600382094
         FINAL  EVALUATION OF URBAN-SCALE
   PHOTOCHEMICAL  AIR  QUALITY SIMULATION MODELS
                       by

    Kenneth L.  Schere  and Jack H. Shreffler
      Meteorology  and  Assessment Division
   Environmental Sciences Research Laboratory
  Research Triangle  Park, North Carolina. 27711
  ENVIRONMENTAL  SCIENCES RESEARCH LABORATORY
      OFFICE  OF  RESEARCH AND DEVELOPMENT
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
RESEARCH  TRIANGLE  PARK, NORTH CAROLINA  27711

-------
                                 DISCLAIMER

     This  report  has  been  reviewed by  the  Environmental  Sciences  Research
Laboratory, U.S.  Environmental  Protection  Agency,  and  approved for  publica-
tion.   Mention  of  trade  names  or commerical  products  does  not  constitute
endorsement or recommendation for use.
                                AFFILIATION

     Dr. Shreffler  and Mr.  Schere are  on assignment  to  the Meteorology  and
Assessment  Division,  Environmental   Sciences   Research  Laboratory,  from  the
National Oceanic  and Atmospheric Administration, U.S. Department  of Commerce.
                                      n

-------
                                   ABSTRACT

     The research study discussed here is a continuation of previous work whose
goal was to determine  the  accuracy  of several  selected urban photochemical air
quality simulation models  using  data from the  Regional  Air Pollution Study in
St. Louis.  The report summarizing the  initial  work on this project discussed
four models  and used  a  sample  of  10 days  for testing.   This  work continues
testing on  three  of  the  models with  an increased sample  size  of  20 days.

     The models  evaluated  here  are:   The  Photochemical  Box Model  (PBM)  de-
veloped in-house by EPA, The  Lagrangian  Photochemical  Model  (LPM)  developed by
Environmental  Research and  Technology, Inc., and The Urban Airshed Model (UAM)
developed by Systems Applications, Inc.  Emphasis in this report is directed at
the ability of the models  to  reproduce the maximum 1-hour ozone concentrations
observed on 20  days  selected from nearly "2  years of data.   The  PBM,  LPM, and
UAM have been evaluated using  statistical  methods and graphical  techniques and
all show potential  as air quality management tools.

     The standard  deviation of  the  differences between observed  ozone maxima
and predicted  concentrations  at  the  same place  and time  ranged  from  0.04 to
0.06 ppm for maxima of 0.16 to 0.26 ppm.   This  measure of uncertainty should be
recognized by  decision-makers using  these models  in regulatory  and  planning
processes.

     This  report covers a period from March 1981 to February 1982.

-------
                                 CONTENTS
Abstract	iii
Figures   	vi
Tables    	"	ix

     1.  Introduction 	    1
     2.  Conclusions and Recommendations   	    5
     3.  The Data Set	  .  .  .    8
     4.  Photochemical Box Model   	12
             Overview	12
             Results	13
             Conclusions   	21
     5.  Lagrangian Photochemical Model  	   39
             Introduction  	   39
             Method   	39
             Results	43
             Conclusions   	46
     6.  Urban Airshed Model   	   63
             Introduction  	   63
             Results	65
             Conclusions   	.-   74

References  	93
Appendices

         A.   PBM - Time series for remaining test days	95
         B.   PBM - Summary of  statistical  results	126
         C.   LPM - Time series for remaining test days	132
         D.   UAM - RS and  RO plots for remaining test days	179
         E.   UAM - Contour plots for remaining test days	228
         F.   UAM - Summary of  statistical  results	237

-------
                                  FIGURES





Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
A-l
A-2
A-3
A-4
A-5
A-6
A-7
A-8
A-9
Map of the St. Louis area 	 ~ 	
Schematic drawing of the PBM domain 	
Map of the PBM domain 	
PBM simulation results for Day 142 of 1975 	
PBM simulation results for Day 178 of 1975 	
PBM simulation results for Day 207 of 1975 	
PBM simulation results for Day 212 of 1976 	
PBM simulation results for Day 275 of 1976 	
PBM frequency histograms for all data 	
PBM residual vs. obs. cone, plots for all data ....
LPM simulation results for Day 231 of 1975 	
LPM simulation results for Day 251 of 1975 	
LPM simulation results for Day 226 of 1976 	
LPM simulation results for Day 237 of 1976 	
LPM simulation results for Day 275 of 1976 	
Schematic drawing of the UAM domain 	
UAM simulation results for CO on Day 142 of 1975 . . .
UAM simulation results for N02 on Day 142 of 1975 . . .
UAM simulation results for 03 on Day 142 of 1975 . . .
UAM contours of max. 03 field on Day 142 of 1975 . . .
UAM simulation results for CO on Day 207 of 1975 . . .
UAM simulation results for NOg on Day 207 of 1975 . . .
UAM simulation results for 03 on Day 207 of 1975 . . .
UAM contours of max. 03 field on Day 207 of 1975 . . .
UAM simulation results for CO on Day 195 of 1976 . . .
UAM simulation results for N0£ on Day 195 of 1976 . . .
UAM simulation results for 03 on Day 195 of 1976 . . .
UAM contours of max. 03 field on Day 195 of 1976 . . .
UAM simulation results for CO on Day 275 of 1976 . . .
UAM simulation results for M02 on Day 275 of 1976 . . .
UAM simulation results for 03 on Day 275 of 1976 . . .
UAM contours of max. 03 field on Day 275 of 1976 . . .
PBM simulation results for Day 182 of 1975 	
PBM simulation results for Day 183 of 1975 	
PBM simulation results for Day 184 of 1975 	
PBM simulation results for Day 209 of 1975 	
PBM simulation results for Day 221 of 1975 	
PBM simulation results for Day 230 of 1975 	
PBM simulation results for Day 231 of 1975 	
PBM simulation results for Day 251 of 1975 	
PBM simulation results for Day 159 of 1976 	
	 11
...'.. 23
	 24
	 25
	 27
	 29
	 31
	 33
	 35
	 37
	 48
	 51
	 54
	 57
	 60
	 76
	 77
...... 78
	 79
	 80
	 81
	 82
	 83
	 84
	 85
	 86
	 87
	 88
	 89
	 90
	 91
	 92
	 96
	 98
	 100
	 102
	 104
	 106
	 108
	 110
	 112

-------
A-10  PBM simulation results for Day 160 of 1976	114
A-ll  PBM simulation results for Day 195 of 1976	116
A-12  PBM simulation results for Day 211 of 1976	118
A-13  PBM simulation results for Day 225 of 1976	120
A-14  PBM simulation results for Day 226 of 1975	122
A-15  PBM simulation results for Day 237 of 1976	124
 C-l  LPM simulation results for Day 142 of 1975	134
 C-2  LPM simulation results for Day 178 of 1975	137
 C-3  LPM simulation results for Day 182 of 1975	140
 C-4  LPM simulation results for Day 183 of 1975	143
 C-5  LPM simulation results for Day 184 of 1975	146
 C-6  LPM simulation results for Day 207 of 1975	149
 C-7  LPM simulation results for Day 209 of 1975	152
 C-8  LPM simulation results for Day 221 of 1975	155
 C-9  LPM simulation results for Day 230 of 1975	158
C-10  LPM simulation results for Day 159 of 1976	161
C-ll  LPM simulation results for Day 160 of 1976	164
C-12  LPM simulation results for Day 195 of 1976	167
C-13  LPM simulation results for Day 211 of 1976	170
C-14  LPM simulation results for Day 212 of 1976	173
C-15  LPM simulation re'sul ts for Day 225 of 1976	' .  . 176
 D-l  DAM simulation results for CO on Day 178 of 1975	180
 D-2  UAM simulation results for NOg on Day 178 of 1975	181
 0-3  UAM simulation results for 03 on Day 178 of 1975	182
 D-4  UAM simulation results for CO on Day 182 of 1975	183
 D-5  UAM simulation results for N02 on Day 182 of 1975	184
 D-6  UAM simulation results for 03 on Day 182 of 1975	185
 D-7  UAM simulation results for CO on Day 183 of 1975	186
 D-8  UAM simulation results for N02 on Day 183 of 1975	187
 D-9  UAM simulation results for 03 on Day 183 of 1975	188
0-10  UAM simulation results for CO on Day 184 of 1975	189
0-11  UAM simulation results for N02 on Day 184 of 1975	190
D-12  UAM simulation results for 03 on Day 184 of 1975	191
D-13  UAM simulation results for CO on Day 209 of 1975	192
D-14  UAM simulation results for NOg on Day 209 of 1975	193
D-15  UAM simulation results for 03 on Day 209 of 1975	194
0-16  UAM simulation results for CO on Day 221 of 1975	195
D-17  UAM simulation results for HO? on Day 221 of 1975	196
0-18  UAM simulation results for 03 on Day 221 of 1975	197
D-19  UAM simulation results for CO on Day 230 of 1975	198
D-20  UAM simulation results for N02 on Day 230 of 1975	199
D-21  UAM simulation results for 03 on Day 230 of 1975	200
D-22  UAM simulation results for CO on Day 231 of 1975	201
D-23  UAM simulation results for N02 on Day 231 of 1975	202
D-24  UAM simulation results for 03 on Day 231 of 1975	203
D-25  UAM simulation results for CO on Day 251 of 1975	204
D-26  UAM simulation results for NOg on Day 251 of 1975	205
D-27  UAM simulation results for 03 on Day 251 of 1975	206
D-28  UAM simulation results for CO on Day 159 of 1976  ........ 207
D-29  UAM simulation results for M02 on Day 159 of 1976	208
D-30  UAM simulation results for 03 on Day 159 of 1976	209

-------
D-31  UAM simulation results for CO on Day 160 of 1976	210
D-32  UAM simulation results for N02 on Day 160 of 1976	211
D-33  UAM simulation results for 03 on Day 160 of 1976	212
D-34  UAM simulation results for CO on Day 211 of 1976	213
D-35  UAM simulation results for M02 on Day 211 of 1976    	214
D-36  UAM simulation results for 03 on Day 211 of 1976	215
D-37  UAM simulation results for CO on Day 212 of 1976	216
D-38  UAM simulation results for N02 on Day 212 of 1976	217
D-39  UAM simulation results for 03 on Day 212 of 1976	218
D-40  UAM simulation results for CO on Day 225 of 1976	219
D-41  UAM simulation results for NOg on Day 225 of 1976	220
D-42  UAM simulation results for 03 on Day 225 oT 1976	221
D-43  UAM simulation results for CO on Day 226 of 1976	222
D-44  UAM simulation results for N02 on Day 226 of 1976	223
D-45  UAM simulation results for 03 on Day 226 of 1976	224
D-46  UAM simulation results for CO on Day 237 of 1976  ........ 225
D-47  UAM simulation results for N02 on Day 237 of 1976	226
D-48  UAM simulation results for 03 on Day 237 of 1976	227
 E-l  UAM contours of max. 03 field on Day 178 of 1975  ....'.... 229
 E-2  UAM contours of max. 03 field on Day 182 of 1975	229
 E-3  UAM contours of max. 03 field on Day 183 of 1975	230
 E-4  UAM contours of max. 03 field on Day 184 of 1975	230
 E-5  UAM contours of max. 03 field on Day 209 of 1975	231
 E-6  UAM contours of max. 03 field on Day 221 of 1975	231
 E-7  UAM contours of max. 03 field on Day 230 of 1975	232
 E-8  UAM contours of max. 03 field on Day 231 of 1975	232
 E-9  UAM contours of max. 03 field on Day 251 of 1975.	233
E-10  UAM contours of max. 03 field on Day 159 of 1976	233
E-ll  UAM contours of max. 03 field on Day 160 of 1976	-.234
E-12  UAM contours of max. 03 field on Day 211 of 1976	234
E-13  UAM contours of max. 03 field on Day 212 of 1976	235
E-14  UAM contours of max. 03 field on Day 225 of 1976	235
E-15  UAM contours of max. 03 field on Day 226 of 1976	236
E-16  UAM contours of max. 03 field on Day 237 of 1976	236
                                    viii

-------
                                   TABLES

Number

   1  Summary of Meteorological Conditions on 20 Test Days   	  10
   2  Predicted and Observed 03 Maxima for the PBM,	20
   3  Initial Pollutant Concentrations Assigned in the LPM   	  42
   4  Predicted and Observed 03 Maxima for the LPM	47
   5  Predicted and Observed 03 Maxima for the UAM	71
   6  03 Regression Statistics for the UAM	  .  .  .  73
 B-l  Summary of Model  Performance Statistics from the PBM	128
 C-l  Temporal  Correlation Coefficients for the LPM	133
 F-l  Summary of Mean Concentration Statistics from the UAM	238
 F-2  Summary of Analysis of Trends Data from the UAM	243
 F-3  Summary of Concentration Maxima Statistics from the UAM  	 246
                                      IX

-------
                                 SECTION 1

                                INTRODUCTION

     This report covers  a  final  evaluation of selected urban  scale  photochem-
ical models which could be applied in regulatory  situations.   A report (herein-
after  referred  to as  Report I)  on  a preliminary  evaluation  was prepared  by
Shreffler and Schere (1982).  In  the interest  of brevity,  material  in Report I
is not repeated here unnecessarily.   Referenced documentation on the  models  and
Report I are  essential  to  fully understand the  procedures used  in the simula-
tions.

     The Regional  Air  Pollution  Study  (RAPS)  was  conducted in the  St.  Louis
region over  the period  1974-1977  (Schiermeier,  1978).   RAPS was designed  to
provide a  comprehensive  data set for  the testing and evaluation of numerical
air quality simulation  models on an  urban scale.   While the RAPS field measure-
ments  were  in  progress  EPA surveyed  the. available, -state-of-the-art,  photo-
chemical  air  quality  simulation  models, and  selected 3 for evaJuation.
Under  contract,  the  builders adapted  their models  to  the RAPS data  base  and
were  supplied  3 days  of data for  test simulations.   In  addition,  there  was
a  need to  look  at a  simple  box-model  approach which  was not embodied  in  an
existing model.   A box  model  was therefore constructed by  EPA.

     The following models were investigated in  the evaluation program.

     Photochemical  Box Model  (PBM)  - a single cell  Eulerian model constructed
     by EPA.

     Lagrangian Photochemical Model  (LPM)  - a multi-level  parcel model  deve-
     loped by Environmental  Research and Technology, Inc.

     Livermore  Regional  Air Quality  Model  (LIRAQ)  - a single-level  Eulerian
     grid model  developed by Lawrence Livermore Laboratory.

-------
     Urban Airshed Model (UAM)  -  a  multi-level, Eulerian grid model  developed
     by Systems Applications,  Inc.

     LIRAQ,  described  by  MacCracken  et al .  (1978), was  dropped  from the
evalution because of a  series  of  technical  and logistics problems  detailed  in
Report I.  LIRAQ was unable to  produce  significant ozone  levels  for St.  Louis,
but no specific, correctable error  could be identified.  Aside  from  the tech-
nical  problems with the model, which are  unresolved and unevaluated,  LIRAQ did
not seem well suited as  a general-use model  easily  transferable to  new locales.
No conclusions can be made from this effort  about  the  use of LIRAQ by special-
ized user groups at specific  locales.

     Some of the other conclusions reached from the preliminary  work  described
in Report I  are  that (a) the PBM performs best in near-stagnation conditions,
(b) the LPM  needs modification of the fixed-box formulation  for  the air  parcel
and (c) the UAM tends to consistently underpredict  ozone.   It was also general-
ly found  that the variability between  specific 1-hour predicted  and observed
concentrations at a particular location  tends to be substantial.

     The  PBM,  LPM and  UAM were  again  tested extensively for  this  study and
corrections  of  obvious  errors  or deficiencies  were  made.   However,  no  effort
was made  to  adjust  or  tune  the  model   predictions  to observed  concentration
values.  The  prevailing  philosophy behind the  evaluation  effort  was to use the
models in an  off-the-shelf mode, much as  they  eventually  would be  applied by a
user in a  regulatory  situation.   However, great care  was taken  in preparation
of data sets  and model executions. Although  model  assumptions vary, effort was
made to use data in similar manners  in  all models.  Data preparation and  actual
execution of the models was accomplished solely  by  the  authors at  EPA.,  The
goal of the evaluation was  to provide a  fair and objective determination  of the
accuracy  of  a set of photochemical  models  when tested in an operational mode
against a comprehensive urban data base.

     Photochemical air quality models are designed to  predict ozone concentra-
tions and thereby provide  a means to assess  beforehand the  effects of changing
emission levels in an urban area.   Emission changes may result from addition  of

-------
new  sources  or controls  on existing  sources designed  to  reach ozone  levels
defined by the National  Ambient Air Quality Standards  (MAAQS).   In either case,
knowledge of  the effects  of emissions  changes  would  allow prudent  economic
decisions on plant siting or control equipment use.   The numerical  air quality
model synthesizes knowledge  about the emissions, transport  and  transformation
of pollutants into a workable computational framework,  but the  central  problem
                                                                             •
is how to be  reasonably certain that the model  is correct in  its projections.

     The only  reasonable  method of determining  accuracy is  to test the  model
against an extensive observational  data  base.  Providing such a  data  base  was
the purpose behind the  RAPS.  Since the level  of  the ozone maximum for  each  day
is of  paramount  importance  relative to  the MAAQS,- the  comparison  between  its
observed value and the model prediction  at  the same time and place  would be of
central   interest.   The  method  of  evaluation  consists  of selecting  a set  of
test days,  executing  model   simulations  and computing  residual  concentrations
(observed-predicted).   The specific outcome of the evaluation is  a presentation
of information on residuals  under the given circumstances.   Conclusions  about
model acceptability  require  further assumptions and judgements.

     The residual  reflects a composite effect of  input errors,  input uncertain-
ties and  model  errors.    Errors,  if not compensating,  will  tend to produce a
non-zero average residual over  many cases.    If  the average  residual is  nearly
zero, we might be willing to accept that the  model  is appropriately formulated
for predicting absolute  levels  of high ozone in the area where  the observations
were taken.   The  case-by-case  variation  could be ascribed to  uncertainties  in
the  input.   The  acceptance  of  the model for absolute  prediction leads  rather
naturally to  accepting  that it will  respond  to  emissions changes  in  the same
manner as the  real-world  system.   Finally, we are likely to conclude  that  the
model  is  transferable  to  another location,  provided  the  meteorological  and
chemical  modules are judged sufficiently flexible to adapt  to the new  condi-
tions.   This line of reasoning  underlies  the  decision-making process attendant
to model  use.

     This report details the  performance of 3 completed models using  the
RAPS data base.  Input  information  and model  components were carefully checked

-------
before proceeding.   Effort was directed at ascertaining that individual modules
were doing what  they  were purported to do, but not to  evaluate  their  accuracy
separately.  Improvements are perhaps possible in  some  of  the modules,  but this
is  left  to the  future.   This  report  indicates  the accuracy of the  finished
models used  in  an  operational  mode with  the best information available  at
present.

-------
                                  SECTION 2

                       CONCLUSIONS AND RECOMMENDATIONS

     This  report presents  the second,  and  final,  evaluation of  urban-scale
photochemical air quality simulation models by EPA's Meteorology and Assessment
Division.  The 3 models included in the analysis'span a wide range in complexi-
ty and sophistication but are  all  based  on numerical  solutions to mass-conser-
vative  equations.    They  are  selected  from  the  general  categories  of  box,
trajectory  (Lagrangian)  and  grid  (Eulerian)  models.   The first  evaluation,
described  in  Report I, gave  detailed  results from the  model  runs on  10  days
selected  for  simulation  from  the RAPS  data  base.    The sample  size  for  this
evaluation  is  20 days; an  additional  10 days were combined with the  original
sample.   Some changes in the  model  algorithms, including  the data preproces-
sors,  from the  time  of  the  original  testing on  the  10-day sample  made  it
necessary  to  perform  the  original  simulations  over.   Emphasis  in the  model
performance evaluations is  placed on ozone, although -results  for other pollu-
tant species  are also discussed.   Conclusions on the performance  of- specific
models did  not change appreciably  from  those listed in  Report  I although our
confidence in them has increased because of the consistency in  results  obtained
from both studies.

     The  PBM predictions  for  maximum  03  for the average  of the  monitoring
stations within  the model domain  were  generally on  the high side.  The average
03  residual  implied  a  23%  overprediction over all  test  days.    However for
the  5  stagnation-type days  where  the  maximum observed  03  occurred  within the
PBM  domain,  the  average  overprediction of 8*  was considerably better  than for
the entire sample.  Only a slight tendency towards overprediction was indicated
for  the  LPM.   The  biases  of the  residuals  were  relatively small,  11%  of the
average observations  at Level-1 and  only 2.5%  at  Level-3.  The standard devia-
tions of  the  LPM residuals  were the  highest among  the three models tested, but
nearly halved from the initial  tests described in  Report I.  The large  variance
in the residuals might be expected since  the  LPM generates a  prediction which
likely is the most  specific  to  a  particular place  and time- Model  predictions

-------
for maximum 03 by  the  UAM in a specific  sense  (at the same time and  location
as  the  observed maximum)  were consistently  low for  all  evaluations  with  an
average 32" underprediction  over  the sample.   Originally it was thought  that
spurious  numerical  diffusion  in   the  advection  component  of  the  model   was
responsible for the low  UAM  predictions.   However the  underprediction  problem
still  persisted after  the original advection algorithm was replaced with  one
containing much less inherent numerical  diffusion.   If the time  and  location of
the model  predictions  are  not constrained  to ^be  the same as  those  for  the
maximum observed  03,  the average 'model  bias  for the  20 days  implied  a  4%
overprediction.  This  excellent agreement  might  suggest that the uncertainty in
specifying  a  wind  field  for a grid  model   like  the  UAM  could  lead to  large
apparent errors in the model  results.

     The  choice  of which  particular model  to  use in a  specific  application
involves not only the accuracy of  the model  but  also  the resources  required to
operate it.  The models  test'ed here have  resource  requirements  correlated  with
their level of complexity.   In terms of  man-months  needed to  set  up  a  single
day simulation and computer  time expended (minutes of CPU on a  UNIVAC  1100/82)
the approximate requirements are:

          PBM	0.15 man-month	 1 minute CPU
          LPM	0.20 man-month	10 minutes CPU
          UAM	0.50 man-month	110 minutes CPU

     The  3  models  discussed  in  this report have all  shown themselves to be
acceptable tools for analysis of urban ozone air quality.  The  specific config-
uration of  an  application along  with the quantity and quality  of related  data
and resources available  to the user must  all  be  considered in  the final  selec-
tion  of a model.   For  an  indication  of average  03   air  quality in  an  urban
area  under  stagnation  conditions  or as a screening method  for  a more complex
model  the  PBM  is  appropriate.  The choice  of a trajectory model,  such  as the
LPM,  or a grid model,  like  the UAM,  might well  be decided by resource  require-
ments or by the number of proposed simulations. - In  any event,  the user of any
of  these  models must  have a strong scientific background and  exercise extreme
care  in implementing  the air quality simulations.  For an air  quality analyst
making  regulatory  decisions  based,  in  part, on  model  results  it should be

-------
understood  that  substantial  variability  can  exist  between  a  single  1-hour
predicted and observed concentration at a particular location.

     These models are  now  being  made available to  EPA's Office  of  Air Quality
Planning  and  Standards  for further  statistical  and  sensitivity  testing  and
ultimately for use in  their regulatory  decision-making process.  Because model
development is an evolving  area  of research it is  very  likely that subsequent
"improved" versions of the models  tested  here  will become available.   A  per-
formance  test with  benchmark results, as  described and tabulated  in  this
report, now exists for future  use in urban air quality  model  comparisons  with
any subsequent versions of the models.

-------
                                SECTION 3

                              THE DATA SET

     As an  integral  part  of  the RAPS  a  network  of  25 surface  stations  was
established in  and around  the  St.  Louis  region.   These  stations,  shown  in
Figure  1,  comprise  the- Regional  Air Monitoring  System (RAMS).   The  RAMS
stations continually  monitored various  meteorological  variables  as well  as
ambient concentrations of  pollutant gases.   Schiermeier  .(1978)  provides
details on  the  instrumentation.   In addition to the  RAMS,  upper air balloons
were released  each hour from urban  and  rural sites  to provide  wind profiles
for modeling purposes.

     An essential  element in the data set for an air  quality modeling exercise
is  a complete  source  emissions  inventory  for  all  gaseous   and  particulate
species over the domain of interest.  Littman (1979)  summarizes the comprehen-
sive data gathering and archiving of the RAPS emissions inventory.  The inven-
tory is divided into  point source  and  area source  classes with points defined
as sources  emitting at least 0.01 percent of  the  total emissions of a  pollutant
for the whole  air.quality  control  region.   Line source emissions are included
in the area source class.   The area  source emissions data are subdivided onto a
horizontal  grid of 1989 cells with variable spatial resolution, dependent upon
the  emissions  source  density.   Grid resolution  varies  from 1 km  in the more
dense  areas  to 10 km  in  the  sparse areas.   The  temporal resolution  of the
emission inventory is one  hour.  Criteria pollutants  included  in the  inventory
are  TSP (total  suspended  particles),  S0£   (sulfur  dioxide),  NOX   (nitrogen
oxides), THC  (total  hydrocarbons),  and CO   (carbon  monoxide).   Furthermore,
total hydrocarbons are broken down  into the  component  classes of non-reactives,
olefins, paraffins, aldehydes, and aromatics.  Examples of the gridded spatial
distribution of selected pollutants  are  shown  in  Report  I.

-------
     This  report  presents  the  results of  model  simulations  for 20  individ-
ual  days  chosen  from  the  RAPS  data  base.   These  days, 11  from 1975 and  9
from  1976,  account  for  some of  the  higher  03 measurements  observed in  the
RAPS  surface  monitoring  network.    Maximum  hour-average  single station  03
values all  exceeded  0.16  ppm on  the 20 days.   Table 1  lists  the basic  char-
acteristics  of the  days:   wind  speed  and  direction,  temperature,  and  solar
radiation computed as averages over 0700-1359 CST.   Also listed are the maximum
mixing heights  for each  day  as determined from the upper air  sounding  data.
Higher 03  levels  generally  occur when  a  prevailing high  atmospheric  pressure
system exists  over the  area with little associated  cloud cover,  and  represent
the  situations  conducive  to production of 03 from locally generated  precursor
emissions.  Both the  dates and Julian day  numbers  for the 20  days are  listed in
Table  1  and  either  form may  be used  throughout  this  report to reference  a
particular day.

-------
TABLE 1.  A SUMMARY OF WIND SPEED (WS),  WIND DIRECTION (WD),  TEMPERATURE,
          SOLAR RADIATION (ALL WAVELENGTHS), AND MAXIMUM AFTERNOON MIXING
          HEIGHT (Mh)  FOR THE 20 DAYS EXAMINED.*
Date
Day
(Julian)
WS
(m/s)
WD
(deg)
Temp
Solar
(ly/min)
Max MH
(m)
—
5/22/75
6/27/75
7/01/75
7/02/75
7/03/75
7/26/75
7/28/75
8/09/75
8/18/75
8/19/75
9/08/75
6/07/76
6/08/76
7/13/76
7/29/76
7/30/76
8/12/76
8/13/76
8/24/76
10/01/76
. 142
178
182
183
184
207
209
221
230
231
251
159
160
195
211
212
225
226
237
275
1.1
0.4
1.4
1.4
1.8
1.0
2.0
0.4
1.6
1.3
1.8
1.0
1.3
2.3
0.3
1.7
2.3
1.1
1.3
0.6
224
245
70
15
324
139
18
88
167
168
181
129
284
145
251 -
205
253
273
110
222
29
29
29
30
30
26
30
26
27
28
25
25
27
28
26
30
29
30
28
22
1.12
0.96
0.99
0.92
0.85
0.98
0.98
0.98
0.96
0.95
0.89
1.06
1.01
1.02
0.53
0.82
0.70
0.86
0.82
. 0.78
1504
1822
. 2606
2488
1875
1477
1909
1195
1488
1052
1797
1972
1772
1853
1706
1304
730
1427
2124
527
         *Meteorological  variables (except MH)  are network averages over
          the period 0700-1359 CST.
                                      10

-------
Figure 1.  The St. Louis area with locations of the
           RAPS surface stations.
                       11

-------
                                  SECTION 4

                           PHOTOCHEMICAL BOX MODEL

INTRODUCTION

     The Photochemical  Box Model  (PBM),  a  single cell  Eulerian air  quality
simulation  model,  simulates  the transport and  chemical  transformation of  air
pollutants  in  smog-prone  urban  atmospheres.   The model's  domain is set in  a
variable volume,  well-mixed  reacting  cell  where  the  physical  and  chemical
processes responsible for 03 generation  by HC and  NOX  precursors  are mathemat-
ically  created.    These  processes  include  the  transport and  dispersion  of
pollutant species through the cell, the  injection  of primary  precursor species
by emission  sources,  and  the chemical  transformation  of the  reactive  species
into  intermediate  and  secondary products.  They are schematically  illustrated
in Figure 2.

     To apply the model to the  St.  Louis RAPS  data base, the  horizontal  length
scale of the  cell  was  set at 20 km and  the  vertical scale  was time  dependent,
proportional  to the depth  of the mixed layer.  The model  domain was  chosen such
that  it included most major emission  sources on  either side of the  Mississippi
River.   Source  emissions  were  assumed  to be distributed  uniformly  across  the
surface face  of  the cell.  Twelve  of the RAMS surface  monitoring  stations as
well  as  one upper  air  sounding location  were  located within the cell  bound-
aries.   This configuration  of  the model  domain  is illustrated  in  Figure 3.

     The steps  involved  in  the method  for  performing a  PBM simulation  are
described  in the  first  evaluation  report by  Shreffler  and Schere  (1982),
Report  I.   There  are 3 data preprocessors that must be  exercised prior to the
model  simulation  itself.    They  access  the RAPS  data  base  and  retrieve  the
appropriate parameters required by the PBM.
                                      12

-------
     The framework of the model itself has not changed from the time of Report
I to the present.  There was a change made,  however,  in  the method of choosing
the upwind  HC  boundary  condition for the  model.   Originally,  if  the  HC data
were invalid from the chosen  upwind  monitoring  locations at a particular hour
an alternate algorithm  for  obtaining these data was  invoked.   It  was  a ratio
technique based  on  the  average non-methane hydrocarbon  (NMHC)  to  CO fraction
over the model  simulation  period.   In this  way  the  boundary  condition for HC
could be obtained  indirectly  from the more  reliable  CO  observations.   It was
found however,  that this method occasionally  gave spuriously large HC values at
the upwind  boundary.   In a simulation with  significant advection  through the
model  domain  this had  the  potential for  creating  an overly  reactive  mix of
photochemical  pollutant  species in the box  that would  lead to an overprediction
of  03.    In the  simulations   presented  here,  if  a  particular  HC background
concentration   was  invalid,  a  value of  0.05 ppmC  for  NMHC  was  substituted.
This is  a  low-level background  value  for this  quantity that  borders  on the
threshold limits of  the monitoring  instrumentation.    In practice  this change
caused only small  differences  in most of  the  simulation results,  although in
some cases  it  had a noticeable  effect.   Background concentrations of species
other than 03  were generally very small,  while  those of  03 were  often a
significant fraction of  the  total  03  burden in the model  domain.

     PBM simulations, were  conducted for  each  of the  20 test  days  listed in
Table 1, including a repetition of the original  10 days  described in Report I.
Although  results for  03 predictions  on  all  days  are  summarized,  detailed
descriptions are  presented  for  only 5  days.   Simulation  results  for pollu-
tants CO,  N02,  and  03  for  Julian days  142,  178, and  207 of  1975  and  212 and
275 of 1976 are emphasized.

RESULTS

     Among the  20 selected  days  for  model  simulations from  the RAPS data base
are 5 in which the maximum  observed  afternoon  03 concentration occurred
within the PBM  domain.   These were days when  stagnation conditions existed with
light and  variable  winds  near the surface.  Very little of the primary emis-
sions moved out of  the  modeled  area.   During  this type  of episode,  a single

                                     13

-------
cell box model would  be appropriate for air quality  simulation,  since  most of
the pollutant mass  stays  within  the model  domain.  On  the  other  modeled days,
the maximum 03 level  occurred  farther downwind, outside of  the PBM's boundar-
ies.   Simulations were  also  conducted  on  these days  even  though maximum  63
measurements within the model  domain  were  somewhat lower than those  downwind.

     Results are  described in detail  for  those 5  days where the maximum  ob-
served  03  concentration  occurred within  the model  domain.   Report  I  high-
lighted 2 of these  days as well   as a  case  of stronger advection.  -The latter
case was used to  demonstrate the  'inertia1  effect  of  numerical  box models that
can lead to problems  where large temporal  concentration gradients  exist.   The
interested reader is referred to  the earlier report for details.

     The  PBM  simulated  concentrations  represent  spatial   averages  over  the
volume  of   the  box.    The  appropriate  observed concentration with  which  to
compare predictions is the average over all  12 monitoring sites within the box.
For  the 5  days  detailed here this  concentration  for  03  exceeded the 1-hour-
National Ambient  Air Quality Standard  (NAAQS)  of 0.12 ppm  at least once during
each simulation period.   These observed  average concentrations are represented
by  the open circles  in  the concentration  time series  described below~.   The
observed range of concentration values among the 12 sites within the PBM domain
is  another  feature  to  consider.   The highest and  lowest measurements  are
described by  the  dashed lines in the  time  series.   A  coarse  measure of model
performance  is  made  by  observing  whether  the  PBM  predicted  concentrations
(shown  as   solid  lines),  fall  within  the  range  of  observed  concentrations
(between the dashed  lines)  and preferably as  close as  possible  to the average
observed concentrations (shown as circles).

     The first simulation day  considered is Day 142-75.   Synoptic meteorolog-
ical features on  this day included an expansive high-pressure system centered
over western  Virginia with a  general  windflow from the southwest  in the mid-
Mississippi  Valley area.   Wind speeds  in St.  Louis, however, were light and on
the  order of  1.0 m s~*.   Skies  were  completely clear.  The maximum afternoon
hour-average  03   measurement  occurred  at  noon  at site 101  in   downtown  St.
Louis.   According to  the convention  for  time  referencing  employed  here,  the
1200  hour-average refers  to  the  average  concentration from 1200 to  1300 h.

                                       14

-------
    Figure 4a shows  the  time series for CO on Day  142.   The model  predictions
fall within  the  'envelope1  of  observed  concentrations and  actually  track  the
average observed values  quite  well.  Carbon monoxide  is  included in  the model
evaluation analysis because  it  reacts very  slowly  in  the  atmosphere and may be
considered virtually  inert  for the  time  scales  discussed here.  As  such,  any
problems with the emissions or the advective and  dispersive portions of the PBM
would be  reflected  in  the CO predictions, whereas  the complex chemical  inter-
actions of  a more  reactive pollutant  could mask  problems  with transport or
emissions.

     Figure  4b  shows the time  series for  N02  on  the same  day.   These model
predictions  are  consistently greater than  the average observations throughout
the simulation but  generally fall  within  the  concentration  envelope.   Figure
4c  displays  the  63  time series.    PBM  predictions  track the  average observa-
tions through mid-morning and  then  diverge  near noon when  observed  03  levels
begin to  decrease while  the predictions peak one hour later at 1300  CST.   The
model-predicted  peak 03  concentration  reached  0.148 ppm  compared  with  the
average observed  peak  of 0.137  ppm, an 8*  overprediction.    This  is somewhat
better than  the  28% overprediction cited in Report I  for Day  142.   The model
predictions  fall  mostly  within  the concentration envelope except for the last
few hours of the simulation.  Background concentrations  of  03 on Day 142 were
on the order of 0.06- ppm.

     The  synoptic-scale meteorological features on  Day 178-75  show  the eastern
third of  the U.S.  covered  by  a  large high-pressure system.    Wind  speeds were
light and directions variable  near  the  surface  as  well  as  at upper  levels of
the atmosphere.   In  St.  Louis surface  winds  were generally  less  than  1.0  m
s~l.  Skies  were hazy most  of  the day with few clouds reported.  Solar insola-
tion levels were high throughout the simulation  period.  The  maximum hour-aver-
age 03  measurement  occurred at  site 112  in  the  western  part  of  the  model
domain at 1400 CST.   Background concentrations  of  03  on  Day 178 were approxi-
mately 0.08 ppm.

     The  time  series for CO on  this day is presented in Figure 5a.   The  PBM
underpredicts CO consistently  throughout the period although  most  predictions
                                      15

-------
do fall within the  concentration  envelope.   The most serious  underpredictions
occur in the earliest hours of simulation where the  differences observed  among
the stations themselves are greatest.   The large departures at the upper edge
of the concentration envelope at that time illustrate this.  These high values
tend to bias the  average within  the model  domain  to the  high  side.

     The NOg  time  series in Figure 5b  shows the  predictions to match the
observations throughout the period except for the, observed  peak at 0700 CST.   A
phenomenon  may  be  occurring  here  similar  to that  which  biased the  CO  peak
toward higher values.  Nitrogen dioxide,  however, is predominantly a secondary
pollutant  and  should  therefore  not  demonstrate  fluctuations ' as  large  as  a
primary pollutant like  CO.   Predicted  values for 03 for  Day  178  are seen  in
Figure 5c.   They lag behind the observations  during  the morning hours and peak
two hours  later  in  the afternoon.  The  rapid morning  rise of the 03 observa-
tions may  be due  to a  rather large entrainment  rate of 03 aloft.   If the as-
sumed rate of rise  of the mixed layer  in  the PBM is  too slow for this particu-
lar day  the predictions would tend to  lag the observations as they do  here.
The model-predicted peak of 0.173  ppm was 8.8* greater  than  the observed  value
of 0.159 ppm.  This is  an  improvement  over  the 38* overprediction indicated in
Report I.

     On Day  207-75  .a high-pressure system had  settled into  the northeastern
quarter of the  U.S. after moving  rapidly southeast  from  the  western Canadian
provinces.   Winds in the St. Louis area  quickly  dropped to near calm  following
a frontal  passage the previous day.   Skies were  clear  throughout the  day.  The
maximum hour-average observed  03  concentration  occurred  at site  113 at 1400
CST,  in  the northwestern portion  of the  PBM domain.   The background value  of
03 on this day was generally about 0.07 ppm.

     Model  predictions  of CO shown  in Figure  6a  agree  quite  well with  observa-
tions for  Day 207.   Again  there are several  large positive anomalies  along the
upper edge of the CO observed concentration  envelope,  similar  to those  occurr-
ences on other modeled  days.   In  Figure  6b  the  model-predicted N02  concentra-
tions peak at  the  same hour and  follow  the  same trend as the observations  in
the time series  shown,  although  the predicted values are  consistently  greater
                                       16

-------
than those observed throughout the simulation period.   With  one  exception they
remain within the N0£ observed concentration envelope.

     Model-predicted 03 values on  Day  207 rise more rapidly  than  the  observa-
tions would  suggest in Figure  6c.  The average observed 03 peak is a  sharp
one  occurring  at  1400 CST  at 0.140  ppm,  while  the  predicted  peak  is  more
plateau-like with a high point at 1500  CST of 0.145 ppm,  a 3.6% overprediction.
Possible reasons  for  the  model's  more rapid  generation  of  03 during  the
morning  hours  include  an excessive  entrainment  rate  from  aloft or an  overly
reactive assumption concerning the mix of organic hydrocarbon species on this
day.

     A weak  stationary front separating  2  similar zones of  high  pressure  was
situated' near  St.  Louis on  Day  212-76.   The front' was  aligned mostly  in  an
east-west direction and had  drifted  north of the  area during the  previous  day
enabling  the  southern-most  air  mass  to  dominate.  Winds  at St. Louis  were
mostly from the south and southwest at  speeds of  1-2 m  s~l during the  simulation
period.    A  layer  of  high   clouds  persisted throughout  much of  the  day  and
decreased  the  available insolation  by  10  to 20%.   The maximum  hour-average
03  measurement occurred  at  noon  at  site  108,  located  in  the  northeastern
corner of  the  model  domain.  The  background 03  concentration on  this day  was
0.10 ppm.

     The time  series for model-predicted concentrations  of CO and  N02 for  Day
212 are shown in Figures 7a  and b,  respectively.   In both  cases the predictions
track the  observations well  through most of the  period  except  for the  early
hours when  the peak  observed concentration  is  underpredicted.    The 03  time
series in  Figure 7c  shows  good agreement between  domain-averaged  observations
and model predictions.   The  model-predicted concentrations rise more quickly in
the  early  hours  and decline more  slowly  in the  later  hours  of  the simulation
than  the observations.   The  average   observed  03 peak  occurred  at  noon  at
0.127 ppm and  the maximum estimated  concentration  was  0.121  ppm  at 1400  CST,  a
4.7% underprediction.   It  appears that less than 20%  of the observed 03  was
produced locally on Day 212  in St. Louis  since the background levels  were only
slightly less than  the  highest afternoon values.
                                      17

-------
     Day 275-76 is considered to be the most  extreme  case  of stagnation during
the 1975-76  RAPS  project.   The  synoptic  meteorological  conditions  revealed  a
high-pressure  region  over the  Texas-Arkansas area with  northwest  flow  aloft
over St. Louis.   Near the surface however, winds were very  light  and variable
in direction with little transport through the area.   Skies  were clear all  day
and a  strong  subsidence  inversion over  the  city  held  the  mixed depth to  a
maximum of  just over 500 m  above ground during  the  afternoon.   The greatest
hour-average 03 measurement  occurred  at  site  102,  just  north  of  downtown,  at
1400 CST.   Background 03 levels on Day 275 were approximately 0.06 ppm.

     The time  series  for CO on  this  day is  presented  in  Figure  8a.   Overall
concentration  levels  substantially exceed those  on  the  other modeling  days
presented here.  Report I discusses the implications of attempting to apply the
PBM to periods of extreme atmospheric  stability  as  occurred  at the very begin-
ning and end of the simulation period  on Day 275.  The net effect is a substan-
tial  underprediction  of  the  peak observed  concentration  during  the  stable
period.   This  can  be seen  in  the CO  time  series.   Both N02  and  03 concen-
tration levels, seen  in Figures 8b and c, are  also  considerably  greater than
those  on  the  other  days.   The  model-predicted  N02  concentrations  follow  the
average  observed   concentrations  well   and   show   slightly  smoother  temporal
gradients  than those observed.    Simulated  03  concentrations  also  display  a
trend  similar  to  observed 03.   The  afternoon predictions  actually track  the
upper boundary of the concentration envelope better than  they track the average
observed concentrations.   The estimated  03 maximum is 0.223 ppm  at 1500 CST,
while  the   average  observed  peak  is   0.183  ppm  at 1400  CST,  a  21.9 percent
overprediction; nearly the same as that in Report I for Day 275.

     Model   performance  on  these  5  days  indicates the PBM can  simulate  the
average  urban   air  quality  for  selected  pollutants in  relatively  stagnant
conditions.  The weakest  performance  seems  to be with the primary pollutant CO
because of  the non-homogeneous character  of its  emission  and dispersal.  While
the areas  of maximum observed 03  on  the  remaining 15  days  listed  in Table 1
did not occur  within  the PBM domain  the  model was  exercised for these days as
well in order  to provide a consistent statistical summary with the other models
discussed  in this  report.   In  the  tables and  figures  that  follow data from
model  performance on  all 20 days are included.                         :

                                      18

-------
     One significant  aspect  of  photochemical  transformation omitted from these
discussions on  PBM  simulation of an urban  atmosphere  is  the presence of vola-
tile organic compounds,  or hydrocarbons (HC), in the  smog  mixture.  While the
chemical kinetic  mechanism  in  the  PBM does  account for HC  interactions  with
other reactive  species,  a  discussion of the details  of the mechanism is beyond
the scope of this paper.

     Table 2 summarizes the 03 maxima predicted and PBM domain-average observed
maximum  concentrations  and  an  elementary  statistical  analysis of  the  results
for all  20  modeled days.   The  "specific" model predictions  correspond to the
same hour as the  observed  maximum,  and  the "independent"  predictions represent
the peak at  any hour of the  simulation.   The specific and independent predic-
tions may not  coincide,  indicating  a phase lag  between  observed  and predicted
03  peaks.    This  lag often  appears when  the maximum  observed 03  within the
model  domain occurs before noon.  Statistics on the residual concentration, AC,
have been computed for both the specific and independent predictions.  Both the
average  signed residual  and  absolute  residual,  and  standard deviation, are
presented for  the  analyses.    The  average residual  is  negative  in both the
specific and  independent cases,  indicating an overprediction  of  03,  although
the specific value  is   one-half  the independent  value.    The values of the
average  absolute  residuals  are  both  different  from  the  average  signed  resi-
duals.   This  implies that there  were  underpredictions as  well  as overpredic-
tions  among  the  individual   days.   The  magnitude of  the  standard deviation
(s.d.)  is slightly  greater than the average  residual.  The discrepancy on Day
251 accounts for  a  large portion  of the s.d.   The  average value of the model's
overprediction  for  03 over  all  20  days is 23%, compared to  the  31%  overpre-
diction  for  the original  10  days cited in  Report  I.   It should be noted  that
the average  value  of the overprediction  over  the  5 days  where  the  maximum
measured 03  occurred within the PBM  domain   is 7.5%,  considerably less  than
that for the full  set of days.

     Although the analysis in Table 2  summarizes one  aspect of model  perform-
ance,  it is clearly not  a  complete  picture of the  accuracy or precision of the
PBM.  To more fully appreciate  PBM  performance,  the  model  can be  viewed from a
broader  statistical  perspective.   Figure  9 presents frequency  histograms  (RS)
                                      19

-------
TABLE 2.  A SUMMARY OF PREDICTED  AND  OBSERVED 03 MAXIMA FOR THE PBM

Julian
date*
142 (1975)
178
182
183
184
207
209
221
230
231
251
159 (1976)
160
195
211
212
225
226
237
275
Hour
(CST)
12
• 14
13
13
13
14
10
13
13
10
15
13
14
15
15
12
12
15
11
14
Observed
at 4-meters
(ppm)
0.137
0.159
0.094
0.115
0.114
0.140
0.108
0.132
0.072
0.084
0.084
0.125
0.162
0.151
0.094
0.127
0.111
0.144
0.115
0.183
PBM
Specific
(ppm)
0.125
0.162
0.128
0.070
0.182
0.144
0.087
0.125
0.096
0.059
0.205
0.203
0.166
0.132
0.084
0.114 '
0.088
0.171
0.132
0.216
Predicted
Independent
(ppm)
0.148
0.173
0.140
. 0.097
0.182
0.145
0.122
0.125
0.119
0.101
0.205
0.210
0.170
0.137
0.088
0.121
0.092.
0.171
0.156
0.223
aTwenty days selected from 1975 and 1976.
     For the data displayed above:
       AC = Obs - Specific
       IT = -0.012
 s.d.UC) =  0.039
     TACT =  0.029
      AC = Obs - Independent
      1C = -0.024
s.d.UC) =  0.036
         =  0.031
                                  20

-------
of CO,  N02 and  03 residual  concentrations,  the observed minus  predicted
model results, over all test days.  An  ideal  distribution would be narrow with
a peak at the 0.0  residual.  The  3  histograms  shown  display  a  range  of  distri-
butions.   The RS  for CO  is skewed  toward  the underpredictive side while
03 is  skewed  toward  the  overpredictive.    The nitrogen  dioxide RS  is more
symmetric than the others  but  does  contain a few outliers on  the underpredic-
tive side.  All  3 distributions show a peak at or very  near 0.0.

     A plot of the residuals versus observed concentrations  (RO)  should  indic-
ate whether a bias in the  predictions displays  any trend with  the magnitude of
the measured concentrations.  Figure 10 shows RO plots  for the' same  species as
presented in  Figure  9.   For CO  and  N02   there  appears to   be  a  trend  toward
greater underprediction with higher ambient concentrations.   The trend  is more
clear with  CO.    Ozone  however,  displays  no particular trend  with observed
concentration, but the PBM  generally overpredicts  ambient  values  above 0.05
ppm.

     The PBM has been  exercised on  a  variety of days taken  from the RAPS data
base.  Prevailing meteorological conditions occurring some of  the days may not
have been compatible with  the environment  for optimum model results.  Using the
PBM  in  a  regulatory  capacity  could  be  restricted  to worst  case  stagnation
conditions where the  model  seems to  perform  best.   Time series for pollutant
species on test days other than those discussed  here are displayed in Appendix
A.   Also,  tabulated  model  performance  statistics  for  data  on mean  concentra-
tions,  analysis of trends,  and  concentration  maxima  from the  PBM are provided
in Appendix B.

CONCLUSIONS

     A thorough evaluation  of  the performance  of  the  PBM  has been  completed.
This  analysis is a major portion of the documentation  of the model's abilities
for air quality simulation.  Evidence shows that the model is  a useful  tool in
assessing urban  air quality for photochemically  reactive pollutants,  especially
in stagnation conditions.   The PBM is  relatively  simple to  use  and its data
requirements  are  far  less  stringent  than most other numerical   air  quality
simulation models.   Areas  of further study  that  may be  pursued  include:
                                      21

-------
(1)   the  hysteresis problem  during  advection conditions,  discussed  in
     Report I;

(2)   the  relationship  between the  average 03  concentration observed
     within  the model  domain and the maximum  03  level  observed  at a
     single station;

(3)   the  continued  testing and refinement $f the  chemical  kinetic mechan-
     ism  within the PBM;

(4)   model  sensitivity to variations in selected  parameters  such  as
     initial  and  boundary  concentrations, initial cell depth, emissions,
     wind speed and solar radiation.
                                22

-------
                    RISING MIXED
                      HEIGHT
                              1      ENTRAPMENT OF
                                    POLLUTANTS ALOFT
Figure 2.   Schematic drawing  of the PBM domain.
                       23

-------
                                                     10
                                              SCALE, Km
Figure 3.  Map of St. Louis area depicting extent of PBM modeling
           domain (boxed area), city of St. Louis (shaded area),
           RAPS monitoring stations (circles), and radiosonde
           ascent site (triangle).
                              24

-------
                    PBM SIMULATION-750522
   a.
   Q_
   o
   o
          I  I I  1  I I  1 f  I  I I  1 I  I I  I  I I  I I  I  I I  I I  I  I
               7.5
 10.0     12.5     15.0
TIME,  HOURS (CST)
17.5
20.0
Figure 4a.   PBM simulation results for Day  142 of 1975.  Time
            series of average observed (circles), range of
            observed (dashed lines),  and  model-simulated (solid
            line) hour-average concentrations of CO.
                             25

-------
                    PBM SIMULAT10N-750522
          I J I  I 1  I. J__ L I . I  I I  I I  I I  I  t I  I I  I I  I  I
               7.5      10.0     12.5      15.0     17.5
                      TIME,  HOURS  (CST)
  20.0
Figure 4b.  PBM simulation  results for N02 - Day  142 of 1975.
           Keys to figure  described in 4a.
                   PBM  SIMULATION-750522
         • -*^'
         ^*5r9—**^ I  i I  1 1 J_ 1 L  t i  | | ' t i  |  I i  j 111 II
               7 A      10.0      12.5     15.0     17.5
                      TIME,  HOURS  (CST)

Figure 4c.  PBM simulation  results for 03  - Day  142 of
           Keys to figure  described in 4a.
                            26
 20.0
1975.

-------
   Q.
   Q.
   O
   o
                    PBM SIMULATION-750627
               i  I i  i i i ^—r—r"i i  l'*7~n~T~
        5.0
Figure 5a.
    7.S      10.0     12.3     15.0
           TIME,  HOURS (CST)
17.5
20.0
PBM simulation  results for Day 178 of 1975.   Time
series of average  observed (circles), range  of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                             27

-------
                    PBM SIMULAT10N-750627
    Q.
    Q.
    CM
    O
          I  1 l__..i_.. I ..!_ J i  1 Ll  J 1.1 II  I I  if I  I	I  J_ I  1 ._!_.!. J	
     0.00
M>
                       10.0     12.5     15.0     17.5     20.0
                      TIME,  HOURS (CST)
Figure 5b.   PBM simulation results  for N02 - Day 178 of 1975.
            Keys to figure described  in  5a.
                    PBM SIMUIAT10N-750627
                       10.0     12.5     15.0     17.5
                      TIME,  HOURS (CST)
                                               20.0
Figure 5c.   PBM simulation results  for  03 - Day 178 of 1975.
            Keys to figure described  in 5a.
                            28

-------
     0.
     CL
     o
     o
          ,»--\
                     PBM SIMUL4TION-750726
           i i  i i  i i  i i  t  i i  i   i i  i i  i i  i i  i  i i  i



7J      10.0     12.5     15.0
       TIME,   HOURS (CST)
                                               17.5
20.0
Figure 6a.   PBM  simulation  results for Day 207 of 1975.   Time
            series  of  average observed (circles), range  of
            observed (dashed lines), and model-simulated (solid
            line) hour-average concentrations of CO.
                             29

-------
                    PBM SIMULATION-750726
                I  i  i i  i j  i T  v"i i  T i  i i  i  r
          i  i i  i *  i i  i i  r i r**T~r*i t  r~~7~t*~i~TT*'i  i •  i <  i
     0.00
7 A     10.0     12.5     15.0     17.5
       TIME.   HOURS (CST)
                                                      20.0
Figure 6b.  PBM simulation  results for N02 - Day  207 of 1975.
           Keys to figure  described in 6a.
                    PBM SIMULATION-750726
    Q.
    Q_
                                IlllfJlllllilli
     0.00
        10.0      12.5     -15.0     17.5
       TIME,  HOURS  (CST)
                                                      20.0
Figure 6c.   PBM simulation results for 03 - Day 207  of 1975.
            Keys to figure described in 6a.
                            30

-------
Q.
CL
o
o
                 PBM SIMULATION-760730
                                     A
                    -•Kr i  iN—I—r-f-TT-l—
        54
            7.5      10.0     12.5     15.0
                   TIME,  HOURS (CST)
17.5
20.0
Figure 7a.
        PBM simulation  results for Day 212 of 1976.  Time
        series  of  average observed (circles), range  of
        observed (dashed lines), and model -simulated (solid
        line)  hour-average concentrations of CO.
                         31

-------
      .100
      .075
                    PBM  SIMULATION-760730
    o.
    CL
      .090
    O
      .025
     9.000
                   ! I  1 I    I I  I I  f  * III j  I I  I I  I  I I  I I
               o   \
                      ^
                                  o  o
          i  i  i i  I i  i  i i Nf*~i-T~rT"t-r"r"r-r"hT—r-i—i
        5.0      7.5      10.0      12.5      15.0      17.5      20.0
                       TIME,  HOURS (CST)

Figure 7b.   PBM  simulation results for N02 -  Day  212  of  1976.
            Keys to figure described in 7a.
                     PBM  SIMULATION-760730
    QL
    Q.
                           i i  l i  I  i i  ( i  i  i t  r i  1  i i
      0.00
                7A      10.0     12.5     15.0     17.5
                       TIME,  HOURS (CST)
20.0
Figure 7c.   PBM  simulation results for 03 - Day 212  of  1976.
            Keys to  figure described in 7a.
                             32

-------
                     PBM SIMULAT10N-761001
     Q.
     a.
     o
     o
1W.W






7.5




5.0







Z5


1 I I 1 I 1 1 1 1 I 1 1 1 1 1 1 1 1 I I 1 I 1 1 I
' ' 	 , /
' v i
• / * ,'
I V '
• / \ '
i \ i
"/
'/
••'
i
* \ :
_ \ !
\ '
\ (
• \ t
o \ <
x 1
-oo
^. V ' O
s~~~^ o \ .'
-/\^ • . •- 	
^x^ ^ 	
• 	 s ,-T::==:^ 	 °- ' w y u
Y ^-' ^^ 	 .
i i i i 1 i i i i ! t i i i I i t i i 7~ i ~i i i I
i i i i
-

•

•

—
•
-
-
-
—


_

-


~
-
till
         5.0      7.5     iaO     12.5     15.0     17.5
                       TIME,  HOURS  (CST)
20.0
Figure 8a.   PBM  simulation results for Day 275 of 1976.   Time
            series  of  average observed (circles), range  of
            observed (dashed lines), and model-simulated (solid
            line) hour-average concentrations of CO.
                             33

-------
                     PBM  SIMULAT10N-761001
     0.
     o.
                                 1(1(11111111
      0.00
                7A      10.0     12.5     15.0     17.5
                       TIME,  HOURS (CST)
 20.0
Figure 8b.   PBM simulation results for M02 - Day 275 of 1976.
            Keys to figure described in 8a.
                     PBM SIMULAT10N-761001
                               Ililllllilflllll
                 7A     10.0     12.5     1.5.0     17.5
                       TIME,  HOURS (CST)
20.0
Figure 8c.   PBM  simulation results for 03  -  Day  275 of 1976.
            Keys to.  figure described in 8a.

                             34

-------
         RESIDUAL HISTOGRAM   DETERMINED OVER ALL TIMES
    250-1-001
<

Q
CO
LJ
cr
 200+001 •

 150+001 U»

 .100-1-001 ***

 .5001-000

 000

-.500-i-OOO

- 100+001

-.150 + 001

-.200+001

- 250-00'
             ••••***

                    10   15
                                         40   45  50
                                 20  25   30   35
                                    FREQUENCY
                     CARBON MONOXIDE
                 DATA AVAILABLE =   260    DATA  POSSIBLE  =  260


Figure  9a.   Residual  histogram from the PBM simulation results
             for all  20 test  days for CO.
                           35

-------
            RESIDUAL HISTOGRAM   DETERMINED OVER AU_ T:MES
        .750-001
    2
    Q.
    Q.
    g
    on
    U-l
    cx
.500-001 (•


 450-001


.300-001


.150-001

.000


- 150-001


- 300-001 •


-.450-001 -


-.600-001


-.750-00'
                   10   20   30   40   50  60  70  80  90  '00
                                  FREQUENCY
                   NITROGEN DIOXIDE

                DATA AVAILABLE =  260   DATA POSSIBLE  =  260

Figure 9b.   Residual histogram from the  PBM simulation  results
             for all  20 test  days for N02-
             RESIDUAL  HISTOGRAM  DETERMINED OVER ALL TIMES

Q.
0^
_l
r>
Q
on
LU



^^u^uuu
.200-t-OOO
.150+000
.100+000
	 	 ]
j
1
500-001 ~
000

8BSSSSSS9S3SS8ES*
- 500-001 ~— .
- 100+000?
if
-.150+000
-.200+000
- 250+000

•
                           30  40  50  60
                                  FREQUENCY
                   OZONE

                DATA AVAILABLE =  260
                                        30   90   '00
                                 DATA POSSIBLE  =  260
Figure 9c.   Residual  histogram from  the PBM simulation results
             for all  20 test days  for 03.
                                36

-------
                 RESIDUAL VS OBSERVED  CONCENTRATION

2
Q.
Q.
*t
RESIDU>

.200+001
150+001
.100+001
.500+000
.000
-.500+000
-.150+001
-.200+001
- 250+001
000

a a
0 an °cP 3 °
MWM J3D *3ri 0 Q O
jj^jBS^am1 . °°
^ BJ P .3} ^o i01 a a !|li
o
QQ O
a
1 ' 1 ) t 1 1 1 ! ^_


LCCENO
FREO SYM
< i, i> a
< 2. 2> 0
< 3. 3> A
< 4,99> «

100+001 .200+001 .300+001 400+001 500+001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON  MONOXIDE
             DATA AVAILABLE =  260    DATA POSSIBLE  =  260

Figure 10a.   Residual vs.  observed  plot from the PBM  simulation
             results for all  20 test days  for CO.
                              37

-------
                 RESIDUAL VS OBSERVED CONCENTRATION
    s
    a.
    Q
    cn
    UJ
    or
./3U— UU 1
SOC-001
.450-001
.300-001
.150-001
000
- 150-001
-.300-001
-.450-001
-.600-001
-.750-001
000
-i
O
a
Batz»
jtf^yf ,,,,,'
PT ffl O
•
•
.500-001 .100+000 150+000 200+000 250-t
                                                         LEC£NQ
                                                        FREQ STM
                                                       < i. i>  a
                                                       < 2. 2>  0
                                                       < 3-
Figure lOb.
                    OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE =  260    DATA POSSIBLE  «  260

             Residual  vs. observed plot from the  PBM simulation
             results for all  20 test days for N02-
                 RESIDUAL VS OBSERVED CONCENTRATION


a
a
~


c/5

LJ
a:






x^u^uwnj
200+000
150+000

.100+000
.500-001
°


-.500-001
-.100+000
- 150+000

- 200+000
- 250+000

•
.

-
0





^MI Hi ii i n ii i'' I i* "^ !»! "1 f"T*f'l m
i&wrrtmrnvr'F® ^"rjtr^lgr^i tLlSiLj jfl ! ' ' 1
tvi ^jn&- JjiP' ft r^T?Tn.''ii<'!rn^^3^^^Q MJLjsi* c* i

*^L 2_ ^ z-rr^.ifr _ _^^ n jr
- LJ PLPl /TO- ! BIIHI *•" "
UJ el CJ ' *3Ufvn n
Jr 'Jffi*f n
m ™
a OD 3



i r I ,
LtCCNO
FREO SYM
< 1. !> 3
< 2, 2> 3
< J. 3> A
< *,99> «


QQQ 500-001 .100+000 ,150+000 200+000 250+000
                    OBSERVED CONCENTRATION  (PPM)
                  OZONE
            DATA AVAILABLE =  260    DATA POSSIBLE  -  260

Figure lOc.  Residual vs. observed  plot from the PBM simulation
             results for all 20 test  days  for 03.
                              38

-------
                                  SECTION 5

                       LAGRANGIAN PHOTOCHEMICAL MODEL

OVERVIEW

     The Lagrangian  Photochemical  Model  (LPM)  was developed  by  Environmental
Research and Technology, Inc. (ERT) and adapted under contract with EPA for use
with  the RAPS data base.   (The LPM  is  essentially  identical .to  the  general-
use model  named ELSTAR).   The  LPM  envisions  a portion  of the  atmosphere  as
an identifiable parcel  which  can be tracked from  early morning to late after-
noon.   As  the parcel  moves over the  various emissions  sources, pollutants are
assimilated,  vertically mixed,  and  subjected to  photochemical  reactions in the
presence of solar radiation.   The  LPM  is attractive  relative to  grid models  in
that  it  is fairly  simple  to  execute  and uses  a  moderate  amount  of  computer
time.    On   the  other hand,  the LPM  calculates concentrations only within  a
parcel and  not over a complete spatial field.

     Previously, LPM  simulations  were  conducted  for  10  days from RAPS  and
documented   in Report I.   Findings  from  that  study pointed'to  the  need  for
lateral  diffusion  to be  included for the parcel  and indicated the  importance  of
initial  conditions  to  the  final ozone  predictions.   Based  on those findings,
the model  has been modified  to  include a Gaussian-type lateral   spread of the
parcel.   Furthermore,  the manner in which the  initial  condition profiles of NO,
N02 and HC  are  loaded  has  been  modified  from  the  original  recommended method.
A  series  of  simulations  over  20  days  (including  the original   10) has  been
executed,  and results are  summarized in this  section of the report.   Meteoro-
logical  characteristics  of  the test days are listed i.n Table 1.

METHOD

     The LPM  is executed using a series of program modules.   They are METMOD,
EMMOD and  KEMOD,  sequentially  performing  calculations  on  meteorology  and  air
quality,  emissions  and photochemistry.    The  input  and   running  procedures
                                      39

-------
described  by  Lurmann et  al.  (1979)  have been  generally  followed,  with  some
modifications as  discussed  in  Report I.   In addition, running procedures  for
the  expanding  Lagrangian parcel  were established  and the  method  of  setting
initial concentration profiles  was altered.   As done  previously,  the  parcel
start position was set so it would  arrive  at the station  observing the  maximum
1-h ozone value for the  day.

     Following the  initial 10 simulations, attention  was focused  on  the
results for  Day  251  where the ozone  prediction  was  a  factor of 2 higher  than
the  maximum  observed concentration.   This  single  case was  important   in  the
summary results  since the large residual  (observed minus  predicted)  accounted
for  a  substantial  portion  of  the standard deviation of the  residuals over the
10 days.   Based on early  morning  surface concentrations,  the parcel  on  Day 251
was  initially loaded with very  high  reactive hydrocarbon  (RHC)  concentrations,
and  parametric analysis indicated that the initial  loading was  totally  respon-
sible  for  the prediction  exceeding  the  observation.    Accepting  the  initial
condition formulation,  the concept of a fixed parcel  size within the  LPM seemed
a  likely contributor to  the over-prediction.  It seemed more reasonable that,
as a parcel leaves the central,  high-emissions areas,  it should expand so as to
simulate lateral  diffusion.  The net effect would be to decrease concentrations
of  precursors and,  ultimately,  ozone.    Therefore,  under   contract  to  EPA,
Lurmann  (1980,  1981)  implemented an  expanding  parcel  scheme in the  LPM.   The
rate of  expansion was based on  Gaussian plume  ay values.   The meteorological
preprocessor, METHOD, now produces a parcel-width schedule in addition to other
outputs.   The METMOD input allows  the user to  specify a  proportionality  con-
stant  for  the a y expansion and  to choose  a  time  when the  expansion  begins.
To maintain  an  objective  evaluation the proportionality constant was taken as
1,  and the  parcel  started expanding  at  0800  CST  for  all  simulations.   The
expansion  rate  was controlled by  an  internally  computed  stability  class,  but
generally  parcel  widths  of 7-10  km were achieved by  midafternoon (from  an
original 5 km).   The effect on  pollutant concentrations  was  noticeable  but not
entirely  satisfactory.    For  example, the Day  251  maximum   of  0.53 ppm  from
Report  I  (Table  4)  dropped  to  0.37  ppm,  and  the s.d.  ( AC)  over the  10  days
dropped from 0.11 ppm to 0.068 ppm.  However, in the second set of 10 days, Day
237  of 1976  also  had high  initial  concentrations with  resulting  over-predic-

                                      40

-------
tion, so that  the  s.d.  (AC) over 20  days  was an unacceptably high 0.090  ppm.
It was  clear that the expanding  parcel  alone would  not  solve the problem  of
substantial  overprediction in certain cases.

     These results prompted a re-examination  of the  method used for initali zing
the  pollutant  profiles  for HC and  MOX  (only surface measurements were  avail-
able).  As  originally implemented by ERT,  the assumed initial profile  showed
concentration decreasing to  1/2 the  surface  observation at the morning  mixing
height  -  typically 200  m -  and  remaining  constant  at that  value  up  to  the
parcel  top  (the  afternoon mixing  height,   typically  1500 m).    The flaw  in
the  original  scheme  is  that  the  assumed total  mass  above the morning  mixing
height may  be  very large if the  surface concentration  is large.   High  HC  and
NOX  initial  concentrations  are  usually  the  result of  accumulation  of  near-
surface emissions  during  very calm nights.   Therefore, it is  reasonable  that
the  true vertical  extent of the high loading  is only up to the morning mixed
layer  and  that  more-or-less background  levels should  exist above.    Such  a
scenario was  implemented  in  the  LPM for  the final  20 simulations,  with  the
surface, observed  concentrations  of  MO,  N0£ and RHC  assumed constant  to  the
lower boundary of Level-3  in the  parcel  (up  to  150-200  m)  and assumed to be at
the lower default through Levels-4 and 5.   The initial  concentrations  are given
in Table 3,  along  with  the defaults assigned in the upper  parcel.  The  method
of estimating the  individual  hydrocarbon class concentrations in ppm  based  on
the  reactive hydrocarbon measurement  (RHC)  in ppmC  was revised from  the orig-
inal   proposal  by  ERT  (Lurmann    et  al.,  1979).   The  new method,  also  used
in the first 10 simulations,  is:

                       [Alkanes]          =  0.0786   [RHC]
                       [Alkenes]          =  0.0328   [RHC]
                       [Aromatics]       =  0.0248   [RHC]
                       [Formaldehydes]   =  0.0405  '[RHC]
                       [Other Aldehydes]  =  0.0405   [RHC]

     The initial profiles  of 03 were derived  by the same method  as  described
in Report I.  Initial  profiles for CO were derived  in  a manner similar to  that
for  RHC and  NOX described above.   However,  the background value  for  Levels-4
                                      41

-------
TABLE 3.  INITIAL SURFACE CONCENTRATIONS ASSIGNED TO THE PARCEL FOR REACTIVE
          HYDROCARBON (RHC), MO, CO, 03, AMD INITIAL CONCENTRATIONS ASSIGNED
          FOR 03. ALOFT.


                    Day     RHC      NO      CO      03    03 Aloft
                           (ppm)    (ppm)   (ppm)   (ppm)     (ppm)
- --
142
178
182
183
184
207
209
221
230
231
251
159
160
195
211
212
225
226
237
- 275
0.26
0.19
0.52
0.15
0.39
0.11
0.32
0.14
0.12
0.15
1.28
0.22
0.05a
0.05a
0.07
0.10
0.05a
0.41
1.57
0.70
0.03
0.01
0.04
0.01
0.03
0.01
0.02
0.01
0.02
0.01
0.08
0.01
0.01
0.00a
0.01
0.01
0.01
0.01
0.06
0.10
0.44
0.41
0.73
0.18
0.77
0.27
0.73
0.69
0.50
0.48
2.60
0.86
0.32
0.17
' 0.13
0.11
0.32
0.97
2.3
2.82
0.00a
0.00
0.01
0.01
0.01
0.01
0.01
0.02
0.00a
0.01
o.ooa
0.01
0.02
0.03
0.01
0.01
0.04
0.01
0.01
0.01
0.06
0.08
0.12
0.12
0.12
0.07
0.08
0.08
0.06
0.06
0.06
0.11
0.11
0.08
0.05
0.11
0.07
0.09
0.11
0.06
          aDefau1ts:  RHC=0.05, N0=0.0025, 03=0.00'25
                                      42

-------
and 5 was  inadvertently  set to a rather high  0.5  ppm.   The general  effect  of
this oversight  would be to  add several tenths  of a ppm  to the afternoon  CO
predictions.   This should be kept in mind when viewing simulation results,  but,
since CO  has  negligible influence  on 03  production,  there was  little  reason
to repeat the simulations to adjust  for this error.

     Expansion of  the  parcel  implies entrainment of  surrounding air during  a
simulation.  LPM therefore requires an estimate o-f  the  concentrations for  side
boundary conditions.   For  RHC,  subject  to the split given above,  the  assumed
value was  0.063  ppmC.    NO and  NOg were assumed  at  continental   background
values.    Ozone  was set  to  the  observed  midday inflow from upwind  stations  on
that day by  a  procedure  described by Shreffler and Evans  (1982).   This  is the
same 03  value  given to  the upper levels of the  parcel  in  the  initial  profile
and generally varies from 0.06  to 0.12 ppm.

RESULTS

     As  was  done  in Report I,  the  time  series results  for Days  231, 251,  226
and 275  are  presented in  this section.   In  this  way  the  effect  of the  new
initial  conditions  and  parcel  expansion can  be judged.   From  the additional
set of 10 selected days,  Day 237 (1976)  time series  results are  also presented,
since high initial  loadings  of NOX  and HC  were observed on  that day.   Summary
statistics are presented on the  residual (observed  minus  predicted) concentra-
tions at  the  time and position of  the  observed  ozone maxima for all 20  days.
Time series for the remaining 15 days and temporal  correlation coefficients are
presented in  Appendix B.

     Figure lla  presents a  trajectory  map  for  Day  231.    This map covers  a
square 100 km on  a side, and only  9  of  the stations are  shown  to  avoid  clut-
tering the  figure.   These  stations  are (compare  Figure  1)  101 in  downtown,
122 and  114  to  the  north,  109  and  123 to  the east,  118  and 124 to  the  south
and 120 and  125  to the west.  The  location-of the parcel  each  hour  along the
trajectory is indicated.   The start position  is "indicated  by  "S" and  the  start
time is  given  above  the figure.   On Day   231,  the parcel  starts  in a  rural
locale south  of the city.   It moves  over the high  emissions areas at  0800-1100
CST and  continues  to the northwest.  The  03  maximum  was   recorded  at T400 CST
by station 121.
                                      43

-------
     Figure lib  gives  the 63  concentrations  for Day 231.   The RAPS  observed
values are given by a solid line.  The  model-predicted  values at Level-1  (L-l,
understood to be the surface)  are given  by a dotted  line,  and the Level-3  (L-3,
about 300 m in most cases) predictions  are given  by  a  dashed  line.   (The  reader
is reminded that the  time series represent the concentrations  within  a moving
parcel and not at a single point in space).

     The  results  for 03  do  not  differ  greatly  from those  for the  original
test  simulation.    There  is  a  slight   underprediction  up to  the  time of  the
observed  maximum.   Beyond  that time  the prediction greatly exceeds the  ob-
servation,  but,  since  the  parcel  is  beyond  the  measurement network,  this
comparison has little meaning.   The predicted NO and N02 peaks are diminished
somewhat  from  the  original  results, reflecting  the lower initial   loading  and
parcel expansion.   The CO predictions indicate  the  influx of morning emissions
with the  L-l values showing greater peaks.  In  the  afternoon, the  model  parcel
becomes  well-mixed, and  the   decreasing  tail   reflects  the  expansion of  the
parcel (compare with the original test).

     The  trajectory  for  Day  251 is given in  Figure  12a.   The  start  point is
near  downtown  St.  Louis,  and  initial  loading of  the  parcel  with HC and  NO is
high.   Furthermore, winds are  very low during the  first 4  hours  so  that the
parcel accumulates  large  quantities of  pollutants from emissions.   After 1000
CST the  parcel moves more rapidly  and  arrives  at station  122 at 1400 CST.  The
03 time  series in   Figure 12b  shows a  substantial response  to  the  new initial
condition  scheme.   Ozone shows  an initial surge in  the  morning,   but,  as  the
mixed  layer  rises,  the  cleaner air aloft is  mixed  into the  system.   The pre-
dicted peak at 1400 CST is only slightly above the observation.  As seen  in the
original  simulation of  this   day,  the L-3  NO,  N02  and  CO predictions do  a
better  job of tracking  the  observations.  The  effect  of parcel  expansion is
evident  in the tail of the CO predictions.

     As  seen in Figure 13a,  the trajectory for Day 226 moves from the west side
of the  city  to the east,  reaching station 109 at 1300 CST.   It then turns and
moves back into  the city  towards evening.  The  03 predictions for  this simula-
tion  must be deemed excellent.   They   have decreased  from those  in  Report I.

                                     44

-------
The L-3 predictions  for  NO,  MO2 and CO are  also  quite  good.  As discussed  in
Report I, the  L-l  (surface)  predictions of primary pollutants seem  to  respond
in an  unrealistic  fashion  when a new  source  is  abruptly encountered.  Use  of
the L-3 predictions is recommended since they are  essentially buffered from  the
surface sources.

     As shown  in Figure  14a,  the parcel on Day 237 started  about 10  km north-
west of downtown St.  Louis  and moved to the west-by late afternoon.  This  day
was not  treated in  Report I,  but   is  highlighted  here because  it  shows  the
highest RHC and third  highest MO initial  values over all 20  days.   The resul-
ting  03  predictions show  a  rapid  initial  build-up with  a dip  at time  of
inversion dissipation.   The  03  predictions are  reminiscent of  those for  Day
251 in  this respect.   The ozone  prediction greatly exceeds  the maximum  ob-
served.  This  case  again emphasizes the importance of initial conditions.   To
reiterate,   the conditions  assume  that the  average RHC  recorded  by  nearby
stations exists through  the lower 200  m  and over an area  5x5 km  (the parcel
size).  If,  for instance, the  vertical  extent of the  RHC  accumulation were only
to 100 m, the  final  ozone predictions  likely  would be much lower.   The  kind  of
detailed definition  of  the  initial  state  needed for  precise calculation  is
unattainable with  RAPS data.

     The trajectory  in Figure  15a indicates that Day 275 was  a very  calm day.
The parcel   starts  out  southwest of  the city center and  reaches station 102  at
1400 CST.   It  then moves south and  nearly  returns to  its  origin.   Two points
should  be  made concerning  such  calm   conditions.   First,  the  trajectory  is
subject to  great uncertainty,  and all that is really  known  is that the air mass
was nearly  stationary.   Second, the  LPM point-source module  is not  designed  to
function well  when  the wind speed is so low that the  emissions grid  (5x5 k.m)  is
not traversed  in a single hour.   The precise  impact  of  this  problem is  unknown
but may be minimal.  Ozone  predictions  are  somewhat  lower  than those in Report
I  and  in  better accord  with  observations.   MO,  M02 and  C02 predictions  are
quite similar to those in Report I.
                                      45

-------
     Table 4 presents a  summary  of the predicted and observed 03 maxima  and a
statistical   evaluation  of  the  results  for the  20 days  of the  study.    The
prediction refers  to the  model  prediction  at the  time and  position of  the
observed 63  maximum.   The residuals, AC,  are  calculated for both the  L-l  and
L-3 predictions.   The average residual  for L-l indicates  slight  underpredic-
tion, while  the  average  residual  for L-3 is essentially zero.  There  would be
an  inclination  to  choose the  L-3  prediction  because  of  the  aforementioned
tendency to  track  the observations  more consistently.   For both levels,  the
s.d.  (AC)'S  are  identical,  and  the  average   absolute  deviations  are  nearly
so.

CONCLUSIONS

     Compared to results  in  Report I,  the LPM  has  shown considerable  improve-
ment  in  its  predictive  capabilities as  indicated  by reduced s.d. (AC).   This
improvement resulted from  implementation of an  expanding parcel  and  changes in
the method of  loading  initial  conditions.   Both adjustments  reduced  the gross
overpredictions which occurred  when initial surface concentrations were high.
AC has remained encouragingly small  throughout  the  preliminary and  final  simu-
lations.  The  variability on a daily basis may arise  from  input uncertainties
(involving winds   or  initial  conditions)  rather  than  underlying model  weak-
nesses.

     The LPM  has  shown  promise  as an  effective tool to  understand  03 produc-
tion  in  an  urban  region.   The model is  relatively  easy  to  use,  inexpensive to
execute,  and seems  immune  to  various  execution  errors which  tend  to  arise
unexpectedly in complex computations of this sort.
                                      46

-------
      TABLE 4.   A SUMMARY  OF  PREDICTED  AND OBSERVED QS MAXIMA
                FROM THE LPM  FOR  20  DAYS  SELECTED FROM 1975-1976.
Day
Hour
Station
Observed
4-meters
(ppm)
LPM Predicted
Level -1 Level -3
(ppm) (ppm)

142 (1975)
178
182
183
184
207
209
221
230
231
251
159 (1976)
160
195
211
212
225
226
237
275
12
14
12
15
13
14
11
15
13
14
14
15
16
15
15
12
13
13
11
14
101
112
125
124
118
113
118
121
121
121
122
122
115
114
120
108
117
109
120
102
0.20
0.20
0.16
0.21
0.18
0.18
0.21
0.17
0.19
0.23
0.26
0.20
0.22
0.22
- 0.16
0.17
0.17
0.23
0.18
0.24
0.07
0.15
0.25
0.18
0.21
0.11
0.14
0.16
0.18
0.19
0.29
0.18
0.21
0.15
-0.11
0.10
0.08
0.20
0.30
0.26
0.13
0.31
0.24
0.18
0.22
0.13
0.15
0.16
0.17
0.19
0.30
0.18
0.23
0.15
0.12
0.11
0.09 '
0.22
0.29
0.30
    For the data displayed  above:
      AC = Obs  minus  L-l  Pred
      1C = 0.023
s.d.UC) = 0.058
    TATT = 0.052
      AC = Obs  minus  L-3  Pred
      1C" = 0.005
s.d.UC) = 0.058
    TACT - 0.050
                                 47

-------
75231. RAMS  121 AT 1400CST. START 0600
   iW) f^"^™T i  i  j ^ i  i i i  i TC^^"*^^^^^""™^^™"^^
    79
    SO
or  o
                         s  •
                  II i i  i  i  i i  I  i  i ._!_.. i  J
                       50
                    KM
   79       100
Figure lla.  LPM parcel  trajectory for Day 231 of 1975.
                       48

-------
        75231,  RAMS  121 AT 1400CST.  START 0600
   CL
   CL
   uT .2
   o
   fSJ
   o
      .1
     00 >->•
         I  I I  I I  I I  I I  I 1  I I  I I

         	08S
         	  PRED L-1
         	PRED L-3
             •rfS-f-r . I  ....  I .... I  ....  I ....
      54      7A     iaO    12J     15.0    17.5     2aO
                         HOUR, CST

Figure lib.   LPM time series of observed and predicted 03 hourly
             concentrations on  Day 231 of 1975.
    .100
    .075
        75231, RAMS 121  AT 1400CST.  START 0600
    .050
  O
  z
    .025
   0.000
                   i  I i  i i  i r
                                             iiii

                                            OBS
                                            PRED L-1
                                            PRED L-i
             7 A
                     10.0    12.5     15.0
                         HOUR, CST
2ao
Figure lie.  LPM time series of observed  and predicted NO hourly
            concentrations on Day 231  of 1975.
                             49

-------
  es
  o
     .100
    .075
    J550
    .025
        75231, RAMS 121  AT 1400CST.  START 0600
                                       	  OBS
                                       	  PRED L-1
                                       	PRED L-i
   0,000* * i  i '  i i  « '  ' t  t '  i t  i i  i t  t i  i t i  i"n i  i t
      56     7.5     iaO    12J     15.0     17.5    20.0
                         HOUR, CST

Figure lid.   LPM time series of observed and predicted  N02 hourly
             concentrations on Day 231 of 1975.
        75231, RAMS 121  AT 1400CST.   START 0600
    CL
    0.
    o"
    u
         I I  I 1
               I '
        ^ i  i i  [ i  11  j I  i i  i j_ . 1. -i
•  i ,  r j  • .  •

""—- OBS
	PRED L-V
	PRED L-1
      54      7.5     10.0     1X5     15.0
                         HOUR. CST
            20.0
Figure lie.   LPM  time series of observed and  predicted CO hourly
             concentrations on Day 231 of 1975.
                             50

-------
 75251. RAMS  122 AT 1400CST. START 0600
   • Wr r""^-^-^^^^^-^^^M«
    73
    50
    2S
             25
  90

KM
75      100
Figure 12a.   LPM  parcel trajectory for Day 251 of  1975,
                      51

-------
       75251, RAMS 122 AT 1400CST.  START 0600
           I  I I  I I  I I I  I I  t I  I I

          — OBS
          —  PRED L-1
       .	PRED L-3
                              i  i till | j i I  L I  I I
                    10.0     12J     15.0
                        HOUR. CST
20.0
Figure 125.  LPM time series of observed and predicted 03 hourly
            concentrations on Day 251 of 1975.
       75251, RAMS 122 AT 1400CST.  START 0600
   CL
   Q.
                    100     12J
                        HOUR, CST
                                           OBS
                                           PRED L-V
                                           PRED L-i
20.0
Figure 12c.  LPM time series of observed and predicted MO hourly
            concentrations on Day 251 of 1975.
                            52

-------
       ,75251. RA.V1S  122 AT 1400CST.   START 0600
        r.
                       03S
                       PRED L-1-
                       PRED L-3
              7.5
10.0     12.5     15.0     17.5    20.0
    HOUR, CST
Figure 12d.   LPM  time  series of observed and  predicted N02 hourly
             concentrations on Day 251 of 1975.
        75251, RAMS 122 AT 1400CST.  START 0600
                                        — oes
                                        —  PRED L-M
                                        —  PRED L-3
      9U)      7&     10.0    12 J     15.0    17 J     20.0
                         HOUR, CST

Figure 12e.  LPM time series of observed and predicted CO hourly
            concentrations on  Day 251  of 1975.
                            53

-------
 76226.  RAMS 109 AT 1300CST.  START 0600
   'wr^  i  i f""1^^~"i"~ i T^™r""T^""|O"T™'T"~T^^^^^ri""lT"—
    90
                        i '  -"  '  I
                        80
                     KM
75       100
Figure 13a.  LPM parcel  trajectory for Day 226 of  1976.
                        54

-------
        76226,  RAMS  109 AT 13QOCST.  START 0600
            •  • i
            - oss
            -  PRED L-1
           —  PRED L-3
                * t  I  t I  1 I  t t  I I  I
                     10.0     12.3
                         HOUR, l
19.0
17.5
20.0
Figure 13b.   LPM  time  series of observed and predicted 03 hourly
             concentrations on Day 226 of 1976.
     .19
        76226, RAMS 109 AT 1300CST.  START 0600
     .10
  Q.
  0_
     .05
    0.00
                f
                              i | i  i I^T r i i  r T  i i  r

                                       "~~~ 08S
                                       -----  PRED L-V
                                       --  PRED L-i
                     10.0    12.5     1S.O
                         HOUR, CST
       17.5
       2O.O
Figure 13c.  LPM time series of observed  and predicted NO hourly
            concentrations on Day 226  of 1976.
                             55

-------
     .18
        76226,  RAMS  109 AT 1300CST.  START 0600
     .10
   0.
   Q.

   CN~
   O
     .05
    0.00
                                              I I  I I  I

                                             OSS
                                             PRED L-
                                           ; \PRED L-i
               I	I | .._|__. L  I I  I I  t I  I I  I ..._!.... 1
54
                     iao     iis     15.0
                         HOUR, CST
17.5
20.0
Figure 13d.  LPM time series of observed and predicted N02  hourly
             concentrations on Day 226 of 1976.
        76226,  RAMS  109 AT 1300CST.  START 0600
    o
    o
         i  I I  \
                                              I '  ' '  '
                                             OSS
                                             PRED L-1.
                                             PRED L-i
              7J     10.0     1Z5     15.0-
                         HOUR, CST
                                      17.5
       20.0
Figure 13e.
      LPM time series of observed and predicted CO  hourly
      concentrations on Day 226 of 1976.
                            56

-------
 76237. RAMS  120 AT 1100CST. START 0600
    i W* P"^^^^"P""T^"HT"1 i~ T J1^^""^1 n* i ^^^^i|"'^p—]*~T~
    SO
    25
                             o   o
              25
  50
KM
75       100
Figure Ha.  LPM parcel trajectory for Day 237 of 1976.
                       57

-------
        76237, RAMS 120 AT 11OOCST. START 0600
   o.
   Q.
   ul
   O
   NJ
   O .2
         —— 08S
         	  PRED L-1
         	PRED L-3
     00
      54
7.5     10.0     12J     15.0
           HOUR, CST
17.5
20.0
Figure 14b.   LPM time series of observed  and  predicted 03 hourly
             concentrations on Day 237  of 1976.
        76237, RAMS 120 AT 11 OOCST. START 0600
  O
                                           i  i i  i i  i

                                            OSS
                                            PRED L-1
                                            PRED L-i
    0.00
              7.5
       10.0     1Z5     15.0
           HOUR, CST
17.5
20.0
Figure 14c.   LPM time series of observed and predicted NO hourly
             concentrations on Day 237  of 1976.
                             58

-------
       76237, RAMS  120 AT 1100CST. START 0600
                                         I I I  I I  I I
                                       —  OBS
                                       •—  PRED L-1
                                       —  PRED L-i
             7A
        10.0     1ZS     15.0
           HOUR, CST
17.3
20.0
Figure 14d.  LPM time series of observed and predicted N02 hourly
            concentrations on Day 237 of 1976.
       76237, RAMS 120 AT 1100CST. START 0600
    Cu
    a.
    o"
    o
                                           I '  ' '  '
                                          OSS
                                          PRED L-1
                                          PRED L-S
      OL-L-L,
      SJO
. i  ....  I .... r?-. .. I  .... i. ...
7.3    10.0     1Z3     15.0    17.5    2aO
           HOUR, CST
Figure  14e.  LPM time series of  observed and predicted CO hourly
            concentrations on Day 237 of 1976.
                            59

-------
 76275.  RAMS 102 AT 1400CST. START 0700
   1W* p**^^T~"'~ J  * I  i  TT^T""Hr"""^^f^HT" I  V ' ~ "(•••^^^^"•p*
    SO
                        I
                       30

                     KM
75       100
Figure 15a.  LPM parcel trajectory  for  Day  275  of 1976.
                       60

-------
      .4
        76275, RAMS 102 AT 1400CST.   START 0700
   a.
   a.
   ul  .2
   o
      .1
     0.0
              OBS
              PRED L-1
              PRED L-3
                                   /-A
&0     7A      10.0     12J     15.0
                   HOUR, CST
                                            17.S
                              20.0
Figure 15b.   LPM  time  series of observed and  predicted 03 hourly
             concentrations on Day 275 of 1976.
       76275, RAMS 102 AT  1400CST.  START 0700
   Q_
   O.
     .1
    oo
                               I  I I  I I  I
                                             I '  ' '
                                            OBS
                                            PREO L-1-
                                            PRED L-4
      SU)
10.0     12.5     15.0
    HOUR, CST
                                     17.5
                                                   2ao
Figure 15c.  LPM time series of observed and predicted NO hourly
            concentrations on Day  275  of 1976.
                            61

-------
        76275, RAMS 102 AT  1400CST.  START 0700
    Q.
    Q.
     *
    CM
    O
     0.0
                                        iiiiiiri  i

                                        —— 08S
                                        	  PRED L-1
                                        	PRED L-3
              7.5      10.0     1ZS     19.0
                         HOUR, CST
                       17.5
        20.0
Figure 15d.  LPM time series of observed and predicted  N02 hourly
             concentrations on Day 275 of 1976.
        76275,  RAMS  102 AT 1400CST.  START 0700
      10
   Q.
   Q.

   O*
   (J
                                       	  08S
                                       	  PRED L-1-
                                       	PRED L-3J
      0
      flU)
          i  i i  I i  i i  i I  i  i i  i I  i i  i- t  I i  i i  i I  i i
10.0     1ZS     15.0.
    HOUR, CST
17.5
20.0
Figure 15e.   LPM time series of observed  and  predicted CO hourly
             concentrations on Day 275  of 1976.
                             62

-------
                                  SECTION  6

                             URBAN AIRSHED MODEL

INTRODUCTION

     The  Urban  Airshed  Model  (UAM)  is  a  three-dimensional  (3-D), grid-type
photochemical air quality  simulation  model  (PAQSM) developed  by  Systems
Applications,  Inc.  (SAI)  of  San Rafael,  California.   The model structure
consists  of  a latticework  array of  cells,  whose total  volume  represents an
urban-scale domain and in which the physical  and  chemical  processes  responsible
for photochemical  smog ar.e mathematically simulated.   These include the advec-
tion of pollutant species through the  modeling domain,  the species'  entrainment
from aloft by a  growing  mixed layer,  diffusion of material from cell to cell,
the  injection  of primary  source emissions  into- the  modeled volume,  and the
chemical  transformations  of  reactive  species  into  intermediate  and secondary
products.  The horizontal dimensions of each  cell  are constant but cell heights
vary throughout  a model  simulation  as  mixed  layer  depth in the  UAM" changes
accordingly.

     The  area modeled  in the  St.  Louis  application  was  60 km wide and 80 km
long.   Individual  cells were  4  km on  a horizontal  side.   Vertically, cells
totaled 4 layers;  the  bottom 2  layers simulated  the  mixed  layer  and the
top  2  represented the region  immediately above  the  mixed layer.   The domain
of the UAM was centered just  west of downtown St.  Louis and included the entire
metropolitan area.  Figure 16 shows a  horizontal  cross-section of the domain as
applied to  the RAPS  monitoring network.   All  25 monitoring station locations
are given in relation to the grid  pattern.   The  outer  ring contains the bound-
ary  cells  for the  model.    Although  model   simulation does  not occur  in the
boundary cells, they  are an  important part  .of the  advective  transport of the
UAM.   Four  monitoring   stations,  122-125,   lie  outside  of  the  computational
domain  and  are  available for  determining boundary  concentrations.   Generally
boundary  concentrations  for  all   species  except 03   were  small,   near  hemi-
                                      63

-------
spheric background levels.   Ozone levels at the boundary however were sometimes
significant compared to those within the model  domain.   The remaining sites lie
within the  domain,  no 2 residing  in  the  same  cell.   The 3-D grid  model  is  a
sophisticated type of PAQSM and provides both spatially and temporally resolved
concentration predictions.   Thus, the UAM attempts to estimate the hour-average
observed concentration of  a pollutant  species  at each monitoring  site  within
the model domain.

     Details of  the  method  for performing a St.  Louis  simulation with the UAM
are described in Report I.   There  are 12 data  preprocessors in the UAM package
that must  be exercised before the simulation  model  itself.   They  access the
RAPS data  base   and  retrieve the  appropriate  parameters  required by  the  UAM.
Model  simulations begin at 0500 CST and continue through 1700 CST.

     The UAM and its preprocessors remain structurally unchanged from the model
simulations described in Report  I,  except for  one aspect.  An independent test
of  the  numerical  advection  algorithm was described there and  showed  that the
algorithm  in  the model  at  that time  produced excess numerical  diffusion,  a
highly undesirable  trait  for  a  grid  model.    A study  of  the  problem  revealed
that the numerical   advection  routine  in the  UAM  should be  replaced  with  a
sounder method that has the qualities of  (1)  minimizing the amount of numerical
diffusion produced  and  (2)  advecting areas  with  large  concentration gradients
such that the gradients  are not destroyed.  Several  advection algorithms were
identified  and   tested  as   potential  candidates  for  replacing the  existing
method.   The one chosen  was  a  fourth-order version of  the multi-dimensional
flux-corrected transport algorithms described  by Zalesak  (1979).   This  method
produced  far  less numerical  diffusion  and preserved peak  concentration areas
better during advection  than the algorithm  that  had resided  in  the model.  An
added advantage  of  the Zalesak  algorithm is that  the  horizontal  advection in
the model  could  be  performed  in one  step while  the original  routine  needed a
step for each  of the  x  and y  directions, another potential  source of  error.
The penalty encountered by  changing the numerical advection  routine is about a
12% increase  in the  execution time  of  the  simulation model  (on  EPA's  UNIVAC
1180 computer).   The potential for  increased accuracy within the model  with the
new algorithm far outweighs  the  small increase in overall computational  costs.
                                      64

-------
     Model  simulations with  the  revised  UAM were conducted for each of  the  20
test days listed in Table 1.  These test  days include the original  10  presented
in Report I but are now  reflecting  the model  update  described above.   Although
results for 03  predictions  on all  20 test  days  are  summarized, detailed model
results are  described for only  4  of these days.   Simulation results for the
pollutants  CO,  N02 and  03  on Julian days  142  and  207 of 1975,  and 195 and
275 of 1976 are emphasized.

RESULTS

     Among  the  20  days from  the  RAPS data  base selected for  model simulation
with  the UAM  were 4  where the  afternoon  measured  03 maximum  occurred  at
one of the monitoring sites located outside the model  domain.  All other days
showed 03 maxima occurring  at sites within the  domain.  The  UAM simulates air
quality within each grid cell.  Species concentration predictions at particular
RAPS monitoring site locations are interpolated from  predictions at the closest
cells by a  distance-weighted (1/r)  formula.   Detailed  simulation   results are
presented here  for 4  of the  20  days modeled,  and then  summary statistics are
given for 03 alone  over all  the days.                            »

     A UAM simulation provides a wealth  of information including   the spatial
and  temporal  concentration   variations of all modeled  species throughout the
grid.    It  is not  practical here  to  display  the time series  of predicted and
observed concentrations  for all  RAPS  sites,  even  for a few  pollutants on  a
single day  of  simulation.    A more  concise method  of displaying simulation
results is  used here.   It is based on an analysis of  residual concentrations,
observed minus predicted,  over  the 21   RAPS  sites  within  the model  domain
(Bencala and Seinfeld, 1979).  The particular analysis  techniques employed' here
include residual histograms, denoted as  "RS", which  show the  general  distribu-
tion  of  residuals  for  a complete  simulation, and  the  residual   vs.   observed
concentration plot, referred to as the "RO" plot, which  demonstrates  any trend
toward over- or underprediction by the model as  a function  of observed concen-
tration levels.  For a given species on one  simulation  day,  there were  21 sites
for each of 13 hours  of simulation  to account for, or 273 possible  data points.

                                      65

-------
     Since  this  report emphasizes  the various  models'  abilities  to  estimate
03 concentrations an  additional  technique is used for that  species to supple-
ment the statistical  analysis of the  paired  observed  and predicted data.  That
technique  is  the  contour plot  of observed and  predicted  03  concentrations
over the  model  domain  at the hour the maximum 03  level  is observed.   These
plots reveal the  spatial  orientation  of the peak  concentration  areas.  Visual
comparisons can be  made that are not  dependent  on specific  paired data.  This
will help  determine  if the DAM is predicting  the  maximum concentration levels
well but at incorrect areas  of the model  domain.  It would  not  be possible to
determine this situation from the residual concentration  data alone.

     The first  simulation day  discussed  is Day  142  of  1975.    The  synoptic
meteorological  conditions on this day  were described  in  Section  4.  Background
03 concentrations for Day 142 were 0.06 ppm.

     Figure 17a  shows the RS  for  CO  on  this  day.   The  ideal  RS distribution
would have a peak at 0.0 ppm and symmetric tails on either side sharply falling
off to 0.0 frequency.  The figure displays a  shape like this  except for a bulge
in  the  peak toward overprediction.   The  RO plot  in  Figure  17b  shows a trend
toward underprediction, especially evident in the 3 points corresponding to the
highest  observed  concentrations.   Report I discussed the  CO  underprediction
problem  for high observed concentrations.   The  clustering  of  points  in this
plot at  slightly negative values  for  very  low observed  concentrations helps
explain  the bulge  in the RS  on  the  overpredictive side.   Both  figures  are
required  to provide  a total  picture  of the  distribution  of  the residuals.

     The RS for  N02 is presented in Figure  18a.   While  the  distribution peaks
at  the  0.0 residual  it is slightly skewed   toward overpredictions.   In the RO
plot in  Figure  18b  the majority of residuals  fall to  the overpredictive side,
but mostly  for smaller observed concentrations.  At-higher concentrations there
tend  to  be more underpredictions.    In  any  case  a definite  trend  would be
difficult  to  discern here for  most  of the  N02 residuals fall  quite  close to
0.0.
                                      66

-------
     Figure  19a  displays  the  RS for  03 on  Day  142.   The distribution  of
residuals here is clearly skewed  toward overprediction, with  a  peak  at the  0.0
residual.  The RO plot in Figure 19b  shows most of  the overpredicted  concentra-
tions lie in  the  range of 0.050 to 0.125 ppm observed 03.  At  higher  observed
concentrations no clear trend appears.
                                                                  •
     The  peak  hour-average 03  level  observed  was 0.195  ppm at  site 101  at
noon.   The  model-simulated concentration  for the  same  time  and location  was
0.116 ppm,  a 40.5%  underprediction.   Figure  20  presents the  contours of  03
concentration for the model-predicted  and  observed  cases  for the hour beginning
at noon  CST.   In the  model-predicted  case the contours have been objectively
drawn based  on  the 300  regularly  spaced  grid values  for  the lowest  layer  of
cells in  the UAM.   For the  observed  case the contours  are based  on the  25
irregularly  spaced monitoring  locations,  4 of  which  lie outside of the model
domain.    Here the objective algorithm  responsible  for generating the  contours
works best toward the  center  of the domain where  the  density of  sites, repre-
sented  in the figure by  small  crosses,  is greatest.   The credibility of  the
contours  near the edges  is not as high,  especially  in terms  of the extent  of
the  area  circumscribed  by  a  given contour.    In  the figure for  Day  142  the
maximum concentration  areas  appear near the  center  of the domain with magni-
tudes close  to  each  other.   In  the  model-predicted  case  the  highest valued
contour  is  displaced  slightly  northwest  of  the  observed maximum  area and a
relative "valley" has developed in the immediate area  of the  observed  maximum,
allowing the large underprediction to  occur.

     On Day  207-75  a large high-pressure  center was  located northeast of  St.
Louis with  light  winds both at  the  surface and aloft.   Measured wind speeds
were  generally  less than  1.0 m  s'1  near the surface  and directions, were
variable.  Sunny  skies  prevailed  through  the  duration of  the modeling period,
although  haze  persisted during  the  day.   Background 03  concentrations  were
approximately 0.07 ppm.

     Figure  21a shows  the RS for  CO  on  Day  207.   The distribution is nearly
symmetric about  the  0.0  residual  with  a  few  outliers on the  underpredictive
side.   The  RO  plot  in  Figure  21b  reveals  these outliers  as  the residuals

                                      67

-------
corresponding to the  highest  observed concentrations.   Most CO concentrations
on this day  were  less than 1  ppm.  Nitrogen dioxide presents a broader histo-
gram  than  CO in  Figure 22a  that is  symmetric  and centered  slightly  on  the
overpredictive siae.   This  bias  is  probably caused by the great many overpre-
dictions at  observed  N02  concentrations  less  than 0.02  ppm  seen  in Figure
22b.   Aside  from this  phenomenon there  is no clear trend  apparent in the RO
plot.   The  RS  for ozone  in  Figure  23a  presents  a histogram  with  the  rough
outline of a normal  distribution but contains,periodic  gaps throughout.    The
majority of  the  predictions are  less  than the observed concentrations and  the
RO plot in Figure 23b demonstrates that a definite  bias exists  toward underpre-
dicting 03 as observed concentrations  increase.

     The maximum hour-average 03  measurement  for  Day 207 occurred at 1400  CST
at site 113, northwest 'of  downtown St. Louis, at 0.185 ppm.  The corresponding
simulated concentration for this  location was 0.141 ppm, a  23.8%  underpredic-
tion.   The  contours of observed  03 concentrations  at  this  hour in Figure  24b
reveal a steep gradient aligned  from the  northwest  to the  southeast through  the
city  of  St.  Louis.   The  predicted  concentration  field in  Figure  24a shows  a
similar  gradient  but  displaced  slightly  north and west.   The alignment  and
magnitude of the  contours  are, in  fact, very  much alike in the  2 cases.
Possible errors  in  the model's  treatment of the wind field, or the generation
of the wind  field  itself,  may have  contributed  to  the  displacement between  the
peak  areas of  predicted and observed concentrations  and  allowed  the  apparent
underpredictions by the model  at the monitoring  sites.

     The synoptic  meteorological  situation for Day 195  of  1976   contained an
upper  air ridge  over the center of the country with a warm  front  northeast of
St.  Louis.   Winds  near the  surface  were from the  southeast  during the  day
averaging just  over 2.0 m  s~l.   Skies were mostly clear with  occasional  high
level cloudiness.  Background 03 concentrations  were nearly  0.08 ppm.

     CO concentrations  were low on this  day, mostly  less  than  1   ppm.  The RS
for  CO for Day  195  is presented  in  Figure 25a.  "It shows  a  narrow  distribution
of  residuals peaking slightly on the overpredictive  side.   In the RO plot in
Figure 25b this  peak  has been translated  to the many very low  observed  concen-

                                      68

-------
trations often  near  the detectable limits  of the monitoring instruments.   At
higher  concentrations  the model  tends to  underpredict,  especially  at the  2
distinct data  points representing CO  hot spots.   The N02  residual  distribu-
tion  shown  in  Figure   26a  displays  a  normal  symmetric  shape  about  the  0.0
residual.   Most of  the overpredictions occur  at low observed  N02  concentra-
tions and a  trend  toward more severe underprediction with  increasing observed
N02  is  apparent from the RO  plot in Figure  26b.   The RS for ozone  in Figure
27a  is  skewed  toward underpredictions.   Below an- observed 03 level  of  0.1  ppm
the RO plot in Figure 27b shows little  bias in the residuals. Above  this level
there is  a  clear  trend toward underpredicting  at  higher observed  concentra-
tions.   This  explains  the skewness  in the  histogram as  resulting from  the
residuals of observations mostly greater than  0.1 ppm.

     The highest  hour-average 03  measurement on Day  195 was at  site  114  at
1500 CST, north of the  city  of St. Louis, at 0.223 ppm.  The UAM  estimated an
03 concentration of  0.160 ppm at  the  same time and place,  an underprediction
of 28.3*.   The contours of 03  predictions  and observations for this hour  are
shown in Figure 28.   Agreement in the  alignment and location of  the contours
between the  2  cases  is  good,  although  the magnitude of  the maximum predicted
contour is considerably  less than  that  of  the observed.   The advection  problem
discussed in Report  I for Day  195  appears  to  have been corrected here with  the
substitution of the new  advection  algorithm.

     Day 275 of 1976 is  typical of an  extreme stagnation  episode in  St. Louis.
The synoptic meteorological  conditions  on  this day were described in  Section 4.
Background concentrations of 03 were  about 0.06 ppm.

     The histogram  of residuals for CO  is  shown in Figure  29a for Day 275.   The
distribution peaks  slightly  on  the  overpredictive  side  of  the 0.0  residual
although  there is  a pronounced  skewness toward underpredictions,   including
several  extreme cases.   The RO plot  in Figure  29b  displays the  trend  toward
underprediction of  CO at higher observed concentrations.   The extreme stability
of the  atmosphere  on the  morning of  Day  275 is reflected  in  the  highest  CO
concentrations, on the  order  of 10 ppm,  over twice  the level  of  the observed
peak on most of the other modeled  days.
                                     69

-------
     The RS  for M02  in Figure  30a is  normally distributed  about the  0.0
residual  and  the RO  plot  in  Figure  30b  shows  little  bias with  increasing
concentrations for  observed N02  levels  of  0.125 ppm  and  less.  Beyond  this
point  there is  evidence of consistent underprediction.  The  63 residual
analysis  in Figures 31a and b  indicates  good model  performance  in the overall
predictions for the day.  The  residual  histogram has a broad normal distribu-
tion about the 0.0  residual  and the  RO  plot demonstrates no clear trends toward
over- or  underprediction.

     The peak 03 measurement on  Day 275  was 0.244  ppm at site  102,  near
downtown  St.  Louis,  at 1400  CST.   The model prediction  for  this  time  and
location  was 0.220  ppm,  a~ 9.8%  underprediction.  The contour plots of predicted
and observed  03  concentration  for  this  hour  are  shown in Figures  32a  and b.
In both  cases  the  highest concentrations appear  near  the  center  of the model
domain and are of comparable magnitude.  Several pockets of reduced 03 concen-
trations  appear in  the pattern  of predicted contours that break up an otherwise
consistent pattern with the observations.   These appear  to be in the vicinity
of large NOX  emissions  sources that have an  exaggerated  scavenging effect on
nearby 03  predictions.    The  observations  do  not  substantiate  this  effect.
The rather simplistic  treatment of  point  source emissions by the UAM fs probab-
ly a major factor in this  problem.

     Table 5  summarizes the predicted and  observed 03 maxima  and  an elemen-
tary statistical  analysis  of  the  UAM  results  for  all  20 modeled  days.   The
"specific"  model predictions correspond  to  the same location and time as the
observed   maximum,   and  the  "independent"  predictions  represent the  maximum
hour-average  03  generated by  the model  in  any cell of the  lowest  grid layer
at  any time during the simulation.   The independent  maximum is  the greatest
surface  value the  model  can  produce  within  its  domain.   The  specific  and
independent predictions, as seen in  Table 5,  typically do not coincide.  This
indicates  that  the  maximum  03  value produced  by  the model  did  not correspond
either in  space  or time,  or both,  to  the maximum value  observed  in the atmo-
sphere.   Statistics on  the residual concentration,  AC,  as  shown in  Table 5,
were computed  for both  the  specific and independent predictions.   The average
residual  and  standard deviation  are also presented  for  the analysis.   Values

                                     70

-------
   TABLE 5.  A SUMMARY OF PREDICTED AND OBSERVED  03 MAXIMA FOR THE UAM

Julian
date3
142 (1975)
178
182
183
184
207
209
221
230
231
251
159 (1976)
160
195
211
212
225
226
237
275
Hour
(CST)
12
14
11
10
13
14
11
15
13
14
12
14
16
15
15
12
13
13
11
14
Station
101
112
121b
119b
118
113
118
121
121
121
121b
114b
115
114
120
108
117
109
120
102
Observed
at 4-meters
(ppm)
0.195
0.202
0.142
0.171
0.184
0.185
0.209
0.166
0.193
0.233
0.179
0.172
0.221
0.223
0.155
0.170
0.166
0.225
0.176
0.244
UAM
Specific
(ppm)
0.116
0.156
0.083
0.154
0.132
0.141
0.128
0.118
0.095
0.133
0.146
0.142
0.121
0.160
0.143
0.128
- 0.050
0.077
0.119
0.220
Predicted
Independent
(ppm)
0.238
0.243
0.166
0.209
0.234
0.252
0.195
0.149
0.214
0.205
0.178
0.312
0.190
0.174
0.169
0.155
0.073
0.124
0.203
0.246
aTwenty days selected from 1975 and 1976.
bOveral1  maximum at 122,123,124,  or 125;  outside UAM domain.
     For the data displayed above:
       AC = Obs -  Specific
       IT = 0.062
 s.d.UO = 0.035
     "ITr - 0.062
 Obs - Independent
-0.006
 0.053
 0.041
                                 71

-------
for the  average residual  for the  specific  and independent  cases,  0.062 and
-0.006 ppm  respectively,  differ  noticeably.   The independent predictions  show
more promise  toward  maximum 03 simulation.   The average absolute residual  is

-------
  TABLE 6.  A SUMMARY OF LEAST SQUARES REGRESSION STATISTICS FOR
            UAM PREDICTED AND OBSERVED 0a


Julian
dateb
•
142-75
178-75
182-75
183-75
184-75
207-75
209-75
221-75
230-75
231-75
251-75
159-76
160-76
195-76
211-76
212-76
225-76
226-76
237-76
275-76
URBAN
Correlation
coefficient

0.805
0.864
0.875
0.864
0.726
'0.927
0.759
0.795
0.819
0.814
0.821
0.859
0.937
0.947
0.833
0.912
0.753
0.796
0.687
0.844
AIRSHED MODEL:

Slope

0.887
0.867
0.610
0.614
0.470
0.654
0.570
0.516
0.729
0.645
0.618
0.862
0.676'
0.720
0.773
0.632
0.310
0.426
0.432
0.896
03

intercept
(ppm)
-0.0057
0.0183
0.0263
0.0284
0.0436
0.0373
0.0347
0.0433
0.0059
0.0176
0.0232
0.0021
0.0464
0.0296
0.0010
0.0244
0.0504
0.0574
0.0581
-0.0004

Sum of
squared error
( ppm2 )
. 0.186
0.186
0.142
0.193
0.317
0.173
0.192
0.268
0.095
0.152
0.150
0.158
0.247
0.112
0.104
0.189
0.247
0.445
0.383
0.340

Includes observations and predictions for all  times and locations.
bTwenty days selected from 1975 and 1976.
                               73

-------
CONCLUSIONS

     The potential  use of a  grid-type  PAQSM,  such as the UAM, clearly is great,
although model  complexity often makes  the  solution of problems which arise more
difficult.    Trouble-shooting  and test application exercises  have taken place
over the past  3 years  at EPA after  the  adaptation of  the UAM  to  St. Louis
by SAI.   The constituent components  of the model have already been investigated
and upgraded where necessary.   The numerical ^advection  routine  was recently
replaced by one with less inherent pseudo-diffusion.

     The model evaluation discussed here  has pointed toward some  of  the
strengths  and weaknesses  of the  UAM.   In studying the  results  for the 3
pollutant species  focused  on  in  this report it is clear  that while the model
typically produces  a  normal  distribution  of residuals in a simulation, those
estimates  corresponding  to   the  highest  observed  concentrations   are often
underpredicted.   This  is especially  true of primary  species such  as  CO  and
M02, but  has  also  been indicated  in the case of  03 as well.    It is these
higher observed concentrations that  are of greatest interest from a  regulatory
perspective.   For  the case  of  those  species  emitted  from and  affected by
emissions from point  sources,  their simulated  performance results may  improve
with a more physically realistic  treatment of point emission  sources.

     It is quite interesting to note  that the average of  independent residuals
for 03  maxima  over the  20  test  days  indicates an almost exact  match  between
predictions  and  observations.   Studying the  contour maps  on  the  test days
reveals  that the  model usually  produces 03  levels  as  high or  higher than
those observed and the contours  of predicted  and  observed often have  similar
shapes.    However,  the variance  in  these residuals  of  the  20   03  maxima is
quite large and this  is caused,  in  part,  by  the different spatial orientations
of  the  predicted  and observed contours.   Indeed,  •even the paired predictions
and observations typically show poor spatial  correlation.   This is particularly
true for 63.   The spatial  alignment of the  concentration fields  is  principal-
ly governed through transport in the  model.  The' treatment  of winds  in  the UAM
is  comparable  to  most  schemes available  today  for  interpolating a wind field
across  a  grid  (Schere, 1981).   However,  even in  a relatively  simple flow

                                      74

-------
situation such  as St.  Louis,  problems  can  arise.    An error  of only  a few
degrees  in  wind direction  can  lead to  gross  errors in specific  air quality
predictions.  It will  be a major responsibility of any UAM user to substantiate
that the wind processor is  providing a credible simulation of wind flow  in the
application  area.  For complex topographic situations or land-sea interactions
within a model domain, a wind field analysis unique  to  that application may be
required.

     In  summary, the  UAM is a  powerful tool  available for use by  the air
quality analyst already experienced in working with complex simulation models.
Model   results  should  be  carefully  studied  in each  application and  may not
necessarily  be used in  an  absolute sense in  any  given simulation.  An area of
future study that may  be pursued is model  sensitivity to variations in  selected
parameters,  such  as  initial  and boundary conditions,  emissions,  wind fields
and solar radiation.
                                     75

-------
                                    122
                            121
                                       114
125
                     121
                                 l.i
                            112
102
                                    07
                                    06
                                    111
 10
                                       105
    08
    05
    04
     10
                                                   109
                                                           10
117
                                                                      125
  I
        ->  x   Jx = Jv = 4 km
                                       118
                                            12
   Figure  16.   Schematic drawing  of horizontal section of UAM domain
                showing locations  of RAPS monitoring  sites.
                                   76

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL T'MES AND LOCATIONS
                             OAY 142-75


2
Q.
Q.
^~*
_l

CO
LJ
or




.3UVJ+UU 1
	 I
4CO+001 -
300+001 t

.200+001
.100+001

t
K

BH""^^
- 100+001
-200+001
-.300+001
-.400+001
- 500+001

Figure 17b.
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  1 74   DATA POSSIBLE  =  273

             Residual  vs. observed plot for CO from  the UAM
             simulation  results for Day 142-75.
                               77

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                               DAY  1*2 -75

         750-001  i	1	1	1	1	1	1	1	•	1
     0.
     0.
600-001


450-001


.300-001


.150-001 J45"


.000
     |

     Q

     00

     j£  -150-001


        -.300-001


        -450-001


        -600-001


        -.750-001
        •*»*
        •*»*»••*••
        ;««.
                    5   10   15   20   25   30   35
                                  FREQUENCY

                    NITROGEN DIOXIDE
                                       40   45   50
                 DATA AVAILABLE =  195    DATA POSSIBLE
                                                   273
Figure 18a.  Residual histogram for N02 from the UAM  simulation
              results for  Day 142-75.
      2
      Q.
      a
      3
      9

      LJ
      or
 .750-001


 .600-001


 .450-001


 .300-001


 .150-001





 -.150-001


 -300-001


 - 450-001


 -.600-001


 -.750-001
                    RESIDUAL VS OBSERVED CONCENTRATION

                                 DAY 142-75
                                                            < i. i>  a
                                                            < 2. 2>  0
                                                            < 3. 3>  *
                       100-001  .200-001  .300-001  400-001  500-001
                        OBSERVED CONCENTRATION  (PPM)

                     NITROGEN DIOXIDE

               DATA AVAILABLE =  195    DATA POSSIBLE  =  273


Figure 18b.   Residual  vs.  observed plot  for N02  from the  UAM

               simulation results  for Day  142-75.
                                 78

-------
     RESIDUAL HIS'OGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 142-75
         25OOOO
        - 100 -i-OOO j*

        - 150+OOOf-

        -.200^0001-

        -.250+000 I—
                   10  20  30
40   50   60   70
  FREQUENCY
80   90  100
                   OZONE
                DATA AVAILABLE =  21
         DATA POSSIBLE  =  273
Figure 19a.   Residual  histogram  for  03  from the UAM simulation
              results for Day 142-75.
                   RESIDUAL VS OBSERVED CONCENTRATION
                               DAY  142-75
^v
0_
Q^
SIDUAL
LU
o;

200+000
150+000
.100+000
.500-001
.000
- 500-001
- 100+000
- 150+000
-.200+000
- 250+000
000
.

' . jfrA** . . f '
9j^ *&&^E?y9&&<$ rtftP ° * oo ' '
9 Q




LEGEND
FREO 5TM
< i, i> n
< 2. 2> 0
< 3. 3> *
< *,99> *

500-001 .100+000 .150+000 .200+000 250+000
                      OBSERVED CONCENTRATION  (PPM)
                    OZONE
              DATA AVAILABLE =  214   DATA POSSIBLE  =  273

Figure 19b.   Residual  vs. observed  plot for 03 from  the UAM
              simulation results  for Day 142-75.
                               79

-------
   ST. LOUIS   HOT 22.197S    I DDT
   ^eoicTEQ co«et
-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 207-75
Q.
a
~
SIDUAL
LU
or



~I_iU *-UU "
400"**00 I
300+001
.200+001
.100+001
000
-.-100+001
-.200+001
-.300+001
-.400+001
-.500+001
i ii j
'
I
5F'"" " i i i i i i
•
•

•
                   50  100  150  200  250 300 350  400  450  500
                                 FREQUENCY
                   CARBON  MONOXIDE
                DATA AVAILABLE  =  237   DATA POSSIBLE  =  273

Figure 21a.   Residual histogram for CO from the UAM simulation
              results  for Day  207-75.
                   RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 207-75

a
o_
*"
RESIDUAL


.400+001
300+001
.200+001
.100+001
.000
-.100+001
-.200+001
-.300+001
-.400+001
-.500+001

a

-l^ref3 ,,,,,'
pF^ .ii...

•



L£CENO
FREO STM
< i, i> a
< 2. 2> O
< 3. 3> *

             .000      100+001  .200+001  .300+001  400+001   500+001
                      OBSERVED CONCENTRATION  (PPM)
                   CARBON MONOXIDE
              DATA AVAILABLE =  237   DATA POSSIBLE   =  273

Figure 21b.   Residual  vs. observed  plot for CO from the UAM
              simulation results  for Day 207-75.

                               81

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 207-75

Q.
a
"— '
j
3
co
LLt
CC




3UU— UUI
400-001
3CO-001

.200-001
100-001
AfVt
uuu
- 100-001
- 200-001
-.300-001
-.400-001
-.50Q-CQ1
1
.

-»*
J»
*****
HiHfims** , ,
tit « it:: m::i:itti::tht:i.»
»*•»*' li»*»»*»*»*«»-»*-*»
»*•<•*' n» •»*£**•»»**»»»
"••*!*?' '*{' '•**»•»»*»
|!"
*
*
, , i
                      10   15   20  25  JO  35
                                FREQUENCY
                  NITROGEN DIOXIDE
                                                    50
               DATA AVAILABLE =  234   DATA POSSIBLE  =  273

Figure 22a.  Residual  histogram for N02 from the UAM  simulation
             results  for Day 207-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 2G7-75

a.
Q.
n
on
UJ
CC
.400-001
.300-001
.200-001
.100-001
000
- 100-001
-.200-001
-.300-001
-.400-001
-500-001
.000
.
•
a a
_ a
a a o ° a
^gP^-jopa^a^
O
a
a


LEGEND
FREO STM
< i. i> a
< 2. 2> 0
< 2, 3> *
< +,99> <»
.150-001 .300-001 .450-001 .600-001 750-001
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
              DATA AVAILABLE =  234   DATA POSSIBLE -  273

Figure 22b.   Residual  vs.  observed plot for N02 from  the UAM
              simulation results for Day 207-75.
                               82

-------
      RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                              DAY 207-75

Q.
CL
_i
•J

LU
or





1 UU^UWW f
.800-001 .
.600-001 .
.400-001 -
.200-001 -
QOfl
WWW
-200-001 .
-.400-001 •
-.600-001 -
-.BOOrOOl •
- 100+000 •
a
[
t
|m«*
fj jgsjHim*"

: ': i: ': \\ til'****

•
•
5 10 15 20 25 JO iS 40 45 5
FREQUENCY
OZONE
)ATA AVAILABLE = 210 DATA POSSIBLE =
Figure 23a.   Residual histogram for 63 from the  (JAM simulation
              results for Day  207-75.
                  RESIDUAL VS OBSERVED CONCENTRATION

                              DAY 207-75

RESIDUAL (PPM



.800-001
.600-001
.400-001
.200-001
000
-.200-001
-.400-001
-.600-001
-.300-001
- 100+000
Q
Q
^c&i^^yF °i °
l"!0**^ * ° .
a


                                                           LEGEND

                                                          FREQ SYM
                                                        < i, i>  a
                                                        < 2. 2>  O
                                                        < 3, 3>  *
                                                        < *,99>  »
             000      500-001  .100+000  .150+000  200+000  250+000
                      OBSERVED CONCENTRATION  (PPM)

                   OZONE

              DATA AVAILABLE =  210   DATA POSSIBLE  =  273


Figure 23b.   Residual vs.  observed plot for 03  from the UAM

              simulation results  for Day 207-75.
                               83

-------
   ST. LOUIS  JULY  28.1975   I OUT 207)
   'ReOICTEO C8NCENTIBTIQNS flf 03 UN PPfl I
   HOUR  1400-1500
ST. LOUIS  JULY 28.I97S    I OUT 2071
39SEMVCO COMCENTMaTlONS Of 03  'IN
HOUR  1400-1500
            (a)
              (b)
Figure 24.   Contours  of  (a)  predicted and  (b)  observed fields
              of  03 at  hour  of observed maximum  on Day  207-75.
                                  84

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                             DAY  195-75

        500-i-OOl
    CL
    CL
    g
 400+001  •


 .300+001  •
        •

 .200+001  •


 .100+001  L


 000


-.100+001


-200+001


-.300+001


-.400+001


-.500+001
               0   10   20   20   40   50   60  70  30  90  100
                                FREQUENCY
                  CARBON MONOXIDE

               DATA AVAILABLE =  240    DATA POSSIBLE  =
                                                 273
Figure 25a.   Residual histogram for CO from  the DAM simulation
              results for Day  195-76.
    Q.
    CL
    Q

    CO
                 RESIDUAL VS OBSERVED CONCENTRATION

                              DAY 195-76
.3UU^VW 1
.400+001
300+001
.200+001
.100+001
000
-.100+001
-.200+001
-.300+001
-.400+001
-.500+001
000
a
a
.*£ ,,,,,,,'
^*?T i i , i i , ,

•
•
100+001 .200+001 .300+001 400+001 .500+
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE

             DATA AVAILABLE =  240   DATA POSSIBLE  =  273

Figure 25b.   Residual vs.  observed plot for CO from the UAM
              simulation  results for Day 195-76.
                              85

-------
    RESIDUAL HISTOGRAM  DETERMINED  OVER ALL TIMES AND LOCATIONS
                            DAY 195-76
3UU-UU I r
.400-001
^ .300-001
a
^ .200-00!
-! .100-001
Q 000
OH
£J -.ioo-oor
-.200-001
-.300-001
-.400-001
-500-001
»

.
|.
'• |jf*j£t««««***
It!: :l 1||ii||IS|lItl1}|«*«
If?: :: lltf ||*«4«M***MiM4M*iMMSMtWt* '
|j:;t;~.
l.»
-*••«•

                     10   15
40   45   50
                            20  25  20  35
                               FREQUENCY
                 NITROGEN DIOXIDE
              DATA AVAILABLE =   253   DATA POSSIBLE
          273
Figure 26a.  Residual  histogram for N02 from  the UAM simulation
             results  for Day 195-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                            DAY 195-75
2
a.
p^
D
9
Ld
a
.3V.XI— UUI
400-001
.300-001
.200-001
.100-001
-.100-001
- 200-001
-.300-001
-.400-001
-.500-001
000
a
a
i o °a B a a° ;
3 nor a a
Q a a mm
_ fl ""BW rt^^p^wnwn hr^r^

f;.-*,
L£CENO
FREQ SYM
< 1, t> O
< 2. 2> 0
.150-001 .300-001 .450-001 600-001 .750-001
                    OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE =  253   DATA POSSIBLE  =  273
Figure 26b.   Residual  vs.  observed plot  for  N02 from the UAM
              simulation results for Day  195-76.
                              86

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 195-76
       .750-001 ,	1	1	1	1	1	1	1	1	,
                                                     I
    s
    o_


    g
    Q
    LU
    Qi
.600-001
.450-001
300-001
.150-001
nnn
WIAJ
-.150-001
- 300-001
-.450-001
-.600-001
- 750-001
r
•i 4

i:
.


Stt
ia
*• ' '5f**2Tf * ****** I i ~ i i
• VvS
*••
•
-
•
10   15   20  25  30  35
          FREQUENCY
                  OZONE

               DATA AVAILABLE -
                                             40   45   50
          184   DATA POSSIBLE  »  273
Figure 27a.   Residual histogram for 03 from the  UAM simulation
              results for Day  195-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 195-76
    Q
    uo
    UJ
    Qi
       - 750-001
            000     .500-001  .100+000  ,150+000  200+000  250+000
                     OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  184   DATA POSSIBLE  =  273

Figure 27b.   Residual vs. observed plot for 03  from  the UAM
              simulation results for Day 195-76.
                              87

-------
  ST. LOUIS  JULY 13.1 ait    I OUT 1951
  WeQICTEO CBNCENTHHTIONS Of 83  I IN CPB I
  HOUR  t500-1500
ST. LOUIS  JUur 13.1975    I OUT 1951
ossemo CONCCNTRSTIONS or as  it*
HOUR  1SOO-1600
            (a)
              (b)
Figure 28.   Contours  of (a)  predicted and (b)  observed fields
              of 63 at  hour of observed maximum  on Day 195-76.
                                  88

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 275-75
        750+001 |—
                  10
                          JO
                                             80  90  100
               40   50   60   70
                 FREQUENCY
   CARBON MONOXIDE
DATA AVAILABLE =  250    DATA POSSIBLE
                                                        273
Figure 29a.   Residual histogram for CO from  the UAM simulation
              results for  Day  275-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 275-76

J
a^
_,
3
Q
l/>
LJ
rw




. / ^\j^vu '
SOO+001
450+001

300+001
.150+001
000
- 150+001
- 300+001
- 450+001
- 600+001
- 750+001
Q |
a « I
.
a°a B
%-n*d m
H, 3 a H
^-AnS^pg^ a . |
IBKjtgpPaH . . . 1 i 1 — (
P^F^^
•
'

1 i ' ' ' 	 i i i i
                                                         LEGEND
                                                        FREO SYM
                                                       < i. i>  a
                                                       < 2. 2>  O
                                                       < <99>  »
            000      200+001  .400+001  .600+001   300+001   100+002
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON  MONOXIDE
             DATA AVAILABLE =  250    DATA POSSIBLE  =  273

Figure 29b.   Residual vs. observed plot  for  CO  from the UAM
              simulation  results for Day  275-76.
                              89

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY  275-75
        250-OOC i	—	1	
    3
    9
    (S)
        200+000 I-

        150*000 r

        .100+000 j-fM

        .500-001 |-?s,
 .000

-500-001

- lOC+OCOJ-'

-.150+000 I-
        I
-200+0001-
        0    5   10   15   20   25  • 30   35
                          FREQUENCY
           NITROGEN DIOXIDE
                                               0
< 2. 2> O
< 3, 3> »
< *,99> »

.500-001 .100+000 .150+000 200+000 250+000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN  DIOXIDE
             QATA AVAILABLE =  228    DATA POSSIBLE  =  273

Figure  30b.   Residual  vs.  observed  plot  for N02  from the UAM
              simulation results  for Day  275-76.
                               90

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL ~'MES AND LOCATIONS
                            DAY 275-75

s*
0.
•— '
_J
CO
Ul
DC
^•\J "UU J
200+000
.150+000
.100+000
.500-001
000
-500-001


j^
JL
=-
s»
- 100+000 fU





-.150+000

-.200+000

-.250+000
f
.
^

	
                 '0  20
                         30
                             40  50  60  7Q
                               FREQUENCY
                                            ao   90  iQO
                 OZONE
              DATA AVAILABLE =  248
DATA POSSIBLE  =  273
Figure 31a.  Residual  histogram for 03 from the UAM  simulation
             results  for Day 275-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                            DAY 275-76

/•^
2
Q.
a
*"*

g
Ul
a:






200+000

.150+000

.100+000

.500-001
f\f\f\
.UwU
-.500-001
- 100+000

- 150+000
-.200+000
- 250+000
,

-
a
n

DQO u a
0 °g g JJD CD ° .
B| Q Jp^|l |3 —»
ffi^^Sg^iff^ ^^nj^"..^. Q
^ CJBU uuf oO^n0 ^"y "I
iii ta o
tf
Q
a
•








UCCCNO
FREO STV
< i. i> a
< 2. 2> 0
< 3, 3> *
< 4,99> »


000 .500-001 .100+000 .150+000 200+000 250+000
                    OBSERVED CONCENTRATION  (PPM)
                  OZONE
            DATA AVAILABLE =  248   DATA POSSIBLE  =  273

Figure 31b.  Residual  vs.  observed plot for 03 from the  UAM
             simulation  results for Day 275-76.
                             91

-------
   ST. LOUIS  OCT.  l.ir?8    I OUT 273)
   PREDICTED CBNCENT»ST1BNS OF 03  UN Pffl I
   HOUR  t400-1500
ST. LOUIS  OCT.  1.1976   IOBT 27SI
OBSERVED CONCENTRATIONS OF 03  ItN Pf8J
HOUR  MOO-1 SCO
            (a)
             (b)
Figure 32.   Contours  of (a)  predicted  and (b)  observed fields
              of  03 at  hour of observed  maximum  on Day 275-76.
                                  92

-------
                               REFERENCES

Bencala, K.  E. and  J. H.  Seinfeld, 1979:  An air  quality model  performance
     assessment package.  Atmos. Environ., 13,  1181-1185.

Littman, F. E., 1979: Regional  Air Pollution Study—Emission inventory  summari-
     zation.    Report   No.  EPA-600/4-79-004,  U.  S.  Environmental   Protection
     Agency, Research Triangle Park,  NC  27711.

Lurmann, F.,  D.  Godden, A. C.  Lloyd, and R. A. Nordsieck,  1979: A  Lagrangian
     Photochemical  Air Quality Simulation Model. Vol.  I-Model Formulation,  Vol.
     II-Users Manual.  EPA-600/8-79-015a,b (available NTIS).

Lurmann, F.,  1980:  Modification and  Analysis  of the Lagrangian  Photochemical
     Air Quality Simulation  Model  for St.  Louis..   Environmental Research  and
     Technology, Inc.   Document No.  P-A095.    Westlake  Village,  CA.    25pp.

Lurmann, F., 1981:  Incorporation of Lateral  Diffusion in  the Lagrangian  Photo-
     chemical Air Quality  Simulation  Model.  Environmental  Research and Tech-
     nology, Inc.   Document No. P-A748.   Westlake Village, CA.   32pp.

MacCracken, M. C.,  D. J. Wuebbles,  J.  J.  Walton, W. H. Duewer,  and K. E.  Grant,
     1978:  The  Livermore Regional  Air Quality  Model, I.   Concept and  develop-
     ment.  J. Appl.  Met.,  17, 254-272.

Schere, K.  L.  1981: Air  quality model   response  to  objectively  analyzed  wind
     fields.   In Proceedings  of the  Fifth  Symposium on Turbulence,  Diffusion,
     and Air Pollution, American  Meteorological  Society,  Atlanta,  GA,  March
     9-13,  1981.

Schiermeier, F. A.,  1978:  Air  monitoring milestones:  RAPS  field measurements
     are in.  Environ.  Sci. Techno!.,  12, 644-651.

                                      93

-------
Shreffler, J.  H.  and  R.  8.  Evans,  1982:  The  surface ozone  record from  the
     Regional  Air Pollution Study, 1975-1976.  Atmos.  Environ.,  16,  1311-1321.

Shreffler, J.  H. and K. L. Schere, 1982:  Evaluation of Four  Urban-Scale  Photo-
     chemical  Air Quality  Simulation Models.   EPA Report, U.S.  Environmental
     Protection Agency, Research Triangle Park.  NC  (in  press).

Zalesak,  S. T., 1979:  Fully  multi-dimensional- flux-corrected transport  algo-
     rithms for fluids.  J.  Comp. Phys.,  31, 335-362.
                                      94

-------
                                 APPENDIX A

                  PBM - TIME SERIES FOR REMAINING TEST DAYS
                                                  •
     This appendix  contains the  time  series plots  for the pollutant  species
CO, N02  and 03  from  the PBM  test results for those  days that were not  spe-
cifically discussed within  the body of this report.   The  plots presented  here
are from  test days  182-75, 183-75,  184-75, 209-75,  221-75,   230-75,  231-75,
251-75, 159-76,  160-76,  195-76,  211-76,  225-76,  226-76 and 237-76  of the  RAPS
data base.
                                      95

-------
                   PBM SIMULATION-750701
  a.
  CL
  o
  o
                               r
               7A     10.0     12.5     19.0
                     TIME,  HOURS (CST)
                                17.3
20.0
Figure A-la.
PBM simulation results for Day 182  of  1975.  Time
series  of  average observed (circles),  range of
observed (dashed lines), and model-simulated (solid
line) hour-average concentrations of CO.
                            96

-------
     .100
                   PBM SIMUIATION-750701
                                   ............
              7.3
 10.0     12.5     15.0
TIME,   HOURS (CST)
17.5
20.0
Figure A-lb.  PBM  simulation  results for N02 - Day  182 of 1975.
             Keys to figure  described in A-la.
                  PBM SIMULAT10N-750701
  a.
  a.
    0.00
              7'A     10.0     12.5     15.0     17.5
                    TIME,  HOURS (CST)
       20.0
Figure A-lc.   PBM simulation results for 63 - Day  182 of 1975.
             Keys to figure described  in A-la.
                          97

-------
                   PBM SIMULATION-750702
   2
   CL
   O.
  o
  O
       5.0
         10.0     12.5      15.0
        TIME,  HOURS (CST)
17.5
20.0
Figure A-2a.
PBM simulation results for Day 183  of  1975.  Time
series  of  average observed (circles),  range of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            98

-------
                   PBM  SIMULATION-750702
.TOO






.075



^^
^2*
Q.
Q.

.•OSO
CS
0
.025



5.

' \
' \
/ \
' V
' \
1 V
' >
— / \
\
\
\
\
\

• / \ \ '

fc- \. \ r '*<»«•"'
• o \ \ 	 N /
: \ °
- ^v.° * °

••

0 7A 10.0 12.5 15.0 17.5


.

™


—
^

^
"



•

—
-
-
_
.
-
I
20,
                     TIME,  HOURS (CST)

Figure A-2b.  PBM simulation results  for N02 - Day  183 of 1975.
             Keys to  figure described in A-2a.
                   PBM  SIMULAT10N-750702
   Q.
   Q.
   fO
   o
    0.00
                      10.0     12.5     15.0     17.5
                     TIME,   HOURS (CST)
20.0
Figure A-2c.  PBM simulation results for 03 -  Day 183 of 1975.
             Keys  to figure described in A-2a.
                           99

-------
                   PBM  SIMULAT10N-750703
     a.
     o_
     o
     o
             A
                       j
                              .a.
       5.0
  7-5     10.0     12.5     15.0
        TIME.   HOURS (CST)
17.5
20.0
Figure A-3a.
PBM simulation  results for Day 184 of 1975.  Time
series of  average observed (circles), range  of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            100

-------
                   PBM SIMULATION-750703
   a.
   Q.
   CS
   O
     0.00
7.5
                      10.0     12.5     15.0
                     TIME,  HOURS (CST)
17.5
20.0
Figure A-3b.  PBM simulation results for N02 - Day  184 of 1975.
             Keys to figure described in A-3a.
                   PBM SIMULATION-750703
  QL
  Q.
  ro
  O
                         i  i T  i r i TriTiiiiiiiir
    0.00
       5U)     7A     iaO     12.5      15.0     17.5     20.0
                     TIME,  HOURS (CST)

Figure  A-3c.  PBM simulation results for 03  - Day 184 of 1975
             Keys to figure described in A-3a.
                          101

-------
0.
0.
o
o
   2.0
   1.3
   1.0
                 PBM SlMUUVTlON-750728
   QJQ
               iiri|(i!iitiiiiiiiij
                            i i  i 7">r'~ T
       &0
            73      10.0     12.5     15.0
                   TIME,  HOURS (CST)
17.5
20.0
Figure A-4a.
           PBM  simulation results for Day 209 of 1975.   Time
           series  of average observed (circles), range  of
           observed (dashed lines), and model -simulated (solid
           line) hour-average concentrations of CO.
                         102

-------
                    PBM SIMULAT10N-750728
                 I i  i i  i  I 1  i I  1  J I  I 1  1  I i  I I  1  I 1  I I
   a.
   CL
   CM
   o
     0.00
           i  i  i I  t
                   !.....!_. 1 ..!  I 1 I  I I  I J_ . t i I  L II  i_._.J_... _L I  1 L
5.0
               7 A      10.0     12.3     15.0     17.5
                      TIME,  HOURS (CST)
20.0
Figure A-4b.  PBM simulation results for NOg - Day  209 of 1975,
              Keys to  figure described in A-4a.
                    PBM SiMULATlON-750728
                           t  I !  I  T L | |  ill  j I  1	1 I  I
               7.5      10.0     12.5     15.0     17.5
                      TIME,  HOURS (CST)
                                               20.0
Figure A-4c.
      PBM simulation  results for 63 - Day 209 of 1975.
      Keys to figure  described in A-4a.
                            103

-------
   a.
   a.
   O
   o
                   PBM SIMULATION-750809
         i  i  r i  I i  i  i i  I i  i  i i  1 i  i  i"7'*l'"VTT"'i  I I  i  i i
       5.0
Figure A-5a.
7 A      10.0      12.3      13.0      17.5

       TIME,  HOURS (CST)
                                         2aO
PBM simulation  results for Day 221 of 1975.   Time

series of average observed (circles), range  of

observed (dashed lines), and model -simulated (solid

line)  hour-average concentrations of CO.
                            104

-------
   Q.
   Q.
   CN
   O
      .08
      .06
      .04
      .02
                    PBM SIMULATION-750809
     0.00 L-1
          I  I I  I  I I  I I I T i  I I  T  j I  1 T I



               -•*'** Y
                                          I  1 I  I  I I  I 1  1
._!...._L ..i	Ll  I i  1.1 I  t t  I I  I  I I  t I  1  t I  I
       5.0      7A      10.0     12.5     15.0     17.5     20.0
                      TIME,  HOURS (CST)

Figure A-5b.  PBM simulation  results for N02 - Day 221  of 1975.
              Keys  to  figure  described in A-5a.
                    PBM SIMUUTION-750809
      *OJ ft "r - i  i IT  i i  II i i  [  l f  i (T  1 III I  1 " ]~ I  I I ~"i
               7A      10.0     12.5     .15.0     17.5
                      TIME,  HOURS (CST)
                                            20.0
Figure A-5c.   PBM  simulation results for 03  -  Day  221 of 1975,
              Keys to figure described in A-5a.
                            105

-------
  Q.
  CL
  o
  o
                   PBM SIMULXriON-750818
                 1  I I  I
           I
           I
        _  I
       9.0
Figure A-6a.
 7J      10.0      12.5      1S.O
        TIME,  HOURS (CST)
17.5
20.0
PBM simulation  results for Day 230 of 1975.   Time
series  of  average observed (circles), range  of
observed  (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            106

-------
                   PBM SIMULATION-750818
   o.
   a.
   CN
   O
   z
                I **«»**.M«.^***B**«»«_ *»*»•«•_«..«..+ _.*.•»*-.*

          titfltrifiitiiFiifiTitii
     0.00
 10.0     12.5     15.0     17.5

TIME,  HOURS  (CST)
                                                     20.0
Figure A-6b.  PBM simulation  results for N02  - Day 230 of 1975.

             Keys to figure  described in A-6a.
                   PBM SIMULAT10N-750818
              i  i t  i i  i  i i  i i  i i  i i  i <  i  i *  i '  i *  * •*
               JA     10.0     ^2A      15.0     17.5

                     TIME,  HOURS (CST)
                               20.0
Figure A-6c.  PBM simulation results for  03 - Day 230 of  1975.

             Keys to  figure described in A-6a.
                           107

-------
                   PBM SIMULAT10N-750819
  a.
  a.
  o
  o
               I  i  r i  i ^i  i i  i j  i i  i  i i  i i  i  i i
       9U)
Figure A-7a.
7A
 10.0
TIME,
                        15.0
               HOURS  (CST)
17.3
20.O
PBM simulation results for Day  231 of 1975.  Time
series  of  average observed (circles), range of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            108

-------
                   PBM SIMULATION-750819
   Q.
   Q_
              I  I I  I I  I  I !  I I  I j  I  I I  I I  I I  I I  I  I I  I !
    aoo L-1
       &0     7.5     10.0     12.3     15.0      17.5
                     TIME,  HOURS (CST)
20.0
Figure A-7b.  PBM simulation  results for N02  - Day 231 of 1975.
             Keys to figure  described in A-7a.
                   PBM SIMULATION-750819
      .301 i  i r >
                         i i i  i ii i  i i
                                i . i  i i  I i  i i  i I  i  i i  i
                      10.0     1Z5      15.0     17.5
                     TIME,  HOURS  (CST)
20.0
Figure A-7c.  PBM simulation results for 03  - Day 231 of 1975.
             Keys to figure described in A-7a.
                            109

-------
   5
   CL
   Q_
  O
  o
                   PBM S1MULATION-750908
                I '
                \
          \
I
                 \s
               7.5     10.0     12.5     15.0
                     TIME.  HOURS (CST)
                                17.5
                      20.0
Figure A-8a.
PBM simulation results for Day  251  of 1975.  Time
series  of  average observed (circles), range of
observed (dashed lines),  and model-simulated (solid
line) hour-average concentrations of CO.
                           110

-------
     .100
     .075
                    PBM  SIMULATION-750908
   a.
   CL
     .090
   s
     .025
    0.000
                    ir*|  r r r r
                                     r r  j i  i  i r
                                —r'VTT'4~TTT~1 I
       5JO      7.3      10.0     1Z5     15.0
                      TIME,   HOURS (CST)
                                 17.5
20.0
Figure A-8b.   PBM  simulation results  for N02 - Day 251 of 1975,
              Keys  to figure described  in A-8a.
                   PBM SIMUUT10N-750908
                   -T ,  I .  . ,  . 1  r . .  . I  . .  .  .
                      10,0     12.5     15.0     17.5
                      TIME,   HOURS (CST)
                                        20.0
Figure A-8c.
PBM simulation results for 03 -  Day  251 of 1975.
Keys to  figure described in A-8a.
                          Ill

-------
                   PBM SIMULAT10N-760607
    a.
    Q.
    O
    O
         T~ i  i I  I i  r  i r r T r *  i  i i  r i  r T i  i i  i i  i i r
             A
               1A     10.0     12.5     15.0
                     TIME,  HOURS (CST)
                                17.5
20.0
Figure A-9a.
PBM simulation results for Day 159  of  1976.  Time
series  of  average observed (circles),  range of
observed (dashed lines), and model-simulated (solid
line) hour-average concentrations of CO.
                            112

-------
                   PBM SIMULAT10N-760607
   CL
   Q.
   CS
   O
    0.00
       &0


Figure  A-9b.

                      10.0     12.3     15.0     17.5
                     TIME,  HOURS (GST)
                                       20.0
             PBM simulation results for  N02 - Day 159 of  1976,
             Keys to figure described in A-9a.
                  PBM SIMULATION-760607
                     10.0     12.5      15.0     17.5
                    TIME,  HOURS (CST)
                                       20.0
Figure  A-9c.
PBM simulation results  for 63 - Day 159  of 1976,
Keys to  figure described  in A-9a.
                          113

-------
                   PBM SIMUtATION-760608
  a.
  Q_
  o
  o
Figure A-lOa.
               7.S
        iao      12.5      15.0
       TIME,  HOURS  (CST)
17.5
20,0
PBM simulation results for Day 160  of  1976.  Time
series  of  average observed (circles),  range of
observed  (dashed lines), and model-simulated (solid
line) hour-average concentrations  of CO.
                           114

-------
                   PBM SIMULAT10N-760608
               7.5
        10.0     12.5     15.0
       TIME,  HOURS  (CST)
17.5
20.0
Figure A-lOb.
PBM simulation results for N02 - Day 160  of  1976.
Keys to  figure described  in A-lOa.
                   PBM S1MUIXT10N-760608
   o.
   0.
                I  I I  I I  I I  I I  I I  i.^1^ I  I I  I I  I I  I  I I  I I
                 i  i i  i I  i i  i  i I  i i  i i
     0.00
               7J     10.0     12^    15.0
                     TIME,  HOURS (CST)
                               17.5     2aO
Figure A-lOc.  PBM simulation results for  03 - Day 160 of  1976.
              Keys' to  figure described in A-lOa.             .
                           115

-------
     1.00
      .75
                    PBM  SIMULATION-760713
    a.
    a.
      .30
    O
    O
      .25
     aoo
 10.0     12JS     19.0
TIME.  HOURS (CST)
                                17.5
                                                      20.0
Figure A-lla.
PBM simulation  results for Day 195 of 1976.   Time
series  of  average observed (circles), range  of
observed  (dashed lines), and model -simulated (solid
line)  hour-average concentrations of CO.
                            116

-------
      .06
                    PBM SIMULAT10N-760713
      .04
   Q.
   Q.
   cs
   O
      .02
     0.00

                                       , ,, ,


                                    V
                                                  Till
                I ill  li  I 1 I I I  I 1  i	I	T I	I
Figure A-llb.
               7A      10.0     12.5      1S.O      17.5     20.0
                      TIME,  HOURS (CST)

               PBM  simulation results for  N02 - Day 195 of 1976.
               Keys to figure described in A-lla.
                    PBM  S1MUUVT10N-760713
   a.
   Q.
                        I  1 I  I I  I  < 1  I I  j I  I  f I  I
          I I  t  I 1  I t  I 1  1  I I  I I  I
     0.00
               7A      10.0      12.5     15.0     17.5
                      TIME,  HOURS  (CST)
                                                      20.0
Figure A-llc.  PBM simulation  results for 03 - Day 195  of 1976.
              Keys to figure  described in A-lla.
                           117

-------
                   PBM SIMULAT10N-760729
   2
   0.
   0.
   o
   o
          I 1  L 1  I (  I  I I  I I  1 I  I
Figure A-12a.
                       10.0     12.5     15.0
                      TIME,   HOURS (CST)
                                17.5
20.0
PBM simulation results for Day 211  of  1976.  Time
series  of  average observed (circles),  range of
observed  (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                           118

-------
     .I00r-r-r
  Q_
  Q_
  CN
  O
     .075
     .oao
     .025
    0.000
                    PBM  SIMUUVT10N-760729
       5U>


Figure A-12b.
                           A
                            \
                            \
                                                 i  ( |  i

7.5     10.0     12.5     15.0     17.5     20.0
       TIME,  HOURS (CST)

PBM simulation results for N02  -  Day 211 of 1976.
Keys to figure described in A-12a.
                   PBM SIMULXnON-760729
   a.
   a.
        r  •^f'-"—
        &f^^T~\* i i  i i  t I i  i i  i I  i i  i  i I  i i  i  i I  i i  i i
     0.00
               7A     iaO     12.5     -15.0
                      TIME.  HOURS (CST)
                               17.5     20.0
Figure A-12c.  PBM simulation  results for 03 - Day  211  of 1976,
              Keys to figure  described in A-12a.
                            119

-------
                   PBM SIMUUVTION-760812
   Q.
   0.
   O
   o
                      10.0     12.5     15.0
                      TIME,   HOURS  (CST)
                                17.5
20.0
Figure A-13a.
PBM simulation  results for Day 225 of 1976.  Time
series  of  average observed (circles), range  of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            120

-------
   QL
   Q.
      .06
      .06
      .04
      .02
                   PBM SIMULAT10N-760812
1 I f 1 1 1



1II111I1IJI1 J* 1 i I 1
; \
i \
' \
i i i i



     0.00
         XV

                f i  t i  t  I i  i 1  i I  i  t i  i T t i  i  i
       5UJ      7A     10.0     12.5     15.0     17.5
                      TIME,  HOURS (CST)
20.0
Figure A-13b.   PBM simulation results  for N02 - Day 225 of  1976.
               Keys to figure described in A-13a.
                   PBM S1MULATION-760812
   Q.
   Q.
                                i i  i i  i  i i  i iiiiiii
     0.00
                      10.0     12.5     15.0     17.5
                     TIME,  HOURS (CST)
20.0
Figure A-13c.  PBM simulation  results for 03 - Day 225  of 1976,
              Keys- to figure  described in A-13a.
                            121

-------
                   PBM SIMULAT10N-760813
   2
   Q.
   Q.
   O
   O
Figure A-14a.
               7.5
        10.0     12.5     15.0
       TIME,  HOURS (CST)
17.5
20,0
PBM simulation  results for Day 226  of  1976.  Time
series of  average observed (circles),  range of
observed (dashed lines), and model-simulated (solid
line)  hour-average concentrations of CO.
                            122

-------
     .100
     .075 —
                    PBM  SIMULATION-760813
   Q.
   Q.
     .050 -
   CM
   O
   Z
     .025 -
    0.000
               7.5      10.0      12.5      15.0     17.5
                      TIME,  HOURS  (CST)
Figure A-14b.  PBM simulation results  for ^ - Day 226
              Keys to figure described  in A-14a.
                                                      20.0
                                                      of
                   PBM  SIMULATION-760813
                   I  I  I I  I I  I I  I  I I  I I  1 I  !  I I  I I  I I  I
               7.5      10.0     12.5     -15.0
                      TIME,   HOURS (CST)
                                              17.5     20.0
Figure A-14c.  PBM simulation results  for 03 - Day 226 of 1976,
              Keys to figure described  in A-L4a.
                           123

-------
                  PBM  SIMULATION-760824-
CL
a.
O
o






3






2






1,





I ,1 1 M | 1 1 1 1 | 1 1 1 1 | 1 1 1 1 | 1 1 I 1 | 1
I \
1 \
1 \
' 1 \
/ \
-r \
i \
•r \
l'
I O >
r \
• i
- °
i
i
\
0 V
,\ \ x' X
_ / \ N / N

" ' ^^^ \/
• / \ ^^^-^^.^ • o
L/ ^--^-a— ^ o * o
/ N "^
7 ,,,,!.,,, >VTT-i-4-i-^f-r-r ""?>>, 1 ,

™
—

™
.
—

~
-


^
_
-
—

•
.

""•

—
-

1 t
                     10.0     12.5     15.0
                    TIME,  HOURS (CST)
                                              17.5
20.0
Figure A-15a.
             PBM simulation results for Day 237  of  1976.  Time
             series  of  average observed (circles),  range of
             observed (dashed lines), and model-simulated (solid
             line) hour-average concentrations  of CO.
                          124

-------
       .15i  i i  i i
                    PBM SIMULATION-760824

                               12.5     15.0     17.5
                      TIME.  HOURS (CST)
                                     20.0
               PBM  simulation results for ^2  - Day 237 of 1976.
               Keys to figure described in A-15a.
        .201  i i i  i
     a.
     a.
       0.00
         SUJ
                    PBM SIMUUM10N-760824
7A     10.0     124    15.0
      TIME,  HOURS (CST)
2Q.Q
Figure A-15c.  PBM simulation results  for 03 - Day 237 of 1976.
              Keys- to figure described  in A-15a.
                            125

-------
                                 APPENDIX B

                    PBM - SUMMARY OF STATISTICAL RESULTS

     This table  summarizes the  mean concentration  data and  the analysis  of
trends  data  obtained  from PBM  simulations on  all   20 days  for the  species
N02. NOX, 03, NMHC and CO.  Note that the mode*!  internally  estimates a measure
of NMHC from its explicit  structural  classes.   The table also includes summary
statistics on concentration maxima data  for all  20  test days.for the species
N02, 03 and CO.   The statistical  parameters  included  in this table are
(Bencala and Seinfeld, 1979):
                          •
     (a)  mean observed and mean  predicted concentration,

     (b)  mean  residual   (measures  the  average bias  in the predictions  and
          indicates whether  the  model  predominantly over-  or  underpredicts),

     (c)  root mean  square (RMS)  error  (measures  the  average  spread of  the
          residuals, but is insensitive to bias),

     (d)  mean  residual/mean   observed  (measures  the  average  percentage  of
          over- or underprediction), and

     (e)  error band (measures the percentage of predictions that fall within a
          25% band around the observations).

     The analysis of  trends data indicates  whether or not the predictions obey
the  same  fundamental  relationships as  the  observations.   The  parameters con-
tained in this table include:
                                      126

-------
     (a)  correlation coefficient  (measures  the  overall  degree to  which the
          magnitude of  the  predictions increases  linearly  with the magnitude
          of the  observations,  but  is insensitive  to  the extent  of  the in-
          crease) ,,

     (b)  slope (the  linear  least-squares  curve  fit  measure  of  the  average
          increase in the observations  as  the  predictions  increase),

     (c)  intercept  (the  linear least-squares  curve  fit measure  of the bias
          if the  slope  parameter is very nearly one;  otherwise  no independ-
          ent meaning),  and

     (d)  sum of  the squared error  (a relative measure  of  the scatter about
          the linear least-squares  fit  to  the  data).

     Concentration maxima  data are presented for the  species  MOg,  03 and
     CO for each test day.   The  parameters listed  here are:

     (a)  peak  observed  concentration (and the  time of occurrence),

     (b)  peak  predicted concentration  (and the time  of  occurrence),

     (c)  residual  concentration of  the peak,  and

     (d)  residual/observed   concentration (percentage  over-   or   underpredic-
          tion  of  the peak).

     In all  tables a positive  residual  indicates  an  underprediction  by the
model  and a negative residual  indicates an overprediction.
                                      127

-------
TABLE B-l.  SUMMARY OF MODEL PERFORMANCE STATISTICS PROM THE PBMa
                       Mean Concentration Data

Species

N02
NOX
03
NMHC
CO
Mean observed
concentration
(ppm)
0.0285
0.0440
0.0725
0.402
0.369
Mean predicted
concentration
(ppm)
0.0281
0.0412
0.0879
0.350
0.676
Mean
residual
(ppm)
0.00039
0.00284
-0.0154
0.0520
0.193
-
RMS error
(ppm)
0.00931
0.0148
0.0289
0.188
0.356
Mean residual
Mean obs. cone.
(I)
1.4 -
6.5
-21.2
12.9
22.2
Error
band0
(?)
60.0
56.5
45.4
43.1
43.5
 Species
                       Analysis of Trends Data
Correlation
coefficient
Slope
Intercept
  (ppm)
   Sum of
squared error
   (ppm2)
  N02
  NOX
  °3
  NMHC
  CO
   ,876
   ,923
   ,871
   .837
  0.848
0.753
0.832
1.09
0.666
0.599
 0.00734
 0.00976
-0.0230
 0.169
 0.464
   0.0276
   0.0685
   0.269
  12.1
  50.6
                                  128

-------
TABLE 8-1.   (Continued)



                    SUMMARY OF MODEL PERFORMANCE STATISTICS  FROM  THE  PBM*

Concentration Maxima Data
Jul ian
date
142-75
178-75
182-75
183-75
184-75
207-75
209-75
Observed Peak
Species
N02
03
CO
N02
°3
CO
N02
°3
CO
NOg
03
CO
N02
03
CO
N02
°3
CO
N02
°3
CO
• Value
(ppm)
0.0367
0.137
1.31
0.0535
0.159
1.85
0.0408
0.0939
0.979
0.0526
0.115
1.57
0.0792
0.114
2.68
0.0416
0.140
1.24
0.0304
0.108
0.976
Concentration
Time
(CST)
0600 •
1200
0700
0700
1400
0600
0600
1300
0700
0800
1300
0700
0700
1300
0700
0600
1400
0500
0800
1000
0800
Predicted Peak
Value
(ppm)
0.0421
0.148
1.18
0.0456
0.173
1.07
0.0532
0.140
1.68
0.0594
0.0971
1.09
0.0753
0.182
2.11
0.0509
0.145
1.02
0.0406
0.122
1.06
Concentration
Time
(CST)
0700
1300
0600
0600
1600
0600
0600
1600
1500
0600
1700
0600
0600
1300
0500
0600
1500
0500
0700
1300
0600
Residual
(ppm)
-0.0054
-0.0110
0.130
0.0079
-0.0140
0.780
-0.0124
-0.0461
-0.7001
-0.0068
0.0179
0.480
0.0039
-0.0680
0.570
-0.00093
-0.0050
0.220
-0.0102
-0.0140
-0.0840
Residual
Obs. cone.
(*)
-14.7
-8.0
9.9
14.8
-8.8
42.2
-30.4
-49.1
-71.6
-12.9
-15.6
30.6
4.9
-59.6
21.3
-22.4
-3.6
17.7
-33.6
-13.0
-8.6
                                            129

-------
TABLE &-1.   (Continued)



                    SUMMARY OF MODEL PERFORMANCE STATISTICS  FROM  THE  PBM*

Concentration Maxima Data
Julian
date
221-75
230-75
231-75
251-75
159-76
160-76
195-76
Observed Peak.
Species
NO 2
03
CO
N02
03
CO
N02
03
CO
N02
°3
CO
N02
*3
CO
N02
°3
CO
N02
03
CO
Value
(ppm)
0.0453
0.132
1.94
0.0541
0.0717
2.08
0.0394-
0.0842
1.04
0.0667
0.0841
2.85
•0.0938
0.125
2.78
0.0652
0.162
1.56
0.0287
0.151
0.555
Concentration
Time
(CST)
0700
1300
0600
0700
1300
0600
0800
1000
0900
0900
1500
0600
0700
1300
0600
0700
1400
0600
1700
1500
0700
Predicted Peak
Value
(ppm)
0.0469
0.125
1.62
0.0499
0.119
1.72
0.0364
0.101
1.06
0.0461
0.205
1.72
0.0816
0.210
1.98
0.0693
0.170
1.50
0.0260
0.137
0.522
Concentration
Time
(CST)
0700
1300
0500
0700
1500
0600
0700
1300
0600
0700
1500
0600
0600
1200
0500
0600
1500
0500
0600
1600
0600
Residual
(ppm)
-o'.ooie
0.0070
0.320
0.0042
-0.0473
0.360
0.0030
-0.0168
-0.0200
0.0206
-0.121
1.13
0.0122
-0.0850
0.800
-0.0041
-0.0080
0.0600
0.0027
0.0140
0.0330
Residual
Obs. cone.
(%)
-3.5
5.3
15.5
7.8
-66.0
17.3
7.6
-20.0
-1.9
30.9
-143.9
39.6
13.0
-68.0
28.8
-6.3
-4.9
3.8
9.4
9.3
• 5.9
                                           130

-------
  TABLE 8-1.   (Continued)

                      SUMMARY OF MODEL PERFORMANCE STATISTICS FROM THE PBM*

Concentration Maxima Da.ta
Jul ian
date
211-76
212-76
225-76
226-76
237-76
275-76
Observed Peak Concentration
Species
N02
03
CO
N02
03
CO
N02
03
CO
N02
03
CO
N02
03
CO
M02
°3
CO
Value
(ppm)
0.0409
0.0939
0.869
*
0.0609
0.127
1.73
0.0421
0.111
1.08
0.0580
0.144
1.75
0.0832
0.115
2.45
0.108
0.183
4.17
Time
(CST)
1200
1500
0700
0700
1200
0600
1600
1200
0600
0700
1500
0700
0700
1100
0700 '
1700
1400
0700
Predicted Peak
Value
(ppm)
0.0338
0.0878
0.719
0.0528
0.121
1.42
0.0491
0.0919
1.26
0.0489
0.171
1.05
0.0521
0.156
1.05
0.0869
0.223
3.00
Concentration
Time
(CST)
0700
1600
0600
0600
1400
0500
0700
1100
0600
0700
1500
0600
0700
1400
0600
1000
1500
0700
Residual
(ppm)
0.0071
0.0061
0.150
0.0081
0.0060
0.310
-0.0070
0.0191
-0.180
0.0091
-0.0270
0.700
0.0311
-0.0410
1.40
0.0211
-0.0400
1.17
Residual
Obs. cone.
(*)
17.6
6.5
17.3
13.3
4.7
17.9
-16.6
17.2
-16.7
15.7
-18.8
40.0
37.4
-35.7
57.1
19.5
-21.9
28.1

Statistics for Mean Concentration and Analysis of Trends data are developed for all  20 test
 days;
 Statistics for Concentration Maxima are listed for each test day.
 All  concentration values represent an average over the model  domain.

bPercent of predictions within 25% of observations.
                                              131

-------
                                   APPENDIX  C

                    LPM TEMPORAL  CORRELATION COEFFICIENTS
                   AND TIME SERIES  FOR REMAINING  TEST  DAYS

     This appendix  presents  temporal  correlation coefficients for 03, NO,  and
    for  all 20  days and time series  from LPM simulations  for the 15 days  not
presented in the body of the report.

     The temporal  correlation  coefficients  were  computed between  observed  and
predicted series from  the  start  time to the lesser of  1700  CST or the end  of
the run.   These coefficients  for  Levels 1 and  3  appear  in Table  C-l.    The
choice of a termination time can  have significant influence on the correlation.
For instance, the  correlation  for  03 on Day 231 is below the average for  the
20 days, but the simulation was  fairly  successful.  A part of the problem with
the correlation  is  the discrepancy  between prediction  and  observation  after  the
time  of the  observed maximum.    As  explained  previously  for  Day  231, this
discrepancy is  expected because  the modeled  parcel  has moved  away from  the
stations contributing  to the  "observed"  value.   The correlations for NO  and
N02 tend to be  low,  perhaps  reflecting the low  concentrations  and  the  conse-
quent difficulty in obtaining accurate measurements.
                                     132

-------
TABLE C-l.  TEMPORAL CORRELATION COEFFICIENT BETWEEN OBSERVED AND LPM PREDICTED
POLLUTANT CONCENTRATIONS FOR THE PERIOD FROM
LATION TO 1700 CST OR EARLIER TERMINATION.
                                                          HE START OF THE SIMU-
Day
             03

         L-l    L-3
    NO

L-l "  L-3
N02
                                                                   L-l
    L-3
142 (1975)
178
182
183
184
207
209
221
230
231
251
159 (1976)
160
195
211
212
225
226
237
275


0.53
0.52
0.99
0.95
0.89
0.73
0.64
0.82
0.85
0.64
0.67
0.97
0.94
0.94
0.90
0.80
0.11
0.92
0.89
0.90
7 = 0.78
°r 0.21
0.66
0.88
0.98
0.97
0.83
0.78
0.58
0.76
0.83
0.58
0.64
0.97
0.92
0.95
0.89
0.73
0.06
0.83
0.84
0.80
0.77
0.20
-0.03
0.54
0.95
0.90
0.90
0.18
0.53
0.82
0.30
0.79
0.91
0.94
-0.12
0.77
0.50
-0.01
0.13
0.78
0.73
0.45
0.49
0.42
0.34
0.51
0.70
0.39
0.48
0.41
0.17
0.56
0.20
0.15
0.87
0.88
-0.02
0.50
-0.28
0.07
0.34
0.48
0.70
0.88
0.42
0.30
-0.26
-0.39
0.94
0.65
0.77
0.32
0.93
0.77
0.43
0.62
0.84
0.89
-0.09
0.61
0.70-
0.30
-0.36
0.28
0.83
0.66
0.47
0.42
-0.03
-0.71
0.94
0.65
0.81
0.53
0.93
0.85
0.42
0.57
0.92
0.92
0.23
0.44
0.78
0.30
0.07
0.88
0.87
0.87
0.56
0.42
                                      133

-------
    75142.  RAMS  101 AT  1200CST. START 0500
      l(AJ •••p^T^"H^Z^™"^^T^
       SO
       25
0       25
                          50
                        KM
75      100
Figure C-la.   LPM parcel  trajectory  for Day 142 of 1975.
                          134

-------
       75142, RAMS 101 AT  1200CST.  START 0500
                               iiiiiiiiiiiiii
         	 OBS
       .	 PRED L-1
       .	PRED L-3
             i 1 I i  i i  I I  r i  t I  i i  i I  I i  i i  i 1  t i  i i
                    10.0     12.5     15.0
                        HOUR, CST
                       17 J
        20.0
Figure C-lb.  LPM time series of observed and predicted  03  hourly
              concentrations on Day 142 of 1975.
   .100
   .075
   .050
       75142, RAMS 101  AT 1200CST.   START 0500
  0.000
            I  I I  I
                    I I  i I  I I  I I  I I  I I  . I
                        I  , I  I
                   	 OBS
                   —  PRED L-1
                   —  PRED L-i
        i i  i i  I I  1 t  1 I  t 1  I I  I t  I
     5.0
10.0    12.5     15.0
    HOUR, CST
17.5
20.0
Figure C-lc.  LPM time series of observed  and predicted NO hourly
             concentrations on Day 142  of 1975.
                            135

-------
 0.
 CM
 O
    .20
    .15
    .10
    .05
       75142, RAMS  101 AT 1200CST.  START 0500
   0.00
                lIllIillIJtlliTI  I I  I |  III

                                      	  oes
                                      	   PRED L-1
                                      	PRED L-4
       A '\
       /  \ •..
                                   \v
                              t 11 1 1 i  I I 1_ I  1 1  II
             7.5     10.0     12.5     15.0
                        HOUR, CST
                    17,5
20.0
Figure C-ld.   LPM  time  series of observed and predicted  N02  hourly
              concentrations on Day 142 of 1975.
       75142, RAMS 101  AT 1200CST.   START 0500
   Q.
   Q.

   o"
   O
                                        I I  I I  ...
                                       	 08S
                                       —  PRED L-t-
                                       —  PRED L-5.
        * I  I I  I I
II 111  J i  I • t  I L _ t I  I I
     5J)      7A     10,0     12.5     15.0
                        HOUR, CST
                    17.5
20.0
Figure C-le.   LPM time series of observed and  predicted CO hourly
              concentrations on Day 142 of 1975.
                            136

-------
       75178. RAMS  112 AT  1400CST. START 0500
          IW p™T"T~~T~ T ^H""rT^^^TT"^~™T'*"'TOFT™Tl"'"T^^P™1^™^^T1"™
          SO
                 yK-^j o   o
              t  II |  I  |  |  || 111  j  1  I I  1  1  I
0        29        90
                KM
                                       79       100
Figure C-2a.  LPM  parcel  trajectory for Day 178 of 1975.
                              137

-------
       75178, RAMS 112 AT 1400CST.  START 0500
  0.
  O.
  ul
  O
  rsj
  O
    OLO
             OBS
             PRED L-1
             PRED L-3
                       /


                                /


                             /

              /
                    10.0     12.5     15.0
                        HOUR, CST
                              17.5
            20-0
Figure C-2b.   LPM  time series of observed and  predicted 03 hourly
              concentrations on Day 178 of 1975.
     .4
       75178, RAMS 112 AT 1400CST.  START 0500
  o
     .1
                            T
                       T
I  1^ I  I Til  f

	 OBS
	  PRED L-1
	PRED L-4
    0.0«
     54
7.5     iaO     12J     15.0
           HOUR, CST
     17.5
2ao
Figure C-2c.   LPM time series of observed  and predicted NO hourly
              concentrations on Day 178  of 1975.
                            138

-------
       75178, RAMS  112 AT 1400CST.  START 0500
     .4
  Q.
  Q.
   *
  CM
  O
    0.0
          iiiiriiijiiijj
                                            OBS
                                            PRED L-1
                                            PRED L-51
     5U)      7.5     iaO     1ZS     13.0
                        HOUR, CST
zao
Figure C-2d.   LPM  time  series of observed and predicted  N02 hourly
              concentrations on Day 178 of 19.75.
   10.0
       75178, RAMS 112  AT 1400CST.  START 0500
I
O*
O
    2J>
    QJQ
            i  I r i §
                     I  i i  i i
                                           OBS
                                           PRED L-1
                                           PRED L-4
               I | i . L....._!_ I I i	1  I I  I f  1 I  t 1  I I
     &0      7.5     10.0     12.S     15.0
                        HOUR, CST
17.5
20,0
Figure C-2e.  LPM time series of observed  and  predicted CO hourly
             concentrations on Day 178 of 1975.
                            139

-------
    75182. RAMS  125 AT 1200CST. START 0500
      •w ^T-MT"^r"TMT™™^^t  i~" v I  k  TCHT"^^"^^
       90
        o    o
                 29
  30

KM
100
Figure C-3a.   LPM parcel  trajectory  for Day 182 of 1975,
                          140

-------
       75182, RAMS 125 AT 1200CST.  START 0500
       .	PRED L-3  A
               i  i i  i I	i i  i t  I I  1 i  I I  I I  * I  I I  I I  t
&0
                    10.0     12J     15.0     17.5     2aO
                        HOUR,  CST
Figure C-3b.   LPM  time  series of observed and predicted 03 hourly
              concentrations on Day 182 of 1975.
       751-82. RAMS 125 AT 1200CST. START 0500
                                t i  i t  i i  i i  j i  i i  r
                                        -   oss
                                      ......  PRED L-
                                      	PRED L-31
 I
 CL

 O
   0.00
            7A      10.0     1Z3     19.0     17.5     2aO
Figure C-3c.  LPM time series of observed  and predicted NO hourly
             concentrations on Day 182  of 1975.
                            141

-------
        75182, RAMS 125 AT 1200CST. START 0500
                                            OSS
                                            PRED L-1"
                                            PRED L-i
   0.000
                     iao     12.3     15.0
                         HOUR, CST
17.5
20.0
 Figure C-3d.   LPM time series of observed  and predicted N02 hourly
               concentrations on Day 182  of 1975.
        75182, RAMS 125 AT  120QCST. START 0500
  Q.
  0.

  o"
  O
                                            OBS
                                            PRED L-1-
                                       	PRED L-3
      8.0      7A     10.0     12.5     15.0     17.5     20.0
                        HOUR, CST
Figure C-3e.   LPM time series of observed  and  predicted 03 hourly
              concentrations on Day 182 of 1975.            :
                             142

-------
    75183, RAMS 124 AT 1500CST. START OSOO
       • W r™"T^"r ~T~~ i "*T~™J*"" |  | j  V ^  ^ |  Lf 11  in ^  |  | |
       SO
                           I  '  "" '  I
                    I i  L |  I 1 I  i  I  I I  II  1
25       90

       KM
                                   75       100
Figure C-4a.   LPM parcel  trajectory  for  Day  183  of  1975,
                          143

-------
       75183, RAMS 124 AT  1500CST. START 0500
   Q.
   Q.

   UJ

   O
   r>4
   O .1
    0.0
         	oas
         	 PRED L-1
         	PRED L-3
I 1  I I  I  I I I  I  I I  I I  I I  I I  I I  I I  I I  I I  I I
  7.9     10.0     12.9    19.0     17.9    20.0
             HOUR, CST
Figure C-4b.   LPM.time series of observed and predicted 03 hourly
              concentrations on Day 183 of 1975.
       75183,  RAMS  124 AT 1500CST. START 0500
                                            OBS
                                      	  PRED L-1
                                        —  PRED L-1
                     10.0     12.9     19.0
                        HOUR, CST
                                17.9
20.0
Figure C-4c.   LPM time series of observed and  predicted NO hourly
              concentrations on Day 183  of 1975.
                           144

-------
     .06
     .04
  Q.
  Q.

  CM
  o
     .02 -f
    aoo
        75183, RAMS 124 AT 1500CST.  START 0500
          I  I I r I  F I T
                                    1 I  i r i i  jT i  I i

                                       	 OBS
                                       	  PRED L-1
                                       	PRED L-J1
              7.5     iaO    1Z3     13.0    17.3     20.0
                         HOUR, CST
Figure C-4d.  LPM time  series of observed and predicted N02 hourly
              concentrations on Day 183 of 1975.
       75183, RAMS 124 AT 1500CST. START 0500
     1.5
  I
  O
  O
             i r  i i  i r i i  r r i i  i i  i i  i i  i
                                            OBS
                                            PRED L-1.
                                            PRED L-i
                . ...  I .... I  ..
                     10.0     1Z5     15.0
                        HOUR, CST
17.3
20.0
Figure C-4e.  LPM time series of observed and predicted CO hourly
             concentrations on Day  183 of 1975.
                            145

-------
    75184.  RAMS  118 AT 1300CST. START 0500
      • *^ ^^T™"^^HHT""T™"^""TT r T TTO^"^TT^m
       75
       90
       25 -
                 25
  50

KM
75      100
Figure C-5a.   LPM parcel trajectory for Day  184 of  1975.
                          146

-------
        75184, RAMS  118 AT 1300CST. START 0500
               i
              OBS
              PRED L-1
        I	PRED L-3
         i i -**V liiiiliiifititiiiiiiliitt
                     10.0     1Z5     15.0
                         HOUR, CST
                       17 J
        20.0
Figure C-5b.   LPM time  series of observed and predicted 03 hourly
              concentrations on Day 184 of 1975.
    .too
    .073
       75184,  RAMS  118 AT 1300CST. START 0500
  Q.

  O*
    .090
   .025
  0.000
                       i  r i  T j  i i  i i  i i  i i  i
                                      -  OBS
                                      ------  FRED L-1
                                      --  PRED L-i
      fluO
10.0    12.3     15,0
    HOUR, CST
17.3
2O.O
Figure C-5c.  LPM time series of observed  and predicted NO hourly
             concentrations on Day 184  of 1975.
                             147

-------
     .20
     .15
  Q.
  0-  .10
  CN"
  O
  z
       75184, RAMS 118 AT  1300CST. START 0500
                                       	  OBS
                                       	  PRED L-1
                                       	  PRED L-3
    0.00 *-
      5.0
10.0    12.5     15.0
    HOUR, CST
17.5
20.0
Figure C-5d.   LPM time series of observed and predicted N02 hourly
              concentrations on Da^y 184 of 1975.
       75184, RAMS 118 AT 1300CST. START 0500
        i  ii r r i 7 i
                                     i  • •  • •  i '  • •  •
                                       	 OBS
                                       	  PRED L-1
                                       	PRED L-4
     oL_i_a
      54     7.5     10.0     1Z5     15.0
                        HOUR, CST
                      17.5
       2ao
Figure C-5e.   LPM time series of observed and predicted CO hourly
              concentrations on Day 184 of 1975.
                            148

-------
    75297. RAMS 113 AT 1400CST. START 0600
       1(W r"T*"T~^~T" t  i  T^T^T i  V n* i  T"TP"7"""1""^~^"~
       75
       90
                  I  '  '  ' '  I
                          90
                        KM
79       100
Figure C-6a.   LPM parcel  trajectory  for  Day  207  of  1975.
                          149

-------
        75207, RAMS  1 1 3 AT 1 400CST. START 0600
   2  •*
   CL
   Q.

   UJ

   O
   r^i
   O  .1
     0.0
              I I  I I  I I
              OSS
              PRED L-1
              PRED L-3
                          I I  1 I  I F  1 i
                                            i r~ i i i  r
      iO      7J     10.0     12.5     15.0     17.5    20.0
                         HOUR, CST
Figure C-6b.   LPM time  series  of  observed and predicted 63 hourly
              concentrations on Day  207 of 1975.
    .100
    .075
       75207, RAMS 113 AT 1400CST. START 0600
 O
    J325
                        ill! 1  I T  I Illlllflili

                                       —~~"  OSS
                                       	  PRED L-1.
                                       	PRED L-i
0.000 L
   5U>
        J. I 1 L
                     iaO     12.5     15.0
                         HOUR, CST
17.5
20.0
Figure C-6c.   LPM time  series  of  observed and predicted NO hourly
              concentrations on Day  207 of 1975.
                              150

-------
    .100
    .073
  Q.
  Q.
  O
    .023
        75207,  RAMS  113 AT 1400CST. START 0600
                                       ~~~~  08S
                                       	  PRED L-1
                                       	PRED L-1
   OJOOI  ' '  ' '  I '  ' '  ' I  ' '  ' '  I '  i '  ' I  '  ' '  • I  ' '  » '
      5.0      7.5     iaO    12.3     13.0     17.5    2aO
                         HOUR, CST
Figure C-6d.  LPM time series  of observed and predicted N0£  hourly
              concentrations on Day 207 of 1975.
       75207, RAMS 113 AT  1400CST. START 0600
   Q.
   Q.

   O
   O
     0
      30)
                                             I ...
                                            OBS
                                            PRED L-V
                                            PRED L-1
        i  111 [  11 iiliiiiitiii I i  i  i IT I
                                              i  i i  i
7.3     10,0     1Z3     13.0     17.5
           HOUR, CST
20.0
Figure C-6e.
LPM time series of observed  and predicted CO hourly
concentrations on Day  207  of 1975.
                             151

-------
    752Q9. RAMS  118 AT 11OOCST. START 0600
       ItX? p'T^TLTJlT™1illlTT--r' r 1  i  i^nW"^" i  i  r"~T""^™T1™"
       90
                  29
  ao
KM
73       100
Figure C-7a.  LPM parcel  trajectory for Day 209 of  1975.
                           152

-------
   0.
   Q.

   ul
   z
   O
   M
   O  .1
     0.0
        75209, RAMS 1 18  AT 11OOCST.   START 0600
         i i  i i  i i  i i  i  r '  i i '  i i  ' i  ' I  ' '  ' '  i  * »  ' •
         	 08S
         	  PRED L-1
         	  PRED L-3
          "»l-< It  I I  !   f I  f I   1  ! I  I 1  I I  I I  1 I  ! _ I
      5.0      7.5     10.0    12.5     15.0     17.5     20.0
                         HOUR, CST

Figure C-7b.   LPM  time  series of observed and predicted 03 hourly
              concentrations on Day 209 of 1975.
       75209, RAMS 118 AT  110OCST.  START 0600
 Q.
 Q_

 O
                                            OBS
                                            PRED L-1
                                            PRED L-31
   0.00
                     10.0     1ZS     15.0
                        HOUR, CST
17.5
20.0
Figure C-7c.   LPM  time  series of observed and predicted NO  hourly
              concentrations on Day 209 of 1975.
                             153

-------
    .100
        75209, RAMS 118 AT 110OCST.   START 0600
                                             oas
                                             PRED L-1
                                             PREO L-i|
         L__..!__. 1 L_..J_.. .!_ I I  I I  I I  i i  L i
                     10.0     12.5     15.0     17.5    20.0
                         HOUR, CST

Figure C-7d.   LPM time  series of observed and predicted N02 hourly
              concentrations  on Day 209 of 1975.
       75209, RAMS 118 AT  1100CST.  START 0600
   Q.
   Q.

   O*
   O
                                       ~~~~" 08S
                                       	  PRED L-1-
                                       	PRED L-4
                                          1 t  I J  I I  I
      &0     7A      10.0     12.3     15.0
                        HOUR, CST
17.5    20.0
Figure C-7e.   LPM time  series of observed and predicted CO hourly
              concentrations on Day 209 of 1975.
                             154

-------
    75221. RAMS 121 AT 1500CST. START 0600
       iiA/ ^*T"^^™™TOT"i™r^™r'""^^^j^T^»^^
        75
    2  »
                            j-r I«-T <  r
       o   o
                  29
 80

KM
73       100
Figure C-8a.  LPM  parcel  trajectory for Day 221 of 1975.
                          155

-------
     .20
     .15
       75221. RAMS 121 AT  1500CST. START 0600
    0.001-1
i  i i  I i  i i  i
   DBS
   PRED L-1
   PRED L-3
                        I
                                             I
7.5     10.0     12.5     15.0
           HOUR. CST
                                           17.5
                                         zao
Figure C-8b.   LPM time series of observed  and  predicted 03 hourly
              concentrations on Day 221  of 1975.
    .020
       75221. RAMS  121  AT 1500CST. START 0600
                                           1 I  ' '
                                            OBS
                                            PRED L-1
                                      	PRED L-i
  0.000 u
      54
       10.0    12.5     15.0
           HOUR, CST
                                 17.5
20.0
Figure C-8c.   LPM time series of observed  and predicted NO hourly
              concentrations on Day 221  of 1975.
                            156

-------
     .06
     .04
  I
  CM
  O
  z
     .02
    0.00
        75221, RAMS 121  AT 1500CST.  START 0600
                                        ""    08S
                                       	  PRED L-1
                                       	PRED L-J
      54     7.5     iaO     12J     13.0
                         HOUR, CST
 17 A
 20.0
 Figure C-8d.  LPM time series  of  observed and predicted NOg hourly
              concentrations on Day  221 of 1975.
    2JQ
    1.5
 I
 O
 O
    OO
       75221. RAMS  121 AT 1500CST. START 0600
                                      1  ) I  I I  1 i  I

                                      	  08S
                                      	   PRED  L-1
                                      	PRED  L-i
               1_ i _J_ 1 ]_ I. 1. I ll I  !  I t  \ I  I I  I I  t JIL
     54      7.5     10.0     12.5     15.0
                        HOUR, CST
17.5
20.0
Figure C-8e.   LPM time series of observed and  predicted CO hourly
              concentrations on Day 221 of 1975.
                            157

-------
    75230. RAMS 121  AT 1300CST. START 0600
       • W n^T^T^HT"^™ \L i T™™T^^ \  r™"^^^^^^^^T^~P™T™TM"
       90
        9   O
                  I  I  1 I  I  I  1  1 I  I  I  I  I I
                 25
  ao
KM
75       100
Figure C-9a.  LPM parcel trajectory for Day 230 of 1975,
                          158

-------
       75230. RAMS  121 AT 1300CST. START 0600
        r  r t  11  i *  T IT T 1 i  i I  i i  r i  j r i i  r j  IT i
        —— OBS
        	  PRED L-1
       .	PRED L-3
  O.
  Q.

  ul

  O
  N
  O .1
    OO
         \
     &o      7.5     10.0     1^5     15.0
                        HOUR, CST
17.5
zao
Figure C-9b.   LPM  time  series of observed and predicted 63 hourly
              concentrations on Day 230 of 1975.
       75230, RAMS 121  AT 1300CST. START 0600
    .15
    •'«
    .05 -
   0.00
                       l I  I l  J I  I T T
                                           OBS
                                           PRED L-1.
                                           PRED L—3.
                    10.0    12.5     15.0
                        HOUR, CST
17.5
20.0
Figure C-9c.  LPM time series of observed  and predicted MO hourly
             concentrations on Day  230  of 1975.
                            159

-------
    .100
       75230,  RAMS  121 AT 1300CST. START 0600
          1*11111111111
                                             I '  ' '  '
                                            OSS
                                            PRED L-1
                                            PRED L-il
                    iaO     12J    15.0     17.5
  QjQOO
                                           20.0
Figure C-9d.   LPM  time  series of observed and predicted  N0£ hourly
              concentrations on Day 230 of 1975.
       75230, RAMS 121  AT 1300CST. START 0600
   o_
   a.
   o"
   o
                                            I I  . I
                                           OBS
                                           PRED L-1-
                                        —  PRED L-i
     0
     6.0
t t  ill |  1 <  J |  i I  111  1	| 1  I 1  1 1 I  I I  1 I  I i
     7J     10.0     12.S    1S.O     17.5    20.0
                HOUR, CST
Figure C-9e.   LPM  time series of observed and predicted CO hourly
              concentrations on Day 230 of 1975.
                            160

-------
   76159. RAMS 122 AT 15QOCST. START 0500
      •WJ p*^^^^••»•^M^m
      90
                             o   o
                     ,  , I  .  , ,  ,  I ,
29      90

      KM
                                79      100
Figure C-lOa.  LPM parcel trajectory  for Day 159 of  1976.
                         161

-------
      76159, RAMS  122 AT 1500CST.   START 0500
  o.
  o.

  ul
  z
  o
  M
  O .1
   oo
             I
                  T i  I i  r r
         	 OBS

         — PRED L-1

         — PRED L-3
                      II I I  I I  I lilt I  I I  I 1  It t
     SuO     7A     10.0     1Z5     15.0     17.5     2aO
                       HOUR. CST


Figure C-lOb.  LPM time series of observed and predicted 03 hourly

              concentrations  on Day 159 of 1976.
      76159, RAMS  122 AT 1500CST.  START 0500
7J
                   10.0     12.5     15.0

                       HOUR, CST
17.5    zao
Figure C-lOc.
  LPM time series of observed and predicted NO hourly

  concentrations on Day 159 of 1976.
                           162

-------
       76159. RAMS 122 AT 1500CST.  START 0500
                                     »  i i  i i i  i i

                                          OBS
                                          PRED L-1
                                          PRED L-i
                    10.0     12.9     19.0
                       HOUR, CST
                       17.9
      20.0
Figure C-lOd.  LPM time series of observed and predicted  N02 hourly
              concentrations on Day 159  of 1976.
    76159, RAMS 122 AT 1500CST.  START 0500
                                       — 08S
                                       —  PRED L-1
                                       •—  PRED L-i
                             i  i -i i  I i  i i  i  I i  i
           7.5
10.0     12.5     15.0
    HOUR, CST
17.5
20.0
Figure C-lOe.  LPM time  series of observed and predicted CO hourly
              concentrations on Day  159 of 1976.
                          163

-------
   76160. RAMS 115 AT  1600CST. START 0500
      i w* r^T^™^^p^r^"™T^'T^^T™*T*™T^~i^^™T^™riTnnr™T"^^~
      25
                              o   0
I  ....  I
                                o
                                . I
                           i  til  I  i  i i  L
                29
        90

      KM
75      100
Figure C-lla.   LPM parcel trajectory for Day  160  of  1976.
                         164

-------
       76160,  RAMS  115 AT 1600CST.  START 0500
                      I '  ' '  ' I  ' '  '  ' I  ' '  ' '  I '  ' '  '
             OBS
             PRED L-1
       .	PRED L-3                   ^.-	


                                   /J
  O.
  CL

  ul

  O
    0.0
             7.5     iaO     12.5     15.0
                        HOUR, CST
                                      17.5
zao
Figure C-llb.   LPM  time  series of observed and predicted  03  hourly
               concentrations on Day 160 of 1976.
    .04
    .03
 £  •«
    .01
       76160, RAMS 115 AT 1600CST.   START 0500
   0.00

                                        r T^T |  T .  r

                                       	 OBS
                                       — PRED L-1.
                                       — PRED L-1
       y , ,  i 1  ,
5.0      7.5     10.0     12.5     15.0-     17J
                   HOUR, CST
                                                  20.0
Figure C-llc.   LPM time series of observed and predicted NO hourly
               concentrations on Day 160 of 1976.
                           165

-------
   .100
   .075
Q.
Q-.050
   .025
  0.000
      76160, RAMS  115 AT 1600CST.  START 0500
         i I  i I  i i  r r
                         I  I (  I I  I T
                                     	  OBS
                                     	  PRED L-1.
                                     	PRED L-i
       jit i	i j  i  i i I  i i i  i 1  i t  ii  lii i	i 1  i i  i i
                   10.0     12 J    15.0
                       HOUR, CST
17.5
20.0
Figure C-lld.  LPM time series of observed  and predicted N02 hourly
              concentrations on Day 160  of 1976.
      76160,  RAMS  115 AT 1600CST.  START 0500
                   10.0     12.5     15.0
                       HOUR, CST
17.5
20.0
Figure C-lle.   LPM  time series of observed and predicted CO hourly
               concentrations on Day 160 of 1976.
                            166

-------
   76,135. RAMS 114 AT 1500CST. START 0500
     IQOi i  i  i  i |  i  i i  i  | i vi I I  I  I I I  I
      79
      SO
      29
                I  1 J  I  t 1
                                         .  s
               29
  SO
KM
79      100
Figure  C-12a.  LPM parcel  trajectory for Day  195 of 1976.
                        167

-------
       76195, RAMS 114 AT 1500CST.  START 0500
              i
             OSS
             PRED L-1
             PRED L-3
            I  I I  I I  f I  I I  I t  1 I  I I  I !  I I  I I  1
                    10.0     12J     15.0     17.5     20.0
                        HOUR, CST

Figure C-12b.   LPM time series of observed and predicted 63 hourly
               concentrations on  Day  195  of  1976.
    .01
    .01
   0.00
       76195, RAMS 114 AT 1500CST.  START 0500
o
         1  1 1   I
                                 i     I 1  1
                                            I '  ' '
                                           OBS
                                           PRED L-1.
                                           PRED L-i
     SU3      7.3     10.0     12.3     15.0     17.5     2aO
                        HOUR, CST
Figure C-12c.  LPM time series  of  observed and predicted NO  hourly
              concentrations on Day  195 of 1976.
                            168

-------
   .100
   .079
      76195, RAMS 114 AT 1500CST.  START 0500
 I
   .080
 CN
 O
   .025
  0.000
               i  r r i |  * i  r i   i  r
                                           I j  I I  I 1

                                           OSS
                                           PRED L-1
                                           PRED L-4
                         I  t I _j ._.J_ | | I  I t  I j  I I i 1._._!_
     54     7.5      iaO     12J     15.0
                       HOUR, CST
        20.0
Figure C-12d.   LPM time  series of observed and predicted N02  hourly
               concentrations on Day 195 of 1976.
      76195, RAMS  114 AT 1500CST.  START 0500
1
o"
O
   0.0
                        r i  i i  i i  i i
                                     i  I I  I  I i  I I  r
                                        —  OBS
                                     	  PRED L-1.
                                     	PRED L-i
       ill!
            7.5     10.0     12.5     15.0
                       HOUR, CST
17.5
20.0
Figure C-12e.   LPM time series of observed and predicted CO hourly
               concentrations on Day 195 of 1976.
                            169

-------
      76211.  RAMS 120 AT 1500CST. START 0600
         • W ^^^"T^f^j^ —... ( "•—j—.y—^-—••—j—^^^—j—j—»yi—^n^^-^-
         90
o   o
                           40

                          KM
   79       100
Figure C-13a.  LPM parcel  trajectory  for Day 211 of 1976,
                           170

-------
     76211,  RAMS 120 AT  1500CST. START 0600
Q.
Q_

ul

O
M
O .1
  0.0
            OBS
            PRED L-1
            PRED L-3
                         xX~"
        -,    i      ,» i        i
      i  i »j--L-J— »-^«-y* i  I  I I  i  I I  I  I I  I
                                            I  1 I  I  I I
    5.0      7A      10.0     12.5     15.0
                        HOUR, CST
                                17.5
                                               20.0
Figure C-13b.  LPM time series of observed and predicted 03 hourly
               concentrations on Day 211 of 1976.
   .100
   .075
       76211, RAMS  120 AT 1500CST. START 0600
   -030
 o
 z
   .025
  0400
                                      ——  OBS
                                      	  PRED L-1
                                      	PRED L-i
            <   *-...»•
                                i i  i
SJO
7J&
                    10.0     1££     15.0
                        HOUR,  CST
17.5
20.0
Figure C-13c.  LPM time series of observed  and  predicted NO hourly
              concentrations on Day 211  of 1976.
                            171

-------
O.
O
   .100
   .075
   .090
   .025
      76211, RAMS 120 AT 1500CST. START 0600
  0.000
                                    T
                                           OSS
                                           PRED L-1
                                           PRED L-4
          i  i I  i i  i !  I i  i i  i I  i i  i  i I
                                               i i
5UJ      7A     10.0     1Z5     19.0     17.9
                   HOUR, CST
                                                  20.0
Figure C-13d.   LPM  time series of observed and predicted NQ2 hourly
               concentrations on Day 211 of 1976.
      76211, RAMS 120 AT  1500CST. START 0600
  CL
  Q.

  o"
  O
                                      ™~~~ OSS
                                      	  PRED L-V
                                      	PRED L-5.
     6U}     7J      10.0     12.9     15.0
                       HOUR, CST
                                      17.5
20.0
Figure C-13e.   LPM  time series of observed and predicted CO hourly
               concentrations on Day 211 of 1976.
                            172

-------
    76212. RAMS 108 AT 12QOCST.  START 0600
      lyQ .—•yni-ji (  (   j_. .._j_. j j_ •   (- ^ .j  T   (—- |  ( (
       25
                 I  '  '  ' '  I
                    .  .  .  I
                           I  I  I  I I  I  I I  T
                         50

                       KM
75      100
Figure C-14a.  LPM parcel trajectory  for Day 212 of 1976,
                        173

-------
     76212, RAMS  108 AT  1200CST. START 0600
CL
CL
 m
LU
Z
o
M
O .1
  0.0
              I  \ i  t
                     I '
           OBS
           PRED L-1
           PRED L-3
                   i  I t  i i  i  I i  i i  i  I i  i  i i  I i  i  i i
            7A     10.0     1i5     15.0
                       HOUR, CST
                                      17.5
  20.0
 Figure C-14b.  LPM  time series of observed and predicted 03 hourly
               concentrations on Day  212 of 1976.
       76212, RAMS 108 AT 1200CST. START 0600
     .02
  0.
  Q.
    .01
   0.00
                                      	  OBS
                                      —~  PRED L—1-
                                      	PRED L-4
                     1
                       1
5U>     7.3     10.0     124     15.0.
                  HOUR, CST
                                          17^
20.0
 Figure C-14c.  LPM time  series of observed and predicted MO  hourly
               concentrations on Day 212  of 1976.
                           174

-------
    .06
       76212, RAMS  108 AT 1200CST. START 0600
    .04
 0.
 a.
  •
 CN
 O
    .02
   0.00
                                            OBS
                                            PRED L-1
                                            PRED L-51
        till
               1  j j  1 L I I  I 111 t 1 l_ .1	J  I L L \  I i . I. .!_
             7.5     10.0     12.5     15.0
                        HOUR, CST
                                             2ao
Figure C-14d.   LPM time  series of observed and predicted N02 hourly
               concentrations on Day 212 of 1976.
    2JO
    1.5
       76212. RAMS 108 AT 1200CST. START 0600
 I

 o"
 O
    0.0
                                     I '  ' '  ' I  ' '  '
                                      ——  08S
                                      	   PRED  L-1.
                                      	PRED  L-i
&0
                    10.0    12.5     15.0
                        HOUR, CST
17.5
20.0
Figure C-14e.   LPM time series of observed and predicted CO hourly
               concentrations on Day 212 of 1976.
                            175

-------
   76225. RAMS 117 AT 1300CST. START 0600
      i^W "^^"f*"T*™T™'TOiri^^^"T^T^^ii^^
      90
      25
       o*-4-
            j. I I  5 I
                29
  30
KM
100
Figure C-15a.   LPM parcel  trajectory  for Day  225  of  1976.
                          176

-------
    .20
    .15


O.
Q.

ul  .10

O
M
O

    .05
      76225, RAMS 117 AT 1300CST.  START 0600
                     I  ....  I
5U>
7A
                           12J     15.0
                        HOUR, CST
17 J
zao
 Figure C-15b.  LPM time series of observed  and predicted 03 hourly
               concentrations on Day  225  of 1976.
 o_
 O*
   .020
   .015
   .010
      76225, RAMS 117 AT 1300CST. START 0600
   .005
  0.000
                                      	 OBS
                                      	  PRED L-1
                                      	PRED L-i
          x\\
          \g£> I        I       i        1^^"^
       tiriliiiiliiitiiiiilri
     5U)     7 A      10.0     12.5     15.0
                       HOUR, CST
                              17.5     20.0
 Figure C-15c.  LPM time series of observed  and  predicted NO hourly
               concentrations on Day 225  of 1976.
                             177

-------
     .04
     .03
  O.
  Q.  .02

  CM
  O
     .01
        76225, RAMS 117 AT 1300CST. START 0600
    0.001-1-1
            I T  I I  I I  i
                                       i  i i  i I  i i  i i
                                       	  OBS
                                       	   PRED L-1
                                       	PRED L-4
                         i  i i  I i  i i  i I  t i  i  i
      &0      7J     10.0     1Z5     15.0
                         HOUR, CST
                                      17.3
20.0
Figure C-15d.  LPM time series  of  observed and predicted N02 hourly
               concentrations on Day 225 of 1976.
     2JO
       76225, RAMS 117 AT 1300CST. START 0600
 o
 O
                                            OSS
                                            PRED L-1.
                                            PRED L-i
&0
                     10.0     12.5     15.0     17.5     20.0
                         HOUR. CST
Figure C-15e.   LPM  time series of observed and predicted  CO  hourly
               concentrations on Day 225 of 1976.
                            178

-------
                                 APPENDIX  D

               UAM - RS AND RO PLOTS FOR REMAINING TEST DAYS

     This appendix  contains  the residual  histogram (RS)  and residual vs.
observed plots (RO)  for the  species  N02,  03 and CO from the  UAM  statistical
analysis that were not specifically  discussed  within  the  body  of this  report.
The plots presented here  are from test days  178-75,  182-75,  183-75,  184-75,
209-75,  221-75,   230-75,  231-75, 251-75,  159-76,  160-76, 211-76, 212-76,
225-76,  226-76 and 237-76 of the RAPS data  base.
                                      179

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 178-75
        500+001
    Q.
    Q.
        .400+001

        .300+001

        .200+001
    -j   100+001
    <

    9   ooo
    c/>
    {£  -.100+001

       - 200+001

       -.300+001

       -.400+001

       -.500+001
                   5  10  15  20  25  30  35
                                 FREQUENCY
                   CARBON MONOXIDE
                                              40   45   50
               DATA AVAILABLE
                                163    DATA POSSIBLE
273
Figure D-la.   Residual  histogram for CO from the  UAM simulation
               results  for Day 178-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 178-75

2
Q.
Q.
^"'
RESIDUAL



.3UU+UUI
400 +001
.300+001
.200 -i-OOl
.100-1-001
000
- IOOt-001
- 200+001
-.300+001
-.400+001
- 500+001
•
- o
0°
0 ° 008
jji|JiSp&f-M^^C^^^" ' I ' ' '
3
•
•
-
            000
                    100+001  .200+001  .300+001   400+001   500+001
                     OBSERVED CONCENTRATION  (PPM)
                   CARBON MONOXIDE
             DATA AVAILABLE =  163    DATA POSSIBLE   =  273

Figure  D-lb.   Residual  vs. observed  plot for CO  from the UAM
               simulation results for Day 178-75.
                                180

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                              DAY  173-75

        750-001
    3
    Q
    on
 .500-001


 450-001


 .300-001 H


 .150-001 j-g;


 .000


- 150-001


-.300-001


-.450-001


-.600-001


- 750-001
                   5   10   15   20   25   30   35
                                 FREQUENCY
                   NITROGEN DIOXIDE
                                       40  45  50
                DATA AVAILABLE =  224    DATA POSSIBLE
                                                  273
Figure  D-2a.  Residual  histogram for N02  from the  UAM simulation
               results for Day  178-75.
                  RESIDUAL VS OBSERVED  CONCENTRATION

                               DAY  178-75
    Q_
    Q.
    Q
                                                            LEGEND

                                                           FREQ STM

                                                         < 1, 1> 0
                                                         < 2. 2> O
                                                         < J, 3> *
                    .150-001  .300-001  .450-001  600-001   750-001
                      OBSERVED CONCENTRATION  (PPM)

                   NITROGEN DIOXIDE

             DATA AVAILABLE =  224   DATA POSSIBLE   =  273


Figure  D-2b.  Residual  vs. observed plot  for N02 from the UAM

               simulation  results  for Day  178-75.

                                 181

-------
     RESIDUAL HISTOGRAM  DETERMINED  OVER ALL TIMES AND LOCATIONS
                              DAY 178-75
        loo+ooo p—-

        .aoo-oo i
        600-001
        .200-001


        .000
    Q.
    Q.
g

9
C/l
g  -200-001

   -400-001

   -.600-001

   -.800-001

   - 100+000

                                      -t—
               H*
0   5  10  15


    OZONE

DATA AVAILABLE
                          20  25  JO  35
                             FREQUENCY
                                              40  45  50
                                203    DATA POSSIBLE
                                                     273
Figure D-3a.   Residual  histogram for  03 from the  UAM simulation
               results  for Day 178-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY  178-75
    5
    Q.
     C/)
     LJ
     o;
                    .500-001  .100+000  .150+000  200+000  250+000
                      OBSERVED CONCENTRATION  (PPM)

                   OZONE

              DATA AVAILABLE =  203    DATA POSSIBLE  =  273


Figure  D-3b.  Residual  vs. observed plot for 03  from the UAM
               simulation results  for Day 178-75.
                                182

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 132-75

D.
Q.
~
_j
Q
W
Ul
or



400+001
300+001
.200 rOC 1
.100+001
000
-.100+001
- 200+001
-.300+001
-.400.+00'
- 500+001
.
•
•
L


•
•
•
•
   50  100  150  200 250 300 350  400  *50 500
                 FREQUENCY
   CARBON  MONOXIDE
DATA AVAILABLE  =  221    DATA POSSIBLE
                                                       273
Figure D-4a.  Residual  histogram for CO from the UAM simulation
              results  for Day 182-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 182-75


G.
Q_
^
_J

C/}
UJ
or







.400+001

300+001

.200+001
.100+001
000
-.100+001

— ^yu+uu

- 300+001
-.400+001
- 5QQ+QQ1
a
a


O 03
T a 8 °
__90MMT_ fl a , i ,
^rr"



•
•
i i i







LcCENO
fT»EQ SYM
< i. t> n
< 2. 2> »
< J. 3> *


.000 .100+001 .200+001 .300+001 400+001 500+001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE = 221    DATA POSSIBLE  =  273

Figure D-4b.  Residual  vs.  observed plot for CO from the  UAM
              simulation results for Day 182-75.
                               183

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 182-75
        .750-001
        600-001 U
    Q.
    Q.
    9
    en
                              *0  50  60  7Q
                                FREQUENCY
                  NITROGEN DIOXIDE
90  90  100
               DATA AVAILABLE =  213   DATA POSSIBLE  =  273

Figure D-5a.   Residual  histogram  for NO^ from the  UAM simulation
               results for Day 182-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY  182-75

v-s.
a.
0.
RESIDUAL (


.600-001
>50-001
.300-001
.150-001
000
- 150-001
- 300-001
-.450-001
-.600-001
- 750-001
000

a
' a a i? ™, a° °
[•ijffl^T M a a ° a ° q
.sp°" " '°°'°° '
:*
•


uecENp
FREO SYM
< i, i> a
< 2. 2> 0
< 3. 1> *
< *,99> 9

200-001 .400-001 .600-001 800-001 100+000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN  DIOXIDE
              DATA AVAILABLE =  213    DATA POSSIBLE  =  273

Figure D-5b.   Residual  vs.  observed plot  for N02 from the  UAM
               simulation results for Day  182-75.
                               184

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                              DAY  182-75
    2
    Q.
    Q.
    Q
    u5
    ui
    or
 600-001


 .450-001


 .300-001


 .150-001


 000


-.150-001


-.300-001


-.450-001


-.600-001


- 750-001
*
8
•»
,
"1
tit
5*r
£»*«<
t
1
£
t*
1
Eji
i.4
:.
.
'S5JS***** "
: T - **•
: 1 1 : i ff?f r***7***v* i i i i
• *• - ??*?** 1 ! ! 1 1
lfffl!!t::::r
•»*
^

Figure  D-6a.
               10  15  20  25  30  35  40  45  50
                          FREQUENCY

           OZONE

        DATA AVAILABLE =  249   DATA POSSIBLE  =  273


        Residual histogram  for  03  from the  UAM simulation

        results  for  Day  182-75.
                  RESIDUAL VS OBSERVED CONCENTRATION

                              DAY  182-75
    Q.
    Q.
    9
    CO
    LJ
    o;
        - 750-001
            000      .500-001  .100-1-000  .150+000  200+000  250-t-OOO

                      OBSERVED CONCENTRATION   (PPM)

                   OZONE

             DATA AVAILABLE  =  249    DATA POSSIBLE  =  273


Figure  D-6b.   Residual  vs. observed plot for 03 from  the UAM

               simulation  results  for Day 182-75.
                                185

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                            DAY 183-75

s
CL
CL
SIDUAL
Ul
ce



3UU-I-UU 1
400+001
.300+001
.200+001
.100+001
000
-.100+001
-.200+001
-.300+001
-.400+001
- 500+00 1
.
L
fr ,
SBS"^ m*i~maL**L*J~LA" *


•
•
                  10  20  20   40  50  60  70
                               FREQUENCY
                  CARBON MONOXIDE
ao  so  too
               DATA AVAILABLE =  213   DATA POSSIBLE
          273
Figure D-7a.  Residual  histogram for CO from the UAM simulation
              results  for  Day  183-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 183-75


2
CL
O._

_l
O

UJ
a:







.3UU-MJU 1
400+001
300+001

.200+001
.100+001

onn
wuu
- 100+001
- 200+001


-.300+001
- 400+001
- SDQ+QO!
, . i . i i i i i
•
O
o
a o en o °
,„<§& *o %0|f
	 liiMjfy^ Q ^^TJT Q

p"^*"8




•
1 1 t 1 i • 1 I 1








LEGEND
FREO STM
< l. i> O
< 2. 2> 0
< 3. J> *
< *,99> «


.000 .100+001 .200+001 .300+001 400+001 500+001
                    OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  213   DATA POSSIBLE  =  273

Figure D-7b.   Residual  vs.  observed plot for CO from  the UAM
               simulation results for Day 183-75.
                               186

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY  183-75
.600-001
^ .450-001
a
"•" .300-001
^ .150-001
a
ig -.150-001
-.300-001
-.450-001
-.600-001
- 750-001
.
•*»
!$.
|| p^:::: r , ,
• '-u«- i **»
•*
4
*
                      10   15   20  25  30  35
                                FREQUENCY
                  NITROGEN DIOXIDE
40
    45   50
               DATA AVAILABLE =  221    DATA POSSIBLE  =  273

Figure D-8a.   Residual histogram for N02 from the  UAM  simulation
               results for Day  183-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 133-75


^^v
2
a.
Q.
~
_i
z>
Q
CO
UJ
o:






. / JV— UU 1
600-001
.450-001

.300-001
.150-001
000


n o
rS 0°
a-,- %
~3EffiK~f£* QCCCr"^ •• J3
3nrjflpCj cj u Q ^r

fflSEzTy d 	 	 ' ' '
|4£^*~^on^j ~ o DB
- 150-001
-.300-001


-.450-001

-.600-001
- 750-001
Pa
TJJ
a
p

n
I i . i , i i i








LEGEND
FREQ SYM
< i. i> n
< 2. 2> 0
< 3. i> *
< ».99> »


            000     .500-001  .100+000  .150+000 .200+000  250+000
                     OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE  =  221    DATA POSSIBLE  =  273

Figure D-8b.   Residual  vs. observed  plot for N02 from the  UAM
               simulation results  for Day 183-75.

                               187

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 183-75

0,
RESIDUAL



800-001 4
.600-001 •
.400-001 -
.200-001 -
000
-.200-001 -
-.+00-001 •
-.500-001 •
-800-001 •
- 100+000 -
t
t*
*if%i" i i i i ii
***i iili !*********



                      10  15

                   OZONE

                DATA AVAILABLE
20   25   30   35
   FREQUENCY
40   +5   50
  200    DATA POSSIBLE
           273
Figure D-9a.   Residual  histogram  for  03 from the UAM simulation
               results for Day 183-75.
     Q
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 133-75
        - 100+000
             000     500-001  .100+000 .150+000  .200+000  250+000
                     OBSERVED CONCENTRATION  (PPM)
                   OZONE
              DATA AVAILABLE =  200    DATA POSSIBLE  =  273

Figure D-9b.   Residual  vs. observed plot  for 63 from the  UAM
               simulation results  for Day  183-75.

                               188

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 184-75
    Q.
    Q.
    <

    9
    CO
                  10   20   30
                                            80  90  100
Figure D-lOa.
              40  50   60   70
                 FREQUENCY
   CARBON  MONOXIDE
DATA AVAILABLE =  223    DATA POSSIBLE
                                                       273
Residual  histogram for CO from the DAM simulation
results  for  Day  184-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 184-75

a
a
*~s
-j
9
Uj
a:

SOO+UUI
400+001
300+001
.200+001
.100+001
000
- 100+001
- 300+001
- 400+001
-500+001
000
a
300
^^j^jBcDD "o^ «. ^i Q D
JpP^ a1 °, aa
a
-



lECCNO
FBEO SYM
< i. i> a
< 2. 2> 0

.150+001 .300+001 .450+001 .600+001 750+001
                     OBSERVED CONCENTRATION (PPM)
                  CARBON  MONOXIDE
             DATA AVAILABLE: =  223    DATA POSSIBLE  =  273

Figure D-lOb.  Residual  vs.  observed plot for CO from the UAM
               simulation results  for Day 184-75.
                              189

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             QAY 184-75

^^\
3
0_
•— '
g
00
LJ
or



^JU-*" um-i
.200+000
.150+000
.100+000
500-001
000
-.500-001
- 100+000
-.150+000
-.200+000
- 250+000

•
A^

^


•
                  10  20   JO   *0  50  60  70
                                FREQUENCY
                  NITROGEN DIOXIDE
             ao  90  100
               DATA AVAILABLE
239   DATA POSSIBLE = 273
Figure D-lla.   Residual  histogram for N02 from  the UAM simulation
                results  for Day 184-75.
                 RESIDUAL VS  OBSERVED CONCENTRATION
                             DAY 184-75
^*v
Q.
OL
1 —

_i
3
00
UJ
or






^3U+UUU
•200+000
150+000

.100+000

.500-001
OCX"!
.


a a
a 30
- aj8M.g,§°baB *
	 iarfr?!i?j'3. 	

-.500-001
-.lOu+OOO

-.150+000
- 200+000
- JSQ-'-QQQ
*


•

,








L£C£NO
FREO STM
< 1. l> O
< 2. 2> O
< *:»> *


000 500-001 100+000 .150+000 .200+000 250+000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
             DATA AVAILABLE =  239    DATA POSSIBLE  =  273
Figure D-llb.   Residual  vs.  observed plot  for  MOg  from the UAM
                simulation results for Day  184-75.
                               190

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 184-75

2j
0^

^
D

£






^uufuuu r
.800-001
.500-001

.400-001
.200-001
000

-.200-001
-.400-001

-.600-001
-.300-001
- 100+000

-»
• H
liili3!
:1|*|:«!
Hi i
tn:;:i:
|||n£:t
- •?***
i|in
< *
' -••»»•«
:*"
(
3 5




!*.««..*
: !?«!(•*
tiiS<<*
j*« '



















10 15 20 25 20 35 40 45 50
FREQUENCY
OZONE
DATA AVAILABLE = 229 DATA POSSIBLE = 273
Figure
D-12a. Residual histogram for 03 from the DAM simulation
results for Day 184-75.
RESIDUAL VS OBSERVED CONCENTRATION



CL
^

g
Q
i7>
OS






100-t-OOO r


aoo-ooi [
500-001 •
.400-001

200-001
000 >




i |
1° °
- 200-001 5a°
fen °
— AOfl — OO 1 J *ftt» m m n*
*vU Jw 1 '

-.600-001

DAY 184-75

• i i i i i i i
•
o
B-jg&S1 9 M°
..^ESffi*1' M ^ °
'•^^^^FlTfi g
a ^__ jJTOgrr Sri
1 i '.i ' • •rt-^TTTrP*' ' i
_ Jj "rt *^j
n ^r ri"
g I^T3 CT
a^a0 a
cn a o
r ^
OBO^
. o a
-800-001 ^f
w


OC Q •








LECENO
FREO SYM
< i, i> n
< 2. 2> O

< 4,99> *

000 500-001 .100-1-000 .150-t-OOO 200+000 250-1-000
                     OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  229   DATA POSSIBLE  =  273

Figure D-12b.  Residual  vs.  observed plot for 03 from the UAM
               simulation  results for Day 184-75.

                              191

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 209-75

^" X
3
Q.
— '
RESIDUAL



^3U+UU^
.200+002
.150+002
.lOO-t-002
.500+001
.000
-.500+001
-.100+002
-.150+002
-.200+002
-.250+002
.
»
-




•
                                FREQUENCY
                   CARBON MONOXIDE
               DATA AVAILABLE =  230   DATA POSSIBLE  =  273

Figure D-13a.  Residual  histogram for CO  from the  UAM simulation
               results for Day 209-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 209-75
    s
    CL
    0-
    <
    9
    c/i
    UJ
    or
.43V -rvu*
.200+002
.150+002
.100+002
.500+001
000
-.500+001
-.100+002
-.150+002
-.200+002
- 250+002
.
a








LEGEND
FREO SYM
< i, i> a
< 2, 2> 0
< J, 3> *
< 4,99> »

            .0
                    500+001  .100+002  .150+002  200+002  250+002
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  230    DATA POSSIBLE  =  273

Figure  D-13b.   Residual  vs.  observed plot for CO from the  UAM
                simulation  results for Day 209-75.
                               192

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY  209-75
. JVAJ— VU i
.400-001
£L .300-001
a
"-" .200-001

^ .100-001
Q 000
(75
^ -.100-001
-.200-001

-.300-001
-.400-001
-.500-001




< j
•;
; ',
. ",
~ *
' <
*


.

.
t.
• !****

' .i : ' i tiiitHi • • « 1 1 • • i
" '< ' >• T9*9*49$*
' \ ", t *••••-*»»•»»
i < • >. k9
»»**
I
•
-
                      10   15   20   25   30  35
                                FREQUENCY
                  NITROGEN DIOXIDE
40   45   50
               DATA AVAILABLE -  225    DATA POSSIBLE  =  273

Figure D-14a.   Residual histogram for NC>2 from  the UAM simulation
                results for  Day  209-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 209-75
    a
    a
     a
    ^  -.100-001
            .000     .150-001  .300-001  .450-001  .600-001  .750-001
                      OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
             DATA AVAILABLE  =  225   DATA POSSIBLE  =  273

Figure D-14b.   Residual vs.  observed plot for N02  from the UAM
                simulation  results for Day 209-75.
                               193

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS

                             DAY 209-75
    2
    Q.
    Q.
    3
    Q

    (/)
    Ld
    or
^3w»-umj r
.200+000
.150+000
.100+000
.500-001
000
-,500-001
-.100+000
-.150+000
-.200+000
- 250+000
.
•
" 1 *
^Hm^jg.j™.
• t; ; • 'Jlil ijllij tjj**»*****tl***»i*»iiJ»*ti»»J* ' '
.

•

3 5 10 15 20 25 JO 35 40 45 5
FREQUENCY
OZONE
DATA AVAILABLE = 218 DATA POSSIBLE =
                                                        273


Figure D-15a.   Residual histogram  for 03 from the (JAM  simulation

                results for Day 209-75.
                 RESIDUAL VS OBSERVED CONCENTRATION

                              DAY 209-75

•5.
Q.
Q.
>~'
i/5
UJ
or

.200+000
150+000
.100+000
.500-001
000
-.500-001
-.150+000
-.200+000
- 250+000
.
•
° jr>0 o
&"n_'a ° °
^ qt^^hcffifoiSi o o °
—^^— — — ^— iTO I .I3IDB .^H^BdD LIII
BI "TnoD^^^






LEGEND
FREX! STM
< i. i> a
< 2. 2> 0
< 3. J> *
< 4,39> «

            .000     .500-001  .100+000  .150+000  .200+000  .250+000
                     OBSERVED CONCENTRATION  (PPM)

                   OZONE

             DATA AVAILABLE =  218    DATA POSSIBLE  =  273


Figure  D-15b.   Residual vs.  observed plot for 03  from the UAM

                simulation results for Day 209-75.
                               194

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 221-75
        .750*001
    2
    Q.
    Q.
                  10   20   30   40  50  60  70
                                FREQUENCY
                  CARBON MONOXIDE
                                ao
                                    90  100
               DATA AVAILABLE =  222   DATA POSSIBLE  =  273

Figure D-16a.   Residual  histogram for CO  from  the  (JAM simulation
                results for Day 221-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 221-75

'"•X
s
Q.
Q.
~
RESIDUAL

.750+001
.500+001
.450+001
.300+001
.150+001
000
- 150+001
- 300+001
-.450+001
-.600+001
-.750+001
000


a
a
lflg°B 0
'wBJ$^°i 8 i °i , , , , '
Q

.150+001 .300+001 .450+001 600+001 .750^



LEGEND
FREQ SYM
< 1, l> HI
< 2. 2> 0
< 3. 3> A
< *,99> 0
^001
Figure D-16b.
        OBSERVED CONCENTRATION  (PPM)
     CARBON  MONOXIDE
DATA AVAILABLE =  222    DATA POSSIBLE  = 273

  Residual vs.  observed plot for CO  from the  UAM
  simulation results  for Day 221-75.

                  195

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 221-75
3UU-UU 1
.400-001
./— X
5J .300-001
0.
^ .200-001
gj .100-001
Q .000
C/l
^ - 100-001
-.200-001
-.300-001
-.400-001
-.500-001
I
»»*
SI"
i**
ljjlfu:t ,",,.'
iffpll. UWUtI*"
••
»*
«*
                      10   15  20  25   30   35  40  45  50
                                FREQUENCY
                  NITROGEN DIOXIDE
               DATA AVAILABLE »   222    DATA POSSIBLE  =  273

Figure D-17a.   Residual  histogram for N02  from the UAM simulation
                results for Day 221-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 221-75

S-*
0.
0.
*-s
_J
^
Q
C/l
LU
cr







.JUU— W 1
.400-001
300-001
.200-001
.100-001
.000

-.100-001
-.200-001

-.300-001

-.400-001
- 500-001

D fl.
B a a
ir°aaa *
• ^\g*i***\^Jr
| .jft'f'^ 'Jl'?"\ pi n ] I D ' 31
9pc nu^lj^ |Tr>^ a o ' ;u
-K|k a°o° a
n 3 jr
Q
a a
a
CD
II ill







LECENO
FREQ SYU
< i. i> a
< 2. 2> 0

< +!99> «


000 .150-001 .300-001 .450-001 500-001 750-001
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
             DATA AVAILABLE -  222   DATA POSSIBLE  =  273

Figure  D-17b.   Residual vs. observed plot for N02 from  the  UAM
                simulation results  for Day 221-75.
                               196

-------
      RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 221-75
.ASW^VVW
.200*000
^ .150*000
Q.
~ .100+000
J .500-001
Q 000
Co
£ -500-001
-.100*000
-.150*000 •
-.200+000 •
- 250*000 •
'*•**•*****• ..



.
.

r*
I1 f3»f3ir"
iljSf jtxffijjtjs**11***1'**"********

•



>
5 10 15 20 25 JO .55 *0 45 5
FREQUENCY
OZONE
DATA AVAILABLE = 239 DATA POSSIBLE =
                                                      273
Figure D-18a.   Residual  histogram for 03 from the UAM simulation
                results  for Day 221-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 221-75

^^
2
0.
a
**~*s
3
5

t/5
Ul
ac





.^3U*VAAJ
200*000

.150*000

.100*000
.500-001
.

•

B n<*>
crp o —
'•••••^ *? rfr'fiip'ri !'i (*r m
rV»*Wp-^m^fiL«J^J g Q
nnn Ln n^h awfiUWi*T V7F n I i I
SSjjfSff'^ llj
-.500-001
-.100+000

-.150+000
-.200+000
- 250+000

•



< i . . i , i i i i









LEGEND
FREQ SYM
< 1. !> 0
< 2, 2> 0
< 3. 3> »
< +,99> »


                    OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  239   DATA POSSIBLE  =  273
Figure D-18b.  Residual vs. observed  plot for 03 from the UAM
               simulation results  for Day 221-75.
                             197

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 230-75
        .250-1-001
    0.
    Q.
    Q
    00
                  10   20   30
                                             30  90  100
                             *0  50  60   70
                                FREQUENCY
                  CARBON  MONOXIDE

               DATA AVAILABLE =   219   DATA POSSIBLE
                                                       273
Figure D-19a.   Residual  histogram for CO from  the  UAM simulation
                results  for Day 230-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 230-75
RESIDUAL (PPM)
.^au-*uu >
.200 1-001
,150-t-OOI
.100^-001
.500-t-OOO
000
-.500+000
- 100+001
-.150 -i-OOl
-.200+001
-.250+001
.000
Q
O
„ '.
"'^,1, 0=

iffr ° iii ii
3 an
9
LEGEND
FREO SYM
< i, i> a
< 2. 2> O
< 3, 3> *
< *,99> »
.100+001 .200+001 .300+001 400+001 500+001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON  MONOXIDE
             DATA AVAILABLE =  219    DATA POSSIBLE  =  273
Figure D-19b.
                Residual  vs.  observed plot for CO  from  the UAM
                simulation results for Day 230-75.           -
                              198

-------
      RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              CAY 230-75
auu — vu i r
.400-001 •
^^\
5| .300-001 •
Q.
^ .200-001 •
J .100-001 -
Qnnn
'UUU •
(75
y - 100-001 -

-.200-001 •
-.300-001 •
-.400-001 •
-.500-001 L
^
.

•* J2^* ^

tesil*i^ii•»«Il»••l * "* '
IH*****************
• • *•*•••••
!«•»
*
•
                       10   15   20   25   30  35
                                 FREQUENCY
                   NITROGEN DIOXIDE
40   45   50
                DATA AVAILABLE =  231    DATA POSSIBLE  =  273

Figure  D-20a.   Residual  histogram for N02  from the UAM  simulation
                results  for Day 230-75.
                  RESIDUAL VS OBSERVED  CONCENTRATION
                              DAY 230-75
    o.
    Q.
    Q
    CD
    UJ
    cc
        - 500-001
            .000      200-001  .400-001  .600-001  .300-001  100-^000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
             DATA AVAILABLE =  231    DATA POSSIBLE  =  273

Figure D-20b.   Residual  vs.  observed plot  for N02  from the  UAM
                simulation  results for Day  230-75.
                               199

-------
    RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 230-75
. IUU+UUU
.800-001
-^
^ .600-00?
Q.
^ .400-00?
-j .200-001
QflfYl
00
UJ _ 200-001
- 400-001
-.600-001
-.800-001
-.100+000

**
.
^
*••
" f
: HI Utg«;*|:**
- " " * " '??*' '??^??*?T?T^S**
" " '2' ^.^
. ^ .. ,{)
** *
*

                      10   15  20  25   .30   35
                                FREQUENCY
                  OZONE

               DATA AVAILABLE =  204
       40
          45  50
DATA POSSIBLE
273
Figure D-21a.  Residual  histogram for 03 from the  UAM  simulation
               results  for Day 230-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 230-75


^-s
0.
Q.
N— '

Q
on
UJ
o:





. lUU-fUUU
800-001

.600-001

.400-001

.200-001
000
-.200-001
-.400-001

-.600-001
-.800-001
_ lOO-"-000

.
aaa
•
a
a OB
a a g
— 3
- Jn a _
(3 SB ~ a
Sw'feS^0 °a
Q Q ^*] Q
a

•










l£C£NO
FREO STM
< t. i> a
< 2. 2> 0
< 3. 3> *


000 500-001 .100+000 .15a+000 200+000 250+000
                     OBSERVED CONCENTRATION (PPM)
                  OZONE
             DATA AVAILABLE =  204    DATA POSSIBLE  =  273

Figure D-215.   Residual  vs.  observed plot for 03 from the UAM
                simulation results for Day 230-75.

                              200

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 23' -75

:s
o_
^
RESIDUAL



JUU+VU 1
.400+001
.300+001
.200+001
.100+001
.000
- 100+001
-200+001
-.300+001
-.400+001
-.500+001
(
t
I
^ 	
F" 	



                             40   50  60  70
                               FREQUENCY
                  CARBON MONOXIDE
30
               DATA AVAILABLE =  205   DATA POSSIBLE
          273
Figure D-22a.  Residual  histogram for CO from the UAM  simulation
               results  for Day 231-75.
    Q.
    a
    (Sl
    UJ
    or
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 231-75
.400+001
300+001
.200+001
.100+001
000
-.100+001
- 200+001
- 300+001
-.400+001
-.500+001
300

a
a a
__rip*? ,",,,, , "
f^a .
•
•
100+001 .200+001 .300+001 *00+001 .500


LEGEND
FREQ SYM
< i. i> a
< 2. 2> 0
< 3. 3> *
t-001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  205   DATA POSSIBLE  =  273

Figure D-22b.  Residual  vs.  observed plot for CO from the  UAM
               simulation  results for Day 231-75.
                              201

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND  LOCATIONS
                             QAY 231-75
     on
3UU— «U 1
.400-001
.300-001
.200-001
100-001

000
-.100-001
-.200-001

-.300-001
-.400-001
-.500-001
.
.
t
'. !$„.»
" 3S3»«*
i i i izi ", 3* : :t***
; . 'i >lj .. .« J^ 'J >«^f £4 J. J. ^ L , 1
: ii :?i :: i: : :i JKI J|I jn|****'**'** "
: : i SJ IsSl5J**'*w
Ht*tt*
cn

•
.,,,,,,,,
                                 25  30  35
                                FREQUENCY
                   NITROGEN DIOXIDE
               DATA AVAILABLE =  248
                                             40   45  50
                                      DATA POSSIBLE  -  273
Figure D-23a.   Residual  histogram for N02 from  the  UAM simulation
                results  for Day 231-75.
                 RESIDUAL VS OBSERVED  CONCENTRATION
                             DAY 231-75
RESIDUAL (PPM)
400-001
.300-001
.200-001
.100-001
.000
-.100-001
-.200-001
-.300-001
-.400-001
-.500-001
000
«5>°°a"aJ =>

fl i1^1 LL L3 .!• »i taj ^ ^(-W ™ ,«. A [ •• i Q i
IR % 5 ^^r^ff1™ ° a
• (HBTin o I'B^^I n
v^D uD jn Q 	 OQ
y Q OHM L^L Q Q
" oo a 0
UECCNO
FRCO STM
< i. i> a
< 2, 2> O
< 3, 3> A
< +,99> «
.150-001 .300-001 .450-001 600-001 750-001
                     OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE =  2*8    DATA POSSIBLE  =  273

Figure D-23b.   Residual  vs. observed plot for  M02  from the UAM
                simulation results for Day 231-75.            :

                              202

-------
      RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 231-75
lUUfUUW -
.aoo-ooi \-
^ 600-001 -
Q.
^ .400-001 -
J .200-001 -
Q .000
to
J£ -.200-001 -
- 400-001 -
-.600-001
-.800-001 •
- 1 00+000 1
»
.
»f

| tit t**
: II 1: : iii Hi if Ui.«..l ' '
: ! : |||*it£»t»*{S*$ii$ •
. ||j«*<-»>
•
•
                       10  15

                   OZONE

                DATA AVAILABLE
20   25   30  35
  FREQUENCY
                                             40   *5   50
  246    DATA POSSIBLE
273
Figure D-24a.  Residual  histogram for 63 from the  UAM  simulation
               results  for Day 231-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 231-75

r— X
Q.
Q.
^
RESIDUAL

. lUU^VUU
.800^-001
.600-001
.400-001
.200-001
.000
-.200-001
- 400-001
-.600-001
-.800-001
-.100 +000
.000
a
•'. 1.
=B00 °a a o ° B
" Q Q Q Q QQ^_ Q Q
^W^H Q
(3D M 0 	 m'l'l Q, _
fflfci]rT_~fff jfBL n I 1 ' l
H^^^^t^ 'i 'i T* ' *" rfffji^n 1*1 ' ' iii
^B1 ri ff Wfp *^p' —
3X.O a offi^ffl
o Wy 3




LECENO
FTO3 SYM
< l. i> O
< 2. 2> 0
< 3. J> *
< 4,99> «

500-001 .100+000 .150+000 200+000 250+000
                     OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  246    DATA POSSIBLE  =  273

Figure D-24b.  Residual  vs.  observed plot for 03  from  the  UAM
               simulation results for Day 231-75.
                               203

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 251-75

Q.
O.
s"-l'
SI DUAL
Ld
or



./3U+UUI
oCOrOOI
450+001
.300+001
.150+001
.000
-.150+001
-.300+001
- 450+001
-.600+001
-.750+001

-
}
|=~ 	
SfiP88BBB^rr '
* .
•

•
                  10  20   30   40  50  60  70   80   90  100
                                FREQUENCY
                  CARBON MONOXIDE
               DATA AVAILABLE =  204   DATA POSSIBLE  =  273

Figure D-25a.  Residual  histogram for CO from  the  UAM simulation
               results  for Day 251-75.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 251-75
     LJ
' SOD-DO ^ (- °
A50+OC1 -
300+00' - OS °
• c. -^ >•->•• M> ~3.^3Q
"5« *vO " «•«. JMH 3ST H
- 300*001 j-
- 450 -CO' -
-60C-DO' f
lECENO
FREO SYU
< i i> n
< 2. 2> C
< J. J> *
< *,99> »
000 200+001 400-rOQI 500-001 300+00! '00+002
                     OBSERVED CONCENTRATION
     CARBON MONOXIDE
3ATA AVAILABLE =  204.
                                    QATA POSSIBLE  =  273
Figure D-25b.
  Residual  vs.  observed plot for CO  from  the UAM
  simulation  results for Day 251-75.
                              204

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 251-75
         250+000
     S
     CL
     w  -.500-001
                      20  .30  40   50   60  ~ 70
                                 FREQUENCY
                   NITROGEN  DIOXIDE
30   90  100
                DATA AVAILABLE =  215    DATA POSSIBLE
          273
Figure D-26a.   Residual histogram  for N02 from the UAM  simulation
                results for Day 251-75.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 251-75

5
Q.
Q.
RESIDUAL


^3U+UUU
200+000
.150+000
.100+000
.500-001
000
-500-001
-.100+000
-.150+000
-.200+000
- 250+000
000
.
•
0
fffl^0 ,
;



LEGEND
FPEQ STM
< I, !> D
< 2. 2> 0
< 3. 3> *
< 4.99> »

.500-001 .100+000 .150+000 200+000 250+000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN  DIOXIDE
             DATA AVAILABLE =  215   DATA POSSIBLE  = 273

Figure D-26b.   Residual  vs. observed  plot  for N02 from the UAM
                simulation results for Day  251-75.
                               205

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 251-75
    a.
    a.
    9
    CO
J
600-001 •„
.450-001 - :
.300-001 \-'
. i
.150-001 -;
000 - *
.
-.150-001 . :
-.300-001 4
-450-001 •
-.500-001 •*
»
"- 750-001 L
.
lift"
til!!!?


t»»


                      10   15
                   OZONE
               DATA AVAILABLE
                              20  25  30  35
                                FREQUENCY
               40   45   50
  238   DATA POSSIBLE =  273
Figure D-27a.   Residual  histogram for 03 from  the  UAM  simulation
                results  for Day 251-75.
                  RESiDUAL VS OBSERVED CONCENTRATION
                             DAY 251-75
     9
     00
1 3U -UU 1 | . . II
1 if
600 -00' - 3d
450-00' - ',, S^c -_3
___ _ £J ^^8ff Q
3CO-00 h _ JHu JJ ^m
5Q"0°'pc 3Q 3™ a
- JOO-OOl L ~lyp a
- 450-CC1 j-
-6oo-:c'- "
7=Q 00' i ' ' ' - , i
t£C£NO
P9£0 SYM
< '. ! > 2
< 2. 2> C
< 3. 3> *
000 500-00' 'X-t-000 '50+000 200-000 250+000
                   OZONE

              DA"A AVAILABLE
238
OA^A POSSIBLE  =  273
Figure D-27b.   Residual  vs. observed plot  for  03 from the UAM
                simulation results for Day  251-75.
                               206

-------
     RESIDUAL ^ISTOGRAM  DETERMINED OVER ALL ~M£S AND LOCATIONS
                             DAY 159-75

0.
Q.
^~'
SIDUAL
_j
a:



4CO 1-002 j
300+002
.200+002
100*002
000
- 100+002
- 200+002
-.300+002
- 400+002
- 500+002
•
•
*» 1 1 ' L 1 1


•

•
                  50  100 150  200  250 300 350' 400  450  500
                                FREQUENCY
                  CARBON MONOXIDE
               DATA AVAILABLE  =  259   DATA POSSIBLE  =  273

Figure D-28a.  Residual  histogram for CO from  the  UAM simulation
               results  for Day 159-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 159-76

CL
'-'
RESIDUAL


.400+002
300+002
.200+002
.100+002
.000
-.100+002
- 200+002
- 300+002
-.400+002
- 500+002
.000
0

•







LECENO
FREO STM
< i. i> a
< 2. 2> ffl

.100+002 .200+002 .300+002 400+002 .500+002
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE  "
             DATA AVAILABLE = 259    DATA POSSIBLE  =  273
Figure D-28b.
Residual vs. observed  plot for CO from the UAM
simulation results  for Day 159-76.
               207

-------
RESIDUAL HISTOGRAM  DETERMINED OVER ALL PMES AND LOCAPQNS
                        DAY  159-75
/ 3U — UU ' f
SOO-001 •
^ .450-001 -
a
"" 300-00! -

^ 150-001 -
Q 000
00
•£ -.150-001 L

r

.
:!$•*
' :: ill Unfit
its:: Iff !:l|H:::i:::::: ni: .» .
»:«:;«::«*«"*"•' ^ '
3«5?i*
r:.»
:j*.
-300-001 L*
t
-.450-001 1-
-.60C-001 •
" _ 750-001 -
••

i i i i













0 5 10 15 20 25 30 35 40 45 50
FREQUENCY
NITROGEN DIOXIDE
DATA AVAILABLE = 243 DATA POSSIBLE = 273
Figure D-29a. Residual histogram for N02 from the UAM simulation
results for Day 159-76.
RESIDUAL VS OBSERVED CONCENTRATION
DAY 159-76

' 600-001 |- -j
"5 a _ a
£ 450-001 t- o °
Q.
"•" .300-001 •

a
" n a
mm u u o 3
	 3C ffi o _ a 30_
J .150-001 • Jgffi " 0 0 u a ^

on WBT^a' a*"0 %i
w _ 150-001 §
- 300-001 •


- 450-001 •

-.600-001 •
cT °
a

o a
a










LECENO
"^ STM
< i. i> 0
< 2. 2> O
< J. 3> *
< 4,99> «

       000     .500-001  .100+000  .150-1-000  200+000  250+000
                OBSERVED CONCENTRATION  (PPM)
              NITROGEN DIOXIDE
        DATA AVAILABLE =  243   DATA POSSIBLE  =  273
Figure D-29b.
          Residual  vs.  observed  plot for N02 from  the UAM
          simulation results for Day 159-76.
                          208

-------
      RESIDUAL HISTOGRAM  DETERMINED OVER AL:_ TiMES AND LOCATIONS
                              OAY '59-75

2
c.
CL
^
<
Q
on
La
a:




.200+000 j-
150+000
.100+000
.500-001
000
- 500-001
- 100+000

- 150+0001-


H!f!!!!!f««..,


*^
*
»
- 200+000 1-
       10   15

   OZONE
DATA AVAILABLE
20   25   20  35
  FREQUENCY
                                              40   *5   50
                                 234   DATA POSSIBLE
                          273
Figure D-30a.   Residual histogram for 63 from  the UAM simulation
                results for  Day  159-76.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 159-76

CL
CL
v*~ '

|
£





.200+000
150+000

.


.100+000 1-
i
.500 -C0 1
000
- 500-00!
' _ °ff ^POQ m
XpCDi pi? .gfip^n ^' ,^ i5»i^.'iij)fi'>ifc'i^' n "n
a a 30 _°a
D ° ~ °
1 00+000 ^

-.150+000
- 200+000

3
•







FREO SYM
< i. i> a
< 2. 2> O
< 3. J> *

             000      500-001  .100+000 .150+000  200+000  250'+000
                      OBSERVED CONCENTRATION   (PPM)
                   OZONE
              DATA AVAILABLE =  234    DATA POSSIBLE  =  273

Figure D-30b.   Residual vs. observed plot for 63  from the UAM
                simulation results  for Day 159-76.
                                209

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 160-76
        .500+001
    Q.
    a
    a
.400+001  •

.300+001  J

.200+001

.100+001

.000

-.100+001

- 200+001

- 300+001

-.400+001

-.500+001
                   10   20   JO  40  50  60  70
                                 FREQUENCY
                   CARBON  MONOXIDE
                                      30  90  100
                DATA AVAILABLE =  242    DATA POSSIBLE
                                                 273
Figure D-31a.   Residual histogram for CO from  the UAM simulation
                results for  Day 160-76.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 160-76

2
Q.
Q.
N— '
RESIDUAL


3UU+UUI
.400+001
.300+001
.200+001
.100+001
- 100+001
-.200+001
-.300+001
-.400+001
-500+001
000
.
-

«a tf f
.__*aiMtHjraWi 4. , i , , ,
5™^° °
•
•



LECtHO
FREO STM
< t. :> n
< 2, 2> 0

100+001 .200+001 .300+001 400+001 500+001
                      OBSERVED CONCENTRATION  (PPM)
                   CARBON MONOXIDE
              DATA AVAILABLE =  242   DATA POSSIBLE  =  273

Figure D-31b.   Residual  vs.  observed plot for CO from the UAM
                simulation  results for Day 160-76.
                                210

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 160-76

Q.
Q^
_i
*<
RESIDUE



. / 3U — uu > r
.600-001 .
.450-001 -
.300-001 -
.150-001 -
.000
- 150-001 -
-300-001 •
-.450-001 •
-.600-001
-.750-001 L
»
|
: * *«••
• •' if «tt»S*t"t*m
' ". '\'. ii|*iiliijiiit|iiii ^**^**
• 9*?
*


                       10  15  20  25   20   35
                                 FREQUENCY
                   NITROGEN  DIOXIDE
                                              40   45  50
                DATA AVAILABLE =  223    DATA POSSIBLE  =  273

Figure D-32a.   Residual histogram  for N02 from the UAM  simulation
                results for Day 160-76.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 160-76
     Q.
     Q.
     g
     00
     LLJ
     cc
,/3U— UUI
.600-001
.450-001
.300-001
150-001
000
- 150-001
-.300-001
-.450-001
-.600-001
- 750-001
QOO
a
•
a
Q
a
fl
d^3 ° ..-.,.'
^^uag9' ' .
3
.
.500-001 .1004-000 .150+000 200+000 .250+
                                                          IECENO
                                                         FREO sry
                                                        < t. i>  a
                                                        < 2. 2>  O
                                                        < 4.99>  »
                      OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
              DATA AVAILABLE =  223    DATA POSSIBLE  =  273

Figure D-32b.   Residual vs. observed plot for N02 from  the  UAM
                simulation results  for Day 160-76.
                               211

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 160-76

S
Q.
Q.
~
_i
Q
y5
LU
o:








Z3U+VUU
200+000
.150+000

.100+000
.500-001
000

-.500-001
- 100+000
-.150+000
-.200+000
- 250+000

.

•1
tJ?
! • npi! ji!||im!niHpiP:*;"
1 ' " 'i< Iiiziil««xi«»» ^
>"
-

•
• i i i . > i . i
0 5 10 15 20 25 JO 35 40 45 5
FREQUENCY
OZONE
DATA AVAILABLE = 229 DATA POSSIBLE =
                                                       273

Figure D-33a.  Residual histogram  for  03  from the UAM simulation
               results for Day  160-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 160-76
2
Q.
0-^
on
UJ
or
.200+000
.150+000
.100+000
.500-001
000
-.500-001
- 100+000
- 150+000
-.200+000
-.250+000
,000
0 °
m TWfe rV^rSiTOii'ifri^ff'
^ [g . mp il7TUf*''r^T^1 TBTj i ^ ai oT^ i i
•
LEGEND
FREO SYM
< 1. t> O
< 2. 2> 0
< 3. J> A
< »,99> »
500-001 .100+000 .150+000 200+000 .250+000
                     OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  229    DATA POSSIBLE  = 273

Figure D-33b.  Residual vs.  observed  plot for 03 from the UAM
               simulation results  for Day 160-76.

                              212

-------
      RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 211-75
.200+001
^ 150+001
Q.
~~* .100+001
J .500+000
Qnno
.UWU
(J]
£ - 500+000
- 100+001
-.150+001
-.200+001
-.250+001
.
t
-»
ms
• " T"[ i jit iii ijiiitijiiit^*^ ******


•
                      10  15
                                             40   *5   50
              20  25   30  35
                 FREQUENCY
   CARBON  MONOXIDE
DATA AVAILABLE =  233    DATA POSSIBLE
                                                       273
Figure D-34a.  Residual  histogram for CO from the UAM simulation
               results  for day 211-76.
                  RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 211 -76

y^
2
Q.
Q.
*— '

_l
<
D
Q
c/5
Ld
o;








^3U+UU 1
.200+001

.150+001

.100+001

.500+000


QOO
.

O
a
a
nnoa
.^3 n a
fllKHCLw^"
m^SXMIMHni Q
i iMIII'HirUB , ,,iii
jBBB9tt9? a
< 2. 2> O
< 3, 3> *
< *.99> «


.000 .100+001 .200+001 .300+001 400+001 500+001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE  -
             DATA AVAILABLE = 233    DATA POSSIBLE  =  273
Figure D-34b.  Residual  vs.  observed  plot for CO from the UAM
               simulation results for Day 211-76.

                               213

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 21 1-76
./3U-UUI
500-001
X-K
OJ 450-001
Q.
^ ,300-001
J .150-001
Q .000
00
{£ -150-001
-300-001
-.450-001
-.600-001
-.750-001
*
.
XT

• •+*•



                      10   15   20  25  30  35
                                FREQUENCY
                  NITROGEN DIOXIDE
40  *5   50
               DATA AVAILABLE =  218   DATA POSSIBLE =  273

Figure D-35a.  Residual  histogram for N02 from  the UAM  simulation
               results  for Day 211-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 21 1-76

S
Q.
RESIDUAL (F


. /ou— uu i
..SOO-001
.450-001
.300-001
150-001
000
- 150-001
-.450-001
-.600-001
- 750-001
.000
. . i i ' i .o
a
a
o
|!f&'s-1 r ' '
.
•


lECENO
FREO SYM
< i. i> a
<. 2. 2> O
< *]99> »

200-001 .400-001 .500-001 800-001 100-1-000
                     OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE =  218    DATA POSSIBLE  =  273

Figure D-35b.   Residual  vs. observed plot  for  N02  from the UAM
                simulation results for  Day  211-76.

                               214

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 211-75

a
a.
RESIDUAL



2001-000
.150+000
.100*000
.500-001
.000
-.500-00?
-.100+000
-.150+000
-.200-t-OOO
-.250+000


^ • , , ,
pPUl1' 	 -
i


                  10   20  X  *0  50   60   70
                                FREQUENCY
                  OZONE
                             30   90  100
               DATA AVAILABLE =  225    DATA POSSIBLE
                                       273
Figure D-36a.  Residual  histogram for 03 from the UAM simulation
               results  for  Day  211-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 211-76

0_
Q^
RESIDUAL

200+000
150+000
.100+000
.500-001
-500-001
- 100+000
-.150+000
-.200+000
- 250+000
000
.
: ;
' J ^SW18 " °°°° i • ,
Q
a



LECCNO
FREQ STM
< i, i> a
< 2. 2> 0
< 3. 3> *

.500-001 .100+000 .150+000 200+000 250+000
                    OBSERVED CONCENTRATION  (PPM)
                  OZONE
             DATA AVAILABLE =  225   DATA POSSIBLE =  273
Figure D-36b.
Residual vs. observed plot  for  0-
simulation results for Day  211-7(

               215
from the UAM

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                            DAY 212-75
    g
    a
    on
.3UU+UUI
400+001
.300+001
.200+001
.
•
Mfr
.100+001 ff
^ ^_^
MHHB|^lMHHBH^HIM^^^^^^^^^^^^^^_
-.100+001
-.200+001
-.300+001
-.400+001
-500+001
r
•
•

1 i i i i i i i
0 10 20 30 40 50 60 70 30 90 10
                               FREQUENCY
                  CARBON MONOXIDE
               DATA AVAILABLE =  2*6   DATA POSSIBLE  =  273

Figure D-37a.  Residual histogram for CO  from  the  UAM simulation
               results for Day 212-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 212-76
^^
Q.
Q^
RESIDUAL


3UU»UUI
•400+001
300+001
.200+001
.100+001
000
- 100+001
-.200+001
-.300+001
-.400^001
- 500+001
000
.
a
-^AjflrtSpi *"*,,,'
Q
•
•


LEGEND
FREO STM
< i. i> a
< 2. 2> 0
< 3, 3> *
< *,99> »

.100+001 .200+001 .300+001 400+001 500+001
                     OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  246   DATA POSSIBLE  =  273
Figure D-375.   Residual  vs.  observed plot for CO from the UAM
                simulation  results  for Day 212-76.

                              216

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 212-76
400-001
3J .300-001
Q.
^ .200-001
-J .100-001
3
Q 000
(75
}£ -.100-001
-200-001
-.300-001
-.400-001
-.500-001
*>
*

-*
' 4

':"•
;!

•
-
.
••
M ^ •
M 2 [••«•
itll?«*; ii....
:[U:iu....
JS5J.
*•


                      10  15  20   25   30  35
                                FREQUENCY
                   NITROGEN  DIOXIDE
                                             40   45   50
               DATA AVAILABLE =  234
Figure D-38a.   Residual  histogram for
                results  for Day 212-76.
                                      DATA POSSIBLE  =  273

                                           from  the UAM  simulation
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 212-75


•s
^C
Q.
Q.
^"^

_i
Q
00
LJ






.3UU— UU'
.400-001

.300-001

.200-001

.100-001
.000

- 100-001
- 200-001

-.300-001
-.+00-001
- 500—001
a
a
a
o a _
9 0 ° 0
o ™ Q a a
TT .DO n o
j?"Sn Jw *

tjjBJIfT^ iTJL **? ^^
jftlf^jffft^ QDu n n $
•ffigBg™ °
.r?


.a
i i










UCENO
FREO SYM
< i. i> a
< 2. 2> 0
< 3. 3> *


000 200-001 .400-001 .600-001 300-001 .100-1-000
                     OBSERVED CONCENTRATION  (PPM)
                  NITROGEN DIOXIDE
             DATA AVAILABLE = 234    DATA POSSIBLE  =  273

Figure D-38b.   Residual  vs. observed plot for N02  from the UAM
                simulation results for Day 212-76.
                              217

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                            DAY 212-76

•C"
?
Q.
s"""'
_l
Z!
9
or









Figure




a
^
_j
a
ui
'-U
CE








. / 350-001

-600-001
- 750-001
fpy,
\

1 1 1 1 1 1 I 1






LEGEND
FREQ STM
< i. 1> 0
< 2.2> 0
< 3. 3> •
< 4,99> «


QOO .500-001 .100+000 .150+000 200+000 250+000
                    OBSERVED CONCENTRATION  (PPM)
                 OZONE
            DATA AVAILABLE =  216    DATA POSSIBLE  = 273

Figure D-39b.  Residual vs. observed plot  for 63 from the UAM
               simulation results  for Day  212-76.

                              218

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALi_ TIMES AND LOCATIONS

                              DAY  225-76

        .250-1-001
        .200*001
    2
    a
    a.
    g

    Q
    in
       -.200+001


       -.250+001
                  10
                      20  20
           40  50  60  70
              FREQUENCY

CARBON  MONOXIDE
80   90
               DATA AVAILABLE =  253    DATA POSSIBLE
                                      273
Figure D-40a.   Residual histogram for CO from  the UAM simulation
                results for  Day  225-76.
    CL
    Q.
                 RESIDUAL VS OBSERVED CONCENTRATION

                              DAY 225-76
.200+001
.150+001
.100+001
.500+000
000
-.500+000
-.100+001
-.150+001
-.200+001
- 250+001
000
.
. °
o no
« o a a
a
o
a
.500+000 .100+001 .150+001 200+001 250H
    Q

    CO

    a   —-' — fa   a *"                                ]    LECENO

                                                         FREO SYM

                                                        < i. i>  a
                                                        < 2. 2>  O
                                                        < 3, 3>  »
                                                        < 4,99>  «
                     OBSERVED CONCENTRATION  (PPM)

                  CARBON MONOXIDE

             DATA AVAILABLE =  253   DATA POSSIBLE  =  273


Figure D-40b.   Residual  vs. observed plot for  CO  from the UAM

                simulation  results for Day 225-76.
                               219

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                              DAY 225-76

s*
Q.
N— '
_i
(75
UJ
or




.400-001 •**
300-001 3*
.200-001 -a]
.100-001 -;;
000 "

- 1 GO-DO 1 -\':
-.200-001 :::
-.300-001 4
-.400-001 •*
-.500-001 1 —
,
•

: i " ;; «•**•
• : ;; ; ^{*««>i»I>tft
J: : lIl^MiHtis**** * '
| * *•»
>i '


1,1,111,
                       10   15  20  25  30  35
                                 FREQUENCY
                   NITROGEN DIOXIDE
                                              40
                                                      50
                DATA AVAILABLE  =  236   DATA POSSIBLE
273
Figure 0-41a.   Residual  histogram for  N02 from the  UAM simulation
                results  for Day 225-76.
                  RESIDUAL VS OBSERVED CONCENTRATION
                              DAY 225-75
    •s
    a
    a
    g
    a
    (/5
    £  - '""-^'l-ff*;* m  „%„  amtt, u                   1    LEGEND
                                                          FREQ SYM
                                                        < t. i> a
                                                        < 2. 2> O
                                                        < 3, 3> »
       - 500-001
            .000      .200-001  .400-001  .600-001  .300-001   .iOO-t-000
                     OBSERVED CONCENTRATION  (PPM)
                   NITROGEN DIOXIDE
             DATA AVAILABLE =  236   DATA POSSIBLE   =  273

Figure D-415.   Residual  vs.  observed  plot for N02  from the UAM
                simulation results  for Day 225-76.
                               220

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                            DAY 225-76

2
a
a^

<
Q
vi
UJ
or




.l^UTUUVf
.200+000
.150+000

.100+000
.500-001
000

-.500-001
-.100+000
-.150+000
-.200+000
- 250+000

.

*•
: •$*•
§|fe||«:*|.
*i*-Si I it*! •» • • i J • S**"^* JT t • ' 1
**»«»«»•»** ' " ' '

•
-
•
-
1 1 1 , 1 , , 1 ,
                     10  15   20  25  JO
                               FREQUENCY
                  OZONE
              DATA AVAILABLE =  209
                                            40  45
50
                                     DATA POSSIBLE  =  273

Figure D-42a.  Residual  histogram for 03 from the UAM simulation
               results  for  Day  225-76.
                RESIDUAL VS OBSERVED CONCENTRATION
                            DAY 225-76

2
a.
0.
_j
RESIDU/

..43U+UUU
.200+000
.150+000
.100+000
.500-001
000
-.500-001
- 100+000
-.150+000
-.200+000
-.250+000
.000
-
0=
t " CJW!^^
aajjjF^iS&S*!™^





LECCNO
FREO Srtt
< t. t> 0
< 2. 2> 0

.500-001 .100+000 .150+000 .200+000 .250+000
                    OBSERVED CONCENTRATION  (PPM)
                 OZONE
            DATA AVAILABLE =  209    DATA POSSIBLE  =  273

Figure D-42b.  Residual vs.  observed  plot  for 03 from the UAM
               simulation results  for Day  225-76.

                             221

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALi_ TIMES AND LOCATIONS
                             DAY 226-76
a
a^
_i
Q




.500+001
.450+001
.300+00)
.150+OOt
000
-.150+001
-.300+001
-.450+001
-.600+001
- 750+001
	 1
»
"Pill! 	 | ( | '
r".' 	 r '" ' ' .
•
•

                  10  20  20  40   50   60  70
                               FREQUENCY
                  CARBON MONOXIDE
30   90  100
               DATA AVAILABLE =  207    DATA POSSIBLE
          273
Figure D-43a.  Residual histogram for CO from the UAM simulation
               results for Day  226-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 226-75

2
a.
0.
^~ '
RESIDUAL


. / ^U *VW 1
.500+001
.450+001
.300+001
.150+001
000
-.150+001
-300+001
-.450+001
-.600+001
-.750+001
.000

a

&n Q
pP551' i i i . i . i ^
•
•



LEGEND
FREQ 5YM
< 1 1 > Q
< 2. 2> 0
< J. 3> A

.150+001 .300+001 .450*001 600+001 .750+001
                    OBSERVED CONCENTRATION  (PPM)
                  CARBON MONOXIDE
             DATA AVAILABLE =  207   DATA POSSIBLE  =  273

Figure D-43b.  Residual  vs.  observed plot for CO from the UAM
               simulation  results  for Day 226-76.

                             222

-------
     RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 226-76
. /^J—UVM 1-
600-001 -
^ 450-001 •
Q.
^ .300-001 -
^ .150-001 -
Q .000
CO
^ - 150-001 -
-.300-001 -
-.450-001 •
-.600-001 •
-.750-001 .
t
*++

. >. U A
1^: ,« ,,,,,"
i ii I Jflflpilil ""*"**•"**'
: •{•*•*•
:
*
-
                      10  15  20   25   JO  35
                                FREQUENCY
                  NITROGEN  DIOXIDE
               DATA AVAILABLE
                                            40  45  50
233   DATA POSSIBLE
273
Figure D-44a.   Residual  histogram for N02 from  the  UAM  simulation
                results  fo.r Day 226-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 226-75

2
CL
^
_1
n
en
LJ

.600-001
450-001
.300-001
.150-001
.000
- 150-001
- 300-001
-.450-001
-.600-001
-.750-001
.000
0-
a
a aa
ffl fl l*T*H PI
Q 153 ji^ IJT n fj
 Q
< 2. 2> 0
< 3, 3> *
< *,99> 9

.200-001 .400-001 600-001 300-001 .100+000
                    OBSERVED CONCENTRATION  (PPM)
                  NITROGEN  DIOXIDE
             DATA AVAILABLE =  233   DATA POSSIBLE  =  273

Figure D-44b.  Residual  vs.  observed plot for  N02  from  the  UAM
               simulation  results for Day 226-76.

                              223

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                            DAY 226-76
.200+000
^ .150*000
Q.
^ .100*000
J .500-001
Q .000
£ -.500-001
- 100*000
-.150*000
-.200+000
-.250+000
.
{*
;; ijjjmmtM
'.'. "• i ' - 1' $$SI$$
jjpfcS.41*"


•
0 5 10 15 20 25 20 35 *0 *5 5
FREQUENCY
OZONE
DATA AVAILABLE = 1 78 DATA POSSIBLE =
                                                      273
Figure D-45a.   Residual  histogram for 03  from  the UAM  simulation
                results for Day 226-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 226-76

RESIDUAL (PPM)


.200*000
.150*000
.100*000
500-001
000
-.500-001
-.100*000
-.150+000
-.200+000
- 250+000
t
a
a ° °
«8™f^|jp^a
a a rp£,ase ° ^ .^ ^°^'"H
Pr '= u ' '
;
•

LEGEND
FREO STM
< i, t> a
< 2. 2> 0
< 3, 3> *

            000     .500-001  .100+000 .150+000  200+000  250+000
                     OBSERVED CONCENTRATION   (PPM)
                  OZONE
             DATA AVAILABLE =   178    DATA POSSIBLE  =  273

Figure D-45b.   Residual  vs.  observed  plot  for 03 from the UAM
                simulation results  for Day  226-76.

                              224

-------
     RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 237-76
       .500-1-001
                  10  20  30   *0   50  60  70  SO   90  100
                               FREQUENCY
                  CARBON MONOXIDE
               DATA AVAILABLE =  234   DATA POSSIBLE  =  273

Figure D-46a.  Residual  histogram for CO from the UAM simulation
               results for  Day  237-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                            DAY 237-76

Q.
0^
3
RESIDU/


.JWU^WW 1
.400+001
.300+001
.200+001
.100+001
000
- 100+001
-.200+001
-.300+001
-.400+001
- 500+001
000
.
a
a
"C
.•3 gj ° °
. 	 — JSOIj 3Q _
,J_flFJ?T^F 0
rrMMBfiPS^ 11 'II
D
•




LEGEND
FREQ SYM
< i, i> a
< 2, 2> 0
< J. J> *

.100+001 .200+001 .300+001 400+001 500+001
                    OBSERVED CONCENTRATION  (PPM)
                 CARBON MONOXIDE
            DATA AVAILABLE =  23*    DATA POSSIBLE  =  273

Figure D-46b.   Residual  vs.  observed plot for CO from  the  UAM
                simulation results for Day 237-76.
                              225

-------
    RESIDUAL HISTOGRAM DETERMINED OVER ALL TIMES AND LOCATIONS

                             DAY 237-75


       .100+000
    3
    a.
    CL
    I
    Q
                  10  20  JO  40  50   60   70

                                FREQUENCY

                  NITROGEN  DIOXIDE
             ao  90 100
               DATA AVAILABLE
212    DATA POSSIBLE  »  273
Figure D-47a.   Residual  histogram for N02  from the DAM simulation

                results for Day 237-76.
                 RESIDUAL VS OBSERVED CONCENTRATION

                             DAY 237-76


2
a
a.
~
j
9
LU





.1UV+UUU
.800-001

600-001

.400-001
.200-001
000
-.200-001
-.400-001
-.600-001
-.800-001

0

•
a
% ^ a
. Jga _ o1 °
jdB&nSjp % a
jB^^rnffftp"*" g O
«%f a°
o a
a
a

- 100 1 000 ' L- ' ' ' * ' ' -








LCCCNO
F»EO ST«4
< i, i> a
< 2. 2> 0
< 3. 3> *


000 500-001 .100+000 150+000 200+000 250+000
                     OBSERVED CONCENTRATION  (PPM)

                  NITROGEN DIOXIDE

             DATA AVAILABLE =  212    DATA POSSIBLE  =  273


Figure  D-47b.   Residual vs. observed plot for N02 from  the DAM

                simulation  results  for Day 237-76.
                              226

-------
    RESIDUAL HISTOGRAM  DETERMINED OVER ALL TIMES AND LOCATIONS
                             DAY 237-75
200+000
OJ .150+000
a
^ .100+000
5? .500-001
QGIY)
.WWU
in
g -.500-001
- 100+000
-.150+000
-.200+000
-.250+000
.
.

if mijj. .„.
" : ' i :: :: || :i illlll| UftSSS $***••**•

:]: |; lUtin**"*********"
. I •*
-

•
Figure D-48a.
       10   15   20  25  30  35  40   45   50
                 FREQUENCY
   OZONE
DATA AVAILABLE =  230   DATA POSSIBLE   =  273

 Residual  histogram for 03 from the  UAM  simulation
 results  for Day 237-76.
                 RESIDUAL VS OBSERVED CONCENTRATION
                             DAY 237-76


a
a.


<;
a
a
LJ
C£.





.^au+uuu
.200+000
.150+000

.100+000

.500-001

000
-500-001
- 100+000

-.150+000
-.200+000
-.250+000


.

iw.-.jy*
Ttc/iflvjfif '
_a_12p!M6'rnlsj _ a
' • Lj^ujj iro^yTJ ^ ^
^H I'l ^ f PL ffy ' |H O

jSWyBJ? T3 ' ' ' ' Q ' ' '
5^ a














l£CENO
FREO STM
< i, i> a
< 2. 2> 0
< 1. 3> »


                   .500-001  .100+000  .150+000  200+000  250*000
                    OBSERVED CONCENTRATION  (PPM)
                 OZONE
            DATA AVAILABLE =  230   DATA POSSIBLE  =  273

Figure D-48b.   Residual  vs. observed plot  for 03 from the DAM
                simulation results for Day  237-76.

                             227

-------
                                  APPENDIX  E

                 UAM - CONTOUR PLOTS FOR REMAINING  TEST  DAYS

     The contour  plots of  predicted and  observed 03  concentrations  for  the
hour of the observed concentration maximum  are  presented here  for all test days
not  specifically  discussed within  the  body  of  this  report.    The Julian  day
numbers are listed in Appendix D.
                                     228

-------
   ST.  uauia  JUNE 27.1975   IOBT 1791
   ^COICTCO CONCCNTtRTIQNS OF 03  I IN
   HOUR 1400-1500
ST. L3UIS  JUNE Z7.1975    1OflT 1781
OBSERVED CUNCENTHBTIONS or 03  KN
HOUR  1400-1SCO
             (a)
              (b)
Figure E-l.   Contours  of  (a)  predicted and  (b) observed  fields
               of  03 at  hour  of observed maximum on  Day  178-75.
                                              	140
   ST. IOU1!  JULT  1.1975   1OBT 1821
   '•WEOICTED COMCENTKATIONS ar 
-------
                                                            \  \
    ST. LOUIS JULT  2,1975   I CRT
    FUEOICTEO CONCENT*ST 1QN3 OF 03 I IN
    HOUR 1000-1100
ST. U9UIS  JULT  2.1975   I OPT 1831
OBSERVED CONCENTRflTlflNS OF 93 I IN
HOUR  1000-1100
             la)
              (b)
Figure E-3.  Contours  of  (a) predicted and  (b) observed fields
               of 63 at  hour  of observed maximum on Day  183-75.
    ST. LOUIS JULT  3.1975   IDBT 1841
    PREQJCTEO CONCENTHBTtONS OF 03  I IN PPB I
    HOUR 1300-1400
              ia)
ST. LOUIS  JULT  3.197S   I OUT 1841
OBSERVED CONCENTRATIONS. OF 03  I IN I*f81
HOUR  1300-1400
              (b)
Figure E-4.   Contours of  (a) predicted and  (b)  observed fields
                of 03  at hour of  observed maximum  on Day  184-75.
                                  230

-------
    ST. LOUIS  JULT 2S.197S    IOST 2091
    MEOICTEfl CONCCNTKRTIONS 3F S3  I IN '
    HOUR 1100-1200
ST. LOUIS  JULT 28.1975   I OUT 2091
OBSERVED CONCENTHBTIONS OF 83 UN
HOUR  1100-1200
             (a)
              (b)
Figure E-5.   Contours  of (a)  predicted and  (b) observed fields
               of 03 at  hour of observed maximum on Day  209-75.
   ST. LBUIS  BUG.  9.1975    I OPT 2311
                                         ST. LOUIS  RUG.  9.I97S    I OUT 2211
                                         OBSERVED CONCCNTRRTKJNS OF 03  I IN
                                         HQUR  1 SCO-1600
             (a)
              (b)
Figure E-6.   Contours  of (a)  predicted and  (b) observed fields
               of 03 at  hour of observed maximum on Day  221-75.
                                  231

-------
   SI-  LOUIS  BUD. 18.1975    I OPT 2JO I
   PREDICTED C ONCE NT*BT IONS 8F 03  UN PP9 I
        (300-1400
             (a)
ST. LOUIS  AUG. 18.13-75    I OAT 2301
OBSERVED CONCENTRSTION5 OF 03  I IN
HOUR  1300-MOO
              (b)
Figure E-7.   Contours  of  (a) predicted and  (b) observed fields
               of 03 at  hour  of observed maximum on  Day  230-75.
   ST. LOUIS  flUC. 19.1975    I CRT 2311
   PREDICTED C ONCE NT «f»T IONS If 03  I IN PP8 I
   HOUR  1400-1500
             (a)
ST. LOUIS  RUG. 19.191S   I OflT 2311
OBSERVED CONCENTRPTIONS Of 93  I IN Pf 81
HOUR  MOO-1SOO
              (b)
Figure  E-8.   Contours of  (a) predicted and (b)  observed fields
                of 03  at hour of  observed maximum  on Day  231-75.
                                   232

-------
  SI*.  LOUIS SEPT.  J.I 9^5   I OUT 2511
  PREDICTED COHCENTtBTlOM Of 03  I IN PP8 I
  HOUR 1200-1300
ST. LOUIS SEPT.  8.1375    I0«T 2511
OBSERVED CONCEHTMTIOM OF as  UN PPSJ
HOUR  1200-1300
             (a)
               (b)
Figure E-9.   Contours  of  (a)  predicted  and  (b) observed  fields
               of  63 at  hour  of observed  maximum on  Day 251-75.
   ST. LOUIS  JUNE  7.1976    I DOT 1591
   PREDICTED CONCENTHflTlONS OF 03  I IN PPflI
   HOUR  UOO-1SOQ
             (a)
 ST. LOUIS  JUN6  7.1976    I OAT 1S9I
 OBSERVED CONCENTHPT10NS, OC 33  UN PT8)
 HOUR  1400-1500
               (b)
Figure E-10.   Contours of  (a) predicted and  (b)  observed fields
                 of 03  at hour of  observed maximum  on Day  159-76.
                                   233

-------
  ST. LOUIS  JUNE  9.1978   I OUT ISO I
  PREDICTED CONCENTRATIONS OF 33  MM P
  HOUR 1600-1700
                        ST.  LflUtS  JUNE  8.1975   (001 1801
                        OBSERVED CONCENtRSTlONS OF 03 I IN
                        HOUR isoo-noo
             (a)
                                       (b)
Figure E-ll.
Contours of (a)  predicted  and (b)  observed fields
of 63 at hour of observed  maximum  on Day 160-76.
                                                       V
   ST.  LOUIS  JULT 29.1978   I DOT 2111
   PREDICTED CONCENT(ATIQMS OF 03  I IN PP8)
   HOUR 1500-1800
                         ST.  LOUIS  JULT 29.1978   I OPT 2111
                         OBSERVED CONCCMTKHTIONS' OF 03 I IN
                         HOUR  1500-1600
                                                        (b)
Figure  E-12.   Contours of (a)  predicted and (b)  observed fields
                 of 63  at hour of observed maximum  on Day  211-76.
                                  234

-------
   ST. LOUIS  JULT 30.1978    I OPT 2121
   PREDICTED CONCENTRATIONS If 03  I IN f
   HOUR  1200-1300
             (a)
ST. LOUIS  JULT 30.1976   I OUT 212)
OBSERVED CONCENTRATIONS Of 93  I IN
HOUR  1200-1300
              (b)
Figure E-13.  Contours of  (a) predicted and  (b) observed fields
                of 63  at hour of observed maximum on Day  212-76.
             o
   —«0.
   ST. LOUIS  SUO. 12.1976    I OAT 2251
   PREDICTED CONCENT HAT IONS OF 03  UN PPS1
   HOUR  1300-1400
ST. LOUIS  BUG. 12.1975   IQAY 2251
OBSERVED CONCENTRATIONS OF 03  1 IN P
HOUR  1300-MOO
             (a)
              (b)
Figure E-14.  Contours of  (a) predicted and  (b)  observed fields
                of 03  at hour of  observed maximum  on Day  225-76.
                                   235

-------
   ST. LOUIS  BUO. U.197S    I OUT 2281
   PREDICTED CBNC£NT««T!ONS Of 33  I IN F
   HOUR  1300-1400
ST. LOUIS  CUC. 13.1976   I BUT 2261
385EHVEO CONCeNTRSTlONS Of 03 I IM
HOUR  1300-1400
             (a)
              (b)
Figure E-15.  Contours of  (a) predicted and  (b) observed fields
                of 03 at hour of observed maximum on Day  226-76.
   ST. LOUIS  BUO. 24.1975    IQPT 2371
   PREDICTED CONC£NT»BT10NS OF 33  I IN PP8 I
   HOUR  1100-1200
ST. LOUIS  BUG- 24.197B   IOPT 2371
OBSEKVEO C3NCENTHHTIONS af 33 I [N
HOUR  1100-1200
             (a)
              (b)
Figure  £-16.   Contours of  (a)  predicted and (b)  observed fields
                 of 03  at hour of observed maximum  on Day 237-76.
                                   236

-------
                                 APPENDIX F

                    UAM - SUMMARY OF STATISTICAL RESULTS

     Three tables  are presented  in  this appendix.   Table F-l  summarizes  the
statistics for  the mean  quantities on  each of  the 20  days  for the  species
NO, N02, 03,  OLE,  NMHC and CO.  Note that the  observed OLE  (olefin)  value  was
obtained by  a temporally  and  spatially constant  splitting  factor from  NMHC.
This  factor  was based on an  analysis   of  the  composite  of volatile  organic
emissions over the entire St. Louis area.
                                              *
     Table F-2  summarizes the  analysis of  trends data  for the  species  N02,
03  and  CO,  and Table  F-3  presents  concentration  maxima data  for  the  same
species.  The statistical  parameters  included in the 3 tables  are the  same as
those defined in Appendix B,  except for  the following:

     (1)  in  addition to an  overall correlation coefficient  there  are-also  the
          following:

          (a)  average spatial  correlation (average  over  correlation  coeffici-
               ents for  each hour  of  simulation  calculated separately),  and

          (b)  average temporal  correlation (average  over  correlation  coeffici-
               ents for each  location  calculated separately);

     (2)  for  concentration  maxima   data  the  following  descriptions   apply:

          (a)  peak observed  concentration  (along  with  the location and  time of
               occurrence), and

          (b)  peak predicted concentration  (at same location  as peak observed
               concentration, along with the  time  of  occurrence).

                                      237

-------
TABLE F-l.  SUMMARY OF MODEL PERFORMANCE STATISTICS FOR MEAN  CONCENTRATIONS  FROM  THE  UAM.

Jul ian
date

142-75





178-75





182-75





183-75





Mean observed
Species concentration

NO
NO?
03
OLE
NMHC
CO
NO
N02
03
OLE
NMHC
CO
NO
N02
03
OLE
NMHC
CO
NO
N02
03
OLE
NMHC
CO
(pom)
0.0086
0.0134
0.0635
0.0080
0.183
0.481
0.0088
0.0212
0.0920
0.0147
0.334
0.809
0.008?
0.0172
0.0737
0.0118
0.269
0.517
0.0105
0.0227
0.0809
0.0126
0.285
0.655
Mean predicted Mean
concentration
(ppm)
0.0064
0.0181
0.0779
0.0044
0.258
0.479
0.0070
0.0259
0.0850
0.0062
0.384
0.723
0.0029
0.0128
0.0777
0.0041
0.202
0.307
0.0055
0.0224
0.0855
0.0059
0.298
0.466
residual
(ppm)
0.0022
-0.0047
-0.0145
0.0036
-0.0754
0.0016
0.0018
-0.0047
0.0070
0.0085
-0.0494
0.0860
' 0.0052
0.0045
-0.0040
0.0077
0.0669
0.210
0.0051
0.0004
-0.0046
0.0067
-0.0126
0.189
Mean residual Error
RMS error Mean obs. cone. banda
(ppm)
0.0094
0.0111
0.0317
0.0093
0.202
0.416
0.0116
0.0168
0.0314
0.0113
0.265
0.537
0.0105
0.0113
0.0186
0.0100
0.208
0.521
0.0144
0.0162
0.0248
0.0112
0.239
0.559
!a\
\*>l
25.6
-35.1
-22.8
45.0
-41.2
0.3
20.5
-22.2
7.6
57.8
-14.5
10.6
63.4
26.2
-5.4
65.3
24.9
40.6
48.6
1.8
0.3
53.1
-4.4
28.9
(%)
19.7
30.8
42.1
17.0
25.0
33.9
15.0
28.6
46.3
13.2
22.5
31.9
11.0
28.6
' 69.5
87.0
22.5
27.1
17.5
32.1
54.0
7.8
28.7
24.4
                                            238

-------
TABLE F-l. (Continued)



            SUMMARY OF MODEL PERFORMANCE STATISTICS FOR MEAN CONCENTRATIONS FROM THE (JAM.

Julian
date

184-75





207-75





209-75





221-75






Species

NO
N02
03
OLE
NMHC
CO
NO
NO?
°3
OLE
NMHC
CO
NO
NO?
03
OLE
NMHC
CO
NO
N02
03
OLE
NMHC
CO
Mean observed
concentration
(ppm)
0.0118
0.0254
0.0814
0.0253
0.575
0.859
0.0070
0.0134
0.0802
0.0113
0.258
0.378
0.0077
0.0149
0.0698
0.0095
0.216
0.486
0.0097
0.0191
0.0703
0.0138
0.303
0.755
Mean predicted
concentration
(ppm)
0.0044
0.0212
0.0806
0.0131
0.470
0.612
0.0060
0.0182
0.0656
0.0063
0.233
0.377
0.0070
0.0190
0.0616
0.0056
0.289
0.442
0.0073
0.0185
0.0523
0.0063
0.221
0.476
Mean
residual
(ppm)
0.0074
0.0042
0.0008
0.0122
0.106
0.247
0.0010
-0.0048
0.0146
0.0050-
0.0246
0.0002
0.0006
-0.0042
0.0082
0.0040
-0.0736
0.0440
0.0024
0.0006
0.0180
0.0075
0.0919
0.279

RM.S error
(ppm)
0.0179
0.0220
0.0319
0.0208
0.412
0.671
0.0077
0.0079
0.0209
0.0149
0.341
0.331
0.0104
0.0095
0.0265
0.0090
0.242
0.986
0.0153
0.0127
0.0274
0.0128
0.294
0.831
Mean residual
Mean obs. cone.
(*)
62.7
16.5
1.0
48.2
18.4
28.8
14.3
-35.8
18.2
44.2
9.5
0.1
7.8
-28.2
11.7
42.1
-34.1
9.1
24.7
3.1
25.6
54.3
30.3
37.0
Error
band*
(*)
16.5
28.9
48.9
13.6
24.5
25.6
24.2
29.9
51.9
17.9
24.4
35.4
14.7
31.6
50.9
21.4
17.6
39.6
15.9
30.6
38.5
10.5
21.6
30.2
                                           239

-------
TABLE F-l. (Continued)



            SUMMARY OF MODEL PERFORMANCE STATISTICS  FOR  MEAN  CONCENTRATIONS  FROM THE UAM.

Jul ian
date

230-75





231-75





251-75





159-76






Species

NO
NO?
03
OLE
NMHC
CO
NO
N02
°3
OLE
NMHC
CO
NO
NO?
03
OLE
NMHC
CO
NO
NO?
03
OLE
NMHC
CO
Mean observed
concentration
(ppm)
0.0130
0.0188
0.0449
0.0146
0.332
0.550
0.0084
0.0165
0.0518
0.0125
0.284
0.437
0.0198
0.0253
0.0507
0.0193
0.438
0.806
0.0163
0.0275
0.0880
0.0182
0.414
0.968
Mean predicted
concentration
(ppm)
0.0100
0.0215
0.0531
0.0065
0.317
0.468
0.0089
0.0194
0.0530
0.0061
0.337
0.456
0.0141
0.0279
0.0447
0.0087
0.449
0.648
0.0084
0.0230
0.0996
0.0095
0.401
0.693
Mean
residual
(ppm)
0.0030
-0.0026
-0.0086
0.0081
0.0148
0.0821
-0.0004
-0.0029
-0.0012
0.0064
-0.0535
-0.0190
0.0058
-0.0026
0.0061
0.0106
-0.0111
0.158
0.0079
0.0045
-0.0116
0.0088
0.0128
0.275

RMS error
(ppm)
0,0165
0.0091
0.0209
0.0117
0.243
0.420
0.0095
0.0087
0.0221
0.0101
0.209
0.355
0.0307
0.0190
0.0214
0.0166
0.349
0.777
0.0215
0.0130
0.0270
0.0149
0.303
2.46
Mean residual
Mean obs. cone.
(%)
23.1
-13.8
-19.2
55.5
4.5
14.9
-4.8
-17.6
-2.3
51.2
-18.8
-4.3
29.3
-10.3
12.0
54.9
-2.5
19.6
48.5
16.4
-13.2
48.4
3.1
28.4
Error
band3
(%)
18.3
38.5
29.9
17.5
25.9
32.4
16.0
37.1
45.5
14.0
15.4
33.7
19.0
39.1
38.2
13.3
22.2
27.9
17.5
35.8
53.0
15.2
31.5
35.1
                                          240

-------
TABLE F-l. (Continued)



            SUMMARY OF  MODEL PERFORMANCE STATISTICS FOR MEAN CONCENTRATIONS FROM THE UAM.

Jul fan
date

160-76





195-76





211-76





212-76






Species

NO
NO?
03
OLE
NMHC
CO
NO
N02
°3
OLE
NMHC
CO
NO
N02
°3
OLE
NMHC
CO
NO
NO?
03
OLE
NMHC
CO
Mean observed
concentration
(ppm)
0.0091
0.0198
- 0.107
0.0130
0.297
0.606
0.00*2
0.0159
0.0827
0.0093
0.212
0.360
0.0074
0.0208
0.0462
0.0127
0.288
0.472
0.0102
0.0181
0.0664
0.0155
0.352
0.572
Mean predicted
concentration
(ppm)
0.0063
0.0207
0.0890
0.0065
0.325
0.504
0.0038
0.0154
0.0738
0.0033
0.237
0.396
0.0034
0.0166
0.0584
0.0061
0.313
0.473
0.0045
0.0186
0.0665
0.0062
0.295
0.429
Mean
residual
(ppm)
0.0028
-0.0009
0.0176
0.0066
-0.0286
0.102
0.0005
0.0005
0.0089
0.0061
-0.0251
-0.0357
0.0040
0.0043
-0.0122
0.0066
-0.0252
-0.0019
0.0057
-0.0005
-0.0001
0.0093
0.0573
0.143

RMS error
(ppm)
0.0147
0.0132
0.0237
0.0126
0.276
0.440
0.0041
0.0105
0.0179
0.0086
0.179
0.400
0.0090
0.0112
0.0214
0.0118
0.245
0.342
0.0137
0.0109
0.0219
0.0129
0.283
0.397
Mean residual
Mean obs. cone.
(*)
30.8
-4.5
16.4
50.8
-9.6
16.8
11.9
3.1
10.8
65.6
-11.8
-9.9
54.1
20.7
-26.4
52.0
-8.8
-0.4
55.9
-2.8
-0.2
60.0
16.3
25.0
Error
band*
(*)
15.3
28.7
59.0
14.7
22.6
30.2
18.0
41.1
75.0
13.7
31.1
27.1
11.6
38.1
38.7
14.3
30.2
24.5
17.6
31.2
56.9
13.1
30.4
27.6
                                          241

-------
  TABLE F-l.  (Continued)



              SUMMARY  OF  MODEL  PERFORMANCE  STATISTICS FOR MEAN CONCENTRATIONS FROM THE UAM.

Jul ian
date

225-76





226-76





237-76





275-76






Species

NO
NO?
03
OLE
NMHC
CO
NO
NOa
03
OLE
NMHC
CO
NO
NOe
03
OLE
NMHC
CO
NO
NO?
03
OLE
NMHC
CO
Mean observed
concentration
(ppm) '
0.0065
0.0208
0.0625
0.0116
0.264
* 0.479
0.0088
0.0236
0.0806
0.0147
0.324
0.693
0.0115
0.0192
0.0813
0.0225
0.511
0.649
0.0364
0.0507
0.0782
0.0304
0.691
1.74
Mean predicted
concentration
(ppm)
0.0044
0.0221
0.0391
0.0058
0.267
0.427
0.0056
0.0227
0.0545
0.0051
0.272
0.450
0.0052
0.0189
0.0538
0.0068
0.284
0.414
0.0264
0.0519
0.0878
0.0152
0.892
1.44
Mejm
residual
(ppm)
0.0021
-0.0013
0.0235
0.0058
-0.0030
0.0521
0.0031
0.0010
0.0261
0.0096
0.0527
0.244
0.0064
0.0003
0.0275
0.0157
0.226
0.236
0.0100
-0.0012
-0.0096
0.0152
-0.201
0.295

RMS error
(ppm)
0.0076
0.0124
0.0280
0.0110
0.257
0.343
0.0144
0.0155
0.0397
0.0142
0.285
0.570
0.0170
0.0140
0.0355
0.0212
0.448
0.562
0.0432
0.0319
0.0392
0.0197
0.394
1.35
Mean residual
Mean obs. cone.
(*)
32.3
-6.3
37.6
50.0
-1.1
10.9
35.2
4.2
32.4
65.3
16.3
35.2
55.7
1.6
33.8
69.3
44.2
36.4
27.5
-2.4
-12.2
50.0
-29.1
17.0
Error
band3
U)
22.4
35.2
26.8
18.9
23.7
32.4
18.5
27.9
20.2
15.4
33.5
25.1
14.6
30.2
20.4
7.8
24.0
32.1
13.5
35.1
36.7
11.8
25.1
27.2

aPercent of predictions within 25% of observations.
                                            242

-------
TABLE F-2.   SUMMARY OF MODEL PERFORMANCE STATISTICS FOR ANALYSIS OF TRENDS DATA FROM THE UAM

Julian
date
142-75
178-75
182-75
183-75
184-75
207-75
209-75
221-75
Species
N02
03
CO
N02
°3
CO
N02
°3
CO
N02
°3
CO
NOa
03
CO
N02
°3
CO
N02
03
CO
N02
°3
CO
Correlation
coefficient
•0.621
0.805
0.488
0.478
0.864
0.581
0.771
0.875
0.559
0.685
0.864
0.685 -
0.683
0.726
0.770
0.791
0.927
0.669
0.741
0.759
0.185
0.575
0.795
0.415
Average
spatial
correlation
0.689
0.174
0.633
0.598
0.357
0.735
0.518
0.571
0.395
0.466
0.435
0.527
0.549
0.444
0.583
0.734
0.461
0.533
0.536
0.411
0.524
0.612
0.395
0.321
Average
temporal
correlation
0.489
0.850
0.374
0.388
0.913
0.299
0.779
0.897
0.432
0.559
0.879
0-.479
0.521
0.696
0.435
0.682
0.959
0.528
0.682
0.670
0.650
0.615
0.818
0.308
Slope
0.665
0.887
0.246
0.447
0.867
0.275
0.523
0.610
0.172
0.574
0.614
0.326-
0.397
0.470
0.403
0.751
0.654
0.340
1.01
0.570
0.063
0.493
0.516
0.145
Intercept
(ppm)
0.0014
-0.0057
0.119
0.0096
0.0183
0.610
0.0105
0.0263
0.464
0.0099
0.0284
0.503
0.0170
0.0436
0.612
-0.0002
0.0373
0.249
-0.0044
0.0347
0.458
0.0100
0.0433
0.686
Sum of
squared error
(ppm2)
0.0183
0.186
5.330
0.0498
0.186
60.4
0.0380
0.142
75.4
0.0601
0.193
94.1
0.157
0.317
154.
0.0139
0.173
35.8
0.0209
0.192
224.
0.0419 '
0.268
177.
                                              243

-------
TABLE F-2. (Continued)
            SUMMARY OF MODEL PERFORMANCE STATISTICS FOR ANALYSIS OF TRENDS DATA FROM THE UAM

Julian
date
230-75
231-75
251-75
159-76
160-76
195-76
211-76
212-76
Species
N02
°3
CO
N02
°3
CO
N02
°3
CO
N02
03
CO
N02
03
CO
N02
°3
CO
N02
°3
CO
N02
03
CO
Correlation
coefficient
0.824
0.819
0.764
0.772
0.814
0.685
0.600
0.821
0.630
0.887
0.859
0.294
' 0.796
0.937
0.628
0.660
0.947
0.286
0.621
0.833
0.511
0.798
0.912
0.759
Average
spatial
correlation
0.804
0.361
0.694
0.706
0.436
0.694
0.410
0.451
0.616
0.676
0.392
0.556
0.747
0.592
0.368
0.654
0.495
0.485
0.685
0.421
0.519
0.642
0.614
0.639
Average
temporal
correlation
0.606
0.885
0.698
0.552
0.804
0.719
0.576
0.864
0.571
0.732
0.929
0.652
0.637
0.961
0.529
0.321
0.964
0.282
0.373
0.715
0.304
0.538
0.742 -
0.546
Slope
0.774
0.729
0.419
0.790
0.645
0.483
0.585
0.618
0.359
0.740
0.862
0.064
0.535
0.676
0.340
0.457
0.720
0.113
0.415
0.773
0.274
' 0.503
0.632
0.439
Intercept
(ppm)
0.0022
0.0059
0.354
0.0012
0.0176
0.217
0.0090
0.0232
0.574
0.0105
0.0021
0.923
0.0088
0.0464
0.435
0.0089
0.0296
0.315
0.0140
0.0010
0.342"
0.0088
0;0244
0.383
Sum of
squared error
( ppm2 )
0.0188
0.0950
58.6
0.0165
0.152
31.5
0.0635
0.150
159.
0.0546
o;i58
1700.
0.0565
0.247
61.5
0.0335
0.112
40.4
0.0322
0.104
32:i
0.0424
0.189
57.8
                                             244

-------
TABLE F-2.  (Continued)
            SUMMARY OF  MODEL PERFORMANCE STATISTICS FOR ANALYSIS OF TRENDS DATA FROM THE UAM
Jul ian
date
225-76
226-76
237-76
275-76
Species
N02
03
CO
N02
CO
NOg
03
CO
N02
CO
Correlation
coefficient
0.704
0.753
0.540
0.561
0.796
0.607
0.799
0.687
0.734
0.597 -
0.844
0.533
Average
spatial
correlation
0.657
0.377
0.573
0.688
0.110
0.707
0.793
0.354
0.561
0.761
0.239
0.622
Average
temporal
correlation
0.578
0.780
0.588
0.368
0.605
0.516
0.648
0.759
0.640
0.329
0.916
0.410
Slope
0.657
0.310
0.396
0.331
0.426
0.170
0.669
0.432
0.279
0.604
0.896
0.219
Intercept
(ppm)
Q.0063
0.0504
0.310
0.0161
0.0574
0.617
0.0065
0.0581
0.534
0.0193
-0.0004
1.42
Sum of
squared error
0.0333
0.247
30.8
0.0670
0.445
86.0
0.0476
0.383
108.
0.183
0.340
572~.
                                           245

-------
TABLE F-3.   SUMMARY OF MODEL PERFORMANCE STATISTICS FOR CONCENTRATION MAXIMA FROM THE  UAM
Julian
date Specie:
142-75
178-75
182-75
183-75
184-75
207-75
209-75
221-75
N02
°3
CO
NQ2
°3
CO
N02
03
CO
NOa
°3
CO
M02
°3
CO
N02
03
CO
M02
03
CO
N02
°3
CO
Observed
Peak Concentration Predicted Peak
2 Value Location
(ppm) (RAPS site)
0.0458
0.195
4.03
0.0701
0.202
4.04
0.0867
0.142
4.19
0.106
0.171
3.81
0.180
0.184
5.26.
0.0621
0.185
4.26
0.0501
0.209
14.4
0.0726
0.166
5.55
102
101
121
103
112
107
106
121
115
106
119
112
104
118
104
106
113
104
110
118
117
102
121
107
Time
(CST)
1300
1200
1700
0700
1400
0600
0800
1100
1400
0700
1000
0700
0700
1300
0700
0600
1400
0500
0800
1100
1700
0800
1500
0700
Value
(ppm)
0.0654
0.144
0.667
0.0517
0.162
1.22
0.0659
0.103
0.196
0.0955
. 0.154
1.39
0.0795
0.133
2.05
0.0530
0.141
1.63
0.0589
0.128
1.23
0.0478
0.144
1.44
Concentration*
Time
(CST)
1300
1400
1500
0900
1500
1400
0600
1400
1700
0800
1000 -
0700
0600
1400
0500
0600
1400
0500
0900
1100
0800
0500
1400
•0500
Residual
(ppm)
-0.0196
0.0460
3.36
0.0184
0.0400
2.82
0.0208
0.0390
3.99
0.0105
0.0170
2.42
0.101
0.0510
3.21
0.0091
0.0440
2.63
-0.0088
0.0810
13.2
0.0248
0.0220
4.11
Residual
Obs. cone.
m
-42.8
23.6
83.4
26.2
19.8
69.8
24.0
27.5
95.2
9.9
9.9
53.5
56.1
27.7
61.0
14.7
23.8
61.7
-17.6
38.8
91.7
34.2
13.3
74.1
                                            246

-------