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 :
------- |