EPA-600/4-77-046
November 1977
OZONE OVER SAN FRANCISCO
Means and Patterns During Pollution Episodes
by
Kenneth P. MacKay
Department of Meteorology
San Jose State University
San Jose, California 95192
Grant Number R802235
Project Officer
George C. Holzworth
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 recommenda-
tion for use.
11
-------
ABSTRACT
Measurements of meteorological parameters were taken at six
levels and ozone at four levels between 260m and *»73m ASL on the
Mt. Sutro T.V. Tower in San Francisco during the summers of 1974
through 1976. Hourly average ozone concentrations within the
elevated inversion layer at this location exceeded the 8 pphm
(160/^g m~3) National Ambient Air Quality Standard .about 15% of
the time.
High inversion layer ozone concentrations at this site were
associated with high surface concentrations occurring during area-
wide air pollution episodes. These episodes occurred when a lobe
of the Pacific high pressure system penetrated inland and inter-
rupted the normal onshore California monsoon flow, often in late
summer and early fall when the monsoon is normally weakening.
During these episodes, superposition of synoptic scale mi
northeasterly flow and locally produced mesoscale flow caused
easterly or light westerly flows during the late forenoon within
the inversion layer and westerly flow In the late afternoon. In-
land, where the inversion was destroyed from below, inversion
layer and surface generated pollutants were convectively mixed.
This mixing and the wind oscillation recycled pollutants. The
episodes ended when the synoptic situation reverted to one more
normal for the season and pollutants were advected from the area.
At the start of two September, 197^ case studies,ozone rich
stratospheric air appears to have intruded into the subsidence
inversion and may have acted as a photochemical trigger or have
mixed additively to locally generated pollutants. There are some
suggestions of transport from Los Angeles at the start of a third
episode studied.
This report was submitted 1n fulfillment of Grant R802235 by San Jose State
University under the sponsorship of the U.S. Environmental Protection Agency.
This report covers a period from April 1973-August 1977, and work was completed
as of August 1977.
111
-------
CONTENTS
Abstract 111
Figures v1
Tables x111
Acknowledgement x1v
1. Introduction 1
2. Conclusions 2
3. Recommendations 4
4. Instrumentation and Data Analysis 5
5. Mean Summer Behavior 15
6. Case Study of July 22-26, 1974; by Stephen H. Holets 35
7. Case Study of July 22-27, 1975; by Betsy L. Babson 63
8. Case Study of September 3-8, 1974 73
9. Case Study of September 14-18, 1974 84
10. Summary and Discussion 103
References 106
Appendix
Monitoring station locations and histories 108
-------
FIGURES
Number Page
1 Mt. Sutro T.V. Tower 6
2 West-East profile of topography through tower site. . . 7
3 Map of San Francisco Bay Area 8
4 Instrument package 9
5 Data acquisition system block diagram 10
6 Mean maximum temperature for Sacramento and San
Francisco and mean afternoon wind speed for
San Francisco 16
7 Generalized flow pattern of marine air 1n the San
Francisco Bay Area under typical summer conditions. . 17
8 East-West cross-sections of temperature and west
wind component, 10 PST, 15 August, 1962 17
9 Oakland and San Jose State mean morning and afternoon
soundings of temperature, mixing ratio and wind
velocity, July, 1972 19
10 Time-height sections of hourly average temperature;
Mt. Sutro Tower, July, August and September, 1974 . . 20
11 Hourly average temperatures, Mt, Sutro Tower,
September, 1974 21
12 Hourly average wind vectors, Mt. Sutro Tower, July,
August and September, 1974 22
13 Wind roses at Sutro Tower, summer, 1974 23
14 Mean vertical velocity distributions at Mt. Sutro
Tower, July and August, 1974 24
15 Mean number of days during 1970-74 with oxldants
exceeding 8 pphm 26
vi
-------
FIGURES
Number Page
16 Co-cumulative frequency distributions of ozone at
Sutro Tower, 1974, 1975 and 1976 27
17 Mean diurnal variation of ozone, Sutro Tower,
September, 1974 30
18 Distribution of ratio of wind direction for hours
with 0, >_ 8 pphm to those with (L < 8 pphm, Sutro
Tower, summer, 1974 T 30
19 Diurnal distribution of wind direction percent
frequency for hours with 0^ >_ 8 pphm, Sutro Tower,
level 5, summer, 1974 . . 31
20 Co-cumulative frequency distributions of ozone at
Qulllayute, Wa,, and level 1, Sutro Tower,
summer, 1974 32
21 Co-cumulative frequency distributions of CO, Sutro
Tower, 1975 and 1976 33
22 Hourly average CO concentrations, Sutro Tower,
September, 1976 34
23 Mean 700 mb height contours for July, 1974 36
24 700mb height contours for 23-27 July, 1974 36
25 Mean July maximum hourly temperature distribution
for Bay Area 37
26a Maximum hourly temperature distribution, 22 July,
1974 37
26b Same as 26a, except for 24 July, 1974 38
26c Same as 26a, except for 26 July, 1974 38
27a Surface air trajectories, 0400 PST, 22 July, 1974 ... 40
vi 1
-------
FIGURES
Number Page
27b Same as 27a, except for 1300 PST, 22 July, 1974. ... 40
28a Same as 27a, except for 0400 PST, 24 July, 1974. . . . 41
28b Same as 27a, except for 1300 PST, 24 July, 1974. ... 41
29a Same as 27a, except for 0400 PST, 26 July, 1974. ... 42
29b Same as 27a, except for 0300 PST, 26 July, 1974. ... 42
30 Mean high-hour oxldant distribution over Bay Area,
July, 1968-1974 43
31a High-hour oxldant distribution over Bay Area,
22 July, 1974 43
31 b Same as 31a, except for 24 July, 1974 44
31c Same as 31a, except for 26 July, 1974 44
32a Oakland temperature, mixing ratio and wind profiles
for 22, 24 and 26 July, 1974, 04 PST 46
32b San Jose State University temperature, mixing ratio
and wind speed soundings for 22, 24 and 26 July,
1974, 06 PST 46
33a Mean July, 1974, hourly average time-height section
at the Mt. Sutro Tower, San Francisco 47
33b Time-height section of hourly average temperatures,
Mt. Sutro Tower, 22 July, 1974 47
33c Same as 33b, except for 24 July, 1974 48
33d Same as 33b, except for 26 July, 1974 48
34a Mean July, 1974, hourly average west wind speeds,
time-height section, Mt. Sutro Tower 49
V111
-------
FIGURES
Number Page
34b Time-height section of hourly average west wind
speeds, Mt. Sutro Tower, 21 July, 1974 49
34c Same as 34b, except for 24 July, 1974 50
34d Same as 34b, except for 26 July, 1974 50
35a Mean hourly average ozone time section, Sutro Tower,
summer, 1974 51
35b Hourly average ozone time section, Sutro Tower,
23 July, 1974 51
35c Same as 35b, except for 24 July, 1974 52
35d Same as 35b, except for 25 July, 1974 52
36a Variation of five-minute averages of ozone, Sutro
Tower, 24 July, 1974 53
36b Variation of f1ve-m1nute average vertical velocities,
Sutro Tower, 24 July, 1974 54
36c Time-height temperature section, Sutro Tower,
24 July, 1974 54
37 Temperature and ozone profiles, Hayward, Ca.,
24 July, 1974 55
38a Schematic NW-SE cross section of air flow, Inversion
and mixed layer heights and ozone concentrations,
afternoon of 22 July, 1974 57
38b Same as 38a, except for morning of 23 July, 1974. ... 57
38c Same as 38a, except for afternoon of 23 July, 1974. . , 58
38d Same as 38a, except for morning of 24 July, 1974. ... 58
1x
-------
FIGURES
Number Page
38e Same as 38a, except for afternoon of 24 July, 1974. . . 60
38f Same as 38a, except for morning of 26 July, 1974. ... 60
38g Same as 38a, except for afternoon of 26 July, 1974. . . 60
39a Mean maximum hourly surface temperatures, San
Francisco Bay Area, July, 1975 64
39b Maximum hourly surface temperatures, San Francisco
Bay Area, July 25, 1975 64
40 Maximum hourly oxldant concentrations at BAAPCD
measurement stations, 25 July, 1975 66
41 Rawinsonde profiles of temperature, mixing ratio
and wind velocity, Oakland, Ca., July 23, 25
and 27, 1975 67
42 Hourly average horizontal wind vectors. Mt. Sutro
Tower, July 24-26, 1975 68
43 Hourly average horizontal wind vectors, Mt. Sutro
Tower, July 24, 1974 69
44 Time-height sections of hourly average ozone
concentrations, Mt. Sutro Tower, July 24-26, 1975 . , 70
45 National Weather Service 500 mb analysis, 12 GMT,
3 September, 1974 74
46 National Weather Service 700 mb analysis, 12 GMT,
4 September, 1974 74
47a Hourly average oxldant concentrations, 13 PST,
5 September, 1974 77
47b Same as 47a, except for 14 PST 77
-------
FIGURES
Number Page
48 Hourly oxldant concentrations, San Jose,
Uvermore and Hayward monitoring stations,
September 5-6, 1974 78
49 Time-height sections of hourly average temperature,
Sutro Tower, September 4-6, 1974 79
50 Hourly average wind vectors, Sutro Tower,
September 4-6, 1974 81
51 Hourly average ozone concentrations, Sutro Tower,
September 5-6, 1974 83
52 Montgomery stream function, 312K surface, 00 GMT,
14 September, 1974 86
53a Montgomery stream function, 312K surface, 12 GMT,
14 September, 1974 87
53b Potential vortlcity analysis, 312K surface, 12 GMT,
14 September, 1974 87
54a Same as 53a, except for 00 GMT, 15 September, 1974 , . 88
54b Same as 53b, except for 00 GMT, 15 September, 1974 , . 88
54c Northwest-southeast cross section of potential
temperature and potential vortidty, 00 GMT,
15 September, 1974 89
55a Same as 53a, except for 12 GMT, 15 September, 1974 , . 91
55b Same as 53b, except for 12 GMT, 15 September, 1974 . , 91
56a Same as 53a, except for 00 GMT, 16 September, 1974 . . 92
56b Same as 53b, except for 00 GMT, 16 September, 1974 . . 92
57 NOAA B visual satellite picture of California and
eastern Pacific, 1721 GMT, 14 September, 1974. , . . 93
x1
-------
FIGURES
Number Rape
58a Surface 0- concentrations, San Francisco Bay Area,
09 PST, 15 September, 1974 94
58b Same as 58a, except for 15 PST, 17 September, 1974. .. 94
59 Time-height section of westerly winds, Oakland, Ca,,
September 14-19, 1974 96
60 Hourly average wind vectors and temperatures,
Sutro Tower, September 15-18, 1974 97
61 Hourly average ozone concentrations, Sutro Tower,
September 15-18, 1974 100
-------
TABLES
Number Page
1 Instrument Boom Characteristics 5
2 Meteorological Instruments , 9
3 Meteorological Data Recovery - Tower 12
4 Ozone Data Recovery (Hours) , . . 12
5 CO Data Recovery (Hours) 12
6 Frequency Distribution of Hourly Average Ozone
Concentrations, Sutro Tower, 26 June to 24 September,
1974 25
7 Percent Hours Exceeding 8 pphm Oxldants at Selected
Bay Area stations, July-September, 1974 28
8 Number of Hours Exceeding 8 pphm Oxldants at
Selected Bay Area Stations, July-September, 1974
and 1975 28
9 Median and 90th Percentlle Hourly Average 03
Concentrations (pphm), Qulllayute, Wa., 1974 31
10 Surface Oxldant Concentrations, July 22-31, 1974,
at Chlco and Redding 62
11 Maximum Temperatures Selected Bay Area Stations,
September 3-8, 1974 75
12 High-Hour Oxldant Patterns, September 3*8, 1974 76
13 Mean Surface to 5000 Ft. Winds, Oakland,
September 3-8, 1974 80
14 Maximum Temperatures, September 14-18, 1974 85
15 High-Hour Oxldant Patterns, September 14-19, 1974. ... 90
16 Mixing Depth Parameters Estimated from 1600 PST
Oakland Soundings, September 13-18, 1974 95
-------
ACKNOWLEDGEMENTS
The cooperation of Sutro Tower, Inc. and especially chief engineer Harry
Jacobs is gratefully acknowledged. Miles Imada and Edward Jeung and their
associates of the Air and Industrial Hygiene Laboratory of the California
Department of Health, the Air Resources Board, and Hanford Chew of the Bay Area
Air Pollution Control District calibrated the air pollutant monitors. Auxili-
ary data were supplied by a number of agencies; the Bay Area Air Pollution
Control District and Battelle Northwest Laboratories were especially helpful.
The companion tower projects under Drs. Albert Miller and Jindra Goodman and
supported by the National Science Foundation provided help and encouragement.
Many undergraduate and graduate students have contributed to the project.
The assistance of all is gratefully acknowledged.
xiv
-------
SECTION 1
INTRODUCTION
The San Francisco Bay Area is a meteorologically and topographically com-
plex area with rapidly expanding population. Air pollution generated by the
expanding population and economy has become an increasingly noticeable social
problem. Local governments and associations of local governments are asking
the local Bay Area Air Pollution Control District for data, including output
statistics from air pollution simulation models, for use in decisions on land
use and other matters that may effect future air quality patterns.
Realistic simulation of meteorological and air pollution patterns by air
quality simulation models then is necessary for governmental agencies to make
well informed decisions. Most air quality computer models, such as the LIRAQ
ModeV'(McCracken and Suiter, 1975), consider the elevated inversion layer,
which persists over the West Coast for nearly half the year, as an impenetra-
ble lid to pollutants.
Previous measurements over the Bay Area (see Lowell and Miller, 1968;
Miller and Ahrens, 1970; Gloria et al, 1974) and in the Los Angeles area
(Lea, 1968; Edinger et al, 1972; Edinger, 1974) found significant concentra-
tions of oxidants within the elevated inversion layer. These measurements,
however, were made from such moving platforms as light aircraft or balloons.
Construction of the Mt. Sutro TV Tower in San Francisco provided the oppor-
tunity to make measurements from a stationary platform which sticks into the
inversion for most of the summer and fall months.
The results discussed in this report confirm the existence of high con-
centrations of inversion layer ozone over San Francisco. The following sec-
tions discuss the mean behavior of meteorological and air pollution parame-
ters with emphasis on measurements made from the Mt. Sutro Tower, and then
discuss the evolution of four periods of high oxidant concentration.
The results of this project, which was funded by EPA under the title,
"Air Pollutant Background Profiles for Air Quality Simulation Models", should
help in the improvement of such models for the Bay Area and perhaps other
West Coast cities.
-------
SECTION 2
CONCLUSIONS
Hourly average ozone concentrations measured with reference to the pre-
1975 California Air Resources Board (ARB) neutral-buffered KI method of in-
strument calibration exceed the 8 parts per hundred million (pphm) National
Air Ambient Air Quality Standards (NAAQS) about 15% of the time within the
elevated inversion layer at the Mt. Sutro TV Tower over San Francisco during
the late summer months. Ozone concentrations within the marine layer at this
near coastal location exceed the NAAQS during about 3% of the hours. (These
measurements are approximately equivalent to the EPA approved method.) 03
concentrations in the inversion layer measured with reference to the 1975 and
later ARB uv-absorption calibration technique exceed the NAAQS during about
9% of the hours. (8pphm»l60/"g m"3)
High inversion layer ozone concentrations are associated with high sur-
face ozone concentrations measured at Bay Area Air Pollution Control District
(BAAPCD) monitoring stations in inland portions of the area during air pollu-
tion episodes. A lobe of the Pacific high pressure system penetrating inland
interrupts the normal onshore California monsoon. This interruption often
occurs in the late summer and early fall when the monsoon is normally weaken-
ing.
During a high oxidant episode in the Bay Area, the daily average height
of the base of the elevated inversion lowers until the peak day of the episode
when the inversion base will be at or below the 250 m ASL peak of Mt. Sutro.
The normal monsoon flow is re-established and the inversion base rises on sub-
sequent days. Diurnal oscillations of the inversion base, similar to the
average monthly behavior and produced locally, are superimposed on this be-
havior. The base of the inversion is highest in the early morning (04-07 PST)
and lowest in mid afternoon (about 15 PST).
During these episodes, the mesoscale circulation produced by local heat-
ing and cooling is superimposed in northeasterly flow. The winds within the
marine layer at Mt. Sutro are westerly or southwesterly throughout the day.
As the inversion lowers in the late forenoon, winds above about 350 m ASL be-
come northeasterly or weak westerly as synoptic scale flow dominates the weak
onshore breezes. In the late afternoon, winds back to westerly as inland
heating produces onshore flow. The duration and vertical extent of easterly
component winds at the Sutro Tower increases until the peak oxidant day of the
episodes.
The onshore flow that commences in the afternoon occurs both within the
marine layer below the inversion base and within the very warm, dry air above.
-------
Therefore, the structure of the leading edge of the incoming air is not like
the typical sea breeze cold front. The inversion layer air is horizontally
homogeneous and a frontal surface is therefore missing. The leading edge of
the incoming marine layer i»s initially a typical sea breeze front. Ahead of
the leading edge of the marine air, the inversion is destroyed by surface
heating and convection. This allows inversion layer ozone and perhaps other
pollutants to intermix with surface generated pollutants.
Other measurements (see e.g., Miller, 1966; Miller and Ahrens, 1970)
have shown nearly stagnant air and high oxidant concentrations ahead of the
leading edge of the marine air. The exact pattern of the leading edge of the
marine air in the peak day of an episode should then determine which stations
will get the highest oxidant concentrations. At night, cooling at inland lo-
cations sllows dynamic processes to reform the elevated inversion thereby
trapping pollutants aloft. Downslope, down valley and land breezes combine to
transport this polluted layer seaward. During the early morning and forenoon
hours this layer remains near the coast where the inversion is intact and is
brought back inland on the late afternoon breeze. Pollutants for each day of
an episode then build upon successively higher backgrounds.
The monsoon circulation is re-established to end the episode and advect
pollutants eastward from the area where they may contribute to high oxidant
concentrations in the Central Valley or the Sierra Nevada.
At the beginning of one episode studied, there may have been pollutant
transport from the Los Angeles basin. While this conjecture has not been
proven, inspection of National Weather Service 850 m b analysis and available
upper air wind data indicates the possibility of an isentropic trajectory
from the Los Angeles basin to San Francisco. In two other cases studied,
stratospheric ozone intrusion may have contributed significantly either addi-
tively or as a triggering mechanism.
-------
SECTION 3
RECOMMENDATIONS
The evolution of air pollution episodes in the Bay Area should be studied
in more detail. Measurements during a number of 1974 observation periods are
presently being processed (Bay Area Air Pollution Control District, personal
communication). These observations will be used for verification of the
LIRAQ air pollution model (McCracken and Sauter, 1975). The need for further
field measurements for model verification purposes should be evaluated once
available data are processed and computer model runs completed. If further
documentation of the evolution of Bay Area air pollution episodes is desired,
either for control or for model verification purposes, then measurement
periods should last for approximately a week to encompass build up, alert,
and break down periods. Aircraft flights should measure the off-shore extent
of the elevated polluted layer.
The existence and amount of pollutant transport from southern California
to the Bay Area, especially at the start of a Bay Area air pollution episode,
should be investigated, probably with the use of instrumented aircraft.
The possible role of stratospheric intrusion of ozone rich air, again
especially at the beginning of a Bay Area air pollution episode, should be
investigated. Computation of potential vorticity patterns coupled with meas-
urements of radionuclides of stratospheric or upper tropospheric origin could
estimate the significance of this possible source.
An important unsolved transport problem appears to be the fate of Bay
Area generated oxidants and their precursors and their influence on the air
quality of the Central Valley and Sierra Nevada of California. Oxidant mon-
itoring for a complete season in a number of locations suspected of being re-
ceptors of Bay Area smog, coupled with field studies of transport parameters
would provide impetus for the development of large scale transport simulation
models.
-------
SECTION 4
INSTRUMENTATION AND DATA ANALYSIS
THE MT. SUTRO T. V. TOWER
The Mt. Sutro T.V. Tower (Figure 1) is located atop a 254 m hill in the
center of San Francisco (Figures 2,3). The site is dominated by the tower
and the transmission facilities building. A stand of Eucalyptus trees, whose
tops extend to about 14 m above the tower base, borders the site from the
north through west to south sides. The trees are about 30 m from the tower
at their closest point.
The tower consists of three equally spaced legs connected by cross beams
at three levels. It has an "hour glass" shape where the distance between
legs 1s 46 m at the base, 19 m at the 162 m high waist, and 30 m at the 225 m
top platform. Each leg 1s an equilateral triangle 1n cross section, 2.1 m on
a side. Our Instruments are Installed on the west leg in which there 1s an
elevator.
INSTRUMENT SUPPORTS AND METEOROLOGICAL INSTRUMENTS
The meteorological instruments and data signal conditioning electronics
are attached to the ends of booms mounted at six tower levels (Figure 1).
Boom heights and lengths are listed in Table 1. All boom lengths are between
one and four tower-leg-widths long. The four uppermost ends are on a nearly
vertical line (± 0.61 m or ± 2 ft.).
TABLE 1. INSTRUMENT BOOM CHARACTERISTICS
Level
1
2
3
4
5
6
Boom
Above Ground
6.1
45.4
87.8
136.2
178.8
218.7
Height (m)
Above Sea Level
260.4
299.7
342.1
390.6
433.1
473.0
Boom
Length (m)
1.83
2.74
2.74
8.23
8.84
5.18
-------
5 2-m boom at
219-m level-"
8 8-m boom at
179-m level
8 2-m boom at
136-m level
2 8-m boom at
38-m level
? 6-m boom at
45 -m level
I 8 - m boom at
6-m level
Figure 1
Mt. Sutro T.V.
Tower
The lowest boom 1s fixed, while the other five are movable for access
purposes. In order to service the Instruments, the movable booms are lowered
to service platforms. Cables are attached to both top and bottom of each
boom near Its free end and pass through pulleys to small winches located at
the platform levels.
Prior to summer, 1975, stops prevented the booms from being raised be-
yond their horizontal position. Subsequently, potentiometric pendulums were
-------
installed 1n each boom so that deviations from the vertical along two orthog-
onal axes are continuously recorded.
The meteorological instruments are mounted as a unit (Figure 4) which
then may be attached or removed from the boom end. Each package contains
instruments for measuring the following parameters: air temperature, wet-
bulb temperature, u-, v- and w- wind components and pressure. Table 2 lists
the instruments for measuring each parameter.
Tower
WEST
DISTANCE FROM TOWER SITE (KM)
Figure 2. West-East profile of topography through tower site.
-------
SAN FRANCISCO BAY AREA
Figure 3. Map of area. Location 14 is Mt. Sutro, other
locations are Bay Area Air Pollution Control District
Monitoring Stations listed in Appendix.
-------
TABLE 2. METEOROLOGICAL INSTRUMENTS
Parameter
Instrument
Remarks
Air Temperature
Wet-Bulb
Temperature
Wind Components
Pressure
Yellow Springs Instrument
Co. (YSI) Model 705
Thermistor
Same, with wick
Gill Model 27003 Propel lor
Anemometer
Rosemount Model 1201
A2G 8BD Pressure
Transducer
Attached .taoYSI
Thermivolt
Signal Con-
ditioner
Same
Temperature of
housing con-
trolled to
± 3 C
L.J
j'tMI-CflMUM
jWIT IULI tlMPIMATOKI
Figure 4. Instrument Package, dimensions in Inches,
Indicated by letters u, v and w.
Wind component sensors
-------
DATA ACQUISITION SYSTEM
The data acquisition system (Figure 5) consists of a completely self-
contained remote subsystem at each boom, the central data coupler, and re-
cording devices.
Level 6
remote
sub-syst
Remote Sub-system
Level 2
Level 1
10 boom V6 tower
sensors sensors
Control Bus
Data Bus
ASI MODEL 420
Central
Data Coupler
Control In
Output Drive to
tape recorder and/or
telephone modem
Figure 5. Data acquisition system block diagram.
The remote sub-systems receive analog signals from up to 16 instruments;
ten signals from instruments located at the ends of the boom and six from
instruments located inside the tower leg. The signal from an instrument
passes first through a signal conditioning circuit which sets the zero offset
and range. The conditioned signal then passes through an analog to digital
converter and then through a differential line driver circuit for transmis-
sion down the tower. The digital information is bussed in a "daisy-chain"
fashion in series via an individually shielded 32-pair conductor cable to the
central data coupler.
The central data coupler is an Ambient Systems, Inc., Model 420 capable
of interfacing with a minicomputer, magnetic tape recorder, teleprinter or
telephone modem (remote telegraph or similar printer connected to the data
coupler via telephone line). The coupler also controls querying each remote
sub-system either automatically or manually. As signals are received from a
remote sub-system they pass through a BCD interface card where the word format
is assembled and converted into computer code. The coupler also inserts end
of word, end of record or other gaps as required. An additional interface
card provides for other Inputs such as real time from a digital clock.
10
-------
During 1974, data were recorded on a Kennedy Model 1400-1R magnetic tape
recorder at IBM low density of 200 bits per inch. During 1975 and 1976 data
were recorded on a Kennedy Model 9230 which writes at a higher speed and high-
er density of 800 per inch. Telephone modems at the Bay Area Air Pollution
Control District and San Jose" State University allowed each group to query
the tower on an operational basis.
AIR POLLUTION INSTRUMENTATION
Ozone was measured by Dasibi Model 1003 AH Ozone Monitors which operate
on the principal of ultraviolet radiation absorbtion. The monitors were
housed in thermostatically controlled boxes inside of the west tower leg with
0.25 in ID tubing extending outside the tower.
Prior to installation each summer the instruments were calibrated by the
Air and Industrial Hygiene Laboratory (AIHL) of the State of California De-
partment of Public Health (1974 and 1975) or by the California Air Resources
Board (1976). The AIHL calibrations used the ARB method using 2% neutrally
buffered KI solution or by comparison with calibrated Dasibis.
During 1975 a study by the Ad Hoc Measurement Committee of the California
ARB (1975) found that the ARB calibration procedures produced ozone readings
that were high by about 25 to 30%. The EPA method, being similar to the ARB
method, also read high. We decided that for a better data comparison, and
since at the time of the 1975 calibration there was still some uncertainty,
the 1975 calibration was done by the ARB "old" method. The 1976 calibrations
were done using ultraviolet photometry as the absolute standard. No correc-
tions for the different calibration procedures have been applied, however, if
year to year comparisons are desired, a constant adjustment factor of 0.78
should be applied to the 1974 and 1975 data to make them conform to 1976 data.
The initial calibrations performed by the AIHL showed that 30 feet of
0.25 in ID tubing does not affect instrument readings.
Carbon monoxide was measured by ECO-lyser Model 3100 carbon monoxide
analysers. The analyser pumps air through a filter system, humidifier and
electrochemical sensor. The electrochemical sensor produces an electrical
current directly proportional to the CO concentration
from an overall reaction of
2 CO + 02 + C02
Initial lab checks showed that the instruments had a maximum span drift of
- 1 ppm CO day" and a maximum zero drift of - 1 ppm day" . Rise and fall
times averaged about 45 sec for a reading of 90% of a step change in CO.
DATA RECOVERY
While data recovery for the meteorological data package was at an accept-
able average of about 80% (Table 3), significant problems hampered air pol-
lutant data recovery. An hourly average value was considered reliable if at
least six five-minute instantaneous values were recorded. Tables 4 and 5
11
-------
list the numbers of hours of each month for each tower level for which valid
ozone and CO data were recorded.
TABLE 3. METEOROLOGICAL DATA RECOVERY-TOWER (percent)
Year Month June July Aug Sept Oct
1974
1975
1976
100(24th-30th)
52
68
86
75
71
85
88
100(lst-24th)
95
90
-
49
94
TABLE 4. OZONE DATA RECOVERY (hours)
1974 1975 1976
Month Level 1356 34561345
June 2 98 136 0 67 178 - 85 59 144 111 131 166
July 71 411 305 76 401 402 79 201 132 610 287 356 244
Aug 412 327 201 201 581 224 124 32 - 490 539 434 380
Sept 535 343 439 338 115 135 91 147 - 0 357 0 230
Oct --- - 0 33 38 72 Note: DasJ T n° J"s*ru-
ment installed
TABLE 5. CO DATA RECOVERY (hours)
Month Level
June
July
Aug
Sept
Oct
1
139
452
620
575
259
1975
3
120
182
297
474
342
4
0
92
154
216
324
5
210
528
274
312
321
1
69
575
443
597
1976
3
277
408
564
0
4
241
660
612
690
5
222
659
681
690
12
-------
Data Recovery, 1974
During the presence of stratus at the tower, water was often drawn into
the air sampling lines, plugging them and causing a zero reading on the ozone
monitor. In order to test whether meteorological conditions were independent
of whether or not the ozone monitors worked, wind direction frequency distri-
butions were calculated for those hours when the ozone monitors recorded
valid observations and for hours of missing data. If the ozone monitors mal-
functioned in a random manner, then ratios of the frequencies of the two dis-
tributions should be constant over wind direction. The results showed that
the ozone monitors worked relatively more often with easterly winds than with
westerly winds. Since stratus conditions in San Francisco are associated
with westerly winds we can conclude that the data for 1974 are biased towards
clear conditions. Change of the tube configuration seemed to solve this prob-
lem.
Data Recovery, 1975
Data recovery for ozone fell to unacceptable levels in 1975 (Table 4)
due to intermittent failure of 0, readings to be recorded on the magnetic
tape. Telephone modem readings obtained from the central processing unit by
San Jose* State and the Bay Area Air Pollution Control District indicate that
the ozone monitors were working properly during times when no readings were
recorded on magnetic tape. The cause of this intermittent failure is still
baffling. Since all signals are conditioned at the remote sub-station elec-
tronics located at the instrument booms before being transmitted to the cen-
tral processing unit, it would seem that either all signals should be re-
corded or none. The only significant change in the data processing system
between 1974 and 1975 was replacement of the magnetic tape recorder, and so
the problem most likely was in the interface between ozone monitor and data
processing-tape recorder system. In addition, there is some indication that
data recovery was lower at points where measured stray radiation from trans-
mitting antennas was highest (levels 4 and 5). Whatever the problem, slowing
down the scan rate ultimately seemed to solve this problem.
Data Recovery, 1976
Air pollution data recovery for 1976, while better than that for 1975,
rose to the just barely acceptable level (Tables 4 and 5). While the 0- re-
cording problems of the previous year seemed to have been solved, ozone mon-
itors failed at a fairly regular rate. This failure rate was probably due to
the fact that the monitors were in their third season in a rather hostile en-
vironment.
Discussions with technicians of the BAAPCD indicate that one of the
causes of failure may have been the number of times each instrument was
turned off and on. According to their experience an instrument should be kept
running as nearly continuously as possible. Each time an instrument is turned
on causes a power surge. Their technicians, for example, try to remove the
optical path tubes for cleaning while the instrument is kept running. The
location of our instruments, however, required turning the instrument off,
removing 1t from its housing, cleaning the optics at the base of the tower,
13
-------
turning on the Instrument to check readings and then replacing it on the
tower housing. Therefore, the instrument was turned on at least once and
usually twice for each routine maintenance. Troubleshooting usually required
more off-on cycles.
The effects of strong radio and TV radiation and relatively high temper-
atures in aging electronic components can only be guessed.
Miller (1976) also remarks on the difficulty in getting periods when all
instruments worked properly.
14
-------
SECTION 5
MEAN SUMMER BEHAVIOR
The mean summer distribution of meteorological and air pollution param-
eters is governed during the summer and early fall by the buildup and decay
in the strengths of the Pacific High and Central Valley low pressure systems
and the resulting progression of the summer portion of the California monsoon.
Local heating and cooling patterns produce diurnal land-sea and slope winds
which interact with this monsoon flow and with modifications of this flow
produced by synoptic scale pattern changes. This section will review some
relevant aspects of previous studies and summarize measurements of the mean
behavior of meteorological parameters at the Sutro Tower.
SURFACE PATTERNS
Schroeder, et al. (1967) and Root (1960) reviewed the summertime climate
patterns of the central California area. The Pacific Coast monsoon slowly
transports mass onshore in both the marine and inversion layers from late
spring until fall. The Pacific High and the Central Valley heat low produce
a maximum onshore pressure gradient of about 1.9 mb per 100 km during July.
Figure 6, from Root (1960) shows the greatest difference in mean maximum tem-
peratures between Sacramento and San Francisco as well as the strongest after-
noon winds occur in July. These measures of the strength of the monsoon are
nearly as strong in August and decrease significantly in September.
Superimposed on the onshore monsoon flow is a diurnal circulation over
the Bay Area, largely caused by diurnal variations in the Central Valley-Bay
Area temperature differences. This circulation 1s, however, complicated by
local heating and cooling effects caused by the Bay itself and by the com-
plicated topography of the area. Root (1960) shows that summertime surface
wind speeds are generally lightest at about sunrise and strongest in the late
afternoon. Root also shows that the diurnal range in wind speeds along a line
from the Farallon Islands to Sacramento is greatest at the coastline and de-
creases rapidly 1n either direction.
Figure 7 (Root, 1960) shows typical summer afternoon flow patterns and
maximum temperatures for the Bay Area. Marine air reaches the Bay Area pri-
marily through the Golden Gate, but also through smaller or higher gaps in
the Coastal Mountain Range. Upon reaching the Inland side of the coastal
range the main portion of the flow diverges; the northern branch traversing
San Pablo Bay ultimately reaching the central valley via the Carqulnez
Straits. The southern fork flows mainly to the Santa Clara Valley with a
portion branching Into the Livermore Valley.
Fosberg and Schroeder (1966) discuss the interaction of the sea breeze
15
-------
95
9C
73
270
E
£
65
60
SO
Socromento
v Mean Maximum
\ Temperature
I
/Son Francisco t
-Mean 4!30 RM.\-
Wlnd Speed \
20
£
z
10
Jan. F«b. Mar. Apr. May June July Aug. Sept Oct. Nov. Dec.
Figure 6. Mean maximum temperature for Sacramento
and San Francisco and mean afternoon wind speed
for San Francisco (Root, 1960).
with the synoptic situation and illustrate with one case study. When upper
air features showed a ridge or fairly flat contour gradient and a lobe of the
subtropical high at the surface extending inland over the Pacific Northwest,
a "warm sea breeze" occurred. As they describe the situation, a sea breeze
front, characterized by a wind shear line with cool air seaward and warm air
1nUnd, begins to move Inland from the coast after sunrise. By about 11 PST,
however, surface heating modifies the leading edge of the marine air producing
a quasi-stationary temperature gradient zone and a wind shear line which con-
tinues to penetrate Inland. No attempt was made to establish a mechanism for
the Inland penetration of the shear line after the density discontinuity has
16
-------
Figure 7. General-
ized afternoon flow
pattern of marine
air in the San
Francisco Bay Area
under typical sum-
mer conditions.
The observed tem-
peratures at 1630
PST are entered to
illustrate the
cooling effect of
the sea breeze.
(Root, I960)
BARB 5 KNOTS
BARB 10 KNOTS
been removed.
Figure 8 (Fosberg and Schroeder, 1966) shows cross sections, running from
Napa (Station 24, Figure 3) to the coast, of the west wind component and tem-
perature for a time when the wind shift line and temperature gradient were
coincident. Winds were light with slight easterly components above the in-
version base. Ahead of the sea breeze front within the marine layer westerly
wind speeds decreased rapidly.
(mi ltw«f Nopo'
Figure 8. East-West cross sections of temperature (left) and west wind com-
ponent (right) 10 PST, 15 August 1962. (Fosberg and Schroe'der, 1966).
17
-------
UPPER AIR PATTERNS
The July 1972 mean morning and afternoon Oakland and San Jose soundings
(Figure 9) shows typical summertime inversion conditions over the Bay Area.
The base of the inversion over Oakland is highest at 04 PST and lowest at 16
PST. At 04 PST the inversion base over San Jose is lower than over Oakland,
while at 16 PST surface heating has essentially destroyed the inversion layer
over San Jose.
Wind directions in the marine layer over San Jose indicate the typical
surface layer diurnal wind reversal. If the top of the morning marine layer
is defined as the point at which the mixing ratio begins to decrease rapidly,
then the morning marine layer winds are down valley (about 140 deg.). While
the afternoon marine layer is less distinct, there is a nearly constant mix-
ing ratio below about 550 m with a decrease above. Winds in the first about
600 m are upvalley (about 310 deg.). This oscillation is produced by a com-
bination bay-land breeze and slope breeze circulation.
TEMPERATURE
Miller (1976) has reported on some aspects of the mean behavior of the
inversion layer over San Francisco as determined by measurements made from
the Mt. Sutro T.V. tower. The average inversion base height is about 300 m
ASL with a mean temperature Increase above measurement Level 2 of 2.2 C/100 m.
Temperature gradients as large as 12C/20 m have been observed just above the
inversion base.
Figures 10 a, b, c show typical July, August and September temperature
time sections at the Sutro Tower. Isotherms are highest between about 02-06
PST and lowest at 16-18 PST. The inversion base on individual days follows
the same pattern. The mean diurnal temperature waves (Fig. 11) show that the
diumal temperature range is greatest at Levels 5 and 6 (4-6C) and least near
Level 2 (1.5-2C).
WINDS
Figures 12, a, b, c show the average diurnal wind speed fluctuations for
July-September 1974. The maximum wind speeds occur at 18-19 PST and the
minima at about 09 PST at each level. Level 5 at about 100 m above the inver-
sion base has the greatest average wind speed as well as the largest diurnal
range. Both the average speed and diurnal range decrease from July to Sep-
tember. 1975 (Miller, 1976) and 1976 wind patterns (not shown) show similar
patterns.
1974 summer wind roses (F1g. 13) show that winds are from the west 30 to
50 percent of the hours at all measurement levels. Winds at Levels 1-3, typ-
ically in the marine layer are usually slightly south of west, while those at
Levels 4-6, typically 1n the Inversion layer are typically more north of west.
Wind speeds from the westerly sectors average between 4.5 to 6.5 m sec"
except near the surface. Average speeds from the easterly sections are about
1 to 3 m sec .
18
-------
M3 (g/kg)
4 6
28
(297)1
(301)\
(307)|
(313)/
(320)/
("WSJ,/.
322) A
(328j((270)
(330> \
(332)/ |
V(257)
Figure y. Oakland (OAK) and San Jose State (SOS) mean soundings for July
1972. Temperature (T), mixing ratio (MR), wind speed (V) and wind direction
(in parentheses) profiles for morning (upper; 04 PST, OAK; 0630 PST, SJS) and
afternoon (lower; 16 PST, OAK; 12 PST, SJS).
19
-------
500 -i
450-
400
350
300'
2 13 14 |. I* 19 13 14
250
500 i
450
400 -
350- -
300- -
250
TIME (PST)
Figure 10. Time-height sections of hourly average temperature, Mt. Sutro
Tower, July (upper), August (middle) and September (lower), 1974.
20
-------
24
22
20
18
16
14
12
10
\
\
\
/ ^.
\
\
4 -.
3 ^ xs
2/^.
>^-sN--^
12
TIME PST
15
18
21
24
Figure 11. Hourly average temperatures, Mt. Sutro T.V. Tower, September
1974. Numbers at right Indicate measurement level.
VERTICAL VELOCITIES
Figure 14 shows the average vertical velocity profiles for July and
August 3974 (Miller, 1976). Large downward vertical velocities of 40 to 100
cm sec" and upward velocities of 30-60 cm sec" indicate the strong vertical
convergence within the Inversion layer. Miller (1976) attributes the verti-
cal velocity pattern to a quasi-stationary standing wave with its trough nor-
mally lying a short distance east of the tower. He speculates that the wave
has its greatest amplitude about 18 PST, when the inversion base is lowest,
and smallest amplitude about 06 PST, when the inversion base 1s highest.
21
-------
WIND VECTORS SUTRO TOWER
1974
r 5OO
iiiiih 250
68 K> 12 14 16 18 20 22
X X X X X /
/ . / /
IIIIIIIIIIII I
18 20 22
Figure 12. Monthly average vectors of horizontal wind Mt. Sutro Tower,
July (top), August (middle), and September (bottom), 1974.
22
-------
305
LEVEL 3
% FREQUENCY
ond m/MC
303
3-1
% FREQUENCY
and m/sec
LEVEL Z
275
10 ^ % FREQUENCY
and m/s«c
% FREQUENCY
10-| % FREQUENCY
ond m/iec
Figure 13. Wind roses at Sutro Tower, summer 1974.
frequency, dashed lines represent wind speed.
Solid lines represent
23
-------
.120 -100 -K>
40
Figure 14. Mean vertical velocity distributions at Mt. Sutro Tower for July
(upper) and August (lower) 1974. Times of observation (PST) indicated on
profiles.
MEAN OZONE BEHAVIOR
Surface Distribution of Oxidants^
Figure 15 (BAAPCD, 1976) shows the mean number of days with oxidants ex-
ceeding 8 pphm. The greatest number of NAAQS excesses occur in the inland
Livermore Valley with the eastern side of the Santa Clara Valley experiencing
nearly as many excesses.
Ozone Frequency Distributions
Figure 16 shows the co-cumulative frequency distributions of the hourly
average values of ozone concentrations for the summers of 1974 through 1976.
Note that the 1974 and 1975 data were obtained using the "old" ARB standard
of calibration, while 1976 data were obtained under the "new" calibration
method. Some ramifications of this difference will be discussed below.
Table 6 gives values of the cumulative frequency distribution for 1974.
24
-------
TABLE 6. FREQUENCY DISTRIBUTION OF HOURLY AVERAGE OZONE CONCENETRATIONS,
SUTRO TOWER, 26 JUNE to 24 SEPTEMBER, 1974.
...«»*»«************
« PEHCENT OF SAMPLES > STATED CONCENTRATIONS
LEVtL 1 35 6 COUKT
0
5
10
15
20
<*'>
rJO
i.3b
c»0
<«?>
«<*()
^H
cbU
*b5
<-fO
rfb
cttl)
*Hb
r"»0
i«)S
3UO
3Ub
310
31*
J*0
3?*
330
.M*
J»0
J»*
JbO
10(1.00 1
96.96 1
87. 55 1
73.73 1
56.06 \
44.02 1
36.57 1
29.80 1
10.10 1
11.96 1
«. 14 1
4.80 1
3.24 !
2.45 1
1.86 1
l.«7 1
.88 1
.69 1
.29 1
.20 1
.id 1
.20 1
.20 1
.20 1
.20 1
.10 1
.10 1
.10 1
.10 1
.10 1
.10 1
.10 1
.10 1
.1(1
.10
.10 1
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
,lu
.10
.10
.10
.10
.10
.10
.10
.10
.10
.10
0
0
0
0
0
0
0
0
0
0
I 100.00 1
93.70 «
77.11 «
1 66.55 1
55.74 «
1 46. 4/ »
1 41.62 1
36.77 1
1 ?6.30 »
1 19.57 1
1 16.60 »
1 13.36 t
I 11.66 1
I 9.70 1
» e.es >
* 8.00 1
( 7.3i 1
1 5.87 1
1 5.36 1
I 4.34 »
t J.3.9« 1
k 3V. 2« 1
> 30.90 1
k 25.3* 1
1 21.72 1
b 18.76 1
b 16.54 1
k 14.97 1
k 13.77 1
k 12.29 1
k 10.35 1
k 8.50 1
k 6.75 1
k 5.64 1
k 4.71 1
k 4.16 1
k 3.60 1
t 2.87 1
k 2.1J 1
k 1.76 1
t 1.39 1
t 1.11 1
t .92 1
t .83 1
t .65 1
t .46
1 .46 1
i .46
t .37
» .37
t .37
» .37
Ik .28
1 .28
» .20
* .20
1 .10
k .0V
i .09
Ik .09
Ik .09
» .09
ik .0*
» .09
i 0
Ik 0
Ik 0
i 0
Ik 0
K V
i 0
k o
t 0
Ik 0
Ik 0
Ik V
Ik 0
» 0
k 0
» 0
Ik 0
Ik 0
1 100.00 1
1 77.22 H
1 74.80 1
1 73.34 «
1 71.41 1
1 70.76 1
k 68.98 1
1 64.46 1
1 57.03 *
1 47.50 1
1 40.55 1
1 33.60 1
1 28.59 1
1 22.78 1
1 20.68 1
1 18.90 t
1 16.32 1
k 13.89 i
1 12.92 1
k 11.31 1
k 9.21 1
1 7.59 1
k 6.30 !
k 5.33 1
1 4.68 1
k 4.20 1
k 3.39 1
1 2.42 t
k 1.94 1
k 1.62 1
1 1.13 1
k 1.13 1
k 1.13 1
1 .97 1
« .65 1
i .65 1
1 .65 1
1 .65
1 .48 1
II .32
1 .32
1 .16
1 .16
II .16
1 0
1 0
1 0
1 0
II 0
1 0
II 0
II 0
1 0
1 0
1 0
1 0
1 0
k o
k 0
k 0
k o
k 0
k 0
k 0
k 0
k 0
k 0
k 0
k o
k o
k 0
1
1
1
t
1
1
1
1
1
t
i
1
1
k
k
k
k
k
1
k
k
k
k
k
k
k
1
k
k
i
t
k
1
1
1
[
I
1
1
1
I
1
^
II
1
1
1
1
t
1
k
t
1
1
I
k
k
k
k
k
k
k
k
k
k
k
k
«.««**"
CAit COUM 10*0 1»TS »»* *»
«*«««««*****
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/«
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
NX.
N/A
N/A
N/«
»»**
>
»*
25
-------
MAP OF THE
SAN FRANCISCO BAY AREA
1970-74 MEAN DAYS
WITH OXIDANT > .08 ppm
B 1000-2000
2000-3000
>3000
TERRAIN HEIGHT (II.)
122°
30'
Figure 15. Mean number of days during 1970-1974 with oxidants exceeding
8 pphm (BAAPCD, 1976).
For each year the distributions show that concentrations in the inver-
sion layer (Levels 3-6) are higher than those within the marine layer. The
03 concentrations within the inversion layer exceeded the 8 pphm NAAQS be-
tween 6 and 25% of the hours recorded during 1974 and 1975. The distribution
for 1975 overestimates the frequency of hours exceeding the 8 pphm level due
to the recording difficulties mentioned in Section 4. Inspection of the
monthly summaries for that year indicate that during periods of high 03 con-
centrations data were more consistently recorded, possibly indicating that
instrument voltages above some threshold value made it through our electronic
26
-------
oaisnrov
: if) » K) M
i. -. i I ' I
in * to » >
UJ UJUJ UJ
_1 _I-J -I
II II II II
e 4 »
*s
.oS
f~ UJ
u
82
Sg
*I
*fc
UJ
aS
-O
iri i i i i i 1 1 1 r
2 in * ro evj « g
o o aaisnrov »
~ ^ "» m ~ -: °
^ /
(WHdd) 0
a>
O)
in
~o>
r
o
cvj
8
o a
to u
-s|
~ o
O I-
10 S
CM UJ
a.
-O
* 10
0)
3 O
«r i-
<4- 00
0) +J
> m -
r- VO
u ai r>.
M 3
» »r- O
U-T> »-
(WHdd)
27
-------
maze. Thus the 138 measured hours of ozone concentrations exceeding 8 pphm
at Level 5 during 1975 compares favorably with the 163 hours at that level
during 1974, but represents a higher percentage of hours of recorded data.
Inspection of Table 7 below indicates that at least for 1974 the percentage
of the hours exceeding the 8 pphm NAAQS at selected BAAPCD surface stations
is not too different than 1n the inversion layer at Mt. Sutro TV Tower.
TABLE 7. PERCENT OF HOURS EXCEEDING 8 PPHM
OXIDANTS AT SELECTED BAY AREA STATIONS, JULY-SEPTEMBER 1974
Station Livermore San Jose (4th St.) Los Gatos Walnut Crk,
hrs >8 pphm 21 19 16 6.0
Some indication of the consequences of the adoption of new ozone monitor
calibration standards by the California ARB between the summers of 1974 and
1975 can be seen from Table 8 which shows the number of hours exceeding 8
pphm at the same selected BAAPCD surface stations.
TABLE 8. NUMBER OF HOURS EXCEEDING 8 PPHM
AT SELECTED BAY AREA STATIONS, JULY-SEPTEMBER 1974 and 1975
Station Livermore San Jose (4th St.) Los Gatos Walnut Crk.
Hrs >8 pphm (1974)
Hrs >8 pphm (1975)
269
49
233
73
196
71
84
49
The ordinates on the right in Figure 15 for 1974 and 1975 reflect the
calibration adjustment factor of 0.78. This adjustment indicates that the 8
pphm NAAQS was exceeded within the inversion during 3 to 9% of the measured
hours during 1974 and 1976. As mentioned above, percent of measured hours
exceeding 8 pphm during 1975 (11X at Level 4 and 16% at Level 5) are over-
estimations, while the 38 hrs. at Level 4 and 82 hrs. at Level 5 compare rel-
atively well with surface data shown 1n Table 8.
While there are differences 1n detail, there 1s some indication that high
concentrations at Level 5 (about 100 m above the mc»n inversion base) occur
more often than at other levels.
28
-------
The distributions shown in Figure 16 seems to fit the log-normal distri-
bution at concentrations above about 2-3 pphm. A definite break occurs in
some distributions; Level 6 during 1974, for example, could well be described
by two straight lines on the log-normal distribution. Such a plot suggests
sampling from two different populations. Such a distribution may occur due
to the fact that CL concentrations are controlled both by wind speed and tem-
perature. Since wind speeds are ordinarily log-normally distributed and tem-
peratures normally distributed, the 03 distributions could be composed of
these two distributions.
Mean Diurnal Variation of 03
Figure 17 shows the mean diurnal variation of 03 for September, 1974, a
month of consistently valid ozone data. Other months show similar patterns.
During the first six hours of the day 03 concentrations at Levels 1, 3
and 6 fluctuate near 3.5 pphm, while concentrations decrease at Level 5 from
6 pphm to about 5. Near sunrise concentrations at Level 6 increase, those at
Level 5 remain constant and those within the marine layer decrease somewhat,
perhaps due to morning emissions of 03-destroying NO . By about 09 PST con-
centrations at all levels increase until a maximum of 8 pphm is reached near
15 PST.
A secondary minimum occurs in the early evening, the time of minimum
occurring later with height, at about 17 PST at Levels 1 and 3 and about 19
PST at Level 6. A secondary maximum occurs near 21 PST, with Level 5 record-
ing the highest values. Similar nocturnal maxima have been reported by
Teichart (1955, see Geiger, 1965, p. 134) and by Perl (1965, see Reiter, 1971,
p. 149).
03 Variation with Wind Direction
Wind direction frequency distributions calculated for various 03 con-
centration intervals would be biased by the prevailing westerly wind direc-
tion. Figure 18 shows the ratio of percent occurrence of wind direction with
0, concentrations greater than or equal to 8 pphm to the percent occurrence
of wind direction with 0Q concentrations less than 8 pphm for the summer of
1974. 3
At Levels 5 and 6, high ozone occurs most frequently with northerly
through easterly winds. The pollution rose for Level 3 is not as statisti-
cally significant, but is included for completeness. High 0., concentrations
at Level 3 occur primarily with northwesterly directions witn a secondary
occurrence with easterly winds.
A break-down of the approximately 15% of the measured hours above 8 pphm
at the upper measurement levels 1s not statistically warranted. However,
wind direction frequency distributions for hours with 03 above 8 pphm were
calculated for each four-hour period during the day. Figure 19 shows that
high 0, concentrations occur with easterly directions primarily before noon
and with westerly directions primarily between sunset and midnight.
29
-------
8
to
O
\
i i i i i i
i i i i i i I I i i i i i i i
9 12 15
TIME PST
18
21
24
Figure 17. Mean diurnal variation of ozone, Sutro Tower, September, 1974.
10
LEVEL 6
Figure 18. Distribution of the ratio of wind direction for hours with 0-
> 8 pphm to those with CL < 8 pphm. Sutro Tower, Summer, 1974. Calms not
Tncluded.
30
-------
0-3 i
4-7-
C/J
0.
I
2
t-
8-lH
12-15-
16-19-
20-25
i
0
~30 00 90 120 150180 210
WIND DIRECTION-DEGREES
240 270 300 330
Figure 19. Diurnal distribution of wind direction percent frequency for
hours with Oo concentrations ^8 pphm. Level 5, Sutro Tower, summer 1974.
Auxiliary Data - Quillayute, Washington
Ludwick, et al. (1976) reported on simultaneous surface measurements of
03 and radionuclides of upper tropospheric and/or stratospheric origin at
rural Quillayute, Wa. They found that high hour 03 concentrations of 5 to 6
pphm occurred on days of high radionuclide measurements, indicating strato-
spheric origin of 03 at this location.
Co-cumulative frequency distributions of the hourly 0, concentrations at
Quillayute calculated from data supplied by Ludwick show tnat the 90-th per-
centHe concentration (Table 9) for September 1974 was the highest of the
months June-October, 1974.
TABLE 9. MEDIAN AND 90TH PERCENTILE HOURLY AVERAGE 0, CONCENTRATIONS (PPHM)
QUILLAYUTE, WA. 1974 J
Month
Median
June
3.1
4.0
July
2.6
3.5
August
2.1
3.5
September
2.9
4.6
October
2.4
4.0
June-October
2.6
4.0
31
-------
Figure 20 shows that the frequency distributions for measurement Level 1
at the Sutro Tower for the summer 1974 and for Quillayute for the period
June-October, 1974 are similar at low concentrations. At high concentrations,
the effects of pollution are shown by the slightly higher concentrations meas-
ured in San Francisco at the high end of the distribution.
lO-i
5-
4
3
1 H
ft
ol"o.5H
0.1
10 20 30 40 SO 60 70 80
PERCENT HOURS EXCEEDED
90 95
Figure 20. Co-cumulative frequency distributions of ozone at Quillayute, Wa.
(triangles) and Level 1, Sutro Tower (dots), Summer 1974.
High radionucllde and 03 concentrations at Quillayute occurred just prior
to the two September Bay Area case studies to be discussed in Section 8 below.
This coincidence led to preliminary studies of possible contribution of strat-
ospheric 03 to Bay Area air pollution episodes discussed below.
Mean Carbon Monoxide Behavior
CO concentrations 1n the Bay Area are typically lower during summer
months than other times of the year.
Carbon monoxide concentrations measured at the Sutro Tower are approxi-
mately log-normally distributed (Figure 21). The difference in distribution
details between levels 1s probably within the measurement error of the in-
struments over most of the range. However, it appears that at the higher con-
centration end of the distributions there 1s some tendency for concentrations
to be higher at the lower measurement levels, as opposed to the pattern for
03> The distribution for Level 1 for 1975 appears to be anomalously low.
32
-------
20
I 2~
o
u
0.3-
o.i
20-
SUTRO TOWER, 1975
»LEVEL 5
» -LEVEL 4 10-
* -LEVEL 3
-LEVEL I
5
4
3
2-
1-
0.5
10
20 30 40 50 60 70 80
PERCENT HOURS EXCEEDED
"90 OJ 2
SUTRO TOWER, 1976
LEVEL 5
* ' LEVEL 4
10
20 30 40 50 60 70 80
PERCENT HOURS EXCEEDED
90 95
Figure 21. Co-cumulative frequency
and 1976. (1 ppm = 1150 mg m J)
Mean Dirunal CO Behavior
distributions of CO, Sutro Tower, 1975
Figure 22 shows the mean diurnal CO behavior for September 1976, a month
which appeared to have consistently reliable CO data. At Level 1, the morning
and afternoon rush hour peaks appear to some degree. The location of the
tower, however, is far enough away from traffic so that these peaks are not
pronounced. At levels 4 and 5, CO concentrations decrease during the daytime
hours, probably in response to the lowering of the inversion base during the
day.
33
-------
<» = LEVEL 5
* = LEVEL 4
. =LEVEL I
T~
18
I
ao
10 12 14
TIME (PST)
16
22
Figure 22. Hourly average CO concentrations, Sutro Tower» September 1976.
34
-------
SECTION 6
CASE STUDY OF JULY 22-26, 1974*
by
Stephen H. Holets
During the Summer of 1974, four air pollution episodes occurred in the
Bay Area in which Smog Advisories (a forecast of >^ 20 pphm 0_) were issued.
The episode of July 22-26, 1974, analyzed here, has been divided into the 0-
"buildup" (July 22, 23), "peak" (July 24), and "breakup" (July 25, 26) pe- J
riods on the basis of changing meteorological and ozone patterns. This chap-
ter compares mean and episode states of meteorological and ozone parameters.
SYNOPTIC PATTERN
The main 700 mb features for July (Figure 23; Wagner, 1974) are the
trough located near the Pacific Northwest Coast, the ridge over the Central
United States, and the small pressure gradient between the Great Basin and
the East-Central Pacific. The trough along the Pacific Northwest Coast de-
veloped at the end of June with the decline of the mean mid-latitude west-
erlies (Wagner, 1974). Retrogression and weakening of the Eastern Pacific
trough, coupled with retrogression of the ridge over Central United States
and weakening of the Subtropical High, produced weak pressure gradients over
California and a wide area of zonal flow by the end of July (Figure 24).
SURFACE PATTERNS
Temperatures^
Large maximum temperature gradients between the maritime coastal areas
and the sheltered inland valleys of the Bay Area dominate the mean July pat-
tern (Figure 25). Isotherm troughs in the San Pablo Bay and San Francisco
Bay-Eastern Santa Clara Valley show the strong marine air penetration through
the Golden Gate and San Bruno gaps. The coastal range protects the western
part of the Santa Clara Valley from marine air.
Maximum temperatures were above the mean during the July 22-26, 1974 air
pollution episode (Figures 26a-c). The inland valleys heated by 8 F to the
Q "peak" period with equivalent decreases during the "breakup" stage.
* This section has been abstracted from "A Case Study of High Ozone in
the San Francisco Bay Area" submitted in partial fulfillment of the degree of
Master of Science at San Jose State University.
35
-------
Figure 23. Mean 700-mb height contours (dekameters) for July
1974. After Wagner (1974).
Figure 24. 700-mb height contours (dekameters) for 23-27 July
1974. After Wagner (1974).
36
-------
o>
tO
i-
0)
a.
cnu_
^ o
37
-------
to
s_
O)
O-
E W3
r- Od
X
(O C
S O
c
O
21:
=J -p
CT> V)
O)
3
-IJ
rtJ
S-
OJ
i- r
O >,
e -3
E **
r- OJ
X
(O C
s: o
S- S-
3 4->
O) trt
38
-------
Winds
Typical early morning and afternoon flow patterns for the three periods
of the July 22-26 episode are presented 1n Figures 27-29. Relatively strong
flow occurred through the Carquinez straits during the early morning, with
drainage flow in the Llvermore, Diablo, and Sattta Clara Valleys. By the
afternoon, marine air persisted throughout the Bay Area. The 0~ "peak" stage
displayed the highest degree of air stagnation.
Oxidant
Mean 1968-74 high-hour oxidant concentrations surpass the federal stan-
dard in the Central, Livermore, and Santa Clara Valleys (Figure 301).
Maximum high-hour oxidant concentrations during the day prior to this
case study were 9 pphm at San Jose and Livermore. At the beginning of the
"buildup" period, surface oxidant concentration surpassed the federal stan-
dard throughout much of the Bay Area (Figure 31a). On July 24 a Smog Advi-
sory was issued by the BAAPCD for the Livermore and Santa Clara Valleys where
oxidant Increased to about 24 pphm (Figure 31b). A band of middle and high
clouds formed across Central California on July 25, attenuating solar energy
and reducing photochemical smog production. Westerly flow dominated the
"breakup" stage as surface oxidant concentrations decreased substantially.
Only the Smog Advisory region of July 24 surpassed the federal standard
during this stage (Figure 31c).
UPPER AIR PATTERNS
Subsidence Inversion
The Oakland 0400 PST temperature profiles for the July 22-26 episode
indicate a lowering of the inversion by July 24 with lifting thereafter, and
a substantially warmer air mass than the mean (Figure 32a). Similar features
occurred over San Jose (Figure 32b).
The mean diurnal temperature distribution for July (1974) on the Mount
Sutro Tower (Figure 33a) shows that the inversion undulates with a 24 hour
period; lowering from rtildday to 2000 PST, then rising to its highest point
at 0600 PST. Hourly average temperature time-height sections at the Mount
Sutro Tower on July 22, 24, and 26 show high temperatures and a stronger In-
version intensity than the mean, waving in the inversion, and a general
lowering of the Inversion through July 24 and then lifting thereafter (Fig-
ures 33b-d).
Winds
Westerly flow persists during the early morning hours over Oakland in
July while San Jose typically has easterly drainage flow to 600 m with
westerlies above throughout the summer and early fall months (Figures
31a-b). During the afternoon, west winds dominate both locations.
Onshore flow prevails through the San Francisco gap as shown by the
39
-------
CD
O
CO O
gl
at *
0.0
-------
o
o
ro o
cu
t/1
* ^^
* E
r-x
o> in
*"" n
2u
l~3 O1
C
«
o
*»
C O
S- O)
0) co
4-> -^.
4-> E
(O
i"
i ^a
u_ s-
(O
U. Q.
O
o
3
3
CM
O (O
O
S_ II
(O «
Q. U
II
.0
0) -Q
i-
rj i
O)r
r- 3
U. U_
41
-------
o
o
CO O
i ^,
M E
CVJ
o
a. o
-------
C O)
o x: c
r- -P O
»-> !-
3 -P 4->
.a c
4J 0) C
Q. O) C
!-> C X
C O
to -o
-O QJ S-
- C 3
X T- O
s_ a> -c
3 -o o>
o c <-
JC +J
en i/>
r- ^J- i-
3C P»- T-
01 M-
CO 3
O «»-
O) O
i- CM
3 CM O)
U- O +»
r- E 0>
O 3 'p-
E "O -P
.G 0) to
o. c s.
£;'Z'£
s-
P O> (o co
0>f 0) l-
>»
-------
>>
3
VO
CM
Q-
>>
*3
r?
**
CM
s_
o
-p
o.
0)
(J
X
o>
(O
CO
m
-------
mean July (1974) westerly component at the Sutro Tower (Figure 34a). During
the July 22-26 episode, however, the easterly component increased steadily
from July 21-24 and decreased rapidly during the "breakup" period (Figures
34b-d). Although westerly flow occurred on the upper tower levels on July 23
and 24, the speeds were low.
Ozone
In general, hourly average ozone concentrations increased with height at
the Mount Sutro Tower during the Summer of 1974 (Figure 35a). At Level 5 (L5)
the NAAQS was exceeded 149 hours during June 26-September 24, 1974, forty per-
cent of these hours during July 23-25. Figures 35b-d show that ozone concen-
trations increased from 03 "buildup" to the "peak" periods, reaching a maximum
hourly concentration of 25 pphm on L5 at 1300 PST on July 24. By the end of
the "breakup" stage, ozone had decreased to mean values.
On July 24 thre ozone layer aloft lowered to within 88 m of the earth's
surface over San Francisco (Figure 36) and appeared to extend throughout the
subsidence inversion, as exemplified by the ozone profile over Hayward (Fig-
ure 37). The lower part of the subsidence inversion (200-400 m AGL) con-
tained an ozone peak of 28 pphm. Since surface monitoring stations within
the Hayward area recorded less than half the ozone of the polluted layer
above, strong downward diffusion did not occur.
In summary, the trough off the West Coast of the United States and ridge
over Central Canada retrograded, thus establishing weak pressure gradients
between the Eastern Pacific and Great Basin regions. Surface temperatures
and oxidant concentrations increased sharply from mean values during the
"buildup" stage. By the time of the 03 "peak" period, inland valley surface
temperatures were 8-16 F above normal with maximum mixing depths of 1800 m,
the inversion base lowered to the earth's surface, surface winds were rel-
atively light with early morning southeast drainage flow in the lower Santa
Clara Valley and anomalous offshore flow through the Golden Gate Gap, and
maximum surface and inversion oxidant concentrations peaked near 17 pphm
above normal. As westerly flow became persistent during the "breakup" stage,
temperature and oxidant values returned to the mean state.
An ozone layer of 24 pphm in the inversion lowered to within 88 m of the
Sutro Tower base where only 6 pphm were recorded. Continued lowering of the
inversion or its partial destruction with downward mixing would quickly pro-
ject ozone concentrations above the federal standard.
EVOLUTION OF THE JULY 22-26 EPISODE
The evolution of the July 22-26, 1974 air pollution episode is modeled
schematically in Figures 38a-g. Although much interpretation was required,
the model conforms to the available data.
On July 21, the maximum oxidant recorded in the Bay Area was 9 pphm. The
National Weather Service 850 mb analysis (not shown) and available upper air
winds along coastal California indicate the possibility of an isentropfc tra-
jectory from the Los Angeles Basin to San Francisco. This information,
45
-------
28?-
MR (G/KG)
4
T(«C)
V(m/Mc)
Figure 32a. Oakland temperature (T), mixing ratio (MR), and
wind speed (V) soundings for 22, 24, and 26 July 1974, 0400
PST. Wind directions are in parenthesis. Mean soundings are
for July 1972.
289-
MR (G/KG)
20
£12
T CO
V (m/MO)
Figure 32b. San Jose State University temperature (T), mixing
ratio (MR), and wind speed (V) soundings for 22, 24, and 26
July 1974, 0600 PST. Wind directions are in parenthesis. Mean
soundings are for July 1972.
46
-------
30C-
OL-
280
12 M
TIME (PST)
20 22
Figure 33a. Mean July (1974) hourly average time-height tem-
perature (°C) section at the Mount Sutro Tower, San Francisco.
500-
450-
\
300-
I I I I I I
24 . 26 . 28
t iii
V V\\
1 16 I A I I I 1
10 12
TIME (PST)
14 16
Figure 33b. Time-height section of hourly average temperatures
(°C) at the Mount Sutro Tower, San Francisco on 22 July 1974.
Dashed lines represent Interpreted data.
47
-------
500
_ I I I I I I I I I
i I I I I I I I
i%. i^Fmyi
INVERSION BASE
TIME(PST)
Figure 33c. Time-height section of hourly average temperatures
(°C) at the Mount Sutro Tower, San Francisco on 24 July 1974.
Dashed lines represent interpreted data.
BOO
-219
IS IT
TIME (PST)
*
Figure 33d. Time-height section of hourly average temperatures
(°C) at the Mount Sutro Tower, San Francisco on 26 July 1974.
Dashed lines represent interpreted data.
48
-------
SCO
0 2
22
Figure 34a. Mean July (1974) hourly average west wind (m/sec)
time-height section at the Mount Sutro Tower, San Francisco.
900
I I I I I I I I I I I I I I
20 22
TIMEtPST)
Figure 34b. Time-height section of hourly average west wind
(m/sec) at the Mount Sutro Tower, San Francisco on 21 July 1974.
Shaded region represents an easterly component.
49
-------
900
480
400
300
300
0 ID 3D 40 4.850 5D 4jO
219
178
45
10 12 14
TIME (PST)
16 18 20 22
Flgure 34c. Time-height section of hourly average west wind
(m/sec) at the Mount Sutro Tower, San Francisco on 24 July
1974. Shaded region represents an easterly component.
500
8 10 12 14 16 IB 20
Figure 34d. Time-height section of hourly average west wind
(m/sec) at the Mount Sutro Tower, San Francisco on 26 July
1974. Shaded region represents an easterly component.
50
-------
. LEVEL 6 : 219 m AOL / 473 m MSL (619 HOURS)
LEVELS' 178 mASL/432 m MSL (1062 HOURS)
LEVELS' 88 in AOL/342 m MSL (1179HOURS)
LEVEL I 6 mAGL/294 m MSL (I020HOURS)
TIME (PST)
20 22
24
Figure 35a. Mean hourly average ozone time section at the
Mount Sutro Tower, San Francisco (June 26-September 24, 1974),
cc
20
18
tM
14
12
10
8
«
4
2
219m AOL /473m MSL
178m AOL /432m MSL
.. . ftftm AOI /^A9 m UfU
^ OQ m MW_ f O^£ m fHOL.
^
I 1 1 1 1 1 1 1 1
54486
_
/
/
f
j-
/i
/x' -
/^ "^i
-^T
.... ''
1 l" IIIIIIIIJIJI
IO 12 M 16 18 ZO 22 Z
TIME (PST)
Figure 35b. Hourly average ozone time section at the Mount
Sutro Tower, San Francisco on 23 July 1974.
51
-------
i rnr~i r i \ t i T
1 1 1
30-
28-
26
22-
20-
18 -
16-
14
12
10-
8^
6-
4 -"
2
SS-
178 mAGL /432m MSL
88mAGL/342m MSL
6m A6L/2S4m MSL
1 1 ' A M-H-MH--1-1 I*
TIME (PST)
1* L i [ ^o J-
Figure 35c. Hourly average ozone-time section at the Mount
Sutro Tower, San Francisco on 24 July 1974.
i«t
22
20
18
16
14
12
10
8
6
4
2
0
^=-=^>"~^^
1 J '
52
219m A6L /473m MSL
178m AGL /432m MSL
88m AGL/ 342m MSL
6m AGL /254m MSL
A
'^^------^ ^"^.-.,
^-^ > """" ' '''''
\ -../ .,..._
1 ' 1 ' i ' L ' L ' ' ' L ' L ' J ' 1
4 8 8 15 12 W 18 18 20 29
-
^
^
-
...
1
M
TIME (PST)
Figure 35d. Hourly average ozone-time section at the Mount
Sutro Tower, San Francisco on 25 July 1974.
52
-------
"1200
1230
1300
1330
TIME(PST)
1400
1430
Figure 36a. Variation of five minute average values of ozone,
Levels 3 and 5, Sutro Tower, 24 July 1974.
53
-------
j
*
LEVEL I
-0.5
--I.O
1200
1230
1300 1330
TIME (PST)
1400 1430
Figure 365. Variation of five minute average vertical
velocities, Mount Sutro Tower, San Francisco, 24 July 1974.
490-
I I I I
1200 1290
1300 1330
TIME (PST)
I4OO 1430
Figure 36c. Time-height temperature (°C) section at the
Mount Sutro Tower, San Francisco on 24 July 1974. Dashed
lines represent interpreted data.
54
-------
2000
1800
1600
MOO
1200
1000
800
600
400
200
OZONE (pphm)
K) 15 20
25
30
0
20
22
24
26
TEMP (°C)
28
30
32
34
Figure 37. Temperature and ozone profile at Hayward,
California on 24 July 1974, 1450-1525 PST.
coupled with the fact that oxidant concentrations at higher elevations in the
South Coast Air Basin during the prior few days, suggests possible transport
of oxidants from the Los Angeles area.
Baboolal, et al (1975) found indications of aged Los Angeles pollutants
being transported some 200 km northwest to the Santa Yuez Valley. Kauper and
Nemann (1975) have documented over water transport of oxidants within the in-
version layer from the Los Angeles Basin to Ventura, about 100 km to the
northwest. The possibilities of longer distance transport to the San Fran-
cisco Bay, while not proven, seems possible and should be studied.
55
-------
The midafternoon condition during the first day of the "buildup" period
is shown in Figure 38a. The subsidence inversion was present over San Fran-
cisco and extended to the southern edge of the San Francisco Bay where it was
destroyed by surface heating. Northwest flow advected ozone (2-6 pphm) and
its precursors into the Santa Clara Valley. Surface concentrations increased
downstream due to enhanced emissions during advection and photochemical pro-
duction. In the Santa Clara Valley, where the inversion was destroyed, 0,
buildups of 10-15 pphm occurred due to decreased 03 destruction rates, the
accumulation of precursors and high temperatures.
As the inversion reformed at night, ozone below the inversion was
quickly destroyed by the earth's surface (Figure 38b). Within the inversion,
however, 0- concentrations of 5-15 pphm probably persisted as proposed by
Gloria et al. (1974). Under light southeast flow, the pollutant layer was
advected northward with the edge of the layer advancing to the San Rafael -
Mount Tamalpais area by noon. In addition, ozone increases to 10 pphm at the
Sutro Tower show the subsequent advection over San Francisco of Santa Clara
Valley generated 03.
About noon (Figure 38c), winds to 500 m AGL became northerly and the 0-
layer receded back over San Francisco. Ozone concentrations increased by 8
pphm by 1500 PST at the Sutro Tower as photochemistry of the aged pollutant
layer progressed. Once again, ozone and its precursors were advected into
the Santa Clara Valley. Pollutants accumulated in the upper Santa Clara
Valley, however, as southeasterly flow of marine air into the lower Santa
Clara Valley from the Monterey Bay Area established a convergence zone with a
13 pphm ozone gradient between the two air masses.
On July 24, this cycle was repeated (Figure 38d). The inversion re-
formed over San Jose during the night of July 23, becoming surface based due
to radiation cooling and Increased subsidence. Ozone was destroyed near the
earth's surface but persisted above. Southeast flow advected the 03 layer
from the San Francisco Bay and upper Santa Clara Valley to the northern Bay
Area, as suggested by Tower air trajectories. A high ozone pocket was ad-
vected over the Sutro Tower between 0200-0400 PST when concentrations in-
creased from 8.5 pphm to 14.5 pphm (Figure 35c). By 1000 PST the winds veered
to the northeast, producing strong offshore flow. Once again, 03 generated
in the Santa Clara Valley was advected over San Francisco. Concentrations at
the Sutro Tower rapidly increased to a high-hour value of 26 pphm and a peak
of 30 pphm at 1300 PST as photochemistry of the aged pollutant layer pro-
gressed.
By 1400 PST, northwest winds commenced, which advected the pollutant
cloud southeastward down the Santa Clara Valley (Figure 38e). As a result,
03 concentrations on the Tower decreased by 10 pphm in two hours. The ad-
vection of ozone and its precursors, plus photochemical production due to a
large mixing depth, produced 03 buildups of 15-30 pphm in the Santa Clara
Valley. Opposing southeast flow from the Monterey Bay area was conspicuously
absent during this period.
56
-------
DISTANCE (km)FROM SUTRO TOWER S J
Figure 38a. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the afternoon (13-16 PST) of 22 July 1974. Winds recorded
over Oakland are presented vectorially with (+) = 4 m/sec. Region
of 10-15 pphm ozone is shaded.
DISTANCE (km) FROM SUTRO TOWER
Figure 38b. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the morning (04-07 PST) of 23 July, 1974. Winds recorded
over Oakland are presented vectorially with (->) = 4 m/sec. Region
of 5-15 pphm 0- is shaded.
57
-------
DISTANCE (km) FROM SUTRO TOWER5 J
Figure 38c. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the afternoon (13-16 PST) of 23 July, 1974. Winds recorded
over Oakland are presented vectorially with (-») = 4 m/sec. Region
of 10-20 pphm 03 is shaded.
DISTANCE (km) FROM SUTRO TOWER
Figure 38d. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the morning (04-07 PST) of 24 July, 1974. Winds recorded
over Oakland are presented vectorially with (-») = 4 m/sec. Region
of 10-20 pphm 03 is shaded.
58
-------
Except for a four hour period during the early morning of July 25, west-
erly flow prevailed over San Francisco. Coupled with reduced photochemistry
due to the presence of middle and high clouds over the Bay Area, 03 concen-
trations decreased to those values recorded on July 23.
During the morning of July 26, westerly flow prevailed over San Francisco
and the San Francisco Bay (Figure 38f). Opposing southeast drainage flow over
the Santa Clara Valley produced a convergence zone near the southern edge of
the San Francisco Bay. Ozone concentrations at this time had decreased below
10 pphm.
By the afternoon of July 26, westerly flow prevailed through the Bay Area
(Figure 38g). Ozone concentrations northwest of the Santa Clara Valley were
only 5-10 pphm. However, the advection of ozone and its precursors, photo-
chemical production, and low 0, destruction due to large mixing depths pro-
duced 03 buildups of 10-20 pphm in the Santa Clara Valley.
The dominant westerly flow through July 27 advected pollutants out of
the Bay Area. As a result, high-hour oxidants of only 10 pphm occurred over
San Jose, thus signalling the end of the episode. High oxidant concentrations
at Chico and Redding in the Sacramento Valley on 25-27 July hint at possible
transport there from the Bay Area (Table 10).
The primary injection process of 03 into the subsidence inversion was
the destruction of the inversion during the day by surface heating and its
reformation at night by radiation cooling. Secondary pollutant injection
processes may have been the downward flux of stratospheric ozone, aircraft
emissions, and the destruction and reformation of the inversion by gravity
waves, aircraft turbulence, and sea breeze induced convergence zones. At the
beginning of the "buildup" period a 110 knot jet stream at 300 mb was located
in the Gulf of Alaska. As the jet progressed eastward, however, air trajec-
troies reaching the lower troposphere over the Bay Area were from the south-
east to southwest, thus suggesting a minimal stratospheric contribution.
Gravity waves, sea breeze convergence zones, and aircraft also appeared
to provide only small contributions to the inversion ozone buildup. Although
waving of the inversion occurred within the subsidence inversion during July
23 and 24, there was no evidence of the waves "breaking" with exchanges be-
tween the marine and inversion air masses. Although sea breeze convergence
zones occasionally occurred over the San Pablo Bay during this episode,
pollution was light, and therefore could not contribute substantially. Early
morning convergence in the upper Santa Clara Valley was typical during this
episode. Since winds were very light, updrafts penetrating the inversion
appear unlikely.
In summary, under weak pressure gradients, a "sloshing" of air pollution
beneath the subsidence Inversion top, parallel to the NW-SE Santa Clara Valley
axis, effectively trapped pollutants in the area. The Santa Clara Valley was
the region of ozone generation, since ozone and its precursors were advected
there from numerous upwind sources, and 03 destruction rates were low during
the day. Upon reformation of the inversion at night, pollutants became
59
-------
(10)
10-15 0,
(10)
«!tt
»-*> 0, ^
10-15 0,
LM_
TO
10-15 03
Jll)
NW
10 0 10
30
DISTANCE (km) FROM SUTRO TOWER
GILROY
Figure 38e. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the afternoon (13-16 PST) of 24 July, 1974. Winds recorded
over Oakland are presented vectorially with (->) = 4 m/sec. Region
of 15-30 pphm 03 is shaded-
CONVERGENCE
ZONE
(4)
- _ AVERSION BASE _
NW
STANCE (km) FROM SUTRO TOWER
90 GILROY 110
Figure 38f. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the morning (04-07 PST) of 26 July, 1974. Winds recorded
over Oakland are presented vectorially with (-») = 4 m/sec. Region
of 10-15 pphm ozone is shaded.
60
-------
DISTANCE (km) FROM SUTRO TOWER
90"GILROY HO
Figure 38g. NW-SE schematic cross section of air flow, inversion
and mixed layer heights, and ozone concentrations (in parenthesis)
during the afternoon (13-16 PST) of 26 July, 1974. Winds recorded
over Oakland are presented vectorially with (-+) = 4 m/sec. Region
of 10-20 pphm 03 is shaded.
trapped within the stable layer and persisted. The polluted layer was ad-
vected to the northwest during the night and returned during the afternoon.
As westerly flow increased, pollutants were advected out of the valley and 03
concentrations decreased below the federal standard at most locations.
The primary injection process of 0, into the subsidence inversion was the
destruction of the inversion during the day by surface heating and its refor-
mation at night by radiational cooling. Secondary pollutant injection proc-
esses were the lowering of the inversion during the day, and may have included
downward flux of stratospheric ozone, aircraft emissions, and the destruction
and reformation of the inversion by gravity waves, aircraft turbulence, and
sea breeze induced convergence zones.
61
-------
TABLE 10. SURFACE OXIDANT CONCENTRATIONS (pphm) JULY 22-31, 1974
Time
Date/PST
22
23
24
25
26
27
28
29
30
31
Time
Date/PST
oo
Cc.
23
24
25
26
27
28
29
30
31
10
8
8
7
9
7
7
7
7
8
6
10
M
5
6
7
5
8
5
M
7
11
9
8
8
10
8
8
8
7
9
8
11
M
6
6
8
8
9
8
7
10
12
9
8
7
10
8
8
8
8
10
9
12
M
6
6
12
10
11
9
7
13
13
8
8
9
9
9
9
8
8
10
10
13
M
6
6
15
14
13
10
7
14
Chi co (
14 15
8 7
7 7
8 8
9 9
10 10
9 8
8 8
M 7
9 9
10 10
Redding
14 15
5C
0
M M
8 10
7 7
14 14
13 13
12 10
10 10
6 7
13 13
76 m
16
6
6
8
8
12
8
8
6
9
9
(220
16
M
10
7
12
13
8
8
7
13
MSL)
17 18
5 4
6 5
8 6
8 7
12 12
8 9
7 8
6 4
9 9
10 9
m MSL)
17 18
5 5
8 7
8 8
11 10
12 11
11 10
9 9
5 5
11 10
19
2
3
3
4
12
6
7
3
9
6
19
6
7
11
11
9
9
9
5
9
20
1
1
2
4
11
6
5
1
7
6
20
5
6
10
10
9
7
7
6
9
21
2
2
3
7
8
5
5
1
7
8
21
3
4
9
6
7
6
3
7
5
Max. High
Hour BAAPCD
Oxidant
Concentration
18
17
25
15
15
10
10
9
10
9
Max. High
Hour BAAPCD
Oxidant
Concentration
1 O
IB
17
25
15
15
10
10
9
10
9
62
-------
SECTION 7
CASE STUDY OF JULY 22-27, 1975
by
Betsy L. Babson*
This case study occurred almost a year to the day after the one dis-
cussed in the previous section. Following the terminology of Section 6,
July 22-24 are referred to as the "intensification" or "build-up" period,
July 25 the "peak" or "health advisory alert" day, and July 26, 27 as the
"break-up" or "restoration" period. On July 25, five monitoring stations re-
ported oxidant levels of 20 pphm or greater, causing the BAAPCD to issue a
health advisory alert for the Santa Clara Valley. This section compares and
contrasts the evolution of the two July episodes studied.
SYNOPTIC PATTERN
A dynamic low centered in the Gulf of Alaska persisted throughout this
episode. High pressure dominated the western United States above 850 mb, and
a trough was located over the Great Lakes region. During the intensifica-
tion period a lobe of the Subtropical High penetrated into the Pacific North-
west to join with a Great Basin High over the northwestern United States.
On July 25, the Low in the Gulf of Alaska began to fill. At 850 mb the
Great Basin High separated from the Subtropical High as a trough of low
pressure developed in the southwestern United States. A weak front at the
surface approaching from the northwest dissipated as it crossed over northern
California at 12 GMT on July 25.
By July 27 the High over California had weakened. West-northwest winds
prevailed over San Francisco as a heat trough became well established through-
out the Central Valley.
SURFACE TEMPERATURE PATTERN
Mean daily maximum temperatures recorded during July, 1975, were similar
to those of July, 1974, and exhibited a large temperature gradient between
the coastal regions and the Santa Clara and Sacramento Valleys (Figure 39a).
On July 22, temperatures throughout the Bay Area were generally 8°F
above the mean. Within the next three days temperatures increased as much as
*Abstracted from "A Case Study o7 the San Francisco fcay Area Air Pollution
Episode, July 1975" submitted in partial fulfillment of Weather Analysis and
Forecasting class, San Jose State University, Meteorology Department.
63
-------
CTl
cu
o «
>
l/l r
«
3 tO
O OJ
x: i-
r- CO
X
03 O
3E O
CO
r-
O
-0 C
CTl (O
CO $_
U.
CU
s. c:
(O
I
as
cu
Q.
O) in
i-> i^
CD
CU r-
O
(O "
(TJ
CO
x u
to m
E-r-
U
C C
rtj
co «
10
cu cu
s- s-
3 3
O>4->
64
-------
12 degrees, reaching 100°F in the Santa Clara Valley on July 25 (Figure 39b).
During the restoration period, temperature decreases were less rapid than
rates of increase for the initial stages of the pollution episode.
SURFACE WINDS
Surface flow patterns for the July, 1975, air pollution episode were
similar to those presented in Section 6 for July, 1974. BAAPCD stations re-
ported calm conditions throughout the Bay Area at 04 PST except in the vicin-
ities of the Golden Gate gap and Carquinez Straits where relatively strong
flow prevailed during the intensification and restoration periods.
By 16 PST, wind speeds increased, and northwest winds dominated the
Santa Clara Valley.
SURFACE OXIDANT CONCENTRATIONS
During the ozone build-up period of the July, 1975, air pollution epi-
sode, oxidant concentrations reached 12 pphm in the Santa Clara and Sacramen-
to Valleys. On July 25, concentrations exceeded the Federal Standard
throughout the entire Bay Area except at San Francisco and coastal areas. A
Health Advisory Alert was issued in the Santa Clara Valley where oxidant
levels of 20 pphm were reached at the Hayward, Fremont, Alum Rock, Sunnyvale,
Mountain View, and San Jose BAAPCD monitoring stations. The highest average
hourly concentration recorded was 23 pphm at the Hayward station at 14 PST
(Figure 40). By July 27, oxidant concentrations were reduced below the
Federal Standard at all stations except Los Gatos, San Jose, and Alum Rock.
The distribution of oxidants throughout the Bay Area was similar during
both the July, 1974, and July, 1975, air pollution episodes, though concen-:
trations in the Santa Clara Valley were slightly higher in 1974.
UPPER AIR WINDS
Time sections of horizontal wind vectors at the Mount Sutro Tower are
shown for the three periods of the episode (Figure 41). Southwest winds pre-
vailed below the inversion base during the Intensification Period, while flow
was generally from the northwest aloft. Light northeast winds were recorded
at all levels during the morning of the alert day when the temperature in-
version base dropped below the Tower base. Wind speeds increased in the
evening as flow shifted from westerly to north-northwesterly between 14 and
20 PST, and the inversion base rose above 250 meters. By July 27, the in-
version base lifted above 400 meters, and winds at all levels of the Tower
were strong out of the southwest.
West-northwest winds ranged from 3 to 9 m/sec over Oakland at 04 PST on
the alert day (Figure 41). During the previous year, winds were northwest at
the surface and south-southeast aloft, while speeds varied from calm to a
maximum of 3 m sec" . West-southwest winds prevailed within the inversion
during the restoration,periods of both episodes, though winds speeds were
approximately 4 m sec" greater during July, 1975.
65
-------
MAP OF THE
SAN FRANCISCO
JULY 25, 1975
MAXIMUM HOURLY OXIDANT
CONCENTRATIONS (PPHM)
HhALTH ADVISORY ALERT AREA
(I >KOO
HMUMHIiaHT IK)
Figure 40. Maximum hourly oxidant concentrations at BAAPCD Measurement
Stations, 25 July 1975.
A comparison of the horizontal wind vectors for the July 1974 and 1975
alert days (Figures 42 and 43) shows similarities and difference between the
two episodes. Flow was north-northeast during the pre dawn hours of the 1975
alert day, whereas winds were south-southeast during the same period of the
previous year. In both cases winds were light from the northeast between the
early morning and the commencement of strong westerly winds in the mid after-
noon.
Section 6 emphasized the advection of pollutants northward from the Santa
Clara Valley. Figure 42 shows that no such advection occurred during this
period, but that the period of light offshore flow during the forenoon and
early afternoon periods were common during both episodes.
OZONE ALOFT
Time-height sections of 0., concentrations at the Mt. Sutro Tower for the
study period (Figure 44) show 03 peaks within the inversion at heights of 400
to 450 m. During the intensification period of the July, 1975, pollution
66
-------
MIXING RATIO (G/Kf>)
f !
10 i;
'\ T T
24
20
i
i 16
§
! '2
*
8
4-
l»o> )»»*»
5 10 15
TEMPERATURE (°C)
20 25
30 0 2 4 6 8 10 12 W
WIND SPEED (M/S)
MIXING RATIO (fi/KS)
2 4
10 15
TCMPMATURC (*C)
20
25
30 2 4 6 8
WIND IPEED (IH/I)
Figure 41. Rawlnsonde profiles of temperature, mixing ratio and wind speed
and direction; Oakland, CA, July 23, 25, 27, 1975 at 04 PST (top) and 16
PST (bottom).
67
-------
471.7 -
431.7
^3892
s
h-
\
V
UJ
I
298/1
259.
171.7
431.7
3692
6 H 10
If. IB
~20 22
21
29B/4
259:
' '
1 J
C
471.7 -
431.7
> ^ /
I -' I 1 L. L-
6 8 10 I? 14
_J L L_
18 20 22
340.6^
296/3
259.0
\
x-x
?.
J^ 1 1 I
~£ |O~~ 12
TIME CPST)
rt ft fr"
20 22
HOUNT SUTRO TOMER HIND VECTORS
INVERJION BASE HIND SP8ED (H/S),'.,' ' ' 'J
Figure 42. Hourly average horizontal wind vectors. Mt. Sutro Tower.
July 24 (top), July 25 (middle) and July 26 (bottom), 1975.
68
-------
471.7
431.7
<
s
3408
Ul
I
2964
zsao
. /~
\V
x\^
\\\v
i \ / / / " ^
, ; / //
\\\\MV
* \
INVERSION BASE
10 12 14
TIME (PST)
MOUNT SUTRO TOWER WIND VECTORS JULY 21, 197t
iSE WIND SPEED (M/S)j, ' ' ' ' i
16 18 20 22
Figure 43. Hourly average horizontal wind vectors, Mt. Sutro Tower,
July 24, 1974.
69
-------
500
T5 ft ir
TIME (J>Sf)
500
TIME (PST)
Figure 44. Time-height sections of hourly average ozone concentrations, Mt.
Sutro Tower, July 24 (top), July 25 (middle, and July 26 (bottom), 1975.
Units are ppftq,
70
-------
episode, a strong vertical gradient of ozone occurring above the inversion
base dipped down to ground level as the inversion base lowered. On July 25,
when the inversion base was below the Tower, a peak of 17 pphm occurred near
400 meters at 1300 PST. Maximum ozone values occurred at the same level and
time during the alert day of the July, 1974 episode* though concentrations
reached 26 pphm for this case.
The restoration period was characterized by a decrease in ozone concen-
trations aloft. The inversion base rose from 280 meters on July 26 to above
475 meters at 0700 PST on July 27, 1975. Ground concentrations dropped below
0.1 pphm at this time.
CONCLUSIONS
The synoptic pattern of the July, 1975, air pollution episode closely
resembled that of the July, 1974 incident. Weak pressure gradients were
established over the San Francisco Bay Area due to the presence of a Great
Basin High in the Pacific northwestern United States and the Subtropical High
off the California coast.
Surface wind flow was similar to the mean monthly pattern for July during
both episodes. Wind speeds were light throughout the Bay Area in the morning
and Increased in the afternoon as northwest flow dominated the Santa Clara
Valley. Within the inversion layer, north-northwest flow prevailed over Oak-
land and San Francisco at 0400 PST during the July, 1975, incident, whereas
winds were from the south-southeast at this time during the previous year.
According to the "sloshing" effect described in Section 6, a southerly wind
component is necessary for the advection of pollutants northward during the
early morning hours to confine pollutants within the Bay Area. Though this
effect was not observed during the July, 1975 episode, there may have been
appreciable "sloshing" of pollutants within the inversion layer due to a
diurnal shift in the winds.
Surface temperatures during both pollution episodes exceeded the July
monthly means, climbing to over 100°F on the respective alert days. During
July, 1975, the vertical temperature gradient was greater and the Inversion
base lower at Oakland than for the previous year at both 0400 and 1600 PST.
Surface oxidant levels recorded at BAAPCD monitoring stations were simi-
lar for the July, 1974 and 1975 air pollution episodes except for relatively
low concentrations recorded at Livermore during the latter episode. Though
surface temperatures at Livermore were high during July 25, 1975, winds were
extremely light. The low oxidant levels can be attributed to a minimum ad-
vection of ozone precursors from surrounding areas since the Livermore Valley
1s not a major source for ozone and its precursors.
Ozone within the elevated Inversion layer over the Mount Sutro Tower ex-
ceeded 10 pphm between 1000 and 1600 PST on July 25, 1975. Concentrations at
the surface throughout the Bay Area reached maximum values between the hours
1100 and 1500 PST. The largest oxidant value recorded during the alert day
was 23 pphm at Hayward at 1400 PST. The BAAPCD monitoring station at Hayward
1s 260 meters ASL, in the elevated peak of ozone concentration within the 1n-
71
-------
version layer situated over the Bay Area.
If the temperature profile throughout the Bay Area resembled that of Oak-
land at 1600 PST on July 25, 1975, (Figure 47)* high ozone concentrations
could be explained by high production rates near the surface and horizontal
diffusion of ozone from surrounding areas. Though mixing depths would be
small due to the intense surface based inversion, ozone destruction rates
would be less than production rates due to high temperatures and weak vertical
mixing.
Since the height and intensity of the inversion varies considerably over
the San Francisco Bay Area (Ahrens and Miller, 1969), the inversion over the
Santa Clara Valley may have been destroyed due to surface heating instead of
being surface based as at Oakland. This condition would be analogous to the
July, 1974, air pollution episode 1n which Holets found large mixing depths
and low ozone destruction.
Miller and Ahrens (1969) found high concentrations of ozone at the lead-
Ing edge of the marine air where the mixing depth 1s large and temperatures
are high. Ozone buildups along this inversion edge marked the limit of in-
land penetration of polluted marine air in which there is a high production
of oxidants. Figures 39b and 40 show that areas along the coast, where the
marine Inversion persisted, had low ozone concentrations compared to the in-
land valleys. Due to stagnant conditions over the Bay Area, the Seabreeze did
not penetrate far inland. High ozone concentrations produced at the leading
edge of the marine layer "spread out" into the Santa Clara Valley causing
large oxidant levels to accumulate within the valley during the day. In the
absence of appreciable winds, ozone was confined within the valley due to the
mountains to the east and the marine inversion to the west. Assuming the in-
version was reestablished during the evening, maxima of ozone that had been
generated 1n the surface layer were cut off from the surface and persisted
within the stable layer aloft. Concentrations near the surface decreased due
to Increased destruction rates and low production rates during the night. As
the Inversion lowered or was destroyed on the following day, downward diffu-
sion of ozone contributed to concentrations generated near the surface. The
pollution episode ended when temperatures dropped and wind speeds Increased
causing oxidants and their precursors to be advected out of the Bay Area.
72
-------
SECTION 8
CASE STUDY SEPTEMBER 3-8, 1974
As this period opened only San Jose exceeded the NAAQS when it recorded
9 pphm on September 3. During the peak of the episode, Livermore recorded 24
and 28 pphm on September 5 and 6, while 20 BAAPCD stations exceeded the 8 pphm
NAAQS. At the end of the period, no stations in the Bay Area exceeded 8 pphm.
SYNOPTIC PATTERN
During the first ten days of September a 500 mb long wave ridge per-
sisted with its axis roughly along the west coast near 130 W. Just prior to
the episode, the pattern north of 45 N resembled an omega block, leaving the
Pacific Northwest and British Columbia under subsiding northerly flow (Figure
45). On 2 September Quillayute, Wa. recorded,a high hour concentration for
the day of 400 DPM/KSCM (disintegrations min"1 per 1000 standard nT) indicat-
ing stratospheric intrusion on that day, and possibly on 3 September (Ludwick,
et al. 1975, 1976).
As the period opened, a trough moved through the Pacific Northwest pro-
ducing more zonal flow by 4 September (Appendix 2) and decreasing stability
while increasing marine layer depths over the Bay Area. By 5 September the
500 mb contour gradients were weak over most of the Southwest with light 500
mb westerly winds over coastal California. As the episode ended, the high
contour center moved eastward producing moderate 500 mb zonal flow over Cen-
tral and Northern California.
At 700 mb, Figure 46 shows the effects of the short wave trough on 4
September producing fairly strong zonal flow over Central and Northern Calif-
ornia. After the trough passed, weak 700 mb gradients over much of the South-
west produced light flow over coastal California. Increased subsidence pro-
duced by air and, as shown below, lowered the inversion base to the surface.
By 8 September the deepening southward moving low produced zonal flow over
the Pacific Northwest with moderate westerlies over Central California.
Surface analyses (not shown) show moderate onshore pressure gradients
along coastal California on September 3 and 4. By 04 PST on 5 September, the
axis of the California low was situated at or near the coastline producing
offshore pressure gradients. At the end of the period, an approaching surface
front re-established onshore flow along the coast.
SURFACE PATTERNS
Temperatures
73
-------
Si-/-'..
.A-/ 2-5 -^-v.
^
%t-^L S
#N"S:3, j
«;
in t
*
"T^ x >f> aj-
u,. - -^--iF *>-..'"
i? ^-A -by
/
u>
/
/
/
/
74
-------
Table 11 lists the maximum temperatures for selected stations in Calif-
ornia for the period. On September 3, maximum temperatures were near the
monthly average throughout the Bay Area and Central Valley. On September 5
and 6 maximum temperatures were 10F to 15F warmer than the monthly average
throughout the Bay Area and about 5F warmer than the monthly average in the
Central Valley. On September 7, Berkeley and Oakland returned to near aver-
age maximum temperatures. The rest of the Bay Area stations returned to near
average maximum temperatures on September 8. In the Central Valley, central
and southern stations remained somewhat above the monthly average while Red
Bluff decreased to about 3F below the monthly average.
TABLE 11. MAXIMUM TEMPERATURES (°F)
SEPTEMBER 3-8, 1974
Station Date
8
Average
September 1974
Marine Influenced
Berkeley
Oakland
Intermediate
San Jose
70 70 83 78 70 74
68 68 71 80 80 69
76 83 92 92 90 82
69.3
71.5
80.3
Inland
Los Gatos
Livermore
Gilroy
81 90 99 95 90 85
88 95 102 100 98 92
90 89 96 102 102 98
84.9
90.6
88.5
Central Valley
Bakersfield AP
Fresno AP
Sacramento AP
Red Bluff AP
96
95
91
97
96
97
93
98
100
100
98
101
104
101
99
101
102
101
99
101
97
98
93
93
96
94
90
96
.5
.8
.4
.3
Oxidants
On September 3 only San Jose exceeded the 8 pphm Federal Standard with a
high hour oxidant concentration of 9 pphm (Table 12). By September 5 and 6,
Livermore recorded high hour concentrations of 24 and 28 pphm respectively
75
-------
while 20 stations exceeded 8 pphm on September 5, and 18 stations exceeded
that value on September 6. At the end of the period, no stations exceeded 8
pphm.
TABLE 12. HIGH HOUR OXIDANT PATTERNS
SEPTEMBER 3-8, 1974
HighestNumber of BAAPCD
Date Station of Highest Maximum Maximum (pphm) Stations > 8 pphm
3
4
5
6
7
8
San Jose
Livermore
Uvermore
Livermore
San Jose, Los Gatos
Los Gatos
9
14
24
28
15
8
1
9
20
18
7
0
In the early afternoon of the peak day of September 5, oxidant maxima
were found near San Rafael in the north, Hayward 1n the central portion of
the eastern bay shore and San Jose in the south (Figures 47a'). Tongues of
relatively clean air were caused by marine penetration through the Golden Gate
and the Crystal Springs gap. At 14 PST (Figure 47b) the northern maximum
shifted to the Vallejo area, Hayward increased to its maximum of 21 pphm and
San Jose decreased slightly. During the next two hours, concentrations con-
tinue to Increase.
During the post sunset hours of 5 September, five stations (East San
Jose, Hayward, PHtsburg, Walnut Creek and Livermore) recorded 0, concentra-
tions greater than 8 pphm until 21 PST. Figure 48 shows the hourly oxidant
concentrations for Livermore, Hayward and (downtown) San Jose for September 5
and 6. Note that while nighttime concentrations lower to 1 pphm, those at
Livermore and Hayward remain high throughout much of the night of 5/6 Sep-
tember, and Hayward concentrations remain relatively high until midnight of 6
September.
The five stations mentioned above, as well as Los Gatos and Concord
which showed nighttime concentrations almost as high, are all located rela-
tively near mountain slopes. It appears therefore that oxidant rich air from
aloft 1s carried to the surface by downslope gravity winds at these locations.
Daytime concentrations at Livermore and San Jose reach 28 and 21 pphm
respectively while Hayward records a peak of 13 pphm on 6 September. Eighteen
76
-------
s-
0)
+J.O
C E
83
r- Q.
O
RJ >
S- I
0) 00
> Q.
a
o c
1C O
r-
03
O>
3 C
O T-
3; cn
01
(D
S-
3
o
0)
S- J_
O
77
-------
30
28
26
24
_ 22
r 20
~ IB
{2 16
I l4
x
O 12
10
8
024
6 8 10 12 14 16 18 20 22 24 2 4 6 8 10 12 14 16
5 SEPT TIME (PST) 6 SEPT
Figure 48. Hourly oxidant concentrations, San Jose, Livermore and Hayward
BAAPCD monitoring stations, September 5-6, 1974.
BAAPCD monitoring stations recorded high hour oxidant readings exceeding 8
pphm. On 7 September the maximum high hour oxidant recordings were 15 pphm
at San Jose and Los Gatos and seven BAAPCD monitoring stations recorded high
hour values greater than 8 pphm. On 8 September, a Sunday, no BAAPCD read-
ings exceeded 8 pphm.
UPPER AIR PATTERNS
Subsidence Inversion
The depth of the marine layer at Oakland showed the normal diurnal oscil-
lation with maximum depth in the morning and minimum in the afternoon.
Superimposed on this oscillation was a continued lowering of the inversion
base until it reached the surface at 16 PST September 4. The inversion base
remained at the surface until 04 PST, September 6 and then began to lift. A
north-south cross section along the west coast (not shown) shows isentropes
sloping steeply downward from Medford, Oregon to Oakland.
At the Sutro Tower, the inversion base remained above 350 m throughout
September 3. A time-height section of temperature for September 4 (Figure
49) shows the inversion lowering to the surface in the early afternoon. The
inversion remained on the surface until about 10 PST September 5 when rapid
warming near the surface commenced. After an unstable period near noon,
cooling of about 8C in two hours occurred at the surface, and the inversion
re-established at about 15 PST. A similar pattern occurred on September 6;
a strong inversion was based at or near the surface until about 9 PST; between
78
-------
/ /
, 14 i 16 Id . 18 i 16 14
'/' \ Ml
IB , , 18 , 18, 202
6 8 10 12-14 16 18 20 22
TIME (PST)
Figure 49. Time-height sections of hourly average temperature (°C) Sutro
Tower, September 4 (top), 5 (middle) and 6 (bottom), 1974.
79
-------
9 and 11 PST rapid wanning occurred near the surface. A near surface neutral
or slightly unstable layer existed until about 14 PST when near surface cool-
ing commenced and the inversion re-established and remained surface based
until about 07 PST September 7 (not shown). After some oscillations, the in-
version lifted from about 40 m above ground at 17 PST to about 150 m at mid-
night. The inversion base remained at least 100 m above the peak of Mt.
Sutro throughout September 8.
Winds
Mean surface to 5000 ft. winds over Oakland (Table 13) were from the
northwest at speeds ranging from 8 to 11 knots until 16 PST on 4 September
when the direction was from 260°. On September 5 winds were light easterly,
shifting to southwest from the morning of September 6 to the afternoon of
September 7 as speeds increased. By the morning of September 8 winds were
from the northwest.
Wind vectors plotted for each hour at each measurement level show the
possible recirculation of inversion layer pollutants (Figure 50). On 4 Sep-
tember hourly average winds at Levels 5 and 6 are 1 m sec" with often a,
slight offshore component from 08 PST until westerlies of about 5 m sec"
start at 15 or 16 PST. From about 20 PST on 4 September through about 14 PST
on 5 September winds have an offshore component; northeasterly at 3 to 4 m
sec until near sunrise then light easterly continue until westerly winds com-
mence near 16 PST. Westerly winds at Levels 5 and 6 on 6 September are in-
terrupted by southerly winds in the forenoon.
TABLE 13. MEAN SURFACE TO 5000 FT. WINDS
MEASURED AT OAKLAND
Date Time (PST) Direction (Deg) Speed (Kts)
3
3
4
4
5
5
6
6
7
7
8
04
16
04
16
04
16
04
16
04
16
04
280
290
290
260
70
80
195
245
235
265
280
8
11
9
4
4
5
6
9
8
10
12
80
-------
500
450-
400-
350- .
\ , . * V
\
/ i "» -..
,//
\ //
i/V
,-r. "
V.
\ if i' '
250
500
450
SF 400
0 2 4 6 8 10 12 ' 14 16 18 20 22
350
300-
500
4SO-
400^
390
.x-
/ /
i 1 / ' V-
\\\
» / I / / NN
x \
280- | 1 1 1 r1 1 1 1 1 1 1 1-1-1 1 1 1 1 1 r i . ,
0 24 6 » 12 M 16 IB 20 22
TIME (P8T)
MS"
U D
Figure 50. Hourly average wind vectors, Sutro Tower, September 4 (top), 5
(middle) and 6 (bottom), 1974.
81
-------
Ozone
Ozone concentrations at all measurement levels at the Sutro Tower re-
mained below about 4.5 pphm on September 3 (not shown). Data are missing for
the morning of September 4. In the afternoon of September 4 Level 6 measured
9.5 pphm for the hours beginning at 16 and 17 PST while Level 5 reached 8
pphm at 17 PST (not shown). Nighttime values decreased to about 4 pphm at all
levels. Level 1 remained below 4 pphm throughout the day.
Figure 51 shows that 03 concentrations remained below about 4.5 pphm
until about 09 PST, 5 September when concentrations started to increase at all
levels. 03 concentrations increased to peaks of 20 pphm at Level 6 and 23.5
pphm at Level 5 at 14 PST and 25 pphm at Level 3 at 15 PST. The near surface
measurement level recorded a daytime peak of just under 8 pphm at noon. A
secondary peak of 18 pphm at Level 3 and the daily maximum of 12.5 pphm at
Level 1 occurred at 20 PST, while increases to about 12.5 pphm occurred at
Levels 5 and 6 at 21 PST.
On 6 September, 03 concentrations were higher at Levels 1 and 3 than at
Levels 5 and 6 until about sunrise. In the early forenoon, concentrations
were about uniform at the top three measurement levels. The highest concen-
tration is 12.5 pphm at 15 PST at Level 5. A secondary maximum occurred at
22 PST with Level 3 recording 10 pphm. Concentrations at Level 3 remain near
9 pphm until 04 PST, 7 September (not shown) and remain below 8 pphm at all
measurement levels for the rest of the day.
SUMMARY
The high oxidant period of this case study began and ended rather
abruptly. At the beginning of the period a short wave trough was followed by
a short wave ridge to initiate the period. At the close of the period an
approaching upper level trough with associated surface front and increased
westerly winds ended the high oxidant concentrations.
Wind patterns measured at the Sutro Tower indicate that a recirculation
of pollutants occurred within the Bay Area. On 4 September a period of few
hours of offshore winds occurred at the upper measurement levels. A few
hours of westerly winds that afternoon were then followed by offshore flow of
decreasing strength until westerlies returned in the late afternoon of 5 Sep-
tember. While the apparent recirculation pattern is not as readily apparent
on 6 September, it appears that there may have been transport along the NW-SE
axis of San Francisco Bay.
High concentrations of ozone and Be at Quillayute, Wa. about two days
prior to the Bay Area episode combined with subsiding northerly flow suggest
the possibilities of stratospheric contributions at the beginning of the
episode.
82
-------
24-
20
O-LEVEL 6
A-LEVEL 5
v-LEVEL 3
.-LEVEL I
02468 10 12 14 16
TIME (PST)
18 20 22
10 12 14
TIME (PST)
20 22
Figure 51. Hourly average ozone concentrations, Sutro Tower, September 5
(top) and 6 (bottom), 1974.
83
-------
SECTION 9
CASE STUDY OF SEPTEMBER 14-18, 1974
Ozone concentrations in the Bay Area did not reach the 20 pphm BAAPCO
alert level during this period. However, the case period does show some
similarities 1n meteorological patterns to those discussed in previous sec-
tions. In addition, potential vorticity analysis on the 312 K isentropic sur-
face Indicates that air of presumed stratospheric origin penetrated at least
to the 700 mb surface during the early portion of the period.
SYNOPTIC PATTERN
On the afternoon of Friday the thirteenth the synoptic situation was
dominated by a developing omega block. A rather Intense low over the Gulf of
Alaska was followed by.a high pressure system near the northern Washington
border and a trough near eastern Nevada. Figure 52a shows the 312 K 1sentrop-
ic surface at 691 mb over Oakland with northerly winds and a mixing ratio of
0.8 gm/kg. Potential vorticity j&opleths on this surface (not shown0 show
values below 50 (1n units of 10 cm sec °K gm ).
By the morning of September 14 (Figure 53a) the trough over eastern
Nevada had formed a cutoff low. The 312 K isentropic surface lifted to 684
mb over Oakland. A potential vortteity maximum occurred over eastern Nevada
(Figure 53b), with values of 50 to 75 over central California.
Twelve hours later (Figure 54a) the high center had drifted south to
central Oregon while the 312 K Isentropic surface had lifted to 670 mb. A
300 potential vortlcity maximum developed over central Nevada (Figure 54b),
while values of 100 covered most of central California. A value of 97.7 was
calculated for Oakland.
A cross section from Salem, Oregon to Winslow, Arizona (Figure 54c) shows
the 312 K 1sentrope centered in a stable layer sloping down from about 500 mb
over Ely, Nevada to about 680 mb over Salem, Oregon.
At 04 PST on September 15, the synoptic pattern was similar to previous
days (Figure 55a). The 312 K Isentropic surface lowered to 697 mb and a po-
tential vorticity value of 112 was calculated for Oakland (Figure 55a).
Twelve hours later, a stream function trough had developed over central Cal-
ifornia (Figure 56a) while potential vorticity values dropped to about 50
over the area (Figure 56b).
SURFACE PATTERNS
84
-------
A deep marine layer at the beginning of the pe/iod caused many stations
in areas either above the normal inversion base or in areas where the inver-
sion is normally eroded during the day to experience maximum temperatures less
than the average for the month. Maximum temperatures at stations usually in-
fluenced by marine air were near the average for the month (Table 14). Berke-
ley and Oakland, for example, were 1.3 F and 0.5 F respectively, below the
average maximum temperature for the month, while Gilroy and Los Gatos were
10.5 F and 4.9 F cooler, respectively. Central Valley maximum temperatures
were about 5 F cooler than the average for the month.
Maximum temperatures decreased over most of the Bay Area on the 15th and
16th. Inland stations, however, recorded maximum temperatures on the 16th the
same as or slightly greater than those of the 15th. Central Valley warmed
more in the north than the south on the 15th and then increased temperatures
slightly on the 16th.
Bay Area maximum temperatures increases significantly on the 17th, while
those of the Central Valley increased slightly. As the period ended, Bay Area
maximum temperatures decreased somewhat at marine influenced and elevated
stations but remained high in the inland valleys. Central Valley temperatures
decreased somewhat in the north and remained about constant in the south.
TABLE 14. MAXIMUM TEMPERATURES (F). SEPTEMBER 14-18, 1974
Station
Date 14 15 16 17 18
Ave.
Sept. 74 Normal
Marine Influenced
Berkeley
San Jose AP
Oakland
68 64 65 70 66
80 77 74 81 80
71 65 65 70 68
69.3
80.3
71.5
Inland
Los Gatos
Gi1roy
Livermore
80 79 81 -89 82
78 85 85 88 95
84 84 86 98 96
84.9
88.5
90.6
Central Valley
Bakersfield AP
Fresno AP
Sacramento AP
Red Bluff AP
89
89
86
90
89
92
90
98
90 94
94 95
92 94
98 101
94
96
93
99
96.5
94.8
90.4
96.3
85
-------
312°
ISENTROPIC SURFACE .4 SEPT 1974-007.
(Mill AM
Iigure 52 . Montgomery stream function on 312 K
surface, 00 GMT, 14 September, 1974.
86
-------
8.2
to e
T- tt)
u to
r-
tf?
O
(O O
r-
4->CSJ
C r-i
0)
P «
O 0)
Q. U
u>
CSJ
C
O i
o +J
C O.
0)
§ .
en
LO CM
87
-------
o
o
o
o.
9)
O
X
0>
C3
3
O)
IO !>
in o.
o> co
in
s
X
01
£
O)
10 S;
CO S.
Q.
tt)
-------
MUM WWICMUOCA ELY
POTENTIAL vwncmr
BSEPTI9W-OCZ
Figure 54c. Northwest-Southeast cross section of potential temperature (°K,
top) and potential vortlclty (10"IU cm sec °K gm ', bottom), 00 GMT, 15
September, 1974.
Winds
Surface winds along the coast were governed by a Pacific cyclonic cir-
culation at the beginning of the period (Figure 57). Bay Area winds, chan-
neled by topography, were similar to those shown 1n Section 6. Surface
stream line analyses for the remainder of the period (not shown) approximated
those discussed in Section 6.
Oxidants
Prior to the case study period, the 20 pphm MAPCD health effects level
was exceeded on September 10 at Llvermore, Alum Rock and San Jose with a maxl-
89
-------
mum of 26 pphm at San Jose. On this day 17 BAAPCD stations exceeded the 8
pphm NAAQS. High hour concentrations decreased until September 14 when the
maximum recorded in the Bay Area was 11 pphm and five stations exceeded the
8 pphm Federal Standards.
Early morning surface observations for each day of the period were simi-
lar to those at 09 PST September 15 (Figure 58a). Concentrations were 4 to 5
pphm over the southern San Francisco Bay, probably due to reduced destruction
rates over the water surface. Those throughout the rest of the area ranged
from 2 to 3 pphm.
Afternoon oxidant readings increased until 17 September. Figure 58b
shows centers of 16 pphm at Livermore and 14 pphm at Los Gatos on the
afternoon of September 17. The areas of the Santa Clara Valley and Livermore-
Diablo Valleys exceed the 8 pphm NAAQS. Afternoon patterns throughout the
period are similar. Table 15 indicates the extent of violations of the NAAQS
for the period.
TABLE 15. HIGH HOUR OXIDANT PATTERNS. SEPTEMBER 14-19. 1974
Station ofHighestNumber
Date Highest Maximum Maximum (pphm) Stations > 8 pphm
14
15
16
17
18
19
Livermore, San Jose
Livermore
Livermore
Livermore
Livermore, Los Gatos
Livermore
11
10
12
17
14
12
4
3
4
14
6
4
UPPER AIR PATTERNS
The Subsidence Inversion
Heights of the inversion base (Table 16) measured over Oakland at 16 PST
lowered until September 16, remained low on September 17 and then began to
lift on September 18. Afternoon mixing depths were estimated at San Jose and
Livermore from the Intersection of the dry adiabat through the maximum surface
temperature with the Oakland afternoon soundings. Mixing depths over San Jose
lowered through September 16 and raised on September 17 and 18. Intense heat-
Ing in the Livermore valley on September 17 produced a mixed layer 1220 m
deep.
90
-------
in
91
-------
o
o
i-
o
O.
0)
u
X
GO
o
o
Q.
-M
in Q.
0)
D>r-
92
-------
Figure 57. NOAA B visual satellite
1721 GMT, 14 September, 1974.
picture of California and eastern Pacific,
93
-------
OO
D-
s_
o
Q.
0)
O
X
Ln OL
O)
o
o
10
c
IT)
(O
00
c:
o
(O *
s- s_
+J O)
C -Q
o o.
U (U
00
ro
o uo
>
3 ns
en co
94
-------
TABLE 16. ESTIMATED MIXING DEPTH PARAMETERS FROM 1600 PST
OAKLAND SOUNDINGS
Inversion BaseMixing Depth (meters;
DATE (mb) Uyermpre San Jose
13
14
15
16
17
18
930
975
970
1000
1000
980
1018
975
613
582
1219
948
811
853
508
408
443
501
Hinds
A time-height section of the westerly wind at Oakland through the tropo-
pause for the period of study is shown in Figure 59. Easterly winds dominate
through most of the troposphere above about 1000 m. Below about 1000 m ra-
winds at 12-hour intervals show winds with a persistent westerly component.
SUTRO TOWER PATTERNS
A temperature time-height section for 14 September (not shown) shows no
inversion through the daylight hours. A weak inversion was established at
about 20 PST. Winds had a westerly component throughout the day.
Figure 60 shows the variations of hourly average temperature and hori-
zontal wind vectors for the period 15-18 September. The average inversion
height lowers until 17 September and then lifts the next day. Superimposed on
this drecrease in marine layer depth are diurnal oscillations similar to the
average pattern for the month. The inversion base was highest in the early
morning and lowest in late afternoon or early evening. It appears that the
inversion begins to descend from its post-midnight maximum somewhat earlier
each day until the peak day of 17 September.
Marine layer winds were southeasterly throughout the period. Within the
inversion layer, winds were strong westerly or northwesterly from near sunset
until about the time of the lowering of the inversion. As the inversion
lowers, winds veer to the north or northeast. As the episode develops, the
strength and duration of easterly wind components increases until 17 Septem-
ber when wind at the top two measurement levels are northeasterly at about 5
m sec~ for nearly six hours. Near sunset, winds become more westerly as the
inversion base rises.
Ozone concentrations remained below 4 pphm throughout 14 September at all
95
-------
M M M f M 1 1
*»**«
crt
t
at
3
r
i
I
5
u
0)
V)
s
i
«4-
o
O
u
0)
£
o>
96
-------
14 16 18 20 22
\\\ 1\\
i\iK^s
J-250
6 6 10 12 ' 14 16 IB 20 22
TIME (PST)
MS
Figure 60. Hourly average horizontal wind vectors and temperatures (°C),
Sutro Tower, September 15 (top) and 16 (bottom), 1974.
97
-------
WIND VECTORS SUTRO TOWER
17 SEPT 1974
12 14 16 18 20 22
I I
8 » 12
TIME (PST)
250
16 18 20 22
MS''
Figure 60 (contd). September 17 (top) and 18 (bottom).
98
-------
measurement levels.
Comparison of the variations of wind and temperature (Figure 60) with
those of ozone (Figure 61) show evidence of high ozone concentrations located
near the upper edge of the strong isotherm gradient. On 15 September, the
ozone maximum of 7.5 pphm at 03 PST coincides with the dip in the 18C iso-
therm below the measurement level. As the inversion base lifts, concentra-
tions at Level 5 decrease before those at Level 6. During the hours 06-08
PST, while the whole tower is within the marine air layer, there is no dif-
ference in Oo concentration from the top to bottom measurement level. As the
inversion lowers in midmorning, 03 concentrations at Level 6 increase about
an hour prior to those at Level 5 since the former encounters inversion layer
air first. The nocturnal maximum at Levels 5 and 6 also coincide with a mini-
mum in the Inversion base height.
On 16 September, all measurement levels were within or below the strong
isotherm gradient and experience about the same 0, concentrations for the
first half of the day. Near noon, the inversion Base begins to descend and
concentrations at the top two measurement levels begin to Increase. The noc-
turnal maxima at these levels occur near the time of minimum Inversion height.
Analysis on 17 September is hampered by the loss of six critical hours of
data, however the pattern appears similar to those of the previous two days.
Nocturnal 03 minima at the top two levels coincide with maximum inversion base
height. Increases in concentration at these two levels beginning about 08 PST
occur at about the time when the strongest thermal gradients are below these
levels. These Increases appear to be too rapid and too early in the day to be
due to in situ photochemical production.
The pattern on 18 September is again similar, with the secondary minimum
near 14-15 PST coinciding with an upward undulation of the inversion base. As
the inversion lifts above the tower near midnight, there is only a slight
vertical gradient 1n ozone. September 19 concentrations remained below 4 pphm
at all measurement levels.
SUMMARY OF THE SEPTEMBER 14-18 CASE STUDY
The study period begins with a deep marine layer extending to 930 mb on
the afternoon of Friday the 13th. The relatively large marine layer depth is
probably Influenced by an offshore cyclonic circulation, which appears from
satellite photo to be the remnants of a tropical depression. The other main
synoptic feature which will continue to dominate the period 1s an "omega"
block: an Intense low 1n the Gulf of Alaska, a high centered over eastern
Washington and another low centered over Nevada.
As the offshore depression weakens, the base of the elevated inversion
lowers. Upper level air flows around the high and is from the north east over
the Bay Area. On the afternoon of September 15, potential vortlcity analysis
shows what appears to be air of stratospheric origin at the 312 K isentropic
surface, which over Oakland is at about 700 mb.
High hour 03 concentrations at Qulllayute, Ma. were 5 pphm on 13 Septem-
99
-------
I
LEVEL I (6m)
A LEVEL 9 (I78iti
O LEVEL 6 (219m)
J L.
_J
24
10
I
12 14
TIMEtPST]
16
18
20
LEVEL I (6m)
A LEVEL 9 (178m)
O LEVELS (219m)
J L
_J
24
10
12 14
TIME (PSTI
16
ie
20
22
Figure 61. Hourly average ozone concentrations, Sutro Tower, September
15 (top) and 16 (bottom), 1974.
100
-------
10
LEVEL I (6m)
A LEVEL 5 (178m)
9 ( O LEVEL 6 (219m)
i i
-I 1 L I
10
12 ,4
TIME (PST]
16
18
20
22
LEVEL I (6m)
A LEVEL 5 (178m)
O LEVEL 6 (219m)
1 I
10
12 14
TIME (PST]
16
ie
20
22
24
Figure 61 (contd). September 17 (top) and 18 (bottom), 1974.
101
-------
ber and 6 pphm on 14 and 15 September (Ludwick et al., 1975, 1976). These
concentrations, correlated with high radionuclide measurements, suggest air of
stratospheric origin at the surface in rural Washington.
Lack of data hampers analysis of the mesoscale circulation patterns. It
appears, however, that a weakened form of the Pacific monsoon-sea breeze cir-
culation dominates the case study period. Compare the Sutro Tower time height
cross sections of temperature and wind vectors for the individual days of
September 15th through 18th with the means for September 1974 (Figures 60 and
12). Each figure shows that the maximum inversion h ight occurs a few hours
before sunrise with lowering until midafternoon and a relatively constant base
height until a couple of hours after midnight. Westerly wind components are
weakest in midmorning, reach a peak at about 300 m near sunset and then de-
crease until the next morning minimum.
During the period of this case study, however, the synoptic scale off-
shore flow masks the usual monsoon circulation upon which the sea-breeze is
superimposed. This results in winds with an easterly component over the Sutro
Tower commencing in the late forenoon or early afternoon. Near noon of the
September 16 and 18 period the zone of easterly wind component appears briefly
and reaches only down to about 450 m. On September 17, the easterly component
zone engulfs the tower from about 09 PST until sometime in the afternoon.
(Data are missing between 12 and 20 PST on this important day.) These wind
shifts are not reported by the Oakland radiosonde because the time of easterly
winds is less than the 12 hours between observations and/or the intervals be-
tween reported winds are too gross to pick up this flow. The only corrobora-
ting evidence is the surface observations at San Francisco International which
report winds with easterly components from 0655 to 1155 PST on September 17.
If there is an afternoon circulation reversal, wind vector patterns
indicate a recycling of ozone and/or its precursors above the inversion base.
In the Livermore valley, and perhaps other sheltered inland valleys the in-
version is eroded from below and convection through relatively large mixing
depths (4000 ft. on September 17) carry surface pollutants up and mix those
from aloft to the surface. As the cycle repeats, each succeeding day's pollu-
tion builds upon a higher base and maximum concentrations increase.
Stratospheric air which appears to have reached down at least to 700 mb
during the beginning of the period may play a role either additively or as a
trigger in the photochemistry, although its significance is difficult to
assess.
102
-------
SECTION 10
SUMMARY AND DISCUSSION
Ozone concentrations within the elevated inversion layer over San Fran-
cisco exceeded 8 pphm during about 15 percent of the measured hours during
the summer of 1974. (Adjustment to reflect the change of calibration proce-
dures during 1975 reduce this percentage to about 9 percent.) This percent-
age compares favorably with the percent of the summer hours exceeding the 8
pphm NAAQS for oxidants at those Bay Area Air Pollution Control District mon-
itoring stations with most frequent violations. In addition, the majority of
hours of violation occurred during the three 1974 case study episodes dis-
cussed previously; for example, 109 of 164 violation hours at measurement
Level 5 and 64 of 101 hours at measurement Level 6 occurred during these three
1974 periods. We assume that the remainder of violation hours at the upper
tower levels occurred during periods of high surface 0.,.
The average summer wind flow over the Bay Area is onshore, caused by a
combination of the summer Pacific monsoon and a more local sea breeze circula-
tion. At the Sutro Tower, the inversion base is highest near 04 PST and the
winds at that time are westerly or southwesterly at all levels. Near 10 PST,
the inversion base begins to descend and winds veer from southwesterly to
westerly at the lower tower levels or from westerly to northwesterly at the
upper levels. (See Figures 10 and 12). The inversion base is lowest in late
afternoon when onshore flow is strongest.
During the study periods considered, a closed low off the British
Columbia coast with high heights over the Bay Area was the typical 500 mb
pattern. This pattern produced subsiding offshore flow which weakened or
interrupted the monsoon sea breeze induced onshore flow. As the episodes pro-
gressed, the first one or two days of the intensification period experienced
winds with slight easterly components during the period when the inversion
layer subsided. The most severe day of the episode typically showed a deep
layer of easterly winds, often extending to the ground, existing from late
forenoon to late afternoon. This is similar to the measurements by Fosberg
and Schroeder (1966, see Figure 8) showing easterly flow above the inversion
layer and very light flow ahead of their wind shift line.
The superposition of synoptic pattern, Pacific monsoon and local circula-
tions produces a recirculation of pollutants. During periods when the local
circulation is weakest, flow from the large scale pattern carries polluted
air offshore above the Inversion layer. In the late afternoon the sea breeze
flow returns these pollutants onshore. In many cases this oscillating flow is
not measured by rawind soundings because the flow may be relatively shallow
and because the offshore component occurs between observation times.
103
-------
Daytime heating during episode periods destroys the inversion layer in
the inland valleys. Ozone rich inversion layer air can be mixed downward to
add to oxidants generated from the present day's accumulation of precursors.
The "warm sea breeze" model of Fosberg and Schroeder (1966) may describe
the inland penetration of modified marine layer air. Measurements reported
by Miller and Ahrens (1970) show maximum oxidant concentrations at the inland
edge of the inversion base. It would seem, therefore, that the locations of
high surface oxidant concentrations would be on the inland edge of the "warm
sea breeze front" where the inversion has been destroyed and inversion layer
ozone can be mixed downward.
At night surface cooling re-establishes the elevated inversion, causing
Oo below the inversion to be destroyed at the surface while cutting off inver-
sion layer 0, from surface destruction. The nocturnal inversion layer air is
advected coastward by air produced by a combination of local cooling effects.
During the episode, the elevated polluted layer may reach to within a
hundred meters or so of Mt. Sutro. Nighttime high oxidant concentrations at
stations located near mountain slopes probably reflect nocturnal downslope
flow bringing down inversion layer ozone. High concentrations at Hayward
during the night of 5/6 September, 1974 (Figure 48) reflects the 258 m sta-
tion elevation and its location near the mountain slope.
At the end of the episode return of westerly flow advects pollutants from
the area. It is possible that these Bay Area oxidants contribute to high
pollutant levels within the Central Valley on subsequent days.
Remote source contribution.to Bay Area oxidant concentrations is possi-
ble, but hard to document with available data. Baboolal et al (1975) con-
cluded that high ozone levels observed in the Santa Ynez Valley "could not be
wholly attributable to local sources." This Valley is about 200 km northwest
of Los Angeles. Kauper and Niemann (1975) show that, with a shallow marine
layer, over water trajectories of elevated 03 layers moving into the Ventura
county coastal area from the Los Angeles Basfn is the "normal Situation."
Gloria et al (1975) found high 03 concentrations within the inversion layer
over the Pacific Ocean.
Just prior to the July 1974 case study, inspection of NWS 850 and 700 mb
analyses suggest isentropic trajectories from Los Angeles to San Francisco.
Available upper air wind data from coastal locations and the fact that surface
oxidant concentrations at high elevations in the Los Angeles basin reached
about 25 pphm do not contradict this suggestion. Total time of travel from
Los Angeles, if such trajectories were real, would be on the order of 48
hours.
The air quality control implications of the possibility of such long
range transport involve more than the local Bay Area Air Pollution Control
District. A coordinated effort is needed to document adequately the reality
of this transport. This effort would involve instrumented aircraft, a
coastal network of pibal and/or rawinsonde measurements and perhaps tetroons.
Measurements must cover a distance of some 600 km during specific synoptic
104
-------
situations.
The possibilities of a stratospheric contribution to the inversion layer
ozone burden has been discussed here in a preliminary manner. Just prior to
each of the September 1974 case studies imultaneous high surface concentra-
tions of ozone and radionuclides of stratospheric or upper tropospheric origin
were measured at Quillayute, Washington by Ludwick, et al (1975). The coin-
cidence of these atmospheric components indicates intrusion of stratospheric
air. In each case, the synoptic pattern was in the form of an "omega" block
where flow around the upper level ridge could have brought stratospheric air
into the Bay Area. Although hand analysis and computation of potential vor-
ticity distributions contain a number of inherent inaccuracies, such analysis
suggests that stratospheric air penetrated to 700 mb over Oakland at the be-
ginning of the 14-18 September, 1974 period. If true, and since this was not
a particularly intense episode, it is very likely that stratospheric ozone
contributes either additively or as a chemical trigger in similar and more in-
tense situations.
The possible contribution of stratospheric ozone involves broader policy
questions. Synoptic situations producing air pollution episodes in the Bay
Area are also those which produce strong upper tropospheric subsidence. If
the natural stratospheric contribution is purely additive and of the order of
5-6 pphm, then what would just be a smoggy day might become a Health Alert
Advisory (oxidant concentrations >_ 20 pphm). The possibilities of strato-
spheric 03 as a photochemical trigger would require measurements of precur-
sors and detailed synoptic analysis combined with numerical modelling of
photochemical kinetics.
The fate of oxidants and their precursors generated in the Bay Area is a
further subject of needed study. As discussed above, a Bay Area pollution
episode ends as the synoptic situation reverts to the normal summer pattern
and pollutants are advected eastward from the area. Observations along the
pollutant path from the area are needed to determine the effects of Bay Area
pollution on the air quality of the Central Valley and the Sierra Nevada.
Such observations might provide the impetus for extending air quality simula-
tion models to include such long-range transport.
105
-------
REFERENCES
1. Baboolal, L.B., I.H. Tombach, and M.I. Smith, 1975: Mesoscale flows
and ozone levels 1n a rural California coastal valley, 1st. Conf. on
Mesoscale Modeling, Analysis and Prediction, Amer. Meteor. Soc., Las
Vegas, Nev.
2. Bay Area A1r Pollution Control District, 1971: Aviation Effect on
A1r Quality in the Bay Region. San Francisco, Ca.
3. gay Area A1r Pollution Control District, 1976: Oxidant Experience in
the San Francisco Bay Area, 1965-1975. Technical Services Division,
Information Bulletin, March 12, 1976, 11 pp.
4. California Air Resources Board, 1975: Final Report of the ad hoc
Oxidant Measurement Committee, No. 75-4-4, February 20, 1975.
5. Edlnger, J.G., 1973: Vertical distribution of photochemical smog
in Los Angeles Basin. Environ. Sci. Techno!. 7:247.
6. Edlnger, J.G., M.H. McCutchan, P.R. Miller, B.C. Ryan, M.J. Schroeder,
and J.V. Behar, 1972: Penetration and duration of oxidant air
pollution in the South Coast A1r Basin of California. J. Air Poll.
Control Assoc., 22:882.
7. Gelger, R., 1965: The Climate Near the Ground. Harvard University
Press, Cambridge, 611 pp.
8. Gloria, H.R., G. Bradburn, R.F. Reinlsh, J.N. Pitts, Jr., J.V. Behar,
and L. Zafonte, 1974: Airborne survey of major air basins 1n
California. J. A1r Poll. Control Assoc., 24, 645-652.
9. Kauper, E.K., and B.L. Nlemann, 1975: Los Angeles to Ventura over
water ozone transport study. Tech. Rept. under Contract ARB 4-1126.
Metro Monitoring Services, Covlna, Ca. 91723.
10. Lea, D.A., 1968: Vertical ozone distribution in the lower troposphere
near an urban pollution complex. J. Appl. Meteor., 7j252.
11. Lovill, J.E., and A. Miller, 1968: The vertical distribution of ozone
over the San Francisco Bay Area. J. Geophys. Res.,73;5073.
106
-------
REFERENCES
12. Ludwig, J.D., T.D. Fox, and L,L, Wendell, 1976: Ozone and radionuclide
correlations 1n air of marine trajectory at Quillayute, Wa. J. Air
Poll. Contr. Assoc.. 2£, 565-569.
13. MacCracken, M.C. and G.D. Sauter, 1975: Development of an Air
Pollution Model for the San Francisco Bay Area. Final Report to
NSF-UCRE51920, Lawrence Livermore Laboratory, Ca., 215 pp.
14. Miller, A., 1966: Land-Sea Boundary Effects on Small Scale
Circulations. San Jose State College M teorology Department, 97 pp.
15. Miller, A., 1975: Project STABLE. Bull. Amer. Meteor. Soc., 56,
52-55.
16. Miller, A., 1976: Wave Properties in the West Coast Inversion.
San Jose State University, 1976, 95 pp.
17. Miller, A., and C.P. Ahrens, 1970: Ozone within and below the
west coast temperature inversion. Tellus 22:328.
18. Perl, G., 1965: Das Bodennake Ozoniu Arosa Seine, regelmSssigen und
unregelmassigen Schwankungen, Arch. Meteorol. Geophys. Bioklimatol.,
Ser. A 14:449-458.
19. Reiter, E.R., 1971: Atmospheric Transport Processes, Part 2: Chemical
Tracers. USAEC. Div. of Tech. Infor., TID 25314, 382 pp.
20. Root, M.E., 1960: San Francisco, the air conditioned city.
Weatherwise, 13.
21. Schroeder, M.J., M.A. Fosberg, O.P, Cramer, and C.A. O'Dell, 1967:
Marine air invasion of the Pacific Coast: A problem analysis. Bull.
Amer. Meteor. Soc., 413, 11, 802-808.
22. Teichart, F., 1955: Vergleichende Messung des Ozong e haltes der
Luft am Erboden und in 80 m Htthe, Z. Meteorol. 9:21-27.
23. Wagner, J.A., 1974: Weather and circulation of July 1974-heat wave
and drought over the middle third of the Country. Mon. Wea. Rev.,
CII, 10, 736-742.
107
-------
APPENDIX A
Surface Monitoring Station Locations and Histories
Location in
Figure 3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Station*
Vallejo
Pittsburg
Concord
Pleasant Hill-Walnut Creek
Liver mo re
Fremont
Hayward
San Leandro
Oakland
Richmond
San Rafael
San Francisco
San Francisco (East)
Mt. Sutro Tower*
Burlingame
Redwood City
Mountain View
Sunnyvale
San Jose
San Jose (Alum Rock)
Los Gatos
Gi 1 roy
Santa Rosa
Napa
Petaluma
Fairfield
Stockton*
Sacramento*
Observations
Began +
1969 +
1969
1972
1968
1968 +
1962 +
1973
1963
1962 +
1962 +
1962
1962 +
1974
1974
1962 +
1966
1972 +
1973
1962 +
1974
1972
1974
1970 +
1969 +
1971 +
1969 +
*A11 stations under jurisdiction of BAAPCD except 14 (San Jose State Univer-
sity) and 27, 28 (California Air Resources Board).
Indicates stations which have been relocated one or more times within the
given city.
108
-------
TECHNICAL REPORT DATA
(l'li:a\e mad Instruct um\ on the reverse be/ore completing}
1 HI I'fjK I NO
EPA-600/4-77-046
3. RECIPIENT'S ACC6SSIOr*NO.
4 TITLE AND SUBTITLE
OZONE OVER SAN FRANCISCO
Means and Patterns During Pollution Episodes
6. PERFORMING ORGANIZATION CODE
5. REPORT DATE
November 1977
7 AUTHOR(S)
Kenneth P. MacKay
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
San Jose State University
San Jose, CA 95192
10 PROGRAM ELEMENT NO.
1AA603 AD-02 (FY-77)
11 CoTlTRACT/GRANT NO.
R802235
2 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 4/73 - 5/77
14. SPONSORING AGENCY CODE
EPA/600/09
15 SUPPL CMl NTARY NOTE S
16. ABSTRACT
Measurements of meteorological parameters were taken at six levels and ozone at
four levels between 260m and 473m ASL on the Mt. Sutro T.V. Tower in San Francisco
during the summers of 1974 through 1976. Hourly average ozone concentrations within
the elevated inversion layer at this location exceeded the 8 pphm (160 yg m )
National Ambient Air Quality Standards about 15% of the time. High inversion layer
ozone concentrations at this site were associated with high surface concentrations
occurring during area-wide air pollution episodes. These episodes occurred when a
lobe of the Pacific high pressure system penetrated inland. During these episodes,
superposition of synoptic scale northeasterly flow and locally produced mesoscale
flow caused easterly or light westerly flows during the late forenoon within the
inversion layer and westerly flow in the late afternoon. Inland, where the inversion
was destroyed from below, inversion layer and surface generated pollutants were
convectively mixed. This mixing and the wind oscillation recycled pollutants. The
episodes ended when the synoptic situation reverted to one more normal for the season
and pollutants were advected from the area.
Kl Y WORDS ANfi DOCUMENT ANALYSIS
I). IDF N riFIERS/OPEN INDED TERMS
* Air pollution
* Ozone
* Meteorological data
lowers
San Francisco, CA
c. COSATI Field/Group
13B
07 B
04B
13M
I!) nlSfHIHUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
123
20. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
RPA rorm 2220-1 (9-73)
109
------- |