EPA-9O8/ 1-77-OO5
A STUDY OF HORIZONTAL VISIBILITY,
ATMOSPHERIC VERTICAL OPTICAL PROPERTIES ,
SOLAR INSOLATION AT STANTON , NORTH DAKOTA
.US. ENVIRONMENTAL PROTECTION AGENCY
REGION VIII
J860 LINCOLN STREET
^DENVER , COLORADO
80295
NOVEMBER 1977
-------
EPA-908/1-77-005
FINAL REPORT
A STUDY OF HORIZONTAL VISIBILITY, ATMOSPHERIC
VERTICAL OPTICAL PROPERTIES AND SOLAR ISOLATION AT
STANTON, NORTH DAKOTA
by
ROLAND L. HULSTROM*
Program Manager
Martin Marietta Corporation
Denver Division
Denver, Colorado
Contract No. 68-01-3567
DAVID B. JOSEPH
Project Officer
U. S. ENVIRONMENTAL PROTECTION AGENCY
REGION VIII
Office of Energy Activities
1860 Lincoln Street
Denver, Colorado 80295
November, 1977
* Now with the Solar Energy Research Institute,
Golden, Colorado 80401
-------
DISCLAIMER
This report has been reviewed by Region VIII of the U. S. Environmental
Protection Agency, and approved for publication. Approval does not s
that the contents necessarily reflect views and policies of the U. S. Environ-
mental Protection Agency, nor does mention of trade names of commercial
products constitute endorsement or recommendation for use.
This report is available from the National Technical Information Service,
Springfield, Virginia 22161. A limited number of copies are also available
by contacting Ms. Betty Thalhofer, U. S. Environmental Protection Agency,
Region VIII, I860 Lincoln Street, Denver; Colorado 80295.
-------
ABSTRACT
A study to establish methods for the acquisition of baseline
data for horizontal visibility and vertical optical air quality
is described. The study was performed for the Environmental
Protection Agency, Region VIII. The site was Stanton, North
Dakota, which is located in an area where extensive develop-
ment of fossil fuel energy is anticipated. The baseline
measurements consisted of horizontal visibility along three
views, solar radiation and vertical atmospheric spectral
attenuation coefficients. Basic meteorological data were
also taken. These baseline data were established for the
months of April, July, and October; 1976, and January, 1977.
The methods used, results, conclusions and recommendations
are presented.
-------
FORWARD
The work described herein was performed at the Denver Division of the
Martin Marietta Corporation, under EPA Contract No. 68-01-3567,
entitled "A Study to Provide For the Acquisition of Baseline Visibility
Measurements at Stanton, North Dakota". The work was performed during
the period from February 1976 to November 1977-
The Martin Marietta Program Manager and Technical Director was
Mr. Roland L. Hulstrom, who has since taken a position with the
Solar Energy Research Institute, Golden, Colorado. The E.P.A., Region
VIII, Project Officer was Mr. David Joseph. Detailed review and
comments were supplied by Mr. David Joseph, Mr. Terry Thoem, Mr. Donald
Henderson, of E. P. A. Region VIII; and Mr. Lloyd Oldham of Martin
Marietta Corporation. Mr. Samuel Rensser and Mr. William Miles were
responsible for on-site data collection.
-------
TABLE OF CONTENTS
Page No.
Disclaimer ii
Abstract iii
Forward iv
1. Introduction 1
2. Summary, Conclusions, and Recommendations 3
3. Basic Concepts of Vertical and Horizontal Optical Air Quality 13
4. Horizontal Visibility 18
4.1 Techniques/Instrumentation 18
4.2 April Horizontal Visibility 27
4.3 July Horizontal Visibility 28
4.4 October Horizontal Visibility 28
4.5 January Horizontal Visibility 29
4.6 Seasonal Horizontal Visibility Comparisons 29
5. Atmospheric Vertical Optical Attenuation 49
5.1 Techniques/Instrumentation Used 49
5.2 April Optical Attenuation 55
5.3 July Optical Attenuation 56
5.4 October Optical Attenuation 57
5.5 January Optical Attenuation 58
5.6 Seasonal Optical Attenuation Comparisons 59
6. Incident Solar Irradiance 81
6.1 Techniques/Instrumentation Used 81
6.2 April Solar Irradiance 82
6.3 July Solar Irradiance 84
6.4 October Solar Irradiance 85
6.5 January Solar Irradiance 86
6.6 Seasonal Solar Irradiance Comparisons 87
7. Sky Appearance 119
7.1 Techniques/Instrumentation Used 119
7.2 April Sky Appearance 119
7.3 July Sky Appearance 119
-------
TABLE OF CONTENTS (Continued)
Page No.
7.4 October Sky Appearance 119
7.5 January Sky Appearance 119
7.6 Seasonal Sky Appearance Comparisons 122
8. Meteorology 123
8.1 Techniques/Instrumentation Used 123
8.2 April Meteorology 124
8.3 July Meteorology 125
8.4 October Meteorology 125
8.5 January Meteorology 126
8.6 Seasonal Comparisons 126
9. Satellite Imagery 159
9.1 Techniques/Instrumentation Used 159
9.2 April Satellite Imagery 159
9.3 July Satellite Imagery 159
9.4 October Satellite Imagery 159
9.5 January Satellite Imagery 160
9.6 Seasonal Satellite Imagery Comparisons 160
10. Comparison of Horizontal with Vertical Optical
Attenuation and Solar Irradiance 162
10.1 Horizontal Visibility Versus Vertical Optical Aerosol
Attenuation at 0.500 urn 162
10.2 Horizontal Visibility Versus Ratio of Diffuse to Total
Solar Irradiance 163
Appendix 164
-------
1. INTRODUCTION
Recently, a growing awareness of the possibility of man-made air
pollution degrading horizontal visibility led the Congress of the
United States to amend the Clean Air Act. Part of these new amend-
ments, Sec. 169A.(a)(3)(A), calls for a study and report to Congress
to include recommended methods for identifying, characterizing,
determining, quantifying and measuring visibility impairment in
Federal pristine areas, such as national parks. The purpose of
the study reported herein was to establish methods of measuring
horizontal visibility, plus methods for measuring the vertical
atmospheric optical air quality, and for quantifying the solar
insolation environment. The purpose of this study also included
the actual establishing of the baseline horizontal visibility.,
vertical optical air quality, and solar insolation at Stanton,
North Dakota. This site was chosen because it is located in an
area where extensive development of coal reserves is likely to take
place in the near future. Therefore, establishing the current base-
line conditions is necessary in order to determine the future
impairment due to the development and utilization of the area's
coal resources.
The Stanton, North Dakota area map is shown in Figure 1. The
measurements at this site included horizontal visibility,, atmos-
pheric vertical optical attenuation (vertical optical air quality),
incident solar insolation, ambient air pressure, temperature,
relative humidity-, wind direction, wind speed, vertical water
vapor content, and relative dust concentration. Data was collected
on site for four periods of time which were indicative of each
season. The first data collection period was from April 5 to May 1,
1976; the second was from July 12 to August 7, 1976; the third was
from October 4 to 30, 1976; and the fourth was from January 13 to
February 19, 1977. During each data collection period, a total of
-------
32 ) DATA COLLECTION SITE
ichard ton-Hebron
MAP EXPLANATION
U* NUMBCICD AND IKTEBSTATE NUMBLXED HIGHWAY!
Two Ion* Fo'td
VJ
St. Anthony fij
_int
FIGURE 1
STANTON, NORTH DAKOTA AREA MAP — VISIBILITY VIEWS DEFINITION
24 days of data were collected. For each day, data was collected
on an hourly basis from approximately 0800 to 1600 Local Standard
Time (Central). In addition to these data, the North Dakota Depart-
ment of Health collected data consisting of SO2, NOX) total suspended
partlculates, particle size distribution, ozone and sulfation rates,
-------
at various sites near the Stanton site. A list of these measure-
ments and their locations is given in Table 1 of Appendix A.
The following sections will present: (1) the basic concepts concerning
horizontal visibility, vertical optical air quality, solar insolation,
and their inter-relationships; (2) the techniques and instruments used
to make each measurement; (3) the results of the measurements; and
(4) the summary and conclusions of this study.
2. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
Various methods were established and used to quantitatively define both
the vertical (optical attenuation) and horizontal (visibility) optical
air quality at Stanton, North Dakota. These same methods can be used
to make similar determinations of optical air quality at other geograph-
ical sites, and are thereby employed to determine baseline conditions
and the impact of man-made pollution on horizontal visibility and verti-
cal optical air quality. In order to determine the impact of man-made
pollution on horizontal visibility and vertical optical air quality in
Federal pristine areas, baseline studies similar to the Stanton, North
Dakota study have to be performed in order to establish the natural
conditions and the variations under natural conditions.
The vertical atmospheric optical air quality measurements consisted of
the atmospheric spectral (0.380, 0.440, 0.500, 0.640, and 0.880 urn) and
broadband (0.30 - 2.8 um) attenuation coefficients. Vertical atmos-
pheric optical air quality also included the broadband (0.30 - 2.8 um)
direct, total, diffuse, and ratio of diffuse to total solar insolation
as functions of relative air mass. The spectral attenuation coefficients
consisted of both the total attenuation, which includes attenuation due
to molecular scattering, aerosol scattering, and ozone absorption; and
the attenuation due to only aerosols. The horizontal atmospheric
optical air quality measurements consisted of visibility determined by
-------
a measurement of contrast reduction of an object versus the backgtround
sky- The contrast reduction was determined by the photographic method.
Meteorological measurements consisted of temperature, pressure,
relative humidity, wind speed, wind direction, precipitable water vapor,
relative dust concentration and all-sky cloud photography. The general
sky appearance and color was documented with color photography. The
ground-based measurements were complimented with LANDSAT 1 and 2 satellite
imagery to determine if it could readily locate and determine the
source of smoke plumes.
The results and analyses of the horizontal visibility and meteorological
data indicated the following:
1) For any given season/month the horizontal visibility can be quite
variable, depending on viewing direction, and time of day.
2) Marked day to day variations in visibility occur for all seasons/
months. These variations are relatively similar for all views.
3) Variations in horizontal visibility occur on a seasonal/monthly
basis. When the average hourly overall horizontal visibility for
each month, for cloud free conditions, was compared it was determined
that July had the greatest visibility, followed by October, April,
and January. The daily, average, overall view (average of three
views), considering both cloud free and cloud present (clouds
behind target) conditions for April is 34.0 miles for July is 39.9
miles, for October is 32.2 miles, and for January it is 27.2 miles.
If only cloud free conditions are considered the corresponding
values are 34.9, 42.1, 38.9, and 32.3 miles, respectively. If only
cloud present conditions are considered the values are 33.4,
38.3, 26.9, and 24.9 miles, respectively. Hence, the derived
visibility is also dependent upon the sky conditions, with cloud
present conditions having somewhat lower visibilities.
-------
The results and analyses of the vertical optical attenuation coefficients
indicated hourly, daily, and monthly variations. However, the variations
are not necessarily similar for all wavelengths. The days having the
highest overall attenuation also displayed a negative slope attenuation
versus wavelength curve. This type of attenuation versus wavelength
simply means that the shorter wavelengths display greater attenuation
than the longer wavelengths. It has been shown that such a negative
slope distribution is characteristic of a dominance of the smaller atmos-
pheric particles/aerosols. Hence, the days having high attenuation were
caused by the presence of small particles. Conversely, days having low
attenuation values were characterized by a positive slope attenuation vs
wavelength curve; thereby indicating a lack of the smaller particles.
It was not within the scope of this study to perform attenuation versus
particle size distribution analyses in order to infer the corresponding
size distributions. However, it is generally known that particles in
the 0.1 to 1.0 um size range account for the scattering of sunlight. The
results of this study, indication of positive and negative attenuation
versus wavelength slopes, points out the importance of taking multispec-
tral vertical optical attenuation data. The five wavelengths (0.380,
0.440, 0.500, 0.640, and 0.880 um) considered in this study provide what
is considered to be a minimum number of wavelengths. A single wavelength
or two wavelength determination would not be sufficient to define the
slope and characteristic of the attenuation versus wavelength curve.
The old two channel (0.380 and 0.500 um) Volz photometer is considered
to be insufficient.
The average hourly broadband (0.3 to 2.8 um) attenuation coefficient was
examined for all seasons/months. It was determined that July had the
greatest atmospheric attenuation, followed by April, October and January.
The hourly broadband attenuation coefficient for July ranged between
0.265 - 0.370, for April it was 0.245 - 0.315, for October it was 0.148 -
0.215, and for January it was 0.105 - 0.160. Hence, it can be seen that
quite significant variations in broadband atmospheric attenuation occur
throughout the year/seasons. However, this broadband attenuation is due
-------
to both atmospheric scattering (molecules and aerosols) and absorption
(mainly water vapor). In order to isolate the attenuation due to aero-
sols, the 0.640 urn narrow band was analyzed for each season/month. The
results showed July to have the highest attenuation/aerosols (0.105 -
0.155), followed by April (0.113 - 0.120), October (0.055 - 0.075) and
January (0.021 - 0.043). These narrow band aerosol attenuation results
agree with the broadband attenuation results, in terms of the relative
clarity of the vertical atmosphere from season to season. However,
recalling the horizontal visibility results mentioned previously, the
vertical atmospheric air quality results do not indicate the same
results concerning clarity as the horizontal visibility results. The
horizontal atmospheric clarity was greatest in July, followed by October,
April and January. The vertical atmospheric clarity was greatest in
January, followed by October, April and July. The relative horizontal
and vertical atmospheric clarity results agree for October and April,
but they are exactly opposite concerning January and July. The horizon-
tal clarity was greatest during July, while the vertical clarity was
the least during July. This is due to the fact that the vertical and
horizontal atmospheric conditions can be quite different and are not
necessarily relatable. For example, a strong temperature inversion near
the surface can trap aerosols near the surface, resulting in poor hori-
zontal visibility/clarity; however, the temperature inversion is so
shallow that the overall vertical atmospheric clarity can actually be
quite good. This situation may exist quite often in the winter season.
These results indicate the need for measuring both the horizontal ard
vertical atmospheric optical air quality.
The analysis of broadband 0.3 to 2.8 urn, solar insolation data consisted
of seasonal/monthly plots of direct, diffuse, total, and ratio of diffuse
to total insolation versus relative air mass. The direct solar insola-
tion versus relative air mass plots were able to delineate clear days,
hazy days, and cloud present conditions. By plotting all direct solar
insolation measurements versus relative air mass, a hypothetical clearest
-------
baseline day was established. This was performed for each season/month,
which indicated that January was the clearest vertically, followed by
October, July, and April. It is interesting to note that this ranking
of vertical clarity is slightly different than that derived, as shown
previously, by averaging the broadband attenuation coefficient. This
simply indicates the difference between averaging and taking the extreme
clarity. Plots of the diffuse and ratio of diffuse to total solar
insolation versus relative air mass were also able to delineate
atmospheric vertical clarity- However, the influence of clouds on
the diffuse insolation make it a less usable indication of clarity, as
compared to the direct insolation. As with the direct insolation, a
hypothetical clearest day baseline was established for the diffuse and
ratio of diffuse to total insolation by considering the lowest values
at any given relative air mass. By analyzing such plots, it was shown
that the slope of the ratio of diffuse to total insolation versus
relative air mass is directly proportional to the broadband optical
attenuation coefficient. By considering the slope of such plots, the
variable impact of ground albedo on the diffuse insolation was eliminated.
One of the most interesting results was obtained by plotting the total
insolation versus relative air mass for all months. This plot revealed
that despite a widely differing atmospheric clarity (as defined by the
direct insolation versus relative air mass plots), the amount of total
insolation remains nearly constant for a given relative air mass. The
only exception is for January, when snow cover results in higher values
of total insolation at a given relative air mass, due to reflection and
back-scattering toward the ground. This constant relationship of total
insolation versus relative air mass indicates that when the direct beam
is attenuated (by scattering), the diffuse sky insolation increases an
amount which compensates for the loss in direct insolation. In this
manner the total (direct 4- diffuse) remains constant. This total versus
relative air mass represents one of the more important baselines,
because it is indicative of the radiation balance of the atmosphe^^ and
-------
ground. This is related to the nature of the atmospheric aerosols/
particles from the standpoint of their ability to scatter sunlight,
and therefore, increase the diffuse insolation to make up for the loss
of direct insolation. If the atmospheric aerosols/particles were
changed to a man-made soot type aerosol, they would tend to absorb the
sunlight, and therefore, not contribute to the diffuse component, which
would result in a lower amount of total insolation at a given relative
air mass. This would result in less energy input to the ground and
would result in a modification of the heat balance. This would result
in a direct impact of the man-made air pollution on the natural state.
Hence, the total insolation versus relative air mass should be estab-
lished in Federal Class I areas and subsequently monitored to determine
any impact.
The sky appearance from photos indicated smog conditions were apparent
during two days in April, during five days in July, during one day in
October, and during two days in January. In April, the smog was sited
to the northeast; in October, it was sited to the north; and in January,
it was sited to the north and east. Obviously, the smog is to the
north of the Stanton site, which corresponds to the Knife and Missouri
River valleys which contain several sources of air pollution (see
Figure A-l in the Appendix).
The meteorological results showed that January was the coldest month
having an average temperature (during the hours from 0800 to 1600 LST)
of -9.62°C, followed by October with 5.81°C, April with 12.7°C, and
July having the warmest temperature of 25.7 C. January had the highest
relative humidity - average of 78.9%, followed by October, 53.1%,
April, 47.7%, and July, 38.6%. However, in terms of average total
vertical precipitable water vapor, January had the lowest value of
0.35 cm, followed by April, 0.55 cm, October, 0.65 cm, and July, 1.35 cm.
-------
The satellite imagery analyses showed that a smoke plume, from a large
coal fired power plant near Stanton, could be easily detected by simply
doing a 7X enlargement of the original LANDSAT positive trans-
parency. The smoke plume source, width, length, and altitude above
terrain can be determined. A qualitative assessment of the smoke
plume opacity can be determined by comparison of the multispectral
images. The shortcoming of the LANDSAT satellite images is the
fact that it passes over the same ground location only once every
18 days. Therefore, the frequency of sampling is poor. For the
month of January, Jan. 19, a large smoke plume was detected, which
originated from the plants at Stanton. This plume was 7 miles long,
0.4 miles wide, and was 850 feet above terrain. The plume appeared
to be very opaque.
Various correlation analyses were performed in order to determine
any relationship between horizontal and vertical atmospheric optical
air quality. Such relationships would depend on the vertical structure
of the atmosphere in terms of the altitude distribution of aerosols
and water vapor. Hence, such a relationship could be quite variable,
depending on the atmospheric vertical structure. The horizontal
visibility, for an overall view, was compared to the vertical optical
attenuation coefficient for a wavelength of 0.500 jim. For all months,
an inverse correlation was found indicating low aerosol vertical
attenuation when high visibility is present. However, the degree of
correlation varied with the season/month. For the month of April, a correl-
ation coefficient (a value of 1.0 indicates a perfect correlation) of
0.38 was obtained; for July a value of 0.56 was obtained; for October
a value of 0.51 was obtained; and for January a correlation coefficient
of 0.45 was obtained. These values of correlation are not considered to
be very high; but they are high enough to indicate a definite correlation
of good vertical atmospheric clarity with good horizontal atmospheric
clarity. The correlation is not high enough, however, to be able to
measure only the horizontal atmospheric clarity and infer the vertical
clarity.
-------
A second correlation analysis was performed to determine any
correlation between horizontal visibility and the ratio of diffuse
to total solar insolation. For all months, an inverse correlation was
obtained, which indicates that low values of the ratio of diffuse to
total insolation are associated with high visibilities. This correlation
is highly dependent upon season however. For the month of April a
correlation coefficient of 0.27 was obtained, for July a correlation
of 0.24 was obtained; for October a correlation of 0.82 was obtained,
and for January a correlation of 0.81 was obtained. The wide range
of correlation is interesting and perplexing. It could be due to the
particular sky conditions, in terms of clouds present. In other words,
clouds present can significantly affect the diffuse insolation while
not affecting the horizontal visibility. Because cloud cover can be
so variable, the correlation between the ratio of diffuse to total
insolation and horizontal visibility can be quite variable.
Several analyses of data for both vertical and horizontal atmospheric
optical air quality were generated. The question arises concerning
what data, analyses, and methods should be used as baselines to
quantitatively define existing, natural conditions in Federal Class I
air quality areas. The following are the recommended measurements and
baselines.
A. Horizontal Optical Air Quality
1. Average hourly visibility for several views (3), for each
season (minimum of one month), for cloud free and cloud
present conditions.
2. Average hourly visibility for an overall view (average
of all views), for each season (minimum of one month),
for cloud free and cloud present conditions.
3. A plot of the percentage of days having an average daily
overall view visibility above a given level. This curve
should be analyzed, as shown in Section 4.6, to determine
the index of visibility quality.
10
-------
B. Vertical Optical Air Quality
1. For all clear days of each season (minimum of one month)
plots of spectral aerosol attenuation coefficient versus
wavelength. The spectral points should be a minimum of
0.380, 0.440, 0.500, 0.640, and 0.880 pi.
2. For each month, plot of the average hourly optical attentuation
coefficient at 0.640 pi and the broadband, 0.3 to 2.8 pi,
attenuation coefficient.
3. Plots of the direct broadband solar insolation and the ratio
of diffuse to total insolation versus relative air mass for
each season (minimum of one month), for the hypothetical
clearest conditions, as derived in Section 6.6.
4. A plot of total insolation versus relative air mass for each
season (minimum of one month), and for the entire year
(each season).
The above data analyses are considered to be a minimum that is required
to establish the baseline conditions of horizontal and vertical optical
air quality. If possible, they should be expanded to include analyses
of the broadband visible (0.30 to 0.700 pi) and broadband near infrared,
(0.70 to 2.8 pi) solar insolation, as described in items 2., 3., and 4.
under vertical optical air quality. By doing this the effects of
aerosols (0.30 to 0.70 pi) can be separated from the effects of water
vapor absorption (0.70 to 2.8 pi). This can be done by simply using
an RG-8 Schott filter on the pyrheliometer (direct beam measurement)
and the pyranometer (total insolation measurement). This was done at
Stanton, but the scope of this study would not allow the analysis of
the additional data. It is recommended that the Stanton data be
additionally analyzed for the 0.30 to 0.70 and 0.70 to 2.8 pi baselines.
It is generally known that relative humidity can significantly affect
visibility, especially for relative humidities above 7070. In order
to separate the effects of natural relative humidity on visibility
from the affects of man-made pollution on visibility, it is recommended
that the horizontal visibility data be stratified in terms of relative
11
-------
humidity. The regions of relative humidity stratification should be
determined from past experiments (if possible) and future specifically
designed experiments and studies.
Finally, it is recommended that a detailed study be undertaken in order
to quantify the relationships between the human eye visibility,
photographic visibility, and photopic-photometric visibility. All
three of these visibilities are currently being pursued and analyzed
by various agencies and institutions. In order to properly respond
to the amended Clean Air Act, the term "visibility" and its proper
measurement has to be specifically designated. An experiment is
called for, where measurements of all three visibilities are performed
and compared.
12
-------
3. BASIC CONCEPTS OF VERTICAL AND HORIZONTAL OPTICAL AIR QUALITY
This section is provided as a general discussion of vertical and
horizontal optical air quality, and the reasons for measuring the
parameters that were measured in the Stanton, North Dakota baseline
study. The detailed techniques and instrumentation used for defining
vertical and horizontal optical air quality are presented in later
sections.
A general illustration of the applicable interactions of sunlight
with the atmosphere are illustrated in Figure 2. The direct solar
beam, I, is attenuated by the atmosphere due to the processes of
scattering and absorption. The attenuation due to scattering is
made up of that due to molecular scattering and that due to aerosols.
The sunlight that is scattered from the direct beam is back-scattered
out of the atmosphere and is also downward scattered and eventually
reaches the ground. This downward scattered sunlight makes up the
skylight, S. The scattering due to molecules is a very marked
function of wavelength, l/(\4-), such that the shorter wavelengths
are scattered much more severely than the longer wavelengths. This
selective scattering, due to molecules, results in the blue skylight.
Aerosols scatter sunlight less wavelength selective than do molecules.
The exact wavelength dependence of aerosol scattering depends on their
size distribution, index of refraction, and albedo; however, natural
aerosols have been shown to scatter sunlight according to a 1/C1.3)manner.
Since the molecular atmosphere is relatively constant, variations in
attenuation of sunlight due to scattering are caused by variations in
the nature and amount of aerosols. The attenuation of the direct beam
due to absorption is highly wavelength dependent because of the
absorption bands of the various atmospheric constituents. However,
over the 0.30 to 2.8 _um solar spectrum, water vapor accounts for
nearly all of the attenuation due to absorption. This absorption
occurs in the 0.70 to 2.8 ^m region. Because atmospheric water
13
-------
Zenith
Sun
/ \ I , Extraterrestrial
Direct Solar
Insolation
Scattered
Solar
Insolation
9Q, Solar
Zenith Anglk
Ground Level
Insolation
by Absorption
I, Direct Solar
Insolation
"Top" of Atmosphere
/ /
/ /
S, Diffuse Skylight /
Insolation (Sky & Clouds)
/ /
//
~hfrs
Sunlight
/
Target
H, Total Insolation
(H = I Cos 9Q + S)
FIGURE 2 - INTERACTION OF SUNLIGHT WITH ATMOSPHERE
B, = Horizon Sky Luminance
B, = Target Luminance
-------
vapor is quite variable, the attenuation of sunlight in the 0.70 to 2.8 urn
region can be quite variable. Hence, the variable nature of atmospheric
attenuation of sunlight is because of aerosols in the visible portion of
the spectrum (0.30 - 0.70 urn); and because of water vapor in the near
infrared region (0.70 - 2.8 urn).
Mathematically, the atmospheric attenuation of sunlight can be
expressed as follows.
, T - Tsec 9 , ,
I = I e o (1)
o
where I is the intensity of the direct solar beam at ground level,
I is the extraterrestrial intensity ("Solar Constant"), sec 9Q is
the relative path length of the beam through the atmosphere (known
as the relative air mass) and T is the total optical attenuation
coefficient of the atmosphere. The total optical attenuation
coefficient is made up of the components as follows,
T = T"-l--7~-4-"P -4- T -f-T
— '^*'TT/-\'*^ ' f^r\ /ON
m a H_0 Oo CC^ (2)
where T is due to molecules, T is due to aerosols, 77, n is due
m 'a ^2^
to water vapor, Tn is due to ozone, and T,,,, is due to carbon
Uo L/Up
dioxide. As pointed out previously, T and T are the most
3 Flo (J
significant and are also the most variable. The atmospheric
optical attenuation coefficient and the components, shown in
equation 2, are the vertical attenuation coefficients. These
vertical attenuation coefficients are related to horizontal
coefficients at various levels (altitudes) in the atmosphere by
oo
T = f& dZ (3)
m
T = dz (4)
T =
0 /~
=/'
a = a (5)
3
z -*
TC00V/»C02aZ (8)
15
-------
where z is the surface altitude, z is altitude above the surface,
o
and /3, /3 , /3 , are the corresponding horizontal attenuation
rn 3
coefficients at various altitudes.
The diffuse skylight is a complex function of T^, Tg, sec9Q, ground
albedo, and the geometry between the direct beam and the aerosols
and the molecules. However, in a general sense, the diffuse skylight
is dependent upon T and sec 9 (relative air mass). As T increases
and sec 9 increases the greater amount of sunlight is scattered from
the direct beam.
The total amount of insolation at the surface, H, is the sum of the direct
insolation
S. Hence,
insolation on a horizontal surface, IcosB , and the diffuse skylight,
H = Icos9 + S (9)
o
where the solar insolation components, I and S, are determined by the
vertical optical properties.
In the lower atmosphere, near the surface, the horizontal optical
air quality parameter is the horizongal visibility, Figure 2. The
horizontal visibility, V, 3s approximated by:
V = 3.912 (10)
08(°>
where j8g(o) is the near surface scattering attenuation coefficient,
which is given by
+0 (o)
where 0m(o) and 0a(o) are the molecular and aerosol components.
The relationship between the vertical optical air quality and the
horizontal optical air quality is shown when equations 4 & 5 are
compared to equations 10 & 11. As can be seen the vertical optical
attenuation coefficients are simply the integral of the optical
attenuation coefficients as a function of altitude. The horizontal,
near surface, optical properties represents one increment of the
16
-------
integral. Therefore, the vertical and horizontal optical air quality
are not directly related and represent distinct properties of the
atmosphere.
In order to define the vertical and horizontal optical air quality,
the parameters to be measured would be the various optical attenuation
coefficients, as defined previously. The most important parameter,
in terms of detecting any man-made pollution effects, would be the
aerosol attenuation coefficients. Therefore, the result of
changes in the aerosol component should also be determined by the
impact on the solar insolation components. The measurements at Stanton,
North Dakota were designed to establish a baseline for the vertical
and horizontal optical air quality. To do this the following
measurements were selected:
A. Vertical optical attenuation coefficients at various
wavelengths and over the entire solar spectrum.
B. Direct, diffuse, and total solar insolation.
C. Horizontal optical air quality - visibility.
17
-------
4. HORIZONTAL VISIBILITY
4.1 Techniques/Instrumentation - A photographic technique was used used to
measure horizontal visibility. The basis of this technique has been des-
cribed by Middleton (Ref. 1). This technique involves photographing a
distant, black, object on the horizon then measuring the contrast between
the object and its background sky. If the luminance of the black object
is B,, and the luminance of the horizon sky is Bh, then (as shown in
Ref. 1, page 63)
Bb = B, (1 - e-br) (12)
where b is the horizontal attenuation coefficient, due to scattering, and
r is the range to the object. The meteorological visual range, V, is re-
lated to the attenuation coefficient by the approximation (see Ref.l p. 105)
V = 3.912/b (13)
If the distant object and horizon sky are photographed, the resultant film
exposures produced by the horizon sky, E and distant object, E , are
Eb = Bb x t (14)
where t is the exposure time (camera shutter speed). Because the
camera shutter speed is identical for each exposure, it is obvious that
E. B.
(16)
Eh Bh
Combining equations 12 and 13, the ratio E /E is related to meteorolog-
ical visual range, V, by,
V = -r x 3.912/ln(l-E, /E ) (17)
Ref. 1: W. E. Knowles Middleton; Vision Through the Atmosphere. University
of Toronto Press, 1952.
18
-------
where the units of V (miles, kilometers, etc) are the same as used for
r, hence; by using photography to determine the quantity E /E , and
knowing the distance to the target, r, the meteorological range can
be measured. This method/technique, is based upon Koschmieder's
theory, as discussed in Ref. 1, having the following assumptions:
(1) The atmosphere is considered as a turbid medium, containing
a large number of small particles.
(2) Each element of volume contains a very large number of
particles, each of a smaller order of magnitude then the
element itself.
(3) The scattering action of each particle is independent of the
presence of all the others; i.e., multiple scattering is
neglected.
(4) The light scattered from an element of volume will be
considered as coming from a point source of which the
intensity is proportional to the number of particles.
(5) Light rays will be considered as rectilinear, that is to say
atmospheric refraction will be neglected.
(6) All parts of the atmosphere in the horizontal plane are
equally illuminated.
(7) The coefficient of attenuation by scattering, b, is constant
in a horizontal plane, in particular near the surface of the
earth.
(8) The curvature of the earth is neglected, and its surface is
considered as plane, horizontal, and diffusely reflecting.
(9) The linear dimensions of the whole observed object are small
in comparison to its distance from the observer.
(10) As stated previously, the object is assumed to be black.
19
-------
Of these assumptions, No's 6, 7, and 10 are the ones that will be most
likely violated in the real/actual case. The assumption that all parts
of the atmosphere are equally illuminated is subject to violation due
to cloud cover changes. In addition, if clouds exist behind the target
they will cause an erroneous value of Eh to be measured. Hence the
presence of clouds can cause erroneous results. The assumption that
the scattering attenuation coefficient is constant in the horizontal plane
is subject to violation because of localized anomalies such as smoke
plumes, dusty roads, open coal pit operations, or any other localized
source of particulates. The assumption that the object is black is
probably the most serious violation. Totally black objects/targets
simply do not exist in the real situation. If sunlight is reflected
from the target, an erroneous value for E is obtained. It is
interesting to note that if sunlight is reflected by the target, the
resultant calculated meteorological range will be low. This is because
the luminance originating from the target is interpreted as additional
luminance due to additional particles and forward scattering; and
therefore, a lower meteorological range. On the other hand, if a
cloud is present behind the target, and has a greater luminance than
the cloud free atmosphere, an erroneously high value of B, will be
obtained which will result in erroneously high values for meteorological
range. Hence, in the real situation, these errors may tend to offset
each other. The magnitude of the error introduced by reflected
luminance from the object is determined by the object's albedo and
the amount of incident sunlight. The amount of incident sunlight is
determined by the sun angle to the objects surface and atmospheric
attenuation and cloud cover. Similarly, the amount of error introduced
by clouds behind the target is dependent upon the incident sunlight on
the cloud and the cloud albedo.
In terms of absolute accuracy of meteorological visual range, as pointed out
in the above discussion, the photographic method/technique (or any other
20
-------
technique that relies on the measurement of a distant target luminance
versus the background sky luminance) cannot be considered as highly accu-
rate. However, in terms of a relative indicator of visibility for a given
set of conditions, it is valid because the basic measurement is of the
luminance ratio between the target and the background sky. This ratio is
certainly an indicator of visibility because the luminance of the target
will be a function of the luminance produced by the scattering of light
by particulates between the receiver and target. The luminance level of
this scattered light is a function of the number, size and characteristics
of particulates, assuming negligible absorption by the particulates. The
set of conditions includes such items as target albedo, sun angles (time
of day, year), cloud cover, etc. For a given time of year and day, sun
angles are obviously constant; and, if no marked changes occur that change
the target, the target albedo will most likely be fairly constant. The
most likely variable will be cloud cover. For this reason, in the following
reported results, cloud free conditions and cloudy conditions were reported
separately.
It should be pointed out that meteorological range, as determined photo-
graphically, is not identical to visibility as perceived by a human obser-
ver. The main differences are because they are both wavelength sensitive
and the detection of the ratio of E, /E. is in one case determined by a
b h
densitometer and in the other case determined by the eye. The wavelength
sensitivity arises from the fact that the attenuation coefficient, b, is
wavelength dependent; therefore, as shown in Equation 13, visibility will
be wavelength dependent. Since the photographic film (Panotomic X, B&W)
and the eye do not respond identically to all wavelengths, a basic difference
exists between the photographically determined meteorological range and the
eye visibility. Another basic difference exists due to the sensitivity of
the eye to a small difference between E and E versus the comparitive
response of a densitometer. However, despite these differences, a good
correlation should exist between the photographically determined meteoro-
logical range and the human eye determined visibility, because they are both
21
-------
sensitive to the relationship between Efe and E^. The instrumentation
used for photographically determining meteorological range, hereafter
referred to as visibility, consisted of an eight inch diameter/aperture
reflective, telescopic, 2000 num. focal length lens (Celestron 8); and,
a Honeywell Pentax Spotmatic 35 mm camera shown in Figure 3.
FIGURE 3 - HORIZONTAL VISIBILITY DATA COLLECTION
CAMERA SYSTEM
The film used was Kodak Panotomic X B&W. The densitometer used was a
MacBeth Model TD 404 Diffuse Densitometer.
The particular film processing included the use of calibrated step-
wedges (21 steps) on each roll of film. This produced a film density
versus log exposure (D log E) curve for each roll of film. This was
performed both before and after the field mission. This produced curves
by which measured film densities of the target, D , and horizon sky,
D , could be converted to the corresponding exposures, E and E .
n D n
22
-------
An example of such a curve is shown in Figure 4. The critical
relationship necessary in order to perform such a conversion is that
the relationship between film density and the log E be linear, at
least over that range of exposures produced by the target and sky
luminance. As shown in Figure 4 the D log E relationship is linear
over the desired range of density. This was accomplished by
experimentation with the film development time. It was determined
that by using Kodak D19 developer and a developing time of 7 minutes,
a linear relationship was achieved.
The field data collection consisted of photographing three views, as
shown in Figure 1, every hour from 0800 to 1600. The horizontal visi-
bility, using the 2000 mm leas system, was determined for views/targets
towards 1> 2, and 3. Target No. 1 (see Figure 5) is an open coal pit
dump, having a fairly low albedo ( ~ 207»), approximately 15 miles from
the measurement site. Target No. 2 (see Figure 6) is a hillside,
having a soil and grass cover, located approximately 15 miles from the
site. Target No. 3 (see Figure 7) is a hillside, having a similar cover
as No. 2, located approximately 16 miles from the site. View No. 4
(see Figure 8) was taken with a 200 mm lens for sky appearance
documentation only.
The baseline horizontal visibility parameters established were as
follows:
(1) Average hourly visibility for three views
(2) Average hourly visibility for overall view (average of 3 views)
(3) Average daily visibility for three views
(4) Average daily visibility for overall view.
23
-------
Film = Panotomic X, Kodak
Emulsion # - 5060 125 6
Shutter Speed - 1/250
3.0 Developer - D19
" Time - 7 minutes
2.0
E
M
1.0
Range of target and
sky densities
__ _J
Figure 4 -
1.5
Log (Exposure)
D LOG E CALIBRATION CURVE FOR RELATING
FILM DENSITY TO RESULTANT EXPOSURE
3.0
24
-------
VISIBILITY TARGET #1
FIGURE 5 - TARGET #1 - VIEWED THROUGH 2000 ram LENS
VISIBILITY TARGET #2
FIGURE 6 - TARGET #2 - VIEWED THROUGH 2000 mm LENS
25
-------
VISIBILITY TARGET #3
FIGURE 7 - TARGET #3 - VIEWED THROUGH 2000 mm LENS
FIGURE 8 - VIEW #4 - SKY APPEARANCE VIEW - 200 mm LENS
26
-------
4.2 April Horizontal Visibility - The results for the April data collection
mission are shown in Figures 9, 10, 11, and 12. They consist of the
average hourly visibility for three views, the average hourly visibility
for an overall view, the average daily visibility for three views, and
the average daily visibility for an overall view. In all cases, the
cloud free and cloud present conditions are designated separately
because of the previously mentioned influences of clouds. The cloud
present conditions refer to a condition when clouds were on the horizon,
behind the target. As Figures 9 and 10 point out, both the hourly and
daily visibility is dependent upon the view. This is true for both
cloud free and cloud present conditions. The hourly dependence of
view 1 exhibits a general increase, for both cloud free and cloud
present conditions, in visibility throughout the day. This increase
is from approximately 25 miles in the early morning hours to approximately
43 miles in the late afternoon. The hourly dependence of view 2
exhibits a much smaller, compared to view 1, fluctuation for both
cloud free and cloud present conditions. The view 2 visibility is
fairly consistant throughout the day, having a visibility of approxi-
mately 25 to 30 miles. The hourly dependence of view 3 exhibits a
difference between the cloud free and cloud present conditions,
especially for the morning hours. The reason for this is not known,
possibly it is due to a sun angle/illuminance effect, i.e., target 3
is sunlit in the morning and shaded in the afternoon. The average
daily visibility for the three views, Figure 11, exhibits day-to-day
variations. In general, view 2 exhibits less visibility than views 1
and 3. All views tend to exhibit similar day-to-day variations. The
hourly dependence for the overall view, Figure 10, displays a general
increase in visibility, for both cloud free and cloud present conditions,
throughout the day. The morning visibility is approximately 30 miles
and the afternoon values approach 40 miles. The overall view daily
visibility, Figure 12, exhibits day-to-day variations ranging from
20 miles (mist-rain) to 41 miles. In addition, any specific day can
27
-------
display a wide range of visibility.
4.3 July Horizontal Visibility - The results for the July data collec-
tion mission are shown in Figure 13, 14, 15, and 16. As concluded
from the April data, both the hourly and daily visibility is dependent
upon the particular view. View 1 displays a general increase in
visibility throughout the day. Views 2 and 3 display a relatively
consistant visibility throughout the day. For cloud free conditions,
view 1 has visibilities ranging from 35 miles in the morning to 55
miles in the afternoon. View 2 displays visibilities of approximately
40 miles. View 3 displays visibilities ranging from about 40 miles
to a maximum of 48 miles at noon. As with the April results, View 2
displays lower visibilities than views 1 and three. The average
daily visibilities for the three views, Figure 15, shows similar
day-to-day variations for all views, with View 2 having the lowest
visibility. The overall view shows a fairly consistant hourly de-
pendence, Figure 14, ranging from 39 to 47 miles. The overall view
daily visibility, Figure 16, shows day-to-day variations, with a
maximum of 63 miles and a minimum of 25 miles.
4.4 October Horizontal Visibility - The October data collection mission
results are shown in Figures 17, 18, 19, and 20. The plots for the
hourly and daily visibilities for the three views again demonstrate the
dependence of visibility upon the view, hour, and day. View 1 again
exhibits a general increase in visibility throughout the day. Views
2 and 3 again exhibit a relative constant visibility throughout the
day. View 3 also exhibits less visibility than views 1 and 2. The
daily average visibility for 3 views and the overall view daily
average visibility, Figures 19 and 20, exhibit large day-to-day changes.
These day-to-day changes are reflected in a similar manner for all
views. The average hourly visibility for an overall view, Figure 18,
shows a very weak dependence upon time of day. In addition, the
cloud present conditions have lower visibilities than the cloud free
conditions. However, the time dependence is similar for both conditions,
28
-------
4.5 January Horizontal Visibility - The January data collection mission
results are shown in Figures 21, 22, 23, and 24. The January results
exhibit features that have been pointed out for the April, July and
October data.
4.6 Seasonal Horizontal Visibility Comparisons - A comparison of the
daily average visibility, for an overall view, is shown in Figure 25,
for all the seasonal months. As can be seen, July exhibits the greatest
visibility, followed by October, April, and January. However, an ab-
solute comparison is dependent upon time of day. July and October
display increasing visibilities throughout the day, while April and
January display a relatively constant visibility throughout the day.
In addition, an absolute visibility comparison between the various sea-
sonal months is complicated by the fact that the targets were not the
ideal black target, therefore, it is possible that there exist variations
due to sun angle and changing target albedo. The target albedo was
observed to be fairly low («.2) and consistent for the months of April,
July, and October, however the target was snow covered during January.
However, it was possible to locate "dark" areas on each of the targets
for January, thereby greatly reducing the adverse effects of snow
cover. Shown in Figure 26 is another type of seasonal visibility com-
parison. The percent of days, out of the total number of days measured,
that a given daily average visibility, plotted on the x axis is exceeded
is shown for each of the seasonal months. Using this type of comparison,
it is concluded that July had the best visibility, January had the
worst visibility, and April/October had different characteristics.
April had a greater number of days with visibilities greater than 15
to 35 miles; whereas, October had a greater number of days with visi-
bilities between 35 to 50 miles. It is obvious, from Figure 26, that
the overall visibilities at Stanton, North Dakota do not exceed 50
miles except for a very minimal portion of time. The 15 mile lower
29
-------
limit is established by the distance to the targets. There are
several ways by which to interpret the results shown in Figure 26,
in order to assign an overall quantitative index to the visibility
for each season. One possible method is to establish the 50% level
for each season. This level would indicate what visibility was exceeded
over 50% of the days. For January this level is about 25-30 miles;
for April this level is about 35 miles; for October this level is
about 35 miles; and for July this level is 40 miles. However, as
.exhibited by the October and April data, this type of interpretation
may not truly represent the overall visibility quality. In order to
do this some sort of integral evaluation is needed over all levels of
visibilities. This can be achieved by essentially integrating each of
the curves shown in Figure 26. by taking the mid—point value of percent
in each of five-mile increments from 15 to 50 miles. This mid—point
value is then multiplied by the increment, five miles, to obtain
the area under the curve; then all increments (seven) are
totaled to obtain the area under each curve. This area can then be
divided by 100 to give a convenient number index of visibility quality
for each seasonal month. When this is done the visibility quality index
for January is 12.40, for April is 18.9, for July is 24.95, and for
October is 17.95. These numbers indicate the large difference (a factor
of 2) in visibility quality between January and July, and the
similarity of April and October. When these relative rankings are
compared with the overall view average hourly data, shown in Figure 25,
it can be seen that this type of data would indicate that July, October,
April, and January was the proper ranking. This disagreement in
ranking is because the data shown in Figure 25 is for both clouds
present and cloud free conditions. Both types of analysis are
useful for establishing the baseline visibility quality. Further
analyses could be performed on each individual view.
30
-------
60 r
Fig. 9 Average Hourly Visibility for Three Views
50
CO
>
4J
•H
•H
CO
•r-l
cfl
4-1
c
O
t-l
o
EC
30
20
O View 1, Cloud Free
D View 2, Cloud Free
O View 3, Cloud Free
View 1, Clouds Present
View 2, Clouds Present
View 3, Clouds Present
10
10
11
12
13
14
TIME - Local Std. Time (Central)
Stanton, N.D., April 5-May 1, 1976
15
16
17
-------
Lo
S3
N
•r-l
M
o
EC
60
50
40
% 30
,-i
•t-i
co
•rl
>
20
10
Fig.10 Average Hourly Visibility for an Overall View
Cloud Free Conditions
Cloud Present Conditions
_L
10
TIME
11
12
13
14
15
16
17
Local Std. Time (Central)
Stanton, N.D. April 5-May 1, 1976
-------
60
Fig. 11 Average Daily Visibility for Three Views
OJ
c
o
N
•H
50
40
30
20
Mist, Rain
10
-J—4-
i i i i
5 10 15 20 25 30
DATE: April 5-30, May 1, 1976, Stanton, N.D.
O View 1, Cloud Free • View 1, Clouds Present
O View 2, Cloud Free g View 2, Clouds Present
O View 3, Cloud Free % View 3, Clouds Present
-------
60
u>
50
co 40
CO
•H
c
o
N
o
5C
30
20
Fig. 12 Average Daily Visibility for
an Overall View, Versus Day of the Month
1
1
>— Cloud's present conditions
O— Cloud-Free Conditions
T
Example
Max.
Mean
J_ Min.
Mist, Rain
10
I ,
I I I I I I I I J I i_l _J I I I I L
10
15
20
25
30
DATE: April 5-30, May 1, 1976
Stanton, N.D.
-------
60
FIG. 13 AVERAGE HOURLY VISIBILITY FOR THREE VIEWS
PQ
M
CO
z
o
IS]
M
g
ad
50
40
30
20
10
O VIEW 1, CLOUD FREE
D VIEW 2, CLOUD FREE
O VIEW 3, CLOUD FREE
_l I L
• VIEW 1, CLOUDS PRESENT
m VIEW 2, CLOUDS PRESENT
+ VIEW 3, CLOUDS PRESENT
_L
_L
10
11
12
13
14
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., JULY 12 - AUG 7, 1976
-------
60
FIG. 14AVERAGE HOURLY VISIBILITY FOR AN OVERALL VIEW
50
CO
H
a
40
£ 30
g
N!
20
CLOUDS PRESENT
O CLOUD FREE
10
10
11
12
13
14
15
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., JULY 12-AUG 7, 1976
-------
FIG. 15 AVERAGE DAILY VISIBILITY FOR THREE VIEWS
H
M
fj
M
PQ
M
CO
M
2;
o
M
§
eo r
50
30
10
O VIEW 1, CLOUD FREE
D VIEW 2, CLOUD FREE
O VIEW 3, CLOUD FREE
J I_J L
VIEW 1, CLOUD PRESENT
VIEW 2, CLOUD PRESENT
VIEW 3, CLOUD PRESENT
J L
JULY
12
17
22
27
AUG -1
DATE: JULY 12 - AUG 7, 1976, STANTON, N.D.
-------
60
00
50
w
H
3
M
M
CO
1
N
M
g
40
30
20
J_
O CLOUD FREE CONDITIONS
CLOUD PRESENT CONDITIONS
FIG.16 AVERAGE DAILY VISIBILITY
FOR AN OVERALL VIEW,
VERSUS DAY OF THE MONTH
10
I I I I
I I I
1 I I 1 I I I I
I ,
I I
JULY 17
17
22
27
AUG-1
DATE: JULY 12 - AUG ' 7, 1976, STANTON, N.D.
-------
60
u>
C/3
w
(J
I
EH
M
i-q
CO
£
O
EC
50
40
30
20
10
FIG. 17 AVERAGE HOURLY VISIBILITY
FOR THREE VIEWS
O View 1, Cloud Free
D View 2, Cloud Free
^ View 3, Cloud Free
I I
• View 1, Clouds Present
B View 2, Clouds Present
^ View 3, Clouds Presnet
J I I
10
11
12
13
14
TIME - LOCAL STD TIME (CENTRAL)
STANTON, N.D., OCT. 4 TO 30, 1976
15
16
-------
60
FIG. 18 AVERAGE HOURLY VISIBILITY FOR AN OVERALL
50
40
-0-
•P-
o
M
PQ
O
Nl
I
EC
30
20
Clouds Present
©
Cloud Free
10
10
11
12
13
14
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., OCT. 4 TO OCT. 30, 1976
-------
60
CO
w
H
3
M
PQ
M
CO
§
g
50
40
30
20
10
FIG. 19 AVERAGE DAILY VISIBILITY FOR
THREE VIEWS
I I I I I I I I I I I I I I
I I I I
OCT
14
19
24
29
o View 1, Cloud Free
D View 2, Cloud Free
O View 3, Cloud Free
DATE
• View 1, Clouds Present
• View 2, Clouds Present
A View 3, Clouds Present
STANTON, N.D., 1976
-------
H
M
hJ
H
PQ
M
CO
-
IS
§
SI
50
30
Clouds Present
Cloud Free
FIG. 20 AVERAGE DAILY VISIBILITY FOR AN OVERALL
VIEW, VERSUS DAY-Of THE MONTH
,10
OCT 4
14
19
24
29
DATES: October 4-30, 1976
STANTON, NORTH DAKOTA
-------
60
50=
w
PQ
M
CO
1 *°
O
LO
FIG. 21 AVERAGE HOURLY VISIBILITY FOR THREE VIEWS
Q View 1, Cloud Free ^ View 1, Clouds Present
Q View 2, Cloud Free ^ View 2, Clouds Present
° View 3, Cloud Free • View 3, Clouds Present
TIME - LOCAL STD TIME (CENTRAL)
STANTON, N.D., JAN 13 - FEE 9, 1977
-------
-p-
-p-
60 _
50
40
FIG. 22 AVERAGE HOURLY VISIBILITY FOR AN OVERALL VIEW
O Cloud Free
Clouds Present
-------
60
FIG. 23 AVERAGE DAILY VISIBILITY FOR THREE VIEWS
50
Ui
w
g
I
H
M
tJ
l-l
M
00
£
H
O
a
o
ffi
30
20
10-
I I I I
I I I I
I I I I
JAN 13 18
O View 1, Cloud Free
D View 2, Cloud Free
O View 3, Cloud Free
23 28 FEB 2
«
0 View 1, Clouds Present
• View 2, Clouds Present
• View 3, Clouds Present
12
STANTON, N.D.
1977
-------
60
FIG. 24 AVERAGE DAILY VISIBILITY FOR AN OVERALL VIEW
Cloud Free
50
w
PQ
H
W
£
40
20
,10
Clouds Present
I II I I I I I 1 I i .1 1 t I I I I I 1 I
J^l, I t ].... J......J....A.
JAN
23
28 FEE 2
DATE: JAN 13 - FEE 9, 1977, STANTON, N.D.
-------
60 t
CO
w
H
M
m
M
C/3
50
40
30
FIG. 25 AVERAGE HOURLY VISIBILITY FOR AN OVERALL
VIEW, FOR JULY, OCTOBER, APRIL, AND JANUARY,
STANTON, N.D., 1976-77
July
a
20
10
10
11
12
13
14
15
16
TIME OF DAY - L.S.T., STANTON, N.D. (CENTRAL)
-------
100
-p-
oo
July
80
60
40
FIG. 26
20 ._
10
VISIBILITY QUALITY INDEX
CURVES FOR APRIL, JULY, OCT
JAN, STANTON, N.D., 1976-77
I
15
20
25 30
OVERALL VISIBILITY-MILES
-------
5 . Atmospheric Vertical Optical Attenuation
5.1 Techniques/Instrumentation Used - The basic technique used to deter-
mine optical attenuation coefficients was that of solar photometers/
radiometry and Beer's/Lambert's law. This relationship between solar
photometer/radiometer measurements and the atmospheric optical atten-
uation coefficient is as follows:
M = MQe" (18)
where M is the meter reading/intensity of the direct (collimated)
solar beam at the bottom of the atmosphere, MQ is the meter reading/
intensity at the "top" of the atmosphere, m is the relative air mass,
and T is the optical attenuation coefficient. Hence, the optical
attenuation coefficient may be derived by
T = -(In M/M )/m (19)
where M is the measured quantity. The relative air mass may be cal
culated by secG , where 9 is the solar zenith angle (see Figure 2),
J o o
for zenith angles from 0 to 60 (air masses of 1.00 to 2.00). For
greater zenith angles, atmospheric refraction has to be considered by
the following expressions:
If 2.0 < sec9 < 3.0
o
i f) /• r
m = secG 3.93 x 10" sec0o ' p/pQ (20)
If 3.0 < secG < 15.0
o
m = secG - 8.6 x 10"4 secG 2t% p/p (21)
o
where p = site pressure, p = sea level pressure
49
-------
The solar zenith angle may be calculated by the well known equation:
cosG = sin0sin 6 + cos0 cos 5 cos(t) (22)
o
where 0 is the site latitude, 6 is the solar declination, and t is
the solar hour angle. A computer program incorporating all these
parameters was used to calculate 9 for any given time of day at
Stanton, N.D. Obviously, the quantity M cannot be determined
directly. Instead, M is determined by performing measurements of
M versus air mass (times of day), and using this data to extrapolate
to zero air mass to determine M , as illustrated in Figure 27.
The optical attenuation coefficient given in Equation No. 18 is
defined as the total optical attenuation coefficient,T . It is
made up of optical attenuation coefficients due to molecular scatter-
ing, T , aerosol scattering, -7- , and gaseous absorption, T , •
m a at)
Hence,
T= T + r + T (23)
m a ab
where T , is also made up of components due to water vapor, ozone,
a o
carbon dioxide, etc, depending upon the wavelength region being
considered.
The specific optical attenuation coefficients addressed for Stanton,
N.D., were the total and the aerosol coefficients. The total attenua-
tion coefficient was determined as discussed previously (Equation 18),
and the aerosol attenuation coefficient was determined by
Ta= T - Tm - T03 (24)
where T(0^) was determined by models given by Elterman* along with
Tm- The specific wavelengths considered were 0.380, 0.440, 0.500,
*Ref. 2: Elterman, 1970, "Vertical-attenuation Model with 8 Surface
Meteorological Ranges 2 to 13" AFCRL-70-0200, 31 ERP No. 318
50
-------
20
o
z
w
H
g 6
g
P-I
^~i
g 4
CO
FIG. 27 EXAMPLE OF PLOTS TO DETERMINE EXTRATERRESTRIAL
METER READING OF PHOTOMETER/RADIOMETER
10 •
M - meter readings at various air mass
.725 urn
M
M
.650
I
2 3
RELATIVE AIR MASS
51
-------
0.640, and 0.880 /im. These wavelengths are those of the 2 channel
and 5 channel Volz photometers (see Figure 28). The 2 channel Volz
photometer wavelengths are 0.380 and 0.500 /im. The 5 channel Volz
photometer wavelengths are 0.440, 0.500, 0.640, 0.880, and 0.940 um.
The f.i£th channel of Volz 5 channel photometer (0.940 urn).is the
senter of a water vapor absorption band and is used, as discussed
later, to derive total atmospheric water vapor content. The M
values for both Volz photometers were determined by the manufacturer.
and by the Atmospheric Sciences Laboratory of Martin Marietta Denver
Division.
Figure 28 Two and Five Channel Volz Photometers
52
-------
In addition to the Volz spectral photometry determinations of the
optical attenuation coefficients, a broadband absolute radiometer
was used to measure the 0.30 to 2.8 ^im region. This radiometer
(see Figure 29) is manufactured by Eppley labs and has a 5.5 field
of view. It was used to measure the direct solar beam. The field
setup of the instruments discussed above is shown in Figure 30.
COLLIMATOR FOR SPECTRAL
RADIOMETER
Figure 29 Eppley Pyrheliometers and Collimator
for Spectral Radiometer
53
-------
VOLZ PHOTOMETERS
Figure 30 Field Setup of Total Data Collection System
Because absolute solar intensities are used in this case, the M
o
and M terms given in Equation No. 18 are replaced by I and I;
where I is the absolute extraterrestral solar irradiance ("Solar
constant"), sid I is the absolute measured solar irradiance.
The baseline vertical atmospheric optical attenuation coefficients
established were as follows:
(1) Daily average optical attenuation coefficients for the
wavelengths of 0.380, 0.440, 0.500, 0.640, 0.880, and
0.30 to 2.8 micrometers.
54
-------
(2) Daily average aerosol optical attenuation coefficients
versus wavelengths for the wavelengths of 0.380, 0.440,
0.500, 0.640 and 0.880 micrometers.
(3) Average hourly aerosol optical attenuation coefficient at
0.640 mircometers.
(4) Average hourly broadband - 0.30 to 2.8 micrometers -
optical attenuation coefficient.
5.2 April Optical Attenuation The optical attenuation results for April
are presented in Figures 31, 32, 33, 34 and 35. The daily average
total optical attenuation coefficients, shown in Figure 31, represent
three days - April 5, 6 and 8 - of cloud free condition, and two days
April 10 and 12 - of cloud present conditions. The remainder of the
data collection days were overcast, allowing no direct solar measure-
ments. Cloud free conditions represent days that had little or no
clouds in any part of the sky; whereas, cloud present conditions
represent days when there were clouds present in a significant portion
of the sky. However, all measurements were made when the solar disc
was not visibly obscured by clouds. As can be seen in Figure 31,
the various days exhibit somewhat different optical attenuation at
all wavelengths; and all wavelengths exhibit the same relative
attenuation for the various days. The fact that the 380 micrometer
channel has the greatest attenuation, with the 0.440, 0.500, 0.640
and 0.880 following, is due to the wavelength selective scattering
by molecules (Rayleigh) and aerosols (Mie). The 0.3 to 2.8 micrometer
band is a broadband that experiences attenuation due to both molecular
and aerosol scattering but also includes the well known strong
attenuation due to water vapor absorption in the near infrared wave-
lengths, 0.7 to 2.8 jum. Therefore, this band cannot be directly com-
pared to the narrow bands, where molecular and aerosol scattering
dominate. As discussed in section 5.1, the molecular scattering
55
-------
component may be subtracted, along with the ozone absorption component,
to yield the optical attenuation due to atmosphere aerosols.
This aerosol optical attenuation coefficient is most indicative of the
atmospheric clarity. The daily average aerosol optical attenuation
coefficient is shown in Figure 33. It can be seen that the cloud
present conditions result in a significantly higher attenuation,
especially in the shorter wavelengths. This higher attenuation is
most likely due to the presence of thin portions of clouds, especial-
ly cirrus clouds, that obscure the solar disc. The cloud free con-
ditions display similar aerosol optical attenuation, on the order
of 0.11. The average hourly aerosol attenuation at 0.640 ^itn is
shown in Figure 34. No readily apparent hourly dependence exists,
with a representative value of about 0.11. The average hourly
broadband optical attenuation is shown in Figure 35. This para-
meter displays a slight hourly dependence, with the maximum occurring
near the middle of the day. The values range from 0.22 to 0.34.
5.3 July Optical Attenuation - The July optical attenuation results are
shown in Figures 36, 37, 38, 39 and 40. As compared to April, July had
a greater number of sunny days which allowed many more determinations of
optical coefficients, shown in Figure 36, which show some day-to-day
variations for all wavelengths. The designation of partly cloudy
signifies a day where clouds were present but did not obscure the
solar disc. The designation of cloudy signifies a day where clouds
were present and did, at times, obscure the solar disc. The greatest
optical attenuations for clear days occurred on July 12 and Aug. 7.
The daily average aerosol optical attenuation coefficient results
shown in Figure 37, show noticeable daily fluctuations. For example,
at the .380 |im wavelength, the clear days exhibit variations between
56
-------
0.30 to 0.18. In addition, it can be seen that the various wave-
lengths do not necessarily "track" each other from day-to-day. The
reason for this is shown in Figure 38. The wavelength dependence
of the aerosol optical attenuation coefficient, as shown in Figure
38, exhibits changes from day-to-day. For example, the Aug. 7 data
indicates what is known as a "negative" slope wavelength dependence.
This type of wavelength dependence is caused by a dominance of
small particles. The July 17 results indicate a "positive" slope
wavelength response, which is indicative of a dominance of larger
particles. From Figure 38, it can be seen that July 16, 17, 23 and
26 were similar days having low optical attenuation and a dominance
of larger particles. On the other hand, Aug. 6 and 7 displayed
higher optical attenuation with a dominance of small particles.
The average hourly aerosol optical attenuation, for 0.640 pm,
results, shown in Figure 39, display a slight hourly dependence
with the maximum occurring from noon to afternoon. In addition,
large variations can occur for any given hour. For example, at
1600 hours, the variation can be from 0.105 to 0.230 (a factor of
2.2). The average hourly broadband optical attenuation coefficient
results, shown in Figure 40, display a similar hourly dependence,
with maximums occurring in the noon-afternoon period.
5.4 October Optical Attenuation The optical attenuation coefficient
results for October are shown in Figures 41, 42, 43 and 44. A total
of seven (7) days of measurements were acquired. Of these seven
days, no day was completely clear. Cirrus clouds were a very common
occurrence, however, as mentioned previously, measurements were made
only when the solar disc appeared to be free of any cloud obscuration.
The daily average total optical attenuation results, shown in Figure
41, display day-to-day variations. The degree of these variations
is dependent upon wavelength. For example, the 0.440 urn data show
57
-------
much more marked daily variations than the 0.880 jim data. The
reason for this is shown in Figure 42. As with the July data, it
can be seen that "negative" slope/small particle and "positive"
slope/large particle were present. As shown in Figure 42, this
results in large differences in aerosol optical attenuation at
the shorter wavelengths, and smaller differences at the longer
wavelengths. For example, at 0.440 jim there exists a difference
of 0.035 - 0.163 between Oct. 8 and Oct. 21, respectively; whereas,
for 0.880 jim, there exists a difference of 0.07 - 0.11. This
points out the need for measuring the aerosol optical attenuation
at several wavelengths, over a range from short to long, to ade-
quately define the attenuation properties. The average hourly
aerosol optical attenuation coefficient, at 0.640 jim, results
are shown in Figure 43. These results indicate a very minimal
hourly dependence for aerosol attenuation. However, as with the
July results, any particular hour may have a wide range of atten-
uation. Similar results are shown, Figure 44, for the average
hourly broadband attenuation coefficient.
5.5 January Optical Attenuation - The January optical attenuation re-
sults are shown in Figures 45, 46, 47, 48 and 49. The daily average
total optical attenuation, Figure 45, and the daily average aerosol
optical attenuation, Figure 46, display daily fluctuation. This
is especially true for the days of January 17 and 19. As can be seen,
January 17 was a much more turbid day as compared to January 19. The
daily average aerosol attenuation coefficient versus wavelength
results, shown in Figure 47, show that all days except the 17th had
a "positive" slope/large particle characteristic. The 17th was
much more turbid and had a "negative" slope/small particle character.
The average hourly aerosol optical attenuation coefficient, at
0.640 pm, results, Figure 48, show little hourly dependence but wide
58
-------
ranges of values for any given hour. The corresponding broadband
results, Figure 49, show a slight decrease toward the late afternoon.
5.6 Seasonal Optical Attenuation Comparisons - The seasonal comparisons
consist of the broadband optical attenuation coefficient, Figure 50,
and the aerosol optical attenuation coefficient at 0.640 ^im, Figure
51. As pointed out previously, the aerosol optical attenuation co-
efficient is dependent upon wavelength, with the greatest sensitivity
at the shorter wavelengths. The 0.640 ^jm wavelength was chosen to
be compromise between the short and long wavelengths. The average
hourly broadband optical attenuations coefficient, Figure 50, shows
that July had the highest vertical attenuation, followed by April,
October, and January having the lowest attenuation. The 0.640^im
results, Figure 51, also display this seasonal ranking of vertical
attenuation. The range between July and January, for both the broad-
band and 0.640^im wavelengths, is quite large. For the broadband
wavelength, July displays a typical value of 0.37 (at noon), compared
to a value of 0.16 (at noon) for January. For the 0.640 urn band,
July displays a typical value of 0.155 (at noon), compared to a value
of 0.04 (at noon) for January. A comparison of the wavelength de-
pendence of the aerosol optical attenuation coefficient at all months
reveals that when conditions are such that the aerosol optical atten-
uation is low, "positive" slope/large particle distribution are also
present. This indicates that small particles cause the overall
attenuation to fluctuate from day-to-day. When the aerosol attenua-
tion is high, the wavelength dependence of the attenuation indicates a
"negative" slope/small particle characteristic. The various days are
quite markedly defined by their wavelength dependence of the aerosol
attenuation coefficient.
59
-------
1.00
o\
o
C
Ol
4-1
U-l
g
•1-1
4-1
14
1
4J
cd
u
•1-1
4J
a
o
o
H
0.80 _
0.60 _
0.40
0.20
Fig. 31 Daily Average Total Optical Attenuation Coefficients
D
O
A
O
.500/<.m
. 640 jim
. 880/un
- .3 TO 2.8^
i,38C
Clouds Present
Cloud Free
©- -
A^~
O-
— —A,5QC
— — -A - - —
Date: April, 1976
Stanton, N.D.
10
11
12
-------
0.50
0.40
4-1
c
•i-l
o
01
u 0.30
c
o
•H
01
0.20
CO
CJ
•H
i-J
o.
o
2 o.io
o
t-i
0^
380
440
/
/^
/>
88
10
11
12
Date: April, 1976
Stanton, N.D.
-------
N5
c
,
-------
4-1
CL
O
o
03
O
Fig. 34 Average Hourly Aerosol Optical Attenuation
Coefficient for .
0.20
Max.
C
0)
s
01
4J
Majc.
0.10
Mean
Min.
10
11
12
13
14
15
16
17
TIME - Local STD. Time (Central)
Stanton, N.D., April 5-May 1, 1976
-------
0.50
Fig. 35 Average Hourly Broadband (.3 to Z.S^u-m)
Optical Attenuation Coefficient
0.40
4-1
c
01
•H
O
-H
<4-4
y-i
-------
1.0
Ui
.8
w
o
u
£3
w
H
H
H
P-i
O
H
O
H
.6
,
.4
.2
FIG. 36 DAILY AVERAGE TOTAL OPTICAL ATTENUATION COEFFICIENTS
PARTLY CLOUDY
CLEAR / CLOUDY
.380
I I I I I I I I I I I I I L I I I I
1
JULY 12
17
22
27
AUG. 1
DATE: JULY 12 - AUG. 17, 1976 STANTON, N.D.
-------
.5
£
W
.4
fe
w
o
la
P .3
W
H
H
O
l-l
H
O
W
O
.1 -
FIG.37 DAILY AVERAGE AEROSOL OPTICAL ATTENUATION COEFFICIENT
I I I I
lilt
I
till
I i i
jl
I i
JULY 12
17 22 27 AUG 1
DATE - JULY 12 - AUG. 7, 1976, STANTON, N.D.
-------
.5
H
Z
w 4
W
O
O
2
"
S
I
H
O
CO
§
.1
FIG. 38 DAILY AVERAGE AEROSOL OPTICAL ATTENUATION COEFFICIENT VERSUS
WAVELENGTH
CLOUD FREE DAYS
O JULY 17
STANTON, N.D., JULY 12 - AUG 17, 1976
.7
.8
.9
l.n
WAVELENGTH - MICROMETERS,
-------
00
FIG. 39
AVERAGE HOURLY AEROSOL OPTICAL ATTENUATION
COEFFICIENT FOR 0.640 pm
10
11
12
13
14
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., JULY 12 - AUG. 7, 1976
-------
.5 r
w
M
U
H
fn
Pn
W
O
W
H
H
.4 -
°, .2
FIG.40 AVERAGE HOURLY BROADBAND (.3 TO 2.8pm)
OPTICAL ATTENUATION COEFFICIENT
O
1
1
pa
10
11
12
13
14
15
16
TIME - LOCAL STD. TIME (CENTRA^)
STANTON., N.D., JULY 12 - AUG 7, 1976
-------
1.0
H
3
W
• .380 Micrometers
O .440 Micrometers
• .500 Micrometers
D .640 Micrometers
A .880 Micrometers
A .3 to 2.8 Micrometers
.41 DAILY AVERAGE OPTICAL
ATTENUATION COEFFICIENTS
PH
§
u
o
H
H
W
H
H
H
P-i
O
.4
•
.y
--- Q
1 1 1 1 1 1 1 1 1 1 1
i i i
oci 4
14 19 24
DATE: STANTON, N.D., 1976
-------
25 i
H
Z
W
M
U
w
o
CJ>
a
o
w
H
H
H
PM
O
O
CO
O
O 1Q/5
• 10/8
D 10/9
• 10/11
10/21
10/29
20 _
FIG. 42
DAILY AVERAGE AEROSOL OPTICAL ATTENUATION
COEFFICIENT VERSUS WAVELENGTH
,15
,10
05
I
.5
.6 .7 .8
WAVELENGTH - MICROMETERS,
1.0
-------
O
C/3
O
at
w
FIG. 43
AVERAGE HOURLY AEROSOL OPTICAL
'.ATTENUATION COEFFICIENT AT 0.640yUm
0.20 _
w
M
U
P-i
§
0.15 -
0.10 _
0.05 ..
TIME - LOCAL STD TIME (CENTRAL)
STANTON, N.D., OCT. 4 TO OCT. 3Q, 1976
-------
0.50
•vj
ut
44
AVERAGE HOURLY BROADBAND :(0.3 TO 2.8 /am)
OPTICAL ATTENUATION COEFFICIENT
TIME - LOCAL STD. TIME (CENTRAL) STANTON, N.D.,
OCT. 4 TO OCT 39, 1976
-------
1.0
FIG. 45 DAILY AVERAGE TOTAL OPTICAL ATTENUATION COEFFICIENTS
W
h-1
0
—i
-P-
w
o
u
o
M
H
w
H
H
.4
.CLEAR
PARTLY CLOUDY
I— .380
B
p-l
o
-------
.10
W
M
0
§
u
2
O
<
,-J
H
PH
O
23
.08
.06
.04
.02
FIG. 46 DAILY AVERAGE AEROSOL OPTICAL
ATTENUATION COEFFICIENT
PARTLY CLOUDY
— .800
.640 //m
. 500 fjm
.380 ^tn
.440^* in
I I
J 1
1 1 ,
I I 1 I 1 i
I 1
JAN 13 18 23 28 FEB 2
DATE - JAN 13 - FEB 9, 1977, STANTON, ND
-------
.10
FIG. 47 DAILY AVERAGE AEROSOL ATTENUATION
COEFFICIENT VERSUS WAVELENGTH
W
M
CJ
M
fe
En
W
O
o
§
M
I
.08
CLOUD FREE
DAYS
.06
1 JAN 17
3
o
§
sa
,04
.02
JAN 28
I
JAN 22
JAN 19
STANTON, ND
JAN 13 - FEE 9, 1977
I
.4
.5 .6 .7
WAVELENGTH - MICROMETERS,
.9
-------
H
Z
w
M
QJ
ft,
W
O
CJ
2
O
FIG. 48 AVERAGE HOURLY AEROSOL OPTICAL ATTENUATION
COEFFICIENT FOR CLEAR DAYS AT 0.640 u m
.200
.150
0
M
.100
o
w
o
.050
10
11
12
13
TIME - LOCAL STANDARD (CENTRAL)
STANTON, ND, JAN 13 - FEE 9, 1977
14
15
16
-------
FIG. 49 AVERAGE HOURLY BROADBAND (.3 - 2.8/tm)
OPTICAL ATTENUATION COEFFICIENT
oo
w .200
u
W
O
M .150
H
H
H
M
H
Q
1
§
CQ
.100
.050
10
11
12
13
14
15
16
Tim - LOCAL STANDARD (CENTRAL)
STANTON, ND, JAN 13 - FEE 9, 1977
-------
.5 ~
H
Z
W
M
u
M
fa
fa
H
H
g
£
.4 -
Q —
,2 -
.1 ~
FIG. 50 COMPARISON OF HOURLY AVERAGE BROADBAND (.3 to 2.8 /1m)
ATTENUATION FOR APRIL, JULY, OCT., JAN, STANTON, N.D.,
1976-77
10 11 12 13 14
TIME OF DAY - STANTON, N.D., 1976-77, L.S.T. (CENTRAL)
15
16
-------
FIG. 51 COMPAISON OF HOURLY AVERAGE NARROW-BAND (0.640 jam)
ATTENUATION FOR APRIL, JULY, OCT, JAN, STANTON, N.D.,1976-77
H
23
W
-.200
oo
o
w
o
u
o
M
H
w
H
H
H
H
PU
O
July
. 15'0
. l,QO
April
•
Oct
I
I
10
TIME OF DAY
11 12 13 14
STANTON, N.D., 1976-77, L.S.T. (CENTRAL)
15
16
-------
6. INCIDENT SOLAR IRRADIANCE
6.1 Techniques/Instrumentation Used - The absolute quantities of solar ir-
radiance measured consisted of the normal incident direct solar beam,
I, and the total (180°, global) incident solar irradiance on a horizon-
tal surface, H. The wavelength region measured encompasses 99% of the
available solar energy and is the 0.3 to 2.8 urn broadband region. The
two quantities measured are related by
H = I cos9 + S (25)
where S is the diffuse sky irradiance.
The specific baseline parameters established were the absolute quanti-
ties of direct, total, and diffuse solar irradiance versus relative air
mass, and the ratio of the diffuse to total solar irradiance. For a
given solar zenith angle, 9O, the relative air mass, m, these parameters
are determined by the atmospheric optical air quality. In the case of
the diffuse irradiance some dependence exists concerning ground albedo;
but, for a given season/condition, this dependence should be relatively
constant.
The instrumentation used consisted of an Eppley pyrheliometer (see Fig-
ure 29) for measuring the direct solar beam and an Eppley pyranometer
for measuring the total solar irradiance. The pyranometers are shown
in Figure 52. The field location of the pyranometers is shown in Figure
30.
The baseline absolute solar irradiance parameters established were as
follows:
(1) Broadband—0.30 to 2.8 micrometer - direct solar irradiance
versus relative air mass.
(2) Diffuse broadband solar irradiance versus relative air mass.
(3) Total Broadband horizontal solar irradiance versus relative
air mass.
81
-------
HEMISPHERICAL IR FILTER
HEMISPHERICAL WINDOW
Figure 52 Eppley Pyranometers
(4) Ratio of broadband diffuse to total solar irradiance.
(5) Average daily ratio of broadband diffuse to total solar ir-
radiance.
(6) Average hourly broadband ratio of diffuse to total solar ir-
radiance.
6.2 April Solar Irradiance - The results for the April solar irradiance
are shown in Figures 53, 54, 55, 56, 57 and 58. The results for the
direct solar beam irradiance versus relative air mass, Figure 53, dis-
play a baseline for cloud free and cloud present conditions. The clar-
ity of the atmosphere, optical attenuation for the .3 to 2.8 jura band,
controls this baseline. Therefore this baseline is indicative of ver-
tical atmospheric optical air quality. As can be seen, the presence
of cirrus clouds significantly reduce the baseline. This is indicative
of the sensitivity of the measurements as compared to the human eye
82
-------
observations, because the ground-based observer could not detect any
obscuration of the solar disc. Shown in Figure 54 is the baseline for
the diffuse sky, solar irradiance versus relative air mass. Again,
baselines for cloud present and cloud free conditions were evident.
Note that the diffuse sky irradiance is significantly lower for the
clear conditions because of less scattering of the direct solar beam.
The results for the total solar irradiance versus relative air mass,
shown in Figure 55, show the interesting phenomenon of a single total
versus relative air mass baseline despite the atmospheric condition.
This can be explained by recalling that although the cloud conditions
reduced the direct solar beam, they increased the diffuse sky component.
This evidently results in the constant baseline shown in Figure 55.
Such a baseline is important because it defines the solar heat input
to the ground, under the natural conditions. These natural conditions
are defined by the characteristics of the aerosols and cloud particles.
If these particles significantly absorbed the solar irradiance, instead
of scattering into the diffuse sky component, the diffuse irradiance
component would not raise the total irradiance to the level shown for
the natural baseline, Figure 55. Hence, if any man-made aerosols that
have a low surface albedo are injected into the atmosphere, the total
solar irradiance versus relative air mass baseline would be altered.
The baseline would most likely shift downward, indicating a reduction
in total solar irradiance, at a given relative air mass. Shown in Fig-
ure 56 is the ratio of diffuse to total solar irradiance versus rela-
tive air mass. As can be seen, this baseline is also indicative of
clear and cloud present conditions. The average daily ratio of diffuse
to total solar irradiance, shown in Figure 57, displays variations be-
tween the clear (April 5,6,8) and cloud present days (April 10 and 12).
The average hourly broadband ratio of diffuse to total solar irradiance,
shown in Figure 58, displays a dependence upon time of day, having a mini-
mum near noon and maximums in the early morning and late afternoon. This
is due to the dependence on relative air mass, shown in Figure 56. At
83
-------
low relative air masses the ratio of diffuse to total solar irradiance
is a minimum because the path length, traveled by the solar beam, is a
minimum, and the total solar irradiance is a maximum. The shorter path
lengths result in a lower amount of diffuse sky irradiance because of
a lower amount of direct solar irradiance scattered from the beam. The
lower relative air masses correspond to higher total solar irradiance
because they are indicative of higher sun angles (elevation) which re-
sult in a greater amount of direct solar irradiance on the horizontal
surface.
6.3 July Solar Irradiance - The July solar irradiance results are shown in
Figures 59, 60, 61, 62, 63 and 64. The broadband direct solar irradi-
ance versus relative air mass, Figure 59, distinctly identify very
clear conditions, July 16, hazy conditions on August 6 and extremely
hazy conditions for August 7. The marked effects of haze on the dir-
ect solar irradiance versus relative air mass is obvious. This points
out the usefulness of the direct solar irradiance versus relative air
mass as a quantitative identification of vertical atmospheric optical
clarity/quality. For example, if one considers a relative air mass of
o
2.0; the clear conditions have a direct irradiance of 850 w/m ; the
hazy conditions have a value of 780 w/m^; and the very hazy conditions
fy
have a value of 660 w/m . It should be pointed'out that the broadband,
.3-2.8 ^/m, direct solar irradiance is also sensitive to attenuation by
water vapor absorption in the near infrared region. However if one
compares the results shown in Figure 36, it can be seen that both the
broadband and narrowband results point out the significantly higher attenua
tion resulting from the hazy conditions on Aug. 6 and 7. The diffuse solar
irradiance relative air mass results, Figure 60, show the significantly
higher amounts of diffuse sky irradiance caused by the hazy conditions,
especially at low values of relative air mass. For example, at a rela-
tive air mass of 1.15, the hazy conditions have a diffuse irradiance
of about 105 w/m2 whereas the clear conditions have values ranging from
84
-------
64 to 80 w/m2. The broadband total solar irradiance versus relative
air mass, shown in Figure 61, results show a very well defined curve,
with no separation of clear versus hazy conditions. This result is
identical to the April results, again pointing out that the reduction
of the direct solar beam is countered by an increase in diffuse irradi-
ance. This results in the constant total solar irradiance versus rela-
tive air mass. The ratio of diffuse to total solar irradiance results,
shown in Figure 62, also point out the differences between clear and
hazy conditions. The average daily ratio of diffuse to total solar
irradiance, shown in Figure 63, shows the relative clarity of the var-
ious days. The cloud present conditions have higher ratios of diffuse
to total solar irradiance because of the large contribution of the
clouds to the diffuse sky irradiance. The various ratios for cloudy
days is an indicator of the amount of cloudiness, and not the amount
of haze. For clear conditions, the relative clarity/haziness of the
days is easily established. The days of July 16, 17, 23 and 26 were
days having approximately the same clarity; whereas, Aug. 6 and 7 had
much higher ratios indicating more haze. The average hourly broadband
ratio of diffuse to total solar irradiance, shown in Figure 64, also
shows the differences between clear and hazy conditions.
6.4 October Solar Irradiance - The October solar irradiance results are
given in Figures 65, 66, 67, 68, 69 and 70. The month of October had
a high occurrence of cirrus clouds, making the solar irradiance data
very scattered and uncertain concerning the establishing of clear con-
ditions. The broadband direct solar irradiance versus relative air
mass results, shown in Figure 65, show a high degree of scattering of
data points. The ground observers reported no apparent hazy conditions,
as with the July data; however, they frequently reported high cirrus
clouds. Therefore, the range of values shown in Figure 65 (and the
remainder of the Oct. results) are most likely due to changing cirrus
conditions. In order to establish the clear conditions one would have
85
-------
to assume that the highest values of direct irradiance represent
clear conditions. The diffuse solar irradiance versus relative air
mass results, shown in Figure 66, also show a high degree of scatter,
probably caused by cirrus clouds. Again, the clear baseline is estab-
lished by assuming no cirrus influence. The total broadband horizontal
solar irradiance versus relative air mass results, Figure 67, again
show a fairly constant relationship despite the influence of cirrus
clouds. As pointed out previously, there seems to exist a natural
mechanism where the total solar irradiance as a function of relative
air mass is kept the same by increases in the diffuse sky irradiance
whenever the direct is attenuated by thin clouds and/or haze. The
ratio of diffuse to total solar irradiance versus relative air mass,
Figure 68, showsa range of values due to probable cirrus influences.
The clear baseline is assumed to be representative. The average daily
ratio of diffuse to total solar irradiance results, Figure 69, show
the relative clarity of the days. October 5 was the clearest day and
October 29 was the least clear day. The influence of cirrus clouds is
probable, but not known exactly. It is possible that October 5 repre-
sents a true clear day. The average hourly broadband ratio of diffuse
to total solar irradiance results are shown in Figure 70.
6.5 January Solar Irradiance - The January solar irradiance results are
shown in Figures 71, 72, 73, 74, 75 and 76. During January the ground
was snow covered. Because snow has such a high albedo, it effects the
diffuse sky irradiance by reflecting the direct solar irradiance, the
reflected direct gets backscattered by the atmosphere. This results
in higher values for the diffuse irradiance. The broadband direct
solar irradiance versus relative air mass, Figure 71, shows the clear
baseline and the influence of cirrus clouds. The diffuse solar irradi-
ance versus relative air mass results, shown in Figure 72, show a marked
difference between clear and haze/cirrus conditions. This difference
is amplified by the ground snow cover, because of the strong backscatter
86
-------
of the reflected direct solar irradiance by the haze and/or cirrus
clouds. The total solar irradiance versus relative air mass, Figure
73, also show the impact of cirrus clouds. For all the other months
given previously, the total solar irradiance versus relative air mass
has been a well defined relationship. The deviation for January is
most likely caused by the combination of the high reflectance snow
cover reflecting the direct solar beam up, then the thin cirrus clouds
reflecting it back down toward the ground. This results in a very
high value for the diffuse irradiance, as shown in Figure 72. The
clear baseline shown in Figure 73, is assumed to be free of cirrus
cloud influences. The ratio of diffuse to total solar irradiance
versus relative air mass, Figure 74, also shows the influence of cir-
rus clouds. The average daily ratio of diffuse to total solar irradi-
ance results, Figure 75, show the influences of haze, cirrus, and clear
conditions. In particular, Jan. 17 was reported to be very hazy by
the ground observer. The average hourly ratio of diffuse to total
solar irradiance results, Figure 76, also show the cirrus versus clear
influences.
.6 Seasonal Solar Irradiance Comparisons - The seasonal comparisons of the
solar irradiance results are shown in Figures 77, 78 and 79. The analy-
sis shown in Figure 77 is the comparison of the direct solar irradiance
versus relative air mass, for each of the seasonal months. For a given
relative air mass, the seasonal month that has the highest level of
direct solar irradiance has the lowest atmospheric attenuation. It
is obvious that there exists a high degree of variation of atmospheric
attenuation for the various seasons/months. The analysis shown in
Figure 77, indicates that January had the lowest atmospheric attenua-
tion, followed by October, July-, and April. It should be pointed out
that these curves, Figure 77, represent a sort of hypothetical best
clear day for each season/month because (as previously shown in Fig-
ure 53, 59, 65 and 71) these clear baselines were derived from using
87
-------
the highest values of direct solar irradiance at given relative air
masses for all days. Therefore, the curves (clear baselines) shown in
Figure 77, represent a composite of the clearest conditions existing
for the entire month. When these results are compared to the results
for optical attenuation coefficients, shown in Figures 50 and 51, it
can be seen that the ranking of the months is different. The optical
attenuation coefficient results indicate that January was the clearest
month, followed by October, April, and July. The reason for this is
due to the fact that the optical attenuation results were obtained by
averaging all values for clear day attenuation; whereas, the results
shown in Figure 77 were obtained by the method described above. The
results shown in Figure 77 simply indicate that the best case clear
conditions in April were higher in attenuation than the best case
clear conditions in July. The results shown in Figures 50 and 51
simply indicate that on the average the conditions during April had
lower attenuation than July. Both sets of analyses are useful for es-
tablishing the clear day baseline attenuation; because one set of cur-
ves, Figure 77, represents the absolute clearest conditions observed
over all days, and the other (Figures 50 and 51) represents the aver-
age clear conditions observed. Shown in Figure 78 is the composite
of the seasonal months - total horizontal solar irradiance versus re-
lative air mass. As previously discussed, the total solar irradiance
versus relative air mass is a surprisingly consistent function despite
the varying atmospheric conditions. The January results display a dif-
ference baseline because of the combination of snow cover and cirrus
clouds. The monthly-seasonal diffuse to total ratio versus relative
air mass analysis is shown in Figure 79. The ratio of diffuse to total
solar irradiance is an indicator of overall atmospheric clarity because
as the atmosphere becomes more turbid the diffuse sky irradiance will
be altered and generally increased. Therefore, as the ratio increased
the overall atmospheric clarity would be diminished. The general in-
crease of the ratio with relative air mass is, as discussed previously,
88
-------
due to the total decreases and the diffuse increases because of sun
angle effects. The diffuse to total ratio results shown in Figure 79
represent the hypothetical clearest conditions as discussed for the
direct solar irradiance versus relative air mass results.
It should be recalled that the diffuse sky irradiance, and therefore
the ratio of diffuse to total irradiance, is influenced by the ground
albedo and any cloud cover present. Since the change in ground albedo
for the various months - except January which was snow covered - is
unknown, the actual magnitudes of the ratio versus relative air mass
cannot be directly interpreted as an indicator of atmospheric clarity.
The curves shown in Figure 79 represent the hypothetical clearest
baseline for each of the seasonal months, with the ground albedo of
that period of time. However, it is interesting to consider the slope
of the ratio of diffuse to total versus relative air mass. This slope
is somewhat indicative of the atmospheric optical attenuation coeffic-
ient. For example, a very clear atmosphere would have a low diffuse
sky irradiance and high relative air masses (slant paths) would be re-
quired to substantially increase the diffuse sky irradiance. Hence, a
very clear atmosphere would be associated with a low slope of the type
of curves shown in Figure 79. A very turbid atmosphere would have a
high slope because of the higher concentrations of aerosols causing
much greater scattering as the direct solar beam path length (relative
air mass) is increased. The actual level/magnitude of the line (dif-
fuse to total ratio versus relative air mass) would be determined by
the ground albedo. When the slopes of the curves shown in Figure 79
are calculated one obtains the following results.
Table 1. Comparison of seasonal slope of diffuse to total ratio versus
relative air mass.
Month
April
July
Oct.
Jan.
89
-------
As the resul;s in Table 1 show, April was the most turbid month followed
by July, October, and January was the clearest month. These results
agree exactly with those obtained by comparing the direct solar irradi-
ance versus relative air mass, Figure 77. The curve shown in Figure 80
verifies that the slope of the ratio of diffuse to total solar irradiance
versus; relative air mass is proportional to the atmospheric attenuation
coefficient. The linear regression correlation coefficient for the plot
shown in Figure 80 is 0.97, indicating a very good correlation. The
optical attenuation coefficient, in Figure 80, was calculated by using
the plots shown in Figure 77. As previously shown, the optical attenua-
tion coefficient, T, is given by
In I/I
r = °- (26)
m
where I is the direct solar irradiance at a given relative air mass, m,
and I0 is the extraterrestrial solar irradiance. For each month, a value
for I, at a relative air mass of 3.0, was taken from the dashed (cor-
rected to mean sun-earth distance) lines shown in Figure 77. The value
for the solar constant, Io, was taken to be 1352 W/M2. This resulted
in a broadband optical attenuation coefficient for the hypothetical
clearest day for each of the seasonal months, as plotted in Figure 80.
90
-------
1000 r
\0
01
4-1
.U
a
n)
-------
250
S3
CO
4-1
4J
fl)
s
-------
1000
vD
01
u
e
cfl
•H
-0
ca
M
(-1
ca
o
C/J
c
a
M
O
Ed
O
H
800 ~
600
400
200
X
n
A
Fig. 55
Broadband (.3 to 2.8 fim) Total Horizontal
Solar Irradiance Versus Relative Air Mass,
For Cloud Free and Cloud Present Conditions,
April 5, Cloud Free
April 6, Cloud Free
April 8, Cloud Free
April 10, Cloud Present
April 12, Cloud Present (Cirrus)
Relative Air Mass
Stanton, N.D. - April 5 to May 1, 1976
-------
QJ
O
a
•H
T3
nJ
1-1
t-l
M
O
CO
tH
4-1
O
0)
co
3
•1-1
p
4-1
O
o
•H
4-1
0.30
Fig. 56 Ratio of Diffuse to Total Broadband (.3-2.8 pn) Solar Irradiance
Versus Relative Air Mass
Baseline for Cloud
Present Conditic ^_
\ S •~^^
Max imum
0.20
o.io
Baseline for Cloud Free
Conditions
Minimum
April 5, Cloud Free Conditions
April 6, Cloud Free Conditions
Q April 8, Cloud Free Conditions
A April 10, Cloud Present Conditions
O April 12, Cloud Present Conditions
1
Relative Air Mass
Stanton, ND, April 5 to May 1, 1976
-------
Fig. 57 Average Daily Ratio of Diffuse To Total Solar Irradiance
0.30
T
at
CJ
C
3)
•H
ID
td
l-l
l-l
M
O
t/3
O
4J
0>
CO
3
0.20
Max.
Mean
O
•H
Min. _[_
0.10
10
11
12
Date: April, 1976
Stanton, N.D.
-------
Ol
o
C
-a
nJ
J-l
i-l
H
0.30
- Ma*
\
Fig. 58 Average Hourly Broadband (.3-2.8 urn)
Ratio of Diffuse to Total Solar Irradiance
VO
CTv
o
CO
o
H
O
4-J
01
CO
O
o
•H
Mean
0.20
Cloud Free
Conditions
Cloud Present Conditions
J-
0.10
10
11
12
13
14
15
16
17
TIME - Local Std. Time (Central)
Stanton, N.D., April 5 - May 1, 1976
-------
1000
OS
w
H
to
H
H
o
en
900
800
700
600
FIG. 59 BROADBAND (.3 TO 2.8 pm) DIRECT SOLAR
IRRADIANCE VERSUS RELATIVE AIR MASS FOR
CLOUD FREE CONDITIONS*
VERY CLEAR CONDITIONS
JULY 16
*DATA POINTS FOR
JULY 16,17,23,26,AUG 6, 7
HAZY CONDITIONS, AUG 6
AUG 7,PM
500
T7o~
RELATIVE AIR MASS
STANTON, N.D., JULY 12 TO AUG 7, 1976
-------
120
VO
00
w
H
:ioo
CO
H
H
w
a
80
o
CO
w 60
CO
p
Pn
H
Q
40 -
20
FIG.60 DIFFUSE SOLAR IRRADIANCE VERSUS RELATIVE
AIR MASS FOR CLOUD FREE CONDITIONS
HAZY CONDITIONS, AUG. 6 AND 7
JULY
7/16,17,23,
26
CLEAR CON-.
ditions
2.0
RELATIVE AIR MASS
STANTON, N.D., JULY 12 TO AUG. 7, 1976
3.0
-------
1000 -
w
H
H
H
w
o
2
<
M
O
CO
h4
3
3
O
NJ
a
O
«
800 -
FIG.61 BROADBAND (.3 T002.8 urn) TOTAL
HORIZONTAL SOLAR IRRADIANCE VERSUS
AIR MASS FOR CLOUD FREE CONDITIONS
600
400
200
1.0
2.0
3.0
RELATIVE AIR MASS
STANTON, N.D., JULY 12
- AUG. 7, 1976
-------
O
O..
o .20
O
CO
H
O
H
O
H
W
CO
Pn
s
fa
O
O
1-1
H
.10
• •— HAZY CONDITIONS, AUG 6 AND AUG 7
O O — CLEAR CONDITIONS, JULY 16, 17, 23, AND 26
HAZY CONDITIONS
r. «
CLEAR CONDITIONS
FIG. 62 RATIO OF DIFFUSE TO TOTAL SOLAR
IRRADIANCE (.3 TO 2.8 jam) VERSUS
RELATIVE AIR MASS
RELATIVE AIR MASS
STANTON, N.D., JULY 12 TO AUG 7, 1976
-------
.20
w
o
CO
H
O
H
O
H
CO
fu
En
O
O
M
H
,10
,05
CLOUD PRESENT CONDITIONS
CLOUD FREE,
HAZY CON-
DITIONS
FIG. 63 AVERAGE BAILY RATIO OF DIFFUSE TO
' TOTAL SOLAR IRRADIANCE (.3 TO 2.8 jam)
CLOUD FREE, CLEAR CONDITIONS
I I I I I I I I I I
I I I I I
JULY
12
17
22
27
AUG 1
DATE: JULY 12 TO AUG /, 19/6
STANTON, N.D.
-------
o
NJ
H
O
H
O
H
W
CO
H
O
O
M
FIG. 64 AVERAGE HOURLY BROADBAND (.3 - 2.8 pm) RATIO OF
DIFFUSE TO TOTAL SOLAR IRRADIANCE
.20
CLEAR CONDITIONS
.9
10
11
12
13
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., JULY 12 TO AUG 7, 1976
-------
1000
B
H
3
i
W
900
800
clear baseline
700
H
O
a
600
FIG. 65 BR0ADB41SD '(.3-.10 2.f ^«X DIRECT
SOLAR IRRADIANCE VERSUS RELATIVE
AIR MASS
500
1.0
2.0 3.0
RELATIVE AIR MASS
STANTON, N.D., OCT. 4 TO OCT. 30, 1976
-------
120
h-
EIG. 66 DIFFUSE SOLAR IRRADIANCE VERSUS RELATIVE AIR MASS
100
OT
H
I
W
U
M
I
W
80
60
40
clear
baseiltte
20
2.B
3.0
4.'0
RELATIVE AIR MASS
STANTON, N.D., OCT. 4 TO OCT. 39, 1976
-------
o
Ul
H
a
o
NJ
M
cei
O
800
600
400
200
FIG. 67 BROADBAND (.3 TO 2.8;am) TOTAL HORIZONTAL
SOLAR IRRADIANCE VERSUS AIR MASS
1.0
2.0
3.0
4,9
RELATIVE AIR MASS
STANTON, N.D., OCT. 4 TO OCT. 30, 1976
-------
w
u
0.20
M
Q
fa
O
O
M
H
.05
FIG. 68 RATIO OF DIFFUSE TO TOTAL SOLAR
IRRADIANCE (.3 TO 2.8 ;im) VERSUS,
RELATIVE AIR MASS
o
o\
1
H
O
H
W
c/D
0.15
0.10
Assumed Clear
Baseline
1,0
2.0
3.0
A.O
RELATIVE AIR MASS
STANTON, N.D., OCT. 4 TO OCT. SO), 1976
-------
w
3
.25 r
.20
o
C/5
o
H
W
!/l
:=>
o
M
H
.15
.10
.05
FIG. 69 AVERAGE DAILY RATIO OF DIFFUSE
TO TOTAL SOLAR IRRADIANCE (.3 TO 2.8
OCT 4
I I I i I 1 t 1 | I I I I I I I I i i I t i I i I t I I t
4 m V9 24 2-9
DATE: OCT. 4 TO OCT. 36, 1976
STANTON, N.D.
-------
FIG. 70 AVERAGE HOURLY BROADBAND (.3 TO 2.8 pn)
RATIO OF DIFFUSE TO TOTAL SOLAR IRRADIANCE
o
00
.3 _
.2 _
,1 _
10
11
12
13
14
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D., OCT. 4 TO OCT. 30, 1976
-------
1000 I
FIG. 71 BROADBAND (.3 - 2.8 jam) DIRECT SOLAR IRRADIANCE
VERSUS RELATIVE AIR MASS
900
2
H
W
PQ
H
O
800
700
Clear Conditions
600
Cirrus Conditions
Ground-Snow Covered
500
•*-
2.0
3.0
4.0
5.0
RELATIVE AIR MASS
STANTON, N.D., JAN. 13 TO FEB 9, 1977
-------
w
u
M
I
W
CO
Pn
a
150
FIG.
72 BROADBAND (.3-2.8 >im) DIFFUSE
SOLAR IRRADIANCE VERSUS RELATIVE
AIR MASS
125
Ground - Snow Covered
100
75
50
Haze &
Cirrus
Conditions
Clear
Conditions
25
2.0
3.0
4.0
5.0
RELATIVE AIR MASS
STANTON, N.D., JAN. 13 TO FEE 9, 1977
-------
500 r
CO
£ 400
3
I
w
z
M
300
3
O
CO
200
3
EC
H 100
FIGURE 73 BROADBAND (.3-2.8 jam) TOTAL HORIZONTAL SOLAR
IRRADIANCE VERSUS RELATIVE AIR MASS
Cirrus Skys
Ground-Snow Covered
Clear
Skys
2.0
3.0
4.0
5.0
RELATIVE AIR MASS
STANTON, N.D., JAN. 13 TO FEE 9, 1977
-------
.60
w
M
g
.50
.40
H
O
H
O
H
2? .30
M
O
O
O
s
.20 ~
10
FIGURE 74 RATIO OF DIFFUSE TO TOTAL SOLAR IRRADIANCE
(.3-2.8 n} VERSUS RELATIVE AIR MASS
Cirrus Conditions
Ground-Snow Covered
Clear Conditions
2.0
3.0
4.0
5.0
RELATIVE AIR MASS
STANTON, N.D., JAN. 13 TO FEE 9, 1977
-------
u>
.60 r
w
.50
w .40
o
H
O
H
W
t/D
D
tu
fe
M
Q
O
O
U)
0
i -2°
.10
FIGURE 75 AVERAGE DAILY RATIO OF DIFFUSE TO
TOTAL SOLAR IRRADIANCE (.3:TO 2.8 ;im)
Max.
Horizon
Haae
Ave.
Min. 1 \ -•
Clea
Cirrus
Ground-Sbow Covered
Clear & Scattered Cummulus
Clear
Clear +
Horizon Haze
Clear with
Some Cirrus
I I I I I I i I i
I I I I 1 I 1 i
JAN 13
18 23 28 FEE 2
DATE: JAN 13 - FEE 9, 1977, STANTON, N.D.
-------
.60
FIG. 76 AVERAGE HOURLY BROADBAND (.3-2.8^um) RATIO OF
DIFFUSE TO TOTAL SOLAR IRRADIANCE
w
u
2;
M
Q
.50
Max.
IS
O
H
O
H
W
CO
O
O
Cirrus & Haze Conditions
.40
Ave.
.30
.20
Clear
Sky
.10
10
11
12
13
14
15
16
TIME - LOCAL STANDARD (CENTRAL)
STANTON, N.D., JAN. 13 - FEB. 9, 1977
-------
a
5
i
Jj
•H
Q
1000
FIG. 77 COMPARISON OF DIRECT SOLAR IRRADIANCE
VERSUS RELATIVE AIR MASS FOR APRIL, JULY,
OCT, JAN, FOR CLEAREST CONDITIONS
900
800
700
600
Derived froffl monthly data
— — — — Normalized to mean sun-earth distance
500
1.0
2.0
3.0
RELATIVE AIR MASS
4.0
5.0
-------
1000
800
w
o 600
(31
o
CO
400
200 -
PIG. 78 COMPARISON OF TOTAL SOLAR IRRADIANCE VERSUS RELATIVE
AIR MASS FOR APRIL, JULY, OCT, JAN, STANTON, N.D.,
1976-^77
Low Ground Albedo
Snow Cover
• April
• July
A Oct.
0 Jan.
RELATIVE AIR MASS
-------
FIGURE 79: COMPARISON OF THE RATIO OF DIFFUSE TO TOTAL SOLAR IRRADIANCE VERSUS
AIR MASS FOR APRIL, JULY, OCT, JAN, STANTON, N .D. 1976-77
.25
O
C/3
o
H
O
H
W
Q
O
O
20 -
15 _
10 -
.05
RELATIVE AIR MASS
-------
co
CO
CO
Fig. 80- RELATIONSHIP BETWEEN THE SLOPE OF THE DIFFUSE TO TOTAL SOLAR
IRRADIANCE VERSUS RELATIVE AIR MASS AND THE VERTICAL OPTICAL
ATTENUATION COEFFICIENT FOR THE BROAD BAND RANGE AT A RELATIVE
AIR MASS OF 3.0
.065
.055
• April
• July
A Oct.
• J an.
I
SB
Q
41
CX
O
.045
.035 -
.025
.15
.16
.17
,18
.19
.20
,21
.22
.23
T (Curve 67) .3 - 2.8
-------
7. SKY APPEARANCE
7.1 Techniques/Instrumentation Used - The general sky appearance, in the
horizontal direction, was documented (hourly from 0800 to 1600 LST)
with Kodak Kodachrome color film (35 mm slides) and a Honeywell Pentax
Spotmatic camera equipped with a 200 mm lens. Hourly photographs were
taken for each of four views, see Figure 1. Each of the four views, as
seen through the 200 mm lens, are shown in Figures 81 through 84.
7.2 April Sky Appearance Due to weather conditions, only 18 of the possible
24 days of observation were documented in terms of sky appearance. Of
the 18 observation days, 4 days were visibly hazy, and a very notice-
able smog layer was observed on 2 days. The 4 hazy days were April 5,
12, 13 and 19. The two days having noticeable smog layer, which was
located to the north-northwest of the site, were April 27, and the 28th.
7.3 July Sky Appearance - Of the 24 possible days of observation, 24 days
were documented in terms of sky appearance. Of the 24 observation days,
4 days were visibly hazy, and a very noticeable smog layer occurred on
5 days. Days having a slightly noticeable haze were July 12, 19 and
the 26th. A medium haze was evident on Aug. 6. On July 24 a smog
layer existed to the northeast. On Aug. 2, smog was evident east to
west. On Aug. 3, smog was evident to the north. On Aug. 4 smog was
evident to the north and west. On Aug. 7 a very heavy haze/smog was
evident in all directions.
7.4 October Sky Appearance - Out of the 24 possible days of operation, 24
days were documented in terms of sky appearance. Of the 24 observation
days, 4 days were visibly hazy and a noticeable smog layer occurred on one
day. These days were Oct. 19, 21, 22, and 29. On Oct. 11 a smog layer existed
existed to the north.
7.5 January Sky Appearance - Of the 24 possible days of observation, 21
days were documented in terms of sky appearance. Hazy conditions
119
-------
VISIBILITY TARGET #1
\
Figure 81 View #1 - 200 mm Lens
VISIBILITY TARGET. #2
120
-------
VISIBILITY TARGET #3
Figure 83 View #3 - 200 mm Lens
Figure 84 View* #4 - 200 mm Lens
121
-------
existed on January 13, 17, 21 and February 7. On January 28 and Feb-
ruary 1, smog was evident to the north and east.
7.6 Seasonal Sky Appearance Comparisons - A comparison of the visibly hazy
and smog days for each seasonal month is shown in the following table.
Table 2. Comparison of Hazy and Smog Days
Month No> of Hazy Days No. of Smog Days
April 4 2
July 4 5
Oct. 4 1
Jan. 4 2
As shown in the above table, each month had four days where a haze was
visibly noticeable. These four days represents about 17% of the total
number of observed days. Considering the days where smog was noticed,
July had by far the greatest smog occurrence, a total of 5 days. The
other months had about 2 days of smog occurrence. The document direc-
tion of this smog was reported to be the north, northeast, and north-
west .
122
-------
8. Meteorology
8.1 Techniques/Instrumentation Used - The meteorological parameters mea-
sured consisted of the temperature, pressure, relative humidity, wind
direction, wind speed, dust concentration, and total precipitable water
vapor.
The temperature and relative humidity were measured using a sling psycho-
meter to determine dry and wet bulb temperatures. The pressure was mea-
sured with an aneroid-type engineer barometer. The wind speed was mea-
sured using a hand-held cup anemometer. The wind direction was measured
with a precision Brunton compass and a directional flag. The relative
dust concentration was measured with an integrating, 90 scattering
angle nephelometer, manufactured by Weather Measure Corporation (Digi-
tal Dust Meter). This nephelometer is an electro-optical instrument
that provides an indication of atmospheric dust concentrations in the
range from 0.01 to 500 mg/m^ of air. Light scattered by dust, as it
aspirates past a fixed intensity lightbeam, is measured by a photo-
multiplier tube yielding a count rate directly proportional to dust
concentration. The size of dust particles sampled are determined by
an optical maze that screens out all particles greater than 20 microns
and accepts approximately 60% of 10 micron particles and 100% of 5 micron
or smaller particles. An estimate of dust concentration in mg/nH of
air is made by
C = (R-D)K (27)
where:
C = dust concentration in mg/nr in air
R = counts per minute
D = background counts (known)
K = mg/m^ of dust per count (known)
123
-------
The absolute accuracy of the nephelometer is subject to errors associ-
ated with difficulties of calibrating the instrument with a known aerosol
that closely approximates the aerosol measured on site. However, the
counts per minute will give an accurate indication of relative dust con-
centration. For this reason only the counts per minute are reported be-
cause of a lack of calibration (after repair) of the instrument.
The amount of precipitable water vapor was measured with the Volz 5
channel photometer. The basic method involves the measurement of dir-
ect solar beam intensity in an absorption band of water vapor, 0.940/urn,
and the measurement of intensity in an adjacent "window" or region of
no absorption, 0.880 /urn. The ratio of these measurements is sensitive
to the amount of water vapor present in the atmosphere. The relation-
ship used is
P.W. = K/m [log (q0/q)] 2 (28)
where K and q are constants, m is the relative air mass, and q is the
ratio of the 0.940 and 0.880 solar photometer meter readings.
The baseline meteorological parameters were established for the hours
from 0800 to 1600 L.S.T., on the hour; they were as follows:
(1) Average daily temperature
(2) Average daily pressure
(3) Average daily relative humidity
(4) Average daily wind velocity
(5) Average daily wind direction
(6) Average daily relative dust concentration
(7) Average daily vertical precipitable water vapor
(8) Average hourly vertical precipitable water vapor
•2 April Meteorology - The results for the April meteorological measure-
ments are shown in Figure 85 average daily temperature, 86 - average
124
-------
daily pressure, 87 - average daily relative humidity, 33- average
daily wind velocity, 89 - average daily wind direction, 90 average
daily relative dust concentration, 91 - average daily precipitable
water vapor, and 92 - average hourly precipitable water vapor. The
mean temperature for April (for 0800 to 1700 LST) was 12.7°C. The mean
relative humidity was 47.7%. As shown in Figure 89, the prevailing
wind direction would fall from south to east. The average precipitable
water vapor, for those days when measurements were possible, was approxi
mately 0.55 cm.
.3 July Meteorology - The July meteorological measurement results are
shown in Figure 93 - temperature, Figure 94 - pressure, Figure 95 -
relative humidity, Figure 96 - wind velocity, Figure 97 - wind direc-
tion, Figure 98 - relative dust concentration, Figure 99 - daily pre-
cipitable water vapor, and Figure 100- hourly preciptable water vapor.
The mean temperature (0800 to 1700 LST) was 25.7°C. The mean relative
humidity was 38.67o. As shown in Figure 97 , the wind direction was
quite variable. The average precipitable water vapor, as derived
from Figure 100 , was approximately 1.35 cm.
.4 October Meteorology - The October meteorological results are shown in
Figure 101 - temperature, Figure 102 - pressure, Figure 103 relative
humidity, Figure 104- wind velocity, Figure 105 wind direction,
Figure 106- relative dust concentration, Figure 107 daily precipi-
table water vapor, and Figure 108 - hourly precipitable water vapor.
The mean temperature (0800 to 1700 OST) was 5.8°C, and the mean rela-
tive humidity was 53.1%. As shown in Figure 105 , the general revailing
wind direction was from the northwest to south. The average precipi-
table water vapor, as derived from Figure 108 , was approximately 0.65
cm.
125
-------
.5 January Meteorology - The January meteorological results are shown in
Figure 109 - temperature, Figure 110 - pressure, Figure 111 - relative
humidity, Figure 112 - wind direction, Figure 113 wind velocity,
Figure 114 - relative dust concentration, Figure 115 - daily precipi-
table water vapor, and Figure 116 - hourly precipitable water vapor.
The mean temperature (0800 to 1700 LST) was -9.6°C, and the mean rela-
tive humidity was 78.9%. As shown in Figure 112, the general prevail-
ing wind direction was from the northwest to west. The average pre-
cipitable water vapor, as derived from Figure 116 , was approximately
0.35 cm.
6 Seasonal Comparisons - A comparisons of the various meteorological
parameters is shown in Table 3.
Table 3 Comparison of Meteorological Results
Month
April
July
Oct.
Jan.
T-C°
12.7
25.7
5.8
-9.6
R.H.
47.7
38.6
53.1
78.9
P.W. -cm
0.55
1.35
0.65
0.35
W.D.
S-E
Vari-
bale
NE-S
NW-W
Vis-Miles
34.0
39.9
32.2
27.2
mean temperature ;
R-H. - mean relative humidity;
P.W. - mean precipitable water vapor;
W.D. - wind direction;
Vis - mean daily visibility, overall view.
As can be seen, January had very cold temperatures, high relative
humidity, low precipitable water vapor, and the lowest visibility.
Whereas, July had the high temperatures, low relative humidity,
high precipitable water vapor, and the highest visibility. April and
October were somewhat similar in all parameters.
126
-------
40
30
20
01
>~i
3
i-i
01
CX
6
aJ 10
o.
ID
O
B
-10
Fig. 85
Atmospheric Temperature
Versus Day of Month
(for 0800 to 1700 LSI)
1 1 I
L I 1
1 1 L
Example
Max.
Mean
Min.
1
I
X
I I I I
I -I
1 ' '
A I I 4.
X i
10
15
20
25
30
DATE: April 5-30, May 1, 1976
Stanton, N.D.
-------
720
N3
CO
715
00
Pd
'o 710
01
^
3
tn
S 705
j-j
P-I
o
•H
a.
03
700
695
Fig. 86 Atmospheric Pressure Versus Day of Month
(For 0800 to 1700 LSI")
Example
-r Max.
I I !.]__!. L_l L _J I I I I I I I I L
10
15
20
25
30
DATE: April 5-30, May 1, 1976
Stanton, N.D.
-------
N>
•iD
100
80
60
4-1
•r-l
T3
•t-l
B
PC
40
20
Fig. 87 Atmospheric Relative Humidity Versus Day of Month
(For 0800 to 1700 LSI)
T
/
\ T
\T
\ 1
, V
i f
i
1 1
i
i
i
Ex amp 1 e
, T Max.
Mean
Min.
1
I I I
jlll
I I I I I I I I I
i i i I
10
15
20
25
30
DATE: April 5-30, May 1, 1976
Stanton, N.D.
-------
100 u
Example
Max.
80
Mean
Fig.88 Wind Velocity Versus Day of Month
(0800 to 1700 LSI)
1
Min.
60
OJ
o
s-l
sc
"s
>^
o
o
•o
c
•1-1
40
20
I I i
II ,1
I I
10
15
20
25
30
DATE: April 5-30, May 1, 1976
Stanton, N.D.
-------
o N.J60
z
3
^ NW
H 1NW »
e
j_i
fe W.270
CO
J5 SW,
w
o
i— i
0
^ 8,180
CO
cu
dl
»-l QE-
W] ^>t,
Q
gE,90
• r-l
u
Ol
.H NE,
Q
•D
•S N.O
~~ Fig. 89 Average Daily Wind Direction
for Stanton site. •»•
I
}
! I
— •*•
- . T
T » T ' i
TK T I ] : !
'1 ill'
i,il i i M i
« 0. . -*-
1
I I
-L t
i |
iti| i i i i i i | i i i i i 1 i l i i 1 i i i i i i
10
15
20
25
30
Date: April 5-30, May 1, 1976
Stanton, N.D.
-------
.040
°S
£ .030
i
Concentration
1
10
Q
." .010
-------
0.90 r-
T
o
0,
-------
8,80 _
co
*J w \^ W
0.70
0.60
0
>
o
cxj
* 0.50
0
S
u
0)
,0
cd
•H °'40
a.
•H
u
M
PM
0.30
T
Max.-p
T T
1 T 1
1 ' 1-r '
| 1 T | j T | 1
- *„ | } 1 j , , |
1 T ' 1 1 $ \
1 1 Q T T i
1 ill 1
i 1 ' ! i i
Mini -L 1 -L 1
1
_ Fig. 92 Average Hourly Precipitable Water Vapor J
1 1 1 1 i 1 1 1 1
89 10 11 12 13 14 15 16 17
'TIME - Local Std. Time (Central)
April 5,6,8,10,12, 1976
Stanton, N0D.
-------
40 -
w
w
H
35
30 _
25 _
20
15
FIG. 93 ATMOSPHERIC TEMPERATURE VERSUS
DAY OF MONTH (FOR 0800 TO 1700 LST
I I I I 1 I I I I
JULY
DATE: JULY 12 - AUG 7, 1976
STANTON, N.D.
-------
720
FIG. 94 ATMOSPHERIC PRESSURE VERSUS DAY OF MONTH (0800 TO 1700 LST)
715
lr*'
Oo
710
o
ffi
705
w
700 L-
INSTRUMENT
MALFUNCTION
695
I I I
I I I I
I I 1 I I I I 1 I I I I I I I I
JULY 12
17
22
27 AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
100 h
FIG. 95 ATMOSPHERIC RELATIVE HUMIDITY VERSUS DAY OF
THE MONTH (0800 10:1700 LST)
80 _
60
40
W
20
i l I I
I I 1 I I I I I I I I
I I I I 1 I I I I I I I
JULY 12
17
22
27 AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
100
FIG. 9& WIND VELOCITY VERSUS DAY OF MONTH (0800 TO 1700 LST)
00
80
60
o
hJ
w
40
20
JULY 12
17
22
27
AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
N.360
o
z
0 W,270
e
w
§ SW
o
S..180
SE
£,90
NE
FIG.97 AVERAGE WIND DIRECTION FOR STANTON SITE
I.
I
I I
I I I i I I I I
i» i •
I I I
I I I I
JULY 12
17
22
27
AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
-p-
Q
6.0
5.5
w 5.0
H
4.5
4.0
INSTRUMENT
BEING REPAIRED
FIG. 98 RELATIVE ATMOSPHERIC DUST
CONCENTRATION
I I l
1 I I f
I I
JULY 12
17
22
27
AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
2.5
2.0
o
PH
o 1.5
CM
s
fc,
o
o
I
H
M
P-i
M
O
1.0
.5
FIG. 99 AVERAGE DAILY PRECIPITABLE
WATER VAPOR
I I I 1 I 1 I I I
I . . . . I
I I I 1 I I i I I I I I
JULY 12
17
22
27
AUG 1
DATE: JULY 12 TO AUG 7, 1976
STANTON, N.D.
-------
2.5
FIG. 100 AVERAGE HOURLY PRECIPITABLE WATER VAPOR (CM)
BJ
O
I
O
CM
nd
fn
O
S
o
w
hJ
m
IS
H
PL.
M
U
2.0 f-
1.5
1.0 h
.5
10
11
12
13
14
14
L6
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D.
-------
30
FIG. 101 ATMOSPHERIC TEMPERATURE VERSUS
DAY OF THE MONTH (FOR 0800 TO 1700 LST)
20
10
w
PM
w
H
-10
i i I i I ill
OCT 4
14 19
DATE: OCT 4 TO OCT 30, 1976
STANTON, N.D.
24
29
-------
720
EIG. J0£, ATMOSPHERIC PRESSURE VERSUS DAY OF THE MONTH (6800 TO 1700 LST)
715
o
«
fn
o
710
70S
CO
w
w
a!
PM
700
695
I .... I
I I I I I I I I I I I I ( 1 I L I I I I I I
t'4
19
24
2f
DATE: OCT. 4 TO OCT 3®, 1976
STANTON, N.D.
-------
100 i-
80
60
H
M
Q
s
£
M
H
40
20
FIG.103 ATMOSPHERIC RELATIVE HUMIDITY
VERSUS mY OF THE MONTH
(0800 TO 1700 LST)
I I II J II III I J I I I I 1 1 I I I I I I I I I
OCT
14
19
24
29
DATE: OCT. 4 TO OCT. 30, 1976
STANTON, N.D.
-------
loo
80
60
H 40
H
i-l
$
Q
S5
* 20
FIG.104 WIND VELOCITY VERSUS DAY
OF THE MONTH (0800 TO 1700
t . I ,1 X._l ._L_1 I -I
OCT 4
14
19
24
DATE: OCT. 4 TO OCT. 3§D, 1976
STANTON, N.D.
-------
N.360 -
o
§ W.270
W
en
u
o
u
g
M
Q
Q
SW
8,180
SE
E,90
NE
N,0
FIG. 105 AVERAGE WIND DIRECTION FOR STANTON SITE
OCT. 4
14
19
23
Z9
DATA: OCT. 4 TO OCT. 3$, 1976
STANTON, N.D.
-------
12
10
CO
o
u
FIG. l6'6 RELATIVE ATMOSPHERIC pUST CONCENTRATION
I 1 I I 1 I
I I I I
I .... I
I I I I
I . I . 11
OCT A
14
DATE: OCT. 4 TO OCT. 8*, 1976
STANTON, N.D.
-------
2.5
2.0
FIG. 10.7 AVERAGE DAILY PRECIPITABLE WATER VAPOR
g
O
CM
ns
s
u
CQ
3
M
C^
M
O
1.5
1.0
.5
\
J_L
OCT
14
19
24
29
DATE: OCT. 4 TO OCT. 30, 1976
STANTON, N.D.
-------
2.5
2.0
EIG. 108 AVERAGE HOURLY PRECIPITABLE
WATER VAPOR (CM)
g
e
1.5
1.0
M-
M
fX.
M
U
2
.6
10
U
13
14
15
16
TIME - LOCAL STD. TIME (CENTRAL)
STANTON, N.D.
-------
+10
0
-10
o
i
H
2
w
w
H
-20 -
-30
FIG. 109 ATMOSPHERIC TEMPERATURE VERSUS DAY OF THE
MONTH (FOR 0800 TO 1700 LSI)
Max.
Ave.
Min.
-40
I I 1 I I I I I I I 1 I I I I 1 I I I I 1 I 1 I I I I I i
JAN 13
18
23
28 FEB 2
DATE: JAN. 13 - FEB 9, 1977, STANTON, N.D.
-------
720
7-15
705
7QQ . _
695
FIG. 110 ATMOSKffiKtC PRESSURE VERSUS DAY Of TEE
MONTH (FOR 0800 TO 170|0 LST)
Min
1 I I I I I I 1
• I
1
J I
I 1 I I 111
JAN 13
1& 23 28 FE3 g
: JAN 13 - FEB 9, 1977* STANTON, N.D.
-------
100 _
80
60
Q
g
£
M
H
40
\
-T. Max
Ave
.. Min
FIG. Ill ATMOSPHERIC RELATIVE HUMIDITY VERSUS DAY
OF THE MONTH (FOR 0800 TO 1700 LST)
20
I I I I
JAN 13
19
23
28
FEB 2
DATE: JAN. 13 - FEB 9, 1977, STANTON, N.D.
-------
N.W.
EL W.270 _
I s...)-
Si S,l®~-
SE
o
CJ
E.9.Q -__
N.E.
N,0
FIG. 112 AVERAGE WIND DIRECTION VERSUS DAY OF THE MONTH (FOR 0800 TO 170|0 LST)
JAN
1 i 1 L I I I 1 I t I 1 1
18 23 28 FEB 2
;. JAN 13 - FEB 9, 1977, STANTON;
-------
100 «_
FIG. 113 WIND VELOCITY VERSUS DAY OF THE MONTH (FOR 0800 TO 1700 LST)
80
s
I
H
M
O
w
Q
'Z.
60
40
20
-r -P Max
I I I I
I I 1 I ,. I 1 I 1 1 I 1 1 1 ..
Ave
JAN
1« 2J 28 FEE 2
DATE: JAN 13 - FEB 9, 1977, STANTON, N.D.
-------
7.5
Ul
7.0
5.5 I
5.0
FIG. 114 RELATIVE ATMO{ff>HERIC DUST
CONCENTRATION VERSUS DAY
OF Tip: MONTH.
Max _.
Ave
6.0 I Min I
/ i
I I I I I I 1 I f i. I 1
JAN 13 1^ 23 18 FEE 2
DATE: JAN 13 - FEE 9, 1977, STANTON, N.D.
-------
FIG. US AVERAGE DAILY PRECIPITABLE WATER VAPOR VERSUS DAY OF THE MONTH
BS
o
K
w
H
13
o
u
w
PQ
h-H
M
o
w
.75
.50
.25 •-
T Max
Ave
-*• Min
I L
ill!
till
JAN
1-8 2^ 28- FE6 2
DATE: JAN 13 - FEB 9, 1977, STANTON, N.D.
-------
a
CQ
PM
1.0
erf
p .75
w
H
.50
.25 I
FIG. 116 AVERAGE HOURLY PRECIPITABLE WATER VAPOR
Max _
Ave
Min J_
10
11
12
13
14
15
16
TIME-LOCAL STANDARD (CENTRAL)
STANTON, N.D., JAN 13 - FEE 9, 1977
-------
9. SATELLITE IMAGERY
9.1 Techniques/Instrumentation Used - The instrumentation used consists
of the LANDSAT satellite systems. These systems image a 100 x 100
mile surface area, every 18 days, in four spectral bands (.5 to .6,
.6 to .7, .7 to .8 and .8 to 1.1 micrometers). This satellite
imagery that was used consists of 9 x 9 in. positive, black and
white, transparencies, ordered directly from the EROS data center.
Two bands were considered, the .5 to .6 and .7 to .8 jam bands.
These positive transparencies were enlarged by a factor of approxi-
mately 7X to give the maximum spatial resolution possible, as shown
in Figure 117. The technique used was to simply photointerpret the
images to detect any smoke plumes or pollution clouds, and determine
their source, size, and opacity. The opacity of the smoke plume can
be qualitatively estimated by comparison of the .5 to .6 pm and .7
to .8 urn images. For example, a thin plume will exhibit a different
appearance in the shorter .5 to .6 ^im wavelengths than the longer
.7 to .8 urn wavelengths, because of the greater opacity in the shorter
wavelengths. An opaque/dense plume will exhibit a similar appearance
in both the short and long wavelengths because of its high opacity.
9.2 April Satellite Imagery - Imagery was obtained for April 3 & 12 inspec-
tion of this imagery, in the .5 to .6 ^im and .7 to .8 pm bands, revealed
no detectable plumes and/or pollution clouds.
9.3 July Satellite Imagery Imagery was obtained for Aug. 7. No detec-
table plumes were present.
9.4 October Satellite Imagery - Imagery was obtained for October 9. No
detectable plumes were present.
159
-------
9.5 January Satellite Imagery - Imagery was obtained for January 17. As
shown in Figure 117, a very noticeable smoke plume was detected.
Although Figure 117 is for the 0.5 to 0.6 urn band, the 0.7 to 0.8 band
displayed the same opacity, indicating that the plume was fairly
dense. The plume is approximately 7 miles long, 0.4 miles wide, and
is at an altitude of approximately 850 ft. above terrain. The altitude
of the plume is easily determined by a knowledge of the sun angle and
the length of the plume shadow. The time of the LANDSAT overpass is
approximately 10:30 AM LST.
9.6 Seasonal Satellite Imagery Comparisons - Because the LANDSAT satellite
passes over a given site so infrequently, it is difficult to make
any seasonal comparisons.
160
-------
POLLUTION SOURCE
,"
.tl^TW_
' "
" ^,pT. ft--^ri^r&^s5gj«r,; .DIRECTION
&' £/ • '^feW*^^^-^--V *"•---
S" . #» •. •«•„.# >.r^ fe^V^^*^' ,|-'- * „;. -.?*• f .>:-•
*^V**«*7-^M^^-* "* Jf .:&-$••-' •'-•'•- ' •* '
~w v^^^^^^^f ',J^ ;!:::i':f ;l
Figure 117 LANDSAT' IMAGE (VISIBLE, .5-.6 urn) OF SMOKE PLUME NEAR
STANTON, N.D., 7X ENLARGEMENT OF 9 x 9 IN. POSITIVE
TRANSPARENCY, DATE-JAN. 17, 1977; TIME-10;30 LST
161
-------
10. COMPARISON QF HORIZONTAL VISIBILITY WITH-VERTICAL OPTICAL ATTENUATION
AND SOLAR 1RKAD1ANCE
The relationships between the vertical optical attenuation properties
and the horizontal visibility are dependent upon the vertical atmos-
pheric profiles of aerosols and water vapor. Because of this, the
relationships between the vertical optical attenuation and horizontal
visibility could be quite variable and quite possibly unrelated. For
example, a strong, shallow temperature inversion could confine high
concentrations of aerosols and water vapor near the surface, greatly
reducing horizontal visibility. Whereas, the vertical attenuation
would not be affected nearly as much.
10.1 Horizontal Visibility Versus Vertical Optical Aerosol Attenuation @ 0.5 up
The vertical aerosol attenuation coefficient at 0.500 urn was compared
to the overall horizontal visibility for the four seasonal months.
The results of such an analysis is shown below:
Table 4 r (.500 pm) versus Visibility
Linear Correlation Analysis
Equation
V = 45.3 - 71.8 T
a
V = 68.0 - 193.6 T
a
V = 52.0 - 181.3 T
a
V = 36.9 - 212.5 r
a
As can be seen a fairly good correlation exists between the vertical
attenuation and horizontal visibility. This correlation depends on
the seasonal month, with the best correlation being in July, followed
by October, January and April. This dependence on season is probably
due to the particular relationships existing, during those seasons,
between the vertical and horizontal atmospheric properties.
162
Month Correlation Coefficients (r )
April
July
Oct.
Jan.
0.38
0.56
0.51
0.45
-------
10.2 Horizontal Visibility Versus Ratio of Diffuse to Total Solar Irradiance
As previously mentioned, the ratio of diffuse to total solar irradiance
is one indicator of atmospheric clarity because it indicates the
scattered diffuse skylight, which is determined by the amount of a
aerosols and relative air mass. A linear correlation analysis was
performed for the visibility versus ratio of diffuse to total solar
irradiance. The results of this analysis are shown below.
Total 5 Correlation Between Visibility & Diffuse to Total
Solar Irradiation Ratio
Month Correlation Coefficient Equation
April 0.27 V = 50.38 - 83.5 (D/H)
July 0.24 V = 61.5 178.4 (D/H)
Oct. 0.82 V = 78.3 - 25.3 (D/H)
Jan. 0.81 V = 57.9 - 110.7 (D/H)
In general, all months of data show that visibility is inversely
proportional to the ratio of diffuse to total solar irradiance.
For the seasonal months of April and July this correlation coeffi
cient of 0.27 and 0.24, respectively. However, for the months of
October and January, the correlation is extremely good, correlation
coefficients of 0.82 and 0.81, respectively. This wide variation
in the correlation coefficient verifies the fact that, as pointed
out previously, the relationship between the vertical atmospheric
optical air quality and the horizontal visibility may or may not be
directly correlated because of the variable vertical structure of
the atmosphere.
163
-------
APPENDIX A
164
-------
• POLLUTION SOURCES
A MONITORING SITES
Figure A-l Location of Pollution Sources and North Dakota
Department of Health Monitoring Sites
165
-------
North Dakota Air Quality Monitoring Sites
Natural Resource Development Area
SITE
1. Stanton
2. Washburn
3. Garrison
4. Beulah
5. Beulah
6. Dunn Center
Dickinson
TYPE
Rural
Rural
Rural
Residential
Rural
Rural and
Commercial
POLLUTANTS SAMPLED
so2
NOX
SCv-NCv, gas bubbler
TSP (High Volume)
Particle size distri-
bution (cascade
impactor)
Oxidants (Oo)
Sulfation Rate
S02-N0?, gas bubbler
TSP (High Volume)
Sulfation Rate
TSP (High Volume)
TSP (Membrane -filter
Sulfation Rate)
S02-N02, gas bubbler
TSP (High Volume)
Sulfation Rate
S02-N02, gas bubbler
TSP (High Volume)
Sulfation Rate
TSP (High Volume)
TIMES -SAMPLED
Continuous
Continuous
24 hr. every 6 days
24 hr. every 6 days
24 hr. every 6 days
Continuous
Monthly
24 hr. every 6 days
24 hr. every 6 days
Monthly
24 hr. every 6 days
24 hr. every 6 days
Monthly
24 hr. every 6 days
24 hr. every 6 days
Monthly
24 hr. every 6 days
24 hr . every 6 days
Monthly
166
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-908-1/77-005
3. RECIPIENT'S ACCESSION NO.
TITLE AND SUBTITLE
A Study of Horizontal Visibility, Atmospheric Vertical
Optical Properties and Solar Insolation at Stanton,
North Dakota
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Roland L. Hulstrom
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
MARTIN MARIETTA CORPORATION
Denver Division
Post Office Box 179
Denver, Colorado 80201
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency, Region VIII
Office of Energy Activities
1860 Lincoln Street
Denver, Colorado 80295
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A study to establish methods for the acquisition of baseline data for horizontal
visibility and vertical optical air quality is described. The study was performed
for the U. S. Environmental Protection Agency, Region VIII.
The site was Stanton, North Dakota, which is located in an area where extensive
development of fossil fuel energy is anticipated. The baseline measurements
consisted of horizontal visibility along three views, solar radiation and
vertical atmospheric spectral attenuation coefficients. Basic meteorological
data were also taken. These baseline data were established for the months of
April, July, and October, 1976, and January, 1977. The methods used, results,
conclusions and recommendations are presented.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Visibility
Air Quality
Solar Insolation
Atmospheric Attenuation
Visual Range
Visibility
Air Quality
8. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
172
20. SECURITY CLASS (This page}
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION i s OBSOLETE
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, it used, in smalle:
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date oj
approval, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR(S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Insert appropriate code.
15. SUPPLEMENTARY NpTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented'at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote releasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. &20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (Rev. 4-77) (Reverse)
------- |