Application of
Photochemical Models
Volume III
Recent Sensitivity Tests
and other Applications
of the LIRAQ Model

-------
/if y ! i
APPLICATION OF PHOTOCHEMICAL MODELS
Vol use III
Recent Sensitivity Tests and Other Applications of the LIRAQ Modelv
prepared by
Association of Bay Area Governments
Hotel Claremont
Berkeley, California 9470b
in association with
Bay Area Air Quality Management District
San Francisco, California
Lawrence Livermore Laboratory
Livermore, California
Systems Applications, Inc.
San Rafael, California
prepared for
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
EPA Project Officer: John Summerhays
Contract No. 68-02-3046
Final Report, December 1979

-------
PREFACE
This document is one of four volumes intended to provide information
relevant to the application of photochemical models in the development
of State Implementation Plans. The reports are particularly directed
toward agencies and individuals responsible for preparation of
non-a tta i nnient plans and SIP revisions for ozone. The four volumes are
titled as follows:
Application of Photochemical Models
Volume 1 - The Use of Photochemical Models in Urban Ozone
Studies
Volume II - Applicability of Selected Models for Addressing
Ozone Control Strategy Issues
Volume III - Recent Sensitivity Tests and Other Applications
of the LIRAQ Model
Volume IV - A Comparison of the; SAI Airshed Model and the
LIRAQ Model
This work is to a large extent based on the photochemical modeling
experience gained in the San FranciSco Bay Area in support of the 1979
Bay Area Air Quality Plan. The following individuals made significant
contributions to this work:
Association of bay Area Governments
Bay Area Air Quality Management District -
Ronald Y. Wada
(Project Manager)
M. Jane Wong
Eugene Y. Leong
Lewis H. Robinson
Rob E. DeMandel
Tom E. Perardi
Michael Y. Kim
Lawrence Livermore Laboratory
Systems Applications, Inc.
William H. Duewer
Steven D. Reynolds
Larry E. Reid
The authors wish to express their appreciation to John Surtmerhays, EPA
Project Officer in the Source Receptor Analysis Branch of OAQPS, for his
thoughtful review and conments on earlier drafts of this report.

-------
TABLE OF CONTENTS
Ti tle	Page
1.	SUMMARY	,	1
2.	INTRODUCTION 	 3
3.	MODEL SENSITIVITY TO SPATIAL RESOLUTION IN THE 5
EMISSIONS DATA BASE 	
Purpose 	 5
Methodology 	 6
Results 	 7
Conclusions	10
References	11
4.	SENSITIVITY OF SHORT-TERM AMBIENT N02 CONCENTRATIONS
TO REDUCTIONS IN HC AND NO EMISSIONS	13
Purpose arid Background	13
Methodology	13
Prototype Day Selection	13
Baseline and Sensitivity Scenarios	14
Resul ts	•	15
Preliminary Evaluation of Model
Performance	
Emission Sensitivity Results	15
Results of Previous Simulations	19
Conclusions	19
References	25
b. MODELING THE DOWNWIND EFFECTS OF CHANGING BAY AREA
HC AND NOx EMISSIONS	27
Purpose	27
Methodology 	27

-------
TABLE OF CONTENTS (Continued)
Ti tie	Page
The Modeling Region 	27
Meteorological Input Fields 	27
Emissions Inventory 	29
Initial and Boundary Conditions 	29
Results 	 33
Model Performance in the
Central Valley 	 33
Comparability of b x 5 km and
8 x B km Baseline Simulations	34
Sensitivity of Central Valley 03 to
Uay Area Emissions	34
Summary and Conclusions 	34
Recommendations	44
References	45
Appendix A: Ozone, Hydrocarbon and Nitric Oxide Concentration Maps for
Different Source Inventory Distribution Patterns.
Appendix B: MASCON-Generated Mass Flux and Inversion Base Height Maps
for the 5 November 1976 Prototype Day and Selected Maps and
Time Histories of 03 and M02 for the 1985 Baseline and
Sensitivity Simulations.
Appendix C: Selected Mass-Flux Streamline ana Inversion Base Height
Fields for the 20 August 1973 and 26 July 1973 Prototype
Days.

-------
LIST OF FIGURES
Fi9ure No.	Uescnption	Page
3-1	Regional Ozone Maxima from L1RAQ Simulations Using
Three Spatial Distributions of 1975 Emissions	9
5-1	Enlarged 160x160 km Long-Range Transport Study Area
and 100x100 km Region Used in Previous Bay Area LIRAQ
Simulations	28
5-2	Spatial Distribution of 24-hour Mean Emission Rates
for Nitric Oxide, 197b Emissions Inventory	31
5-3	Hourly Variation of HC and N0X Emissions in
Sacramento/Stockton Area	32
5-4	1300 PST Surface Ozone Concentration Field, 20 August
1973 Meteorology, 1975 Emissions, Standard Gridded
Region with 5x5 km Grid Cells	35
5-5	1300 PST Surface Ozone Concentration Field, 20 August
1973 Meteorology, 1975 Emissions, with the Expanded
8x8 km Gridded Region	36
5-b	1300 PST Surface Ozone Concentration Field, 26 July
1973 Meteorology, 1975 Emissions, Standard Gridded
Region with 5x5 km Grid Cells	37
5-7	1300 PST Surface Ozone Concentration Field, 26 July
1973 Meteorology, 1975 Emissions, with the Expanded
Modeling Region of 8x8 km Grid Cells	38
5-b	1300 PST Surface Ozone Concentration Field, 26 July
1973 Meteorology, 1985 Baseline Emissions	39
5-9	1300 PST Surface Ozone Concentration Field, 26 July
1973 Meteorology, 60% HC Reduction from 1985 Baseline
(Bay Area Only)	40
5-10	1300 PST Surface Ozone Concentration Field, 26 July
1973 Meteorology, 43% Hydrocarbon and 20% NO
Reduction from all 1985 Baseline Emissions, Boundary
Values, and Initial Conditions	41
5-11	1300 PST Surface Ozone Concentrations, 2b July 1973
Meteorology, 60% Hydrocarbon Reduction from all 1985
Baseline>Emissions, Boundary Values, and Initial
Conditions	42
l/

-------
LIST OF TABLES
[able No.	Description	Page
J-1	inventory Summaries and Ozone Predictions	8
4-1	Comparison of Observed vs LIRAQ-Simulated Hourly
Averages of Uzone (pphm), 5 November 19 7 0
Meteorology, 197b Emissions	16
4-2	Comparison of Observed vs LIRAQ-Simulated Hourly
Averages of N02 (pphm), 5 November 1976 Meteorology,
19/b Emissions	17
4-3	Sensitivity of Bay Area 03 Maxima to HC and NO
Emission Reductions, Based on LIRAQ Simulations Using
198b Emissions and 5 November 197b Meteorology	18
4-4	Ozone Sensitivity to Emission Reduction Strategies in
Selected Bay Area Sub^egions, 1500 PST, 5 November
19/b Meteorology	.	20
4-b	Sensitivity of Bay Area N02 Maxima to HC and NO
Emission Reduction Strategies, Based on LIRAQ
Simulations Using 198b Emissions and 5 November 1976
Meteorology	21
4-b	N02 Sensitivity to Emission Reduction Strategies in
Selected Bay Area Subregions, 1/0u PST, 5 November
19/b Meteorology	22
4-7	Effect of NO and HC Emission Reductions on Bay Area
OS and NO2 Maxima (Based on LIRAQ Simulations Using
26 July 1973 Meteorology and 1985 Emissions
Projections)	23
b-i	Source Inventory Totals for the Sacramento/San
Joaquin Valley Area Used in LIRAQ Transport Studies	30
5-2	Observed vs Predicted Ozone in the Central Valley	33

-------
1. SUMMARY
A series of photochemical modeling experiments, using the LIRAQ model
applied to the San Francisco liay Area, was performed to investigate: a)
model sensitivity to the spatial resolution of the gridded emissions, b)
the impact of future HC and NO controls on the future Bay Area N02
levels, and c) the future impact of Bay Area HC and NO reductions on 03
in adjacent downwind valleys.
The emission resolution experiments consisted of three simulations
wherein emissions were smoothed over a) 5 x 5 km areas, b) 10 x 10 km
areas, and c) distributed according to population. It was found that
simulated 03 concentrations are sensitive to emission distribution
patterns. Changing from 5-km to 10-km resolution changed the 03 maximum
concentration by up to 10%. When emissions were distributed
proportional to population substantial changes occurred in the timing
and magnitude of the 03 maxima. The results suggest that short-cut
methods should not be used for source inventory distribution.
The short-term NO^ experiments consisted of simulating 03 and N02 fields
under meteorological conditions favoring high N02 buildups. Model
performance in approximating observed space/time distributions of 03 and
N02 on the prototype day was judged to be adequate. Sensitivity runs
were made using three combinations of HC and NO reductions from a
projected 19H'j inventory. The results suggest that HC control is the
most effective strategy for both 03 and N02. Control of NO tends to
increase local 03 and decrease N02 slightly. N02 was generally less
sensitive to precursor reductions than was 03.
The long range transport experiments consisted of modifying LIRAQ to
simulate an expanded 160 x 160 km region that included the Bay Area
"source" region and portions of the Sacramento and San Joaquin Valleys
as a "receptor" region. We investigated the sensitivity of downwind 03
to varying HC and NO emissions from projected 1985 values. A base year
simulation with 1975 emissions yielded reasonable 03 concentrations in
the downwind study region on both meteorological prototype days used.
The emission sensitivity results showed that for the (26 July 1973)
prototype day, downwind 03 was strongly influenced by specification of
initial and boundary conditions. It was not possible to assess the
downwind effects of Bay Area emissions changes from the simulations that
were run in this study. It is believed that such assessments are
feasible with proper selection of meteorological prototype conditions
and with reasonably accurate estimates of initial and boundary
parameters.
-1-

-------
2. INTRODUCTION
L1RAQ is a grid-based photochemical model that has been in use for
several years in the San Francisco Bay Area. The 1979 Bay Area Air
Quality Plan was based on LIRAQ ozone modeling results using historical
and projected emission inventories. During that effort, several issues
arose which clearly needed more examination than deadlines permitted.
These issues included:
•	Effectiveness of HC control measures vs. NUx control measures
in reducing ozone and N02 concentrations.
•	Long range transport -- effects of local control strategies on
downwind receptors, including receptors located outside the
standard 100 km by 100 km modeling area.
•	Degree of resolution required in the source inventory to
produce acceptable modeling results; possibility for savings
through use of inventory shortcuts.
t Effects of controlling mobile sources vs. effects of
controlling stationary sources.
•	Sensitivity of model ozone results to hydrocarbon reactivity
classes — number of classes, distribution of inventory betwen
classes, and effects of control measures affecting only
certain classes.
These issues are of interest to the modeling community and to many
people involved in air pollution control and air quality planning.
The first three issues were of sufficient interest to EPA that a
contract was awarded through the Association of Bay Area Governments
(ABAG) so that the local modeling group could investigate the importance
and effects of certain actions. The experimental plan emerged through
an iterative process involving all of the cooperating agencies: EPA,
ABAG, the Bay Area Air Quality Management District (BAAQMD), Systems
Applications, Inc. (SAI), Lawrence Livermore Laboratories (LLL), and the
California Air Resources Board (ARB).
The final experimental design included three separate sections: 1)
emission inventory patterns, 2) short-term ambient N02 concentrations,
and 3) long range transport of pollutants.
Effects of emission inventory patterns are addressed in Section 3 of
this report. Three different inventory patterns were tested for each of
two different years. The 1975 Bay Area inventory represents a baseline
case with relatively stringent controls on emissions of organic
compounds. A 198b inventory year was also studied, where very stringent
and comprehensive control measures were assumed. The three emissions
patterns used for each year were: 5x5 krn grid resolution, 10 x 10 km
resolution, and pure population-based emissions distribution (over a 5 x
5 km grid).
-3-

-------
For the short-term NOZ question, <\ new prototype meteorology day was
developed for the LlRAl) model runs. Previous meteorological data had
been derived from historical 
-------
3. MODEL SENSITIVITY TO SPATIAL RESOLUTION IN THE
EMISSIONS DATA BASE
PURPOSE
There is a great deal of interest in the degree of data resolution
required to produce acceptable photochemical modeling results. This
interest is based primarily on cost considerations. Because
photochemical models are very data intensive, the choice of spatial and
time scales has a critical effect on the total resources that will be
required for data collection, input file preparation, computer demands,
and output analysis. The photochemical version of LIRAQ has been run
with 5-km grid squares in a 20 by 20 array, for a total grid area
coverage of 100 km by 100 km. The emissions patterns must be defined,
for each pollutant of interest, over the modeling area.
This experiment was designed to test the effects of different techniques
for spatial allocations of emissions. Because it is impossible to
locate and measure each source of pollutants in a modeling region, some
approximations are required in the preparation of emissions inventories.
The modeler normally receives an aggregated (regional, annual average)
source inventory, which must be allocated over space and time to fit the
model data requirements. Some techniques for spatial and temporal
resolution have been published by EPA (1974) and Perardi et al_. (1979).
This project does not consider the technical merits of one or another
technique. We are interested only in comparing the model outputs to see
first if there are any noticeable differences, and second, if less
expensive techniques can provide acceptable results.
Three patterns of emissions distributions were tested for this project:
1.	"b x 5." This is the standard disaggregated inventory
used in most LIRAQ runs, including the 1979 Bay Area Air
Quality Plan. Point, area, and mobile sources were
distributed as nearly as possible to their actual
locations in 5 x 5 km grid squares* over a 100 x 100 km
region.
2.	"lu x 10." For these runs emissions were averaged over
10 km x 10 km grid squares—four times as large as the 5
x 5. The preparation of a 10 x 10 inventory would be
somewhat less expensive than a 5 x 5, and there would be
a further savings potential in computer time for model
execution. As a first estimate, computer time would be
one-fourth as much for the same size area, or an area
four times as large could be run for the same time.
*Actually most of the data were first compiled by 1 x 1 km grid squares
for LIRAQ I.
-5-

-------
"Pop." In these inventories the model area inventory
totals, w ere 
  • 0.1 ton/day), 2) distribution of small point sources and area sources by association with 19 employment and land use categories, and 3) addition of mobile source emissions derived from a travel model and trip tables. The process has been described in some detail by Perardi et al. (1979). Stationary source emissions are actually maintained on a~T km grid basis, for use with some (non-reactive) LIRAQ applications. These are aggregatedd to 2 km or 5 km grids, as necessary. Only the 5 km grid has been used for photochemical modeling, because of computer space limitations, with a large number of grid squares and large number of equations to be solved for each grid square and each time step. The 10 x 10 km yrid square inventory was derived from the existing 5x6 by averaging groups of four 5x5 emission rates into one 10 x 10 emission rate. The model runs still used a 5 x 5 calculation grid but emissions were introduced on a 10 x 10 grid for the entire simulation period. The LlRAq model was "later modified to perform calculations on a larger grid size (8 x 8 km) and those results are discussed in Section 5 of this report, briefly, it appears that the model is more sensitive to emissions resolution than to calculation grid size. The population-distributed inventory was prepared by allocating the total emissions for all pollutants over the Bay Region on the basis of population density. The base year population and projections were provided by the local COG, the Association of Bay Area Governments (1977). The original census data covered about 1,000 census tracts over an area of about 20,000 km2. The populated areas are actually only about b,000 km2, the balance being essentially uninhabited (bays, tidelands, marshes, mountains). Regional emissions were distributed proportional to population over the 6,000 km2 of developed or developable land on a 5 x 5 km grid basis. The three types of inventories were prepared for each of two years--a 1975 base year and a 1985 control strategy scenario with large reductions in the hydrocarbon inventory. -6-

  • -------
    RESULTS
    1 ho aggregated inventory totals and ozone results are shown in Table
    J-L. The three different inventories tested (b x b, 10 x 10, and pop.)
    did produce significantly different ozone predictions, especially for
    the 19/b base year. For the 19bb projections, with comprehensive
    control strategies and 50X hydrocarbon reductions, the differences were
    less notable.
    The modeling output of fundamental interest is the regionwide ozone
    maximum, in the second column from the right. This is the highest 03
    concentration predicted anywhere in the modeling region at any time of
    the day. This value changed by about 25% over the three 1975 base year
    runs, and about 15'i. in the 198b runs. The basic 03 patterns are similar
    in contour shapes because the same meteorology data were input in each
    run, but the timing, magnitude and spatial resolution of ozone were all
    affected by the emissions inventory changes. The precursor inventory
    totals of HC and NO were essentially equal for a given year; only the
    distribution patterns were changed.
    Figure 3-1 shows the regionwide high ozone, as a function of time, for
    the three 1975 inventories tested. The baseline run, with standard 5 x
    b km inventory, produced the highest 03 prediction of .20 ppm at 1300
    hours. The 10 x 10 has the same timing with a .18 ppm maximum. We
    attribute the difference to a smearing of emissions over the larger
    areas, with lower resulting cell concentrations of precursors, and lower
    reaction rates. Figures A4 - A9 in Appendix A show the hydrocarbon* and
    NO concentration patterns for 0900 hours. These early precursor
    concentration maps are closely related to emission patterns, which are
    not directly available in graphic form. Comparing A4 and A6 shows for
    example the striking dilution of the ban Francisco high HC "fingerprint"
    by the change from S x 5 to 10 x 10 emission resolution. This dilution
    and spreading of anissions takes place over the entire area, and is more
    evident in areas where high-emissions squares are bordered by
    1 ow-emissions neighboring squares in the 5x5 format. Thus, San
    Francisco emissions were substantially diluted, while Oakland and San
    Jose highs were less affected.
    The difference between 5x5 and Pop. ozone production is even more
    striking. The high value drops from .20 to .15 ppm and occurs three
    hours later. In the latter case there is a notable redistribution of
    emissions. NOx emissions, particularly, are removed from normal point
    source locations and transferred to (residential) population areas. In
    the baseline inventory some area sources are distributed on population
    or population-related variables. Less than 40* of organics and less
    than 15% of MOx is population related in the baseline 5x5 inventories.
    In the Pop. inventory, by contrast, 100'fc of each pollutant was
    distributed by population. The result is that all point source, area,
    airport and vehicle emissions are transferred to population centers.
    Cities such as San Francisco, Oakland and San Jose tend to remain as
    ~Actually a refpV~eserTtative LIRAQ hydrocarbon group, "HC2", corresponding
    to alkanes and less reactive aromatics.
    -7-
    

    -------
    Table 3-1. inventory Summaries and Ozone Predictions
    | Inventory	Model Inventory [L ]	Model Predictions
    Year Pattern	Totals (tons/12 hr) kegionwide	Time of
    j	Maximum	Occurence
    	
    		...
    _ 	....	
    HC2 m
    NO
    03 (ppm)
    (hour)
    1975
    5 x
    5 km
    521
    202
    .20
    1300
    
    10 x
    10 kin
    514
    1%
    AH
    1300
    
    Pop.
    Di str.
    526
    202
    .15
    1600
    198b
    5 x
    5 km
    2b0
    172
    .13
    1500
    
    10 x
    10 km
    25tt
    168
    .13
    1500
    
    Pop.
    Distr.
    259
    169
    .11
    1500 i
    _i
    L. The model inventory is the sum of emissions in the model area for
    the critical 12 hour period from 4 am, when the simulation begins,
    to 4 pm, when the latest ozone maximum occurs. The model area is
    the specific 100 km by 100 km square grid (within the Greater Bay
    Area inventory region) chosen for LIRAQ model runs.
    2. HC2 is one of the three hydrocarbon reactivity classes used in the
    LIRAQ model. It is representative of total organic emissions and
    includes about 7U« of the total HC mass. Slight variations in the
    inventory totals for HC2 (and NO) are due to rounding errors from
    the model inventory summation readouts. Precursor totals for a
    given year are essentially equal for the three different source
    inventory patterns.
    

    -------
    .'DO
    180
    160
    140
    120
    100
    • 'i x ') km (briscl irif)
    ¦ 10 x 10 km omissions
    A P o p u 1 a t i o n - d i <; t r i b i j t; e d
    emissions
    Data points show highest.
    IJU	I I » u > Til WW II I	I.	j
    instantaneous ozone concentration / ¦'
    (parts per billion) predicted in
    the modelinq region.
    80
    1
    JULY 26, 1973 Meteoroloqy
    I	i	l	l	l	I	I	I	i	i	»	'	i	i
    06 08 10	12 14	1 fj ]'¦
    HOUR (PST)
    Figure 3-1. Regional ozone maxima from LIRAQ
    simulations usinq three soatial
    distributions of 1975 emissions
    -9-
    

    -------
    high emission areas and may oven increase, especially in UOx. The large
    NOx emissions from power plants and major industries along the Carquinez
    Straits would be moved to more densely populated areas.
    In an area ot low lU./NOx ratios, this has the effect of increasing the
    NUx content of precursor parcels and lowering the? expected downwind
    ozone. It is possible that the long range effects would be to increase
    03 formation, but this did not occur within the 100 x 100 km study area.
    For the L985 series, the results are not so clear cut. The difference
    between the 5 x S and 10 x 10 inventory runs is very small: . 1J0 ppm
    for the 5 x b, compared to .132 ppm for the 10 x 10. The results are
    presented to three significant figures in order to show that the 10 x 10
    result actually came out higher. This is the opposite of the 1975
    result and was not expected. The reversal itself is so small as to be
    insignificant, but the disappearance of the 1975 difference is of
    interest. The reduced 198b inventories resulted in smaller differences
    overall (.11 to .13 ppm from Pop. to 5 x 5, compared to .15 to .20 ppm
    for the corresponding 1975 patterns).
    In general, models are more credible and useful as they treat more
    explicitly the relevant physical phenomena. The results of this study
    show that the model does respond to the degree of physical reality in
    the inventory. Thus there is a motivation to incorporate as much
    reality as possible in the source inventory. The final choice of
    inventory resolution will be determined by a balance of several factors.
    Among these are the degree of detail available in existing source
    inventories, project time and budget, size of the region to be modeled,
    and cost of required computer services.
    CONCLUSIONS
    •	Simulated ozone concentrations from the LIRAQ model are sensitive
    to the spatial distribution patterns of the NOx and HC emission
    i nventori es.
    •	A change in the resolution of a gridded inventory, from 5 km to 10
    km cell length, can produce changes up to 10^ in the predicted
    ozone high.
    •	Distributing all omissions proportional to population substantially
    changed the timing and magnitude of the maximum estimated ozone
    concentrations.
    •	LIRAQ model performance, with respect to predicting regionwide high
    hour-ozone, was significantly better with more realistic source
    i nventories.
    •	The results of this work suggest that short-cut methods of source
    inventory distribution should not be used.
    -10-
    

    -------
    REFERENCES
    L. Association of Hay Area Governments, "Summary Report, Provisional
    Series J Projections," Berkeley, California, March 19/7.
    2.	Perardi, T. I., et al., "Preparation and Use of Spatially and
    Temporally Resolved "Emission Inventories in the San Francisco Bay
    Region," Journal of the Air Pollution Control Association, Vol. 29,
    No. 4, pp. 3S8-364, April 19797	
    3.	U.S. Environmental Protection Agency, "Guidelines for Air Quality
    Maintenance Planning and Analysis," Volumes 8 and 13, 1974.
    -11-
    

    -------
    4. SENSITIVITY OF SHORT-TERM AMBIENT N02 CONCENTRATIONS
    TO REDUCTIONS IN HC AND NO EMISSIONS
    PURPOSE AND BACKGROUND
    The goals of this study are: 1) to evaluate the LIRAQ photochemical
    model's usefulness for assessing effects of HC and NO emission controls
    on ambient NO^ concentrations over time periods of one hour to one day,
    and 2) to assess the sensitivity of future (1985) ambient N02
    concentrations to HC and NO emission changes under meteorological
    conditions that produce high N02 levels.
    LIRAQ was useful in the development of the Bay Area's non-attainment
    plan for 03. Sensitivity analyses using projected future emissions
    revealed that, while HC reduction is effective for controlling 03, NO
    reduction tends to increase 03 locally. Because of time and budgetary
    constraints and because the Federal N02 standard has never been exceeded
    in the Bay Area, we did not conduct any N02 modeling studies.
    Models like LIRAQ usually simulate conditions over periods of a day or
    less and are therefore not well suited for evaluating impacts on
    long-term standards like the Federal one-year N02 standard. Over short
    periods, however, photochemical grid models are appropriate.
    LIRAQ simultaneously tracks several of the primary pollutants including
    CO, NO and three classes of reactive hydrocarbons. It also treats some
    of their photochemical derivatives including 03 and N02. LIRAQ had
    never been applied to the meteorological conditions (usually in early or
    mid-autumn) when the highest N02 levels occur. Also, the model's past
    performance for N02 was not as good as it was for 03. This result is
    not surprising when one considers that N02 performance was a secondary
    rather than a primary criterion in the development of the model
    chemistry.
    The EPA is considering the adoption of a short-term N02 standard.
    California's current one-hour standard of 0.25 ppm is sometimes exceeded
    in parts of the Bay Area. Therefore it is prudent to evaluate future HC
    and NO controls in terms of their impacts on both 03 and N02. Then it
    will be possible to optimize strategies for meeting air quality goals
    for both pollutants.
    METHODOLOGY
    Prototype Day Selection
    Selection of a prototype day was based upon two criteria. First, we
    wanted a day with widespread, sustained high N02 levels in the Bay Area.
    Second, the day had to be in 1974, 1975 or 1976 to ensure that the 1975
    emission inventory was fairly representative of actual emissions on that
    day.
    -13-
    

    -------
    Ihe day chosen, b November 197b, was the fourth day of a seven-day
    e> isode of extremely restrictive dispersion conditions during which
    da.ly regional N02 maxima ranged between 0.24 and 0.30 ppm. In fact,
    when the duration, intensity and spatial extent of very high levels of
    NOId, particulates, CO and S02 are considered, this was perhaps the Bay
    Area's most severe air pollution episode of the decade. Oxidant levels
    were only moderate, however, and reached 0.13 ppm on 5 November. This
    is an unusually high value for the Bay Area so late in the year.
    Gridoed, mass-consistent flow and inversion base height fields for LIRAO
    were generated from analyses of a large number of surface wind
    measurements, winds and temperature aloft from the Oakland National
    Weather Service station, solar radiation data from several BAAQMD sites,
    and a few SODAk soundings provided by SRI International. The MASCON
    code preprocessed these data to yield the necessary mass flux and
    inversion base height fields required by LIRAQ. Examples of these
    preprocessed fields are shown in Appendix B, Figure series B-l and B-2.
    Baseline and Sensitivity Scenarios
    A series of simulations was performed 1) to assess LIRAQ's ability to
    reproduce the space/time distributions of 03 and N02 observed on 5
    November 1976, and Z) to evaluate future (1985) sensitivity of N02 to
    changes from projected HC and NO emissions. A "verification" run was
    required because LIRAQ had never been run using 5 November 1976
    meteorological conditions. The 197b emission inventory was used—no
    inventory for 1976 was available. Actual differences in emissions
    between the two years are throught to be small, however.
    The 1985 future-year simulations consisted of a baseline run with
    projected 1985 emissions plus three sensitivity runs with the following
    emission reductions:
    •	Strategy 1: 50% reduction in hydrocarbon emissions;
    •	Strategy 2: a 5% reduction in hydrocarbon emissions and 25%
    reduction in NO emissions;
    •	Strategy 3: b0% reduction in hydrocarbon emissions and 25%
    reduction in NO emissions.
    This matrix of runs was selected so that we could evaluate N02 and 03
    responses over a wide range of precursor control scenarios. The percent
    reductions were applied uniformly in space and time throughout the
    gridded modeling region. Boundary parameters for the 1985 base case
    were reduced from 1975 levels in proportion to emissions; initial
    conditions were the same in all runs.
    

    -------
    KESUL1S
    Preliminary Evaluation of Model Performance
    Before one can judge the validity and implications of the sensitivity
    experiments with any confidence one must first determine whether the
    model does a reasonable job of transforming primary emissions to
    concentrations of their photochemical derivatives. Tables 4-1 and 4-2
    show observed and simulated one-hour averages of 03 and N02 at eleven
    BAAQMD monitoring locations throughout the modeling region.
    Overall, the observed and simulated 03 values agree very well. The
    predicted and observed regionwide maxima only differed by one pphm.
    LIRAQ correctly predicted high concentrations in San Jose, moderate
    values in Fremont and Livermore and very low values in Concord. Phasing
    of the hourly values was also good. For example, the time of predicted
    vs. observed maxima differed by more than one hour at only one station.
    Only once did an hourly value of the predicted 11-station maximum differ
    by as much as 3 pphm from observations. Model performance was poor at
    Pittsburg, where simulated concentrations were highly boundary-condition
    dependent because of inflow along the northern boundary during much of
    the day. Vallejo was similarly affected. Elsewhere, LIRAQ
    underpredicted at San Rafael and Redwood City and overpredicted at San
    Francisco and Burlingame. However, considering the difficulty of
    photochemical modeling in an air basin as geographically complex as the
    Bay Area, model performance for 03 is regarded as encouragingly good.
    Agreement for N02, while not as good as that for 03, was improved when
    compared with past performance of LIRAQ for other meteorological
    scenarios. Duewer, crt al. (1978) found that, for the 20 August 1973 and
    26 July 1973 prototype 
    -------
    Table 4-1
    Comparison of Observed vs LIRAQ-Simulated Hourly
    Averages of Ozone (pphra) , 5 November 1976 Meteorology,
    1975 Emissions
    HOUR BEGINNING (PST)
    STATION (MAP SYMBOL)
    
    04
    05
    06
    07
    08
    09
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    MAX
    San Francisco (DSF)
    OBS.
    0
    0
    0
    0
    0
    0
    1
    1
    1
    1
    1
    0
    0
    0
    0
    0
    0
    0
    1
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    2
    3
    4
    4
    3
    1
    0
    0
    0
    0
    0
    0
    4
    Burlingame (DBU)
    OBS.
    0
    0
    0
    0
    0
    0
    0
    1
    1
    ->
    C.
    2
    o
    1
    0
    0
    0
    0
    0
    2
    
    LIRAQ
    0
    0
    0
    0
    0
    r\
    \J
    1
    1
    2
    4
    5
    4
    1
    0
    0
    0
    0
    0
    5
    Redwood City (DRC)
    OBS.
    1
    1
    1
    1
    2
    3
    3
    r-
    D
    7
    8
    9
    7
    4
    3
    3
    3
    3
    2
    9
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    1
    2
    3
    4
    5
    4
    2
    0
    0
    0
    0
    0
    5
    San Jose (DSJ)
    OBS.
    0
    0
    0
    0
    1
    1
    3
    4
    9
    10
    11
    3
    4
    1
    1
    1
    1
    1
    11
    
    LIRAQ
    0
    0
    0
    0
    0
    0
    1
    3
    9
    12
    11
    9
    5
    1
    0
    0
    0
    0
    12
    Fremont (DFR)
    OBS.
    1
    1
    1
    1
    1
    1
    2
    4
    6
    6
    7
    6
    3
    1
    1
    1
    1
    1
    7
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    2
    3
    5
    6
    6
    5
    4
    2
    1
    0
    0
    0
    6
    Richmond (DRM)
    OBS.
    1
    1
    1
    1
    1
    1
    2
    3
    2
    2
    2
    1
    1
    1
    1
    1
    1
    o
    3
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    2
    3
    4
    4
    3
    2
    1
    0
    0
    0
    0
    0
    4
    San Rafael (DSR)
    OBS.
    0
    0
    0
    1
    1
    2
    3
    3
    4
    4
    4
    2
    1
    1
    1
    1
    1
    1
    4
    
    LIRA Q
    0
    0
    0
    0
    0
    0
    1
    1
    2
    2
    2
    2
    1
    0
    0
    0
    0
    0
    2
    Vallejo (DVA or DVT)
    OBS.
    0
    0
    0
    0
    2
    3
    3
    3
    3
    4
    5
    5
    2
    0
    0
    0
    0
    0
    5
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    1
    2
    2
    2
    1
    1
    0
    0
    0
    0
    0
    0
    2
    Pittsburg (DPT)
    OBS.
    0
    0
    0
    0
    3
    -
    1
    5
    6
    7
    7
    3
    0
    0
    0
    0
    0
    0
    7
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    1
    0
    0
    0
    0
    0
    0
    0
    0
    0
    0
    0
    1
    Concord (DCO)
    OBS.
    0
    0
    0
    0
    0
    1
    1
    1
    2
    1
    1
    1
    0
    0
    0
    0
    0
    0
    2
    
    LIRAQ
    0
    0
    0
    0
    0
    1
    2
    2
    2
    2
    1
    1
    0
    0
    0
    0
    0
    0
    2
    Livermore (DLI)
    OBS.
    0
    0
    0
    0
    0
    1
    2
    4
    5
    5
    5
    4
    2
    1
    0
    0
    0
    0
    5
    
    LIRAQ
    0
    0
    0
    0
    1
    2
    4
    6
    6
    6
    6
    5
    4
    3
    2
    1
    0
    0
    6
    11 STATION MAXIMUM OBS,	111133359 10 11 9433332 11
    LIRAQ 000012468 12 11 9532100 12
    

    -------
    Table 4-2
    Comparison of Observed vs LIRAQ-Rimulated
    Hourly Averages of NO2 (pphm), 5 November 1976 Meteorology,
    1975 Emissions
    
    
    
    
    
    
    
    HOUR
    BEGIN!
    JING
    (PST)
    
    
    
    
    
    
    
    STATION (MAP SYMBOL)
    
    04
    05
    06
    07
    03
    09
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    2 0
    
    \ « * v
    San Francisco (DSF)
    OBS .
    7
    8
    8
    8
    9
    14
    16
    15
    10
    14
    13
    11
    12
    11
    12
    11
    9
    T
    J
    16
    
    LIRAQ
    5
    8
    11
    14
    19
    34
    37
    23
    27
    27
    26
    25
    23
    19
    16
    16
    16
    16
    37
    Burlingame (DBU)
    OBS .
    9
    10
    3
    9
    12
    10
    9
    13
    10
    —
    — —
    21
    21
    15
    14
    14
    15
    12
    21
    
    LIRAQ
    4
    5
    6
    7
    *7
    4
    3
    11
    14
    16
    18
    21
    25
    30
    30
    29
    29
    29
    29
    3 0
    REDWOOD CITY (DRC)
    OBS.
    -7
    Am
    2
    2
    2
    3
    3
    10
    12
    0
    10
    6
    8
    12
    15
    14
    13
    11
    0
    15
    
    LIRAQ
    5
    5
    6
    7
    7
    9
    10
    11
    11
    11
    12
    14
    18
    20
    19
    17
    17
    
    2 0
    SAN JOSE (DSJ)
    OBS .
    10
    7
    8
    10
    13
    17
    22
    24
    13
    	
    — —
    16
    19
    23
    26
    27
    2 C
    ' ?
    27
    
    LIRAQ
    6
    8
    11
    11
    12
    15
    23
    29
    29
    24
    20
    20
    24
    27
    25
    24
    25
    23
    29
    FREMONT (DFR)
    OBS.
    4
    5
    4
    4
    5
    9
    9
    9
    7
    7
    9
    10
    12
    13
    12
    12
    11
    
    13
    
    LIRAQ
    4
    5
    6
    6
    6
    7
    9
    9
    7
    7
    9
    12
    16
    14
    12
    11
    12
    12
    16
    RICHMOND (DRM)
    OBS.
    6
    6
    5
    £
    6
    6
    n
    i
    8
    /
    6
    7
    A
    rt
    9
    9
    3
    5
    
    
    9
    
    LIRAQ
    3
    3
    4
    4
    5
    6
    9
    8
    9
    9
    10
    10
    9
    9
    6
    5
    5
    
    10
    SAN RAFAEL (DSR)
    OBS.
    5
    6
    6
    8
    10
    7
    3
    4
    	
    3
    4
    S
    10
    9
    9
    9
    
    
    10
    
    LIRAQ
    3
    "4
    6
    6
    7
    8
    11
    12
    13
    13
    14
    15
    16
    13
    11
    10
    10
    
    16
    VALLEJO (DVA or DVT)
    OBS.
    7
    7
    7
    6
    6
    —
    —
    —
    —
    —
    3
    3
    7
    9
    11
    0
    0
    a
    0
    
    LIRAQ
    2
    2
    4
    4
    4
    5
    5
    4
    4
    3
    2
    2
    3
    2
    2
    2
    2
    
    5
    PITTSBURG (DPT)
    OBS.
    7
    7
    7
    8
    9
    4
    4
    6
    4
    3
    4
    4
    3
    13
    11
    11
    9
    -1
    13
    
    LIRAQ
    3
    4
    4
    4
    4
    5
    7
    3
    3
    7
    6
    6
    6
    6
    5
    4
    1
    4
    Q
    CONCORD (DCO)
    OBS.
    4
    4
    4
    5
    7
    9
    10
    7
    6
    9
    9
    9
    9
    10
    9
    8
    3
    7
    10
    
    LIRAQ
    3
    4
    4
    4
    4
    4
    4
    4
    5
    5
    6
    6
    6
    6
    6
    5
    6
    £
    6
    LIVERMORE (DLI)
    OBS.
    3
    3
    3
    4
    6
    7
    7
    7
    6
    6
    6
    7
    9
    9
    9
    8
    7
    7
    Q
    
    LIRAQ
    3
    3
    3
    4
    4
    5
    5
    4
    3
    3
    2
    1
    2
    2
    2
    3
    4
    <
    5
    11 STATION MAXIMUM
    OBS.
    10
    10
    9
    10
    13
    17
    22
    24
    19
    14
    13
    21
    21
    23
    26
    27
    2 6
    19
    27
    
    LIRAQ
    6
    8
    11
    14
    19
    34
    37
    29
    29
    27
    26
    25
    30
    30
    29
    29
    29
    29
    37
    

    -------
    Table 4-3.
    Sensitivity o) Bay Area U-j maxima to nc and
    NO emission reductions, based on LIRAQ simulations
    usimj 1985 emissions and 5 November 1976 meteorology
    Time (PST)	Ozone Concentration (ppb)
    
    1985
    Baseline
    50% HC
    Reduction
    25% HC & 25% NO
    Reduction
    50% HC & 25% NO
    Reduction
    0800
    11
    11
    10
    11
    0900
    24
    24
    23
    24
    1000
    43
    43
    42
    43
    1100
    56
    56
    55
    56
    1200
    64
    64
    65
    64
    1300
    72
    65
    71
    68
    1400
    96
    66
    77
    68
    1500
    104
    63
    85
    65
    1600
    80
    57
    72
    58
    1700
    54
    48
    56
    50
    1800
    41
    37
    41
    39
    1900
    28
    26
    28
    23
    i
    2000
    19
    18
    20
    19
    2100
    13
    12
    14
    |
    13
    2200
    ; 9
    9
    10
    9
    -18-
    

    -------
    03 highs developed (see Figure B-3) the results are similar except in
    the Livermore Valley. Table 4-4 shows that precursor controls strongly
    affect 03 in and near the Bay Area's major population centers.
    Simulated 03 in the Livermore Valley, however, was primarily determined
    by initial and boundary conditions and was insensitive to controls (see
    also Figures B-5 and B-7).
    Tables 4-5 and 4-b show equivalent analyses of N02 sensitivity to HC and
    NO reductions. The results indicate that HC control is also effective
    for reducing N02. Control of NO also reduces N02, but to a much smaller
    degree. We also note N02 is generally less responsive to precursor
    controls than is 03. The somewhat different results for the Santa Clara
    Valley N02 "high" are suspect because the high center appears to have
    drifted out of the modeling region by 1700 PST in two of the simulations
    with emissions reductions (see Figures B-4, B-6 and B-8 for more
    details). The concentrations in the Santa Clara Valley may also have
    been more influenced by boundary conditions than those in the central
    Bay Area.
    Results of Previous Simulations
    The LIRAQ sensitivity runs that were performed for the 1979 Bay Area Air
    Quality Plan (ABAG, et al., 1979) using the 26 July 1973 prototype day
    yielded results for 03 and N02 that are comparable to those reported
    here. Table 4-7 summarizes these results. Again, N02 is much more
    sensitive to HC control than to NO control. The comparability of the
    N02 sensitivity results is interesting in that the two prototype days
    were meteorologically quite different and the resultant N02
    concentrations and time profiles were also quite different at most
    locations. We also note that, except for the Livermore Valley, 03 was
    somewhat more sensitive to HC controls in the 5 November simulations
    than in the 26 July simulations.
    CONCLUSIONS
    1. Model performance for the 5 November 1976 "N02 day" was excellent
    for 03 and, in many N02-prone locations, adequate for N02. Because
    the model is able to reproduce observed N02 concentrations fairly
    well it is reasonable to select it as an instrument for evaluating
    the sensitivity of N02 to precursor emissions.*
    *NeverthelessTTT02"~p'erformance was only a secondary consideration in the
    development of LIRAQ's photochemical reaction set. Thus further
    investigation of the model's ability to simulate N02 response to
    emissions (where smog results would provide the basis for comparison)
    would be desirable.
    -19-
    

    -------
    Table 4-4. Ozone Sensitivity to Emission Reduction Strategies*
    in Selected Bay Area Subregions, 1500 PST,
    5 November 1976 Meteorology
    a) Maximum 03 Concentration (ppb)
    Bay Area	- - 1975 ~ ~'19B!T
    Subregion		B_as_eJ_ine	 Basel ine Strategy 1 Strategy 2 Strategy 3
    San Francisco 126 pbb 83.6 ppb 14.9 ppb 69.2 ppb 26.9 ppb
    Santa Clara VI. 13b pbb 104.3 ppb 21.0 ppb 85.0 ppb 37.8 ppb
    Livermore Valley 70.5 ppb	68.6 ppb 62.9 ppb	67.4 ppb 64.8 ppb
    b) Percent 03 Change From 1985 Baseline
    Bay Area Subregion 		Strategy 1	Strategy 2 Strategy"?
    San Francisco Bay -82% -171 -68%
    Santa Clara Valley -80% -19% -64%
    Livermore Valley	- 8%	- 2%		-6%
    *Strategy 1: 50% reduction in hydrocarbon emissions
    Strategy 2: 25% reduction in hydrocarbon emissions and 25% reduction
    in NO emissions
    Strategy 3: 50% reduction in hydrocarbon emissions and 25% reduction
    in NO emissions
    -20-
    

    -------
    Table 4-5.	Sensitivity of Bay Area NOo maxima to HC and NO
    emission reduction strategies*, based on LIRAQ
    simulations using 1985 emissions and 5 November
    1976 meteorology
    NUo concentration ippoj
    Hour (PST)
    1985
    Baseline
    Strateav 1
    Strateav 2
    Strateav 3
    0400
    48
    48
    48
    48
    0500
    68
    70
    66
    66
    0600
    93
    94
    81
    81
    0700
    154
    155
    125
    125
    0800
    175
    174
    140
    140
    0900
    220
    186
    170
    149
    1000
    371
    221
    278
    199
    1100
    287
    183
    218
    165
    1200
    272
    182
    221
    178
    1300
    280
    204
    230
    198
    1400
    268
    219
    211
    205
    1500
    280
    220
    216
    200
    1600
    322
    220
    249
    200
    1700
    311
    220
    247
    220
    1800
    299
    200
    236
    220
    1900
    287
    200
    228
    200
    2000
    276
    200
    240
    200
    2100
    294
    205
    240
    »
    L
    180
    2200
    297
    210 j 243
    200
    ~Strategy 1: 25% reduction
    Strategy 2: 25% reduction
    NO emissions
    Strategy 3: 50$ reduction
    emissions
    in hydrocarbon emissions
    in hydrocarbon emissions
    in hydrocarbon emissions
    and 25% reduction in
    and 25% reduction in NO
    -21-
    

    -------
    Table 4-6. NO2 Sensitivity to Emission Reduction Strategies*
    in Selected Bay Area Subregions, 1700 PST,
    5 November l^'/b Meteorology
    a) Maximum N02 Concentration
    Bay Area	197$	1 955
    Subregion 		B as eJJno	Baseline Strategy 1 Strategy 2	Strategy 3
    S. F. Peninsula 397 pbb	311 ppb 190 ppb 247 ppb	181 ppb
    Oakland/East Bay 318 pbb	282 ppb 173 ppb 225 ppb	173 ppb
    Santa Clara VI. 359 ppb		303 ppb	 230 ppb	246 ppb _ 230 ppb
    b) Percent N02 Change From 1985 Baseline
    Bay Area Subregfon 		strategy	2	Strategy 3
    S. F. Peninsula	-39%	-21%	-42%
    Santa Clara Valley	-39%	-20%	-39%
    Livermore Valley	-24%	-19%	-24%
    ~Strategy 1: 25/i reduction in hydrocarbon emissions
    Strategy 2: 257, reduction in hydrocarbon emissions and 25% reduction
    in NO emissions
    Strategy 3: 50% reduction in hydrocarDon emissions and 2 5% reduction
    in NO emissions
    -22-
    

    -------
    Table 4-7. Effect of NO and HC emission reductions on Bay Area
    O3 and NO? maxima (based on LIRAU simulations using
    26 July 1973 meteorology and 1985 emissions
    projections*)
    Emission
    Changes
    Resultant Concentration Changes
    % HC Reduction
    % NO Reduction
    % O3 Change
    % NO2 Change
    0
    40
    +33
    -14
    40
    0
    -55
    -30
    40
    20
    -36
    -35
    80
    0
    -80
    -55
    80
    40
    -70
    -60
    Source: Waterland, etal., 1978.
    * Hore, an earlier 1985 baseline inventory was used that included regionwide
    emissions of 782 tons/day of HC and 725 tons/day of N0X (as NC^)- Elsewhere
    in this chapter a revised 1985 baseline inventory, with 835 tons/day HC and
    721 tons/day N0X was used. The change resulted from a revision in the EPA
    emission factors for mobile sources.
    

    -------
    Z. Based on Bay Area LIRAQ simulations using projected 1985
    emissions, we conclude that:
    a)	HC controls are effective for both U3 and N02;
    b)	NO controls increase 03 levels and produce relatively
    small reductions in N02; and
    c)	N02 is less sensitive to controls than 03.
    -24-
    

    -------
    REFERENCES
    1.	Association of Bay Area Governments, Bay Area Air Quality
    Management District, and Metropolitan Transportation Commission,
    "1979 Bay Area Air Quality Plan," January 1979.
    2.	De Mandel, R. E., L. H. Robinson, J. S. L. Fong and R. Y. Wada,
    "Comparisons of EPA Rollback, Empirical/Kinetic, and
    Physicochemical Oxidant Relationships in the San Francisco Bay
    Area," J. Air Pollution Control Assoc., 29, pp. 352-358, April
    1979.
    3.	Duewer, W. H., M. C. MacCracken and J. J. Walton, "The Livermore
    Regional Air Quality Model: II. Verification and Sample
    Application in the San Francisco Bay Area," J. Appl. Meteor., 17,
    pp. 273-311, March 1978.
    4.	Waterland, L. R., K. J. Lim, K. G. Salvesen, R. M. Evans, E. B.
    Hi gginbotham and H. B. Mason, "Environmental Assessment of
    Stationary Source NOx Control Technologies—Second Annual Report,"
    Accurex Corporation Report TR-78-116, July 1978.
    -25-
    

    -------
    5. MODELIMG THE DOWNWIND EFFECTS OF CHANGING BAY AREA
    HC AND NOx EMISSIONS
    PURPOSE
    The LIRAQ photochemical simulations used in the development of the 1979
    Bay Area Air Quality Plan (Association of Bay Area Governments, et al .,
    1979) indicated that local NOx reductions might actually increaseUay
    Area ozone concentrations. An inverse relationship between ozone and
    NOx is possible in regions with low HC - NOx ratios. Before making a
    final judgment on the desirability of NOx controls, however, it is
    advisable to evaluate the effects of local emission changes on downwind
    air basins.
    The work described in this section was designed to:
    1)	Expand the LIRAQ modeling region such that transport can
    be simulated from the Bay Area to the north and east as
    far as Stockton and Sacramento.
    2)	Evaluate model performance in the expanded modeling
    region under the contrasting meteorological conditions of
    26 July 1971s and 20 August 1973.
    j) Assess the sensitivity of ozone in the Central Valley
    (i.e., the Sacramento and San Joaquin Valleys) to Bay
    Area emissions under the restrictive meteorological
    conditions of 26 July 1973. (This was the prototype day
    used in the development of the 1979 Bay Area Air Quality
    Plan.)
    METHODOLOGY
    The Modeling Region
    In order to simulate transport from the Bay Area to the Central Valley
    the LIRAQ modeling region was expanded to a 160 x 160 km area that
    included the central Bay Area "source" region and Sacramento and
    Stockton as potential receptor locations. Because of computer
    limitations the LIRAQ-2 photochemical model is currently limited to 400
    grid cells. Therefore it was necessary to increase the grid cell size
    from the 5 x S km used in previous work to 8 x 8 km. The old 100 x 100
    km region and the new 160 x 160 km region are shown in Figure 5-1.
    Meteorological Input Fields
    It was also necessary to modify the inputs to the MASCON preprocessor
    for the two prototype days that were simulated. The MASCON code
    calculates mass-consistent gridded fields of mass flux and inversion
    base height. It also generates solar flux fields interpolated from a
    limited number of pyranometer measurements made at BAAQMD stations.
    -27-
    

    -------
    Figure 5-1
    123'
    Enlarged 160 x 160 km Long-Kange Transport
    Study Area and 100 x 100 km Kegion Used in
    Previous Hay Area IjTRAU .'>imu J ations .
    (The 8-Km colli; used in the enJarged study
    are not yhown.)
    3°'	l22'	3°'	4300
    IS
    £
    
    r?
    4120
    
    10 km i ' ' 1 'I "
    500
    520
    540 560 580 600 620
    UTM coordinate (easting) — km
    640
    660
    Legend:
    Enlarged 160 x 160 km Study Area.
    ,77771 100 x 100 km Region where most previous LIRAQ
    runs were made.
    -28-
    

    -------
    Unexpected, time-consuming problems associated with the generation of 8
    x B km MASCON files oh the Lawrence Berkeley Laboratories (LBL) computer
    forced postponement of the long range transport simulations until late
    in the contract period.
    Emissions Inventory
    Gridded emission inputs for all previous Bay Area LIRAQ simulations were
    limited to the nine county jurisdictional area of the BAAQMD. The
    expanded modeling region includes a large area outside of this
    jurisdiction. In order to generate gridded emissions for the exogenous
    region, county totals (in tons/day) for the five major pollutants
    (particulates, HC, NOx, S02 and CO) were obtained from the California
    Air Resources Board (CARB) (Bradley, 1979). County source inventories,
    provided for both 1975 and 1985, are shown in Table 5-1.
    Because of contractual time constraints and limited data availability a
    simplified method was used for spatial distribution of emissions.
    First, undevelopable land areas (e.g., mountain ridges, marshland, bays)
    were identified from U.S. Geological Survey maps. Next, county
    population totals were distributed over the developable land.
    Population projections for 1985 were obtained from the California
    Department of Finance (UOF, 1978). The emissions were distributed into
    1 x 1 km grid cells based on population. Finally, the 1 x 1 km gridded
    emissions were aggregated into 8 x 8 km cells. As an illustration of
    the resultant emissions pattern, Figure 5-2 shows the 24-hour
    distribution of nitric oxide emission rates for 1975.
    The temporal distribution of emissions was based primarily upon datii
    supplied by Systems Applications, Incorporated (SAI). In 1978 SAI
    prepared hourly emission estimates, from 0500 to 1800 PST, of both
    stationary and mobile sources in the Sacramento area (Reid, 1979).
    BAAQMD personnel used this and other information to develop the profiles
    of hourly variations in mobile, stationary-source and total emissions
    shown in Figure 5-3. The temporal distribution of emissions within the
    Bay Area was the same as that used in previous simulations. A detailed
    discussion of the preparation of gridded emissions in the Bay Area is
    presented by Perardi, et al_., 1979.
    Initial and Boundary Conditions
    As a result of expanding the modeling region, it was necessary to modify
    the background pollutant concentrations specified along the southern and
    eastern boundaries for the 26 July 1973 prototype day. These values are
    important because:
    1)	there was inflow along portions of these boundaries for
    part of the simulation period,
    2)	inflow along the northern half of the eastern boundary
    directly affects pollutant concentrations in the regions
    of greatest interest (Sacramento/Stockton), and
    

    -------
    Table 5-1. Source inventory Totals for tho Sacramento/San Joaquin
    Valley Area Used in LIUAQ Transport Studies
    a) 197b
    f
    1 - .Pollutant
    Area —„
    Emissions in Tons/Day
    Part. HC NOx S02 CO
    Sacramento County
    San Joaquin County1
    Solano County2
    Yolo Counvy
    33.9 132.0 76.6 5.6 567.0 i
    40.5 57.8 34.6 10.2 210.0
    9.4 7.6 6.3 .4 35.3
    27.8 23.0 15.1 2.1 104.0
    Total
    Bay Area AQMD
    111.6 220.4 132.6 18.3 916.3
    133.6 748.0 567.0 654.6 1,842.5
    Grand Total
    245.2 968.4 699.6 672.9 2,758.8
    b) 1985
    Emissions in Tons/Day
    i Area ;•
    Part.
    HC
    
    NO
    X
    so2
    CO
    
    	 		
    Sacramento County !
    San Joaquin County1
    36
    .8
    92
    .9
    70
    .5
    6
    .5
    205
    .6
    48
    .7
    45
    .3
    34
    .9
    15
    .4
    134
    .0
    1 Solano County2 i
    11
    .0
    5
    .6
    6
    .5
    
    .4
    24
    .0
    Yolo County 1
    31
    .9
    19
    .0
    16
    .0
    2
    .7
    79
    .4
    Total |
    128
    .4
    162
    .8
    127
    .9
    25
    .0
    566
    .4
    Bay Area AQMD !
    147
    .8
    557
    .6
    455
    .0
    647
    .9
    2,192
    .4
    Grand Total 1
    1
    276
    .2
    720.
    4
    582
    .9
    672
    .9
    2,758
    .8
    *80% of total county emissions (20% are estimated to be
    outside of the study area).
    Emissions only from the portion of the county outside of
    the BAAQMD.
    -30-
    

    -------
    ••300.
    
    «?*0.--
    E
    ::: 'il-fM:
    mm ..
    <-• 		
    ¦mmm.
    
    
    
    " " •:> ::: ;
    
    
    
    
    
    
    M 160.--
    'I.* ¦ J* • • . • • « »	. •>
    11*0
    »	S - _"r '
    <.)¦,. I i V-1— I ¦ I | t	| j • i i 1 i i 1_1 4- j
    E-W UTM coordinate (km)
    
    X-. Ll^fe' ^
    II
    ::::. : J	•*
    i:::::::::: :~tl::
    : *i
    1
    
    ~4-
    -i—i-
    _J—L_i -i. i. .
    Figure 5-2. Spatial distribution of 24-hour mean emission rates tor
    Nitric Oxide, 1975 emissions inventory (units: grams per
    second per grid square, and each grid square is 64 km )
    

    -------
    12
    11
    10
    9
    l/)
    C
    o
    Lf)
    .£ 7
    ra
    I a
    00 ,
    c
    a;
    " 4
    CJ
    -•	-•	•'
    :nd
    Stationar/ Sources
    • Mobile Sources
    J	1	1	I	I	I	I	t	I	I	I	I	I	I	I	1	I	I	I	I	L
    J	L
    9 10 11 12 13 14 15 16 1
    Tine (PST)
    1-
    21 22
    Figure ij-3. Assumed hourly variation of hydrocarbon and NOx emissions in Sacramento/Stockton Area.
    •:n\v: I'ro 1 iminary data supplied by Systems Applications, Inc.
    

    -------
    3) the expanded region excludes the heavily populated Santa
    Clara Valley to the south, and inflow from this region
    contains higher pollutant concentrations than flow into
    the southern boundary of the previous 100 x 100 km
    modeling region (see Figure 5-1).
    Based on air monitoring data from Sacramento and Stockton on 26 July
    1973, the hydrocarbon background parameters for the eastern boundary
    were increased from those used in simulations with the 100 x 100 km
    region. Background parameters for all species were increased at the
    southern boundary by factors of from 2 to 6. (Ouewer, et al., 1978,
    includes a detailed discussion of the specification of boundary values
    for earlier LIRAQ simulations.) The same modified boundary values were
    used in both the 1975 and 1985 baseline simulations. All but two of the
    emission sensitivity simulations used these same boundary values.
    For 26 July 1973, initial (0600 PST) pollutant concentrations were based
    on monitoring data from Stockton and Sacramento. It is not known
    whether data from these urban locations are representative of the
    Central Valley region as a whole, much of which is agricultural.
    Initial conditions in the Bay Area were essentially the same as in
    earlier simulations. The same initial conditions were used in both the
    1975 and 1985 baseline simulations. All but two of the emission
    sensitivity simulations used these same initial concentration fields.
    RESULTS
    Model Performance in the Central Valley
    In 1973 there were just two ozone monitoring sites, in Sacramento and
    Stockton, within the expanded 160 x 160 km modeling region from which
    model performance in the Central Valley could be evaluated. Table 5-2
    shows that for the two prototype days simulated, 20 August 1973 and 26
    July 1973, the 1975 baseline runs produced high-hour ozone maxima that,
    in general, agreed reasonably well with observations. At Sacramento,
    hourly values for the 26 July 1973 tracked observations quite well. In
    Stockton, however, LIRAQ predicted peak ozone levels that were lower and
    later in the day than the observed ozone levels. A discussion of
    previous verification studies of these two prototype days within the Bay
    Area is found in Duewer, et al^., 1978.
    Table 5-2
    Observed vs. Predicted Ozone in the Central Valley
    Maximum Hourly Average (ppb)
    Prototype Day Sacramento Stockton
    	LIRAQ Obs. LIRAQ OFsT
    26 July 1973	110 120 110 150
    20 August 1973	70 90	90 80
    -33-
    

    -------
    Comparability of b x b km_ jind H x 8 km baseline Simulations
    In order to determine the degree to which simulationr. with the enlarged
    region and H km cells agree with earlier local simulations with 6-km
    cells, LIRAQ runs with both grids were compared for two prototype days.
    Generally the region common to both grids displayed similar, but not
    identical, ozone concentration fields. This can be seen by comparing
    Figures 5-4 ana 5-b, which depict 1300 PST ozone contours from
    simulations with the standard and expanded modeling regions,
    respectively, on the relatively well-ventilated 20 August 1973 prototype
    day. Similarly, Figures 5-6 and 5-7 provide a comparison for the more
    stagnant and polluted conditions of 2b July 1973. Differences between
    these two fields are on the order of 10 percent, and are probably due to
    both the change in grid size and to the modified boundary conditions
    that were needed for this prototype day. Similar differences were found
    between the 5 x 5 km and 10 x 10 km simulations discussed in Section 3.
    Sensitivity of Central Valley Ozone to Bay Area Emissions
    A series of simulations was performed using the expanded modeling region
    in an attempt to assess the sensitivity of Central Valley ozone to
    changes in Bay Area precursor emissions. The baseline simulation
    represented a 198b control strategy scenario with large reductions in
    Bay Area hydrocarbon emissions. Initial and boundary conditions were
    the same as those used in the 1975 baseline simulations.
    Figures 5-8 and 5-9 show sample output from the 1985 baseline simulation
    and from a simulated reduction of Bay Area hydrocarbon emissions by 60%
    from the baseline levels. The two ozone fields differ only slightly
    over the Bay Area in that neither the pattern nor the location of the
    'fingerprint' changed, and the peak ozone value (0.06) shifted to the
    northwest but did not significantly change in magnitude. The two ozone
    fields are nearly identical in the Central Valley. The same can be said
    for the 1975 ana 1985 baseline simulations (see Figures 5-7 and 5-8) and
    for several other simulations in which precursor emissions were changed
    significantly. Clearly the effects of boundary and/or initial
    conditions overwhelm the effects of Bay Area emissions in this series of
    simulations.
    Two additional runs were made in which emissions were reduced throughout
    the modeling region and in which proportionate reductions were made in
    initial and boundary values. Sample output from each run, one
    representing a 43% hydrocarbon and 20% NOx reduction from the 1985 base
    case and the other representing a 60% hydrocarbon-only reduction, are
    shown in Figures 5-10 and 5-11. Ozone concentrations were significantly
    different in those runs, but it is difficult to assess the relative
    impacts of reduction in emissions, initial conditions and boundary
    conditions.
    SUMMARY AND CONCLUSIONS
    The completion of an operational version of LIRAQ-2 for the 160 x 160 km
    expanoed modeling region, complete with 1975 and 1985 emissions
    -34-
    

    -------
    >~220
    ~oco
    HI 80
    ~OL,
    HI 70
    ¦P >s
    •MC
    tsi
    Soul hern boundary gf 8 km ejfoanqjjpd
    ~*SJ \
    modeling £ regio:
    O	OOOOOOOOOO
    foa-mujt-cocDo-aiw
    in m in id ,n in ,n - 
    -------
    «
    4J
    e
    •H
    •O
    U
    O
    O
    o
    -Portion of the gridded region
    shown in Figure 5-4
    
    iE-01,
    tta e
    * 99 9
    *20 0
    W.f
    »•
    E-W UTM coordinate (kjr>)
    Figure 5-5. 1300 PST surface ozone concentration field, 20 August 1973
    meteorology/ 1975 emissions, with the expanded 8 x 8 km
    gridded region.
    

    -------
    (CM
    822E-01
    0.1035
    -o
    01 r
    0. 769E--0I
    0.1226
    -01,
    CO
    Southern Boundary of 3 x 6 kr. expanded
    modeling region
    4?« 3
    «ro 0
    E-W UTM coordinate (km)
    Figure 5-6. 1300 PST surface ozone concentration field, 26 July 1973
    meteorology, 1975 emissions, standard gridded region with
    5 x 5 km grid cells.
    

    -------
    
    ; *11 ¦1 <
    t
    CO
    I
    • $0Q£.
    0!
    I 1
    'M '
    ! «r*
    ;" J T .
    •' » -dp*
    E
    CD
    4->
    #Q
    c
    -o
    o
    o
    u
    S *?90 - -
    00
    I
    \
    /  ( H;
    \ <=> * »¦ i s : I i i I w / > i /'
    ~ ¦ ' 1 ' a/ •"»»f	 (
    ? ,
    \ 2 1-1. i a / ;
    —*	VLWCN gl I /—v
    ' x '-"Wl \
    •T \\ / '\ I Jpk \
    \V ¦'	\
    IT	' ' ••
    -l \ SU /; V,
    •»hv C"; ) * 'n / l M r ., ,
    -^ + ii(U^JLt
    • IMH	' +PPr< \
    C.6CVE
    .1	— *. 3	iiJ	3	*M J	J	tO 0
    E-W UTM coordinate (km)
    Portion of the gridded regior
    shown in Figure 5-6
    Figure 5-7. 1300 PST surface ozone concentration field, 26 July 197 3
    meteorology, 1975 emissions, with the expanded modeling
    region of 8 x 8 km cells. (The 'fingerprint' ozone low near
    Pittsburg is associated with the quenching effects of large N0X emissions
    from nearby power plants.)
    

    -------
    £
    ^ UK - •
    *0# 0	•	14# 0	*t«190.0	|M	«40.»	til t
    E-W UTM coordinate (km)
    Figure 5-8. 1300 PST surface ozone concentration
    field# 26 July 1973 meteorology, 1985
    baseline emissions.
    -39-
    

    -------
    "rt'> 1	WI1 0	Sh* 9	so o	MO O	AM d	m o	4	*49 9
    E-W UTM coor^iinatq (km)
    Figure 5-9. 1300 PST surface ozone concentration
    field, 26 July 197 3 meteorology,
    60% HC reduction from 1985 baseline
    (Bay Area only).
    -40-
    

    -------
    £
    j*:
    CD
    +->
    C
    •r—
    "O
    s-
    o
    o
    o
    00
    I
    o,ieoe 01
    -pb
    3191 -0
    
    H
    0,306 E
    ¦>«•<> P	540
    910 0	400.0	470 0	4«l 0
    E-W UTM coordinate (km)
    Figure 5-10. 1300 PST surface ozone concentration
    field, 26 July 1973 meteorology, 43%
    hydrocarbon and 20% NO reduction from
    all 1985 baseline emissions, boundary
    values and initial conditions.
    -41-
    

    -------
    0.S16E-0I
    T3
    z:
    IE-01
    •in
    12
    »« i
    E-W UTM coordinate (km)
    Figure 5-11. 1300 PST surface ozone concentrations,
    26 July 1973 meteorology, 60% hydrocarbon
    reduction from all 1985 baseline emissions,
    boundary values and initial conditions.
    -42-
    

    -------
    inventories for adjacent areas north and east of the Bay Area,
    represents an important addition to our photochemical modeling
    capabilities. Unfortunately the simulations to date did not provide a
    definitive assessment of the downwind impact of Bay Area precursor
    emission changes. Valuable modeling information was gained however, and
    is summarized as follows:
    1.	Central Valley ozone concentrations resulting from
    simulations using 1975 emissions and two 1973 prototype
    days were reasonably close to observed levels on those
    days.
    2.	Model performance does not appear to have changed greatly
    as a result of the coarser (8-km) grid and revised
    boundaries of the enlarged 160 x 160 km modeling region.
    This was determined by comparing parallel simulations
    (same emissions, same meteorology) using the 100 x 100 km
    and 160 x 160 km modeling regions. In the subregion
    common to both regions ozone fields were generally
    similar for each of two prototype days. Significant
    differences did occur locally near parts of the southern
    boundary of the enlarged modeling region and near the
    large NOx sources in the Pittsburg area.
    3.	Initial and boundary conditions dominated some of the
    1985 simulations to the extent that Central Valley ozone
    was unaffected by large percentage changes in Bay Area HC
    and NOx emissions under 26 July 1973 meteorological
    conditions. Inflow along the eastern boundary may have
    produced a significant boundary influence in the
    Sacramento area. The fact that simulations with large
    differences in emissions and fixed initial and boundary
    conditions produced similar ozone values throughout the
    modeling region suggests that initial conditions (or
    perhaps conditions at the upper boundary) dominated the
    production of ozone.
    4.	The accuracy of the boundary and initial-condition values
    used in the simulations is not known. The values are
    reasonable in that they are based on observations (albeit
    sparse) taken on 26 July 1973. The simulations using
    1975 emissions produced reasonable results and thus there
    is no obvious evidence of large errors.
    5.	For the 1985 simulations initial and boundary values of
    some pollutants should be reduced because anthropogenic
    emissions will be lower than they were in 1973. It is
    not necessarily true, however, that the reductions should
    be proportional to reductions in the emissions
    inventories. It is quite possible that, as controls
    become more and more effective in the future, an
    increasingly significant fraction of ambient
    -43-
    

    -------
    concentrations of some pollutants will be due to natural
    or other sources that are not accounted for in current or
    projected emission inventories.
    fa. It will be necessary to conduct further studies before we
    can assess adequately the downwind effects of Bay Area
    emissions.
    RECU WEND AT IONS
    The authors believe that LIRAQ, now operational over an extended
    modeling region, can be an effective tool for evaluating the effects of
    transport from the Bay Area. Such an evaluation could best be
    accomplished by a program of study that includes the following elements:
    1.	Past meteorological records should be surveyed to find
    those meteorological prototype days having the greatest
    potential for Bay Area/Central Valley interaction. One
    would hope to find days on which there were high 03
    levels in the Valley and general west-to-east flow over
    the modeling region. It is expected that the day(s)
    selected would be less influenced by inflow along the
    eastern boundary than was 26 July 1973. It may be
    desirable to conduct a multi-day simulation. The Bay
    Area's impact downwind is likely to be greatest on the
    day after peak ozone levels in the Bay Area when the
    onshore pressure gradient intensifies enough to induce
    transport of air into the Central Valley.
    2.	For future simulations it will be important to make
    initial and boundary values as accurate as possible. It
    is hoped that improved estimates will be possible as a
    result of extensive Central Valley field studies
    conducted by Pacific Gas and Electric Company 1n the
    summer of 1979.
    3.	Finally, for any given prototype day, systematic tests
    should be conducted of model sensitivity to varying
    emissions, boundary conditions and initial conditions.
    -44-
    

    -------
    REFERENCES
    1. Association of bay Area Governments, Bay Area Air Quality
    Management District, and Metropolitan Transportation Commission,
    "1979 Bay Area Air Quality Plan," January 1979.
    t. Bradley, Richard, California Air Resources Board, private
    communication, 1979.
    3.	California Department of Finance, "1978 Population Estimates of
    California Cities and Counties," Report 78 E-l, Sacramento, May
    1978.
    4.	Duewer, W. H., M. C. MacCracken and J. J. Walton, "The Livermore
    Regional Air Quality Model: II. Verification and Sample
    Application in the San Francisco Bay Area," J. Appl. Meteor., 17,
    pp. 273-311, 1978.
    5.	MacCracken, M. C. and G. D. Sauter, editors, "Development of an Air
    Pollution Model for the San Francisco Bay Area," Final Report,
    Lawrence Livermore Laboratory Rep. UCRL-51920.
    6.	Perardi, T. E., M. Y. Kim, E. Y. Leong and R. Y. Wada, "Preparation
    and Use of Spatially and Temporally Resolved Emission Inventories
    in the San Francisco Bay Region," J. Air Poll. Control Assoc., 29,
    pp. 358-364, 1979.
    7.	Reid, L. E., private communication, Systems Applications,
    Incorporated, San Rafael, CA, 19/9.
    -45-
    

    -------
    APPENDIX A
    Ozone, Hydrocarbon and Nitric Oxide Concentration Maps
    for Different Source Inventory Distribution Patterns
    

    -------
    

    -------
    t>00.- •
    *01
    0.1035
    • HI
    Oil
    0.769Er01
    -Oh
    SURF cMcEN CONTOuUfllF
    119 •
    190.f
    OZONE
    Figure A-l. 1300 PST surface ozone (ppm), 1975
    5 x 5 km emission pattern.
    

    -------
    •era
    •0*J
    %
    i .... I i i sr\ i i |dts*T*» l
    SURF V&fflcEN CONTOurf?'bF
    OZONE
    Figure A-2. 1300 PST surface ozone (ppm) , 1975
    10 x 10 km emission pattern.
    A-2
    

    -------
    ¦01*
    ».»# 0
    SURF V&tfcEN CONromU? i>F
    OZONE
    Figure A-3. 1600 PST surface ozone (ppm), 1975
    pop. emission pattern.
    A-3
    

    -------
    V.
    %
    0.545
    Q.565E-0
    0.2595 • v
    L
    0.626E
    
    SURF '
    -------
    I •.
    4.2BSE-a».
    0.208E-02
    *o»f H
    0.780E-0I
    0.102E-
    0.I06E-02
    H
    O.69SE-02
    0.993E-03
    . L
    0.37IE-02
    -f-
    0.92U-03
    
    " ' SURF VWcEN CONTOUJ? fef
    NITRIC OXIDE
    Figure A-5. 900 PST surface NO (ppm), 1975
    5 x 5 km emission pattern.
    A-5
    

    -------
    te-oi
    •«MW
    «l«0. - •
    Or 628E-0
    •Ml
    SURF VMcEN CONTOud?
    *11 I
    
    ALKANES
    Figure A 6. 900 PST surface alkanes (ppm), 1975
    10 x 10 km emission pattern.
    A-6
    

    -------
    £M6?e-or •
    l'^94E-Q*«|
    0.129E02
    Ml
    E-02
    •in - ¦
    SURF '(f&licEN CONTOUlUf &F
    NITR'C OXIDE
    Figure A-7. 900 PST surface NO (ppm), 1975
    10 x 10 km emission pattern.
    A-7
    

    -------
    0/4071
    'f°oo'r.
    »10 0
    *«l) 0
    SURF VWcEN CONTOUff?f ?)F
    ALKfcNES
    Figure A-8. 900 PST surface alkanes (ppm),
    1975 pop. emission pattern.
    A-8
    

    -------
    0.I06E-O2
    ,«oco
    «•«
    0.356E-02
    •mi
    0, I09E-02
    SURF cWcEN CONTOurf? t)F
    NITRIC OXIDE
    Figure A-9. 900 PST surface NO (ppm), 1975
    pop. emission pattern.
    A-9
    

    -------
    . 289E
    CT?*89E
    ' SURF (fWcEN CONTOtllJif I*
    OZONE
    figure A-10. 1500 PST surface ozone (ppm), 1985
    5 x 5 km emission pattern.
    A-10
    

    -------
    
    
    
    Figure A-ll. 15^0 PS1? surface ozone (pnn), 10^5
    10 x 10 Vm emission pattern
    A-ll
    

    -------
    ¦0*1
    • OAM
    0.S09E-G!
    		
    twc
    
    ~ OP
    V -—
    u. . . .	.. . I >wr< , }.»>»
    SURF C?)NCEN CONTOURS? %F
    430 0
    OZONE
    Figure A-12, 1500 PST surface ozone (ppm), 1985
    pop. emission pattern.
    A-12
    

    -------
    APPENDIX B
    MASCON-Genervited Mass Flux and Inversion Base Height Maps fo
    the 5 November 1976 Prototype Diy
    Selected Maps and Time Histories of 03 and N02 for the 198
    Baseline and Sensitivity Simulations
    

    -------
    DAY 310, METHOD I, 3 HOUR.. 5KH
    Al
    •ou
    ¦m
    411*.'
    ADJUSTED FLUX FIELD
    NUMBER OF ITERATIONS » 55
    Figure B-la. Mass Flux streamline computer
    analysis for 0530 PST,
    November 5, 1976
    J3-1
    

    -------
    DAY 310. METHOD !. 3 HOUR. 5KM
    			K—i				H	,	-I		
    :V.« , £\
    \.1»»/ J *DI»
    r'rcvJL-^. t
    *ms	\
    
    49#. • |!|.l SW.» (M.I III.I (M.t
    ADJUSTED FLUX FIELD
    NUMBER OF ITERATIONS = 62
    Figure B-?b. Mass Flux streamline computer
    analysis for 0830 PST,
    November 5, 1976
    B-2
    

    -------
    DAY 310. METHOD I. 3 HOUR, 5KM
    
    H
    4270.
    u
    /
    4IM.
    4199.
    4IM.
    4M.C IIM INI
    •Tl.l Ml IIM Ml Mi.9
    ADJUSTED FLUX FIELD
    NUMBER OF ITERATIONS «
    Figure B-lc. Mass Flux streamline computer
    analysis for 1130 PST,
    November 5, 1976
    B-3
    

    -------
    DAY 310, METHOD !. 3 HOUR, 5KM
    •Oil
    'VSN
    •OCI
    •u
    •OAK
    •U.I
    i
    ADJUSTED FLUX FIELD
    NUMBER OF ITERATIONS = 67
    Figure B-ld. Mass Flux streamline computer
    analysis for 1430 PST,
    November 5, 1976
    B~4
    

    -------
    DAY 310. METHOD I. 3 HOUR. 5KM
    ItH
    'M '
    • IN.'
    4IM.<
    •0«.
    ¦M.t
    •N.f
    H9.I
    ADJUSTED FLUX FIELD
    NUMBER OF ITERATIONS « 65
    Figure B-le. Mass Flux streamline computer
    analysis for 1730 PST,
    November 5, 1976
    

    -------
    DAY 310, METHOD I, 3 HOUR. 5KM
    1/
    
    -------
    -METHOD I , T?
    T' 	| i ¦ i ' > 1 i ¦ '1'< ¦ ">	I	'
    rKM
    «isa.
    4279.--
    4860.
    4230.--
    4219.--
    4190.--
    4170.--
    4180.- -
    4130.--
    4110.
    4090
    
    
    
    4:40.0 510.0 530.0 550.0 570.0 S90.0 610.0 630.0 CS0.0
    CONTOURS OF INVERSION HEIGHT
    MEASURED ABOVE TOPOGRAPHY (METERS!
    S'.AI E= 5.0 KM
    Figure B-2a Inversion base height (above topography)
    analysis for 0700 PST, November 5, 1976
    B-7
    

    -------
    KM
    AMT
    - *Wi -
    O
    OF A
    4230.
    4210.--
    
    *GI
    4170.
    ~FMO
    
    •FEA
    r i ¦/. i*
    t.
    i •/./
    '-OfIrOURS OF INVERSION HEIGHT
    Mf-ASMRFI' APOVE TOPOGRAPHY iMf-TFRv
    Figure B-2b. Inversion base height (above topography)
    analysis for 1600 PST, November 5, 1976
    13-8
    

    -------
    REO I 310 76 MET 65 S.I OSOR
    2291
    -01 \^- ¦
    Q'0.5'
    •on
    (5E-01
    0; 1776-1
    ID 0
    »»».»	III.*
    SURF CONCEN CONTOURS OF
    »» o
    OZONE
    Figure B-3a 1100 PST surface ozone (pnn),
    1985 baseline emissions,
    5 November 1976 meteorology
    R-9
    

    -------
    RLG I 310 76 MET 85 S.I QSOR
    -01 .
    0.203E-v
    
    -------
    RE 'j I CUP 7€ MET 35 . I Q;SOR
    •t»i
    0.I7SE-Q&
    .»»>• /
    pi«< j
    SURF rON' BN CONTOURS OF
    OZONE
    Figure B-3c. 1500 PST surface ozone (ppm), 1985 baseline
    emissions, 5 November 1976 meteorology.
    B-ll
    

    -------
    RE
    -------
    REG I 310 78 MET 85 S.I QSOR
    ~
    
    SURF CONCEN CONTOURS Of
    NITROGEN DIOXIDE
    Fiqure B-4a 0700
    1985
    1976
    PST surface NO2 (ppm),
    omissions, 5 November
    meteorology
    B-13
    

    -------
    REG I 310 76 MET 89 S.I OSOR
    239V'
    5 13 0
    iti.o	»mo
    SURF CONCEN CONTOURS OF
    NITROGEN DIOXIDE
    Figure D-4b 0900 PST surface NO2 (ppm),
    1985 emissions, 5 November
    1976 meteorology
    B-14
    

    -------
    REG I 310 76 MET 85 S.I OSOR
    oco
    £-0
    261E-02
    
    uo.
    0E-01 '
    0.250E- 12
    •D»C
    9	t/0.0
    SURF CONCEN CONTOURS OF
    NITROGEN DIOXIDE
    Figure B~4c. 1100 PST surface N02 (ppm), 1985 emissions,
    5 November 1976 meteorology
    B-15
    

    -------
    REG I 310 76 MET 85 S.I QSOR
    3TE-QL
    o.
    253E-02
    
    0. 182E- 12
    3 surf concen contours of
    HO 5
    ~ <0.0
    NITROGEN 010XIDE
    Figure B-4d. 1300 PST surface N02 (ppm), 1985 emissions,
    5 November 1976 meteorology
    B-16
    

    -------
    REQ I 310 76 MET 85 S.I OSOR
    r\-
    . »».¦
    •Oil
    
    0. HIE
    tltt	H».l	11«.«
    SURF CONCEN CONTOURS OF
    ng.»
    NITROGEN DIOXIDE
    Figure B-4
    -------
    REG I 310 76 MET 85 S.I QSOR
    0. 520E-02
    M I
    o!
    0. I88E- 12
    SURF CONCEN CONTOURS OF
    NITROGEN DIOXIDE
    Figure B-4f. 1700 PST surface NO2 (ppm), 1985 emissions,
    5 November 1976 Meteorology
    B-18
    

    -------
    RIG I 318 76 MET 85 S.I OSOR
    'surf CONCIN CONTOUR# OF
    NITROOCN DIOXIDE
    Figure B-4g 1900
    1985
    1976
    PST surface NC>2 (ppm) ,
    amissions, 5 November
    meteorology
    B-19
    

    -------
    REG I 310 76 MET 85 S.I QSOR HC 50 PCR
    •*' .
    •OCP
    •011
    <\l
    0.135E-
    -02
    0.1S0E-0T
    0.964E-02
    SURF CONCEN CONTOURS OF
    MO 9
    OZONE
    Figure B-5a. 1500 PST surface ozone, 50% HC emission
    reduction from 1985 baseline, 5 November
    1976 meteorology.
    B-20
    

    -------
    R I 310 78 MET 65 OSOR HC 25 NO 25 P:R
    ¦0»r
    •i/ei
    KM
    • IM
    •Oil
    0.232E
    lll.l	IW.t
    SURF COtyCEN CONTOURS OF
    Figure B-5b. 1500 PST surface ozone, 25% HC and
    25% NO emission reduction from 1985
    baseline, 5 November 1976 meteorology.
    B-21
    

    -------
    R I 310 76 MET 85 QSOR HC 50 NO 25 PCR
    ~DVT
    • N'
    O
    at _
    ~ocn
    .01-.
    ~011
    o
    JOSE-01
    SURF CONCEN CONTOURS OF
    410.9
    OZONE
    Figure B-5c. 1500 PST surface ozone, 50% HC and 25% NO emission
    reduction from 1985 baseline, 5 November 1976
    meteorology.
    B-22
    

    -------
    REG I 319 76 MET SS 5.1 OSO* HC 50 PCR
    
    L
    18E-02
    X
    
    
    0.188E-" >2
    lll.l	till	_ IM.«
    SURF CONCEN CONTOURS OP
    NITROGEN DIOXIDE
    Figure B-6a 1700 PST surface NO2 (ppm),
    50% HC emission reduction
    from 1985 baseline, 5 November
    1976 meteorology
    B-23
    

    -------
    R I 310 79 MET 85 OSOR HC 25 NO 25 PCR
    "•lift +¦
    0.5I6E-02 -¦
    
    •on
    I960
    0.189E- 12
    ~oil
    SURF CONCEN CONTOURS OF
    NITROGEN DIOXIDE
    Figure B-6b 1700 PST surface N02 (ppm),
    25% HC and 25% NO emission
    reduction from 1985 baseline,
    5 November 1976 meteorology
    B-24
    

    -------
    R t 310 76 MET 85 QSOR HC 50 N' 25 PCR
    •fio
    0.516E-02
    etc
    "Iff.
    I*
    m«W CONCEN CONTOURS OP '
    NITROGEN DIOXIDE
    Figure B-6c 1700 PST surface NO2 (ppm),
    50% HC and 25% NO emission
    reduction from 1985 baseline,
    5 November 1976 meteorology
    B-25
    

    -------
    /O0 r
    oa
    nj
    CT>
    PC
    CL,
    Oi
    c
    o
    RS
    +->
    c
    0)
    o
    c
    6
    CL>
    c
    o
    M
    o
    o Baseline
    X -50% HC
    -25% NO
    A -501 HC; -25% NO
    November 5, 1976
    Meteorology
    Hour (PSTJ
    Figure B-7a. Hourly ozone concentrations at San Jose (DSJ)
    from LIRAQ simulations using 1985 baseline
    emissions and three combinations of HC and NO
    reductions.
    

    -------
    /oor
    30
    ° Baseline
    X -50% HC
    ~ -25% HC; -25% NO
    A-50% HC; -25% NO
    txf
    I
    rsj
    ft4
    Cu
    g
    •H
    (0
    4->
    S
    IJ
    c
    §
    o
    N
    o
    60
    0
    A
    0
    A
    X
    8
    X
    ?
    
    ~
    g
    S
    §
    
    a—»¦
    a
    _i	
    November 5, 1976 meteorology
    
    _S	8
    og
    /0
    &	/4-
    Hour (PST)
    
    &
    ZO
    22
    Figure B-7b. Hourly ozone concentrations at Livermore (DLI) from
    L1RAQ simulations using 1985 baseline emissions
    and three combinations of HC and NO reductions.
    

    -------
    3S0
    O Baseline
    X -50% HC
    0 -25/25
    A -501 HC/25S NO
    *>o
    a-—o--
    u
    CM
    * N
    ja	a	~
    November 5, 1976 meteorology-
    Hour (PST)
    Figure B-8a.
    Hourly NC>2 concentration at San Francisco
    simulations using 1985 baseline emissions
    tions of HC and NO reductions.
    (DSF) from LIRAQ
    and three combina-
    

    -------
    ~	a
    
    ¦o-
    p-
    0 Baseline
    X -50% HC
    ~
    A-50 HC/25 NO
    November 5, 1976 meteorology
    /O
    08
    Hour (PST)
    Figure B-8b.
    Hourly NO2 concentrations at Burlingame (DBU) from LIRAQ
    simulations using 1985 baseline emissions and three
    combinations of HC and NO reductions.
    

    -------
    O Baseline
    X -501 HC
    200
    -C
    100
    November 5, 1976 meteorology
    00
    /O
    Hour (PST)
    Figure B-8c.
    Hourly NO2 concentrations at Oakland airport (AOA) from LIRAQ
    simulations using 1985 baseline emissions and three combinations
    of HC and NO reductions.
    

    -------
    JOr
    O-
    >rH
    
    Hour (PST)
    Figure B-8d Hourly NO2 concentrations at San Jose (DSJ) from LIRAQ
    simulations using 1985 baseline emissions and three
    combinations of HC and NO reductions
    

    -------
    APPENDIX C
    Selected Mass Flux Streamline and Inversion Base Height Fields
    for the 20 August 1973 and 26 July 1973 Prototype Days.
    

    -------
    / ~opr.
    *29
    H.58
    / ~ ASN
    4.26
    
    4.81
    4.19
    4 12
    E*03
    Ai>('UST 'iA'jf STUDY
    <1
    REG I ON
    A
    ¦ l>
    ~ VSN
    ~ AHA,'
    ~N£P
    •OSR
    *ISH
    ~DNA
    ~ ANA
    »t5RH
    ~NAM
    ~DCO
    ~01 A"
    >7
    /
    7 ~UCB¦
    , U ~MAG- . '
    ~MP®
    ; ' ~DBF
    ^ ^ -Vpcp ./ /
    ¦ . .	• ~AHHOHW ~pps /
    ¦	'A
    ~obv
    ~ as*
    > i
    ~ AHM
    ~AW .
    »0«e
    ~ PNC
    ¦¦~or p.
    'S*LLH''
    
    
    
    \ ;
    V
    v \
    ~OLO
    
    \ >W
    
    
    h
    
    
    \
    <
    \
    \
    \
    ~re*
    
    
    ~ NMH
    	«>
    
    \
    
    •-
    -»•
    
    
    
    
    \
    
    
    /
    
    
    ~OGL^ '' - . i
    "h.
    	L—
    —1	
    l	
    
    	
    J-
    . J.
    ,...i	_L —J	J
    *
    IT
    to
    r- o
    0> o
    -
    
    w # in to
    tfi
    in
    in
    in in
    in io
    
    10
    
    -------
    AUG'rj CASr STUDX
    REGION 1
    4 . 10
    't , "I
    'i . ?B
    4 .ai
    4 .26
    4.25
    H . 5W
    H.23
    4.22
    4.81
    4.?0
    4.19
    4.10
    4.17
    4. 16
    4. 15
    4.14
    4.13
    4 . ie
    4.11
    4 . 1(1
    C'03
    4 .09
    J( V	S.'ODlft
    •.¦¦¦V.^ \ . - >PSa :¦' vAC .V . ; >-7 {'
    ¦ i#"
    \ •
    ;-^X ¦ <-x ,
    ^ , a(k«i. )
    Mi 'v;v*,N\\
    I 1 V,V i \ - X > I
    i;
    v
    !(XSV rLl. v> .oik ' 1 -
    1\ l^VSN , ,,
    •	X ' | ; ' N
    " • ,
    ."(!¦¦ VAN* „
    / I' '-' ' i
    J-
    TIME
    8:30.0
    *UO eo 1973
    I \
    1 -
    c;
    
    X : ; /
    ¦ f\Y Vjr'" /
    1 +*$R y .. j '•
    I. ^dsr * '-'J
    ft /
    ~MftE
    V
    ~DPT.-
    ~NAN i \
    , I *osr JA'li^. {)/!- ' [0 ^''•-nvAC^ f ' s
    - V8 .¦:	"M0-n0 \\\ \\
    •'//a'-. ^
    ' li O i ^ .
    I
    Vv' f^NC - . I /'
    *DRC./ VX \>X \
    \	^ )V]/0
    rm
    J	l . „ i	L.
    o ~ f\j tn
    u \W c \<% \
    . l' -;/	\ cr-v r;\«; )rcy))
    V V/.-OHx \	
    iDiriioioininiAi/)inu}(ou)i0iou)(D
    CONTOURS OF IMVERSICN HEIGHT
    MEASURED ABOVE TOPOGRAPHY
    (METERS)
    SCALE- 5.0 KM
    Fig. C-lb.
    Inversion base height (above topography) analysis for 0830,
    August 20, 1973.
    C-2
    

    -------
    AUGUST CA5E STUDY - REGION 1
    H. 30
    'ASM
    1.21
    ~UCB
    *~.19
    ~ MAG - y
    ~GO A.
    ~LLH
    v. 16
    ~OLA.
    ~NMM
    ~F£A ••
    H . I I
    A
    4.10
    £~03
    H.09
    
    	1	X
    m <0
    ID
    o>
    m
    r-
    o
    s
    ADJUSTED FLUX FIELD
    TIME	MASS CONSISTENT SCHEME: LEAKY
    14:30.0
    auo 20 1973	NUMBER OF ITERATIONS = 57	scale- s.o km
    Fig. C-lc. Mass flux streaallne computer analysis for 1430, August 20,
    1973.
    C-3
    

    -------
    'i . AO
    4 .29
    4.2B
    4 . c?7
    )'
    / 1 H 11'"> I ( A'.i 'jfU.lV Ki.GION 1
    I \ i
    I' tm), ) \
    ¦» fJy (h/,A? "
    < ,dsa'
    <
    '' \ (\
    \ '¦-. 1 ! '
    ~ASM
    • AMC
    ~ AMT
    ~ ASE
    4.85
    4 .24
    4.?3
    4 .25
    4.21
    4.2C
    4.19
    M il
    4. 17
    4. 16
    4. 15
    4.14
    4.13
    4.12 1
    4.11
    4.10 ;•
    E *03 !
    VM8^
    ' "l \
    v;.. "
    ~hftE
    i
    ~ v3n
    . ~OPE
    ~ DNA
    / '
    ~ A '*k
    ~Df A
    ~ ATI
    ^:
    1 r-J
    ~	AHA
    ~	KtjiR
    '.~UpR
    •OVA
    UV;,
    
    ¦HSH.
    , *A0F ,
    -•OC.O
    . »QPT
    ~ NAN
    V' „ „ . '~UC8 I I
    ~ MA£ | , i
    ~ HH0 '	' \
    •DSf ~ A AN 'j
    / ~NSD
    iuia
    i ;
    *G0'lWl «PCP
    •A^diOHW ~PPS
    I I
    V'iV.
    1»LLH
    ~ asI
    ~ASF
    ~OBU
    \
    \
    ~AHM	*ASC
    / ~ DRC
    j \ ' 'L ' .
    . Tcv;' \
    t: •' .tnc
    ~DFR* 
    7'v ;
    »f 1Q'
    ,' ' /
    ' •-* i
    ; U-Vs
    ' 1 / ' /
    ¦vs.
    Jabh
    /
    / •-* '¦~OLG
    ~ MPpV ' ' r-"'
    •y .	/' '
    V \ I
    i.	. i i
    ' ''¦'
    T ! ME
    IM:30 . J
    AUG 20 197 5
    CONTOURS OF INVERSION HEIGHT
    MEASURED ABOVE TOPOGRAPHY
    (METERS)
    SCALE- 5.0 KM
    Fig. C-ld. Inversion base height (above topography) analysis for 1430,
    August 20, 1973.
    C-4
    

    -------
    ««. 30
    •» .i?n
    4 .C-8
    H ?7
    4.P6
    4.?5
    4. ?4
    4 ?3
    4.f£
    4.81
    4.20
    4. 19
    4 . 18
    4.17
    4. 16
    4. 15
    4 14
    4.13
    4 . 12
    4.11
    4.10
    E»03
    4.09
    /¦
    /
    ~ M^O
    \
    \
    JULY CAf.l* STUDY - REGION
    \
    f
    •roi
    \
    ~FMJ \
    \
    
    \
    I'DSA
    
    W
    WsN '0NA
    c -
    .a*m / t<
    \ »AMC \
    /• I h '
    V \
    - " *A5I
    ~ AMI
    r
    > '
    \ \
    \'
    ~DC#
    \
    ^MftE A_.
    *A(JA
    I
    \
    ~ rtv*
    ~ofa ,A}r
    I
    'a i
    "H I
    «-MRV
    4-
    
    \
    \ ~ AHA \
    \ *NSR	~ ISH
    / 0SR J " *ABF
    /	»DRW
    t'
    ~DPT
    L\
    
    )RJF	-toccy
    ~i$n
    »HAG
    ~ucu
    ! »Dl/
    
    -fry
    ¦'1
    ~OSF
    H ' ''
    • ¦/
    "J
    	*1_~ ¦ "AStr
    
    tr :*•
    
    ."ft!
    / •*
    
    /
    ~ Nf.B	<0.DSL«PCP
    • ~	\	*AVki
    . ~AHUOMU 'PPG/
    ~ ASF	4 !
    ' ~pBU " N ' ~-*PNt
    \	\ ~OFR I
    \SC '' -t-
    \«DRC	f"
    •?
    "3llm~
    ~AHM
    ^iO(Dr^cDOko»fu^j,iPa3
    0) o
    . nj
    to to u> id to u>
    "	ADJUSTED FLUX FIELD
    MASS CONSISTENT SCHEME: LEAKY
    NUMBER OF ITERATIONS = 57
    Fig. C-2& Mass flux streamline computer analysis at 0830 on July 26, 1973.
    TIME
    8:30.0
    JULY 2G 1973
    SCALE- 5.0 KM
    C-5
    

    -------
    JULY CASE STUDY ' Rf-GION 1
    n.?9
    >1 ?Q
    *.?6
    4.25
    W .SH
    M .53
    4.?2
    4.51
    M .20
    4 . 19
    '~.19
    M . 17
    4. 16
    4 . I S
    4 . IH
    4.13
    4 . 12
    't. 11
    w. io
    £~03
    4 09
    ~ Af>M
    » AHA
    ~ FBI
    <>
    /
    ~*(1C
    \
    w
    ~ASF
    M
    ' !<1'\iif/
    ¦ "~AMI
    1 J
    o
    kN MVU
    * L/H. If/
    j ')
    ~tav>'
    ~?rA
    I
    /
    ~AT I
    +MRV
    /
    
    WDPT
    ./ . *ISH
    _ 4 . ) ".io J ') 'T;::/' ' /'
    ' \V r-^ /I	V I *01#, ¦ ( 		
    '11) <¦—> *MagV f° V r_ .—.-t-. • ^ ¦ ---- jnrj.ijj
    '1 ¦ ',5'7"\osr •«« V (J (i - - OWv r~ "'
    , K\k..:	i
    ~ AS*
    -I
    \
    ~ ASF
    y ~OBU
    . ^Ht4pnM >»PPr /'//. I
    \ '^Vribr / -
    • asc \ . 1
    r\\ v^a '1
    / \ 'x ~ ofov ' ,' ( i
    ( -'' I /'V. - V
    "5?
    o
    < 3
    ~o*v. \ ' L
    Cj
    .'V.
    R (
    .~M^P
    ^ ^ARH
    ~D|\C^ j>J j
    \
    " uDiDa3ir»
    CONTOURS OF INVERSION HEIGHT
    MEASURED ABOVE TOPOGRAPHY
    (METERS)
    SCALE« 5.0 KM
    Fig. C-2b Inversion base height (above topography) analysis at 0830 on
    July 26, 1973.
    C-6
    

    -------
    JUL Y CA'.")L :uliv • HF. GI ON 1
    4 . 30
    4.29
    4.88
    4.?7
    >».a6
    4.?5
    4.?4
    4
    4.52
    4 ,ai
    4.?0
    4.19
    4. IH
    4. 17
    4. 16
    4. 15
    4 . 14
    4. 13
    4 . 12
    4.11
    .... -I'
    if
    /
    ' '»/;%¦
    ¦ ./ * / /
    ~ » MJ	/ / /
    ,,*asn / / j
    . -V
    ~riV
    ^	?y ? ^ *
    ' */ /'///tfff /-£ /
    
    ,'tiSA ;
    /
    rt*- -¦
    *MBO
    3»-
    
    i
    ~ DPE / A /
    .< 'V .' ~ ANA
    ~or a
    ~ AMT
    / *n^®	:.r,0/psttpcp
    '	i-
    ~ AHWuhW • «Pf)?
    • V*
    ~LLH
    /
    ~ ASf
    •OUU
    ' WV
    1
    •AHM \ «ASC -
    ~ OHC
    h
    /
    /
    4 /
    •>i /
    \
    ~iwv
    I j f>
    
    Japh.
    /
    4-
    ••..V
    • 0W5
    >,«MPP	\	/
    ~ftA""
    >fjMH
    4.IC - /	, /	•••ft, V.	y
    ~03 Up	t\ 4	*0Gt
    I, 09 L-.	L 	1	 J-- - i	 i	I	I	i	I	1	I	I	1	J	
    f
    I
    J	I	1-
    oio — iMfij-mior-aiaio — «j m * ir> ic
    . (\j	.	.	.	.	.	.	.	•	.	.	•	.	•	•	•	•
    j,ou>>iPtoif)intf)iPiflifiint0i0u)40
    -------
    '( .in
    •i
    4 )]/.
    K "\ \\\\ lv'r;:M% , .
    i	\ W „ ~		 ' Y	^ x,
    \ V-H • >. v
    0\M 'V
    v- "Cd *•'. v.	* »' ilXf^vyvK i, V v/ t.
    /T f. IV- \\ \ /- , ¦ 'A )¦ ¦¦!¦! A \ /.
    >\
    ! ' tijU
    M'iVi
    ~ Af,n
    I | ~ AMC
    r it,. i r
    ~ A5F
    AMT
    OSA
    .1 /"
    \
    I
    ¦Xi
    v
    -~v
    
    
    /
    /
    n c-rt::^:'V
    i	i
    + DVA I	; /
    [fV(
    / <-MRV
    ! /
    .' \ 'As»-y nry /	, / /V i
    T:iTO
    u(v/i>w< '• //v/ -«r
    I
    J ~MP8
    (' /-
    ~NSB
    -A\ \ :>, .. /,W
    i r^m
    1 \	r
    ~ASF •
    ~rwu
    \\ M\\	/'- !}}. \
    \^JAS^iXu- •
    "A T u }bio<
    M I
    ' iw.rj
    »AHM
    -'~A
    \
    \ 'Yj *orh <••_
    \tV,
    V
    "?J
    \ N ^	I '•	v \	V \
    .	|	~DMV 'v N ' V/ J to <. } \ \ \ 1) \
    VcVx-sTHs^1
    / -1	\»DLg/ ,¦-"/( r>
    ->-'/ 1 \
    /
    tr
    s a in to
    T I ME
    14:30.0
    JULY 26 1973
    ~hpP O / / N.' , 1 .
    \;Tr'^. vffi
    ,
    j- (Ou)«ouj
    CONTOURS OF INVERSION HEIGHT
    MEASURED ABOVE TOPOGRAPHY
    (METERS)
    SCALE• 5.0 KM
    Fig* C-2d Inversion base height (above topography) analysis at 1430 on
    July 26, 1973.
    C-8
    

    -------
    TECHNICAL REPORT DATA
    (I'h tisr Ifihl Illk/jlf lHiin on llir /ri'irvr hc/nrf l oini'li-liny,}
    1 IIM'l <1 1 1 N( < J
    LPA-4L>0/4-79-027
    4. Illlt AND :
    -------