EPA-600/3-78-039
April 1978
VISIBILITY IN THE SOUTHWEST
An Exploration of the Historical Data Base
by
John Trijonis and Kung Yuan
Technology Service Corporation
2811 Wilshire Boulevard
Santa Monica, California 90403
Grant No. 803896
R.B. Husar, Principal Investigator
Washington University
St. Louis, MO 63130
Project Officers
John C. Butler
Office of Planning and Evaluation
Washington, D.C. 20460
and
William E. Wilson and Thomas G. Ellestad
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, NC 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Laboratory,
U.S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
ABSTRACT
The historical data base pertinent to visibility in the Southwest is
analyzed. The data base includes over 25 years of airport visibility ob-
servations and more than 10 years of NASN particulate measurements. The
investigation covers existing levels of visibility, long-term trends in
visibility, and visibility/pollutant relationships.
Although still quite good, visibility in the Southwest has deterio-
rated over the past two decades. The haze levels in the Southwest appear
to be mostly the result of secondary aerosols, especially sulfates. These
conclusions are verified by decreases in sulfates and increases in
visibility during the 1967-1968 industry-wide copper strike.
This report was submitted in fulfillment of grant 802815 by Technology
Service Corporation under the sponsorship of the U.S. Environmental
Protection Agency. This report covers a period from March 1977 to November
1977, and work was completed as of November 1977.
m
-------
CONTENTS
ABSTRACT iii
FIGURES vii
TABLES ix
1. INTRODUCTION AND SUMMARY 1
Summary of Conclusions 2
Limitations of the Analysis 5
Future Work 7
2. DATA BASE PREPARATION AND DATA ANALYSIS METHODS 9
Airport Weather Data and NASN Pollutant Data 9
Frequency Distributions of Visibility Data 17
Analysis of Visibility/Pollutant Relationships 25
Allowances for Meteorology 31
3. EXISTING VISIBILITY LEVELS 35
Visibility in Urban and Nonurban Areas 35
Geographical Patterns in Visibility 37
Consistency with Recent Field Programs 41
4. HISTORICAL VISIBILITY TRENDS 42
Year-to-Year Visibility Trends 42
Net Changes in Visibility and Extinction, 1954-1971 ... 56
5. VISIBILITY/POLLUTANT RELATIONSHIPS 61
Phoenix, Maricopa County Data 61
Phoenix, NASN Data 70
Salt Lake City, NASN Data 72
Analysis of Other Data Bases 74
Discussion and Generalization of Results 75
-------
CONTENTS
6. THE COPPER STRIKE OF 1967-1968 82
Copper Production and Visibility at Tucson 83
Regionwide Changes in Sulfates During the Copper
Strike 85
Regionwide Changes in Visibility and Extinction
During the Copper Strike 90
Changes in Extinction Compared to Changes in
Sul fate 96
Analysis of Meteorology 97
REFERENCES 101
APPENDICES
A. List of NCC Weather Sites in the Southwest 104
-------
FIGURES
Number Page
1 Airport weather stations used in the Southwest study .... 12
2 NASN monitoring sites in the Southwest 16
3 Cumulative frequency distributions of visibility at
urban locations IS
4 Cumulative frequency distributions of visibility at
nonurban locations 19
5 Historical visibility trends at Salt Lake City 32
6 Historical visibility trends at Denver 33
7 Geographical distribution of median visibilities (in miles) . 38
8 Geographical distribution of (best) 10th percentile
visibilities (in miles) 39
9 Geographical distribution of (worst) 90th percentile
visibilities (in miles) 40
10 Long-term visibility trends at Phoenix 43
11 Long-term visibility trends at Tucson 14
12 Long-term visibility trends at Denver 45
13 Long-term visibility trends at Salt Lake City 46
14 Long-term visibility trends at Fort Huachuca 47
15 Long-term visibility trends at Prescott 48
16 Long-term visibility trends at Winslow 49
17 Long-term visibility trends at Colorado Springs 50
18 Long-term visibility trends at Grand Junction 51
19 Long-term visibility trends at Pueblo 52
20 Long-term visibility trends at Ely 53
vn
-------
FIGURES
Number Page
" 11 i ii i d» i i.
21 Long-term visibility trends at Cheyenne 54
22 Extinction coefficient versus TSP in Phoenix 63
23 Extinction coefficient versus RH in Phoenix 64
24 Extinction coefficient versus Sulfate in Phoenix 65
25 Extinction coefficient versus Nitrate in Phoenix 66
26 Extinction coefficient versus BSOL in Phoenix 67
27 Normalized light scattering by aerosols as a function of
particle diameter 76
28 Long-term increase in haze correlated with copper
production 84
29 Changes in number of hazy days at Tucson during 1967-1968
copper strike 86
30 Changes in sulfation rate at Tucson during 1967-1968 copper
strike 86
31 Variations in pollutant levels at Tucson during 1967-1968
copper strike (NASN data) 87
32 Seasonally adjusted changes in sulfate during the copper
strike compared to the geographical distribution of
smelter SOV emissions 89
/\
33 Frequency distribution of sulfate concentrations during
the copper strike compared to seasonal average distri-
bution, Grand Canyon and Mesa Verde data combined 91
34 Seasonally adjusted percent changes in visibility during
the copper strike 93
35 Seasonally adjusted percent changes in extra extinction
during the copper strike 94
36 Seasonally adjusted changes in extinction during the
copper strike, units of [10^ meters]"^ 95
vm
-------
TABLES
Number Page
1 Airport Weather Stations Included in the Southwest
Visibility Study 12
2 Data for Visibility/Pollutant Studies 26
3 Median, Best-Case, and Worst-Case Visibility Levels at
Seventeen Locations in the Southwest 36
4 Net Percent Changes in Visibility, 1953-1955 to
1970-1972 57
5 Net Percent Changes in (Extra) Extinction, 1953-1955
to 1970-1972 59
6 Results of Uni-Variate Regressions for Phoenix, Maricopa
County Data 62
7 Intercorrelations Among Independent Variables at
Phoenix, Maricopa County Data 69
8 Average Extinction Budget for Phoenix, Maricopa County Data. 7C
9 Correlations Among All Variables at Phoenix, NASN Data ... 71
10 Average Extinction Budget for Phoenix, NASN Data 72
11 Correlations Among All Variables at Salt Lake City, NASN
Data 73
12 Average Extinction Budget for Salt Lake City, NASN Data. . . 74
13 Estimates of Extinction Coefficients per Unit Mass 77
14. Predictions of Average Visibility at National Park Sites . . 80
15. Average Extinction Budgets for National Park Sites 81
16. Changes in Sulfate Levels During the Copper Strike
Compared to Seasonal Averages 88
17. Changes in Visibility and Extinction During the Copper
Strike Compared to Seasonal Averages 92
-------
CHAPTER 1
INTRODUCTION AND SUMMARY
*
Perhaps the most important air quality concern in the Southwest United
States, where the high mountains, rugged terrain, and excellent visual range
produce many exceptional vistas, is the issue of visibility. It is a common
public opinion in the Southwest that the high visibility levels have already
deteriorated, and there is concern that future population growth and energy
development will lead to further visibility degradation. These concerns are
reflected in the 1977 Clean Air Act Amendments which contain provisions for
the prevention of significant deterioration and the protection of visibility
in federally designated Class I areas.
In response to the growing interest in visibility, several field programs
and modeling studies have been initiated in the Southwest. These studies will
help provide the data and analytical tools necessary to include visibility
considerations in future air quality planning.
We can also further our understanding of the visibility issue by analyz-
ing existing data bases. A potential wealth of information is offered by
over twenty-five years of airport visibility observations and more than ten
years of NASN (National Air Surveillance Network) particulate measurements.
The purpose of this report is to explore the historical data base in an attempt
to answer several key questions concerning visibility in the Southwest. The
*
For the purposes of this study, the Southwest is defined as Arizona,
Colorado, New Mexico, Utah, eastern Nevada, and southern Wyoming.
1
-------
questions we address are as follows:
What are existing visibility levels in urban and nonurban parts of
the Southwest? What are the statistical distributions and geograph-
ical patterns of visual range?
What trends have occurred in visual range over the past 25 years?
What is the spatial scale of the historical visibility changes?
Based on regression models, what are the key contributors to haze in
the Southwest? Is fugitive dust important to light extinction, or do
secondary aerosols (e.g. sulfates)dominate visibility reduction?
Recognizing that copper smelters were the major source of SOX emissions
in the Southwest during the late 1960's, what changes occurred in sul-
fate levels and visibility during the July 1967 - March 1968 industry-
wide copper strike? Do these changes agree with the predictions of
regression models?
This report is organized in six chapters. The present chapter provides a
statement of purpose and a summary of conclusions. Chapter 2 describes the
data bases that are used and the statistical methods that are applied. The
remaining four chapters sequentially deal with the four sets of questions
listed above.
SUMMARY OF CONCLUSIONS
Existing Visibility Levels (Chapter 3)
Visibility in the Southwest tends to be quite high. Median visibility
ranges approximately from 30 to 55 miles at urban locations and from
65 to 80 miles at nonurban locations. Best 10th percentile visibility
ranges from 45 to 70 miles among urban airports and from 90 to 115
miles among the nonurban airports. For comparison, we note that an
atmosphere consisting of air molecules alone would exhibit a visual
range of approximately 160 miles due to blue-sky (Rayleigh) scatter
by the air molecules.
Three nonurban airports exhibit lower visibilities than the values
listed above for nonurban locations. Of these three exceptions, one
airport almost qualifies as an urban location; another is within ten
miles of a copper smelter; and the third is within 150 miles of sev-
eral large copper smelters.
-------
Worst-case 90th percentile visibility ranges approximately from 10 to
40 miles among both urban and nonurban locations. Unlike 10th per-
centile and median visibilities, the 90th percentile appears to be
dominated by special meteorological events such as fog and precip-
itation.
No obvious large-scale geographic patterns exist for visibility with-
in the Southwest. With the three exceptions noted above, visibility
tends to be quite good at nonurban airports throughout the study
region.
The visibility percentiles estimated from the airport data agree very
well with the results of recent special field programs. The airport
data indicate that median visibility is 65 to 80 miles in nonurban
areas; field studies using photographic photometry and integrating
nephelometry find median visibilities of 60 to 80 miles in remote
areas of the Southwest.
Historical Visibility Trends (Chapter 4)
Although still quite good, visibility in the Southwest appears to
have deteriorated significantly over the past two decades. A study
of trends at 4 urban and 8 nonurban airports indicates a distinct
worsening of visual range from the early 1950's to the early 1970's
at all locations but one. The 10th percentile, median, and 90th
percentile visibilities all show a decrease on the order of 10 to 30%.
The historical decrease in visibility becomes even more notable when
one considers that blue-sky scatter (which of course remained con-
stant) is a substantial fraction of extinction in the Southwest. The
visibility trend data indicate that extra extinction (above-and-beyond
blue-sky scatter) increased on the order of 20 to 70% from the early
1950's to the early 1970's.
Although visibility is not uniform over the Southwest, and although
quantitative visibility trends are not identical at all locations,
indications are that visibility has deteriorated on a large spatial
scale throughout the Southwest in urban areas, nonurban areas,
and even very remote areas. From the conclusions of later chapters,
we deduce that the historical decrease in visibility is most likely
due to increases in secondary aerosols, the result of growth in SOx,
N0x» and possibly hydrocarbon emissions from copper smelters, power
plants, automobiles, and other sources. The large spatial scale in-
volved in the visibility deterioration would be due to two factors:
(1) growth of emission sources in many parts of the region, and (2)
the tendency of secondary aerosols to be spread widely because of the
mixing and transport that occurs during the time required for aerosol
formation.
-------
Visibility/Pollutant Relationships (Chapter 5)
It is possible to complete visibility/pollutant regression models
using two data bases for Phoenix and one for Salt Lake City. The
multiple regressions between airport visibility data and Hi-Vol
particulate data attain total correlation coefficients as high as
0.87 in Phoenix and 0.81 in Salt Lake City.
Blue-sky scatter accounts, on the average, for approximately 17% of
total extinction in Phoenix and for approximately 13% in Salt Lake
City. The regression models for the three data bases indicate that
secondary aerosols, particularly sulfates, dominate extra extinction
(above-and-beyond blue-sky scatter). Among the three regression
models, estimates of average contributions to extra extinction range
from 32% to 53% for sulfates, 23% to 37% for nitrates, and 0% to 35%
for the remainder of TSP.
The extinction coefficients per unit mass for sulfates, nitrates, and
the remainder of TSP that we estimate by the regression models are in
agreement with other values in the published literature and with known
principles of atmospheric physics. In particular, there is agreement
that secondary aerosols, which tend to reside in the .1 to 1 micron
size range, exhibit one order of magnitude greater extinction coef-
ficient per unit mass than fugitive dust, which tends to have a much
larger particle size.
By using generalized extinction coefficients per unit mass for sulfates,
nitrates, and the remainder of TSP, our visibility/pollutant model can
be extended to remote areas. Based on NASN data from three national
park locations, we predict that average visibility in remote areas
should be approximately 70 to 80 miles; this is in agreement with ob-
served visibility. Blue-sky scatter accounts for approximately 45% of
total extinction in remote areas. We conclude that two-thirds of the
extra extinction at the national park sites is due to sulfates, with
one-sixth due to nitrates and one-sixth due to the remainder of TSP.
Although they appear to dominate visibility reduction, sulfates and
nitrates constitute only a relatively small fraction of total aerosol
mass. Sulfates and nitrates account for about 10% of total aerosol
mass in urban areas and 20% of total aerosol mass in nonurban areas
of the Southwest.
The Copper Strike of 1967-1968 (Chapter 6)
In the late 1960's copper smelters were the dominant source of region-
wide SOX emissions in the Southwest. During the nine-month, industry
wide copper strike of 1967-1968 sulfates dropped by 38 to 76% (com-
pared to seasonal averages in surrounding years) at the five NASN sites
located within 70 miles of copper smelters. Sulfates also dropped 60%
at Grand Canyon and 57% at Mesa Verde; these two national parks are
-------
located approximately two to three hundred miles from the main group
of smelters.
Visibility improvement during the strike ranged from approximately 5
to 25% at locations within 150 miles of copper smelters. The decrease
in extra extinction (calculated from the visibility changes) was 10
to 30% at those locations.
Our visibility/pollutant regression models for Phoenix and Salt Lake
City are verified by changes during the strike. Predicted decreases
in extinction (or increases in visibility), based on the sulfate re-
ductions and the regression models, agree with actual decreases in
extinction during the strike. The visibility/pollutant model extend-
ed to nonurban areas does not appear to be verified by the strike
changes. Several possible reasons for this discrepancy are discussed.
t The sulfate reductions and visibility increases during the strike are
statistically very significant. An examination of meteorological data
for the strike period indicates that weather patterns were not notably
different from normal. If anything, pollution potential appears to
have been slightly greater than average during the strike. We con-
clude that meteorology was not a major factor in producing the ob-
served air quality improvements.
LIMITATIONS OF THE ANALYSIS
We would expect a priori that the main limitation to our analysis would
be the quality of the airport visibility data. The principal use of the visi-
bility observations is air traffic control; for this purpose, low visibilities
(i.e. less than 7 miles) are most critical. We would not expect extreme care
to be taken with observations of visual range on the order of 20 to 100 miles
(which constitute the preponderance of actual visibilities in the Southwest).
In actuality, we found most of the visibility data to be of good quality.
The quality of the data is evidenced by the consistency of the cumulative
frequency distributions from one airport to another. The frequency distribu-
tions of airport visibility have also been shown to be consistent with the re-
sults of special field programs using photographic photometry and integrating
nephelometry. The quality of the data is further evidenced by the high cor-
relations (from 0.68 to 0.87) obtained between airport visibility measurements
*
5
-------
and Hi-Vol participate data. This surprisingly good data quality may be, in
part, due to our airport survey; observation practices were screened before
airports were selected for the study.
In our analyses of historical visibility trends, a question arises con-
cerning the possibility that errors may have been introduced by changes in
airport personnel, observation sites, and/or reporting practices. A careful
survey was conducted at each airport in an attempt to eliminate such errors.
If any undocumented procedural changes have occurred which affect visibility
trends, it is expected that they would introduce random errors and would not
bias our overall conclusions. Because of the consistency of the downward visi-
bility trends at various airports, we are confident in our conclusion that
visibility has deteriorated significantly over the Southwest during the past
two decades.
The regression models relating visibility to particulate measurements
involve several limitations. These limitations, discussed in Chapter 2,
include the following:
spatial nonhomogeneity of the atmosphere and consequent differences
between measured pollutant levels at the Hi-Vol site and average
pollutant levels over the visual range.
statistical difficulties introduced by intercorrelations among the
independent variables.
the possibility that the independent variables may act as surrogates
for pollutants that are not included in the analysis.
potential errors in measurement techniques for sulfates, benzene
solubles, and (especially) nitrates.
In spite of these potential limitations, the results of our regression models
are consistent with the published literature, known principles of aerosol
physics, and air quality changes during the 1967-1968 copper strike. These
consistencies lend credence to our conclusions concerning the causes of haze
6
-------
in the Southwest.
Perhaps the most important limitation in the visibility/pollutant analysis
involves the treatment of nitrates. Nitrate measurements are known to be sub-
ject to interferences, and nitrates may be partially acting as surrogates for
related photochemical contributors to haze that are not included in the anal-
ysis. In order to account for these problems, it may be best to regard the
nitrate variable as representing not only nitrate aerosols but also related
photochemical pollutants such as NOg and secondary organic aerosols.
FUTURE WORK
In this investigation of the historical data base, we have sought answers
to very basic questions: What are existing visibility levels in the Southwest?
Has visibility changed significantly over the past 25 years? What are the main
contributors to haze in the Southwest? There are other more detailed questions
that may be answered by analyzing the historical data. The potential infor-
mation available in over twenty-five years of airport data and more than ten
years of NASN measurements should not be neglected in future studies of
visibility.
In characterizing existing visibility levels, we have examined only the
overall, yearly frequency distribution of visual range. One could very easily
disaggregate the data and examine seasonal, weekday/weekend, and diurnal pat-
terns in visual range. Visibility observations could be correlated with
meteorological data to determine how meteorological parameters or weather pat-
terns affect visibility. Correlations could be run among visibility measure-
ments at various sites to ascertain the spatial scale of day-to-day visibility
changes.
-------
Similarly, the analysis of historical visibility trends could be per-
formed in more detail. The trends could be disaggregated according to seasons}
wind conditions, or meteorological classes. The trend analysis would benefit
by the application of more sophisticated statistical methods. It would also
be interesting to investigate the possible occurrence of long-term
meteorological changes.
We have not been able to find coincident data on visual range and aerosol
composition in remote areas. When such data become available, visibility/
pollutant regressions should be performed to test our conclusions concerning
the main contributors to extinction in nonurban areas.
-------
CHAPTER 2
DATA BASE PREPARATION AND DATA ANALYSIS METHODS
The objectives of this report are to document the historical trends of
*
visibility in the Southwest and to characterize the relationships between
visibility levels and pollutant concentrations. Before presenting our find-
ings, it is worthwhile to summarize the data bases and statistical methods
which serve as the foundation for those findings. This chapter describes
the data bases used and the analysis methods applied.
AIRPORT WEATHER DATA AND NASN POLLUTANT DATA
Two types of data are used in this study: airport weather data (includ-
ing measurements of visibility or visual range) and National Air Surveillance
Network (NASN) particulate data. The airport weather data provide infor-
mation on historical changes in visibility. The airport data and NASN data
are combined to investigate the relationship between visibility and pol-
lutant levels. Before any analyses were performed on these data sets, tele-
phone surveys were conducted at each airport and pollutant monitoring site
to uncover potential problems in the data.
Survey of Airport Weather Stations
The visibility data presented in this report consist of daytime "pre-
vailing visibility" observations made by airport meteorologists. According
to National Weather Service procedures, prevailing visibility is defined as
the greatest visual range that is attained or surpassed around at least half
the horizon circle, but not necessarily in continuous sectors (Williamson,
*
In this report, the terms "visibility" and "visual range" will be used
interchangeably. Both will refer to the distance at which a black ob-
ject can just be distinguished against the horizon sky.
-------
1973). Daytime visibility is measured by observing markers (e.g. buildings,
mountains, towers,etc.) against the horizon sky; nighttime visibility measure-
ments are based on unfocused, moderately intense light sources. Airport
meteorologists perform visibility measurements each hour. In recent years,
only the readings from every third hour are entered in the National Climatic
Center (NCC) computerized data base.
NCC has compiled computerized data records for 105 airport weather stations
*
in Arizona, Colorado, New Mexico, Utah, Nevada and Wyoming. At only 45 of
these stations do the computerized records cover a sufficient time period to
be useful in this study, and only 35 of those stations are located in areas
of prime interest. Appendix A provides a listing of the weather stations and
distinguishes the 35 sites of potential utility to this study.
A detailed telephone survey was conducted with the meteorologists at
each of the 35 airports. The purpose of the survey was to ascertain the over-
all quality of the visibility measurements, the utility of the measurements
for historical trend studies, and the usefulness of the measurements for vis-
ibility/pollutant analyses. The questions contained in the survey were as
follows:
What are the farthest daytime visibility markers in various dir-
ections? (List directions and marker distances.) Over what per-
centage of the horizon does the observer have an unobstructed view
to distant markers?
Do the daytime markers generally meet the criteria of a black ob-
ject against the horizon sky?
Are visibility measurements made during the night as well as dur-
ing the day? If so, what are the farthest markers at night?
Are the observations made at ground level? If not, at what height
above the ground?
t How many members does the observation team include? Has there been
a major discontinuity in the observation team?
*
Weather data are also available at other airports, but these data are
not in computerized form.
10
-------
t Have the observation techniques, reporting practices, or observation
site changed significantly in the last 30 years?
Are the visibility measurements significantly affected by very localized
pollution sources?
(If applicable) Are the visibility observations generally representative
of the air mass and visibility at the nearby NASN site?
Does the meteorologist have any comments or recommendations with regard
to our trend studies or visibility/pollutant analyses?
The telephone survey led to several important discoveries. For instance,
we found that nighttime and daytime visibility measurements are not compatible
in the Southwest. Typically, the most distant daytime visibility marker
is on the order of 30 to 100 miles, while the most distant nighttime marker
is on the order of 10 to 20 miles. We decided to restrict all of our analyses
to the four daytime measurements ( i.e., 8 AM,11 AM, 2 PM, and 5 PM).
We also found that the farthest daytime marker at some locations was only
at a distance of 10 to 40 miles and that this farthest marker was reported
over 90% of the time. Using these sites to characterize visibility would
be similar to conducting a traffic study with a speedometer that did not
read above 15 mph. Thus, we eliminated many sites on the basis that the far-
thest marker was at too short a distance.
At a few locations we found that the observation site or reporting
practices had changed significantly during the period of interest. Some
of these sites we eliminated from our analysts; others we retained by restrict-
ing our analysis only to periods of consistent observation procedures.
Our survey of airport weather stations led us to select seventeen locations
for the study; these locations are illustrated in Figure 1 and listed in
Table 1. Table 1 also summarizes the types of analyses to which the data
11
-------
M
CO
t/1
OJ
3
O
O)
o
0)
CO
CO
c
O
r
-M
03
QJ
(O
0)
s-
o
Q.
12
-------
>-
h-
00
>-
1
1 1
1
KH
CO
1 1
oo
1 1
1
oo
UJ
nr
h-
ra
O
00
UJ
_i_
I
2:
i)
Q
UJ
O
1
1 1
oo
o
1 1
r
Jf:
oo
C£.
UJ
-_
1
1
<:
UJ
3
I
Qi
0
d.
o:
i i
<:
i i
LU
I
CO
1
oo
UJ
00
1
z:
**
1 1
0
oo
UJ
CL-
>
1
J"
4^
o
1 1
S-
O O)
CD CL
M- Q-
4- 0
UJ <_>
^
(O
^
O CO
Q. Q.
">i'-C
i- E
f O
r- *r~
O +J
r- ro
CO i
r- O)
> a:
co
-o
cu
S- >>
1 -M
r
i i
0 S
i 'r
S- tO
O !-
CO
r- C
3: -
CO
c
a>
-M
rO
D-
r +->
ro -
O r
C"" ^")
Q-'i-
ro co
S- -r-
CT> ^>
O
0> c:
CJ3 -i-
X X X X X XX
X X X X X
xxxxxxxx
xxxxxxxx
N -r-
1- i-
s- <: o
ff r O
0
N <'-*> O "
p- ^-~ . ^-- O
S- . ) "CO
<£ S_ 4-> to O
" re N c: c , i- CD -i S_ Q-'i
ZJJ^et+jS-Q-rO-U
-C 00 C O
(_> »,^, OO C
rtJ -»-> « O 3
3 X -M S T3 O
DC T- O C O rO S-
c o o r- s_ cu -o
-PC
s-ooucr era
U.Q.Q-H-SOQC'J
xxxxxxxx
X X
XX XX
xxxxxxxxx
.;
0)
2T
* r~"
* S, (O
c +->
(O . . ZO
i-srs:
S- «
(0 >>-£= -
O O -C T- ^-> >,
r ^jT «« «v (^j Q^ *"") ^g
o -^ o c +J
o 'toTJOr^cu >*
> ra s- -t-> -it $- a>
N rjj CT) O O} ** rO d) C
o z ai g> c >>_i > c
i > O" T fO O CD
> to ro S- O>r C CU
Z3 r"^ /O r~" (& ^3 fO QJ ^~
f\ ^J m_J
-------
are applied. All seventeen stations are used to characterize present geographical
patterns in visibility; data from certain stations are not useful for some of the
other analyses.
We have classified the Phoenix, Tucson, Denver, and Salt Lake City air-
ports as urban, and the rest of the airports as nonurban. Here, urban is de-
fined as "in or near a city with population exceeding 150,000."*
Survey of NASN Monitoring Sites
There were several locations (Grand Canyon, Phoenix, Tucson, Denver,
Mesa Verde, Ely, Albuquerque and Salt Lake) where we hoped to link NASN pollutant
data with airport visibility data in order to study the visibility/pollutant
relationship. For these locations, we contacted the local monitoring agencies
which operate the NASN samplers. The purpose of these contacts was to assess
the utility of the NASN TSP (total suspended particulates) data for visibility/
pollutant studies.
The survey of the NASN TSP monitoring sites included the following
questions:
a How long has the TSP Hi-Vol been operated? Has it been relocated?
t What is the height of the Hi-Vol above the gound? Is the sampler
exposed to air flow in all four directions?
Is the sampler exposed to significant local sources of dust (e.g.,
unpaved roads)?
The urban/nonurban classification is complicated by the fact that most of the
airports are on the outskirts of the cities or towns. The selection of a cut-
off point at a city population of 150,000 is empirical; this is the cut-off
point at which we observed a significant change in median visibilities. Our
"nonurban" category actually includes a range from "suburban" to "nearly remote'
With only three exceptions, however, all the nonurban sites have about the
same visibility frequency distribution; this visibility frequency distribution
appears to be the same as the one observed at very remote locations of the
Southwest (see Chapter 3 for a discussion).
14
-------
Is the NASN site representative of area-wide pollution levels?
In particular, is it representative of the air mass at the NCC
site?
Are there any suggestions or comments in regard to our visibility/
pollutant studies?
From our survey of airport observers and NASN monitoring agencies we
found that several locations were not appropriate for the visibility/pollutant
studies. The main reasons were either that the airport and NASN sites were
not representative of one another or that the visibility data were deficient.
Phoenix and Salt Lake City were the only two sites where conditions for the
study appeared quite good. Several other locations (See Table 1) were poss-
ibilities that at least merited an attempt to analyze the data.
The NASN pollutant data were also used to investigate air quality
changes during the 1967-1968 industry-wide copper strike. Figure 2 shows
the NASN locations that provided data for documenting the effects of the
copper strike.
Initial Data Processing
For each of the airport locations studied, complete tapes of all surface
weather data were obtained from NCC in the CD-144 format. These tapes were
processed to extract data for the four daytime hours (8 AM, 11 AM, 2 PM, and
5 PM). With data for these hours, we formed a "processed visibility data
base" for each location; this data base included the date, hour, visibility,
relative humidity, and special notations (storms, liquid precipitation, frozen
precipitation, fog, blowing dust, smoke, haze, etc.).
For some years, the original NCC tapes contained data for every hour
rather than every third hour. For consistency, we extracted the same four
daylight hours in all years.
15
-------
1/5
Ol
13
O
GO
01
c.
c
O
GO
CM
(U
s_
16
-------
The nationwide NASN data for TSP, sulfate, nitrate, benzene solubles,
etc. were obtained in tape form from EPA's SAROAD data bank. We reorganized
the original EPA data to create a "processed pollutant data base." For each
site, this data base listed the date and the various pollutant measurements
in a consistent, easy to access, format.
In order to investigate visibility/pollutant relationships for certain
locations, the "processed visibility data base" was combined with the "processed
pollutant data base" at these locations. The resultant data base listed, for
each day, the 24 hour average pollutant concentrations and the daytime averages
of visibility and relative humidity.
FREQUENCY DISTRIBUTIONS OF VISIBILITY DATA
Because of the nature of the reporting methods, visibility data are
most appropriately summarized by cumulative frequency distributions of the
form "percent of time visibility is greater than or equal to X miles."
Figures 3 and 4 present recent cumulative frequency distributions for all the
sites studied. Figure 3 is for urban locations; Figure 4 is for nonurban sites.
When analyzing cumulative frequency distributions for visibility, it is
important to use only those visibilities that are routinely reported by the
observer. For instance, it is not uncommon to see the following type of
situation:
*
When an airport observer reports a visibility of X miles, this usually
means that visibility is at least X miles, not that visibility is exactly X
miles. This method of reporting is especially significant in the case of the
farthest marker. Airport observers in the Southwest do not report visibilities
greater than their farthest marker even if visual range obviously extends
beyond the farthest marker.
17
-------
60 -
80-
-------
100-
100,
80-
Cheyenne 1970-1972
Grand Junction 1970-1972
Pueblo 1966-1968
20
10% 20 30 40 50 60 70 80 90 100X
Cumulative Frequency (percent)
Figure 4. Cumulative frequency distributions of
visibility at nonurban locations.
19
-------
120.
100.
8Cf-
'i 60-
i/i
s-" 20'
100-
80-
-------
100 J
80 -
OJ
I
60 -
?0 -
10
OJ
.a
l/t
80'
60
40'
20-
80 -
60 -
40 -
20 -
Winslow 1970-1972
~i 1
107o 20 30 40 50 60 70 M) 90 100%
Fort Huachuca 1968-1970
IQ% 20 30 40 50 60 70 80 90 100%
Ely 1970-1972
20 30 40 50 60 70 80 90 100%
Cumulative-Frequency (percent)
Figure 4. Cumulative frequency distributions of
visibility at nonurban locations. (Cont'd)
21
-------
80-
J=_
£ 40-
100-
^ 80-
V>
1 60-
£ 40-
Wendovcr 1970-1972
1C%
60 70 80 90 100%
Las Vegas 1970-1972
10% 20 30 40 50 60 70 8r. 90 100%
Dugway 1967-1969
eT sb' 45 55" 66" To So 5o u
Cumulative Frequency (percent)
Figure 4. Cumulative frequency distributions of
visibility at nonurban locations. (Cont'd)
22
-------
Farmington 1970-1972
20 30 40 50 60 70 80 90 1002
Cumulative Frequency (percent)
Figure 4, Cumulative frequency distributions of
visibility at nonurban locations. (Cont'd)
23
-------
Visibility % of Time Reported Cumulative Frequency
80 miles 10% 10%
70 miles 10% 20%
65 miles .1% 20.1%
60 miles 9.9% 30%
50 miles 10% 40%
In this case, the 65 mile recordings produce a "kink" in the cumulative
frequency distributions. It is obvious, in this case, that the 65 mile
visibilities are not routinely reported, but they happened to be recorded
a few times by a member of the observation team. In our analysis of frequency
distributions for visibility data, we took care to use only those visibilities
that are routinely reported.
The graphs in Figures 3 and 4 illustrate a property that we found to be
nearly universal among the sites studied. The cumulative frequency distribution
tends to be nearly linear at the higher visibilities (i.e., the lower percentiles).
In many cases we have used this property to calculate the 10th percentile of
visibility even if the actual recordings started at a higher percentile (e.g.,
the farthest marker might be reported 20% or 30% of the time). This calculation
was done by linear extrapolation of the cumulative frequencies for the two
farthest markers. The extrapolation is indicated by the dashed lines in Figures
3 and 4.
In this report, historical trends in visibility are based on changes in
visibi1ity percenti1es. In most cases we use the 10th percentile (best conditions),
the 50th percentile (median), and the 90th percentile (worst conditions). This
method of reporting visibility trends differs from the traditional method
24
-------
Each of the components of B should be directly proportional to aerosol
or gas concentrations (assuming other factors such as light wavelength, aero-
sol size distribution, particle shape, and refractive index remain constant).
In polluted urban air, it is thought that aerosol light scattering (B .)
scar
tends to dominate over the other contributions to the extinction coefficient
(Charlson, 1969).
Slight transformations are also performed on the independent variables.
SULFATE and NITRATE are defined as 1.3 S0| and 1.3 N0§ (White and Roberts 1975) in
order to account for the mass of the cations (presumably ammonium) associated with
the measured values of $04 and NO" . The variable, TSP - SULFATE - NITRATE,
*
is used to represent the non-sulfate, non-nitrate fraction of TSP.
Multi-Variate Regression
When several independent variables (TSP, RH, SULFATE, and NITRATE) are
affecting a dependent variable (B) it is important to perform a multi-variate
analysis that can separate out the individual impact of each independent
variable, discounting for the simultaneous effects of other independent
variables. Uni-variate analyses, based on simple one-on-one relationships,
can lead to spurious results because of intercorrelations among the independent
variables. For instance, in some cases we found that TSP apparently correlated
with B only because TSP is correlated with SULFATE and NITRATE which are in_
turn significantly related to B.
An appropriate tool for multi-variate analysis is multiple regression.
Following the procedure of Cass (1976) and White and Roberts (1975), we perform
multiple linear regressions of the form:
When benzene solubles are also included in the analysis, this variable
is defined as TSP-SULFATES-NITRATES-BSOL.
27
-------
B = a + b,(TSP-SULFATE-NITRATE) + b, RH + bo SULFATE + b. NITRATE . (3)
1 e. o q
These regressions are run stepwise, retaining only those terms which are
significant at a 95% confidence level. The regression coefficients (bj,
b3, and b4) represent the extinction coefficient per unit mass for each
pollutant species, in units of (104 meters)"1/(yg/m3).
We also perform regressions which include relative humidity effects in
a nonlinear manner. Cass (1976) indicates that light scattering by a sub-
RH a
micron, hygroscopic aerosol might be proportional to (1 - y7yQ) , where the
exponent « is expected to occur in the range -0.67 to -1.0. To
account for this type of effect, we attempt regressions of the form
B = a + b TSP-SULFATE-NITRATE + fa SULFATE + b NITRATE (4)
loo ^o "loo
Average Extinction Budget
The regression equations can be used to compute the fraction of visi-
bility loss, on the average, that is due to each pollutant species. These
calculations are best illustrated by an example.
Using the Maricopa County pollution data, our regression equation for
Phoenix reduces to the following:
B = .22 + .037 SULFATE + .051 NITRATE , (5)
with a total correlation coefficient of 0.87. The average value for the
extinction coefficient at Phoenix is .85 [10^ meters]'1, corresponding to a
visibility of 29 miles. Using Eq. (5), the average extinction (haze) level
at Phoenix can be disaggregated as follows:
28
-------
Average Sulfate Average Nitrate
t t
.85 = .15 + .07 +^.037 (10.1 yg/m3)/v.Q51 (5.1 ug/m3)v
y Y /
/ t
Blue- Re- Contribution Contribution
light mainder of sulfates of nitrates
scatter of .22
by air constant
molecules term
or .85 = .15 + 0.7 + .37 + .26 (6)
Equation (6) indicates that, on the average, 44% of the extinction is from
sulfates, 31% is from nitrates, 17% is from air molecules, and 8% is un-
accounted for.
Alternately, we could examine only the extra extinction above-and-beyond
the blue-light scattering by air molecules. In the Phoenix example, the extra
extinction is .70, (0.85 - 0.15), with 53% from sulfates, 37% from nitrates,
and 10% unaccounted for.
Limitations of the Regression Approach
There are several limitations in using regression equations to estimate
the contribution of various pollutants to visibility loss. These limitations
are best explained with reference to the Phoenix example.
Our analysis implicitly assumes that the air mass is of uniform compo-
sition over the entire visual range. In actuality, we would expect that the
average pollutant concentrations over the entire visual range are lower than
the concentrations measured at the downtown Phoenix monitor (Moyers et al. 1977)
The relatively high pollutant measurements may cause us to underestimate the
extinction coefficients per unit mass for sulfates and nitrates.
29
-------
Conversely, the statistical regression equation may overstate the importance
of sulfates and nitrates if these variables are correlated with other pollutants
which are not included in the analysis. Our intuitive opinion is that this problem
is greater for nitrates than for sulfates. Nitrates may in part be acting as a
surrogate for other related photochemical pollutants: secondary organic aerosol
and nitrogen dioxide.
Potential errors in measurement techniques also raise a caution flag.
Artifact sulfate (formed by S02 conversion on the measurement filter) may cause
us to underestimate slightly the extinction coefficient per unit mass for sulfates.
In cases where we have included benzene soluble organics, we should note that
the extraction efficiency for secondary organic aerosols is poor with the benzene
method (Grosjean 1974). Perhaps the greatest measurement concern, however,
involves nitrates (Spicer and Schumacher 1977). Because of potential difficulties
in nitrate measurements, it might be best to regard the nitrate variable as a
representation of ammonium nitrate, nitric acid, N02» and other related photo-
chemical products rather than a measure of nitrate aerosols only.
A final difficulty in the regression analysis is the problem of colinearity,
i.e. the correlations that exist between the "independent" variables (sulfates,
nitrates, TSP, and relative humidity). Although multiple regression is designed
to estimate the individual effect of each variable, discounting for the simul-
taneous effects of other variables, the colinearity problem can still lead to
distortions in the results. These distortions should not be too great in our
study, however, because the intercorrelations among the independent variables are
not extremely large (typically 0.3 to 0.5).
Although the regression models are subject to several limitations, it should
be noted that the results of these models have proven to be quite reasonable.
30
-------
Chapter 5 demonstrates that our results are consistent with the published
literature and with known principles of aerosol physics.
ALLOWANCES FOR METEOROLOGY
Special meteorological events, such as fog or precipitation, can have
substantial effects on visibility. It is important to discount for these
special weather conditions in the visibility/pollutant regression studies.
Regression, based on minimization of squared errors, is very sensitive to
outliers in the data, and the extremely low visibilities (extremely high
extinction coefficients) associated with special weather conditions can dominate
the results of the regressions. Accordingly, we have eliminated all days with
foc^or precipitation in our visibility/pollutant regression analysis.
We have also investigated the effect of sorting for meteorology in
our analysis of historical visibility trends. In this case, we have
eliminated all hours with RH>70%, fog, or precipitation. Figures 5 and 6
present some typical results.
As indicated in Figures 5 and 6, sorting for special meteorology has
a very slight impact on 10th percentile visibility, a slight impact on 50th
percentile visibility, and a substantial impact on 90th percentile visibility.
From several test cases, we have concluded that sorting for meteorology will
have essentially no effect on our conclusions concerning historical trends
(changes) in 10th and 50th percentile visibilities. Sorting for meteorology
can sometimes have significant effects on conclusions concerning trends in
90th percentile visibility.
The historical trends presented in this report (Chapter 4) will be
based on all the data, without sorts for special meteorology. The main
31
-------
Hours of Fog, Precipitation, and High RH Deleted
All Hours
CO _
50 -
40 -
30 -
£ 20 _|
-Q
10 J
to
O)
10th Percentile
N
50th Percentile
90th Percentile
1950
1955
1960
1965
1970
YEAR
Figure 5. Historical visibility trends at Salt Lake City
32
-------
Hours of Fog, Precipitation, and High RH Deleted
All Hours
O)
100 -
90 -
80 -
70
60
50
30 -
20 -
10
\ 10th Percentile
t
\
50th Percent!le
90th Percent!le
1950
1955
1960
1965
I ' '
1970
YEAR
Figure 6. Historical visibility trends at Denver.
33
-------
caveat that should be kept in mind is that the trends in 90th percentile
visibility might be affected by special weather events such as fog or pre-
cipitation. The trends in 50th and 10th percentile visibilities should not
be affected by such weather events.
In future studies, it may prove interesting to perform more detailed
analyses of visibility trends; the data could be sorted by season, relative
humidity, wind direction, precipitation conditions, etc. Our trend analysis,
based on all measurements each year, is intended to document only the gross,
overall history of visibility in the Southwest.
34
-------
CHAPTER 3
EXISTING VISIBILITY LEVELS
A very basic issue that needs to be resolved is "what are existing visi-
bility levels in the Southwest?" Specifically, we would like to quantify
visual range on average days, "best-case" days, and "worst-case" days; we would
also like to know if any large-scale geographic patterns exist in visibility
within the Southwest. These questions can be answered by an analysis of air-
port visibility measurements.
VISIBILITY IN URBAN AND NONURBAN AREAS
Figures 3 and 4 (in Chapter 2) presented recent cumulative frequency dis-
tributions for visual range at four urban locations and thirteen nonurban lo-
cations. From these figures, one can easily read the 10th percentile (best),
50th percentile (median), and 90th percentile (worst) visibilities for each lo-
cation. These percentile visibilities, rounded to the nearest five miles to re-
flect uncertainties, are listed in Table 3. It should be remarked that some of
these visibility percentiles have been estimated by linearly extrapolating cum-
ulative frequency distributions; some of the other percentiles have been esti-
mated by comparing the "upper-end" of the frequency distribution among locations.
Table 3 indicates that visibility in the Southwest tends to be quite
*
good, at least relative to visibility east of the Mississippi River . Median
visibility is on the order of 30 to 55 miles in and near urban centers
*
Median visibility in the Northeast is approximately 10 miles at urban
locations and 12 miles at nonurban locations (Trijonis and Yuan 1977).
35
-------
TABLE 3. MEDIAN, BEST-CASE, AND WORST-CASE VISIBILITY
LEVELS AT SEVENTEEN LOCATIONS IN THE SOUTHWEST
LOCATION
URBAN
Phoenix, Ariz.
Tucson, Ariz.
Denver, Col .
Salt Lake City, Utah
NONURBAN
Fort Huachuca, Ariz.
Prescott, Ariz.
Wins low, Ariz.
Colorado Springs, Col.
Grand Junction, Col.
Pueblo, Col.
Ely, Nev.
Las .Vegas, Nev.
Alamogordo, N.M.
Farmington, N.M.
Dugway, Utah
Wendover, Utah
Cheyenne, Wyo.
YEARS
OF DATA
1970-1972
1970-1972
1970-1972
1970-1972
1968-1970
1970-1972
1970-1972
1964-1966
1970-1972
1966-1968
1970-1972
1970-1972
1968-1970
1970-1972
1970-1972
1970-1972
1970-1972
VISIBILITY PERCENTILES
10th% (Best)
60 miles
65*
*
70
45
70 miles
**
no
**
no
*
115
*
no
90
**
70
80*
*
95
**
no
*
90
**
no
*
100
50th% (Median)
35 miles
55
50
30
50 miles
**
80
**
80
75
80
75
*
50
60
65
**
80
65
**
80
65
90th % (Worst)
20
40
10
10
30
40
25
20
40
20
20
35
35
40
20
40
15
miles
miles
**
Estimated by linearly extrapolating the cumulative frequency distribution
for this site
t
Estimated by comparing frequency distribution to distributions for other
locations
36
-------
(those in which population exceeds 150,000) in the Southwest. Median
visibility tends to be around 65 to 80 miles at most sites away from large
urban centers. For comparison, we note that an atmosphere containing air
molecules only (with absolutely no aerosols and with no light-absorbing
gaseous pollutants) would exhibit a visual range of approximately 160 miles.
The only nonurban airports exhibiting median visibilities less than
65 miles are Las Vegas (60 miles), Ely (50 miles), and Fort Huachuca (50 miles).
The Las Vegas airport almost qualifies as an urban location; Ely is within
10 miles of a copper smelter; and Fort Huachuca is within 150 miles of several
copper smelters.
Best 10th percentile visibility ranges from 45 to 70 miles among the
four urban locations and from 90 to 115 miles at most nonurban locations.
Again, the three exceptions among the nonurban locations are Las Vegas, Ely,
and Fort Huachuca, with 10th percentile visibilities of 80, 70, and 70 miles,
respectively.
Worst-cast 90th percentile visibilities range from 10 to 40 miles at
the urban locations and from 15 to 40 miles at the nonurban locations. The
site-to-site variations in 90th percentile visibility are not always consis-
tent with the site-to-site variations in 10th and 50th percentile visibility.
Unlike the 10th and 50th percentiles, the 90th percentile appears to be domi-
nated by special, localized meteorological events such as fog and precipitation.
GEOGRAPHICAL PATTERNS IN VISIBILITY
Figures 7, 8, and 9 present maps of 50th, 10th, and 90th visibility per-
centiles, respectively. The visibility percentiles are plotted at the locations
of the airports; circled numbers represent urban locations.
37
-------
CO
O)
co
QJ
C
-------
E
c:
in
OJ
O)
OJ
O
O)
Q.
OJ
o
o
3
jQ
O
i
-C
Q.
fO
CD
O
0)
00
3
O1
cn o
O) re
CO C-
o
39
-------
OJ
OJ
O
OJ
Q.
O
CT>
S-
o
O
O
JQ
ll
(T3
O
Q.
fO
CD
O
OJ
cu
40
-------
The main conclusion evident in Figures 7,8, and 9 is that no dominant
large-scale geographic patterns exist for visibility within the Southwest.
With the three exceptions noted earlier, visibility tends to be quite good
at nonurban sites throughout the study region. There is, of course, an ob-
vious difference between urban and nonurban locations.
CONSISTENCY WITH RECENT FIELD PROGRAMS
The quality of the airport visibility data is evidenced by the consis-
tency between the airport observations and special field programs. Based on
the airport data, we have concluded that median visibility in nonurban areas
of the Southwest is 65 to 80 miles. Using photographic photometry, Roberts
et al. (1975) found a median visibility of 69 miles in the Painted Desert,
Arizona. The C-b Oil Shale Venture (1977) reported a median visibility of
79 miles based on photographic photometry measurements in a remote area of
northwest Colorado. Using an integrating nephelometer, Tombach and Chan (1977)
determined a median visibility of 58 miles in remote northeast Utah.
The airport data indicate that 10th percentile visibility ranges from
90 to 115 miles in nonurban areas of the Southwest. Roberts et al. (1975)
measured a 10th percentile of 98 miles in the Painted Desert; the C-b Oil
Shale Venture (1977) reported a 10th percentile of 115 miles in remote NW
Colorado. The only inconsistency is that Tombach and Chan (1977) found a
10th percentile of 146 miles (very near Rayleigh scattering) in remote NE
Utah.
41
-------
CHAPTER 4
HISTORICAL VISIBILITY TRENDS
It is a common public opinion in the Southwest that haze
levels have increased substantially during the past two decades. This per-
ception may be due to actual increases in pollution, or it may be due to
nostalgia. The airport observations offer a unique opportunity to check
whether or not visibility really has deteriorated. This chapter uses the
airport data to document changes in visibility from the late 1940's to the
early 1970's.
The trend study described here examines only gross, overall changes
in yearly visibility indices. In future work, it may prove useful to in-
vestigate visibility trends in finer detail. The trend analysis could be
stratified by season or by meteorological class; air quality and emission
changes could be documented in order to help explain visibility trends at
each location; more work could also be done to identify changes in observa-
tion practices that might have affected apparent trends. The present study,
however, is directed only at answering the general question: "Has visibility
changed in the Southwest during the past two to three decades? If so, by
approximately how much?"
YEAR-TO-YEAR VISIBILITY TRENDS
Figures 10 through 21 illustrate historical visibility trends at four
urban locations (Figures 10 to 13) and eight nonurban locations (Figures 14
to 21). Trends are presented for the 10th percentile (best visibility),
42
-------
-* Yearly Values
- Three-Year Moving Averages
70_
60 _
ji> 50 _
ji^
$ 40-
I 3°-
>
20 _
10_
10th Percent!le
50th Percent!le
90th Percentile
Ii|iiii|ii 1i|iiii|iiii|i|
1948 1950 1955 1960 1965 1970 1972
Year
Figure 10. Long-term visibility trends at Phoenix.
43
-------
80
70 -
60
50
40-
30-
20-
101
_ Yearly Values
.» Three-Year Moving Averages
90th Percent!le
1 [ I T
1955
| I I I I | till
I960 1965 1970 1972
1948 1950
Year
Figure 11. Long-term visibility trends at Tucson.
44
-------
« Yearly Values
' Three-Year Moving Averages
100.
90-
80-
70-
a) 60-
50.
40 _
30 _
20 _
10th Percentile (Estimated)
A^xA
50th Percentile
90th Percentile
I ' I
1948 1950
1955
1960
Year
1965
i I i I
1970 1972
Figure 12. Long-term visibility trends at Denver.
45
-------
80.
70.
60 _
to
* 50_
i
E
£> 40 _
1 30.
r-
5*
20.
10J
Yearly Values
Three-Year Moving Averages
10th Percentile
50th Percentile
90th Percentile
1948 1950
1955
1960
1965
' 1 i |
1970 1972
Year
Figure 13. Long-term visibility trends at Salt Lake City.
46
-------
Yearly Values
Three-Year Moving Averages
80
70
60
i/i
^50-
i
5 40j
E 30 I
20\
10-
10th Percentile (Estimated)
90th Percentile
1948 1950
1955
1960
1965
1970
Year
Figure 14. Long-term visibility trends at Fort Huachuca.
47
-------
- - Yearly Values
Three-Year Moving Averages
I/)
O)
90 -
80
70 _
60
50
40 _|
-0
30 _
20
10
75th Percentile (Estimated)
A
90th Percentile
i > i i I i i i I I i I 1 I I I i^ |Ii
1948 1950 1955 1960 1965 1970 1972
Year
Figure 15, Long-term visibility trends at Prescott,
48
-------
- Yearly Values
- Three-Year Moving Averages
90i
80
70
60
50
40
30
20 -
101
80th Percent!le (Estimated)
90th Percentile
I I'
1948 1950
1955 1960
Year
1965
I I I | I |
1970 1972
Figure 16. Long-term visibility trends at Winslow,
49
-------
-Q
r-
I/)
130-
120_
110.
100_
90_
80_
70_
60
50
40
30
20
- Yearly Values
Three-Year Moving Averages
10th Percentile
\ (Estimated)
50th Percentile
V
/90th Percentile
I
T
1965
1948 1950 1955 1960
Year
Figure 17. Long-term visibility trends at Colorado Springs.
50
1970 1972
-------
. Yearly Values
Three-Year Moving Averages
120.
100-
90-
80_
; 70_
' 60_
£ 50_
CO
40_
30 _
20-
10th Percentile
(Estimated)
50th Percentile
A
90th Percentile
f
1948 1950
1955
1960
1965
n I ' I
1970 1972
Year
Figure 18. Long-term visibility trends at Grand Junction
51
-------
Yearly Values
Three-Year Moving Averages
100
90 _
80 _
70 _
"> 60 _
OJ '
£50-
.f! 40 -
30 _
20 _
10 _
1 ^ I
1948 1950
10th Percent!le
(Estimated)
"* 50th Percentile
..90th Percenti le
F
1965
1955 1960
Year
Figure 19. Long-term visibility trends at Pueblo.
1970 1972
52
-------
80
70 _
60
50
Yearly Values
Three-Year Moving Averages
_a
£ 30
20
80th Percentile (Estimated)
90th Percentile
1950
1955
1965
1960
Year
Figure 20. Long-term visibility trends at Ely.
1970
1975
53
-------
- Yearly Values
- Three-Year Moving Averages
140-
130-
120
110-
100-
90-
-».
3 80-
>>70-
60-
50.
40-
30-
20
10
10th PercentHe (Estimated
50th Percenti le
90th Percenti le
I I 7 I i i i i i i I i i i I i i i | i
1948 1950 1955 1960 1965 1970 1972
Year
Figure 21. Long-term visibility trends at Cheyenne.
54
-------
50th percentile (median visibility), and 90th percentile (worst visibility).
For each year, the percentiles are computed from cumulative frequency plots,
such as Figures 3 and 4, based only on those visibility markers that are
routinely reported. In cases where the data do not permit computation of
the 10th and 50th percentiles, trends are presented for the 75th or 80th
visibility percentiles.
The basic period for which trend data are available is 1948 to 1972.
For several sites, this period was shortened due to unavailability of data,
relocation of observation sites, or changes in observation procedures (e.g.,
changes in visibility markers used).
As evidenced by Figures 10 to 21, many of the sites show some improve-
ment in visibility from the late 1940's to the middle 1950's, followed by
deterioration from the middle 1950's to the early 1970's. The improvement
in the late 1940's and early 1950's most likely occurred because of the
switch to cleaner fuels (from coal to oil and gas) during that period. This
improvement has been noted earlier in the work of Holzworth (1962).
All locations but one show a downward trend in visibility after the
middle 1950's. From analyses presented later in Chapters 5 and 6, we suspect
that much of the deterioration in visibility is related to increases in
secondary aerosols (e.g., sulfates and nitrates). The increases in secon-
dary aerosols would be due to growth in SO , NO , and possibly hydrocarbon
A A
emissions from copper smelters, power plants, motor vehicles and other sources.
A very unusual feature in Figure 19 (for Pueblo, Colorado) deserves
special comment. Pueblo is the only location that does not exhibit a definite
*
We call these aerosols "secondary" because they tend to be formed from
gas-to-particle conversion. However, they may also be partly "primary"
(directly emitted) in the sense that the gas-to-particle conversion occurs
before emission into the atmosphere.
55
-------
downward trend in visibility after the middle 1950's. The 10th percentile
visibility at Pueblo decreases somewhat; the 90th percentile visibility re-
mains about constant; but the median visibility increases substantially!
There appear to be two possible explanations:
Subtle changes may have occurred in reporting practices which
resulted in higher visibilities being recorded when visibility
was in the range 40 to 80 miles. The changes in reporting prac-
tices would be "subtle" because, in examining the data, we could
not find any obvious variations in marker selection, and, in in-
terviewing the observers, we could not document any procedural
modifications.
Local emission reductions may have improved median visibility
levels at Pueblo, while 10th percentile visibility deteriorated
due to emission increases on a regionwide scale. Emission re-
ductions did occur during the period at the largest industry
in Pueblo, CF&I Steel. However, these reductions were not very
great, and it does not seem plausible that they would have lead
to a substantial visibility improvement (Pearson 1977).
Neither explanation is very satisfactory, and the unusual trends in the
visibility data for Pueblo remain somewhat of a puzzle.
NET CHANGES IN VISIBILITY AND EXTINCTION, 1954 TO 1971
Table 4 summarizes net changes in visibility since the middle 1950's.
Visibility has decreased by about 10 to 30 percent at most sites from 1954
to 1971; the deterioration is evident in all three visibility percentiles.
There is considerable variation in the visibility trends from one
location to another. Some of this variation represents actual site-to-site
differences in the rate of visibility deterioration. Some of the variation
We did examine data for one other airport (Las Vegas, McCarran Inter-
national) which exhibited an increasing trend in visibility. The airport ob-
servers at Las Vegas considered this to be an artifact produced by personnel
changes, (Taylor 1977). Thus, we did not include visibility trends for Las
Vegas in this report.
56
-------
TABLE 4. NET PERCENT CHANGES IN VISIBILITY,
1953-1955 TO 1970-1972
LOCATION
URBAN
Phoenix, Ariz.
Tucson, Ariz.
Denver, Col.
Salt Lake City, Utah
NONURBAN
Fort Huachuca, Ariz.
Prescott, Ariz.
Winslow, Ariz.
Colorado Springs, Col.
Grand Junction, Col.
Pueblo,1^ Col.
Ely,1" Nev.
Cheyenne ,"^Wyo.
CHANGES IN
Best (10th
Visibility
0%
N,A,
-22%
-24%
-12%
N.A.
N.A.
-17%
-9%
-9%
N.A.
-28%
THREE- YEAR AVERAGES,
%) Median
Visibility
-23%
-22%
-13%
-27%
-27%
-25%*
**
-17%
-12%
-4%
+35%
**
-42%
-23%
1954 TO 1971
Worst (90th %)
Visibility
-20%
-11%
-29%
-19%
-28%
-21%
-27%
-17%
-3%
0%
-33%
-19%
**
N.A.
Trends for these sites are extrapolated from data covering most,
but not all, of the period 1954-1971.
t
75th percentile visibility instead of median visibility.
t
80th percentile visibility instead of median visibility.
Not available.
57
-------
may also be due to errors in estimating trends at individual locations
(e.g., errors induced by undocumented changes in observation procedures).
Considering the potential for errors, we are more confident of the overall
conclusion (that visibility has generally deteriorated about 10 to 30%)
than we are of the exact visibility changes at the individual locations.
Another way of expressing visibility trends is to compute changes in
the extinction coefficient. Here it is useful to examine only "extra" ex-
tinction, the fraction of extinction above and beyond the constant contri-
bution from blue-sky (Rayleigh) scatter. Given visibility, V in [miles],
extra extinction is computed according to the expression
^- - 0.15
with units of [10 meters]" .
Table 5 summarizes the net changes in extra extinction since the middle
1950's. The increases in extra extinction are relatively greater than the
decreases in visibility because we have subtracted out the constant contri-
bution from blue-sky scattering. At most sites, extra extinction has increased
on the order of 20 to 70% from 1954 to 1971.
One of the most interesting features of the visibility decrease (or ex-
tinction increase) is the large spatial scale that appears to be involved. The
evidence points to the conclusion that visibility has decreased throughout the
Southwestin urban areas, nonurban areas, and even very remote areas. The
ubiquitous nature of the visibility deterioration is indicated by the
following:
58
-------
TABLE 5. NET PERCENT CHANGES IN (EXTRA) EXTINCTION,
1953-1955 TO 1970-1972
LOCATION
URBAN
Phoenix, Ariz.
Tucson, Ariz.
Denver, Col.
Salt Lake City, Utah
NONURBAN
Fort Huachuca, Ariz.
Prescott, Ariz.
Winslow, Ariz.
Colorado Springs, Col.
Grand Junction, Col.
Pueblo,"1" Col.
Ely,"1" Nev.
Cheyenne t Wyo.
CHANGES IN
Best (10th
Extinction
0%
N.A.
+64%
+49%
+25%
N.A.
N.A.
+91%
+39%
+26%
N.A.
+152%
THREE- YEAR AVERAGES,
%) Median
Extinction
+35%
+50%
+22%
+50%
+43%
*
+67%
**
+37%
+24%
+8%
-46%
+97%**
+52%
1954 TO 1971
Worst (90th %)
Extinction
+20%
+17%
+43%
+25%
+48%
+38%
+50%
+21%
+4%
0%
+57%
+29%
^Trends for these sites are extrapolated from data covering most,
but not all, of the period 1954-1971.
75th percentile visibility instead of median visibility.
**
N.A.
80th percentile visibility instead of median visibility
Not available.
59
-------
Nearly every station exhibits a decreasing trend in visual range.
The twelve stations studied include urban and nonurban locations
scattered over the Southwest.
Even though most of the nonurban airports would not qualify as
extremely remote locations, we have found (in Chapter 3) that
present visibility at the nonurban airports is essentially the
same as present visibility measured at remote locations (Roberts
et al. 1975; Tombach and Chan 1977; C-b Oil Shale Venture 1977).
Since visibility has deteriorated at nonurban airports, we deduce
that visibility has also deteriorated in extremely remote areas.
Otherwise we would have to accept the improbable conclusion that,
twenty years ago, visibility at nonurban airports was significantly
greater than visibility in remote areas!
At the nonurban airports, the most frequently used markers are on
the order of 50 to 100 miles. Thus, the scale of the measurements
is itself quite large.
As noted earlier, we hypothesize that much of the visibility deteriora-
tion is due to increases in secondary aerosol concentrations. The large
spatial scale in visibility deterioration is consistent with this hypothesis.
Secondary aerosol concentrations, produced by chemical transformation of
gaseous pollutants, tend to be widely spread because mixing and transport
occur during the time required for aerosol formation. The increased
haze in nonurban areas could be the result of diffuse secondary aerosols
(e.g., sulfates and nitrates), the products of transported, diluted, and
aged gaseous emissions (e.g., SOX and NOX). As will be evident from the
results of Chapter 5, even very small increments in secondary aerosols
can produce noticeable deterioration in the high visibilities characteris-
q
tic of the Southwest; our results indicate that a 1 yg/m increase in
secondary aerosols would reduce a visual range of 100 miles to 85 miles.
60
-------
CHAPTER 5
VISIBILITY/POLLUTANT RELATIONSHIPS
Before control strategies can be planned for maintaining or improving
visibility in the Southwest, we must identify the atmospheric components that
contribute to visibility reduction. This chapter relates airport visibility
measurements to Hi-Vol particulate measurements in order to gain insight as to
the causes of haze in the Southwest. The analyses are based on statistical
techniques discussed in Chapter 2; the basic procedure is to regress daily es-
timates of extinction coefficient against TSP, sulfate, nitrate, and relative
humidity. Potential limitations in this procedure are also discussed in Chapter 2.
PHOENIX, MARICOPA COUNTY DATA
Data for analyzing the visibility/pollutant relationship in Phoenix were
made available through the courtesy of the Maricopa County Health Department.
The data base consisted of visibility and relative humidity measurements taken
by the National Weather Service at Phoenix Sky Harbor Airport, and TSP, $04,
N03, and benzene soluble (BSOL) measurements taken by the Maricopa County Health
Department at the central Phoenix station. The air quality monitor 1s located
approximately two miles northwest of the airport.
*
The data covered the years 1973-1974. Because of the standard procedure of
intermittent Hi-Vol sampling, only 106 days during the two years provided complete
Data were later supplied by Maricopa County for the period 1968 to 1976.
Over these nine years, there appeared to be several step jumps in the composition
data (sulfate, nitrate, and organics), possibly due to changes in Hi-Vol filter
type (Layden 1977). Data for the years 1973-1974 were internally consistent and
were consistent with the NASN data. We therefore restricted the analyses to the
1973-1974 period.
61
-------
data for all five variables. Several of these days were eliminated because fog
and/or precipitation was reported by the airport observer. This left 87 days for
the statistical study.
The type of data used for the study is summarized in Table 2 (page 26). The
sulfate and nitrate measurements were made by the colorimetric and 2-4 xylenol
methods, respectively.
Uni -Variate Analyses
As a first step in the data analysis, we examined the simple relationship
between extinction coefficient (B) and each of the independent variables (TSP,
**
RH, SULFATE, NITRATE, and BSOL) one at a time. Figures 22 through 26 show scatter
plots of B versus each of the other variables. Included in the plots are (least-
square) regression lines. The regression equations and the correlation coeffi-
cients are summarized in Table 6.
TABLE 6. RESULTS OF UNI-VARIATE REGRESSIONS FOR PHOENIX, MARICOPA COUNTY DATA
Correlation t
Variable Coefficient.R Regression Equation Significance level ,F
TSP
RH
SULFATE
NITRATE
BSOL
.31
.34
.81
.71
.31
B =
B =
B =
B =
B =
.41 +
.54 +
.35 +
.36 +
.60 +
.0025
.012
.050
.097
.024
TSP
RH
SULFATE
NITRATE
BSOL
9
11
159
87
9
t
F statistic of 4 corresponds to 95% significance level.
Two other days were eliminated from the data base: 12/6/73 when benzene
soluble organics were inordinately high, and 1/23/74 when sulfates and nitrates
at the central Phoenix site were an order of magnitude greater than values re-
ported at other Maricopa County stations. Including these two days in the anal-
ysis would have lowered our total correlation coefficient from 0.87 to 0.81.
As explained in Chapter 2, extinction coefficient is 24.3 * visibility.
SULFATE and NITRATE are 1.3 SOjj and 1.3 NO^, respectively.
62
-------
AilliaiSIA
in
10 O
C\J if>
o
in
oo
o
o
co
0)
o
in
CM
co
E
o>
CO
UJ
5
Q.
oo
10
(L)
O
o
o
H-4
ex.
O
UJ
O
UJ
00
oo
01
r
u
iE
a;
o
o
c
o
+J
o
a.
oo
in
CVJ
o
o
o
o
X
UJ
CJ
evj
co
co ce.
o
LO
o
in
o
CO
O
CM
O
*
O
I-
oi] '9 'lN3IDIJd300 NOI10NI1X3
63
-------
AilliaiSIA
LO
LO
in
CM
o
LO
o
o
- o
o
oo
c
CO
o
0.
o
<£>
o
LO
>-
I
tt
a
LU
ce.
3
S-
0)
4J
C
u
_ o
c
O
O
C
0£.
CM
iI
O
+
LO CO
II II
ca a:
O
co
o
CM
-O
CO
CM
QJ
o
IT)
o
CO
o
*
CM
'9 '
O
O
NOI10NI1X3
64
-------
AimaisiA
un
LO
CM
O
in
co
o
Lf)
O
CO CO
II
CQ
II
oc.
o
LO
_ o
o
co
o
CM
o
CU
O
Q-
LU
f.
U.
_l
3
to
en
CO C
Ol
CO -r-
o
Ul 0)
I O
< o
CO !-
(J
c
CM
o>
3
O
CO
o
CM
O
o
'9 '
NOI10NI1X3
65
-------
AlIliaiSIA
LO
in
CM
en
o
+
U3
CO
II
CQ
o
LO
I
O
in
CM
o
CM
in
O
CO
o
CM
o
o
CO
O)
i co
o
o
c
O)
o
Q_
S-
cu
CO O)
r-
«-H O
O
in
CM
I-
'8 '
NOI10NI1X3
66
-------
AiHiaiSIA
If)
CNJ
o
Lf)
.. o
c
cu
o
co
E
01
co
o
CO
CO
10
3
VI
01
(V (J
.0 O -r-
Csl «»-
Ol
O
o
CO
O
CO
c
o
o
CO
CO
CVJ
o
1Z. U
Ul C
M -r-
LU
CO
o
X
LU
10
CsJ
o>
o
10
ro
il
co
II
o:
o
un
o
co
o
C\J
o
o
NOI10NI1X3
67
-------
Table 6 indicates that all five variables are significantly correlated with
B at a 95% confidence level. SULFATE and NITRATE show especially high correla-
tions and significance levels. That a positive relationship exists between B and
SULFATE (or NITRATE) is also evident by a cursory examination of the scatter
plots (Figure 24 and 25).
Figures 24 and 25 also reveal one of the main limitations of the visibility/
pollutant regressions for the Southwest. Most of the B values correspond to
quite high visibilities (20-60+ miles); only three days in the data base exhibit
visibility less than 14 miles. At high visibility levels, errors in measuring
visual range tend to be greater because of the sparsity of distant visibility
markers; this may account for the relatively greater scatter at low B values.
Also, at high visibility levels, the extinction coefficient may vary more with
location over the visual range. An implicit assumption in our analysis is that
the extinction coefficient is spatially uniform out to the most distant markers.
The high correlation between B and SULFATE (or NITRATE) is quite suprising in ..
light of these potential errors.
Multi-Variate Analysis
As explained in Chapter 2, it is important to conduct a multi-variate an-
alysis that can separate out the individual impact of each independent variable,
discounting for the simultaneous effects of other independent variables. Uni-
variate analyses can lead to spurious conclusions because of inter-correlations
among the independent variables. In particular, TSP, RH, or BSOL might correlate
with B only because each of these variables is correlated with SULFATE and NI-
TRATE (see Table 7) which in turn are correlated with B.
68
-------
TABLE 7. INTERCORRELATIONS AMONG INDEPENDENT VARIABLES
AT PHOENIX, MARICOPA COUNTY DATA
TSP
RH
SULFATE
NITRATE
BSOL
TSP
1.00
-.25
.44
.34
.51
RH
-.25
1.00
.31
.38
.34
SULFATE
.44
.31
1.00
.56
.45
NITRATE
.34
.38
.56
1.00
.43
BSOL
.51
.34
.45
.43
1.00
Following the procedures outlined in Chapter 2, we performed a multiple
stepwise regression of the form
B = a + bjT + b2RH + b3SULFATE + b4NITRATE + bgBSOL
where T = TSP - SULFATE - NITRATE - BSOL
The variables T, RH, and BSOL did not retain significance in the multiple re-
gression; the resulting equation was
B = .23 + .037 SULFATE + .051 NITRATE (7)
with R = 0.87, ESULFATE = 81, and FNIJRATE = 32. The coefficients, ..037 and
.051, represent extinction coefficients per unit mass for sulfates and nitrates,
respectively, in units of (10 meters) /(yg/m3).
A regression model incorporating relative humidity effects in a nonlinear
manner (Equation (4), page 28) was also attempted. This model did not improve
on the correlation obtained with Equation (7).
Average Extinction Budget
As explained in Chapter 2, Equation (7) can be used to estimate the fraction
of visibility loss, on the average, due to each pollutant species. Substituting
in average values for SULFATE and NITRATE in Equation (7), we arrive at the re-
sults listed in Table 8,
69
-------
TABLE 8. AVERAGE EXTINCTION BUDGET FOR PHOENIX, MARICOPA COUNTY DATA
Category
SULFATE
NITRATE
BLUE-SKY SCATTER
UNACCOUNTED FOR
Contribution to
Total Extinction
44%
31%
17%
8%
Contribution to ±
Extra Extinction
53%
37%
-
10%
Above-and-beyond blue-sky scatter.
PHOENIX, NASN DATA
The visibility/pollutant analysis for Phoenix was repeated using NASN par-
ticulate data for the years 1966 through 1974. The Phoenix NASN site is located
near the Maricopa County monitoring site, two miles northwest of the Sky Harbor
Airport.
The type of data used and the averaging times are summarized in Table 2
(page 26). The NASN sulfate and nitrate measurments are made with the color-
imetric and the reduction/diazo coupling methods, respectively.
After eliminating days with fog and/or precipitation, there were 202
NASN sampling days providing complete data for TSP, SO^, and NO^. The basic
correlations in the NASN and airport data are summarized in Table 9.
Multi-Variate Analysis
As occurred with the Maricopa County data, stepwise multiple regression
with the NASN data retained only sulfates and nitrates as significant contribu-
tors to extinction. The resulting equation was
B = .42 + .027 SULFATE + .030 NITRATE (8)
70
-------
with R = 0.68 FSULFATE = 85 and FNITRATE = 27. The regression coefficients
(extinction coefficients per unit mass) for SULFATE and NITRATE are about
one-third lower than those obtained with the Maricopa County data.
A regression model incorporating relative humidity effects in a nonlinear
manner was also attempted. This model failed to improve upon the linear re-
gression exemplified by Equation (8).
TABLE 9. CORRELATIONS AMONG ALL VARIABLES
B
TSP
RH
SULFATE
NITRATE
B
1.00
.40
.19
.62
.48
TSP
.40
, 1.00
.00
.38
.55
RH
.19
.00
1.00
.16
.23
AT PHOENIX, NASN DATA
SULFATE
.62
.38
.16
1.00
.36
NITRATE
.48
.55
.23
.36
1.00
Average Extinction Budget
The extinction budget obtained from Equation (8) is summarized in Table 10.
Because of a weaker relationship between the pollutant data and visibility data,
we were not able to account for as much of the extinction with the NASN data as
we were with the Maricopa County data.
71
-------
TABLE 10. AVERAGE EXTINCTION BUDGET FOR PHOENIX, NASN DATA
Contribution toContribution to t
Category Total Extinction Extra Extinction
SULFATE 25% 32%
NITRATE 19% 23%
BLUE-SKY SCATTER 20%
UNACCOUNTED FOR 36% 45%
Above-and-beyond blue-sky scatter.
SALT LAKE CITY, NASN DATA
The visibility/pollutant analysis for Salt Lake City was conducted using
observations from the International Airport and pollutant measurements from the
downtown NASN site. The airport is located about six miles northwest of down-
town, on the outskirts of the city and near Great Salt Lake,
The data for the analysis covered the years 1966 through 1972. Eliminat-
ing days with fog and/or precipitation left 130 NASN sampling days with com-
*
plete pollutant data. The correlations in the NASN and airport data are sum-
marized in Table 11.
Multi-Variate Analysis
A stepwise multiple linear regression with the Salt Lake City data yielded
the equation
B = -.53 + .004 T + .016 RH + .036 SULFATE + .130 NITRATE
where T = TSP - SULFATE - NITRATE.
*
Two other days were also eliminated: 1/19/66 when the meteorologist remarked
that visibility was obscured by low clouds just after a snowstorm, and 11/19/69
when the meteorologist remarked that visibility was fair at the airport but
poor downtown. Including these two days in our analysis would lower the total
correlation coefficient from 0.72 to 0.69 in the multiple linear regression.
72
-------
Unlike the Phoenix results, all four variables retained significance at a 95%
confidence level. The total correlation coefficient was 0.72, and the F-statistics
were 7, 23, 6, and 12 for T, RH, SULFATE, and NITRATE respectively.
The regression model including nonlinear RH effects attained an even better
correlation, R = .81. The resulting equation was
.0022 T .024 SULFATE .057 NITRATE
B = .14 +
(10)
(1 - .01 RH) (1 - .01 RH) (1 - .01 RH)
Our findings indicate that relative humidity is a more important factor in
Salt Lake City than in Phoenix. This is physically plausible because relative
humidity is greater in Salt Lake City than in Phoenix. Extinction is much more
sensitive to changes in relative humidity at higher relative humidity levels
(RH > 60%) because of the hygroscopic and/or deliquescent properties of sulfates,
nitrates, and other aerosols.
TABLE 11.
B
TSP
RH
SULFATE
NITRATE
Average Extinction
CORRELATIONS
NASN DATA
B
1.00
.54
.48
.50
.59
Budget
AMONG ALL
TSP
.54
1.00
.24
.45
.61
VARIABLES AT SALT
RH
.48
.24
1.00
.24
.25
SULFATE
.50
.45
.24
1.00
.59
LAKE CITY.
NITRATE
.59
.61
.25
.59
1.00
Table 12 presents an average extinction budget based on Equation (10). The
results indicate that sulfate, nitrate, and the remainder of TSP each account
for about one-third of the extinction above-and-beyond blue-sky scatter.
73
-------
TABLE 12. AVERAGE EXTINCTION BUDGET FOR SALT LAKE CITY, NASN DATA
Contribution toContribution to
Category Total Extinction Extra Extinction
SULFATE 30% 34%
NITRATE 27% 31%
REMAINDER OF TSP 30% 35%
BLUE-SKY SCATTER 13% :
Above-and-beyond blue-sky scatter.
ANALYSIS OF OTHER DATA BASES
An attempt was made to derive visibility/pollutant relationships at five
other locations: Denver, Tucson, Grand Canyon/Prescott, Grand Canyon/Winslow,
and White Pine/Ely. Not unexpectedly, these attempts met with little success be-
cause the airport was not representative of the NASN monitoring site and/or be-
cause relatively poor resolution was available in the visibility measurements.
^ulfate (in two of the cases) and nitrate (in one of the cases) showed margin-
j
ally significant correlations with extinction, but the relationships were not
strong enough to allow estimation of extinction coefficients per unit mass.
7«
The Denver and Tucson airports are located about 7 and 10 miles from
the downtown NASN sites, respectively. Prescott and Winslow are each 90
miles from Grand Canyon; Ely is 40 miles from the White Pine monitor.
**
At a 95% confidence level.
74
-------
DISCUSSION AND GENERALIZATION OF RESULTS
Considering the potential errors in the data bases, the regression studies
for Phoenix and Salt Lake have been quite encouraging. The results indicate
that visibility reduction in Phoenix is significantly related to sulfate and
nitrate concentrations. Because the remainder of TSP in Phoenix consists large-
ly of fugitive dust (Richard et al. 1977), and because the remainder of TSP is
not significantly related to extinction in a multi-variate analysis, we con-
clude that visibility in Phoenix is not highly related to fugitive dust.
These conclusions are not surprising in light of known principles of
atmospheric physics. Sulfates and nitrates are secondary aerosols that tend
to occur in the particle size range of 0.1 to 1 micron. Fugitive dust particles
reside in a size range considerably larger than 1 micron. As shown in Figure
27, light scattering per unit mass of aerosol exhibits a pronounced peak in the
0.1 to 1 micron range, around the wavelength of visible light. Even though
sulfates and nitrates (including the cation) account for only 9% of the aerosol
mass in Phoenix, it is not unreasonable for them to dominate the light scat-
tering and visibility reduction.
The regression analysis also indicates that sulfates and nitrates are the
main contributors to haze in Salt Lake City. In Salt Lake City, however, the
remainder of TSP is significant, contributing about one-third of the visibility
reduction.
Comparison With Published Literature
Table 13 compares our estimates of extinction coefficients per unit mass
with other results published in the literature. Fairly good agreement is evident.
75
-------
o
d
o
K
O
if)
O
to
o
OJ
o
I
>^.
C)
in
0'
ro
0
CM
O
O
fe
^.
lO
Q
^^
IO
j;
o
S-
o
r-
E
t-
d)
4J
d>
(O
p-
Q
0)
"o
r-
M
S-
O)
to LT>
r-
T3 r
O) X
n (U
OT3
r- C
4J .p.
S_
to d)
cx >
«*- T>
o o
(O
C J_
0 «»-
I- d)
+J S-
o
c <«-
=> o
l+_
(0
(O d)
p_
W 0
tO !-
4->
CO J_ .
n (O ^ ^
O Q.LT)
to r^.
Si CT>
IO
0) 0
(O -r- S-
i- d>
>> d> TJ
J2 JT C
Q. 03
D> to i
c -a
i- >> d>
s. i? -
dl-r- i.
4-> tO U_
»-> C
10 d» -O
0 -0 C
to 3 4->
i- J-
i J- d>
O J3
O l|- O
d) CC.
N T3
i- ttl "
r- 4-> 01
(O =}+->
1
o
I
00
o
- 1
to
o
1
^
O
1
CM
o
1
c
ro
'Q
CM
en
76
-------
CO Q.
E oo
--^ I
CT
3- «t-
- o
i-H S-
1 CU
- "* "O
CO E
S- !-
CU to
-P E
CU CU
E cc.
o
r 1
oo
oo oo
«=C 1 CO
z: z cu
LU -P
1 i i to
i i O S-
Z3 U_ -i-
o: LU
LU O
CL. 0
(/*) ^^
I- 0
Z. I-H 00
UJ H- CU
I-H 0 -P
CJ 'Z. tO
I i t I <4-
U_ I i
U_ X 3
LU LU OO
o
z
o
1 1
H-
0
-« z:
1 O
X > i
LU t
Ll_ O
0 0
_J
oo
LU
1
1
0
oo
*
r-l t-H ^- «*
O O O O
O O O O
V V
^c
LO co co o
O O rH t I
*
^~ CO ^" il*
o o o o
*
^
-p
s~~+. «r~
fO *'*"" ** C«3
P to
(C -P O)
T3 (O -^
X X T3 to
r- . !- _l
E >> E Z
CU -P 0) OO -P
O E O < r-
O. Q_ CO
C
r-
-o
CU
r-
r
CL
CL
tO
00
r»
CU
-o
o
E
E >.
0 -o*
CO 4J
CO tl
CU
C7^ -
cu t~~
Q£ ^_>
LO
O
O
.
O
I
rO
O
«4_>
o
E
O
t 1
o
CO CO
tO O)
0) J
fO -P
3
cu o
-P CO
i cu
CU -E
S- -P
^-**
<£>
CO fv.
E
O f-H
r ^»^
4-*
to .
r-~ r**
3 (0
O
r 4^
to cu
CJ
e~
< o
ro (O
0 JQ
>- E
P CU
0) CO
S- S-
0 ^
cu
C^ >s
h- .Q
i i
O
o
.
o
n
(0
o
-p
o
E
O
pM
«0
o
-p
o
E
CO
E
to
1
D_
P
to
CU
s-
CJ3
,-~,
CO
f"^
CT>
i-H
M-
<^«
3
S-
o
o
o
3
-o
E
rO
E
CU
CT
>O
~r*
^.^
CO
O
-p
CO
p
CO
3
Q
LO
r-l
O
UO
O
^
O
00
cu
r .
cu
CT
E
^^
00
o
_J
^"^
LO
|"^,
cn
i-H
00
P
r S-
cu cu
TJ JD
§O
a:
^Z T3
O E
r- 10
CO
oo cu
cu -P
S-T-
CTJE
0)3
K
00 ^1-
O O
O 0
*
V
^c
rH LO
O O
V
*
<£> CTi
r-l O
00
cu
r
cu
E
^^
CO
o
J_ J
**-+.
Cf>
r*^
^^
t-H
00
00
to
o
-**
(»»
cu
-a
o
E
C
O
r-
CO
CO
CU
S.
o^
cu
a:
u
r
to
o
p
o
E
.
O
r
10
o
-p
o
c
o
i-H
0
p
LO
o
1
It-
CD -P
cu
I
0 S-
co CU
O E
S- 0
cu en
tO CT
tO
i 3
O ^"r-H
i n:
(O Z i
o
t
cu
^
*D
.
o
O"
^
L~
CD
c
1
-~
J*^
I
cu
o
o
E
E
O
CO
00
cu
i.
CT
CU
*T"
cc
S-
-------
In particular, there is agreement that the extinction coefficients per unit mass
for sulfates and nitrates are an order of magnitude greater than the extinction
coefficient per unit mass for the remainder of TSP.
Our estimates of extinction coefficient per unit mass tend to be lower
than other estimates found in the literature, especially for sulfates. Two
facts could account for these differences. First, relative humidity is low in
Phoenix and Salt Lake City. Lower relative humidity should lead to a smaller ex-
tinction coefficient per unit mass for sulfates and other hygroscopic particles.
Second, visibility measurements at Phoenix and Salt Lake City are made over
rather large distances. The pollution concentrations measured at the downtown
monitors might be higher than the average concentrations over the entire visual
range. This effect could artificially lower the estimate of extinction coef-
ficient per unit mass.
That even higher extinction coefficients per unit mass are found in the
literature makes our results appear conservative. That is -- based on published
extinction coefficients, we would have concluded that there is more than enough
sulfate and nitrate in Phoenix to account for all of the observed light scattering.
Extension to Nonurban Areas
It is unfortunate that data are not available for completing regression
models in nonurban areas. However, using the urban models for the Southwest,
we can attempt to extend the analysis to nonurban areas. The extension is made
by using general extinction coefficients (derived from Table 13) for sulfates,
nitrates, and the remainder of TSP.
Daytime average RH is 27% in Phoenix and 41% 1n Salt Lake City. In compar-
ison, daytime average RH is 53% in downtown Los Angeles.
78
-------
Based on the models for Phoenix and Salt Lake City (and tempered slightly by
other published results listed in Table 13), our estimates of extinction coeffic-
ients per unit mass appropriate to the Southwest are as follows:
sulfates: .03 to .06 (104m)"1/(yg/m3)
nitrates: .04 to .09 "
remainder of TSP: .001 to .005 "
These coefficients can be applied to NASN pollutant data at three national park
sites (Grand Canyon, Mesa Verde, and White Pine) to predict visibility levels at
those locations. Since the three national parks in question are in the more arid
parts of the Southwest, we chose relatively low coefficients-- .04 for sulfates,
.05 for nitrates, and .002 for the remainder of TSP.
Table 14 presents recent NASN data for the three national park sites and
lists predictions of average visibility based on our estimates of extinction
coefficients for each aerosol component. The predicted average visibilities
agree very well with the median visibilities, 65 to 80 miles (see Chapter 3),
actually observed in nonurban parts of the Southwest.
The data and assumptions that serve as the bases for Table 14 also can be
used to estimate the fraction of extinction contributed by each aerosol component;
Table 15 presents an average extinction budget for each of the three national
park sites. Because average visibility is quite good (65 to 80 miles) at these
locations, blue-sky scatter accounts for a significant fraction of total ex-
tinction (~40 to 50%). If our assumptions on extinction coefficients for the
aerosol species are correct, sulfates account for the dominant part (~two-thirds)
of the extra extinction. The other one-third is about equally divided between
nitrates and the remainder of TSP.
Note that 0.15 (lO^eters)'1 is added to account for blue-sky scatter.
79
-------
f-
oo
UJ
I
H <
00
a:
Q.
i
^^
z
o
1 1
1
z
=C
r
i i
ca
i i
oo
i i
>
UJ
CD
Q£
UJ
«g;
1 1
O
OO
z
0
1 1
1
o
I-H
o
1 1 1
111
a:
CL.
.
^~
. i
UJ
CO
1
1
>- CM
H- 0
I-H O
_J
1 1
ca +
oo z
I-H LO
=* o
oo
UJ
CD «* +
o
Q +
UJ
1 LO
O i
i i «
O
LU II
0- >
P"
a. z
^-» 00
l/> | .
O LL.
r~- o co
en
UJ '
Q
' II *
< O
LU OO
_1 I
^C ^t oo
=3 U-
Z ZD -
< oo
^p^
o
f ' '<
I
o
0
_J
to
O)
r = =
'i
en en en
oo
E
****. ~
0)
^3.
O LO i-H
CM -! rH
OO
E
0)
~1
LD r^. *3"
VO LO ^~
CD O C3
OO
E
Q)
^
CM LO CO
00 00 CM
O
>> O) CU
c -o c
(0 i- -r-
<_> eu o-
-a cu
C (0 «->
CU
i- -O
C C
3 >
ro
CU CU
a. &-
C -P
CU
i- S-
C
M- CM
CU O
O O
O
C T3
O C
r- (O
O "
c to
r- CU
X. ro
CU S-
+-> !-
ro C
-4-> i-
O
C 4-
o
1- LO
(-> O
Q.
E
3 >
to to
ui CU
ro 4J
CU t-
c- >
+-> 3
0 S-
T3 4
CU
ro O
CO
4-
80
-------
TABLE 15. AVERAGE EXTINCTION BUDGETS FOR NATIONAL PARK SITES1
CATEGORY
CONTRIBUTION TO
TOTAL EXTINCTION
CONTRIBUTION TO
EXTRA EXTINCTION'
Sul fates
Nitrates
Remainder of TSP
Blue-Sky Scatter
Sul fates
Nitrates
Remainder of TSP
Blue-Sky Scatter
Sul fates
Nitrates
Remainder of TSP
Blue-Sky Scatter
36%
9%
12%
43%
40%
8%
9%
43%
37%
7%
7%
49%
Grand Canyon
64%
16%
20%
--
Mesa Verde
70%
14%
16%
--
White Pine
72%
14%
14%
__
t
Based on the assumption that extinction coefficients per unit mass are .04
for sulfates, .05 for nitrates, and .002 for the remainder of TSP.
*
Above-and-beyond blue-sky scatter.
81
-------
CHAPTER 6
THE COPPER STRIKE OF 1967-1968
The data presented in Chapter 4 point to the conclusion that visibility
has deteriorated in the Southwest over the past two decades. Although visi-
bility levels are not uniform over the Southwest, and although quantitative
trends in visibility are not the same at all locations, the pattern of de-
teriorating visibility appears to be ubiquitous. Indications are that visual
range has decreased in urban, suburban, and remote areas.
The analyses of Chapter 5 Indicate that extinction above-and-beyond
blue-sky scatter is dominated by secondary aerosols, sulfates and to a lesser
degree nitrates. This conclusion is reasonable on physical grounds because
secondary aerosols tend to occur in the particle size range of .1 to 1
micron and because particles in that size range are extremely efficient at
scattering visible light.
From these conclusions we hypothesize that the deteriorating trend in
visibility is related to increases in secondary aerosols, the result of
growth in SO , NO , and possibly hydrocarbon emissions from copper smelters,
A X
power plants, motor vehicles, and other sources. We further hypothesize that
the large spatial scale in visibility deterioration is due to two factors:
(1) growth in sources throughout the region and (2) the tendency for secon-
dary aerosols to be spread widely because of the mixing and transport that
occurs during the time required for aerosol formation.
A unique opportunity to test our conclusions and hypotheses is pro-
vided by the industry-wide copper strike of July 1967 to March 1968. During
82
-------
the late 1960's and early 1970's, copper smelters accounted for about 80
percent of the SOV emissions in the Southwest (NEDS 1976) . The nine-month
X
shutdown during a labor strike allows us to test the impact of a substantial
reduction in regional SO emissions. We would expect to find decreases in
A
sulfate and increases in visibility on a large spatial scale.
COPPER PRODUCTION AND VISIBILITY AT TUCSON
The relationship between copper-smelter emissions and visibility was
first noted in data for Tucson, Arizona (Lockwood and Hartmann 1970; Hartmann
1972, 1976, 1977). Tucson is located in the southeast quadrant of Arizona,
an area containing several copper smelters (see later Figure 32). The
smelter nearest to Tucson is at a distance of about 30 miles.
Figure 28 compares trends in frequency of hazy days at Tucson (for
six month periods) with trends in yearly copper production for the state of
Arizona. Figure 28 indicates that the frequency of haze, increased from
about 15% in 1957-1959 to about 80% in 1969-1972. The long-term increase in
haze appears to be very well correlated with trends in copper production.
In particular, during the first six months of the 1967-1968 copper strike,
haze incidence dropped to 25%. An improvement in haze levels is also notice-
able after 1973; Hartmann (1976) attributes this improvement to partial con-
trol of smelter SO emissions in the early 1970's and to control of auto-
J\
motive emissions.
The Southwest is defined as the four corner states plus Mpvada and
Wyoming. Note that copper production accounts for less than 1% of regional
NOx emissions and less than 10% of regional particulate emissions from con-
ventional (non-fugitive-dust) sources.
83
-------
COPPER PRODUCTION (THOUSANDS OF TONS)
o
o
o
o
co
I
o
o
o
o
to
I
o
o
in
I
o
o
I
I I I I I I
co
to
co
to
to
to
to
in
to
to
ro
to
C«J
to
o
to
en
tn
oo
tn
in
o
o
3
O
O
a.
0)
Q.
Q.
O
O
en
^O *"""^
OJ
re re
Ol !->
S- $_
S- ro
O OZ
o
0) O
N r*»
re en
.C t-H
c c
O
C O
O _J
03
C\J
OJ
3
cn
o
o
o
00
o
to
O
CM
(iw 09 > AiniaisiA) SAVQ ..AIVH,,
d4
-------
A more detailed illustration of haze levels at Tucson before, during,
and after the 1967-1968 strike is presented in Figure 29. The step-function
change in haze incidence during the strike is quite obvious. Hartmann (1972)
attributes the improvement in visibility during the copper strike to reduc-
tions in sulfur oxide emissions; Figure 30 demonstrates that sulfation rate (a gross
measure of S02 concentrations) dropped dramatically in Tucson during the strike.
Hartmann also noted that TSP levels did np_t change significantly during the strike.
In order to check Hartmann's conclusions about changes in pollution
levels during the strike, we plotted bi-monthly averages of TSP, NOI, and
S(L at Tucson from 1966 to 1969 using the NASN data. Figure 31 confirms
that sulfates did show a pronounced drop during the strike; TSP and nitrates
exhibited no significant change.
REGIONWIDE CHANGES IN SULFATES DURING THE COPPER STRIKE
It is of interest to examine regionwide changes in sulfates and visi-
bility during the nine-month copper strike. Using the NASN data, it is
possible to document sulfate changes at ten locations in the Southwest.
Table 16 indicates that most of these locations did experience a substantial
drop in sulfates during the strike as compared to (July - March) seasonal
averages for surrounding years.
Figure 32 compares the spatial distribution of sulfate changes during
the strike to the spatial distribution of smelter SOV emissions. Substantial
A
decreases in sulfate occurred at the five locations (Maricopa County, Phoenix,
Tucson, Ely, and Salt Lake City) that are within 12 to 70 miles of copper
smelters. More significantly, sulfates appeared to drop by about 60% at
Grand Canyon and Mesa Verde; these national park sites are located about
two to three hundred miles from the smelter areas.
85
-------
100
o
V
|2 50|_
4
O
O
oe
SHELTERS ON
SMELTERS OFF
SMELTERS ON *
J A S 0 N 0
63/66
0 H 0 J F M
66/67
JASONDJFM
67/66
J A S 0 N D J FM
66/59
JASONOJFM
69/70
Figure 29.
Changes in number of hazy days at Tucson
during 1967-1968 copper strike (Hartman 1972).
§
O
o
o
8.10
o
o
§.05
LU
£t
O
b 0
2 JFMAMJJASONDJFMAMJJ ASONOJFMAMJJASON OJFMAMJJASOND
=J (966 1967 . 1968 1969
I
SMELTERS ON
1
1
SMEUERS OR
I
SMELTERS ON
_L
Figure 30. Changes in sulfation rate at Tucson
during 1967-1968 copper strike (Hartman 1972).
06
-------
150 -,
CO
100 -
Q-
00
I
co
o
50 -
3 -
2 ~
STRIKE
i
12
10
8
6
4
2
1966
1967
1968
1969
Figure 31. Variations in pollutant levels at Tucson
during 1967-1968 copper strike (NASN data, bi-monthly averages)
37
-------
TABLE 16. CHANGES IN SULFATE LEVELS DURING THE COPPER STRIKE
COMPARED TO SEASONAL AVERAGES*
LOCATION
ARIZONA
**
Grand Canyon
Maricopa County
Phoenix
Tucson
COLORADO
Denver
**
Mesa Verde
NEVADA
Las Vegas
**
White Pine
NUMBER OF
SURROUNDING
YEARS OF DATA
4
4
6
4
4
4
6
4
AVERAGE SULFATE
DURING
STRIKE
1 . 00 yg/m3
1.60
2.17
1.68
3.62
.80
2.20
.47
AVERAGE
SULFATE
OTHER YEARS
2.51 yg/m3
4.23
5.73
5.05
4.41
1.85
4.52
1.92
PERCENT
CHANGE
-60%t
-62%t
-62%f
-67%t
-18%
-57%f
-51%*
-76%*
NEW MEXICO
Albuquerque
3.37
3.45
-2%
UTAH
Salt Lake City
4.37
7.10
The season of the strike was July through March.
**
National park locations
Significant from zero at 95% confidence level
88
-------
to
c
o
E
00 LU
S- X
O) O
O. 00
a.
o s-
O O)
4->
0) r
j= aj
*-> e
to
en
C <4-
i- O
s-
3 C
Q O
tO -Q
M- -i-
3 4-»
l/> CO
r
C Q
CO (O
CD U
CT-i-
c ^:
T3 O
0) 0)
4-> C3
CO
3
89
-------
The sulfate decreases at Grand Canyon, Mesa Verde, and most of the other
locations easily pass tests for statistical significance at a 95% confidence
level. That the changes at Grand Canyon and Mesa Verde are significant can be
*
seen qualitatively in Figure 33. The statistical distribution of sulfate
levels during the strike appears very different from the statistical distri-
bution before and after the strike.
To put some of the above observations in better perspective, we note that
Los Angeles basin produces haze at distances greater than 75 miles from the main
source areas (Blumenthal et al 1974). The entire Los Angeles basin emits on the
order of 500 tons/day SOY and 1500 tons/day NOV (AQMP 1976). Although copper
A A
smelters emit negligible NO , they do produce very large amounts of SO . Dur-
/\ X
ing the late 1960's, the largest smelter emitted more than 1300 tons/day of SO ,
X
and the group of smelters in southeast Arizona emitted over 5000 tons/day of
SOV (NEDS 1976; Oliver 1977).**
X
REGIONWIDE CHANGES IN VISIBILITY AND EXTINCTION DURING THE COPPER STRIKE
Table 17 and Figure 34 summarize changes in visibility during the copper
strike. Visibility improved at almost all locations, with the largest improvements
occurring near and downwind (north) of the copper smelters in southeast Arizona
and near the copper smelters in Nevada and Utah. The nine locations showing sta-
tistically significant improvements are all within 150 miles of a copper smelter.
Table 17 also lists percent changes in extra extinction (above-and-beyond
blue-sky scatter) during the copper strike. These percent changes are
plotted in Figure 35. Figure 36 shows the changes in extinction coef-
*
Note, the monitoring dates at Grand Canyon and Mesa Verde were usually
not coincident.
**
Note that sulfur oxide controls were installed at several of these
smelters in the early 1970's.
90
-------
0)
-o
S-
OJ
10
00
a;
s- nj
00 O
i.
a> o c co nj
r- > t-
H oo O C O)
rtj <1J -i- O Q.-P
3T ^ -P 00 Q. (O i
1 3 (0 O 1
rL s- x» > 1 ! 00
°il l§ 1 ^ 1
C -i- S- O" S- CM
.° h 0 2! o 3 ^»
^3 co -P U.M-CO- _J
3 <0 F
as-
r- Q) i/> >|r
S- O.-P V
-p a. c
r- C_) E 1
Q (1)
V i- 1
>>-C 3 '
O -P (/) |
3 C E
CT-r-
0) S- 'd'
S- 3 CO
~ _. J
L n L
1
1 _
n
1
o o o o
'-J- CO CM i-l
VO
O
vo
CM C
0
Lf) !
QC.3
1 ' 1
.6 4.4
concentrations dm
sonal average distr
combined.
CO OJ to
P QJ (O
a) oo -P
CO r O T3
E 3 -P
-^ 00
LU C Q-
CM 1 O £ tO
ff ! O co
U_ -POO)
_i 3 s:
P i- ro
00 -P
c 00 C
o -o o
CM >>
Ll_
39NVM Q3IdI33dS NI SI
3WI1 dO !N30cl3d
91
-------
TABLE 17. CHANGES IN VISIBILITY AND EXTINCTION DURING THE
COPPER STRIKE COMPARED TO SEASONAL AVERAGES
VISIBILITY^ DURING**
LOCATION PERCENTILE* STRIKE**
URBAN
Phoenix, Arizona
Tucson, Arizona
Denver, Colorado
Salt Lake City, Utah
NONURBAN
Fort Huachuca, Ariz.
Prescott, Arizona
Winslow, Arizona
Grand Junction, Col.
Ely, Nevada
Las Vegas, Nevada
Alamogordo, H.M.
Farmington, N.M.
Dugway, Utah
Wendover, Utah
Cheyenne, Wyo.
50th
75th
50th
50th
50th
75th
80th
50th
85th
50th
50th
75th
50th
75th
50th
47.4 miles
62.1
52.6
31.8
56.0 miles
65.7
52.5
81.6
26.8
61.7
72.3
67.3
64.6
60.9
71.2
VISIBILITY PERCENT
OTHER CHANGE IN
YEARS*** VISIBILITY
36.8 miles
50.4
50.3
29.8
48.1 miles
57.0
48.6
78.4
22.3
60.5
68.7
69.6
62.9
51.7
72.9
+29X+
+23Xt
+ 5%
+ 7J+
+16X+
+15%f
+ 8%f
+ 4*
+20Xf
+ 2%
+ 5X*
- 3X
+ 3%
+18«f
- 2X
PERCENT
CHANGE IN
EXTRA
EXTINCTION
-29%f
-27%f
- 6%
-tt*
-20%f
-20%f
-m+
- 82
-19%+
- 3%
-«*
+ 6%
- 4%
-22%f
+ 4%
We attempted to use the 50th percent!le for all locations. In some cases we
were forced to use higher percentiles to avoid extrapolation of the frequency dis-
tributions.
**The strike season was July - March and included 1100 3-hourly daytime obser-
vations.
For all cases but Alamogordo, Farmington, and Dugway we were able to use six
surrounding years (6600 measurements) to determine seasonal averages.
Significant from zero at 95% confidence level. In calculating significance
we assume that the 4 measurements each day are totally dependent but that the
days are independent. This assumption is, of course, rather crude.
92
-------
CU
10
s_
cu
Q_
Q-
O
O
CU
S-
3
-o
I?
CO
O)
a>
c
1C
O
0)
u
CU
Q_
-o
CU
-o
rtj"
(O
c:
o
10
to
cu
n
cu
=9
CT)
93
-------
05
.*:
s_
o>
Q.
0-
O
O
-------
to
s_
0)
CM
O
1
,
s-
-»->
in
fe
1 *>
1 QL
I O.
1 O
(
} 0)
I
.C
4->
01
C
r
*
3
T3
ti-
cs
CM
CO
<0
CM
1- O
O !-
4- 4->
(0
^H O
O
OJ
i.
3 H-
O> O
r
U. CO
OJ £
(U (0
Q.
fO
JD
i-
rD :
O
3 O
c o o
O Lf> i-l
Z CM
Ml
^ *»
to
CD
r
i
kE.
LU
_1
<:
o
OJ
o
c
X
01
(LI
en
c
to
^
o
T3
O)
+J
T3
CO
CO
C
O
O)
to
a;
O)
95
-------
ficient in absolute terms, units of [10 meters] . Again it is evident
that the greatest improvements occurred within 150 miles of the copper smelters.
The spatial scale of the visibility impact during the strike (apparently
on the order of 150 miles away from the smelters) does not agree with the
spatial scale of the sulfate changes during the strike (apparently on the order
of 300 miles away from the smelters). Some of the potential reasons for this
discrepancy will be covered in the discussion at the end of the next section.
CHANGES IN EXTINCTION COMPARED TO CHANGES IN SULFATE
The reductions in sulfate during the copper strike offer us an oppor-
tunity to check the results of our regression studies (Chapter 5). For
instance, using the Maricopa County data we concluded that sulfates account
for 53% of the extra extinction in Phoenix (Table 8). Since sulfates decreased
62% in Phoenix during the strike (Table 16), we would predict that extra ex-
tinction should decrease by 33% (.53 x 62%). The actual decrease in extra
extinction was 29% (Table 17)...quite good agreement!*
Similarly, based on our regressions using the Phoenix NASN data, we
would predict that the decrease in extra extinction at Phoenix during the
strike should be 20% (.32 x 62%). Again this is not far from the actual
decrease of 29%. For Salt Lake City, our regression model would lead us to
predict a 13% (.34 x 38%) decrease in extra extinction during the strike. The
actual decrease was 8%.
The copper strike provides a completely independent check of the Phoenix
(Maricopa County data) regression model because the strike occurred in
1967-1968, while the regression model is based on data for 1973-1974.
The Phoenix/NASN and Salt Lake City/NASN regression studies are based
on data from 1966-1974 and 1966-1972 respectively, and thus include the
strike period. The strike still should provide an independent test,
however, because the regression models are dominated by order-of-magnitude,
day-to-day variations over several years and should be relatively un-
affected by the lesser changes during the nine month strike.
96
-------
The extension we made of our regression results to nonurban areas
(Table 15) does not fare as well in this test. Based on the sulfate changes
during the strike, we would predict that extra extinction should decrease by
38% (.64 x 60%) at Grand Canyon and by 40% (.70 x 57%) at Mesa Verde. The
actual decrease in extra extinction was only 11 to 20% in north/central
Arizona, and extinction appeared to Increase slightly at Farmington, New
Mexico. There are four plausible explanations for these discrepancies:
At the nonurban airports in northern Arizona and northern New
Mexico we were forced to use 75th or 80th percentile visibility
to estimate changes in extra extinction during the strike. It
is very possible that we would have found greater changes in ex-
tinction during the strike had data been available for median
(50th percentile) visibility. The 75th and 80th percentile visi-
bilities reflect extinction produced by weather events (fog and/or
precipitation) more so than do the median visibilities.
The changes in sulfate at Grand Canyon and Mesa Verde during the
strike are based on only 2 samples per month. Part of the ob-
served drop in sulfates may be due to statistical error. If our
estimates were off by one standard error, the sulfate changes at
Grand Canyon and Mesa Verde would be on the order of 45 - 50%
rather than 60%.
The extension of our regression results to nonurban areas could be
in error. We estimated that sulfates account for about two-thirds
of the extra extinction in nonurban areas. A rudimentary error
analysis of our assumptions (as listed on page 79) indicates that
sulfates may account for a smaller fraction, possibly less than
one-half, of the extinction in nonurban areas.
Farmington, New Mexico is near a very large power plant. It is
possible that visibility at Farmington is more dependent on this
local source than it is on regionwide sources.
ANALYSIS OF METEOROLOGY
A question naturally arises as to whether the changes in sulfates and
visibility during the copper strike could be due to unusual meteorology.
97
-------
Possibly the nine month copper strike coincided with a period of weather
conducive to good visibility, or with a period of extremely low pollution
potential. We have examined this question and have arrived at the conclu-
sion that favorable weather was not a major factor contributing to the
observed decreases in sulfates and extinction. In fact, pollution potential
during the strike appears to be slightly greater than in surrounding years.
Statistical Significance of the Changes
Many of the sulfate and visibility changes during the strike are
statistically significant from zero at very high confidence levels. With
crude significance tests we find "t" statistics as great as 8 for the sul-
*
fate changes and as great as 12 for the visibility changes. The statistical
significance of the changes can also be appreciated by inspection of the raw
data (see for instance Figures 29 and 33).
If meteorology were to account for these highly significant changes
during the nine-month copper strike, we would expect that weather patterns
would have been notably unusual. To the contrary, inspection of weather
data for the period (Smith 1977; Zeldin 1977) indicates that weather patterns
were not remarkably different from normal. The small anomalies that did
occur indicate that pollution potential might have been slightly higher
during the strike than under normal conditions (Smith 1977).
Korshover Analysis
Stagnating anticyclones often result in very low wind speeds and re-
stricted vertical mixing and are known to be associated with heavy air
*
t - 1.7 for 95% confidence level and t - 3.9 for 99.95% confidence
level.
98
-------
pollution in urban areas. Korshover (1976) has investigated pollution po-
tential for the eastern United States by counting the occurrences of stag-
nating anticyclones. An extension of Korshover's analysis to the Southwest
(Niemann, 1977) may provide us with an indication of pollution potential dur-
ing the copper strike.
Niemann1s results indicate that 20 stagnating anticyclones occurred
in the Southwest during the nine month copper strike; this compares to a
long-term seasonal average of 17.1 stagnating anticyclones. The 1967-1968
strike period included more stagnating anticyclones than any of the six sur-
rounding nine month periods (1964-1965 to 1970-1971) and ranked as the fifth
highest period among the nineteen years (1957 to 1976) included in Niemann's
study. If we accept the Korshover method as a measure of pollution potential,
we conclude that pollution potential during the 1967-1968 copper strike was
slightly higher than normal.
Stratification by Meteorological Class
Sorting the data by meteorological class can explicitly account for
the effects of meteorology. We have attempted this type of meteorological
stratification with the sulfate data for Grand Canyon. Sorting by the
occurrence of stagnating anticyclones indicates a 51% decrease in sulfates
during the strike. Sorting by conditions most conducive to transport from
the smelters in southeast Arizona indicates a 71% decrease in sulfates dur-
ing the strike. These results are in basic agreement with the 60% reduction
calculated using all of the sulfate data at Grand Canyon.
Stratifying the sulfate data by meteorological class creates a problem
of small sample size. NASN sulfate data are collected once every fourteen
99
-------
days; if we attempt to sort the data 1n meteorological classes, we are left
with very few samples. Thus, the statistical significance of the sulfate
changes at Grand Canyon using the sorted data (t ~ 3) is smaller than the
significance level using all the data (t ~ 6), even though the absolute
level of the change (-60%) agrees in all cases.
In future work it may prove interesting to conduct further sorts of
the visibility and sulfate data according to meteorology. However, our
present results concerning the statistical significance of changes during
the strike and concerning meteorology during the strike provide us with
confidence that the copper strike did produce obvious reductions in sulfate
and increases in visibility.
100
-------
REFERENCES
AQMP (Air Quality Maintenance Plan), Preliminary Emissions Inventory and Air
Quality Forecast 1974-1995, Final Report of the Boundary and Forecasting Com-
mittee of the Air Quality Maintenance Planning Policy Task Force for Southern
California, April 1976.
Blurnenthal, D.L.. W.H. White, R.L. Peace, and T.B. Smith, "Determination of
the Feasibility of the Long-Range Transport of Ozone Precursors", EPA-450/
3-74-061, 1974.
Cass, G.R., "The Relationship Between Sulfate Air Quality and Visibility at
Los Angeles", Caltech Environmental Quality Laboratory Memorandum No. 18,
Pasadena, August 1976.
C-b Oil Shale Venture, Environmental Baseline Program for Oil Shale Tract
C-b, Nov. 1974- Oct. 1976, Final Report, Available through Area Oil Shale
Office, Grand Junction, Colorado, 1977.
Charlson, R.J., "Atmospheric Visibility Related to Aerosol Mass Concen-
tration, A Review", Environmental Science and Technology. Vol. 3, No. 10,
pp. 913-918, October 1969.
Green, C.R. and L.J. Battan, "A Study of Visibility Versus Population Growth
in Arizona", Journal of the Arizona Academy of Science. Vol. 4, Arizona Uni-
versity Press, pp. 226-228, October 1967.
Grosjean, D,, Post Doctoral Atmospheric Chemist, California Institute of
Technology, Personal Communication, April 1974.
Hagen, L.J. and N.P. Woodruff, "Air Pollution from Duststorms in the Great
Plains", Atmospheric Environment, Vol. 7, pp. 323-332, 1973.
Hartman, W.K., "Pollution: Patterns of Visibility Reduction in Tucson",
Journal of the Arizona Academy of Science, Vol. 7, pp. 101-108, October 1972.
Hartman, W.K., "Air Quality. Pollution Forecast and Application to Planning",
Final Report to the City of Tucson Department of Planning, Planetary Science
Institute, Tucson, Arizona, March 1976.
Hartman, W.K., Planetary Sgience Institute, Tucson, Arizona, Personal Com-
munication of data, April 1977.
Holzworth, G.C. and J.A. Maga, " A Method of Analyzing the Trend in Visibi-
lity", Journal of the Air Pollution Control Association, Vol. 10, No. 6,
pp. 430-435, December 1960.
101
-------
Holzworth, G.C.i'Some Effects of Air Pollution on Visibility In and Near
Cities", Symposium: Air Over Cities. SEC Technical Report A62-5, pp. 69-
88, 1962.
Keith, R.W., "A Study of Low Visibilities in the Los Angeles Basin, 1950-
1961", Los Angeles Air Pollution Control District, Air Quality Report
No. 53, 1964.
Keith, R.W., "Downtown Los Angeles Noon Visibility Trends, 1933-1969",
Los Angeles Air Pollution Control District, Air Quality Report No. 65, 1970.
Korshover, J., "Climatology of Stagnating Anticyclones East of the Rocky
Mountains, 1936-1975", NDAA Technical Memorandum ERL ARL-55, Air Resources
Laboratory, Silver Spring, Maryland, March 1976.
Layden, J., Air Quality Division, Maricopa County Health Department, Phoenix,
Arizona, Personal Communication, September 1977.
Lockwood, G.W. and W.K. Hartman, "Visibility Variations at Tucson, Arizona
and Kitt Peak National Observatory", Publications of the Astronomical
Society of the Pacific. Vol. 82, pp. 1346-1350, 1970.
Miller, M.E., N.L. Canfield, T.A. Ritter, and C.R. Weaver, "Visibility
Changes in Ohio, Kentucky, and Tennessee from 1962 to 1969", Monthly
Weather Review, Vol. 100, No. 1, pp. 67-71, January 1972.
Moyers, J.L., L.E. Ranweiler, S.B. Hopf, and N.E. Korte, "Evaluation of
Particulate Trace Species in the Southwest Desert Atmosphere", Environmental
Science and Technology, Vol. II, pp. 789-795, August 1977.
NEDS (National Emission Data System), 1973 National Emissions Report.
EPA-450/2-76-007, EPA Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina, 1976.
Neiburger, M., "Visibility Trend in Los Angeles", Southern California Air
Pollution Foundation Report No. 11, 1955.
Nieman, B., Meteorologist, Teknekron, Berkeley, California, Personal Com-
munication of data, July 1977.
Oliver, W., EPA Region IX, San Francisco, California, Personal Communication
of data, September 1977.
Pearson, E., Air Quality Engineer with C F & I Steel, Pueblo, Colorado,
Personal Communication, September 1977.
Richard, G. et al., An Implementation Plan for Suspended Particulate Matter
in the Phoenix Area. Vols. 1-4, TRW Environmental Engineering, EPA Contract
No. 68-01-3152, April 1977.
102
-------
Roberts, P.M., J.L. Gordon, D.L. Haase, R.E. Kary. and J.R. Weiss, "Visibility
Measurements in the Painted Desert", Paper No. 75-26.1, Presented at the
68th Annual Meeting of the Air Pollution Control Association, June 1975.
Robinson, E., "Effects of Air Pollution on Visibility", in Air Pollution.
A.C. Stern, editor, Vol. 1, 2nd edition, Academic Press, New York, 1968.
Smith, T., Meteorologist, Meteorology Research Inc., Altadena, California,
Personal Communication, April 1977.
Spicer, C.W. and P.M. Schumacher, "Interferences in Sampling Atmospheric
Particulate Nitrate", Submitted to Atmospheric Environment, Battelle
Columbus Laboratories, 1977.
Taylor, F., Meteorologist, McCarran Airport, Las Vegas, Nevada, Personal
Communication, September 1977.
Tombach, I.H. and M.W. Chan. "Physical, Chemical, and Radiological Character-
ization of Background Particulate Matter in Northeastern Utah", Paper
No. 77-48.6, Presented at the 70th Annual Meeting of the Air Pollution
Control Association, June 1977.
Trijonis, J. and K. Yuan, "Visibility in the Northeast", Report in Preparation
at Technology Service Corporation under Contract to EPA Office of Research
and Development, to be released in December 1977.
Ursenbach, W.O., A.C. Hill, W.H. Edwards, and S.M. Kunen, "Atmospheric
Particulate Sulfate in the Western United States", Paper No. 76-7.5,
Presented at the 69th Annual Meeting of the Air Pollution Control As-
sociation, June 1976.
Waggoner, A.P., A.J. Vanderpool, R.J. Charlson, S. Larsen, L. Granat, and
C. Tragardh, " Sulfate-Light Scattering Ratio as an Index of the Role of
Sulphur in Tropospheric Optics", Nature, Vol. 261, No. 5556, pp. 120-122,
May 1976.
White, W.H., P.T. Roberts, and S.K. Friedlander, "On the Nature and Origins
of Visibility-Reducing Aerosols in the Los Angeles Basin", Caltech Working
Paper, Pasadena, California, March 1975, (Note that this figure also
appears in Characterization of Aerosols in California (ACHEX), by G,M.
Hidy et a! ."T
White, W.H. and P.T. Roberts, "The Nature and Origins of Visibility-
Reducing Aerosols in Los Angeles", Paper No. 75-28.6, Presented at the
68th Annual Meeting of the Air Pollution Control Association, June 1975
(Note that this work also appears in Characterization of Aerosols in
California (ACHEX), by G.M. Hidy et aTT
Williamson, S.J., Fundamental Air Pollution, Addison Wesley, Reading,
Massachusetts, 1973.
Zelden, M., Meteorologist, Technology Service Corporation, Santa Monica,
California, Personal Communication, June 1977.
103
-------
APPENDIX A
LIST OF NCC WEATHER SITES IN THE SOUTHWEST
= Sites of Prime Interest for this Study
ARIZONA
SERVICE STATION
PERIOD
NO. REELS
A
F
W
W
A
W
A
W
W
N
A
F
W
W
A
A
W
W
W
A
0 Chandler /Williams
Douglas/Bisbee
Flagstaff/WBO
§ Flagstaff/Pullian
Fort Huachuca
Gila Bend
Gila Bend
Payson
Phoenix/Sky Harbor
Phoenix/Litchfield Pk.
Phoenix/Luke
Prescott/Mun.
Tucson/Int.
Tucson
Tucson/Davis Monthan
Tucson/Marana
Winslow/Kun. (24 obs 1968)
Yuma/Int.
Yuma/WBO
Yuraa/Sig
* 01/49-12/70
11/48-12/54
01/48-01/50
* 01/50-12/75
0 10/54-12/70
11/48-12/54
* 09/68-12/70
* 07/48-05/52
01/48-12/75
* 04/48-09/66
04/51-12/70
01/48-12/75
10/48-12/75
01/48-10/48
01/49-12/70
09/42-09/45
01/48-12/75
09/48-12/75
* 01/48-09/48
01/55-12/62
4
1
2
3
2
1
1
1
3
2
3
3
3
1
3
2
3
3
1
1
COLORADO
SERVICE STATION
PERIOD NO. REELS
A
F
W
W
W
A
A
N
F
A
W
F
W
W
F
AF Academy
Akron/Washington Cnty.
Alamo sa
Colorado Springs/Peterson
Denver/Stapleton
Denver/Lowry
Denver /Buckley
Denver
Eagle/Cnty.
Ft. Carson/Butts
t Grand Junction/Mun.
La Junta/Mun.
Pueblo
Pueblo /Mem.
Trinidad/Las Animas
* 11/67-12/70
01/48-12/54
* 01/48-12/75
07/48-12/75
01/48-12/75
01/49-06/66
03/61-12/70
* 10/47-03/59
01/48-12/75
* 09/66-12/70
01/48-12/75
01/48-12/64
01/48-06/54
07/54-12/75
# 01/48-09/61
1
1
3
3
3
2
1
2
3
1
3
2
1
2
2
104
-------
NEW MEXICO
SERVICE STATION
PERIOD
NO. REELS
F
A
W
A
F
W
A
F
F
F
W
F
F
A
F
A
F
T7
F
W
A
W
F
W
W
F
F
Acomita
Alamogordo/Holloman
Albuquerque/Int.
Albuquerque /Kir tland
Carlsbad/Cavern City
Clay ton/Mini.
Clovis/Csnnon
Columbus
Eagle
t Farmington/Mun.
Gallup/Sen. Clarke Fid.
Grants/Milan
Hobbs/Lea County
Las Cruces/White Sands
Las Vegas /Mun.
Melrose Gun Range //
Otto
Raton/Crews
Rodeo
Ro swell /Mun.
Ro swell/Walker
Rosvell/Industrial Air Center
Santa Fe/Mun.
Silver City/Grant Cnty.
Truth or Consequences/Mun.
Tur.umc ar i /Mun .
t Zuni/Intermediate Fid.
07/48-04/53
01/49-12/70
07/48-12/75
01/49-01/52
07/48-12/54
* 07/48-12/75
11/51-12/70
07/48-12/54
07/48-05/50
02/52-12/75
* 01/73-1 2//3
05/53-12/54
* 07/48-12/54
* 01/49-12/62
07/48-12/64
* 07/63-12/70
07/48-12/54
07/48-08/53;
* 11/55-11/68
07/48-12/54
* 07/48-01/69
01/49-03/67
01/73-12/75
07/48-12/54
05/60-08/68
05/50-12/75
07/48-12/75
02/49-01/73
1
3
3
1
1
3
3
1
2
3
1
1
1
2
2
2
1
1
2
2
2
1
1
1
3
3
3
105
-------
UTAH
SERVICE STATION
PERIOD
NO. REELS
F
F
F
A
F
y
F
W
W
A
W
F
F
A
t Bryce Canyon
Cedar City/Kun.
Delta
Dug-./ay PG/Michales
Fairfield
Hanlrsvil le
Lucin
t Kilford/Hun.
Ogclen/Mun.
Ogden/Hill
Salt Lake City/ Int.
St. George
0 Wcndover/Auxiliary
Fendover
11/48-12 /75
11/48-12/75
11/48-12/54
// * 12/49-12/70
11/48-07/50
11/49-12/54
11/48-03/50
* 07/48-12/75
ff 01/48-12/54
01/49-12/70
01/48-12/73
11/48-12/54
# * 03/50-12/75
01/49-11/49;
11/56-10/57
3
3
1
4
1
1
1
3
1
4
3
1
3
1
WYOMING
SERVICE STATION
PERIOD
NO. REELS
W
W
W
F
F
W
ft
F
F
F
W
F
CasperAJardell Fid.
Casper/Air Ter.
Cheyerne/1-hm.
Douelas
Ft. Bridger
Lander/Hunt
Laracie/Gen. Brees
Moorcroft
Rawlins/Mun.
Rock Springs/Hun.
Sheridan/Cnty
Sinclair
01/48-
03/50-
01/48-
01/48-
0.1/48-
01/48-
01/48-
01/50-
01/55-
01/48-
01/48-
01/48-
03/50
12/75
12/75
12/54
12/54
12//5
12/54
07/52
12/64
12/75
12/75
-02/51
1
3
3
1
1
3
3.
1
1
3
3
1
106
-------
NEVADA
SERVICE
STATION
PERIOD
NO.REELS
F Battle Mountain/Lander Cnty.
U Elko/Mun.
W Ely/Yelland Fid.
N Fallen
F Fallen
A Indian Springs
W Las Vegas
1\T t Las Vegas /KcCarran Int.
A Las Vegas/Ncllis
F Lovelock/Derby
A Mercury/Camp Mercury
VI Reno/Ir.t.
A Reno/Stead
F ' Tonopah/Mun .
W Winnenucca/WBO
W Winnemucca/Kun.
N Yucca Flats
11/48-12/54
it 01/48-12/75
* 01/48-12/48;
* 01/53-12/73
* 03/45-04/46;
* 01/49-12/7^
* 11/48-12/54
* 09/63-06/64;
* 01/65-12/70
01/48-12/48
12/48-12/75
01/49-12/70
' 11/48-12/75
02/51-06/53:
03/54-05/54
01/49-12/75
08/52-03/66
04/51-12/7-
* 01/48-02/49
* 09/49-12/75
12/61-12/7*,
1
3
3
4
1
1
1
3
3
3
1
3
2
2
1
3
2
107
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1 REPORT NO.
EPA-600/3-78-039
2.
3. RECIPIENT'S ACCESSIOI»NO.
4. TITLE ANDSUBTITLE
VISIBILITY IN THE SOUTHWEST
An Exploration of the Historical Data Base
5. REPORT DATE
April 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
J. Trijonis and Rung Yuan
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Technology Service Corporation
2811 Wilshire Boulevard
Santa Monica, California 90403
1C. PROGRAM ELEMENT NO.
1AA603
11. CONTRACT/GRANT NO.
803896
AG-17 (FY-77)
12 SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTF, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Tnl-pr-im 3/77 - 11/77
14. SPONSORING AGENCY CODE
EPA/600/09
15. SUPPLEMENTARY NOTES
This research was supported under EPA, grant 803896 to Washington Univeristy,
R.B. Husar, Principal Investigator.
16. ABSTRACT
The historical data base pertinent to visibility in the Southwest is analyzed.
The data base includes over 25 years of airport visibility observations and more
than 10 years of NASN particulate measurements. The investigation covers existing
levels of visibility, long-term trends in visibility, and visibility/pollutant
relationships.
Although still quite good, visibility in the Southwest has deteriorated over
the past two decades. The haze levels in the Southwest appear to be mostly the
result of secondary aerosols, especially sulfates. These conclusions are verified
by decreases in sulfates and increases in visibility during the 1967-1968
industry-wide copper strike.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
* Air pollution
* Aerosols
* Sulfates
* Visibility
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI l;ield/Group
Southwest
13B
07D
07B
I'J fJlS miBUTION STATEMENT
KKLKASK TO I'llIU, 1C
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
118
2O SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
108
------- |