-------
64
If the scattering coefficient corresponding to each of the modes for
the different aerosols sampled by Whitby and Sverdrup (see Table 3) is cal-
culated, the accumulation mode is the dominant scattering mode, with the
coarse mode contributing a small amount and the nuclei mode a negligible
amount. For the clean continental background aerosol (second entry in Table
3), we calculated that the total scattering coefficient (b ,) equaled
_4 _i scat
0.23 x 10 m , which is very close to the minimum scattering coefficient sug-
gested by the data shown in Figure 9 (Chapter II). The accumulation mode
contributes 0.09 x 10 m (39 percent of bc^a.), the coarse mode 0.04 x
A 1 scat . -1
10 m (17 percent of b$cat), and Rayleigh scattering 0.10 x 10 m (43
percent of bsca+)- In the clean background case, the contribution of the
accumulation mode to the extra extinction (b ) above Rayleinh was 69 per-
cent of the total. This clean background aerosol corresponds to a visual
range of 170 km, or 105 miles (3.912/b .). For the average background
" _A -i
aerosol (third entry in Table 3) the computed b is 0.57 x 10" rrf ,
scan
corresponding to a visual range of 69 km. In this case, the accumulation
4 -1
mode contributed 0.27 x 10" m (46 percent of b ,), the coarse mode
4 i scat 4 i
0.21 x 10 m (36 percent of bscat)» and Rayleigh 0.10 x 10 m (17 per-
cent of b,.,.). These calculations indicate that for background conditions
scat
the accumulation mode is a larger contributor to light scattering than the
coarse mode but that the coarse mode is a nonnegligible component of the
scattering coefficient of the background atmosphere. This situation strik-
ingly contrasts with that of polluted urban atmospheres, in which the
accumulation mode causes more than 90 percent of the total scattering coef-
ficient. We should also note that the average background presented in
Table 3 is not as clean as the average background in the nonurban western
United States, as Figure 9 suggests.
In our visibility models, we have used the size distributions given in
Table 3 for specifying ambient background (second and third items in Table 3)
and plume aerosol (eighth item). We then calculated optical properties of
the aerosol (using Mie theory) from the computed concentrations of coarse and
accumulation mode (sulfate, nitrate, and associated cations and water)
aerosol.
-------
65
D. ATMOSPHERIC OPTICS
In the atmospheric optics component of the visibility models, the
light scattering and absorption properties of the aerosol and the resultant
light intensity for various illumination and viewing situations are computed.
The details of these calculations are given in Appendix B; the major points
are summarized in this section to give the reader an overall view of the
process.
1. Calculation of the Scattering and Absorption Properties
After the concentrations of the pollutants are specified by the trans-
port and chemistry subroutines, their radiative properties must be deter-
mined. For N02, the absorption at a particular wavelength is a tabulated
function (Nixon, 1940) multiplied by the concentration. For aerosols, the
procedure is more complicated, however.
In general, a particle's ability to scatter and absorb radiation at
a particular frequency is a function of size, composition, shape, and
relative humidity. Because we wanted to be able to alter the size distri-
bution of both primary and secondary particles, we needed to be able to
compute the effect of particle size on the wavelength dependence of the
extinction coefficient and the scattering distribution function. The only
rational method of making this computation is to use the solution of
Maxwell's equations for scattering by a sphere, the so-called Mie equations.
To verify that these calculations were appropriate for atmospheric aerosols,
we compared them with the empirical correlations of scattering to mass and
found substantial agreement, as discussed below.
The calculations were performed using an IBM subroutine written by
J. V. Dave (Dave, 1970). The required inputs are the particle size param-
eter (ratio of the circumference to the wavelength of radiation), the index
-------
66
of refraction (real and imaginary part), and the number and location of the
scattering angles (between 0° and 180°). The output is the scattering and
absorption cross sections and the Stokes transformation matrix (Van De Hulst,
1957), which can be simply converted to the scattering distribution assuming
randomly polarized light. The scattering and absorption properties per par-
ticle are then summed over the particular size distribution in such a way
that as the size distribution changes so do the radiative properties.
Different types of empirical correlations have been made in recent years
relating particle scattering properties to particle mass. The property mea-
sured has been either the volumetric scattering coefficient, as measured by
an integrating nephelometer, or the extinction coefficient, calculated from
the observed visual range. Among recent discoveries is the conclusion that
the scattering properties of urban atmospheres correlate much better with
the submicron accumulation mode concentration than with total mass. A
second important development is that for most of the United States and
Europe, sulfates are generally a significant fraction of submicron accumu-
lation mode mass.
Many studies of the empirical correlations of scattering coefficients
to sulfate mass have been performed. Table 4 summarizes some of these
measurements. Similar tables have appeared in other reports, such as that
by Trijonis and Yuan (1977). The correlation coefficients for these rela-
tionships have been very high (0.7 to 0.9), supporting the dominant role of
sulfate in scattering. Of particular importance is the use, by many
researchers, of visual range data (e.g., airport visibility) derived from
actual, though somewhat imprecise, visual perception of objects in the atmo-
sphere. Thus, it appears that sulfates play an important role in visibility
impairment.
Calculations of scattering-to-volume ratios (see Figure 20) reveal that
the maximum theoretically possible value is about 0.06, as expressed in the
-------
67
TABLE 4. ESTIMATES OF EXTINCTION COEFFICIENTS PER UNIT MASS
Extinction
Coefficients
COO4 m~V(Mg/m3)]
Source
Regression models
(Trijonis and Yuan, 1978)
Regression models
(Trijonis and Yuan, 1977)
Dust storms
(Hagen and Woodruff, 1973)
Regression model
(White and Roberts, 1975)
Regression model
(Cass, 1976)
Calculations for a model aero-
Location
Chicago
Newark
Cleveland
Lexington
Charlotte
Columbus
Salt Lake City
Phoenix
(county data)
Phoenix
(NASN data)
Great Plains
Los Angeles
Los Angeles
0
Sulfates
0.04
0.03*
(0.02)
0.06*
0.08
0.07*
0.06
0.06*
0.11
0.11*
0.12
0.13*
0.04
0.04*
0.04
0.03
NC
0.07
0.16
0.09*
.05-0.10
Nitrates
(0.00)
(0.00*)
(0.00)
(0.00*)
(0.00)
(0.00*)
(0.00)
(0.04*)
(0.00)
(0.00*)
0.09
(0.06*)
0.13
0.10*
0.05
0.03
NC
0.05
(0.00)
0.05*
NC
Remainder
of TSP
(0.000)
(0.000*)
0.026
0.014*
(0.000)
(0.000*)
(0.000)
0.019*
(0.001)
(0.000*)
(0.000)
(0.001*)
0.004
0.004*
(0.000)
(0.000)
0.001
0.015
0.008
(0.004*)
NC
sol of (NHjpSOjj at 70% RH
(Waggoner fit al., 1976)
Regression model
(Waggoner et al., 1976)
Southern Sweden 0.05
NC
( ) = not significant at the 95 percent confidence level.
NC = not calculated.
*Based on nonlinear RH regression model, with insertion of average RH.
Source: Trijonis and Yuan (1978).
-------
68
*
units used in Table 4. However, the reported empirically determined
values range from about 0.03 to 0.1. This discrepancy was resolved by
including the mass of liquid water associated with the sulfate. Investi-
gators have found better correlations with scattering-to-mass ratios that
depend on relative humidity. These fits are of the form:
(49)
1 - RH
where c represents the scattering-to-mass ratio at 0 percent relative
humidity. The range of these values is between 0.02 and 0.04, which is
in agreement with theoretical values shown in Figure 20."*"
To account for the effects of relative humidity, we simply added the
amount of water absorbed by the sulfate particles:
MaSSSulfate + MassCation + MassWater
Then we used a formula from Winkler (1973) to account for the mass of water
as a function of relative humidity. Finally, we compared the dependence of
the scattering-to-mass ratio on relative humidity determined by Cass and
Trijonis with calculations using the following assumptions:
This was computed from the maximum value of 0.08 x 10"^ m~'/ym3/cm3 in
Figure 20 assuming the sulfate was associated with ammonium ion as
(NH4)2S04:
cm3 132
W
A
l .8 g(NH4)2S04A10 yg 96 gS04
= 0.06 x 10"4 m"1/(yg/m3 SOj)
1 If we. use an accumulation mode bscat/V = 0.06 x 10~4 nH/ym3/cm3, we obtain
bscat/(vg/m3 S04) ranging from 0.034 to 0.046 x 10~4 m~Yyg/m3 depending
on whether the sulfate is H2S04 or (NH4)2S04.
-------
69
> Aerosol with a lognormal size distribution and a mass
median diameter of 0.2 ym and a geometric standard
deviation equal to 2.0 at RH = 0 percent.
> Index of refraction equal to 1.5 - Oi (typical, non-
absorbing aerosol).
> Density equal to 1.8 g/cm .
> Light of 0.55 pro wavelength.
> Sulfate as NH4HS04 (molecular weight of 115).
Figure 21 shows the striking agreement between our calculations and the
dependence of scattering-to-mass ratio on relative humidity observed by
Trijonis in the Southwest. This agreement gives support to our calculation
method for the scattering properties of the secondary aerosol.
The computational procedure used in the visibility models then takes
the input size distribution, assuming that the particles are spherical
with an index of refraction of 1.5 - Oi, and computes the scattering pro-
perties of the aerosol as a function of wavelength from the Mie equations.
The background size distribution properties are taken from the Whitby and
Sverdrup model of clean continental aerosols (see Table 3). The accumula-
tion mode in the plume is assumed to be the size of those measured in the
plume downwind of the Labadie power plant near St. Louis. The properties
of the sulfates and other accumulation mode particles are assumed to change
with relative humidity, as discussed above. A more complete inclusion of
relative humidity would require a modification of the refractive indices and
a recomputation from the Mie equations. This modification could easily be
done later if desired.
The limitations of the process of specifying the radiative properties
of aerosols are the usual ones: uncertainties in the size distribution,
nonspherical particles, and ambiguities in mean refractive indices. How-
ever, the close agreement shown for the sulfate scattering-to-mass values
suggests that the errors are not large.
-------
70
0.14
0.12
0.10
o
GO
CO
I, 0.08
2 0.06
o
CO
0.04
0.02
0.20
0
0 CALCULATED VALUES
*=> TRIJONIS AND YUAN (1977)
I
0.40 0.60
Relative Humidity
0.80
FIGURE 21. RATIO OF LIGHT SCATTERING TO MASS AS A FUNCTION
OF RELATIVE HUMIDITY
-------
71
2. Calculation of Light Intensity
p
The light intensity (watts/m /steradian) at a particular location in
the atmosphere is a function of the direction of observation n and the wave-
length >. Calculation of the light intensity in a medium follows from the
radiative transfer equation. This equation is a conservation of energy
statement that accounts for the light added to the line of sight by scat-
tering and the light lost because of absorption and scattering. Approxi-
mations and solution techniques applicable to planetary atmospheres have
been discussed by Hansen and Travis (1974) and Irvine (1975).
The physical situation that we are concerned with is shown schemati-
cally in Figure 22. To compute the spectral light intensity at the
observer, we sum (integrate) the scattered and absorbed light over the
path, r, associated with the line of sight n. The resultant general
expression for the background sky intensity at a particular wavelength is
'T
e' dT'
where
r
T = the optical depth (T E^ b . dr, where
is the extinction coefficient),
w = the albedo for single scattering (w = bscat/bext
where b . is the scattering coefficient),
-»• n) = the scattering distribution function for the
angle n1 -> jj,
I = the spectral intensity at T' from direct and
diffuse solar radiation.
-------
72
o.
oo
D_
cc
o
oo
5
CJ
OJ
C3
-------
73
Equation (51) is valid for the usual continuum, no refraction, random
polarization assumptions.
The intensity seen by an observer in direction n of an object at dis
tance R is:
WB) -
) dfi1 e"T dT' . (52)
'=4TT
Equations (51) and (52) then completely describe the spectral intensity
of the background and an object. Once these two quantities are known, the
visual effects of the intervening atmosphere can be quantified. In evaluat-
ing Eqs. (51) and (52), we encounter two main difficulties: First, the
quantity in the integral is a fairly complicated function, and accurate
specification is tedious. Second, the atmosphere is inherently inhomogen-
eous, and thus, the radiative properties w, p are somewhat complicated
functions of r and n. Approximations are therefore necessary. Appendix B
outlines in some detail the approximations we have used; we present only a
summary here. The approximations we used are the following:
> Plane parallel atmosphere.
> Two homogeneous layers.
> Average solar flux approximation.
> Average diffuse intensity approximation.
The equation for the background intensity at the surface becomes, for a
direction y, »
-------
74
_ Tdi f /n 0°° \ / r i\
WOD !av U - e ) ' (53)
and for the intensity in the direction of an object in the planetary bound
ary layer,
where
o' , Pnn(o) = the average albedo and phase function
respectively,
Tnn = ^^e °Ptica^ depth of the path in the boundary
layer,
F_ , I = the average solar direct intensity and diffuse
s ,av av
intensity, respectively,
I . , I = the intensities from the upper atmosphere and
S Kjr 0
object, respectively.
The exact definitions of the terms are given in Appendix B.
Thus, the background intensity and the intensity in the direction of an
object at distance R from the observer can be computed given the following
inputs:
> Background radiative properties (e.g., size distribution,
visual range).
> Solar zenith angle.
> Scattering angle.
> Direction of observation, n, 4>-
> Planetary boundary la^er height.
-------
75
The intensities, including the effects of air pollution, are computed
from essentially the same formulae with the radiative effects of the pollu-
tants included in the background atmosphere. In the regional model, the
intensity for a given optical path is calculated from an integration of the
concentration through the cells of the grid model. For an initial approxi-
mation, we used the expressions for a homogeneous atmosphere.
In the plume model, it was necessary to treat the plume as a homogen-
eous layer with an optical depth and mean properties
-------
76
> Coloration of objects.
- Brightness
- Hue and saturation.
> Contrast and color difference between two objects.
- Black object and horizon sky (to calculate
visual range).
- Haze layers.
- Plume and background.
The perception of an object such as a distant mountain results from
changes in light intensity, coloration, or both. Visual range is defined in
terms of differences in light intensity (contrast) between a distant black
object and the horizon sky. Contrast is also a useful concept for charac-
terizing the appearance of plumes and haze layers. However, a plume may
be perceived against a background as the result of a color change unaccom-
panied by a change in light intensity (i.e., with no contrast). Ue there-
fore need a means of characterizing the perception of changes in both the
intensity and the coloration of light. We discuss the different means of
characterizing visibility impairment in the following subsections.
1. Visual Range
Visual range is defined as the farthest distance at which a black
object can be perceived against the horizon sky. As we have noted in
Chapter II, the threshold of perception of differences between the light
intensity of two objects has been characterized by a liminal contrast.
The value of the liminal contrast is commonly taken to be 0.02, as first
suggested by Koschmieder in 1924 (Middleton, 1952). However, the liminal
contrast is a function of the observer and his state of mind (e.g., fatigue,
attentiveness) as well as the intensity of the background lighting. Under
the best conditions, the liminal contrast may be as low as 0.005 (Committee
on Colorimetry, Optical Society of America, 1963). The Federal Aviation
Administration assumes a value of 0.055. Based on an experiment using 10
-------
77
observers and a total of 1000 observation hours, Middleton (1952) reported
a median of 0.03 and a mode of 0.02 for the liminal contrast. For the pur-
poses of standardization, it is reasonable to describe the perception of a
"standard observer" and to select and use a single value for the liminal
contrast. We used the Koschmieder value (0.02) for our calculations.
It is clear then that the observation of distant targets such as moun-
tains is not an accurate measurement of strictly defined visual range, i.e.,
the farthest distance at which a black object is distinguishable from the
horizon sky by a standard observer where liminal contrast is 0.02. This
is true not only because of the variability in the contrast threshold, but
also because distant markers such as mountains are usually not perfectly
black.
The contrast between two objects is defined as:
- I2(x) '
If the two objects are the same color [i.e., I,(A)/Ip(x) is constant
over 0.4 < A <0.7 pm], then the contrast at all wavelengths will be the
same. However, if the objects have different colors, then C is a function
of wavelength. For the calculation of visual range, we evaluate the con-
trast at a wavelength of 0.55 um, which is at the middle of the visible
spectrum and is the wavelength to which the human eye is most sensitive.
The intrinsic contrast of a black object (I, =0) against the horizon sky
(\2 - Ih) is -1; the visual range is the distance at which this contrast
is reduced by the light scatter and absorption of the intervening atmo-
sphere to -0.02. Thus, visual range can be evaluated by computing con-
trast iteratively as a function of distance from the observer until it
drops to -0.02. This approach is necessary if one is dealing with a non-
homogeneous atmosphere.
-------
78
For a homogeneous atmosphere, however, the calculation of visual
range is analytic, using the Koschmieder relationship:
ext
For the computation of visual range through a homogeneous atmosphere con-
taining an optically thin plume, Latimer and Samuelsen (1978) suggested
the following simplified approach:
r =
3.912 - £n -^ - T I . (57)
v bext-0
The second term in the brackets is necessary to account for the effect of
light absorption caused by plume NCL on the contrast between the horizon
sky (IhD) and the black object seen through the plume. It can easily be
shown that the effect of the plume on visual range is significantly less
when the plume is discolored by NOo Uhp/Ih < 1), and greater when the
plume is bright Ohp/Ih > 1). As a corollary, it is also true that with
increasing distance between the observer and the plume, the impact of
plume N02 on visual range increases as plume coloration decreases. This
somewhat surprising result was confirmed in the sensitivity analysis of
the plume visibility code.
2. Contrast of Haze Layers and Plumes
Contrast can be used to characterize the perceptibility of a haze
layer or a plume against a background—the sky, a cloud, or a distant
mountain. A plume would be visible if the absolute value of the contrast
between it and the background were greater than a threshold or liminal
contrast. Figure 23 is a photograph of a plume illustrating plume con-
trast. The plume is clearly visible against the mountain because the
plume light intensity is greater than that of the mountain. Thus, the
contrast of the plume against the mountain can be calculated using Eq. (45):
-------
79
o
<
Q.
U.
o
ro
C\J
-------
80
C . J^L-J^o . (58)
H m
The plume is also visible against the horizon sky, perhaps mainly
because of the color change, but also because of contrast:
(59)
The magnitude and the sign of the contrast of a haze layer or plume against
a background is therefore a useful way to characterize visibility impair-
ment. Positive contrasts refer to plumes brighter than the background,
whereas negative contrasts refer to plumes darker than the background. We
do not have any experimental data for liminal contrast (the barely percep-
tible threshold contrast) in the case of a plume against a background. The
same liminal contrast used to define visual range (0.02) could be used to
define plume visibility. However, it seems likely that the liminal contrast
for plumes is greater than 0.02 because in many cases the boundary between a
plume and the background is not distinct owing to the nature of plume dilu-
tion. It would be useful to carry out some experiments with several
observers and plume views to determine the liminal contrast for an average
observer.
Contrast of plumes can be evaluated at several different wavelengths;
we used 0.55 ym for the evaluation of plume contrast. However, plume con-
trast may be greater at the blue end of the visible spectrum. Latimer and
Samuelsen (1975, 1978) used the ratio of plume to background intensities
at the blue end (x = 0.4 ym) and at the red end (x = 0.7 ym) as a means of
characterizing the wavelength-dependent plume contrast and plume coloration
with respect to the background. This blue-red luminance ratio is defined
as:
I (0.4 ym)/I. (0.4 ym) C, (0.4 ym) + 1
D = JL, _ ' " _ _ P _ (60)
K Ip(0.7 ym)/Ih(0.7 ym) Cp(0.7 ym) +1
-------
81
The use of the luminance ratio in conjunction with the plume contrast at
0.55 ym is a simple way of characterizing plume color. When R > 1, the
plume is more blue than the background; when R < 1, the plume is redder
(or more yellow-brown); when R = 1, with C (0.55 pm) > 0, the plume is a
brighter white than the horizon, and with C (0.55 pm) < 0, the plume is
a darker grey. We discuss more sophisticated methods of quantifying color
in the next subsection.
3. Color
The color associated with a given spectral light intensity distribution
is due to processes occurring in the human eye. The retina has three dif-
ferent frequency sensors that convert signals into color sensations by means
of the brain. The system operates so that an object that reflects half blue
light and half yellow light is identified not as yellow-blue, but rather as
a new color, green. This attribute of the eye-brain system gives rise to
another mode of detecting an object, that of color change or discoloration.
Thus, an object can be perceived because it has a different brightness from
that of the background (contrast) or because it has a different color (so-
called color contrast). Gases and particles in the atmosphere can give rise
to coloration by their scattered light (blue sky or white clouds) or by
altering the color of objects seen through them (brown coloration due to NO?).
The chromaticity diagram was developed to quantify the concept of color.
In such a diagram, the spectral distribution of light is weighted with three
functions corresponding to the detectors in the eye. For any distribution
of light, there are three numbers, which define a point in space. Next, the
projection of the point onto a unit plane (x + y + z = 1) is computed. The
result is a two-dimensional surface called a chromaticity diagram (see
Figure 24). Monochromatic light forms the outside of the surface, and white
light is located in the center. Any color can thus be represented by its
chromaticity coordinates (x,y), which are defined by:
-------
82
0.80 —
0.70 —
.0.555
>,0.4C—
0.00
GREENISH YELLOW
YELLOW
A58^-ORANGE YELLOW
.585
°-49 Y GREENISH
BLUE
0.00 0.10
REDDISH ORANGE
0.70 ym
0.70
FIGURE 24. CHROMATICITY DIAGRAM
-------
83
X=X+Y+Z ' y X+Y+Z
where
X = / I(A) x dA
A
Y =/ I(x) y dA
=J
I(A) Z dA
A
and I (A) is the wavelength distribution of light and x, y, z are the three
weighting functions. The weighting functions (called tristimulus values)
are shown in Figure 25.
Horvath (1971) and Husar and White (1976) computed chromaticity coor-
dinates of atmospheric scattered or transmitted light and showed that the
light would be distinguishable from white light for various sun angles,
aerosol properties, and NfL concentrations. Since the chromaticity dia-
gram does not differentiate between differences in intensity (e.g., between
yellow and brown or between white, grey, and black), chromaticity coordinates
must be used in conjunction with a descriptor of light intensity for a com-
plete specification of color. Thus, if we establish a color solid by taking
the two-dimensional chromaticity diagram and adding a third dimension per-
pendicular to this plane to represent brightness, we have a means of com-
pletely specifying by three coordinates the color and intensity of a color.
Figure 26 is a drawing of such a color solid. The brightness in such
a coordinate system is usually specified by the value of Y [see Eq. (61)]
or by a parameter (L*), which is directly proportional to the subjective
perception of brightness and is related to Y as follows:
-------
84
O)
3
a
3
I/)
•r—
$-
400
500 600
Wavelength, x (nm)
Source: Judd and Wyszecki (1975).
FIGURE 25. SPECTRAL TRISTIMULUS VALUES x(x), y(x), z(x)
-------
85
L* = 25 Y1/3 - 17
(62)
L* is used in quantifying color differences and is simply the parameter
called "value" in the Munsell color system multiplied by 10.
VALUE
(brightness)
CHROMA
(saturation)
Source: Munsell Color Company (1976).
FIGURE 26. REPRESENTATION OF A COLOR SOLID
The Munsell color system is the most widely used means of specifying
colors. In this system, colors are arranged in order by value (brightness),
hue (the shade of color, for example, yellow, red, green, blue), and chroma
or saturation (the degree of departure of a given hue from a neutral grey of
the same value). By specifying a given hue, value, and chroma, one can
obtain a sample color chip from the Munsell Book of Color that corresponds
to the specification. By this means, the objective specification of color
-------
86
(L*,x,y) can be related to the subjective perception of color by visually
examining the color paint chip. ASTM Standard D 1535-68 (American Society
for Testing and Materials, 1974) is the reference method for converting
objective color specifications (L*,x,y) to the Munsell hue, value, and
chroma notations by which a colored paint sample can be selected. We
used this method to convert light intensity (Y or L*) and chromaticity
coordinates (x,y) calculated by the plume and regional visibility models
to Munsell notation to be used by a commercial artist in illustrating
atmospheric discoloration. We discuss this process in Chapter IV.
4. Color Changes
The final step in the quantification of visibility impairment is the
specification of color differences—differences both in chromaticity (x,y)
and brightness (Y). In 1976 the Commission Internationale de 1'Eclairage
(CIE) adopted two color difference formulae by which the perceived magni-
tude of color differences can be calculated. Color differences are speci-
fied by a parameter A£, which is a function of the change in light intensity
or value (AL*) and the change in chromaticity (AX,Ay). AE can be consid-
ered as a distance between two colors in a color space that is transformed
in such a way that equal distances (A£) between any two colors correspond
to equally perceived color changes. This suggests that a threshold (A£Q)
can be found to determine whether a given color change is perceptible.
Since the CIE could not decide between two different proposed formulae
for AE, both were adopted in 1976 as standard means by which color differ-
ences can be specified. These color differences, which are labeled
AE(L*U*V*) and AE(L*a*b*), are calculated as follows:
AE(L*U*V*) * [(AL*)2 + (AU*)2 + (AV*)2]
where
-------
87
L* = 116 (Y/YO)I/S - is,
U* = 13W*(u - U0),
V* = 13W*(v - v0),
and u and v are defined as
- 4X _ 6Y
U / u ', l rw i l-> \ s V
and u~, v0 as
(X + 15Y + 3Z) ' v (X + 15Y + 3Z)
4XQ 6YQ
U0 " (XQ + 15YQ + 3ZQ) ' V0 " (XQ + 15YQ + 3ZQ)
Similarly,
AE(L*a*b*) = [(AL*)2 + (Aa*)2 + (Ab*)2]
where L* is defined as above and
Ky \1/3 , v \l/3-|
— I - (~\
A0/ \rO' J
.0 " (zn)
In these equations, the tristimulus values X0, YO, Zfi define the color of
the nominally white object-color stimulus. In our atmospheric discolora-
tion calculations, we used values of Xfl, Y,,, Z- corresponding to the
reflected intensity from a perfectly diffuse reflector normal to the direct
solar beam. Calculations are normalized such that Yn = 100.
To determine the liminal or threshold (just perceptible) value of AE,
we computed AE for two color fields with identical chromaticities
(AU* = AV* = Aa* = Ab* = 0) and with a contrast of 0.02 (i.e., Y2 =
as:
-------
V3, ,v J/3
- 0.981/3^ -
3) - 0.78 (
Thus, for a bright horizon (say, Y, = TOO), we obtain a threshold or liminal
AE equal to 0.78. This value can be compared with a AE = 10, which is the
difference between two colors having identical Munsell hue and chroma but
with values differing by 1. Thus, AE can be used as indicator of atmo-
spheric discoloration: AE's less than 1 would be imperceptible, those between
1 and 10 would be detected as a discoloration by most people, and the sever-
ity of discoloration would increase with increasing AE. More work is clearly
necessary to determine what the standards of atmospheric discoloration
should be.
-------
89
IV THE OUTPUT OF VISIBILITY MODELS
This chapter discusses the outputs from several sample calculations
using our regional and plume visibility models. The models are described
from the viewpoint of a person who will be faced with regulatory, siting,
and design decisions based on the output of such models. We provide
samples of graphic display alternatives that can be used to translate
quantitative descriptions of visibility impairment into color samples,
perspective views, artist's renderings, and color television video displays,
These display techniques enable the user to understand the meaning of
visibility impairment models. Further details of the models and sample
outouts are given in Appendices D, E, F, and G.
The following models are illustrated by examples:
> Plume visibility model.
- Emissions from a hypothetical 2250 Mwe coal-fired
power plant meeting New Source Performance Standards.
- Emissions from a large copper smelter in Arizona.
- Emissions from a large coal-fired power plant in
Arizona.
> Plume/terrain perspective and color graphic displays--
emissions from a large coal-fired power plant in Arizona.
> Regional visibility model.
- 1976 and 1986 SOV and NOY emissions from sources in
A A
the Northern Great Plains.
1973 SOV emissii
A
and New Mexico.
- 1973 SO emissions from copper smelters in Arizona
A
-------
90
A. THE PLUME VISIBILITY MODEL
The plume visibility model predicts the visibility impairment
resulting from emissions from a single source, such as a power plant or
smelter . The model calculates the reduction in visual range caused by
the plume for several observer locations, and it also calculates plume
color, plume contrast, and color changes to determine whether the plume
can be distinguished by an observer. In this latter regard, the plume
model differs from the regional model, which calculates the visual effects
of a relatively homogeneous atmosphere. The plume model quantifies the
coloration and appearance of a plume in comparison with the homogeneous
background atmosphere and thereby characterizes the perceptibility of the
plume. The logic flow, program structure, and data requirements of the plume
visibility code (PLUVUE) are presented in detail in Appendix D. In this
section, we illustrate and discuss sample outputs of the model.
The user of the model must provide the source emission parameters,
ambient meteorological conditions, ambient air quality, and background
aerosol size distribution parameters. Exhibit 1 lists the parameters of
the sample calculation done for the hypothetical 2250 Mwe coal-fired power
plant, which was assumed to emit particulates, S02> and NOX at the maximum
rates permitted by the EPA's New Source Performance Standards. The user
must also select the dispersion coefficients (oy, az) to be used to com-
pute plume dilution as a function of downwind distance. The code has
subroutines for Pasquill-Gifford and for TVA dispersion coefficients, and
it will also accept values entered by the user.
After computing the initial plume dilution and N0£ formation during
plume rise from the stack to the location of final plume rise (1.2 km
downwind), the code calculates pollutant concentrations within the plume
and parameters characterizing plume visual impact at distances from 1.2
out to 350 km downwind of the source. Exhibit 2 presents an example of
the pollutant concentration parameters that are printed out at each down-
wind distance. In that exhibit, both the plume increments and the total
-------
91
a
E
a
5
SB
74
-s
I I
• i
^ «
g
i !
b. fc
I
5?
i i **
§ 5 S
I
c
g
i
t
i
*
3
B P s
W®
••w
c;
I
a
E S
H- O
rs z
Qu O
i— o
O h-
Q CO
o z:
z: <
LU O
s:^
Z3
-------
92
I
-1
i P
i I
K
S>
I i
T3
-------
93
nt- «"-
**
»h^
O
O
O
o.
ace
n ic
D.
O
UJ
o
»i
e
gg
> >•
> i* ww Wrt tr tj co rs rort $ *~
"&i ** C5C5 »•*- COK1 <"<> V®
1J C4d W P5 C4 W CO C? W W 41 **5
>e coco -- -- -- coco i-
H
E
I
UJ
I g
a »
II
fw-
OC
-a. e£
4» »
«, . . . «,
' & *9 9 N ^ COCO
vW^*0 OiC cOrt
I 55 SS si 9S 55 II
I
I
I
P
CO
I—I
:c
x
13
!>-N3
J* S
5£«im££
JJ-rfeg ^
£^l
nr
«i
S
CL,
>•
-------
94
ambient concentrations are displayed. The mole ratios of sulfate to
total sulfur and NC^ and nitrate to total nitrogen are also displayed, as
are the concentrations of ozone and the plume ozone deficit resulting from
NCL production. The plume increment and total particle scattering coef-
ficient (bsp at 0.55 ym) are printed out, as is the percentage of b con-
tributed by secondary aerosol (SO^, NO^). Note that in the example given
in Exhibit 1, at 10 km downwind, sulfate--which is assumed to form at the
rate of 0.5 percent per hour—contributes 46-1 percent of the plume scat-
tering coefficient, the remainder of the scattering is due to the emitted
primary particulate matter (fly ash).
Exhibit 3 provides the first of the three visual effects printouts
that the user can choose to display for each downwind distance. These
computations are done for sight paths through the plume center and can be
done for ground-level sight paths as well. Visual effects can be dis-
played for scattering angles 0 selected by the user (22°, 45°, 90°, 135°,
and 180°); only 180° (back scatter) calculations are shown in Exhibit 3.
Visual effects are calculated as a function of assumed observer location
relative to the plume. Observer location is specified by the distance
(along the sight path) between the observer and the plume at distances
that are 2, 5, 10, 20, 50, and 80 percent of the background visual range
and at four azimuthal angles with respect to the plume centerline
(a = 30°, 45°, 60°, and 90°).
In Exhibit 3, the first parameters printed out are the visual range
rv and the percentage reduction from background visual range. The follow-
ing two columns are the light intensity parameters Y and L*, described in
Chapter III. The chromaticity coordinates (x,y) are then displayed,
followed by two columns showing the differences in light intensity
(AY, AL*) between the plume and the background sky (without clouds). The
negative values in this example indicate that the plume is darker than
the background sky. The plume contrast (at X = 0.55 ym) is shown next,
succeeded by the blue-red ratio. The change in chromaticity coordinates
between the plume and the background sky is shown next (AX,Ay). Positive
-------
95
§n n P?
. . .
>•
-»*CI ««S« —
-* eo g ^ « ee n N r- •»
-------
96
Ax's and Ay's indicate a shift toward yellow-brown, and negative values
indicate a shift toward blue relative to the horizon. The final two
columns are the CIE color difference values AE(L*U*V*) and AE(L*a*b*).
To understand how the values in Exhibit 3 can be used to characterize
plume color, consider an observer's sight path that is perpendicular to
the plume (a = 90°) at a distance rp/rvQ = 0.02. An L* of 82.23 indicates
that the plume is bright, but not as bright as the horizon because AY,
AL*, and C are all negative. The chromaticity coordinates (0.3181,0.3253)
used in conjunction with L* (which is 10 times the Munsell "value") specify
the Munsell color notation, which in this case is 2.5 Y 8.912/0.6, a
weakly saturated yellow, essentially grey. The blue-red ratio of 0.9194
also indicates a slight, but perhaps not visible, yellow discoloration.
However, the contrast of -0.1611 and the AE values of 10.6 and 6.8 indicate
that the plume would be visible because it is darker than the background
horizon sky.
Exhibit 4 shows the visual effects of the plume for nonhorizontal
sight paths when viewed against a background of blue sky. Note that
visual effects are calculated for the permutations of a (azimuthal angle
relative to the plume centerline) and elevation angle 8 (15°, 30°, 45°,
60°, 75°, and 90°). These calculations indicate that the plume is more
distinctly visible against the blue sky background than it was against
the horizon sky (Exhibit 2). Note also that the plume is much brighter
than the blue sky background because AY, AL*, and C are all positive.
Exhibit 5 completes the characterization of plume visibility at a
given downwind distance by comparing the light intensity of the plume with
white (representative of a white cloud or snowbank), grey, and black back-
grounds at various distances from the observer behind the plume. The
plume appears somewhat darker and bluish in front of the white object
(REFLECT = 1) and brighter than the black objects (REFLECT = 0) at close
distances. The plume appears slightly darker than black objects at long
distances because the apparent light intensity of the black object distant
from the observer approaches that of the horizon.
-------
97
3««e)«te + «aei*n««
®co«. — (8«w + NN + -+4<
c?**p)gioc3egcoi'?M
j
H
A
O
M
o
0.
I
o;
o
u_
oo
v 9 OIO
< i£''#1*
U « CO CO IT) » 9 <) * 03 C5
^ d f> t«- QCO » ••
I/O
S
I
|
g
* V ••* S
u CO «9 «• <» IO * « ^»b- K5 * * C- >
-j E w— ' ' ' '<>i- '-
o
------- An error occurred while trying to OCR this image.
-------
99
The large amount of output required to characterize the visual impact
of a plume at a given downwind distance is necessary because of the large
number of possible observer locations, sight path orientations, and back-
ground objects. We have an optional printout table for plume visual
effects for only one given observer location for user-contributed values
of rp and a. This output option is useful for describing a plume for an
artist's rendering, as discussed in Section B.
We have designed a computer plot package to display plume visual
effects for horizontal sight paths as a function of distance. Examples of
these computer plots are shown in Figures 27 through 35. Four parameters
were selected that most easily characterize (with numbers) the visual im-
pact of a plume:
> Percentage reduction in visual range.
> Blue-red ratio (plume color relative to background).
> Plume contrast (plume light intensity).
> AE (L*a*b*) (plume perceptibility).
We have selected these plots to show the visual effects seen by an observer
situated at distance rp = 0.02 x ryQ and with a horizontal sight path per-
pendicular to the plume center! ine.
Figures 27 through 32 show the results of a sensitivity analysis to
determine the effect on plume visibility impairment of:
> Plume diffusion.
- Distance downwind (1.2 km < x < 350 km).
- Atmospheric stability (Pasquill C, D, and E).
> Scattering angle (0 = 45°, 90°, 180°).
> Sulfate formation (0 and 0.5 percent per hour).
(0 and 0.7 lb/106 Btu NOX emission rate).
The effect on visual range of the 3.4 ton/day primary particulate
emission rate assumed here (less than the 0.1 lb/10 Btu standard in order
-------
100
60.0
SQ.O
40.0
30.0
20.0
10.0
0.0
1.1
B
ii.o
a
j
JJ0.9
j
a
0.8
0.7
0.1
-o.i
U
1-o.a
-0.3
15.0
10.0
S.O
0.0
(1! NBRHRL NBX EMHSiaNS; 0.5 PERCENTER ?;..,_--
(2) MBRMfiL HBX EMI3SI0NS; N« SULFHTE F0RHHTI0N
(3i HB HBV EMISSIONS: o.s PERCENT .'HI? SULFHTE FT
«) NB NBX EHISSIBN3: N8 SULFflTE F£lPMCiTI8U
I ill
10 20 40 60
OfMNHIND DISTANCE (KM)
100
20-
FIGURE 27. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING
ANGLE OF 45° AND STABILITY CLASS C
-------
101
60.0
z 50.0
r°'°
,,,30.0
SIlO.O
0.0
1.1
= 1.0
a
ui
^ 0.9
(11 NBRHSL N0X EHI33I0NS: 0.5 "ERCENT •-is 5!.!,_-;7- -jif|.!;T;ti|,,
(21 N8RMRL NBV EMI5SI0NS; N8 3ULFSTE FBRMPTI«-tj /
13) N« NBS EHIS31BNS: 0.5 PERCENT/HR TilJLffiTE "T -TlflN
(4) MB NST' EMfSSIBN?: KB SULFfiTE
0.8
8:1
-0.0
-0.2
-0.3
JS.O
10.0
ui
a
S.O
0.0
10 20 40
DMNHINO OIBTRNCE (KM)
100
200
FIGURE 28. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 45° AND STABILITY CLASS D
-------
102
60.0
(I) NBPHRL NBX EHI3SIBNS! 0.5 FEfCEfiT.'KS SJAF;,-::
(21 NBRHBL NBX EHISSIBNS: NB SULFflTE FBRMST;?'.!
(3) H8 N8X EHISSIBNS; 0.5 PERCENT/HR SULFflTE ff
(41 NB NBX EHISSIBNS: NB SULFflTE FBPHRTI3N
z 50.0 -
40.0 -
30.0 -
20.0 -
10.0 -
0.0
1.1
1.0
d.9
0.8
-0.0
-0.3
15.0
to.o
5.0
0.0
I I I I i I I
6 10 20 40 60 100
DWNMIND DISTANCE (KM)
FIGURE 29. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 45° AND STABILITY CLASS E
-------
103
60.0
x SO.O
5 20.0
u
2 10.0
0.0
1.1
£ i.o
i 0.9
0.8
8.7
.1
. -0.0
8-o.t
-0.2
(1) MBRMRL NBV EHI33I3N3; 0.5 PERCENT/MR SULFflTE FBRHRTiaH
(2) NBRHRL NflX EHISSI8NS; H9 SULFflTE F0RKHTI0N
(31 N0 N8)t EMISSI8N5: 0.5 PERCENT/HR SULFflTE FBRHSTIBN
It) NB NBX EHISSIBNSi NB SULFflTE FBRHRTIBN
-0.3
1S.O
10.0
6.0
0.0
10 20 40 60
(MMNHtNO DISTANCE (KM;
100
ZOO
FIGURE 30. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 90° AND STABILITY CLASS C
-------
104
60.0
50.0
40.0
SO.O
20<0
UlO.O
0.0
1.1
I 1.0
(1) NBRMflL MBS EHI35IBN5: 0.5 FERCENT/HR SULrfiTE F3f»KJ
(2) NBRMflL NBX EHIS5UNS: MB SULFflTE FBRMflTISN
(3) NB HVt EHISSIBNS: 0.5 PERCENT/HR SULFfiTE FBRHflTiaN
(4i NB NBX EHISSIBNS: NB SULFRTE FBRMSTIBN
0.8
0.8
8:1
-0.0
1-0.1
• -0.2
-0.3
15.0
1 i J I 1^ 1
10.0
I
5.0
0.0
10 20 40 60
MHNHIW OI6TANCE (KH)
100
200
FIGURE 31. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 90° AND STABILITY CLASS D
-------
105
60.0
50.0
£30.0
>
5 20.0
u
Kio.o
0.0
I.I
I
: 1.0
I
, 0.9
II! NBRHflL NBX EHI3SI8NS; 0.5 PERCEHT/HR SULFfl'E FOf rinT
(2) NBRHflL N8J EHISS18NS; N8 3ULFRTE FBRHBTIBH
(3) NB N8X EHI3SI8N5: 0.5 PERCENT/MR 3ULFHTE FBRH^TIBN
(4) NB HBX ENIS3IBNS; MB 3ULFSTE F0RMBTI8N
0.8
-0.0
I -0.1
-0.2
-0.3
15.0
I I I I I I
10.0
u
I
5.0
0.0
10 20 «0
DWNHINO DISTRNCE (KM)
100
200
FIGURE 32.
CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 90° AND STABILITY CLASS E
-------
106
60.0
50.0
10.0
ao.o
20.0
lO.O
0.0
1.1
(1) N0RHHL N0X EHI3SI0N3J 0.5 PERCENT/HI! %F«Tc
(2) N0RMHL NBX EMI3SI8HS: N8 3ULFATE F0RHRTI0N
O) MB NBX EMISSIONS; o.s PERCENT/HI! SULFRTE Fen
(4) M0 NflX EMISSI0NS: N8 SULFRTE FBRHATI0N
S
I'
0.8
8.7
.1
-0.0
! -o.i
• -0.2
10.0
S.O
0.0
10 20 40
OWWHIHD DISTflNCE (KH)
100
200
FIGURE 33. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 180° AND STABILITY CLASS C
-------
1U7
60.0
xSO.O
§40.0
o
£
.,30.0
>
5 20.0
£ to.o
0.0
1.1
§ 1.0
£0.9
0.8
.-0.0
[-0.1
• -0.2
•0.3
15.0
(1) NBRHSL NB» EMISSIBNS! 0.5 FERCENT/HR SULFfiTt FBRHRT18N
(2) HBKHRL N0X EHISSIBN3; Ng 3ULFHTE FBRMBTIBN
(31 Ng N8X EM1SSI8NS! O.S PERCENT/HR SULFflTE FBRHflTIgN
(4) N8 N8X EHISSIBNS; N8 SULFfiTE FBRMBT10N
to.o
ki
I
s.o
0.0
10 tO 40
DMNNIND DISTANCE (KM)
100
JOO
FIGURE 34. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 180° AND STABILITY CLASS D
-------
108
so.o
z SO.O
^ 40.0
^30.0
>
5 20.0
s
£ 10.0
o.o
1.1
I 1.0
(1) NBRHflL N8I EHI33I8N5: 0.5 FERCENT/HR SULFSTe FSifwaT
(21 ttflRHflL N8T EM1S3IBN5: N0 3ULFRTE FBRHATI0N
(3) NB M«X EMISSIBN5: 0.5 PERCENTER 3ULFBTE FPRHflTISH
(4) NB NBX EMISSIONS; NB SULFPTE FBRHRTIBN
g
A
0,8
-0.0
1-0.1
• -O.Z
to.o
5
s
E.O
0.0
10 20 <0
DWMMINO D1STRMCE (KM)
60
100
£00
FIGURE 35. CALCULATED PLUME VISIBILITY IMPAIRMENT
FOR A HYPOTHETICAL 2250 Mwe COAL-FIRED
POWER PLANT WITH A LIGHT SCATTERING ANGLE
OF 180° AND STABILITY CLASS E
-------
109
order to meet the 20 percent opacity standard) is apparent in these figures,
particularly for stable atmospheric conditions (Pasquill E). However, the
significant reduction in visual range caused by primary particulate is
noticed only at short downwind distances. All these effects are for sight
paths through the plume centerline. At large downwind distances, the
effect of sulfate on visual range becomes very Significant, particularly
for stable conditions. Indeed, the reduction in visual range increases
with increasing downwind distance between 60 and 350 km downwind.* The
scattering angle has a small but not negligible effect on visual range:
Visual range increases with increasing scattering angle. In other words,
the reduction in visual range is greatest for forward scattering.
Plume coloration and contrast are indicated by the values of the blue-
red ratio, plume contrast (at A - 0.55 pro), and the CIE color difference
parameter AE(L*a*b*). The effect of NO^ on plume coloration becomes clear
when the curves (Nos. 1 and 2) that assume normal NO emissions are com-
X
pared with those curves (Nos. 3 and 4) that assume no NOX emissions.
Yellow-brown coloration, as indicated by blue-red ratios less than 1.0, is
stronger with N02 than without. Note that the effect of sulfate (Curve 3)
on color is very small for all scattering angles; however, sulfate has a
significant effect on plume contrast, increasing plume brightness at small
scattering angles and decreasing plume brightness at large angles. Compar-
ing Curves 1 and 2, we can see that the pronounced coloration caused by N02
during stable conditions (E stability) is reduced by light scattered by
sulfate. If we use AE(L*a*b*) as an overall indicator of the perceptibility
of the plume (because of both plume contrast and color changes), we find
that N02 has the most pronounced effect on plume visibility at significant
downwind distances. Sulfate has a significant but smaller effect, and'
primary particulate has the least effect. Perhaps the most significant
result of these calculations is that plume visibility impairment (both a
* A cautionary comment is in order here. For this evaluation, the wind
speed was assumed to be 5.0 m/s; thus, almost 20 hours would be re-
quired for emissions to be carried 350 km. It is unlikely that stable
atmospheric conditions will persist that long.
-------
no
reduction in visual range and an atmospheric discoloration) increases with
downwind distance, suggesting that a significant impact could occur hun-
dreds of kilometers from the source.
B. PLUME/TERRAIN PERSPECTIVE MODEL
To supplement the quantitative description of plume visual impact
described in Section A, we developed a Perspective Terrain Viewing Program
(PTVP). Using computer graphics, this program is capable of displaying
views of plumes and background terrain with the perspective of the human
observer or camera situated at user-specified positions. These plume and
terrain perspective scenes can be used in conjunction with the quantifi-
cation of plume visibility impairment discussed in the previous section to
provide an understanding of the subjective impact of the computer predic-
tions. In addition, these scenes, along with computed Munsell color nota-
tion, can be used by a commercial artist to produce color renderings of
the visual impression of the background atmosphere and the plume for
various assumed emission conditions.
To use the PTVP, the user must provide the following information:
> Boundaries of the region in which the facility is situated
or is to be constructed must be identified, and terrain
within it must be digitized. (The U.S. Geological Survey sup-
plies digitized terrain elevations.)
> Design parameters of the facility that affect the effluent
plume characteristics must be determined. Among these are
stack height, flue gas temperature, and flow rate.
> Representative meteorological conditions must be specified;
important parameters include wind speed and direction,
ambient temperature, lapse rate, atmospheric stability
category, and the height of the inversion layer, if one
is expected to exist.
-------
Ill
> Observation parameters must be decided upon. Among these
are the location of the observer with respect to the
facility, the direction in which he is viewing, and the
field of view of the "conceptual camera" he is using to
record the scene. The last of these parameters is required
because the PTVP uses the lens characteristics to reproduce
the optics of a camera, "recording" by means of computer
graphics the appropriate camera film image. Lens "size"
as measured by the cone-angle of the field of view may
range from a few degrees (telescopic lens), to from 40° to
50° (standard lens), to 90° ("wide-angle" lens), to 180°
("fish-eye" lens).
The use of the PTVP can best be illustrated by means of an example.
Digitized terrain elevation data were obtained from the U.S. Geological
Survey for an extensive portion of the Southwest. From this data base,
a 50 x 50 km portion of terrain immediately west of Page, Arizona, was
isolated. A computer-generated plot of the terrain is presented in
Figure 36. Among the prominent geographical features contained within.
that region are the Vermilion Cliffs, the Marble Canyon through which the
Colorado River flows, and the Paria Canyon.
In this sample terrain grid, a plume from a hypothetical power plant
was displayed and viewed from several different observer vantage points.
In the example, the power plant stack is 775 feet high, plume rise was
determined using values typical of a large coal-fired power plant, winds
were light and headed slightly south of due west (compass heading of 2556
and meteorological conditions prevailed that are typical of Pasquill-Giffo'rd
Stability Category E (stable). The plume was assumed to be Gaussian, with
its "envelope" defined by the locus of la dispersion coefficient values.
The hypothetical observer in this example flew around the power plant
observing the power plant plume. Using a "camera" having a wide-angle lens
(with a 90° field of view), the observer took a series of pictures. The
-------
112
00
O
LU
CL.
OO
i
LU
Q_
a:
o
u.
oo
o
oo
ca
o
to
-------
113
location of each picture is shown in Figure 36. The sequence of pictures
is presented in Figures 37 through 51. All views are directed at the
power plant, except for Figure 39, which is aimed west toward the Vermilion
Cliffs.
C. COLOR DISPLAY TECHNIQUES
The most complete and realistic display of predicted visibility impair-
ment, particularly atmospheric discoloration, is a plume-terrain perspective
view in color, with accurately specified and rendered colors calculated from
the plume and regional models. We investigated two methods of displaying
atmospheric discoloration and plume visual imp-set:
> A color illustration, drawn or painted by a commercial
artist, using Munsell specifications for plume and
background color and plume and terrain perspective
views.
> A color video display, based on a photograph of a view
from a vista, computer-enhanced to display a plume or
homogeneous atmospheric discoloration on a color
television.
Both of these methods, presented schematically in Figure 52, use the spec-
tral intensity I(A) calculated by the visibility models for specific lines
of sight as a base.
These two display methods are the most technically difficult, time-
consuming, and expensive output options available for visibility models,
but they may be the only ways of giving the user of models an understand-
ing of the calculated visibility impairment. Without the aid of these
color display techniques, it is very difficult to translate numbers
describing visual impact into an observer's actual visual impression.
-------
114
c_>
c
o
o:
co
UJ
a;
-------
115
5
o
o
cc
CO
ro
UJ
-------
116
CO
o
5
o
o
CsL
LU
O1
CO
uu
01
-------
117
s
o
o
or
o
-------
118
LO
Z
o
o
e;
-------
119
UJ
I—t
>
C\J
*3-
UJ
OL
Z3
CD
-------
120
r-.
o
i—i
h-
o
_i
o
co
UJ
-------
121
CO
2TT
o
O
O
o:
-------
122
-------
123
•a:
O
o
o:
CD
-------
124
5
o
o
C£
Lu
-------
125
CVJ
CJ
o
o
c;
UJ
CO
^j-
UJ
cz
-------
126
co
o
o
01
UJ
o:
-------
127
-------
128
-------
129
d.
CO
a:
o
o
es
o zr en 3E
U-l I-H 2= 0.
CO
tn
CJ
i—i
u_
-------
130
1. Color Illustration
This technique is a synthesis of the regional and plume visibility
model outputs and the Perspective Terrain Viewing Program (PTVP) illus-
trated in Section B. The process requires the following steps:
> The source characteristics and location are selected,
and the digital terrain data are obtained.
> The Perspective Terrain Viewing Program is used to
generate a plume terrain view for a given set of
meteorological conditions (wind speed, wind direction,
stability category).
> The plume visibility model is run for the same set of
source and meteorological conditions. The coloration
of various sight paths through the plume and the back-
ground sky are predicted.
> Specific areas of the plume terrain view are assigned
the appropriate Munsell color notation and associated
color chips.
> A commercial artist colors the plume terrain view.
The color chips are used as a reference check on the
artist's color display.
To illustrate the capabilities and limitations of this technique, we
have constructed a test case that compares our predictions with the actual
visual impact of a power plant plume. The comparison demonstrates the
need for carefully documented studies of the accuracy of the model.
A test case should have the following attributes:
> A large point source with a visible effluent impact.
> Documentation of the source emissions (NO , S09,
}\ £
primary particulates).
> Location near a nonurban area of great aesthetic value.
-------
131
For our test case, we selected a power plant in northern Arizona for
which we had color photographs documenting the visual impact of the plume.
The source emissions and dispersion conditions have been studied, but
these data were not available to us. However, we were able to estimate
the source emission conditions from other similar power plant sources.
The photograph we selected is reproduced in Figure 53, which clearly
shows the brown coloration caused by the power plant plume. We chose this
example because of the clearly apparent brown coloration. However, upon
closer inspection, one notices that the lighting conditions are somewhat
unusual because the photograph was taken very early in the morning. The
sky is less blue than normal and is slightly yellowish at the horizon. The
long shadows of the river canyons are visible, indicating a very low sun
angle. As noted in Chapter III and Appendix B, the diffuse component
(multiple scattered light) of the solar intensity becomes significant
near sunrise and sunset. Since the diffuse component is hard to model
correctly, particularly in extreme situations like this, this photograph
represents a difficult test case.
For our estimation of the source characteristics and meteorological
conditions, the Perspective Terrain Viewing Program was used to generate
a plume terrain view, which is shown in Figure 39. A comparison of Figures
39 and 53 shows that the terrain and plume locations are rather faithfully
reproduced. Although the resolution of the photograph is much greater than
that of the computer graphics algorithm, the resemblance is clear. The
distant mountains on the horizon (dark blue on the right-hand side of the
photograph) are not plotted because they were outside the terrain boundar-
ies of the program for this particular case. The plume boundaries are
plotted at la (a ,a ) concentration values. The vertical extent of the
plume in the photograph is less than it is in the computer plot, suggest-
ing that the actual plume oz was less than a Pasquill E stability. Plume
concentration measurements would be required to substantiate this assump-
tion, but the usefulness of the PTVP is clear.
-------
132
srs:
C-
0£
1-0
O LU
O- 00
I— o
M
U. i— l
Si
O LU
o :n
re o
CO
LO
-------
133
The plume visibility model was then run for our estimated source and
meteorological conditions. The Munsell color notation and associated
color chips for various parts of the plume terrain scene are shown in
Figure 54. After comparing the color chips and the original photograph,
we concluded that the colors are reasonably close. The yellowish color
of the horizon and the brownish color of the plume are reproduced, demon-
strating that the plume model is capable of predicting the brownish color-
ation and displaying it correctly to a user of the model. The sky color
is approximately the correct saturation and brightness.
The results of the final step of having a commercial artist paint in
the correct colors are shown in Figure 55. The artist was never shown
the original photograph; he had to rely on the color predictions from the
plume visibility model. Unfortunately, we are less satisfied with the
results of this step than with the previous two. The problems in this
step appear to be that:
> It is difficult to paint and blend the correct colors
to maintain fidelity to the predictions.
> The spatial resolution necessary to produce a realis-
tic scene is also difficult and requires a large amount
of time.
> The artist has a natural tendency to paint what he thinks
the plume should look like.
Despite these difficulties, we believe the technique has promise and should
be pursued, though more work on this step is needed.
Overall, we are encouraged with the results of the comparison. The
significant findings were that:
> The Perspective Terrain Viewing Program can generate
a realistic plume terrain scene.
-------
134
UJ
"O
Q-
UJ
I— a:
UJ o
(/)
Qi
LU i
is
-------
135
00
CL
l—o
<— J
u3o
0:0
OLU
C/)Q
ctro
02:
a.
o >
Is
cc
^
C3
-------
136
> The plume visibility model can predict the plume
coloration correctly.
It is important to emphasize that these findings are somewhat preliminary,
and testing must continue to verify computer models. In addition, more
information must be gathered so that other tests can be conducted.
2. Color Video Display
In addition to the color illustration described in the preceding
section, we investigated the possibility of using computer-generated color
display facilities. This technique was originally developed for proces-
sing photographic data, particularly satellite data, and it requires
special equipment, including:
> A color densitometer to digitize a color photograph
in three colors.
> Image enhancement software to allow manipulation of
the digitized information.
> A color video display unit and supporting software.
Although these facilities were not directly accessible, we were fortunately
able to acquire the assistance of researchers at Los Alamos Scientific
Laboratory who have used this technique. The Los Alamos personnel utilized
this technique to predict visibility impairment from power plant plumes
(Williams, Wecksung, and Leonard, 1978).
We sent Los Alamos the test case photograph (Figure 53), which was
then digitized into three colors. Then the plume was removed from the
digitized photograph by interpolating the sky intensity from the horizon
below the plume to the sky above it. Next we gave the Los Alamos per-
sonnel the results from our plume model for specific locations in the
plume. These intensities were displayed on the color video screen,
and a photograph was taken. The results are shown in Figure 56. This
-------
137
o
QL
S
i—i
—I
-------
138
figure is then an illustration of the color video display technique. The
photograph can be compared to the test case photograph (Figure 53).
Because of compatibility difficulties between our computer output and the
Los Alamos facility, the results in Figure 56 should be considered as a
qualitative indication of the technique.
The color video display is a very powerful technique that gives the
user a rendition of a photograph with the plume effects superimposed. The
technique is somewhat cumbersome, however. Specifically, the difficulties
are the following:
> The hardware is expensive, available only at specific
locations, and not available on a dedicated basis.
> It is difficult to transfer information from the ori-
ginal film to the final produced photograph without
introducing errors. In other words, a quantitative
measure of the color fidelity of the process is not
possible at present. This difficulty is due mostly to
the problems involved in film processing.
This description of the difficulties of the color video technique is not
meant to be a criticism of the Los Alamos personnel and their work. It is
simply a listing of problems that must be faced in using the technique on
a routine basis.
D. THE REGIONAL VISIBILITY MODEL
We have modified the Northern Great Plains regional grid model (Liu
and Durran, 1976) so that it has the capability to compute regional con-
centrations of N02 and sulfate. We show in this section how these pollu-
tant concentrations can be used both to display visual range isopleths
for the region and to characterize atmospheric discoloration at specific
locations (Class I areas) within the region. We summarize the results of
sample calculations using 1975 and 1986 SO and NO emissions from point
A X
sources in the Northern Great Plains. Also, using 1972 SOp emissions from
-------
139
the copper smelters in Arizona and New Mexico, we constructed a hypotheti-
cal situation by assuming that these sources were located in the Northern
Great Plains. The objective of that task was to study the effect of the
large SO emission rates from the copper smelters (6000 tons per day) on
X
regional visibility using the existing Northern Great Plains regional
model. The significant impact of copper smelter SO emissions on visi-
X
bility in the Southwest is indicated by the results of the data analysis
described in Appendix A and the regional model calculations reported in
this section. This impact suggests the need for a regional visibility
model for the Southwest capable of handling the transport and diffusion
of copper smelter emissions as well as power plant emissions in complex
terrain.
Figures 57 through 60 show the isopleths of S02, N02> and sulfate
concentrations and visual range calculated using the regional grid model.
N0? concentrations were calculated using the technique described in
Chapter III from total NO emissions assuming a background ozone concen-
X
tration of 0.020 ppm. Sulfate concentrations were calculated from SOX
emissions using a pseudo-first-order rate constant of 0.5 percent per
hour and assuming negligible primary sulfate emissions. Visual range was
calculated from the Koschmieder relationship using the following value for
the extinction coefficient:
bext = (°'24 + °-04[s04' in M9/m3])(lO"4 m"1)
With the assumed background SO^ concentration of 1.5 ug/m , this
expression gives bgxt = 0.30 x 10"4 m'1 , which corresponds to a visual
range of 130 km. The bscat-to-mass ratio used here (0.04 x 10"4 nr1/
yg/m3) is appropriate for sulfate aerosol in the accumulation mode at
average relative humidity, and it is the average reported by Trijonis
and Yuan (1977) for the Southwest.
The calculations of visual range (Figure 60) indicate that anthro-
pogenic emissions from point sources within the Northern Great Plains
-------
140
en
£
— I— 10
r--
oo a: CTI
z o •—
o u_
z a:
i— i a.
LU a.
O et O
o LU
o: I—
C3 OO
o z:
oo z:
a: o
LU LU O
O O
Z O
O O
:r z
I •-( c^
LU z:
LU CO O
oo o
— i to z
-------
141
!:;'•
o
o.
0
O)'"
o.
td
0
r*"
o.
03
O.
Ml
O.
O.
o.
«M
O.
""*
or, oo u.'.
„- -"'*"
: "
"
•
.
•
•
P ..•"""<•
^?
,. -•'" ..- ' •.. t
:\ ".-•;""" \ [
.' .• * * " jt
\ *. '•••••.
''•». -( •-„.., /
ft/ f
'.,. , /
*--o: """ - ' •'"
- t— *'• — , .-•*' '
- g <
;g .v-
:'•'*"""""
- cc
z
cc
" ^ .••'""
• x _.-•"
/•' .. •''
V, ..' ..
.* • .' *
' .*' ** •** f "'^t
. **" »*
•>:.!•••""' >
.• *" .•''
*' /*'
*'' .•-'""
• ' .»••••"***
' '.3
00
^y
T
1—
x:
^
0 SE 0
| 111 .W—
1 Ivi-f ^^
a: x:
ZD 1 — U.
-0 °..°
— • 1— 1
•n 00 O
— 1 CO Z
«=C I-H O
Z S 0
O UJ
_ t— < — 1
•2 C3 ^D o r^
H-. O O
a LU uj
uj oo i—
OS
.0 W
"" rs
CJ3
H!
06
QL 08
OS
oe
01
-------
142
06
09
OL
09
OS
ot
OS
OZ
01
o
o. -
o.
O>
o.
ID
O. .
:•-•§
»'tt i
B-!2
i
a
tn
o. .
Ift
-•• ----------
O.
(O
i—i
<: en H- o:
i ra «c Q-
CC O UJ fS.
•=>^&.
O CD VO
:c a:
I LU Z Z
LU a. o: o
LU LU
O
_l C£.
^C I1'
Z O-
o
i—i UO
CD •
LU O
a:
a
LU CD
o
•z. o
o
LU r~-
h- T
o
z o
oo
z: u_
O O
O i— « l~l
a o
LU oo
0£ 00
Q- —
LUt-
-------
143
06
03
o. .
en
o.
CD
o.
<0
o. .
ir>
Oi
9-Hf
09
05
Ob
oe
o?
01
••••;> -on, \
ii
ct
o
- z
*
o
6-
--.__„„.... 12D
021
"T»B •'"T
•ft
IT
CO
bJ
130
06
08
09
OS
a
o
3
s
*o.§
. .o
,4.0
Ok
06
0?
01
o
/:.o
«
. .o
. .a
21!
LU CC
gj
Q
UJ O
o: oo
O. r—
o
10
UJ
Q:
LU •—i
ar o;
fc S=
Q_. *^
o
UJ 2:
re o
oo
00 O
z o
o r^-
f—l r—
OO I
oo o
>-> o
-------
144
cause only a small reduction in visual range, even with the increased
emissions projected for 1986. Maximum visibility impairment (i.e.,
minimum visual range) occurs in only a small part of the region. Visual
range is predicted to be reduced from the background 130 km to about
100 km, approximately a 25 percent reduction in visual range. This
small reduction would be difficult to measure, particularly using visi-
bility target observations, and to separate from visibility impairment
caused by natural sources. The calculated reduction in visual range in
the Northern Great Plains resulting from 1975 emissions is even smaller.
The results of calculations for 1975 emissions are presented in Appendix G.
Anthropogenic visibility impairment is significant in the Southwest,
however, as indicated by the analysis of visual range data presented in
Appendix A and summarized in Appendix B. In southeastern Arizona, copper
smelters emit large quantities of SO ; in 1972, before pollution abatement
A
equipment was installed, copper smelters emitted an average of 6000 tons
of SO per day, which is more than an order of magnitude larger than the
/\
1975 S02 emissions from point sources in the Northern Great Plains (430
tons per day) and three times the projected 1986 SO emissions (1990 tons
A
per day) in the Northern Great Plains. Although we have not developed a
regional model for the Southwest capable of handling the complex flow pat-
terns caused by rugged terrain, we examined the impact of the copper
smelter SO emissions on visual range by locating hypothetical copper
A
smelter emissions sources in the middle of the Northern Great Plains grid.
The hypothetical emission sources had characteristics identical to the
copper smelters located in Arizona and New Mexico that were operating in
1972.
We then ran the grid model for the meteorological conditions of the
Great Plains to evaluate the impact of the high SO emission rates from
A
the copper smelters. We summarize the results of this calculation for a
given time period in one of the test simulations in Figures 61 through 63.
The state boundaries have been removed from these plots to emphasize that
-------
145
c»
*— «~
o
o.
o.
0.
CO
o.
o.
UJ
o.
o.
«*•
o.
OO
o_
C*J
o.
06 08 OL 09 OS Ofr 06 01 01
.... •' .- .-•••"-.. *..'"'.
..-• - ? .', ,' •; \ •'..
'<•' '.'•'''
• •"' .' .' •'' ..'
" ' .' _ .' ..' ' ,, .-'" .. ••
: X" ."••'////>"., •"'" / :
,.-'' • • f <••'' •' ' . '
•: ' i ••'•'/
.' .' / ; / 1 ''
'. . '' /' ..•>' .-' •'.' /'' !
.' .' / ? ,'' '
•' ••' ' t* ' •''
/,',•''
'"//''' '
-•"' •*' / •'* ' * r*"*****! •"" »
' * '•*"** "*"' ' ' WT '
; /*/.•"""" >/ / :
•l • •".*,•* .
.. *' .-••"* ."•'* .•• '' .*".*' '"' •'*
" \\T\'- * -""!."•-• '''"..-'" '''-'? w^ .-' .-.--
--••"""""-*" ^ " • ' ' *fll';^ ''-f I • '"/' t»"T"'i"tt*'
*"....."" T •• " " .•''"<£' ,- ' " " ' j »fc5 .*' "*" "•• •'•'•' ' "5i»*' *
: "**••. •'--"' >/•'•' ••.'•>:-;;' ••-"' ;
' ' f'\^Lr'e t ' • ^"^ .'"•','
' *-' V '• «?*"••;•,:'' "' .-'"'
• " ••''.-'':' -.;.;:•• V '' .-,>'''
- " ' -"_•".' •"" ' •'! " •**' >^"'*
..... .... . .-•''_.,._ - ' ,;, %(•-'." ',.•' • ^ "
.. '" % '•''"' y.:'"- • ',."' '
,»•• '" . • ' ."/'..•"-'*" ,<•*' •' *". * -•' .•"'
* " - \V '' * '' ' ' "'"' ' fe .*" • ' ' •*'
" **-. (M :- •-'"!'*''.- ' . " .-*" t' *
'" ' . ' • .— --"*'" • '*' .* ' .'" •'
*'•-., * "' .-""" • '' • ' • *
"*'* ..,- • - * ,. '*\ '"•',''
,•-' ^'
. .-'•'. v ..
*•?> .-"',-v .-•" ,'' . '''
. ' f --••" ' ,..-*'* >
/'^ . ^ . ' ' •''*
/-'* »
* ''
-.'_.'"
-
,...•''' ,-•-••• " I — .......
/ . ' " ..-• * "" "*'•. ..^
"^ ...-'" ..,. .•-•" "*•— • / v
[ ^ ^ ..--•" ..,..•• * "'""- .. "•""
..-••'
(
"
- / ^
06 08 QL 09 OS 0> 06 OZ 01
~JU
..
o
•^
o
.0
.0
0)
.0
00
.0
r-
.0
— i
OO _J — 1
E •— <
to z
O «3
III ; [ ^^
f n c/o ^t
LU sr i
S> LU t— "
•a: ce
I O£ CX
rv LU O
LU »— i O
LU r—
Q 3C
LU 1 — LU.
t— O O
CJ Q-
t— i >- (/)
C3 DC 2T
LU O
fV^ ~Z^ >-H
Q- O h-
__:
UD
LU
O£.
CD
I-H
U.
-------
146
06 08 Oi 09 OS
Ob
oe
QZ 01
o
— '
o
0-
o_
o.
03
O.
^
0.
'
' / /' .-•'£"
-
* •' I*
/ / /' •'''''
/
' •'
{**
tfr
•'',"':
/
'" ' .' ,!
. P
*v
/
«' *x
''*.'"
.-"".•''•'••
"" ••" " ,.••
-- -- 4 -•""" / N
" .- " • -.. .-' .''' *
•••-._. -
(t) •
/ .-*' • -
W .-•"'' ,•'•'*"''
.'* . '**' *'"
""""-. .* ' .••' .••'"' ,;
„•' ,..'•" .''
{
'.
.,••'
- ... ' ..--"" 'i
\ .'' ..-'*' f
.' -'* •'*' i
''
t"
' ^ W.''
.
-..•'"'
•
.
-
•
i ;
06 08 OL 09 OS Ok 06 OZ 01
r— fV)
x -^ Z
01 O
O f~*
co o o co
^y 1 1 1 121 y^*
O CO O i— i
t-i co co
eC 00 MD
i D: i— i r^-
.0 GZ. rD sr CTI
vo ra o LU i —
o :c
•y C^ 1
i a: LU i— <
LU LU (— C£
LU O. _1 Q.
-O fV LU cf
"* 31 1— 21
1— Z CO U3
LU
_i c_> o: z
1
*^^ <~^/ t^^ 1
1— Z LU O
0 •-< x; o
O ^D O r—
-2 UJ CO D.
°* oi co >- u_
Q_ et 3; O
.0 ^
LU
a:
rs
U-
-------
147
c;
(Zl
o
o.
o.
O)
o.
00
0.
o.
to
o_
in
o.
<**
o.
W>
o.
rg
O.
06 08 Oi 09 OS Ob 0£ 0? Ot
•'.-.'/:- .''"'VV' \X", 'r
•.<-'.', ! ••' .: ' '.• 'c0 r
• ^«! ,' .' . .' \ •*! ,1
. ' .' ' " ' ' • ' / •'
'• •'' •'& •'•'.-'" ...'" ,'"
• ,• . :'Q> •' •' •' .•'" I'
1 $/// '/// o
: "N •' • i .' iS.- .•' <""•
; /••''/ ••'!/' >^/ ""
/ / / i'" •'.•**.•'
.' . * / .• •' •
,• ' * ~ /f ••'
• '•/'/ *
. .- . .-' * ,-' i
,• .• / .' ' j -
.' *s •' f / > 1.
• ' -V / •' '' i
••' 'i^5 • / •' ••' L
". /:'*?/'! i I / ;_
; ^ (j . '...•',.••/ //// i;
..-••"' v '.-•*' •' ..•' -•' • V - !•
-•-.•-.::• - •-• '••-•.:".-" - • \"^r ..--, .-''.••'/ / f
* •:'":: l;;--:-"- ••- -• iX .-- •- "" ,'' /!»* '.
- ...;—•--..-- «v / / _>v "- c5
>"""*... " " % *'-.' ..'' / .-•"' 2
(5j * i *' * /* * ™
. , —,-* •- •*• '* •' . t' Cj J m
"n • . • ' •'' .' ' / •'' ^- (
: "" V''"-''-''/ •••••-""" ^" ' :
***/•' •' i \ ••"'' t ' 1
'1 j' i *. '-• X *•"
1 "•( \ '.' \* -•''
" ' • / / 1-0 . •' ,.-•''
^' i ,-' ••' •o* » .•' ' •*
>' .;' '' ,'4' i''
//'.'.•'
, '^i'-' •' /''
. ' '.^ . ,•'''
'''•''•'
'*•• . ''••''••<&> "
VV' -•' "*
"••' .'
.'
/*'"
^'
•
/
/
.-•''
..." /'"""' *
--•—'*"'' t'' \
.-.^— «*™***-""~" 3 30 '" *"'"" .*' .
06 08 Oi 09 OS Ot 06 OZ 01
-x ;c o
i— 2: Q- OO
< >- 2:
x a: z
o
o -fZ.0
oo o
0 Q CO f—
.0 -^>
^4 ^—
O " U-
O UJ O
o: (—
C3 et OO
^ cc 2:
.0 0 0
O) ef 2: >— '
CQ 0 (—
<£ 1— O
2 ^
•o 2: o; o
•D __ O
s: u i
oo uj o oo
OO I— •— i ^
•° ^ U- O <
^-Z3 o a.
E oo o;
^1 ° *"
.0 3 (— UJ
UJ UJ O UJ CC
CD 3C S C3
cC C£ Di 2t
a: uJ o cc
o U_ \ ' '
"10 i< i— oo ^
Z3 2* 2= CC
OO UJ O O
t— * C-> >— < 2T
> CC OO
0 UJ 00 UJ
•• i ex i— -i re
< 2: H-
z: uo uj
o • zc
i — i O CC i — i
CD UJ
-O I i I CC _J I^~
Q UJ CTi
Q -z. n »-
UJ cC 00
1__ |
r~~ —J
0 o E cc K-I
"AJ '""' -^ Lx* CC
" Q Q. a.
UJ O Q- ct
cc co o
O- r~ C_5 **O
.0
oo
to
UJ
CC
C3
t— •
-------
148
this is a hypothetical test case. We assumed a 0.5 percent per hour sul-
fate formation rate in these calculations and negligible primary sulfate
emissions. The results of a sensitivity analysis using different sulfate
formation rates are summarized later in this section and are presented in
detail in Appendix G.
Figure 61 shows that the maximum Sf^ concentrations occur at short
distances downwind of the hypothetical smelters, which are located in the
center of the region. However, as shown in Figure 62, the maximum S0|
o ^
concentrations (greater than 8 u9/m ) occur hundreds of kilometers down-
wind of the smelters. The impact of this sulfate on visual range is shown
in Figure 63. The worst visibility (less than 60 km) occurs in one small
area in the upper middle portion of the region. The increased sulfate
concentrations and resultant decreases in visual range occur in two direc-
tions from the sources as a result of a change in wind direction that
occurred on the day before the simulation. The calculated concentration
maps for other time periods of this simulation period are presented in
Appendix G.
The impact on visual range of assuming different sulfate formation
rates (0.3 and 1.0 percent per hour) is indicated in Figures 64 and 65.
Note that with the reduced sulfate formation the minimum visual range is
70 km, and with the increased sulfate formation rate it is 40 km, which
compares with the minimum visual range of 60 km computed for the base case
of 0.5 percent per hour.
These sample calculations of reduced visual range cannot be compared
directly with the observational data from the Southwest because these cal-
culations were based on Northern Great Plains meteorological conditions.
However, the results agree qualitatively with some of the conclusions of
the data analysis summarized in Chapter II: namely, SOX emissions from
copper smelters in the Southwest can cause a significant reduction in
visual range even at locations several hundreds of kilometers downwind.
The predicted maximum SO? concentrations in these simulations, ranging from
3
8 to 16 vig/m, agree qualitatively with measured maxima in Arizona. The
-------
149
o
-!'
o
o.
«-*
o.
at
o.
CD
o.
fw
o.
ID
0.
in
o.
«r
o.
Ol
o.
'' .*•-"' i'
•'. // *•*' /'' / o -
.-' / / / '•' m .
/ f '
/ /
; .• .• /
.'* -'* •' f
,' / / .'
.' ,.' •• f
1* .'' / •'' "
'„'*•"*
,' / / /
£, ••" ••' .'
$/ II \
.' /'* / ' "
'" ' •""*'• f^J f *
.-•'' 0^ •'* • ""• *'
.M***" * \ .. ' (X •' l&
-• - — --•"".",-."•• • ""-••"'' ^ / >
'tfiljV-- . .. ,-•"*"" .•' /* /
' • ' OA ** '* ' / /* * f*.
.-.. **"*•'•--.„., / .•' ^ f aC
?' '*. ***** /' /' .'' / *^ "
„ j J •'/'/• t .
/ ' / .- -•**' 1
''" " .'* / , ' •"' •*" / *
: -•---•' ////" 1 •
'• ^ i& .•' / .*' / •
/ c> » ^0 ' ' •*' •
'. i "£ '""'* .-•''' "'S
.••'•.. O \
•, 9 }
• •*
' I f'' . '' "
"•"/•'.'
• / ' ^* •' .' •'*
.• ' .•' .-•*'
,••' .• V ••'"'
- , • ' / ' .'l^^
' ' '.''' "\
'•';''•''
:''.'' '.
p.-
-
f'
/
/
*•''
,.''&{•*.__ . ^- .— '-••" .•••'"'\-
06 08 QL 09 OS Oft 06 02 01
E
I
X
o
-— *
a
.0
•••
.a
o>
o
ID
•g
.O
«
.O
U>
.O
*
.0
w>
,0
0
^
LU
CD
Z U-
«r zr o
c2 o
oo
_J Q Z
=3! LU O
^ 00 i— >
i— i CQ i— <
> 0
-2T
Q LU O
Z t— O 00
~™N —•«• ^^
~^ ^4^ ^—
O C£. 1 t— i
rv «3C eC
CD z: cj _i
:*£ o i— i a.
CJ i— i CD
<: I— O (—
CQ <=C _ J ef
2: 0 LU
=1 C£ Di Di
O O CD
CD U_ LU
'ZL i — z;
i — i iii 1 1_ i rv
^- i 5- iij
^^ P^ ac^ I-*-!
r> «=c n:
oo LI_ a; i —
co _i o cc:
=t rs u_ o
oo 2:
• oo
^--.a; 2: LU
E rs o a:
• — -ac oo
oo z
LU CcT ' — ' > — i
CD LU s;
z: a. LU UD
as H- CK: 01
«Z LU i—
_l LU (—
cj o re i—
i— l CO h- 1
a i— o o
LU Q- O
C£ U_ >- <^-
Q. O 3C r-
•
VO
LU
CD
HH
1 i
-------
150
• -,
%-t _
— <
o
o.
O-
o>
0.
eo
O-
o.
0.
o.
o.
o.
'
'• '••"'''•.',•*? ' j "f
- .' '••'.' : •<£•' ''.' •""' .••' f
' t-' ^ • - "* ' C^ •-' * " *"
/ 'Cvatv" . ' ' ^^ ..-"'" Q
r ' -'*Y' :'. ''•"•', / /' •*
t- f\7 '. / -' " •' '* •'" t-
**• i -' ; •' .* " . •' L
• •• ':.£>• ..•'•>/,§> i-
•' ' • v' •''••'•' /"* t
''.''.•*•' i"
^, /••'•' /"V/ r
*/'"**' **V ,',' V
/••/.-•'.••"•'/ r
: :•;•••'•"•''/•$•.•'/.•' I
.- •'' " > 'C1 .- ' $*•• ' .' '* 'Qi i i
. ... "" '"—'", ->- '1 ••'..*' -' ••/', "V '•— ^ i
"* '•" " 1)}" '•- • "" ' ' .. • '"' • ,• ' , t •' \
- '" 1" * •t?..".,'. •-•- """ "' -' *'/'/ , ' a
' *'-- crfi ., ^' •*••' .'' J A*' .••'"'.• Cf
' t?^ .. * " ". ^J^ •' / ' ' *1^ ,-•'' •* *"
* — . / / ,•' / / ^' ••* *v «'f f
} / • •' '• '••'',•''..-• ' -vl? /' f
•' .•' - - _. ,' .•' ..-•' t-' f
" -•' • ' • w* ' .-'*' -•'* ' . ' t
•* .•"'.'" • .'*' .' •'*' ' . • "'1 ••*' i-
"-••. .—• ,• .V r> •' • ,-'' _.""' ,; ,.' i.
- .''/.''•'•$? ''"" V' •'•'"' l
" • •'/'•' i • *. \ • t\y '*•
' • ,''*)'"'" * ""'*" *' .••' S '..
'.t \ \& '.:."""' /,;<•••' -
t *'""'. } • ' •'*' .'•''
') £"• '' '" •'.'"'
- .••'"" '\ I ^ ••' . .•'.*$••"'' \
: .' : .-' ' . 130 -, ».
iiiiiiiiiiiiiiil'ii'iiiiiiiii -*-'**f'* i I i i i i i t t i i i i i i.* i i i i i
E
O
X
o
. "~«
*^
o
.0
.0
•»
.0
QO
.0
.O
vD
.O
I/)
.O
«•
.O
en
.0
.0
w*
O <-i
UJ UJ o
CO T*
QO
uj o
O CO «rf-
py ^f __
•^> DO
O U_
a: "o
e; uj
«r. §o
cc i— i
^ i
<; o 1-1
i— i r~i
=i<£o
— ^ CJ
I) O _l
co o z
<; uj i— i •— i
"«£ O -J
•^-Li ID-
E-JO
NX "*"^ fV ^.^
CO UJ cj;
1 — UJ
UJ C£. UJ C£
2: o
•=t r: o: z
a: o en
(V 1 i 1 .J
• uj co
-------
151
measured mean and maximum SOT concentrations in 1972 and 1973 in several
locations in Arizona are summarized in Table 5. Note that mean sulfate
3
concentrations ranged from 2.2 to 12.5 ^g/m , and maxima ranged from 5.8 to
41.0 yg/m3.
TABLE 5. MEAN AND MAXIMUM 24-HOUR-AVERAGE
SULFATE CONCENTRATIONS MEASURED
IN ARIZONA IN 1972-1973
[so]
_ Location Mean Maximum
Ajo 12.5 41.0
Douglas 9.1 25.9
Flagstaff 2.6 5.8
Grand Canyon 2.2 6.0
Phoenix 6.5 34.5
Superior 5.2 22.1
Tuba City 2.6 9.7
Tucson 6.4 22.0
The highest concentrations occurred in the immediate vicinity of
smelters (e.g., Ajo, Douglas, and Superior) and in cities (e.g., Phoenix
and Tucson), where, presumably, the rate of sulfate formation would increase
because of the elevated concentrations of reactive species in polluted
urban air. The high sulfate concentrations measured near the smelters may
be the result of primary SOT emissions or the rapid conversion to sulfate
of emitted S02 in the initial stages of plume dilution. The measured
maximum sulfate concentrations in nonurban areas of Arizona distant from
smelters (e.g., Flagstaff, Grand Canyon, and Tuba City) ranged from 6 to
10 pg/m , in qualitative agreement with the regional model calculations.
-------
152
Our approach to the calculation of atmospheric discoloration on the
regional scale has been to compute color parameters for particular loca-
tions (e.g., Class I areas) using the NOp and S07 concentrations obtained
using the regional model. In these calculations, we assumed homogeneous
pollutant concentrations within the mixed layer, and we computed optical
effects for several different sight paths. Exhibit 6 shows the output
from an example of these calculations for a location with background con-
3 3
centrations of 3.4 ug/m of sulfate, 30 yg/m of coarse mode particulate,
and no NCL. For different scattering angles 3 and sight path elevation
angles 6, the following specifications of color and color change (similar
to the parameters characterizing plume impact) are printed:
> Optical thickness (T).
> Light intensity (Y and L*).
> Chromaticity coordinates (x,y).
> Change in light intensity between the given background
atmosphere and a Rayleigh (no particles) reference
atmosphere for the given Q and 6 (AY,AL*)
> Contrast between the given background atmosphere and
the reference at three wavelengths [C(>K ^ = 0-40,
0.55, and 0.70 ym].
> Blue-red ratio between given background and reference.
> Changes in chromaticity coordinates (AX,Ay).
> Color difference parameters [AE(L*l!*V*) and AE(L*a*b*)].
With these parameters, the color of the background sky at a given loca-
tion in a region can be specified and Munsell color notation can be deter-
mined from the values of L*, x, and y. The change in light intensity and
color between the given location's sky and the reference atmosphere is spe-
cified by contrast values AY, At*, AX, Ay, and A£. These differences do
not have the same meaning as the corresponding parameters for plume impact
because the observer compares the light intensity and coloration of the
-------
153
i>:
m
a.
a
1
£ K
OH
sS
«
£ s;
;r*T l-c
SSSSSSS
> I- K * co w •? N ; «• co a co
«• * s * ® I- *
N * aIt-1- * *
«• i- co «• n *
9 0* 4' ff1 s ft f-
* « n tf»»« - * »• n <
ei — »ee t-(• i- «c « — «»«• eo ec
*> *i •»• CO
> Sco-
i
i
— rao — M— N c* * cc co t» » M * j *)« «r es
e - i- + 1-
— i- IT * fc
§« «*
wwwN S? * *>""£:§
— »•
n M
flM
« ^ «. «. S' « -5'
- «• n «cj - -
9 !M M N * M -
CO CO CO CO CO CO « (5 k' M M M N
<- «-*s-NCOt-
? co •• ep c* fr o »p ff- r- w i^ * *
> N N M — — — ® M M C-t M f 1 IS
!•«'««'»»«' *"""""'
.. £jS®*S«<
C&QC^C&t^^^ *^^0>{v
§»«*«* * I- « — N N « rt — * * w t> * *
nowt^i^'c oMft»n»o«ic CO******
i £
5 -
W 1- 1- » ES
Nf- *• r> * i
M 5' N «
ff SO C
1- 0- M 5' N « 4
O CD
05 1* «' N « cc « ^ T:
* N * — N *? l»
«• N » «• «• + «
- S 1- t- » ^6 -0
CM- 0- * M - l-
SR n c c> M M
Nme1-*^ —
* «• «• c- e- »• e-
1
5
»C9<>S*CfJ ^•^h-M*®C-
np*«5*«r5 *«h-*«*-*
3rt^»«o*-« (fc^ — c-*>*o^
>«K5^-*f * — *•*
1- * n N * * *
e- *.**«!- 0»
— » W O 8 N N
*p fr- (6 *O Cj C5 * i *
99>O*9'C&9tM CD*r
SMMNNMS S«
renncin £^e
* b- 9* 9 <» -•
0» N «• S i 0
D 04 JC C* W ^ C4
I ^8:
g
g
i i i i i i i i i i i i i i i i i i i
- * rt S M r-
i CO * ® « « «
i i t i t » i
t 1 1 1 t t
*««**rt« —mneoMMM
•* CO « C5 W
•» M — e-oCD*
— M *9®K9COeO£'
••51 CO^'C^NCOO'f
- 9- «> * * CO CO ft
row® N»»N
— * 4 <> » M N
CO N * J1 ?l S (5
t* o> * •• c- co co
CO N M CJ N C:l M
cjSeo*-* ^ ®
fr CO *
N * n t» »• M —
<> *• *• i* •* i^ 10
»««** + •*
fl 0) 0) P» N N M
»n9>ce4-s
§N N 9- CO N 1-
*>«•*"#•* *
M N M N N «
p «• ce * i- n
J P? »P ^ t^ *t CO
J - ^ W ^ « I*
-«!-
* n co
05*Ne-
»!>•«««»*
* P5 N CO * «
C* (*• *^ *^ ^ *^ ^
?SS!33c3«2
•• w (* ft* t* w w
Sn'I'^aNl-
*««-—-
6r5**«M»l-
*«« — ••••
& .^,
a -c
«««»
- n + «
t^c* ••CO**!*^ -^
LU
o;
««
rs
o
—I Q
LU
> LU
o o:
o
I—
«=C u_
o
oo
I—
o
U.
LU
O
CO
-------
154
plume and the background simultaneously; a relatively sharp line of demarca-
tion separates the plume and the background. For the homogeneous regional,
background atmosphere, the observer compares the given atmosphere with a
recollection of a clean atmosphere. An exception would be an observer on
a mountain looking down on a homogeneous mixed layer from a position where
he can compare the color of the mixed layer with the clean air above it.
These characteristics of the color differences for the homogeneous,
regional background atmosphere must be kept in mind when interpreting the
values of color difference parameters for determining the perceptibility
and the significance of atmospheric discoloration. Further work is neces-
sary to identify the threshold values of the color difference parameters
for the homogeneous atmosphere.
-------
155
V RECOMMENDATIONS FOR FUTURE WORK
As stated in Chapter I, we have followed a pragmatic approach to the
development of models to predict anthropogenic visibility impairment. Our
goal has been to develop predictive tools helpful in:
> Writing the report to Congress on visibility for setting
policy and promulgating regulations.
> Evaluating the impact of proposed sources and making siting
and pollution control decisions.
> Determining the requirements for retrofitting pollution
abatement equipment on existing sources.
This chapter recommends additional work that we believe is necessary
in the near term to refine and test the models, to assess the impact of
proposed visibility regulations, to improve color display techniques, to
develop a regional visibility model for the Southwest and other regions,
and to verify the predictions of the models by comparing them with field
measurements.
Figure 3 in Chapter I illustrates the potential uses for visibility
models in environmental policy and regulatory decisions, emission source
siting, and pollution control. Assessment of the extent of existing or
past visibility impairment can be accomplished through measurements using
such methods as (1) visual range and coloration observation by trained
personnel, (2) photographic documentation of visual range and atmospheric
discoloration, (3) telephotometry, (4) integrating nephelometry, and (5)
transmissometry. However, estimation of the extent of future impairment
(e.g., impairment caused by new sources or by the abatement of existing
sources) requires a scientifically based prediction capability that can
provide estimates of visual range and atmospheric discoloration.
-------
156
Visibility modeling will clearly play an important role in determining
a rational definition of "significant visibility impairment," in setting
environmental policy (regulation promulgation, new source and existing
source retrofit reviews, and long-term goals), and in determining pollution
abatement and siting requirements on a case-by-case basis. It is expected
that visibility impairment rather than ground-level air quality will be-
come the dominant air quality issue and will have a significant influence
on siting and pollution abatement decisions, particularly in the West.
The following sections outline the work that we believe is necessary
to support the EPA's efforts in visibility regulation promulgation. These
recommendations are presented in the order of their urgency. In our view,
further testing of models and analysis of the impacts of visibility regu-
lations should be done as soon as possible.
A. IMPACT ANALYSIS IN SUPPORT OF REGULATION DEVELOPMENT
The most urgent requirement for the application of visibility models
is the development of regulations. Modeling work will be necessary to
determine siting constraints on new sources and requirements for pollution
abatement, both for new and existing sources, that will be imposed by pro-
posed visibility regulations.
We have drawn two conclusions of major regulatory importance in our
initial applications of visibility models:
> The sulfate formed from S02 emitted from such sources as
smelters and power plants may cause significantly reduced
visual range at locations hundreds of kilometers away from
the sources. Indeed, the magnitude of the visibility im-
pairment may increase with increasing distance downwind
from the source, thereby making identification of cause
and effect more difficult.
-------
157
> NO emissions from large coal-fired ppwer plants may cause
X
perceptible yellow-brown plumes and atmospheric discolora-
tion more than 100 km downwind, particularly during stable
atmospheric conditions. Control of participate and S02
emissions will make the discoloration more prominent by
reducing the masking effect due to light scatter.
The implications of these conclusions for siting and control are
obvious. Impacts at large distances from emissions sources must be con-
sidered in siting studies. Although the impact of power plant emissions
on visual range will be reduced by controlling SO emissions, NO control
x x
is needed to reduce the yellow-brown discoloration that is caused by N0?.
In the analysis of the impact of visibility on industry, considera-
tion must be given to:
> The magnitude and the spatial and temporal extent of im-
pairment for various sources, ambient conditions, and
geographical locations.
> The siting constraints imposed on new sources.
> The pollutants that must be controlled.
> The degree of control required to reduce visibility im-
pairment to acceptable levels compared with the capabi-
lity for, feasibility of, and cost of implementation of
various pollutant control technologies.
> The appropriate regulatory policy to deal with visibi-
lity impairment (i.e., emission standards, ambient air
quality standards, or some standard of visual range and
atmospheric coloration).
B. MODEL REFINEMENT AND TESTING
Further work is recommended to test and refine the visibility models
in the near term, including:
-------
158
> Further testing of the models through sensitivity analyses.
> Incorporation of more sophisticated gas-to-particle and
aerosol growth algorithms in the code.
> Further assessment of the subjective visual impact of and
human threshold response to light intensity and color changes.
> Refinement of color display.
We limit our discussion here to work that should be done in the near
term to support the development of visibility regulations. In the future,
when a complete set of measurements are available (e.g., from EPA's VISTTA
program), visibility models should be verified. Measurements needed to
validate models include source emission rates, primary particulate size
distribution, meteorological conditions, plume dimensions, plume and ambi-
ent chemistry, aerosol size distribution and chemical composition, scat-
tering and absorption coefficients, solar direct and diffuse intensity, and
spectral light intensities and color photographs for several lines of sight,
Model validation is discussed in Section C.
1. Model Testing
As we noted in Chapter IV and demonstrated in Appendices E and G, we
have started to test the models by applying them to different emission and
ambient conditions to test their sensitivity to various input parameters,
including:
> Atmospheric stability (rate of dilution)
> Background ozone concentration
> Solar zenith angle
> Scattering angle
> Observer location and sight path orientation
> Background object light intensity and color
> Pollutant emission rate.
We recommend that more sensitivity analyses be performed with the
plume model for a variety of emission sources, meteorology, ambient
-------
159
conditions, and viewing conditions to evaluate the model qualitatively.
The results of this sensitivity analysis could be displayed in graphical
and tabular form so that they could be used by environmental engineers in
regulatory actions, impact analyses, siting studies, and design.
Further parametric analyses should be performed to evaluate the sen-
sitivity of model results to:
> Primary, secondary, and background aerosol size distribution.
> Ratio of diffuse to direct solar flux.
> Ratio of [NOg] to bscat-
> Locations of the background object and the plume relative
to the observer.
In a manner analogous to the ozone isopleth diagram, it may be
possible to characterize on an isopleth diagram the impact on visibility
of a range of combinations of the precursor pollutants. By plotting
contours of constant value for some visibility-related objective function,
these precursor mixtures which lead to the same visibility conditions may
be identified. Among the candidate objective functions are the contrast,
visual range, blue-red ratio, and AE.
Another sort of visibility isopleth diagram might be constructed to
characterize general regional visibility. Instead of HC and NO as
/\
"precursors," sulfate and nitrogen dioxide could be viewed as "indices" of
visibility degradation. By plotting concentrations of SCL and N0? along
the axes, one could determine isopleth lines that correspond to constant
objective function values.
One of the chief values of the ozone isopleth diagram is that it
provides an easily computed estimate of the reduction in precursor emis-
sions from current ambient levels required to reach the NAAQS. If the ob-
jective function chosen for use in the visibility isopleth diagrams were
the same as that employed in setting the federal standard, these diagrams
-------
160
might have a similar use. Required reductions in either SCL/NO or SCL/NO?
C. A *T t.
might be directly estimable. If the development and use of visibility plots
were shown to be both feasible and reliable, they might prove to be impor-
tant tools as promulgation and implementation of a federal visibility
standard occurs. This could be of particular significance for state and
local agencies, having limited resources and expertise, since they are re-
quired to incorporate visibility considerations in State Implementation
Plans.
2. 6as-to-Partic1e Conversion and Aerosol Growth
Currently in the visibility models gas-to-particle conversion (SOp to
SO^ and NO to NO^) is treated in a simple manner through the use of pseudo-
T* X O
first-order rate constants. Secondary aerosol is assumed to form in the
accumulation mode with properties observed by Whitby and Sverdrup (1978),
in the Labadie plume. Although this is a first approximation, it is a
reasonable assumption for modeling purposes.
We recommend that further work be done to identify the reaction
mechanisms effecting the conversion of SOp and NO to sulfates and nitrates
t~ A
and typical concentrations of reactive species in various nonurban areas
(Class I) in the United States. Reactions with the following species should
be considered:
OH'
»
>
and 0^ (in clouds)
I
RO-
HO*
NH3
Through evaluation of the concentrations of reactive species in non-
urban areas and resultant formation rates, appropriate formation rates can
be selected by the user or computed in the code.
-------
161
We recommend that an aerosol growth model be studied for possible
incorporation in the plume model. Such a model would compute the equilib-
rium particle size distribution as sulfate and nitrate form and water con-
denses onto the particle surfaces. We would determine if such a growth
model would improve the existing model sufficiently to justify its use.
3. Assessment of Color Impact Thresholds
We have incorporated into the visibility models the most recent
methods for quantifying color differences developed by the CIE in 1976
[AE(L*U*V*) and AE(L*a*b*)]. More work is necessary to determine what
standards for atmospheric coloration should be used (if any) in the vis-
ibility regulations. The AE's appear to be reasonable parameters to
characterize color changes associated with pollution; however, more work
is needed to determine what AE values mean subjectively in various cases
and what perceptibility-threshold and acceptability-threshold values should
be adopted in the analysis of atmospheric discoloration.
4. Refinement ofColor Display
In few instances is the display of model results so important as it
is in the prediction of visibility impact. A considerable number of sep-
arate lists of information are required in order to characterize a single
scene. However, the human eye and brain together are able to assemble and
integrate all this input, synthesizing it to a final impression of visual
impairment. It is the subjective judgments based on these impressions
that constitute "visual impact" of the most fundamental sort.
Consequently, the practical utility of a model depends on its ability
to collapse its predictions into a similarly simple and usable format.
It is for this reason that model predictions in this study have been ex-
pressed not only by means of specific visibility-related parameters, but
also through artist renderings of entire scenes with colors and intensi-
ties of sky and pollutant determined by model predictions.
-------
162
Several advantages have been achieved as a result. Predictions can
be assimilated more easily. Judgments about visual impairment are faci-
litated. The visibility impact projected to result from construction or
alteration of a facility can be more readily presented to and evaluated
by policy-makers and the general public.
While considerable progress has been made in this study to develop
suitable means for displaying model predictions, the following continued
efforts seem to hold the promise of substantial payoff:
> Additional studies could be conducted of the feasibility of
using artist-produced color illustrations to represent
visibility model predictions.
> A study could be performed of the feasibility of using
artist renderings of plume and atmospheric coloration pre-
dictions "overlayed" onto actual photographs of terrain.
This "photo-montage" technique has been used successfully
by the U.S. Forest Service's MOSAIC land use assessment
program.
> A study could be undertaken of the comparative accuracy
and acceptability of each of the above two display tech-
niques, as well as with the color video approach used by
workers at Los Alamos.
> The Perspective Terrain Viewing Program (PTVP) could be
linked to the plume visibility prediction model (PLUVUE).
By doing so, one could first display the terrain as seen
from a specified location, select a point whose colora-
tion was desired (as expressed in chromaticity coordinates,
perhaps), and calculate directly the visibility predicted
at that point. In this way, use of the visibility model
would be much more tightly integrated conceptually with
terrain views.
-------
163
C. MODEL VALIDATION
In this section we discuss our preliminary thoughts on a model vali-
dation effort, including:
> The type of measurement program that is needed.
> The specific measurements that should be made.
> The type of analysis of measurements and model pre-
dictions needed to assess model performance and to
provide direction for model refinement.
1. The Type of Measurement Program
No attempt should be made to validate SAI's regional visibility model
at this time. Rather, efforts should be aimed at providing a comprehen-
sive set of measurements downwind of a point source so that the plume vis-
ibility model can be validated. Information obtained from the point source
measurement program will be useful later in regional model validation and
refinement.
A large, coal-fired power plant should be selected for the measurement
program. The visibility regulations required by the Clean Air Act Amend-
ments of 1977 are likely to affect power plants, particularly in the wes-
tern United States, more than any other single class of emissions source.
Although copper smelters emit large quantities of SO , which has been shown
A
to significantly affect visual range in the Southwest, most of those sources
are exempted from the requirements of Section 169A on visibility protection
because they are more than 15 years old.
The power plant that is selected for measurement should have the fol-
lowing attributes:
> Pollutant emissions should be easily measurable and
should be relatively constant during the measurement
program.
-------
164
> Participate emissions should be well controlled using
state-of-the-art abatement equipment, such as efficient
electrostatic precipitators or wet scrubbers, so the
plant is representative of modern coal-fired power plants.
A major objective of the measurement program is not to
measure the visibility impairment caused by large emission
rates of primary particulates from older plants, but to
assess the visibility impairment caused by secondary aero-
sols (i.e., sulfates and nitrates) and NCL. Large emission
rates of primary particulate might interfere with the mea-
surement of secondary aerosol generation and light scattering.
> Sulfur dioxide (SO^) emissions should not be controlled by
scrubbers so that a significant amount of SC^ is available
for conversion to sulfate.
> The power plant should be located in the western United
States and should be isolated so that the plume can easily
be identified, tracked, and measured without interference
from plumes from other sources.
The emphasis of the measurement program should be on the visibility
impairment caused at far downwind distances. This contrasts with the
objectives of most air quality monitoring programs, which are designed to
determine the maximum ground-level pollutant concentrations, which gener-
ally occur within 20 to 30 km of the source. Visibility impairment appears
to be a long-range air pollution problem because it is caused by secondary
pollutants (NOp, sulfates, and nitrates) that are formed relatively slowly
in the atmosphere. Preliminary calculations show that the maximum reduc-
tions in visual range occur hundreds of kilometers from power plants and
maximum plume discoloration due to NO^ occurs during stable conditions 40
to 100 km downwind. Visibility impairment at distances 100 km or more
downwind of proposed or existing emissions sources will be the controlling
factor in determining the amount of pollution control equipment that must
be retrofitted on existing sources and in evaluating what the siting and
-------
165
emission constraints on new sources will be. Therefore, long-range impacts
must be measured so that the visibility model can be validated.
A crucial part of the long-range tracking and measurement of plume
visibility impairment will be the measurement of upper air transport winds
at frequent intervals and several locations so that accurate, real-time
plume trajectories can be calculated. These trajectories will be needed
to help track the plume and to identify the plume location relative to
fixed observation locations. Also, the upper air wind data can be used
later in conjunction with National Weather Service measurements to calcu-
late back trajectories and air parcel histories, so that potential sources
of background aerosol and trace gases can be identified. For example,
during a measurement program at a point source in a nonurban area of the
Southwest, there might be several days when air originating from an urban
area or from a copper smelter complex would get transported to the measure-
ment area. One could thus, with little additional expenditure, supplement
the information regarding point-source plume impact with information
regarding the regional impact of distant sources.
An attempt should be made to measure plume visibility impairment dur-
ing stable meteorological conditions. Greatest visibility impairment,
according to model calculations, occurs during stable conditions (e.g.,
Pasquill E or F). However, even greater impacts might occur during stag-
nant conditions or in locations where there are flow reversals (e.g.,
drainage flows) that could cause a build-up of pollutants in a confined
area. A study of climatological records could be carried out prior to
the measurement program so that periods of the year most likely to have
stable or stagnant conditions could be selected. For example, in the
Southwest stable conditions occur most frequently in the winter.
2. Specific Measurements
A large number of measurements will be required to validate the plume
visibility model. The necessity of each measurement can be appreciated
by examining Figure 13 (Chapter III), which shows the schematic logic flow
-------
166
diagram of the visibility models. Most air quality measurement programs
and models are concerned only with the first two elements of the visibility
model—the emissions and the atmospheric transport, diffusion, and removal
processes. The desired measurements and output are time-averaged pollutant
concentrations at given ground-level locations. However, in the visibility
model and in measurements to validate visibility models, the desired result
is a light intensity, perceived by an observer at a given ground-level
location, which is affected by air pollutants some distance away from the
observer.
Thus, the most important single measurement necessary for the valida-
tion of visibility models is of the spectral light intensity I(A) for spe-
cific observer locations and lines of sight. The spectral light intensity,
a strictly physical parameter, can be translated to visibility-related
psychophysical parameters, such as luminance (Y), chromaticity (x, y),
contrast (C), and the color difference parameter (A£), by weighting the
light intensity by the spectral response characteristics of the three
different light sensors of the human eye. These psychophysical parameters
are directly related to what an observer sees and are necessary and suffi-
cient for quantitying visual range and atmospheric discoloration. The
multiwavelength telephotometer is the only instrument with which we are
familiar that can directly measure these psychophysical parameters. By
equipping the instrument with color filters corresponding to the spectral
response of the three light receptors of the human eye, one can measure
tristimulus values (X, Y, and Z) for a given line of sight and calculate
Y, x, y, C, and A£. The telephotometer can also be used to determine
visual range by measuring the contrast between a distant mountain and
the horizon sky. Since the light intensity of several lines of sight can
be measured with a single telephotometer at one location, three or four
ground-based telephotometer measurement stations might be sufficient for
a measurement program at a single point source. Station locations might
be in a preferred transport direction (in stable conditions) at distances
from the source of 20, 50, 100, and 150 km. If possible, some of these
-------
167
stations might be at vista points in Class I areas. Telephotometer sta-
tions can be moved fairly easily, depending on the transport of the plume,
to the most strategic locations and could be located on high terrain where
views in many directions are possible. It would be valuable to have sta-
tions on opposite sides of a plume so that the same sight path could be
sampled from different angles to study the effect of the scattering angle e.
One telephotometer could be constantly moved to obtain views of the plume
from several angles and distances to characterize fully the effects of
plume-observer geometry. For example, measurements of the plume could be
made at several locations along a road or highway to evaluate the effect
of the plume-observer distance and the angles between the line of sight
and the plume centerline and between the line of sight and the horizon.
Each telephotometer operator should take color photographs of the
scenes that he is measuring with the telephotometer for later documentation
of contrast, atmospheric coloration, and the positions of the plume and the
sampling aircraft. At the edge of the camera's field of view in each photo-
graph, a color test strip should be placed (in direct sunlight, if possible)
so that the quality of the development of the color film can be controlled
and checked. In cases of forward scatter (e < 90°) where the sun is in
front of the camera, the color test strip cannot be placed both in the field
of view and in direct sunlight. In such cases, the test strip can be photo-
graphed separately at some interval. It may be possible to maintain and
check color film development quality by calibration using the color test
strip only once per role of film. The feasibility of cross-checking the
color photographs with the multiwavelength telephotometer measurements
should be evaluated. Color time-lapse movies from strategic vista points
could also be taken.
To link the measured spectral intensity to air pollution, one must
know the aerosol and NOp concentrations along the specific sight paths.
Thus, airborne measurements of NOp concentration, scattering coefficient
(b ,), and aerosol size distribution will be necessary. Attempts should
-------
168
be made to make aircraft traverses of the plume as close to the lines of
sight used in the telephotometer measurements as possible. It is essen-
tial to measure the location of the plume relative to telephotometer
stations since coloration will depend on the proximity of the plume to
the telephotometer. Airborne measurements could be supplemented and
checked with correlation spectrometer measurements of N0? burden. Since
the plume optical depth due to NCL is directly proportional to the N0~
burden, this measurement would be valuable. The correlation spectro-
iiieter should be considered as an optional supplement to airborne
monitoring.
Direct and diffuse solar flux should be measured using a pyrheliometer/
pyranometer combination. The occurrence of cloud cover should be docu-
mented. The location of the sun should either be measured or calculated
so that the solar zenith angle and the scattering angle for all telephoto-
meter and photographic lines of sight can be calculated later. Alterna-
tively, scattering angles could be measured at the time of measurement.
It is imperative that the line of sight of each light intensity measure-
ment be specified and recorded; measurements of the scattering angle,
solar zenith angle, sight path azimuth, sight path elevation angle,
observer location, plume location, and plume dimensions fully describe
each line of sight.
An important part of the measurement program should be the determi-
nation of the production site, chemical composition, size distribution,
and causes of secondary aerosol production, particularly at large down-
wind distances. The secondary aerosol production rate could be determined
by calculating [SO=]/[S02], [NO^/LNO^, [bscat]/[S02], and [bscat]/[NOx].
Each of these ratios will increase with secondary aerosol formation. The
measured aerosol size distribution, chemical composition, mass concentra-
tion, and scattering coefficient should be cross-checked using Mie theory
and accounting for the cations and liquid water associated with sulfates
and nitrates. Hypotheses regarding the mechanisms of secondary aerosol
formation should be tested by looking at the time-dependent rate of
-------
169
aerosol formation. If possible, measurements of background and plume
ammonia and radical concentrations should be made to identify possible
fundamental reactions effecting gas-to-particle conversion. Some attempt
should be made to measure the refractive indices, including the imaginary
(absorption) components, of the background, fly ash, and secondary aerosols.
The conversion of NO to NCL in the plume should be measured. Ozone
concentrations in the background air and in the plume and ultraviolet radi-
ation should be measured to test the validity of the steady-state assump-
tion used in calculating plume N02 production.
Plume dispersion parameters (a , o2) should be calculated from the
measurements of peak plume concentrations. The wind field at the plume
centerline should be measured by pibal releases from several locations at
hourly or three-hourly intervals. The plume position and transit times
based on measured wind speed and direction at plume height should be com-
puted on a real-time basis and should be compared with actual plume position.
Real-time calculations of plume position in the field would be used by the
pilot and ground-level observers to determine the location of the plume,
and to direct the airborne plume measurements at night and at far down-
wind distances. These calculations could also be used in conjunction with
weather forecasts to relocate ground-based stations to optimize plume
impact measurement. Vertical temperature gradients should be measured
during the aircraft flights.
Finally, the emissions from the power plant must be measured accur-
ately. If possible, the following measurements should be made throughout
the measurement program: mass emission rates or flue gas concentrations
of S02, NO, N02» and fly ash, in-stack opacity, flue gas volumetric flow
rate, flue gas temperature, and flue gas oxygen concentration. It would
be desirable in simplifying the measurement program if the power plant
operated at constant capacity throughout the measurement program.
-------
170
3. Data Analysis, Assessment of Model Performance, and Model Refinement
Data collected during the measurement program must be reduced and com-
piled in a format useful for providing input data for the visibility computer
simulation models. After the plume visibility model is run, the calculated
SC>2> NO, N02, 0^, SO^, and NO" concentrations, scattering coefficients,
spectral light intensities, and visual effects (reduction in visual range,
plume perceptibility, and atmospheric discoloration) should be compared with
measurements.
Model calculations should be made based on the measured values of:
> Emission rates
> Upper air wind speed and direction
> Plume dilution (a , a )
> Secondary aerosol formation rates
> Aerosol size distributions
> Ambient conditions
> Geometry of sun, plume, and observer.
Calculations and measurements of the following parameters could be
compared:
> Pollutant concentrations
> [N02]/[NOX]
> bscat
> bscat/mass ratios
> Visual range
> Luminance (Y)
> Chromaticity (x, y)
> Perceptibility (A£)
> Contrast.
-------
171
In addition, SAI's plume perspective and color display techniques could be
used to create color renderings of certain vistas for comparison with color
photographs.
In comparing model calculations and measurements necessary for model
validation it is important to consider the errors that can occur in:
> Measurement of input parameters
> Measurement of output parameters
> Model formulation.
Since we want to test and validate the model formulation, it is essen-
tial that errors in measurement of input and output data be minimized, error
bounds established, and that all parameters necessary for defining model
input and output be measured.
Model performance can be evaluated using:
> Correlation coefficients.
> Differences between measured and calculated values: Either
mean or root-mean-square, and either absolute or relative
differences.
> Ratio of measured to calculated values.
> Regression statistics.
> Qualitative comparisons.
A thorough model evaluation may identify directions for model refinement.
Limits on model applicability and accuracy may also be established. After
further model refinement and development based on the information gained from
the comparison with measurements, it could be advantageous to test the model
again using another set of measurements, possibly from another emissions
source, and allowing no intermediate fine tuning of the model input parameters,
-------
172
D. FURTHER DATA ANALYSIS
We have obtained an extensive data base with which further analyses
of anthropogenic visibility impairment can be made, including:
> Nearly 500 station-years of National Weather Service (NWS)
visibility and meteorological data.
> About 10 station-years of National Park Service visibility
data.
> Holzworth mixing depth and mixed layer wind speed for all NWS
upper air stations in the United States for 1960 through 1964.
We recommend that further data analysis be coupled with the develop-
ment of a regional visibility model for the Southwest. The objectives of
this analysis would be to determine through an analysis of upper air flow
trajectories typical transport wind fields which could be used for re-
gional calculations. Temporal and spatial variations in visual range
could be studied in conjunction with calculated trajectories to determine
the transport of emissions from source areas to clean nonurban areas (for
example, southern Utah). Also, inferences could be made as to the rate of
sulfate formation and removal by studying trajectories, ground-level sul-
fate measurments, and visual range observations.
In this report, visual range has been shown to be correlated with
many variables. Correlations with meteorological variables as well as
diurnal and seasonal variations have been explored. However, these rela-
tionships have been studied one variable at a time, and no attempts have
been made to elucidate the simultaneous effects of many variables. In
statistical terms, only univariate analyses have been made thus far,
though multivariate analyses are needed.
A question of great interest is what combinations of meteorological
conditions are associated with poor visibility? This question could be
investigated using a multivariate classification technique, such as linear
-------
173
discriminant analysis [e.g., Gnanadesikan (1977), Morrison (1976)], which is
a method that assigns multivariate observations to one of several classes.
This is done by finding planes in the multidimensional space of the inde-
pendent variables (in this case, the meteorological variables) that opti-
mally divide the observations (visual ranges) into different classes (e.g.,
visual range between 80 and 100 km). Discriminant functions for subsets
of the data can be determined, for example, for different years or for the
period of the copper strike. Differences between discriminant functions
would indicate that the conditions associated with various visual ranges
were changing. Misclassification rates should be evaluated for each dis-
criminant function to indicate its utility as a predictor.
Another technique that could be applied to the data set to elucidate
the dependence of visual range on other variables is multiple regression.
However, since many of the independent variables that would be used in the
regression (e.g., the meteorological variables) are likely to be highly
correlated, some dimensionality-reducing method, such as factor analysis,
should be applied to these variables first. In this way, some smaller
number of uncorrelated surrogates could be generated, and the multiple re-
gression could be made more meaningful and useful. Alternatively, a vari-
able selection method such as Cp analysis (Daniel and Wood, 1971) could be
used in conjunction with the regression.
Intervention analysis can be applied to determine whether a sudden
shift in conditions was associated with a corresponding change in visual
range (Box and Tiao, 1965, 1975). This technique would be applicable
to quantifying the changes in visual range that may have occurred during
the copper strikes. This time series method allows for the nonindependence
of successive observations in evaluating a change in level of a series of
observations. (A t-test of the difference in levels before and after the
event would be invalid because of the dependence between successive ob-
servations.) The analysis proceeds by calculating a function of the ob-
servations, with a known statistical distribution, which estimates the
shift in level. Straightforward statistical inference then gives the
-------
174
level of significance of the observed shift. Application of time series
techniques should enable both an analysis of trends in visual range and
sudden shifts due to events such as the copper strike.
E. DEVELOPMENT OF A SOUTHWEST REGIONAL VISIBILITY MODEL
For several reasons, the Southwest is likely to be the first region
in which visibility regulations, which are to be promulgated by August
1979, are implemented. First, there are a large number of mandatory
federal Class I areas in the Southwest, including national parks of such
obvious scenic value as the Grand Canyon, Bryce Canyon, Canyonlands, and
Arches. Second, the existing visual range in clean areas of the Southwest
is probably the best of anywhere in the contiguous United States. Accord-
ing to our analysis of visibility data (see Appendix A), there are several
locations in the Southwest, notably in northern Arizona, Utah, Colorado,
and New Mexico, where visual range is often greater than 160 km (100 miles).
Indeed, based on nephelometer measurements in Bryce Canyon reported by
Charlson (private communication, 1978), visual range at times may approach
the Rayleigh scattering limit of 390 km (240 miles) in the Southwest.
Third, significant energy development is planned for the Southwest, par-
ticularly in Utah and Colorado. Several large coal-fired power plants
are currently being proposed to be located at sites in the Southwest, in-
cluding Harry Allen (2000 Mwe), Intermountain Power Project (3000 Mwe),
Warner Valley (500 Mwe), and Garfield (2000 Mwe). Fourth, several large
coal-fired power plants are currently in operation in the Southwest, some
of whose plumes have been observed from scenic vistas in national parks
such as Bryce Canyon and Mesa Verde. These plants include Four Corners
(2175 Mwe), Mohave (1500 Mwe), Huntington Canyon (800 Mwe), Navajo (2300
Mwe), and San Juan (1500 Mwe). Finally, several very large emissions sources
are located in the Southwest or are in the prevailing upwind direction from
the Southwest. These sources include the copper smelters, whose current
aggregate SO emissions are more than 3000 tons per day, and the metropolitan
/\
areas of Phoenix, Tucson, Las Vegas, Salt Lake City, and Los Angeles, from
which pollution may be transported to Southwest mandatory Class I areas.
-------
175
Since spectacular scenery in the Southwest is enhanced by generally
excellent visibility and since considerable development of coal resources
is planned or has already occurred, a regional visibility model that can
study and answer the following questions is needed:
> Will a proposed power plant have an impact on visual range
in mandatory Class I areas located 100, 200, 500, or even
1000 km away? Will yellow-brown haze be visible? What sit-
ing alternatives exist, and how much pollution abatement is
required?
> Does an existing power plant's plume have a significant im-
pact on visual range and atmospheric color in national
parks? Where? How much? How often? What pollution
abatement equipment is required (i.e., particulate, SOX,
or NOX control)?
> Does the existing copper smelter complex have to control
SO emissions further to reduce visibility impairment?
A
Where does SO get transported, how much is removed by
A.
natural atmospheric processes, and how much is converted
to sulfates?
> Are sulfate, nitrate, and organic aerosol*, emitted from
urban areas, such as Phoenix, Salt Lake City, or even Los
Angeles, transported to mandatory Class I areas in the
Southwest? Do they have a significant impact on visibility?
These questions have potentially significant technical, socioeconomic, and
political implications. For example, if anthropogenic pollution is found
to be the cause of significant visibility impairment in the Grand Canyon,
the question to be answered is which combinations of sources contribute--
urban areas, the copper smelters, or a nearby coal-fired power plant?
A Southwest regional visibility model should be developed to point toward
the answers to these questions and to provide guidance for critical decisions.
-------
176
REFERENCES
Alkezweeny, A. J., and D. C. Powell (1977), "Estimation of Transformation
Rate of S02 to $04 from Atmospheric Concentration Data," Atmos.
Environ.. Vol. 11, pp. 179-182.
American Society for Testing and Materials (1974), "Standard Method of
Specifying Color by the Munsell System," Dl535-68, Philadelphia,
Pennsylvania.
Arizona Department of Health Services (1977), "First Annual Report on
Arizona Copper Smelter Pollution Control Technology, Phoenix,
Arizona.
Atkinson, R., R. A. Perry, and J. N. Pitts, Jr. (1976), "Rate Constants
for the Reactions of the OH Radical with N02 (M = Ar and Ne) and
S02 (M = Ar)," J. Chem. Phys., Vol. 65, pp. 306-310.
Barrie, L., and H. W. Georgii (1976), "An Experimental Investigation of
the Absorption of Sulfur Dioxide by Water Drops Containing Heavy
Metal Ions," Atmos. Environ.. Vol. 10, pp. 743-749.
Bassett, H., and W. G. Parker (1951), "The Oxidation of Sulfurous Acid,"
J. Chem. Soc.. Pt. 2, pp. 1540-1560.
Bauer, E. (1973), "Dispersion of Tracers in the Atmosphere: Survey of
Meteorological Data," Paper P-925, Institute of Defense Analysis,
Arlington, Virginia.
Baulch, D. L., D. D. Drysdale, and D. G. Home (1973), Evaluated Kinetic
Data for High Temperature Reactions. Volume 2--Homogeneous Gas
Phase Reactions of the Hg-^-O;? System (CRC Press. Cleveland, Ohio).
Beilke, S., and G. Gravenhorst (1977), "Heterogeneous S02 Oxidation in
the Droplet Phase," International Symposium on Sulfur in the
Atmosphere, Dubrovnik, Yugoslavia.
Beilke, S., D. Lamb, and J. Mueller (1975), "On the Uncatalyzed Oxida-
tion of Atmospheric S02 by Oxygen in Aqueous Systems," Atmos.
Environ.. Vol. 9, pp. 1083-1090.
Benarie, M., A. Nonat, and T. Menard (1973), "Etude de la Transformation
de L'Anhydride Sulfureaux en Acide Sulfurique en Relation avec les
Donnees Climatologiques, dans un Ensembly Urbain in a Caractere
Industriel, Rouen," Atmos. Environ.. Vol. 7, pp. 403-421.
-------
177
Benson, S. W. (1968), Thermgcheniical Kinetics (John Wiley & Sons, New
York, New York).
Bergstrom, R. W. (1973), "Extinction and Absorption Coefficients of the
Atmospheric Aerosol as a Function of Particle Size," Beitra'ge Zur
Physik der Atmosphare, Vol. 46, pp. 223-234.
Bergstrom, R. W., and J. T. Peterson (1977), "Comparison of Predicted
and Observed Solar Radiation in an Urban Area," J. Appl. Meteor.,
Vol. 16, pp. 1107-1116.
Bergstrom, R. W., and R. Viskanta (1974), "Spherical Harmonics Approxima-
tion for Radiative Transfer in Polluted Atmospheres," Progress in
Astronautics and Aeronautics, Vol. 35, pp. 23-40 (MIT Press,
Cambridge, Massachusetts).
Boris, J. P., and D. L. Book (1973), "Flux-Corrected Transport--!: SHASTA
A fluid Transport Algorithm That Works," J Computational Phys.,
Vol. 11, pp. 38-69.
Box, G.E.P., and G. C. Tiao (1975), "Intervention Analysis with Applica-
tions to Economic and Environmental Problems, J. Amer. Stat. Assoc.,
Vol. 70, p. 70.
(1965), "A Change in Level of a Non-stationary Time Series,"
Biometrika, Vol. 52, Nos. 1 and 2, p. 181.
Briggs, G. A. (1969), "Plume Rise," Office of Information Services, U.S.
Atomic Energy Commission, Washington, D.C. (National Technical
Information Service TID-25075).
Brimblecomb, P., and D. J. Spedding (1974), "The Catalytic Oxidation of
Micromolar Aqueous SOg--!. Oxidation in Dilute Solutions Containing
Iron (III)," Atmos. Environ., Vol. 8, pp. 937-945.
Businger, J. A., et al. (1971), "Flux-Profile Relationships in the
Atmospheric Surface Layer," J. Atmos. Sci., Vol. 28, pp. 181-189.
Calvert, J. G., et al. (1978), "Mechanism of the Homogeneous Oxidation
of Sulfur Dioxide in the Troposphere," Atmos. Environ., Vol. 12
(in press).
(1977), "Mechanism of the Homogeneous Oxidation of Sulfur
Dioxide in the Troposphere," International Symposium on Sulfur
in the Atmosphere, 7-14 September 1977, Dubrovnik, Yugoslavia.
Cass, G. R. (1976), "The Relationship Between Sulfate Air Quality and
Visibility at Los Angelas," EQL Memorandum No. 18, California
Institute of Technology, Pasadena, California.
Chamberlain, A. C. (1966), "Transport of Gases to and from Grass and
Grass-like Surfaces," Proc. Roy. Soc.. Vol. A290, pp. 236-260.
-------
178
Charlson, R. J., and N. C. Ahlquist (1969), "Brown Haze: N02 or
Aerosol?," Atmos. Environ., Vol. 3, pp. 653-656.
Charlson, R. J., N. C. Ahlquist, and H. Horvath (1968), Atmos. Environ.,
Vol. 2, p. 455.
Chen, T., and C. H. Barren (1972), "Some Aspects of the Homogenous
Kinetics of Sulfite Oxidation," Ind. Eng. Chem. Fund., Vol. 11,
pp. 446-470.
Cheng, R. T., M. Corn, and J. I. Frohliger (1971), "Contribution to the
Reaction Kinetics of Water Soluble Aerosols and SO? in Air at ppm
Concentrations," Atmos. Environ., Vol. 5, pp. 987-1008.
Committee on Colorimetry, Optical Society of America (1963), "The Science
of Color," Washington, D.C.
Conner, W. D., and J. R. Hodkinson (1972), "Optical Properties and Visual
Effects of Smoke-stack Plumes," AP-30, Office of Air Programs,
Environmental Protection Agency, Research Triangle Park, North
Carolina.
Covert, D. S., R. J. Charlson, and N. C. Ahlquist (1972), "A Study of
the Relationship of Chemical Composition and Humidity to Light
Scattering by Aerosols," J. Appl. Meteor., Vol. 11, pp. 968-976.
Cox, R. A., and S. A. Penkett (1972), "Aersol Formation from Sulphur
Dioxide in the Presence of Ozone and Olefinic Hydrocarbons," J. Chem.
Soc., Vol. 68, p. 1735.
(1971a), "Photo-Oxidation of Atmospheric S02>" Nature. Vol. 229,
pp. 486-487.
(1971b), "Oxidation of Atmospheric S02 by Products of the Ozone-
Olefin Reaction," Nature, Vol. 230, pp. 321-322.
(1970), "The Photo-Oxidation of Sulphur Dioxide in Sunlight,"
Atmos. Environ., Vol. 4, p. 425.
Crutzen, P. J., and J. Fishman (1977), "Average Concentration of OH in
the Northern Hemisphere Troposphere, and the Budgets of CH., CO,
and HC," Geophys. Res. Letters, in press.
Daniel, C., and F. S. Wood (1971), Fitting Equations to Data (John Wiley
& Sons, New York, New York).
Dave, J. V. (1970), "Subroutines for Computing the Parameters of the
Electromagnetic Radiation Scattered by a Sphere," IBM System 360
Program 3600-17.4.002
-------
179
Davis, D. D., W. Heaps, and T. McGee (1976), "Direct Measurements of
National Tropospheric Levels of OH Via an Aircraft Borne Tunable
Dye Laser," Geophys. Res. Letters, Vol. 3, pp. 331-333.
Davis, D. D., G. Smith, and J. Klauber (1974), "Trace Gas Analysis of
Power Plant Plumes Via Aircraft Measurements: (h, NOX and SO?
Chemistry," Science. Vol. 186, pp. 733-736.
Durran, D. D., et al. (1978), "A Study of Long Range Air Pollution Problems
Related to Coal Development in the Northern Great Plains," Systems
Applications, Incorporated, San Rafael, California.
Egan, B. A., and J. R. Mahoney (1972), "Numerical Modeling of Advection
and Diffusion of Urban Area Source Pollutants," J. Appl. Meteor.,
Vol. 11, pp. 312-322.
Ensor, D. S., L. E. Sparks, and M. J. Pilat (1973), "Light Transmittance
Across Smoke Plumes Downwind from Point Sources of Aerosol Emissions,"
Atmos. Environ., Vol. 7, pp. 1267-1277.
EPA (1976a), "1973 National Emissions Report," EPA-450/2-76-007, National
Air Data Branch, Office of Air Quality Planning and Standards,
Environmental Protection Agency, Research Triangle Park, North
Carolina.
(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.
ER&T (1977), "Development of an Atmospheric Model for Sulfate Formation,"
Document No. P-1534, Environmental Research & technology, Incorporated,
Santa Barbara, California.
Ettinger, H. J., and G. W. Royer (1972), "Visibility and Mass Concentra-
tion in a Nonurban Environment," J. Air Poll. Contr. Assoc.. Vol. 22,
PP. 108-m.
Federal Power Commission (1976), "FPC Form 67: Steam Electric Plant Air
and Water Quality Control Data for the Year Ending December 31, 1975,"
Washington, D.C.
Forrest, J., and L. Newman (1977a), "Further Studies on the Oxidation of
Sulfur Dioxide in Coal-Fired Power Plant Plumes," Atmos. Environ..
Vol. 11, pp. 465-474.
(1977b), "Oxidation of Sulfur Dioxide in the Sudbury Smelter
Plume," Atmos. Environ., Vol. 11, pp. 517-520.
-------
180
Forsythe, G. E., and W. R. Wasow (1960), Finite-Difference Methods for
Partial Differential Equations (John Wiley & Sons, New York,
New York).
Foster, P. M. (1969), "The Oxidation of S02 in Power Station Plumes,"
Atmos. Environ., Vol. 3, pp. 157-175.
Freiberg, J. (1975), "The Mechanism of Iron Catalyzed Oxidation of S02
in Oxygenated Solutions," Atmos. Environ., Vol. 9, pp. 661-672.
(1974), "Effects of Relative Humidity and Temperature on Iron-
Catalyzed Oxidation of S02 in Atmospheric Aerosols," Environ. Sci.
Techno! . , Vol. 8, pp. 731-734.
Friedlander, S. K. (1977), Smoke, Dust and Haze: Fundamentals of Aerosol
Behavior (John Wiley & Sons, New York, New
Fuller, E. C., and R. H. Crist (1941), "The Rate of Oxidation of Sulfite
Ions by Oxygen," J. Am. Chem. Soc., Vol. 63, pp. 1644-1650.
Galbally, I. F., (1974), "Gas Transfer Near the Earth's Surface," Advances
in Geophysics. Vol. 18B, pp. 329-340.
Gartrell, F. E., F. W. Thomas, and S. B. Carpenter (1963), "Atmospheric
Oxidation of SO? in Coal -Burning Power Plant Plumes," Am. Ind. Hyg.
J_., Vol. 24, pp. 113-120.
Gerhard, E. R., and H. F. Johnstone (1955), "Photochemical Oxidation of
Sulfur Dioxide in Air," Ind. Eng. Chem., Vol. 47, p. 972.
Gnanadesikan, R. (1977), Methods for Statistical Data Analysis of Multi-
yariate Observations (John Wiley & Sons, New York, New York).
Grosjean, P., and S. K. Friedlander (1975), "Gas-Particle Distribution
Factors for Organic and Other Pollutants in the Los Angeles
Atmosphere," J. Air Poll. Contr. Assoc., Vol. 10, pp. 1038-1044.
Hagen, L. J., and N. P. Woodruff (1973), "Air Pollution from Duststorms
in the Great Plains," Atmos. Environ.. Vol. 7, pp. 323-332.
Halow, J. S., and S. J. Zeek (1973), "Predicting Ringelmann Number and
Optical Characteristics of Plumes," J. Air Poll. Contr. Assoc..
Vol. 23, pp. 676-684.
Hanson, J. E., and L. D. Travis (1974), "Light Scattering in Planetary
Atmospherics," Space Science Reviews, Vol. 16, pp. 527-610.
Heisenberg, W. (1948), "Zur Statistischen Theorie der Turbulenz," Z_.
Phys . , Vol. 124, p. 614.
-------
181
Hidy, G. M., et al. (1975), "Characterization of Aerosols in California
(ACHEX)," Science Center, Rockwell International, Thousands Oaks,
California.
Hodkinson, J. R. (1966), "Calculations of Colour and Visibility in
Urban Atmospheres Polluted by Gaseous N02," Air Water Poll. Int. J.,
Vol. 10, pp. 137-144.
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.
Horvath, H. (1972), "Author's Reply: On the Brown Colour of Atmospheric
Haze," Atmos. Environ. , Vol. 6, PP- 143-148.
(1971), "On the Brown Colour of Atmospheric Haze," Atmos.
Environ., Vol. 5, pp. 333-344.
Husar, R. B., and W. H. White (1976), "On the Color of the Los Angeles
Smog," Atmos. Environ., Vol. 10, pp. 199-204.
Irvine, W. M. (1975), "Multiple Scattering in Planetary Atmospheres,"
Icarus, Vol. 25, pp. 175-204.
Jarman, R. T., and C. M. DeTurville (1969), "The Visibility and Length
of Chimney Plumes," Atmos. Environ., Vol. 3, pp. 257-280.
Jerskey, T. N., and C. S. Burton (1977), "The Prediction of Visibility
on a Regional Scale: The Status of Visibility Models and Measure-
ments," El77-23, Systems Applications, Incorporated, San Rafael,
California.
Johnstone, H. F., and D. R. Coughanowr (1958), "Absorption of Sulfur
Dioxide from Air: Oxidation in Drops Containing Dissolved Catalysts,"
Ind. Enq. Chem., Vol. 50, p. 1169.
Johnstone, H. F., and A. J. Moll (1960), "Formation of Sulfuric Acid in
Fogs," Ind. Eng. Chem., Vol. 52, p. 861.
Judd, D. B., and G. Wyszecki (1975), Color in Business Science and
Industry (John Wiley & Sons, New York, New York).
Junge, C. E. (1963), Air Chemistry and Radioactivity (Academic Press,
New York, New York]"!
Junge, C. E., and T. G. Ryan (1958), "Study of the S02 Oxidation in
Solution and Its Role in Atmospheric Chemistry," Quart.- J. Roy.
Meteor. Soc.. Vol. 84, pp. 46-55.
Katz
, M. (1950), "Photoelectric Determination of Atmospheric S02 Employ-
ing Dilute Starch-Iodine Solutions," Anal. Chem.. Vol. 22, p. 1040.
-------
182
LACAPCD (1971), "Profile of Air Pollution Control," Air Pollution Control
District, County of Los Angeles, Los Angeles, California.
Larson, T. V., N. R. Horike, and H. Harrison (1977), "Oxidation of Sulfur
Dioxide by Oxygen and Ozone in Aqueous Solution: A Kinetic Study
with Significance to Atmospheric Rate Processes," Atmos. Environ.,
in press.
Latimer, D. A., and G. S. Samuel sen (1978), "Visual Impact of Plumes
from Power Plants," Atmos. Environ., Vol. 12, pp. 1455-1465.
(1975), "Visual Impact of Plumes from Power Plants," UCI-ARTR-
75-3, UCI Air Quality Laboratory, School of Engineering, University of
California, Irvine, California.
Leighton, P. A. (1961), Photochemistry of Air Pollution (Academic Press,
New York, New York).
Levy, A. H., D. R. Drewes, and J. M. Hales (1976), "S02 Oxidation in
Plumes: A Review and Assessment of Relevant Mechanistic and Rate
Studies," EPA-450/3-76-022, Environmental Protection Agency,
Research Triangle Park, North Carolina.
Liu, M. K., and D. Durran (1977), "The Development of a Regional Air
Pollution Model and Its Application to the Northern Great Plains.,"
EPA-908/1-77-001, Systems Applications, Incorporated, San Rafael,
California.
(1976), "On the Modeling of Transport and Diffusion of Air
Pollutants over Long Distances," ER76-55, Systems Applications,
Incorporated, San Rafael, California.
Liu, M. K., M. A. Wojcik, and D. Henderson (1978), "Numerical Experiments
of the Effect of Lateral Diffusion on Mesoscale Pollutant Dispersion,"
Summer Computing Conference, 24-26 July 1978, Newport Beach,
California.
Matteson, M. J. (1978), "Capture of Atmospheric Gases by Water Vapor Con-
densation on Carbonaceous Particles," Conference on Carbonaceous
Particles in the Atmosphere, 20-22 March 1978, Lawrence Berkeley
Laboratory, Berkeley, California.
Matteson, J. J., W. Stober, and H. Luther (1969), "Kinetics of the
Oxidation of SOo by Aerosols of Manganese Sulfate," Ind. Eng. Chem.
Fund., Vol. 8, pp. 677-687.
McKay, H.A.C. (1971), "The Atmospheric Oxidation of Sulphur Dioxide in
Water Droplets in Presence of Ammonia," Atmos. Envi ron.. Vol. 5,
pp. 7-14.
Megaw, W. J. (1977), "Thin Layer Brown Haze," J. Aerosol Sci., Vol. 8,
pp. 21-29.
-------
183
Middleton, W.E.K. (1952), Vision Through the Atmosphere (University of
Toronto Press, Toronto, Canada).
Miller, J. M., and R. G. DePena, (1972), "Contribution of Scavenged
S02 to the Sulfate Content of Rain Water," J. Geophys. Res.,
Vol. 77, pp. 5905-5916.
Morrison, D. F. (1976), Multivariate Statistical Methods (McGraw-Hill,
New York, New York).
Munsell Color Company (1976), "Munsell Book of Color," Baltimore,
Maryland.
Newman, L., J. Forrest, and B. Manowitz (1975), "The Application of an
Isotopic Ratio Technique to a Study of the Atmospheric Oxidation
of Sulfur Dioxide in the Plume from an Oil-Fired Power Plant,"
Atmos. Environ., Vol. 9, pp. 959-968.
NGPRP (1974), "Atmospheric Aspects Work Group Report," Northern Great
Plains Resource Program, Denver, Colorado.
Niki, H. (1974), "Reaction Kinetics Involving 0 and N Compounds," Can.
J. Chem., Vol. 52, pp. 1397-1404.
Nixon, J. K. (1940), "Absorption Coefficient of Nitrogen Dioxide in the
Visible Spectrum," J. Chem. Phys.. Vol. 8, pp. 157-160.
Noll, K. E., P. K. Mueller, and M. Imada (1968), "Visibility and
Aerosol Concentration in Urban Atmospheres," Atmos. Environ.,
Vol. 2, pp. 465-475.
Nordlund, G. G. (1975), "A Quasi-Lagrangian Cell Method for Calculating
Long-Distance Transport of Airborne Pollutants," J. Appl. Meteor.,
Vol. 14, pp. 1095-1104.
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," Advances in Geophysics. Vol. 18B, pp. 137-150.
Novakoy, T., S. G. Chang, and A. B. Marker (1974), "Sulfates in Pollu-
tion Particulates: Catalytic Oxidation of S02 on Carbon Particles,"
Science, Vol. 186, pp. 259-261.
Orel, A. E., and J. H. Seinfeld (1977), "Nitrate Formation in Atmospheric
Aerosols," Environ. Sci. Techno!., Vol. 11, pp. 1000-1007.
Payne, W. A., L. 0. Stief, and D. D. Davis (1973), "A Kinetics Study of
the Reaction of H02 with $03 and NO," J. Am. Chem. Soc.. Vol. 95,
pp. 7614-7619.
-------
184
Penkett, S. A. (1972), "Oxidation of SC>2 and Other Atmospheric Gases by
Ozone in Aqueous Solution," Nature, Phys. Sci., Vol. 240, pp. 105-106.
Penkett, S. A., and J. A. Garland (1974), "Oxidation of SO? in Artificial
Fogs by Ozone," Tell us, Vol. 26, pp. 284-290.
Pilat, M. J., and D. S. Ensor (1971), "Comparison Between the Light
Extinction Aerosol Mass Concentration Relationship of Amtospheric
and Air Pollutant Emission Aerosol," Atmos. Environ., Vol. 5,
pp. 209-215.
Roberts, E. M,, et al. (1975), "Visibility Measurements in the Painted
Desert Through Photographic Photometry," Engineering Bulletin 47,
Dames & Moore, Santa Barbara, California.
Roberts, P. T., and S. K. Friedlander (1975), "Conversion of $03 to
Sulfur Particulate in the Los Angeles Atmosphere," Environ. Health
Perspectives, Vol. 10, pp. 103-108.
Rodhe, H. (1972), "A Study of the Sulphur Budget for the Atmosphere over
Northern Europe," Tell us. Vol. 24, pp. 128-138.
Samuels, H. J., S. Twiss, and E. W. Wong (1973), "Visibility, Light
Scattering and Mass Concentration of Particulate Matter,"
California Air Resources Board, Sacramento, California.
Sander, S. P., and J. H. Seinfeld (1976), "Chemical Kinetics of Homogeneous
Atmospheric Oxidation of Sulfur Dioxide," Environ. Sci. Techno!.,
Vol. 10, pp. 1114-1123.
Schroeter, L. C. (1963), "Kinetics of Air Oxidation of Sulfurous Acid
Salts," J. Pharm. Sci.. Vol. 52, pp. 559-563.
Schwartz, S. E., and L. Newman (1978), "Processes Limiting Oxidation of
Sulfur Dioxide in Stack Plumes," Environ. Sci. Techno!., Vol. 12,
No. 1, pp. 67-73.
Scott, W. D., and P. V. Hobbs (1967), "The Formation of Sulfate in Water
Droplets," J. Atmos. Sci.. Vol. 24, pp. 54-57.
Scriven, R. A., and B.E.A. Fisher (1975a), "The Long-Range Transport of
Airborne Material and Its Removal by Deposition and Washout--
I. General Considerations," Atmos. Environ., Vol. 9, pp. 49-58.
(1975b), "The Long-Range Transport of Airborne Material and
Its Removal by Deposition and Washout--!!. The Effect of Turbulent
Diffusion," Atmos. Environ., Vol. 9, pp. 59-68.
Sellers, W. D. (1965), Physical Climatology (University of Chicago Press,
Chicago, Illinois).
-------
185
Smagorinsky, J. (1963), "General Circulation Experiments with the Primitive
Equations: I. The Basic Experiment," Mon. Wea. Rev.. Vol. 91,
pp. 99-164.
Spicer, C. W., and P. M. Schumacher (1977), "Interferences in Sampling
Atmospheric Particulate Nitrate," Atmos. Environ., Vol. 11, pp. 873-876.
Thorn, A. S. (1972), "Momentum, Mass and Heat Exchange of Vegetation,"
Quart. J. Roy. Met. Soc., Vol. 98, pp. 124-134.
Trijonis, J., and K. Yuan (1978), "Visibility in the Northeast: Visibility
Trends and Visibility/Pollutant Relationships," Technology Service
Corporation, Santa Monica, California.
(1977), "Visibility in the Southwest: An Exploration of the
Historical Data Base," Technology Service Corporation, Santa Monica,
California.
Urone, P., and W. H. Schroeder (1969), "S02 in the Atmosphere: A Wealth
of Monitoring Data but Few Reaction Rate Studies," Environ. Sci.
Techno!.. Vol. 3, pp. 436-445.
Urone, P., et al. (1968), "Static Studies of Sulfur Dioxide Reactions in
Air," Environ. Sci. Technol., Vol. 2, pp. 611-618.
van der Heuvel, A. P., and B. J. Mason (1963), "The Formation of Ammonium
Sulfate in Water Droplets Exposed to Gaseous S02 and NH3," Quart. J.
Roy. Meteor. Soc.. Vol. 89, pp. 271-275.
Van de Hulst, H. C. (1957), Light Scattering by Small Particles (John
Wiley & Sons, New York, New York).
Waggoner, A. P., and R. J. Charlson (1971), "Simulating the Color of
Polluted Air," Applied Optics. Vol. 10, pp. 957-958.
Waggoner, A. P., R. J. Charlson, and N. C. Ahlquist (1972), "Discussion:
On the Brown Color of Atmospheric Haze," Atmos. Environ., Vol. 6,
pp. 143-148.
Waggoner, A. P., et al. (1976), "Sulfate-Light Scattering Ratio as an
Index of the Role of Sulfur in Tropospheric Optics," Nature.
Vol. 261, pp. 120-122.
Weir, A., et al. (1976), "Factors Influencing Plume Opacity," Environ.
Sci. Techno!., Vol. 10, pp. 539-544.
(1975), "Measurement of Particle Size and Other Factors Influ-
encing Plume Opacity," International Conference on Environmental
Sensing and Assessment, 14-19 September 1975, Las Vegas, Nevada.
-------
186
Whitby, K. T., and G. M. Sverdrup (1978), "California Aerosols: Their
Physical and Chemical Characteristics," to be published in the
ACHEX Hutchinson Memorial Volume, Particle Technology Laboratory
Publication Number 347, University of Minnesota, Minneapolis,
Minnesota.
White, VI. H. (1977), "NOX-03 Photochemistry in Power Plant Plumes:
Comparison of Theory with Observation," Environ. Sci. Techno!.,
Vol. 11, No. 10, pp. 995-1000.
White, W. H., and P. T. Roberts (1975), "The Nature and Origins of
Visibility Reducing Aerosols in Los Angeles," 68th Annual Meeting,
Air Pollution Control Association, 15-20 June 1975, Boston,
Massachusetts.
Whitten, G. Z., and H. Hogo (1977), "Mathematical Modeling of Simulated
Photochemical Smog," EPA-600/3-77-001, Systems Applications,
Incorporated, San Rafael, California.
Williams, M. D., and R. Cudney (1976), "Predictions and Measurements of
Power Plant Plume Visibility Reductions and Terrain Interactions,"
Proc. of the Third Symposium on Atmospheric Turbulence Diffusion
and Air Quality, American Meteorological Society, pp. 415-420.
Williams, M. D., M. J. Wecksung, and E. M. Leonard (1978), "Computer
Simulation of the Visual Effects of Smoke Plumes," Society of Photo-
Optical Instrumentation Engineers Technical Symposium, 30 March 1978,
Washington, D.C.
Winkelmann, D. (1955), "Die Electrochemische Messung du Oxidations-
geschwindigkeit von Na?S03 Durch Gelosten Sauerstoff," Z. Elektrochemie,
Vol. 59, pp. 891-895.
Winkler, P. (1973), "The Growth of Atmospheric Aerosol Particles as a
Function of the Relative Humidity--!I. An Improved Concept of
Mixed Nuclei," Aerosol Sci., Vol. 4, pp. 373-387.
Yanenko, N. N. (1971), The Method of Fractional Steps (Springer-Verlag,
Berlin, Germany).
-------
187
GLOSSARY
accumulation mode--Submicron particles in the range 0.1 to 1.0 ym, which
are formed from smaller nuclei by coagulation or by direct conden-
sation of gases. This particle mode is most effective per unit mass
in light scattering. Secondary particles such as sulfates, nitrates,
and organics are found predominantly in this size range. Since coag-
ulation is relatively slow for particles larger than 1 ym, particles
in the accumulation mode are generally removed from the atmosphere by
precipitation and surface deposition before they grow larger.
aerosol—A suspension of fine solid or liquid particles in air or a gas.
In the context of visibility impairment, aerosol means a suspension of
nuclei, accumulation mode, and coarse particles (and gaseous pollutants)
in ambient air. Aerosol is also sometimes used to signify the liquid
or solid particles themselves.
anthropogenic—Caused directly by man or indirectly by man's technology
(e.g., anthropogenic pollutant emissions from combustion sources such
as automobiles and boilers).
a'ppearance--The subjective visual impression or aspect of a thing observed
by a human (e.g., the appearance of distant mountains or a plume).
atmospheric discoloration—An imprecise term describing the change in color
of the sky or distant mountains or clouds observed through the atmos-
phere due to natural or man-made pollution. The term implies that
there is an atmospheric color or set of colors that can be defined
as natural and not discolored. Examples of atmospheric discoloration
are white, grey, yellow, brown, or black haze or plumes.
background—An object being viewed by an observer (e.g., sky, cloud, or
mountain). Light reflected (if any) from the object and light scat-
tered and absorbed by the atmosphere along the sight path or line of
sight between the object and the observer determines the color per-
ceived by the observer. The nature of the background affects the
apparent coloration of a plume or haze layer.
back scatter—Situation where the sun is behind the observer (e>90°).
chroma—The numerical index in the Munsell color notation system that
describes the degree of saturation or the departure from grey. A
chroma of 0 is grey; a chroma of 2 is slightly colored; a chroma of
6 is highly colored.
-------
188
chromaticity diagram and coordinates—A mathematical description of the
relative spectral distribution (hue and saturation) of a given color.
Chromaticity coordinates uniquely describe the position of a given
color on a two-dimensional plot called a chromaticity plot. Chroma-
ticity coordinates provide information only on the relative mix of
light of different wavelengths, not the overall light intensity or
luminance. For example, chromaticity coordinates are not sufficient
to describe the difference between yellow and brown (a dark yellow).
C.I.E.--Commission Internationale de 1'Eclairage, or International Commission
on Illumination, a scientific organization responsible for setting stan-
dards for light and color measurement and specification.
Class I areas—Areas such as national parks, wilderness areas, and national
forests that are afforded the most stringent air quality protection
both in terms of the significant deterioration air quality increments
and the visibility protection and restoration mandates of the Clean
Air Act (Section 169). Mandatory Class I areas are those areas desig-
nated as such by the legislature; however, Class II areas may be redes-
ignated as Class I.
coagulation—The process of particle growth resulting from particle collision,
coarse particles—Particles larger than 1 ym, caused principally by grinding
and mechanical process (e.g., soil dust).
color difference parameter (A£)—An index quantifying the difference between
two colors, in terms of light intensity differences as we'll as chroma-
ticity differences, transformed such that equal values of AE correspond
to equally perceived differences. The parameter is useful to character-
ize the overall perceptibility of haze layers resulting from color
differences.
color solid or volume—A three-dimensional representation of color. A color
can be located in a color volume by specifying Munsell hue, value, and
chroma or by specifying overall light intensity or luminance (Y) and
chromaticity coordinates (x, y). The color difference parameter (A£)
may be visualized as a distance between two points in a color volume
which has been transformed such that"equal distances correspond to
equally perceived color differences.
contrast—The fractional difference in light intensities of two colors. The
contrast defined at specific wavelengths of light is useful in charac-
terizing color changes. The contrast between a black object and the
clear, horizon sky is used in defining visual range.
diffuse radiation—See "multiple scattering."
elevation angle (g)--Angle between the horizontal and the line of sight
(sight path).
-------
189
extinction coefficient (bext)"The derivative of transmitted light intensity
with respect to distance along a sight path of light attenuation due to
light scatter and absorption. The extinction coefficient is a function
of wavelength and depends on concentrations and characteristics of
aerosol particles and absorbing gases such as NO;?. The extinction
coefficient is the sum of the scattering coefficient (bscat) anc' the
absorption coefficient (baDS).
fly ash—Primary particulate matter emitted from furnaces and boilers,
usually consisting chiefly of silica with traces of oxides of metals
such as aluminum and iron. Plume opacity at the top of a stack is usu-
ally caused by fly ash.
forward scatter—Situation where the sun is in front of the observer so
that direct solar radiation is scattered less than 90° into the
observer's line of sight (e<90°).
hue—Index in the Munsell color notation system characterizing the dominant
coloration (e.g., red, green, or blue).
integrating nephelometry—A technique that measures the scattering coeffi-
cient of a small volume of air within a chamber.
Koschmieder relationship--The mathematical expression for calculating the
visual range in a homogeneous atmosphere:
r = 3.912
v bext
liminal contrast—The contrast that is barely perceptible. This contrast
threshold will depend on the observer and lighting conditions, but a
liminal contrast of 0.02 is common and is used in the definition of
visual range.
line of sight (or sight path)—The line connecting the observer and the
observed object. Particles and gases in the atmosphere along this
line will affect the perceived color of the object by absorbing light
and by scattering light into and out of the line of sight.
luminance (Y)--The overall light intensity within the visible spectrum,
weighted by the photopic response of the human eye.
Mie scatter!ng--The theory describing scattering of electromagnetic
radiation by spherical particles of diameters of the same order as
the wavelength (A) of the radiation. Rayleigh scattering theory
covers scattering by particles with diameters much shorter than A.
multiple scattering—Radiation that has been scattered more than once.
Single scattering results when direct solar radiation is scattered into
-------
190
the line of sight. Multiple scattering occurs when direct solar
radiation is scattered at least once by gases and particles not in
the line of sight or is reflected from the surface of the earth or
clouds before being scattered by particles and gases along the line of
sight. Diffuse radiation has been scattered or reflected at least
once, while direct solar radiation, as its name implies, is radiation
directly from the sun.
Munsell color notation—A system of describing a color quantitatively by
reference to three indices (hue, value, and chroma). The Munsell Book
of Color displays color paint chips at specific intervals of hue, value,
and chroma.
nuclei mode—Particles less than 0.1 ym in diameter, which are not effective
in light scattering, but grow by coagulation to the accumulation mode.
opacity—A term characterizing the optical thickness of an aerosol layer,
usually used to characterize smoke plumes in or near the stack. Opacity
is usually expressed in percent and is defined as
Opacity = 1 - Transmittance ,
where transmittance is related to the optical thickness T as follows:
Transmittance = e~T
optical thickness (T)—The integral of the extinction coefficient of an aero-
sol between two points along a given line of sight.
particulate matter—Small solid or liquid particles, consisting of many mole-
cules, that are suspended in air.
perceptibility—As used herein, the characteristic of an object that makes
it visible to a human observer. Perceptibility results from differ-
ences in light intensity and color between two objects. For example,
a distant mountain is perceptible because it is darker than the back-
ground sky. Air pollution is perceptible if color differences exist
between a plume and a background, a haze layer and a capping layer, or
between a haze and a recollection of a clear day. (See also "liminal
contrast").
phase function—See "scattering distribution function."
pollutant flux—The total mass of a pollutant species in a plume passing
through a plane perpendicular to the plume centerline per unit of time.
primary particulate—Particles emitted directly from an emissions source
(e.g., fly ash).
-------
19.1
Rayleigh scattering—The theory describing scattering of radiation by
molecules or particles much smaller than the wavelength of the radi-
ation, resulting from a dipole interaction with the electric field
of the radiation.
saturation—See "chroma."
scattering angle—The angle between the vector describing the unscattered
radiation on an object and the vector along the line of sight between
the object and the observer.
scattering distribution function (or phase function)--The function describing
the direction in which radiation is scattered. For aerosols the scat-
tering distribution function is largest in forward scatter (e<45°),
which explains why haze layers are bright when the sun is in front of
the observer.
2
solar flux--The intensity of solar radiation (watts/m ) incident on a given
plane perpendicular to the solar rays.
spectral light intensity—The light?intensity along a particular line of
sight at wavelength A (watts/m /steradian). The spectral light inten-
sity can be considered the increment of radiant energy of wavelength A
to A + dA passing through an elemental area dA within a solid angle dw
along the given line of sight:
telephotometry--A technique for measuring the light intensity of a distant
object using a photometer coupled to a telescope. By measuring the
differences in light intensity between the clear horizon sky and dis-
tant mountains, one can estimate the visual range.
transmissometry—A technique of measuring the transmission of light through
the atmosphere by which the overall extinction can be determined.
tristimulus values—Indices that describe a given color by indicating the
amount of red, green, and blue light needed to match the color. Tris-
timulus values (X, Y, Z) are keyed to the wavelength responses of the
three color sensors in the human eye. They can also be translated
into chromaticity coordinates (x, y).
value—The index in Munsell color notation related to brightness (luminance
or overall light intensity).
visibility—See "visual range."
visibility impairment—A reduction in visual range, the presence of atmos-
pheric discoloration, or both. The Clean Air Act Amendments of 1977
-------
192
refer to "any" and "significant" visibility impairment, terms which
have yet to be defined in regulations.
visual range (or visibility)--The distance at which a large, black object
is barely visible when viewed against the horizon sky. For calculations,
it is convenient to use a more strict definition of visual range, i.e.,
the distance at which the contrast between a black object and the clear
(cloudless) horizon sky is reduced to 0.02. When calculating contrast,
one should use overall light intensity or luminance (Y) or, as an approx-
imation, the spectral light intensity at X = 0.55 ym, which is the mid-
point of the visible spectrum and the wavelength to which the human eye
is most sensitive.
zenith angle—The angle between the solar beam and the vertical at a given
location on Earth.
-------
193
TECHNICAL fJEPOHT DATA
({'lease read linlruc nuns un il r rcrcnc liclnrc cotni'lvtittfl
1. REPORT NO. 2.
EPA-450/3-78-nOa,b,c
4. TITLE AND SUBTITLE
THE DEVELOPMENT OF MATHEMATICAL MODELS FOR THE
•PREDICTION OF ANTHROPOGENIC VISIBILITY IMPAIRMENT
7. AUTHOR(S)
D. A. Latimer, R. W. Bergstron, S. R. Hayes, M. K. Liu,
J. H. Seinfeld, G. Z. Whitten, M. A. Wojcik, M.d. Hillye
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Systems Applications, Incorporated
950 Northgate Drive
San Rafael, California 94903
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Waterside Mall
401 M Street, S.W.
Washington, D.C. 20460
3. RECIPIENT'S ACCESSION-NO.
6, REPORT DATE
November 1 978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
» EF78-68A,B,C
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
EPA 68-01-3947. and 68-02-2593
13. TYPE OF REPORT AND PERIOD COVERED
Rnal Report: 10/77 to 9/78
14. SPONSORING AGENCY CODE
EPA-OPE/OAQPS
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report describes a nine-month study to recommend and develop models that pre-
dict the contribution of man-made air pollution to visibility impairment in federal
Class I areas. Two models were developed. A near-source plume model based on a
Gaussian formulation was designed to compute the impact of a plume on visual range
and atmospheric coloration. A regional model was designed to calculate pollutant
concentrations and visibility impairment resulting from emissions from multiple
sources within a region with a spatial scale of 1000 km and a temporal scale of
several days. The objective of this effort was to develop models that are useful
predictive tools for making policy and regulatory decisions, for evaluating the
impacts of proposed new sources, and for determining the amount of emissions reduc-
tion required from existing sources, as mandated by the Clean Air Act Amendments
of 1977. Volume I of this report contains the main text; Volume II contains the
appendices; Volume III presents case studies of power plant plume visual impact for
a variety of emission, meteorological, and ambient background scenarios.
17. KEY WORDS AND DOCUMENT ANALYSIS
». DESCRIPTORS
Air quality modeling
Visual range
Atmospheric discoloration
Power plants
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
b.lDENTIFlERS/OPEN ENDED TERMS
19. SECURITY CLASS (TltisKvpon)
UNCLASSIFIED
20. SECURITY CLASS (Tliit pancj
UNCLASSIFIED
c. COSATI i icid/Group
Vo. 'II--494J
22. PRICE
Vol. III--91
(PA farm 2220-1 (9-73)
-------
3 <£
0) Q)
33OO
CT -5 -£
y) — —
CD o n
Q3 fD ttl
>z
m > TJ m TI ~
§ 3 5.0,
I a
-------