning of the near-surface winds to
those at 850 millibars (mb) will be examined.
The data used in the analyses are from the Power Plant
Stack Plumes in Complex Terrain study conducted by GEOMET, Inc.
for the U.S.Environmental Protection Agency. The purpose of
the study was to collect a reliable data base for meteorological
studies in complex terrain and to examine the effects of complex
terrain on the dispersion of power plant plumes. The study was
conducted in the mountains of southwest Virginia along the
Clinch River in the vicinity of Carbo, Virginia.
The terrain within the study area is highly complex.
Elevations vary from 1500 feet MSL to better than 3500 feet MSL.
The major valley axes extend approximately from southwest to
northeast. Most of the study area is covered by forest.
The power plant used for the study is the Clinch River
Steam Plant. The plant has a maximum generating capacity of 712
megawatts (MW). The daily-average minimum output is 491 MW
between 3 and 4 a.m.. The average daily maximum output is 661
MW between 11 a.m. and 12 noon. The plant burns coal, releases
the exhaust through two 453 foot stacks, and is considered a
base-load plant.
The Power Plant Stack Plumes in Complex Terrain study
spanned a sixteen month period, June 1976 through September
1977, during which meteorological and pollutant data were
collected at eight fixed stations. For brevity the names of the
fixed stations will be referred to as:
88
-------
Castlewood Castle (8 km WSW)
Tower Tower ( 3 km NE)
Nash's Ford Nash (11 km ENE)
Lambert Lambert ( 8 km E)
Munsey Munsey ( 4 km SE)
Hockey Hockey ( 6 km SE)
Johnson Johnson ( 6 km S)
The approximate distance and direction of the seven stations
from the Clinch River Plant are noted.
The data collected at each fixed station are listed in
Table 1. All data are stored as hourly average values, with the
exception of the hourly peak values of the pollutants, on
magnetic tape.
Upper-level data consist of 850 and 700 mb winds from three
stations surrounding southwestern Virginia; namely, Greensboro,
NC; Huntington, WV; and Nashville, TN. The winds were
decomposed into east-west (U) and north-south (V) components,
and distance-weighted values then interpolated to the study
area. Winds were considered parallel to the valley axis if
within a 45° sector of the mean valley bearing (60°-240°).
Perpendicular winds were those within a 45° sector centered on
1 50* or 330°. Wind direction at 850 and/or 700 mb fell within
these four sectors 64.6% of the time included in the study
period; NW perpendicular and SW parallel winds accounted for
60.1% of all winds directions. Directional sectors for surface
winds were 60° wide in order to allow for some terrain
deflection and frictional effects.
Results presented in the next section were obtained during
the one-year period October 1976 through September 1977.
RESULTS AND CONCLUSIONS
The description of the diurnal cycle of slope and valley
winds by Defant is simplified in using a symmetric valley. The
orientation of a valley as well as the angle of its slopes and
the complexity of its surface will complicate the slope and
valley flows. However, the basic physics of Defant's descrip-
tion is good, as can be seen in the day and night wind roses for
a valley (Nash) site from the Power Plant Stack Plumes in
Complex Terrain study (Figure 11.2).The Nash site islocated in
a valley with higher elevations to the northeast (Figure 11.3).
The wind roses show the tendency of the wind directions to re-
verse from down-valley at night (Figure 11.2a) to up-valley
during the day (Figure 11.2b). In addition, the up-valley winds
tend to be faster than the down-valley winds. This may simply
reflect the fact that daytime winds are generally faster than
89
-------
nighttime winds from all directions, as appears to be the case
here.
The surface winds in complex terrain regions become further
complicated by the channelling effects of the terrain (Start,
1975; Egami, 1974; Nappo, 1977). Channelling effects can be
observed readily at the ridge sites of the Power Plant Stack
Plumes in Complex Terrain study. Figure 11.4 depictsthe night
and day wind roses for the Hockey site. The terrain surrounding
the Hockey site is shown in Figure 11.5. The wind roses show
the preference of winds to come up through the hollows on the
side of the ridge and, in particular, up through the wide hollow
to the south-southwest of the site. This is observed both day
and night. In addition, the speeds of the winds tend to be
faster when the winds blow through the hollows.
The analysis schemes used to examine the interaction of the
surface and upper-level winds during periods when the upper-
level winds were parallel and perpendicular to the valley axis
worked well. In general, the terrain influence appeared more
significant when the upper-level winds were perpendicular to the
valley as opposed to parallel to the valley. Tables 11.2 and
11.3 present many of the relationships described further below.
In the latter table poor (< 25%) and good (> 60%) agreement
refer to the percentage of the 12-hour periods centered on the
two radiosonde times that the hourly surface winds were in
agreement with the sector direction of the upper-level winds.
In this way something of the short-term persistence of the
surface wind is introduced as another factor in categorizing
results.
Examinations of the periods when the upper-level winds were
parallel to the valley showed fair to good correlation of wind
directions between valley stations and the upper levels. The
valley station wind directions were from the same directional
sectors as the upper level winds about 50% of the time. The
agreement was found to be stronger during the day than at night
suggesting a stronger thermal influence on the local winds
during the night than during the day. Stations located in broad
valleys, on the side of valley slopes, or on ridges had
significantly lower percentages of agreement. A 30m tower
located in a valley had wind directions within ±45° of the 850mb
wind directions 85% of the time during the day. At night the
agreement dropped to 76.2% of the time. The tower located on
the ridge had lower agreements: 80.0% during the day and 58.3%
at night. Intuitively, the magnitude of the agreements
decreased when considering sectors 60" ( A9 = ±30°) wide or when
veering cases alone were considered ( A9 = +45°, +30°). Faster
wind speeds at 850mb tended to improve the agreement in the wind
directions between the surface and upper levels during the 1200
GMT radiosonde release periods, but were not as strong a factor
90
-------
during the 0000 GMT radiosonde release periods.
In general, the agreement of wind directions between the
surface and upper levels was lower when the upper level winds
were perpendicular to the valley axis instead of parallel to the
valley axis. During the perpendicular flow periods, surface
stations located on ridges had better agreement of wind
directions with the upper levels than the valley sites. Again
the agreements were better during the day than at night. The
.wind directions at the tower located on the ridge were within
±45° of the upper level wind directions 78.9% of the time during
the 0000 GMT radiosonde releases and only 33.3% of the 1200 GMT
radiosonde releases. The valley site tower had corresponding
values of 71.4% during the 0000 GMT releases and only 9.9% of
the the 1200 GMT releases. Similar to the parallel flow
periods, faster 850mb wind speeds improved the agreement between
the 850mb wind directions and the 30m level wind directions at
Hockey and Tower during the 1200 GMT radiosonde release times,
but did not appear to be a strong factor at the times of the
0000 GMT radiosonde releases.
The results of the analyses give insight into the extent
of the influence of complex terrain on surface winds. Slope and
valley winds apparently account for a large portion of the
observed surface winds. Typical valley winds are commonly
observed in narrow valley locations and are less frequent in
broad valleys. The vertical extent of valley winds is fairly
deep. Defant (1951) suggested the valley winds extend to the
height of the flanking ridges. During periods when the upper
level winds are parallel to the valley, however, the wind
directions may be uniform up to the 850mb level. Slope winds
affect locations on valley slopes and ridges. The vertical
extent of slope winds on the ridges appears small with some data
suggesting about 30m.
In the analysis that compared the difference in wind
direction between the 850mb wind and the corresponding hourly-
average 30m level winds at Hockey and Tower (Table 11.4), the
seperation of the results into the two radiosonde release times
( 0000 GMT and 1200 GMT) was significant. The variation between
the release times is attributed to the day to night differences
of the Tower and Hockey winds. Since most of the release
periods studied occurred during the winter and spring, this
assumption appears appropriate. The 0000 GMT (0700 LST)
radiosonde release which was compared to the 0700-0800 hourly
average winds at Hockey and Tower is believed to represent
nighttime conditions. Although sunrise usually occurs by
0700 LST, the solar elevation is not large enough to affect the
flow in the valleys due to the shading by the surrounding
ridges; thus, the nocturnal valley wind circulations continue
considerably past the time of sunrise. The length of the delay
until the sun will heat the valley surface depends on the
91
-------
orientation of the valley, the height of the surrounding ridges,
and the season. The argument to use the 0000 GMT (1900 1ST)
radiosonde release is weak for the late fall, winter, and early
spring months. During the other months, sunset occurs near to
or later than 1900 LST; therefore, the daytime valley wind
circulations are still present and the 0000 GMT radiosonde
release represents the daytime conditions. During the late
fall, winter, and early spring, the sun sets considerably before
1900 LST. Therefore, the nocturnal valley wind structure is
developing; however, the up-valley wind from the daytime may
still be present and therefore be partially representative of
the daytime wind field structure. Consequently, the 0000 GMT
radiosonde release generally represents the daytime cases and
the 1200 GMT radiosonde release represents the nighttime cases.
The appropriateness of using the 30m level wind directions
at the Hockey or Tower sites to predict the direction of plume
travel will depend on the time of day and the direction of the
850mb winds. Provided the plume direction of dispersion is
needed more in a qualitative rather than a quantitative sense,
the Tower and Hockey 30m winds can be used in some cases. The
Hockey and Tower winds are inappropriate at night when the 850mb
winds are perpendicular to the valley axis. During the day,
with the same 850mb winds, the Hockey 30m level winds will give
a fair to good indication of plume travel. During all periods
when the 850mb winds are parallel to the valley, the 30m level
winds at Tower will give a good indication of plume travel. The
daytime winds, however, will be more representative than the
nighttime winds. In addition, the faster the 850mb winds the
better the Tower and Hockey winds will represent the direction
of plume dispersion.
Numerous studies of the Power Plant Stack Plumes In
Complex Terrain data should be undertaken to determine the basic
structure of wind fields in complex terrain. Other
studies could readily include the correlation of winds with
ground level pollutant concentrations, the temperature field in
complex terrain, the distribution of stability categories, and
the change in wind speeds with height within the confines of the
valley. In all, the data are of good quality and can be used to
identify many basic influences of terrain on the wind field.
92
-------
d.
Figure 11.1 Variations in the Windfield
Within a Symmetrical Valley Due to
the Uneven Heating of the Variably
Sloped Terrain (Defant, 1951).
93
-------
a. Night
(0200-0400)
10. &
0-5 -(60.3E- °'9
b.O
.5
.9
2.3
3.2
15.1
-'. 1-3 3-5 5-10 >iO
Wiad S?*«d (s/s)
19.7
5.3
7.6
b. Day
(1400-1600)
Figure 11.2 Night and Day Wind Roses
at a Valley (Nash) Site.
1 u
-t Contours: 100's at t»
Figure 11.3 Terrain Surrounding the
Nash Site.
94
-------
a. NighC (0200-0400)
U.6
11.3
19.3
O.J-l l-l 3-5 5-13 >10
'•'lad Sp««d (*/»)
16.6
b. Day
(1400-1600)
Figure 11.4 Night and Day Wind Roses,
ootaur*: 100's at '*«t
Figure 11.5 Terrain Surrounding Che Hockey Site.
95
------- An error occurred while trying to OCR this image.
-------
c
CU
o
c
1-1
o
CJ
09
T3
C
3 aa
CU C
•2s
VJP^
3 CU
w >
CU
o a.
c a.
CU 3
3
O*«J
01 jS
H3
CN
CU
^
J3
H
o
(0
on
CU
o w c
rft /l% /11
fv \U **J
-P at?
c cv^
CU 3 0
u c
U TJ -H
»inp*vo
r*^ r^ vo c^ ^^ ^^ f^ ^D
O'^'HOCNO-^'-^
sO"~*cNc^oo•••••
in *~^ o *o o o^ CN ^
csininfTim«-4in'^*
or-aoin^'^'ino
\o oo c^ co ^^ f^ c^ c^
f*icNCN^rcN^r^ifi
CU CU
*o ro
•H -H
cn cn
>,>,>,>>,>, >,
cucucucucucucucu
rH *™H *™4 fH • i"H O^ »H C7*
^i— |i— 4rH^fOiH'O
(0 ID >>.>>CS>Q5
S o
O f*i
CU > U >i >>
--( U CU CU : cu^i
cnacncEUSU
(00n33(0000
CJS^ZSJSc' —
97
-------
00
C
3 Z
^*N »Cj *
9< CO T3
O CO 13 q
vo q o co
/\ O •*•*"
^i fll di
"O O £•* rH
O CU r-l
O U co 03
Q 05 03
T3 rS CX
Ct3 «H
co q co 35
-4 qcn
__ r^ .-4
/•^ »£ ^«
$4 CU
in r-i M JTJ
CN cu co u
V > r-l
>-' CU 3 <4M
i-J O O *-\
V4 -^ •
O Vl t> CO CO
CU CU 0) 00 O
a. a. co -H
q 33 !-i 4J )H
1-4 cu q cu
•o cu cu au
co q u
C 03 3 U 05
O Z CU -4
i-l CU O. CO
JJ CJ "O ^i
CO CO q 05 r-4
•^4 4-4 CO CO CO
u i-i q
CO 3 r-l -O CO
> en cu cu
i— 1 4-) t-l
•u (4-1 1— i q co
JS O CO QJ ,-4
00 W co 3
•r4 4-1 CQ CU O
z q cu )-( "-i
cu a. -a
053: q
u cu en co cu
cu 4J a.
CO 00 JS Q CU
a <; 4J *— ' a.
«
^
cu
r-4
CO
5-1
U
(0
.H
3
U
tJ
CU
o
a
2
Z
r-4
r^
iH
(0
s_,
(0
0*
3:
en
'O
0
0
o
u
o
0
a*
T3
0
O
O
u
0
0
CU
JJ
Crr
W
e M z
CU O
CU vO
Oi- >|
< (Q
d
! [
4J JC
Cn
W
CU « -H
S * Z
cu in
cu CN
W^
< fll
Q
jJ
JJ JZ
q o
CU —^ -i-t
e »« z
CU 0
CU ^O
W A
< "~* «
C
JJ
JJ ff.
q o
CU — .-H
S X Z
cu m
cu CN
u v
< fl
o
—
\J
CQ
•*
0 -r4
•H U
JJ O
fl U!
JJ
CN »-4
c*- m
e *
<^ rn
f*"4 ^*^
>1 >f
0) 0)
<-* p-4
^^ ^™1
fl fl
JJ
u
cu
ff. n
•fl B
fl fl
Z *4
•*
vO
0
O
0>
^r
rH
0>
ro
0
•
CO
r-i
O
vO
CN
O
TJ"
rH
O
O
>,
CU
rH
^H
fl
U
CU
3
0
£-1
vO
0
1-^
en
T
ro
CN
V
ro
CN
oo
•
ro
CJv
^4
f^
•
CN
ro
m
vO
ro
• CU
a»
^3
an
>r
CU
a
o
X
ro
•q*
ro
ro
,_,
CN
U1
m
CN
O
•
^*
rH
O
vO
CN
O
vO
,_«
O
o
rH
>,
CU
r-4
fl
S
0
ro
U
a
3
0
e-
vO
o
rH
O
a
•*
ro
CN
0
s:
^*
>
in
rH
rH
ro
CN
rH
•
ro
CN
CN
Tj
•H
O
rO
^
OJ
U
o
•***
98
-------
3
00 CO
G
CO T-4 CO
l-i WjS
CO 3 4J
30
C
CO T3 -^
C CX
O i4 4->
• »^ i
i-l3i-l
u 3
O J3
co e oj
W O CO
•** m co •
Q oo co co
CO "O
•O CO«-4 O
C J3 co -^
i-i -u pi V4
3 % o -o
CO c 3
J< ^N O O
CJ -^*J
TJ Vj
"O CO CO CO
C CO OS i-4
cd co 3
V4 H O
H 002 1-4
CO CO O T3
3 Q C
O O CO
H T3 O Ou
CO CN U
0) U i-4 CO
S co 0.
*^J • \ *v«
" U ^
E-t •«•* C 3
"O to Z
U-i C
O •-" H *O
S C
0) CO O CO
an r;
CO 4-1 O i—l
U o CO
C C O p-i
CO -^4 O f-l
0 j£ CO
i-J 4-J CO W
CO i-4 .C ro m
0 mo
in • •
•«* vO CN
+ in r»
0 CN O
O
en in •v
+ 1 >O T
O 00
in • •
•v r- o
-h| 00 CO
II
CD
<3
^
U 0)
0) X
3 CJ
0 0
e- E
<-»
0)
r-t
.-t
f^
(TJ
U
fl
a.
3
cn
O *4 O
o
<^> O\ CK
1 f-t
0 ^00
in • •
T ej\ en
1 CN
0 ^oo
O
en ej\ en
+| CN
0 -4 en
in • •
T ,
u w a>
•• m :
r^ 3 0
300
0 £t E
•H
T3
C
MI
1)
a
w
0)
a<
2
2
99
-------
REFERENCES
Bhumralkar, C.M., 1975: Numerical experiments on the
computation of ground surface temperature in an atmospheric
general circulation model. J. Appl. Meteor., 14, 246-1258.
Blackadar, A.K., 1957: Boundary layer wind maxima and their
significance for the growth of nocturnal inversions. Bull.
Amer. Meteor. Soc., 38, 283-290,.
and K. Buajitti, 1957: Theoretical studies of diurnal
wind variation in the planetary boundary layer. Quar.
Jour. Royal Meteor. Soc., 83, 486-500.
Bonner, W.D. and J. Paegle, 1970: Diurnal variation in boundary
layer winds over the south-central United States in summer.
Mon. Wea., Rev.. J.9, 199-205.
Bowling, S.A., Benson, C.S., 1978: Study of the sub-arctic heat
island at Fairbanks, Alaska. U.S. Environmental Protection
Agency, Report Number EPA-600/4-78-027.
Bross, Irwin D. J., 1958: How to use ridit analysis.
Biometrics, 14, 18-38.
Carlson, T.N. and F.E. Boland, 1978: Analysis of urban-rural
canopy using a surface heat flux/temperature model. J. Appl.
Meteor., _17, 998-1013.
Clarke, J.F., 1969: Nocturnal urban boundary layer over
Cincinnati, Ohio. Mon. Wea. Rev., 97, 582-589.
Deardorff, J.W., 1978: Efficient prediction of ground surface
temperature and moisture with inclusion of a layer of vegeta-
tion. J. Geophys. Res., 83, 1889-1903.
Defant, F., 1951: "Local Winds." Compendium of Meteorology.
American Meteorological Society, Boston, Mass.
Duckworth, F.S. and Sandburg, J.S., 1954: The effect of cities
upon horizontal and vertical temperature gradients. Bull.
Amer. Meteor. Soc., 35, 198-207.
100
-------
Egami, R.T.,: 1974 Diffusion study in the vicinity of Mahave
Generating Plant. Symposium on Atmospheric Diffusion and Air
Pollution. American Meteorological Society, Boston, Mass.
Faulkenberry, D.G. and C. D. Craig, 1976: Visibility trends in
the Willamette Valley, 1950-71. Third Symposium on
Atmospheric Turbulence, Diffusion, and Air Quality, 547-552.
Forchtgott, J., 1949: Wave streaming in the lee of mountain
ridges. Bull. Met. Czech. , _3_, Prague.
Frank, N.H., 1974: Temporal and spatial relationships of
sulfates, total suspended particulates and sulfur dioxide.
Presented before the Annual Meeting of the Air Pollution
Control Association, Denver, CO.
GEOMET, Inc.. Monthly Technical Progress Report, No. 34,
October 1971TTGEOMET Report Number E-746.
Gleeson, T.A., 1951: On the theory of cross valley winds
arising from differential heating of the slopes. J. Met., {J,
398.
Hage, K.E., 1977: Research in urban climatology of the
University of Alberta. McGill University, Montreal, Dept of
Geography. Climatological Bulletin Number 22, 25-29.
Hall, P., Jr., C. E. Duchon, L. G. Lee, and R. R. Hagan, 1973:
Long range transport of air pollution: A case study, August
1970. Mon. Wea. Rev.. 101 , 404-411".
Heffter, J.L., A. D. Taylor, and G. J. Ferber, 1975: A
regional-continental scale transport, diffusion and
deposition model. NOAA TM ERL ARL-50, 29 pp., Silver
Springs, Maryland, 1-16.
Hering, W.S. and R. Borden, 1962: Diurnal variations in the
summer wind field over the central United States. J. Atmos.
S_ci., 19, 81-86.
Hester, N.E., R. B. Evans, F. G. Johnson, and E. L. Martinez,
1977: Airborne measurements of primary and secondary
pollutant concentrations in the St. Louis urban plume.
International Conference on Photochemical Oxidant and Its
Control, Proceedings.Volume I, EPA-600/3-77-001a, p. 259,
U.S. EPA, Research Triangle Park, N.C., January 1977.
Holton, J.R., 1967: Diurnal boundary layer wind oscillation
above sloping terrain. Tellus, 19, 199-205.
101
-------
Holzworth, G. C., 1959: A note on visibility at Sacramento,
California. Mon. Wea. Rev., 87, 148-152.
, 1962: Some effects of air pollution on
visibility in and near cities. SEC Technical Report No.
A62-5, 69-88.
and J. A. Maga, 1960: A method for analyzing the
trend in visibility. J. Air Poll. Con. Assoc., 10, 430-435.
Howard, L., 1883: Climate of London Deduced from Meteorological
Observations. 3rd Edition. London, Harvey and Darton.
Husar, R.B., N.V. Gillani, J.D. Husar, C. C. Paley, and P. N.
Turcu, 1976: Long range transport of pollutants observed
through visibility contour maps, weather maps and trajectory
analysis. Third Symposium on Atmospheric Turbulence,
Diffusion, and Air Quality, 344-347.
Idso, S.B., J. K. Aase, and R.D. Jackson, 1975: Net radiation-
soil heat flux relations as influenced by soil water content
variations. Bound. Layer Meteor., 2, 113-122.
Koch, R.C., 1978: Power Plant Stack Plumes in Complex Terrain-
Description of an Aerometric Field Study, Final Draft..
GEOMET, Inc., Gaithisburg, Maryland.
Korshover, J., 1975: Climatology of stagnating anticyclones
east of the Rocky Mountains, 1936-1975. NOAA Tech. Mem. ERL
ARL-55, U. S. Department of Commerce, 27 pp.
Landsberg, H.E. , 1968: Micrometeorological temperature
differen-tiation through urbanization. From the proceedings
of tne Symposium on Urban Climates and Building Climatology,
Brussels, Oct. 1968 (Volume 1).
Lettau, H.H., 1967: Small to large-scale features of boundary
layer structure over mountain slopes. Proceedings of the
Symposium on Mountain Meteorology, E. R~Reiter and J. LT
Rasmussen,eds., Atmosphere Science Paper No. 122, Department
Atmosphere Science, Colorado State University, Ft. Collins,
Colorado, pp. 1-74.
Lowry W. P., 1972: Urban effects on the atmosphere—Who in the
world cares? Preprint volume of the Conference on Urban
Environment and Second Conference on Biometeorology, A.M.S.,
Philadelphia, Pa., Oct. 31 - Nov. 2, 1972.
Ludwig, F.L., 1968: Urban temperature fields. From the
proceedings of the Symposium on Urban Climates and Building
Climatology, Brussels, Oct., 1968 (Volume I).
102
-------
, 1979a: Vertical profiles of photochemical pollutants in
CtTe" vicinity of several urban areas. Proceedings Fourth
Symposium on Turbulence, Diffusion and Air Pollution,Reno,
Nevada,15-18 January 1979 (American Meteorological Society,
Boston, Massachusetts).
, 1979b: Assessment of vertical distributions of
photochemical pollutants and meteorological variables in the
vicinity of urban areas. Draft final report for SRI Project
No. 6869, January 1979.
Lyall, I.T., 1977: London heat island in June and July 1976.
Weather, J32, 296-302.
Lyons, W. A., 1977: Mesoscale air pollution transport in
southeast Wisconsin. Report No. EPA-600/4-77-010,
Environmental Sciences Research Laboratory, Research Triangle
Park, NC.
Martin, F. P. and Powell, G. L., 1977: The urban heat island in
Akron, Ohio. Proceedings of the Conference on Metropolitan
Physical Environment, 94-97.
Miller, M. E., T. L. Canfield, T. A. Ritter, and C. R. Weaver,
1972: Visibility changes in Ohio, Kentucky, and Tennessee
from 1962 to 1969. Mon. Wea. Rev., 100, 67-71.
Musaelyan, S. A., 1964: Barrier waves in the atmosphere.
•Isreal Program of Scientific Translations, Jerusalem.
Nappo, C.J. 1977. Mesoscale flow over complex-terrain during
the eastern Tennessee Trajectory Experiment (ETTEX).
J. Appl. Meteor.. 16, 1186-1196.
Nickerson, E.G. and V.E. Smiley, 1976: Surface layer and energy
budget parameterization for mesoscale models. J. Appl.
Meteor., J.4, 297-300.
Oke, T.R., 1977: Research in urban climatology at the
University of British Columbia. McGill University, Montreal,
Canada, Dept. of Geography, Climatological Bulletin, No. 22,
Oct. 1977, 1-10.
and G. B. Maxwell, 1975: Urban heat island dynamics
in Montreal and Vancouver. Atmos. Environ., 2> 191-200.
Renou, E., 1862: Differences de temperature entre Paris et
Coisyle-498. Societe Meteorologique de France, Annuaire,
1_0, 105-109.
Shreffler, J.H., 1979: Heat island convergence in St. Louis
during calm periods. J. Appl. Meteor., 18, 1512-1520.
103
-------
Spirtas. R. and H. J. Levin, 1970: Characteristics of
Particulate Patterns 1957-1966. U.S. Department of Health,
Education, and Welfare, Public Health Service, Raleigh, NC.
Start, G.E., 1975: Diffusion in a canyon within complex
terrain. J. Appl. Meteor., 14, 333-346.
Summers, P.W., 1965: An urban heat island model: Its role in
air pollution problems, with applications to Montreal.
Presented at the First Canadian Conference on
Micrometeorology, Toronto, Canada, April 12-14, 1965.
Trijonis, J. and K. Yuan, 1977: Visibility in the southwest.
Environmental Protection Agency, Office of Planning and
Evaluation, Office of Research and Development, Research
Triangle Park, N.C.
U. S. Environmental Protection Agency, 1974: Health
Consequences of Sulfur Oxides -A Report from CHESS 1970-1971.
Report No. EPA-650/1-74-004, Human Studies Laboratory,
Research Triangle Park, NC.
, 1975: Position Paper on Regulation of Atmospheric
STulfates. Report No. EPA-450/2-75-007, Strategies and Air
Standards Division, Research Triangle Park, NC.
Wexler, H., 1961: A boundary layer interpretation of the
low-level jet. Tellus, 13, 368-377.
White, W.H., D. L. Blumenthal, J. A. Anderson, R.B. Husar, and
W.E. Wilson, Jr., 1977: Ozone formation in the St. Louis
urban plume. International Conference on Photochemical
Oxidant and Its Control, Procedings.Volume I,
EPA-600/3-77-001a, p. 237, U.S. EPA, Research Triangle Park,
NC, January 1977.
Wilson, W.E., R. J. Carson, R. B. Husar, K. T. Whitby, and D.
Blumenthal, 1977: Sulfates in the Atmosphere -a Progress
Report on Project MISTT. Report No. EPA-600/7-77-021,
Environmental Research Laboratory, Research Triangle Park,
NC.
Wood, J.L., 1971: Nocturnal urban heat island in Austin, Texas.
Texas University, Austin, College of Engineering, Atmospheric
Science Group, Report No. 28, May 1971.
104
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
STUDIES IN AIR QUALITY METEOROLOGY AT
NORTH CAROLINA STATE UNIVERSITY
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
WALTER J. SAUCIER, TED L. TSUI
GERALD F. WATSON, ALLEN J. RIORDAN
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
DEPT. OF MARINE, EARTH AND ATMOSPHERIC SCIENCES
NORTH CAROLINA STATE UNIVERSITY
RALEIGH, NORTH CAROLINA 27650
10. PROGRAM ELEMENT NO.
CDTA1D/03-0260 (FY-83)
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY —RTP.NC
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSOMT^ *G*EWCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report is comprised of summaries of eight investigations into diverse areas
of air quality meteorology resulting from a cooperative research effort by graduate
students and faculty of the atmospheric sciences program of North Carolina State
University and the staff and facilities of the EPA Meteorology and Assessment
Division, Research Triangle Park, North Carolina. Research topics include:
Meteorological analysis of the St. Louis RAPS data -
Climatology of the urban heat island.
Ozone variability in the urban boundary layer.
Surface energy budget over concrete, blacktop, and soil.
Atmospheric visibility and suspended particulates -
Nation-wide trends in visibility from 1955 to 1972.
East Texas air transparency related to haze, smoke, and dust.
Synoptic-scale variability in suspended sulfates.
Mesoscale wind structure over complex terrain -
Diurnal variation in winds across the Appalachian Mountains during an air
stagnation period.
Surface winds in mountainous terrain inferred from 850 mb rawinsonde data.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
3. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
-------
|