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).
------- |