EPA-600/4-8,-070
August 1981
LONG-RANGE TRANSPORT AND TRANSFORMATION OF S02 AND SULFATE
Refinement, Application, and Verification of Models
Grant No. R 805271
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
EPA-600/4-81-070
August 1981
LONG-RANGE TRANSPORT AND TRANSFORMATION OF S0? AND SULFATE
Refinement, Application, and Verification of Models
by
Teizi Henmi and Elmar R. Reiter
Department of Atmospheric Science
Colorado State University
Fort Collins, Colorado 80523
Grant No. R 805271
Project Officer
George C. Holzworth
Meteorology and Assessment Division
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Science Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
-------
ABSTRACT
Improvements have been made in the long-range transport model of SCL
and sulfate. Trajectories of the mean wind of the mixing layer are cal-
culated using wind and temperature sounding data. Heights of the daytime
and nighttime mixing layers are determined by temperature sounding data.
Trajectories of the mean winds of the nighttime ground-based stable layer
and of the daytime mixing layer, and trajectories of the mean wind of the
layer between daytime mixing height and nighttime stable layer height are
taken into consideration. Interpolation schemes of concentrations along the
trajectories into grid points, as well as the interpolation scheme of pre-
cipitation rate at meteorological stations into the midpoint of trajectory
segments, have been modified in order to save computing time. This model is
developed to calculate particularly the 24-hour concentration distributions
of SO and sulfate.
The model has been applied to calculate the distribution patterns of
concentrations and deposition amounts of SCL and sulfate over the area be-
tween 35°N and 45°N and between 75°W and 95°W for the dates of January 25
and July 11, 1976.
Although statistically significant correlations between the observed
concentrations and calculated concentrations are obtained both for SCL and
for sulfate, the results also show the limitation of applicability of the
model to the calculation of 24-hour average concentrations.
A climatological model of long-range transport of SCL and sulfate has
been also developed in this research project in order to calculate long-term
distributions of S02 and sulfate concentrations as well as the acidity of
precipitation. The budget of sulfur over eastern North America can also be
studied.
The model has been applied to calculate the distribution patterns of
monthly-averaged concentrations of S0? and sulfate for the months of January
1977 and March 1979 over the area between 35°N and 55°N and between 62°W and
95°W. The calculated concentrations for the month of January 1977 were
compared with the observed concentrations, resulting in statistically sig-
nificant correlations between the observed concentrations and the calculated
concentrations. For the month of March 1979, calculated pH values were
compared with those observed at 17 stations located over the eastern United
States. Highly statistically significant correlation coefficients between
these variables were obtained.
i i i
-------
According to the calculation of sulfur budgets, major portions of
sulfur emitted over eastern North America were removed by wet and dry de-
position over the region. The contribution of transported sulfur from the
United States to Canada is substantial for the sulfur over Canada. On the
contrary, transported amounts of sulfur from Canada to the United States
were small. The fraction of outflow amount of sulfur to the Atlantic Ocean
was substantially smaller than the previously estimated amount by Galloway
and Whelpdale (1980).
From empirical studies of precipitation chemistry data, statistically
significant correlations between observed pH and the contents of S0~ and N03
ions were found. Inclusions of NO /NO- transport and removal processes in
X *j
the long-range transport model would have improved the predictability of
precipitation acidity.
IV
-------
CONTENTS
ABSTRACT iii
FIGURES vi i
TABLES x
ACKNOWLEDGMENT xi
1. Introduction I
2. Summary and Conclusion 3
3. Improvement of the Model for 24-Hour Average Concentrations ... 5
3.1 Mixing Layer Height 5
3.2 Interpolation Schemes of Concentrations for Grid Points . . 5
3.3 Increase of the Number of Trajectories Per Day 7
3.4 Removal and Transformation Terms 7
4. Results of Model Applications for 24-Hour Average Concentrations
of S02 and of Sulfate 10
4.1 Results and Discussion 10
5. Development of a Climatological Model of S0? and Sulfate
Transport 23
5.1 Introduction 23
5.2 Model 23
6. Results of Climatological Model Applications to Eastern North
America 34
6.1 Introduction 34
6.2 Results 38
6.3 Comparison of Calculated Results with Observations 38
-------
6.4 Mass Budget of Sulfur Over the Northeastern United States . 57
7. Investigation of a Prediction Method of the Acidity of
Precipitation 62
7.1 Introduction 62
References 74
-------
FIGURES
Number Page
1 Scheme of calculating 24-hour average concentrations.
Here M is the number of days and K is the number of
trajectories in a day 8
2 Locations and intensities of SO- emission sources
(x 103 ton/year) 11
3 Distribution of SO,, concentrations (ug/m ) at the surface
level on (a) January 25, 1976 and (b) July 11, 1976 12
3
4 Distribution of sulfate concentrations (ug/m ) at the
surface level on (a) Janaury 25, 1976 and (b) July 11, 1976 . 13
5 Distribution of 50^ deposition amounts due to dry and
wet deposition on (a) January 25, 1976 and (b) July 25,
1976 15
2
6 Distribution of sulfate deposition amounts (kg/km ) due
to dry and wet deposition on (a) January 25, 1976 and (b)
July 11, 1976 16
2
7 Distribution of S0? deposition amounts (kg/km ) due to
precipitation on (a) January 25, 1976 and (b) July 11, 1976 . 17
2
8 Distribution of sulfate deposition amounts (kg/km ) due
to precipitation on (a) January 25, 1976 and (b) July 11,
1976 18
9 Relationship between calculated and observed S0? concentra-
tions on (a) January 25, 1976 and (b) July 11, 1976 19
10 Relationship between calculated and observed sulfate con-
centrations on (a) January 25, 1976 and (b) July 11, 1976 . . 20
Ha Average values of the dispersion parameter of I and its
s\
standard deviation in the x direction due to the meandering
of trajectories as a function of time 27
VI 1
-------
Number
lib
12a
12b
13
14
15a
15b
16a
16b
17a
17b
18a
18b
19a
19b
20a
20b
21a
21b
22a
22b
Same as Fig. lla, except for y direction
Average values of the dispersion parameter of a and its
/s
standard deviation due to vertical wind shear
Same as Fig. 12a, except for a
Division of the region into four major areas. For the
U.S. the grid points are indexed by 1, Canada by 2, the
Great Lakes by 3, and the Atlantic Ocean by 4
Locations and intensity of SO- emission sources
3
(x 10 ton/year)
Monthly precipitation amounts (mm) for January, 1977 ....
Distribution of monthly precipitation amounts (mm) for
March 1979
3
Distribution of SO- concentrations (pg/m ) for January 1977 .
3
Distribution of SO- concentrations (ug/m ) for March 1979 . .
Distribution of sulfate concentrations (ug/m ) for January
1977
Distribution of sulfate concentrations (ug/m ) for March
1979
2
Dry deposition amounts (kg/km ) of SO^ for January 1977 . . .
2
Dry deposition amounts (kg/km ) of SO- for March 1979 ....
2
Dry deposition amounts (kg/km ) of sulfate for January 1977 .
2
Dry deposition amounts (kg/km ) of sulfate for March 1979 . .
2
Wet deposition amounts (kg/km ) of SO- for January 1977 . . .
2
Wet deposition amounts (kg/km ) of SO- for March 1979 ....
2
Wet deposition amounts (kg/km ) of sulfate for January 1977 .
2
Wet deposition amounts (kg/km ) of sulfate for March 1979 . .
pH distribution for January 1977
oH distribution for March 1979
Page
28
29
30
32
35
36
37
39
40
41
42
43
44
45
46
47
48
49
50
51
52
VI 1 1
-------
Number Page
23 Observed concentration distribution of SCL for January,
1977. [From Bhumralkar et al. (1980).] 53
24 Observed concentration distribution of sulfate for January,
1977. [From Bhumralkar et al. (1980).] 54
25 Calculated S0« concentration versus observed SO- concentra-
tion for January 1977 55
26 Calculated sulfate concentration versus observed sulfate
concentration for January 1977 56
27 Diagram showing the relationship between observed pH and
calculated pH 59
28 Concentrations of SO. ions and NO, ions in precipitation
water, (A) data from "Atmospheric Turbidity" and "Pre-
cipitation Chemistry Data for the World" and (B) data
from Hales and Dana (1979) 63
29 Same as Fig. 28, except S0~ ions and NH, ions 64
30 Same as Fig. 28, except NO, ions and NhL ions 65
31 Emission rates of S0? and that of NO . Data obtained
from EPA 67
32 Observed pH and pH calculated using Granat's (1972)
theory. "Atmospheric Turbidity" and "Precipitation Data
for the World", 1974, were used" 69
33 Observed pH and pH calculated from -log,Q[2SO~].
(A) "Atmospheric Turbidity" and "Precipitation Chemistry
Data for the World". (B) Data from Hales and Dana (1979) . . 70
34 Observed pH and pH calculated from log1Q (2[SO^] + [N0~]}.
(A) "Atmospheric Turbidity" and Precipitation Chemistry
Data for the World", (B) data from Hales and Dana (1979). . . 71
IX
-------
TABLES
Number Page
1 Correlation coefficient between the observed concentrations
and the calculated concentrations 21
2 Sulfur budget of region 21
3 Site name, geographical location and monthly average pH for
March 1979 58
4 Sulfur mass budget for January 1977 60
5 Sulfur mass budget for March 1979 61
6 Correlation coefficients 62
7 Correlation coefficients 68
8 Correlation coefficient and regression parameters of the
relationship between pH, -log^JCNOg] + 2[SO~]} and
-log10{2[Su=]} 73
-------
ACKNOWLEDGMENTS
Part of the calculations reported in this document was conducted on the
EPA computer. Mr. Adrian Busse of the Environmental Sciences Research
Laboratory was very helpful in running our program on the computer. Thanks
are also due to the project officer, Mr. G.C. Holzworth, for giving us con-
structive suggestions.
The trajectory calculation portion of our models in this report is an
expansion and modification of an original scheme developed by Mr. J.L.
Heffter, Air Resources Laboratories, NOAA at Silver Springs, Maryland.
XI
-------
SECTION 1
INTRODUCTION
There is growing public, scientific, and governmental concern over the
long-range transport and transformation of sulfur oxides and nitric oxides
and other industrial pollutants, and their subsequent consequences for human
welfare and for the environment, notably the increase of acidity of precipi-
tation. Over the northeastern United States and Scandinavia, precipitation
with high acidity, which is caused by the scavenging removal of such pol-
lutants as SO-, sulfate and nitrate, has been observed. The average pH of
the Adirondack lakes has already dropped from 6.5 in the 1930's to 4.8 now;
more than 90 of those lakes are completely fishless (Science News, 1979).
In the last several years, our major research efforts have been to
develop a long-range transport model of S0? and sulfate which, as input,
uses the data observed routinely at weather stations, and to apply the model
for studying the air quality of the eastern United States. Our model can be
classified as a Lagrangian, forward-trajectory model. The most prominent
feature of the model is that trajectories of mean winds of the nighttime
stable layer and daytime mixing layer, and trajectories of the mean winds of
the layer between daytime mixing height and nighttime stable layer height
are taken into consideration. The model has been applied to calculate 'the
geographical distributions of 24-hour average concentrations of SOp and
sulfate over the region between 35°N and 45°N and between 75°W and 95°W,
which encompasses the Ohio River basin.
Although statistically significant correlations between observed con-
centrations and calculated concentrations are obtained both for SOp and for
sulfate, the results also showed the limitation of applicability of the
model to the calculation of 24-hour average concentrations. The model can
be used to calculate overall patterns of distribution, but it is improper to
simulate minor details of such patterns. The details of our model and the
results of its application were reported by Henmi and Reiter (1979) and
Henmi (1979 and 1980).
Further improvements of the model have been made for the calculation of
24-hour average concentrations. Details of improvements are described in
Section 3. The results of application of the improved model are given in
Section 4.
It is possible to calculate long-term (monthly, seasonal or annual)
average concentrations by repeatedly using the model developed for the
-------
24-hour average concentrations. However, due to the large amount of
computing time required for this process it is not practical to do so.
Realizing this disadvantage, we have developed a model which is suit-
able for calculating long-term average concentrations. This model has been
applied to compute the distribution of average acidity of monthly precipita-
tion over the region of eastern North America and to estimate the budget of
sulfur over the region. This climatological model and the results of its
applications are described in Sections 5 and 6, respectively.
In order to find empirical relationships between the acidity of precip-
itation and the concentrations of ions, we have analyzed available data of
precipitation chemistry. The results of the analysis are described in
Section 7.
-------
SECTION 2
SUMMARY AND CONCLUSION
In this report we have described the improvements made in our
long-range transport model of SO,, and sulfate (Henmi and Reiter, 1979).
Trajectories of the mean wind of the mixing layer are calculated using the
wind and temperature sounding data. The heights of the daytime and night-
time mixing layer are determined by temperature sounding data. In order to
save computing time, interpolation schemes of concentrations along the
trajectories into grid points, as well as the interpolation scheme of pre-
cipitation rate at meteorological stations into the midpoint of trajectory
segments have been modified, resulting in substantial savings of computing
time. Additional improvements include the change of the magnitude of re-
moval and transformation terms and the increase in the number of trajector-
ies per day.
The model has been applied to calculate the distribution patterns of
concentrations and deposition amounts of SO,, and sulfate for the dates of
January 25 and July 11, 1976. Correlations between the calculated con-
centrations of S02 and sulfate and the observed concentrations of these
compounds are examined. Although statistically significant correlations
between the observed concentrations and the calculated concentrations are
obtained both for SO^ and for sulfate, the results also show the limitation
of applicability of the model to the calculation of 24-hour average concen-
trations. The model can be used to calculate the overall patterns of dis-
tribution, but it is improper to simulate the minor details of such pat-
terns. This limitation may be mainly due to the following facts:
(1) Annual emission data of S0?, instead of the daily emission data,
were used as input data.
(2) Small emission sources were neglected. Thus, the local effects
due to these sources were not included in the distribution pat-
terns.
(3) Measurements of SO- and sulfate may be subject to errors. Fur-
thermore, measurements may reflect local rather than regional
effects.
-------
(4) Model parameters, such as dry and wet deposition velocities and
the transformation rate of SCL to sulfate, may need further im-
provement.
(5) There may be inaccuracies in trajectory calculations. During the
periods of calculations for January 25 and July 11, 1976, the
region was under the influence of frontal systems. Trajectories
leaving the area near the fronts may have been inappropriately
calculated.
(6) The model does not consider the effect of relatively short-range
dispersion, which in s,ome cases may have significantly affected
the observed concentrations.
In order to calculate long-term distribution of S0? and sulfate concen-
trations, as well as the acidity of precipitation, a climatological model of
long-range transport of SO- and sulfate has been developed. The model can
be applied to study the budget of sulfur over eastern North America.
The model has been applied to calculate monthly-average concentrations
of SO- and sulfate for the months of January 1977 and March 1979. The cal-
culated concentrations of SO- and sulfate for the month of January 1977 were
compared with the observed concentrations over the eastern United States,
resulting in statistically significant correlations between the observed
concentrations and the calculated concentrations. For the month of March
1979, calculated pH values were compared with those observed at 17 stations
located over the eastern United States. Highly statistically significant
correlation coefficients between these variables were obtained.
The calculation of sulfur budgets showed that major portions of sulfur
emitted over eastern North America were removed by wet and dry deposition
over the region. The contribution of transported sulfur from the United
States to Canada is substantial for the sulfur over Canada. On the con-
trary, transported amounts of sulfur from Canada to the United States are
small. The fraction of outflow amount of sulfur to the Atlantic Ocean was
substantially smaller than the previously estimated amount by Galloway and
Whelpdale (1980).
From empirical studies of precipitation chemistry data, it was con-
cluded that there were statistically significant correlations between ob-
served pH and the contents of SOT and NO- ions. Therefore, it was possible
to predict the acidity of precipitation from our model. Inclusions of
NO /NO- transport and removal processes in our long-range transport model
X O
would have improved the predictability of precipitation acidity.
-------
SECTION 3
IMPROVEMENT OF THE MODEL FOR 24-HOUR AVERAGE CONCENTRATIONS
MIXING LAYER HEIGHT
The model developed previously used the wind data for 1974 provided by
the Air Resources Laboratories, NOAA at Silver Springs, Maryland. In order
to run the model using wind and temperature sounding data for the years 1975,
1976 and 1977, which were obtained from the Air Resources Laboratories, the
computer programs of the model had to be modified substantially. In the
previous model, the daytime mixing layer height and the nighttime stable
layer height were defined from climatological data, and trajectories of the
mean wind of the layers were calculated. In contrast, the present version of
the model incorporates vertical temperature profiles along a trajectory to
determine mixing layer depths over which average transport winds are cal-
culated. The top of the daytime mixing layer is defined as the base of any
nonsurface-based temperature inversion. A maximum inversion height is chosen
as 3000 m. If no inversion occurs below 3000 m, this height is used for the
top of the afternoon mixing layer and winds are averaged over that layer.
The top of the nighttime stable layer is defined as the top of the surface-
based temperature inversion. If no surface based inversion occurs, 500 m is
substituted as the top of the nighttime stable layer.
INTERPOLATION SCHEMES OF CONCENTRATIONS FOR GRID POINTS
In the model previously developed (Henmi and Reiter, 1979), more than 90
percent of computing time was spent on interpolation schemes of concentra-
tions along the trajectories into grid points. In order to save computing
time, the computer program has been modified so that when the concentrations
of S09 and sulfate along the trajectory become lower than certain values
3
(i.e., 0.5 ug/m ), the loop containing the interpolation schemes of concen-
trations along trajectories into grid points was bypassed. The result of a
test run of the model showed that the distribution patterns of S0? and sul-
fate were not different from those obtained without bypassing the loop.
An additional modification of the model has been made in the interpola-
tion scheme of precipitation rates observed at meteorological stations into
the midpoint of each trajectory segment. In the previous version of the
model (Henmi and Reiter, 1979), an interpolation routine (i.e., triangle
method) had been used. In the present model, the interpolation scheme used
is a combination of inverse distance-squared weighting and the fitting of a
linear surface so that the main advantage of each is retained - namely the
-------
dominance of observations which are spatially close to the interpolation
point, as well as the general distribution obtained from a number of obser-
vations. This scheme was originally used by English (1973).
The interpolation scheme consists of fitting a linear surface to the
midpoint of a trajectory segment, with inverse distance- squared weighting
applied to the observations. The surface chosen is the one for which the
expression
I = I* + by. + c
d d d d'
applied to a number of precipitation rates observed at meteorological sta-
tions, has a least-square solution such that
N
I-* r. - (ax. + by. + c) = minimum (2)
i=l df 1 n 1
In these equations
d is the distance of a meteorological station from the midpoint
of a trajectory segment,
x and y are the longitudinal and latitudinal distances of a
meteorological station from the midpoint of a trajectory segment,
r is the precipitation rate at a meteorological station, and a, b,
c are coefficients.
For each of the N chosen observations one can write
*
d. - d. d. d.
where i = 1 to N.
To interpolate the precipitation rates for the midpoint of trajectory
segments, the present scheme chooses such rates observed at six meteorologi-
cal stations which are located close to the midpoint of a trajectory segment.
Interpolation into areas of no rainfall often leads to negative values, which
are then set to zero. According to the precipitation field analysis by
English (1973), six values give the most acceptable results, a pattern simi-
lar to what an analyst might produce by hand. A number less than six often
leads to discontinuities and irregularities in the fields, whilst a number
more than six tends to produce too much smoothing and also consumes more
computer time.
The application of the above scheme has reduced the computing time of
concentration calculations to 30 percent of that of the previous version of
the model .
-------
INCREASE OF THE NUMBER OF TRAJECTORIES PER DAY
In the previous model, trajectories from each source area were cal-
culated every 12 hours, starting at 6 and 18 CST, and each trajectory was
pursued for 48 hours. In the present model, in order to increase the ac-
curacy of calculations, four trajectories per day starting at 6, 12, 18 and
24 CST from each source area are calculated, and each trajectory is pursued
for 48 hours. A trajectory is composed of a series of three-hour segments.
In order to calculate 24-hour average concentrations, we must take into
consideration the contributions of trajectories which started up to 2 days
earlier. In Fig. 1, the scheme for calculating 24-hour concentrations is
shown. Trajectory segments drawn in thick solid lines were used to calculate
the 24-hour average concentrations.
REMOVAL AND TRANSFORMATION TERMS
Dry Deposition
Recent results of dry deposition velocity estimates for SOp and sulfate
were reviewed by Garland (1978 and 1979). He concluded that the dry deposi-
tion velocity of S02 was about 0.8 cm/sec and that of sulfate was 0.1 cm/sec.
In accordance with these recent results, we use in the present model I cm/sec
for the dry deposition velocity of S0« and 0.1 cm/sec for that of sulfate.
Wet Deposition
In the previous application, it was assumed that (K/X) was 5 x 10 for
both SOp and sulfate. Here, K is the concentration of pollutants in rain-
water and x the concentration of pollutants in air. The subscript v means
that the ratio (K/X) is formed on a volume basis. Recent observational
results (Garland, 1978 and Hales and Dana, 1979) show that S02 removal by
rain is about an order of magnitude less efficient than sulfate removal.
Therefore, we use the following scavenging velocities:
v = (£) . p = i x 104 x P for SO (4)
x v
v = (-) P = 1 x 105 x P for sulfate. (5)
* v
Here P is the precipitation rate (cm/hr). It has been reported that the
sulfate content in precipitation varies depending on the season with a sum-
mertime peak (Pack, 1978 and Hales and Dana, 1979). The mechanism of pre-
cipitation formation and the condition of the environment may change with
season. Rain can bring down dissolved gases. In the ice phase processes of
precipitation formation, the growth of ice is a purification process and
dissolved gases will be released back into cloud water. This feature can not
be included in the present model and field studies on the role of ice in the
-------
to
r~
^
C
03
"O
CO
CO--
00
ro
0
C\J
Ul
o>
O)
ui
C
o
ra
+j
c
0)
(J
c
o
u
OJ
O)
(0
>
18
m
5-
U
-
0)
E
HI
Ol
-------
scavenging processes will be needed for clues to an understanding of the
seasonal variation of the amount of sulfate in precipitation.
Transformation
The transformation rates of SCL to sulfate depends on sunlight intensity
and also appears to depend on water vapor concentration, back-ground ozone
levels and the extent to which the plume has mixed with background air (Husar
et a!., 1978 and Wilson, 1978). Husar et al. reported that during the noon
hours of summer the transformation rate is 0.01 ~ 0.04 hr and below 0.005
hr during the night. In accordance with these values and based on the
results of preliminary calculation we select 0.05 hr during the daytime and
0.001 hr during the night in the present model.
-------
SECTION 4
RESULTS OF MODEL APPLICATIONS FOR 24-HOUR AVERAGE
CONCENTRATIONS OF S02 AND OF SULFATE
In this section we describe the results of model applications. We
calculated the geographical distributions of the 24-hour average concentra-
tions of SOp and of sulfate over the region between 35°N and 45°N and between
75°W and 95°W, which encompasses the Ohio River basin. Part of the computa-
tions was conducted on the EPA computer. The dates chosen for computations
were January 25, and July 11, 1976 on which the observed concentrations of
SO- and sulfate at stations located in the region were abundant.
INPUT DATA
5
Sixty point sources of S0? whose emission rate is more than 10 ton/year
were taken into consideration. The geographical locations and the emission
rates are shown in Fig. 2 (Clark, 1979). The emission of S02 from these
sources contains about 90% of the total emissions over the area.
Upper-air data, which were prepared by ARL, NOAA and purchased from the
NMC, contain upper air winds, temperature, and heights from rawinsonde sta-
tions for North America from the surface to 500 mb. Station identification
information, including average terrain height at each station, and observed
meteorological data were recorded for four observation times per day (00, 06,
12, and 18Z).
The precipitation rate data for 81 stations located in the region were
used to calculate the wet deposition amounts.
RESULTS AND DISCUSSION
The distributions of 24-hour average concentrations of S0? are shown in
Figs. 3a and b. Figure 3a is for the date of January 25, 1976 and 3b for the
date of July 11, 1976. On January 25 high concentrations are found in the
northeastern part of the region. On the other hand, the southeastern part of
the region was covered by high concentrations of S0? on July 11. Predominant
winds over the region on these dates were respectively from the southwest and
from the northwest.
Figures 4a and b contain the distributions of sulfate concentration at
the surface level. Again, high concentrations of sulfate can be found in the
10
-------
o:
01
c
o
CO
o
in
o>
u
J-
3
O
1/1
c
o
in
in
QJ
eg
O
ui
0)
to
c
O)
-p
c
in
c
o
ro
U
O
CM
11
-------
44
42
94
90 88 86
82 80 78
94
44
42
90 88 86 84 82 80
78
Fig. 3. Distribution of S0~ concentrations (ng/m"3) at the surface level on
(a) January 25, 1976 and (b) July 11, 1976.
12
-------
A. 94
44 -
42
44
42
94
92
9° 88 86 84 82 80
78
B. 94
44 -
42
94
44
88 86
84 82
80
76
Fig. 4. Distribution of sulfate concentrations (ug/m ) at the surface level
on (a) Janaury 25, 1976 and (b) July 11, 1976.
13
-------
northeastern part and in the southeastern part of the region on January 25
and July 11, respectively.
In both Fig. 3 and Fig. 4, it can be seen that the western part of the
region was covered by clear air.
The distributions of deposition amounts of SO- due to dry and wet de-
positions are shown in Figs. 5a and b. Figures 6a and b contain the distri-
butions of deposition amounts of sulfate due to dry and wet depositions.
It can be seen from Figs. 7a and b that the removal of SCL due to pre-
cipitation is inefficient. At the present time we are not certain how ap-
propriate these results are due to the nonexistance of observed data of SCL
content in precipitation water. In the present model, the precipitation
scavenging velocity is given by
A
v = 1 x 10 x P cm/sec fc^
s (b)
where P is the precipitation rate (cm/hr). The validity of this assumption
will be examined in the future.
Figures 8a and b show the distribution of wet deposition amounts (kg/
2
km ) of sulfate. These amounts are substantial compared with those of SCL.
In order to verify the performance of the model, concentrations of SCL
and sulfate calculated for the surface level were compared with concentra-
tions observed at stations located in the region. The 24-hour average con-
centrations of SCL and sulfate from observed data were supplied by EPA. The
observed concentrations of SCL are plotted against the calculated ones in
Figs. 9a and b. Similar figures for sulfate are given in Figs. lOa and b.
From these figures it can be seen that the calculated SCL concentrations
are generally too low compared to the observed concentrations, and that the
caluclated sulfate concentrations are generally too high. These facts clear-
ly indicate that the transformation rate from SCL to sulfate used in the
present calculations is too large.
Correlation coefficients between the observed concentrations and the
calculated concentrations are given in Table 1.
14
-------
44
94
92
90 88 86 84 82 80
76
44
42
94
Fig. 5. Distribution of SCL deposition amounts due to dry and wet deposi-
tion on (a) January 25, 1976 and (b) July 25, 1976.
15
-------
A. 94
44 -
44
42
94
76
76
44
42
84 82 80
78
76
Fig. 6. Distribution of sulfate deposition amounts (kg/km ) due to dry and
wet deposition on (a) January 25, 1976 and (b) July 11, 1976.
16
-------
44
42
40
83 86 84 82
94
84 82 80
78
76
44
94
76
Fig. 7. Distribution of S0? deposition amounts (kg/km ) due to precipita-
tion on (a) January 25, 1976 and (b) July 11, 1976.
17
-------
A. 94
44 -
42
44
42
94
84 8S 80
-\ 44
42
40
94
76
Fig. 8. Distribution of sulfate deposition amounts (kg/km^) due to precipi-
tation on (a) January 25, 1976 and (b) July 11, 1976.
18
-------
SJ-
sr
II
t-
2
UJ
O
u_
u_
UJ
o
o
H o
h-
UJ
?J IT
=> O
T5 O
i i i I
O
cd
<
»
*
.
4111
O
N
<
H
-|
-
_
_
-
_
-
-
'.
.
...,., .. «»»«
» » i
i
* * f*^Vt
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ' I 1 , I ,2m ' \
O O O O O o C
(O If) T IO OJ
(£ui/DUJ)2os JO NOIlVdlN30NOD Q3AJd3SaO
00
H
3P
^.
UJ
o
~*
u_
CO LL.
fs» UJ
0) 0
o
in z
CM g
>u_
p^
cc u ....
< UJ
D CE
< 0
-3 O
* 1
*
«
t * 4 -
t 1 J J. j t 1 tlllt \1\\\ l.t»IJl«4IIJi«\|-lw^> l^^ j
O O O O O o C
CD in *r ro oj
(£UJ/6UJ)20S JO NOIlVyiN30NOO Q3Ad3S80
o
in
o
0
ro
O
cvj
O
3
O
in
0
O
to
8
o
_
3
2
O
5
cc
1
2
UJ
0
^
-C.
o
o
Q
UJ
^J
rr
o
^J
0
^f
0
1-
^1
cc
t-
UJ
o
o
Q
LJ
D
y
^r
o
fO
E
X.
CM
O
CO
U_
O
ro
e
o>
E
OJ
O
CO
u.
o
-o
C
fO
r-
cn
r-H
ID
CNJ
^.
i.
3
C
(O
"5
(0
C
0
Ul
C
o
-p
s_
+J
C
cu
C
o
(J
c\
o
o
0)
s_
10
o
o
C
T3
0)
fO
^
u
(J
C
0>
OJ
+J
Ol
Q-
r
I/I
C
o
r
-P
(0
Qi
,
cn
O)
j
CT>
r 1
l 1
^
3
i^
^^
c5
19
-------
A.
z
O
30
__
^
E
*
o>
3 20
Z UJ
Q
S
en
CD
O
I0
JANUARY 25, 1976
CORRECTION COEFFICIENT
= 77
B.
u.
O
Z
O
30 -
< E
cc \
I- o>
SI 20
Z UJ
UJ
en
CD
O
10
__
10 20 30 40 50
CALCULATED CONCENTRATION OF SULFATE
(mg/m3)
JULY II, 1976
CORRECTION COEFFICIENT = .85
Fig. 10.
0 10 20 30 40 50
CALCULATED CONCENTRATION OF SULFATE
(mg/m3)
Relationship between calculated and observed sulfate concentrations
on (a) January 25, 1976 and (b) July 11, 1976.
20
-------
Table 1. Correlation coefficient between the observed concentrations
and the calculated concentrations.
January 25 July 11
S02 r=0.68 (rc=.202, N=162) r=0.44 (rc=.176, N=216)
Sulfate r=0.77 (^=.205, N=155) r=0.85 (r^O.278, N=86)
In this table, r is the correlation coefficient, r is the critical
value for correlation coefficients with 99.9% confidence levels and N is the
number of data pairs. Although the calculated concentrations are statisti-
cally related to the observed concentrations for SO,, as well as for sulfate,
the values of the correlation coefficients are not significantly high. In
particular, the correlation coefficients between the observed SCL concentra-
tions and the calculated concentrations are disappointingly low. This low
correlation can be partially attributed in addition to the inappropriate
value of the transformation rate to the following fact: in the pre-
vious application (Henmi and Reiter, 1979), the data taken at stations
classified as "rural", "remote", and "suburban-residential" were used for
comparison of S0? concentrations. Such a screening procedure of the data was
not conducted in the present study.
Sulfur budgets over the region for January 25, 1976 and July 11, 1976
were calculated as Table 2.
Table 2. Sulfur budget of the region.
January 25, 1976 July 11, 1976
Sulfur Emission 24,492 tons
Removal by Wet Deposition
as S02
as Sulfate
Removal by Dry Deposition
as SCL
as Sulfate
Total Deposition of Sulfur
Amount Exported Across the
Boundaries of the Region
90 tons
718
808
9,240
927
10,167
10,975
13,517
21
129 tons
1,886
2,015
6,406
1,089
7,495
9,510
14, 982
-------
Table 2 shows that more than 50 percent of sulfur is exported across
the boundaries of the region for both dates. Significant differences of the
present calculation from the previous estimates (Henmi and Reiter, 1979) are
found in the removal due to precipitation. In the previous calculation,
about 30 percent of sulfur was removed by precipitation, whereas less than 10
percent is thus removed in the present calculation. The change of the wet
removal formula in the model is the major reason for this drastic decrease in
the removal amount.
22
-------
SECTION 5
DEVELOPMENT OF A CLIMATOLOGICAL MODEL OF S02 AND SULFATE TRANSPORT
INTRODUCTION
One of the purposes of this project was to develop a climatological
model of S0« and sulfate transport and transformation. Our intention was to
develop a model which was economical in computing time and which could be
applied to the studies of the distribution of acidity of precipitation over
the area of eastern North America and of the transport and deposition of
sulfur across the national boundaries between the U.S. and Canada.
In this section, we describe the details of the climatological model
developed.
MODEL
Trajectory Calculations
In this climatological model, trajectories of air parcels are computed
(using average wind speeds and directions of the mixing layer) four times a
day from each source for 30 days. The locations of the endpoints of the
trajectory segments are calculated at 3-hour intervals. Each air parcel is
tracked up to 3 days. In the model developed for calculating 24-hour average
concentration (see Section 3), the vertical temperature profiles along a
trajectory were considered to determine mixing layer depths over which aver-
age transport winds were calculated. Furthermore, trajectories of average
winds of nighttime and daytime mixing layers, as well as trajectories of
average winds of the layer between daytime mixing height and nighttime mixing
height, in which pollutants were trapped during the nighttime, were taken
into consideration.
In contrast, trajectories of average winds of the daytime mixing layer
defined from climatological data (Holzworth, 1972) are calculated in this
model. The reasons for this simplification are as follows: The error of
trajectory calculations due to the use of the climatological height of the
mixing layer over the area of interest, instead of the use of mixing height
determined from soundings along the individual trajectories may be cancelled
out when the average locations of numerous trajectories are calculated, and a
substantial savings can be realized in computing time.
The coordinates of the endpoints of the trajectory segments can be
expressed as follows:
23
-------
V
Y«_i jAt (8)
where At is the time interval, U and V are the velocity components in X
(west-east) and Y (south-north) direction, j=l , 2 .... N are trajectory
endpoints at the time of t=At, 2At, ...., NAt from the start of the cal-
culation, £ is the label for each trajectory. In the application described
in the next section, At, and N are 3 hours and 24 hours respectively, and the
total number of trajectories from each source is 120 (4x30).
The coordinates of the average trajectory of air parcels leaving from
each source area are expressed as:
Y. = I Y. (10)
J L £=1 J*
where L is equal to 120 in the present application.
It is assumed that S0? and sulfate emitted from the source are trans-
ported and transformed along the average trajectory and that they are dis-
persed into the X- and Y- directions from the average trajectory as described
in the next section. Mixing in the vertical direction is assumed uniform
throughout the layer.
Horizontal Dispersion
In the model the long-term average plume from a pollutant source is
approximated by a series of puffs. Horizontal dispersion of pollutants from
the center of a puff can be expressed as:
D . = I . + a - (11)
x,J x,j x,j
D . = I . + a . (12)
y,j y,j y,j
Here, D . and D . are the total dispersion of pollutants in the x and y
x > j y » j
direction, respectively. I . and I . are the dispersion due to the mean-
x > j y > j
dering of each trajectory, and a . and a . are the dispersion due to the
x > j y > j
vertical shear of wind. These dispersion parameters can be written as fol-
lows:
24
-------
z .. . -^i
x,j " ' L (13)
(Y.u - V2
y.j " ' L (14)
a . and a . are described by
x > j y > j
j .
a . = I a .-At (15)
x,J j=1 u,j
a = 1 a -At (16)
y > j ---.-i v > j
where
-------
In this model, the dispersion due to instantaneous mixing is neglected. The
dispersion parameters I . and I . and a . and a . are calculated for the
x > j y»j x, j y»j
plumes from each pollutant source.
In the application described in the following section, 120 trajectories
(4 trajectories per day for 30 days) from each source were calculated for the
duration of 72 hours. In application of this climatological model, trajec-
tories from 72 major sources of S02 over the region of eastern North America
were computed using wind data for the month of January, 1967. In Figs, lla
and lib, and Figs. 12a and 12b, the averages of I and I , and of cr and a
x y x y
for trajectories from 72 major sources are shown. In these figures, thick
lines represent the average values and thin lines are the standard devia-
tions.
In comparing Figs, lla and lib, it can be seen that the dispersion due
to meandering of the trajectories is substantially greater in the x direction
than in the y direction. Furthermore, Figs. 12a and 12b show that the dis-
persion due to wind shear is comparable to that due to meandering of the tra-
jectories.
Concentration Distribution
It is assumed that in a horizontal direction pollutants in a puff are
dispersed according to a Gaussian distribution. Therefore, the horizontal
distribution function of SCL and sulfate is expressed by
Ph(x-x,, y-y.) = 2 p D
n J j 2 Dx Dy ux uy
where x-x., and y-y. are the distances in the x- and y-directions, respec-
J J
tively, from the center of a puff, and D and D have been defined by Eqs.
x y
(11) and (12) in the previous section. A similar approach has been taken by
Sheih (1977).
SO- and sulfate are transformed and deposited along a mean trajectory
according to the following equations.
dQl Vol + Vsl
i r y-1- ->J-T n i/n fT>\
J * ^-.-K.1^-. \C-t-)
dt u h
10,, V n + V
5 ^ t _ rt , O (/ /^ ^^*5^
h 2
where Q, and Q2 are the mass of SCL and sulfate, V , and V 2 are the dry
deposition velocities, V , and V are the wet deposition velocities due to
precipitation, K is the transformation rate, and h is the height of the mix-
ing layer. The subscripts I and 2 stand for SO^ and sulfate.
26
-------
c
o
u
O)
in E
r
m -P
"O
c
c
o
X +->
lA) U
'+- 3
O M-
P
tu
03
i-
-------
ro
CO
to
in
CO
ro
3
O
LU
oo
c
O
(J
O)
s_
O
Q.
OJ
(J
X
o>
Ol
O
O
co
O
O
in
o
o
o
o
ro
o
8
o
o
co
(0
CU
e
00
.0
en
u_
28
-------
u
-p
S-
0)
o
-p
d)
3
fO
(O
10
O
(£> JC
ro ^,
LU
cvj
oo
o
o
(O
o
o
10
o
o
o
o
ro
o
o
C
O
p
m
OJ
a
a
s_
m
T3
c
fO
S-
ai
-P
ai
E
(O
S-
(T3
Q.
C
O
V)
i-
a>
a.
o
HI
o
OJ
> ro
Ol
HI -C
D) en
ro
1 ^^
OJ d
IT3
CsJ
C35
29
-------
co
8
O
O
CM
O
O
C\J
to
CO
to
in
* .
a)
i_
D
O
co x:
to ^,
LU
00
s-
o
O
X
CD
to
CM
cn
in
ro
Ol
E
itJ
CSJ
O)
30
-------
Concentrations of S0» and sulfate, C, and C~, can therefore be expressed
by ILL
C1=TT' PhCx-Xj, y-yj> (24)
C2 = F- Ph(x-xj, yyj) (25)
The details on dry deposition velocity, wet deposition velocity, and trans-
formation rate have been described in Section 3 and elsehwere (Henmi, 1979
and 1980; Henmi and Reiter, 1979).
Concentrations at the grid point (x, y) are calculated by
N M
C,(x,y) =11 C (26)
N M
C (x,y) =11 C (27)
^ j=l 1=1 ^1J
where C,(x,y) and C2(x,y) are the concentrations of SOp and sulfate at the
grid point (x,y), respectively, the subscript i represents the source number,
and j stands for the number of puffs. In this study, N is 24 and the total
number of sources, M, is 72.
Deposition Amount
Concentration distributions of S0? and sulfate are calculated over the
region between 35°N and 53°N and between 62°W and 95°W. In order to estimate
the deposition amount of sulfur, the gridded regions are divided into four
major areas as shown in Fig. 13: the United States, Canada, the Great Lakes
and the Atlantic Ocean.
The dry and wet deposition amounts of sulfur, 0 , and D , for each grid
point (x,y) are calculated as
C,(x,y) C (x,y)
D Cx v") = V - + V - (28}
Vx'y; gl 2 g2 3 U°J
C-,(x,y) C?(x,y)
W-y> = vsi ~^Y- + Vs2 -V- (29)
The total amount of sulfur deposition for each region is, therefore, ex-
pressed by
Dd T = J Dd dS = ZDd(x,y) -Ax-Ay (30)
DwJ = J Dw dS = !Dw(x,y) -Ax-Ay (31)
31
-------
v X
N
- x
-. .x ^
X >.
X x *X'
' "" . " '" x ""' X XX" "-' ", N
, . - > . > -
N -N. -^ X -\ -V -v >. .^. X ->-. ^s -V.
X X- -. - . -. X N, -.X 'N % ~v N
^ -X^V^X t '^^"
'
.,
^M
^
,<
~ ~
'f
-. X
>-*
'
~
"*
_,
'
-. - -r - ~ -
"^
'_
"" E»
\ T
> J J^_
! £ '-" "-"- ' --j
- - -
. ,
0)
X
OJ
-a
c
O)
4-i
c
o
Q.
i-
O)
dJ
00 JD
ID c
03
QJ OJ
-C U
-U O
S- U
O -r-
U_ -P
c
fD
en -p
C
^ fQ
E
O >,
O *"
4j ai
cn
01
s_
O)
a>
ai
o
i- (0
vn "D
r- (O
> c
r- (0
Q C_>
32
-------
where, dS is the area element and Ax and Ay are the grid intervals in the x-
and y-directions, respectively.
Acidity of Precipitation Due to Sulfate
As will be described in Section 7 acidity of precipitation is a very
complex problem. It has been shown that there are statistically significant
correlations between observed pH and the content of sulfate in precipitation.
For this preliminary application of the model, we assume, admittedly crudely,
that the acidity of precipitation is expressed by the wet deposition amount
of sulfate divided by the precipitation amount.
The pH value at grid point (x,y) is given by
pH = - log102[SO;j] = - log10{2(IVs2 C2(x,y)/P(x,y)} (32)
where [SOT] is the molar content of SOT in precipitation water (mole/ liter),
and P(x,y) is the precipitation amount per unit area.
33
-------
CHAPTER 6
RESULTS OF CLIMATOLOGICAL MODEL APPLICATIONS TO
EASTERN NORTH AMERICA
INTRODUCTION
Using the model described in Section 5, we calculated the geographical
distributions of the monthly average concentrations and deposition amounts of
S0? and of sulfate over the region between 35°N and 53°N between 62°W and
95°W, which encompasses eastern North America. Furthermore, the distribu-
tions of acidity of precipitation, and of the deposition amount of sulfur
over the regions of the United States, Canada and the Great Lakes were cal-
culated. Finally, mass budgets of anthropogenic sulfur for the regions were
estimated.
The months chosen for computations were January 1977 and March 1979.
For the month of January 1977, the calculated concentrations of SO,, and
sulfate are compared with those observed in the region. For the month of
March 1979 the calculated pH values are compared with those observed at
stations of the National Atmospheric Deposition Program (Natural Resource
Ecology Laboratory, 1980).
INPUT DATA
Seventy major sources of S0? whose emission rate is more than 10 ton/
year are taken into account. Figure 14 shows the geographical locations and
the emission rates of these S02 sources. This figure was composed from the
emission inventory for the U.S. by Clark (1979) and from the emission in-
ventory for Canada by Voldner et al. (1980).
The monthly precipitation amounts, for approximately 700 stations lo-
cated in the region, were used to calculate the wet deposition amounts. The
distribution of monthly precipitation for January 1977 and March 1979 are
shown in Figs. 15a and b, respectively.
Upper air data, which consist of winds, temperature, and heights from
rawinsonde stations for North America from the surface to 500 mb, were used
to calculate trajectories from each source. The magnetic tapes, which were
prepared by ARL, NOAA and purchased from the NMC, contain observed meteoro-
logical data for four observation times per day (00, 06, 12, and 18Z).
34
-------
(0
c
o
CO
o
CD
u
S-
3
o
c/5
C
o
I
I/)
to
0)
o
OO
IO
C
O)
c
(T3
i/)
C
o
-p
(O
o
o
35
-------
01
TO
c
ro
in
-P
O
m
c
o
m
u
O)
a.
ro
ID
36
-------
CT1
P-.
cn
u
TO
s-
o
in
-p
c
3
o
TO
c
o
-p
TO
u
01
S-
Q.
>i
.C
-(->
c
o
c
o
5-
-P
c/)
Q
JD
un
O)
37
-------
RESULTS
In the following presentation of figures, A is for the case of January
1977 and B for the case of March 1979.
Distributions of the monthly average concentrations of SCL and sulfate
are shown in Figs. 16 and 17, respectively. It can be seen that for both
months, high concentrations of SO- and sulfate are found over the Ohio river
basin and the northeastern region adjacent to Lake Huron where major sources
of SOp are located.
Distributions of dry deposition amounts of S0« and sulfate are shown in
Figs. 18 and 19. Similar figures for the distributions of monthly wet de-
position amounts are shown in Figs. 20 and 21. Finally, Figs. 22a and b show
the distribution of precipitation acidity due to sulfate.
Statistical comparisons of the calculated results with those observed
will be described in the next section.
COMPARISON OF CALCULATED RESULTS WITH OBSERVATIONS
Calculated concentrations of SO^ and sulfate for the month of January
1977 are statistically compared with the observed concentration data compiled
by Bhumralkar et al. (1980). Figure 23 and Fig. 24 are cited from the report
by them. These data were compiled from the SURE data and from the Storage
and Retrieval of Aerometric Data (SAROAD). The SURE air quality data were
compiled by the Environmental Research and Technology, Inc. (ERT) for the
Electric Power Research Institute (EPRI).
Figures 25 and 26 show the relationship between observed concentration
and calculated concentrations for S0? and sulfate, respectively. The cor-
relation coefficients for S02 and sulfate are 0.538 (0.258) and 0.477
(0.318), respectively, where the numbers in parenthesis are critical values
for correlation coefficients with 99 percent confidence level.
Again, although the calculated concentrations are statistically related
to the observed concentrations of S0? as well as for sulfate, the values of
the correlation coefficients are not significantly high.
For the month of March 1979, observed data of SO- and sulfate concentra-
tions are not available at this stage. pH values of precipitation observed
at stations of the National Atmospheric Deposition Program (NADP) are com-
pared with those calculated by the climatological model. The precipitation
chemistry data of NADP are recorded on weekly intervals, so that the monthly-
average pH value is calculated as a geometric mean of weekly values of pH.
The monthly-average pH is calculated using the data observed during the
period between February 27 and April 3.
38
-------
01
TO
3
C
TO
'-3
i.
o
c
o
+J
(0
S-
-p
c
Ol
u
c
o
u
C\J
o
o
^
+J
3
.a
f
1-
^->
1/5
O
39
-------
en
r^
CTl
u
fO
S-
o
D5
3.
in
c
o
fO
s_
-p
c
(U
u
c
o
u
CM
O
to
c
o
I
-p
.a
s_
U1
Q
40
-------
01
3
C
m
S_
o
cn
3.
1/5
C
o
TO
S_
+J
C
01
(J
C
o
u
0)
+J
03
3
ISt
C
O
I/)
.f
o
m
r---
O)
41
-------
01
U
5-
m
Ol
in
c
o
-1-1
fD
S-
(J
c
o
U
(T3
4-
3
Ul
-------
C
(0
1-
o
-------
tj)
r^
01
u
s-
(TJ
S-
o
CM
o
C\J
tn
4->
c
^
o
03
C
O
to
o
Q.
CU
a
-Q
00
44
-------
S-
o
QJ
in
-P
c
3
o
(0
c
o
to
o
Q.
O)
(Ti
45
-------
O1
r--
u
s-
re
S-
o
0)
-p
in
4-
O
CsJ
05
c
3
O
03
C
O
o
Q.
OJ
TO
D5
46
-------
cn
3
c
m
S-
o
CM
O
CM
Ul
-p
C
3
O
E
(0
C
O
O
CL
Ol
T3
(0
O
CM
47
-------
en
r--
CT1
.c
u
TO
i-
o
o
to
CM
l/l
+J
c
3
O
(O
Irt
o
Q.
-------
m
3
C
m
OJ
-P
-------
u
S-
o
it-
's
CNJ
cn
c
3
O
fD
c
O
to
O
Q.
O)
T3
O)
CM
OT
50
-------
C
fO
"3
C
o
S-
+->
I/)
r
T3
3C
Q.
03
CNJ
CM
51
-------
en
!_
O
c
O
t
+J
o
z
Q.
CNJ
52
-------
MEASURED
Figure 8. S02 concentrations (/ag/m3) for January 1977.
Fig. 23 Observed concentration distribution of SCL for January, 1977.
[From Bhumralkar et al. (1980).]
53
-------
Figure 9. SO^concentrations
for January 1977.
Fig. 24 Observed concentration distribution of sulfate for January, 1977.
[From Bhumralkar et al. (1980).]
54
-------
0>
g
o
100
90
80
70
60
50
OJ
B
30
20
10
r = 0.538
10 20 30 40 50 60 70 80 90 100
Observed S02 Concentration (/j.g/m3)
Fig. 25 Calculated SCL concentration versus observed S0? concentration for
January 1977.
55
-------
r» 0.477
5 10 15
Observed Sulfate Concentration
20
Fig. 26 Calculated sulfate concentration versus observed sulfate concentra-
tion for January 1977.
56
-------
For the data from this period, 17 stations were available within the
region of our interval. In Table 3 the location of stations and the monthly
pH values for each station are given.
Using the pH values given in Table 3 and the calculated pH values shown
in Fig. 22b, the diagram showing the relationship between the observed pH
values and the calculated pH values are obtained as Fig. 27. For this
diagram, the correlation coefficient is 0.825 (0.575), where the number in
parenthesis is a critical value for a correlation coefficient with 99 percent
confidence level. From this figure, it can be seen that the acidity calcu-
lated based on sulfate content in precipitation is slightly higher than that
observed. However, from the highly correlated relationship it can be con-
cluded that, with further improvement of the model, the prediction of acidity
of precipitation can be made more accurately than the present results.
MASS BUDGET OF SULFUR OVER THE NORTHEASTERN UNITED STATES
Mass budget of sulfur emitted from the major sources over the region is
estimated using the climatological model. Tables 4 and 5 are for the months
of January 1977 and March 1979, respectively. From these tables the follow-
ing can be seen:
(1) The budgets are different between the two months studied. This
is due to the different meteorological conditions for two periods.
(2) During the month of January 1977, the inflow amounts of sulfur
from the U.S. to Canada are only a small fraction (0.23) of the
total amount of sulfur over eastern Canada. On the contrary, a
substantial fraction (0.40) of sulfur was imported from the U.S.
during the month of March 1979.
(3) Transported amounts of sulfur from Canada to the United States are
small. Less than 0.03 of the total sulfur over the northeastern
United States is the sulfur imported from Canada.
(4) Major portions of sulfur were removed by wet and dry deposition.
(5) The fraction of outflow amounts to the Atlantic Ocean were sub-
stantially smaller than the fraction estimated by Galloway and
Whelpdale (1980). They estimated that approximately 0.26 and 0.21
of the total sulfur were transported to the ocean, respectively,
from Canada and the United States. Our estimation shows that the
fractions of less than 0.03 and 0.1 are blown out to the ocean.
(6) The Great Lakes received approximately 0.05 of the total sulfur
emitted.
The residue of the budget consists of sulfur transported out of the
boundaries and errors of calculation.
57
-------
Table 3 Site name, geographical location and monthly average pH for March
1979.
Site
Name
Bondvi lie
Dixon Springs Ag. Cts.
Wellston
Marcel! Exp. Forest
Lamberton
Hubbard Brook
Huntington Wildlife
Lewiston
Coweeta
Piedmont Research Stn.
Clinton Crops Res. Stn.
Finley (Raleigh)
Fin ley (Raleigh)
Delaware
Cal dwell
Wooster
Kane Exp. For.
Parsons
Lat.
40°03'
37°26'
44° 13'
47°30'
44°15'
43°57'
44°00'
36°08'
35°03'
35°40'
35°01'
35°44'
35°33'
40°17'
39°45'
40°46'
41°33'
39°06'
Long.
88°22'
88°40'
85°51'
93°28'
95°19'
71°42'
74°13'
77°10'
83°27'
80°34'
78°17'
78°41'
78°41'
83°04'
81°31'
81°56'
78°46'
79o3g,
Monthly
Average pH
4.30
4.47
4.53
4.70
5.64
4.46
4.33
4.82
4.76
5.13
4.94
4.76
4.69
4.33
4.39
4.52
3.96
4.44
Treated as one station.
58
-------
5.8
5.6
5.4
5.2
5.0
o
0>
4.8
o
i
"
4.6
4.4
4.2
4.0
3.8
r =0.825 (rc(a=0.01) = 0.575)
y=l.345+0.746x
3.8 4.0 4.2 4.4 4.6 4.8
pH-calculated
5.0 5.2
Fig. 27 Diagram showing the relationship between observed pH and calculated
PH.
59
-------
Table 4 Sulfur mass budget for January 1977.
Canada
U.S.A
Input
Emissions
Inflow to U.S. from
Canada
Inflow to Canada from
U.S.
8
1.4132 x 10" kg
.42711 x 108 kg
1.84031 x 108 kg
8
7.4586 x 10 kg
.21187 x 108 kg
7.67047 x 108 kg
Output
Wet Deposition
Dry Deposition
Outflow to Ocean
Outflow from Canada
to the U.S.
Outflow from the U.S
to Canada
To the Great Lakes
In the Air
Residue
50705 x 108 (27.55%) 1.9987 x 108 (26.06%)
.61692 x 108 (33.52%) 2.71407 x 108 (35.38%)
,06148 x 108 ( 3.34%) .77192 x 108 (10.06%)
21187 x 108 (11.51%)
.10802 x 10° ( 5.87%)
.10686 x 108 ( 5.81%)
11.6122 x 108 (87.60%)
.22811 x 108 (12.40%)
.42711 x 108 ( 5.57%)
.35029 x 108 ( 4.57%)
.47945 x 108 ( 6.25%)
6.74154 x 108 (87.89%)
.92893 x 108 (12.11%)
60
-------
Table 5 Sulfur mass budget for March 1979.
Canada
U.S.A
Input
Emissions
Inflow to U.S. from
Canada
Inflow to Canada from
U.S.
8
1.41328 x 10" kg
.94141 x 108 kg
2.35469 x 108 kg
8
7.4586 x 10" kg
.16623 x 108 kg
7.62483 x 108 kg
Output
Wet Deposition
Dry Deposition
Outflow to Ocean
Outflow from Canada
to the U.S.
Outflow from the U.S
to Canada
To the Great Lakes
In the Air
.89768 x 108 (38.%12) 1.9024 x 108 (24.95%)
.76484 x 108 (32.48%) 2.0117 x 108 (26.38%)
.04735 x 108 ( 2.01%) .6313 x 108 ( 8.28%)
.16623 x 108 ( 7.06%)
.94141 x 108 (12.35%)
.10482 x 108 ( 4.45%) .5397 x 108 ( 7.08%)
.1763 x 108 ( 7.49%) .4419 x 108 ( 5.80%)
2.15722 x 108 (91.61%) 6.46842 x 108 (84.84%)
Residue
.19747 x 108 ( 8.39%) 1.15641 x 108 (15.16%)
61
-------
SECTION 7
INVESTIGATION OF A PREDICTION METHOD OF THE ACIDITY OF PRECIPITATION
INTRODUCTION
Our studies on the acidity of precipitation have been centered on
finding empirical relationships between the acidity of precipitation and the
concentrations of ions, using the data reported in "Atmospheric Turbidity and
Precipitation Chemistry Data for the World", "Global Monitoring of the En-
vironment for Selected Atmospheric Constituents" (WMO and NOAA, 1976, 1977,
1978), the data reported by Hales and Dana (1979), and the precipitation
chemistry data obtained at the National Atmospheric Deposition Program net-
work (Natural Resource Ecology Laboratory, 1980). The first data consist of
precipitation samples collected on a monthly basis, the second data of
samples collected on a short-term basis at the precipitation chemistry net-
work operated in the METROMEX region surrounding St. Louis, and the third
data of samples collected on a weekly basis. The purpose of our study was to
examine whether the acidity of precipitation can be calculated as an output
of our long-range transport model.
RESULTS
Figures 28, 29, and 30 show the relationships between major ions SOT,
N03 and NH^ in precipitation. In these figures, (A) stands for the data
obtained from WMO and NOAA (1976, 1977, 1978), and (B) is for the data re-
ported by Hales and Dana (1979). The correlation coefficients for these ions
are given in Table 6.
TABLE 6
Variable Pair WMO and NOAA Data Hales and Dana
SO^ - N0~ .299 (rgg = .092) .538 (rgg = .141)
4
N03 "
NHj
NH*
b
.155 (rgg =
.142(rgg=
.092)
.092)
.407 (rgg =
.511 (rgg =
.141)
.141)
As can be seen from this table, all the relationships are statistically
correlated. The strongest, most consistent correlation for these two dif-
ferent data samples is that between SOT and NO-.
62
-------
c
C
QQ
(
(
C
<
_ _
-
"
f
' ' ' f .
*', .'*«**
: >>:#&'.. :
' ' .' {. ';*'.**,
. .. ) '..'<; ".
r -.- ,?:, '
.v . IT ' . . *
,!';..-" ''' -
' /
: -
~ I
~
_
1 1 LLI 1 1 1 1 1 1 1 1 1 1 1 1
) O O C
5 0 - -
>
J9i!i/9ioiu9_oi x [£QN]
-1)1111 t M 1 I I I 1 1 1 I I I I f 1 J ( I i
-
-
"
' , ( .
.
4 ,' ' " . . ' '
. !
: ' , -
, « . . ^ , , _
';*': '.": -.. * . ' . .-
*.''. "^.f. V'i^y-. -.!". _
k - - v:^:i- : ' =
i ' \-:;V'!i.;)-;;:i;' =
~ * ". °
.-. , ... .-./ .'.
-
1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 i 1 1 1 1 1 1 1 1 * i .
3 O O <
i g
J3»ii/3iouj9_oi x [SON]
U 03
r- C
i- ro
oj a
f^
a. TJ
tn c:
o t/1
0 J-
s- o
a. M-
C rO
j_3
r* 4^
(O
M °
8 | £
O ^J
I «.S
5 i
Z a,
r*
A_»
T3 0
>- C
£ (0 c
2 >, ^ ':
^ c *-*
o o re
£ -r- -P
>
C__ j-i
* ^^
0 +J -r- .
_- c -a ^
2 a) T- en
u JD r^.
C S-
O 3 rH
C_3 ( ^
00
C\J
CT1
63
-------
o
o
o
o
o
o
o
o
o
o
o
E
Q '6*
CO
o
o
o
o
o ^
"5
e
1C
b
l/l
c
o
a
c
ns
c
o
o
OO
Q.
O)
U
X
0)
00
CM
0)
TO
OO
OsJ
64
-------
o
o
o
0
o
o
(D
1
Q
C
o
o
o
o
O
0
O)
o =
o ^
JD
o
E
i
1-1 '"i
2 o
J3|I|/9|OLU9_OI =< [J
TD
C
ro
in
c
o
-P
Q.
0)
U
X
O)
00
CM
(O
-------
The observed correlations between ions in precipitation may arise from
two main physical processes. The first of these is the dispersion of a plume
in which all components will be diluted by clean air during the transit from
source to removal area. Consequently, concentrations of these species in
precipitation will be high or low, depending on the degree of dispersion
before removal. The second process involves the rate of precipitation which
affects the concentration of all species scavenged.
In Fig. 31, the relationship between emission rates of SCL and that of
NO from sources over the United States are given. Although the transfor-
mation processes from S0? or NO to SO. or NO-, respectively, are not well
understood, this figure supports the first of the two physical processes
mentioned above.
Granat (1972) has proposed a model for calculating acidity of rainfall
from the concentrations of individual ions as follows:
a = 2([SO] - [N]) + [N0] - [NH] (33)
[N+3) . (34)
-, 10.0 n,+ -1,
a ] ' 457- CNa])
e = a - 2b (35)
where: a = amount of available acid
b = amount of base expressed as moles of carbonate
e = excess acid in moles per liter (-e = alkalinity).
This model assumes that all sodium in rainfall is of marine origin. The
2- + 2+ 2+
fraction of the SO. , K , M , and C ions assumed to be of marine origin is
g a
equal to the concentration of sodium times the ratio of the respective ion in
sea water to the concentration of sodium in sea water. The model is appar-
ently intended to represent the net acidity or alkalinity of rainfall assum-
ing it to be caused by an aqueous solution or suspension formed from sea
salts, sulfuric acid, nitric acid and ammonia. Implicit in the model is the
assumption that the ratio of Cl and Na is the same in rainfall as in sea
water.
The relationship between alkalinity, e, and pH is
K,K?
-e = -4r - [H 3 (36)
[H ]
and pH = Iog10 [H+] (37)
66
-------
1000 c r
ro
2,00
X
o
o
LJ
o: 10
z.
o
CO
CO
QJ
1.0
I i i i i i i 11 i i i i i i 111 i i i i i M i
1.0 10 100 1000
EMISSION RATE OF S02 x I03 ton/year
Fig. 31 Emission rates of S02 and that of NO . Data obtained from EPA.
67
-------
-12 +
where K,K? = 6.30 x 10 , and [H ] is the concentration of hydrogen ion in
mole/liter.
Granat (1972) shows that the model is reasonably accurate in predicting
the net acidity or alkalinity of rainfall. The pH values were calculated
from the above equations using the data in WHO and NOAA (1976).
In Fig. 32 the pH values calculated from theory are plotted against the
pH values observed. In this figure data obtained in the United States,
Canada and Europe during the year 1974 were used. As can be seen, no appar-
ent relationship could be found between the observed pH and calculated pH.
Noticing that the annual average pH in rainfall is presently less than
4.5 over most of the eastern United States (EPA, 1979) and assuming that SOT
and NCL are the two major contributors to acidity, the relationships between
observed pH and the concentrations of these ions were studied.
In Fig. 33, observed pH values are plotted against pH values calculated
from - log-,Q 2[SO~]. Data whose observed pH was greater than 5.0 were ex-
cluded from this figure. (A) contains the data from WMO and NOAA (1976,
1977, 1978). (B) uses the data from Hales and Dana (1979).
The relationships between observed pH and the pH calculated from
-log [2[SO~] + [N0~]} are plotted in Fig. 34.
In Table 7 the correlation coefficients are given.
TABLE 7
Variable Pair WMO and NOAA Data Hales and Dana (1979)
Observed pH
-Iog10 2[SO~] .383 (rgg = .135)
Observed pH
-Iog1() (2[SO~] + [N03]} .421 (rgg = .130)
.295 (rgg = .148)
.288 (rgg = .148)
Although there is considerable scatter in the data points shown in Figs.
33 and 34, there are statistically significant correlations between these
variables.
68
-------
8
OBSERVED PH
6 5
7
o
c
o
~n
33
O
m
O
8
Fig. 32 Observed pH and pH calculated using Granat's (1972) theory.
"Atmospheric Turbidity" and "Precipitation Data for the World",
1974, were used.
69
-------
o
o"
fO
in
rO
X
0.
Q
UJ
> 0
UJ
to
m
0
in
sj-
o
CD in
C
o"
m
hO
Q.
Q
UJ
> 0
ce *
LU
GO
CO
O
in
t
o
< in
[Ios]26°i-iAJoyj aaixnnoivo Hd
>n o in o in o
ro ^r ^r m in to
< i i i i
. "_ t .
* -h '
. ' ' .%*.'. .* .. .
:. *'. V1"'.-'.1 ' ' ':
. '.!*.,. .
'" ;/ '.' "'' ':'-. '.
'' '-. '
~ . t ..'.' "
; .
i i ' i i i
[SOS2] &°I-WOHJ 031VinO~IVD Hd
3 in o m o in c
T (O sj- ^r m in u
^ ! , ] j | -
"..'. '. !
. .
* ,
. i > t*-^ , . . ._ % « ' -
' ' - ' " ' .
V--.- ..i. ...»: i
. ..-..:.. . .".;.._.. , ... ir ( - -.
.. . 1.1... .. ...^ »......_
i , ..'.'... \ . '.i - -...
., j * ^ ', .
. -',*!
; ' '. '. '. ,'t > -I
1 1 1 ' 1 1
1
a.
u
Ol
i.
Q_
<*J
c
>
.)
a
e~\ /^s
t- (71
1 01
^
f«>
S- 03
QJ C
yj ^
Q. Q
o -a
E C
.)-> re
S
re a;
, JZ
3 -I-"
're o
u 4-
T 03
a. +£
re
Q
* *
C -^.
re j_^
Q. !-
E
TJ Ol
01 -C
> O
5-
0) C
ui O
0 -t^
70
-------
{[5oNH£os]2}&°i-wodd
c
0^
Kl
in
rO
X
Q.
Q
LU
> 0
LU
CO
03
O
in
T
0
CD in
C
c"
10
in
fO
X
Q.
Q
LU
> 0
cc ^
s^
r°i ro
10 C
o ro
E 0
4->
< TJ
= C
5 l^
*>M^ CJ
^
p; e
i ro O
1 -
1 ' ro
X -S
"g"g
^j *.
f"""' ^
D5 O
o s
3 "~ -
u -P
in
^_ r*
- |
C
ro c
_ o
31 -r-
a. +j
fD
-a +J
> p
^. .p
OJ
On
k^-.
71
-------
Further studies of the empirical relationship between pH and -log,~
(3[SOT]} and -log{2[S07] + [NO-]} were conducted using the data obtained at
the stations of NADP network. For our studies, the data taken during the
period between July 1978 and June 1979 at 17 stations over the eastern United
States were used.
In Table 8, correlation coefficient r and regression parameters A and B
of the relationship between observed pH and ~log,Q{3[SOT]} and -log{2[S07] +
[N03]} are given. Here A and B are coefficients of y = AX + B, where y is
the observed pH and x is the corresponding variable. [SOT] and [NCL ] are
the concentrations expressed in time of mole/liter.
The following can be seen from Table 8:
(1) In general the correlation coefficients between observed pH and
-log1Q{[NO-] + 2[S07]} are higher than those between observed pH
and -log,g{2[S07]} , suggesting the importance of NO., ion in the
calculation of pH.
(2) The degree of correlation is dependent upon locations where data
are taken.
(3) Among 17 stations studied, the data of 10 stations are statistical-
ly correlated with 95 percent confidence levels between pH and
-log10{2[SO~] + [N03]}, and 11 stations between pH and -Iog10
From the above empirical studies, the following can be concluded:
Admittedly crude, it is possible to predict the acidity of precipitation
from the model outputs of SO, and NO- ions. In Section 6, it has been shown
that the pH values calculated from [SO,] content in precipitation are highly
correlated with the pH values observed. Inclusion of NO /NO., in the long-
/\ O
range transport model will improve the predictability of acidity of precipi-
tation.
72
-------
z
Q.
C
Ol
cu
-p
cu
.0
Q.
r-
.C
in
o
r-
P
ro
r
CU
£_
CU
.c
-p
*4
o
I/)
S-
cu
-p
cu
E
TO
s-
TO ""-<
Q. i i
It ^Jt"
C 0
O OO
r- 1 1
10 CM
in -~~
cu o
S- rH
cn en
Ol O
i. t
i
^3
c -o
fO C
(O
-P
C i~~*
CU I I
r || ^"
U 0
r- OO
M l I
4- CM
CD
0 +
(J
I I
C 1 OO
0 O
r- 2
-p 1 1
fO i
0
CU r-l
5- Cn
S- 0
O r-
c_> t
CO
Ol
r
O
ro
r-
*
**"^
1 1
o
oo
CM
cn
o
r
1
o
c
rO
a.
/-^
p i
II *3-
O
OO
1 t
CM
^.
I I
1 OO
0
2
cn
o
r
1
a
c
ro
I
Q.
a.
r"
in c
C (V
O 0)
r 5
4-* 4-^
03 CU
^ f^]
^J
a:
<4_
<*-
cu
o
o
. s_
^.
S-
o
o
10
S-
0)
p
cu
E CO
ro
s_
ro
Q-
C
O
r-
tn
tn
ai c to
a o i
C X r
o -i- ai
CO Q S
K £
cn r^
oo r^-
00 U3
, .
0 0
00 ID
to t£>
d o
CM 00
cn «3-
r-i rH
cn o
LO CM
^" CO
d d
00 rH
oo OO
cn en
d d
r*** CD
*
o o
QJ
lf_
r-.
__
r-
o ^
Q
O
+J
a cn
J^ C
(^ *r~
Q +J
__Q CI
3 r3
m z
^- -h-
iH r- 1
rH i-H
r^s p^.
, ,
o t£> o cn
P**« rH
rH CS1
. .
o o
CO CO
ft. r>-
d d
r-t '"J
rH ^
4^-
S «
r^ r i i^
d?M°CM
t ,
O O
. , 00
CM (^
rH rH
00
d rt
CM
O
i I I
C
+J c
00 +J
OO
t ) *
fc. w
«*4 QJ
"5 X5
fll L£>
u/
vn
« 2
C_j
C -P
O ro C C
4-* +-' O O
i/l QJ £ -f-J
r- (U "D C
5 ? ^ *r<~
CU 0 !- r
1 O d. o
X *f* 4^
rH «q- CM
ID CO LD
OO CO ID
...
O O O LD
CM
OO
.
0
o CM r^.
IO CM CO
LD rH £
O rH o
o r-. «*
CM OO CM
CM o CM
i
LD ID m
CM cn CM
*3" CO «^- pH
* ' " CM
f"^ t^ f~~^ .»
.
O
CM CM CM
r^** CM r^«
O r-l o
CO 2^
1-1 ° rH
x s r->
r~ _c~
cn cn
cu cu
ro ro
~^ ~^
CU i
i- r
^> ^^ TJ QJ
^ ^ 5 3*
r- i ^ oo
in ^ oo
o ° d
OT * S
cn . ">
r-i ""* 0
"*~ O
0 LD 2 OT
*"' TL» r^ r^«»
. | QQ '
'^ CO
^4- - . ****
d° o
5 o
LD iv^ OO
cO . cn
5° o
LD
OO . LD
0 rH
CM °
jj ^
U- OO
f^
s- m-1" m
cu c c
-P O O
in cu -P in
O C t- S-
O ro O ro
3 ^ a: a.
cu cn
r
J3 -P
ro ro
r-
J- -P
ro C
^ TO
U
cn *^
C M-
5 c
c cn
o >-
Q. in
in
01 >,
i. r
O ro
U U
r"
CU -P
.C in
P !-
-P
in ro
r- +J
to
X
in
"D cu
C +J
rO ro
(J
Z >-
OO Q.-O
f-
CU ^ -r-
S "~*
ro ^
r- ^^j
C! ^
, ^ ^
cu
c -c -t-
CU 3
> x:
r- --P
cnoa T-
5
CD +
ro ^s CU *
in j_ r
C .. fo Ol
0 " £ >
P" !
^J >) . +J
TO O
-P <4- ,j OJ
to O "tl Q.
t/>
14~- LA QJ
° *-> 'u ^
c 1 E «
O (j i
+J '
ro 4_ u CU
<-> CU ^
0 o S
r- u 0 +J
"- c
r- ... -P CU
ro {_ ro (j
U m f~~" S
r- CU CU
x: ^-Q.
a. ca s- "
ro o in
i. "O u cn
cn c
O ro CU "U
CU -C C
CD < r- ro
. * .
r-H CM OO
. .
CU
.p
o
"Z.
73
-------
REFERENCES
Bhumralkar, C.M. , R.L. Mancuso, D.E. Wolf, R.A. ThuiTMer, K.C. Nitz and
W.B. Johnson, 1980: Adaptation and Application of a Long-Term Air Pol-
lution Model ENAMAP-1 to Eastern North America. Final Report to EPA
Contract 68-02-2959, SRI Project 7760, SRI International, Menlo Park,
CA.
Clark, T.L., 1979: Gridded Annual Pollutant Emissions East of the Rocky
Mountains. EPA-600/4-79-030. United States Environmental Protection
Agency, Environmental Sciences Research Laboratory, Research Triangle
Park, NC 27711.
English, E.J., 1973: An Objective Method of Calculating Area! Rainfall.
Meteorological Magazine, 102, 292-298.
EPA, 1979: Research Summary, Acid Rain, EPA-600/8-79-028, Office of Re-
search and Development, U.S. EPA.
Galloway, J.N. and D.M. Whelpdale, 1980: An Atmospheric Sulfur Budget for
Eastern North America. Atmosspheric Environment, 14, 409-417.
Garland, J.A., 1978: Dry and Wet Removal of Sulphur from the Atmosphere.
Atmospheric Environment, 12, 349-362.
, 1979: Dry Deposition of Gaseous Pollutants. Paper Presented at the
WMO Symposium on the Long-Range Transport of Pollutants and its Relation
to General Circulation Including Stratospheric/Tropospheric Exchange
Processes, Sofia, 1-5 October, WMO-No. 538.
Granat, L. , 1972: On the Relation Between pH and the Chemical Composition
in Atmospheric Precipitation. Tel 1 us, 24, 550-560.
Hales, J.M. and M.T. Dana, 1979: Precipitation Scavenging of Urban Pol-
lutants by Convective Storm System. Journal of_ Applied Meteorology,
18, 294-316.
Henmi, T. , 1979: Long-range Transport Model of S02 and Sulfate and its
Application to the Eastern United States. WMO Symposium on the Long-
Range Transport of Pollutants and its Relation to General Circulation
Including Stratospheric/Tropospheric Exchange Processes. Sofia, 1-5
October, WMO-No. 538.
74
-------
, T. , 1980: Long-range transport model of SO^ and sulfate and its
application to the eastern United States. Journal of Geophysical
Research, 85, C8, 4436-4442.
and E.R. Reiter, 1979: Long-Range Transport and Transformation of
S02 and Sulfate. EPA-600/4-79-068, United States Environmental Protec-
tion Agency, Environmental Sciences Research Laboratory, Research
Triangle Park, NC 27711.
Holzworth, G.C., 1972: Mixing Heights, Wind Speeds, and Potential for Urban
Air Pollution Throughout the Contiguous United States. AG-101, Office
of Air Programs, Environmental Protection Agency.
Husar, R.B., D.E. Paterson, J.D. Dusar, and N.V. Gillani, 1978: Sulfur
Budget of a Power Plant Plume. Atmospheric Environment, 12, 549-568.
Natural Resource Ecology Laboratory, 1980: National Atmospheric Deposition
Program Data Report. Colorado State University, Fort Collins, CO.
Pack, D.H., 1978: Sulfate Behavior in Eastern United States Precipitation.
Geographical Research Letters, 5, 673-674.
Science News, 1979: Acid Rain in the Spotlight, 116, p 244.
Sheih, C.M., 1977: Application of a Statistical Trajectory Model to the
Simulation of Sulfur Pollution Over Northeastern United States. Atmo-
spheric Environment, 11, 173-178.
Voldner, E.G., Y. Shah and D.M. Whelpdale, 1980: A Preliminary Canadian
Emissions Inventory for Sulfur and Nitrogen Oxides. Atmospheric En-
vironment, 14, 419-428.
Wilson, W.E., 1978: Sulfates in the Atmosphere: A Progress Report on Pro-
ject MISTT. Atmospheric Environment, 12, 537-548.
WMO and NOAA, 1976: Atmospheric Turbidity and Precipitation Chemistry Data
for the World, 1974. Environmental Data Service.
and , 1977: Global Monitoring of the Environment for Selected
Atmospheric Constituents, 1975. Environmental Data Service.
and , 1978: Global Monitoring of the Environment for Selected
Atmospheric Constituents, 1976. Environmental Data Service.
75
-------
TECHNICAL REPORT DATA
(Please retd Jaumcnons on the reverse before completing1
1 REPORT NO
EPA-600/4-81-070
3. RECIPIENT'S ACCESSION NO.
4 TITLE AND SUBTITLE
LONG-RANGE TRANSPORT AND TRANSFORMATION OF SO, AND
SULFATE *
Refinement, Application, and Verification of Models
5. REPORT DATE
August 1981
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
Teizi Hennii and El mar R. Reiter
B. PERFORMING ORGANIZATION REPORT NO.
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Colorado State University
Fort Col ins, Colorado 80523
10. PROGRAM ELEMENT NO
CCVN1A/01-0337 (FY-81)
11. CONTRACT/GRANT NO.
R-805271
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Environmental Sciences Research Laboratory - RTF, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
TYPE OF REPORT AND PE
Final 5/80-3/81
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
16 ABSTRACT
A long-range transport model of S02 and sulfate for twenty-four-hour concen-
tration distributions was refined and applied to calculate distribution patterns
of concentration and deposition of S02 and sulfate over the area between 35°N and
45°N and between 75°W and 95°W for January 25 and July 11, 1976. The calculated
concentrations and the observed concentrations were compared.
A climatological model of long-range transport of S02 and sulfate was also
refined to calculate average monthly distributions of S02 and sulfate concentrations
as well as the acidity of precipitation due to sulfate and the budget of sulfur over
eastern North America. The model has been applied for the months of January 1977 and
March 1979 over the area between 35°N and 55°N and between 62°W and 95°W. The
results are described.
Empirical studies of precipitation chemistry data were conducted and indicate
that inclusion of NO /NOl in the long-range transport model is important to improve
the predictability of precipitation acidity.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATi Field/Group
13 DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY C,LASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (t-73)
76
------- |