EPA - 9O8X1 - 77 - OO1

JUNE 1977
  THE DEVELOPMENT
    OF A REGIONAL
AIR POLLUTION MODEI
 AND ITS APPLICATION
  TO THE NORTHERN
    GREAT PLAINS
   US. ENVIRONMENTAL PROTECTION AGENCY

       REGION VIII

     DENVER . COLORADO 8O295

-------
                                  EPA-908/1-77-001
Final Report
THE DEVELOPMENT OF A REGIONAL AIR POLLUTION
      MODEL AND ITS APPLICATION TO THE
           NORTHERN GREAT PLAINS
                     by

                Mei-Kao Liu
              Dale R. Durran

 (With Contributions from Mark J. Meldgin)

     Systems Applications, Incorporated
             950 Northgate Drive
        San Rafael, California  94903
          Contract No. 68-01-3591

              SAI No. LT77-48
               Project Officer

              Donald Henderson

Environmental  Protection Agency, Region VIII
         Office of Energy Activities
             1860 Lincoln Street
           Denver, Colorado  80203

                 July 1977

-------
                             DISCLAIMER

     This report has been reviewed  by  the U.S.  Environmental Protection
Agency, Region VIII and approved  for publication.  Mention of trade
names or commercial products  does not  constitute endorsement or recom-
mendation for use.

     This document  is available to  the public  through  the National
Technical  Information Service,  Springfield,  Virginia   22151.

-------
                               ABSTRACT
      This report describes a regional  scale air pollution model  and its
application to existing and proposed energy developments in the Northern
Great Plains.  The objective of this study was to examine the air quality
impacts roughly 100 to 1000 kilometers  from these point sources of emis-
sions.

      The regional model  is composed of two interconnected submodels:
 a mixing-layer model and a surface-layer model.  The mixing-layer model
 is designed to treat transport and diffusion above the surface.   The
 major feature of this model is the assumption that the pollutant distribu-
 tion is nearly uniform in the vertical direction.   This assumption per-
 mits adoption of a simplified form of  the general  atmospheric diffusion
 equation.  The compelling reason for this choice is that the vertical
 diffusion term in that equation is shown by dimensional analysis to be
 about 100 times greater  than the transport term.  The model  for  the
 surface layer (which is  embedded in the mixing layer) is designed to
 calculate pollutant fluxes to the ground.  For emissions from elevated
 sources or distant ground-level sources, most of the pollutant mass is
 contained in the mixing  layer.  The removal  processes thus consist of
 the diffusion of the pollutants through the surface layer to the ground
 and absorption or adsorption at the ground.   A unique feature of the
 surface-layer model is its ability to  incorporate  the diurnal  variation
 in surface temperature resulting from  daytime heating and nighttime
 cooling of the ground.  This variation affects the vertical  pollutant
 distribution through atmospheric stabilities, and  consequently,  affects
 the rate of surface uptake of pollutants.

      The regional model  was designed to predict concentrations of pri-
                                                                          o
 mary and secondary pollutants averaged over areas  of approximately 100 km

-------
with a temporal resolution on the order of 3 hours.  This model was thor-
oughly tested via sensitivity analysis.  The responses of the model were
consistent with expectations based on physical  reasoning.  This model was
exercised for all combinations of two emissions inventories (for 1976 and
1986) and three meteorological scenarios (a strong-wind winter case, a
stagnation spring case, and a moderate-wind summer case).  The predicted
SCL and sulfate concentrations are generally greatest in spring, inter-
mediate in winter, and lowest in summer.  From these preliminary results it
appears that neither the 1976 nor the 1986 emissions as estimated in this
study will cause SCL or sulfate concentrations  significantly higher than
background values at locations far from emissions sources.

-------
                          ACKNOWLEDGMENTS
     We would like to sincerely thank a number of individuals  whose  kind
and able assistance has been indispensable in carrying out this  project.
Terry Thoem, George Boulter, and Dave Maxwell in the Office of Energy
Activities of the Environmental Protection Agency (Region VIII)  provided
the necessary emissions data, and David Joseph, also from the  EPA Region
VIII, furnished all of the pertinent air quality and meteorological
measurements.  The analysis of the meteorological data was performed,
under a sub-contract with us, by Loren Crow.

     We would also like to take this opportunity to  express our  appreciation
to several of our colleagues:  Shep Burton, Terry Jerskey, and Phil  Roth  for
many stimulating discussions and constant encouragement;  Tom Myers and
Gary Lundberg for their help on computations; and Eric Mathre, Ron Rice,
and Bob Frost in the preparation of the data  base.

-------
                                   VI
                               CONTENTS

DISCLAIMER	    11
ABSTRACT	ill
ACKNOWLEDGMENTS 	     v
LIST OF ILLUSTRATIONS	    ix
LIST OF TABLES	    xi
   I   INTRODUCTION 	     1
  II   COAL DEVELOPMENT IN THE NORTHERN GREAT PLAINS  	 .     5
       A.  Potential Sources of Energy  	     6
       B.  Coal Mining Methods	     7
       C.  U.S. Coal Reserves	     8
       D.  Possible Uses of NGP Coal   	    14
       E.  Scenarios for Use of NGP Coal   	    16
       F.  Air Quality Impacts from the Use of NGP Coal	    18
PART A   DEVELOPMENT OF A REGIONAL AIR POLLUTION MODEL FOR THE
         SIMULATION OF POLLUTANT TRANSPORT AND DIFFUSION OVER
         LONG DISTANCES	    23
 III   OVERVIEW	    24
  IV   REVIEW OF PREVIOUS STUDIES 	    26
       A.  Swedish Studies	    26
       B.  Norwegian Studies   	    27
       C.  Finnish Studies	    28
       D.  Danish Studies	    29
       E.  British Studies	    30
       F-  Studies in the United States	    32
   V   MAJOR ATTRIBUTES OF LONG-RANGE DISPERSION MODELING 	    36
       A.   Transport and Diffusion	    36
       B.   Removal  Processes   	    42
           1.   Dry Deposition	    43
           2.   Wet Deposition	    44

-------
   V   MAJOR ATTRIBUTES OF LONG-RANGE DISPERSION MODELING  (continued)
       C.  Chemical Transformation  	   46
       D.  Subgrid-Scale Problems 	   47
  VI   DEVELOPMENT OF A REGIONAL AIR POLLUTION MODEL  	   52
       A.  The Mixing Layer Model	   56
           1.  The Model Equations	   56
           2.  The Numerical Method   	   61
       B.  The Surface Layer Model  	   64
           1.  Dry Deposition on Surfaces	   64
           2.  The Formulation of a Surface Deposition Model   ....   66
 VII   SENSITIVITY OF THE REGIONAL AIR POLLUTION MODEL  	   74
       A.  Horizontal Eddy Diffusivity  	   75
       B.  Mixing Depth	   81
       C.  Prescription of Dry Deposition	   81
       D.  Surface Reaction Rate	   88
       E.  S02/Sulfate Conversion Rate  	   88
VIII   SUMMARY AND CONCLUSIONS FOR PART A	   96
PART B   APPLICATION OF A REGIONAL AIR POLLUTION MODEL
         TO THE COAL DEVELOPMENT AREAS IN THE NORTHERN
         GREAT PLAINS	   97
  IX   OVERVIEW	   98
   X   COMPILATION OF THE DATA BASE	102
       A.  Emissions Data	102
       B.  Meteorological  Data	107
       C.  Surface Data	110
       D.  Air Quality Data	116
  XI   AIR QUALITY ANALYSIS	119
       A.  Winter	121
       B.  Spring	121
       C.  Summer	146
       D.  Air Quality Impacts	146

-------
                                  vm
 XII   SUMMARY AND CONCLUSIONS FOR PART  B	159



APPENDICES


       A   AN ANALYSIS OF NUMERICAL METHODS 	  16°



       B   COMPILATION OF SIMULATION RESULTS  	  174


                                                                      272
REFERENCES   	

-------
      IX
ILLUSTRATIONS
1
2
3

4
5

6
7

8
9

10
11

12

13

14

15

16

17
18
The Northern Great Plains 	
Schematic Illustration of Scales of Motion in the Atmosphere. .
Schematic Illustration of the Modeling Region in the Regional
Air Pollution Model Developed in This Study 	
Vertical Distribution of S0~ and SO^ Over Central Germany . . .
Schematic Illustration of Diurnal Variations in Surface
Deposition 	
Schematic Illustration of the Surface Layer 	
Horizontal Eddy Diffusivity as a Function of Traveling Time
and Plume Spread 	
Predicted S0? Concentrations for the Base Case 	
Predicted S0? Concentrations for Reduced
Horizontal DTffusivity 	
Predicted S0? Concentrations for Reduced Mixing Depths 	
S02 Deposition Velocities (in mn/sec) Calculated with e as
Prescribed by the Alqorithm of Owen and Thompson
S02 Deposition Velocities (in mm/sec) Calculated with 6 as
Prescribed by the Alqorithm of Thorn 	
Predicted $63 Concentrations for Reduced Surface
Reaction Rate 	
Predicted SO- Concentrations for Increased S02/Sulfate
Conversion Rate 	
Predicted Sulfate Concentrations for Increased S0?/Sulfate
Conversion Rate 	
Energy Conversion Facilities Scheduled for Completion
before 1936 	
Point Sources in the Northern Great Plains in 1976. . .
Point Sources in the Northern Great Plains in 1986. .
4
37

54
58

68
70

76
77

79
82

84

86

89

92

94

99
105
106

-------
19    Wind Measurement Networks in the Northern Great  Plains	108

20    Temperature Gradients and Exposure Classes at Idaho
      Falls, Idaho	Ill

21    Vertical  Thickness of the Modeling Region (in Meters)  	   112

22    Vegetation in the Northern Great Plains	H5

23    EPA S02 Monitors in the Northern Great Plains	H?

24    Winds at 850 Millibars Altitude During 27-31  January  1976  ...   122

25    Predicted S02 Concentrations for Winter Case	   132

26    Winds at 850 Millibars Altitude During 4-7 April  1976  	   135

27    Predicted S02 Concentrations for Spring Case	   143

28    Winds at 850 Millibars Altitude During 9-12 July 1975  	   147

29    Predicted S02 Concentrations for Summer Case	   153
                                                o
30    24-Hour-Average S0£ Measurements  (in yg/nr)  in the Northern
      Great Plains	   155

31    Predicted Concentration  Distributions  Using  the Upstream
      Difference Scheme  	  163

32    Predicted Concentration  Distributions  Using  the SHASTA
      Method	164

33    Predicted Concentration  Distributions  Using  the Egan and
      Mahoney Method	167

34    Variation of Amplification  Factor  |r|  as  a Function of a for
      e = 0.6	172

-------
                                 TABLES
  1    Mineral  Resource Terminology Adopted  by  the  Interior
      Department	
  "    Reserve Base of Coals in the Western United  States  by  Sulfur
      Content	    11

  ^    Characteristics of Three Northern Great Plains  Coals and
      Illinois Basin Coal	    13

  •",    Projected Production and Use of Northern Great  Plains  Coal
      in 1985	    17

  5    Estimated Emission Rates of Various Trace Elements  from
      Coal Combustion	    19

  G    Projected Increases in Statewide Emissions from Power  Plants
      and Coal Gasification Plants, 1974-1985, for Various Development
      Scenarios	    20

  7    Hydrocarbon and Oxide of Nitrogen Emissions  in the  Los Angeles
      Basin and the  Northern Great Plains	    21

  8    Mean Seasonal and Annual Morning and Afternoon  Mixing
      Heights and  Wind Speeds  for the Northern Great  Plains  	    40

  9    Comparison of Physical  Processes Pertinent to  Long-Range
      Pollutant Transport 	    41

10    S0? Removal  Processes	    42

11    Deposition of SO- onto  Vegetation	    44

12    Downwind Distance Traveled  by a Puff as a Function  of
      Atmospheric  Stability 	    60

13    Ps  do-Diffusivity in Advective Transport for a  10  Kilometer
      Grid and vAt/Ax =  1/2	    63

14    Surface  Resistance Measurements  for  S0? 	   100

15    Point  Sources  Emitting More  than  10,000 Tons of  SO  per Year
      in 1976	X	   103

-------
                                   xn
16    Point Sources Emitting More than 10,000 Tons of SO  per Year
      in 1986	?....-•   IW

17    Surface Roughnesses for Various Vegetation Types  	  •  •

18    Periods Chosen for Air Quality Analysis	  .  •  •   120

19    S0? Emissions and Areas of Ohio and the Northern Great
      Plains	
                                                                       -| ro
20    Significant Deterioration Increments for SO^	

21    Effective Diffusion Coefficients in the x-Direction  for the
      First Test Problem	   lbb

22    Estimated Computation Time Required To Follow a Plume  for
      750 km	    lb8

-------
                           I   INTRODUCTION


     The energy crisis, dramatically thrust into the  national  and  inter-
national scenes by the oil embargoes of 1973,  has probably  become  one
of the most challenging problems facing our society.   In  the  search  for
solutions to the problem, a variety of energy  sources have been pro-
posed—nuclear and solar energy, coal and  oil  shale, and many others.
However, it is only in the course of resolving  the myriad of problems
associated with the development or application  of these new sources of
energy that a painful realization has emerged:   the shortage of energy
is neither a short-term nor a isolated problem.  The raw materials
required in the development of new energy  sources, including renewable
sources, will become  increasingly scarce.  Obviously, concerns about
resource availability, as well as a wide range  of social, economic,
and environmental problems, will have to be carefully analyzed before
a rational approach for solving the long-range  energy problem can be
formulated.

     For the near future, the vast amount of accessible coal  reserves
in the U.S.  and the serious problems currently  plauging alternative
energy sources easily make coal one of the more attractive candidates
for coping with the energy problem.   In addition to simply being a
source of energy, coal is also an ideal substitute source of  petro-
chemical feedstocks.   It is thus interesting to note that of  the seven
goals set by President Carter in his April, 1977 address  to the nation
on energy, increasing coal production by about  two-thirds to  more than
one billion tons per year by 1985 is the only goal that is  not directly
related to energy conservation.

     Clearly the use of coal, particularly on a large scale,  will  pose
problems.   The most severe one appears to be the degradation  of our
air environment.   Direct combustion  of coal will undoubtedly  produce

-------
enormous amounts of air pollutants.   For example,  a  large power plant
without pollution control  equipment  typically emits  several  hundred
tons of sulfur dioxide per day,  as much as the entire Los Angeles
metropolitan area.  Uses of coal  other than direct combustion,  by  new
energy technologies such as coal  gasification for  example, may  also
generate large amounts of air pollutants.   Furthermore,  coal  mining,
transport of coal to power-demand population centers (an alternative  to
alleviate local environmental problems), and other activities related
to various modes of coal development can all generate appreciable
amounts of air pollutants.  As a  result of these emissions,  significant
deterioration of air quality in the  vicinity of major coal users is
an immediate problem.   The short-range air quality impact in adverse
meteorological conditions is generally characterized by extremely
high pollutant concentrations of short duration within several  kilometers
from a major emission source.  This  problem has been studied extensively.
For pollutants with relatively long  half-lives that  are emitted from
tall stacks, a different air pollution problem arises because of long-
range transport of these pollutants  and their derivatives.  On a time
scale of the order of several days and a spatial scale of several
hundred kilometers, the conversion of sulfur dioxide to sulfates,  for
example, becomes important.  Elevated sulfur dioxide and sulfate levels
may lead to a variety of environmental problems such as impacts on
ecological systems, reductions in visibility, and  acid rain.   In view
of the severity of these problems, characterizing  the long-range trans-
port of air pollutants has recently attracted considerable attention.

     The Northern Great Plains contains one of the world's largest
known coal reserves.  Immense deposits of coal* exist in northeastern
* Coal can be basically classified into four types:  lignite, sometimes
  referred to as brown coal; bituminous and subbituminous, known as
  soft coals; and anthracite, or hard coal.  Each type of coal has a
  different range of carbon and hydrogen content.  Eastern bituminous
  coal, from states such as West Virginia and Pennsylvania, generally
  has a higher sulfur content by weight than Western coals from states
  such as Montana and Wyoming, which are primarily subbituminous with
  some lignite.

-------
Wyoming, eastern Montana, and western  North  Dakota  (Figure  1).  As part of
an effort to achieve energy independence,  many  large  coal-fired power plants
using locally mined coal  have already  been built  in this  area.  Many more
power plants and coal  gasification  plants  are being built or planned.  (This
development  is  reviewed  in Chapter II.)  The Northern Great Plains  is
largely undeveloped at present,  and current  ambient concentrations of air
pollutants are low.  Consequently,  the stringent  Federal  regulations for pre-
venting significant deterioration  (Federal Register,  1974,  1975) apply to the
area.  It is clear that a careful  study of the  impact of  coal developments on
air quality is urgently needed.

      Under the sponsorship of Environmental Protection Agency,  this
 project has been initiated to address the general  problem  of maintenance
 of air quality in the Northern Great Plains.   The  primary  emphasis  of
 this project is to study the impact of SO-  emissions from  multiple
 point sources at large distances (on the order of  several  hundred  kilo-
 meters).   According to the original  plan, an existing  dispersion model
 suitable for assessing air quality impacts  at  large  distances  was  to
 be selected and adapted for the Northern Great Plains.   A  careful  review
 of the various models currently available revealed,  however, that  none
 of those models was adequate for handling multiple sources and  chemical
 reactions on the temporal and spatial scales of interest to the  present
 project.   Instead a new long-range transport model was developed.   A
 detailed  discussion of the development of this model can be found  in
 Part A of this report.  Subsequently, this  model was applied to  the
 Northern Great Plains to examine the impact of coal  developments.   The
 result of this application is described in  Part B  of this  report.

-------
7 - -.
                            CANADA
                                             NORTH DAKOTA
                 FIGURE 1.  THE NORTHERN GREAT PLAINS

-------
      II   COAL DEVELOPMENT IN THE NORTHERN  GREAT  PLAINS
     Vast amounts of coal underlie the Northern Great Plains  of the
United States.  This coal is being mined and will  continue to be mined
because its energy is needed.  The main issues concerning the use of
this coal are how it should be used and what level  of environmental
impact is acceptable.  This chapter is intended  to  provide perspec-
tive on these issues, particularly with respect to  air quality.

     There is no question that coal is needed.   In  1974 the total  U.S.
consumption of energy* was 7.31  x 10   BTU.   Petroleum liquids  and
natural gas produced in the U.S.A. accounted for about 60 percent of
the total.  At current rates of production,  U.S.  reserves  of  natural
gas and petroleum would be depleted in four  to  eight  decades.   These
estimates are highly uncertain because of the possibility  of  new dis-
coveries and the difficulty of quantifying known reserves.  In  addi-
tion, current rates are unlikely to persist.   Larger  and more easily
accessible deposits are generally extracted  first,  so further produc-
tion will  become more difficult  and costly.   Production of natural gas
in the United States has  declined since  1972, and oroduction of petroleum
will probably begin to decline after 1985 (Benedict,  1976).  These
declines may be reversed  temporarily by  deregulation  of the price  of
natural gas, extensive drilling  in Alaska and on  the  continental  shelf,
and by new production techniques such as  CO 2  injection, but the  con-
clusion is clear—the United States must look elsewhere for energy.
* By the First Law of Thermodynamics, energy is conserved, not consumed.
  "Energy consumption,"  as  used  here, means degradation of energy from
  concentrated forms into waste heat at near-ambient temperatures.

-------
A.   POTENTIAL SOURCES OF ENERGY

     To satisfy its demand for  energy,  the United States has placed greater
reliance on energy sources other than domestic  fossil  fuels.   For  example,
in 1950 only 12 percent of the  U.S.  consumption of petroleum liquids was
imported, but by 1974 39 percent  was.   The first nuclear power plant in the
United States was opened in 1957,  and by  1974 nuclear power supplied roughly
3 percent of the total  energy consumption.   Both of these sources have disad-
vantages.  Petroleum imports cause a  balance of pyaments problem and are
not politically controlled by the  United  States.  The safety of conventional
nuclear power plants and the storage  of nuclear wastes are matters of contro-
versy.  Aside from the safety issues, the planned and operating plants would
consume the estimated U.S.  reserves of  natural uranium concentrates in less
than 100 years (Hubbert, 1971:  Benedict,  1976).  Fast breeder reactors (cooled
by liquid sodium) are much more efficient; they could probably meet the pre-
sent demand for electricity for many years.  However, development of most
types of breeders in the United States  has been halted by President Carter
because their wastes can be reprocessed into atomic weapons.

      Other potential domestic  sources of energy include hydroelectric
 power, direct solar radiation, geothermal, wind, and tidal power, oil
 shale and tar sands, and fusion.  If  any of these sources were as
 economical as petroleum and natural gas, they would already have been
 developed.  Such is the case  for water power; most of the suitable
 sites in the U.S.  are either  in use or reserved for recreation and
 wilderness preservation.  Direct solar radiation provides an enormous
 amount of  power, but the costs of gathering and concentrating it with
 present technology are too high.   Geothermal, wind, and tidal power
 are  being  used on  a small  scale in favorable locations, but these
 resources  are not of sufficient magnitude  to  solve  the U.S.  energy
 problem.   Oil  shale and tar sands contain enormous amounts of oil;
                                                                   12
 proven U.S.  reserves of oil  in oil  shale are estimated at 2.3 x 10
 barrels,  ClnterTechnology Corp.  1971), or 12 times U.S.  petroleum
 reserves.   Separating  the hydrocarbons from the shale or  sand is expen-
 sive,  however.   In addition, current techniques  for producing oil from

-------
 shale use 1  barrel  of fresh  water  for  each  barrel  of oil produced, and
 the  material  left after the  oil  is  extracted takes up a greater volume
 than the original oil  shale.   The  energy  available from fusion is enormous.
 The  fusion of the estimated  world  reserves  of minable lithium-6 with
 deuterium, which appears  to  be the  most practical  fusion reaction, would
 provide roughly as  much energy as  the  total  world  supply of all fossil
 fuels (Hubbert, 1971).   But  fusion  is  not expected to be a commercial
 energy source before  the  next  century.  In summary, none of these energy
 sources is expected to  reduce  our dependence on petroleum and natural
 gas  in the next few decades.

      The remaining  energy source is coal.   Coal  is versatile--!'t can
 be converted  to gaseous or liquid  fuels or  petrochemical feedstocks, or
 it can be burned directly to generate  electricity.  (Note  that nuclear
 power is efficient  only for  generating electricity.)   The  technology
 to mine coal  is well  developed, and promising  new  technologies  are being
 investigated.   Finally,  the  United  States has  very large coal  resources.
                                                    12
 The  "identified resources" of  the  U.S. are  1.7 x  10   short tons
 (Averitt, 1974), or roughly  3.4 x  1019 BTU.

 B.    COAL MINING METHODS

      Before discussing  coal  resources, it is helpful  to  consider how
 coal  is mined.   Coal  generally occurs  in  layers or seams.   These seams
may be  25 feet thick or more.  Coal is  mined by both subsurface and surface
 techniques.   In subsurface mininq,  the  roof  of the  mine must be supported
 (at least temporarily).   In some techniques, such as the traditional  room-and-
 pillar method, the roof is supported by leavinq 30  to  50 percent of the coal
 in place.  In longwall mining,  coal  is  sheared  and  removed  from a lonq  face
underuround, and the roof is  supported  by  hydraulic .lacks.   As  coal  is  removed
 the hydraulic lacks are advanced, and  the  roof  behind  is allowed tn collapse.
Lonnwall mlhinq recovers more of the coal  in a  seam than most  subsurface
techniques, but at present it is applicable  only to certain types of rock
strata.   In the United States the averaqe  recovery  factor for  coal  from all
types of subsurface mines is  57 percent (Nephew, 1973).

-------
     Surface mining techniques include strip mining  and augering.   In
strip mininq the material  above the coal  seam,  or  the  overburden, is
cleared by heavy excavating equipment  and the coal is  removed.  Then
the overburden I rum an adjacent area ii moved  into the cleniiMl -lira,
exposing more coal.  This  process  is repeated until  the overburden  is
too thick to be handled economically.   After mining  (and before reclamation),
a surface-mined area is generally  covered with  piles of broken overburden,
or spoils, and at one end  of the area  is  a nearly  vertical wall where the
stripping was stopped.   The recovery factor for coal is typically 80 percent
for strip mining and 50 percent for augering (Nephew,  1973).  Surface mines
are generally safer and more productive (in tons of  coal per day per employee)
than subsurface mines,  and they have steadily increased in importance.  The
percentage of total U.S. coal  production  obtained  from surface mines increased
from 22 percent in 1950 to 50  percent  in  1974 (Nephew, 1973; Nehring and
Zycher, 1976).

 C-   U.S. COAL RESERVES

      Estimate  of coal  reserves vary widely.    In some studies the  mini-
 mum thickness  of subbituminous coal that is  considered economically
 minable is  3 meters, in other studies it is  1.5 meters, and in some
 studies minability is  ignored.  Some studies  include all  coal within
90  meters of the  surface, others  include all  coal within 1800 meters.
All estimates are  based on extrapolation from  limited geologic data.
Finally, what is being estimated  differs. Coal deposits  that are or
may be  minable are  generally  termed reserves;  resources commonly
include all  coal, whether presently minable or not.  Reserves,  however,
are divided differently in different  estimates and  sometimes  reserves
are called resources.  The terminology adopted by the U,S.  Department
of the Interior is given  in Table  1  (EPA,  1976a, p. 143).

     The distribution of  coal  in the western United States is given
in Table 2.  In this table  "total   reserve base" is equivalent  to

-------
                 TABLE 1.    MINERAL RESOURCE TERMINOLOGY ADOPTED
                            BY THE INTERIOR DEPARTMENT
 DEFINITIONS:

 Resource - A  concentration of naturally occurring solid, liquid, or
      gaseous  materials in or on  the  earth's crust in such form that
      economic extraction of a commodity is  currently or potentially
      feasible.

 Identified Resources  - Specific  bodies  of mineral-bearing material
      whose location,  quality,  and  quantity  are known from geologic
      evidence supported by engineering  measurements  with respect to
      the demonstrated category.

                              TOTAL RESOURCES
l

Economic
Paramar-
ainal
IT3 ID
E C
.O ••-
1/1
IDENTIFIED
Demonstrated
Measured
Indicated
Inferred
RESERVES


* *
RESOURCES
UNDISCOVERED
HYPOTEHTICAL
(in known
districts)
*
SPECULATIVE
(in undis-
covered
districts)

h
_o
to
O)
0
•1 —
E
O
c
o
o
o>
it-
CD
	 01
O)
cn
01
c
	 10
fO
Ol
S-
u
c
            Increasing degree of geological assurance

Undiscovered Resources - Unsoecified bodies of mineral-bearing material
     surmised to exist on the basis of broad geologic knowledge and
     theory.

Reserve - That portion of the identified resource from which a usable
     mTneral and energy commodity can be economically and legally
     extracted at the time of determination.  The term ore is also used
     for reserves of some minerals.

-------
                                    10
                            TABLE 1  (Concluded)


 The following definitions for measured, indicated,  and inferred are
 applicable to both the Reserve and  Identified-Subeconomic resource
 components (see chart).

 Measured - Material  for  which estimates of the  quality and quantity
      have been computed, within a margin of error of less than 20
      percent, from analyses and measurements from closely spaced and
      geologically well-known sample sites.

 Indicated - Material  for which estimates of the quality and quantity have
      been computed partly from sample analyses  and  measurements and
      partly from reasonable geologic projections.

 Demonstrated - A collective term for the sum of materials in both measured
      and indicated resources.

 Inferred - Material  in unexplored but identified deposits for which
      estimates of the quality and size are based on geologic evidence
      and projection.

 Identified-Subeconomic Resources -  Known deposits not  now minable econ-
      omical ly.

 Paramarginal  - The portion  of subeconomic resources that  (a) borders on
      being economically  producible  or (b) is not commercially available
      solely because  of legal  or political  circumstances.

 Submarginal - The portion of subeconomic resources  which  would require
      a substantially  higher price (more than 1.5 times  the price at the
      time of  determination)  or a major cost-reducing advance in technology.

 Hypothetical  Resources -  Undiscovered materials  that may  reasonably be
      expected to  exist in a  known mining district under known geologic
      conditions.   Exploration  that  confirms  their existence and reveals
      quantity and  quality will  permit their  reclassification as a Reserve
      or identified-subeconomic resource.

 Speculative Resources -  Undiscovered  materials  that may occur either in
      known  types of deposits  in  a favorable  geologic setting where no
      discoveries have been made,  or  in  as  yet unknown  types  of deposits
      that remain to be recognized.   Exploration  that confirms their
      existence and reveals quantity  and  quality  will permit their re-
      classification as reserves  of  identified-subeconomic resources.

Source:   EPA (1976a).

-------
                                         11
                   TABLE 2.  RESERVE BASE  OF COALS IN  THE WESTERN
                             UNITED STATES BY SULFUR CONTENT
                                (in 106 short tons)

State
Alaska
Arizona
Arkansas
Colorado
Iowa
Kansas
Missouri
Montana
New Mexico
North Dakota
Oklahoma
Oregon
South Dakota
Texas
Utah
Washington
Wyoming
Less than
1.0 percent
11,457.0
173.0
81.0
7,476.0
1.6
0
0
101,646.9
3,575.5
5,389.0
275.0
1.5
103.1
659.8
1,968.5
603.5
33,912.0

1.1 to 3 percent
184.0
177.0
463.0
786.2
226.7
309.3
182.0
4,115.3
793.5
10,325.5
326.6
.3
287.9
1,884.7
1,546.8
1,265.4
14,657.4
 *Total *
167,324.5
37,531.5
Greater than
3 percent

     0
     0
    46.0
    47.3
 2,105.9
   695.6
 5,226.0
   502.6
      .8
   268.7
   241.4
    0
    35.9
   284.1
    49.4
    39.0
 1,701.1

11,244.1
                                                                 Unknown
                                                                 Content
                                                              Total*
                                                              Reserve Base
0
0
74.0
0
549.2
383.2
4,080.5
2,166.7
27.5
15.0
450.5
0
1.0
444.0
478.3
45.1
3,060.3
11,645.0
350.0
665.7
14,869.2
2,884.9
1,388.1
9,487.3
108,396.3
4,394.8
16,003.0
1,294.2
1.9
428.0
3,271.9
4,042.5
1,954.0
53,336.1
                                                                 18,323.0
                                                             234,412.4
* Totals may not  add due to  rounding.

Note:   Total reserve base for  the entire U.S. is  4.37 x 10    tons, 2.00  x  10
        tons of which have less  than 1  percent sulfur.

Source:   Bureau of  Mines (1975).

-------
                                   12
 "demonstrated reserves" as shown in Table  1; that  is, coal believed
 to exist in  thicknesses and at depths similar to that being  mined  on
 the  basis of preliminary geologic and engineering  evaluations.   No
 allowance is made  for  recoverability factors of roughly  50  percent for
 subsurface mining  and  80 to 90 percent for surface mining.   Montana,
 North Dakota, and  Wyoming account for 41 percent of the  demonstrated
 reserves and 70 percent of the demonstrated reserves having  less than
 1  percent sulfur in the entire U.S.   (The importance  of  the  sulfur
 content  is discussed in Section F.)   These figures  include the Wasatch
 and  Fort Union formations, which are generally  considered the Northern
 Great  Plains coal  field,  and minor coal fields  in  western Montana  and
 Wyoming.

     Regardless of how it is estimated or labeled,  the coal   in the
 Northern Great Plains  can provide an enormous amount of  energy.  Accord-
 ing  to Nehring and Zycher (1976, p.  20), the "most probable  estimate
 of ultimate  strippable resources [in the Northern Great  Plains]  is  equal
 to 26 times  U.S. energy consumption in 1974."  By "strippable resources"
 they mean coal in  seams more than 1.5 meters thick  lying under less
 than 90 meters of  overburden.  The known strippable resource in Montana,
 North Dakota, and  Wyoming is 7.89 x  10   tons*,  or  45 percent of the
 total U.S.  strippable  resource, and  its average heat content is esti-
 mated to be  7.9 x  10   BTU/lb, which  corresponds  to  lignite or sub-
 bituminous  coal.    Almost three-fourths of the coal  contains  less than
 0.6  Ibs sulfur/10  BTU.  The deposits in Montana and Wyoming generally
 have a higher heat content and lower sulfur content than the North
Dakota  deposits  (Nehring and Zycher, 1976).

     The  coal in  the Northern Great  Plains differs  from eastern coal
 in many respects.   Table 3 gives some properties of three coals
from  the  Northern  Great Plains and an average coal  from the  Illinois
Basin,  which  is  typical of many Eastern bituminous coals.  Note that
* For comparison, the total  U.S.  production of coal  in 1972 was 5.97 x 10
  tons,  which was valued at  almost 4.5 billion dollars.

-------
                      TABLE 3.   CHARACTERISTICS  OF THREE NORTHERN  GREAT  PLAINS
                                 COALS  AND  ILLINOIS BASIN COAL
                                                 Sulfur
Ash
Northern Great Plains
Coal I
Coal II
Coal III
Illinois Basin
Average of
82 coals
Heat Content Percentage Percentage Moisture
[BTU/lb (dry)] lbs/106 BTU (dry) lbs/106 BTU (dry) (percent)

9511 0.76 0.72% 21.7 20.6% 27.0%
11708 0.42 0.49 6.15 7.2 29.2
9838 1.46 1.44 12.6 12.4 36.8

12750 2.75 3.51 8.85 11.28 10.02
Source:  EPA(1976a).
                                                                                                            U)

-------
                                   14
 these NGP coals  are  lower  in heat content and sulfur content and higher
 in moisture content  than Illinois Basin coal.  Even on a heat equivalent
 basis, NGP coal  has  much less sulfur than Illinois Basin coal.  For
 reasons discussed  below, this low sulfur content is the main reason for
 development of NGP coal.

 D.   POSSIBLE  USES OF NGP  COAL

      One  of the  main issues in the use of Northern Great Plains coal
 is how it should be  used.  As mentioned above, coal is a versatile fuel;
 it can be used in  the following ways:

      > Burning  to generate electricity
      > Conversion to low- or medium-BTU gas
      > Conversion to synthetic natural gas
      > Conversion to liquid fuels or petrochemical feedstocks.

 These uses  are discussed in turn below.  Note that each use can take
 place either near the mine or at a distance.

      Burning coal to generate electricity is its most common use.  For
 example,  63 percent of the coal  mined in the U.S. in 1973 was burned
 in coal-fired steam electric generating plants (FEA, 1975).  When coal
 is burned, about 95 percent of the sulfur in it is converted to gaseous
 sulfur oxides (Smith, 1966).   Federal and state limits on sulfur emis-
 sions  are one of the major  forces behind the development of coal gasi-
 fication and liquefaction processes.

     Coal  gasification  and  liquefaction involve unavoidable energy
 losses, but they remove much of the sulfur and ash, converting coal
 into clean-burning  forms.   Dried  coal typically has a hydrogen to car-
 bon ratio  of about  0.8.   For comparison, crude oil has a ratio of
 about 1.1  and natural gas  has  a  ratio of about 4.0, so coal conver-
sion requires a source  of hydrogen,  usually steam.

-------
                                   15
      Coal  gasification is not new; it was widely used in the United States
 until  the  1930s, when natural gas became available, and is used at pre-
 sent  in  many  foreign countries.  In modern coal gasification processes,
 pulverized coal and steam are heated together under pressure.  Heat is
 produced in part by adding to the steam some air or oxygen, so the coal
 can burn slightly.  The product is a gas consisting of CO, H^, CH., CO-,
 HpO,  HoS,  other organic gases, and N~.  H~S can be removed efficiently,
 so the product gas is low in sulfur.  The heat content of such gas is low,
 perhaps  100 to 200 BTU/scf if air is used and 250 to 500 BTU/scf if
 oxygen is  used, compared to 1000 BTU/scf for natural gas.  At this stage
 the synthesized gas cannot be piped long distances economically, so it
 may be either burned near the plant for electricity generation, or con-
 verted to  high-BTU gas by "shift conversion" (i.e., CO + H^O ->- COo + H~)
 and catalytic methanation.  Commercial coal gasification processes have
 efficiencies  of 80 to 90 percent in producing low- or medium-BTU gas and
 60 to  70 percent in producing high-BTU gas (Tillman, 1976).

     Coal  liquefaction processes are less well  developed than coal gasi-
 fication.   Liquefaction is carried out for a variety of reasons:   to
 remove sulfur and inorganics before combustion, to produce petrochemical
 feedstocks  or substitutes for crude oil, or to  produce fuel-grade methanol.
The processes currently proposed include pyrolysis, solvent refining,
and catalytic hydrogenation at high temperatures and pressures.   A great
deal  of  research in coal  liquefaction is being  carried out in this
and other countries,  some pilot plants have been built,  and  one plant
operating in South Africa, but gasification is  generally expected to be
more  important than liquefaction in the short term.

     Coal from the Northern Great Plains could  be transported economically
by either rail or slurry  pipeline.   So-called "unit trains,"  which often
contain  100 hopper cars,  travel  as  units from mine to point  of use and
back,  and seldom uncouple.   Unit trains are commonly used at  present in
the Northern Great Plains.   In slurry pipelines, finely  pulverized coal
is mixed with approximately an equal  weight of  water and pumped.   Slurry

-------
                                  16
pipelines are claimed to transport coal  at  roughly  one-half  the cost of
transport by unit trains, but none has yet  been  built  in  the Northern
Great Plains because railroad companies  have  not granted  permission
to  let  pipelines cross their rights-of-way  (C&EN, 1977).

E.   SCENARIOS FOR USE OF NGP COAL

     Future development of the coal in the  Northern Great Plains depends
on  many factors, including leasing policies for  public  and Indian lands,
environmental regulations, and the price of imported crude oil.  Con-
sequently, forecasting coal production and  use is complex.  The Northern
Great  Plains Resource Program (1974) assembled two  forecasts of interest
here,  based on scenarios of most probable development and extensive
development.  The mining and use of coal in these scenarios is summarized
 in  Table 4.   Nehrincj and Zycher  (1976) suggest that  the projections
of  the  most probable scenario will be exceeded because  coal production
 under  contracts already signed nearly equals  these  projections, and few
 contracts  are signed more than five years before initial  delivery.

      The present report deals primarily with  atmospheric  sulfur dioxide
 and sulfates  in the Northern Great Plains,  but to provide a broader
 perspective we briefly discuss other impacts  of  projected large-scale
 use of coal.  A major hindrance  to development of NGP coal is scarcity
 of water.   A  coal-fired power plant with evaporative cooling requires
 roughly 4 tons of water for each ton of  coal  burned.  (Dry cooling is
 much less efficient.)  High- and low-BTU coal gasification require
  roughly 1.0 and 0.1 tons of H^O  per ton  of  coal  if  air  cooling is used
  extensively (NGPRP, 1974, pp. 129-130; Radian Corp., 1975, p. B-19).
  The extensive development forecast of  the NGPRP  thus calls for use of
         ft                                    5
  3.1 x  10  tons of water per year, or 2.3 x  10 acre-feet. For compari-
  son, the mean annual flows of the two  major rivers  in  southeastern
  Montana and northeastern Wyoming, the  Tongue  River  and  the Powder River,
  are 3.0 x 105 and 3.3 x 105 acre-feet, respectively (Nehring and  Zycher,
  1976).  Coal  development will therefore  require  extensive use of  ground-
  water, or pipelines on the order of 100  miles in length to transport

-------
                               17
       TABLE 4.   PROJECTED PRODUCTION AND USE OF NORTHERN
                  GREAT PLAINS COAL IN 1985

                    (10  short tons per year)

                       (a)  Most Probable
State
Montana
North Dakota
Wyomi ng
Total
Production
74.2
49.1
60.5
183.8
HCI
<0.05
1.0
<0.05
1.0
EG-M
11.0
19.1
4.2
34.2
EG-0
41.0
9.0
40.8
90.8
CG
22.2
20.1
15.5
57.8
(b) Extensive Development
State
Montana
North Dakota
Wyomi ng
Total
Production
150.9
89.9
133.4
374.2
HCI
<0.05
1.0
<0.05
1.0
EG-M
11.0
19.0
4.2
34.2
EG-0
77.0
10.0
78.8
165. C
CG
62.9
59.9
50.4
173.2
Key:   HCI = household,  commercial,  and  industrial
      EG-M = electricity generation  near mine
      EG-0 = electricity generation  out  of state
        CG = coal  gasification  to  produce high-BTU  gas.

Source:   NGPRP (1974);  Nehring  and Zycher (1976).

-------
                                   18
water to coalfields from larger rivers in nearby drainage basins, such
as the Yellowstone, Big Horn, and Missouri Rivers.  Water availability
is only one phase of the problem; water rights, interstate water com-
pacts, and other legal requirements must also be dealt with.

     The land area to be used for coal development activities is
extensive.  For example, known strippable coal deposits in Montana
and Wyoming occupy an area equal, to the combined areas of Delaware and
Rhode Island (Nehring and Zycher, 1976).  The area disturbed by a
single coal mine in the Northern Great Plains producing 3 x 10  tons
of coal per year for 30 years is 15 to 178 square miles, depending on
the thickness of the coal seam(s) being mined (Edwards, Broderson, and
Hauser, 1976).  Unless reclaimed, mined areas may become large sources
of fugitive dust.  Various types of reclamation are currently being
carried out at coal mines in the Northern Great Plains (EPA, 1976a),
but they are hindered by the low average rainfall  in the region.

F.   AIR QUALITY IMPACTS FROM THE USE OF NGP COAL

     Air pollutants emitted from coal mining, transportation, burning,
and conversion include various trace compounds, nitrogen oxides, par-
ticulates, hydrocarbons, carbon monoxide, and sulfur oxides.  Trace
compounds include both chemical elements present in small  amounts and
complex hydrocarbons that are formed or released during coal burning
and gasification.  Trace elements found in coal that are hazardous in
excessive (though small) amounts include arsenic, beryllium, cadmium,
fluorine, lead, mercury, and selenium (Magee, Hall, and Varga, 1973;
Kaakinen, Jorden, and West, 1974).   Except for selenium, which is
enriched in coal by a factor of ten, these elements are contained in
coal in roughly the same concentrations as in the earth's crust.  A
portion of these elements may enter the atmosphere after being subjected
to a hot oxidizing atmosphere in a  coal burner (see Table 5), or
they may enter the water supply by leaching from ash, mines, or spoils.
Trace compounds may also cause environmental problems.  Many carcin-
ogenic organic compounds have been identified in emissions from
industrial boilers and output from coal gasification plants.  At

-------
            TABLE  5.  ESTIMATED EMISSION  RATES  OF  VARIOUS  TRACE
                     ELEMENTS FROM COAL  COMBUSTION

            Trace  Element       Emissions rate (1bs/10  ton)
              Arsenic                       2.9
              Beryllium                     3.7
              Cadmium                       1.0
              Manganese                     1.0
              Mercury                       0.4
              Nickel                         3.0
              Vanadium                       0.5
             Source:  EPA (1973).

present  the potential degree of  hazard of these compounds in the
environment is unknown.  A thorough review of both trace elements and
trace  compounds is given by Radian Corp. (1975, Vol. Ill, App.  D).

     The major air pollutants from coal utilization, namely nitrogen
oxides,  hydrocarbons, particulates, carbon monoxide, and sulfur oxides,
have been studied far more extensively than trace compounds.  Forecasts
of the emissions of these pollutants from coal-fired power plants and
coal gasification plants in the Northern Great Plains are presented by
NGPRP  for the two scenarios mentioned above (most probable and  extensive
development), and a scenario based on information derived from  state
agencies, utility companies, newspaper articles, and so on, which we
will term "planned development."  These estimates, listed in Table  6,
are based on many assumptions, including the attainment of Federal New
Source Performance Standards; in general they indicate maximum  or worst-
case emissions (NGPRP, 1974, p.  122).   Table 6 also  lists these  emis-
sions as percent increases over total  statewide emissions in 1972.

     Estimated increases in  emissions  of particulates and carbon monoxide
from coal utilization are small  fractions of  current statewide emissions.
Since these emissions come from  point  sources,  it  is possible that they

-------
          TABLE  6.   PROJECTED  INCREASES  IN EMISSIONS  FROM POWER PLANTS AND  COAL GASIFICATION
                     PLANTS, 1974-1985, FOR VARIOUS DEVELOPMENT SCENARIOS
                                                   Emissions  Increase for Given Scenario (In 10  tons/year)*
Most Probable
Pollutant
Participates


Sulfur Oxides


Nitrogen Oxides

Hydrocarbons


Carbon Monoxide


State
Montana
North Dakota
Wyoming
Total
Montana
North Dakota
Wyoml ng
Total
Montana
North Dakota
Wyoming
Total
Montana
North Dakota
Wyoml ng
Total
Montana
North Dakota
Wyoming
Total
Power
Plants
9.5
11.2
9.3
31.0
114.1
135.1
111.3
360.5
66.5
78.7
64.9
210.1
2.0
2.4
2.0
6.4
6.8
8.0
6.6
21.4
Coal
Gasification
5.7
3.8
3.8
13.3
64.6
43.1
43.1
150.8
31.1
20.7
20.7
72.5
261.2
174.1
174.1
609.4
2.3
1.5
1.5
5.3
Percent
Increase^
6
13
9
43
202
212
89
69
81
147
156
173
1
2
2
Extensive Development
Power
Plants
9.5
11.2
9.3
31.0
114.1
135.1
111.3
360.5
66.5
78.7
64.9
210.1
2.0
2.4
2.0
6.4
6.8
8.0
6.6
21.4
Coal
Gasification
15.3
11.5
11.5
38.3
172.2
129.2
129.2
430.6
82.9
62.2
62.2
207 3
696.6
522.4
522.4
1741.4
6.1
4.6
4.6
15.3
Percent
Increase1
9
20
14
69
300
329
135
98
120
390
465
514
1
2
3
Planned Development
Power
Plants
7.1
24.5
11.0
42.6
85.3
294.4
132.0
5)1.7
49.7
171.7
77.0
298.4
1.5
5.2
2.3
9.0
5.0
17.4
7.8
30.2
Coal
Gas1 fication
0
1.9
1.0
3.8
0
21.5
21.5
43.0
0
10.4
10.4
20.8
0
87.1
87.1
174.2
0
0.8
0.8
1.6
Percent
Increase1'
3
23
8
21
359
210
45
126 ro
o
82
1
82
07
1
3
2
* Emissions associated with coal mining are not Included.

* Percent  Increase over  total  emissions In state In 1972.

Source: NGPRP (1974).

-------
                                   21
may degrade air quality near the sources.   (Note  that  these emissions
estimates do not include the impacts of mining, which  may  be a large
source of particulates.)

     For hydrocarbons and nitrogen oxides, the precursors of photochemi-
cal oxidant, emissions  from coal utilization substantially increase the
total statewide emissions.  Note that in all three scenarios coal gasi-
fication produces 90 percent or more of the hydrocarbon emissions from
coal utilization.  Some perspective on these emissions may be gained by
comparing them with emissions in the Los Angeles  Basin, as given in
Table 7.  The Los Angeles Basin is roughly 2,000  sq. mil.  in area;
eastern Montana, North  Dakota, and Wyoming encompass 267,000 sq. mi.
The NGPRP report  provides information on the composition of hydrocarbon
emissions from gasification plants.  In view of the end product, it
is possible that these  emissions are largely methane,  which is relatively
unreactive in photochemical oxidant production.

        TABLE 7. HYDROCARBON AND OXIDE OF NITROGEN EMISSIONS IN THE
                 LOS ANGELES BASIN AND THE NORTHERN GREAT PLAINS
                                (103 tons/year)

                                        Northern  Great Plains  (1985)
                                        Most Probable     Extensive
      Species       Los Angeles Basin    Development    Development
  Hydrocarbons            950*              615             1750
  Nitrogen Oxides         400+              280              420
  * Data for 1972 from Trijonis and Arledge (1975).
  * Data for 1973 from LAAPCD (1974).

     Perhaps the most serious air pollution problem from coal utiliza-
tion is emission of sulfur oxides.  Table 6 shows that coal  utiliza-
tion in the Northern Great Plains will substantially increase state-
wide emissions of sulfur oxides.  This is ironic because the low  sul-
fur content of NGP coal is the prime motivation for mining it.  Sulfur

-------
                                   22
oxides cause damage to vegetation and  the  respiratory system.  In addi-
tion, it is believed  that  they can cause acid rain as much as 1000 km
downwind from sources.   Because  of these effects, the EPA and individual
states have established strict controls on SO  emissions from power
plants. The EPA standard fs 1.2  Ibs  SO/106 BTU input, or 0.6 Ibs S/106 BTU.
                                      A
Thus subbituminous coal with a  heat content of 8000 BTU/lb and a sulfur con-
tent greater than approximately  0.5  percent can be burned only if some
method is employed to recover sulfur compounds.   Much NGP coal meets the
EPA standard, but most coal from the eastern U.S.  does not.  Sulfur recov-
ery methods include flue gas desulfurization (FGD), or scrubbing, after
burning and coal gasification, liquefaction, solvent refining, and wash-
ing before burning.  The feasibility and  costs  of using these methods are
matters of controversy (ES&T, 1976).

-------
                   23
               PART A

      DEVELOPMENT OF A REGIONAL
AIR POLLUTION MODEL FOR THE SIMULATION
OF POLLUTANT TRANSPORT AND DIFFUSION
         OVER LONG DISTANCES

-------
                                     24
                           III    OVERVIEW


      At  present,  emissions from new coal-fired electric generating plants
 are  regulated  by  the New Source performance Standards promulgated by the
 Environmental  projection Agency (EPA, 1975).  Allowable ambient concen-
 trations of many  pollutants are also specified 1n various Federal and
 state standards,   Furthermore, since most of the coal reserves are
 located  In,  largely undeve]oped areas where current ambient concentrations
 are  low, more  stringent federal regulations for prevention of significant
 deteriorate  Pf  air quality apply (Federal Register, 1974, 1975).  In
 order to meet  these statutes, differept air pollution control techniques,
 ranging  frpfl] direct clean-up a!t the stac|< to indirect methods such as
 tall  stacks apd the Supplementary Control System,* will have to be con-
 sidered,  If Indirect control strategies are adopted, they will relieve
 air  quality problems 1n the Immediate vicinity pf pollutant sources, par-
 ticularly u^der worst-cflse meteorology,   These control strategies do not
 reduce the emissions of pollutants, however, they are just released at
 greater  heights anc| probably piore uniformly.  Cpnsequently, primary pol-
 lutants  pap be expected tp have longer residence times, and a net degra-
 dation of «|1r quality can be expected at large distances from the sources.
 Longer residence times for primary pollutants 1n the atmosphere also
 promote  the formation of secondary pollutants.  This effect can be seen
 1n many  critical environmental  problems  that have been discovered recently,
 such as  the observation of high SUlfqte  levels, the Increase of acidity in
 rain, the reduction in visibility 1n  many pristine regions, and the obser-.
 vatlon of plevated oxidant concentrations over rural or semi-rural areas.
These Imposing problems have led  to research on air quality problems at
 large, reglpnal scales,
  The Supplementary Control  System (SCS) is a time-variable emissions
  control  scheme based pn load curtailment or fuel  switching during
  meteorological conditions  of low dispersion.

-------
                                    25
     Considerable effort has been expended in the past few years in
attempts to obtain a quantitative understanding of long-range transport
and to develop mathematical models for predicting air quality impacts.
A review of previous studies pertinent to modeling of long-range air
pollutant transport is presented in Chapter IV.  A close examination of
these models revealed that none of the models is adequate for handling
multiple sources and chemical reactions on the temporal  and spatial scales
of interest in this project.  It was thus decided that,  instead of adapt-
ing an existing model as originally planned in this project,  a new regional
air quality model would be developed.  Part A of this report  is devoted
to the description and discussion of this model.  To provide  a general
background for modeling, various physical processes pertaining to the
long-range transport of air pollutants are delineated in Chapter V.  The
model developed in this project adopts a grid modeling approach and is
composed of a mixing layer model and a surface layer model.   The develop-
ment of this model and its components is described in Chapter VI.   The
model results appear to be affected by a number of physical parameters.
To explore the effects of varying these parameters on the model  predic-
tions, a sensitivity test of the model was carried out as discussed in
Chapter VII.  Part A closes with a brief chapter of summary and conclu-
sions on the development of the model.

     The regional air quality model developed in this project was  applied
to the Northern Great Plains to examine the impact of coal  development
in that area.   A detailed description of this application is  given in
Part B of this report.

-------
                                    26
                   IV    REVIEW OF PREVIOUS  STUDIES


     A variety of mathematical  models have become available for predicting
the spread of air pollutants from point,  line,  or areal  sources.   Most of
these models were developed to  address problems  characterized by spatial
scales on the order of 100 km or less.  Only a  few modeling studies have
focused on the simulation of pollutant transport over long distances
(approximately 1000 km); these  are discussed below.

A.   SWEDISH STUDIES

     Following an early study by Reiquam  (1970), Rodhe (1971, 1972) appears
to have been the first to suggest a model  that  considers the variation of
surface deposition with travel  distance from elevated industrial  sources.
The model was used to compute the atmospheric sulfur budget for northern
Europe.  Rodhe found that the anthropogenic sulfur emissions in this area
outweigh natural emissions.  His results  also show that  the dispersipn of
sulfur has a continental character; i.e.,  sulfur is  transported,  on the
average, more than 1000 km before it is removed  at the surface.  His model
yields an estimated atmospheric residence time  for anthropogenic  sulfur of
two to four days.  On the basis of this study,  Rodhe dramatically concluded
that about half of the sulfur measured in Sweden originates from foreign
industrial emissions, and the other half  is caused by Swedish emissions
and a natural background.

     To clarify the relative roles  played by different physical processes
in determining the residence time of atmospheric pollutants, Bolin and
Granat (1973) and Bolin, Aspling, and Persson (1974) used a one-dimensional
model  describing the balance of vertical  diffusion,  sources, and  sinks.
Particular emphasis was given to assessing the  importance of rainout, washout^
*
 Washout, often referred to as  precipitation scavenging, designates the
 process whereby pollutants are collected by falling raindrops.  Rainout
 designates the process whereby pollutants are  first absorbed by  a cloud
 and then brought to the ground by rain.

-------
and dry deposition.  The results of the model calculations show that the
residence time is strongly dependent on the deposition velocity, surface
roughness, and turbulent intensity near the surface.   For low-level
emissions (~ 20 meters), the height of emission also has an important
effect on residence time,  but it becomes less important as the height
of emission increases.

B.   NORWEGIAN STUDIES

     In Norway, a modeling effort for the long-range transport of air
pollutants was undertaken by Nordb and his associates (Nordb, 1973;  Nordb,
Eliassen, and Saltbones, 1974) in connection with the OECD* project, "Long
Range Transport of Airborne Pollutants" (Ottar, 1973).  Nordb's model is
based on the two-dimensional, time-dependent, atmospheric diffusion  equa-
tion that includes sources, sinks, and chemical transformations.  Only
two pollutant species, SO- and HLSO   were considered in that study.  Both
surface depositions and chemical transformations were parameterized; the
former were characterized by linear decay, and the latter by both a  linear
and a quadratic term.  The distribution of pollutants in the vertical dir-
ection was assumed to be homogeneous between the surface and the inversion
layer, which was taken to be 2000 m in the model calculations.   The
observed winds on the 850 mb surface were used to estimate the horizontal
wind distributions in this layer.  The modeling region was divided into
two-dimensional cells, and the governing equations were cast into finite
difference form and were solved numerically.  Two grid systems were  tested
in this study:  A Cartesian coordinate and a polar coordinate consisting
of eight sectors.   Nordb found that numerical diffusion caused by the
truncation error of the finite difference scheme is very pronounced  in
the Cartesian approach, so he selected the sector approach for computing
the concentration  fields.
Organization for Economic Cooperation and Development.   In 1973, the
member countries included the United States, Canada, Australia, New
Zealand, and 19 Western European countries.

-------
                                     28
     The predictions of the sector approach were compared with those
obtained from the moment method developed by  Egan and Mahoney  (1972a,b).
The latter method was found to be more suitable for reducing numerical
diffusion.  In addition to the above numerical transport model, Nordo,
Eliassen, and Saltbones (1974) also developed a trajectory model for
analysis purposes.  In the trajectory model, pollutants  are uniformly
distributed in the vertical direction, but the thickness of the mixing
layer may change with position and time.   The trajectories were used to
follow the location of an air mass bounded by a triangle or a polygon in
the horizontal plane.  During the transport process the  deformation of
the polygon, which may shrink or expand,  was compensated for by vertical
displacement so that the mass continuity  requirement was satisfied.

     This trajectory model was further developed by Eliassen and
Saltbones (1975).  They used their model  to estimate the rate of decay
and transformation of SCL and SO, by comparing observed  and predicted
concentrations.  In the model  calculations,  48-hour isobaric trajectories
were computed from analyzed wind fields on the 850 mb surface.   The com-
puted trajectories arrived at the sampling sites four times a day. and
positions along a trajectory were given every half-hour.  The results of
this study show that the S09 decay rate due to dry deposition is on the
               -51
order of 2 x 10   sec  , corresponding to an atmospheric residence time
of approximately 12 hours.  The rate of SO^  transformation to sulfate was
found to be an order of magnitude smaller than the decay rate for dry
deposition.

C.   FINNISH STUDIES

     An alternative approach for modeling long-range pollutant transport
was taken by Nordlund (1973, 1975) of Finland.  His model, also of the
trajectory type, consists of an array of  air columns (or cells) that flow
into the emissions area.  A cell is allowed to shrink in a convergent flow

-------
                                    29
and to expand in a divergent flow.  At the same time, the height of the
cell also changes so that its volume remains unchanged.  The transport of
the cells was calculated using the advective scheme based on the moment
method (Egan and Mahoney, 1972a, b).  Although lateral diffusion was also
considered following the method of Smagorinsky (1963), it was noted that
the effect is only marginal.  However, the model predictions were found to
be most sensitive to the following four parameters:

     >  Emissions rate
     >  Height of the mixing layer
     >  Rate of pollutant removal
     >  Wind velocity.

Nordlund applied this model  to northwestern Europe for two different three-
day periods; the calculated concentrations agreed relatively well with
measurements.

D.   DANISH STUDIES

     In Denmark, Prahm and his colleagues (Prahm, Buch, and Torp, 1974;
Prahm, Torp, and Stern, 1976) have studied the problem of long-range
transport of atmospheric pollutants.  On the basis of sulfate measure-
ments and trajectory analysis, they showed that sulfur pollutants can be
transported more than 500 to 1000 km over the Atlantic.  The uncertain-
ties in the trajectory analysis, however, made it difficult to trace the
air masses.   Consequently, they examined various numerical  techniques
suitable for long-range air  quality modeling (Christensen and Prahm, 1976).
Nineteen different numerical  methods were examined including the Egan-
Mahoney method (1972a,b) and the pseudo-spectral  method (Fox and Orszag,
1973).   They concluded that  the pseudo-spectral method is the most accur-
ate solution procedure for Eulerian models.

-------
                                   30
E.    BRITISH STUDIES

     Smith (1970)  was credited with formulating a trajectory model  using
surface wind data  to compute the distribution of pollutants emitted from
Great Britain.   His work was followed by an extensive effort by Scriven
and Fisher (1975a,b), who developed a variety of models to address  ques-
tions related to the long-range transport of air pollutants.  Adopting a
trajectory approach, they first showed that the two-dimensional  steady-
state diffusion equation can be used to derive an integral equation for
the concentration  distribution as a function of the transverse and  ver-
tical distance  from the source.   They then developed a box model  that
accounts for pollutant removal  by washout and dry deposition in  a lin-
early expanding plume.  The effect of variable inversion height was also
considered in this simple approach (Scriven and Fisher, 1975a).   The
following general  conclusions were reached from an analysis of the  model
results:

     >  Decay distances of several  hundred kilometers are possible
        when rain  is absent and when the inversion height is on  the
        order of 1 km, assuming that the ratio of mean wind speed,
        u, to deposition velocity,  v , is 500 or more.   This is  in
        qualitative agreement with the results obtained by Scandi-
        navian  investigators (Rodhe, 1971, 1972; Nordo, 1973; Nordb,
        Eliassen,  and Saltbones,  1974; Eliassen and Saltbones, 1975).
     >  For a fixed velocity ratio, u/v , the travel  distance is
        proportional to the inversion height.   Thus low-level inver-
        sions cause short travel  distances unless the major part  of
        the emissions rises above the inversion.
     >  At a fixed distance from a large area source emitting at  a
        constant rate, there is  a maximum received concentration  in
        the absence of rain as other meteorological conditions vary.
        This maximum concentration corresponds to a maximum rate  of
        deposition that is independent of deposition velocity and
        falls off  inversely with the distance.

-------
                                  31
        Moderate  rainfall  (1  mm per hour)  reduces  travel  distances
        considerably.   Washout  dominates  deposition  while it is
        raining,  but not  on  a long-term basis  (e.g., annual
        average).
        In the absence of rain, sulfate aerosol  travels much
        greater distances than  S02 because the aerosols are  removed
        from the  atmosphere  principally by washout and rainout  rather
        than deposition.   Thus  their half-life is  much longer.
        Annual  average ambient  concentrations  and  deposition rates
        are orders  of magnitude smaller than  "in-plume" values.
        Typically,  large  industrial  areas  emitting S02 at rates  of
        hundreds  of tons  of  SC^ per hour  give  rise to dry deposition
        rates hundreds of kilometers away  that are at most a small
        fraction  of one gram of sulfur per square  meter per  year.
     A more sophisticated model was also developed by Scriven and Fisher
(1975b) to assess the accuracy of the simple box model  discussed above
and to investigate the buffering effect of diminishing  atmospheric  tur-
bulence as an emissions plume approaches the earth's surface.  The  model
is based on the time-dependent, one-dimensional  diffusion equation,
which follows a wind trajectory.  The solution was written in terms  of
Green's function.  The results of the model calculations  show that,  at
most distances of interest,  the predicted ground-level  concentrations
are lower than those computed from the simple box model.   Consequently,
the mean travel distance  (or average residence time) is  greater.  Fisher
(1975)  subsequently applied  this model to study the  deposition  of
sulfur over Great Britain, Sweden, and the rest of Europe.  His  con-
clusion, based on model calculations, was that only  approximately 6
percent of the total annual  deposition of sulfur over rural  Sweden  can
be attributed to  high-level  sources in the United Kingdom.

-------
                                   32
F.   STUDIES  IN THE UNITED STATES

     Dickerson, Crawford, and Crandall (1972) carried out a study in the
United States concerning the modeling of long-range transport of pollu-
tants.  This  study, motivated by two previous Russian works (Petrov, 1971;
Izrael, 1971), was concerned with the long-range transport, diffusion, and
deposition of radioactive substances from a Russian nuclear cratering
experiment.  An interesting feature of this study is the inclusion of a
method for computing wet deposition of tritium as a function of precipi-
tation rate,  storm cloud depth, and absolute humidity.  Computed plume
centerline concentrations, surface concentrations, and tritium deposition
were reported to be in good agreement with airborne and surface measure-
ments over Japan.  The need for further understanding of the transport,
diffusion, and deposition processes and for developing predictive capa-
bility over regional and extended scales was also discussed by Knox (1974).

     Recently, a model similar to Nordo's (1973) was developed by Miller,
Galloway, and Likens (1975) of the Air Resources Laboratories, National
Oceanic and Atmospheric Administration, for the study of common air pol-
lutants.  Heffter, Taylor, and Ferber (1975) also developed a regional-
scale transport model, based on the trajectory approach, that incorporates
both dry and wet deposition.   This model  is a part of a global model  for
computing long-term pollutant concentrations.  This model  was applied by
Lamb and Whitten (1975) to assess the impact of SCL emissions from
Illinois on the air quality of the northeastern United States.  More
recently, a box model, similar to that of Machta (1966), was developed
by Draxler and Elliott (1977)  of the Air Resources Laboratories.

     As part of an investigation to provide more data on atmospheric
pollutant loadings over the Upper Great Lakes, McMahon, Denison,  and
Fleming (1976) developed a long-range air pollution model  operating  on
a  daily time scale.   This simple model  was adopted from a  circular box
approach proposed by Slade (1967), but modified to operate on  a daily

-------
                                   33
basis and to account for wet and dry deposition.  Their analysis showed
that model predictions were very sensitive to the deposition velocity and
the washout coefficient.  They also concluded that background levels
resulting from natural sources can be significant in the overall balance
of the pollutant budget.

     To assess the transport and deposition of sulfur dioxide over the
continental United States, Fox (1975) adopted a trajectory model similar
to that of Scriven and Fisher (1975a).  He reported gross estimates for
the SOp concentration levels of the ambient air in the United States that
he deemed to be reasonable.

     Under the sponsorship of Federal Republic of Germany, Johnson, Wolf,
and Mancuso (1975) of the Stanford Research Institute demonstrated the
feasibility of developing an air quality budget model for central Europe.
The model tracks many "puffs" of S02, which are released at 12-hour
intervals from each grid cell containing areal sources.   These puffs
are transported according to the 850 mb wind field and are tracked every
three hours.  SCL emissions were assumed to be uniformly mixed in the
vertical direction and a simple Fickian diffusion, with a diffusivity
increasing linearly with time, was invoked for the lateral direction.
Moreover, exponential decay relationships were used to account for both
dry and wet deposition.   The authors stated that the results from pre-
liminary model calculations provided rough but reasonable estimates for
sulfur dioxide fluxes across international boundaries and amounts of
sulfur dioxide removed by deposition processes within individual
countries.

     More recently, several  field measurement programs were initiated to
examine the long-range transport of air pollutants.   A few of the more
well-known ones are

     >  MISTT--Midwest Interstate Sulfur Transformation  and
        Transport Project (White et al., 1976).

-------
                                     34
     >  SURE--Sulfate Regional Experiment (Hidy, long, and
        Mueller, 1976).
     >  MAP3S--Multistate Atmospheric Power Production Pollution
        Study (MacCracken, 1976).
     >  Northeast Oxidant Transport Study (Bufalini and Lonneman,
        1977).

 In conjunction with these studies, many regional airshed models were also
 proposed.  For example, as part of the Sulfate Regional Experiment a three-
 dimensional grid model was developed (Rao, Thomson, and Egan, 1976).  The
 Egan-Mahoney moment method (Egan and Hahoney, 1972a,b) was adopted for
 solving the atmospheric diffusion equations for sulfur dioxide and sulfate.
 The effect of surface deposition was parameterized in terms of a simple
 boundary condition at the ground surface, and chemical transformations
 between S0? and sulfate were grossly represented by a first-order reaction.
 This model was applied to an air pollution episode (October 3, 1974) over
 northeastern United States.   The results appear to compare favorably with
 the measurements collected at the AIRMAP Network of Environmental Research
 and Technology, Inc.  In another study, Rao, Lague, and Egan (1976) devel-
 oped a one-dimensional Lagrangian model.  The pollutant mass in each box
 was assumed to be well-mixed in the vertical direction.  From their sensi-
 tivity analyses they concluded that more accurate estimates of chemical
 reaction rates relative to surface removal rates are clearly important.

     Another regional  airshed model (Wendell, Powell, and Drake, 1976) is
 being developed by the Battelle Pacific Northwest Laboratories for the
 MU Histate Atmospheric Power Production Pollution Study (MAP3S) (HacCracken,
 1976).   This model  is  based  on a trajectory approach, and utilizes a scheme
 proposed by Wendell  (1972).   A power law is used to prescribe the horizontal
diffusion as a function of distance from the source.  Pollutant removal by
dry deposition,  precipitation scavenging, and chemical reactions is included
via simple linear relationships.  As part of a continuing program, the
effect  of the precipitation  pattern on pollutant removal  and the effect of

-------
                                   35
wind shear on the regional air pollution distribution are also being
examined.

     The cursory review presented above is intended only as an overview
of previous work in the modeling of pollutant transport over long dis-
tances.  In the next chapter, we delineate what we view to be the major
attributes of long-range transport models.

-------
                                    36
                     V    MAJOR  ATTRIBUTES  OF
                 LONG-RANGE  DISPERSION  MODELING
     A variety of long-range  air  quality models were discussed in
Chapter IV.   These  models differ in  data  requirements and model objec-
tives, and use various modeling approaches or formulations.  Moreover,
they place different degrees  of emphasis on the treatment of the many
physical processes pertinent  to the long-range transport of air pollu-
tants.  It thus seems important at this juncture to delineate the major
attributes of long-range dispersion modeling.

     Some atmospheric processes play  an important role in the dispersion
of air pollutants on large spatial scales, and others are important on
small spatial scales. The interactions among these processes, and the
overlapping influences of them on the eventual pollutant distributions,
are very complex (Fortak, 1974).   A classical example, shown in Figure 2,
is the effect of atmospheric  turbulence of different scales on pollutant
transport and dispersion.  The following sections discuss physical and
chemical phenomena that are unique to long-range air pollution modeling.

A.   TRANSPORT AND DIFFUSION

     The spatial  and temporal scales  of interest to the present study
are on the order of  several hundred kilometers and several days.  As
shown in Figure 2, the atmospheric motions important on these scales
range from mesoscale convection to synoptic-scale cyclonic waves.

     Changes  of wind speed and direction in the lowest layer of the
atmosphere are the result of  many competing physical processes.  The

-------
                            37
io6
1 week —
1 day —
IO4
1 hr. —
IO3
IO2
1 min. 	
10
-
i
Atmospheric
Turbulence
L, J
i t
1
Planetary
Waves
(Extra-
Domain of Interest 	 | ironical
to Regional Cyclones,
Air Quality Studies^^ . Anti-
(Hurri-
canes) '
i Cumulonimbus- con-^T" • s" s*
'vection (Land-Sc* ^ 	 	 _, ^^ -^
jBreezes and Moun- /- <, ^ ^
'tain-Valley Winds)] /^a4^
Cumulus' s*-^ ^
Convec- j ^^f^^
tion 	 ^ ^s
1 ^^ ^
/ s*
^* *^ Circumference
s ^ of Earth
, ,1
m 10m 100m 1km 10km 100km 1000km 10,000km 100,000km
I
Microscale
Meteorological
Study
Conventional Urban
Airshed Study
1 H
' Regional
Airshed Study
                 Characteristic Lateral Length Scale
FIGURE  2.   SCHEMATIC ILLUSTRATION OF SCALES OF  MOTION
            IN THE ATMOSPHERE

-------
                                    38
 interaction between  the  synoptic-scale air motion and the surface
 boundary layer usually produces  complex flow patterns.   These patterns
 change diurnally and seasonally.   They also vary spatially if nonuniform
 terrain or inhomogeneous  heating is  present.   The terrain of the North-
 ern Great Plains,  with the  exception of the Black Hills  in western  South
 Dakota, can be characterized as  flat.   This condition simplifies air
 quality modeling because  it eliminates  complicated flow  patterns such
 as valley winds and  drainage flows.   More  interesting to the long-range
 transport models,  as  pointed out by  Pasquill  (1974),  is  the fact that
 the prevailing wind  flow on this  scale  will  have a characteristic
 frequency which coincides with or  is  larger than that of the "spectral
 gap" in the longitudinal  velocity  spectrum of the atmosphere (Van der
 Hoven,  1957).   Minor  topographic  features  can sometimes  lead to  high
 surface concentrations under special  flow  situations.  For example,  accord-
 ing  to  a  tracer study carried  out by Heimbach, Super and  McPartlana
 (1975), the highest SCL concentration in the vicinity of  the Colstrip
 Power Plant near Billings, Montana,  is  observed  at a  hill  about  350
 meters  higher  than the plant and 20  kilometers downwind.   Obviously.
 this result is  due to the impingement of the  plant's  emissions plume
 upon the  hill.  Clearly,  both mesoscale and microscale flow patterns
 are  important  in determining ground-level  concentrations  of pollutants.

      Aside  from the dominant atmospheric motions,  divergence  in  the
 synoptic  and mesoscale horizontal wind  regimes leads  to  vertical  air
 motions.  Vertical  currents, which give rise  to  the phenomenon known
 as Ekman  pumping, are also generated by viscous  forces in  the boundary
 layer and can be particularly  large in regions of  complex  terrain.
Although  the vertical velocities  generated by these processes have a
magnitude of only 1 to 10 cm/sec, they can have  significant effects on
the net transport of air  pollutants (Liu and Seinfeld, 1975).  Accurate
estimates of the vertical  components of the wind vectors on this scale
are extremely difficult to obtain.  Thus,  in all of the long-range dis-
persion  models discussed  above,  horizontal  wind  fields were prescribed

-------
                                   39
                                     *
from upper air pressure distributions ,  and no vertical  velocity com-
ponents were specified.

     Over the Northern Great Plains,  wind fields  are  often  strongly  influ-
enced by the high pressure system west of the Rocky Mountains.  The
magnitude of the prevailing westerlies is governed by the location and
strength of the Pacific High.  Holzworth (1972) calculated  the average
wind speed and the mixing-layer depth within the mixing  layer from a
five-year record of upper air observations at National Weather Service
stations.  The portion of that data pertinent to this study is repro-
duced in Table 8.  We note that in the Northern Great Plains wind speeds
and mixing depths are generally lower in winter than  in  other seasons,
and thus it is expected that the greatest potential  for  air pollution
episodes in this region should occur in  winter.

     Relative to horizontal transport by wind, vertical  diffusion plays
a completely different role than lateral diffusion in determining the
fate of air pollutants at large distances.   This  can  be  seen from a
simple analysis.  According to Table 9,  the following two ratios can
be formed:
                                                K
                         Lateral Diffusion       H
                       Horizontal  Transport     UAX

                                              K      2
                      Vertical  Diffusion   _   v /Ax\
                     Horizontal  Transport     UAX\Az /

where
           U = characteristic wind speed,
          Ax = characteristic length in the lateral direction,
          Az = characteristic length in the vertical  direction,
*
 These are typically derived from the 850 mb pressure surfaces.

-------
            TABLE 8.   MEAN SEASONAL  AND ANNUAL MORNING AND  AFTERNOON MIXING HEIGHTS
                        AND WIND SPEEDS  FOR THE  NORTHERN GREAT PLAINS*
Winter

S-.aVcn
La-.:*',
Wy:-'-;
Glas"-,
Kcr-.a-a
Grea: ri;is.
Hcr-.2-a
B1S-2r;<,
N:rt- lakota
Ra-'- "•'•/
$;„•." :a'
-------
                                    41
         TABLE 9.  COMPARISON OF PHYSICAL PROCESSES PERTINENT
                   TO LONG-RANGE POLLUTANT TRANSPORT
                              Mathematical    Characteristic
          Physical Process   Representation       Value	
            Horizontal            u —            U —
            Transport               3x              Ax
            Lateral            _3_/K  3_c\        K    AC
            Diffusion          3x \ H 3x/         H     n
            Vertical           _3_L  3c\        .,    Ac
            Diffusion          3z \ V 3zl         v
          Ku = horizontal eddy diffusivity,
           H
          K  = vertical eddy diffusivity.
Eddy diffusivity in the horizontal  direction is known to vary not only
with lateral scale and altitude, but also with latitude (Czeplak and
Junge, 1974).  Furthermore, the zonal and meridional components of the
large-scale eddy diffusivity can be shown to be different in magnitude
(Kao, 1974).  According to Heffter  (1965) and Randerson (1972), a value
     A  O
of 10  m /s appears to be the median horizontal diffusivity for the
spatial and temporal scale of interest.   Vertical eddy diffusivity is
a strong function of height and atmospheric stability.  For the present
                       2  2
analysis, a value of 10  m /s can be viewed as representative (Pasquill,
1974).  Thus, using a 10 m/s average wind, and Ax - 100 km, Az = 100 m,
the above two ratios become

                         Lateral Diffusion      ,n-2
                                             ~ I u     ,
                       Horizontal  Transport

                        Vertical  Diffusion   ^  ,Q2
                       Horizontal  Transport  "~

The implication of this analysis  is  that while  vertical  diffusion  is
overwhelmingly dominant,  lateral  (or horizontal) diffusion  is  also mar-
ginally important in pollutant transport over  large  distances.

-------
B.   REMOVAL PROCESSES


     Over a  travel  distance  of,  say,  1000 km, more than half of the

total mass of most  pollutants  is removed by various removal  processes.

For sulfur dioxide, the rough  estimates in Table 10 provide  a ranking

of the importance of each  removal  process.


     It is thus clear that the following three processes should be

included in  models  of long-range S02  transport:
     >  Dry deposition

     >  Rainout and washout
     >  Photochemical  reactions  (if significant NO  and HC are present)
                                                  A


The first two processes  are discussed below and the third in the next

section.
                    TABLE  10.   S02 REMOVAL PROCESSES
              Process
       Photochemical  Reaction
       (S02/Clean Air)

       Fog

       Photochemical  Reaction
       (S02/NOx/HC)

       Dry Deposition

       Rainout and washout
Rate of Removal of SO,
  (Percent per Hour) i

         0.03


           2

         1-10


         1-10

          12

-------
                                    43
1.    Dry Deposition

     The most extensive outdoor areas available for the deposition of
S0£ are the oceans, vegetation, and soil.  In towns and cities, build-
ing materials must also be added to this list.  In his study of the
atmospheric sulfur cycle, Junge (1963) estimated that the direct up-
take of SO- and hydrogen sulfide (HoS) by soil and plants is 7x10
tons per year, with a similar amount being absorbed by the sea.  Junge
compared this with an industrial release of 4 x 10  tons of SO,, per
year and a biological release of H9S from the soil, sea, and coast of
       7                          
-------
                                   44
            TABLE 11.  DEPOSITION  OF  S02  ONTO  VEGETATION
          Velocity  of
Plant     Deposition
Alfalfa
Mustard
Barley
Several
2.5
0.7
1.5
2.0
                                 Method Used
                              Rate of removal of   Hill (1971)
                              SOp by leaves
                              Analysis of S in     Spedding (1969)
                              leaves
                              Analysis of S in     Spedding (1969)
                              leaves
                                                   Eriksson (1966)
plant roots have an adequate supply of water and when the relative
humidity of the atmosphere is high.  Under conditions that wilt leaves
the stomata are closed.   A further factor influencing the opening or
closing of stomata is the concentration of atmospheric SOp.  At SOp
concentrations greater than about 0.4 ppm, the closing of the stomata
is increased (Katz, 1949; Mansfield and Heath, 1963).  Field observa-
tions of this effect were reported by Martin  and  Barber  (1971).
2.   Met Deposition

     Rainout and washout have  long  been  considered  to  be major  sinks
for atmospheric S02-   It has been speculated  that these physical mecha-
nisms are responsible for the  occurrence of  "acid rain."  The efficiency
of rainout and washout in removing  SOp from  the  atmosphere  generally
depends on three factors:

-------
                                     45
      >  The amount of clouds.
      >  The efficiency of the consumption mechanisms of clouds
         and raindrops.
      >  The frequency of rains.

      The absorption of gases by cloud droplets, known as rainout,
depends on the chemical composition of the droplets.  Much further work
is needed to provide a quantitative understanding of this process, but
there is evidence of a rather rapid transformation of sulfur dioxide
into sulfuric acid in clouds as long as the pH of the cloud droplets is
significantly greater than 4 (Brosset, 1973).  But clearly the most impor-
tant factor in the overall efficiency of rainout in removing SO,, from the
atmosphere is the frequency of rains (Rodhe and Grandell, 1973).  On the
basis of rain statistics from Stockholm, Rodhe and Grandell showed that
even with very effective transfer of SO^ into cloud droplets—and ulti-
mately into rain drops—the average residence time for SOp in the atmos-
phere would be about 40 hours in winter and 90 hours in summer if rainout
were the only removal mechanism.  These values are approximate, of course,
and would certainly be different in another climatic region.   Rodhe and
Grandell (1973) derived the distribution function for the probability of
rainout of a pollutant released at an arbitrary instant (see also Bolin and
Rodhe, 1973).   Although a more precise characterization of rainout might
well be important, it has been generally assumed that removal  of pollutants
by precipitation can be described adequately by a characteristic mean
residence time, and that the amount of pollutant removed by rainout at any
one place is proportional  to the concentration of that pollutant in the air.

      The capture of gases and particles by falling raindrops is called
washout.   Typically,  the duration of washout is relatively short compared with
that of rainout.   However, pollutant concentrations at the cloud level  are
generally much  lower than  those near the ground in the presence of an emis-
sions plume.  Thus,  rainout and washout can be of similar importance in
the acidification of rain.   The uptake  of S02 by rain depends on physical

-------
                                    46
parameters, such as rainfall  intensities and raindrop size distributions,
and on chemical characteristics, such as the presence of oxidizing agents
in the atmosphere and the chemical  composition of the raindrops.   Other
factors may also be influential.  For example, Li and Landsberg (1975)
found that the extent of acidic washout from a plume has a notable depen-
dence on wind speed.   Models  of washout generally reduce asymptotically
to two limiting cases.   These cases are mass-transfer-limited (i.e.,
irreversible washout) and chemical-reaction-limited (i.e., equilibrium
washout).  A recent study by  Dana,  Hales,  and Wolf (1972) suggests that
under typical atmospheric conditions washout is often mass-transfer-
limited.

C.   CHEMICAL TRANSFORMATION

     Most pollutants undergo  a variety of chemical changes in the
atmosphere.  The chemical reaction  of most interest to this study is
the oxidation of sulfur dioxide to  sulfate.   Sulfate is found in
particulate matter primarily as sulfuric acid  (H^SO,), ammonium bisul-
fate  (NH.HSO.), and ammonium sulfate [(NH.)? SO.].  Atmospheric sulfur
dioxide  (SO^)  is both reactive and soluble.  It can thus participate
in many  homogeneous and  heterogeneous chemical reactions, and many
mechanisms have been proposed for its oxidation to sulfate.  Although
these complex  reactions  are not currently well understood, it is  gen-
erally  thought that near a source sulfur dioxide inhibits the production
of ozone and the formation of photochemical smog.  Downwind of a  source,
on the other hand, photochemically initiated free-radical interactions
of sulfur dioxide and nitrogen oxides are thought to produce secondary
pollutants such as ozone and sulfuric acid.

     Basically, SO^ can  be converted to other pollutants in two ways:

     >  Gas phase reactions lead to the formation of sulfur trioxide
         (SO.,), which rapidly combines with water to give sulfuric
        acid (H2SO.).  The H2SO. molecules formed in the qas phase can

-------
                                  47
        then dissolve in existing droplets or serve as nuclei  for
        clusters of water molecules.   In the presence of hydrocarbons
        and nitrogen oxides, SCL can  be oxidized in the atmosphere
        at appreciable rates (on the  order of 5 percent per hour),
        forming SO,.  Reactions of SCL with oxygen-containing  free
        radicals, principally OH, and with oxidized products of ozone-
        olefin reactions generally account for most of the gas phase
        conversion of SC^ to SO.,.
     >  Sulfur dioxide dissolves in aerosol droplets where it  is sub-
        sequently oxidized to sulfate (SO,).  The oxidation requires
        a catalyst.  Two types of catalysts have been identified and
        studied—dissolved NH_ and metal salts.  The catalytic oxida-
        tion of S0? in solution is known to be promoted by ammonium
        ions (NHt) and by metal ions, such as Fe+3 and Mn+^.  NH^ is
        essential to the oxidation of S02 in solution because  it
        buffers the solution, permitting effective absorption  of S0?
        from the gas phase.  The absorbed S0« then forms sulfurous
        acid and sulfite ions.  The solution chemistry of this system
        seems to be reasonably well understood (Scott and Hobbs, 1967;
        Miller and de Pena, 1972).

     The above discussion is largely  qualitative.  Quantitatively the
chemistry of SO,,, particularly in complex systems, is not well under-
stood, although some advances have been made recently.  For example,
Liu et al. (1976) developed a kinetic mechanism for the chemistry of
the hydrocarbon-nitrogen oxides-SO- system.  This kinetic mechanism,
based in part on data from smog chamber experiments, has been  used
in studying the chemical reactions occurring in power plant plumes.

D.   SUBGRID-SCALE PROBLEMS

     On the subgrid scale, modeling of large point sources at  long
distances presents certain unique problems.  Compared to emissions of
oollutants from areal sources (generally related to transportational or

-------
                                   48
residential use of fossil fuels), emissions from point sources such
as power plants, refineries, and other industrial facilities possess
several distinct physical characteristics.  The most obvious ones can
be stated as follows:

     >  The emissions from point sources are generally more
        concentrated.
     >  The emissions from point sources are almost invariably
        released at greater heights.
     >  The emissions from point sources are most often buoyant.

These characteristics distinguish the point source air pollution prob-
lem from that associated with areal  emissions.   Perhaps the most
prominent difference between point and areal source models is the ques-
tion of spatial resolution.  Due to the disparity in spatial scales
appropriate to each, conventional grid models—even the most sophisti-
cated ones—have difficulty in properly treating the transport and dis-
persion of point source emissions in the immediate vicinity of the
stack.  This is probably the reason why the Gaussian formula has been
used so extensively for point sources in the past, despite its many
known deficiencies.  Because the emissions from a point source are
buoyant and are released into the atmosphere at great heights, an accu-
rate prediction of the impact—in particular, the impact on ground-
level concentrations—will require knowledge of not only the height to
which the plume will eventually rise (the effective plume height), but
also the effects of plume interaction with the ground surface, particu-
larly if the terrain is not flat.

     The special characteristics of point sources pose a variety of
problems in modeling.  Foremost of these, perhaps, is the problem of
predicting plume rise.   Although there is no lack of plume rise for-
mulas (Liu et al., 1976), they are generally empirically based.

-------
                                    49
Because of the different terrain, meteorological, and emission con-
ditions under which data were collected to derive these formulas, it
is not uncommon for the predictions of plume rise formulas to vary by
more than a factor of two.

     To compound the problem of estimating plume rise, plume behavior
is critically affected by the vertical structure of the atmosphere.
In ttie case of a surface layer capped by an elevated temperature inver-
sion, which is generally associated with the worst air pollution epi-
sodes, a number of possible plume configurations may take place.  A
buoyant plume can penetrate an elevated inversion if the plume is
"strong" and the inversion is "weak", but the plume can be entirely
tranned underneath the temperature inversion if the opposite is true.
During transient conditions, such as those associated with the daily
heating of the surface layer or the development of a diurnally varying
land-sea breeze along coastal areas, the gradual entrainment of a
plume into the surface layer gives rise to plume fumigation, which
typically produces the greatest ground-level comcentrations.  All of
the phenomena described above are intimately connected with the pre-
diction of plume rise.

     Other problems related to the effective plume height can be
equally important.  One of these is concerned with wind shear   Ideally,
to minimize the error in model predictions, one should use the measured
wind speed at the height of the pollutant cloud.  This does not pose
a major problem in the modeling of ground-based areal sources because
surface wind data can generally be considered as representative and
are readily available.  In the case of a buoyant plume, however, the
effective plume height is not always known a priori.  Furthermore, to
measure the wind speed at that height is not a trivial matter.  The
current practice is to use the measured wind speed at the stack height.
Any attempt to correct this deficiency clearly requires knowledge of
the vertical profile of the horizontal wind.  Many Gaussian models

-------
                                   50
have achieved this by simply adopting a power-law wind profile for
the conversion of the measured wind speeds at the stack height to those
at the effective plume height.

     The importance of other, more complex aspects of the plume-wind
shear interaction should not be disguised by the  simple discussion pre-
sented above.  For example,  under the conditions  of a local  surface
wind—such as the drainage flow or sea breeze—imbedded in a synoptic-
scale flow of the opposite direction, a drastic change in wind pro-
files may be responsible for the occurrence of such anomalies as  bifur-
cation of the plume (Liu et  al., 1976).
     Also related to the elevated nature of point sources is the prob-
 lem of the impact of the plume on the topography.  Depending upon the
 relative heights of the plume and the ground surface and the vertical
 structure of the atmosphere, it is conceivable that the plume can
 either be lifted above or impinge upon the surface.  The occurrence
 of either should depend in general on whether the kinetic energy of
 the air stream approaching an obstacle is greater or smaller than the
 potential energy required to lift it over the obstacle, which is in
 turn dependent upon atmospheric stabilities.  Thus, the conditions
 that are conducive to plume impingement are light winds and stable
 atmosphere.  However, the physical processes governing the occurrence
 of impingement phenomena are extremely complex, and have only very
 recently received the attention of air pollution researchers.

     For reactive pollutants, certain features are also unique to the
 point source problem.  Because the emissions from power plants are
 rich in nitric oxide, ozone entrained from the ambient air is generally
 completely depleted within the plume in the vicinity of the stack.
 This phenomenon has been frequently observed and is well documented.
 At large downwind distances, depending upon the ambient hydrocarbon

-------
                                   51
and nitrogen oxide levels, secondary pollutants can be formed in some
situations (Liu et al., 1976).  Thus, in the modeling of reactive pol-
lutants, it is important to assess the interactions of the plume with
urban or rural background emissions.

-------
                                   52
       VI   DEVELOPMENT OF  A REGIONAL AIR  POLLUTION  MODEL


     It had been originally conceived that in  this  study a  suitable
regional air pollution model was to be selected  and adapted  for applica-
tion to the Northern Great Plains.   The selection of a  model  must  of
course be based on its ability to include the  attributes discussed in  the
previous chapter, so that the  effects of various sources on  air quality
can be predicted with reasonable accuracy.   For  the present  study, the
following model attributes appear to be particularly pertinent:

     >  The ability to handle  a multitude of emission sources.
     >  An adequate treatment  of pollutant transport over
        large distances.
     >  An adequate treatment  of pollutant depletion processes.
     >  Provisions for including chemically reactive pollutants.

Other important considerations include computational  requirements  and
availability and resolution  of the  data base.

     As discussed  in Chapter  IV, a  variety of regional  models  have been
developed recently and are available for estimating concentrations of  air
pollutants at large distances  from  the sources.  These models generally
fall into the following four categories:

     >  Box models (e.g.,  Johnson,  Wolf,  and Mancuso, 1975)
     >  One-dimensional  models (e.g.,  Bolin, Aspling, and Persson, 1974)
     >  Gaussian models (e.g., Scriven and Fisher,  1975b)
     >  Numerical  models  (e.g.,  Rao,  Thomson,  and Egan,  1976).

A careful  examination of  all these  models  revealed  that  the model  devel-
oped by Rao,  Thomson, and  Egan (1976)  appeared to be closest to  satisfying

-------
                                   53
the model attributes listed above.  This model, however, suffers from the
following two deficiencies:

     >  It does not contain a sufficiently detailed algorithm for
        the prescription of surface deposition.  For the present
        application, which deals primarily with elevated emission
        sources, the diurnal variations in deposition rates are
        expected to be quite important.
     >  This model has unfortunately retained the vertical  dimen-
        sion.  As discussed in the previous chapter, the inclusion
        of this dimension is unnecessary and obviously imposes a
        severe computational burden.

     In viev/ of these deficiencies, it was decided during the course of
this study that a new regional air pollution model be developed.  As
shown in Figure 3, this model is composed of two interconnected submodels:

     >  A mixing layer model
     >  A surface layer model.

     The mixing layer model is designed to treat transport  and diffusion
above the surface.  A grid approach is adopted in this project in order
to facilitate the handling of multiple sources and complex  chemistry.
The major feature of this model  is the assumption that pollutant distri-
bution is nearly uniform in the vertical  direction.   With this assumption,
a simplified form of the general  atmospheric diffusion equation can be
invoked.

     The surface layer model is designed  to calculate the pollutant fluxes
lost to the ground.   The surface layer, a shallow layer immediately above
the terrain, is embedded within the mixing layer.   For pollutants origi-
nating from either elevated sources or distant ground-level  sources, most
of the pollutant mass is contained in a layer aloft, i.e.,  in the mixing
layer.   The removal  processes consist of  the diffusion of the pollutants

-------
SURFACE LAYER fh
                                                                     TOP OF THE
                                                                     MIXING LAYER
                                                                             H  V MIXING
                                                                                 LAYER
                                                                                                          01
                                                   GROUND  SURFACE
           FIGURE 3.   SCHEMATIC ILLUSTRATION OF THE MODELING REGION IN THE
                      REGIONAL AIR POLLUTION MODEL  DEVELOPED IN THIS STUDY

-------
                                   55
through the surface layer to the ground,  followed by absorption or adsorp-
tion at the atmosphere-ground interface.   A unique feature of the surface
layer is its diurnal variation  in  surface temperature,  which is a result
of daytime heating and nighttime cooling.   This  variation affects the
vertical pollutant distribution through  atmospheric stabilities, and
consequently, affects the rate  of  surface uptake of pollutants.

     These submodels are discussed  separately in  the following sections.
It should be emphasized, however,  that because of the limited scope of
this study, we attempted only to develop  the basic and most desirable
elements of an ideal  regional air  quality model.   A number of important
issues were not addressed,  including:

     >  Predictions from the regional  air quality model  in its
        present form are unlikely  to be applicable within, say,
        a few kilometers downwind  of a major emission source.
        Thus, subgrid-scale  concentration distributions,  as
        discussed in Chapter V must be dealt with on a differ-
        ent level.   Models of this  type have been discussed in
        a recent report  by Liu et  al.  (1976).
     >  The only pollutant removal  process treated is dry
        deposition  on the surface.   Other important  removal
        processes such as rainout  and  washout are not consid-
        ered.  Unless these  processes  are included,  the  present
        model is, strictly speaking,  applicable only during
        periods of  no precipitation.
     >  The treatment of chemical  reactions is limited to a
        first-order overall  reaction  between SCL  and sulfate.
        Although no constraint except  computational  time  imposes
        any problem,  the inclusion  of  complex chemistry  awaits
        the development  of a kinetic model  capable of simulating
        chemical  transformations during nighttime and the
        effects of  natural emissions of hydrocarbons.

-------
                                  56
A.   THE MIXING LAYER MODEL

     The mixing layer model is designed  to treat  the transport  and  diffu-
sion of air pollutants over long distances.   The  model  formulation  is
discussed  in Section  1.  As  stated  earlier,  the  grid approach  was
adopted in the  present  study.   There  are  a  number  of significant advan-
tages to the grid approach—it  is very  versatile,  and it can easily
handle time- and space-varying  emissions  and meteorological  variables,
complex chemistry, and  surface  sinks.   But  there is one major  disadvan-
tage associated with  this  approach; pseudo-diffusion associated with the
numerical  solution of the  governing equation can be overwhelming.   An
accurate scheme must  thus  be found  for  the  simulation of the advection
term.  The selection  of an appropriate  numerical method is  discussed in
Section 2.

1.   The Model  Equations
                                                                    *
     Uithin the framework of the  so-called  gradient-transport theory,
the concentration distributions of N  reactive  species can be described
by the atmospheric diffusion equation of  the following  form (Monin and
Yaglom, 1971):
   SC^     8C.      3C.      3C.      /  3C.\      /  3C.
   —-— + N 	 + V	 + W 	 = -—CS<	—I + —— I V	-
   3t      3X      3y      32    3X\A3X  /   3y\y3y
                                  3 '-  3C-'
                                  3Z
                               +  S^c.)     i = 1, 2,  .... N    ,   (1)
*
  The gradient-transport  theory,  analogous to molecular diffusion theory,
  states that a  pollutant flux  in the direction of decreasing concentra-
  tion is established  as  a result of turbulent fluctuations.  The magni-
  tude of this flux  is assumed  to be proportional to the gradient of the
  average concentration.   The limitations of models based on the gradient-
  transport theory,  also  known  as K-theory, were examined by Corrsin (1974).

-------
                                    57
where c^ denotes concentration for pollutant species i, u, v, w, and K ,
K , KZ represent wind speeds and turbulent eddy diffusivities in the x,
y, and z directions, respectively, and R and S are the chemical  reaction
and source (and/or sink) terms.

     One of the major simplifications in the present model is the assump-
tion of vertical homogeneity in the concentration distribution.   One of
the reasons for this choice is that the vertical  diffusion term, based on
the dimensional analysis shown above, is about 100 times greater than the
transport term, and the horizontal diffusion term is only a fraction of  the
transport term.  Thus retaining the vertical variation terms  in  the dif-
fusion equation will compound difficulties in the numerical solution of
the governing equation, without necessarily improving the accuracy of the
model's predictions.   As  shown in Figure 4, measurements of the vertical
distributions  of  sulfur compounds over central Germany (Georgii, 1970)
show  that  in  these  remote areas  the profiles are fairly uniform beneath
the temperature inversion.   Similar observations were also reported by
Rodhe  (1971)  in southern  Sweden.   Thus it does not seem necessary to
include the vertical  dimension in the model.
     Assuming that the concentration distribution  in  the  vertical  is  nearly
uniform below the base of the temperature inversion,  a  vertically  averaged
concentration can be defined as
                                 rH
                           i  = FT/   ci
                                     c..dz
                                 '0
where H is the height of the inversion  base.   Performing  the  same opera-
tion on Eq.  (1) and imposing the appropriate  vertical  boundary conditions,
one obtains

-------
   2800

   2400
o  2000
§!  1600
o

"^  1200

|!   800

    400
                        S02


                       \\\ \ \ TEMPERATURE
                                  INVERSION
 10         20
      ug/Nm3

(a)  25 February 1967
                                         30
                                             2800

                                             2400

                                           § 2000
                                           o

                                             1600
                                                           O
Ol
o
<  1200
1/1

S   800

-------
                           59
                           ,  /  3cH\
                        + MV)+ M'r'z--^ + s^
 3C-    _3C.   _3C.     /  3C.
 3t     3X     3y    3X \ *3X

                             •»•  D-C(D)    i = 1,  2,..., N             (3)

where CQ.  is  the  background concentration of  species i, u and v are the
                                                   H
vertically averaged horizontal wind components  (u = fn udz/H,
-    H                                                  -
v = /Q vdz/H),  D  is the  two-dimensional  divergence [D   (3u/3x)
+ (3v/3y)], and c(D)  is  a step function  defined by

                              for    D  > 0    ,
                             i                                     (4)
                              for    DiCl

In the derivation of  Eq. (3),  the following assumptions were made:

     >  Deviations from the average concentration, c.,  in the
        vertical  direction are small.
     >  The vertical  velocity  at  the top  boundary  is approxi-
        mately  given  by
>  The diffusive flux of pollutants at the top  boundary
   is negligible.
>  The following relationships hold for the reaction and
   source/sink terms:
                                                               (6)

                                                               (7)

-------
                                   60
     One of the problems encountered in the present model formulation is
 the  disparity of scales in the treatment of emission sources.  Since the
 preponderance of sulfur dioxide emissions in the area of interest comes
 from isolated point sources, the spatial scales associated with these
 sources and the grid spacings adopted in the mixing layer model are cer-
 tainly not commensurate.  In order to resolve this subgrid-scale problem,
 a  special algorithm was developed in which the emissions are first
 treated as puffs.  These  puffs  are  emitted  from each major point source at
 regular time  intervals  and  tracked  downwind along their separate trajec-
 tories.  The  horizontal spread  of each  puff is calculated according to
 the  Gaussian  formula  (Turner,  1969).  When  the width of a puff reaches
 that of one grid cell,  the  emissions contained in that puff are released
 into that cell.  Table  12 lists  typical  downwind distances at which the
                 *
 width of the  puff  equals 10  km.  It is  apparent that,  particularly under
 stable conditions, the  puff can  travel  a  few grid cells before it is
 picked up by  the mixing layer model.

              TABLE 12.   DOWNWIND DISTANCE TRAVELED BY  A PUFF
                        AS  A FUNCTION OF ATMOSPHERIC STABILITY

                       Stability        Downwind  Distance
                       Category          Where 4a  = 10 kmt
                          A         '         13.3 km
                          B                   17.7 km
                          C                   25.8 km
                          D                   42.8 km
                          E                   59.0 km
                          F                   88.5 km
*
  The width has  been  chosen  to  be 4a, within which the puff contains more
  than 95 percent  of  the  pollutant mass.
f From Turner (1969);  a adjusted for a one-hour sampling time.

-------
                                     61
2.   The Numerical Method

     The solution of Eq. (3) with appropriate initial conditions and
boundary conditions surrounding the modeling region requires a numerical
method.  Since the transport of pollutants on this scale is dominated,
as demonstrated above, by horizontal advection,  the problem of numer-
ical diffusion arises in the discretization processes.  That is, the
numerical solution tends to smooth any sharp concentration profiles as
the pollutants are advected downwind, even when the horizontal diffusiv-
ity is zero.  We  investigated and compared the accuracy and computing time
requirements of three finite difference methods for solving simplified
forms of Eq. (3):

     >  The upstream difference method
     >  The SHASTA method
     >  The Egan-Mahoney method.

The upstream difference method is the simplest of the three.   It is also
well-known and widely used (Forsythe and Wasow,  1960).  The SHASTA
method (^harp And Smooth Transport Algorithm) was developed by Boris and
Book (1973).  The method proposed by Egan and Mahoney (1972a,b) has the
distinctive feature that the first and second moments of the  mass distri-
bution in each cell are also calculated.   The performance of each method
was examined using hypothetical  situations.  Based upon considerations of
both accuracy and computing speed, the SHASTA method appeared to be most
suitable to the needs of the present study and was thus selected for
treating the horizontal  advection terms.   (Details of the numerical
analysis and selection are given in Appendix A.)  In the following para-
graphs we present a brief description of the numerical method used in the
mixing layer model.

     Let the continuous  variables be represented on a grid with mesh
widths AX and Ay so that x-.  = x(iAx.jAy).   Define the operators

-------
                                  62
                                                                  (8)
                                    Q(2)
                                     *
                                                                  (9)
Then our numerical method is given  by  the following three fractional  steps
(Yanenko,  1971):

Ste[3 1 — x-direction
                      ^'U + q^'u +  (r - d) At]cn
          c -
         c".c"-oc     ,                              (10)

Step 2--y-direction
                                                              'c**
                    + (Q(2)v  + Q(2)y .
          c+ = c"" -g-D^'D^'cT                                (11)

-------
                                    63
Step 3—point sources

               = c+ + S    ,                                       (12;
                      KxAt        K At
          where a, =  —*— , an =  •*n
r is the chemical reaction rate, and d is the surface deposition rate.
The stability and accuracy of the scheme are analyzed in detail for the
constant velocity case in Appendix A.  The advection terms are treated
with at least second-order accuracy while the fractional ized scheme as
a whole is accurate to the second order in space and to the first order
in time.

     In order to estimate the accuracy of the numerical  method adopted,
in Table  13  we  give the effective psuedo-diffusivities produced  by the
model on the ten-kilometer grid with an optimum stepsize.   For the pre-
sent problem, the pseudo-diffusion generated appears to be small  when
compared with the physical diffusivity in the horizontal plane, which  is
estimated to be on the order of 10 m /sec (Randerson, 1972).  A more thor
ough analysis of the problem of psuedo-diffusion is presented in
Appendix A.

           TABLE 13.  PSEUDO-DIFFUSIVITY IN ADVECTIVE TRANSPORT
                      FOR A 10 KILOMETER GRID AND  vAt/Ax  =  1/2

                              Wave Number    Pseudo-Diffusivity
               Wave Type         (m~l )       _ (m^/sec) _
                                60TT/106          2.5 x 103

                                307T/106          1.6 x 102
                                                     40

-------
                                  64
     On the other hand, computational  stability is  guaranteed when
                             AX
                                                                    (13)


                                                                    (14)
With a ten-kilometer square grid  cell,  the most  restrictive stability con-
straint derives from the advection terms [Eq. (13)]  if  K  and  K  are  less
       5   ?
than 10  m /sec.  For higher horizontal diffusivities,  Eq. (14)  becomes
more stringent.  The time  step used in the mixing layer model  has  been
chosen in  such a way that  these conditions are  always satisfied.   Thus
accurate and  stable solutions were obtained for the mixing layer model.
B.   THE SURFACE LAYER MODEL

     Pollutants are removed from the atmosphere via both dry and wet
deposition.  Only dry deposition at the earth's surface was considered
because of the limited scope of this study.  The importance of surface
deposition on pollutant concentrations at large distances has been well
established (e.g., Bolin et al., 1973, 1974; Scriven and Fisher, 1975a,b).
Thus an indispensable element in the regional air pollution model is the
treatment of pollutant depletion processes near the surface.  In this
section, we describe the surface layer model, beginning with a discussion
of previous studies on surface deposition, followed by a description
of the approach adopted in this study.

1.    Dry Deposition on Surfaces

     In most studies, removal  of pollutants by the ground surface is
generally characterized by

                             F = Vdc    ,                         (15)

-------
                                  65
where F is the mass flux  to the  surface, c  is the concentration measured
at an unspecified reference height,  and V.,  having the units of velocity,
is commonly referred to as  the deposition velocity.  In this expression,
the deposition velocity is  viewed  as a proportionality constant whose
magnitude is established  empirically.  The  surface deposition is governed
by many complex physical  processes,  which depend primarily upon:

     >  The state of atmosphere  near the ground
     >  The types and configurations of the  surface.

For example, Bolin, Aspling, and  Persson (1974) noted that for a perfect
sink of a particular gas,  in which  all molecules of that gas reaching the
surface are absorbed,  the  ground-level concentration is zero and the
deposition velocity  is theoretically infinite.  In this case the flux is
diffusion-limited.  Consequently, the simple concept of the deposition
velocity  is generalized.

     In analogy with electrical  circuits, surface deposition was treated
in terms of resistance to mass transfer (Owen and  Thompson,  1963;
Chamberlain, 1966).  The  transfer  of gases from the atmosphere to a  sur-
face is described by three  resistances in parallel:

     >  The resistance to momentum transfer, r .
     >  The excess resistance to mass or heat transfer, r, .
     >  The resistance at the ground surface, r .

The total  resistance,  R,  which is  defined as the reciprocal  of the deposi-
tion velocity, is then given by


                       V'SK***'T   '

Within the framework of the surface  boundary layer (Owen and Thompson,  1963)

-------
                                  66
where u(z) is the vertical wind profile and u* is the friction velocity.
The deviation between momentum and mass/heat transfer is characterized by
where 0 is dependent on the surface roughness, a Reynolds number appropri-
ate to the flow in the roughness layer, and the ratio of the kinematic
viscosity of air to the molecular diffusion coefficient of the pollutant
gas.  This correction is necessary because the process of mass transfer
is generally less efficient than that of momentum transfer, resulting in
a nonzero concentration of the gas at the surface.  Based upon a study of
the heat transfer to roughened glass plates, Owen and Thompson (1963)
suggested

                                          0'8    ,                  (19)

where u^., ZQ, v, and D are the friction velocity, surface roughness, kine-
matic viscosity, and molecular diffusivity, respectively, and a is an
empirical constant determined by the shape of the roughness elements.  In
further investigations by Chamberlain (1966) and Thorn (1972), little
functional relation was found between 3 and ZQ.   Thus, Thorn proposed
                      = a,
where a-|  and a~ are empirical  constants primarily determined by the sur-
face roughness elements.

2.    The Formulation of a Surface Deposition Model

     For pollutants originating from either elevated sources or distant
ground-level  sources,  most of  the pollutant mass is contained in
the mixing layer.  The removal processes, as discussed above, consist of

-------
                                   67
diffusion of the pollutants through the surface layer to the ground and
absorption or adsorption at the atmosphere-ground interface.  As illus-
trated in Figure 5, the diurnal variation of temperature in the surface
layer affects the vertical pollutant distribution through atmospheric
stabilities, and consequently, affects the rate of surface uptake of
pollutants (Hogstrom, 1975).  As a result, an algorithm that can account
for these variations must be included as part of the surface layer model.

     The surface layer model developed in this study for the prescription
of pollutant fluxes is similar to those discussed by Bolin and Granat
(1973) and Galbally  (1974), but  has  been  extended  to include:

     >  Diabatic atmospheric conditions
     >  Nonlinear surface reactions.

We favor this approach over the relatively simple resistance approach
primarily because the latter is restricted to linear surface reactions,
which may not fit all situations of interest.  For example, Hill (1971)
observed that the adsorption of ozone by leaves does not vary linearly
with concentration at high concentration levels.

     In the model, it is envisioned that the transfer of pollutant gases
from the atmosphere to a surface is accomplished via three stages
(Sehmel, Sutter, and Dana, 1973; Galbally, 1974):

     >  The gases are transported to a laminar sublayer just
        above the surface primarily by turbulent diffusion.
     >  The gases are transported through this laminar sub-
        layer primarily by molecular diffusion.
     >  The gases interact by adsorption or chemical reaction
        with the surface.

-------
c
o-
o
a.
4) ^
U IA
                      Stable
               Temperature
             Concentration
        vNeutral
         .Unstable
Temperature
  Temperature
                                                                 JC
                                                                 en
Concentration
Concentration
 12 p.m.
                             6 a.m.
            12 a.m.
                                         6 p.m.
                                                 Time
CTi

CO
         FIGURE  5.   SCHEMATIC ILLUSTRATION OF  DIURNAL  VARIATIONS IN SURFACE DEPOSITION

-------
                                    69
     Thus, as shown in Figure 6, the surface layer is divided into two
parts:  the turbulent layer and the laminar sublayer.  In the turbulent
layer, after the atmosphere reaches an equilibrium state, the atmospheric
diffusion equation becomes
with the following boundary conditions,
                          c = c"    at z = h    ,

                      Kv(i)= F    at z = 2o    •
where c is the cell -averaged concentration in the  mixing layer,  F is the
pollutant flux across the turbulent layer-laminar  sublayer interface,  and
ZQ is the height of the surface roughness element.  The vertical  diffu-
sivity 1C. can be prescribed as

                                           -
where
          k = von Karman constant (= 0.35)
         u* = friction velocity
          z = height
          L = Monin-Obukhov length.

This formula is the result of the similarity theory for the constant-flux
surface layer (Businger et al . , 1971).   For the neutral  case,  the <(>-
function equals unity.  For the stable and unstable cases,  the ^-function
is  greater and less than one,  respectively.   The following empirical
expressions for the ^-function  were  proposed by Businger et al .  (1971)
based on observational data:

-------
Surface
Layer
Turbulent Layer
                            Laminar Sublayer
                                                                                 Concentration
                                                        Height
                                            	_  Velocity
                                                       Concentration;
                                                       Velocity
                                                                                                                -vl
                                                                                                                o
                        FIGURE 6.   SCHEMATIC  ILLUSTRATION OF THE SURFACE LAYER

-------
                                  71
     For the stable case  (L > 0)
                         *s(f) = ]  +  4-7(r)
                                                               (23)
     For  the unstable case  (L < 0)
                                r        r1/4
                         „(£) • f  -  15
-------
                                     72
Across the laminar sublayer, it is assumed that the pollutant flux can be
written as

                            F = eu*(cQ - cs)    ,                   (29)

where cn and c  denote the concentrations at the interface and the surface,
respectively, and 8, analogous to the Stanton number in heat transfer, is
the inverse of a dimensionless resistance for the laminar sublayer.  If it
is further assumed that mass and momentum are transferred in an identical
manner in the turbulent layer, but differently through the laminar sublayer,
then the relationships established by Owen and Thompson (1963) and Thorn
(1972) discussed above can be used:
             ,    /  zn\0.45.  0.8
           B   = aK-^-)    (jy)         (Owen-Thompson)     ,        (30)


                              2/3
                                                                    (31)
     To complete the description of the surface layer model,  a boundary
condition is required at the surface.   Uptake of air pollutants occurs
by chemical  reaction with,  or catalytic decomposition within  either the
soil or vegetation or by these processes at their surfaces.   These pro-
cesses are generally dependent on the  gas concentration  at the surface.
A general equation for the  gas loss per unit area per unit time can be
written as (Benson, 1968),

                             F - Y C*     ,                           (32)
where F is the pollutant flux,  y  is  a  reaction  rate  constant,  and c  the
concentration of the gas at the soil or vegetation surface.  The exponent,
a, denotes the reaction  order.  Eliminating  cn  and c  from Eqs.  (28),
(29), and (32),  the following transcendental  equation  is  obtained for F,
                     I.F  +      .p     . -,  0     ^                   (33)

-------
                                    73
where
                      I =
                          BU,
 f
J
                                  0
Although the reaction order is most likely to be 1, closed-form solutions

can be found for the cases of a = 1, 2, and 3,
        F =
                   I + -
                       Y
                          21
                                  1/2
                                                  a = 1
                                                  a = 2
                                                  a = 3
                                   (34)
where
                    A+ = 3 <4- ±
                                         27Yr
                                                1/2
                    1/2
     It is interesting to note that these formulas  reduce  to  that  of

Chamberlain (1966) or Galbally (1974)  for the special  case of (1)  a

first-order surface reaction and (2) a neutrally stratified atmosphere.

-------
                                74
   VII   SENSITIVITY OF THE REGIONAL AIR  POLLUTION  MODEL
     In the process of model  development,  the study  of  the  sensitivity
of the model plays a vital  role.   Through  systematic  variation  of  input
parameters within the range of physical  reality,  the  sensitivity study
serves as a vehicle for examining the responses  of the  model  under
controlled but realistic conditions.   The  purpose of  carrying out  such
a study is to assess the relative importance  of  various physical para-
meters to the predictions of the  model.

     In order to test the sensitivity of the  regional air pollution
model developed in this study, we selected as a  base  case four
typical days in Spring (as  represented by  the meteorological  patterns
of 4 April 1976 through 7 April  1976) with emissions  projected  for
the year 1986.  A detailed  description of  the meteorological  and emis-
sions data associated with  this  case  can be found in  Part B of  this
report.  After the base case was  chosen, parameters  in  the  base case
were varied one at a time and the regional  air pollution model  was
exercised.  The parameters  studied in this project include:

     >  Horizontal eddy diffusivity
     >  Mixing depth
     >  Prescription of dry-deposition algorithms
     >  Surface reaction rate
     >  SOp/sulfate conversion rate.

A discussion of the sensitivity  of model predictions  to each  of these
parameters follows.

-------
                                  75
A.   HORIZONTAL EDDY DIFFUSIVITY

     Horizontal spreading of the plume by turbulent diffusion in the
atmosphere is expected to play an important role in long-range trans-
port of contaminants.  Dispersion of air pollutants at the mesoscale
depends upon a number of variables.  For example, Kao and Henderson
(1970) investigated the relative diffusion of particles in six dif-
ferent synoptic-scale flow configurations.  It is, however, well known
that for pollutants released at lower levels the plume spread is a
function of travelling time.  As shown in Figure 7, the range of equiv-
alent horizontal diffusivities pertinent to the temporal  and spatial
scales of interest to the present study is

                   105 m2/sec > Ku > 103 m2/sec                    (35)
                                 n
                         4  2
with a median value of 10  m /sec, a number used in the base case.

     To test the effect of horizontal diffusivity on air quality pre-
                                              42          32
dictions, we lowered the base case value of 10  m /sec to 10  m /sec.
The results of the base case simulation are shown in Figure 8* for the
morning hours (2:00-5:00) and afternoon hours (14:00-17:00) on the
fourth day of the base case, 7 April 1976.  The corresponding results
of the simulation with the reduced diffusivity are shown in Figure 9.
A comparison of these figures shows that, as expected, the maximum
concentrations and the impact areas are significantly larger for the
lower diffusivity.  It is clear that this is one of the most important
parameters in the determination of concentrations at long distances.
Unfortunately, it is also one of the most uncertain ones.   Thus a
separate effort will be made to search for a better way to prescribe
this parameter.
* In these figures isopleths are drawn for concentrations of 2n,
  where n = 0, 1, 2,  ....

-------
                 PREDICTIONS     )
                    CURVE AND BOUNDS
                 (HAGE ET AL., 1966)
               DOMAIN OF INTEREST
                 TO THIS STUDY
1.0
                                                                                                 CD
                                 Time (sec]
             FIGURE 7.  HORIZONTAL EDDY DIFFUSIVITY AS A FUNCTION
                        OF TRAVELING TIME AND PLUME SPREAD

-------
                          77
3»     Ut      59      60
                                  70     80
90
100     lie  x 10 km
             (a)  200-500 MST 7 April  1976

                      (1986 emissions)
FIGURE ».   PREDICTED SO? CONCENTRATIONS FOR THE BASE CASE.
           Isopleths at 1, 2, 4, ..., ng/m3; plume
           maxima in boldface.

-------
                                     78
                                                                   100     110
I I  I I I  I I I  I I  I I I  I I I  I I  I I I  I
                     (b)  1400-1700 MST 7 April 1976



                              (1986 emissions)






                           FIGURE 8  (Concluded)
                                                                  100     no   x  10 krr

-------
                                         79
       30     40
        I I  I ,1 I  I ..
                                        60
 70     80
jj-l I I  I I
                                                             90
                                                            H-

          12
:  (    v/   \
                                       "3-OflKOTP
                                          '\  .\ ^
                                         -  \ V,\S\
                                             \ \\\';\
                                              \\ \ • ». V
                                              \ i \ ••. •, \
                                                    •
                                            ?   /\L\\
                                                      :
                                      i COLORRpfl'
                .
                15
              1  1 t  1 1 1
     10
20     30
                     r-'  rer
                                 50      60
        80      90
                                                                          \
                                                                          no
km
                            (a)   200-500  MST  7  April  1976;
                                 Ku  =  IO3 m2/sec.
                                  n
                FIGURE 9.   PREDICTED S02 CONCENTRATIONS FOR REDUCED
                           HORIZONTAL DIFFUSIVITY.  Isopleths at 1,
                           2, 4, ..., pg/iri3; plume maxima in boldface.

-------
                                    80
10     20     30      40      50      60      70     80     80      100     110
1»      20     3*     40     50     60     70     80      90      100     110  X 10 km
                   (b)  1400-1700 MST 7 April 1976;
                        K  = 103 m2/sec.
                          FIGURE  9   (Concluded)

-------
                                   81
B.   MIXING DEPTH

     Vertical ventilation of air pollutants is restricted within the
mixing layer, the top of which is generally defined by the base of
inversion.  As discussed in Part B, in the application of the regional
air pollution model to the Northern Great Plains, the seasonal average
mixing depths in the afternoon as estimated by Holzworth (1972) were
used.  For the base case, the afternoon mixing depth (for spring)
varies from 1,500 meters to 2,800 meters in this region.  These esti-
metes are comparable with those measured in northern Europe (Georgii,
1970; Rodhe, 1971).  In order to examine the effect of the mixing depth
on predicted concentrations, the base case values were uniformly
decreased by a factor of two.  The results for the two three-hour
periods are presented in Figure 10.  It can be seen from a comparison
with the base case results (Figures 8a and 8b) that the concentrations
increase appreciably for lower mixing depths, particularly during
the afternoon.

C.   PRESCRIPTION OF DRY DEPOSITION

     As discussed in the previous chapter, two prescriptions—one
proposed by Owen and Thompson (1963), one by Thorn (1972)--are available
for prescribing the 3 factor in the surface deposition model.  The
two algorithms have different functional forms for the dependent
variables.

     Figures 11 and 12 show the predicted deposition velocities for
1400-1700 MST 4 April 1976 and 200-500 MST 5 April 1976 calculated
using 3 as prescribed by Owen/Thompson and by Thorn.  Davis et al.
(1976) reported that the Black Hills in South Dakota are a strong
sink for atmospheric pollutants;  Thorn's prescription of 3 appears to
produce deposition patterns consistent with these measurements.  On
the other hand, Shepherd (1974) observed that the process of S02
deposition onto vegetation is often surface-limited; the deposition

-------
                                      82
    It
           20      30
           50     60      70      80     90     100     110
                       \  \
                        \
                                •:
                             '

.)

i
24

       \.
28
          16
          .$>
                  -N DOKOTP
                                        ^«
                                          v-
                                     :;,.,.   C^VS—S.
                                       3 DBKOTP
       \ \
""•-.       \
     f'X  \ \
  \  \ \ \ \
/   \  XN\\
(      \   \\ V
                                                                     \
                                                                       \
                                      NEBI?PSKn
                             r-:,a
                            • Is-*"''
                                                       54
               29
                                       T
                                                               9
                                                               a-
   10     20     30     H0      50      60      70     80     90      100     110xK)
                  (a)   200-500 MST 7 April  1976; mixing depths
                       one-half of base  case values
     FIGURE 10.  PREDICTED S02 CONCENTRATIONS  FOR REDUCED MIXING  DEPTHS.
                 Isopleths at 1, 2, 4,  ...,  yg/m3;  plume maxima  in boldface.

-------
                                         83
      It      29     30     UO      50      60     70     80
'1  /" "&/^2
..'   /
  34
                                                        lJ.^-4  I I I
                                                                        100     110
                                                              \


                                                                    \
                                           \    \\ \
                                            \    VV V
                                         COLORflOO
                                       -.':\
                                       \i?;
       10     20
                    30
                                           60     79     80      90
100     no  x  10 km
                (b)  1400-1700 MST 7 April  1976;  mixing depths
                     one-half of base case  values
                             FIGURE 10  (Concluded)

-------
                                       84
           >fc^        /
            I  I illii	14 t t-rf*t I
      !•     20      30     i|0     50      60     79     80      80     100    110   x ]Q y
                                                                        I    1  o-i
                                                                        mm  1-2

                                                                        Ml  2-3

                                                                              >  3
                        (a)  1400-1700 MST  4  April  1976



FIGURE 11.   S02 DEPOSITION VELOCITIES (IN  mm/sec) CALCULATED WITH 3 AS PRESCRIBED

            BY THE ALGORITHM OF OWEN AND THOMPSON

-------
                                        85
10      20      30      H0      60      60     70      60     80      100     110
 I  I I  I !  I
      -••Tl
      •'• •'-
I I  l  I I I
                           HONTRNR
                           A
                         /   \
                        /      \



                       Vl
!l-.!|.!.a.M!.!|.!*!.
                                                 .:

                                                      *
                                                        r
        ^ '  ' '  I
                                       ,

                                     POJ-ORflDO
10     20      30     10      50     60      70     B0      80     100     11C
                            (b)  200-500 f1ST 5 April  1976
                                                                                     10 km
                                FIGURE  11   (Concluded)

-------
                                   86
                                                 80      80
                                                               1*0     110
                  (a)  1400-1700 MST 4 April 1976
                                                                     a  °-2
                                                                           2-4
                                                                           4-6
                                                                           6-8
                                                                           > 8
FIGURE 12.   S02 DEPOSITION VELOCITIES (IN mm/sec) CALCULATED WITH 6
            AS PRESCRIBED BY THE ALGORITHM OF THOM

-------
              87
                                                      0 kn
                                                 D  0-2
                                                 3  2-4
                                                 1  4-6
                                             LJ  6-8
(b)   200-500 MST  5  April  1976
    FIGURE  12   (Concluded)

-------
                                    88
velocities generated by Thorn's algorithm for prescribing 6 also seem
to duplicate such behavior.  Thus Thorn's formulation was selected for
the base case and for use in the model application studies described
in Part B.

D.   SURFACE REACTION RATE

     Concentrations at large distances are apparently affected by the
rate of depletion of pollutant at the surface in the course of its
journey.  In order to test the sensitivity of model predictions to
the surface reaction rate, the base value for k ,  which is 1  cm/sec,
was decreased to 0.1 cm/sec.  The results for the two three-hour
periods are shown in Figure 13.  A comparison with the base case results
(Figure 8) reveals that although the predicted concentrations near the
                                                        3
sources are almost unchanged, the area within the  2 yg/m  isopleth is
approximately doubled.  In the base case computed  deposition  velocities
are limited by either diffusion or surface reactions, depending upon the
time of day and the underlying terrain, but with the lower S0~/sulfate
conversion rate (R = 0.1  cm/sec), the deposition velocity is  always
limited by surface reactions.  Consequently the lower rate leads to
transport of pollutants to greater distances.

E.   S02/SULFATE CONVERSION RATE

     It was stated earlier that one of the major concerns in  the devel-
opment of the present regional-scale model is the  ability to  predict
sulfate distributions, because of the variety of problems apparently
associated with high atmospheric sulfate concentrations.   Reduction
in visibility and increase of acid rain are only two examples.  As
discussed in Chapter V,  the rate of conversion of  gaseous S02 to
sulfate depends upon a number of physical and chemical parameters.
Humidity and the presence of other reactive pollutants are probably
among the most influential ones.  The S02/sulfate  conversion  rate has
been reported to be as low as 0.1 percent per hour and as high as
10 percent per hour.   For a largely undeveloped area with relatively

-------
                                      89
     1*     20     38     i»0     59     60     70     80
                       \
                                 -t-
                              nONTPHO
                                1.

       (V
                        13
                              HYOniNG
                    \
                          12
r
:  15
                                       3 DfiKOTP
                                             \
                                            21
                                                     28
                          90
                                        ! 10
                                                                                  . .O
                                                                                    CO
     10     20
                                 50      60
                  80     90
100     no  x 10 km
                    (a)  200-500 MST  7 April  1976;  surface
                         reaction rate =  0.1  cm/sec
         FIGURE 13.  PREDICTED  S02  CONCENTRATIONS FOR REDUCED SURFACE
                     REACTION RATE.   Isopleths  at 1,  2,  4,  	
                     plume maxima in  boldface.

-------
                                             90
        10
                      30
                                     50
   I  I I  I I I  I I  I I  I I  I I I  I I  I I  I ..I I  I, I  1 I  I
                                                   70      80
                                                                         100
s. .
CD

»_
(O
                                     X
             j-y"
                                 HTOnlNQ
at
a-
  \     \
O-
OJ
   • 18
          2

                                          N. N'~ OflKOTfl
                                           3 OPKOTR
                                               \
                                           NEBRflSlft)





                                                  '

                                           COLORflDO
                     T. i T  1 t  i
                                                                                 /
                                                                                       . -S
                                                                                         00
                                                                                       .  O
                                                                                         f^
                                                                                       .  Q
                                                                                         (U
                                                                                       . ,s
                                                                                        U5
        10
               20     30     U0     50     60      70      80      90     100     l!d  X 1
                        (b)  1400-1700 MST 7 April  1976;  surface
                             reaction  rate = 0.1 cm/sec
                                FIGURE 13  (Concluded)

-------
                               91
clean air and generally low relative humidity, such as the Northern Great
Plains, only a low conversion rate can be justified.  Thus a conversion
rate of 0.3 percent per hour was selected for the base case.  The predicted
sulfate concentrations were extremely low and are shown in Figure 14.   In
the sensitivity study, a higher value of 3 percent per hour was used.   The
calculated S02 and sulfate concentrations are presented in Figures  14
and 15.  It is interesting to note that the distributions of S02 and
sulfate are entirely different.  S02 is a source-oriented pollutant and
S02 emanating from a number of major emission sources is  clearly visible.
As a product of chemical reactions, sulfate is not easily linked to
identifiable sources.  The maximum predicted sulfate concentration, using
a 3 percent per hour conversion rate, is approximately 20 ug/m .  This
level  would exceed the 10 ug/m^ limit which is being considered by  EPA
for the standard.  Also, it has been shown that as little as 1-2 ug/m^
$04 concentration will reduce visibility significantly.

-------
                                92



—
CD



*
as
s-

 /
^^L..X
o ^"~<5
^•v^^r^ s-— — -
'^^"•-^"""""---^ T

20 30 40 50
60 70 80 9« 100 HO
,1 , 1 1 1 , , L , , , , 1 . . , , 1 i i i i j | ,^,- , ; V-T-rT-

. N DRKOTR '• \
I
-^^..... |
^"X. ~s'~.'~"*'^ *
v-*-% "~*'-J5-. i
% %._ ..^:*'-X.« ^
^=^:i "t-"":::-.''-^";!>::'-:r». , |
"^^i^iL-i^;;' ;:j-"iii^->:io f j
"'v:s%:, r 1 :
~ 1 \
3 DflKOTP \ / !
'• **\
': ]
\ X
'i -. \ \ ;


( \\ ) ;

NEBIJflSKH h \ \ V^r— ^ :,
i A \ \ ^ -
'• ' ''•• :- i s
\l ^J L \ -
Hot \ X :
i --i "•-, \
/' .--\ ^i) 'H V
/' ! Viffi; X I
/ / ! V X V
: /»•» -. V
28 x • }
60 70 80 90 100 1!0 x 1(



.«
CD



Q
IB
.Q
S
(O

Q
to
Q




n

®
fli

s
km
            (a)   200-500 MST  7 April  1976;  SOz/sulfate
                 conversion rate  =  3  percent  per  hour
FIGURE 14.   PREDICTED S02 CONCENTRATIONS FOR INCREASED SO?/SULFATE
            CONVERSION RATE.   Isopleths  at 1,  2,  4,  ....  pg/m3;
            plume  maxima  in  boldface.

-------
                                          93
      1*      20
            y
- .'
                             -^ 'BTOMlNG
             12
            (
  \
--•   1-
y-^7-' 17
                                          3 DflKOTP

                             ';
                                                                90      190     !!J

                                                                I I  I I  11. I I  I l.T
                            y
                           /
                                 \-
      10
                    30
50
70     80
           90
100     110  x 10
kn
                 (b)  1400-1700 MST 7 April  1976; S02/sulfate

                      conversion rate = 3  percent per hour.
                              FIGURE 14   (Concluded)

-------
                                 94
                                                60      80
                                                                          10 kn
              (a)  200-500 MST 7 April 1976; S02/sulfate
                  conversion rate = 3 percent per hour
FIGURE 15.   PREDICTED SULFATE CONCENTRATIONS FOR INCREASED S02/SULFATE
            CONVERSION RATE.   Isopleths at 2,  4, 8,  ...  yg/m3.

-------
                                          95

 ;

• \
             20
                    30
           Jt
60
                                         60
              70
90

                               HYOfllNG
                            \
                                        •'N  DRKOTR   ;'
                                                  /
                                                   \
                                                    \    \
                                         8  DRKOTP
                                       j/COLORflDO
               110
                                         60
                                                                      100
                                               x  10 km
                   (b)   1400-1700 MST  7  April  1976; S02/sulfate
                        conversion rate  =  3 percent per  hour
                            FIGURE 15   (Concluded)

-------
                                   96
           VIII   SUMMARY AND CONCLUSIONS FOR PART A
     Part A contains a review of previous  studies  pertinent  to  the
transport of air pollutants  over large  distances  (ca.  100 to  1000 km),
followed by a delineation  of the major  attributes  governing  the distri-
bution of atmospheric pollutants on  this scale.  The development of a
regional air pollution model  accommodating these major attributes is
then described.   This model  is primarily intended  for  the prediction of
                                                                   2
pollutant concentrations averaged over  areas of approximately 100 km
with a temporal  resolution on the order of 3 hours.  Two unique fea-
tures of this model  are the  assumption  of  homogeneous pollutant distri-
butions in the vertical and  the  incorporation of a model of diurnally
varying surface  deposition.   This model was thoroughly tested via
sensitivity analysis.   The responses of the model are consistent with
expectations based on physical  reasoning.

-------
                      97
                   PART B

APPLICATION OF A REGIONAL AIR POLLUTION MODEL
        TO THE COAL DEVELOPMENT AREAS
        IN THE NORTHERN GREAT PLAINS

-------
                                    98
                             IX    OVERVIEW
     The Northern Great Plains currently enjoys some of the cleanest air
and possesses some of the richest coal  deposits in the United States.
The U.S. energy program includes mining this coal  and using it for elec-
tric power generation or synthetic fuel production.   Such activities are
certain to adversely affect air quality in the Northern Great Plains.   In
Part B of this report we examine this impact by applying the regional  air
pollution model discussed in Part A.

     Figure 16 shows the locations of proposed energy conversion plants
scheduled for completion before 1986.  These plants  are scattered over a
large area containing many types of terrain.  A few  are located in the
Rocky Mountains, where pollutant dispersion modeling would be more diffi-
cult, but fortunately most of the facilities of interest to the present
study lie in the plains of Montana, Wyoming, Colorado,  and North Dakota.

     In Part A of this report, the development of  a  regional  air pollution
model was described.  This model is composed of two  interconnected sub-
models, a mixing layer model  and a surface layer model.   The mixing layer
model is designed to treat the transport,  diffusion, and chemical  reac-
tions of pollutants by numerically solving the two-dimensional  atmospheric
diffusion equation
 9C.      3C,      3C,    a  /    9C.  \       /   3C.
                   *-             £  K    1
 3t    " 3X    v  3y    9X \'XH 3x /    3y \NH 17

                                + Rn.  +  Si  -  (c.  -  c) • D

                                1=1,2    .                        (36)

-------
                                        99
       10     2»     39     lie     60      60      70      60     60     100    110

CD"

 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 i 	 iii i i i i
20 30 U0 50 60 70 80 90 100 IIP X 1C

"CD

.s
CO


Q


.®
1£>

"in

s

.Q
CO

0
cu



13


k
FIGURE 16.   ENERGY CONVERSION  FACILITIES SCHEDULED FOR COMPLETION  BEFORE  1936

-------
                                   100
For this study  i =  1 for S02 and i = 2 for sulfate,  although  the model-
ing approach  can be extended to handle more complex  chemistry.

     The surface layer model, which is embedded in the mixing layer
model, is designed  to calculate the pollutant fluxes lost  to  the ground
due to dry deposition.  As shown in Part A of this report,  for  linear
surface reaction the pollutant flux to the ground surface,  or the  sur-
face removal  rate,  can be expressed as
                  F-c/y-*     m^^\                   07)
where u+ is the  friction velocity and r  = I/Y is the resistance  to
       11                              s
deposition at the  ground surface.  The surface removal  rate  generally
varies linearly  with concentration unless the concentration  is  so high
that saturation  effects take place (Hill, 1971).   Measured values of
r  for deposition  of S0? on grass appear in Table 14.
         TABLE  14.   SURFACE RESISTANCE MEASUREMENTS FOR  S02
                          Surface
                         Resistance
      Type of Surface      (sec/m)     	Reference	
    Grass (3 cm)/summer       80       Shepherd (1974)
    Grass (3 cm)/winter      300       Shepherd (1974)
    Grass (ZQ = 0.5 cm)      150       Garland et al.  (1974)
    Grass (9-13 cm)           75       Owers and Powell (1974)

-------
                                 101
     As discussed in Part A, two different formulas, one by Owen and
Thompson (1963) and one by Thorn (1972), were examined for the prescrip-
tion of 0.  A comparison of these formulas in the sensitivity analysis
revealed that the formula proposed by Thorn appears to yield more
realistic results.  As a result, this formula was adopted in this study.

     In the next chapter (X),  the compilation of emissions and meteo-
rological data for the Northern Great Plains is described.  The regional
air pollution model was exercised for three different meteorological
patterns and two emissions scenarios.  Based upon these simulations,
the impact of energy conversion plants on air quality is analyzed in
Chapter XI.  The application of the regional  air pollution model  to the
Northern Great Plains is summarized in Chapter XII.

-------
                                  102
                 X  COMPILATION  OF THE  DATA  BASE


     The application of the  regional  model  requires  extensive  data,
which can be divided into  four general  categories:

     >  Emissions data
     >  Meteorological data
     >  Vegetation data
     >  Air quality data.

Considerable effort was required  to collect and analyze  input  data for
the Northern Great Plains  modeling  exercise.   This section  is  devoted
to the discussion of this  task.

A.   EMISSIONS DATA

     In l.ho NGP 8(> percent of the t.otol  SO   emissions .ire attributable.
                                         X
to point sources (LPA, 19/61)).  Future energy  development should  increase
this figure, so only point source emissions were  included in our model.
The point source inventory was assembled by the EPA  Region  VIII  office
in Denver from the most recent complete  base-year emissions data for
each state, either 1973 or 1975.  The emissions data were obtained from
permit application data provided  by plant engineers, or  from data pro-
vided to an individual state by a hired  contractor.  Emissions estimates
projected for future sources were drawn  from the  following:  (1) "Existing
and Proposed Fuel Conversion Facilities  Summary"  (EPA, 1976c), (2) Northern
Great Plains Resource Program: Atmospheric Aspects  Workgroup  Report", (NGPRP,
1976), and (3) "FPC Form 67:  Steam Electric Plant Air and  Water Quality
Control Data for the Year Ending  December 31, 1975"  (FPC,  1976).

     Tables 15 and 16 and Figures 17 and 18 summarize the  emissions  data.
The tables list point sources in  the 1976 and 1986 inventories that  emit

-------
                     103
TABLE 15.  POINT SOURCES EMITTING  MORE THAN
           10,000 TONS OF SOX  PER  YEAR IN 1976
Source
Dave Johnston, WY
Ideal Basic Industries
Naughton, WY
Exxon, MT
Milton R. Young, ND
Stanton, ND
Leland Olds, ND
Hayden, CO
MW
750
-
710
-
240
167
650
180
Grid
Location
(40,32)
(46,7)
(4,20)
(20,64)
(77,78)
(76,81)
(76,81)
(30,5)
SOX Emissions
(tons/year) % Control
31 ,000 50% control #3
25,800
21,700
17,300
16,000
15,600
14,500
14,200

-------
                                      104
             TABLE 16.   POINT SOURCES EMITTING MORE THAN 10,000
                        TONS OF SOV PER YEAR IN 1986
                                  /\


                                                        SOX  Emissions
                   Source        MW     Grid  Location     (tons/year)           % Cont.
Gerald Gentleman, NB
Craig, CO
Naughton, WY
Col strip, MT
Pawnee, CO
Coal Creek, ND
Wyodak, WY
ANG, ND Antelope Valley
Coyote, ND
Jim Bridger, WY
Dave Johnston, WY
Milton R. Young, ND
Ideal Basic Industries,
American Natural Gas,
ND
Peoples Gas, ND
Exxon, MT
Stanton, ND
Leland Olds, ND
Hayden, CO
Laramie River, WY
1300
1520
15101
20602
1000
1000
660
880
880
2000
750
688
-
-
-
167
650
430
1500
(78,13)
(28,6)
(4,20)
(35,65)
(59,3)
(78,82)
(45,47)
(73,80)
(73,80)
(19,18)
(40,32)
(77,78)
(46,7)
(73,81)
(66,81)
(20,64)
(76,81)
(76,81)
(30,5)
(48,23)
» - i *s 	 /
98,200
87,900
50,000
50,000
50,000
42,500
38,400
37,400
37,400
34,700
31 ,000
31 ,000
25,800
21,500
21,500
17,300
15,600
14,500
14,200
11,000
NA
50
-
38
-
-
20
55
503
404
NA
NA
NA
N;,

-
-
83%
1   Unit 4 & 5 may not be built, equivalent units  may be built in Utah
2   Units 3 & 4 (700 MW each)  may not be constructed
3   Control on Unit No. 4 only
4   Control on Unit No. 2 only

-------
                             105
CD"
r> .
(0
03°
Sh.
LO
a.
fll~
»_
K, , . _. j
20 30 40 50 60 70 80 90 100 11(5
i i T .n • i i i ; i . , , , -, i -, i -T"T~. ! , ; i
'. riONTRNP
I,
: «7 ^
I 13
: ^36

10
Source: EPA
o *
23
0-
17
N OQKOTR \ _,
' V
* *52°8 1
1* * " " :
V
/
3 DflKOTO ^ /
* *•
tj
C-
NEBRP3KQ ^V>>~~~~>'x^ L "
N
4o „. COLOKPDO t
24 >4)3 :
20 30 U0 50 60 70 80 90 100 110 X l6
(1^77).
"CD
.9
CD
'LO
en
<\i
kn
FIGURE 17.   POINT SOURCES IN THE NORTHERN GREAT  PLAINS  IN  1976.
            Numbers represent emissions  in kg/min;  small diamonds
            represent emissions  of less  than  10  kg/min  or
            6000 tons/year.

-------
                                          106
       i«      2*      30     Ue     50     60     7«     80
       * w      .       .       1       I       t       I      _'
                                                                            :io
                               "'  "'

o>"
OD
I?"
CD'
ID"
m"
o.
-i T -r . -i'^i • i 'i-. T ^"v »-S • ri'-n-i . T~r i v , .
;
.
- 47 ^4
: 13
tf
: * «
" * -^ *52
.<- 23*
19
'. 101 *
- * *58
* c. *


*•
	 	 __^_™___ - -
<•
*
C-
' . | | | I | | 1 1 4 t 1 1 1 1 1 1 1 1 _l. 1 1 f 1 --|_L 1 1 1 .1 1 J 1 J^lL.-I -
                                                         vr 1-1 v-i-r->
                                                         25
                                                                                    .e
                                                                                    CD
                                                       164
       1*     20      30


Source:  TPA (1977).
                                                                            110  x 10 km
           FIGURE 18.  POINT SOURCES  IN  THE  NORTHERN GREAT PLAINS IN 1986.
                       Numbers represent emissions in kg/min; small diamonds
                       represent emissions  of less than 10 kg/min or
                       6000 tons/year.

-------
                                  107
more than 10,000 tons of SO  per year.  The figures show the locations
of all grid cells that contain sources in the 1976 and 1986 inventories.
Cells where the sum of all point source emissions exceeds 6,000 tons per
year are marked by dark diamonds and the strengths of emissions are noted.
Of the total SO  emitted by point sources, 97 percent was assumed to
be SOp and 3 percent sulfate.

B.   METEOROLOGICAL DATA

     The long-range transport model requires several different meteorolog-
ical inputs.  These include:  vertically averaged horizontal winds,
surface wind speeds, afternoon mixing depths, and a measure of the
thermal gradient near the ground.  These data were compiled for three
meteorological episodes:

     >  Strong wind winter case based on data for 27-31 January 1976.
     >  Stagnation spring case based on data for 4-7 April  1976.
     >  Moderate wind summer case based on data for 9-11  July 1975.

     The winds in the mixing layer determine how pollutants move  after
they are emitted, so characterization of these winds is crucial to the
modeling exercise.  The winds for the three test cases were calculated
by Mr. Loren Crow, a consulting meteorologist, under subcontract  from
SAI.  He computed a set of wind vectors for the 30-point coarse grid
shown in Figure 19.  These coarse grid wind fields were constructed at
six-hour intervals to represent vertical averages through a layer 500
to 1500 feet above the terrain.  The wind data were derived from:

     >  The geostrophic winds associated with the 850 and 700 millibar
        maps available every 12 hours from the U.S. Weather Service.
     >  Twice daily measurements from the eight U.S. Weather Service
        rawinsonde stations shown in Figure 19.
     >  Twice daily pibal measurements taken on alternate days by the
        EPA monitoring network at the stations shown in Figure 19.

-------
                                                              O  National Weather
                                                                 Service  Surface
                                                                 Wind  Station

                                                              D  National Weather
                                                                 Service  Rawinsonde
                                                                 Station

                                                              A EPA Pibal  Station
                                                                                        o
                                                                                        CO
FIGURE 19.  WIND MEASUREMENT NETWORKS IN THE NORTHERN GREAT PLAINS

-------
                                  109
     After the 12-hour maps were completed additional  maps at inter-
mediate six-hour intervals were generated by consulting three-hour
surface wind maps to estimate probable changes in the  upper air flow.
Finally, these six-hour coarse grid wind vectors were  linearly inter-
polated in time and bilinearly interpolated in space to produce three-
hour-averaged winds for the entire 120 x 100 grid.

     Surface wind speeds are required for deposition calculations  in
the surface layer submodel.  Mr. Loren Crow collected  hourly surface
wind data from the National Weather Service stations shown in Figure 19.
The surface wind vectors were averaged over three-hour intervals and
then a value for each grid cell was interpolated according to the  fol-
lowing prescription:
                              rii
-------
                                  no
Idaho Falls from January 1955 through May 1958.  These data, averaged
for the months of January, April, and July, are plotted in Figure 20.
As expected, the gradient is closely linked to the incoming solar radi-
ation and shows both diurnal and seasonal variations.  This information
is incorporated in the regional  model through the dimensionless variable,
exposure class (Liu and Durran,  1977).   The second set of vertical axes
in Figure 20 gives exposure class as a  function of the time of day.

     The afternoon mixing heights determine the thickness of the modeling
region, and hence the amount of  dilution due to vertical  diffusion.   Mix-
ing height data for the Northern Great  Plains are virtually unobtainable.
In the regional model we used the seasonally averaged afternoon mixing
heights shown in Figure 21  (Holzworth,  1972).  It is  unfortunate that
particular data for our three episodes  are not available, but  since  the
afternoon mixing depths only approximately represent  the  depth of the
layer above the ground through which most mesoscale transport  occurs,
the seasonal averages are probably adequate.

C.   SURFACE DATA

     Surface deposition rates are influenced by vegetation and ground
cover.  We have already noted that available data are insufficient to
distinguish the surface resistance of a pine needle from  that  of a
blade of grass.  Moreover,  the different geometries of pine trees and
grasses generate different amounts of mechanical  turbulence, thereby
promoting different rates of deposition.   Figure  22 shows the  modeling
region divided into six different vegetation types.  The  divisions
reflect differences in potential  natural  vegetation (Kuchler,  1966)  or
current land use (Marschner, 1950).   The surface  roughnesses  (without
zero plane displacement)  associated  with each vegetation  type  appear
in Table 17; they  were estimated from a self-consistent summary of
experimental data  compiled  by Sellers (1965).

-------
   January
             /Tlv
                                        April

                                                                      a
                                                                      5
July
Source:   DeMarrais  and  Islitzer  (1960).
              FIGURE  20.  TEMPERATURE GRADIENTS AND EXPOSURE CLASSES AT  IDAHO FALLS, IDAHO

-------
                                 112
•\      \	
      20     39     U0     50     60     70     80      90      100    110 X 10 l
-------
                                   113
 It      29
19
       39      39
U9     59     69     79     89




    (b)   April  1976






FIGURE 21  (Continued)
89
100     110x10 km

-------
                                           114
-U
     1»     20      30      40

i~Vi'r*!"y-iri~i rv.TV i "r~i '""KJ"
               N
                                    50
                                     (c)   July 1975
                                  FIGURE 21 (Concluded)

-------
                                           115
                                                            WSG •:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:•:
Key:

ASG  alfalfa,  small grains
DS   sagebrush,  steppe
FG   corn,  soybeans, oats
FW   pine,  fir,  spruce
SG   alfalfa,  hay
WSG  wheat,  barley, flax
                 FIGURE 22.  VEGETATION IN THE  NORTHERN GREAT PLAINS

-------
                                 116
        TABLE 17.  SURFACE ROUGHNESSES FOR VARIOUS VEGETATION TYPES

                                                  Surface Roughness
     Vegetation Code*                                   (cm)	
            SG           alfalfa, hay                   2.4
            DS           sage brush, steppe             2.6
            ASG          alfalfa1, small grains          15
            WSG          wheat, barley, flax            22
            FG           corn, soybeans, oats           75
            FW           pine, fir, spruce              283
     * Code used in Figure 22.

 D.    AIR QUALITY DATA

      Initial  and boundary  pollutant  concentrations are  the remaining
 inputs  required  by  the  regional model.  The ideal way to generate
 such inputs  is from air quality measurements, but this  requires a
 dense modeling network  throughout the region and along  its borders.
 Measurements  taken  at a point are not strictly equivalent to the
 volume-averaged  concentrations used  in grid modeling; the problem
 is  especially serious in long-range modeling because the grid cells
 are large.  Each cell in the regional model represents a layer 1000
 to  3000 meters thick above a surface of 100 square kilometers.   The
 pollutant concentration measured at a single location in such a cell
 could certainly be much different from the actual average concentra-
 tion in that cell.  In particular, measurements taken at urban  loca-
 tions in the Northern Great Plains are unsuitable for input to  or
 validation of the regional  model.   The most useful  measurements for
 regional  modeling are those gathered at ten S02 monitoring  sites
established by the EPA at the rural  locations  shown in Figure 23.
Two of these stations  had continuous  SO- monitors,  and the  rest
took a  24-hour-averaged  measurement  every  six  days.

-------
                                           STANTON  WASHBURN
                                          MOTT
NORTH  DAKOTA
                                 BUFFALO
                                 BELLE  FOURCHE
                                                      SOUTH  DAKOTA
FIGURE 23.   EPA  S02 MONITORS IN THE NORTHERN GREAT  PLAINS

-------
     These ten stations can be of some use when assessing the accuracy of
model predictions, but they simply do not provide enough data to deter-
mine initial and boundary conditions.  Rasmussen, Taheri, and Kabel (1974)
                                                               3
estimated a general  background S09 concentration of 1  to 4 yg/m  and
                                     o
Georgii (1970) measured 0.5 to 2 yg/m  over Colorado.   In the regional
model a background S0? concentration of 1.5 yg/m  was  used for both
initial and boundary conditions.  McMullen, Faoro, and Morgan (1970)
suggested an average nonurban sulfate concentration of 2.5 yg/m .   A
comparison of the estimates of Georgii and Rasmussen et al.  suggests that
background S09 concentrations in the Northern Great Plains are somewhat
                                        3
lower than many rural  sites, so 1.5 yg/m  was also taken for initial and
boundary sulfate concentrations.

-------
                                 119
                  XI    AIR  QUALITY ANALYSIS
     For the assessment of the impact of coal  development  on  air quality
in the Northern Great Plains, three meteorological  patterns were selected.
The selection was based on considerations of meteorological and air  quality
conditions of interest and data availability.   The  three cases probably
represent typical situations for winter, spring,  and summer in this  area,
as shown in Table 18.  For each meteorological  pattern, the regional  air
pollution model was exercised for two emissions scenarios:

      >   Scenario  I—emissions  in 1976
      >   Scenario  II--emissions  in  1986.

A  complete  list  of the SOp  concentrations predicted for the three  cases
and two  emission  scenarios  (a  total of  six simulations) is presented
pn isopleth contour maps  in Appendix B.  The isopleth contour intervals
are 2n yg/m , where n = 0,  1,2 ...  The model  is started  from a constant
                                 3
initial  concentration of  1.5 yg/m  SO^, so the first several  plots
in each  series show the concentration field building up to a  quasi-
equilibrium.  Maps of point source locations in 1976 and  1986 are
included as clear overlays  in a pocket  inside the back cover; they
 fit over the  maps in  Appendix  B to show the  locations of these sources
 relative to their plumes.   Because the  initial SO-  concentration assumed
                          3
 in the model  was 1.5  pg/m ,  the outer  isopleth around each plume is gen-
 erally  the  2  yg/m contour.

-------
                                 120


         TABLE 18.   PERIODS CHOSEN FOR AIR QUALITY ANALYSIS

          Case          Season          	Dates	
            1           Winter          27-31  January 1976
            2           Spring          4-7 April  1976
            3           Summer          9-11  July  1975
                                                                    3
     The isopleth maps  show that  concentrations  greater  than  16  yg/m
are rarely predicted over more  than  one  cell  by  the  1976 emissions  inven-
tory; in the 1986 case  predictions rarely  exceed 32  yg/m .  Table  19
provides a comparison of total  SO emissions  within  the  study area  with
SO  emissions in the State of Ohio.   Ohio  generates  thirteen  times  the
  A
SO  emissions in one-tenth of the study  area  of  the  Northern  Great
  X                                                                 o
Plains.   Twenty-four-hour-averaged concentrations  exceeding 150  yg/m
have been measured at many locations in  Ohio  (EPA, 1976b),  so the  4 and
      o
8 yg/m  predictions shown in Appendix B  seem  reasonable.  The three
cases listed in Table 18 are analyzed in more detail  in  the following
sections.
             TABLE 19.   S02  EMISSIONS AND AREAS OF  OHIO
                        AND  THE  NORTHERN GREAT PLAINS
             State
        Eastern Montana
        Nebraska
        North Dakota
        South Dakota
        Wyoming

          Total
        Ohio
Total SOX Emissions
(tons/year)
43,000
55,000
80,000
3,000
70,000
251,000
3,347,000
Area
(sq. miles)
98,000
77,000
71,000
77,000
98,000
421 ,000
41,000

-------
                                 121
A.   WINTER

     The winter case, characterized by low mixing depths and a strong,
relatively constant wind from the northwest, provides favorable condi-
tions for long-range transport.  Selected 850 millibar maps for the
winter case meteorology are shown in Figure 24.  Figure 25 shows the
predicted S02 concentrations for the 1976 and 1986 emissions inventories
33, 48, and 67 hours after the beginning of the simulation.  The great-
est cell-averaged concentration at the head of each plume is given in
    3
yg/m .   The predicted SCL concentrations in the southeastern corner of
                                        3
the modeling region are less than 1 yg/m  in Figures 25a, b, e, and f.
This value is lower than the initial and boundary S02 conditions.   Con-
centrations decrease below background when an air parcel moves extended
distances without encountering significant emissions.  The diagonal
corridor from southeastern Nebraska through northwestern South Dakota
to the Canadian border is free of major SO- emissions.  Pollutant  par-
cels blowing down this corridor experience surface deposition and
chemical decay losses but no emissions loading, hence SO- concentrations
may be depleted below the initial concentrations.   A shift in the  winds
(Figures 25c and d) can eliminate such regions.  The 1986 case (Figure
25d) reveals SOp transport to great distances.   Plumes from Col strip,
Montana and Wyodak, Wyoming merge and travel into Sutherland, Nebraska
to link with the Gerald Gentleman plume, and the 2 yg/m  isopleth  from
the North Dakota developments extends well  into Iowa.

B.   SPRING

     Unlike the winter case, the spring case conditions are  favorable
for the retention  of  pollutants within the  Northern Great Plains.   The
850 millibar maps  for the  spring case, given in Figure  26,  show a
stagnant  high  pressure system  lingering over the region.  The  result-
ing winds  in the mixing  layer  are  light and variable.   Figure  27,
which  indicates pollutant  concentrations 24, 33, and  51  hours  after
the start  of the simulation, shows  a reversal  in the  mixing  layer

-------
,       \    x     r

 W^H^iif^T
                                                                                         ro
                                                                                         ro
                   (a)   500 MST 27 January 1976


FIGURE  24.  WINDS AT 850 MILLIBARS ALTITUDE DURING  27-31 JANUARY 1976

-------
                                                                        ro
                                                                        oo
(b)   1700 MST  27  January 1976
    FIGURE  24  (Continued)

-------
(c)   500  MST  28  January 1976
   FIGURE  24  (Continued)

-------
                                                                          ro
                                                                          en
(d)   1700 MST 28 January 1976
    FIGURE 24 (Continued)

-------
                                                                          ro
                                                                          CTl
(e)   500 MST 29 January 1976
   FIGURE 24 (Continued)

-------
(f)   1700 MST  29  January  1976
    FIGURE  24  (Continued)

-------
S x? -^
                                                                                                     IN3
                                                                                                     co
                             (g)  500 MST 30 January  1976
                                FIGURE 24 (Continued)

-------
(h)   1700 MST 30 January 1976
    FIGURE 24 (Continued)

-------
                                                                         GO
                                                                         O
(1)   500 MST 31 January 1976
   FIGURE 24 (Continued)

-------
(j)   1700 MST J1 January 1976
    FIGURE 24 (Concluded)

-------
                                                                                                                GO
                                                                                                                ro
                                                                                                     110 ii 10 In
(a)   1400-1700 MST 28 January 1976;  1976 emissions     (b)   1400-1700 MST 28 January 1976; 1986 emissions
         FIGURE 25.   PREDICTED S02 CONCENTRATIONS  FOR WINTER CASE.   Isopleths at  1,  2,  4,  ...,  yg/nT;
                     plume maxima in boldface.

-------
                                                                                                I»«   110
                                                                                                               CO
                                                                                                               CO
                                                                                                1*«   no • 10 I
(c)   500-800 MST 29 January 1976; 1976 emissions      (d)   500-800 MST 29 January 1976;  1986 emissions
                                         FIGURE  25   (Continued)

-------
  It    2«   3t
                     E«   6»    7»   8«   9»    !•»   I'.i! I 10
                                                                                                                      oo
                                                                                                                      •pa
                                                                                                90   I a*   lie « 10 !•
(e)  800-1100 MST  30 January 1976;  1976 emissions        (f)   800-1100 MST 30 January 1976;  1986 emissions
                                             FIGURE 25   (Concluded)

-------
        13^480 [ s'lS     v   \
                                                                                          U)
                                                                                          en
                    (a)   500 MST 4 April  1976


FIGURE 26.   WINDS  AT 350  filLLIBARS ALTITUDE  DURING 4-7 APRIL 1976

-------
 A"*    I   V
/ N     \>
                                                                                                       GO
                              (b)   1700 MST 4 April 1976
                                 FIGURE 26 (Continued)

-------
                11 477    19-02-.
(c)   500 MST 5 April  1976
  FIGURE 26 (Continued)

-------
                                                                          Co
                                                                          co
(d)  1700 MST 5 April 1976
   FIGURE 26 (Continued)

-------
                                                                       CO
(e)   500 MST 6 April  1976
  FIGURE  26 (Continued)

-------
(f)   1700 MST  6  April  1976
   FIGURE 26 (Continued)

-------
         1L425 i^-
(g)   500  MST  7 April 1976
  FIGURE  26  (Continued)

-------
(h)   1700 MST 7 April  1976
   FIGURE 26 (Concluded)

-------
I*    2»   It
                  5»    ED    7»   8«
                        I i i I I I I I I i I i
2«
I I I I
 3«
T+T
                                                                       lit    5«    et   7»    8«   9»
                ntmiflxn   N Doroin
      .
   13 "
     I I I I I I I I ' I I I ' ' ' I ' I I ' I I I I I I1
                                                       20
                                                                    24
                                                                          Htonr((8
                                                                2  —7=i2°>r"
                                 I I I I 1 I I I I I I
                                                                                    7   23
                                           X
                                                                                        68
 •aigSH1    'e^-^- 3  cou
   ^!T'     ~—-^ ...x w .*—™r
  .Sfr^.'.M
     3*   lit   S«    6»
                                                                                                                   00
              «*    6«
                                               1H . 10 ta      |f    29    3(
                                                                                6«   It    a«   90
 (a)  500-800  MST 5 April 1976;  1976  emissions          (b)   500-800 MST 5 April  1976; 1986 emissions
                    FIGURE 27.    PREDICTED S02 CONCENTRATIONS FOR SPRING CASE.   Isopleths
                                  at 1, 2, 4,  ...,  yg/m3;  plume maxima  in boldface.

-------
           30   K0
           I I I I I I I I
                     5»
                           60
                                     60
                                                                10
2«   30
I I I I I I I I
                                                                                                   89    90    100
                   nOHTONO   N DKKOTP
               .^3-
                  '\
       | i i i i | i i i i |
                                                                                       „ \COLORRDO-,
                                                                                                   35

                     50    8«    70    80    90    100   110 i 10 hi      |«    20    3«    q«    53    60    70    60    90    I CD   110 X 10 in
(c)   1700-2000  MST 5  April  1976;  1976 emissions           (d)   1700-2000  MST 5  April  1976; 1986 emissions
                                                FIGURE  27  (Continued)

-------
     g«
          3*

         •H-T
 144
-Hi
 6*
T-H
 7»
•H-i
 9*

T+f
                                10 l'''
[»   2«   38   *•   S»   6«

I I I.I I I.I I I I I I	I
                                                                                                 88    90    l««   110
                                                                                       9 OOKOTP   \.   \

                                                                                                                _)
                                                                                                                   x> r
                                                                                                                   1
!•    ?«    3>    UK    5*    6<    7«    $•    »•    1M   110 i 10
                                                                        3«    U<    5«    61    7t    8»    9i>
(e)  800-1100  MST 6  April  1976;  1976  emissions          (f)   800-1100 MST  6  April  1976;  1986 emissions
                                              FIGURE 27   (Concluded)

-------
                                 146
wind pattern.  Depleted areas (dotted regions) again appear in the
emissions-free corridor.

C.   SUMMER

     The 850 millibar maps for the summer case appear in Figure 28.
Figure 29 gives S02 concentrations 33 and 51 hours after the start of
the simulation.  It shows slow westerly flow in Wyoming and Colorado
and a strong southerly flow through the Dakotas.

D.   AIR QUALITY IMPACTS

     Certain behavior is  common to all  three cases.   The predicted
1986 S09 concentrations are seldom more than double  the 1976 values,
                                   3
but the area impacted by  the 2 yg/m  isopleth increases dramatically
(see especially Figures 27e, f and 29c, d).   The deposition rate
showed considerable temporal and spatial  variation (see Figure 12);
in all cases deposition rates were generally lowest  in the early
morning and highest in the late afternoon.   Predicted S0? concentra-
tions reflected this;  they were generally highest at dawn and  lowest
at dusk.

     Ten monitoring stations measured S02 at rural sites in the NGP.
Most stations measured only one 24-hour-average concentration  in each
multi-day episode.   These data are displayed in Figure 30; most of the
                                      3
measurements were less than the 4 yg/m  noise limit  of the instruments
(as were most of the model predictions).   These data agree qualitatively
with the model  predictions.

     The EPA significant  deterioration increments for Class I  and Class II.
regions are given in Table 20.  Currently the entire Northern  Great Plains
is a Class II region.   However, it has been proposed that some areas be
reclassified as Class I.   In our simulations the Class I increments were
exceeded by the 1986 energy developments only near plant stacks; Class II
increments were never violated.  It should be noted, however,  that in view

-------
1—MA;   \
     r-.n/  \L   S
                     (a)  500 MST 9 July 1975



    FIGURE Zo.  WINDS AT 850 MILLIBARS ALTITUDE DURING 9-12 JULY 1975

-------
                                                                       CO
(b)  1700 MST 9 July 1975
  FIGURE 28 (Continued)

-------
r  ,.x
\    -V^ c^V ,  I
                                 (c)  500 MST 10 July 1975
                                  FIGURE 28 (Continued)

-------
(d)   1700  MST  10  July  1975
   FIGURE  28  (Continued)

-------
(e)   500 MST  11  July  1975
  FIGURE 28 (Continued)

-------
(f)   1700 MST 11  July 1975
   FIGURE  28 (Concluded)

-------
l«   it    3t

 (a)   1400-1700 MST 10 July 1975;  1976 emissions       (b)  1400-1700 MST 10 July 1975; 1986 emissions
                     FIGURE  29.    PREDICTED  SOo  CONCENTRATIONS FOR SUMMER CASE.   Isopleths
                                  at  1,  2, 4,  *~...,  ug/rri3;  plume maxima  in boldface.

-------
 U    2«   3(    II •    6«   S»    7«    89    9«   !••   US I 101.
                                                                     3«   l|«   B»    6«    7»   8«    9>    181!    119 I 10 lr»
[c)   800-1100 MST 11  July 1975;  1976 emissions
(d)   800-1100  MST 11 July 1975; 1986 emissions
                                            FIGURE 29  (Concluded)

-------
                                                             0/17*      0

                                                            STANTON  WASHBURN
                                                            MOTT
                                                   BELLE FOURCHE
* Second value is the maximum three-hour  average  obtained  from a continuous

  site; NM means no measurement was  made.


                                        (a)   27-31  January 1976
                                                                                     NORTH DAKOTA
                                                                        SOUTH  DAKOTA
                                                                     monitor at the same
                                                                                                            01
                                                                                                            en
                                                     3>
FIGURE 30.   24-HOUR--AVERAGE S02  MEASUREMENTS  (IN  yg/mj)  IN  THE  NORTHERN  GREAT  PLAINS

-------
                              STANTON  WASHBURN
                                0        0
                             MOTT
                    BUFFALO
                       NM

                    BELLE  FOURCHE
NORTH  DAKOTA
                                         SOUTH  DAKOTA
(b)   4-7  April  1976
FIGURE 30  (Continued)

-------
                                STANTON  WASHBURN
                                 0/0*       14
                                MOTT
                                  0
                       BUFFALO
                           0

                       BELLE  FOURCHE
                             0
 NORTH  DAKOTA
SOUTH DAKOTA
  (c)   9-11  July  1975
FIGURE 30  (Concluded)

-------
                                 158
of the approximations invoked in the present model formulation, liujhiT uiu'o
tainties are associated with model predictions near major emissions sources.

     The regional model also predicted sulfate concentrations for the
same episodes.  The 0.3 percent per hour conversion rate for SCL to
sulfate selected for this investigation did not result in significant
sulfate production in the Northern Great Plains.  Consequently,
                                                          3
sulfate concentrations were largely masked by the 1.5 yg/m  initial
and boundary concentrations.  As noted in Chapter VII, sulfate concen-
trations are increased considerably by a faster conversion rate of
3 percent per hour.
       TABLE 20-   SIGNIFICANT DETERIORATION INCREMENTS FOR S02
                                      S02 Increment
                Averaging
                  Period           Class I      Class II
                One year               2           15
                24 hours               5          100
                3 hours               25          700
         Source:  Federal Register (1974, 1975).

-------
                                  159
            XII   SUMMARY AND CONCLUSIONS FOR PART  B
     The regional  air pollution model  described  in  Part A was applied
to the Northern Great Plains to assess the  air quality impacts of exist-
ing and proposed energy developments  utilizing coal  resources in that
area.  Emissions inventories were prepared  for the  years 1976 and 1986.
Three meteorological  scenarios, a strong-wind winter case, a stagnation
spring case, and a moderate-wind summer case, were  selected for the
impact analyses.  Model simulations were carried  out for each combina-
tion of emissions  inventory and meteorological scenario.  Sulfur dioxide
and sulfates were  considered.   In general,  the predicted impacts are
greatest in spring, intermediate in winter,  and  lowest in summer.  From
the present preliminary results it appears  that  neither the 1976 nor
the 1986 emissions as estimated in this study are likely to cause pol-
lutant concentrations significantly higher  than  background values at
locations far from the emissions sources.    Also> in our  simulations
the Class  I  increments were exceeded  by the 1986 energy developments  only
near  plant  stacks; Class II increments were never violated.

-------
                160
           APPENDIX A
AN ANALYSIS OF NUMERICAL METHODS

-------
                                  161
                             APPENDIX A
                AN ANALYSIS OF  NUMERICAL METHODS


     One of the major decisions in the development of the long-range
dispersion model is the selection of a suitable numerical  method  for
solving the model equations described in  Chapter VI.   Therefore,  at the
outset of this project, an effort was made to carry out  a comparative
study of different numerical  methods with  respect to  accuracy  and  effi-
ciency.  Three methods were examined:

     >  Upstream differencing
     >  The SHASTA (Sharp and Smooth Transport Algorithm)  method
     >  The Egan-Mahoney method.

As discussed in Section 1, the SHASTA method appeared to be  the best for
the present application and was thus chosen.   A detailed analysis  of this
method can be found in Section 2 of this  appendix.

1.   COMPARISON OF THREE NUMERICAL METHODS

     Mesoscale atmospheric transport is dominated by  advection, so in the
horizontal  direction the numerical method  selected  for the present pro-
ject must be able to treat the pure advection case  without generating
excessive numerical  diffusion.  As a test  we compared three  numerical
methods for the solution of a two-dimensional  advection  problem with a
constant wind on a 40 x 40 grid of 25-kilometer squares:

                          ct  + (uc)x + (vc)y =  0    .              (39)

The wind was a uniform 25 and 12.5 km/hr  in  the x and y  directions,
respectively.   A point source yielding a cell-averaged concentration

-------
                                  162
of 20 ug/m3 was located  on  the  upwind boundary.  The background concen
                  3
tration was 2 yg/m ;  physical diffusion was zero.



     The first method tested was a  fractional step upstream differenc-

ing method:




                   c* - (i - .     +     -       -     = vAt/Ax.  As shown in Figure 31, this
       x               y
method is highly inaccurate.  (Since physical diffusion is zero any

plume spread is due  to numerical diffusion.)  The lowest-order error

term in an individual  fractional step is
In the first test  case  [Figure  31 (a)], 0=1.  Therefore this term
                                        A

in the first fractional  step  is always zero.  In fact the transport


in the x-direction is indeed  exact, hence the plume appears to chop


off abruptly as  expected.   Figure  31 (b)  shows the same simulation with


o  = 1/2,  for which the  effective  numerical diffusion in the x-direction


is 4.3 x 104 m2/sec.
     Figure 32 shows the performance of a fractional step version of

the SHASTA method:
          c   =  cn   +  aD(1)cn  +  1 +2*   n(1)n(1>n
          Cij    Cij  +  VO  Cij +    +      D   D
                     — D^
                     8  u+

-------
DISPERSION CF -  5 I \GI_E  PLU-E
DISPERSION  OF  fl SINGLi
  B T
                     UNIT:
       (a)  oy = 1/2, a  = 1/4
            A       y
         (b)   a  =1,0=  1/2
              A      y
      FIGURE 31.   PREDICTED CONCENTRATION DISTRIBUTIONS USING THE UPSTREAM DIFFERENCE SCHEME

-------
DISPERSION  OF  ?  SINGLE  PLUMS
DISPERSION  OF fl  SINGLE PLUME
               	 -0-  10ft
                                                                        ppm
                                                                                         o
        (a)  ax = 1/4, ay = 1/8
         (b)  ay = 1/2, a  = 1/4
             A       y
            FIGURE 32.   PREDICTED CONCENTRATION DISTRIBUTIONS USING THE SHASTA METHOD

-------
          ?ij = C*.. + Oy[
                                                                   (42)
The computed plume profile is reasonably contained in a corridor with


a constant width of six cells and is relatively independent of a  and
                                                                A

o .   The lowest-order truncation errors in a single fractional step of


the SHASTA method are
                            3c       4c


                           3X       3X





 where



                                    oo    ,2

                                    t2u2 - ^~
 and
                           4322      4
                    k  =  U At    U AtAx    3AX
                           24        48     192At
 Numerical error is generated by dispersive errors from the c    term,
                                                             A A A

 and diffusive errors  from the cxxxx term, so we cannot characterize


 psuedo-diffusion by the simple coefficient in the c   term.  When o  = 1/2,
                                                    XX              X

 the c    term vanishes and the error becomes purely diffusive and thus easy
      A A A

 to analyze.




        Consider the two equations
                             - Vxx
   which have the solutions
                                   .  v

                                   lwx
                           c(t)  = e'WA e  u



                                           W4t
                           c(t)  = e'WA e

-------
                                  166
When a  = 1/2 the effective coefficient of psuedo-diffusion generated
      x               „
by SHASTA is thus -k0u> .   This is wavenumber-dependent; in our test
                                                 fi          1
case the shortest wave has a wavelength of 20u/10  (meters)   and is
affected by a numerical diffusion of 1.2 x 10  m /sec.  All other waves,
being longer, diffuse more slowly.  Table 21 compares the numerical
diffusion associated with upstream differencing and the SHASTA method
for various wavelengths in the first test problem.  Note that these
results are dependent on grid size, so decreasing the cell width will
decrease the numerical diffusion.  The computations for the SHASTA
method are based on the assumption that o  = 1/2.  Although Figures 31
                                         X
through  33 indicate that SHASTA is less sensitive to a  and a  than
the other methods, this is still an optimal  condition; the entries
in Table 21  are not worst case.  Similarly, the estimates for the
upstream differencing method are not upper bounds either;  as a
                                                              A
decreases both methods generate greater errors.
         TABLE 21.   EFFECTIVE DIFFUSION COEFFICIENTS IN THE
                     x-DIRECTION FOR THE FIRST TEST PROBLEM.
                     a  = 1/2, AX = 25 km.
                      A
         Wave
             Effective Diffusion Coefficient
                         (m2/sec)
Wavenumber
  (m-1)
 20^/106
                        5TT/10
Upstream
Differencing
4.3 x 104
4.3 x 104
4.3 x 104
SHASTA
1.2 x 104
3.0 x 103
7.5 x 102
     The third method tested in the present study was the two-dimen-
sional Egan-Mahoney method, which computes the pollutant concentration
and the first and second moments of that concentration in each grid
cell.   By calculating these subgrid-scale details the Egan-Mahoney

-------
DISPERSION  OF  fl SINGLE  PLUME
DISPERSION  OF  fl SINGLE  PL'
                   - - UNIT: - ppm
                    'MIT:  ppm
        (a)  ox =  1/2, ay = 1/4
          (b)  ox = 1,  oy = 1/2
       FIGURE 33.  PREDICTED  CONCENTRATION DISTRIBUTIONS USING THE EGAN AND MAHONEY METHOD

-------
                                   168
method achieves considerable accuracy.   Details of the method can be
found in Egan and Mahoney (1972a,b) and Pedersen and Prahm (1974).
Figures 33(a) and 33(b) show that in the Egan-Mahoney solution with
a  = 1/2 the plume corridor is 5 cells  wide; with ax = 1 the width is
just 3 cells.  However, as with upstream differencing, the x-direction
transport is essentially exact when a  = 1  so that the high quality
                                     A
solution shown in Figure 33(b) must be  interpreted with care.

     The authors are not aware of any estimates of the numerical  diffusion
associated with the Egan-Mahoney method.  Unlike the methods discussed
above, the numerical error in this method is dependent upon the con-
centration itself.  The method will follow a 10 pg/m  spike through a
zero background concentration without generating any numerical diffusion,
              o                                            3
but a 110 pg/m  spike cannot be followed through a 100 pg/m  background
concentration without considerable diffusive error.  In our application
background concentrations should be low, so that, as Figure 33 indicates,
the Egan-Mahoney method should be suitably accurate for the present
application.

     Table 22 shows the relative speed of each method.  The Egan-Mahoney
method produces a better solution than  SHASTA, but is an order of magni-
tude slower, and hence much more expensive.  In terms of overall  effi-
ciency, it appears that the SHASTA method possess the blend of speed
and accuracy most suited to our application.
            TABLE 22.   ESTIMATED COMPUTATION TIME REQUIRED
                        TO FOLLOW A PLUME FOR 750 km
Numerical Method
Upstream differencing
SHASTA
Egan and Mahoney
* Number of
a Steps
1
1/2
1
30
60
30
Required
Computing Time
0.187 sec
0.845 sec
11.0 sec
       * ax = uAt/Ax.  The a's associated with each method are optimal
         for that method.  AX is 25 km.
       t On a CDC 7600 computer.

-------
                                  169
2.   ANALYSIS OF THE SHASTA METHOD

     In Section 1 of this appendix we examined the psuedo-diffusion
associated with three numerical methods.  Based on considerations of
accuracy and computation time, we selected the SHASTA method for use
in our model.  Now we will focus on an analysis of the stability and
formal accuracy of SHASTA for the simplified constant wind case.

a.   Accuracy

     Under constant wind conditions the Step 1 difference scheme [see
Chapter VI, Eq.(lO)] may be written as the following one-step scheme:

           -£_   LL   ±\ rn     + t  4. J_   5e   3c2
           16 " 16 " 64/ci+2,j   \a   16 " 8     4
                 31 4. ,   9    lle               1    5e
                 32 + A - 2a --8
                      :2    n  n
                 ^16   16   647  i-2,j

where

          e = UAt/AX
                     2
          a = k At/AX
               X- ^ At     .
Substituting the true solution into Eq. (45), we obtain c..,  c
c"j.i -•' c" i -;> ar>d c" 9 n- by Taylor series expansion about c^..   {At
the moment we are considering this fractional  step  individually,  not
the scheme as a whole, so we assume c**. = c1-.[(n + l)At]}.   The result
may be simplified and expressed:

-------
                                  170
         3t    8X     X 9X2
                                                                   (46)

Evidently Step 1 (and Step 2) are accurate to first-order in time,
second-order in space.   But again, the problem is dominated by hori-
zontal advection; if our Step 1  equation is reduced to:

                             |°_+ u||= o                          (47)

Equation (46) becomes
                        ? \  1    / A.  1    9    7      d \  4
              ..22   AX \ 9 C ^ (U At    U AtAX    3AX   j 3 C
                    "^"I    ~24 --- 48—
                                                                   (48)
This is second-order in both space and time;  in the special  case where
e = 1/2 it is third-order.   Thus, the important advection terms in
Eq. (1) of Chapter VI should be handled with  acceptable accuracy.

     The entire three-step method, being a fractional- step formulation,
is inherently only accurate to the first order in time.  Steps 1 and
2 are second-order in space; Step 3 has no spatial  discretization errors,
so the overall three-step spatial accuracy is second-order.

b.   Computational Stability

     Assume that the solution to Eq.  (39)  may be expanded in a Fourier
series and that a separation of time and space variables is possible.
A typical Fourier component may be written
                                         ,                         (49)

where w is called the wavenumber.   We define the amplification factor as

-------
                                  171
                        r = <|;(t + At)Mt)     .                   (50)

When A - 0 the solutions to fractional steps  1 and  2  should  be  nonincreas
ing; the stability requirement is thus |r|  <_  1 for  all wavenumbers.
Substituting Eq. (49) into Eq. (45), we find

                     2     \
                   ^o~ + W ) cos 2wAX -- jp- Sin
                    o   oc. i             4
(if -
\6d.
                   2a +   +   H COS u,AX +     - 2a  -          -   (51 )
Figure  34 shows  |r| as a function of uAx for e = 0.6 and various
values of a.  We are assured of stability whenever e <. 0.6 and a <_ 0.15.
The condition on a may be relaxed by tightening that on e, but there
is no advantage to relaxing this condition.  Advection dominates dif-
fusion in the horizontal so that the most restrictive time-step con-
straint derives from the requirement that e <. 0.6.

     The chemistry and removal term, A, can also affect the performance
of the method.  Formal stability will not be lost by adding this undif-
ferentiated term (the Strang Perturbation Theorem), but more is required.
When XAt < 0 the solution decays in time, and we would like to ensure
that our numerical method has the same  property.  The addition of the
chemistry term to Eq. (51) adds AAt to  the real part of r.  This real
part is at a minimum when kAx = -n .  From Eq. (51) we calculate that
|r| < 1 requires

                       -AAt < £ - 3e2 - 4a    .                  (52)

In the most restrictive case, e = 0.6 and a = 0.15, we have -AAt < 0.07.
This is easily satisfied in our model since surface deposition and
chemical reactions occur at relatively  slow rates when compared with
atmospheric transport across a 10 km grid cell.

-------
                          172
                                                      = 0.05
    3TT/2
3ir/2
           a = 0.10
       a = 0.15
    3TT/2
3TT/2
FIGURE 34.   VARIATION OF AMPLIFICATION FACTOR  r| AS
             A FUNCTION OF  a  FOR e  = 0.6

-------
                                   173
     In conclusion, it was shown that the SHASTA method  as  modified
in the present study is stable and accurate,  and is  capable of producing
acceptable results at reasonable cost.

-------
                 174
            APPENDIX B
COMPILATION OF SIMULATION RESULTS

-------
                                  175
                             APPENDIX  B
                COMPILATION OF SIMULATION  RESULTS
     The  long-range air pollution model developed in this project  was  used
in six simulations:

     1.   27-31  January 1976 meteorology, 1976 emissions;  pp.  176-195
     2.   27-31  January 1976 meteorology, 1986 emissions;  pp.  196-215
     3.   4-7  April  1976 meteorology, 1976 emissions; pp.  
-------
                                 176
1.   27-31  JANUARY  1976 METEOROLOGY: 1976 EMISSIONS

-------
                  3*     U*     S«     6«     71     a»
                  I 11  I I I I	
                            nONTHNP   N DPKOTO
                           ••a
                            HTOniNB
                                     3 OOKOTO
                                      NEBRQ3KQ
                                      COLORBOO
      I*
                                I  I  I  I I

                   38     Y •     54     51     73     84     9'     1««
                                                                                                   3*
                                                                                                                5«
                                                                                                             nONTQNO   N
                                                                                                                             '6     8«     9*
  5.        4.

*-^.'-,*.     ''.
                                                                                                            X  .2
                                                                                                                      3 OflHOTB
                                                                                                                      NEBRflSKB
                                                                                                                      COLOROCO


                                                                                                                   S- 2
                                                                                      1*     2*     3(     U*
                                                                                                                      6«
                                                                                                                                          9«
S02   CONCENTRflTIONS  IN UG/M3 FOR  THE  HOUR  500-800  MST ON 7bOil^       S02  CONCENTROT 1 ONS IN  UG/M3  FOR THE HOUR
                                                                                                                                            Mbi

-------
                                                                                                                             100    110
           2«    30    44     5»    60    78     3D
S02  CONCENTRfiTIONS  IN  UG/M3  FOR THE HOUR 1100-1400 MST  CM  ~SJl!"    S02  CONCENTPflTIONS IN UG/M3 FOR THE HOUR  1^00-1700  MST  ON 73012;

-------
      1*    2«    3*    »•     54     54     74    34    94    144    ;:J
                                                                              14     2*     3*
                                                                                                     S«    6»    7«
S02  CONCENTRATIONS  IN UG.M3 FOR  Th£  HOL'P 1700-2000 .1ST  ON  7^012'    S02   CONCEN 7RPT! ONS  IN  UG/M3  FOR THE HOUR  2000-2300  MST  01

-------
     10     20     30     U0     SO     60     70     30     30     100    110
                                                                                    10     20     30    •"»     50     60     70
                                                                                                          rtONTQNQ   N DDKOTR
                                                                                                                                      90     M0
                                                                                 9
                                                                                 I*-.
                                                                                                   5<.
                                                                                                          MtbfllNS
                                                                                                                  3 DfiKOTH
                                                                                                          4«--,
                                                                                                                   COUORBOO
      10     20     30    U9     50     60     70
                                                        90     100     110
                                                                                                                                                           00
                                                                                                                                                     r
                                                                                    10     20    30     U0     S0     60     70     30     90     100
S02   CONCENTRPT10NS  IN UG/M3 FOR  THE  HOUR -100-200 MST ON 7E012S     S02   CONCENTRflTICNS IN  'JG/M3 FOR THE  HOUR 200-=30  MST ON  78;:I5

-------
       If    2*     3(     J«
                                       6*
a-
                            ' ' 2
                                      3 OOKOTB
                                      NEBRP3KQ
                                      COLQPODO
                                      ,2
                                                   8«     9*
                                                                                                  3»
                                                                                                               5*     6«
                                                                                                            HONTONR    N OBKOTQ
                                                                                                                                  3*     3*
                                                                                        2
                                                                                                            HTtinlNS
                                                                                                                     3 DflKOTO
                                                                                                                     NEBRA3KB
                                                                                                     3
COLOROOO
•2
        l«     H     3»     4«     Sf     61
                                                          9«
                                                                                                  3*
  S02   CONCENTRATIONS  IN UG/M3 FOR  THE HOUR  5^0-800 flCT  ON rj_::c       S02   CONCENTRATIONS IN  UG/M3  FOR THE  HOUR S00-J190  MiT ON  7.:0123

-------
                   38
                                                                                                        «»     s*
                                                                                                                           7«     80
                            HTOHIMS
        2 '
                                     M DfiKOTB
                                     3 DflKDTS
                                     NEBRR3KA
COLDRflDO


 2
                                                         .1
            2»     3«
                               S2    68     79     8«     98     100     1 ; -1
                                                                                                           MONTBNR   N OflKOTB
                                                                                           6.
                                                                                              'X.
                                                                                                           HYOHIM8
                                                                                            10
                                                                                                                    3 OflKOTB
                                                                                                                     NE8RB3KQ
                                                                                                             7    '-.  (JOLOROOO

                                                                                                                   '
                                                                                                                               14
                                                                                                                                                             00
                                                                                                                                                             INJ
                                                                                     It     28     3»
S02   CONCENTRPT10NS IN  UG/f3  FOR THE  HOUR  1100-1100 MST ON  ~£'21!£     302   CONCENTRfiTIONS IN  UG/M3 FOR THE  HOUR  1^00-1700 MST  ON  7G012S

-------
      If     21     3*     »«    5«     68     78     e«     9»     14*
                                    N OBHOTQ
                                                 10
                                                                                                      u«     5J    5J     79     8«     9*     1'*    IIJ
                                    3 OfiKOTB
                 2
                                   • COIORHDO
      1*     2»     31
                               S«     6*
                                                                                                         nONTRNfl
                                                                                                       11
                                                                                                                    :flKOTB
                                                                                                                  3 DflKOTP
                                                                                                                             10
                                                                                                                                                         CO
                                                                                                                                                         00
                                                                                   1*     2»
                                                                                                            S«    5*     7«     8*     8*
S02   CGNCENTF^-: ;'NS  !N UG/M3 FOR  THE HOUR  1790-2000 MST  ON "oOlIS     302  CONCENTRRTIONS  IN UG/M3  FOR ~^E  HOUR 2000-2300  MST ON  r^:^

-------
s •
    10
                   3»     40     5 1)     6«     7»
                             HONTRNB
                                                        90     ic0
                       5_
                             HTYOHINB'-
                              2 ~—!
         '••... ""••C*
                       -2     6
                                     COLORflOB
                                                13
             2«     30    U0
                                      60     70
                                                        30     130    11.1
                                                                                                30     U0     30     60     70    30     90     100
 S02   CONCENTRRTIONS  IN UG/M3 FOR THE  HOUR  -100-200 MST  CN 7'0123     S02   CONCENTROTIONS IN  'JG/H3 FOR  THE HOUR 200-500 NST ON  7eOiSS

-------
                               S«    Sf     7»
                            HONTRNH   N DPKOTH
                                                                     IIJ
                                                                                                              5«     6*
                                                                                                                                       9*
   10
                                     3 ORKOTR
                              6      COLGRRDO
                                                10




                                                7-
                   3«
                                            7*     8»     9«
                                                                                                           MONTflNR
                                                                                                           HTOIflNS
                                                                                                                     5HKOTH
                                                                                                                    3 DHKOTfl
                                                                                                                    NEBRB3KO
                                                                                                                                           3 .T
                                                                                                                    COLORROO
                                                                                     it     2*
                                                                                                              s«     6*
SOI   CONCENTROTIONS  IN  UG/M3  rOR  'HE HOUR  500-300  MST ON  7t0120      S02   CONCENTRATIONS IN  UG/M3 F0=? "HE  HOUR  800-!!^^ M5T  ij

-------
10     20     30     U0    50    80    70
                     nONTRNfl 1  N DflKOTfl
                                                 80     100
                              NEBRfl8KB
                                         1O
                                         «S^
                                         ,-•=« r-^
         2 -"-._
                              COLORflOO
!•     2«     319

                        SJ    6U    70     80     9>
                                                                                                          5»     70     30    30     100    110
S02  CONCENTRfiTIONS IN UG/M3 FOR  THE HOUR
                                                    M3T ON  7;f31Ia    S02   CONCENTRflTIONS  IN  UG,^3 r:?  THE HOUR  1^00-:700 MST  ON  7601IS.

-------
                            5*     6»
                                                     9*     !«•    MO
3k--.
                                  N DBKOTfl
                                  3 DOKOTO
                    .2
   1*    2«     3*
                            54    6»     It     »»
                                                                                      2«     3*    4*     5»     6«     7*     B*     9«     I'*
                                                           IM    Hi)

                                                                                                       •\
                                                                                                              3 DQKOTQ
                                                                                                              NEBRB3KB
                                                                                                       6     - COtOSPDO
                                                                                                         s*
S02   CONCENTRPTIONS  IN UG.M3  FOR THE  HOUR  :
                                                        MST  ON "BO 129     3d   v.oNCENTRPT I ONS IN  UG/M3 FOR  THE HOUR  2000-2;03 MiT  ON

-------
10    20    30     40    50     60
302  CGNCENTRflTIONS IN UG/M3 FOR THE HOUR  -100-200  MST  ON
                                                                302  CONCENTRATIONS  IN UG/M3 FOR THE HOUR  200-500  MST ON 760130

-------
      1*
                              Si     60    70     et     it
                           HOHTONB    N DflKOTa
                                                                                                                                                11J
                           NYWUN6
                                    3 OBKOTP
                                      .J.
                                    NEBRflSKQ
                                    COLOMOO'-,
                              s«
                                                 a»
                                                                                  1)    2*     3*     It)     Si     S)     7)    e»    9)
S02   CONCENT = C-:;NS  ;N  UG/M3 FOR THE  HOUR 500-»0o ^s*  :n  7bi?: .-•.«       502  CONCENTRATIONS  IN  UG/MS FOR  THE HOUR  80^-1100  n;~  :N "e

-------
      10    20    30    110    so
                                                                                                                  79    Si)    90    100
     10    20     38     1*0    50     50     7»
                                                     99     10«    110
                                                                               10    20    30    V0     50     60     70     80     90     100    !:.'
302  CONCENTRflTIONS  IN  UG/H3 FOR  THE HOUR  1100-14-00 MST  ON 760130    S02   CONCEMTRflTlONS  IN UG/M3  FOR  THE HOUR  1400-1700  MST ON 7=^1:?

-------
       1*
                    3*
s •
                                 S«
                                               7«     9»
                              NONTBNA   N DAKOTA
                        6
                        *L
                              .
                         V   )\
             13
                                                   VT -
                                        3 DflKOTB  ' '
                                                                         111
                                                                                                                                                      l«»
                                        HEBRQ3KB
                                        COLORADO
                    3*
                                                                                       14
                                                                                                                 10NTPNQ   N DAKOTA
                                                                                                 n

                                                                                                /I
                                                                                                 I       y
                                                                                                                                      13
                                                                                                                          3 DAKOTA
                                                                                                                           NEBRASKA
                                                                                                                           COLORADO
                                 5*     5<
                                                                                                 It
                                                                                                                           61     73
  SOS   CONCENTRATIONS IN  UG/'M3  -OR  THE  HOUR  1700-it\V' "$*  C.'i  7t-;iDii.i       SC2   CONCENTRMT: ONS  IN .'0. M3 FOR  THE  HOUR  1000-i30iJ  1-T  Ofl

-------
     10     20     38     140     50     68     70     80     90     180
     II    20    30     U0     50     60     70     60     90     100    110
                                                                                l«     20     30     110     50     60     70     60     90     100    11?
S02  CONCENTRPT10NS  IN UG/M3 FOR  THE HOUR  -100-200 MST  ON 7S0I31     302   CONCENTRPTIONS IN UG/M3 FOR THE  HOUR 200-500  MST ON  7&2\-..\

-------
            it     3«
                                      6»     7«
                                                          3»      \»e
                                                                                       !•
                             nONTBNB
                             HroniNe
    - 2
 ii   ill
                                      3 ORKOTQ
                                      NE6RH3RH
                              7    •
                                      COLORODO
                                                  a.1
                                                          1 ,
                   3»     UK     S«      6*     71
                                                                                                           4*     S«     6*      7(     8*     8*
                                                                                                              HONTBNR    N DOKOTB
                                                                                                                                   10
                                                                                                             v2
                                                                                                              HTONIN8
                                                                                                                       3 DBKOTfl
6'      ' COLORRDO

 '•':.-.      ''.2
                                                                                       1*     2*     3*
                                                                                                                 S«     6>     7*     80      8*
SOS   CONCENTRPT10NS  IN UG/M3  FOR  THE HOUP  S^U-33(? HST  ON  760 if,]        S02   CONCENTRfiTl ONS IN  UG/M3  FOR  THE HOUR 8^0- 1J <>i> MST ON

-------
              48     56     64     73     98     98     130    110
  A
                  lOHTQNfl    N OOKOTB
3V , •'-.,
                  HTOHING
          2V.
               I-
                           3 DflKOTB
                           NEBRRSKR
                           COLORflDO
                                                                           18
                                                                        10
                                                                                                 HTtmlNB
                                                                                          \ 2
                                                                                            3,
                                                                                                          3 ORKOTfl
                                                                                                          NE8RB3KH
                                                                                                        '.. 'COLORflDO
 29     30     U8     58     b»
                                                                           1»     28    38     U8     58     68     78     88     98     18«
S02   CONCENTRfiTIONS  IN UG/M3  rOR  THE HOUR 1100-1400  M3T ON
                                                                     S02   CCNCENTRflTIONS IN  UG/M3  FOR  THE HOUR 11*00-1730  ri£T  ON 7SOJ3J

-------
 1*     2»
I I  I I I I  I i
             3*     49
                               54     S«     7»      34
                               i I	Ill	1 I I I  I I I  I
                                                           lit    11.'
                       nONTBNO   N OflKOTO
                       MTOHINB
                                3 OHKOTO
 If     2«     3*     U*     54
                                       It     84
S02   CONCENTRPT10NS  IN  UG/M3  FOR  THE HOUR
                                                        MST OH  7C?:"1

-------
                                  196
I.  27-31 JANUARY 1976 METEOROLOGY; 1986 EMISSIONS

-------
            2*     3*
                              5*    6«     7«     at     9«
                           RONTRHR   N DRKOTR
                           MTOniNB
                       .JO

                                    9 DRKOTR
                                    NEBRQ8KP
                 19
COLORPOO
 P

      1*     2*     3(     <(«    5«     6(     7*     9*     8«     1«*
                                                                                                           S«     6»    7»     8«     9«
S02   CONCENTROTIONS  IN UG/M3  FOR THE  HOUR S00-800 MST  ON 76012',        S02  CONCENTROT10NS  IN UG/M3  FOR THE  HOUR 800-1100 Mt>T  ON "C3127

-------
     1«
                                                                            1*    2(    3«     U«    5«    6«     7»    8*    8*     lit
                                                                                             u*     se    s*    ?»    e*     at
502  CONCENTRfiTIONS IN UG/M3 FOR THE  HOUR  1100-1400 MST ON 76012"    SQ2  CONCENTRflTIONS  IN UG/M3 FOR THE HOUR  1400-1700 MST ON 7e

-------
                                                                                 2«
                                                                                            U*     S<    6*    7*     8«    9*    !»•    110
S02  CONCENTRBTIONS  IN UG/M3 FOR THE HOUR 1700-2000 MST ON  760127   S02   CONCENTRPT10NS IN UG/M3 FOR THE HOUR 2000-2300  MST  ON 7E0127

-------
                                                                                                          9*.
                                                                                   -, ,  , , ,\V-*?\	, ,••, V. ..1«... ^
     1*    2*     3*     4*     S«    6«    7*    8*    9*
                                                                                    3*     3*
                                                                                                     S*    6*    7*     8«     9B     1M    114
S02  CONCENTRPTIONS IN UG/M3  FOR THE HOUR  -100-200 MST ON  76012S     S02  CONCENTROTIONS IN UG/M3  FiOR THE HOUR  200-500 MfT ON  76012S

-------
     1*    2*    3f     »«•     5«     S*     7»     8«     9*     !••
     1*    I*    3*    1*    5«    S«     7»     8*     9*     14*
                                                                                                      S«    6*    7*    8*    9*     !•«    1:»
SOS   CONCENTRRTIONS  IN UG/M3 FOR  THE HOUR 500-800  MST ON 760113       S02  CONCENTRATIONS IN UG/M3  FOR THE HOUR  800-J100 MST  ON •'6011?

-------
                                                                                       3<     U«    S*    6«     7«    90     9*     !•»
                                                                                \    \*T\V..7      \ \   \ JJ ':
                                                                                 \ . . \K-:;>'f	A,\ , .'&.~i. . .
S02  CONCENTRflTIUNS  IN UG/M3 FOR THE HOUR 1100-1400  MST ON 760128    S02  CONCENTRfiTIONS IN UG/M3 FOR THE HOUR  1400-1700 MST ON 7C0128

-------
                                                                                                                                1M    II*
     1*    2*     3»
                                                                                   2«    3*    «•     S«     6*    7«    8*     9«     !••    Ill)
S02  CONCENTRRTIONS IN UG,M3 FOR  THE  HOUR 1700-2000  MST ON 76^123    S02  CONCENTRATIONS  IN UG/M3 FOR  THE  HOUR  2000-2300 MST ON 7C012S

-------
      1*     20     30     10     50     60     79
                                                      90     190    110
                                                                                 It    20    3*    U0    50    60    70    9*    9*     100
      10    20     30     V0     50     60     70     30     90     100    113
                                                                                 10     20     30     U0     50     60     70     80     90     100    110
302   CONCENTRATIONS IN UG/M3 FOR THE  HOUR -100-200 MST ON  763J29      S02  CONCENTROTIONS  IN  UG/M3 FOR  THE HOUR  290-500  MST ON 760129

-------
     1*
     l«
                3»    »«     Si    g«    7*     8*    9*    !••
S02  CONCENTRflTIONS  IN  UG/M3 FOR THE HOUR 500-800  MST  ON  "fOJJ9       S02  CONCENTRfiTIONS  IN  UG/M3  FOR  THE HOUP 300-J100 MST  ON

-------
                                                                                                 54    S«    7»     8*    98    1»8
302  CONCENTRfiTIONS IN UG/M3  FOR  THE  HOUR  1100-1400  MST  ON 760129    S02  CONCENTRPTIONS IN UG/M3 FOR THE HOUR  1400-1700  MST  ON  7S3129

-------
     !•     2«    3*    It*     S«    6»    71     8*     8*
                                                                                                   S«    S«     7«     8»    8*    !*•    119
302  CONCENTROTIONS IN UG/M3 FOR  THE  HOUR 1700-2000 MST ON  760119    S02   CONCENTRPT1ONS IN UG/M3  FOR  THE HOUR 2000-2300  MST ON 760120

-------
S02  CONCENTRHT10NS IN UG/M3 FOR THE HOUR -100-200 M3T  ON  7bijl30      302   CONCENTRfiTIONS IN UG/H3 FOR THE HOUR 200-500 (1ST ON 760130

-------
                                  6«    7<     3*    8*    !••    lltf
     1*     2*     3«    U»
S02  CONCENTRPTIONS IN UG/M3 FOR THE HOUR 500-800 MST ON  780130
S02  CONCENTROTIONS IN OG/M3  FOR  THE HOUR 800-1100 MST ON 763130

-------
                             50    60    70    89    9»    189
                                                                                10     20     30     U0     50     60     70     80     90
                                                                                                     flOHIQNB
 8^
.$i
-,--
                                                                                                     «>QflIN8
                                                                                                             N OBKOTB

                                                                                                             3 OPKOTP
                                       *X\V\ \ \\
                                          \\\  \\\

                                           \\\\\
                                                                                                                                                    no
                                                                                                                                                    o
                                                                                U
                                                                                                              6*     71     80    99    10«    110
S02  CONCENTRRTIONS  IN UG/M3  FOR THE HOUR  1100-1400  MST ON 7t31;0     S02   CONCENTRATIONS IN UG/M3  FOR THE HOUR  1400-1700 MST ON  7601S9

-------
                                                                                                           6*    7*     8*     g*     !*•    1U
                                                                                                                                                \3
                 3»
                             S*    6*     '    8»     9»
                                                                              1*    2*    3*     t«     S*    6*    7*    8«
                                                                                                                                   !*•    HO
S02  CONCENTRPT10NS IN UG/M3  FOR THE HOUR  1700-2000 MST ON  7601:0    S02   CONCEN FRHTIONS  IN  UG/M3 FOR THE HOUR  2000-2300 MST  ON ~:-^

-------
                                    6«7»3»at     \tt    lie
      1*    2«    3»    It*     5«     6*     7*
                                                                             TTTT"H i i  i I i i  i i I rr-rr+T~rT*rfr*
                                                                                                      HONTRNR   N DRKOTQ
                                                                                                                                  e«
                                                                                                                                 i-H

                                                                                                                               r ''•'•-. \ \\ '.\  /   '"•'
                                                                                                                                \\N\ '•- '-. - ""V
                                                                                                                                   •-. '--
S02   CONCENTROTIONS IN UG/M3 FOR  THE HOUR  -100-200 MST  ON 760131      S02   CONCENTRfiTlONS IN  UG/M3 FOR  THE HOUR  200-500  MST ON  763131

-------
      I*
                                                                                                         5»     6«     7«     8»     9»     1M    114
                             S«     6«     7«     8*     9*     !«•
      1*     2*     3*     ««     51     6»     7*     8*     9«     !••    110
                                                                                 !•     2«     3*     14*     5»     6»     71     81     9«     !••    Ill
302   CONCENTRPT10NS IN OG/H3 FOR THE  HOUR 500^-800 MST ON  760131
S02  CONCENTRPT10NS  IN UG/M3- FOR THE HOUR  800-1100  MST ON 7601:1

-------
                                                                                                                           100    113
302  CONCENTRQTIGNS IN UG/M3 FOR THE HOUR 1100-1400 MST ON 7S01SJ    S02  CONCENTRRTIONS IN UG/M3 FOR THE HOUR 1400-1700 MST  ON  760131

-------
           2«     3*     Ht     St     S«    7f    3*    9*
                                                                 lit
     !•    2*    3»     U«    5*    6(    7*    3*    8*    !••    11*
S02  CONCENTRPT10NS  IN UG/M3  FOR THE  HOUR 1700-2000  MST ON 7601?!

-------
                                  216
3.   4-7 APRIL 1976 METEOROLOGY;  1976  EMISSIONS

-------
                                                             1«*    11*
      1*     2*     3*    4«     5*     6*     7*     6>    9(     !•»
                                                                                               3«    U«     5«     6«     7«     B«     a«	IM




                                                                                                         nONTBNQ    N DHKOTfl
                                                                                                         NToniNe
                                                                                                      20_
                                                                                       15
                                                                                                                  3 DflKOTQ
                                                                                                                  NEBRB3KO
                                                                                                                  COLORBDO
                                                                                   1*     2*     3*    U(     S*     6«     7*    8*     9(
S02   CONCENTRPTIONS IN UG/M3 FOR  THE HOUR  500-300  MST ON  76040^       S02  CONCENTRPT]ONS  IN UG/M3 FOR  THE HOUR  800-1100 MST  CN  ~£

-------
     1*     20     30     U0    50    S«    79     80     90     100
                                                                             If     20     30    l»0    50     50     70     30    90     100    11J
S02  CCNCENTRPTIONS IN UG/M3  FOR THE HOUR  1100-1^00 MST ON 760401+   S02   CONCENTRPTIONS  IN UG/M3 ~3R THE  HOUR  1400-1700  MST ON 7

-------
     !•    2«    It     U«     5«     6»    7*    3«    9*     !••
                                                                             1*
                                                                            T-H
                                                                                         3»
                                                                                                    5*   . 5J
                                                                                                                      3*     9«     [«•   !!•
S02  CONCENTRfi::JNS IN UG/M3  FOR THE HOUR  1700-2000 MST ON 760U0U    302  CONCENTRPTIONS IN UG/M3  FOR  THE  -CU*? 2000-2300  MST ON 760uO

-------
302  CONCENTRflTIONS IN UG/M3 FOR  THE  HOUR  -123-200  MST  ON  760405      S02   CONCENTRATIONS IN UG/M3 FOR THE HOUR 200-500 MST  ON  76.:-.Ji

-------
      10
                                                                                    10     it     34     U0     50    60     7»     80    90     100    11J



                                                                                                         nONTRNR   N  DPKOIO
                                                                                                        '.' '.2
                                                                                                                  3 OflKOTfl
                                                                                                                  COLORflOO
                                                                                                                                                         ro
                                                                                                                                                         ro
                                                                                          20     3»     U4     50    60     74     34     90     104
S02   CONCEUTRfiTIONS  IN UG/M3 FOR  THE HOUR  £k.'0-300  MST ON 7i.C'*0£       S02   CONCENT-": :NS  IN  UG/M3  FOR THE  HOUR  800-1100  MST ON  ~:'J^O£

-------
                                                                                 1*    20     36
                                                                                                                                 90     '.81)
                                                                                                                                 .2
                                                50     98     180    1!'
                                                                                 1»     28     38     48     58    68
                                                                                                                           88    90     '.?»    11J
302  CONCENT^aTTViS  IN UG/MS  FOR THE  HOUR 1200-HI00 MST  ON  7i?01 ^t    302   CONCENTROTIONS  IN UG/T13  FOR *^E -".'.m  U00-1700 MST  -.s

-------
      14     24     24     '4*
                            PIONTflNfl
                            NYORIN8
        2 .._
                                      6*     74     94     94     !«4
                                     N 09KOTQ
                                     9 DBKOTO
                                     NEBRPIKQ
                               8
                                     COLORADO
                                                                                             24     34     44     54     64
                                                                                                             HONTQNP
                                                                                                                                    34     9*
                                                                                                      -~ 5
                                                                                                             HTOrtlNB
                                                                                                                      3 OflKOIQ
                                                                                                                      NEBRR3K3
                                                                                                                      COLORUPO
                         41     S<     6*      7«     30     9*     !»•    M-'
                                                                                                   3»
                                                                                                                5*     6*
                                                                                                                                   a*     8*      14*
S02   CONCENTPOTIONS  IN UG/M? FuR  THE  HOUR  1700-20^0 M£l  MN ~t040S     S02   CONCENTRATIONS IN  UG/M3  FOR  THE HOUR  2000-2300  MST ON "^ :''.';

-------
      1*     2U     38     43    5J
                                                                               18     28    3(S    U8    5i5    68     70     88     9<
                                                                               11     £4     38     18     59     68    78    38    98    188
SOS  CONCENTRPTIONS  IN JG-"r TOR  THE  HOUR -130-200  MST ON 7£.;'U06      S02  CONCENTRflTlONS IN  UG/M3 FOR  THE TiOUR 290-500  MST  ON 7604Q6

-------
                 u«    5»    6»    7«     98
It     2t    3*
S02  CONCENTRfiTIONS IN U&/M3 FOR  THE  HOUR 500-300 MST  ON  7 ."
                                                                   S02   CONCENTRfiT I ON3  IN  L!G-'M3 FOR THE HOUR  800-1130 MST ON ~ •-. ,'U ?.v

-------
S02  CONCENTRATIONS  IN  UG/M3 FOR THE HOUR  1100-14-00 MST ON 7tO^Ci?    S02  CONCENTRRFIONS  IN  UG/M3 FOR THE ,-iO..R  :-^0-1730 MST ON 7bi'4a!£

-------
     1*     a*     3*
                              5*
                                          ?•    89     9«
                                                                                  1*
                                                                                                          5»     6»     7»
                           MTOniNS
                               1	' CILCJiOO
            it     3«
                                                                                                        tlONTQNR '
                                                                                                                N &PKOTO
                                                                                                                3 OflKOTB
                                                                                                           6    COLORMiO
                                                                                                           s»     6*
                                                                                                                             e»    at     \n    .:-'
302   CONCENTRflTIONS  IN UG'M3  FOR *i£ HOUR  1700-2000 MST  ON r-^'UgO     S02  CONCENTRPTIONS IN  UG/M3 FOR  THE HOUR  2000-2300 MiT  ;N

-------
      10     a«     30     '49     50     60     70     80     94
                                       I. I I I I  I I ll. I I  I I    II
                             MONTHNP    N  O.fiKOTB    ;''j
             7     4
                             UTOH1NG
                        _2
                      14.'"
   *-,
                                      NEBRH3KB
                                           :  2
                                      COLORWO
                                                                                             20     30     44     54     64     70     60     90
                                                                                                                       N DUKO.T-fl
                                                                                                                           '•-2
                                                                                                              HT01IN6
                                                                                                                      -8-OflKOTP
                                                                                                                            X.
                                                                                                              6 — __  .
                                                                                                                       COLORflOO
            59
                         U3     Si3     60     71      83
                                                                                       It     29     3«
                                                                                                                 5J     60     70
                                                                                                                                           90     100
S02   CONCENTRfiTICtIS  IN UG/M3 FOP THE  HOUR -100-2C0  ,^SI ON  7b0U07      S02   CONCENTRflTIONS IN  UG/M3  FOR  THE HOUR 200-500  MST  ON 7

-------
^. i

2
                      MTcniHe
                                     60     7«     80    8«
                                     —t-r^	1  ! i i ,  i I i i  i i I  i

                                     *. OflKOTfl

                                     ..    2
                               NEBRBSKfl
                              •COL9RRDO
                         5«     6«     7»
                                                                              It     2*     3*
                                                                                                       S«
                                                                                                              6«     74     3»     -*     !•«     1IJ
                                                                                                  I,'
                                                                               1*
                                                                                                 13
                                                                                                 '
                                                                                                                ,2
                                                                                                                          ^ 7
                                                                                                      7^     ;CDLORflDO


                                                                                                      ~'--i5—    1
                                                                                           3«     U*
                                                                                                              6«     7*
S02   CONCENTRPTIONS  IN  UG/M3  FOR THE  HOUR 500-800  ,1ST ON  ^
                                                                         S02   CONCENTRfiT I ON£  IN UG/M3 FOR  THE  HOUR 800-1'. JC ^ST  ON "C 'HO

-------
      ID
     r-H
            213     33     a?     54
 • 6
                                   3 DftWJTB
             2             12      COLORflDO
                  -7       /?-•-.
                                   sa    7a    30    919     i •>.«    i
302  CONCENTRPTIONS IN UG/M3  FOR THE HOUR  1100-1H00 MST ON  ri04-;7    S02  COMCENTRPTIONS IN UO/M3  FOR THE  HO'..R :-j;'-1700 ^57 ON  "CO'-O^

-------
   1*     2»
7
         14
        V  1
                3*

               T-K
                                          It     33
                1
                          nONIPNC.
                          HTOBIN6
                                          1
                                   3 OflHOTO
                                   NEBRPSKP
                                   COLORflOO
                                                                                                                ro
                                                                                                                CO
S02   CONCENTROT10NO  IN  UG • M3  FOR  THE
                                               1700-290i5  M£" ON

-------
                                   232
4.  4-7 APRIL 1976 METEOROLOGY;  1986 EMISSIONS

-------
       1*
s
   .21
                                  S»     6*     71     8»     9*
                               nONTBNfl    N OBKOTfl
                                                                                         1*
                                                                                                      3*     4»     S«     6*     7»     8«
                                                                                                                                                    1M
MTOHIN8
 11
                            12
                                                ^  _-;:•*,.:'
                                                  «K5i
                                        3 DflKOTn
                                        HEBR03KO
                                                    37
                   34
         COLDBQOO
         21
        19     2*     3*      U«     5*     6»     7*     e«     3«      !«•
                                                                                                                flONTBHfl
                                                                                                                MTOHIN8
                                                                                                               10 <
                                                                                                                         N OfllCOTB
                                                                                                                         3 DAKOTA
                                                                                                                         NEBRB3KB
                                                                                                                                      119
COLORADO
AS
                                                                                         1*     2*     3*     U«     S«     S«     78     8«     9*     '.«*     1 • •'
  S02   CONCENTRPTIONS  IN  UG/M3  FOR  THE HOUR 500-8i30 MST ON  7:'"^:',       302   CONCENTRfiTIONS IN  UG/M3  FOR  THE HOUR 800-1100 MST ON  7;:ueu

-------
     l»    2«     3*     49     50     69    78    80    89    194    !11
                                                                              1»     20     39    149    S«
                 3*     U«     5»     69    79    30
                                                                                                                                   iae    no
S02  CONCENTRflTIONS IN UG/M3 FOR THE HOUR 1100-1400 MST  ON 7cO^>^    302  CONCENTRRTIONS IN  UG/M3 FOR THE HOUR  1400-1700  (1ST ON 76Q4C'Li

-------
     1*     2*     3*
     1*    2»    3*     U(     54     6*
                                              e*    9*     i««
                                                                               It     2*     3«    U«    S*     S8     7*     6»     9«     !«•    MO
S02  CONCENTRRT10NS  IN  UG/M3 FOR  THE  HOUR 1700-2000 MST  ON ~ > :'-' HI     S02  CONCENTRBTIONS IN UG/M3 FOR  THE HOUR 2000-2300 MST ON  ?

-------
     1*     20     30    U0    50     60     70    80    90     1C0
                                                                                                   50    S0     7«     80    80    100
302  CONCENTRPTIONS IN UG/M3 FOR THE  HOUR -100-200 MST ON  760405     S02   CONCENTRfiTIONS  IN  UG/M3  FOR THE HOUR 200-500  MST ON

-------
     K    2*    3»    U«    5*    6«     74     8*    9*    1M
                                                                             1*    2«     3«
                                                                                                    5«    6*    7*    3*    a*     !*•
           2*     3«     II*     6«    6*    7«
302  CONCENTRPTIONS  IN  UG/M3 FOR THE HOUR 500-800 MST ON 7C0405       S02   CONCENTRPTIONS  IN  UG/M3 FOR THE HOUR  800-1100  MST  ON 7a:~:5

-------
      1*    2*    38    lie    59    68    7*    88    9»
                                                                                                       58     6«     7»     8»	9»
                                                                                                                                           110
                             S«     61     70     89     99     114    !1.1
     1*    it    3*
                                                                                                 "»»    5»    68    7«    88    9«    1198    11D
802  CONCENTRRT10NS  IN UG/M3 FOR  THE HOUR  1100-1400 MST  ON  783^35    302   CONCENTROT10NS IN UG/M3  FOR THE HOUR 1^00-1700 MST ON  760405

-------
     !•     2*    3«    »•    S*    6*     7(     a*     8«     1*«
                                                                             1*    2*    3(     U*     5* •    6«    7*    B*    8«    !••
                             S*    8«    7*    8»     9«     IM
                                                                                                                            8*     !••
S02  CONCENTROT10NS IN UG/M3 FOR THE HOUR  1700-2000 MST ON  7e0405   S02   CONCENTRPTIONS  IN  UG/M3 FOR THE HOUR  2000-2300  MST  ON 7S04C5

-------
     10     20     30     H9     59    60    70    89    90    100    113
                                                                                10     20     39     U»     50     60     70     80     90     100
     1*    29    39    49    59    69    70    39    99    109    110
                                                                                10     20     30     U0     50     60     70     60     90    100    110
S02  CONCENTRPTIONS  IN UG/M3  FOR THE HOUR -100-200  MST ON 7S0406    S02  CONCENTRflTlONS  IN  UG/M3 FOR  THE HOUR  200-500  MST ON 763106

-------
                            5*     6*    71    a«     8»     1»«    110
                 3*     U*    S*    6«     7«     3*    8«
                                                                                                    S«     6*    7*    94    9*     UO
302  CONCENTRRTIONS IN UG/MS*FOR THE HOUR 500-800  MST  ON 760UJ£      S02  CONCENTRQTIONS  IN UG/M3 FOR THE  HOUR 800-1100 MST  ON 7£

-------
10     20     30    U0    =«     60     70    60
                                              at
                                                                                                                                  no
            39    U«    S»
                                                                        1*     20    30    40    50     60     70    30    B0    100
S02  CONCENTRflTIONS IN UG/M3  FOR  THE HOUR 1100-1400  MST ON
                                                                   S02  CONCENTRflTIONS IN UG/M3 FOR  THE HOUR 1400-1700  (1ST ON 730436

-------
                3«     4«    51    6«     7»     8«    9«
           2«     3*    1«    S»     6*    7«
S02  CONCENTRRT10NS IN UG/M3 FOR THE  HOUR  1700-2000 fIST  ON
S02  CONCENTRPT10NS IN UG/M3 FOR  THE  HOUR  2000-2500 MST ON  ~'?.1UO

-------
     19    2*    39    U«    59    69
                                               3»     90     109    110
                                                                                10    30    3t    H»    50    60    70    80     90     100    ::'
           2*    3»
                             59    6»    7«
                                                     99
                                                                                19     29    39    U«    54    6»    79    38    90
S02  CONCENTRflTlONS  IN UG/M3  FOR THE HOUR  -100-200 MST  ON 780^7     S02   CONCENTRfiTIONS IN UG/M3  FOR THE  HOUR 200-500 MST ON  76C'-07

-------
     1*
                                                                            1*     2*    3*    «•     5»     6«    7t    8*     9«    !•«    MJ
S02  CONCENTRPTIONS IN UG/M3 FOR THE HOUR  500-800 MST ON 7??'- r      SOS  CONCENTRPT]ONS  IN UG/M3 FOR THE HOUR  800-1100 MST ON 7oO

-------
     1*    20     31
                             50     Si
                                              80    90
                                                                             10
                                                                                   Z«     39     U0     50     60    70    80    90     :•'<>
     1«    2*    3*
                                                                                                                       30     90     :,;
S02  CONCENTROTIONS  IN  UG/M3 FOR THE HOUR  1100-1400  MST  ON 750^-07    S02  CONCENTROTIONS IN UG/M3  FOR  THE HOUR  U00-1700 MST  CN

-------
     It
    T-H
 2«
T-H
                3*    U»     5*     S«
                24
 8«
T-H
                                   tXlKOTB
                                             Hi
                                 3 OSKOTO
                                 MEBRflSKn
                                 COLORQDO
                 3*
                                  6«     7C     8*
 g<
T-H
                                                         :ae
                                                                  V,
                                                         :»
                                                                                                        ro
302  CONCENTRRTIONS IN UG/M3 FOR  THE HOUR  1700-2000 MST ON  "

-------
                                  248
5.   9-11  JULY 1975 METEOROLOGY;  1976  EMISSIONS

-------
                               S»     6*     7«
                                                        9«     lit
            i/
          4 ::.' '
 ' 11
                            tlONTPNd
                            HTOniNB
                         12
ooroTp
                                     COLORflOO
      :*     2'     3«
                                                                                           10
                                                                                 17
                                                                                                           nONTQNR
                                                                                                                      OP^OTQ
                                                                                                             6      COLORADO



                                                                                                            '» 1
                                                                                                              54     6*
S02   CONCENTRPTJOMi IN  UO/M3 FOR THE  HOUR 503-800  MST ON  ?5i7^0
                                         SOS   CONCENTPCTIONs IN  UG/M3 FOR THE  HOLR  b-V-; : .V  M£T rij  r^":'

-------
      '*'     i»     3»     ua     s*     en     7»
  10
            1°,
            -'-'1
                                                          Si     KJW    n:
                        13
                       f?fr
3 OflKOTH
                                      NEBRflSKfl
                                      COLORBDO
            2«     3fl     tie      50     60
                                                         9J     1J9
                                                                                      ID     ?«     39     UU     59     6J
                                                                                                             HONTPNP
                                                                                                             NYQniHG
                                                                                3 OPKCTO
                                                                                                  4...
                                                                                                                      COUGRfDO
                                                                                                       2 -'
                                                                                                   3D
                                                                                                                Sii    Sd
302   CCSCE1!' PnTIONS IN  UG/M3  FOR  THE HOUR n3d-JHe0  UST  ON "£ )7S;-     302  CONCENTRPT1ON3 IN  UG/M3  FOR Tr£  HOUR  m.00-:"'30  MST  OM 750^,?.^

-------
     1*     2»     3*     -•     S»     6»     7«     8i»     3«
                            nONTANfl
                                     N DAKOTA
                                     NEBRASKA
                                     COLORADO
                                                         9«
                                                                                     '*
                                                                                            'p
                                                                                                           MTOHING
                                                                                      ----—55     2
                                                                                                                    3 DRKOTQ
                                                                                                                    NEBR03KP
                                                                                                                    COLORADO
                                                                                           2»
                                                                                                                                        3*     1J*     ' 1J
S02   CONCENTRPTlOriS :N  Uo, M3 FOR THE  HOUR  1730-2k>0k> MST  ON 7;. 071-.-1     S02  CONCENTRPT IONS  IN UG-'hS FOR  THE HOUR  2000-2300 MST  ON

-------
            M
==,6
                  3»     >ta     50     Si     79
                                     N DflKOTfl
                                                               .*»     11 iJ
                            HYOniNG
                                     3 OBKOTfl
                                     OOUORflOO
                                                                                    — 2
                                                                                                       a a     sa     sa
                                                                                                          nONTRNfl   H DOKOTP
                                                                                                                                30     90
                                                                                                       i-13
                                                                                                                   NEBRB3KH
                                                                                                                   COLORPDO
                                                                                                                                                           PO
                                                                                                                                                           en
                                                                                                                                                           ro
            20     30
                               50     60     -a
                                                                                    U     20     30     <4»    5a     63     70
502   CCNCENTRfiTIONS IN UG/M3 FOR  THE  HOUR -130-20(3  MS' CN  7.-;07i9      S02   CONCENTRflTIONS  IN UG/M3 FOR  THE HOUR 223-503  MST ON  7?J"iO

-------
s



si
      1*     2*     3«     U4     54     64     7?     34     94      144



                            nONTflNB   N OOIOTO
                                                                                       11
                                                                                                    3*
              .3.3
                        16
                       *A
                              NYOniNG
                           17
                                       a OHKOTR
                                       NE8PP3KQ
                                       COLOPPDO
                                 54     6J
                                                                                      i.—-
                                                                                       2
                                                                                                                 5«     6»     7*     8*     9*      It*
                                                                                                              nONTRNfl    N POKOTO
                                                                                                              MTOntNS
                                                                                                                       COLOPODO
                                                                                        I*
                                                                                                                              7i9     31     9*
S02   CONCENTRATIONS  IN  UG/M3 FOR  THE HOUR ;^t-Sj0 MST  ON
                                                                                  S02   OONCENTRPT I ONS IN  UG/M3  FOR THE  HOUR  SOO-ilPO  MST ON  7C .;" ; :

-------
           23     30     U0     50     Eii     70
                            HONTflNB    ,'l DflK'jTB
                     12
                                     3 DDKOTR
                                     NEBRH3KB
                                     OOLORPOO
                — -4    2—-
                                                                                     10     20     30
                                                                                                                                       90

                                                                                                                    3 DflKOTQ
                                                                                                                    NEBRH3Kfl
                                                                                                                    COLORROO
            26     30    Hi     50     60     70
                                                                                     10
                                                                                                 30     «0     Si*     60     78     8i)     30     1»13
S02   CONCENTRPTIONS  !N UiJ/M3 FOR THE  HOUR  1100-1409 MS"  ON "^3/13    S02   CONCENTRATIONS IN  UG >13  ^OR THE  HOUR  1400-1700 MST ON  "jTl-.

-------
             13
                                      6<
                            nONTRNB    N OMOTQ
                                                         g«     Hi    !lil
                                                                                                                                   3*
                       5

                      ,1)'.
                       '.    •  HTflniNS

                       V
 •-. 7
                                      S DRKQTQ
                                      NE3RR3KR
                                      COLOCROO
            2*
                         U4     5*     6<
                                                                                                            nONTBNQ   N DPKOTP
ll.
I :
    .  MYoniwg
                                                                                                         11
                                                                                                        --*.
                                                                                                                     3 ORKOTR
                                                                                                                     NEBRRSKfl
                                                                                                                                 I    I
                                                   5J     3d
                                                                                      :•     n     3*     4*     5*     6*     ?a     8»     j»     ice     . •>
S02   COHCENTKPT10NS  IN UG/M3 FOR  THE HCoS1 1^30-2000  MST  ON 750710    S02  CONCENTRfiT 1ON3  IN UG/M3 FOR  THE  HOUR 2000-2300 MST  or.  "0":^

-------
     10    30     3
-------
s



st
       1*    2«     3*     U»     S«     6«    78     38     H
                              HONTSNO
                                                                                                                                                      ll»
                              NTOniNO
                                       NE3RR3KH
                                       COLORPOO
              2*     3»
                                 S«
                                                                                             13
                                                                                                            HONTRND
                                                                                                       '-      MTOHINS
                                                                                                                     N DQKOTtl
                                                                                                                      3 OPKOTP
                                                                                                                      COLORRDO
                                                                                       It     It     34     14 •     S<     B«     78
                                                                                                                                                !••    lit)
  S02   CONCENTRflTIONS IN  UG/M3 FOR THE  HOUR  500-300  MST ON  7St>';]       S02   CONCEN'R-.'! ONS IN UG/M3 FOR THE  HOUR  800-1130 MST  ..N 75j"ll

-------
302  CONCENTRATIONS IN UO/M3  FOR  THE  HOUR  1100-1400 MST ON 7Z'i7\l    S02  CONCENTRfiTI ONS IN UG/113 FOR THE HOUR  : 40)3-1700 MS" ON

-------
3-
       i*
      T
                             nONTQNQ
                                            It
                                           i I  i i
                                                  34     9*
                                     3 WKOTtt
                                     coiceaoo
                                                                                                                   ro
                                                                                                                   en
                                                        9»     \t>    11.'
  S02  CONCENTRPT10MS IN UG/M3 FOR  THE  HOUR  1700-2000 MST  ON "5J".:

-------
                                  260
6.   9-11  JULY 1975 METEOROLOGY;  1936  EMISSIONS

-------
                     3«
                                                                         11 J
                                                                                                            U«
a-
              is
                               nONTflNO
                               MTOHIN6
                            12
                                             • •   [('.•„•. •  .

                                             ' ••".  "SVV"\ '
                                                   V ».-  -
                                        3  DflKOTQ
                                        COLOBQOO


                                        '0
^
 2
                                                                                               Ifl
                                                                                                                (lONTBNfl
                                                                                                                HTOHIN6


                                                                                                                20
                                                                                                 A
                                                                                                                         N DBKOTB
                                                                                                                         3 DflKOTB
                                                                                                                         NEBRB3MJ
COLORBDO

13
                                                                                                                                                                  ro
                                                                                                                                                                  CT
                     3»      VII     Si
                                                            90     IDC
                                                                                         1*
                                                                                                      3<     UD     5«     6«
                                                                                                                                      a*
S02   CONCENTRRTlONe IN  UG/M3  ruK  THE HOUR e
                                                              (1ST ON  7S07 )9      S02   CONCENTRPT I ON?  IN UG/M3 FOR  THE  HOUR  300-1100 MSI  ON

-------
 — 75
       i '-Jai
                            MONTBNB   N OQKOTB
                            UTOnlNG
                      i  '-•>'
                        -~
                                          «    J?A
                                      NESRQ3SS
                                      COLOPBCO
                  r-19
                                                                                -,..25
                                                                                                        44
                                                                                                       T-H
                                                                                                     17
                                                                                                            MTOnlNB
 619
T-H
 7»
-H-i
 80
T-H
                                                                                                                                        sa     ia»
                                                                                            	~- 12
                                                                                                           .~-9
                                                                                                                     3 DBItOTB
                                                                                                                     NEBRQ3KQ
                                                                                                                     COLORflDO
                                                                                                                                  ' ^2
                                                   at     at
                                                                                                  39     u»     50     6e     7«     31
302   CONCENTRPTIONS  IN UG/MS FOR  THE HOUR :i30-;^0a  MST ON 7t07jy    S02  CONCENTRPTIONS  IN UG/M3 FOR  THE  HOUR  14-00-1700 MST Of!  7507.;;.

-------
       1*
                   3t
  2 -
                             nONTONfl    N DflKOTH
                      17
 HTORIN6

 13

.fa
          6«
         i-H
                                           £   J'^.

                                           \\ P'*':-.

                                              *  \\
                                      9 DHKOTR
                                      NEBRQ3KR
                                                 34
                                      COLORflGj
      1*     2«
                         4«     56     6«     7«     8.     9*
                                                                      110
                                                                                            2*     3«
                                                                                           I  I I I  I I I  I
                                                                                                                      6»     7f     3«     9«
                                                                                                            nOMTflNR   S DflKOTQ
                                                                                                            HTOtlINQ
                                                                                                      -7 11
                                                                                                                     3 BQKOTQ
                                                                                                                                 28
                                                                                         COLORflDO

                                                                                         .9
                                                                                      \i     it     3«
S02   CONCENTPPTIONS  IN UG/M3 FOR  THE HOUR  1700-2000 MST  Or; '33~:i    S02   CONCEMTPPTIONS  IN  UG/M3  FOR  TH£  HOUR  2300-2309 M£T  on

-------
      It     it     3*     U»     5U     60     70
                            MONTQNQ
                                                   if.     90     Ida    11'
             14
                      26

                     ../I
     '"I""
                            14
                                     3 OQKOTB
                                     NEBRflSKB
                                                 34
            2*     3»
                                                                                           -H-
                                                                                                       -i a     sa
                                                                                                           rlONTRNR  I  N DQKOTR
23
         t-3
       ,-ta
                                                                                                          X
                                                                                .17
                                                                                                           ---A

                                                                               L	T
                                                                                                                    3 OflKOTB
               _5        CQLQRSDO

               ~;:      10      ,
                _  ^ ff>      '
                                                                                                              53     68
302   CONCENTRflTICNS !N  UG/M3  FOR THE  HOUR  -100-200  ^' ON  75e7j3      302  CONCENTRPT: JN3 IN  UG/M3  FOR THE  HOUR  200-503 HCT  :~i 'f^i.?

-------
        1*     it
s-
                                 54     64     74     84     94
                              flONTHNB   N DUKOTfl
                                                                                                    3«
                                                                                                                 5«     60
                                                                                                                                    30     9*
                        49
      — 2
I    KYQnlNG

!   21
   19
     '
                                        3 OQROTQ
                                        NEBRR3KR

                                       COLORRDO
        1*     a*     3«     4(     5*     64     7»     fit     9»     184    • ;.)
                                                                                              34
                                                                                                              HTOrlNS
                                                                                                                       N DRKOTR
                                                                                                                       3 DRKOTH
                                                                                                                       NEBRS3KO
                                                                                                                       COLORHDO
                                                                                        1*     2*     3*
                                                                                                                 5<    6<     7»
  302   CONCENTRATIONS  IN UO/M3 FOR  THE HOUR 500-800 MST ON  7S0710       S02   CONCENTRATIONS IN  UG/M3  FOR THE HOUR  330-1330 MST ON  750" 1J

-------
     10    20     30     4i9     50     60     70     3J     9«     100
                                                                                  l»    it    3»    40    50     68     70     30     90     ll>«

     10    Zt    3»     40     58     60     70     30     30     1190    110
                                                                                              30     U0     50     60     70     80    90     100
S02  CONCENTRRTIONS IN UG/M3 FOR  THE HOUR  1:50-1490  MST OM 75071J    302   CONCENTRATIONS  IN  UG/M3 FOR  THE HOUR  lLHD£-;7i30 MST  CN  "J7!J

-------
        H
                    3     »      S»
                                                   34    90     \J«
s-
                             HONTQNQ   N OflKPTQ
              1?
?
                                      9 DQKOTfl
                                      NEBPB3KH
                                      COLORBDO



                                      17
                                                                                     It     24     3«     44     54    68     78     3*
                    3*
                                S«     S»     7(     38    98     108    no
                                                                                      18     28     38     U8    54     68     78     84     98     188
  S02  CONCENTRPTIONS  IN  UG,M3  FOR  THE HOUR  1-00-2000  MST UN  7^.1710     302   CONCENTRPTI CMS  IN UG/M3  FOR THE  HOUR  2000-2300 McT  ON 7TT~i

-------
     U     20     30
                                                  3D     98     1^19    11
                                                                                           it     aa
                      U
                      ill
                      ,-- ;s10
                      '
  is     •;
  ^-

P"
                                          5    14


                                          ^   ft
                                                    •T
                                                 18
                             7       COLORflOO


                            -'-       9
      10     24     30     41     -.t     5D
 6
 i^
. n if
                                                                                                              se     se     ?a
                                                                                                           nONTQNB    N DwKOTQ
                                                                                                                                                    11J

                                                                                                           NTOHIN9


                                                                                                            10
                                                                                          	^m    /
                               $    14
                          3 DRKOTft ;
                                                                                                                    COLORADO
                                                                                     .1     zt    3»     tie     sa     sa     79     30     8J
302   CONCENTRPTIONS IN  UG/*3 r'R  "HE  HOUR -100-20Q  MST OH  750^11       S02  70NCENTRPTIONS  IN  UG/M3  FOR  THE HOUR  200-500 MST  .IN 7;-07il

-------
s
s-
        1*     2*     3*     U«     51*     6*     7»
                                      N DPKOTB
          _
                                           5    13
                                      9 OPKOTO-;
                                      NEBRflSKO
                                                          8*     .iJ
                                                          '.2
                                      COLORADO
                                S4     6*
                                                                                                              5«
                                                                                                    10

                                                                                                    iff.
12  ,
                                                                                                                                IS
                                                                                                                   COLOfOOO


                                                                                                                    9
                                                   S»    88     !««    | l?
                                                                                                  3»
                                                                                                              S(     61     70    8«     9«
                       C
                                                                                                                                                           PO
                                                                                                                                                           o>
                                                                                                                                                           vo
  S02   CONCENTRATIONS  IN UG/M3  FOR THE  HOUR  530-800  MS' SN  ^'rj!       302  CONCENTRATIONS IN  UG/^3 FOR  THE HOUR 800-1100 MST  ON  "£''"1

-------
  17

&
            39
                  39
                 T-H
                               59     6a
                            MTOHIN6
                            6
             32 ;TKV-^.
                     "
                                    N DflKOTfi
                                     3 DflKOTQ   '
                                    NEBRHSKfl
                                                 16  •
                                                 A  '
COLORADO

9
            2*     39
                               S»     60     78
                                                                                          2«     30
                                           r-Jl
                                                                                                                         79     30
                                                                                                                                      9*
                                                                                                                                                   1!J
                                                                                                   '•   '—>
                                                                                                   j    ." 5  :
                                                                                                                   N DOKOTB
                                                                                                                        4    11
                                                                                                                   3 DPKOTO
                                                                                                                               IS
                                                                                                 39     U8    5*     59     70     89     90     1»9    110
S02   CONCENTRflTlONS  IN UG/M3 FOR  THE HOUR 1100-l«t00 MST  ON -^711     302  CONCENTRfiTIONS  IN UG/HS  ^OR  THE HOUR  1430-1700 MST ON  7337]

-------
      1*     it     3*     4»     54      6*     7»     S«
                             MONTONfl
                      12
                             MTOnlNB
                          • t-
                                       I  I I I  I I
                                      N DflKOTfi
                                      3 DBKOTH
                                      NEBRflSKR
                                      COLOKBDO
                                                        \a
                                                         1
                                                 19,
                                 I ' I  I ' I
                                                  'f.t.
                                                                                                                      PO
      1*     2»     36     M«     S»     6»     ?«     8*     9«     \»*     I'll
S02   CONCENTRRTIONS IN  UG/H3  FOR  THE HOUR 17^0-2000 M«T ON

-------
                                    272
                             REFERENCES
Averitt, Paul (1974), "Coal Resources of the United States," Bulletin 1412,
     U.S. Geological Survey, Washington, D.C.

Benedict, Hanson (1976), "U.S.  Energy:  The Plan That Can Work," Technol.
     Rev., May 1976, pp. 53-59.

Benson, S. W. (1968), Thermochemical  Kinetics  (John Wiley & Sons, New York,
     New York).

Bolin, B., and L. Granat (1973), "Local  Fallout and Long Distance Transport
     of Sulfur," Ambio, Vol. 2, pp.  87-90.

Bolin, B., and H. Rodhe (1973), "A Note  on  the Concepts  of Age Distribution
     and Transit Time in Natural Reservoirs,"  Tell us, Vol.  25, pp.  58-62.

Bolin, B., G. Aspling, and C.  Persson (1974),  "Residence Time of Atmospheric
     Pollutants  as Dependent on Source Characteristics,  Atmospheric Diffusion
     Processes,  and Sink Mechanisms," Tell us,  Vol.  26,  pp.  185-194.

Boris, J. P., and D. L. Book (1973),  "Flux  Corrected Transport—I.   SHASTA,
     A Fluid Transport Algorithm That Works,"  J.  Comput.  Phys.,  Vol.  11,
     pp. 38-69.

Brosset, C.  (1973), "Airborne  Acid,"  Ambio,  Vol.  2, No.  1-2, pp.  1-9.

Bufalini, J. J., and W. A.  Lonneman  (1977),  "Proceedings of Symposium on
     the 1975 Northeast Oxidant Transport Study," EPA 600/3-77-017,
     Environmental  Protection  Agency, Research Triangle  Park, North Carolina.

Bureau of Mines  (1975), "The Reserve  Base of U.S. Coals  by  Sulfur Content,"
     IC8680 (East of the Mississippi) and IC8683 (The Western States),
     Washington, D.C.

Businger, J. A., et al. (1971), "Flux-Profile  Relationships in the  Atmospheric
     Surface Layer," J. Atmos.  Sci.,  Vol.  28,  pp. 181-189.

Chamberlain, A.  C.  (1966),  "Transport of Gases to and from  Grass and  Grass-
     Like Surfaces," Proc.  Roy. Soc., A.  290,  pp. 236-260.

           (1960),  "Aspects of the  Deposition  of Radioactive and Other Gases
     and Particles,"  Int.  J.  Air  Poll..  Vol.  3,  pp.  63-88.

-------
                                      273
Chemistry and Engineering News [C&EN](1977), "Outlook for Coal:  Bright,
     but with Problems," pp. 24-31, 14 February 1977.

Christensen, 0., and L. P. Prahm (1976), "A Pseudospectral Model  for
     Dispersion of Atmospheric Pollutants," J. Appl.  Meteor., Vol. 15,
     pp. 1284-1294.                         	  	

Corrsin, S. (1974), "Limitations of Gradient Transport Models in  Random Walks
     and in Turbulence," Adv. in Geophysics, Vol.  ISA, pp. 25-60.

Czeplak, G., and C. Junge (1974), "Studies of Interhemispheric Exchange in
     the Troposphere by a Diffusion Model," Adv. in Geophysics, Vol.  18B,
     pp. 57-72.                                           	

Dana, M. T., J. M. Hales, and M. A. Wolf (1972), "Natural Precipitation
     Washout of Sulfur Dioxide," BNW-389, Atmospheric Sciences Dept.,
     Battelle-Pacific Northwest Laboratories, Richland, Washington
     (NTIS PB-210 968).

Davis, B. L., et al.  (1976), "A Study of the Green Area Effect in  the  Black
     Hills of South Dakota," Atmos. Environ.. Vol. 10, pp. 363-370.

DeMarrais, G. A., and N. F.  Islitzer (1960), "Diffusion Climatology  of the
     National Reactor Testing Station," Report IDO-12015, Idaho Falls
     Operation Office, U.S.  Atomic Energy Commission, Idaho Falls, Idaho.

Dickerson, M. H., T.  V. Crawford, and W. K. Crandall  (1972), "Long-Range
     Transport, Diffusion, and Deposition from a Russian Nuclear  Excavation
     Project," UCRL-51281, Lawrence Livermore Laboratory, Livermore,
     California.

Draxler, R. R., and W. P. Elliott (1977), "Long-Range Travel of Airborne
     Material Subjected to Dry Deposition," Atmos. Environ., Vol.  11,
     pp. 35-40.

Edwards, R. G., A. B.  Broderson, and W. P. Hauser (1976), "Social, Economic,
     and Environmental Impacts of Coal  Gasification and Liquefaction  Plants,"
     IMMR14-GR2-76, Institute for Mining and Minerals Research, University
     of Kentucky, Lexington, Kentucky.

Egan, B. A., and J. R. Mahoney (1972a), "Numerical Modeling of Advection
     and Diffusion of Urban Area Source Pollutants,"  J. Appl. Meteor.,
     Vol. 11, pp. 312-322.

	 (1972b), "Applications of a Numerical Air Pollution Transport
     Model to Dispersion in the Atmospheric Boundary Layer," J. Appl.
     Meteor., Vol. 11, pp. 1023-1039.

Eliassen, A., and J.  Saltbones (1975),  "Decay and Transformation  Rates of
     S02 as Estimated from Emission Data, Trajectories and Measured  Air
     Concentrations," Atmos. Environ..  Vol. 9, pp. 425-430.

-------
                                       274
Environmental Protection Agency [EPA](1977), unpublished data supplied by
     Mr. Terry Thoem, EPA Region VIII, Denver, Colorado.

	 (1976a), "Surface Coal  Mining in the Northern Great Plains of
     the Western United States," OEA 76-1, EPA Region VIII, Denver,
     Colorado.

	 (1976b), "Monitoring and Air Quality Trends Report, 1974,"
     EPA-450/1-76-001, Environmental Protection Agency, Washington, D.C.

	 (1976c), "Existing and Proposed Fuel Conversion Facilities
     Summary," TS-5, EPA Region VIII, Denver, Colorado.

	 (1975), "Standards of Performance for New Stationary Sources,"
     Code of Federal Regulations,  §40, Part 60.

	 (1973), "Emission Factors for Trace Substances," EPA-450/2-73-001,
     Research Triangle Park, North Carolina.

Environmental Science & Technology [ES&T](1976), "How To Make Coal Burn
     Cleaner," Vol. 10, pp. 16-17.

Eriksson, E. (1966), Handbuch de Pflanzenernahrung und Dungung, Vol.  2,
     No. 1, p. 774.

Federal Energy Administration [FEA](1975), "Coal Conversion Program—Final
     Environmental Statements," Washington, D.C. (also NTIS PB-250 104).

Federal Power Commission [FPC](1976), "FPC Form 67:  Steam Electric Plant
     Air and Water Quality Control Data for the Year Ending December 31, 1975

Federal Register (1975), Vol. 40,  No. 33, pp. 7042-7070, 18 February  1975.

           (1974), "Air Quality Implementation Plans--Prevention of
     Significant Air Quality Deterioration,"  Vol.  39, No.  235,  Part III,
     pp. 42510-42517, 5 December 1974.

Fisher, B.E.A. (1975), "The Long Range Transport of Sulphur Dioxide,"
     Atmos. Environ., Vol. 9, pp. 1063-1070.

Forsythe, G. E., and W. R. Wasow (1960), Finite-Difference Methods for
     Partial Differential Equations (John Wiley & Sons, New York, New York).

Fortak, H. G. (1974), "Mathematical Modeling  of Urban Pollution," Adv. in
     Geophysics. Vol. 18B, pp.  159-172.

Fox, D. G. (1975), "Modeling Atmospheric Effects—An Assessment of the
     Problems," Proc. of the First International Symposium on Acid
     Precipitation of the Forest Ecosystems,  Ohio State University.

Fox, D. G., and S. A. Orszag (1973), "Pseudospectral Approximation to
     Two-Dimensional  Turbulence," J. Comp.  Phys., Vol. 11, pp.  612-619.

-------
                                      275
Galbally, I. E. (1974), "Gas Transfer Near the Earth's Surface," Adv.  in
     Geophysics, Vol. 18B, pp. 329-340.                          ~

Garland, J. A., et al. (1974), "Deposition of Gaseous Sulphur Dioxide to
     the Ground," Atmos. Environ.. Vol. 8, pp. 75-79.

Georgii, H. W. (1970), "Contributions to the Atmospheric Sulfur Budget,"
     J. Geophys. Res.. Vol. 75, pp. 2365-2371.

Hage, K. D., et al.  (1966), "Particle Fallout and Dispersion in the
     Atmosphere," Final Report SC-CR-66-2031, Aerospace Nuclear Safety,
     Sandia Corporation, Los Alamos, New Mexico.

Heffter, J. L. (1965), "The Variation of Horizontal Diffusion Parameters
     with Time for Travel Periods of One Hour or Longer," J. Appl.  Meteor.,
     Vol. 4, pp. 153-156.

Heffter, J. L., A. D. Taylor, and G. J. Ferber (1975), "A Regional-Continental
     Scale Transport, Diffusion, and Deposition Model," Technical  Memorandum
     ERL ARL-50, National Oceanic and Atmospheric Administration,  Air
     Resources Laboratories, Silver Springs, Maryland.

Heimbach, J. A., A.  B. Super, and J. T. McPartland (1975), "Colstrip
     Diffusion Experiment," Dept. of Earth Sciences, Montana State
     University, Bozeman, Montana.

Hidy, G. M., E. Y. Tong, and P. K. Mueller (1976), "Design of the  Sulfate
     Regional Experiment (SURE)," EPRI-EC-125 Volume 1, Electric Power
     Research Institute, Palo Alto, California.

Hill, A. C. (1971),  "A Sink for Atmospheric Pollutants," J. Air Poll.  Contr.
     Assoc., Vol. 21, pp. 341-346.

HbgstrBm, U. (1975), "Further Comments on the Long Range Transport of  Air-
     borne Material  and Its Removal by Deposition and Washout," Atmos.
     Environ.. Vol.  9, pp. 946-947.

Holzworth,  G.  C.  (1972), "Mixing Heights, Wind Speeds, and Potential for
     Urban  Air Pollution Throughout the Contiguous United States,"  AP-101,
     Office of Air Programs, Environmental  Protection Agency, Research
     Triangle Park,  North Carolina.

Hubbert, M.  K.  (1971),  "Energy Resources,"  in Environment:  Resources,
     Pollution & Society, W.  W.  Murdoch, ed., pp. 89-116 (Sinauer  Associates,
     Incorporated, Stamford, Connecticut).

InterTechnology Corporation (1971),  "The U.S. Energy Problem,"  Vol.  II,
     Appendices—Part B, Appendix S, "Technology of Alternative Fuels,"
     Warrenton, Virginia (also NTIS PB-207 519).

-------
                                    276
Izrael, Yu. A. (1971). "Radiation Conditions in the Zone of Long-Range Fallout
     from Underground Nuclear Cratering Explosions," The State Committee for
     Uses of Atomic Energy, Moscow, U.S.S.R., presented at the Third Stage Soviet-
     American Technical Talks on the Peaceful Uses of Nuclear Explosions,
     Was.hington, D.C.

Johnson, W. B., D.  E. Wolf, and R.  L.  Mancuso (1975), "Feasibility of the
     Air Quality Budget Concept," Stanford Research Institute, Menlo Park,
     California.

Junge, C. E. (1963), Air Chemistry and Radioactivity (Academic Press, New
     York, New York).

Kaakinen, J. W., R. M. Jorden, and R.  E.  West (1974), "Trace Element Study
     in a Pulverized Coal-Fired Power  Plant," Paper 74-8, 67th Annual
     Meeting of the Air Pollution Control  Association,  Denver, Colorado,
     June 1974.

Kao, S. K. (1974),  "Basic Characteristics of Global Scale Diffusion in the
     Troposphere,"  Adv. in _Gepphysi_cs, Vol.  18B,  pp. 15-32 (Academic Press,
     New York, New York).

Kao, S. K., and D.  Henderson (1970), "Large-Scale Dispersion of Clusters
     of Particles in Various Flow Patterns," J.  Geophys.  Res., Vol.  75,
     pp. 3104-3113.

Katz, M. (1949), "Sulphur Dioxide in the  Atmosphere and Its Relation to
     Plant Life," Ind. and Eng. Chem., Vol.  41,  pp. 2450-2465.

Knox, J. B. (1974), "Numerical Modeling of the Transport, Diffusion and
     Deposition of Pollutants for Regional and Extended Scales," J.  Air
     Poll. Contr. Assoc., Vol. 24,  pp. 660-664.

Kiichler, A. W. (1966), "Potential Natural  Vegetation,"  Sheet #90 (map),
     U.S. Geological Survey, Washington,  D.C.

Lamb, R. G., and G. Z. Whitten (1975), "An Assessment of the Impact of
     Illinois Sulfur Emissions on the  Air Quality of the Northeastern
     United States," EF75-63, Systems  Applications, Incorporated, San
     Rafael, California.

Li, T.  Y., and H. E. Landsberg (1975), "Rainwater pH Close to  a Major Power
     Plant," Atmos. Environ., Vol.  9,  pp.  81-88.

Liu, M. K., and D.  R. Durran (1977), "On  the Prescription of the Vertical
     Dispersion Coefficient over Complex  Terrain,"  Joint  Conf. on Applica-
     tions of Air Pollution Meteorology,  American Meteorological  Society
     and Air Pollution Control Association,  28 November-2 December 1977,
     Salt Lake City, Utah.

-------
                                     277
Liu, M. K., et al. (1976), "The Chemistry, Dispersion, and Transport of
     Air Pollutants Emitted from Fossil Fuel Power Plants in California,"
     EF76-18, Systems Applications, Incorporated, San Rafael, California.

Liu, M. K., and J. H. Seinfeld (1975), "On the Validity of Grid and
     Trajectory Models of Urban Air Pollution," Atmos. Environ., Vol. 9,
     pp. 555-574.                               	

Los Angeles Air Pollution Control District [LAAPCD](1974), "Profile of
     Air Pollution--1974," Los Angeles, California.

MacCracken, M. C. (1976), "Multistate Atmospheric Power Production Pollution
     Study (MAP3S)," UASG 76-11, Lawrence Livermore Laboratory, Livermore,
     California.

Machta, L. (1966), "Some Aspects of Simulating Large Scale Atmospheric
     Mixing," Tell us, Vol. 18, pp. 355-362.

Magee, E. M., H. J. Hall, and G. M. Varga, Jr. (1973), "Potential  Pollutants
     in Fossil Fuels," EPA-R2-73-249, Environmental Protection Agency,
     Research Triangle Park, North Carolina.

Mansfield, T. A., and O.V.S. Heath (1963), "An Effect of 'Smog1  on Stomatal
     Behavior," Nature, Vol. 200, p. 596.

Marschner, F. (1950), "Major Land Uses in the U.S.," revised by J.  R.  Anderson
     (1967), U.S. Government Printing Office, Washington, D.C.

Martin, A., and F. R. Barber (1971), "Some Measurements of Loss of Atmospheric
     Sulphur Dioxide Near Foliage," Atmos. Environ., Vol. 5, pp.  345-352.

McMahon, T. A., P. J. Denison, and R. Fleming (1976), "A Long-Distance Air
     Pollution Transportation Model Incorporating Washout and Dry Deposition
     Components," Atmos. Environ., Vol. 10, pp. 751-761.

McMullen, T. B., R. B. Faoro, and G. B. Morgan (1970), "Profile of Pollutant
     Fractions in Nonurban Suspended Particulate Matter," J. Air Poll. Contr.
     Assoc.. Vol. 20, pp. 369-372.

Miller, J. M., J. Galloway, and G. E. Likens (1975), Proc.  of the  First
     International Symposium on Acid Precipitation of the Forest Ecosystem,
     Ohio State University.

Miller, J. M., and R. de Pena (1972), "Contribution of Scavenged Sulfur
     Dioxide to the Sulfate Content of Rain Water," J. Geophys.  Res.,
     Vol. 77, pp. 5905-5916.

Monin, A. S., and A.  M.  Yaglom (1971), Statistical Fluid Mechanics:  Mechanics
     of Turbulence, Vol. 1 (MIT Press, Cambridge, Massachusetts).

-------
                                    278
Nehring, Richard, and Benjamin Zycher (1976),  "Coal  Development and Govern-
     ment Regulation in the Northern Great Plains:   A Preliminary Report,"
     R-1981-NSF/RC, The Rand Corporation,  Santa Monica,  California.

Nephew, E. A. (1973), "The Challenge and Promise of Coal," Technol. Rev..
     December 1973, pp. 20-29.

Nb'rdlund, G. G. (1975), "A Quasi-Lagrangian Cell Method  for Calculating
     Long-Distance Transport of Airborne Pollutants," J.  Appl.  Meteor..
     Vol. 14, pp. 1095-1104.

	 (1973), "A Particle-in-Cell  Method  for Calculating  Long Range
     Transport of Airborne Pollutants," Technical Report No.  7, Finnish
     Meteorological Institute.

Nordo, J. (1973), "Meso-Scale and Large-Scale  Transport  of Air  Pollutants
     Proc. Third International Clean Air Congress,  B105-B108,  Dusseldorf,
     Federal Republic of Germany, VDI-Verlag.

Nordo, J., A. Eliassen, and J. Saltbones (1974), "Large-Scale  Transport of
     Air Pollutants," Adv. in Geophysics,  Vol.  18B,  pp.  137-150.

Northern Great Plains Resource Program [NGPRP](1974), "Atmospheric Aspects
     Work Group Report," Denver, Colorado.

Ottar, B. (1973), "The Long Range Transport of Air Pollutants," Proc.  Third
     International Clean Air Congress,  B102-B104, Dusseldorf,  Federal  Republic
     of Germany, VDI-Verlag.

Owen, P. R., and W. R. Thompson (1963), "Heat  Transfer Across  Rough Surfaces,"
     J. Fluid Mech., Vol.  15, pp. 321-324.

Owers, M. J., and 0. W. Powell (1974),  "Deposition Velocity of Sulphur Dioxide
     on Land and Water Surfaces Using a 35S Tracer Method," Atmos.  Environ.,
     Vol. 8, pp. 63-67.

Parker, N. A., and B. C. Thompson (1976),  "U.S. Coal Resources  and Reserves,"
     FEA/B-76/210, Federal Energy Administration, National Energy Information
     Center, Washington, D.C. (also NITS PB-252 752).

Pasquill, F. (1974), "Limitations and Prospects in the Estimation of Disper-
     sion of Pollution on a Regional Scale," Adv. in Geophysics,  Vol.  18B,
     pp. 1-14.

Pedersen, L. B., and L. P. Prahm (1974), "A Method for Numerical  Solution of
     the Advection Equation," Tellus, Vol. 26, pp.  594-602.

Petrov, V. N. (1971), "Effect of Atmospheric Parameters  on the Diffusion and
     Fallout of Radioactive Products from Clouds Traveling Great Distances,"
     State Committee for Uses of Atomic Energy U.S.S.R., Moscow,  presented
     at the Third Stage Soviet-American Technical Talks  on the  Peaceful
     Uses of Nuclear Explosions, Washington, D.C.

-------
                                     279
Prahm, L. P., H. S. Buch, and U. Torp (1974), "Long-Range Transport of
     Atmospheric Pollutants over the Atlantic," Symposium on Atmospheric
     Diffusion and Air Pollution, pp. 190-195 (American Meteorological
     Society, Boston, Massachusetts).

Prahm, L. P., U. Torp, and R. M. Stern (1976), "Deposition and Transformation
     Rates of Sulphur Oxides during Atmospheric Transport over the Atlantic,"
     Tell us, Vol. 28, pp. 355-372.

Radian Corporation (1975), "A Western Regional Energy Development Study,"
     "Executive Summary" and "Volume III:  Appendices," RC# 10U-064, Austin,
     Texas.

Randerson, D. (1972), "Temporal Changes in Horizontal Diffusion Parameters
     of a Single Nuclear Debris Cloud," J. Appl.  Meteor., Vol. 11,
     pp. 670-673.

Rao, K. S., J. S. Lague, and B. A. Egan (1976), "An Air Trajectory Model
     for Regional Transport of Atmospheric Sulfates," Preprints, Third
     Symposium on Atmospheric Turbulence, Diffusion, and Air Quality.
     19-22 October 1976, Raleigh, North Carolina  (American Meteorological
     Society, Boston, Massachusetts).

Rao, K. S., I. Thomson, and B. A. Egan (1976), "Regional Transport Model
     of Atmospheric Sulfates," 69th Annual Meeting of the Air Pollution
     Control Association, Portland, Oregon.

Rasmussen, K. H., M. Taheri, and R. L. Kabel (1974), "Sources and Natural
     Removal Processes for Some Atmospheric Pollutants," EPA-650/4-74-032,
     Environmental Protection Agency, Washington, D.C.

Reiquam, J. (1970), "Sulfur:  Simulated Long-Range Transport in the
     Atmosphere," Science, Vol. 170, pp.  318-320.

Rodhe, H. (1972), "A Study of the Sulphur Budget  for the Atmosphere over
     Northern Europe," Tellus, Vol. 24, pp. 128-138.

	 (1971), "Measurements of Sulfur in the Free Atmosphere over
     Sweden, 1969-1970," Report AC-12, Institute  of Meteorology,
     University of Stockholm, Stockholm,  Sweden.

Rodhe, H., and J. Grandell (1973), "On the Removal  Time of Aerosol
     Particles from the Atmosphere by Precipitation Scavenging," Report
     AC-20, Institute of Meteorology, University  of Stockholm, Sweden.

Scott, W. D., and P. V.  Hobbs (1967), "The Formation of Sulfate in Water
     Droplets," J. Atmos. Sci.. Vol. 24,  p. 54.

Scriven, R.  A., and B.E.A. Fisher (1975a), "The Long Range Transport of
     Airborne Material and Its Removal by Deposition and Washout--!.
     General  Considerations," Atmos. Environ., Vol. 9,  pp. 49-58.

-------
                                 280
           (1975b), "The Long Range Transport of Airborne Material and
     Its Removal by Deposition and Washout--!I.  The Effect of Turbulent
     Diffusion," Atmos. Environ., Vol. 9, pp. 59-68.

Sehmel, G. A., S. L. Sutter, and M. T. Dana (1973), "Dry Deposition
     Processes," in "Pacific Northwest Laboratory Annual Report for
     1971, to the USAEC, Division of Biomedical and Environmental
     Research, Vol. II:  Physical Sciences, Part 1, Atmospheric Science,"
     BNWL-1751, pp. 150-153, Battelle Northwest Laboratories, Richland,
     Washington.

Sellers, W. D. (1965), Physical Climatology (University of Chicago Press,
     Chicago, Illinois).

Shepherd, J. G. (1974), "Measurements of the Direct Deposition of Sulphur
     Dioxide Onto Grass and Water By the Profile Method," Atmos.  Environ.,
     Vol. 8, pp. 69-74.

Slade, D. H. (1967), "Modeling Air Pollution in the Washington, D.C.  to
     Boston Megalopolis," Science, Vol.  157, pp. 1304-1307.

Smagorinsky, J. (1963), "General Circulation Experiments with the Primitive
     Equations:  I.  The Basic Experiment," Mon. Wea.  Rev.,  Vol.  91,
     pp. 99-164.

Smith, F. B. (1970), "A Contribution to  the Estimation of Pollutant Dosages
     Arising from a U.K. Source Using a  Simplified Trajectory Method,"
     Internal Meteorological Office Memorandum.

Smith, W. S. (1966), "Atmospheric Emissions from Coal  Combustion,"
     Publication AP-42, Public Health Service,  U.S. Department of Health,
     Education, and Welfare, Washington, D.C.

Spedding, D. J. (1969), "Uptake of Sulphur Dioxide by  Barley Leaves at
     Low Sulphur Dioxide Concentration," Nature, Vol.  224, pp. 1229-1231.

Thorn, A. S. (1972), "Momentum, Mass and  Heat Exchange  of Vegetation,"
     Quart. J. Roy. Meteor.  Soc., Vol. 98, pp.  124-134.

Tillman, D. A. (1976), "Status of Coal Gasification,"  Environ. Sci.
     Techno!., Vol. 10, pp.  34-38.

Trijonis, J. C., and K. W.  Arledge (1975), "Impact of  Reactivity  Criteria
     on Organic Emission Control Strategies in  the Metropolitan Los Angeles
     AQCR," TRW, Incorporated, El  Segundo, California.

Turner, D.  15. (1969),  "Workbook of Atmospheric  Dispersion Estimates,"
     999-AP-26, U.S. Public Health Service, Cincinnati, Ohio.

Turner, D.  B., J.  R. Zimmerman, and A. D.  Busse (1973), "An  Evaluation
     of Some Climatological  Dispersion Models," Appendix E of "User's
     Guide for the Climatological  Dispersion Model," EPA-R4-73-024, Environ
     mental Protection Agency, Research  Triangle Park, North Carolina.

-------
                                   281
Van der Hoven, I. (1957), "Power Spectrum of Horizontal  Wind Speed in
     the Frequency Range from 0.0007 to 900 Cycles per Hour," J.  Meteor.,
     Vol. 14, p.  160.
Wendell, L. L. (1972),
     Determined from a
     pp. 565-578.
"Mesoscale Wind Fields and Transport Estimates
Network of Wind Towers," Mon.  Wea.  Rev.,  Vol.  100,
Wendell, L. L., D. C. Powell, and R. L. Drake (1976), "A Regional  Scale
     Model for Computing Deposition and Ground Level  Air Concentration
     of S02 and Sulfates from Elevated and Ground Sources,"  pp.  318-324,
     Preprints, Third Symposium on Atmospheric Turbulence,  Diffusion, and
     Air Quality, 19-22 October 1976, Raleigh, North  Carolina (American
     Meteorological Society, Boston, Massachusetts).

White, W. H., et al.  (1976), "Midwest Interstate Sulfur Transformation
     and Transport Project:  Aerial Measurements of Urban and Power Plant
     Plumes, Summer 1974," EPA-600/3-76-110, Environmental  Protection
     Agency, Research Triangle Park, North Carolina.

Yanenko, N. N. (1971), The Method of Fractional  Steps (Springer-Verlag,
     Berlin, Germany).

-------