INTERSTATE  AIR  POLLUTION

                                             STUDY
 B!-STATE Dl VEOf'MINT
        AGENCY


ST. LOUIS  DF I'AITM-ItlT OF
 HEALTH AND HOSP  IMS


ST. LOUIS  -  I) I* ISI'IN OF
 AIR POLLUT  ON  :(HfROL
  PHASE  n PROJECT REPORT
  EAST SI. IOUI!  "AIR
   POLLUTION  CCH1R5L
      COMMISSION
    ST.  LOUIS COLIN Y
   HEALTH  Di:PARrHEf.hNl
    OF PUBLT, H-/_"H

 CHAMBER OF COMMEK'l E OF
 METROPOLITAN S'   iOUIS
MFTFOROIOGY AND  TOPOGRAPHY
 ILLINOIS Alk PiiLi.l TION
     CONTROL BO^Ri)
         DHI W
  PUBLIC  HEM.Th SCI VICE

-------

-------
          INTERSTATE AIR POLLUTION STUDY
               PHASE II  PROJECT REPORT
          V.   METEOROLOGY AND TOPOGRAPHY
                       prepared by

                      D. O.  Martin
                     P. A. Humphrey
                       J. L. Dicke
            U.S. Environmental Protection Agency
            Region 5, Library (5PL-16)
            230 S. Dearborn St-eet,  Room 1670
            Chicago,  IL   60604
U. S.  DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE

                  Public Health Service

   Bureau of Disease Prevention and Environmental Control
          National Center for Air Pollution Control
                     Cincinnati, Ohio

                       April 1967

-------
      Copies of this report are available from the cooperating agencies listed on
the cover of this report and from the National Center for Air Pollution Control,
1055 Laidlaw Avenue, Cincinnati, Ohio 45237.

-------
                                   FOREWORD
     The Interstate Air Pollution Study was divided into two phases.  Phase I, a
general study of the overall air pollution problems in the St. Louis - East St. Louis
Metropolitan area, was conducted to determine specific activities  that would re-
quire further study in Phase II of the project.  The effort -was  divided into two
phases to provide a logical stopping point in the event that interest and resources
for proceeding further might not materialize.  The necessary  impetus did continue,
however,  and the Phase II operation was also completed.

     The Phase I operation  resulted in a detailed report, designed  primarily for
use of the Executive Committee members and their agencies in making decisions
concerning the Phase  II project operation.  A Phase I summary report was also
prepared; it received  wide  distribution.

    Numerous papers, brochures, and reports were prepared during Phase  II
operations,  as were some 18 Memorandums  of Information and Instruction concern-
ing the project.  All of these documents were drawn upon in the preparation of the
Phase II project report.  The Phase II project report consists  of eight separate
volumes under the following titles:

              I.  Introduction
             II.  Air  Pollutant Emission Inventory
             III.  Air  Quality Measurements
             IV.  Odors - Results  of Surveys
              V.  Meteorology and Topography
             VI.  Effects of Air Pollution
            VII.  Opinion Surveys  and Air Quality Statistical Relationships
           VIII.  Proposal  for an Air Resource Management Program
                                      111

-------
                                 CONTENTS



GENERAL CLIMATOLOGY	    1

TOPOGRAPHY 	    1
METEOROLOGICAL PARAMETERS	    1
     Stability	    1
     Wind	    3
     Winds Aloft - 1,100 Feet Above St. Louis	    3

AIR  POLLUTION CLIMATOLOGY   	    3
URBAN AND TOPOGRAPHIC EFFECTS	   15
     Urban Heat Island Effects	   15
     Topographic Effects	   19

METEOROLOGICAL INSTRUMENTATION IN THE ST. LOUIS AREA ....   21

RESEARCH ASPECTS OF  THE STUDY	   21
     Tracer Studies	   23

REPRESENTATIVENESS OF METEOROLOGICAL
  CONDITIONS DURING THE SURVEY	   31
     Surface Wind	   31
     Wind Data for Pollution Roses	   34
     Temperature	   38
     Precipitation	   39
     Visibility	   39
SUMMARY	   40

REFERENCES	   41

SUPPLEMENTARY READING LIST	   42

DATA SOURCES	   42

-------
                     V.  METEOROLOGY AND TOPOGRAPHY
GENERAL CLIMATOLOGY

     The St.  Louis area experiences the meteorological conditions typical of most
Great Plains cities that are not influenced by major natural features such as moun-
tains or large bodies of water.  The winters are fairly brisk and stimulating as a
rule and are seldom severe.  Almost every winter has several periods of mild
weather and occasional short spells of extreme cold.  Summers are generally
warm,  whereas spring and fall are characterized by moderate temperatures.

     Marked variations in day-to-day conditions are provided by changes from
warm,  moist air flowing up from the Gulf of Mexico during  some periods to cool,
dry  air from the Northern Plains during others.


TOPOGRAPHY

     The St.  Louis area is gently rolling, generally with gradual undulations rather
than sharply defined ridges  and valleys.  Although there are a few rises and drops
of 100 feet in elevation over a short distance, the area within  a 25-mile radius of
downtown St. Louis is generally free of major orographic features that strongly in-
fluence  meteorological variables.   In general, elevations  range from 480 feet  above
sea level in the  downtown St. Louis area to about 550 feet at Lambert Field, 12
miles away,  with a slight ridge rising to 600 feet in between.  A flat area known as
the American Bottoms,  surrounded by a crescent-shaped bluff rising to an  average
of 640 feet above sea level,  lies on the east side of the Mississippi River across
from St. Louis.  The boundary bluff extends from Alton, Illinois,  on the north, to
10 miles east of the Mississippi at East St. Louis, to within 3 miles of the  river
opposite the  mouth of the River des Peres on the south.

     The average elevation of the Mississippi River is 400 feet above sea level, and
of the American Bottoms, about 420 feet above  sea level,  about 60 feet below the
main commercial area of St. Louis.  At times, stagnation of a shallow pool of cool
air in this basin during the night enhances radiation fog formation under clear skies
in the early morning.   These conditions  are generally dissipated within a few hours
after sunrise.  Similar conditions exist in the Missouri River channel  to the -west
and north of  St.  Louis, though the Missouri River bottom area is relatively small
in extent.

METEOROLOGICAL PARAMETERS

Stability

     Stability may be simply described as resistance to change.  In the atmosphere
it maybe measured by the vertical variations (lapse  rate) of temperature.  Be-
cause of the  interaction of pressure,  temperature, and volume, a parcel of air

-------
lifted through the atmosphere experiences an increase in volume (hence a decrease
in density) and a decrease in temperature.  If the parcel is insulated from its  sur-
roundings or is moved quickly enough to prevent exchange of heat with its surround-
ings,  its temperature decreases at the  rate of  5. 4°F for each  1, 000-feet increase
in elevation (0. 98°C for each 100 meters).  This condition is termed the adiabatic
lapse rate.  If, when released at its new position, the air mass is warmer than its
surroundings,  it continues to rise.  If it is cooler than its surroundings it will tend
to descend. If this tendency to rise or  descend is such as to return the air mass to
its level of origin,  the atmosphere is stable; but if the tendency is to move it farther
from  its level of origin,  the atmosphere is unstable.  Thus, if the temperature in
the atmosphere decreases with height more rapidly than the adiabatic lapse rate,
the atmosphere is unstable.
    If the temperature decreases  less rapidly  or even increases with height, the
atmosphere is  termed  stable.  Special terms are applied to these conditions.  A
decrease of temperature with height more rapid than adiabatic is  described as
super adiabatic and is an unstable condition.  A decrease of temperature at the
adiabatic lapse rate is called adiabatic, of course, and is a neutrally stable condi-
tion.  A condition wherein the temperature does  not vary with  height is  termed
isothermal (iso means equal)  and is quite stable.  An increase of  temperature with
height is an inversion, which is most stable.   Conditions between adiabatic and
isothermal are sometimes called "lapse" or slightly stable. Each of the terms
"superadiabatic, " "lapse," and "inversion" covers a range of  conditions.  The
boundaries separating  these ranges are the adiabatic and isothermal conditions.
These variations  in lapse rate and stability are illustrated in Figure 1.
   1000
 
    500
 UJ
                          i    r    i    r
         UNSTABLE
STABLE
                                                   cr
                                                   LJ
                                       I
         I
VERY STABLE
                    -6  -5  -4  -3  -2   -I    0
                         TEMPERATURE,°F
                        Figure 1.   Lapse rates and stability.

-------
     A stable condition inhibits the creation of eddies and thus the diffusion of air
pollutants,  whereas an unstable  condition supports the creation of eddies and ac-
celerates the diffusion of air pollutants.   These effects of variations in stability
may be seen in smoke plumes, as illustrated in Figure 2.


Wind

     The term  "wind"  includes both direction and speed.   The principal effect of
wind direction  is readily apparent since it determines the part of a city that may
be affected  by  the transport of pollution from given sources or areas.  Wind speed
and atmospheric stability largely govern the degree of dilution of effluent before
reaching a given location downwind from a pollution source.  Figures 3 through 6
are surface wind roses for particular months which characterize each of the four
seasons.  The  radial bars on the wind roses indicate the  direction from which the
wind blows; for example, the southeast bar  represents the wind that blows from the
southeast. Figure 7 illustrates monthly average wind speeds at the surface and at
1, 100 feet above St. Louis  (500 meters above sea level) during the years 1950-1959,
as well as seasonal average wind speeds within this layer.


Winds  Aloft - 1, 100 Feet Above St. Louis

     The wind roses in Figures 8  through  15 for 500 meters above mean sea level
(1, 100 feet  above St.  Louis) were prepared from Weather Bureau pilot balloon
(PIBAL) observations  at Lambert Airport. * Again the radial bars on the wind
roses indicate  the direction from which the wind blows.   This level was considered
high enough to  be generally uninfluenced by any surface features, and thus to
present a picture of smooth airflow over the area.  During the -winter months,
1, 100 feet is generally near the top of the mean maximum mixing layer and the
airflow does not show very much diurnal variation in speed or direction; however,
during the other months,  turbulence from surface heating changes the wind regime
with time of day so that at this level daytime wind speeds are generally lower than
nighttime speeds.

AIR  POLLUTION CLIMATOLOGY

     The general weather conditions  and climatology that  have specific application
to this study are summarized in Figures 16 and 17.   Table  1 summarizes and
compares several climatological parameters for St. Louis as well as other large
cities.

     The principal effect of temperature on air  pollution is that space heating is
required in  cold weather.  Heating requirements are measured in degree-days.
The heating degree-day is based on the premise that when the daily mean tem-
perature falls below 65 °F,  space heating is generally required.  The number  of
heating degree-days is defined as the number of degrees  the  mean temperature is
below 65 °F.  Emissions from space-heating equipment are very nearly proportional
to the number of degree-days; hence, degree-day data may be used to estimate
daily and seasonal variations of pollutants emitted from such sources.

    The annual distribution of degree-days is presented in Figure 16 with the
monthly average maximum and minimum temperatures.   Also presented in Figure
16 are the average number of days with thunderstorms, average number of days

-------
         WIND
                                                       .
     TEMPERATURE
                       UNSTABLE CONDITION (LOOPING)
      TEMPERATURE
                   MODERATELY STABLE  CONDITION (CONING)
      TEMPERATURE
                      VERY STABLE  CONDITION (FANNING)
      TEMPERATURE
                   STABLE BELOW, UNSTABLE  ALOFT (LOFTING)
      TEMPERATURE'
                 UNSTABLE BELOW,  STABLE ALOFT (FUMIGATION)
Figure 2.   Schematic  representation of stack gas behavior
             under various  conditions of vertical stability
             (Dry adiabatic lapse  rate shown by  dashed line).

-------
Figure 3.  January surface wind    SPEED CLASSES    Figure 4.   April surface wind rose
     rose for Lambert Airport, CALM^-«=BHZZF        for Lambert Airport,
     1951-60.                       0-3 4-78-12 13.18-19      1951-60.
                                   OCCURRENCE,%
   Figure 5.   July surface wind rose
        for Lambert Airport,
        1951-60.
Figure 6.   October surface wind
     rose for Lambert Airport,
     1951-60.

-------
            25

            23
                   O—O
AVERAGE 500-meter (M S L)
WIND SPEED , NIGHTTIME
AVERAGE 500-meter ( M S L )
WIND SPEED, DAYTIME
AVERAGE SURFACE WIND SPEED , ALL HOURS
AVERAGE WIND SPEED BY SEASONS,
SURFACE-TO-500-meter LAYER
                        M   A   M   J    J    A
                                     MONTH
                                 N
        Figure 7.   Wind speeds at Lambert Airport, averages for  1950-59.
measurable (0. 01 inch or more) precipitation,  and the normal monthly amount of
precipitation.  Thunderstorm days are of interest since a thunderstorm is associated
•with unstable conditions and strong vertical motions, which are effective in carrying
pollutants aloft.  Precipitation alone is less important,  though there is some ten-
dency for rain (or snow) to -wash (or  scour) pollutants out of the atmosphere.  The
number of thunderstorm days is greatest in the summer season.  Spring and early
summer are the best seasons for removal of pollutants from the atmosphere by
wash-out, in view of the high number of days with rain.

     Stability in the lowest layer of the atmosphere varies in a diurnal pattern,
normally being stable at night and unstable in the afternoon.   The height to which
the unstable layer develops is defined as the maximum mixing depth, and,  taken
with wind speed,  represents a volumetric indication of the dilution of air pollutants,
since the scale and applications  are generally area-wide.  Holzworth has computed
mean maximum mixing depths from vertical temperature profiles  (from radiosonde
observations) for all places in the United States -where data were available.  These
•were interpolated for St.  Louis from his maps  (since the nearest radiosonde station
to St.  Louis  is  Columbia, Missouri).  Figure 17  shows the monthly mean maxi-

-------
Figure 8.   January wind rose
    1, 100  feet above surface at
    Lambert Airport,  daytime,
    1950-59, 504 observations.
SPEED CLASSES Figure 9.   April wind rose 1, 100
               feet above surface at Lambert
               Airport, daytime,  1950-59,
               540  observations.
             i
OCCURRENCE,%
   Figure 10.  July wind rose 1,100
       feet above surface at  Lambert
       Airport,  daytime, 1950-59,
       440 observations.
        Figure 11.  October wind rose
            1, 100 feet above surface at
            Lambert Airport, daytime,
            1950-59, 578 observations.

-------
„.     ITT         • j
Figure 12.  January wind rose
    1,100 feet above surface at CA
    Lambert Airport, night-
    time, 1950-59, 496
    observations.
SPEED CLASSES

>-=«=>
0-34-78-12 13-18*19
   mph
    20  30 40
     Figure 13.  April wind rose
         1, 100 feet above surface at
         Lambert Airport, night-
         time, 1950-59, 523
         observations.
 Figure 14.  July wind rose  1,100
     feet above surface at  Lambert
     Airport, nighttime, 1950-59,
     502  observations.
Figure 15.  October wind rose 1, 100
    feet above surface at Lambert
    Airport, nighttime,  1950-59,
    568 observations.

-------
                                          j 'NOIlVlldlDBMd
                            o  o  o  o   o   o
                            m  p  m  p   in   p
                            ^f  ^-'  ro  ro   e\i   oj
                                                   o   o   o   o
                                                   m   o   m   o
                                                   --do
o

I
0.
o
  •g
   cr
   UJ
   m
   5
          o:
          w
          o
     —  v>
     M
     S  fe
     ^  15
     t  3
  *S's  i  S^
   "i Q  a:  uj o
Ib
_ <
                evi   —
                                                             —  o
                                SAVQ
                        ooooooooooo
                        Ooor~-iBin
-------
              Mduu'Q33dS  QNIM
                                                        39V83AV
                            OJ  —
                                        O)    00
                                •Illllllllllllllllllllllll
 >
 0  D
 Y/////////////////M^
             I    1    1    I    I    I
             oooooooo   oooooo
             tom^-focvj   —   oo>   oo   N   (o   m   ^j-   to
                                                                        fi
                                                                        o
                                                                        a
                                                                        i—i

                                                                        0>
                                                                        10

                                                                        0)
                                                                        t)
                                                                        C!
                                                                        fX

                                                                           N   ID   m   ^-   ro
 punojD
                                  'Hld3Q 9NIXIW WflWIXVIfl NV3IAI
10

-------
              Table 1.  CLJMATOLOGICAL NORMALS FOR ST. LOUIS
                        AND FOUR OTHER LARGE CITIES
City
St. Louis
Chicago
Atlanta
Denver
Dallas
Temperature, °F
Mean
maximum
66. 0
59.3
72. 2
63. 5
76. 6
Mean
minimum
46. 6
40.8
52.2
36. 1
56.4
Record
high
115
105
103
105
111
Record
low
-23
-23
- 9
-30
- 3
Degree-
days
4699
6310
2826
6132
2272
Yearly
precipi-
tation,
inches
36.80
32.72
49. 16
14. 20
34.42
Cloud
cover
(tenths)
6.0
6.2
5.7
5.2
5. 1
Mean hourly
wind speed,
mph
9.3
10. 1
9.6
9.6
10.8
Percent
possible
sunshine
56
58
60
69
66
mum mixing depths for St. Louis associated with the average wind speed and with
the number of hours visibility at Lambert Field is restricted to less than 7 miles.
The large monthly variations in mixing depth are quite evident, -with lowest values
during the season of maximum heating.  The  combination appears to be related to
restricted visibility.

     The product of maximum mixing depth and average monthly wind speed at or
near the surface gives a qualitative idea of atmospheric dilution for an area.
Table 2 illustrates this idea for the St. Louis area by using the average wind speed
from the surface to 1, 100 feet above ground.   This  concept indicates that the
atmosphere is least able to dilute pollutants during  the fall and winter season when
dilution is  most necessary, though its application is limited to afternoon conditions.

     The wind rose in Figure  18 illustrates the wind directions associated with
visibility of less than 3 miles at Lambert Airport due to all causes during 1941-
1950.   Visibility restrictions to less  than 3 miles were present in 11 percent of
the weather observations during this  10-year period at the airport.   Since fog is
normally associated with the lower wind speeds, it  appears that visibility of less
than 3 miles attributable to nonfog conditions, primarily smoke and haze,  occurs
most frequently -with -winds coming from the general direction of the  city of
St. Louis (from the southeast) and Granite City (from the east-southeast or east).
The frequency of such visibility restrictions -with -winds from the east-southeast is
about twice that expected from examining the  surface wind roses (Figures 3,  4,
5, 6).

     Frequency of temperature inversion is another useful  clirnatological parameter
for describing atmospheric dilution.  Hosier's^ data in  Table 3 show seasonal per-
cent frequency of inversion at or below 500 feet above ground. The percentage of
total time during which inversions occur is highest  in fall; therefore, the time
during which pollutants can mix freely into a deep layer is a minimum during the
fall.  The very high frequency of nighttime inversions in summer and fall suggests
that early morning pollutant levels will average much higher than early evening
levels.  Figures 19 to 22 illustrate  the percentage of low wind speed classes
(7 mph or less)  recorded at 9 p.m.  and 3 a.m. at the surface and are typical of
conditions beneath  an inversion.  These low wind speeds show the poor dilution
capabilities near the surface  at night,  especially -when considered -with  the high
inversion frequencies in summer and fall.  Also, morning fumigation conditions
occur when pollutants aloft in stable layers of air are mixed down-ward  by turbulence
                                                                                11

-------
                Table 2.  ATMOSPHERIC DILUTION FACTOR
Month
Dec.
Jan.
Feb.
Mar.
Apr.
May
Jione
July
Aug.
Sept.
Oct.
Nov.
Average
wind speed, mph
15.8
16.2
16.8
17.4
17. 1
15. 1
13.7
12.3
12.2
14.2
15. 1
15.6
Mean
maximum
mixing height,
meters
460
460
530
1, 000
1, 060
1,200
1, 040
1,400
1,250
1, 000
800
700
Product
(mph x
meters)
7,400
7,400
8, 900
17,400
18, 100
18, 100
14,400

15, 200
14,200
12,800
10, 900
Seasonal
product
average
Winter,

7,900
Spring,

17,900
Summer,

15, 600
Fall,

12, 600
Relative
dilution
factor


1.0


2. 3


2.0


1.6
       Table 3.  PERCENT FREQUENCY OF INVERSION OCCURRENCE
                           AT COLUMBIA, MO.
Season
Winter
Spring
Summer
Fall
9 p. m. 9 a. m. 6 p. m. 6 a. m.
(Central standard time)
53
54
84
80
38
4
5
24
27
1
5
20
52
67
78
66
Totala
time
31
31
35
43
Maximum
61
62
86
85
         Percent total time.
         Percent of dates on which at least one observation showed an
         inversion.
12

-------
                                     N
W
                                                        OCCURRENCE,%
   Figure 18.  Wind rose for visibility of less  than 3 miles from all causes at
               Lambert Airport, 1941-50.

from surf ace-heating and ground-pollutant concentrations are increased considerably
for 1 to 2 hours.

     Interpolation from Korshover's -work,   indicates that St. Louis experienced
nine cases -when high pressure systems -with very low wind speeds stagnated over
the area for  4 or more consecutive days  during the period 1936-1956.  These
cases resulted in 40 stagnation days,  which occurred from April through October.
Such stagnation days are usually associated with  low -wind speeds and restricted
vertical diffusion.

     Air pollution potential is defined as a set of weather conditions conducive to
the accumulation of air pollutants in the atmosphere over a period of time.  It is
a condition that can be forecast -with reasonable confidence when applied to large
areas and a reasonable persistence.   Miller and  Niemeyer  indicate that Z high
                                                                              13

-------
  Figure 19.   Winter low-speed
     surface wind roses, 9p.m.  CALM
     and 3 a. m. , Lambert
     Airport,  1951-60.
SPEED CLASSES
             Figure 20.  Spring low-speed
                surface wind roses, 9 p.m.
'o-S4-T8..ziS-i8>.8     and 3 a-m> >  Lambert
                Airport, 1951-60.
                                  OCCURRENCE,%
   Figure 21.   Summer low-speed sur-
      face wind roses, 9 p.m.  and
      3 a.m., Lambert Airport, 1951-60.
        Figure 22.  Fall low-speed surface
           wind roses,  9 p.m. and 3  a.m. ,
           Lambert Airport, 1951-60.
14

-------
air pollution potential days occurred in St. Louis from August I960 through July 1961.
Their criteria for high air pollution potential are: (1) surface wind speeds less than
8 knots,  (2) winds at no level below 500 mb  (about 18, 000 feet) greater than 25 knots,
(3) the existence of subsidence below 600 mb  (about 14, 000 feet), and  (4) persis-
tence of these  conditions for 36 hours or more over a minimum area equivalent
to a 4-degree latitude - longitude square.  Since July,  1961,  these criteria -were
met only for a 3-day period,  November 30 - December 2, 1962, and for 1 day in
October 1964,  when St.  Louis was on the boundary of such an area on the 15th and
16th.  Forecasts of high air pollution potential areas are issued daily over the
Weather Bureau Service C teletype circuit by the Weather Bureau Research Station
in Cincinnati,  Ohio,  and may be obtained from the Meteorologist-in-Charge at
Lambert Airport,


URBAN AND TOPOGRAPHIC EFFECTS

Urban Heat Island Effects

     Metropolitan areas impose their  own effects upon the atmosphere.   The city
acts as a heat  island, or heat  reservoir,  by  storing up heat by absorption from the
sun's rays during the day  and  releasing the stored heat at night.  The heat island
has two direct effects.   Nighttime temperatures remain higher than in the surround-
ing countryside; and  the nighttime inversion,  which generally is based at the ground,
is lifted to some distance  above  ground over the city itself.   St.  Louis is no ex-
ception to this rule.  The  heat island effect is illustrated in Figures 23 through 26,
in which the average minimum temperatures for each of 4 months are mapped for
several sites in and around the city.  Note that  in each case the  minimum tempera-
ture at the KM OX-TV tower averages higher than at any of the outlying sites.
Table 4 illustrates the  inversion effect; average early morning temperatures for
each of 4 months are given for various  levels on the KMOX-TV tower.  In each
case the average minimum temperature at 125 feet was lower than the average
temperature at 250 feet, and in three of the four cases,  lower than the average
temperature at 455 feet.  In July and  April the average condition was an inversion
for both layers; in October, an isothermal layer above an inversion; and in January,
a stable (between neutral and isothermal) layer  above an inversion.

     Although the existence of  these urban "heat islands"  has  been -well documented" ~ ^
and can be explained quite easily,  their dimensions and total  effect have not been de-
fined.   When the multitudinous variations affecting them are  taken into account, a
simple model likely to  provide such information is not possible.  Their vertical ex-
tent has been estimated' at about three times the average height of buildings, al-
though this may be suspect in  the case of a small cluster of tall  buildings such as
occurs in the downtown  area of larger cities.

     The factors involved in producing a heat island are many.  Heat itself may be
supplied by absorption  of solar radiation during the day and reradiation at night, or
by space heating during colder months.  Low •wind speeds are essential if the thermal
stratification characteristic of a heat island  is to be  established and maintained. The
presence  of buildings reduces -wind speeds.  Clear skies  intensify the effect since
radiation  from the countryside permits rapid cooling at night -while radiation from
building walls  reflects back and forth to be absorbed eventually by the air, perhaps
mostly by conduction from the •warm surfaces.
                                                                              15

-------
                                                                                  SO  Ut  HO
 Figure 23.  Average minimum tem-
               perature for July 1963.
  Figure 24.  Average minimum tem-
                perature for October
                1963.
     HS-HAZELWOOD SCHOOL SP-STATE POLICE STATION
     LA-LAMBERT AIRPORT   KT-KMOX-TV TOWER
     LS-LINDBERG SCHOOL  fwEW-RADIO STATION (Oct only)
A OBSERVED MINIMUM TEMPERATURE AT 125ft ABOVE
  GROUND
B. REDUCED TO GROUND LEVEL USING AVERAGE LAPSE
  RATE 125 TO 250 ft
C. REDUCED TO GROUND LEVEL USING ADIABATIC
  LAPSE RATE
                                                                             500- SIO  S!0  MO  S«
 Figure 25.   Average minimum tem-
               perature for  January
               1964.
  Figure 26.   Average minimum tem-
                perature for April 1964.
16

-------
            Table 4.  EARLY MORNING MINIMUM TEMPERATURES
                  AT 125-FOOT LEVEL ON KNOX-TV TOWER
           (Adiabatic Temperatures are Derived From 125-Foot Data
              for Comparison With Observed Data at Other Levels)
Height above street
Temperature, °F
1963
July
Oct.
1964
Jan.
Apr.
455 feet    Average, measured
              adiabatic
250 feet    Average, measured
              adiabatic
125 feet    Average, measured
125 to 250 feet
125 to 455 feet
31
26
23
16
21
19
20~
22
69. 5  60. 7   31. 1
66.0  58.6   29.7
69.3  60.7
67.1  59.7
67.8  60.4
                                                                 52.9
                                                                 50.6
                                                          31.7   52.7a
                                                          30.8
                                                          31.5
51.7
52.4
                                              Number of Inversions
             Data from 1 day •were missing.

     The heat island itself may be described as a "bubble" of air with a neutral to
isothermal lapse rate or weak inversion inclosed in a general inversion, with the
inversion layer deep enough to inclose the entire bubble.  This is a subjective
description,  illustrated in Figure 27, and should be verified when possible by
actual measurements.  Some such measurements are available, but are not
                                             o
sufficiently detailed for a definite description.

     The circulation within the heat island itself must be described by inference
                                                      12
since suitable measurements are not available.  Arnold   has already provided
a reasonable suggestion for  the St.  Louis area.   The simplest model would be
based on two assumptions,   (1) a symmetrical warm area,  warmer at the center
than at the edges, and  (2) no overriding wind flow.   In this model the warm air
would rise in the central area,  flow outward along the surface of the bubble,  lose
heat to the surrounding atmosphere, sink to ground  level at the outer edges,  and
return to the central area along  the ground from, all directions.  This is also shown
in Figure 27.   There would be some exchange with the  surrounding  atmosphere  due
to mixing at the interface, but this would be minimal.  This flow pattern •would
permit pollutants  emitted in the  central areas to eventually reach intermediate
areas from the opposite direction.  In reality,  variations in terrain and heat sources
(a factory might be a "hot spot") -would introduce many modifications to this pattern,
but it should  be expected that pollutants emitted at almost any source within the
bubble would reach the central area,  and that average concentrations of any pollu-
tant emitted within the bubble would increase •with time until the condition is
terminated or the source eliminated. A very light wind would perhaps  distort the
bubble without  destroying it.

     Termination of such a heat island,  or rather the effect of its closed circula-
tion, would come  about either upon general warming of the surrounding area suf-
                                                                              17

-------
                                                                                                                                     n!
                                                                                                                                     m
                                                                                                                                     O
                                                                                                                                     4->
                                                                                                                                     C
                                                                                                                                     

                                                                                                                                     B
                                                                                                                                     cu

                                                                                                                                     o

                                                                                                                                     B

                                                                                                                                     CUD

                                                                                                                                     -.-I

                                                                                                                                     ^
                                                                                                                                     o
                                                                                                                                     ,T!
                                                                                                                                     CO
                                                                                                                                     rt
                                                                                                                                     0)
                                                                                                                                     CD

                                                                                                                                     SI
                                                                                                                                     00
18

-------
ficient to create a neutral or unstable condition,  or upon an increase in wind speed
sufficient to break down the interface between the more stable and less stable layers.


Topographic Effects

     Topography also has an effect on weather conditions.   Low-lying areas have
consistently lower minimum temperatures than surrounding hills.  As air cools and
becomes more dense, it settles  into low places.   When low rural or suburban areas
are next to a metropolitan area, this effect becomes dramatic.  The American
Bottoms across the Mississippi  River from St.  Louis,  is such a low lying area.
Inversions in this area are of great frequency and intensity.  This local effect has
been well discussed, for the fall season, by Dr.  G. R. Arnold in his doctoral
dissertation. ^

    Dr. Arnold set up an observing  site for both temperature and wind at the WEW
radio transmitter site, bet-ween  East St. Louis and  Caseyville, Illinois, about 5
miles east of the Mississippi River, and made observations from September 6 to
December 6, 1963,  inclusive. During this period there were  only six dates on
which no inversion -was detected in observations made at either dawn or dusk.

     Ten-degree or greater  inversions (where the temperature aloft was at least
10°F warmer than at the surface) occurred on  45 percent of the mornings and
15°F inversions, on 25 percent  of the mornings.  On one occasion an inversion  of
27 °F was measured at dawn, and on another,  a 17°F inversion was measured at
dusk.

    Most of these inversions •were represented by cold air in the Bottoms.  The
elevation of the surrounding bluff and hills averages about 200 feet above the
Bottoms.  Of all the inversions  Dr.  Arnold measured,  about 81 percent of the
average temperature increase noted at  500 feet occurred in the first 200 feet, or
•within the  confines of the valley.

     The minimum temperature  at the WEW site averaged about 7°F lower than  at
the Customhouse  in St.  Louis; only  once was it as much as 3°F warmer, whereas
on 27 occasions it was 10°F or more colder.  (December data were not included. )

    Evidence of a "count erf low" (wind reversal) was found in both meteorological
data and soiling index measurements of suspended particulates.  Although "There
are no principal sources of  smoke either south or east of WEW, " by far the greater
number of soiling index  samples showing high pollutant values were obtained -with
•winds from those directions.  Visual observations  of smoke from burning dumps
or open fires showed the counterflow also.  Figure  28 is reproduced from Arnold's
treatise to illustrate the counterflow, which is similar to the heat island circulation
postulated above.

    The juxtaposition of the cold pool of air in the American Bottoms and the  heat
island of the City of St. Louis thus produces complex wind patterns, the effect of
which are to retain effluents in the city  and bottom area during night and early
morning hours. Dr.  Arnold presents seasonal evidence of a quasi-closed circula-
tion within these areas.  Unfortunately year-round data are not available, but there
is little doubt that this condition  occurs, in greater  or  lesser degree,  in all seasons
of the year.
                                                                              19

-------
                                                                                                    O
                                                                                                    s
                                                                                                   B
                                                                                                   •*->
                                                                                                    o
                                                                                                   PQ

                                                                                                    rt
                                                                                                    u
                                                                                                    (U
                                                                                                    s
                                                                                                    fn
                                                                                                    (1)
                                                                                                    I-l

                                                                                                    O
                                                                                                   oo
                                                                                                   (M
                                                                                                    d
                                                                                                    ao
                                                              I	I
                                   '13A31 V3S  3A08V NOI1VA313
20

-------
METEOROLOGICAL INSTRUMENTATION IN THE ST. LOUIS AREA

    Figure 29 shows the location of'Weather Bureau Offices,  Cooperative Climato-
logical Stations,  and special meteorological instrumentation sites in the St.  Louis
area.  During the atmospheric dispersion field experiments, special equipment and
instruments were used to determine more accurately the character  of the atmosphere
at low levels.  Balloon-borne  instruments were used to measure the variation of
temperature with height, and pilot ballons were used to measure the variation of
wind with height.

    Table 5 lists the location  of the Interstate Air Pollution Study meteorological
sites  and the instrumentation at each.

RESEARCH ASPECTS OF THE STUDY

    In connection with the Interstate Air  Pollution Study, detailed dispersion studies
were  conducted by the Meteorology Section, Laboratory of Engineering and Physical
               Table 5.  METEOROLOGICAL INSTRUMENTATION
               FOR ATMOSPHERIC DIFFUSION RESEARCH STUDY
Location
KMOX-TV tower








Hazelwood High School

Missouri State Police Radio Tower -
Daniel Boone Expressway near
Ballas Road
Lindbergh High School

Forest Park
Height, ft
125


250

393
455


6
10 ft above
roof
6
50
6
10 ft above
roof
3
Instrumentation
Temperature sensor
Aerovane (wind direction
and speed)
Anemometer (speed only)
Temperature sensor
Anemometer (speed only)
Anemometer (speed only)
Temperature sensor
Aerovane
Anemometer (speed only)
Hygrothermograph
Aerovane
Hygrothermograph
Aerovane
Hygrothermograph
Aerovane
Fluorescent particle
dispensing equipment
                                                                               21

-------
                      Location

             1.   Alton Dam No. 26.
             2.   E.  St. Louis Airport
             3.   St.  Charles
             4.   St.  Louis River
                  Forecast Center

             5.   St.  Louis University
             6.   Valley Park
             7.   Webster  Groves
             0.   Lambert  Airport
             X.   USWB City Office

             A.   KMOX TV  Tower
             B.   Hazelwood High School
             C.   State  Pol ice Radio Tower
             D.   Lindbergh High School
             E.   Forest Park
            Instrumentation

Recording rain  gauge  and  temperature
Precipitation and  temperature
Recording rain  gauge  and  temperature
Precipitation and  temperature
Precipitation  and  temperature
Precipitation
Precipitation  and  temperature
Complete weather  records
Complete weather  records

Instrumented for  air  pollution  study
Instrumented for  air  pollution  study
Instrumented for  air  pollution  study
Instrumented for  air  pollution  study
Dispensing site for  tracer  study
  Figure 29.   Climatological stations  and  specially instrumented meteorological
                 sites in St.  Louis area.
                                                ^
22

-------
Science, United States Public Health Service.  The objective of this study was to
describe quantitatively the dispersion of material emanating from various locations
near and within an urban area,  and to relate the dispersion to the  causative
atmospheric flow and turbulence.

    The dispersion  studies were of two types involving different techniques.  In
one study,  fluorescent (tracer) particles •were released under various meteoro-
logical conditions and their dispersion over the downwind area -was measured.  In
the other study,  an inventory of sources of sulfur dioxide was made and the dis-
tribution of this pollutant over the city was measured and correlated with mete-
orological  data.

    Meteorological  measurement sites in Figure  29  provided continuous data, which
has been reduced and placed on automatic data-processing cards.   When these data
are combined with the tracer dispersion data,  a mathematical atmospheric disper-
sion model can be developed.  The model itself is an expression relating  the vari-
ous pollution sources throughout the area,  both by location and strength,  and vari-
ous meteorological parameters (principally wind direction,  wind speed, and sta-
bility to the distribution of pollutants). Once completed and tested, the model will
be useful in planning for the future, in estimating concentrations  to be expected at
any place,  and in estimating the effect of new sources or changes  in existing sources.

    Work  is not  yet completed on dispersion studies.  When it is, separate tech-
nical papers and reports  will be prepared.


Tracer Studies

    Studies in which tracer materials are deployed in the atmosphere have been
conducted  at various times and places over many  years.  From these  studies,
atmospheric diffusion coefficients have been determined with a reasonable degree
of precision.  Most  of these studies have been over open  and level countryside,
and their applicability to  the relatively complicated features  of an urban area is
not straightforward.

    The initiation of the Interstate Air Pollution Study made  it possible to conduct
tracer studies over  an urban area at minimum expense because of the facilities
and equipment on hand.  Additional meteorological instrumentation was installed
on the  KMOX-TV tower downtown, at Lindbergh High School,  Missouri State Police
Station C,  and Hazelwood High School--all on the  periphery  of the urban area.
Samplers were placed along seven arcs  centered on two release points,  Forest
Park for use •with westerly •winds, and the Knights of Columbus Building at Grand
and Potomac for southeasterly -winds.  These sites and the sampling arcs are
shown  in Figure  30.   Over 300 sampling sites •were selected; 30 to 50 were used
in each experiment.   They included trees,  utility  poles, public buildings, residen-
ces, and commercial structures.

    The tracer material  used -was zinc cadmium sulfide,  a fluorescent compound.
Particle diameters ranged from 0. 5 to 5 microns and averaged  about 3 microns (about
1/10,000 inch).   Particles of this size settle so slowly and respond so readily to ran-
dom molecular motions in the air that their dispersion may be  considered the same
as that of a gas for studies of this type.  Samplers were of three types:  (1) roto-
rod samplers,  consisting of an H-shaped collector rod  kept  spinning by a battery-
                                                                              23

-------
      LEG END
      PARKS
      AIRPORT
      SAMPLING SITES
      METEOROLOGICAL SITES
      FOREST PARK
      GRAND AND POTOMAC
      AVENUES
            Figure 30.  Sampling site locations used in tracer study.

powered motor (uprights were about 1/70 inch broad and 1-1/5 inches high,  of
metal -with a light coating of silicone grease,  and spinning on a radius of about one
and one-fifth inches at 2, 400 rpm); (2) membrane filter samplers, in which air
is drawn through a membrane filter at a predetermined rate by an electrically-
powered vacuum pump; and  (3) a drum pulsed-sampler, in which air is drawn through
a slit-orifice at a predetermined rate and impinges upon a metal drum (coated lightly
with silicone grease),  which rotates by 3-degree steps at variable but controlled in-
tervals.  The  first two samplers provide estimates of total exposure to the cloud of
24

-------
particles (dosage), whereas the drum provides an estimate of the density of the cloud
at various times as it passes over the sampling point.

     In addition to the fixed meteorological network (Figure 30),  a wind recorder
•was operated and pilot  balloon observations were taken at the tracer release site.
The wind recorder provided a continuous  record of -wind speed and direction near
the ground, and the pilot balloon observations provided a measurement of the
vertical variation of  -wind speed and direction at a particular time.  Transponder-
equipped tetroons*' •were released from the tracer release site during each experi-
ment and followed by use of the Weather Bureau radar at Lambert Airport.  Tet-
roons float at a predetermined height above ground and show visually •where a
parcel of air is going.  Thus  they enabled the research crew to follow the cloud
of invisible particles.  The transponder is an electronic device that enables a
radar operator at a considerable distance to follow the tetroon by eliminating
interfering radar echoes from buildings and other fixed objects.   The radiosonde
provides a record of the  rate at which temperature and relative humidity change
•with altitude,  and thus  an estimate of atmospheric stability at the time of observa-
tion.  Other meteorological data were obtained during many releases by radiosonde
ascents,  either free  or tethered from the roof of the Federal Building at 12th  and
Market Streets.   Tethered balloons were  also used to carry Rotorod samplers to
elevations of about 1, 000 feet to obtain  vertical profiles of the tracer cloud  during
17 releases.

     Seven series, totaling 43 experiments, were conducted between May 1963  and
March 1965.

     Limited assays of  the sampling results •were made during each test series to
correct  any obvious deficiencies in the  sampling  layout or procedures.  After each
test series, all samples  -were sent to Metronics, Inc. in  Palo Alto, California, for
actual sample  assay.  From  3 to 6 months was required to complete the sample
counting and  checking.   These assay results,  expressed as number of particles per
sample,  •were  analyzed together •with  the meteorological data to determine the  dis-
persion patterns  obtained under various meteorological conditions.

     Examples of the results for two quite different conditions •were selected for il-
lustration.  The ninth experiment, a release from Forest Park,  was  run during
midday on September 12,  1963.  The  wind was blowing steadily and briskly  from the
northwest, and the pattern of concentrations •was the reasonably familiar elongated
ellipse (Figure 31), with the concentrations decreasing rapidly with distance from
the source.  In this case, all the various  measurements of wind direction (the three
outlying  sites, two levels on the KMOX  tower, the pilot balloon observations,  the
surface •winds  recorded at the release site, and the tetroon trajectory) were within
15 degrees of the mean direction of the  tracer.  The measured wind speeds, allow-
ing for local  differences  due to differing instrument exposures, were also consis-
tent and  gave a good  indication of the  speed of travel of the tracer.  Figure  32  shows
the cross-plume  concentration of particles in this experiment.  The mean path of
the tetroon clearly indicates the centerline of the cloud of particles.   Figure 33
shows the time required for the cloud of particles to reach each sampling arc in the
same experiment,  and  the duration  of its  passage at each arc.   These data are
from specific drum-type samplers, and the multiple peaks shown for the first two
samples are undoubtedly clue  to the  shifting of the cloud pattern jn fluctuating winds.
The time required for the tetroon to reach the sampling arc corresponds well -with
the time at which measurable concentrations were first in evidence.
                                                                             25

-------
  SEPTEMBER  12, 1963, 1215-1315 C.D.T.
  TRACER  RELEASE
  3XIO~'°  Lower limit of significant data
    FOREST
     PARK
N
                                                            A
                                                                 X 10
                                                                     - 10
                                                                   -10
                                                              3XIO~9
                                                                - 8
Figure 31.  Pattern of relative concentration, in grams per cubic meter, for a
           daytime tracer experiment with steady winds (Concentrations based on
           emission  rate of 1 gram per second).


    To use dispersion data, a term must be found to describe the distribution of
pollutants in a cloud or plume.  Such a term is the statistical standard deviation
(
-------
   SEPTEMBER 12,1963, 1215-1315 C.D.T.
   TRACER RELEASE
   TETROON  1230-1330 C.D.T.
   ARCS  ARC-I

    2, 3
                                                       LEGEND
                                                    •  ROTOROD
                                                    +  MEMBRANE FILTER
                                                    *  DRUM
                     120       130      140       150       160       170

                     AZIMUTH FROM RELEASE SITE, degrees

Figure 32.  Crosswind relative dosage distributions along three sampling arcs,
           and mean direction of tetroon travel.

                                                                    27

-------
                               17.577
                                                                      600
      1,6 OO
      1,400
                                            SEPTEMBER  12, 1963
         "0        10       20       30       40       50      60
            TIME FROM FIRST ARRIVAL OF CLOUD AT SITE, minutes
Figure 33.  Sequences of dosages at three sites at approximately the same azimuth
            from tracer release site (Abscissa is  adjusted for travel time to each
            sampling site according to mean tetroon speed).
28

-------
            SEPTEMBER  12,1963
       1,000

      o>
      1
     £500

     o
     <
     UJ
     Q
        200
     V)
         100
                              TRAVEL  DISTANCE, km
                             I
5  6
7 8 9  10
i  i  i  i
                                     4       6    8   10
                               TRAVEL TIME, minutes
        20
  Figure 34.  Measured crosswind standard deviations (ay) and derived vertical
              standard deviations (
-------
   SEPTEMBER 18,1963, 2100-2200 C.D.T.
   TRACER RELEASE
   IXIO"9 Lower limit of significant data
N
                                                      HAZELWOOD HIGH
                                                                    •- 2335
               LAMBERT AIRPORT
                                                                       2330
  STATE PATROL.
                                                                    - .  2315
                                                            GRAND AND
                                                            POTOMAC
  LINDBERGH HIGH
                                                         2310
                          2305
Figure 35.  Pattern of relative concentration, in grams per cubic meter, for an
            evening tracer experiment with light and shifting winds (Arrows
            indicate mean wind directions at  meteorological sites during time of
            tracer release.  Thin solid lines  indicate time when surface winds
            shifted from easterly to southerly, interpolated from meteorological
            sites).
 30

-------
 Two aspects of the travel of the tracer may contribute to this apparent "loss" of
 material.  First, an appreciable transport of the material within the southerly
 flow aloft must have occurred before the cloud at the surface passed; the sampling
 results indicate that the passage of the cloud aloft was of a duration comparable to
 the transit time of the surface cloud.  Second, the earlier portion of the tracer
 release probably was carried westward around the end of the sampling arcs, and
 subsequently was carried northward a few miles •west of the most distant arc,  so
 that very little  dosage from this portion of the release was observed anywhere.

     With such a complex pattern of flow,  there  is no "standard" diffusion equation
 for which coefficients can be derived.  Indeed, it is impossible to define a "mean
 •wind direction" or a "mean •wind speed, " and  it  appears that the sampling results
 rather poorly define the geometry of the entire tracer cloud.   Weather situations
 similar to this  case are of more than routine  interest, however, because situations
 like this are the type that lead to high concentrations of air pollutants.  Any at-
 tempts to trace sources of pollutants with such a  complex flow pattern are likely
 to be frustrating and inconclusive.
REPRESENTATIVENESS OF METEOROLOGICAL CONDITIONS DURING THE
SURVEY

     Since dispersion and, to some extent, emission of atmospheric pollutants  are
strongly dependent upon meteorological conditions, in a study of air pollution levels
one must know whether meteorological conditions are representative  of those usu-
ally encountered.   With this knowledge one can estimate the concentrations that may
be expected under average  conditions and under other conditions as •well.

     Data used in the folio-wing discussion are from surface observations at Lam-
bert Field and from temperature sensors mounted on the KMOX-TV transmitting
tower in downtown St. Louis.  Data are for the period beginning July 1, 1963,  and
ending June 30, 1964.


Surface Wind
     The average wind speed for the year of the  study was 10. 0 mph, 0. 7 mph in
excess  of the normal.  Of 9 months showing an excess over normal, 4 were at
least 10 percent above,  and 3 of these -were around 20 percent above normal.  The
remaining 3  months were below normal, although in no case by as much as 10 per-
cent.  The monthly variations of average -wind speed and  the normal values are
shown in Figure 36.  With the exception of March,  April,  June,  and July, wind speeds
were such that average concentrations of pollutants for the month should be con-
sidered representative,  as far as the effect of wind speed alone is considered.
For  these 4 months,  however, the measured values should be  considered to be
below average.

     Wind roses are shown for July and  October  1963,  and for  January and April
1964 in Figures 37 through 40,  which correspond to the normal roses in Figures
6, 5, 3, and 4, respectively.  In each case an excess frequency  of winds from the
southeasterly to southerly directions  is noted, as  is a deficiency in winds from the
northwesterly to northerly directions and,  except  in January,  a lower than normal
frequency of calm hours.  In many cases a greater-than-normal frequency of winds
of the higher speed groups can be noted. Since the east-to-west and west-to-east
                                                                              31

-------
    14

    13
  Q.
  E
 LJ
 UJ
 Q.
    10

     9

     8

     7
1      I
                      1964
    NORMAL (ISS
           J	I
                               I  ,--..
 x"-NORMAL (1951-1960)
'    i	i	i	i
                      M
                            J     J
                            MONTH
               N
            Figure 36.  Average hourly wind speeds,  Lambert Airport.
transport of pollutants is of interest because of the interstate air movement, the
normal wind roses and wind roses for the study period were compared by cardinal
sectors with some emphasis on the east and west sectors (Table 6).   The frequen-
cy of all  winds in these two sectors was very near normal except for deficiencies
in October  1963 (west) and January 1964 (east). The frequency of low wind speeds
(7 mph or less),  however, was much below normal except for westerly winds in
January  1964; the frequency of speeds of over 7 mph was near normal for westerly
winds and appreciably above normal for easterly winds.   Since pollutant concentra-
tions generally bear an inverse relationship to wind speeds,  it appears that for

     Table  6.  RELATIVE FREQUENCY OF WIND BY DIRECTION AND SPEEDa
Direction
c
Speed
January 1964
April 1964
July 1963
October 1963
Calm


4. 1
0.2
0. 1
0.7
North

All
0.5
0.5
1.0
0.7

> 7
0.5
0.4
2.0
0.6

< 7
0.4
0.7
0.5
0.8
East

All
0.8
1.0
1. 1
1.1

> 7
1.2
1.3
1.8
1.4

< 7
0.3
0.4
0.6
0.8
South

All
1. 4
1.2
1.2
1.6

> 7
1.5
1.4
1.7
1.8

< 7
1.0
0.2
0.4
1.2
West

All
1.0
1.0
1. 1
0.7

> 7
1.0
1. 1
1.1
0.8

< 7
1.0
0.5
0.6
0.7
 Tabular figures are the ratio of occurrences in the specified month to the average
 number of occurrences in the same month during the period 1951 - I960.
"Directions are grouped by quadrants:  N includes NNW and NNE; E includes NE
 through SE; S includes SSE and SSW; and W includes SW through NW.
cSpeeds included are: All, all reported values not including calm; > 7, all occur-
 rences of over 7 mph; and _< 7, all occurrences of 1 to 7 mph, inclusive.
 32

-------
Figure 37.   July 1963 surface
   wind  rose for Lambert
   Airport.
        Figure 38.   October 1963
           surface wind rose for
           Lambert Airport.
   Figure 39.   January 1964 surface
      wind rose for Lambert
      Airport.
Figure 40.   April  1964 surface
   wind rose for Lambert Airport
                                                                         33

-------
these 4 months, at least, east-west or •west-east transport concentrations should
have been below average.

    In evaluating the wind roses for January and April, an additional factor must
be considered.  On January 1, 1964, the method of recording wind direction at
airport stations •was changed from entry by compass directions to entry by tens of
degrees in azimuth.  In adjusting these  data, recorded for 36-directions, to the
16-point wind rose of previous  data, a  slight bias in favor of the cardinal direc-
tions  (N, E,  S, W) is introduced at the expense of adjacent directions.   This bias,
however,  is much too small to account for the excessive frequency of south •winds
in January and April,  or of the west wind in January (Table 7).


Wind  Data for Pollution Roses
    Since atmospheric pollutants are carried by air currents,  relating their occur-
rence or concentration to wind speed and direction is a common practice, usually
illustrated in the form of a "pollution rose. " When the site at -which -wind data are
obtained is considerably removed from either sources or receptors of air pollution
a question arises as to its applicability.  In the Interstate Air Pollution Study,
measurements are available at Lambert Airport,  Hazelwood High School, the State
Police radio tower on the Daniel Boone Express-way,  Lindbergh High School, and
the KMOX-TV tower  at two levels  (Table 5). Of these only the airport location
can be considered permanent.

    Of the five locations listed,  the most centrally located  is the KMOX-TV tower.
The two wind direction instruments on the tower are at 125 and 455 feet above
street level (Table 5).  Unfortunately neither instrument produced  a complete
record.  For comparison with Lambert Field data wind roses for both levels on
the KMOX-TV tower  were prepared for the  months of July and October 1963 (Fig-
ures 37-40), and January and April 1964 (Figures 41-48).   The months of October
and April are particularly deficient,  with only 480 and 350 hours of data,  respec-
tively, for the upper  level and only 649 hours of data for the lower level in April.
The other data are complete enough, however,  to provide a qualitative rose for
comparison.  The two highest speed  classes used for airport data are combined
into one for the  tower data.  The roses are  generally similar with  one pronounced
exception.  Compared with October airport  data,  the lower tower level shows a
very marked excess  of winds from west through north-west, and an equally marked
deficiency of winds from east-southeast through south.

    A more detailed  and valid comparison can be made by use of Table  7,  where
total occurrences from each direction are expressed in percent of  valid data for
each of the three records.  Because  of the method of recording data,  a slight bias
is introduced.  Lambert Airport data for July and October  1963 are recorded to 16
points; all other data used here are recorded to 36 points and converted to 16 points
by combinations that  group three points into one for the cardinal directions and two
points into one for other directions.   The presence of this bias should be recog-
nized, though its effects are small enough to be disregarded here.

    Of more interest where very local circulations are concerned  is a difference
in pattern between the two tower levels.  For the  2 months  with relatively complete
records,  the lower level shows an excess of winds from the easterly quadrant and
a deficiency from the -westerly quadrant (Table 8). Though based on incomplete
34

-------
ft
w
•5
O
H
 i

X
O
  W
co 2
Q r


gn

U r3
W H
OH U
o
  w
f-1  I


§§
Q DH
P £
h
O
   cq
o
U
•8
H




vO
O--
'H
rH
ft






vO
0s
i — i
fj
rt
1-1





00
• — i
O




00
sO
l-H

1-5



"£
o
ft

«|H


n
CD
|
H
h
0
o
rH
• H



rH
D
>
O
H

O
ft
• H

1)
O


o
rH
• H
\
*
O
FH



(M r^-j f^J rO xO O*^ ^O CD co xO uo \o ^3 ^^ ^O r^l ^D
r-H rH i— ( r-H




O
(M
r^-

rH i — 1 ^
^) I-H f\] co "^t4 LO co i — i LO ro uo co r** o^ o co o o
(Nj i — 1 r— 1

co •— i o o «~H co oo ^* E — ^-O ^o "^ to f\j oo ^ r*-
r-t t-H r-H



o
$

"*&
^*
f*


J^rH^^O^CO^OO^^ g

r^ "^f CO ^5 »~^ "~H ^^ f\] CO CO t*~ CT^ »~H CO ^O CO CO >~ •*
^H (M f-H r-H

(S]
1 — 1
ro f^J rO ""H co Is* *^ LO sO t*** LO ^O "^ cO LO ^^ LO ^*
r-H i-H rH ^*
J LO "^ CO LO CO f\l CO LO OO CTV f\] LO fM OO O CO rH 0s-
rH i — 1 rH r-H CO

t-^ 0s OO ^f CO CO r-H rO f\J OO rH f\J OO CO O ' — 1 ^ O
,-H f\] ,—(

o
00
i-H i-H i-H *^h
rjLOLOCOLO^LOOOCOOOt^-OOf^lLOMtNJCOO
rH i — 1



£^ oo ^^ t*~ ^O LO ^^ ro *~-i r*J ^^ oo LO "*J"* co co co ^^
rH fM rH




r^ £H £H W H H CO CO CO CO CO r^ r^ ^ r^ S O
t-



1 — 1
00
to
fi
o
•H
-M
rt
r-H ^
n) 
-------
 Figure 41.
July 1963 wind rose
for lower level of
KMOX-TV tower,
734 observations.
Figure 42.
October 1963 wind
rose for lower level
of KMOX-TV tower,
739 observations.
                                             MILES PER HOUR

                                    20  40  60  80  IOO 120  140
                                   ^^mit^S^imm^f^imim^SSi
                                    NUMBER OF OBSERVATIONS
                                           I
 Figure 43.  January  1964
             wind rose for
             lower level of
             KMOX-TV tower,
             743 observations.
                                     Figure 44.
              April 1964 wind
              rose for lower
              level of KMOX-
              TV tower, 649
              obs ervations.
36

-------
                                           Figure 46.   October 1963 wind
                                                       rose for upper
                                                       level of KMOX-TV
                                                       tower,  480 obser-
                                                       vations.
Figure 45.
July 1963 wind rose
for upper level of
KMOX-TV  tower,
731 observations.
                                                    MILES  PER HOUR

                                         20  40   60   80  100  120  140
                                         S^J^t^2-—^J—-1
                                         NUMBER  OF  OBSERVATIONS
 Figure 47.  January  1964
             wind rose for
             upper level of
             KMOX-TV tower,
             712 observations.
                             Figure 48.
April 1964 wind
rose for upper
level of KMOX-
TV tower,  350
observations.
                                                                       37

-------
                     Table 8.  NUMBER OF OCCURRENCES
                 OF WIND DIRECTION FROM EAST AND WEST
                        QUADRANTS, KNOX-TV TOWER


040° - 140°
220° - 320°
July 1963
Lower
215
138
Upper
182
164
Jan. 1964
Lower
89
268
Upper
36
299
data for the upper level, an even more pronounced excess of -west to northwest
and deficiency of south and south southwest winds existed at the lower level in
October 1963, according to Table 7.   These effects are almost certainly due to
the presence of the city and its action as a heat source.


     The value of a pollution rose is dependent upon both its  purpose and the data
used in preparing it.   To relate a specific source and a specific receptor, wind
data should be obtained in the immediate vicinity from an instrument so located as
to indicate motion between source and receptor.  Even then, care must be taken to
avoid possible eddies induced by nearby buildings or strong  heat sources.  Wind
roses (Figures 41-48) and tabular comparisons (Tables 7 and 8) indicate that  the
tower data are influenced by local effects.  Whether these effects  influence only
one or both levels cannot be determined from the data.   The presence of the tower
itself creates eddies  reflected in the records as erroneous direction and/or speed
information.  These errors vary with both speed and direction and cannot be ac-
curately assessed.  The instruments at Lambert Airport, on the other hand,  are
well away from any terrain or heat island effects.   In addition, the data are un-
interrupted and there is assurance of continued operation.   The records should be
as representative of the general flow as can be obtained.  Thus, for the overall
study, and for future reference,  the use of Lambert Airport -wind  data for con-
structing pollution roses is preferred.


     Wind records from the other three  locations are from temporary installations
and also have periods of missing data.  In addition,  they are no more centrally
located with respect to the areas of important sources and most affected receptors
than is the airport.  They were intended for use in the tracer studies  (as  were the
KMOX-TV tower records); and although of possible application in  a neighborhood
problem,  they are not suitable for use for overall pollution  roses.
Temperature

    Monthly degree-day totals for the study period are compared with normal
totals in Figure 49.  In most months of this study heating requirements  appeared
to be near normal.  In October,  requirements were about 65 percent below normal;
however, December  of 1963 was the coldest December of record in St. Louis, and
heating requirements were about 145 percent of the normal for the month.
38

-------
LJ
cr
100

 80

 60
cr
UJ  40
o.
    20

     0
                                               MEAN MAXIMUM
                                               MEAN  MINIMUM
                                                  NORMAL
                                               DEGREE-DAYS
                                                          SON
                                                            1963
                                                                         D
1,400

1,200

1,000
      CO
      >-
 800  Q
       i
      LJ
      LJ
 600  cr
      LJ
      Q
 400

 200

   0
      Figure 49.  Monthly mean maximum and minimum temperatures, and monthly
                  normal degree-days at Lambert Airport from July 1963 through
                  June 1964.
     Precipitation

         Rain and snow are of interest in air pollution because of their  effectiveness in
     reducing the pickup of dust particles from the earth's surface by the wind.  A
     minor effect is the -washout or scrubout of some particulate by rain or snow as they
     fall.  Days with measurable precipitation and total monthly precipitation amounts
     are shown in Figure 50.   The number of thunderstorm days is also shown.  The
     occurrence of  a thunderstorm signifies unstable air, usually at  or  near the surface,
     and consequently good dispersion conditions.

         The study year as a whole was below normal in all three categories, total
     amount of precipitation,  number of days with thunderstorms, and number of days
     •with rain.  The summer months were deficient in total precipitation and number of
     days -with thunderstorms, -with September  1963 and May 1964 also deficient in days
     •with rain.  February, March, and April 1964, -were above normal  in total precipita-
     tion; April v/as the only month above normal in all three categories.  Other months
     •were near normal in days with rain  and days with thunderstorms, but December 1963
     and January 1964 -were both deficient in total  precipitation.


     Visibility

         Visibility,  the horizontal distance at which an object  can be seen, is sometimes
     used as an indicator  of air pollution; however, the frequency of visibilities restricted
     to less than 3 miles from all causes  (including rain,  snow, and fog) during the year
     of study,  was only 6  percent compared with a normal of 1 1 percent.  During only
                                                                                  39

-------
    6
    5 -
O
c
O
s
I-  3
Q_
O
LJ
01
    0
                                            D  DAYS WITH THUNDERSTORMS
•                                               DAYS WITH PRECIPITATION
                                               > 0.01 inch
                                                                              20
                                                                              15
                                                                              10
                                                                                 O
                                                                                 en
                                                                                 LU
                                                                                 GO
                                                                    N
                                                                         D
        JFMAMJ        JASO
                     1964                               1963
Figure 50.  Monthly precipitation totals  at Lambert Airport for July 1963 through
            June  1964.
4 months, November,  January, February, and March, did as many as 10 percent of
all hours  show visibilities restricted to less than 3 miles by all causes.  Thus
visibilities -would  indicate that particulate concentrations were  somewhat below nor-
mal at Lambert Airport.
SUMMARY

    Inspection of  Figures 36 and  49 indicates  that 5 months of the period from July
1963 through June 1964 could not  be considered  representative.  July,  April,  and
June were much above normal in  wind speed,  therefore,  diffusion conditions for'the
respective months were better than average.  With an average  wind speed slightly
above normal and heating requirements much  below normal in October, measure-
ments  of pollutants resulting from space heating would be expected to be below
normal.   Since space heating requirements are  normally lower in October than in
winter months,  this departure may not be significant in the total pollution loading of
the atmosphere.   December, with an average  wind speed slightly below normal,
•was much above normal in heating requirements and therefore  in pollutants  result-
ing from  space heating.   In all other months,  both wind speed and heating require-
ments  were near normal.
 40

-------
                                 REFERENCES
 1.   U.S. Weather Bureau.  Winds aloft summary and parameters.  500M MSL.
         1963.

 2.   U.S. Weather Bureau.  Local climatological data supplements.

 3.   Holzworth, G. C.   Estimates of mean maximum mixing depths  in the contig-
         uous United States.   Monthly Weather Review, 92:235-242.  May 1964.

 4.   U.S. Weather Bureau.  Terminal Forecasting Reference Manual.  Lambert -
         St. Louis Airport.  Jan.  1959.

 5.   Hosier,  C. R.  Low-level inversion frequency  in the contiguous United States.
         Monthly Weather Review, 89:319-339.  Sept.  1961.

 6.   Korshover, Julius.  Synoptic climatology of stagnating anticyclones.  Robert
         A.  Taft Sanitary Engineering Center.  Cincinnati, Ohio.   SEC TR A60-7.

 7.   Miller, M. E. and  Niemeyer, L. E.  Air pollution potential forecasts - a year's
         experience.  Journal of the Air Pollution Control Association,  13:205-210.
         May 1963.

 8.   DeMarrais,  G. A.  Vertical  temperature difference observed over an urban
         area.  Bulletin  of the American Meteorological Society, 42:548-554.
         Aug.  1961.

 9.   Duckworth, F. S. and Sandberg,  J.S.   The effect of cities upon horizontal and
         vertical temperature gradients. Bulletin of the American Meteorological
         Society, 25:198-207.  May 1954.

10.   Landsberg, H. E.   The climate of towns in man's role in changing the earth.
         Thomas, W.L. , Jr.   ed.  University of Chicago Press.   1956.

11.   Mitchell, J.M.,  Jr.  The temperature  of cities.  Weatherwise, 14(6):224-229.
         Dec.  1961.

12.   Arnold, G.R.  Local inversions,  air  currents, and smoke pollution in Cahokia
         Bottom. Doctor of Science dissertation.  Washington University.  St. Louis,
         Missouri.  June 1964.

13.   Pack, D. H.  Air trajectories and turbulence statistics from  weather radar
         using tetroons and radar  transponders.  Monthly Weather Review,  90:491-
         506.  Dec. 1962.

14.  Hoel, P. G.  Introduction to  mathematical statistics.  John Wiley and Sons,
        Inc.  New  York.  1963.  p. 74.
                                                                               41

-------
                       SUPPLEMENTARY READING LIST
 1.   Barnum, D. C.  The influence of several meteorological parameters on the
        smokiness in the St. Louis area.  St. Louis University.  Thesis  for M.S.
        degree.  1962.

 2.   Hanson,  D. M. and Koontz, I. A.  A study of restrictions to visibility at
        Lambert Field, St. Louis.  (Preliminary report 1952) U.S. Weather
        Bureau manuscript.

 3.   Murine,  V. S.  Atmospheric pollution and weather in the St. Louis,  Missouri,
        area.   St.  Louis University.  Thesis for M.S. degree.   1952.

 4.   Schueneman,  J. J.  A report of the atmospheric concentrations of SO  in
        St.  Louis  - 1950.  Industrial Hygiene Section, Division of Health.  City of
        St.  Louis.
                                DATA SOURCES
 U.S. Weather Bureau.
Climatography of the United States Number 30-23
(Summary of Hourly Observations 5 years 1950-1955)

Climatography of the United States Number 60-23
(Climates of the States - Missouri 1959)

Climatography of the United States Number 82-23
(Decennial Census of United States Climate 1951-1960

Climatic Summary of the United States - Supplement for
1931-1952 Missouri.  Series Number 11.

Climatological Data (State of Illinois)

Climatological Data (State of Missouri)

Surface Wind Summary and Parameters (In preparation
at National Weather Records Center)
                         U.S. Environmental Protection Agency
                         Begion 5, Library (5PL-16)
                         230 S. Dearborn  Street, Room  1670
                         Chicago, IL    60604
42

-------