-------
-------
StOU
?. ft!
1
a
.
TABLE OF CONTENTS
Page
Visibility Task Force Representatives -.... iii
Summary; of* FJ ndi ngs and Recommendati ons ...;...........
Introduction. .......,..:........ .*..;. 1
*v
. Summary;'of Findings'..... ... 1
1^3 .{Cha'pacterlzat ion ".of Current Regional Visibility.-...
$S . ;.-Conditions.: and Historical Trends 3
\sj Sources of Haze and Projected Trends ......................... 3
jy ...Value- of Risibility .>..........;... -11
' Alternative .Control Strategies 15
I ;Q Publi'c/Commerits .'................;-...- 17
OQ ;.', - ' .
Im Recommendations "....-.'.. 18
-v. ,-" ' .'
-J\ -.: Research .Needs 18
|*tp - Policy Analysi s;, .v..........: 19
Interim Regulatory and Legislative Considerations.'............. 20
References ;......... 23
I Appendix A>; - Characterization of Current Regional Visibility, A-l
Trends and'Pollutant Visibility
-x Relationships '
| Appendix B'i Projecting Future Regional Trends 8-1"
» Appendix C. Value of Visibility C-l
- Appendix D. Alternative Control Strategies for Regional Haze 0-1
I Appendix E.:' Research Needs ." E-l
Appendix F.- -Summary of Public Comments F-l
I
J
U.S. Environmental Protection Agency
Library, Boom 2404 PM-211-A . .
40i;M Street, S.W. ' '" : ''
^ Washington, DC 20460 : '
1
-------
I
I
1
-------
1
_
1
1
1
1
1
1
1
1
1
1
1
J
1
I
1
VISIBILITY
ENVIRONMENTAL PROTECTION AGENCY
. Office of ATr and Radiation:
John Bachmarm
Jim D.ieke .
Tom Pace
Bruce Polkowsky
Vivian Thomson
Dwaine Winters
Office of General Counsel:
Dorothy Patton
Sara Schneeberg
Regional Offices:
William Ballinger, Region III
Dave Bray, Region X
Butch Rachal, Region VIII
James Wilburn, Region IV
NATIONAL PARK SERVICE
David Joseph
Bill Malm
Brian Mitchel
U.S. FOREST SERVICE
Jim Byrne
Charles ficMahon
BUREAU OF LAND MANAGEMENT
Stan Coloff
iii
TASK FORCE REPRESENTATIVES
Office of Policy, Planning and Evaluation.
Alex Cristofaro
Tom Lareau
Sidney Worthington
Office of Research and Development:
Al Galli
Marc Pitchford
Charles Rodes
William Wilson
DEPARTMENT OF ENERGY
Roger Morris
Ted Williams
TENNESSEE VALLEY AUTHORITY
Larry Montgomery
DEPARTMENT OF DEFENSE
Richard Boubel
-------
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
.DEVELOPING LONG-TERM STRATEGIES'FOR REGIONAL HAZE:
FINDINGS AND RECOMMENDATIONS OF THE VISIBILITY TASK FORCE
INTRODUCTION
This report has been prepared by an interagency task force composed of
professional staff members from a number of EPA offices, the National Park
Service, the Forest Service, the Bureau of Land Management, the Department of
Energy, the Tennessee Valley Authority, and the Department of Defense. The
task force was established by EPA's Deputy Administrator in January 1984 to
examine the issue of developing long-term strategies for visibility impairment
from pollution-derived regional haze and to recommend a long-range (5-10
year) program to address it.
The task force report consists of this summary of findings and recommenda-
tions plus independent appendices covering the major subject areas listed
below.
A. Characterization of current regional visibility conditions
B. Projecting future regional visibility
C. Visibility values and other criteria for evaluating alternative
strategies
D. Alternative regulatory strategies
E. Research needs
F. Public comments to the task force
These appendices primarily reflect the perspectives and judgments of the
individual authors and contributors listed on each appendix and have not .
undergone any formal peer-review process. They are not the official position
of the agencies and offices represented on the task force. The task force
also has prepared an interim research needs document (March 1984) and a
Federal Rejjster notice and supporting .materials soliciting public comment on
regional haze issues (November 1984). In addition, two draft contractor
analyses have been prepared in support of the task force (SAI, 1984; SAI, 1985).
SUMMARY OF FINDINGS
Historically, visibility impairment has been among the most frequently
reported effects of air pollution. The Clean Air Act of 1970 authorized .
protection of visibility generally through the ambient standards (Sections 108-
110) and other programs intended to protect public .welfare (Section 302h).
In the 1977 amendments to the Act, Congress called' for special protection of
visibility in certain Federal lands such as national parks and wilderness
areas under Sections 169A(a)(l) and 165, and established "as a national goal,
-------
the prevention of any future, and the remedying of any existing, impairment
of visibility in mandatory class I Federal areas which impairment results
from manmade air pollution."
It is useful to refer to two categories of visibility impairment:
1) visible plumes of smoke, dust, or colored gas that obscure the sky relatively
near their, source of emission, and 2) regional haze, which reduces visual
range and the contrast of form, color, and texture and which can occur on a
geographic scale ranging from a large urban area to multistate regions. In
some transition cases, individual plumes or hazes can appear as bands or
of discoloration.
Haze is the result of light scattering and absorption by particles
(mostly fine particles smaller than 2.5 urn) and to a much lesser extent, by
r""N02. The composition of haze particles varies with time and location but
1 usually includes "secondary" particles (e.g., sulfate, nitrate, and organic
\ species formed in the atmosphere from "precursor" gases including SOg, NOX,
\ and VOC), carbonaceous materials (elemental carbon and some organics), condensed
(atmospheric water and other trace components. Important sources of hazes
include natural sources (blue sky scatter, fog, dust, forest fires, volcanoes,
sea spray, and biologic sources) and anthropogenic sources of sulfur oxides,
particles, nitrogen oxides and volatile organ-ics. Humidity and other
meteorological variables have a major influence on the nature of pollution-
derived haze. The long-range transport and atmospheric transformations of
precursor gases and haze producing particl.es add substantially to the complexity
of source receptor relationships.
State and local visibility regulations and the first phase of Section
169A visibility requirements (1980, 1984) have focused on readily identifiable
sources of visible plumes. Decisions on programs for regional haze - both
for class I areas and in the context of ambient air quality standards - have
been deferred for several related reasons:
1. the complexities associated with multiple pollutants, long range
transport, and atmospheric transformations;
2. the need for improved scientific and technical information; and
3. the absence of any coordinated examination of how regional haze
programs might be integrated with current or future programs for related
pollutants (e.g., SOX, NOX, PM, VOC) and problems (e.g., acidic deposition,
regional oxidants).
Indeed, much remains to be learned about the origins and composition of
regional haze, source receptor relationships, perception, and values through
continued research. Nevertheless, a substantial body of data and analytical
capability exists at present. When understood in light of the uncertainties,
these can provide useful information on long-term .'approaches to regional haze
problems. The following discussion outlines major findings in each of the
key areas addressed by the task force. These. findings form the basis for the
recommendations that are presented in the concluding section.
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Characterization of Current Regional Visibility Conditionsand Historical Trends
Because active visibility monitoring networks have only recently been
established, human observations of visibility at airports have been the main
data used to examine the geographical distribution and trends in the intensity
of regional haze. These data have recognized limitations, but are useful in
distinguishing overall trends and geographical distributions. The airport
data show a major difference between East and West with substantially higher
visibilities in the Rocky Mountains and desert southwest (Figure 1). Analysis
of trends from 1948 to the present show clear shifts in seasonal airport
visibilities in various regions. In the northeast quadrant of the U.S.,
wintertime, visibilities appear to have.improved, while the intensity and
areal extent of summertime regional hazes worsened from 1948 to 1972, with
sorae improvement since 1972. In the Gulf Coast and South Atlantic States,
visibility appears to have deteriorated in all seasons. Current median
visibility (Figure 2) in the East ranges from 8 to 25 miles (13 to 40 km)
with summer as the season of worst visibility. Based,on the available information
anthropogenic particulate pollution appears to dominate eastern regional
haze. Estimates of natural background visibility in the Appalachians range
between 30 to 50 miles (50 to 80 kilometers), while actual summertime visibility
tends to.be 1/3 to 1/5 that distance (Trijonis, 1981; Ferman et al., 1981).
More detailed information on regional visibility in the West, derived
mostly from the National Park Service network, is displayed in Figure 3.
Present day visibilities in some areas of the West often approach the
theoretical limits set by air molecules, corresponding to a visual range
on the order of 400 -km (240 miles). In such conditions, even small additions
to regional fine particle loadings can result in a perceptible degradation
of visibility. The airport data from 1948 through 1976 indicate mixed trends
in the southwest with a general decrease from the early 1950's to the early
1970's and some improvement thereafter, especially in southern Arizona. In
California, airport visibility declined in most areas from 1948 through 1966
and improved thereafter, but overall trends differed in various locations.
Trends are also mixed in the Pacific Northwest and the Northern Plains.
A view of regional visibility conditions in class I areas is afforded
by the preliminary subjective judgment of the Federal Land Managers response
to a 1978 analysis (Figure 4). In that analysis, about one-third of the
managers reported undesirable conditions and/or the need to evaluate suspected
anthropogenic impacts, usually haze ascribed to various source mixes.
Sources of Haze and Projected Trends
Information on sources of visibility impairment comes from fine particle
and visibility monitoring (Table 1), receptor modeling, emissions inventories,
predictive models. Although current data do not permit precise quantitative
conclusions, the major pollutants and sources involved in various regions
can be identified. In the East, important sources, of haze producing pollution
include sulfur oxide emissions from utility and industrial coal and oil
combustion; primary carbonaceous particles from these and other industrial
sources, heating, automobiles and fires; and organic emissions from various
sources. Because sulfates currently may account for 40 to 80% of regional
haze in much of the East -- with less known about remaining fine particle
-------
« a. vn t/i
t- re fli
it) W OJ I-
S =
i- O) *J HI O
4-1 31 .M in
= C ro
'«'- .C a)
-31-
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
<'Y i
(if** »«! MdtM
(*rth«*t
FIGURE 2. Observed median visibilities (miles) at selected airports
for 1982 and historical isopleths (1974-76). Top Figure, annual
median; Bottom Figure, summertime median (SAI, 1984).
-------
I
FIGURE 3. Isopleths of standard visual-range (km) over the western
United States for the summer of 1982. Squares reflect data from
NPS, teleradiometer network. Circles reflect airport visibility
readings by human observers (Malm et al., 1984).
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
8
C <- * » >
» <^ u * wl
W -0 C 5
> *j 3! * 4J
». o "3
O 4! >
> -» C
in «>*
4J M wl -3
<« 4t 4* >i
^ C 3 't
tS >/i .= S
HI 40 11 3
Ol i- £ u
c a.c c -
u"~ 3 C
a c v
iJ? *** * *
o w u a e*
>, o^ii ^ £
s ii «
y u
.a 5 U
i- -3 C I/I O
fl 3 *
> <- 3.= 4J
^, >, o a M
,2 » ^- -5
w ^ ^ ^ a ~ ~
'"" "^ * '^ O
IQ a. s u a e
O> i/l 9 31 9» o
u 31 a, £,
« »
-* ui ** S «^ c
t. a. in g -o 5
^Q (9 O U **> U
> e 2 £
t/i ^^» 3 a
«i ii « « a.
§& "Si . u-
S C w 3
«. win S u
*y *j 41 u s
o t» * a ^
t« « -S 3
§ - ^ « t
5 S S il
^ 01 -J ~ S
Q U fl >,4I
Ji 1 1) *
U 3 > i
2 jj^S"
SA V
^ u 4)
Ii J*3
iill-
~ '"3. "3 -
sl'2« *
a ^ -j "3
'_iJ 3 ) ^ u)
^f 3i 2
^ i e 'c S
>
Qj ^1
00 3J
2 ^
^ J^
u
>, vrt
U .C 3
A
3 "5
v) O
11
ff>
c.
z
v,
o
U tfl
fl! -^
3 -j a
«S 3
t) VI W
1
S
'S
e
9
u
3
- <^ « » «
* s s - g
§-
s s s s
«c «< ft i <:
a s s ns
/~~N
(M r>. CO \ 38 \ VI CX) -^ ^ '
»; <^^ ^^ ^» ^cj ^
52 "r'a lr'° 5^ ""*
"* ** "* *rf U U IQ IQ
e v.^uti 33 e
aw OCO«. 3 5
J 9.130.41 C 2
C O ^ -^ k "3 >^h o 4 wi
s i- «.««9oi z e ._
* ** eg ^ 19 dl r<*» i 1) a
-> 11 _J » *J J= 9
w* O o SS
1 1 1
l~ O 3H M Ot O
,
a S
?' i 2J o ,**i a
JT» t^ -jj ^ ^^
^ 1 1
4rf A 4)
0) T3 3 X
^ c B^aiM i/i'Q^vU
^& o ^>t* c^ e oic^ac
=> » >3*3 * S ° 5 5> | c
' 3 B 3 in .a i J ^» ^ 5 '
UO tl) N <-> O O w i <^
w> Z , = a U j ae uii-^ »
3
O
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
mass -- and because of the relationship of regional sulfate loadings to the
acid deposition phenomenon, the task force has placed most attention on
projecting future trends and control options for regional sulfur oxide emissions.
Historical emissions analysis show.that the long-term trend of increasing
SOX emissions was halted with the onset of Federal, state and local control
programs in 1970-72 and that from 1974 to 1983, SOg emissions nationally
declined by about 23% (EPA, 1984). Projected trends for emissions in 31
eastern States indicated that without further control, a modest (10%)
increase in emissions is possible through 1995. Assuming no major changes
in other haze related pollutants, this would probably lead to little overall
change in eastern haze. Depending upon factors that are difficult to
predict with confidence, in the long term future (2020 to 2030) substantial
decreases in SOX emissions might occur through improved new source controls
and retirement of existing facilities. Implementation of potential acid
deposition or other control programs could result in significant decreases
sooner.
The draft contractor analysis of eastern visibility (SAI, 1984) used
a regional air quality model (RTM-LT) to examine visibility impacts of
alternative SOX rollback scenarios, including several acid rain programs.
Factors such as uncertainties in emission projections and treatment of important
meteorological, transport, and transformation processes impose significant
limits on the results. A detailed review of the study sponsored by the
Utility Air Regulatory Group (UARG) suggests that the results may overstate
visibility improvements. Comments from ORD task force members suggest components
of the analysis may understate improvements. The analysis is currently
undergoing external peer review. The available results do suggest, however,
that:
1. large (30 to 50%) reductions in SOg emissions in 31 eastern states could
result in substantial improvements (15 to 33% or more) in annual and summertime
j/isibility in large portions of the East;
C" 2. smaller emissions reductions (e.g., £20%) spread over 31 states or
I more targeted acid deposition approaches are less likely to result in perceptible
( improvement over current baselines; and
<-- 3. simultaneous reductions in other visibility impainjvg pollutants
\ (fine particles and possibly VOC and NOX) might markedly improve the
\ perceived effectiveness of SOX reductions. Conversely, Increases in such
) pollutants could negate some of the benefits associated with SOX control.
In the West, important sources of regional haze vary with location but
can include smelters^ urban "plumes" consisting of parti/cles derived from
numerous area and point sources, energy development activities, fugitive
dust, and controlled burning. Preliminary efforts to apportion sources by
receptor and predictive modeling are in reasonably close agreement, but
important uncertainties remain, particularly in identifying the importance
of nitrates, organics, and carbon and identifying the relative influence of
various source regions (Table 2). Under current regulatory programs, preliminary
estimates indicate relatively modest growth through 1995 in total emissions
in six southwestern states from these categories, with' a potential substantial
-------
10
TABLE 2. MODELED CONTRIBUTION BY POLLUTANT AND SOURCE CATEGORY TO ANNUAL
AVERAGE ANTHROPOGENIC-DERIVED HAZE IN THE GRAND CANYON (SAI, 1985). Results
fliustrate model outputs. Although In rough agreement with empirical studies
of anthropogenic source influence (Appendix A), such predictions are extremely
uncertain. At relatively clean sites such as this, "blue sky" scatter by
air molecules is larger than the average anthropogenic extinction.
Percent (%) Contribution
Source
Category
Utilities
Smelters
Oil Industry
Other Industrial
Sesidential/Connn
Gasoline Vehicles
Diesel Vehicles
Other Mobile
Evaporative
Fugitive Dust
fres Burns/F.P.
Wood Stoves/F.P.
Extraregional Trans
SG4
£
15
10
3
0
1
1
2
0
0
0
0
e
.9
.2
.2
.9
.9
.1
.3
.7
.0
.0
.0
.0
.5
ND3
7.
0.
1.
2.
1.
6.
4.
1.
0.
0.
0.
0.
0.
9
0
3
4
3
9
4
8
0
0
0
0
0
QHG
0
0
0
0
0
0
0
0
0
0
0
0
0
.0
.0
.0
.0
.0
.3
.1
.0
.2
.0
.0
.0
.0
Other Coarse
Soot EW-2.5 PM-10
0.0
0.0
0.0
0.0
0.0
0.6
0.8
0.0
0.0
0.0
2.3
0.1
0.0
1.1
0.9
0.1
0.1
0.1
0.6
0.1
0.0
0.0
7.0
1.1
0.0
0.0
0.1
0.1
0.0
0.1
0.1
0.4
0.0
0.0
0.0
2.6
0.0
0.0
0.0
N02
1.5
0.0
0.2
0.3
0.2
0.8
0.5
0.2
0.0
0.0
0.0
0.0
0.0
Total
17.5
16.2
11.7
6.3
2.6
10.7
7.3
4.8
0.2
9.6
4.0
0.1
8.5
1C1M.
50.7 25.9 0.8 4.3 11.0 3.6 3.6 100.0
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
decrease in U.S. smelter emissions (Pechan, 1985). The major near-term
increase in emissions of concern to regional haze is< from new smelter construction
in Northern Mexico, but some growth is also projected in primary particle
emissions fron mobile and other area sources.
In addition to projecting emissions for this region, the draft study
done for the task force used a regional model (RIVAO) to examine potential
visibility impacts under a "base" and two alternative emissions growth scenarios
(SAI, 1985). The preliminary nature of the modeling tools and large uncertanties
in current and projected emissions, meteorological and other inputs, and
treatment of dispersion, chemistry, and removal cannot be overemphasized.
Nevertheless, the available results can provide useful insights. Figure 5
displays the change in regional visibility by 1995 predicted by this exercise
under current regulatory programs. The analysis suggests that, with
implementation of current programs and the expected low growth in energy
development, no large decrements are likely in regional visibility in the
southwest for the near term. In the longer-term future (2UOO and beyond),
however, continued population increases and growth in energy development
offer the potential for more significant increases in visibility impairing
pollution and more serious impacts on the regional visibility resource.
Value of Visibility
Regional haze can affect public welfare in essentially two areas:
1) the subjective enjoyment of the environment {aesthetics, general well *'
being) and 2) transportation. Evidence of such effects can be drawn from
studies of perception and social awareness of air pollution and scenic beauty,
economic studies, and visibility/air transport relationships. Key examples
are illustrated in Table 3. The available information indicates that the
public is concerned about visibility and is willing to pay for maintaining or
improving air quality in a variety of contexts. Our ability to calculate this
value in comprehensive and reliable quantitative terms is, however, limited.
A number of economic studies of visibility values in urban and non-urban
contexts have been examined in detail by the task force (Appendix C). These
have been used to provide estimates of economic benefits associated with
strategies that improve visibility (Table 4). They suggest that visibility
benefits might pay for a substantial fraction of control costs, even in the
East. Major uncertainties in such estimates must be considered. These
derive from:
1) available survey approaches (contingent valuation) nay provide biased
estimates and cannot be verified for a non-market good like visibility;
2) competing claims for multiple public goods may lower estimated visibility
values;
3) methods for extrapolation of results to smaM changes in average visibility,
over tine and consideration of the full temporal distribution of visibility
changes are quite limited by current information on perception; and
4) options and existence values are not well understood.
-------
12
Figure 5. Percent change in modeled annual visual range from 1980 to 1995
(base scenario). The predictions suggest that with implementation of current
programs as planned, regional visibility in 1995 will not be substantially
different from current levels in most of the southwest. The estimates are,
however, sensitive to uncertainties in emissions, transport and transformations
meteorological data and other model related factors (SAI, 1985).
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13
TABLE 3. SUMMARY OF QUALITATIVE EVIDENCE FOR VISIBILITY RELATED VALUES
Effect of Increased
Visibility
Affected Groups
Supporting Observation
Aesthetic:
Perception
Decreased perception of
air pollution
a) urban settings
Substantial percentage of general
population; groan areas
b) natural settings
Improved view of
night sky
Outdoor recreatlonlsts, campers,
tourists
Amateur astronomers, other
star watcners '
Perception of air pollution 1n Los Angeles
significantly related.to visibility for
all averaging times (Flachburt and
Phillips, I960)
Perceived visual air quality significantly
related to particle scattering (Mladleton
et al.. 1984)
Visual air quality related to awareness of
haze for Grand Canyon visitors {Ross
et al., 1984)
Scenic Beauty estimates related to contrast
transmittance (Malm et al., 1981)
Decrees* in star brightness by fine
particles (Leonard et a 1.. 1977)
Economic
Studies
Increased property values HOB* owners
Enhanced enjoyment (user
or activity values) of
environment in:
a) urban settings
b) natural settings
Options values; main-
taining or increasing
opportunity to visit
less Impaired natural
and urban settings
Existence values,
maintaining pristine
environments
Urban dwellers
Outdoor recreatlonlsts.
campers, residents of
non-urban areas
Outdoor recreatlonlsts.
campers, tourists
General population
Property values related to perception of
air pollution, hence visibility (Srooksnirt
et al., 1979; Thay and Trijonis. 1981
Unman, et al., 1981)
Willingness to pay for Increased visi-
bility in several eastern and western
urban areas (Brooksnlre et al., 1979;
Tolley et al., 1984; Rae, 1984)
Willingness to pay for preservation,
improvement and accept compensation
for degradation in national parks
(Rae et al.. I960; ftae et al., 1982)
Aggregate of activity values In Itera-
tive bidding studies suggests Importance
of options values (Rowe and Chestnut, 1981)
Existence values may far outweigh activity
or user values (Rowe and Chestnut, 1981)
Psycno- enhanced enjoyment in
logical natural areas (user and
Studies options)
Less concern over
perceived health
effects
Outdoor recreatlonlsts,
campers, tourists
Seneral population,
urban areas
Visual air quality attributes ran* nign
among visitors to both Mesa Verde and Srand
Canyon. Knowing park resources protected
most valued psychological attribute
(Ross et al.. 1984)
About 2/3 of bid for improved visibility
1n Los Angeles was related to concern
over potential health effects (Srooksnire
et al.. 1979)
Transports- More efficient, lower
tion risk operations, visual
approach permitted
Increased opportunity
to operate aircraft
Increased opportunity
to conduct aircraft
and weaoons testlno
Airport users, operators
General aviation aircraft
(non-instrument capable
pilots, aircraft)
000 testing facilities
1n Desert Southwest
Visual approaches permitted *nen visi- ",
blllty >3^5 riles; airport specific
(FAA. 1980a)
Visual flight rules permitted when visi-
bility >3 miles (FAA, 1980b)
Basis for location of facilities 1n part
due to high visibility
-------
14
I
. TABLE 4. ECONOMIC VALUATION OF EASTERN HAZE REDUCTION (Appendix C)
4A. EASTERN URBAN SURVEYS (CONTINGENT VALUE STUDIES). Based on interviews of urban resident!
using photographs to depict large visibility changes. While subject to potential limitations
and biases (see text), they are the best means for economic appraisal of some non-market
environmental goods like visibility. Data used to fit exponential bid function for valuing I
smaller visibility changes in the control scenarios below. ~
Study
Change in Visual Range
Valued (miles)
Willingness to Pay
S/mile/yr/household
J
Chicago; six
Cities (Atlanta,
Boston, Cincinnai,
Miami, Mobile,
Washington)
(Tolley et al.,
1979)
Cincinnati (Rae, 1984)
Survey 1
Survey 2
10 to 5 (annual)
10 to 20 (annual)
.10 to 30 (annual)
8 to 25 (most of time)
11.5 to 16.5 (most of time)
13-51 (avoid decrement)
6-26
4-17
24-34
79-110
I
I
I
I
I
4B. ESTIMATES OF VISIBILITY BENEFITS ASSOCIATED WITH $02 CONTROL SCENARIOS. Even modest
bids and changes in visual range can lead to substantial benefits estimates when aggregated I
over the urban population of 31 eastern states. Results for such changes are, however,
highly sensitive to assumptions about the extent to which people can perceive such changes.
SOX Scenario
(ICF. 1984)
Annualized Costs
(Billion S/yr)
Visibility
Change*
Alternative Benefit Estimates
(Billion $/yr)2
i
Low Bier+ T5*
"perception"
.threshold
Mean Bid; "percep-
tion" threshold
15% 5*
High Bid + _
No percept iB
threshold
10 State
"8" million
ton 31 State
"12" million
ton 31 State
1.7
3.7
9.2
5%
11%
16%
0.2
0.6
0.4
1.3
0.4
1.1
1.6
1.3
2.9
4.1
I
I
I
^-Predicted annual increase in visual range averaged over 31 states based on regional
modeling. May over (or under) estimate improvements. -
2Low bid ($6/mile/household-yr) and mean bid ($12)based on Tolley et al., data. High bid
($40) from Rae. None include options or existence values* Alternative treatment of thresholds
for the Utility Air Regulatory Group leads to substantially lower estimates than shown here
(Zanetti et al., 1985; Rudd, 1985). .
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
15
Alternative Control Strategies
In considering potential long-term regulatory strategies for regional
haze, it is useful to treat the regional haze phenomena in the East and
West separately. Regional haze characteristics, intensity, sources, and
associated air quality problems are sufficiently different between West and
East that distinct control program goals and approaches should be considered
within each region. Furthermore, specific control approaches could be
tailored for certain source categories, sub-regions, or urban areas in both
East and West.
. Given the complexity and variability of regional haze, both from a
pollution and public perception standpoint, it is questionable whether any
single regulatory approach will have the flexibility needed to provide
efficient and effective control of regional haze. Rather a combination of
approaches - - some designed specifically to protect visibility, some
implemented to meet other air quality objectives - - may be more appropriate.
The task force examined a number of alternatives for limiting or
reducing emissions of precursors of regional haze. The various approaches
can be categorized into three classes with respect to legislative authorities:
1. those capable of being implemented using the existing legislative
and regulatory authorities in the Clean Air Act;
2. those requiring modifications to existing legislative authorities; -
and
3. those that rely on addition of new legislative authorities.
The approaches also can be categorized according to environmental goals:
1) reduction of regional haze as a primary objective, 2) regional haze
control as a partial or supplemental objective, or 3) reduction of some
other adverse effect. The degree to which haze reduction is used to justify
regulatory action, then, forms a second conceptualizing dimension.
These two categorizations are used to organize alternative strategies
in Table 5. To the extent possible, each of the approaches listed within
the matrix has undergone a preliminary qualitative examination with respect
to appropriate criteria including cost-effectiveness, equity, ease of
understanding, enforceability, flexibility, compatibility with existing air
pollution control programs, and -acceptability to affected parties (Appendix
D). As noted above two.major approaches - various SOg rollbacks and targeted
acid deposition scenarios were examined more quantitatively in the contractor
analyses.
Based on its examination of strategies as they might be applied to the
somewhat distinct problems associated with regional haze in the East and
the West, the task force offers the findings listed below.
1. Maintaining current regulatory programs will likely result in
approximately the present level of episodic regional haze in the East
-------
1
TABLE 5. MATRIX OF REGULATORY OPTIONS* 1
1
1
AUTHORITY
A. Existing
B. Modified
Existing
C. New
GOAL "- '
i
I. Haze Principal ly
1) Secondary NAAQS -
fine particles, sulfates,
extinction, other
2) "Phase II" visibility
programs-class I areas
'
1} Codify integral
vistas
2) Modifications to
16 9A
3) Modify PSD
4) State based programs
,
1) Visibility
Standards
II. Haze Secondarily
1) PSD Requirements
2) Interstate Pollution
Control
3) International Air
Pollution
4) NSPS, BACT/LAER major
source categories
1) Modify Interstate
Pollution Control
2) Regional Secondary
NAAQS; NAAQS
"Criteria"
3) Episode Control
4) Revised NSPS
requirements
5) Requirements for
"Nontraditional"
Sources
1) SC-2 Rollback
2) Age Based Controls
1 1 1. Other Object! v«
1) Primary NAAQta
PM, S02 |
2) Motor Vehicl<=
Emissions
Standards
1
~, ..,.-.>..,
trading
2) "Reasonable |
Effort" for
SNAAQS _
1
3) Modify PSO "
1
i
II
!
1
1!
!
I
Ii
i
^Options are described and' discussed in Appendix D. 1
1
1
1
1
-------
I
17
2. The analysis of $03 control strategies suggests that the benefits of
improved visibility estimated for SOg roll backs are uncertain, but might be
\ substantial. Even at the higher end of the uncertainty range, however, they
^-are unlikely to equal or exceed control costs by themselves. Some of the
more "targeted" strategies designed to control acid-deposition produce visibility
improvements that are-limited in extent and magnitude. If accelerated reductions
of 502 emissions are judged appropriate, the most effective long-term regulatory .
approach would be to develop cost-effective regional strategies designed to
maximize all of the known and anticipated multiple environmental benefits of
$03 control, including haze reduction. Implementation of such a.strategy is
best accomplished with new legislation.
3. Although less timely and more difficult to implement than a mandated
control strategy, if it is decided that accelerated reduction of eastern haze
is warranted, consideration should be given to developing a fine particle
SNAAQS. Outside groups may attempt to force action on such a standard in any
case, and in the absence of new legislation mandating regional S02 reduction,
a standard could guide long-term sulfur oxides and other control programs by
establishing an acceptable target. Some modifications to current Clean'Air
Act authorities might improve the design and implementation of SNAAQS for
fine particles, especially if haze control is the major objective.
4. A supplemental (or alternative) long-term strategy, for reducing S02
is control of sources-based after a defined source life-time. This could act
to ensure and accelerate reductions projected from retirement (or control) of
higher emitting existing -sources and their replacement with cleaner new
facilities.
5. Protection of western visibility may be effected well enough without
new legislation. Assuming implementation of current regulatory programs, the
western analyses suggests that delaying development of new regional haze
.programs for several years while awaiting improved source-receptor information
would probably not result in unacceptable or irreversible degradation. As
new research results improve our ability to predict regional impacts of
emission source limitations, "phase II" regulations can be developed under
the existing section 169A. Protecting visibility in class I areas while
implementing existing NAAQS, PSD, and mobile source controls will provide
substantial protection against impairment from haze throughout the West.
Public Comments
Oral comments were received from interested parties at two public meetings
and written comments were also submitted in response to a Federal Register
notice (Appendix F). The meetings were well attended and the discussion,
particularly at the Denver meeting, was informative and productive.
Among the points made by various commenters are the following:
1. A concensus exists among industry and environmental groups that
EPA should expand research on regional haze and facilitate coordination
among public and private sector research.
-------
I
18
2. A consensus also exists on the desirability of integrating visibility
programs with other air quality management programs and evaluating multiple
benefits of particular strategies.
I
|^i V yi UHl J TVIVIl W i* I JCI UIF v^ w "-* « » v^j IMUI i i* ^j t«.mv^ i i t* |*« I vy ^ i uuij
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
19
work should also be closely coordinated with related programs, particularly
acidic deposition.
2. EPA should maintain or encourage research in the following areas:
i) studies of human perception and improved indices of visibility;
ii) development of improved monitoring;
iii) characterization of current regional visibility conditions
and pollutant/visibility relationships in representative regions;
iv) investigation of important source categories of visibility impairment;
v) improvement of emissions inventories and models for predicting
visibility impacts; and
vi) studies on the value of visibility.
These general needs are discussed in Appendix E. Near-term research priorities
for EPA budget planning are more fully documented in the interim needs statement
and have already been factored into FY-85 and FY-86 budget planning.
3. EPA, in cooperation with other groups, should provide a periodic
summary of current information in these key areas for research planners and
for policy makers. This function could be accomplished in association with the
research coordination work.
Policy Analyses
Current understanding of regional haze, values, and source-receptor
relationships permits the identification and some analysis of alternative
emission control approaches and regulatory strategies. Nevertheless, important
information gaps remain. This state of knowledge suggests the following
approaches for developing a long-term haze strategies:
1. wait for new research efforts to address key unknowns before
evaluating strategy alternatives in detail;
2. recognizing the limitations and uncertainties, conduct in-depth
analyses of key strategy alternatives using available tools;
3. use regional haze impacts analyses to help guide decisions on
regulatory or legislative alternatives whose principal goal
is not related to, regional haze; and
4. use regional haze impacts analyses to develop and establish
regulatory approaches whose principal goal is regional haze control.
The choice on the most appropriate course -- particularly for the normative
judgments involved in the use of results in options 3 and 4 must be made by
the responsible decision maker, after considering the available data and the
-------
2Q
nature of the current and projected problem. The task force has found both
the eastern and western modeling analyses to be of significant value for
answering and raising important policy and research questions. Task force
recommendations with respect to future analyses, are listed below.
1. EPA should encourage development and evaluation of more advanced
policy analyses tools for urban and regional scale haze. Following the peer
review of the present analyses, OAR, ORD, and OPPE should develop consensus
criteria for conducting haze related benefits analyses using readily available
tools and specifying relevant uncertainties. Coordinated efforts to enhancing
current capabilities should be promoted. Improved approaches as needed for
dealing with perception and predicting episodic as well as seasonal and
average changes.
regional
i)
ii)
Based on the
haze control
examination of the alternative strategies, the following
alternatives should be examined in future policy analyses:
Alternative fine particle and sulfate NAAQS, with a focus on the
eastern U.S., plus a typical western urban area. This approach
should examine projections and controls for non-sulfur related fine
particles as well as 862;
Accelerated NSPS, age based control options. DOE has already projected
emissions for such alternatives. The timing of extent and value of
possible visibility benefits should be examined. The regional
effects of planned and alternative control of mobile sources, particuarly
diesels, should also be examined further;
iii} Promising alternative cost-effective S02 rollbacks in the East as
they are developed; and
iv) Alternative offset, emission limit policies for smelters, urban
areas, new sources, and prescribed burning in the West. This would
follow up the projections for current scenarios with an examination
of various control approaches.
In conducting such analyses, where possible, the work should include or
facilitate calculation of non-visibility air quality related benefits associated
with each strategy.
3. Analyses of acid deposition control strategies, by EPA and NAPAP, and
other regulatory and policy analysis of strategies that might effect visibility
on a regional scale should^ where possible, include or facilitate analysis of
visibility benefits.
Interim Regulatory and Legislative Considerations
As noted above, the available projections suggest that regional visibility
in both the East and West is likely to be stable'for the next decade - the
East experiencing continued episodic regional.haze and the West maintaining
relatively good current conditions. Smaller areas within these regions may,
however, experience visibility degradation depending upon local emissions and
meteorology. In this context, the task force makes the following recommendations
regarding legislative initiatives and regulations:
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
21
1. Decisions on .legislative recommendations or regulatory policy that
would affect acid deposition or regional sulfur oxide emissions in the East
also affect the occurrence of regional haze. Although actions might not be
taken solely on the basis of reducing eastern haze, policy makers should
recognize this linkage.
2. Although projections suggest development of comprehensive "phase II"
haze programs under Section 169A can be deferred in the West, this finding is
predicated on vigorous enforcement of current regulations, particularly those
designed to attain the current NAAQS. For example, major $03 emissions
decreases are projected from U.S. smelters under current programs. It is
also desirable to continue negotiations to achieve controls of the two
large Mexican smelters.
3. In the interim, ongoing and planned implementation of current visibility
regulations under Sections 169A and 165 should, wherever possible, proceed in
such a way as to permit an orderly transition to subsequent phases. Specific
recommendations for achieving this are made in the areas listed below.
i) Documentation of current impairment by regional haze. EPA will
promulgate a monitoring strategy and new source review requirements for many
States in June. Background information from such monitoring will include
information on regional haze episodes; the monitoring strategy should facilitate
accurate recording and analysis of such impairment.
ii) Criteria for impairment. EPA and the Land Managers have not developed
specific, quantitative criteria for determining whether impacts of new sources
are adverse. This will eventually lead to implementation problems. In
development of these criteria, definitions applied to plumes may have implications
for future programs on regional haze and so should be specified carefully.
iii) Resolution of integral vistas. EPA is currently acting to resolve
issues on integral vistas (views from within the class I area to points
outside the park) with the land managers. Inclusion of such views might
extend the effectiveness of regional haze programs in some cases, but make
little difference in others.
iv) Long-term strategy coordination. EPA should issue guidance on how to
examine the effect of ongoing implementation programs such as NAAQS and PSD
on visibility. In some cases, ongoing strategies may conflict with visibility
goals. For example, the current long-term strategy for ozone control in Los
Angeles encourages diversion of residential and other growth to the desert
(where a number of class I areas.are located) to relieve ozone, concentrations
in the western part of the valley. Interim guidelines and policies should
be examined for possible inclusion in the upcoming proposal dealing with
long-term strategies.
v) Cumulative impacts of new sources. If many new sources-have
/located near a class I area (e.g., development of a oil field), the total
impact of these sources-may have been underestimated if all permits were
reviewed independently. This could lead to pockets of haze in some western
areas, even if overall southwestern visibility is not greatly affected.
Policy and guidelines are needed in this area.
-------
22
I
I
vi) Prescribed burning policy. Some progress has been made to incorporate
visibility impairment into the decision process used to plan prescribed
burning schedules, but continued coordination among EPA, land managers and I
states is needed to develop guidelines for long-term implementation plans.
Currently such policy could address reasonably attributable impairment and _
recognize the unique natural and temporary characteristics of prescribed I
burning. In the long-term, large scale coordination of burning should be B
examined.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23
.REFERENCES
Brookshire, D.S., R.C. D'Arge, W.D. Schulze (1979). Methods Development for
Assessing Air Pollution Control Benefits. Vol. II: Experiments on Valuing
Non-Market Goods: A Case Study of Alternative Benefit Measures of Air-
Pollution Control in the South Coast Air Basin of Southern California.'
EPA-600/5-79-001b, U.S. Environmental Protection Agency, Washington, D.C.
Dzubay, T.G., R.K. Stevens, C.W. Lewis, D.H. Hern, W.J. Courtney, J.W. Tesch,
M.A. Mason, (1982). Visibility and aerosol composition in Houston, Texas.
Environ. Sci. Tecnnol., 16:514.
EPA (1979). Protecting Visibility: An EPA Report to Congress. Office of Air
Quality Planning and Standards, Research Triangle Park, N.C. EPA-4 50/5- 79-008.
EPA (1982). Review of the National Ambient Air Quality Standards for Particulate
Matter: Assessment of Scientific and Technical Information OAQPS Staff
Paper. EPA-450/5-82-001. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, N.C.
EPA Visibility Impairment from Pollution; Public Meetings of Interagency Task
Force on Visibility. 49 FR 44770-44772. November 9, 1984.
FAA (1980a). 14 CFR 91.105.
FAA (1980b). Air Traffic Service Performance Measurement System for Major
Airports. Nov. 1975-1980, Washington, D.C.
Ferman, M.A., G.T. Wolff, and N.A. Kelly (1981). The Nature and Sources of
Haze in the Shenandoh Valley/Blue Ridge Mountains Areas. JAPCA 31: 1074.
Flachsbart, P.G., and S. Phillips (1980). An index and model of human response
to air quality. Air. Poll. Cent. Assoc. 30:759-768.
Groblicki, P.O., G.T. Wolff and R.J. Countess, (1981). "Visibility-reducing
species in the Denver 'brown cloud'--!. Relationships between extinction
and chemical composition," Atmos. Environ., 15(12) :2473.
Hasan, H. and T.G. Dzubay, (1983). Apportioning light extinction coefficients
to chemical species in atmospheric aerosol. Atmos. Environ., 17:1573.
Husar, R.B. (1985). Progress Report on Cooperative Agreement 810351. Prepared
for Atmospheric Sciences Research Laboratory. U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
ICF (1984). Analysis of Alternative Emissions Reduction: 9-8-12 Million Ton
Reductions and 10/12 State Reductions. Prepared for Office of -Policy Planning
and Evaluation, U.S. Environmental Protection, Washington, D.C. September.
Leonard, E.M., M.D. Williams, and J.P. Mutschlecner (1977). The visibility
issue in the Rocky Mountain West. Prepared by Los Alamos Scientific Laboratory
for the Dept. of Energy, preliminary draft report.
-------
24
I
I
Loehman, E., 1984. "Measurement of Air Quality Benefits from Survey Data,"
Staff Paper, Department of Agricultural Economics, Purdue University, January.
Macias, E.S., J.O. Zwicker, J.R. Ouimette, S.V. Hering, S.K. Friedlander,
T.A. Cahill, G.A. Kuhlmey, and L.W. Richards, (1981). "Regional haze case
studies in the Southwestern U.S.--I. Aerosol chemical composition," Atmos. I
Environ., 15(10/11):1971. "
Malm, W., K. Kelley, 0. Molenar and T. Daniel, (1981). "Human perception of
visual air quality (uniform haze)," Atmos. Environ., 15(10/11):1875.
Malm, W. and J.V. Molenar (1984). Visibility Measurements in National Parks
in the Western United States. JAPCA 34: 899. J
Middleton, P., T.R. Stuart and D. E]y (1984). Physical and Chemical Indicators
of urban visual air quality. Atmos. Env. 18:861-870.
Pechan, E.H. (1985). Baseline Emissions Inventory and 1995 Emission Projections
for PSD and Visibility Task Force Analyses, Prepared for Office of Policy Analysis,
U.S. Environmental Protection Agency, Washington, D.C. |
Rae, D., Hausman, J., Stankunas, A., and Vitka, S., (1982). Benefits and _
Costs of Improving Visibility: Case Studies of the Application of the
Contingent Ranking Methodology at Mesa Verde and Great Smoky Mountain National "
Parks, Electric Power Research Institute, Research Project 1742, October.
Rae, D.,/Charles River Associates, (1984). Benefits of Visual Air Quality in |
Cincinnati Results of a Contingent Ranking Survey. Draft Report. Electric
Power Research Institute, Project #1742, August.
Ross, D.M., G.E. Haas, R.J. Loomis, and W.C. Malm (1984). "Visibility
impairment and visitor enjoyment," Paper No. 84-61P.1, Presented at 77th
Annual Meeting and Exhibition of the Air Pollution Control Association, San
Francisco, Ca. June 24-29, 1984.
Rowe, R.D., R.C. d'Arge, and D.S. Brookshire (1980). An experiment on the
economic value of visibility. J. Environ. Econ. Manag. 7:1-19.
Rowe, R.D., and L.G. Chestnut (1981). Visibility Benefits Assessment
Guidebook. EPA-450/5-81-001, U.S. Environmental Protection Agency,
Research Triangle Park, N.C.
Systems Applications, Inc. (1984). Visibility and Other Air Quality Benefits
of'Sulfur Dioxide Emission Controls in the Eastern United States. Draft
contract report prepared for Office of Policy, Planning, and Evaluation, U.S.
Environmental Protection Agency, Washington, D.C. September 21, 1984.
Systems Application, Inc. (1985). Modeling Regional Haze in the Southwest:
A Preliminary Assessment of Source Contributions. Revised draft contractor
report prepared for U.S. Environmental Protection Agency, National Park
Service, U.S. Department of Energy. February 28, 1985.
-------
Thayer, M., and Trijonis, J., (1984). Visibility Benefits Analysis draft
report for California Air Resources Board, Sacramento, Ca., February.
Trijonis, J. K. Yuan and R.B. Husar (1978). Visibility in the Northeast:
Long-Term Visibility Trends and Pollutant Visibility Relationships,
EPA-600-3-78-075, U.S. Environmental Protection Agency, Research Triangle
Park, N.C.
I
I
I
Tolley, G., Randall, A,, Blomquist, G., Fabian, R., Fishelson, G.,
IFrankel, A., Hoehn, J., Krumrn, R., Mensah, E., and Smith, T., (1984).
Establishing and Valuing the Effects of Improved Visibility in Eastern United
States, Draft Final Report, U.S. Environmental Protection Agency, Office of
Research and Development, Washington, D.C., March.
I
I Trijonis, J. (1981). Existing and Natural Background Levels of Visibility
and Fine Particles in The Rural East, EPA-450/4-81-036. Office of Air
Quality Planning and Standards, U.S. Environmental Protection Agency, Research
Triangle Park, N.C.
White, W.H. and P.T. Roberts, (1977). On the nature and origins of visibility-
reducing aerosols in the Los Angeles Air Basin, Atmos. Environ., 11:803.
I
I
I
I
I
I
I
I
I
I
I
Wolff, G.T., N.A. Kelly and M.A. Ferman (1982). Source Regions of Summertime.
Ozone and Haze in the Eastern U.S. Water, Air, Soil Pollution 18:65. -'
Wolff, G.T., M.A. Ferman, N.A. Kelly, D.P. Stroup, M.S. Ruthkosky, (1982).
The relationship between the chemical composition of fine particles and :
visibility in the Detroit metropolitan area. J. Airfoil. Control Assoc.,
32:1216.
Zannetti, P., I. Tombach, and R. Drake (1985). Critique of the Draft Report
"Visibility and Other Air Quality Benefits of Sulfur Dioxide Emission Controls
in the Eastern United States," prepared for Utility Air Regulatory
Group, AeroVironment, Inc. (AV-FR-85/503), February.
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX A. CHARACTERIZING CURRENT REGIONAL VISIBILITY TRENDS,
AND POLLUTANT/VISIBILITY RELATIONSHIPS
by: William Main, NPS
William Wilson, EPA/ORD
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
APPENDIX A
TABLE OF CONTENTS
List of Figures
List of Tables
1.0 Introduction
1.1 Defining Regional Haze
1.2 Monitoring
2.0 Monitoring Results
2.1 Regional Visibility and Trends from Airport
Observations
2.2 Regional Haze in the East
2.3 Regional Haze in the West
2.3.1 Visibility Measurements
2.3.2 Particle Measurements
2.4 Sources of Visibility Impairment
2.5 Modeling
2.5.1 Receptor Models
2.5.2 Predictive Modeling
References
Page
111
iv
1
1
4
5
5
7
7
7
20
23
32
32
35
37
-------
iii
A-l
A-2
A-3
A-4a
A-4b
A-5
A-6
A-7
A-8
A-9
A-10
A-ll
A-12
A-13
A-14
A-15
A-16
A-17
A-13
A-19a
A-195
LIST OF FIGURES
3 Modes of Visual Degradation
Geographical Distributions and Trends in U.S. Haze
Visibility Reductions in Wash., D.C. and St. Louis
Trends of U.S. Coal Consumption
1950's Seasonal Coal Consumption
Trends in Median Extinction Coefficient
Visibility Trends for N.E. States
Three Year Mean Annual Visibilities in California
Summertime Visibilities in Central California
Three-year Means of Adverse Visibilities in California
Summertime Visual Range in Grand Canyon National Park
Five Year Average Visibility at Grand Canyon and Big
Bend National Parks
Isopleths of g Over Western U.S.
Isopleth Map of Visiblities in Calif.
Avg. Composition of Fine Particles in Northern Great Plains
Distribution of Coarse Soil, Fine Mass, Fine Sulfate, and
Other Fines in Western U.S.
Contour Plots of Fine Sulfur for Spring and Summer 1983
Time Plot of Nonsoil Potassium
Mean Chemical Composition of Fine Particulate Mass
of the Denver Aerosol
Eight-Day Average of Daytime Fine- and Coarse-Fraction
Mass Concentrations -
Eight-Day Average of Daylight Light Scattering and
Extinction Coefficients
Page
3
6
8
9
9
10
11
13
14
15
17
18
- 19
21
22
24
25
26
27
27
27
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
IV
LIST OF FIGURES (continued)
A-20 Typical Extinction Budget for Rural SW
A-21 Typical "Theoretical" Extinction Budget for Rural SW
A-22 Isopleth Plot of Source Contribution-Function at
Grand Canyon
Page
31
31
33
LIST OF TABLES
A-l Percent Extinction Budget .for Several Cities
30
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX A. CHARACTERIZING CURRENT REGIONAL VISIBILITY TRENDS, AND
POLLUTANT/VISIBILITY RELATIONSHIPS.
1.0 INTRODUCITON
1.1 Defining Regional Haze
Regional.haze is a complex smorgasbord of man-made and natural gases
and aerosols. Regional haze that can be attributed to man-made pollutants
has been identified not only in urban environments but also in the remote
areas of the continental United States.(1,2) Research on regional haze
has emphasized the technical, economic and health aspects of the problem.
However, haze also affects human behavior.(3) It affects the amount of
time we spend outdoors, and the extent of our physical activity. It affects
recreation patterns, task performance and interpersonal relationships.(3)
The natural atmosphere is composed primarily of nitrogen and oxygen
along with some trace gases such as argon and hydrogen. Natural and
man-made aerosol that are perceived as haze are condensed water vapor
(water droplets), wind-blown dust, primary and secondary aerosols.
Primary aerosols are those emitted from a source as particulate matter,
while secondary aerosols are airborne dispersions of particles formed by
atmospheric reaction of gaseous "precursor" emissions. Gases are converted
into aerosols through very complex reactions that have only recently been
understood. The gas to aerosol conversion process takes place by essen-
tially two processes; condensation and nucleation.(4) Condensation involves
gaseous vapors condensing or combining with existing small nuclei, usually
referred to as condensation nuclei. The small condensation nuclei may have
their origin in sea salts or from combustion processes. Gases may also
interact and combine with droplets of their own kind and form larger aerosols.
This process is referred to as homogenous nucleation. Once aerosols are
formed they can grow in size by a process called coagulation. In coagulation,
particles essentially bump into each other and "stick" together. Aerosols
also may result from reactions in cloud droplets and subsequent evaporation
of droplet water.
Man-made-fine particles and gases such as sulfur oxides, nitrogen
oxides, and hydrocarbons are emitted from urban and industrial sources. Im-
mediately upon emission many of the gases begin to convert to aerosols of
various composition and sizes. Sulfur oxides convert into sulfuric acid and
ammonium sulfate, nitrogen oxides convert to nitric acid and ammonium nitrate
and hydrocarbon's became organic aerosols.(4) Near (within 0-100 km) a source
region such as an urban center, haze is a mixture of gases, primary and
secondary aerosol.(5) After the pollutants are transported hundreds of
kilometers, regional haze is primarily made up of fine primary and secondary
aerosols such as ammonium sulfate and organic aerosols.(6) t -
Perception of regional haze involves air quality effects on light, the
psychophysics of the eye-brain system, and ultimately psychological reactions.
It is therefore important to investigate the meaning and metrics for
visibility. Should visibility be defined strictly in terms which concern
-------
themselves with exact measurements of illumination, threshold contrast and
precisely measured distances? Or is visibility more closely allied with
value judgments of an observer viewing a scenic vista?
Historically, "visibility" has been defined as "the greatest distance at
which an observer can just see a black object viewed against the horizon
sky."(7) An object is usually referred to as at threshold contrast when the
difference between the brightness of the sky and the brightness of the object
is reduced to such a degree that an 'observer can just see the object.
However, visibility is really more than being able to see a black object
at a distance for which the contrast reaches a threshold value. Coming upon a
scenic vista an observer does not ask, "How far do I have to back away before
the vista disappears?" Rather, the observer may comment on the color in the
vista, on the amount of snow cover resulting from a recent storm system, on
the contrast detail of nearby geological structures or on shadows cast by
overhead clouds.
Visibility is more closely associated with the atmospheric conditions
which allow appreciation of the inherent beauty of landscape features.
Visitors to class I areas are interested in being able to see and appreciate
the form, contrast detail, and color of near and distant features. It is the
ease with which an observer can see and appreciate scenic vistas under various
atmospheric and illumination conditions.(11,12,13,14)
Generally, regional air pollution manifests itself as a uniform haze
that masks both natural and urban landscapes. The effect that uniform
haze has on a vista is to reduce the contrast of form, color, and texture.
Distant mountains and structures can no longer be seen, while detail and
color in nearby vistas become washed out. Even cumulus clouds, which
have sharp edges on clear days seem to fade into a continuum on hazy
days.
Regional haze can also appear as a layered haze under certain atmospheric
conditions. A layered haze is a distribution of pollutants which results in a
spectral (color) discontinuity between that haze and either a sky or terrain
background. However, pollutants which are emitted from nearby sources and may
be reasonable attributable to that source will also manifest themselves as
layered haze while regional hazes, which are usually well dispersed, have a
uniform appearance.(12) Schematic diagrams of regional and layered haze are
shown in Figure A-l.
Visibility monitoring programs have shown that in excess of 90% of
1 the time the ease with which scenic vistas can be seen and appreciated is
\ impaired at all monitoring locations within the continental United States.
Even in remote areas such as Grand Canyon National Park, visitors are
sometimes denied the pleasure of seeing the opposite canyon rim or the
great canyon depths because of poor visibility.-(13) At Yoseroite National
Park, smoke from prescribed fires sometimes obscures the view of the massive
cliffs and domes that have made the park internationally famous. In the
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
o
** i
C 4-)
01 3
*|«« fm**
o f
O) 4->
,
> Q.
O gn «/i
O -M O «B
E -f- 4J t.
01 (O Q
O -0
en 0) OJ-""-
T3 OJ in £
<0 -C 3
I- 01 -U .
0>+J C O.
4) C O
>> 3 S- r
> -M ^ *
C 0) O trt
10 i O ifl
£ * C
O 3 -r
3 8J Q 01
r*« £
o o QJ
+J dl 2
M .c E
Q. i. 01
v> o ~a
>i O <*- O
(O E -^ -^
S 4J C i-
IQ 3
-c 3 m
I
3 »- (U *J
C5 C X »-
^- 3 i- S
-------
enacting legislation for establishing Shenandoah National Park, congress spoke
of being able to see the National Capital on a clear day.(14) Yet today the
National Capital is never visible from Shenandoah.
1.2 MonitojMng
An understanding of the relationship between visibility and visibility
reducing aerosols requires a monitoring program with the capability of
measuring both particulate concentrations and visibility related indices such
as change In vista contrast, color, texture or atmospheric extinction.
Historically, atmospheric particulate concentrations and composition have
been routinely measured with particle samplers that do not selectively measure
particles of a known size range. However, recent developments in particulate
monitoring allow for a more accurate determination of mass concentration and
composition as a function of particle size. Most importantly, current
particle monitoring systems allow for the selection and chemical analysis of
particles (<2.5 urn) that are primarily responsible for visibility reduction (15),
Visibility may be mesaured by a variety of techniques. These include
1) human observer, 2) transmissometer, 3) telephotometer or teleradiometer,
4) photography, 5) and scattering by integrating nephelometer plus absorption
by particles collected on a filter. Teleradiometry is most useful under
ideal viewing conditions, i.e. no clouds on the horizon and black targets (16).
Unfortunately, many applications of interest do not involve ideal conditions.
However, for relatively low visibility conditions, artificial targets may
be used with teleradiometers to yield precise measurements (17).
The first four techniques integrate visibility over the viewing path.
The fifth technique is a point mesurement but is especially useful since
it can be related to point measurements of aerosol properties. Human
observer and instrumental techniques agree reasonably well when a contrast
ratio of 5% is used (rather than the theoretical 2% for ideal conditions) (18).
Visibility monitoring has evolved from human observer-based measurements
to the use of complex automated electro-optical instruments. Until recently,
monitoring of visibility has been associated with establishing the furthest
distance an observer could see a dark target (usually a tree-covered knoll
or mountain) or appropriate size on the horizon. The National Weather Service
has sponsored a nationwide program located primarily at airports to record
"observer determined visual range" (7). Data from this network has been
used to investigate visibility trends as far back as 1948.
More recently, the National Park Service has established 35 long-term
(greater than five year) visibility monitoring sites at various remote loca-
tions throughout the continental United States (19). Other Federal and state
agencies have operated shorter term programs, as have a number of indus-
tries.(18,20,21) Most monitoring programs operate* at a.minimum, a teleradio-
meter measuring sky-target contrast at 550nm (green wavelength), a 35mm camera
system and a .size selective fine particulate monitor. In some cases where
commercial power and environmental enclosures are available, integrating
nephelometers have also been used.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In the last few years, some progress has been made toward automating
measurements and toward instrument configurations that can operate on solar
power for extended periods of time in remote, harsh environments. Also, new
types of "scanning" radiometers have been developed to measure radiance at
various wavelengths and at many points within a vista.(12,22) Although these
types of radiometers are capable of collecting information on texture and
color reductions and on haze layers, they have not been used in a routine
monitoring program because theoretical formalisms have not been adequately
developed to handle the millions of pieces of data produced by these instruments,
All instruments currently used in visibility monitoring have limitations.
Although teleradiometers measure characteristics of the scene in a way that
relates directly to the human eye-brain system, if it is desired to interpret
contrast data in terms of extinction or visual range a whole series of
assumptions must be made about inherent contrast, illumination and uniformity
of aerosol distributions.{16,23) An integrating nephelometer makes a point
measurement, does not measure absorption, underestimates scattering of large
particles and, quite possibly, modifies the aerosol as it is drawn through the
sampling chamber.(18)
Furthermore, particle samplers currently in use may underestimate nitrate
concentrations and other volatile aerosol components. For instance, Appel et
al. estimate that as much as 40% of the organics are lost during the sampling
process.(26) During summer months when Southwest desert environments experience
temperatures in excess of 100°F, Appel et al. suggest that current sampling
techniques allow for as much as 90% of the nitrates to be reevaporated back"
into the atmosphere.(25)
There is a pressing need to develop instruments that measure the
atmospheric extinction and scattering characteristics as they exist in the
atmosphere, independent of meteorological and vista conditions. An "open air"
integrating nephelometer measuring atmospheric scattering in conjunction with
a long path transmissometer measuring atmospheric extinction would meet these
needs. .
There is also a need to assess the role'of volatile aerosols on visibi-
lity reduction. If volatile aerosol scattering is significant, there will be
a need to design new particle monitoring devices that better characterize
those aerosols with-low vapor pressures.
2.0 MONITORING RESULTS
2.1 Regional Visibility and Trends from Airport Observations
The most extensive monitoring data base for any air pollution-related
parameter is the human observer visual range measurements of the National
Weather Service (NWS). Several hundred weather stations in the US and
Canada record visual range every hour. A computerized data base exists-
from 1948 to the present.'- Daily noon visibility isopleths for the US,
covering the period from 1948 to 1978, have been calculated (26). These
isopleths have been used to study regional trends in visibility from 1948 to
1978 (27,28,29). More recently, these analyses have been extended through
1983 (Figure A-2).
-------
o
's
' M t
" 'f» A
O E
01 O)
oi *« in
» = >
« » ~ si 01
O> = E fc.
*j ^ B
> £. *n - o
* ** +* .
x e> E o>
O< «n x «j<
* 0> i. Oi i_
W W .fi *rf «B
* o a> w E
U U U Vi OJ
e> e c ~ «
«"*-*> 2
o t. e »
^ Oi **
» Ol ft) »- VI
t- = «: TT X e -S
a.- » u « e s
> -r -e w
« « t At tfl «< k.
.« m « ^
*> ^ b. - Ol HI
= = ic S t.
6> ic « £ O
Ol ) < ~ wi in i
I
I
I
I
I
I
I
IT! =. f * S
C = = 01 . O-
« u « «- * «« e
""^S^-ssut!
& > «« a s^se!soie
«« ^> ^. m ^
«« u E = w «
e s £ s x 0
-es.«e-s«.= =
S ««i-h»vifa4 3
« - P» S M
W "8 01 C
e _ «B « ^ > ^s
C £ >> s
S M «i -B <«
** w & i_ t; =
- ». » ai «* u e ww M VI A! e
= w> c 4. w
*- e E >
0t . w 5 'e
S«' +t * tn ^
^ » v> ** « x
I
I
I
I
I
I
_
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2.2 Regional Haze in the East
Eastern regional haze can be observed in satellite .photographs. In
the summer haze blobs can be observed that cover several states. These .
areas of low visibility maintain some cohesiveness for several days and can
be tracked as they undergo transport for 1000's of kilometers. Patterns of
haze observed from satellites agree fairly well with isopleths of NWS
-visual range less than 6 miles (28,30).
Between 1948 and 1972 visibility reduction changed from a local
to a regional phenomena (Fig. A-3)(28). At an urban airport, Washington
National, low visibility has changed from being associated with low wind
speeds to no relationship with wind speed indicating that low visibility no
longer results from a build-up of urban pollution during stagnations. At 'a
rural airport, Lambert Field, St. Louis, low visibility was initially
associated with wind flow from the city but now shows no relationship with
wind direction. These observations indicate that reduction in visibility
is not associated with local urban sources but with a regional distribution ,
of pollution.
The NWS visual range data from 1948 to 1978 has been used to analyze .
visibility trends on a.regional and seasonal basis (22,28,29) to compare
regional visibility trends with regional coal usage (Figure A-4a,b).
Visibility isopleths for 1979-1982 have been prepared recently and are .1
currently being analyzed with earlier data for trends and patterns from 1948 ;".,
to 1982. Some observations are fairly obvious and can be made by observation .
of the isopleths shown in Figure A-5. In the winter season there has been "|
an improvement in visibility in New England and the north central U.S.
However, there has been degradation of visibility along the south east and
gulf coasts and in south central U.S. In the spring season after 1970
there has been a degradation of visibility in the entire eastern U.S.,
especially along the gulf coast and the south and central east coast. The
most dramatic changes, however, are evident during the summer season. A
region of modest visibility in the northeastern U.S. (bscat « 0.24-0.30)
during 1948-1952 steadily expanded and became worse until the entire eastern
U.S. and southeastern Canada were affected. The fall season shows significant
improvement in the north central industrial areas but degradation along the
gulf coast.
Improvements in the mid- and northeastern U.S. during the 72-82 period
may be associated with reduced industrial activity during^the 80-82 recession
(Figure A-6). Efforts are currently underway for both modelling and statistical
analyses of these relationships..
2.3 REGIONAL HAZE IN THE WEST
2.3.1 Visibility Measurements
In spite of inherent limitations in instrumentation currently used to
measure visibility, monitoring programs are yielding a wealth of information.
Airport observations of visibility are useful sources of trend information in
areas where visibilities are comparable to the distances to available targets.
-------
8
HRSMINCTQN O.C.
WINTER
EXT. ROSE LaitBEflT
ST. LOUIS. HO.
b.
» 10 « 20 28 30
HINO l^ltO. KNOTS
sorr )Ef I II - "75 was due to remote' rather tha
sources, (b) A haziness rose for Lambert Airport in St
sss-ssi sr
local,
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
FIGURE A-4a. The trends of U.S. coal consumption in
contiguous7areas'show strong regional dependence. Electric
utility/coal consumption is represented by the dark shading (16)
M A M J 4 ABONO
FIGURE A-4b. In the 1950's, the seasonal coal consumption
peaked in the winter, primarily due to increased residential
and railroad use. In 1974, the seasonal pattern of coal use
was determined by the winter and summer peak of utility coal
use (16).
-------
e
I
CM
MEDIAN EXTINCTION COEFFICIENT
1948-52
60-64
70-74
Do. 17-0.
km1
o.Z7-0.32
»0.32
78-82
|
I
FIGURE A-5. Trends In median extinction coefficient (3.9/visual range}
from airports in the Eastern U.S. Data are normalized to a standard
relative humidity using an empirically derived correction (4).
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
j
n
I-M *-OI MJCM
«
* I
338
I-B »-oi
-------
I
12
I
Trijonis et al.(28) examined airport visibility data for trends in visual air
quality in the Rocky Mountain Southwest over the time period from 1948 to
1976. They found that in the late 1940's to the early/mid 1950's, visibility
trends were mixed with some geographic areas showing a slight improvement and
a lesser number of areas showing a slight deterioration. From the early/mid
1950's (1953 to 1955) to the early 1970's (1970-1972), most areas indicated a |
drop in visibility of approximately 1U to 30 percent. From the early 197U's
(1970-1972) to the middle 197U's (1974-1976), visibility generally tended to _
increase by about 5-10 percent, especially at those sites in or near Arizona. I
Trijonis also examined airport data at 67 sites in California.(32) Plots
of long-term trends in median visibility (for all data with no sorting for
meteorology) at the 19 study sites reveal that visibility trends in California |
tend to split into two general sub-periods, divided at approximately 1966.
Before 1966, nearly all locations exhibit deteriorating visibility, with _
especially large visibility decreases occurring in and near the Central
Valley. After 1966, nearly all locations have displayed improving visibility. B
Over the entire two and'one-half decades from 1949-1951 to 1974-1976, the
major areas experiencing a net improvement in visibility were the
central/coastal parts of both the South Coast Air Basin and the San Francisco
Bay Area Air Basin. The densely populated, central/coastal portions of these
metropolitan regions underwent net improvements in visibility on the order of
10 to 40%. Slight improvements in visibility also seem evident in north-
eastern California and southern Oregon from 1949-1951 to 1974-1976. The major
areas that experienced net deterioration in visibility on the order of 10
to 30* are the San Joaquin and southern Sacramento Valleys, the South
Central Coast Air Basin, the inland part of the South Coast Air Basin, and the
Southeast Desert Air Basin.
Other California visibility trends are reported by California Air
Resources Board (CARB).(33,34,35) Their discussion centers around "adverse"
and "superior" visibility. They define "adverse" visibility as having
concentrations of visibility-reducing particles sufficient to reduce prevailiny
visibility to less than 10 miles when the relative humidity is less than
70%. Visibilities that are 30 miles or more are termed "superior".
The trend of the composite superior visibility for two pristine stations,
Mt. Shasta and Bishop, is shown in Figure A-7 . It shows a gradual (about
1% per decade) decrease in average superior visibility occurrences in these
pristine areas of California. The 20-year (1958-77) overall summertime
trend of adverse visibilities in the great Central Valley (Red Bluff,
Sacramento, Stockton, Fresno, Sakersfield) shows little change during the
period 195d-1967 and marked improvement "si nee then. The improving trend is
statistically significant at the 1% level. These data are presented in
Figure A-8. As shown by the trend line, the occurrence of adverse summertime
visibilities in the Valley decreased from about 13% to less than 4% from
1967 to 1977. -
The overall summertime trend of adverse midday visibilities in the
coastal plains of California is presented in Figure A-9. As shown, the
occurrence of adverse visibilities during this period decreased in the larger
urban areas throughout this 20-year period. A similar decrease is also
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13
too-I
C.
3
(A
96-
99 -
I I i
1999 SO 61
FIGURE-A-7. Three year mean annual occurrence of superior midday
visibilities in the pristine areas of California (1958-1977).
-------
14
to--
I .9--
10-
LEAST 3OUAPKS
'<*
MOT
i t t
70 Tl Tl 73 T4 73 71 77
FIGURE A-8. Summertime occurrence of adverse midday visibilities in
the Great Central Valley of California.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
r s -0.55
NOT SIGNIFICANT
55
MID-YEAR
FIGURE A-9. Three-year means of percentage occurrence of adverse
visibilities based on June-August (summer) data from (») San Francisco,
Oakland, Los Angeles, and Long Beach and from (o) Salinas, Santa Maria,
Oxnard, Riverside, and San Diego. In the above figure,' r is the
correlation coefficient and S.L. is significance level.
-------
ib
I
I
evident in the smaller urban areas since 1969, while an increase is evident
prior to that time. The all-season data show similar trends. The increase in
adverse visibility in the smaller urban areas during the 1960's appears to be
associated with the high rate of population growth during that period. The
decrease since 1969 in all areas appears to be associated with a slowing in
the rate of population growth and the effect of the various air pollution
controls applied to mobile and stationary pollution emission sources.
The decade-to-decade change of adverse visibilities for all-season data
is similar to the trend shown by the summertime data. The larger urban areas
show a decade-to-decade decrease in adverse visibility ranging from 7 to 12%
(San Francisco, Oakland, Los Angeles, and Long Beach). This change is _
consistent with the steadily decreasing trend shown by the summertime data I
(Figure A-9). The decade-to-decade change for the smaller urban areas, however,
shows a net increase in adverse visibility ranging from 2 to 14% (Salinas,
Santa Maria, Oxnard, Riverside, and San Diego). This change is consistent
with the trend shown by the summertime data (Figure A-9).
Data from the recently i'mplemented (fall 1979) National Park Service
visibility monitoring network shows summertime visibility in the Colorado I
Plateau area, an area that encompasses Grand Canyon, Zion, Bryce Canyon,
Canyonlands, and Mesa Verde National Parks to be decreasing from 1978 to 1981,
and then increasing in 1982 and again in 1983.(19) Grand Canyon summertime
visibility presented in Figure A-10 is representative of this short-term trend. |
The National Park Service monitoring network has also yielded interesting «
seasonal trends. Figure A-ll, a plot of seasonal visibility averaged over five
years, shows that at Grand Canyon and Big Bend National Parks visibility is
best during winter months, worst during summer season and intermediate during
spring and fall. This trend is representative of seasonal variation at most
monitoring sites in the western United States. I
Seasonal patterns in median 1:UO p.m.visibility was also investigated in
the Trijonis California visibility study.(32) It is found that the seasonal
pattern in visibility is not uniform throughout California. The seasonal
patterns are usually consistent, however, within individual air basins and
major geographical sections of California. I
Nearly all locations in southern California and along the central coast
-- the South Coast, San Diego, Southeast Desert, South Central Coast, and
North Central Coast Air Basins -- exhibit minimum visibility during the spring
or summer (especially the summer), and maximum visibility during the fall and
winter. Nearly all locations in the San Joaquin Valley, Sacramento Valley,
and San Francisco Bay Area Air Basins display minimum visibility during the I
fall and winter and a distinct maximum during the spring; a similar pattern
exists in the Northeast Plateau Air Basin (as well as some locations in the
North Coast and Lake Tahoe Air Basin), except that maximum visibility is
usually displaced from the spring to the summer. -
The NPS monitoring program also shows a definite spatial distribution _
of visibility.(19) Figure A-12 shows isopleths of SVRg over the western I
United States for summer of 1982. The highest visual range, SVRg = 180 km,
occurs in the northern parts of Nevada and Utah and the southern portion
I
I
-------
1
1 - »
1
1
1
200
1
Ul
o
1 5
x
J
1 I
>
Io
CE
S
c
2
1 £
1
100
:5:$:$S5$">:":':-;
&:$:»*:$::-:$!*:$ :-:-S-*S:":->S:-:":-!
Iflliill Ililixt llllllll
"X"X*t%*X*X"-*X*I 'XvX*X*X*!*X*X* C*X*Xv***v**X*X*
^x^ilSS IJiSx^^S: Ix^Hixlll
$:xj;-§:?§x*i"x" yx-Xs-x-x-x-x- S-*.§::::i-x:x-x::; w'xvS'x-i'x'x
jx-i-i-Sx:;:-:-:-:-:-:: $:§:$:₯?$§:$§: ;:$:::::;::x-::::::xv: §iS:$:S"-'-x'?:S
Sp^x&x?! Jx-x^lS^xl H^^|:i>l x^S^x>^::
j-X'X'x-xXv'x"; x-x*x"XvX:Xx: Sfx'i'iWft'StXx xXxi-x-x-xtX-
&mm%& KW*&&& jMm&m x":S:xxSx§i::
m&&y*m imm&y& pxx$ssw %$&$&&? mzmzm
fftKx-xVx-x-xJ S-xyrvXwX-x- fe::XvX::xK:%::::: fxwxWxtttf «::%%-x-XvX-x
x^x^xwb iHv::::::::x%:::::::: K:::X:?X$X:X:X: ft:::::::::::::::::::::::x:: S:>:Svi:::i:::S:::::::
I'X*XvX*XvXvH l"X'X*X*X'X"X'X tvy'X*'*'*XvX'X" 't'xXxv-XvX'x*; J*X'XvX%vX*X*-
>X*X"X*X"X*X%"»'' 'X'X'X'X'X'X'X* "X*X"X*I'."«"»fI">*»"»* ."*»**" "*'%".*»""*«". %"X*X*I"I*X*X%"X'
§:J:§:::-!::::::$5: Si-SI'Svi-JxJ:: f§x£$:-:-:§?$::£ x$?x$:-:-!xSx; ?::$:§:??:%?:?
I'X-X'X'X-X'X'X' I'XvX-X-X'XvX K-X.X-X-X'X'X'X x*X-TXxX;X;Xy X-X'X'X-X%*X4X
1979 1980 1981 1982 1983
YEAR
1
1
FIGURE A-10. Summertime standard visual range for Grand Canyon
National Park.
1
1
1
-------
18
300
c
_J
<
5 200
Q
cr
tO
CO
Grand Canyon National Park
Big Bend National Park
Summer
Fa i
Winter Spring
FIGURE A-ll. Five year averaged seasonal visibility at
Grand Canyon and Big Bend National Parks.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
19
°f SVfV°Ver the western United States for
-------
I
20
of Idaho. The region with the next highest visual range, SVR = 165 km,
corresponds to that geographic area that contains Grand Canyon, Bryce
Canyon and Canyonlands National Parks, commonly known as the Colorado
Plateau. Southern Arizona, New Mexico and the "Front Range" area of the
Rocky Mountains have SVRg = 140 km. The lowest visual range is found in
California near the populated urban areas and in areas where there is
significant industrial and agricultural activity.
In addition, the NPS monitoring network has documented layered haze
trapped by stable air masses at Bryce Canyon and Mesa Verde National Parks.
On winter mornings, portions of Navajo Mountain as seen from Bryce Canyon are
completely or partially obscured as much as 80% of the time that data was
collected. Recent analysis suggests that this haze layer tends to be a result
of local emissions into the Glen Canyon area.
Tn'jonis has investigated spatial variations in visibility throughout
California in more detail.(32) Using 1:00 p.m. data for 1974-1976 at 67 loca-
tions, a detailed isopleth ma,p, shown in Figure A-13, illustrates the geographi-
cal patterns of visibility in California. The map reveals that the spatial
gradients of visibility in California are far more severe and complex than
those observed anywhere else in the United States. Some parts of California
exhibit among the best visibilities in the nation, while other parts
experience among the worst visibilities in the nation.
The clearest air in California occurs along the Nevada border. One area
along the border, Death Valley National Monument and the mountainous areas
immediately northwest, experience median visibility exceeding 70 miles. This
area is on the fringe of a large region in the desert/mountain southwest
United States which exhibits the highest visibilities in the nation.
Median .visibility is also quite good, 45 to 7U miles, in the plateaus and
mountains of northern California, the mountains of central-eastern California,
the desert near the Arizona border, and the Vallecitos Mountains east of San
Diego. To the west of all these areas, very sharp gradients occur, with
visibility falling to less than 15 miles along the entire coastline except the
far northern coast near Oregon, where median visibility falls to less than 25
miles.
Two significant pockets of poor visibility occur between the coast and
eastern California. Median visibility is less than 15 miles in the large area
consisting of the central/southern San Joaquin Valley. Visibility is less
than 10 miles in the center of the Los Angeles basin.
2.3.2 Particle Measurements
Fine particles (smaller than 2.5 urn) are generally responsible for a major
share of visibility impairment. The pie diagrams in Figure A-14 show the
average composition of fine particles at sites 'in the northern Great Plains,
Arizona, Texas and VIrginia.(36) The largest contribution is from sulfates,
especially at Shenandoah. Soils contribute around 10% to 20% of the fine mass
in the west but much less at Shenandoah.(36) More detailed studies have shown
that these fine soil particles are larger than 1 um. Whereas smoke tends to
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
21
V - 10 miles
10 miles < V * .15 miles
15 niles < V - 25 miles
25 miles < V - 45 miles
45 miles < V - 70 miles
70 miles < V
FIGURE A-13. Shaded isopleth map for median 1 p.m. visibilities
in uanTorni
-------
22
al
Theodore Roosevelt
etal
Big Bend
FIGURE A-14. Average composition of fine particles (smaller than
2.5 um) at sites in the northern Great Plains, Arizona, Texas and
Virginia.
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23
contribute less than 10% of fine mass in the Southwest, it can contribute in
excess of 60% of the mass in the Northwest.(37) The component labeled "other"
includes organic material, nitrates and water incorporated in the particles.
Figure A-15, "bubble" maps for summer 1983, shows the geographic distribution
of coarse soil, fine mass, fine sulfate, and fine nonsulfate. Shenandoah's
sulfate value is shown for comparison purposes. The component labeled "fine
other" shows the nonsulfate portion of the fine mass, including smoke.
Glacier and Grand Teton had consistently high nonsulfate in the fall and
winter, while Yosemite had high values in summer and fall.(36)
The contour plots of fine sulfur, shown in Figure A-16, indicate that over
a season there are well defined regions of relatively constant sulfate concen-
tration. The values in Oregon and northern California are comparable to those
measured at the cleanest sites in the world, while southern California,
Arizona, New Mexico, western Texas and the western Great Plains have the
highest sulfur concentrations in the western United States. It should also be
noted that sulfur concentrations are highest during summer and lowest during
winter months.
The time plot of fine nonsoil potassium (K) shown in Figure A-17, for sites
in Arizona, New Mexico and Utah shows a remarkably regular pattern of impact
every June. These particles are associated with smoke, and while the sources
are still uncertain, they may represent transport from Mexico.(36)
Composition of aerosols in western urban centers differ considerably from
what is found in both eastern (urban and rural) and remote western national
parks and wilderness areas. Figure A-18, the composition of Denver fine
aerosol that is _< 2.5 ym, is typical of most western cities(6). C0 and
Ce refer to organic and elemental carbon respectively. In urban settings,
sulfates are a smaller fraction of total aerosol mass, while organic carbon
and nitrates make up a significantly greater fraction of the mass than they
do in remote areas. It should be pointed out that the total mass concentration
in urban settings is considerably higher than in class I areas. Mean
concentrations of fine particulate mass associated with the pie diagram
shown in Figure A-7 was 39.5 ug/m^, whereas mean fine mass concentrations
in the rural west are 5-10 ug/m^. However, sulfate concentrations are similar
in western urban and rural areas.
2.4 SOURCES OF VISIBILITY IMPAIRMENT
Visibility data, when combined with particle composition and concentration,
allow for developing an understanding of which of the many atmospheric
constituents are responsible for visibility reduction (lignt extinction
budgets). Because different size particles reduce-visibility with varying
degrees of efficiency, it does not automatically follow that an aerosol
species making up a certain fraction of total mass will be responsible for
that same fraction of visibility reduction. For instance, cearse windblown
dust may constitute 60% of total mass and yet be responsible for only 20% of
the visibility reduction.(36,15)
Two techniques have been used to date to develop light extinction
budgets"theoretical" (or "first principle") and "statistical". Each has
its own advantages and disadvantages.(38)
-------
24
COARSE SOIL
FINE SULFATE
FINE MASS
FINE OTHER
AVERAGE SEASONAL CONCENTRATION IN MICRCCRAMS/M--3
FOR SITE IS PROPORTIONAL TO DIAMETER Oc CIRCLE
FINE OTHER - FINE MASS - FINE SULFATE
SHCNANDQAH SULFATE
o
FIGURE A-15. Geographic distribution of coarse soil, fine mass,
fine sulfate, and fine nonsulfate for summer 1983. Shenandoah's
sulfate value is shown "for comparison purposes.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
cr
o
3s
r-
LJ r^
IB fe
z co o
yj ** O.
f L3 ^
tr
o
u ^ 3
CJ « 3
2 cr w
i w. y
UJ
>- <
25
en
CL
CO
§?
tr
tf! 3
fc1 S
-i
c 2
T3
fO
en
a.
i/i
i.
3
5? i?
ec
^ a
z * a
Ut (R "^
O
*Q.
o
+j
c
o
u. a
o w
O <
ft u
O *- W
r < -I
U9 U Ul
">
-------
26
FINE NON-SOIL K
3 DAY MEAN
n
137911980 ISQOllSai ISaiilSez' 1592il983
AUC 1373 . TQ NCV 1383
FIGURE A-17. Time plot of fine nonsoil potassium (K) for sites in
Arizona, New Mexico and Utah.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
27
Ry Ash & Soil
.2-
Ncn-SO, Sulfur-2.5%
FIGURE A-18. Mean chemical composition of fine particulate mass
of the Denver aerosol.
TUMI
1.T*
eauta riucneM. IT j *
FIGURE A-19a. Eight^day average of daytime
fine- and coarse-fraction mass concentra-.
tions apportioned by chemical species in
Houston from September 11 to 19, 1980.
FIGURE.A-19b. -Eight-day average of
daytime light scattering and extinction
coefficients apportioned by chemical
species in Houston from September 11
to 19, 1980.
-------
I
28
The theoretical technique requires the size distribution and chemical
composition of each aerosol species to calculate the aerosol extinction effi- B
ciency. This technique assumes that the size distributions of all species are I
properly measured and the properties of the mixture are known. Aerosol conpo-
sition data are generally incomplete: specifically, current monitoring tech-
niques do not properly account for water and other volatile aerosol compo-
nents, nor do they allow for complete specialization of the various aerosol |
components.(24,25)
The second approach used for developing extinction budgets requires a I
longer-term data base. It is listed here as the statistical technique, in
that it seeks to fit measured light extinction coefficient and species mass
concentration data to a linear relationship (linear regression analysis). As
with the theoretical approach, assumptions about aerosol optical properties I
are required. If the assumptions are satisfied, multiple regression analysis
applied to measurements of the extinction coefficient and the masses of
individual chemical species may be expected to yield reasonable estimates of I
the average species extinction efficiency.
A number of studies have employed multivariate statistical techniques to I
examine relationships between atmospheric extinction and various atmospheric "
pollutants.(5,15,39-44) Fewer studies have been designed to gather data
necessary to theoretically calculate extinction budgets and still fewer
experiments have been designed to gather the detailed aerosol data required |
to make theoretical extinction to mass calculations and at the same time
conducted over a long enough time period to allow for a statistical _
treatment.(38)
Whether the analysis has employed a theoretical or statistical approach
it is clear that man-made aerosols are responsible for a significant portion
of visibility reduction. Many studies have demonstrated a high correlation |
between sulfate, fine particle mass and light scattering (15,45,46,47,48). This
is not unexpected. Fine particles are much more effective than coarse particles m
in scattering light (46). Sulfate is more effective than other aerosol
components in scattering light because of the large amount of water associated
with it. As relative humidity increases sulfate particles grow into the size
range which most effectively scatters visable light (46). Carbon aerosol
or soot is also important because it both scatters and absorbs light (13). 8
Nitrate may be important in areas where there is sufficient ammonia to
neutralize all.the sulfuric acid aerosol (45,49). Organic aerosol may also
be important in some situations.
An example of the kind of study that has attempted to identify all
components of the aerosol and calculate the light extinction associated
-with each component (41) depicted in Figure A-19. Sulfate was found to be
\ the most important visibility reducing component. Sulfate becomes even
\ more important when the ammonium ion and water associated with it are
included. This study may have underestimated the. volatile components-water,
^organic, and nitrate-because of difficulties in measuring them. Carbon is
very important in urban areas in the winter because of increased use of _
wood and coal in home heating. The.importance of sulfate will increase in
rural locations and durina low visibility eoisodes (501. In aeneral. the
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
29
Importance of primary aerosols such as carbon soot, lead particles, road
dust, etc. will decrease at increasing time or distance from the point of
emission while the importance of secondary aerosol such as sulfate, nitrate
and organic will increase. This is because although both types of aerosols
are removed by deposition processes, secondary aerosols continue to be formed.
Table A-l is a summary of extinction budgets from several studies. In spite
of city variations, sulfates and carbonaceous material seem to consistently
be major contributors to urban extinction. For instance, it was found that
in Denver 20% and 12% of the extinction is associated with sulfate and
organic carbon while 38% is associated with elemental carbon.
Recent studies in western rural areas including Grand Canyon, Bryce
Canyon, Canyonlands and Theodore Roosevelt National Parks, and China Lake in
the Mohave Desert also yielded data that, when analyzed using regressional
techniques, indicated sulfur aerosol to be the dominant light scattering
component. These studies show sulfate aerosol was a major contributor to
visibility reduction, responsible for 30% to 80% of the atmospheric scattering
associated with particulates with scattering-or extinction-to-mass ratios
between 0.004 and O.OlOkm'Vug/m3. A typical extinction budget for these
areas is shown in Figure A-20. Twenty percent of the extinction is due to
coarse mass and only 17% to other fine mass. Other fine mass is made up of
organic carbon, elemental carbon and nitrates.
The major difference between urban and rural extinction budgets is the j
importance of nitrates. Ammonium nitrate typically is responsible for more :
extinction in the urban setting. The extinction budgets obtained through
regression analysis should be compared to those obtained theoretically. :
Figure A-21 shows a typical "theoretical" extinction budget for rural Southwest.
Thirty and 33% of the extinction is attributed to organic carbon and ammonium
sulfate respectively. Coarse material, primarily dust, contributes 15%, while
12% of the extinction is attributed to soot (forest fires).
Another study designed to allow for theoretical as well as statistical
determination of the extinction budget was carried out in the Mohave Desert
(China Lake).(38) Results of the theoretical calculations associated with
this experiment showed that 32% of the extinction was associated with
sulfates, 19% with organics, 11% with crustal material and 39% unaccounted
for, while regression analysis suggested that 50%,,12%, 18% and 16% of the
extinction budget was associated with sulfates, organics, crustal, and
unaccounted for, respectively.
Notice'that regression analysis typically suggests sulfates are more
efficient scatterers and are responsible for more of the extinction than does
the theoretical approach. Regression analysis also suggests that-organics are
a small part of total extinction while the theoretical approach indicates
organic carbon to be responsible for as much as 30% of total extinction.
A very plausible explanation for inflated extinction to sulfur scattering
efficiencies is that current particle monitoring techniques do not adequately
capture volatile aerosols. If volatile aerosols such as water, ammonium
-------
00
w«
a
V*
X.
hi
v
ai
2
0)
*
14
u
o e
"- JS *
*j w
£ 3
§
x
LU
01
ai
a.
«
UJ
ca
o
e<4
I.
0«
*
O
-I.- s? B« s: || j||
x ||| sis Us «J t:g?r-l
SwSS 2 '
3 *
2: >« i i
CD
*^22 ' '
« ^
I!^S5 | | ffp"
*- * J A-A-S
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
31
20%
COARSE
MASS
63%
SULFATE
17%
OTHER
FINE
MASS
FIGURE A-20. Typical extinction budget for rural Southwest.
(Grand Canyon, Arizona).
(NHJ,SO4 (Ammonium Sul(ate)
Organic Carbon
Soot
Fine Crustal
Undefined Fine
Coarse (Soil)
FIGURE A-21. Typical "theoretical" extinction budget for
rural Southwest.
-------
32
nitrate, organic aerosols,, etc. are correlated with sulfate concentrations and
are not captured by particle samplers, the result of a multiple linear
regression analysis would be to artificially inflate bex^/[S] ratios. A very
pressing research need is to explicitly examine the contribution of volatile
aerosols to the overall extinction budget.
2.5 MODELING .
2.5.1 Receptor Models
Once particles that contribute to visibility reduction are identified, it
is of interest to determine their origin. Diagnostic receptor-oriented models
have evolved as a clear alternative to source-oriented dispersion
models.(51,52,53,54,55) Receptor models start with the measurement of a spe-
cific feature of the aerosol at the receptor, and after the fact calculate
contributions of specific source types and/or source location. Historically,
receptor models have been thought of in terms of chemical mass balance (CMB)
analysis (or modifications thereof). If successfully applied, CMB achieves
source apportionment but does' not identify locations of source or source type.
Recent developments in statistical treatment of back trajectory analysis
techniques form another type of receptor modeling which identifies locations
of various source types.(56).
-- Difficult questions arise as to the appropriateness of the various
^techniques. For instance, CMB requires among other assumptions, that source
[emission composition is known and that aerosols are non-reactive.(56) If
sources that impact a receptor are hundreds of kilometers distant, identifica-
tion of emission composition from all sources is impractical. Furthermore,
over these distances, aerosols are created, change and are lost. The reaction
mechanisms and rates for this evolution are not well understood. This is not
to say that CMB cannot be used, but rather the conditions necessary for a
specified degree of confidence must be clearly defined. Within proximity to
the source, CMB has been used quite successfully.
On the other hand, back trajectory techniques have been applied quite
successfully in remote areas such as the Colorado Plateau and western North
Dakota region. Figure A-22 is an isopleth plot of the source contribution
function (SCF) for "extreme" sulfur concentrations at Grand Canyon.(57)
"Extreme" sulfur concentrations are defined to be those days when the sulfur
concentration is 1 standard deviation greater than the mean. The geographic
domain shown in Figure A-22 is subdivided into 1/2 by 1/2 degree grid cells.
Each isopleth line corresponds to those grid areas that are associated with
the same number of back trajectory end points. The numbers associated with
each isopleth line is the probability that each grid area along the isopleth
line will contribute to extreme sulfur contributions at Grand Canyon relative
to the grid area that is most likely to contribute high concentrations. For
instance, grid areas along the 0.73 line are 0.73 times as likely to contribute
to high sulfur as the grid that has the highest source contribution value.
Based on this analysis, the major source of sulfate at Grand Canyon
appears to be California's south coast air basin, not the copper smelter
region in Southern Arizona as once assumed. Only under certain meteorological
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
33
s_
3
3
VI
X
O)
u
c
J2
O
u
01
01
O O
*>£
o
I/I
1 (3
a: at
=) o
u_ u
sanuivn
-------
34
I
I
conditions do emissions from a smelter reach class I areas in the Colorado
Plateau area.
Similar analysis for low sulfur days shows those source areas (or non-
source areas) that correspond to air masses that arrive at Grand Canyon with _
low sulfur concentrations. To the north of Grand Canyon are those remote
areas where there is an absence of urban centers or significant industrial
activity.
Current thought concerning industrial siting criteria associated with new I
energy development suggests that extraction and processing of natural
resources such as coal, tar sands, and oil shale should be carried out close
to actual deposits. However, trajectory analysis suggests that the only days I
that Grand Canyon has clean air is when the air originates from those same
areas that contain significant amounts of coal and tar sands. Development of
industry that emits even small amounts of pollutants in either southern Utah I
or southern Colorado could impair the only remaining clean days that occur at
Grand Canyon National Park.
Because back trajectory techniques rely on sparse meteorological data ' |
gathered by the National Weather Service, these techniques can only be
expected to perform satisfactorily when applied to pollutant impacts that have
their origin hundreds of kilometers distant from the receptors. They probably
will perform poorly if impacts are due to sources that are less than 100km
from the receptor site.
There is a need to understand the scale over which CMS and trajectory |
techniques are valid. When should analysis rely on CMB or trajectory
techniques or when can they be used in combination? Ideally, new analysis _
techniques will evolve that will incorporate strong points of both approaches
into one analysis technique. ;
One recent analysis did combine extinction budget analysis with receptor
modeling techniques to determine a source extinction budget for a number of |
source categories at Lake Tahoe.(42) Results show a large seasonal variation
in visibility impacts by wood smoke associated with residential heating. In a _
similar analysis, Albuquerque's overall contributions to extinction are as I
follows: mobile sources, 45.2%; wood burning, 51.6%; natural gas, 1.7%, and "
crustal, 1.5%. During the day, mobile sources dominate wood burning 78.2%
compared to 19.0%, while during the night, wood burning dominates the mobile
sources, 57.0% compared to 40.6%.(58)
Similar results were found in Denver, where no single source dominates
visibility reduction.(5) Combustion sources, as a group, however, account for
more than 80% of the visual range reduction (VRR). The largest contributor to
VRR is motor vehicles which account for 27% VRR. Within this category, diesel
trucks, which account for only 4% of the vehicle miles traveled (VMT), con- I
tribute 12% of the total VRR. These diesel truck particulates are dominated I
by elemental carbon' which, by far, is the most effective visibility-reducing
species. Light-duty catalyst- vehicles which account for the largest part of
the VMT (50%), contribute only 5% of the total VRR.
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
35
The second largest source category is coal combustion, which contributes
25% of the VRR. This source appears to be responsible for most of the sulfate
(75%) and is the largest (39%) source of nitrate in Denver. Both of these
species are secondary particulates, existing primarily as ammonium salts.
Based on a radiocarbon-dating technique and emissions testing, it is
estimated that wood burning contributes 18% of the VRR during the winter in
Denver. This source is the largest single source of both elemental carbon
(39%) and organic carbon (29%) particulate in Denver.
Combustion of natural gas is estimated to account for 12% of the visual
range reduction. These particulates, too, are essentially .all carbon. Fuel-
oil combustion contributes 8% of the VRR. The most important species from
this source are elemental carbon, organic carbon and sulfate.
2.5.2 Predictive Modeling
Whereas receptor models tend to be used in a diagnostic sense--that is to
determine the cause of visibili'ty impairment--predictive models are used to
determine visibility impacts of new proposed sources. Typically, visibility
predictive models allow calculations of pollutant dispersion, chemical
reaction between pollutants and between pollutants and "background"
atmospheric constituents, and the effect that pollutants have on the ability
to see a scenic vista.
y
Current models allow for prediction of the visual impact of plumes.
(59,60,61) These models are typically built on Gaussian plume dispersion
algorithms with subroutines" to calculate chemical conversion of gases to ,
optically active aerosols, and radiative transfer for specific lines of sight.
These models have been evaluated by comparing model calculations and field
measurements as part of a multi-agency funded research program. The
evaluations indicate that plume visibility model performance is comparable
to that of standard plume dispersion models and that most of the uncertainty
in plume visibility models is in the specification of plume dispersion.
Plume visibility models and existing screening tools are inadequate to
calculate cumulative regional effects. Regional visibility models are
designed for this purpose.(62,63) These models utilize regional emissions
and meteorology as input and calculate transport, dispersion, wet and dry
deposition, chemical conversion, and resulting optical effects on a regional
scale.
Several regional models have been applied to predict haze. Corre-
lations with various indicators of experimental- data vary from 0.4 to 0.8
(64,65). Problems in visibility models include those common to .regional
models plus the problems of going from an aerosol concentration to visibility.
Transport
trajectories.
- For realistic transport, models' should use multilayer
-------
36
Relationship of Visibility to Aerosol Parameters Aerosol extinction can
alculated reasonably well from aerosol mass;
to be able to do this for individual components.
Visibility Parameters -- There is no agreement as to which visibility
parameters should be used when averages need to be calculated for comparison
with data or for policy consideration-mean, median, upper quartile, number of
days with visibility less than some value, or some other parameter.
Model Type Eulerian models permit use of detailed .chemistry; Lagrangian
models allow better depiction of source-receptor relationships. Monte Carlo
models have advanced a useful compromise for some applications. However,
there is no universal agreement of the best model or type of model to use.
Current models do not address certain transitional cases that produce
ground based and layered hazes. During winter months, stable atmospheric
conditions can persist for periods of between two and ten days. For
instance, during winter months, ground based layered haze builds up in
Glen Canyon (Lake Powell) to a point where portions of Navajo Mountain, as
viewed from Bryce Canyon, disappear. .Without adequate models, new source
review processes cannot account for this potentially significant visibility
impact. Consequently, there is a pressing need to develop predictive models
that allow for determining visibility impact resulting from ground-based haze
layers that are formed by pollutants emitted in stagnant air mass conditions.
I
Dispersion Regional dispersion is governed by wind sheer and veer. The
wind field derived from the twice daily NWS radiosondes is inadequate. _
Realistic dispersion requires more experimental detail, empirical parameter!- I
zation, or theoretical wind field modelling to provide accurate interpolation.
Volatile Components .Organic aerosol, nitrate aerosol and aerosol-bound
water are not adequately measured by existing technology.
I
Carbon Carbon soot and organic aerosols are significant components. _
Emission inventories for soot, primary organic aerosols and aerosol forming "
organic vapors are inadequate. Chemical mechanisms and rates for the forma-
tion of secondary organic aerosols are not known well enough to predict
organic aerosol formation. There is still some lack of agreement over appro-
priate technique for measuring aerosol absorption. |
Size Distribution Aerosol dynamic models, especially for individual M
components, need further development.
Relative Humidity Effects '-- Sulfate aerosols grow in size as relative
humidity increases but the point of deliquescence depends on composition. As
relative, humidity decreases the point of crystallization (loss of water and
change to solid) occurs at a lower RH the deliquescence (hysterisis). The RH
at which crystal crystallization occurs may be very low (<10%) and depends on
composition.
be calculated reasonably well from aerosol mass; however, it would be desirable I
-------
i:
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
37
REFERENCES
1. Hering, S.W., J.L. Bowen, J.G. Wengert, and L.W. Richards, "Characterization
of the regional haze in the southwestern United States," Atmos. Environ.,
15{10/11):1999 (1981).
2. Flocchini, R.G., T.A. Cahill, L.L. Ashbaugh, R.A..Eldred and M. Pitchford,
"Seasonal behavior of particulate matter at three rural Utah sites,"
Atmos. Environ, (in press). .
3. Evans, G.W. and S.V. Jacobs. "Air pollution and human behavior," J. of
Social Issues, 37{1):95-125.
4. Friedlander, S.K, Smoke, Dust and Haze. John Wiley & Sons, 1977.
5. Groblicki,*P.J., G.T. Wolff and R.J. Countess, "Visibility-reducing
species in the Denver 'brown cloud1I. Relationships between extinction
and chemical composition," Atmos. Environ., 15(12):2473 (1981).
6. Macias, E.S., J.O. Zwicker, J.R. Ouimette, S.V. Hering, S.K. Friedlander,
T.A. Cahill, G.A. Kuhlmey, and L.W. Richards, "Regional.haze case studies
in the Southwestern U.S.I. Aerosol chemical composition," Atmos.
Environ., 15(10/11):1971 (1981).
7. Guide to Meteorological Instrument and Observing Practices, Fourth Ed.,
Secretariat of the World Meteorological Organization, Geneva, Switzerland
WMO-No. 8.TP. 3, 1971.
8. Malm, W., K. Kelley, J. Holenar and T. Daniel, "Human perception of visual
air quality (uniform haze)," Atmos. Environ., 15(10/11):1875 (1981).
9. Malm, W.C., K.K. Leiker and J.V. Molenar, "Human perception of visual air
quality," J. Air Poll. Control Assoc., 30(2):122 (1980).
10. Latimer, D.A., H. Hogo and T.C. Daniel (1981). "The effects of
atmospheric optical conditions of perceived scenic beauty," Atmos.
Environ., 15:1865 (1981).
11. Middleton, P., T.R. Stewart and R.L. Dennis (1983b). "Modeling human
judgments of urban visual air quality," Atmos. Environ., 17:1015 (1983b).
12. Malm, W.C., A. Pitchford, R. Tree, E.G. Walther, M.'Pearson and S.F.
Archer, "The visual air quality predicted by conventional and scanning
telephotometers and integrating nephelometers," Atmos. Environ., 15(11):
(1981)
13. Visibility seasonal data summaries from National Park Service visibility
monitoring program, available from: Dr. William C. Malm, Air and'Water
Quality Division, National Park Service, 301 S. Howes, Fort Collins, CO
80521.
14. Shenandoah National Park enacting legislation.
-------
38
15. Waggoner, A.P., R.E. Weiss, N.C. Ahlquist, D.S. Covert, S. Will and R.J.
Charlson, "Optical characteristics of atmospheric aerosols," Attnos.
Environ., 15(10/11):1891 (1981).
16. Allard, 0., I. Tombach, "The effects of non-standard conditions on
visibility measurement," Atmos. Environ., 15:1847 (1981).
17. Ellestad, T.G. and R.E. Speer. (1981) Application of a Telephotometer
to Visibility Measurements in the Eastern United States, Atmos. Environ.
JJ5_, 2443-2449.
18. Tombach, I.H., E.W. Allard, R.L. Drake, R.C. Lewis, C.V. Mathai, B.M.
Muller, Western Region Air Quality Studies Interim ReportVisibility and
Air Quality Measurement Program--1981-1982, technical report to be
published by EPRI, 1985.
19. Malm, W.C. and J.V. Molenar, "Visibility measurement in National Parks in
the western United States," J. Air Poll. Control Assoc., 34(9):899
(1984).
20. Malm, W.C. and E.G. Walther, "A review of instruments measuring visibility
related parameters," EPA-600/4-80-016, February 1980.
21. Mueller, P.K., ^'Visibility research," EPRI (Electric Power Research
Institute) Journal, 9(5):50 (1984).
22. Evans, W.E. and W. Viezee, EPRI Automated Telephotometer: Field Test,
Color Measuring Capability, and Data Analysis, EPRI Project Report EA-
2386, Project 1630-10, May 1982.
23. Malm, W.C. "Considerations in the measurement of visibility." J. Air
Poll. Control Assoc., 29(10):1042 (1979).
24. Appel, 8. "Sampling of carbonaceous particles in the atmosphere," Atmos.
Environ., 17:1787 (1983).
25. Appel, B. "Sampling nitrates in ambient air," Atmos. Environ., 15:283
(1981).
26. Patterson, O.E., J.M. Holloway and R.8. Husar. (1980) Historical Visi-
bility Over the Eastern U.S. Daily and Quarterly Extinction Coefficient
Maps, EPA-600/3-80-43a,b,c,d,e,f.
27. Husar, R.B., O.E. Patterson, J.M. Holloway, W.E. Wilson and T.G. Ellestad.
(1970) Trends of Eastern U.S. Haziness Since 1948. Proceedings of
4th Symposium on Atmospheric Turbulence Diffusion and Air Pollution,
Reno, Nevada.
28. Husar, R.B. and D.E. Patterson. (1980) Regional Scale Air Pollution:
Sources and Effects, Annals of the New York Academy of Sciences 338,
399. :
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
39
29. Husar, R.B., J.M. Holloway and D.E. Patterson. (1981) Spatial and
Temporal Pattern of Eastern U.S. Haziness: A Summary, Atmos. Environ.
15, 1919-1928.
30. Husar, R.8. (1985). Progress Report on Cooperative Agreement 810351.
Prepared for Atmospheric Sciences Research Laboratory. U.S. Environmental
Protection Agency, Research Triangle Park, N.C.
31. J. Trijom's and K. Yuan, "Visibility in.the Southwest: An exploration of
the historical data base," EPA-600-3-78-039, U.S. Environmental
Protection Agency, Research Triangle Park, N.C.
32. Trijonis, J. "Visibility in California," J. Air Poll. Control Assoc.,
32:165 (1982).
33. Duckworth, S. and J.J.R. Kinney, "Visibility trends in the pristine areas
of California, 1958-1977," California Air Resources Board, Technical
Services Division, P.O. Box 2815, Sacramento, CA 95812, September 1981.
34. Duckworth, S. and J.J.R. Kinney, "Visibility trends in the coastal areas
of California, 1958-1977," California Air Resources Board, Aerometric
Analysis Branch, P.O. Box 2815, Sacramento, CA 95812, December 1980.
35. Duckworth, S. and J.J.R. Kinney, "Visibility trends in the Great Central
Valley of California, 1958-1977," California Air Resources Board,
Aerometric Analysis Branch, P.O. Box 2815, Sacramento, CA 95812,
October 1978. .
36. 1983-1984 Annual Report to the National Park Service for Particulate
Monitoring and Data Analysis, NPS Contract Number USDICX-0001-3-0056,"
produced by Air Quality Group, Crocker Nuclear Laboratory, University of
California, Davis, CA 95616.
37. Washington State Visibility-Study 1982 Final Report, produced by R.W.
Beck and Associates for the Washington Department of Ecology, Air
Programs, Rowesix, Olympia, WA 98504; May 1983.
38. Ouimette, J.R. and R.C. Flagan, "The extinction coefficient of
multicomponent aerosols," Atmos. Environ,, 16:2405 (1982).
39. Hasan, H. and T.G. Dzubay, "Apportioning light extinction coefficients to
chemical species in atmospheric aerosol," Atnos. Environ., 17:1573
(1983).
40. Wolff, G.T., M.A. Ferman, N.A. Kelly, D.P. Stroup, M.S. Ruthkosky, "The
relationship between the chemical composition of fine particles and
visibility in the Detroit metropolitan area," J. Air Poll. Control
Assoc., 32:1216 (1982). . -
41. Dzubay, T.G., R.K. Stevens, C.W. Lewis, D.H. Hern, W.J. Courtney,.J.W.
Tesch, M.A. Mason, "Visibility and aerosol composition in Houston,
Texas," Environ. Sci. Techno!., 16:514 (1982).
-------
40
42. Pitchford, M., D. Allison, "Lake Tahoe visibility study," J. Air Poll.
Control Assoc., 34:213 (1984).
43. White, W.H. and P.T. Roberts, "On the nature and origins of visibility-
reducing aerosols in the Los Angeles Air Basin," Atmos. Environ., 11:803
(1977).
44.. Henry R.C. and G.M. Hidy, "Multivariate analysis of particulate, sulfate,
and other air quality variables by principal components - 1. Annual data
from Los Angeles and New York," Atmos. Environ., 13:1581 (1979).
45. Appel, B.R., Y. Tokiwa, J. Hsu, E.L. Kothny, E. Hahn and J.J. Wesolowski.
(1983) Executive Summary to Visibility Reduction as Related to Aerosol
Constituents, CA/DOH/AIHL/SP-29.
46. Covert, O.S., A.P. Waggoner, R.E. Weiss, N.C. Ahlquist and R.J. Charlson.
(1979) Atmospheric Aerosols, Humidity, and Visibility, in Character
and Origins of Smog Aerosols {Hidy, ed.), John Wiley & Sons Inc.
47. Leaderer, B.P., R.L. Tanner, P.J. Lioy and J.A.J. Stolwijk. (1981)
Seasonal Variations in Light Scattering in the New York Region and Their
Relation to Sources, Atmos. Environ. 15", 2407-2420.
48. Lewis, C.W. (1981) On the Proportionality of Fine Mass Concentration
and Extinction Coefficient for Biomodal Size Distributions, Atmos.
Environ. 15, 2639-2646.
49. Crume, R.V., D.L. Fox and T.6. Ellestad. (1984) Scattering Efficiencies
of Sulfate and Nitrate Aerosol in an Outdoor Smog Chamber. To be published,
50. Ferman, M.A., G.T. Wolff, and N.A. Kelly (1981). The Nature and Sources
of Haze in the Shenandoh Valley/Blue Ridge Mountains Areas. JAPCA 31:
1074.
51. Henry, R.C., C.W. Lewis, P.K. Hopke, H.J. Williamson, "Review of receptor
model fundamentals,"
52. Dzubay, T.G., "Chemical element balance method applied to dichotomous
sampler data," Ann. N.Y. Acad. Sci., 338:126 (1980).
53. Core, J.E., P.L. Hanrahan and J.A. Cooper, "Air participate control
strategy development: A new approach using chemical mass balance
methods." In: Atmospheric Aerosol: Source/Air Quality Relationships,
Macias and Hopke, Eds., ACS Symp. Series No. 167, American Chemical
Society, Washington, D.C.
54. Stevens, R.K. and T.G. Pace, "Overview of the Mathematical- and Empirical
Receptor Models Workshop (Quail Roost II)," Atmos. Environ., 18:1499
(1984).
55. Gordon, G.E., W.R. Pierson, J.M. Daisey, P.J. Lioy, J.A. Cooper, J.G.
Watson II and G.R. Cass, "Considerations for. design of source
apportionment studies," Atmos. Environ., 18(8):1567 (1984).
-------
I
156. Watson, J.G., "Overview of receptor model principles," J. Air Poll.
Control Assoc., 34(6):619 (1984).
157. Ashbaugh, L.L., C. Malm and W.Z. Sadeh, "A methodology for establishing
the probability of the origin of air masses containing high pollutant
concentrations," Number 83-104.13, Proceedings of the 76th Annual Meeting
of the Air Pollution Control Association, (1983).
58. Zak, B.D., W. Einfeld, H.W. Church, G.T. Gay, A.L. Jensen, J. Trijonis,
I M.D. Ivey, P.S. Homann, C. Tipton. "The Albuquerque winter visibility
study, Volume 1. Overview and data analysis," Report No. SAND84-0173/1.
June 1984. Sandia National Laboratory, Albuquerque, NM 87185.
159. Elthgroth, M.W. and P.V. Hobbs, "A numerical model for power plant plumes
including impacts on atmospheric visibility," unpublished manuscript.
160. Latimer, D.A., R.W. Bergstrom, S.R. Hayes, M.K, Liu, J.H. Seinfeld,
G.Z. Whitten, M.A. Wojcik and M.J. Hillyer, "The development of mathematical
models for the prediction of anthropogenic visibility impairment,"
. EPA-450/3-110a,b,c.
61. Chan, L.Y., M.D. Williams, D.H. Nochumson, R.J. Lewis, "User's manual for
the Los Alamos Visibility Model (LAVM)," Los Alamos Scientific Laboratory,
Los Alamos, New Mexico.
62. Nochumson, D.H. and M.D. Williams, "Copper .smelters and atmospheric :
I visibility in the Southwest seasonal analysis," Paper 83-108.12, 76th
Annual Meeting of the Air Pollution Control Association, Atlanta,
Georgia, June 19-24, 1983.
63. Latimer, D.A., H. Hogo, R.G. Ireson, R.E, Morris, P. Saxena, "Development,
* application and evaluation of regional visibility models," Paper 84-
115.4, 77th Annual Meeting of the Air Pollution Control Association, San
Francisco, California, June 24-29, 1984.
I
I
I
I
I
/
I
I
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX B. PROJECTING FUTURE REGIONAL VISIBILITY
by: Sidney Worthington, EPA/OPPE
Vivian Thomson, EPA/OAR
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX B*
TABLE OF CONTENTS
1.0 Introduction
2.0 Eastern Study
2.1 Description
2.2 Study Results
2.3 Major Uncertainties
3.0 Western Study
3.1 Description
3.2 Study Results
. 3.2.1 Emission Projections
3.2.2 Source Apportionment
3.2.3 Visibility Projections
References
Page
l'
1
1
2
5
7
7
8
8
11
13
22
-------
APPENDIX B
FIGURES
B-l 1980-Annual Average Visual Range
B-2 Contributions of $04 to Total Anthropogenic Light
Extinction
B-3 1980 to 1985 Change in Annual Visual Range
B-4 1995 (Base) to 1995 (Low Emission) Change in
Visual Range
B-5 1995 (Base) to 1995 (High Emission) Change in
Annual Visual Range
APPENDIX B
TABLES
B-l Eastern S02 Emissions
B-2 Modelled Annual SOg Reductions from 1980 to 1995
B-3 Improvement in Yearly Median Visibility
B-4 1980 Regional Emissions
B-5 1995 Regional Emissions
B-6 Changes in Regional Emissions by Source Category
B-7 Contributions to Light Extinction in Urban and Nonurban
Atmospheres
Page
15
16
18
19
20
3
. 4
6
9
10
12
14
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX B. PROJECTING FUTURE REGIONAL VISIBILITY
1.0 INTRODUCTION
If we are to devise a sensible strategy to protect the air
from visibility degradation, we need to estimate how large future
emissions will be, where they will occur, and how they will affect
visibility. To that end EPA has performed two major studies.on
future regional visibility. One looks at the 31-state region in
the eastern U.S. that has been proposed as an acid rain "control
region. The other looks at a six-state region in the western
U.S. that contains numerous Class I areas with pristine air
q u a 1 i ty .
Although these two studies do not cover every state in the
U.S., they represent all types of visibility problems we are
likely to encounter here. Ideally, we would like to study every
region, but budgets are not adequate to allow us to cover such
a broad area and still maintain the quality of analysis we want.
2.0 EASTERN STUDY
2.1 Description
As explained in Appendix A, visibility degradation in the
eastern U.S. consists mostly of regional haze, a sort of soup,
made up of pollutants from many sources that travel long distances
_and generally obscure-the atmosphere. The largest single
"contributor to visibility degradation in the east is sulfates,"
secondary pollutants formed in the atmosphere from primary emissions
of sulfur dioxide (S02). Sulfates cause more than half of. the
light extinction in the East.
-? The eastern study looks at (1) current visibility, (2) expected
visibility in 1995 if environmental regulations 'remain as they
are now, and (3) expected visibility in 1995 if environmental
co.ntrols on sources of S02 emissions change.
' The study was performed by Systems Applications, Inc. (SAI)
jaf San Rafael, California, with two air quality dispersion models.
The Regional Transport Model - Long Term (RTM-LT) is a long-term
Eulerian model that does not consider near-source dispersion of
spatial scales less than the size of an .80x80 .km grid square.
The Regional Impacts on Visibility and Acid Deposition (RIVAD)
model is a plume-segment Lagrangian model that evaluates regional
source-receptor relationships.
Visual range estimates are obtained from sulfate concentrations
calculated by RTM-LT and RIVAD using theoretically and empirically
-------
2
I
I
derived formulas. The study assumes that concentrations of
nonsulfate species remain constant at 1977/1978 levels in all
future control scenarios. Thus,"only sulfate concentrations I
change in the control scenarios. Our results are limited by this
assumption since the concentrations of other visibility-impairing
pollutants may well change between the two study periods.
Evaluation of model performance showed that up to 60 percent
of the variance in regional air quality is explained by the models
used in this study. The evaluation was not capable of determining |
the ability of the models to predict specific source-receptor
relationships, like the effect of emissions from a specific power _
plant on a specific National Park.
Estimates of emissions from most sources come from the
National Emission Data System (NEDS) Snapshot Emission Inventory,
developed for the National Acid Precipitation Assessment Program |
(NAPAP). However, because E.H. Pechan and Associates has more
complete and accurate emission and stack data for electric utilities,
SAI replaced the NEDS file with the Pechan file for that source.
ICF, Inc., produced future emission estimates for the control
scenarios.
SAI introduced seasonal variations in emissions by developing |
seasonal factors, one for each of four seasons, for utility,
industrial, and area sources using the 1978 EPRI/SURE data base. «
Emissions from sources outside the modeling region and ' *
emissions from sources within the region that are transported
outside and then returned with changing wind direction are .treated I
by assumed boundary conditions. I
2.2 Study Results
Table B-l presents S02 emission estimates for 1980 and 1995
with no change in the regulatory status quo. Very little change
in eastern U.S. and Canadian emissions is expected. Emissions I
in the entire region in 1980 were 24.1 million tons and in 1995
are projected to be 24.7 million tons. The differences in visual
range calculated from the differences in emissions are concomitantly
quite small. They ape only about + 2 percent. {See Table A-2). |
This change is highly unlikely to be perceptible on an annual
basis. Consequently, overall eastern regional visibility in 1995
will be about the same as it is today, unless a new emission
control program is implemented.
I
program is implemented.
In addition to estimating the change in visibility degradation
in 1995 under the status quo, SAI examined several-S02 emission
control scenarios that would improve visibility as well as other
sulfur-related pollution problems such as acid rain. Reductions _
in S02 emissions ranged from 3.8 to 12.0 million tons per year.
PTable 2 describes those scenarios. As expected, increasing
mprovements in visibility accompany further reduced S02 emissions.
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 8-1.. EASTERN S02 EMISSIONS
(103 tons/year)
State
Alabama
Arkansas
Connect!cut
Del aware
District of Columbia
Florida
Georgi a
Illinois
Indi ana
Iowa
Kentucky
Loui si ana
Maine
Maryl and
Massachusetts
Mi chi gan
Mi nnesota
Mississippi
Missouri
New Hampshire
New Jersey
New York
North Caroli na
Ohio
Pennsylvania
Rhode Island
South Carolina
Tennessee
/ Vermont
^p Vi rgii,ni a
West Virginia
Wisconsin
Total Eastern U.S.
Eastern Canada
TOTAL
1980
751
102
69
108
10
1061
807
1455
2140
333
1142
300
95
341
327
949
243
278
1388
92
272
943
612
2660
1978
15
323
1053
^"355
1128
640
21,976
2,108
24,084
1995
695
203
- 70
. 93
7
1268
1005
1257
2145
315
969
485
99
385
331
961
274
438
1383
63
369
855
651
2797
1948
7
387
1146
^415
1172
858
23,058
1 ,625
24,683
-------
SCENARIO
10-State
12-State
12-State +
25% reduction
in Canadian
emi ssions
31-State
8-nti 1 1 ion ton
reduction
31-State
1 2 - mi 1 1 i o n ton
reduction
4
TABLE B-2
Annual
S02
Reduction
TARGETED from 1980
STATES (106 tons)
ME, VT, NH, MA, 3.8
CT, RI, NY, PA
WV, OH
' Same as 10-state, 5.2
plus IN, MI
Same as 12-state, 5.6
plus eastern
Canada
31 states east of
and borderi ng on 8.0
Mississippi River
31 States east of
and bordering on ' 12.0
Mississippi River
-
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table B-3 shows percent improvements in visibility in
1995 for each of the emission reduction scenarios. The improvement
p~i n visual range is usually greatest in the middle Atlantic states
( LDelaware, Maryland, Virginia, and West Virginia), followed
closely by New Jersey, Pennsylvania, Kentucky, North Carolina,
New York, and Ohio. Visibility improvements for individual states
range from none for Missouri in the Id-state control option to 32
percent in Maryland, Delaware, New Jersey, and Pennsylvania in
the 12-mi11 ion-ton control option.
The important question to ask is whether these changes in
vjsibility are perceptible. For the 10-state control scenario,
none of the time and state-averaged percent improvements in visual
range reaches the 15 percent threshold of day-to-day perceptibility.
In this case, the value of the S02 reductions, in terms of visibility
alone, would be near zero.
The costs of S02 reduction in the
control scenarios are presented below:
U.S. for each of the
Scenari o (bi 1
10-state
12-state
12-state
+25% Canada .
31-state
8 million tons
31-state
12 million tons
Annual i zed
Costs
lions of 1982 S/year)
2.0
2.4
2.4
4.3
9.8
$/ton
S02
Removed
349
348
348
371
667
As we increase the stringency of S02 controls in each
scenario, increasingly more states' do reach perceptible
changes in visibility and do reap positive economic benefits
from controls. For the 12-millipn ton-per^year control scenario,
18 of the 31 states in the modeling region achieve perceptible
improvement in summer visibility.
2.3 Major Uncertainties
It is
associ ated
projection
processes,
percepti on
study is c
the study .
1985) cone
of visibi1
important to note the limitations and uncertainties
with the eastern analysis. Uncertainties in emisison
s, treatment of meteorological, transport and transformation
pollution-visibility relationships, and treatment of
"thresholds" all impose significant limitations. The
urrently undergoing peer revi'ew. A detailed review of
for the Utility Air Regulatory Group (Zanetti et a!.,
luded that the methods used "tended to provide estimates
ity improvements and economic benefit that were greater
-------
e
~
o
I/I
in
OJ
u
4)
01
I
0.
>
e
>e
1
>>
at
utdd SZ'O
utdd 05*0
(UJ5|)
o.
i
CD
UOIILJ.UI 21
nns IE
ui
nns zi
oi
S66T
0861
. CM in *. ifl f» » S» o» . sa N CM « 01 va « a CM
tv0i m cv « « rg (0 in in in to v w CM a
v \« v n;a m vo
» CM » « ie « in «M » m n m « o»
n M r» CM CM P,
.-.« CM
~ cvno * »in r> »o cst» « ie * CM
M« .v
CM N
M IM M
CM (T) v> <
Ztt. X W -I U
tf
< at <
i/i z
z > z o
o
j ^ tj (j
X
-------
I
I
I
I
r
I
I
I
I
I
I
I
I
I
than one might derive using unbiased estimates of emissions
reductions, nonlinear atmospheric chemical reactions, the
extinction efficiency of sul f ate-contai ni ng participate matter,
intrinsic contrasts of targets, the range of human thresholds of
response to visibility changes, and the economic value of visibility
We thus conclude that 'the SAI visibility improvement percentages
tend to be close to the upper limit of our confidence range and,
compared to our best unbiased estimate, they over-estimate
visibility improvement by 50% for both the 8 MTPY and the 12 MTPY
scenarios." The SAI authors have reviewed this critique and,
while agreeing with some points, reject its major findings and
feel their original analysis does not overstate visibility
improvements. Based on the examination of the contribution of
sulfates to eastern haze in Appendix, it could be argued that
some components of the SAI work understate visibility improvements,
while others.may overstate sulfate reductions. Resolution of the
implications of the uncertainties and potential biases following
the peer review is of major importance. Based on the assessments
to date, uncertainties in dealing with perceptibility of visibility
changes appear.more influential than the remaining uncertainties
combined.
3.0 WESTERN STUDY
3.1 Descri ption
The study projecting regional visibility in the West focuses
on the six states of California, Nevada, Utah, Arizona, Colorado
and New Mexico. These states contain over half of the Class I
areas in the country, including the Golden Circle where many
areas enjoy pristine air quality.
This study thus omits the Pacific Northwest, an area that
is subject to periodic -visibility degradation from controlled and
uncontrolled fires. Controlled burning.by farmers to manage
soils and by foresters to manage forests and forestry, along with
uncontrolled wild fires, emit large amounts of smoke. This smoke
consists mostly of fine particles that, are very efficient at
obscuring the atmosphere. Hence, the Pacific Northwest suffers
severe episodes of visibility degradation. An ongoing study
(PANORAMAS) is investigating the frequency, extent, and sources
of haze in that region. Other regions of the west contain
fewer class I areas and do not report substantial haze problems.
The objective of the western study was to model current
(1980) and projected (1995) visibility conditions in the six
western states mentioned above and to calculate different source
and pollutant contributions to visibility degradation. While the
eastern study examined only the effects of sulfur dioxide emissions
on visibility, the western analysis looked at the effects of
volati1e organic compounds (VOC), oxides of nitrogen (NOx), and
particulate matter (PM) as well.
-------
8
I
I
In the first phase of the study, E.H. Pechan and Associates,
Inc. developed and refined detailed emission estimates for 1980 '
and 1995. Pechan used the NAPAP emission inventory as the primary |
source of data for total PM, S02, NOx, and VOC. Where appropriate,
however, data from other sources were substituted for the NAPAP _
data. For example, up-to-date EPA emission estimates for copper I
smelters and woodstoves were substituted, as were USDA Forest
Service estimates for particulate emissions from wildfires and
controlled burning. j|
Where possible, Pechan broke the four pollutant classes into
those constituents which have the greatest /impact on visibility. «
Total VOC emissions were partitioned into individual species, and jl
the individual species emissions were then grouped into reactivity
classes (e.g., olefins, aromatics, etc.). Primary sulfate emissions
were estimated, and PM emissions were broken down into particle
size distributions of less than 2.5 urn, between 2.5 and 10 um, |
and greater than 10 um. , :
To project emissions to 1995, Pechan employed the Environmental I
Trends Analysis Model (ETAM), a model which they originally
developed to assess the environmental implications of alternative
energy futures. The method used to project emissions varied
among the different source categories examined. For example,
utility projections were made by using data on planned utility
installations (e.g., unit size, fuel type), and by linking these
data to various assumptions regarding service life, capacity g
factors, and other relevant parameters. Smelter emission projections
consisted of two scenarios: one scenario assumed that smelters
would control S02 emissions by 1995, the other assumed that 1995
emissions would equal current emissions. The USDA Forest Service
provided growth estimates and emission factors for wildfire and
controlled burning PM emissions. Area source emissions were
scaled up or down according to population projections. Emissions |
from refineries and other major industrial point sources were
scaled by projected changes in industry earnings. _
The western analysis is preliminary and has undergone only
limited review to date. Substantial uncertainties exist in
emissions inventories and projections, modeling tools and inputs,
and extent of verification to date. The results should be |
viewed in light of these uncertainties and limitations.
3.2 Study Results I
3.2.1 Emissions Projections
Tables 8-4 and B-5 present emission figures by -source for |
input to the western visibility model.
Next to fugitive dust, which is always the overwhelming I
contributor to total particulate emissions, burning and motor m
vehicles are the primary sources of PM. The greatest uncertainty
-------
TABLE B-4. 1980 regional emissions. (SAI, 1985}
REGIONAL
Source
Type
Utilities
Smelters
Bef ineries * . .
Other Industrial
Besidential/Ccnan
Gasoline Vehicles
Diesel Vehicles
Cther Mobile
Evaporative
Fugitive Cust
fres Burns/F.F.
Weed Stoves/F.3.
TSTKL
1980 EMISSION 1DTALS
(1000 flsns/2r)
Farticulates
1 .Total
72.9
5
58
.0
.7
172.5
163
640
56
30
0
4785
793
17
6801
.4
.5
.7
.6
.0
.2
.2
.7
.5
PM10
67
4
' 51
142
126
417
41
28
0
2153
, 713
, 17
3763
.0
.7
.9
.2
.2
.3
.4
.5
.0
.2
.7
.6
.3
SCOT
0
0
0
0
0
10
11
0
0
0
317
6
345
.0
.0
.0
.0
.0
.3
.9
.0
.0
.0
.2
.1
^
Gaseous Precursors
SCx
469
1405
364
190
52
30
33
79
0
0
. 1
0
2626
.6
.2
.3
.3
.2
.4
.6
.1
.0
.0
.3
.1
.1
NCx
585.
4.
115.
276.
255.
617.
404.
172.
0.
0.
37.
0.
2469.
0
5
6
2
6
3
0
7
0
0
7
7
3
vcc
4.3
0,1
159.7
222.3
308.7
1011.1
74.3
114.1
888.2
0.0
213.4
1.2
2997.9
*In this and other tables, "refineries" represents emissions from
all petroleum industry sources, including refineries and oil-field
operations.
-------
10
TABLE B-5. 1995 regional emissions. (SAI, 1985)
REGIONAL
Source
1995 EMISSION TOTALS
(1000 Tons/Xr}
Particulatea
type Total
Utilities
$m*iters
Befineries
Other Industrial
Besidential/Conoa
Gasoline Vehicles
Diesel Vehicles
Gther Mobile
Evaporative
Fugitive Dust
Pres Burns/F.F.
Wood Stoves/F.2.
TCTRL
38.
9.
70.
266.
136.
716.
173.
40.
0.
3
6
3
4
4
6
9
2
0
4735.2
1189.7
38.3
7519.3
JM10
31.9
9.0
62.2
219
139
453
123
37
0
2153
1070
38
4338
.7
.7
.9
.2
.4
.0
.2
.6
.1
.3
SOCT
0.
0.
0.
0.
0.
16.
0
0
o
0
0
7
26.0
0.0
0.0
0.0
475.3
13.3
531.9
Gas
SPOT
SCx
485.3
1420.2
377.1
310
53
32
65
103
0
0
2
0
2850
.5
.7
iri*
.2
.5
.0
.0
.0
.2
.5
is Pr<
seursors
NCx
893.7
7.3-
112.3
424.2
289.3
291.0
263.5
227.2
0
0
56
1
2567
.0
.0
.6
.5
.7
VCC
7.2
0.1
195.1
333.1
339.7
278.3
69.3
150.1
871.2
0.0
320.2
2.6
2567.5
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
In the emissions inventory lies in the category of prescribed
burning and wildfires. Substantial variation in participate
emission factors from various agencies exists for burning
activities. For modeling purposes, SAI used the U.S. Forest
Service emission factors of 100 Ib/ton consumed by wildfire and
50 Ib/ton consumed by prescribed burning. In addition, they
partitioned particulate emissions into 70 percent < 2.5 um; 20
percent between 2.5 and 10 um; and 10 percent > 10 um.
Approximately 70 to 80 percent of the region's VOC emissions
and 60 to 70 percent of the region's NOx emissions occur in the
major urban areas and come from motor vehicles and various
residential, commercial, and industrial sources located in those
areas.
Copper smelters contribute 54 percent to the region's loadings
of S02 with utilities and petroleum industry sources contributing
most of the rest. In spite of the projected 51 percent reduction
in U.S. copper smelter emissions, total regional copper smelter
S02 emission are expected to remain fairly constant because of
the projected significant increases in S02 emissions of 680,000
tons per yer at two smeltrs in northwestern Mexico.
Table B-6 summarizes the percentage changes from 1980 to 1995
in regional emissions by source category for each chemical species.
Particulate emissions for all source categories were projected
to increase from 1980 to 1995, except utilities, for which
particulate emissions were projected to decrease by approximately
50 percent. This decrease is largely attributable to particulate
controls at the Four Corners power "plant. Total PM-10 and soot.
emissions were projected to increase by 15 and 54, respectively.
Total S02 and NOx emissions would increase by only 9 and 4 percent,
respectively. Primarily because of hydrocarbon controls on new
motor' vehicles, total regional VOC emissions were projected to
decrease by 14 percent between 1980 and 1995.
3.2.2 Source Apportionment
Fine particles, including sulfates, nitrates, organics, arid
elemental carbon (soot) contribute most of the anthropoge'ni c
visibility impairment in the Southwest. Per unit mass, fine
elemental carbon (soot) is most effective in causing visibility
impairment because it both scatters and absorbs light. Sulfate
and nitrate are also very effective because they are often
associated with liquid water within the aerosol droplet, which
increases the effective scattering area of the aerosol. Organic
and elemental carbon aerosol is emitted directly from combustion
sources, while some organics and most sulfate and nitrate are
formed in the atmosphere from direct emissions of VOC, S02, NOx.
The contribution of nitrate aerosol is difficult to predict
because nitrate can exist as either nitric acid vapor, which does
not impair visibility, or as ammonium nitrate aerosol, which very
-------
12
TABLE B-6. Percentage changes in regional emissions
by source category. (SAI, 1985)
REGIONAL
Source
Type
DtUities
Smelters
Refineries
Other Industrial
Residential/Conn
Gasoline Vehicles
Diesel Vehicles
Other Mobile
Evaporative
Fugitive Dust
Pres Burns/F.F.
Wood Stoves/F.P.
TCTAL
PERCENT CHANGE IN EMISSIONS
FROM 1980 1C 1995
Farticulates
Total
-47
91
19
54
10
11
215
31
0
0
50
116
10
.5
.1
.7
.4
.7
.9
.4
.2
.0
.0
.0
.4
.6
FM10
-52
91
19
54
10
8
197
31
0
0
50
117
15
.4
.2
.7
.4
.6
.8
.6
.2
.0
.0
.0
.1
.3
SCOT
0
0
0
0
0
62
US
0
0
0
50
117
53
.0
.0
.0
.0
.0
.0
.4
.0
.0
.0
.0
.9
.9
Gaseous Precursors
SCx
3
1
3
63
2
6
94
30
0
0
52
217
8
.5
.1
.5
.1
.9
.6
.0
.8
.0
.0
.7
.6
.5.
NCX
52
73
-2
53
13
-52
-34
31
0
0
50
127
4
.8
.9
.9
.6
.4
.9
.8
.5
'0
.0
.2
.2
.0
vcc
68.3
81.6
' 22.2
49.3
10.0
-72.4
-7.3
31.5
-1.9
0.0
50.0
122.1
-14.4
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13
efficiently scatters light. The vapor/aerosol distribution
depends on the ambient temperature, relative humidity, and
concentrations of sulfate, nitrate, and ammonia. Although ammonia
concentrations in the Southwest are variable and uncertain, there
appears to be enough ammonia to form nitrate aerosol. Nitrate
aerosol, however, is unstable and can volatilize at high ambient
temperatures.
The modeling results suggest that much of the visibility
impairment can be explained by the known anthropogenic emissions
in the region in many nonurban areas of the Southwest, however,
natural sources including blue-sky Rayleigh scatter, terpene-
related organics, and soil dust contribute more than half of
total extinction. Table B-7 lists the relative contributions
of various chemical species to visibility degradation in urban
and nonurban areas.
Figure B-l presents current annual average visual range
throughout the study reg'ion as mod'eled by SAI. The best visibility
is in the northern and .eastern parts of the area while the worst
visibility occurs in California around large urban areas.
Sulfate is the species that contributes most to regional
visibility impairment in the Southwest. In most areas sulfate .
aerosol contributes 50-60 percent of anthropogenic visibility -
impairment. In California and in the area in which most of the
copper smelters are located and during summertime episodes
throughout the region, sulfate is predicted to contribute even
more (70-80 percent, See Figure B-2).
The next most important species is nitrate aerosol, which
contributes up to 25 percent of anthropogenic extinction in many
parts of the region. However, nitrate's contribution is very
uncertain because it is volatile and difficult to predict.
Although anthropogenic organic and elemental carbon aerosol
appears to be a significant contributor to urban haze, this
analysis suggests that these species are a relatively small
contributor to regional haze (less than 10 percent). Biogenic
organics (terpenes) appear to contribute more to visibility
impairment than do anthropogenic organics in most parts of the
regi on.
3.2.3 Visibility Projections
For future projections of western visibility, under the base
scenario in most of the region, 1995 sulfate concentrations will
be approximately the same as or slightly lower than 1980 concen-
trations. Significantly increased nitrate aerosolconcentrations
are predicted in large portions of the Four Corners states because
of increased NOx emissions from urban areas and power plants.
Soot concentrations are predicted to approximately double in much
of the region because of assumed increases in the use of prescribed
burning and diesel vehicles. Total PM-10 concentrations would
-------
TJ
ID
43
t_
a
1C
"a.
c
0
u
X
0>
JJ
4C
o
VI
0)
v>
VI
3
o
IB
O
I*
O 9^
3 *
-O
£ <:
c*"'
o
u
4J V
e t.
U Ol
W 4=
L. CL
0) VI
°"S
2 =
43
Lkl (_
_) 3
i i
e
o
. *
**
u
c
x
LU
u
V
01
o
e
*o
1
u
01
a.
§
u
4J
x
LU
O
tow
(to
O
i
c
ID
3
C
£
C
jQ
4V
|
*
e
ig
L.
81
U
14
O O "^ °i 0 °i
Lf> -^
O en eg v r«- cn
« ^ 0 ee »* « «
? ^ "* ~ 2
f) QO ff) ^j pj ^ ^)
<-» a o i.
eu vi « O
jt
u o o o t- a *j
3 C trt Z LU O. Lrt O
CD tZ
o o .
g ^ ^ esi o
~*
^ 0 -: '*: * o
^ CM csi o en o
to m m 10 IT* o» o
O LO i*> o «*> O O
V *« o
CM vo «n LO o o
CM O * CM «M Ol O
CO o
4^^
«M
0 i ,,
1
^L _ _ ._
VI 0)
ot u e -e
^J ^Z W
-^ CU i A O
+* t, in o 4J . ^
o . * 3 a. e -a.
<-> 0,^0 4J t/i e
43 t_ 43 ^ C;
5 * Z £ = °*
"* u'o'c "* u £ IS
(C CO «S JC *J *J
5 S £ 5
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
15
LT>
CO
o
0
u
V
c
«c
£_
9
3
V)
3
/»
C
i
CO
-------
16
u
c
a
a
a
e
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
17
increase somewhat in the region. The maximum $02 concentration
in the region in 1995 will occur in Mexico because of significant
increases in Mexican smelter emissions. (Decreases in U.S.
smelter S02 emissions are expected to be totally offset by.
increases in Mexican smelter emissions.) Because patterns of wet
and dry sulfur deposition will move south, impacts in the sensitive
areas of the Rocky Mountains are predicted to decrease slightly.
Anthropogenic light extinction in California will decrease from
1980 to 1995, but in the rapidly growing areas of the Southwest
it is predicted to increase, primarily as a result of increased
NOx and soot emissions. However, these projections are highly
uncertain because of uncertainties in current and future soot
emission inventories and nitrate aerosol model input information.
Figure B-3 shows that regional visibility is projected to
improve somewhat (2-20 percent) in California and Nevada but is
projected to decline by roughly 5-10 percent in the Four Corners
states. These s u b r e g i o n,a 1 patterns are the-result of local
decreases and increases, respectively, in S02 and NOx emissions
that are projected to cause- 70-80 percent of regional man-made
haze.
As a variation on the 1995 base scenario, SAI evaluated the
impacts associated with a low regional S02 emission scenario for
1995 by starting with the 1995 base-case emissions but assuming
an 80 .percent S02 emission cutback at the two Mexican smelters
and the two largest coal-fired power plants (Mohave and Navajo).^
Sulfate concentrations in southern Arizona would drop from 2.5 t~o
1.5 ug/m3, or about 30 percent. Nonurban sulfate concentrations
in national parks and wilderness areas in Utah, Arizona, and
Colorado would drop 5-15 percent. Anthropogenic light extinction
'decrease nearly 20 percent, resulting in
in visual range. More typically, regional
only slightly (<5 percent) compared to
(See Figure B-4).
in souther Arizona would
a 10 percent improvement
visibility would improve
the base 1995 scenarios.
Another variation on the 1995 base scenario was a high emission
case where, in addition to the 1995 base emissions, SAI assumed
that U.S. copper smelter emissions were not controlled in the
period from 1980 to 1995 and that a large synfuel industry was
developed in Utah and Colorado. The latter assumption is unrealistic
for 1995, but reflects potential increases in the long-tern future.
The nuch larger U.S. copper smelter S02 emissions would result in.
regional sulfate increases of 5-20 percent. Wet sulfur deposition
in the Rocky Mountains would be 20 percent larger than the 1995
base, at an annual rate of 3 kg/ha sulfur. Light extinction in
much of the region would be increased significantly, with large
increases (up to 16 percent) in southern Arizona and New Mexico,
and largest increases (up to 40 percent) in the oil shale development
areas of Utah and Colorado. Visua.l ranges would be decreased
significantly from the 1995 base by 2-8 percent in the southern
Arizona area to 10-16 percent in the oil shale development area.
(See Figure B-5).
-------
18
e
o
t.
o
3
cn
is
c -
IQt-i
cr>o
? u
j: -c
u r
u
-j u
c cn
4J
U CJ
U W
4; «
I
QO
U.
L.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
19
00
cu
cr»
3 O
en
u
>
c u
3 Cfl
-------
20
O CO
i. en
s
t>
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
21
On the basis of this analysis, S02 and NOx emission controls
would have the greatest potential beneficial effect on regional
haze in the Southwest. Although there is concern that the
projected increase in the use of diesel vehicles and wood stoves
will increase local urban haze, these source categories appear to
be relatively insignificant contributors to visibility impairment
on a.regional scale. However, more work is needed to develop and
refine soot emission inventories and regional soot concentration
estimates. The analysis of impacts of controlled burns and
wildfires is highly uncertain; however, over the long run,
burns appear to be relatively smal 1 contributors to haze compared
to anthropogenic sulfate and nitrate aerosol.
-------
3.
4.
22
REFERENCES
ICF, Inc., Analysis of Alternative Emission Reduction
Strategles: Four/Eight/Twelve Mil 11 on Ton Reductlons
and Ten/Twelve State Reductions, prepared for Envlron-
mental Protection Agency, October, 1984.
Pechan, E.H. & Associates, Inc., Baseline Emissions
Inventory and 1995 Emission Projections for PSD and
Visibility Task Force Analyses, prepared for U.S.
Environmental Protection Agency, Office of Policy
Analysis under EPA Contract No. 68-01-6543, February,
1985.
Systems Applications, Inc., Visibility and Other Air
Quality Benefits of Sulfur Dioxide Emission Controls
in the Eastern United States, prepared for U.S. Environ-
mental Protection Agency, Office of Policy Analysis,
SYSAPP-84/165, September 21, 1984.
Systems Applications, Inc., Modeling Regional Haze in
the Southwest: A Preliminary Assessment of Source Con-
tributions, Revised Draft Report, prepared for U.S.
Environmental Agency, National Park Service, U.S.
Department of Energy, SYSAPP/85-038, February 28, 1985.
Zannetti, P., I. Tombach, and R. Drake (1985). Critique
of the Draft Report "Visibility and Other Air Quality
Benefits of Sulfur Dioxide Emission Controls in the
Eastern United States," prepared for Utility Air Regulatory
Group, AeroVironment, Inc. (AV-FR-85/503), February.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX C. VALUE OF VISIBILITY
Perception and Psychological Values
by: William Malm, NPS.
Economic Benefits on visual Air Quality
by: Thomas Lareau EPA/OPPE
Douglas Rae, Consulting Economist
Visibility Effects on Aircraft and Related Operations
by: William Ballanger EPA/Region III
-------
-------
I
I
I APPENDIX C
TABLE OF CONTENTS
1.0 Introduction 1
J 2.0 Perception and Psychological Values 1
2.1 Perception 1
2.2 Psychological Values in Natural Settings 4
References for Section 2 7
3.0 Benefits of Visual Air Quality 9
1 -
I 3.1 Introduction 9
3.2 Quantification of Visual Air Quality Benefits 9
I 3.2.1 Visibility Values 10
3.2.2 Methods to Measure Visibility Values 10
3.2.2.1 Contingent Valuation Survey Approaches 10
3.2.2.2 Hedonic Property Value Approach 14
3.3 Economic Benefit Studies of Visual Air Quality
« in Urban Areas 15
* 3.3.1 Chicago Urban Visibility Study 16
3.3.2 Six Cities Study 17
3.3.3 Cincinnati Visibility Study 19
I 3.3.4 Review of California Visibility/Air Quality 21
. Studies
3.3.4.1 South Coast Air Basin Study 21
3.3.4.2 The San Francisco Bay Area (SFBA) Study 22
I 3.3.4.3 California Property Value Visibility - 23
Study
13.3.4.4 Conclusions from California Visibility 24
Studies
I
I
-------
3.5.1 Theoretical and Empirical Considerations 3D
I
I
Page
3.4 Value of Visual Air Quality in Class I Areas 24 |
3.4.1 Grand Canyon/Southwest Parks Study 24 M
3.4.2 National Park Service Study 26
3.4.3 EPRI Parks Study 28
3.4.4 Six Cities Study 29
3.5 Comparison of Benefits of Visual Air Quality I
in Urban and Class I Areas . 29
I
3.5.2 Comparison of Visibility Benefits in
Urban Areas ' 33
3.5.2.1 Derivation of Urban Visibility- 34
Benefits Estimates _
3.5.2.2 Low and High Estimates 35
3.5.3 Comparison of Class I Area Visibility Studies 35
3.6 Quantification of Dollar Benefits of Improvements in
Visual Range 43 «
References for Section 3 45
4.0 Visibility Effects on Aircraft and Related Operations 49
4.1 Significant and Major Visibility Reductions 49
4.2 Effects of Visibility Reductions on Safety 49 |
4.3 Major Visibility Reductions 50
4.4 Significant Visibility Reductions 50
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX C. VALUE OF VISIBILITY
1.0 INTRODUCTION
C Regional haze can affect public welfare in essentially two areas:
) the subjective enjoyment of the environment (aesthetics, general well
eing) and 2) transportation. Evidence on such effects can be drawn from
studies of perception and awareness of air pollution and scenic beauty,
assessment of psychological values, economic studies,, and visibility
requirements for air draft operation and related activities. Key examples
are illustrated in Table 1. Detailed discussions prepared by task force
members in each of these major areas are presented in the following sections.
The available information taken together indicates that the public is
concerned about visibility and is willing to pay for maintaining or improving
air quality in a variety of consequences. This ability to capture this
value in comprehensive and reliable quantitative terms is, however, limited.
2.0 PERCEPTION AND PSYCHOLOGICAL VALUES
2.1 Perception
A number of visibility "perception" studies have been carried out over
the past 50 years. Investigators in the 1940's, 50's and 60's primarily
concerned themselves with determining the contrast threshold of the human
eye as a function of background illumination, target size, and time searching
for the target. An important result of this research is that threshold
contrast is constant over a wide range of background luminance. However,
as early as 1947, Hecht reported variations in contrast threshold as a
function of size and shape. They found that square objects are more easily
seen than line objects. Taylor investigated the effect of target size in
some detail. He found that for targets viewed in daytime and 1/2 degree or
more in angular size, the threshold contrast ws 0.025 (1/3-sec. research
time)(l). More recently a number of investigators have investigated in
detail the threshold contrast sensitivity of the human eye brain system to
a grating pattern whose brightness varies in a sinusoidal fashion. Results
of this work show that the eye brain system is most sensitive to spatial
frequencies of around 2 to 3 cycles per degree (2).
Whereas work discussed in the previous paragraph has emphasized
detection thresholds, other researchers have concentrated their efforts in
establishing the change in image appearance required to just noticeable
difference (JND) in .image sharpness. This work has primarily been detected
toward incorporating results of basic psychophysical measurements into
models which will predict the change in display modulation transfer functions
(MTF) required to evoke a one JND in display image sharpness. An integral
component of the JND calculation is the establishment of the image mean square
luminance fluctuation, termed the image modulation depth, Henry has suggested
that JND's and modulation depth may be appropriate visibility indices that
incorporate all of the information content in a scenic vista (3).
-------
Aesthetic:
Perception
Economic
Studies
Psycho-
logical
Studies
Transporta-
tion
-
TABLE t. SUMMARY
Effect of Increased
Visibility
Decreased perception of
1r pollution
a) urban settings
b) natural settings
Improved view of
night sky
Increased property values
Enhanced enjoyment (user
or activity values) of
environment 1n:
a) urban settings
b) natural settings
Options values; main-
taining or Increasing
opportunity to visit
less Impaired natural
and urban settings
Existence values.
maintaining pristine
envl romnents
Enhanced enjoyment In
natural areas (user and
options)
/
Less concern over
perceived health
effects
More efficient, loner
risk operations, visual
approach permitted
Increased opportunity
to operate aircraft
Increased opportunity
to conduct aircraft
and weapons testing
2
OF QUALITATIVE EVIDENCE FOR VISIBILITY
Affected groups
Substantial percentage of general
population; urban areas
Outdoor recreatlonlsts, campers.
tourists
Amateur astronomers, other
star watchers
Home owners
,
Urban dwellers
Outdoor recreatlonlsts.
campers, residents of
non-urban areas
Outdoor recreatlonlsts.
campers, tourists
General population
Outdoor recreatlonlsts.
campers, tourists
General population,
urban areas
^
Airport users, operators
l
General aviation aircraft
(non-Instrument capable
pilots, aircraft)
DOO testing facilities
1n Desert Southwest
RELATED VALUES
Supporting Observation
I
M
Perception of air pollution 1n Los Angeles
significantly related to visibility for _
all averaging times (Flachburt and
Phillips. 1980)
Perceived visual air quality significantly
related to particle scattering (Nlddleton
et at., 1984) _
Visual air quality related to awareness of m
haze for Grand Canyon visitors (Ross
et al.. 1984)
Scenic Beauty estimates related to contrast*
transmUtance (Halm et al., 1981) |
Decrease 1n star brightness by fine
particles (Leonard et a!., 197?) mm
Property values related to perception of m
air pollution, hence visibility {Brookshire
et al., 1979; Thay and THjonls, 1981
Lohman. et al.. 1981)
. I
em
Willingness to pay for Increased v1s1-
bllfty in several eastern and western m
urban areas (Brookshire et al., 1979;
Tot ley et al.. 1984; Rae, 1984)
Willingness to pay for preservation.
Improvement and accept compensation
for degradation in national perks
(Rae et al., 1980; Rae et al., 1982)
Aggregate of activity values In itera- 1
live bidding studies suggests importance
of options values (Rowe and Chestnut, 19811
I
Existence values may far outweigh activity
or user values (Rowe and Chestnut, 1981) fj
Visual air quality attributes rank nigh
among visitors to both Mesa Verde and Grand
Canyon. Knowing park resources protected
most valued psychological attribute
(Ross et al.. 1984) |
About 2/3 of bid for Improved visibility
1n Los Angeles was related to concern fj
over potential health effects (Brookshire
et al.. 1979)
Visual approaches permitted when visi-
bility >3-S miles; airport specific _
(FAA, 1980a)
Visual flight rules permitted when visi-
bility >3 miles (FAA, 19806)
I
Basis for location of facilities 1n part
due to high visibility
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Determination of detection thresholds and JND's are statements about
changes in information content in an image. However, a JND change in image
sharpness is not necessarily a good indication of a change in judged image
quality. For instance, a change in 10 JND's in a scene with low overall
contrast may be judged to have the same change in image quality as a 10 JND
change in a high contrast scene.
A just noticeable difference" in the appearance of a vista will be
dependent on whether haze appears as layered or uniform. Layered haze can be
thought of as any confined layer of pollutants that results in a visible
spectral discontinuity between that layer and its background (sky or landscape)
while uniform haze exhibits itself as an overall reduction in air c.larity.
Quantifications of the perceptible effect of these two types of impair-
ment is really quite different. The eye is much more sensitive to a sharp
demarcation in color or brightness than it is to a uniform change in bright-
ness or color over a period of time. Layered haze falls into the first
category in that the layer of haze is observed against some background (sky or
landscape element) while uniform haze falls into the second.
Important features of layered haze are its color, size and edge.
sharpness. Since the eye-brain system is more sensitive to objects with
certain angular or spatial frequencies and to objects with sharp edges, the
calculation of a JND must incorporate these features.; A JND calculation must
also account for differences in threshold between when a layered haze can just
be seen and when a change is taking place in a layer that is already quite
visible.
Whereas the "geometry" (size, shape and color) of layered haze is impor-
tant to its perceptibility, so is the "geometry" of vistas important to its
sensitivity to increases in uniform haze. Visual effects of pollutants are
greatest on vistas with landscape features that produce a retinal image that
subtends approximately 0.3 to 0.5 degrees.(2) Therefore, the combination of
scenic element size and observation's distance become important variables to
integrate into a JND calculation for uniform haze.(3) Research also ,suggests
that a JND resulting from uniform haze is a function of the vista contrast
squared and atmospheric contrast transmittance squared.(4)
Some progress has been made in integrating the perceptible effects of
layered haze and vista geometries into JND calculations. Models have even
been proposed to combine perceptual effects of images-made up of various
spatial frequencies with inherent color of haze layers and landscape features.(5)
Yet no formalism has emerged that comprehensively addresses all of these
complex issues. The ability to quantify the amount of pollution that
constitutes visibility impairment remains ripe for.further research.
In recent years a number of studies have concentrated on establishing
relationships between Judgments of .image quality of natural scenes and various
atmospheric and vista, parameters such as mountain/sky contrast-,- solar angle,
extinction coefficient, sky color and percent cloud cover (6,7,8). Latimer
et al. had observers judge scenic beauty (SBE) and visual air (VAQ) quality
of a- number of eastern and western national park vistas as they appeared ,
under a variety o,f illumination and-meteorological conditions (7).
-------
The results of their study were mixed and in some cases contradictory. In
an early report they conclude "To different extents for different vistas,
ratings of VAQ and SBE both increase with increasing visual range. In a
later study, they conclude "Ratings of SBE of a given vista^were independent
of visual range unless there was a dominant distant landscape feature in
the landscape scenery." Sirice the visual range calculation "normalizes"
out specific unique characteristics of vistas, these results are; not
surprising. These studies did conclude that changes in illumination did
have a considerable effect on SBE ratings.
Middleton et al. examined relationships between VAQ and various
atmospheric parameters (8). They also concluded that illumination was
important to VAQ judgments, and were able to show at one site that there is
a good correlation between VAQ and the natural logarithm of (b$CAT) tlle
atmospheric scattering coefficient.
Visibility parameters used in the Latimer and Stewart studies were
simple indices that were unable to account for the unique viewing conditions
and qualities associated with various vistas. Change in illumination
manifests itself as changes in vista color and texture, while different
viewing distances require observers to "look through" a greater optical
depth that in turn reduces contrast; of form, texture, and color.
Malm et al. examined the relationship between VAQ and contrast, a
parameter that does account for variations in observer-vista distance (6,9).
They used various southwestern national park scenes under a variety of
illumination, meteorological and air quality conditions. Analysis of the
data showed that under fixed illumination and meteorological conditions,
apparent vista contrast of the most distant vista element was a good
prediction of VAQ judgments. The study also showed that changes in foreground
color (due to change in illumination), addition of clouds or snow cover
caused the VAQ ratings to be higher, but did not cause the sensitivity of
VAQ to change in vista contrast to change, i.e., slope of VAQ versus apparent
vista contrast remained the same but is translated upward as foreground
color increased. The study also presented a model of human perception of
VAQ. The model suggests that ratings of VAQ are proportional to the sum of
the fraction of each scenic element subtended by various landscape features
multiplied by the atmospheric transmittance between that landscape feature
and observer. It was shown that when a single landscape feature that is
void of textural detail dominates the perceived change in visual air quality,
the model predicts a linear relationship between VAQ and the apparent
contrast of the landscape feature (contrast of form).
Although SBE and perceived VAQ (PVAQ) appear to be good indicators of
image quality for uniform haze, little work- has been done to examine
which of many psychophysical indicators best described perceptions of
layered haze much less, the relationship between those indicators and air
quality. .
2.2 PschyoTogical Values in Natural Settings
Once the relationships between atmospheric, haze and psychophysical
indices such as JND.'s, SBE.'s and PVAQ are understood, it is still necessary to
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
understand the value, whether it be psychological, sociological or economic,
that visitors place on a perceived change in visual air quality.
A method which has been used to assess the value of good visual air
quality is recreation demand theory. Briefly, this theory professes that the
demand for, or value placed on, a recreational experience is a function of
demand for activities, for attributes, and for certain psychological satisfac-
tions.(10) The satisfactions which flow from an experience are highly
dependent on the activities and the presence and quality of the attributes
supplied. In the context of class I areas, this means that if good visual air
quality is an important attribute, and if existing visual air quality does not
match what is demanded, the visitor will be less than fully satisfied with the
experience.
Research conducted at Grand Canyon and Mesa Verde National Parks assessed
the relative importance of visibility related attributes in comparison with
other park attributes both within and between parks. It might be hypothesized
that visibility attributes are of lesser importance at a park with a cultural
and historical perspective such as Mesa Verde than at a nature-oriented park
like Grand Canyon.
Cluster analysis ratings of relative importance of 24 attributes revealed
that individual attributes collapsed into five clusters of attributes for each
park. A "naturalness" cluster, which included clean, clear air was the most .,
important attribute cluster at both parks. A view-related cluster was the t
second most important cluster at Grand Canyon and the third most important at.
Mesa Verde.(11)
The fact that the naturalness attribute cluster was the most important
cluster at both parks which have different underlying themes has led to a
hypothesis-that "naturalness" might be a universal trait of.National Parks.
In the context of this study, visitors felt that "naturalness" was more
important to their recreational experience than the specific attributes for
which the parks are famous. The significant part of this finding, in terms of
visibility protection, is that these visitors felt that clean, clear air was a
part of that naturalness.
Cluster analysis of importance ratings of certain satisfactions which
might accompany the recreational experience being sought at these parks showed
that an "existence" cluster was rated as most important-at both parks. This
cluster .is comprised of two satisfaction items, one of which pertains to
knowing that the park resources are being protected. When combined with
previous results, this indicates .that visitors derive satisfaction from
knowing that park resources are being protected and they feel that the
naturalness of the park (which includes clean, clear air) is the most
important of those resources.(12)
Before reduced visibility can have an impact on the overall recreational
experience being sought, it must be salient to park visitors. That is, they
must not only be able to perceive it, but they must also be aware of it as it
occurs in the context of the overall recreational setting. Regression
analysis between visitor awareness of haze at the Grand Canyon and a physical
measure of the actual visibility at the time of the interview showed that
-------
changes in visitor awareness of haze corresponded to actual changes in physi-
cal air quality. As visibility improved, visitors said they thought it was
less hazy, and when visibility became worse, they thought it was more hazy.
Furthermore, at Grand Canyon 84% of the respondents indicated they saw haze,
and 82% of the respondents at Mesa Verde said it was hazy. Of special in-
terest is that 80% of the Grand Canyon respondents and 83% of the Mesa Verde
respondents who said they saw haze thought the haze occurred naturally, yet
other research has shown that at least 50% of the haze is man-made.(12)
Having established that "naturalness", including clean, clear air, is
most important to park visitors, and that visitors are aware of haze, it
remains to be determined whether degraded visual air quality, or "hazy" air,
affects the visitor's ability to enjoy park resources. Analysis was also
performed to examine relationships between visitor ratings of 1) awareness of
haze, 2) rating of enjoyment of the view, 3) the negative impact haze had on
overall park enjoyment, and 4) satisfaction with visibility. Results of the
analysis showed that as visitors say they become more aware of haze, they also
say their enjoyment with the view decreases, their overall park enjoyment
decreases, and their satisfaction with visibility decreases.(13)
"7 In summary, these studies show that naturalness (which includes clean,
clear air) is the most important type of attribute, knowing that park re-
sources and values are being protected is part of the most important satisfac-
tion cluster associated with their recreational experience, that at least 80%
of the respondents said they were aware of haze, and that at least 80% of
those aware of haze thought it was natural. Other research, however, has
shown that 50% or more of the haze at these areas is not natural. Put more
succinctly, visitors to these parks think the visibility is being protected
from man-caused sources of degradation, when the fact is that at least one-
na'lf. of the haze they see is man-caused. Given the above, plus the fact that
more than 70% of visitors at both parks rated the views as excellent or
perfect, one could speculate that the high view ratings may be due, in part,
to a largely uninformed public enjoying the parks with the false impression
that the park resources are being protected.
Another approach to quantification of visitor experience hypothesizes
that the visitor will choose between various park attributes, including
viewing of scenic vistas, in such as way as to yield the highest state of
psychological and physical well-being. It is assumed that the park visitor
will try to maximize personal enjoyment of park resources and will allocate
accordingly the amount of time spent on various park activities. It is
assumed that the visitor would not rationally choose any activity that results
in increased psychological stress or greater discomfort.
The amount of time that park visitors would allocate to various
activities, specifically viewing scenic vistas as a function of air pollution,
can be determined through the use of a contingent ranking technique that
assesses how much time a visitor is willing to give up in exchange for in-
creases in visibility. The approach uses an ordered logit probability-method
similar to that used to describe consumer demand. It also inherently takes
into account the frequency, spatial extent, intensity, and duration of
visibility irapairment.(9)
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Results show that, in some cases, visitors place a high value on small
changes .in visual air quality. In general, the studies show that
t"***s
To Vistas containing medium distance (10-30km) highly colored landscape
v~ features are very sensitive to small increases in air pollution
o Vistas which lack color and texture (shadowed) but with scenic
elements at many varied distances show little sensitivity to changes
in air quality
o Visitors were most sensitive to changes in air quality during clean,
clear winter months.
/
o Vistas with distant large dominant landscape features are most
sensitive to increases in air pollution
o The value that visitors in one study placed on vistas is least during
summer months. At Grand Canyon National Park visitors would spend 5.4
additional hours of driving time to see the vistas during winter
months as opposed to the summer season.
o Observer sensitivity and values could not be significantly related to
socio-economic variables such as household income, education level,
residence, age, etc.
REFERENCES FOR SECTION 2
1. Taylow, J.H. "Use of visual performance data in visibility prediction,"
Appl. Opt. 3(5):562(1964).
2. Campbell, F. and Lamberti, "Contrast and spatial frequency," Scienific
American, 231(106) (1974).
3. Henry, R.C., "The human observer and visibility - Modern psychophysics
applied to visibility degradation." In: View on Visibility - Regulatory
and Scientific, Air Pollut. Control Assoc., Pittsburgh, PA (1979).
4. Carlson C.R. and R.W. Cohen, Image Descriptors for Dispays: Visibility of
Displayed Information. RCA Laboratories, Princeton, NO, July 1978.
5. Faugeras, O.D., Digital Color Image Processing within the Framework of a
Hunan Visual Model, IEEE Trans. Acoust., Speech, Signal Processing ASSP-
27, pp. 3oO-393 (1979).
6. Malm, W.C., K. Kelly, J. Molenar and T, Daniel, "Human perception of
visual air quality (uniform haze), Atmos. Environ., 15:1875-1890 (1981).'
7. Latimer, D.A., T.C. Daniel, H. Hogo and O.H. Hern, "Effects of air
quality and visibility on human perception of scenic beauty." Final
report to .the American Petroleum Institute, Systems Applications, Inc.,
San Rafael, CA (1982).
-------
8
8. Middleton, P., T.R. Stewart and R.C. Dennis, "Modeling human judgement of
urban visual air quality," Atmos. Environ., 17, 1015-1021 (1983).
9. Malm, W.C., P. Bell, 6.E. McGlothin, "Field Testing a methodology for
assessing the importance of good visual air quality," Proceedings of the
77th Annual Meeting and1Exhibition of the Air Pollution Control
Association, June 24-29, 1984, San Francisco, CA.
10. Driver, B.L. and P.J. Brown, "The opportunity spectrum concept and
behavioral information in outdoor recreation resource supply inventories:
A rationale." In: Proceedings of the National Workshop on Integrated
Inventories of the Renewable Natural Resources. January 8-12, 1978,
Tucson, AZ.
11. Ross, D.M., R.J. Loocnis, "Assessing the value of a visit to a National
Park." Report No. 7 for a project entitled "Assessment of Visibility
Impairment on Visitor Enjoyment and Utilization of Park Resources,"
submitted to'Air and Water Quality Division, National Park Service,
Washington, D.C., 1985.
12. Ross, D.M., G. Haas, R.J. Loomis, W.C. Malm, "Visibility impairment and
visitor enjoyment" Proceedings of the 77th Annual Meeting and Exhibition
of the Air Pollution Control Association, June 24-29, 1984, San Francisco,
CA.
13. Ross, D.M., W.C. Malm, "An assessment of visitor enjoyment and values at
Grand Canyon and Mesa Verde National Parks." Proceedings of the Air
Pollution Effects on Parks and Wilderness Areas, May 20-23, 1984, Mesa
. Verde National Park, Colorado.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.0 ECONOMIC BENEFITS OF VISUAL AIR QUALITY
3.1. Introduction
Air quality affects both our health and well-being.
Economists have attempted to measure the value of air quality,
including its aesthetic visual component, by examining data on
residential property values and by asking valuation questions
directly in a hypothetical or 'what if1 context. . Visibility
values have been quantified for households engaged in day to day
activities in an urban context and in the context of household
visits to recreation (Class I) areas, where the clarity of scenic
vistas may be an important element to the overall experience.
This section is divided into five parts: Part 3.2 describes
benefit measures and the methods used by economists to quantify
visual air quality benefits. Part 3.3 reviews selected studies
of the benefits of visibility in urban areas. Part 3.4 reviews
selected studies of visibility benefits in a recreational context
for visitors to Class I areas. Finally, Part 3.5 summarizes the
estimated benefits of visual air quality in urban and Class I areas
and develops a range of dollar benefits that are appropriate for
assessing benefits of. policies to reduce emissions.
3.2 Quantification of Visual AirQuality Benefits
Economists generally perceive most things as having value. *
A measure of value is the well-being, or what economists call
utility, derived from the consumption of a good or service. The
utility derived from visibility is difficult to measure because,
like other public goods, it is not traded on'a market and has no
explicit price. Economic benefit measures attempt to put a
monetary value on changes in an individual's utility by measuring
the change in income that would yield the same change in well-
being as the change in visual air quality.
The benefits'of a change in visual air quality can then be
quantified by methods -that attempt to elicit a maximum willingness
to pay (WTP) an amount (forego income) or willingness "to accept
(WTA) compensation (receive income) to secure or avoid a change
in visual air quality.* Measuring WTP requires a reference or
base level of utility, and either a payment or compensation measure
can be used. In evaluating visual air quality the exist ing.level
is usually taken as the reference point so that improvements are
measured in terms of willingness to pay and degradations in terms
of willingness to be compensated.
*In economic theory benefits are measured by consumer's surp-lus,
which is the difference between maximum WTP and expenditures.
However, since visibility is a public good supplied at zero price,
expenditures are zero and maximum willingness to pay is equal to
consumer's surplus .
-------
10
3.2.1 Visibility Values
There are three types of visual air quality values commonly
addressed: activity values, option values, and existence values.
Activity value (also called user value) is the value associated
with the present or future use or enjoyment of the visibility at
a site, such as a residence or scenic area. The value of visi-
bility in contributing to the enjoyment of some activity is
reflected in market decisions, such as choice of housing location,
and can be approached through the use of both market and non-market
methods.
Option value is the value of an option to preserve some level
of air quality level at a site in anticipation of potential future
use or activity. This value is separate from the value of future
use. WTP to insure a level of air quality for future use represents
an option price that includes both a present (discounted) value of
future use plus an option value. Freeman (1983) argues that for
most utility functions the'option value is unlikely to exceed 5
percent of activity values. Option values for visibility can be
measured only by the use of non-market survey methods.
Some individuals may value the preservation or existence of
a resource, such as visual air quality, even though they do not
intend to use the resource themselves or to participate in an
activity at the site. Such non-use values may be tied to the
philanthropic goal of preserving the quality of the experience
for future generations. Since by definition non-use values are
not linked to market decisions, these values can only be measured
by surveys of consumer behavior.
3.2.2 Methods to Measure .Visibility Values
Data for estimating demand for changes in visual air quality
have come from both consumer surveys and market transactions.
Because air quality affects both health and welfare only surveys
can separate aesthetic values, such as visibility, from other
effects on human health, materials, and crops. Nevertheless,
market data from housing location or property value studies have
been used to attempt to estimate the total air quality value,
including the value of visibility.
3.2.2.1 Contingent Valuation Survey Approaches
Most visibility benefit studies have used a contingent
valuation (CV) approach. CV methods require survey respondents to
reveal market information based on a hypothetical market context.
The di_rect_ questi on method asks respondents the maximum they would
be will ing to pay Tor the minimum they would be willing to be
compensated) to change the level of an environmental good or
-------
I
I
I
I
I
J+
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
amenity. The contingent ranking method asks respondents to rank
order alternatives thatincludedifferent levels of environmental
amenities and changes in household income. Estimates ..of the
contribution of each attribute to the ranking can then be used to
reveal a trade-off between income and the environmental amenity.
3.2.2.1.1 Direct Question Method
Conceptually, the direct question WTP approach is quite simple.
A WTP or WTA value is obtained by asking.the maximum value that
respondents are willing to pay or the minimum amount they are
willing to accept. This value is then used to estimate a .demand
curve as a function of individual or household characterstics, such
as income, age, location (distance to site), etc.
A number of oral and written direct question formats have been
compared, and the WTP format appears to affect the reliability and
magnitude of benefits (R'andall, et al.f 1981). The oral bidding
format uses an iterative technique that increases or decreases the
asking price until a maximum WTP or minimum WTA is obtained ana the
respondent refuses to bid. Oral bidding has been subject to biases
attributable to different starting bid levels and to the -skills of
different interviewers. The open end bid, administered either*in
oral or written form, asks for a single maximum or minimum bid.
This format has been especially subject to protest behavior in-the
form of zero bids and refusals to bid. The payment card and
checklist range formats ask the respondent to indicate a particular
value or value range that approximates a maximum (or minimum) WTP. -
Randall et al., (1981) found that of the six formats tested the
payment card elicited the highest average bids due to fewer zero
bids while the checklist range yielded the lowest average bids.
More recent experiments with contingent valuation formats have
revealed that initial WTP bids do not yield a true maximum, and
rebid mechanisms that accept or reject the maximum b-id ha^ve been
found to.yield significant increases in WTP (Schulze, et al., 1984).
3.2.2.1.2 Contingent Ranking Method
In the contingent ranking method the benefit trade-off is
measured by the change in household income that compensates for a
change, in vi'sual air quality from some base condition, -such as the
average or median visual ranye. Respondents are given a set of
alternatives - each 'specifies a visual air quality level, income
gain or loss, and level of other attributes. The ranking of these
altern-ati ves from most preferred to least preferred provides a set
of data from which the relative weights on visual air quality and
household income can be quantified by a probability choice
estimator.* Numerically, if improving 'average urban visual range
"A comprehensive mathematical treatment is given in Beggs, et al., 1981'
-------
I
12
I
from 10 to 20 miles is weighted twice as much as a $100 increase
in household costs (decrea
improvement is worth $200.
in household costs (decrease in household income), then that I
This contingent ranking method avoids many of the potential
biases attributable to interviewers and the sequencing of bids. |
It also may be more resistant to strategic biases, since it may
be less obvious how to manipulate the result. Survey design, _
however, requires considerable care to develop cost trade-offs I
that do not give rise to perfect lexicographic ordering (ranking
solely on the basis of one attribute), and there is clearly a
limit on the complexity of the trade-offs that respondents can
manage. |
The ranking method appears to provide consistent and _
statistically reliable estimates in comparisons of four or five I
alternatives composed of an environmental quality and a cost
attribute. However, in more complex experiments with eight or nine
alternatives and three or more attributes the ranking results have
been imprecise and difficult to interpret. I
3.2.2.1.3 Comparison of Contingent Valuation Methods
In general, the contingent ranking method seems to give higher,
estimates of benefit trade-offs than the direct question methods. I
For example, Desvouges, et al . (1983) found that for a change in
water quality from 'boatable to swimmable1 the ordered logit
estimator yielded a trade-off value of about $ 50, and the direct
question approach yielded a value that ranged from about $ 4-31, |
depending on the format. In a study of benefits of avoiding odors
from diesel vehicles Rae (preliminary draft, 1984) also found that _
the ranking method generated larger benefits. To reduce the number
of contacts with a strong diesel (truck) odor by one per week the
average household was willing to pay about $ 15-18 per year in the
ranking study and about $ 2-4 per year in the direct question study.
It has been suggested that this difference in results is because |
the ranking method yields a.true reservation price while the direct
question method does not. This hypothesis remains to be proven, _
but there does appear to be an important difference in the magnitude
of benefits between the two contingent valuation methods. *
3.2.2.1.4 Issues in Contingent Valuation Studies of Visual Air Qualitji
The use of contingent valuation in estimating benefits of
visual air quality necessitates developing a survey instrument. All I
surveys require careful attention to survey design to avoid biasing
the results due to different starting points {Boyle et.al., 1985),
payment vehicles (Brookshire, et al., 1970), and information levels
(Rowe and Chestnut, 1983). But of greatest concern in contingent '
valuation surveys is potential for hypothetical bias which occurs if
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
* «
(*
13
Individuals respond differently to a survey than they would in an
actual market choice situation. Results obtained by Bi'shop, et al .
(1984) in comparing a simulated market experiment that used real
cash payments with a contingent valuation survey suggest some upward.
bias in WTP survey bi'ds. The average respondent indicated a WTP of
$ 32-43 to obtain a deer hunting permit in a hypothetical survey,
but was willing to pay only $ 19-25 when allowed to actually
purchase a hunting permit for cash. Similarly, the mean WTA (sell)
in a hypothetical survey was about $ 830 compared to a mean of about
$ 1200 for accepting actual cash offers. The hypothesis that mean
WTP survey bids were equivalent to cash transactions was rejected at
the .10 level of significance in all but the sealed bid experiment.
In a follow-up experiment (Bishop, et al., forthcoming 198b) using
an accept or reject format,' the researchers found a mean WTP survey
bid of $ 25 and an actual mean payment of $ 31. In this experiment
the null hypothesis of no significant difference was- accepted. A
tentative explanation for tne different results may be due to the
larger uncertainty in th"'e first experiment. Only four permits were
available to be bought and sold, and the authors suggest that this
may have caused sellers to overstate WTA in the hopes of a windfall
and buyers to understate bids in hopes of a bargain. The more
recent experiment was conducted with many more permits, and
respondents were confronted with take it or leave it offers so that
uncertainty was not a major problem. Thus, there is mixed evidence
on whether hypothetical bias is an important problem in CV survey
research. More apparent is the large difference in magnitudes
between willingness to pay and willingness to accept compensation,
and until markets for environmental goods are better understood the
income-limited WTP measure appears less prone to error.
Valuing visibility requires the use of some visual 'materials
and descriptors in order to illustrate the changes being considered.
Severa.l studies have referred to a sli.de or photograph as an average,
median, or typical condition and asked respondents to value large
ch.anyes in visual range -- on the order of 50 to 200 percent.
Visibility, however, is a distribution of daily, even hourly, events,
and describing visibility by reference to a mean or median condition
may oversimplify the problem, especially in urban areas where
respondents are familiar with the range of visibility cond-itions.
Two studies (Loehman, et al. 1981 and Rae/CKA, EPRI draft, 1984)
have described these distributions in terms of the number or per-
centaye of days with different visibi1ity levels and found that
respondents were able to value relatively small changes in the
number of days of clear or hazy visibility that were equivalent to
a change in mean visual range of less than 15 percent.
Contingent valuation surveys require a payment vehicle, such
as a utility bill or entrance fee, and the time period of these
hypothetical payments may affect the results. Payments can be
specified on a one time, annual,- or monthly basis. Advertisers
know that many buyers are more eas'ily persuaded by "low montnly
payments" of S 10 compared to a larger lump sum or annual payment
-------
14
of $ 120. There may be a hidden upward bias in payment vehicles
that require larger numbers of smaller payments. On the other
hand, respondents with cash flow constraints may bid less when
asked their WTP in large annual or lump sum payments.
A final concern is that surveys that value one environmental
good, say visual air quality, may be capturing more generalized
values for all air quality, or even all environmental quality. Two
studies evaluated the effects of presenting visibility in the
context of other competing claims on scarce household resources.
Randall, et al. (1981) found that the value of preserving visibility
at the Grand Canyon as a single environmental good was significantly
larger than when visibility at the Grand Canyon was included with
visibility improvement in the Chicago area and the eastern United
States. When valued as a single good preserving visibility at Grand
Canyon was worth about $ 90 per household per year compared to $ 15
per household per year when valued as the third component of a three
good visibility package. Similarly, Rae/CRA (EPRI draft, 1984)
found that preserving visual air quality in the Smokies was worth an
average of about $ 60 per .househol d annually when presented as
single good, but average household WTP declined to about $ 20 when
presented in a competing context with 10 other 'good causes' Rae
attributed much of this reduced WTP to a change in value for the
non-use component of total benefits, but more research is needed to
evaluate the effects of benefit estimation in a multiple good
context.
3.2.2.2 Hedonic Property Value Approach
Property value studies have been generally preferred by
economists for valuing environmental goods because they are based
on real market preferences. Despite their limitations in measuring
the aesthetic or visual air quality component, property value studies
do provide a measure of willingness to pay for a given change in
overall air quality and therefore provide an upper bound to estimates
of visual air quality benefits from contingent valuation studies.
Hedonic price theory, as developed by Griliches (1971) and
Rosen (1974), suggests that consumers select the particular bundle
of attributes they desire when purchasing a differentiated market
good, such as housing and that consumers derive utility from these
attributes. The amounts of the different attributes in a composite
good determine the prices that consumers are willing to pay for that
good. The value of each attribute can then be estimated by regres-
sing property value against housing characteristics to obtain a
hedonic or implicit price function. The hedonic price function
enables the analyst statistically to hold all attributes constant
and to answer the question of how differences in air quality or
other attributes would affect prices of property if each residence
were identical in every other respect.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
15
The regression coefficients represent the household's marginal
implicit price, MIP, or willingness to pay for a unit change in air
quality. However, public policy questions typically require estima-
tion of the household's willingness to pay for non-marginal changes
in air quality. This requires a second step, in the hedonic approach,
which is to estimate the demand or willingness to pay (WTP) function
for air quality at the residence. This is the relationship between
the marginal implicit price of air quality and the level' of air
quality at which the household will choose to locate. It must also
account for the effects of household income, household size, and
other household characteristics and property attributes that
influence the household's demand for air quality.
The application of the hedonic property value technique has
encountered a number of practical and theoretical problems. Among
the practical problems are the requirements for detailed data on
large numbers of properties and the costs of collecting these data.
In. addition, three important theoretical problems have emerged in
applying the hedonic method to property values: instability of
results across specifications, multicollinearity among variables,
and endogeneity of the marginal implicit price of air quality. In
empirical applications minor changes in variables specified or
number of observations in the data set have caused key air quality
variables to become insignificant (Brookshire et al. 1979).
Hulticol1inearity among variables has also made it quite difficult
to separate the effects of different pollutants from each other and
from other variables that affect property values. Principal
components analysis has been used to separate correlated variables,
but the resulting coefficients are not easily interpreted (Thayer
and Trijoriis, 1984).' Finally, the MIP function appears to be
endogenous to the model, and thus estimation is possible only by
assuming an underlying functional form (Mendelsohn, 1980).
3.3 Economic Benefit Studies of Visual Air Quality in Urban Areas
As of January 1985 there exist six studies of visibility
benefits in urban areas, although three remain in draft form. Three
studies have quantified the benefits of improving visibility in
eastern cities using a contingent valuation .approach. Three other
studies in California urban areas provide estimates of visibility
and air quality benefits using both the contingent valuation a.nd
property value methods. '-Below we review these studies to estimate
a value for visibility improvement in the eastern United States.
-------
16
3.3.1 Chicago Urban Visibility. Study*
The Chicayo visibility experiment (Randall, et al., 1981)
evaluated household willingness to pay for visibility improvement
and willingness to pay to avoid visibility degradation in the
Chicago urban area. A contingent valuation survey was designed to
compare six oral and written direct question formats. The survey
was conducted in the summer of 1981, and usable information was
obtained from 273 repondents. A composite set of photographs of
three Chicago urban vistas was used to portray poor, median, and
good visibility conditions. A non-specific payment vehicle was
then used to elicit a maximum monthly willingness to pay (WTP) per
household to achieve an improvement in visual air quality or to
avoid a reduction in visibility. The five scenarios are listed
below:
Chicago area
Chicago area
Chicago area
Chicago plus eastern
Chicago plus eastern
vi si bi1i ty at Grand
Tne six different formats generated a wide range of benefit
estimates, as shown in Table 1. Formats using a payment card
generated the fewest zero bids and protest bids and the largest
do lar benefits. The written checklist format yielded less than
(rll Jh\m?9n11t!!d! °f benefits obtained by the expanded payment card
((.£) that included information on typical expenditures for public
and private goods. For the three urban visibility increments the
results can be summarized as follows:
Measure
WTP-degradation
WTP-i mprovement
WTP-improvement
a.
b.
c .
d.
e .
WTP
WTP
WTP
WTP
U.S.
WTP
U.S.
(degradation )
(1
(i
(i
(i
P
Canyon
mprovement
mprovement
mprovement
mprovement
)'
)
)
lus preventi
**
9
9
9
9
9
n,g
to
to
to
to
to
a
4 mi 1
18
30
18
18
mi
mi
mi
mi
decli
es -
1
1
1
1
es
es
es
es
ne i
<>
n
Visibility
(mi les )
9 to 4
9 to 18
9 to 30
WTP
($/year)
$ 121-283
$ 109-324
$ 152-456
WTP
U/mile)
$ 24-57
$ 12-36
$ 7-22
Note: These bids
dol1ars ).
are all based on solid core data (1331
*This study is included as an appendix to the later Six Cities study
Results of this scenario will be discussed later in Section 4, Vis-
ibility Benefits in Class I Areas.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
17
The average household willingness to pay for all respondents
(all six formats) to avoid a 5 mile reduction in visibility from 9
to 4 miles was $ 227 using only the "solid core" sample of 221
respondents. Improvement from "9 to 18 miles in visual range yielded
an annual average bid of $ 218 per household, and improvement from 9
to 3U miles resulted in an annual average bid of $ 271 per
household.
In evaluating these benefit results it should be remembered
that visual air quality was the only component of air quality
valued, and it is possible that these visibility benefits include
some measure of health or other air quality concerns. Nor was there
a test for the effects.of other competing claims for scarce household
resources on household WTP for urban visibi1ity. The authors did
test for this effect in comparing values for preserving visibility
in Grand Canyon from a previous CV study, and found a substantial
decline in household -WTP. It should also be noted that the authors
did not test for order effects in the bids, which proceeded in
order of increasing visual range. Evidence from other studies
(Rowe and Chestnut, 1983) suggests .that benefit magnitudes may be
affected by the'starting point and the direction of bidding.
Despite these cri
drawn from this study.
significant amounts of
from 9 to 18 miles and
is very much dependent
to avoid a reduction i
WTP to improve visibil
narios to improve visi
a nonlinear WTP curve
range increases.
tici sms,
First,
annual
from 9
several important conclusions can be
respondents were willing to pay
income to improve typical visibility
to 30 miles. Second, the WTP benefit
on the choice of CV format. Third, the WTP
n visibility is significantly larger than the
ity. The estimates of WTP for the two sce-
bility appear fairly consistent and suggest
with decreasing marginal benefits as visual
3.3.2 Six Cities Study
The Six Cities study (Tolley, et al., EPA draft, 1984, 1985)
was a major effort to apply the results of .the Chicago urban
visibility study to six other eastern cities. The WTP survey
utilized an expanded payment card.and a rebid mechanism to value
changes in visual air quality. A total of 792 respondents were
surveyed in 1982 in six eastern cities -- Washington, D.C.,.Atlanta
Mobile, Boston, Miami, and Cincinnati,-- and 528 complete surveys
were used in the analysis.
Visual air quality was portrayed with sets of photographs of
three different urban scenes that showed poor, median, and good
visibility conditions. The same set. of generic urban, visibility
photographs was shown to all respondents; regardless of city.
Monthly bids were elicited for five scenarios:
-------
a,
b,
c,
d,
e
WTP
WTP
WTP
WTP
WTP
States
(degradati on)
{improvement)
(i mprovement)
(Improvement)
(improvement)
18
10 to 5 miles
10 to 20 miles
10 to 30 miles
10 miles eastern United States
10 miles eastern and western United
Visibility in the east was portrayed by photographs of poor, median,
and good visibility in the Smoky Mountains, and visibility in the
west was portrayed by photographs of poor, median, and good vis-
ibility in the Grand Canyon.
The average household bids varied greatly across the six cities,
as shown in Table 2. Average annual household bids to avoid a five
mile reduction in typical visibility ranged from $ 57 in Cincinnati
to $ 232 in Washington, D.C. To improve from 10 to 20 miles average
bids ranged from $ 57 in Cincinnati to $ 238 in Washington, D.C. A
larger improvement from 10 to 30 miles elicited average bids that
ranged from $ 64 in Cincinnati to $ 303 in Washington, O.C. For the
eastern and total U.S. scenarios average annual household bids ranged
from $ 74-358 and $ 80-422, respectively.
A number of linear and non-linear specifications relating visual
range to WTP values were tested, and visual range was found to be
highly significant in explaining differences in WTP. The linear
specification found that for a 1 mile improvement in urban visual
range the average household was willing to pay $ 14.20 per year. The
researchers found that the best non-linear specification was a
negative exponential of the form
BIDj = A[l-exp (-B)(VR) ]
where VR is visual range resca
and A and B are parameters to
through the origin, increases
as VR increases. The paramete
maximum WTP is approached and
mic characteristics of respond
parameter, A, is a rotational
determines the position of the
bid. In this case the effects
visibility are included as ind
the rotational constant.*
led so that the existing level is zero,
be estimated. The function passes
monotonically, and implies reduced WTP
r, B, determines how rapidly the
is estimated as the sum of socio-econo-
ents and city-specific effects. The
constant, or shift parameter, that
function and the magnitude of the
of non-local {eastern and western)
ependent (dummy variable) terms in
Empirical estimation found that differences in socio-economic
and city effects explained away much of the difference in the raw WTP
bids. For a 1 mile improvement in local visual 'range from the base
I
I
I
I
I
*Alternatively, dummy variables for non-local visibility could
have been specified independent of the rotational constant. Tests
of significance could then be applied to determine the appropriateness
of the specifications.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
19
(10 miles) average household'WTP varied from $ 12.78 per year in
Cincinnati to $ 15.9U per year in Washington, U.C. Similarly,
annual household WTP for a 10 mile improvement in local visual
range varied from about $ 115 in Cincinnati to $ 139 in Washington,
D.C. Non-local household WTP bids varied from $ 68 to 82 per year
for a 10 mile improvement in eastern visual range and about $ 32
to 39 for a 10 mile improvement in western visual range.
-The reliability of the data depends on the adequacy of survey
procedures. One concern is that the photographs used were'of
generic urban scenes, and the failure to provide site-specific
photographs may be a source of bias (probably downwards) in esti-
mates of WTP for improvements in visual range in each city.
Similarly, the photographs of the Great Smokies and Grand Canyon
do not accurately portray typical vistas in the east and the west,
and as a consequence the values for visibility improvement in the
east and west may be biased upwards. In addition, the appropriate-
ness of including non-local visibility effects, much of which are
probably non-use values, in the shift parameter, remains to be
demonstrated. - ;
These surveys in six eastern cities provide the largest amount
of data on household WTP for improved visual air quality. In
addition, tne authors have succeeded in demonstrating that large
variations in the raw data bids are mostly explained by socio-
economic and city-specific differences so that the resulting effect
of changes in visual range on WTP bids is fairly consistent across
the six city sample. Finally, the non-linear specificatTon captures
the effect of diminishing WTP for increases in visual range.
3.3.3 Cincinnati Visibility Study
The Electric Power Research Institute (EPRI) commissioned a
benefits study of visual air quality in Cincinnati using the con-
tingent ranking and the direct question methodologies. (Rae/
Charles River Associates, 1984). A survey pre-test was conducted
on 147 respondents in March 1982, and a final survey collected
data from 314 respondents in June 1982. The report has undergone
several drafts to refine the estimates and evaluate the methodology,
the most recent being August 1984.
The contingent valuation survey
choice of alternatives that included
fuel costs and visual air pollution.
linked with changes.in home heating
derived synthetic gas and oil woul.d
home heating systems. Lower heating
that contained pollutants, especiall
were associated with clean -fuel. A
a WTP measure for an improvement in
saving) in heating cost represented
limited by the annual cost of fuel)
presented respondents witn a
different levels of household
Changes in visibility were
costs'by suggesting that coal-
soon be available for use in
costs were associated with fuel
y sulfur. Higher heating costs
heating cost increase represented
visual range, and a reduction (or
a WTA measure (with compensation
for a redaction in visual range.
-------
20
WTP
WTA
Pre-test
(S/mile)
$ 22-32
$ 81-132
Final Survey
($/raile)
$ 71-108
$ 103-154
I
I
Visual air quality was portrayed by slides of Cincinnati that
depicted haze, median, and clear conditions. The pre-test used an I
older set of 35 mm. slides with visual ranges of about 3, 8, and 25
miles. An updated set of 50 mm. slides with visual ranges of about
3, 11.5, and 16.5 miles was used in the final survey. The three I
visibility slides were projected simultaneously on one screen.
Changes in visual range were presented using two different
descriptors: one where visibility conditions were stated to occur I
'most of the time' and a second where information was provided to
indicate the annual percentage of days typical of each of the three
visibility conditions. The results of the survey pre-test and the
final survey are compared in Table 3. |
Using the first descriptor, an improvement in visual range from m
median to clear 'most of the time1 yielded an annual household WTP
(increase in fuel costs) of $ 381-544 in the pre-test and $ 343-516
in the final survey. For a degradation from median to haze 'most of
the time' WTA compensati'on (saving in fuel costs) amounted to $
387-635 in the pre-test and $ 884-1324 in the final survey. The |
range of benefits is attributable to different specifications. On a
per mile basis the WTP estimate is higher in the final survey while _
the WTA measure yields fairly similar per mile estimates in botn
surveys. . ^
Measure
I
I
Note: These estimates are in 1982 dollars.
Using the probability distribution descriptor, respondents in I
the pre-test evaluated a simple two attribute trade-off between
visibility and fuel costs and obtained fairly precise estimates.
Annual household WTP amounted to $ 217 to improve from an existing .1
distribution (27 percent clear) to a distribution with 80 percent
clear days. To improve from a distribution with 27 percent clear to
40 percent clear was worth $ 142. Annual compensation of $ 106 and $ I
268 was required to accept a. reduction in visibility to distributions
of 15 and 8 percent clear days.
T.he final survey added health effects as a third attribute. |
Health effects proved highly significant, but the reliability and
significance of the visibility attributes were eroded.* On a per
mile basis visibility benefit estimates from the pre-test were
generally less than in the final survey, where the-range is quite
wide and the results reflect considerable uncertainty.
*The best specifications evaluated the likelihood of the orderiny
on only the first three out of a total of nine alternatives ranked. _
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
21
Measure Pre-test Final Survey
WTP: Existing to Best $ 23 $ 35-102
Existing to Good $ 59 $ 36-204
WTA: Existing to Poor $ 38 $ 169-227
Existing to Worst $ 57 . $ 123-184
Note: Estimates are in 1982 dollars.
Respondents also registered a direct question bid (checklist
format) for a change from median to clear visibility 'most of the
time'. In the pre-test repondents were willing to pay an average of
about $ 85 per household per year for an improvement of about 8 to 25
miles while in the final survey they were willing to pay about $ 143
for an improvement of about 11.5 to 16.5 miles. For a one mile
improvement in visual range these benefits equal about $ 5 (pre-test)
and S 30 (final survey) 'and are considerably smaller than the benefit
estimates obtained from the ranking method.
The Cincinnati study generally confirmed that visibility is an
important component of the air quality good and that respondents were
willing to pay significant amounts of income in fuel costs to obtain
better visual air quality. A significant household WTP for visual
air quality was demonstrated in both contingent ranking and direct .
question surveys. The magnitude and significance of these benefit
estimates were affected by the use of different sets bf slides,
different visibility descriptors and the inclusion of health effects,
and some results tended to be unstable across specifications.
3.3.4 Review of California Visibility/Air Quality Studies
There have been three other major studies- that have attempted to
value changes in urban visual air quality. Two studies compared the
CV and property value approaches in the Los Angeles and San Francisco
areas. In addition, a more recent study reestimated air quality
benefits from property value data using the extinction coefficient as
a measure of air quality.
*
3.3.4.1 South Coast Air Basin Study
This study (Brookshire, et al., 1979, 1980) used data from
1977/78 property values and a 1978 direct question oral bidding
survey of residents to estimate benefits of improved air quality.
In both analyses homes were grouped into poor, fair, and good air
quality. Empirical estimates of-the hedonic price function found
that a 0.01 ppm decrease in annual average NGx was associated with
a $ 2010 per unit increase in residential property values (1977/78
dollars), and that a 1 ug/cubic meter decrease in annual average TSP
-------
22
was associated with about a $ 300 (1977/78 dollars) increase in
property values. Estimates of household willingness to pay yielded
annual (discounted) benefits of about $ 600 (1977 dollars) to improve
from poor to fair (about 20 percent) and about $ 440 (NOx) to $ 580
(TSP) to improve front fair to good (about 25 percent).*
In the contingent valuation survey a total of 345 respondents
were asked how much they would be willing to pay for air quality
improvement, including an aesthetic visibility component and a
chronic and acute health component. An iterative oral bidding
format was used to elicit willingness to pay as an extra charge on
their monthly electric bills. The authors reported an average
household WTP of $26 per month ($ 312 per year) for bids to improve
visual air quality from poor to fair (2 to 12 mile visual range) and
fair to good (12 to 28 mile visual range). Results, however, varied
with different payment vehicles, time of compliance, and order of
questions. For example, aesthetic visibility improvement accounted
for 34 percent of the total benefit or about $ 105 per year over all
bidding sequences, but it ranged from under 10 percent when valued
as the last benefit element to over 70 percent when valued as the
first. The annual benefit estimate of $ 312 per household
represents an improvement in visual range of 10-16 miles so that
benefits per mile vary from about $ 6.50 to $ 10.50 per household
(1978 dollars).
3.3.4.2 The San Francisco Bay Area (SFBA) Study
Loehman, Boldt, and Chaikin (1981) also compared_the results
of a property value study and a contingent valuation survey that
quantified benefits of air quality changes in the San Francisco Bay
area (SFBA). The property value study (1978 data) utilized house-
hold level data and found significant negative effects for increases
in ozone and a Pollution Standards Index'(PSI), a composite measure
of pollution that captures both concentration levels and frequency
of occurrence. Non-linear specifications yielded the best results,
and marginal willingness to pay was found to increase with increases
in pollution and with income. Average annual household willingness
to pay (1978 dollars) amounted to $ 45-98 to reduce ozone 30 percent
and $ 86 to reduce the PSI 30 percent. These annual average house-
hold WTP benefits, which include the effects of health, visibility,
and ottter air quality components, are considerably lower than the
estimates obtained in the SCA3 study, but this could be partially
due to the lower pollution levels in the San Franci sco .area.
*A reanalysis (Brookshire, et al. 1980) using.a slightly
reduced number of properties (634 instead of 719) and a different
form of the hedonic price function generated a similar range in
annual benefits.
I
e
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23
A contingent valuation survey of 412 respondents was conducted
in the San Francisco area in early summer 1980 and obtained WTP bids
for improved air quality using a written checklist format * This
SFBA study pioneered the use of distributions as descriptors for
both visibility and health variables The visual aesthetic com-
ponent of air quality was characterized by.the number of non-pol-
luted days (over 10 miles visual range), moderate days (6 to 10
miles visual range), and poor days (0 to 5 miles visual range).
Willingness to pay to improve from a mean visual range of 16.3 to
18.6 miles amounted.to about $ 71 per year (1980 dollars).**
Visibility accounted for about 43-64 percent of the total health
and aesthetic air quality values. For a 30 percent improvement
in air quality the CV survey yielded an annual average household
WTP of about $ 51 compared to $ 45-98 for the property value study.
However, comparison of benefits for households in each air quality
locale revealed considerable differences between the CV and property
value results.
3.3.4.3 California Property Value Visibility Study
Thayer and Trijonis (draft, 1984) used disaggregate data from
7871 homes sold in 1973/74 and 1978/79 and measures of light
extinction to assess the effect of air pollution in the- Los Angeles
and San Francisco areas. In all specifications light extinction had
a significant negative association with property value, but the size
of the coefficient varied considerably with the definition of
extinction - with and without sea haze.
Important differences were also implied in the separate WTP
functions. In the San Francisco area, where median visibility is
about 17 miles, marginal WTP was found to increase as the level of
pollution (as measured by light extinction) increased. In the more
polluted Los Angeles area, however, marginal WTP was found to
decrease with increases in pollution. A 10 percent improvement in
visibility yielded average annual household benefits (1978/79
dollars) of $ 57-153 in Los Angeles and $ 115-128 in San Francisco
with the range of estimates attributable to different forms of the
hedonic price function. In San Francisco ozone, which accounts for
most of the air pollution effects on health, is independent of
extinction, and the authors were able to estimate separate
coefficients for both air pollutant variables. " They found that a 10
percent reduction in ozone increased housing values by 1.03-1.3
percent compared to T.4-2.5 percent for a 10 percent reduction in
extinction. In San Francisco extinction accounts for mostly
aesthetic effects so that visibility accounts for up to 57-66
percent of total air quality values.
*The authors cal-1 it a payment card, but it appears equivalent
to the checklist used by Randall, et al. in Chicago.
**Mean visibility was calculated by assuming visual
8, and 20 miles for the three air quality conditions
ranges of 3,
-------
2.4
3.3.4.4 Conclusions from California Visibility Studies
The three California studies reviewed above all
there was a significant WTP for improved air quality
and San Francisco and that the aesthetic
quality accounted for a'bout one-third to
benefit. This relationship is used to compute estimates of the
visibility benefits, which are summarized in Table 4.
found that
in Los Angeles
visual component of air
two-thirds of the total
The annual household WTP benefits (1984 dollars) per mile
improvement in visual range from the California studies are not
dissimilar to the estimates obtained from the visibility studies
conducted in eastern urban areas. The contingent valuation results
range from about $ 10-16 per mile improvement in the SCAB study to
about $ 40 per mile in the SFBA study. When converted to 1984
dollars, the property value studies yield annual household
visibility benefits in a wide range of abou.t $ 23-165 per mile
improvement. The recent work by Thayer and Trijonis yields a
large annual household WTP" value for aesthetic air quality benefits
in San Francisco that ranges from about $ 112-122 per mile improve-
ment in visual range.
3.4 Value of Visual Air Quality in Class I Areas
Visibility in class I areas results from both regional haze and
discrete plumes. Early attempts to quantify the economic value of
visibility were in response to aesthetic and environmental insults
(plumes) attributable primarily to existing and proposed power plants
(see Rowe and Chestnut, 1983). More recent economic benefit studies
have focused on the problems of regional haze.
Several studies have employed contingent valuation surveys t
quantify the value of visual air quality in Class I areas. Schul
et al. (September 1983) conducted a-contingent valuation oral
bidding survey at four urban locations to measure willingness to
for visibility at Grand Canyon and other Southwest Parks.
MacFarland, Malm, and Molenar (1982) used a contingent valuation
oral bidding format to quantify user (visitor) values of improved
visibility at Grand Ca.nyon, Mesa Verde, and Zion National Parks.
Rae (1982a,b) used contingent ranking and direct question written
checklist methods to'measure visi tor wi11ingness to pay for improved
visibility at Mesa Verde.'and Great Smoky Mountain National Parks.
In addition, the Six Cities study, which was discussed in Section 3,
evaluated WTP bids for improved visual air quality in both eastern
and western areas of the United States. A detailed review of these
studies is presented in the following sections.
3.4.1 Grand Canyon/Southwest Parks Study
The authors (Schulze, et al., 1983) designed and implemented a
WTP survey with an oral bidding format to measure user and non-user
to
ze,
pay
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
t:
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
25
benefits of visual air quality. Regional visibility at Grand Canyon
was portrayed by a panel of photographs that included three vistas
and five different air quality conditions. Similarly, visibility at
Southwest Parks was portrayed by a regional composite that included
vistas from Grand Canyo.n, Mesa Verde and Zion National Parks, each
with five different air quality conditions. These photographs were
evaluated in terms of frequency of occurrence and grouped from A
(about 7th percentile) to C (median) to E (99th-percent!1e). Bids
to improve visibility for users were
increase in entrance fee ($ 2.00 per
to preserve visibility were elicited
monthly electric bills.
elicited in terms of an
vehicle), and bids by non-Users
in terms of an increase in
A total of 450 questionnaires were completed in 1980 in
Albuquerque (115), Los Angeles (127), Denver (110), and Chicago
(98). Respondents who indicated past or expected future visits to
Grand Canyon numbered 166 and were categorized as users. User WTP
for improved visibility 'on the day of a vis.it ranged'from $ 1.68 per
household visit for an improvement from A to B (about 78.to 105 mile
visual range)* to $ 5.15 per household visit for an improvement from
A to E (about 78 to 258 miles visual range), as shown in Table 5.
For the entire Southwest Region users were willing to pay $ 4.29 per
household visit to avoid a reduction from C to B (about 120 to 99
miles visual range). '
All 450 respondents were asked their WTP to avoid a
deterioration in visibility from C to B both at Grand Canyon and for
all Southwest Parks. The average respondent household was willing
to pay an additional $ 5.38 per month in electricity costs to avoid
a reduction in visibility from about 125 to 105 miles at Grand
Canyon and an additional increment of $ 4.58 per month to avoid a
similar reduction in visibility (about 120 to 99 miles) at all
Southwest Parks.
Non-use values are quite large and overwhelm user values in
significance. In fact, the average user visits Grand 'Canyon 0.28
days per year so that an improvement in visibility from B to C
(about 105 to 125 miles) is worth about $ 0.30 per household per
year. In contrast, the preservation value of avoiding that -same
reduction in Grand Canyon visibility amounts to about $ 65 per
per year. Avoiding a similar reduction (about 12U to 99
all Southwest Parks is valued at about $ 120 per household
household
mi 1es ) at
per year.
The magnitude of the estimates of user benefits are consistent
with other studies, but the 'validity of the preservation benefit
estimates has been challenged by the results of otjier studies.
*Visual ranges were calculated from data in Schulze, et al.,
Table 3 (p. 18) and averaged across the.three vistas.
-------
26
Randall, Hoehn, and Tolley (1981) found that WTP for preserving
Grand Canyon visibility was significantly reduced when the Grand
Canyon component was valued as the third element in an air quality
program. In 1980 they found that for Chicago respondents the value
of preserving visibility at Grand Canyon as the sole program was
about $ 90 per household per year, but in a 1981 survey that value
declined to about $ 15 per household per year when Grand Canyon
visibility was evaluated third in a program that included improving
visibility in Chicago and the eastern United States. Rae/CRA (EPRI
draft, 1984) obtained similar results when visual air quality in the
Smokies was compared in a single and multiple good context.
In follow-up research Schulze et al. (1984) tested the.
sensitivity of their findings in the Grand Canyon/Southwest Parks
study to the single/multiple good context. Four different
experiments were designed to introduce budget constraints and to
include other competing local and regional air. quality programs.
The authors found no significant difference in bids to preserve
Grand Canyon visibility for four different contexts. Nevertheless,
Schulze, et al. did not duplicate the Chicago procedure. Whereas
Randall, et al. asked for total WTP for an expanded air quality
program, Schulze, et al. asked whether under new circumstances the
Grand Canyon bid would change. The oral nature of the questioning
is likely to affect the response, since respondents may be reluctant
to change their bid for the sake of maintaining consistency. More
theoretical and empirical research is needed to verify the bits and
pieces of knowledge so far available.
The Grand Canyon/Southwest Parks study
use and non-use benefits of visual air quali
User WTP was measured in terms of increases
yielded a value at Grand Canyon of up to $ 5
to improve visibility from a 7th percentile
a 99th percentile day (about 258 miles). Th
amounted to about $ 65 per household per yea
visibility (about 125 miles) at Grand Canyon
household per year to preserve median visibi
at al.l Southwest Parks. Subsequent studies
about the accuracy and reliability of these
estimates, but this study does suggest that
important in valuing visual air quality in C
3.4.2 National Park Service Study
attempted to quantify
ty in Class I areas.
in entrance fees and
.15 per household visit
day (about 105 miles) to
e non-use benefits
r to preserve median
and about $ 120 per
lity (about 120 miles)
have raised questions
non-use benefit
such benefits could be
lass I areas.
The National Park Service conducted a survey at Grand Canyon
and Mesa Verde National Parks to assess the effect of visual air
quality on visitor enjoyment (MacFarland, Malm, and Molenar, 1982).
The study included an economic valuation component that employed an
oral bidding format to measure willingness to pay for improved
visibility. Approximately 1000 respondents were surveyed at each
site during the summer of 1981.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
V
f :
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
ft
27
Visitors to Grand Canyon NP were asked about their WTP for
visibility improvement at Grand Canyon and their WTP for visibility
improvement over the Southwest Region. A Grand Canyon composite
depicted three different vistas under five different air quality
conditions. A Southwest Region composite included vistas of Grand
Canyon, Mesa Verde, and Zion National Parks under five different air
quality conditions. At Mesa Verde NP visitors were shown vistas of
Mesa Verde, Zion, and Grand Canyon, one park at a time, and asked
their WTP for five levels of improved visibility. The five air
quality levels in each survey represented visual ranye conditions of
about 110, 16U, 210, 270, and 35U kilometers (66, 96, 126, 162, and
210 miles).
Respondents were shown all five slides simultaneously; then they
were asked to compare the worst visibi1ity 'with the next worst and
to indicate their maximum WTP in increased entrance fees to visit
the park under the improved air quality condition. At Grand .Canyon
NP respondents were will'ing to pay an average of about $ 1 per
household visit over the $ 2.00 entrance fee to improve visibility
from 110 to 160 km. visual range. To improve from 110 to 210 km.
was valued at about $ 1.70, from 110 to 270 km. at about $ 2.10, and
from 110 to 350 km. at about $ 2.70 by Grand.Canyon respondents.
The regional composite of Grand Canyon, Zion, and Mesa Verde was
valued slightly less than the slides of Grand Canyon alon'e, which
suggests that Grand Canyon WTP bids may be a proxy for general"
regional visibility improvement.
The survey of respondents at Mesa- Verde elicited independent
bids for improved visual air quality at all three parks. As
expected, for comparable improvements in visual air quality
respondents were willing to pay the most for Grand Canyon; visual
ai'r quality at Zion was valued slightly more that at Mesa Verde.
Mesa Verde NP respondents valued Grand Canyon visibility about the
same as respondents at Grand Canyon NP: an improvement from 110 to
160 km. was valued at about $ 1 per household visit while a larger
improvement from 110 to 350 km. was valued at about $ 2.7b. For
similar visibility improvements household WTP in increased entrance '
fees to visit Mesa Verde was about 10 to 2U cents less than to visit
Grand Canyon. Visual air quality at Zion valued slightly higher
tnan at Mesa Verde.
The authors found that socioeconomic variables nad no important
effect on UT? for visual air quality. The most important variaole
was length of stay, and visitors "were willing to pay more the longer
their intended stay in the park. Overall, this study found that
visitors (users) were willing to pay several dollars per household
in increased entry 'fees .to .vi si t the parks in the Southwest on days
of good vi si bi1i ty .
-------
28
3.4.3 EPRI Parks Study
The Electric Power Research Institute sponsored a contingent
valuation study of benefits of improving visibility at Mesa Verde
and Great Smoicy Mountains National Parks (Rae/CRA, draft 1982).*
survey was given to about 200 visitors at each park in the summer
1981. Both contingent ranking and direct question (checklist
format) were used in eliciting WTP for visibility improvement as
increase in park entrance fees.
A
of
an
Four different visibility slides were shown to respondents at
each site. Slides were projected simultaneously so that respondents
could study and compare the differences in visual air quality. The
Mesa Verde vista showed Hogback Ridge under three visual ranges:
119, 156, and 256 km. (71, 94, and 154 miles) and an intense plume
that obscured the view of the ridge. The slides of Great Smoky
portrayed regional haze conditions with visual ranges of 10, 20 50,
and 100 km (about 6, 12/30, and 60 miles).
The survey design asked respondents to compare these visibility
conditions under two assumptions. In one case the visibility
condition on the day of a visit was known with certainty. In the
second case visibility was not known and was .described as a
probability distribution with each possible condition occurring a
specified percentage of time. The results of this second ranking
were not consistent across all specifications. Consequently, the
value of visibility as a probability distribution is difficult to
interpret and is not reported here.
When visibility conditions were presented as certain, the
rankings of visitors to M'esa Verde indicated a value of about $ 3.4U
per household per visit to improve from a hazy condition of 119-156
km. (71-94 miles) to a clear condition of 256 km. (154 miles). In
this case the intense haze and the median condition could not be
separated statistically, and the two conditions were subsequently
treated as visually equivalent. At Great Smoky visitors rankings
indicated a WTP trade-off of about $ 10 per household visit to
improve visibility from 50 to 100 km. (about 30 to 60 miles). To
improve visibility from 10-20 km. (6-12 miles) to 100 km. (6U miles)
was valued at about $ 15 per household visit.
Respondents were also asked to indicate in a direct question
written checklist format their WTP to insure a clear day on their
next visit. At Mesa Verde this value amounted to about $ 4.2U per
household to insure a clear (256 km.) condition relative to a median
(156 km.) condition (about 154 to 94 miles) on a future visit.
Given a stated probability of a future visit of .456, negligible
option value, and zero discount rate, the average use value amounts
*A paper on this research was presented at the National Park
Service's Visibility Values Workshop, May 1982.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
29
to about $ 9 per household visit.** At Great Smoky, this bid to
i.nsure clear visibility on a future visit amounted about to $ 6 per
household. Given a .769 probability of a return visit, negligible
option value, and a zero discount rate, this amounts to a use value
of about $ 7.75 per household to insure a clear (100 km.) condition
relative to a median (about 30 km.) condition (about 60 to 18 miles).
However, in the case of the Smokies the $ 10 limit on the.checklist
may have biased bids downwards, despite the opportunity to write in
ahigherbid.
This study found that visitors
were willing to pay about $ 3 to 15
major improvement in visual range.
bid appeared to yield values of simi
ranking method, although the results
independent samples. Visual air qua
values per visit at Great Smoky than
for the average length o'f visitor st
Smoky compared to one day at Mesa Ve
benefits per household.
to Mesa Verde and Great Smoky
per household per visit for a
The direct question checklist
lar magnitude to the contingent
were not confirmed on
1ity appeared to yield higher
at Mesa Verde, but correcting
ay (about three days at Great
rde) yields fairly similar
3.4.4 Six Cities Study
The survey design in the Six Cities study, reported above, " '
included WTP bids (expanded payment card and rebid) for a 10 mile
improvement in visual range in the eastern and western United
States. These bids represent both future use and non-use values for
visual air quality outside the immediate urban area. Annual average
household WTP for each of the six eastern cities ranged from about
$68 to 83 for the eastern scenario and about $ 34 to 39 for the
western scenario. The survey employed vistas of the Smokies and
Grand Canyon to portray eastern and western visibility, but these
vistas are probably not typical of the whole area included in the
bid. Nor were these bids presented in a wider multiple good context
of competing claims on household resources. Nevertheless, these
results provide further evidence that there may be an important
non-use component in valuing non-local visibility, especially in
Class I areas.
3 . b. Comparison of Benefits of Visual Air Quality in Urban and
Class I Areas
This report has reviewed visibility benefits research from nine
different urban areas and three National Parks. Contingent .
.*
**A positive discount rate would increase this user value while
the likelihood of more than one future visit would reduce it.
-------
30
valuation studies, including both direct question and contingent
ranking methods, were conducted in all nine cities and the three
national parks Property value studies were conducted 'in Los
Angeles and San Francisco and permit a useful comparison between
the survey-based CV and market-based hedonic methodologies.
3.5.1 Theoretical and Empirical Considerations
Many concerns about visibility benefit studies have been
summarized by Ruud.(1985) for the Utilities Air Regulatory Group.
The UARG report concludes that the evidence of economic benefits
associated with improved visibility is insufficient for policy
prescriptions except to assert that the benefits could range from
zero to a very large value. The principal criticisms are summarized
below:
I
I
I
I
I
I
I
1. The hypothetical'nature of the
of visibility in a single good context
the benefits of visibility improvement.
surveys and the presentation _
causes respondents .to overvalue!
2. The expanded payment card CV format and the method of
adjusting and removing high and low bids lead to an upward bias in
benefit results.
3. There i s .a wide variation in bids across individuals and
across CV formats that implies a very wide confidence interval.
4. The median serves as a better
than the mean, which in small samples
outliers, especially high bids.
indicator of average benefits
is.wore easily influenced by
' As discussed in
contingent valuation
,1.4., the hypothetical nature of
two components: unfami 1iarity
bids because no
because
with the market
overvaluing or
is a more serious
Section 2.2
surveys has
with the market and the possibility of inaccurate
money is actually exchanged. Lack of familiarity
is certainly a concern, but it may lead either to
undervaluing a good. The lack of a money payment
concern, since economists are always more comfortable with direct
market evidence. However, as reported above, the results of com-
paring CV results from a hypothetical survey with the results of
a simulated market experiment in buying and selling hunting per-
mits are mixed. In one study Bishop, et. al. (1984) found that a
hypothetical mean WTP exceeded the actual (real money) mean payment
and that the difference was statistically significant. However,
preliminary results of data collected in 1984 found no significant
difference between the results obtai ned. f rom the hy-pothetical survey
and the actual cash payment in a simulated market. Thus,
hypothetical bias may not be as significant as once feared, and
pending further research there is no empirical basis for adjustiny
results from contingent valuation surveys to account for
hypothetical biases.
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
31
The issue of competing claims on respondents' incomes is
another source of potential upward bias. The Cincinnati and Chicago
studies suggest that this problem is clearly of concern in valuing
non-use benefits in Class I areas, but its applicability to user
benefits is unclear. A'comparison of user WTP, with and without
budget constraint information, to avoid contact with diesel odors
indicates no important differences in results. A survey (Rae, 1983)
in Boston (79 individuals) without a budget constraint yielded
annual benefits of about $ 2.50 and $ 10 (1983 dollars) to avoid
increased weekly contacts with low and high intensity odors.
Preliminary results of a similar study in Philadelphia (Lareau and
Rae, forthcoming) that included WTP for other competing environmen-
tal and charitable goods yielded higher benefit values, about $ 3
and $ 16 (1984 dollars), respectively The expanded payment .card
format that contains budget information may also serve to remind
respondents of their budget constraint. This problem of competing
claims on scarce income is by no means settled, but again there is
mixed evidence as to whether it is an important concern in measuring
user values associated with urban visibility improvement.
The criteria for portrayi
is the subject of considerable
survey respondents may view vi
little market familiarity, in
may be consistency in an indiv
but that an individual may be
or 40 mile visual range relati
Consequently, comparing values
studies requires great care.
associated with this problem.
visual range or a 35 mile visu
visibility range with a 30 mil
ng and interpreting visual materials
research. There is concern that
sibility, a good with which they have
a qualitative manner. That is, there
idual's valuations of median to good,
relatively indifferent to photos of 25
ve to a median 10 mile visual range.
of improved visual range across
However, there is no obvious bias
Respondents may react to a 25 mile
al range as both typical of some wider
e mean.
Some direct question formats have encountered problems with
protest bids, and Ruud suggests that the procedures used in both the
Six City and Chicago studies may bias, the results upwards. These
studies identified zero, protest, and high bids and further
questioned respondents to determine the accuracy of the stated bid.
Protest bids were eliminated, and some very high bids were adjusted
downwards. These corrections appear to have been somewhat ad hoc;
nevertheless the direction of the change is to depress the mean of
the "solid core" data relative to the raw data. A more rigorous
approach to outlyers would have been preferable, but it is not
apparent that such an approach would yield significantly different
benefit estimates than were reported in the "solid core" data.
Tolley, et al. justified using the expanded payment card (C2)
format in the Six Cities study on the basis of a greater percentage
of "solid core" data (mainly attributable to fewer zero and protest
bids), which results in higher WTP. The UARG study points out that
this criterion creates potential upward biases due to format
selection. -However, a strong downward bias is also at work in the
-------
32
Six Cities study in that the photographs depicted generic
backgrounds. These two biases appear to cancel each other.
Average WTP bids in the Six Cities study using the C2 format and
generic photos are about the same as the six formats in the Chicago
study that employed photos of the Chicago urban area.
The UARG report points out that individual visibility bid data
exhibit a wide confidence interval for a given format. However,
econometric evidence suggests a fairly precise relationship between
visual range and WTP for visibility improvement in the Six Cities
data. A linear specification yielded a coefficient on visual range
(miles) of $14.2 with a standard error of $ .60. Thus for a 20 mile
improvement the 95 percent confidence>interval (two standard errors
or $ 14.20+S1.20) would be $ 260 to $ 308, which is actually a
very tight confidence interval. Further econometric analysis using
an aggregate model has provided non-linear specifications with
similarly tight confidence bands.
That different formats yield different bid results is clear
from the Chicago study. If variation across formats is included,
Ruud demonstrates that the confidence.interval increases
significantly. This calculation assumes that choice of format is a
random element in the survey design. The development of different
survey formats has been motivated by the need to avoid certain types
of biasing influences and to capture a full measure of WTP. The
fact that these formats yield different results is to be expected.
It is certainly fair to suggest that the expanded payment card
format used in the Six Cities study has some potential biases and to-
suggest that some other format is more accurate. However, it is not
correct to ignore all the literature on CV format development and to
assert that all formats ever used in previous studies are part of a
distribution that should be treated as a random element in
estimating benefits.
Another important issue is whether the mean or the median of a
sample represents a more accurate measure for computing aggregate
benefits. It should be pointed out that the mean of the population
is the correct measure, but in small samples the mean is easily
biased by extreme values. The appropriate question is at what
sample size is the mean of the sample more accurate than the median.
This question applies primarily to the direct question studies (Six
Cities and Chicago), since the ordered logit estimator used in the
Cincinnati study approximates a median estimate (Hanneman, 1984).
Table 2.shows.that for the Six Cities study with over 500
observations the median amounted to about 72-80 percent of the mean
local visibility bids. However, the standard errors of the means
of the three bids indicate a reasonable level of precision in the
estimate of the mean. For the 10 to 30'mile improvement we can say
with 95 percent confidence that the mean lies between 170.36 and
212.04. The median value in this case lies outside the range of
even four standard errors. Thus, in the Six Cities study with a
sample size over 500 the sample mean is a better approximation of
the true mean of the population than is the sample median.
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
33
3.5.2 Comparison of Visibility Benefits in Urban Areas
The three eastern urban visibility studies are most appropriate
for estimating WTP benefits for improvements in urban visibility.
Table 6 compares in 1984 dollars the WTP benefit estimates for a one
mile improvement in median visual range from.the three visibility
studies in the eastern United States.
The direct question WTP method generated annual average WTP
benefit es t i mates (1984 dol1 a rs ) per household that ranged from
about $ 5-41 for different cities.and across different formats. The
Chicago study tested WTP direct question formats and obtained annual
average household benefits of $ 14-41 per mile improvement in
typical visual range. The highest WTP benefit, $ 41 per.mile, was
achieved with the expanded payment card format. Using an expanded
payment card plus a rebid mechanism, the Six Cities study yielded
average annual household WTP benefit values for each city that
ranged from $ 6-25 per mile improvement in typical visual range.
Estimation of a non-linear visibility value function reduced that
range to about $14-17.50 per mile. The Cincinnati study used a
checklist format and obtained annual average WTP benefits of .about
$5 per mile improvement in the pre-test and about $ 33 per mile in
the final survey.
*
The contingent ranking method was used only in Cincinnati. A
pre-test generated an annual average household WTP (1984 dollars) of
S 24-34 per mile improvement in visual range for a change in
visibility occurring 'most of the time'. In contrast, the
comparable benefits from the final contingent ranking survey, which
used smaller visual range increments, were much higher, $ 78-119.
The range of estimates for each survey is due to specifications that
used different choice sets. When visibility was described as a
.distribution of visibility conditions, the contingent ranking method
yielded larger WTP benefits on a per mile basis. The pre-test
yielded a benefit of $ 25 per mile for a large (84 percent) change
in average visual range' and $ 65 per mile for a small (21 percent)
change in average visual range. The final survey resulted in
substantially higher benefits, but inconsistencies across
specifications yielded considerable uncertainty in the estimates.
Benefit estimates from the California s
valuation and property value, tend to lie in
310-127 per mile for improvements in visual
California contingent valuation studies were
eastern CV visibility studies. Annual avera
one mile improvement in median visual range
SCAB study and about $ 40 in the SFBA study.
the visibility component of total benefits,
property value studies ranged from $ 23 to $
improvement in visual range. A one mile imp
in the South Coast.Air Basin was valued at $
In the San Francisco Bay Area study a one mi
tudies, both contingent
a wide range of aoout
range. The two
i n the same range as
ge household WTP for a
was about $ 10-16 in the
Usi ng 34-50 percent as
WTP values for the three
127 per mile
rovement in visual range
23-54 per household.
le improvement in visual
-------
34
range was valued at $ 49-72 per household.. A more recent hedonic
property value study in Los Angeles and San Francisco that used tne
extinction coefficient as a measure of air quality yielded somewhat
higher household WTP benefit estimates for a one mile improvement in
visual range: $ 37-127 in Los Angeles and $ 38-61 in San Francisco.
The wide ranges for these property value estimates reflect
differences in WTP benefits for the different functional forms
estimated.
The WTP benefit values generated by the direct question formats
used in the Chicago, Six Cities, Cincinnati, and California
contingent valuation studies fall into a range of about $ 5 to $ 41
per mile for 50 to 100 percent increases in typical visual range
The results of the Cincinnati contingent ranking pre-test also lie
within this range. The annual household benefit values from the
property value studies conducted in California range from $ 23-127
per mile, which is wider than the range implied in the contingent
valuation results The property value benefit results depend on tlie
percentage of visibility benefits, and thus the CV results in the
range of $ 5-41 per mile appear more defensible at this time. It
should be noted, however, that several studies found benefits per
mile improvement in visual range that were in excess of $ 5U per
household, and the $ 5-41 range of benefits may be considered
relatively conservative.
3.5.2.1 Derivation of Urban Visibility Benefits Estimates
There is no one study that can withstand all the relevant
criticisms. Rather, it is necessary to choose among the existing
studies, recognizing their limitations. The University of Chicago
studies of Chicago and an additional six eastern cities appear to
offer the most broad-based data set for estimating the value of
improved visibility in eastern areas. Annual household bids (1984
dollars) for the expanded payment card format yielded an average
estimate (seven cities) of about $ 18 for a one mile improvement in
visual range. The simplicity of the linear functional form is
appealing, but a non-linear specification is indicated by the data.
A non-linear function was estimated on data for six cities using the
negative exponential specification presented below:
Bid (i 1932) = 519.7206[1-exp{-0 .02703* VR)]
r-squared a 0.469
where VR is visual range rescaled so that the mean VR
0
This relationship yields a significant parameter for visual
and explains a high percentage of the variance in the estimates.
also explains much of the difference in the raw bids in terms of
differences between individuals and cities. In 1984 dollars
(constant term becomes 571.6926) it predicts that for an average
range
It
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
i
i
i
i
i
i
35
household (in Boston or Mobile) a 1 mile increase*from the mean
(about 10 miles) would 'be worth about $15 and an increase from 19
to 20 miles visual range would be worth about $12. Thus, a best
estimate of urban visibility benefits can be obtained by applying
the relationship below to changes in visual range:
Bid (1984 $) = 571.6926[l-exp(-0.02708*VR)j
3.5.2.2 Low and High Estimates
Results from other studies provide a basis for low and high
estimates. A comparison of the average of the five other formats
in the Chicago study suggests that the expanded payment card
generated average bids that were about 50 to 100 percent higher.
It seems appropriate to use this variation as a basis for a low
estimate. The constant term from the Six Cities study is adjusted
to one-half of the value used in the best estimate, and a 1 mile ,
improvement from the average visual range is worth about $7.50:
The hi gh estimate is
litial Cli
based on the upper bound estimate from
the ini ti aTThicago study, the San Fransisco Bay Area study, and
the pretest -results from the Cincinnati study. The Chicago upper
bound estimate was $41/mile, while the CV portion of the San
Fransisco estimate was $40/mile. A value of $45/mile is. derived .
from the Cincinnati contingent ranking pre-test data using a
semi-log specification to relate bids to changes in visual range.
Thus, there is evidence from three studies that visibility could be
valued as high as $40/mile, which serves as the high estimate.
This benefit is captured in a higher constant tel
exponential function:
Bid (1984 $) = 1497[l-exp(-0.02708*VR)]
high i
jrm in
the negative
These predictors are all based on a mean visual range of 9 to
10 miles, which must be rescaled to zero. This allows changes from
the mean visual range to enter the equation directly as VR. The
resulting benefit estimates are in 1984 dollars.
3.5.3" Compari son of Class I Area Visibility Studies
Class I areas include many unique National Parks and Wilderness
areas. Consequently, it is to be expected that visibility benefits
will differ at different sites. Yet, the user values reported in
these three studies are quite similar after correcting for inflation,
visual range improvement, and time of visit. These studies have
elicited user values for large increases in visual range, and
Table 6 presents user values in 1984 dollars for a 100 mile visibility
improvement at Grand Canyon, Mesa Verde, and Great Smoky Mountains
National Parks. At Grand Canyon these benefit estimates vary from
about $ 2.40 to 7.80 per househol d.-day. At Mesa Verde all but one
of the benefit estimates lie in the range of about $ 2.20 to 5.50
per household-day. At -Great Smoky Mountains the benefit estimates
-------
36
range from about $ 7.75 (probably 'biased downwards) to $12.80 per
household-day. The.remaining differences in visibility values
between eastern and western Class I areas may be explained by
differences in mean visual range. Mean visual range in the rural
west is about 60 miles compared to about 20 miles in the rural
east. There are not sufficient studies to estimate a non-linear
visibility value function for Class I areas, but it appears
reasonable to apply the following user benefit estimates (1984
dollars per household-day) to any 100 mile improvement in
visual range at class I areas:
User Benefits Per 100 Mile Increase in Mean Visual Range
(1984 Dollars/Household-Day)
Low
Medium
High
West
2.50
4.00
7.50
East
$ 8.00
11.00
13.00
suggest that households are
to preserve or improve visual
for
In addition, there is evidence to
willing to forego some income in order
air quality at certain vistas for the use of others. Estimates
the Grand Canyon ranged from about $ 15 to 90 per household per
year, and estimates for the Smokies ranged from about $ 20 to 60 per
household per year. The differences in these estimates were due to
the bidding context: a multiple good context generated the low
estimates and a single good context generated the high estimates.
Using photographs of Grand Canyon and the Smokies to portray western
and eastern visibility, respectively, it was also found that urban
households in the east were willing to pay $ 34 to 39 per year for a
10 mile improvement in western visibility and $ 68 to 83 per year
for a 10 mile improvment in eastern visual range. These values were
in addition to bids for local visibility improvement, but did not
consider competing claims of other non-visibility goods.
pay significant sums of money to
our natural and historical
number of dollars contributed to
That Americans are willing to
preserve the important features of
heritage is evidenced Dy the large
different funds, especially the current Save the Statue of Liberty
campaign. However, what is not clear is whetner that amount of
money would be forthcoming every year and what other charities or
other volunteer causes are receiving less as a result. Until the
answers to these questions are known it is impossible to estimate
non-use value for visibility in non-local or Class I areas. It
should also be noted that for some areas non-use values are likely
to be quite significant in the magnitude of benefits for
improvements in visual air quality.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
1
1
1
1
1
1
I
1
1
1
1
1
1
1
1
1
*
Average Annual Househol d -Wi 1
(Solid Core
WTP Format Observations
-Single Bid (Al ) 37
(Oral)
Accept/Reject (A2) 35
(Oral)
Ural Iterative (A3) 31
Bid
Checklist (8) 34
Payment Card (Cl ) 42
(Public Goods)
Payment Card (C2) 42
(Pub! i c and
Pri vate Goods )
>
Total Sample 221
Notes: Numbers in parenthese
37
Table 1
1 i ngness
Chi cago
Bids -
Change
9 to 4
$ 250
(51)
156
(30)
222
(37)
121
(25)
210
(44)
283
(57)
227
(20)
to Pay for V
1981 dollars)
in Typical Vi
9 to 18
$ 236
'(50)
147
(22)
210
(38)
109
(22)
186
(35)
324
(72)
" 218
(20)
are standard errors
Source: Randall, A., Hoehn, 0., and
of Contingent Markets: Results of a
presented at the American Economic As
1981 .
.
isibility Change in
sual Range
9 to 30
$ 362
(50)
171
(24)
24U
(.39)
152
(29)
\ * "^ /
234
(53)
\ /
456
(115)
271
(28)
of the mean .
Tolley, G., "The Structure
Recent Experiment," a paper
sociation Meetings, December
-
-------
(
Average Annual Househo
City
Atl anta
Boston
Cinci nnati
Mi ami
Mobi le
Washington
Total Observations
Sample Mean
Sample Median
'
38
Table 2
Id Willingness to Pay for Visibility Change in
Six Cities
(City Average in 1982 dollars)
Change in Typical Visual Range
10 to 5 10 to 20 10 to 30 +10 East +10 U.S.
(miles) (miles) (miles) (miles) (miles)
$ -196 $ 188 $ 286 $ 281 $ 353
-145 139 171 ^189 224
- 57 57 64 74 80
- 99 88 104 116 118
-156 168 197 215 238
-232 238 303 358 422
-525 533 535
-150 150 191 205 239
' (181) (182) (241) (NA) (NA)
-120 108 144 NA NA
Notes: 1) Negative values represent WTP to avoid a reduction
in visual range; positive bids represent WTP to
achieve an improvement in visual range.
2) City values are mean bids for each city.
3) Mean and median values are presented for the
entire sample. Numbers in parentheses are
standard deviations from the mean.
Source: Tolley, G.f et al., Establishing and Valuing the Effects of
Improved Visibility in Eastern United States, draft report for U.S.
Environmental Protection Agency, Office of Research and Oevel opment ,
March 1984.
f
1
i^B
1
I
|
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
1
1
1
1
1
1
39
Table
Average Annual Household Benefits
Method
Ci nci nn
(1982 dol
3
.of Changes in
ati
lars) '
Visibility in
Basic Specification
(Pre-test)
(Final Survey)
First Contingent Ranking
WTP:
WTA:
Obse
Second
WTP:
WTA:
medi an to cl ear
( 'most of the time1 )
median to haze
( 'most of the time ' )
rvations:
Contingent Ranking
Existing (27 Perce'nt Clear)
to Best (80 Percent Clear)
Existing (27 Percent Clear)
to Good (40 Percent Clear)
Existing (27 Percent Clear)
to Degraded (15 Percent Clea
Existing (27 Percent Clear)
$ 381-544
(38-97)
$ 387-635
(85-120
140
$ 217
(36)
$ 142
(25)
$ 106
r) (24)
to $ 268
Very Degraded (8 Percent Clear) (35)
1
I
1
1
1
1
No Health Effects to 2 days
lung and eye irritation
No Health Effects to 8 days
lung and eye irritation
Observations:
Di rect
WTP:
Obse rva
Notes :
Source:
Qual ity
report,
Question - Checklist Bid
medi an to clear
t i o n s :
1) The range of estimates is
with different choice sets
2) Numbers in parenthese are
associated with the bottom
-
82
$ 85
(10)
147
based on specif
,
standard errors
and top of the
Rae, D., /Charles River Associates, Benefit
In Cincinnati Results of a C
Electric Power Research Inst
r-
ontingent Ranki
itute, Project
$ 343-516
(50-66)
$ 884-1324
) (102-153)
296
$ 122-358
(73-40)
S 32-184
(49-45)
$ 288-386
(61-93)
$ 417-625
(72-113)
$ 323-498
(73-127)
$ 530-906
(84-182)
296 -
$. 143
(12)
314
i cat ions
f
of estimate
range.
s of Visual Air
ng Survey, draft
#1742, August, 1984
i^^^nH^^Hi
-------
40
. Table 4
Summary of California Visibility Studies
(Average Annual Willingness to Pay Per Household}
Study
WTP for Change
in Ai r Qua1ity
{$ Current) (5 1984) (34-65%
WTP for
VR Change
VR Change
(miles) ($ 1984)
$ 1073 $ 365-697
477
162
2 to 12
12 to 28
2 to 12
12 to 28
$ 23-70
10-16
16.3-21.2 10-19
B rook shi re , et al .
South Coast Ai r Basin
Property Value $ 558
30 % reduction in
TSP - 1977/78
Contingent Valuation 312
30 % reduction in
ambient pol luti on
1978 survey
Loehman , et al .
San Francisco Bay Area
Property Value " 86
30% reduction in
PS! Index - 1978/79
Contingent Valuation 71
14% improvement in
visibility - 1980
Thayer and Tri joni s
Los Angeles "Property 57-153
Value 10% reduction
in extinction - 1978/79
San Francisco Prop- 115-125
erty Value 10% reduc-
tion in extinction
independent of ozone
Note: Rate of inflation for residential properties in California
9.7 % per year. For households values are inflated by the Depart
143
90
49-93
90
16.3-18.6
40
95-254 32-165
191-208 191-208
10-11
32-165
17-18.7 112-122
of Labor's Implicit Price Deflator for Personal Consumption
Expenditures, which rose from 149.2 in 1978 to about 228 in 1984
is
ment
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Source: Bropkshire, D., d'Arge, R., Schulze,
Experiments in Valuing Non-Market Goods
of Air Pollution in the SouthCoast Air
W
and
Thayer,
Alternative Benefit Measures
Basin of Southern California,
M., 19791
EPA-600/6-79-0016; Loehman, E., Boldt, 0.', and Chaikin, K., 1981
Measuring the Benefits of Air Qual.ity Improvements in the San Fancisco
Bay Area, SRI 8962; and Thayer, M., and Trijonts, J., February 1984,
Visibility Benefits Analysis, draft report for California Air Resource
Board.
I
I
I
-------
1
1
1
|
1
1
.
1
I
1
1
1
1
1
1
41
Table 5
Comparison of Willingness to Pay Benefit Esti
Changes in the Eastern United
(dol 1 ars per mile)
Study Benefit
($/mi le)
CHICAGO: Six Direct Question Formats
(Solid Core Bids - 1981 Dollars)
9 to 4 miles 24-57
9 to 18 miles 12-36
9 to 30 mi les 7-22
SIX CITIES: Payment Card and Rebid Format
(1982 Dollars)
10 to 5 mi les ' 11-46
10 to 20 mi les 6-24
10 to 30 mi les 3-15
CINCINNATI: Pre-Test (1982 Dollars)
WTP: Contingent Ranking
7.8 to 25.2 miles (most of time) 22-31
11.4 to 21.0 miles (distribution) 23
11.4 to 13.8 miles (distribution) 59
WTP: Direct Question - Checklist Bid
7.8 to 25.2 miles (most of time) '5
CINCINNATI: Final Survey (1982 Dollars)
WTP: Contingent Ranking
11.6 to 16.4 miles (most of time) 71-1U8
10.9 to 14.4 miles (distribution) 35-102
10.9 to 11.8 miles (distribution) 36-2U4
WTP: Direct Question - Checklist Bid
11.4 to 16.4 miles (most of .time) 30
mates for Vi sibi 1 i ty
States
Estimates
(1984 $/mi1e)
27-64
14-41
8-25
.
13-51
6-26
4-17
24-34 "
25 -
65
5
79-118
43-127
39-226
33
Source: 6. Tolley, et al.-, 1984, Establishing and-Va1uing the Effects
of Improved Vi si bi 1 i ty -i n Eastern Uni ted' States , Office of Research and
Development, Environmental Protection Agency, March 1984; D.
Rae/Charles River Associates, Benefits of Visual Air Quality In
Cincinnati Results of a Contingent Ranki.ng Survey, draft report,
Electric Power Research- Institute, Project #1742, August.
-
-------
42
Table 6
WTP Benefits of Visual Air Quality in Class I Areas
(dollars per household visit)
Observations Visibility Vista
Benefits ($)
(mi res)
Grand Canyon
1980
National Park
Service
1980
EPRI - Parks
1981
166
166
166
166
1000
1000
1000
1000
800
800
800
800
196
193
202
202
201
78
78
78
78
66
66
66
66
66
66
66'
66
71-94
94
6-12
30
18
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
(Cur
105
125
179
258
96
126
162
210
96
126
162
210
154
154
60
60
60
G
G
G
G
G
G
G
G
M
M
M
M
M
M
G.
G.
G.
rent!
.C.
.C.
.C.
.C.
.C.
.C.
.C.
.C.
.V.
.V.
.V.
.V.
.V.
.V.
S.M.
S.M.
S.M.
) (1984/100 miles/day)
$ 1.
2.
3.
5.
0.97
1.35
68
76
90
15
-0.98
-1.60
1.92-2.10
2.
o;
1.
1.
2.
3.
9.
14.
9.
7.
75
85
15
75
55
40
14
94
91
75
$ 7
7
4
3
4
2.80
2.50
2
3
2
2
2
5
17
11
12
7
.80
.40
.85
.60
.10
-3.
-2.
.40
.55
.40
.30
.20
.50
.70
.10
.80
.15
40
75
Notes: 1) The National Park Service Study evaluated visibility at
Grand Canyon in surveys undertaken at both Grand Canyon and Mesa
Verde National Parks. The benefit range presented above reflects
the small differences between the two surveys.
2) Benefits in current dollars per visit are converted to
1984 dollars per household-day. The Consumer Price Index is used
to inflate 1980 and 1981 values to 1984 dollars. The average
visitor length of stay is used to convert the benefits per visit
to benefits per household-day. This value is about 3 days at
Great Smoky Mountain and 1 day at Mesa Verde. Data for Grand
Canyon were not available and this calculation assumes 1 day.
Source: Schulze, W., Brookshire, D., Walthe
Thayer, M., Whitworth, R., Ben-David, S., Ma
J., September 1983, Methods Development for
Benefits Assessment, Volume III, The Benef.it
Visibility in the National Parklands of the
Office of Policy Analysis, Office of Policy
ment, EPA-230-07-83-009, Washington, D.C.;
Value to Visitors of Improving Visibi1ity at
Smoky Mountains National Parks," and MacFarl
W., and Molenar, J., " An Examination of Met
Assessing the Value of Visibility," in Rowe,
(eds.), 1983, Managing Air Quality and Sceni
Parks and Wilderness Areas, Westview Press,
r, E., Kelley, K.,
1m, W., and Molenar,
Environmental Control
s of Preservi ng
Southwest, U.S. EPA,
and Resource Manage-
Rae, Douglas A., "The
Mesa Verde and Great
and, Karen K., Malm,
hodologies for
R. and Chestnut, L.
c Resources at National
Boulder, Co.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
43
Quant i f 1 cati on of Pol 1 ar Benef 1 ts of Improvements 1 n VI suaj
t
An estimate of the benefits of policies that improve visual air
quality requires two pieces of information: the change in visual
range caused by reduced emissions and the valuation of specific
visibility increments. The preceding sections developed a visibility
value function that can be used to value specific increments in
urban visual range. The contractor study (SAI, 1984} described in
Appendix B provides estimates of changes i/i visual range in the
eastern United States for various sulfur reduction scenarios.
The SAI analysis yields a prediction of visual range for 80
regions in 30 eastern states for several different emission reduction
scenarios. 'A visibility value function was then applied to tnese
changes to yield a benefit value per household in each state.* Total
state benefits were computed by summing over the number of households.
Similarly, total benefits for the eastern United States were computed
by summing over all 30 states. A summary of these results is provided
below for three acid rain abatement proposals by applying the low,
best, and high visibility benefits estimates derived above to the
changes in visibility predicted by SAI:
Visibility Benefits of Acid Deposition Abatement in the Eastern
United States
Reduction VR Change Benefits {$ million)
Scenario (S02) (%) Low Midpoint
10 State
8 million tons
12 million tons
4.5 $ 250
11.1 .500
$ 500
1100
15.8
800
1600
4100
Note: These benefits are in 1984 dollars.
Research from perception studies has demonstrated that the human
eye can perceive a change in contrast of 2-5 percent which translates
into a change in visual range of about 5-15 percent. This peception
threshold applies to evaluating specific slides or photographs that
represent median or typical conditions. In urban areas, however,
studies have shown that respondents do value small changes in dis-
tributions of good, median, and bad days that may change the mean
visibility by less than 5-15 percent. Consequently, it is likely
that this concept of a 5-15 percent threshold of perception is more
applicable to visits to Class I areas than to changes in local
visibility. Nevertheless, a medium estimate of benefits is provided
below with perception thresholds of 0, 5, and 15 percent.
*The SAI report, used a semi-log visibility value function
while this report uses a negative exponential visibility value
function taken from the Six Cities study described above.
-------
Benefits of Improvements in Visual Range
(Medium Benefit-Scenario)
Reduction
Scenari o
10 State
8 mi 1 1 i on tons
VC Change
(%)
4.5
11.1
Bene
0 %
$ 500
1100
fits
Th
($ mi
reshol
$
1
1 %
400
100
Hi
ds
ons
1
$
)
.5
0
400
I
I
I
I
I
12 million tons 15.8 1600 1600 1300
Note: These benefits are in 1984 dollars.
A report by the Utilities Air Regulatory Group (AeroVironmen
1985) found much smaller benefits for improving visibility in the
eastern United States. In addition to applying different scienti
and modeling parameters this analysis employed an economic benefi
function based on the Six Cities data whose origin was shifted to
account for a 15 percent threshold of perception. This function
assumed zero benefits at a 15 percent threshold level and fitted
linear and non-linear curves from this point of origin. This Shi
of the origin is inappropriate, since it does not fit the Six Cit
WTP data, which are best fit by a non-linear function through the
origin and which lies significantly above zero at the threshold
poi nt.
visual range
However, no
The development of similar benefit estimates for
improvement in western urban areas is also possible. .i«c»c.
comprehensive modeling analysis is yet available to translate
regional emission reductions into changes in visual range.
A similar valuation analysis could be applied to visitors to
Class I areas. As mentioned above, for short duration visits the
concept of a threshold of perception is applicable. An improvement
in visual range of less than 5-15 percent on a given day would ha
no value to visitors. However, even a sub-threshold improvement
mean visual range could cause some days to improve by more than 5
percent, but this value could be estimated only if a highly detai
modeling analysis were available to predict the full distribution
daily visibility events.
In the case of Great Smoky Mountains National Park, which "I i
on the border between North Carolina and Tennessee, a 12 million
reduction scenario yields an estimated improvement in mean visual
range of 13.9 to 20.7 percent, or about 1.5 to 2.3 miles. If we
assume this increase in mean visibility 'occurs each day of the ye
we can estimate the benefit to visitors to Great Smoky Mountains
National Park. The" Park records -about 9 million visitor days per
year, or about 3 million household days. The value of visibility
eastern Class I areas was estimated in dollars per household-day
100 mile increase in visual range and ranged from about $ 8-13.
t,
fie
t
then
both
ft
ies
e
nt
ve
in
-15
led
of
es
ton
ar,
in
per
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
45
Thus, a 1.5 to 2.3 mile increase in mean visual range at Great Smoky
Mountains National Park would have a user value of about $ 400,OUU
to 900,000 per year. -
Other eastern Class I areas have to be evaluated individually.
A 12 million ton reduction scenario would yield a 23.6 percent
improvement in visual range at Shenandoah NP in Virginia, but only
a 5.0 percent improvement at Acadia NP in Maine. The total bene-
fits at the former can easily be -calculated from visitation data.
Total benefits at Shenandoah NP are less than at the Great Smoky
Mountains NP1 because Shenandoah receives substantially fewer
visitors per year. At Acadia the visibility change is not great
enough to exceed the threshold value so that benefits are probably
negligible. As is evident, the range of benefits per Glass I area
is fairly small and summing over the relatively small number of
Class I areas in the east is expected to yield annual dollar bene-
fits for users that is unlikely to exceed $ 10 million. This
level of benefits is small relative to the magnitude of benefits
for local visibility improvement. However, it should be noted
that there is some evidence that non-use benefits for visual air
quality improvement in Class I areas could be siynificant, but at
this point in time there is not enough agreement on the level of
these benefits to incl.ude them in this summation.
1
REFERENCES FOR SECTION 3
1. Beggs, S., Cardell, S., and Hausman, J ., 1981. "Assessing the
Potential Demand for Electric Cars," Journal of Econometrics, 17:
1-19.
2. Bishop, R.C., and Heberlein, T.A. 1979. "Measuring Values of Extra
Market Goods: Are Indirect Measures Biased?" American Journal of
Agricultural Economics, 61 (December): 926-930.
3. Bishop, R.C., Heberlein, T.A., Welsh, M.P., and Baumgartner, R.M.,
.1984. "Does Contingent Valuation Work? Results of the Sandhill
Experiment," Joint Meetings of the Association of Environmental and
Resource Economists, the American Agricultural Association and the
Northeast Agricultural Economics Council, Cornell University, August
5-8, Ithaca , New York .
4. Bishop, R.C. 198b.-oralcommunication of preliminary results from
a contingent valuation and simulated market experiment at Sandhill
conducted in-the summer, 1984. ...
5. Boyle, K.J., Bishop, R.C., and Welsh, M.P., 1985. "Starting Point
Bias and Contingent Valuation Bidding Games," forthcoming in Land
Economi cs .
-------
46
6. Brookshire, D., d'Arge, R., Schulze, W., and Thayer, M., 1979.
Experiments in Valuing Non-Market Goods: Alternative Benefit
Measures of Air Pollution in the South Coast Air Basin of Southern
California, EPA-600/6-79-0016.
I
I
I
I
7. Brookshire, D., Thayer, M., Schulze, W., and d'Arge, R., 1980.
Valuing Public Goods: A Comparison of Survey and Hedonic .
Approaches," Resource and Environmental Economics Laboratory, I
University of Wyoming, later published in American Economic Review,*
72 (March).
8. Oesvouges, W., Smith, V.K., and McGivney, 1983. A Comparison of |
Alternative Approaches for Estimating Recreation and Related Benefits
of Water Quality Improvements, U.S. Environmental Protection Agencyg
Economic Analysis Division. March.
9. Freeman, A.M., 1979. The Benefits of Environmental Improvement,
Johns Hopkins Press, Baltimore.
10. Freeman, A.M., 1982. "The Size and Sign of Option Value,"
unpublished paper, Bowdoin College, Brunswick, Maine.
11. Griliches, Z., (editor), 1971. Price Indexes and Quality Change.
Cambridge, Massachusetts: Harvard University Press.
12. Hanemann, W.M., 1984. "Welfare Evaluations in Contingent
Valuation Experiments with Discrete Responses," American Journal
of Agricultural Economics, Vol. 66: 332-341.
13. Lareau, T. and Rae, D., 1985. preliminary results of a contin-
gent valuation survey on WTP to avoid contacts with diesel
odor, forthcoming.
14. Loehman, £., Boldt, D., and Chaikin, K . , 1981, Measuring the
Benefits of Air Quality Improvements in the San Fancisco Bay
Area, SRI 8962.
15. .Loehman, £., 1984. "Measurement of Air Quality Benefits from
Survey Data,11" Staff Paper, Department of Agricultural Economics,
Purdue University, January.
16. MacFarland, Karen K., Malm, W., and Molenar, J.,^1982. " An
Examination of Methodologies for Assessing the Value of
Visibility," a paper presented at a National Park Service
.Conference at Keystone, Co., May 1982, published in Rowe,R. and
Chestnut, L. (eds.j, Managing Air Quality and Scenic Resources
at National Parks and Wilderness Areas, Westview Press, Boul der ,
Co., 1983. '
17. Malm, W-. Molenar, J., and Leiker, K., 198U. "Human Perception
of Air Quality,? Air Pollution Control Journal, 30 (February).
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23. Scnulze, W., Srookshire, 0., Walther, E., Kelley, K., Thayer,
M., Vlhitworth, R., Ben-David, S., Malm, W., and Molenar, J.,
1983. Methods Development for Environmental Control Benefits
Assessment, Volume III, The Benefits of Preserving Visibility
I in the National Parklands of the Southwest, U.S. Environmental
Protection Agency, Office of Policy Analysis, Office of Policy
and Resource Management, EPA-23U-07-83-009, Washin-gton, D.C.
' _ September.
I
I
*47
18. Mendelsohn, R., 1980. "The Demand and Supply for Characteristics
of Goods," University of Washington.
19. Mitchell, R., and Carson, T., 1982. An Experiment in Determining
Willingness to Pay for National Water Quality Improvements.
Resources for the Future.
20. Rae, D., 1982. "The Value to Visitors of Improving Visibility at
Mesa Verde and Great Smoky National Parks," a paper presented at
a National Park Service Conference, Keystone Colorado, May, later
published in Robert Rowe and Lauraine Chestnut, eds., Managing Air
Quality and Scenic Resources at National Parks and Wilderness Areas,
. Westvie.w Press, Boulder, Colorado, 1983.
21. Rae, D., Hausman, J., S.tankunas, A., and Vitka, S., 1982.
Benefits and Costs of Improving Visibility: Case Studies of the
Application of the Contingent Ranking Methodology at Mesa Verde
and Great Smoky Mountain National Parks, Electric Power Research
Institute, Research Project 1742, October.
22. Rae, 0., 1983. Benefits of Reducing Odors from Diesel Vehicles-
Results of a Contingent Valuation Survey, a draft report for U.S.
Environmental Protection Agency,' Economic Benefits Staff. January.
23. Rae, D.,/Charles River Associates, 1984. Benefits of Visual Air
Quality In Cincinnati Results of a Contingent Ranking Survey-, draft
report, Electric Power Research Institute, Project #1742, August.
24. Randall, A., Hoehn, J.,-and Tolley, G., 1981. "The Structure of
Contingent Markets: Results of a Recent Experiment," a paper
presented at the American Economic Association Meetings, December.
25. Rosen, S., 1974. "Hedonic Prices and Implicit Markets: Product
Differentiation in Pure Competition," Journal of Political
Economy, 82 (January/February): 34-55.
26. Rowe, R.D., and Chestnut, L.G., 1983. Visibility Benefits
Assessment Guidebook, U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Washington, D.C.
27. Ruud, P.A. 1985. A Review of the SAI Report to EPA, Chapter 7,
for the Utility Air Regulatory Group. February.
-------
48
29
30
31
32
33
34
Schulze, W., Brookshire, 0., et al., 1984. "The National Parks
Visibility Experiment," Methods Development for Environmental
Control Benefits Assessment, (mimeo), U.S. Environmental Protec-
tion Agency, Office of Policy Analysis, Washington, D.C.
Systems Applications, Inc., 1984. Visibility and Other Air
Quality Benefits of Sulfur Dioxide Emission Controls in the
Eastern United States, Volume I, U.S. Environmental Protection
Agency, Office of Policy, Planning, and Evaluation, Office of
Policy Analysis, SYSAPP-84/165, Washington, D.C., September.
Thayer, M., and Trijonis, J., 1984. Visibility Benefits Analysi
draft report for California Air Resources Board, Sacramento, Ca
February.
s,
Thayer, M, Murdoch, J., Trijonis, J., Hagemand, R., 1985. "The
Benefits of Air Quality Improvements," forthcoming Journal of the
Air Pollution Control Association Meetings, Detroit, Mi., June.
Tolley, G., Randall, 'A., Blomquist, G., Fabian, R., Fishelson,
G., Frankel, A., Hoehn, J., Krumm, R., Mensah, E., and Smith,
1984. Establishing and Valuing the Effects of Improved Visibil
in Eastern United States, U.S. Environmental Protection Agency
Office of Research and Development, Washington, D.C., March.
Tolley, G. 1985. "Visibility Value Function," (mimeo), U.S.
Environmental Protection Agency, Office of Policy Analysis,
Washington, D.C., January.
T.,
ity
35. U.S. Environmental Protection Agency, 1979. Protecting
Visibility: An EPA Report to Congress. EPA-450/5-79-OU8.
36. Zannetti, P., Tombach, I., and Drake, R., 1985. Critique of the
Draft Report "Visibility and Other Air Quality Benefits of Sulfur
Dioxide Emission Controls in the Eastern United States," prepared
for Utility Air Regulatory Group, AeroVironment, Inc.
(AV-FR-85/503), February.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
49.
4.0 VISIBILITY EFFECTS ON AIRCRAFT AND RELATED OPERATIONS
4.1 Significant and Major Visibility Reductions
Any analysis of the effect of visibility on aircraft operations must
necessarily distinguish between significant and severe reductions in
visibility. For the purposes of this discussion, a severe reduction in
visibility is one which reduces visible range below 3.miles, or below one
mile within about 1000 feet of the surface. ] Federal Aviation Regulations
allow VFR (Visual Flight Rules) operation when visibility is greater than 3
miles, or greater than one mile within 700. or 1200 feet of the surface. In
general, the 700-foot figure will apply in the vicinity of airports or in
areas of congestion. Visibility reductions which prohibit VFR operation
are considered severe in this context. /
/'
Again for the purposes of this discussion, a significant reduction in
visibility must be defined, but without benefit of specific aviation regulation
defining a visibility limitation. By popular consensus, visibility in the
West would be considered significantly reduced when visibility drops below
about 25 miles, but in the East 25 miles is considered excellent visibility.
25 mile visibility is great enough that it is very unlikely to affect
aircraft operations. 25 miles represents a travel time of 1/4 hour in a
small private plane. Below 10^000 feet, speeds are restricted so that 25
miles represents a travel time of at least 5 minutes in even the fastest
aircraft. This travel time would give more than ample time for accident
avoidance and visual navigation, and so is considered to be equivalent to
unrestricted visibility. Instead, a lower visibility will be selected as a
"significant reduction." This visibility will be selected for its effect
on aircraft visual operations rather than on-an aesthetic perception. For
these purposes, a visibility of five to ten miles is selected based on the
operations of student pilots in cross-country flights. A five to ten mile
visibility is considered adequate depending on the student's capabilities
as determined by the instructor while a lesser visible range is considered
insufficient for cross-country operations by student pilots. This is
because the student will have difficulty in locating his destination airport
and in avoiding other aircraft when visibility is more restricted but still
above VFR limits.
4.2 Effects of Visibility Reductions on Safety
Most pilots and non-pilots assume that reduced visibility degrades the
level of safety of flight operations. This assumption is based on the fact
that there is more difficulty in maintaining visual contact with other
aircraft during reduced-visibility conditions. In addition, the load on
air traffic controllers would increase, especially below 3 miles visibility,
as "special VFR" operations increase reliance on air traffic control to
maintain aircraft separation.
Examination of accident records does not produce immediate evidence of
a direct correlation of reduced visibility conditions with increased
accidents, although impaired visibility is frequently listed as a contributing.
factor in accident reports. Apparently, pilots are more vigilant in reduced
visibility situations and greater use is made of other-than-visual means
-------
50
of maintaining aircraft separation. Reduced visibility appears to transfer
the responsibility for collision avoidance from the pilot to the air traffic
controller. This transfer does not seem to have an immediate effect on
safety but may restrict some flight operations and would certainly increase
controller workload.
4.3 Major Visibility Reductions
The effect of major reductions in visibility is to cause a change from
Visual Flight Rules to Instrument Flight Rules. This change eliminates all
operations by pilots who are not instrument rated and restricts operations
by rated pilots to necessary flights between airports. (Sightseeing or
aerial photography would be precluded). Thus major visibility reductions
would tend to change the make-up of the air traffic and would eliminate
most or all pleasure flying.
Airport data reported in the 1982 EPA Sulfur Oxides and Particulate
Matter Criteria Document indicate major reductions (jc 3 miles) are not
likely to be caused by haze,, except in the summer. In 1971-75 haze apparently
caused visibility in the eastern U.S. to be less than 3 miles 2 to 12% of
the time during the summer (% occurrences with no fog, precipitation, or
blowing materials). The frequency of these episodic reductions slightly
decreased in the ensuing pentade (76-80). In other seasons, it is usually
fog or precipitation related events that produce major reductions less
than 3 miles.
4.4 Signif1 cant Visibi1ity Reductions
The effect of significant visibility reductions are somewhat more
subtle than those of major reductions. VFR operations are possible and are
common under significantly reduced visibility, and accidents do not appear
to be more common under these conditions. However, activities which depend
on long visual range will not occur. These activities include military
activities where visual contact with a "target" are necessary. Officials
at Edwards Air Force Base and the China Lake Naval Weapons Center, both
located in the Mohave Desert, have expressed concern over haze intrusions
that increasingly limit their activities; these include aviation and defense
related research, development, testing, and evaluation and all require
maintenance of excellent visibility. Aerial photography would also be
greatly curtailed or eliminated since picture quality would be degraded.
Finally, student pilot and pleasure flying would be reduced due to the need
for greater visibility for these operations.
Significant visibility reductions are frequently the result of urban
haze. Humidity is often a contributing factor, but the haze would not be
characterized as fog. This range of visibility reduction is often due to
air pollution rather than natural phenomena.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
APPENDIX D. ALTERNATIVE CONTROL STRATEGIES FOR REGIONAL HAZE
by: John Bachmann, EPA/OAR
David Joseph, NPS
Brian Mitchel, NPS
Roger Morris, DOE
Bruce Polkowsky, EPA/OAR
-------
-------
I
I
I
I
I
I
I
ii
TABLE OF CONTENTS
Page
1.0 Introduction 1
2.0 Examination of Control Strategy Alternatives Matrix 3
2.1 Approaches Based on Current Legislative and 3
Regulatory Authorities
2.1.1 Haze as Principal Objective 3
a) Secondary National Ambient Air Quality 3
Standards (SNAAQS)
b) Visibility Protection for Class I Federal 4
Areas (Phase II)
I 2.1.2 Haze as Partial Objective 7
a) PSD Sections 160-169 CAA 7
b) Interstate Air Pollution Control 8
I
I
c) International Air Pollution 10
d) New Source Performance Standards (NSPS), 11
Motor Vehicle Emission Standards (MVES)
2.1.3 Control of Haze Precursors to Meet Other Objectives 12
a) Primary National Ambient Air Quality 12
Standards (PNAAQS)
2.2.0 Modifications of Existing Authorities 14
2.2.1 Haze as Principal Objective 14
a) Codification of Integral Vistas 14
b) Modification of Section 169A 15
c) Modify the PSD Program 17
d) State Based Programs: Six Rocky Mountain 19
States Cooperative Study
I
I
I
I
I
2.2.2 Haze as Partial Objective 20
» a) Interstate Air Pollution Control 20
b) Variable Secondary National Ambient Air 21
Quality Standards (SNAAQS)
c) Episode Related Control 23
d) Revised New Source Performance Standard
_ Requirements 24
e) Requirements for Forest, Range, Agricultural
" and Other Land Management Practices 25
-------
171
2.2.3 Modifications Based on Other Objectives
a) Emissions Trading
b) Reasonable Effort to Attain SNAAQS
2.3.0 New Clean Air Act Authorities
2.3.1 Haze as Principal Objective
a) Visibility Standards
2.3.2 Haze as Partial Objective
a) S02 Rollback
. b) Age Based Control
(existing power plants)
2.3.3 New Authorities Based by Other Reasons
a) Acid Rain
3.0 Major Approaches for Further Examination
3.1 Eastern Haze Strategies
3.2 Western Haze Strategies
3.3 Issues Associated with Current Visibility
Regulations
3.3.1 Immediate Issues
3.3.2 Long-Term Strategies Under Phase I
Page
26
26
29
29
29
30
31
31
32
33
33
34
34
35
35
35
36
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX D. ALTERNATIVE CONTROL STRATEGIES FOR REGIONAL HAZE
1.0 INTRODUCTION
In considering potential long-term regulatory strategies for regional
haze, It is useful to treat the regional haze phenomena in the East and
West separately. Regional haze characteristics, intensity, sources, and
associated air quality problems are sufficiently different between West and
East that distinct control program goals and approaches should be considered
within each region. It may also be useful to consider tailoring specific
control approaches for certain source categories, sub-regions, or urban
areas in both East and West.
Given the complexity and variability of regional haze, both from a
pollution and public perception standpoint, it is questionable whether any
single regulatory approach will have the flexibility needed to provide
efficient and effective control of regional haze. Rather a combination of .
approaches - - some designed specifically to protect visibility, some
implemented to meet other air quality objectives - - may be more appropriate.
The task force examined a number of alternatives for limiting or
reducing emissions of precursors of regional haze. These alternatives
included both strategies (master plans for control) and tactics (detailed *
elements needed to implement strategy) that address this visibility problem.
The various approaches can be categorized into three classes:
1. Those capable of being implemented using the existing legislative
authorities in the Clean Air Act.
2. Modifications to existing legislative authorities that might enhance
(or diminish) their effectiveness for control regional haze.
3. Those that rely on addition of new legislative and regulatory
authorities.
The approaches also can be categorized according to environmental goals:
1) reduction of regional haze as a primary objective, 2) regional haze
control as a partial or supplemental objective, or 3) reduction of some
other adverse effect. The degree to which haze reduction is used to justify
regulatory action, then, forms a second conceptualizing dimension.
These two categorizations are used to organize alternative strategies
in Table 1. To the extent possible, each of the approaches listed within
the matrix has undergone a preliminary qualitative examination with respect
to appropriate criteria including cost-effectiveness, equity, ease of
understanding, enforceability, flexibility, compatibility with existing air
pollution control programs, and acceptability to affected parties. That
examination is summarized in the next section of this appendix. The evaluations
are highly subjective and represent the opinions of individuals based on a
qualitative understanding of general approaches. More detailed specifications
and quantitative analyses are needed for deciding on the potential usefulness
of particular variants. Two major approaches - various 502 rollbacks and
-------
TABLE D-l. MATRIX OF REGULATORY OPTIONS
AUTHORITY
GOAL
I. Haze Principally
II. Haze Secondarily
III.Other Objective
A. Existing
1) Secondary NAAQS -
fine particles, sulfates,
extinction, otheY
2) "Phase II" visibility
programs-class I areas
1) PSD Requirements
2) Interstate Pollution
Control
3) International Air
Pollution
4) NSPS, BACT/LAER major
source categories
1) Primary NAAQS
PM, S02
2) Motor Vehicle
Emissions
Standards
I
B. Modified
Existing
1) Codify integral
vistas
2) Modifications to
16 9A
3) Modify PSD
4} State based programs
1) Modify Interstate
Pollution Control
2) Regional Secondary
NAAQS; NAAQS
"Criteria"
3) Episode Control
4) Revised NSPS
requirements
5) Requirements for
"Nontraditional"
Sources
1) Emissions
trading
2) "Reasonable
Effort" for
SNAAQS
3) Modify PSD
j
I
C. New
1) Visibility
Standards
1) S02 Rollback
2) Age Based Controls
1) Acid Rain
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
targeted acid deposition scenarios have been examined more quantitatively
in the contractor analyses (See Appendix B). In addition, Section 3 of this
appendix provides observations on the major alternatives that deserve further
consideration in policy analysis and discusses outstanding issues in current
visibility regulations that could affect haze control programs in the future.
2.0 EXAMINATION OF CONTROL STRATEGY ALTERNATIVES MATRIX
2.1 Approaches Based on Current Legislative and Regulatory Authorities
2.1.1 Haze as Principal Objective
a) Secondary National Ambient Air Quality Standards (SNAAQS)
Section 109(a)(l)(B) of the Act requires the EPA to prescribe secondary
national ambient air quality standards that specify a level of air quality
which once attained and maintained will "protect the public welfare from any
known or anticipated adverse effects." Under Section 110(a)(2)(A}(ii), the
States are required to submit plans to the EPA that show attainment and maintenance
of SNAAQS in a "reasonable time."
Specific haze-based approaches include developing SNAAQS for fine particles
(<2.5um); components of fine particles such as sulfate; visibility important
precursors such as sulfur dioxide, nitrogen oxides and reactive hydrocarbons;
or atmospheric optical parameters such as extinction or-scattering coefficient.
Different limits could be set for the above, parameters and could apply to
ground-level, point or area-wide concentrations; worst-case short-term events;
or monthly, seasonal or annual averages.
Effectiveness:
This approach can be effective depending on the regulatory goals of the
program and the specific area or region of interest. The secondary standard
could be effective in protecting currently good or excellent visibility, preventing
undesirable visibility impairment or improving existing visibility in those
regions where it is currently impaired. However, a single nationally applicable
welfare standard could not be effective at achieving all such goals in all
areas of the country such as western and eastern urban areas or pristine class
I areas. The extent of visibility impairment and the value people place on
visibility vary widely with affected populations, regions of the country and
settings within each region. A single SNAAQS could not reasonably address all
facets of the visibility problem equally. In particular a SNAAQS that would
maintain current visibility levels in the rural West might require lower than
natural background levels in the East. The level of any SNAAQS might best be
directed at establishing desirable visibility goals for those regions in the
East affected by large scale regional haze of multistate origin and those major
western urban centers affected by haze predominately of local origin. The
visibility in pristine class I areas would have to be protected- via some other
mechanism, e.g., the PSD program, Section 169A, etc.
-------
Simplicity:
The SNAAQS approach is relatively simple conceptually and understandable
to the regulators, industry, environmentalists and the general public, but
would be complext to implement. The decision on selecting an acceptable national
level could be difficult. A'SNAAQS based on PM-2.5 or atmospheric optical
conditions would require a national visibility or fine particle/secondary
aerosol monitoring network and corresponding modeling capability.
Cost:
The cost of design, implementation and enforcement of this approach would
depend on the standard level. In any event, monitoring network implementation
and development of modeling capability may be initially costly. To the extent
substantial additional controls are required, control strategy design,
implementation and enforcement are likely to be costly.
Enforceability:
The enforceability would depend on the implementation alternatives chosen.
The traditional approach of predictive modeling and monitoring would be
significantly hampered by available tools and the long range transport problem.
Flexibility:
This approach is a relatively inflexible one in that it cannot address all
problems and differences between geographic areas and vistas within a given
area. Modifications to the Act {See 2.2.2.b and 2.2.3.b) could add flexibility,
but at a cost to simplicity.
Compatibility:
Regulating regional haze with a SNAAQS would very compatible with existing
approaches for regulating air quality.
Acceptability:
The acceptability of this approach is problematic and depends greatly on
the nature of the standard, its goal, relation to other approaches, and
implementation alternatives. A single national level cannot be expected
to deal with both eastern and western visibility.
b) Visibility Protection for Class I Federal Areas (Phase II)
Section 169A of the Clean Air Act establishes'the national goal of remedying
existing and preventing future impairment of visibility in class I federal
areas resulting from man-made air pollution. The mechanisms within the Section
169A program include 1) the requirement for certain existing sources to reduce
their contribution to existing visibility impairment by installing the Best
Available Retrofit Technology (BART), 2} the long-term strategy that States
must develop~to make progress in meeting the national goal, 3) the review of
new sources' impacts on visibility in federal class I areas, and 4) integral
vistas, which encompass selective protection of visibility beyond class I area
boundaries.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
EPA promulgated regulations under Section 169A.in December 1980. These
requirements dealt with plumes and other pollution that were "reasonably
attributable" to specific sources. {These requirements were referred to as the
"Phase I"'requirements.) In October 1984, EPA proposed to implement parts of
these regulations (monitoring and new source review) in those States that do
not adopt their own implementation procedures.
Action on "Phase II" (regional haze} visibility protection was originally
deferred to permit development of improved scientific and technical bases for
action. The provisions of this section could, with appropriate regulatory
development and technical guidance, be applied to visibility impairment in
class'I areas caused by regional haze in the West. It would be more difficult
to justify the use of this authority as a principal means for addressing haze
in the East.
Effectiveness: '
The effectiveness of this approach relies heavily on how well visibility
research programs can reduce the uncertainties in areas such as predicting the
formation of secondary aerosols under varying meteorological conditions; the
estimation of transport and dispersion parameters in areas of complex terrain;
the predicition of the impact of single or multiple sources on a regional
scale; the determinations of whether incremental changes in contrast or color
will be perceptible; and visibility monitoring.
The application of BART will.be restricted to those major sources for
which there is good evidence linking them to noticable visibility-impacts,
which meet the age requirement (began operation during the period from August*1
1962 to August.1977) and the control of which can be expected to result in a
significant improvement in visibility. The application of BART in phase I will
therefore have little effect on regional haze. Even if applied to regional
haze, the number of sources affected would be limited. BART would not be
applicable to source categories such as agricultural activities, prescription
fires, copper smelters, and most urban pollution (motor vehicle emissions,
space, heating, etc.)
The requirement for a "long-term (10-15 years) strategy for making
reasonable progress toward meeting the national visibility goal may provide
additional authority to implement regional haze presursor. control, particularly,
in the West where a large number of class I areas are located. Specific
mechanisms for such application are not specified in the Clean Air Act, and
substantial regulatory guidance would be required. In its development of a
long-term .strategy, the State can consider additional measures for remedying
existing visibility impairment in mandatory class I areas, such as adopting
additional emission limitations and schedules for compliance for uncontrolled
or poorly controlled sources not covered by BART.
The review of new sources' impacts on visibility in federal class I areas
can provide a mechanism for controlling a new source's contribution to regional
haze. However, the capability for predicting visibility impacts over regional
scales must be developed adequately.
-------
Selective integral vistas protection would provide for greater geographical
coverage of the Phase II regional haze rules and may provide protection for
some areas between class I areas.
Simplicity:
The Phase II approach, if it builds on ongoing SIP programs, could be
relatively understandable to the regulators and affected parties. Given the
complexity of the regional scale phenomena, however, multistate coordination
will be required. Improved predictive modeling and monitoring capability would
greatly enhance the effectiveness of such an approach.
Cost:
The cost of the design and implementation of this program would be
relatively high. Monitoring network implementation and development of modeling
capability can be costly especially in an enforcement mode. An individual
source's contribution to the visibility problem and the predicted effectiveness
of control options must be analyzed prior to enforcement. Such tasks may be
resource intensive.
Enforceability:
The enforceability would depend on the effectiveness of the modeling
approaches adopted and the extent and reliability of the monitoring data.
Flexibility:
This approach is relatively flexible for class I area visibility impairment
problems. Many but not all visibility impairing sources would be regulated
under this approach. The long-term strategy is a flexible mechanism that can
adapt to difficult impairment problems. Urban areas with haze problems, but
which do not impact class I areas would not be subject to control under Phase
II.
Compatibility:
Phase II rules could be made compatible with the Phase I visibility program
and the State Implementation Plan but would represent a substantial expansion
of concerns.and coverage. Section.3.3 discusses Phase I issues that should be
addressed prior to a Phase II program.
Acceptability:
The acceptability of this approach would depend on the general acceptability
of the monitoring data and regional haze modeling approaches. Source/receptor
relationships must be better understood and documented.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2.1.2 Haze as Partial Objective
a) PSD - Sections 160-169 CAA
The PSD program is designed to protect and preserve clean areas of the
United States and its territories by controlling the amount of air quality
deterioration allowed in these clean areas. This strategy is implemented ,
primarily via a preconstruction permit program for major and modified sources
of air pollution and a clean" air area classification system which places clean
air areas into one of three classes, each with progressively more stringent
levels of allowable air quality deterioration. The key provisions within the
PSD program that are of interest to the regional haze problem are: 1) the PSD
sulfur dioxide (S02) and paniculate matter increments for the three classes of
areas; 2) the requirement that PSD sources reduce air pollutant emissions :>y
the application of the Best Available Control Technology (BACT); 3) the PSD
provision that allows States to redesignate areas to class I, the most protective
area classification; and 4) the provisions giving the Federal Land Managers
(FLMs) the responsibility and role in the PSD program to protect the air quality
related values (AQRVs), including visibility, of certain class I federal areas.
The net effect of the PSD program, including the present visibility review
requirements, is to limit regional emission increases for certain major emission
source categories.
Effectiveness:
B
This approach is only partially effective but it does have the potential
for improvement with respect to protecting and enhancing visual air quality.
In order to make the PSD program more effective in protecting visibility, ^
stronger emphasis must be placed on class I area air quality protection as a
major goal of the program with appropriate, supporting policy guidance and
regulatory structure. Several aspects of the PSD program could be improved
through this interpretation of the goals of PSD. These aspects generally fall
into two categories: procedural and reguattory.
The procedural aspects to be considered include: 1) better implementation
of the existing requirements to notify affected FLMs; 2) additional notification
requirements which will involve the FLMs earlier in the permit review process;
and 3) a mechanism to ensure interstate cooperation in regard to notifying and
encouraging participation by State air quality officials and FLMs in States
impacted by PSD permit activities in antoher State. Suggested additional re-
quirements for FLM notification might involve: 1) advance notification
whenever the permitting authority receives such advance notice, such as the
scheduling of a preapplication meeting with the applicant; 2) consultation
with the affected FLMs on the completeness of the permit application as it
relates to impacts on class I areas;, and 3) development of impact/source
s.ize/distance criteria to guide the determination of which FLMs to notify.
The regulatory aspects to be considered include: 1) requiring a cumulative
visibility impact analysis in permit applications that incorporate all existing
and permitted sources of visibility impairing pollutants (or presursors to
those pollutants); 2) develop PSD increments for other pollutants for which
NAAQS exist and which are visibility impairing pollutants or precursors to
those pollutants; and 3} explicitly requiring visual air quality monitoring by
PSD applicants.
-------
Simplicity:
Historically, the PSD program is viewed as quite complex. However, given
the experience gained over the life of the program, it has become more understand-
able to the regulatory agencies, industry, and public interest groups. It
would be somewhat more simple matter to incorporate these additional visibility
aspects into the existing program, than to create a parallel visibility review
program.
Cost:
The additional costs involved could be incremental to the existing costs
of implementing the.PSO program, requiring additional monitoring and modeling.
Enforceability:.
Enforceability would be comparable to that under the existing PSD program.
It will depend upon the legal interpretation of the goals of the PSD program,
the construction of any new regulatory language, the proper implementation of
the program, and establishment of enforceable permit conditions.
Flexibility:
This approach is flexible in that all PSD determinations are made on a
case-by-case basis. New increments, however, would not be flexible.
Compatibility:
This approach is compatible with the existing PSD program; the new
permitting procedures could be handled under current frameworks.
Acceptability:
Acceptability will depend to some extent upon the development of approved
analysis techniques for predicting and monitoring visibility. The procedural
aspects may be reluctantly acceptable due to the increased administrative burden
placed on permitting agencies.
b} Interstate Air Pollution Control
Subsection 110(a)(2)(E)(i)(II) of the Act prohibits one State from
"interfering" with visibility protection plans of another State. Section 126
requires that nearby States be notified of all new, modified and existing
major sources which could so "interfere." Subsection 165(d) (concerning air
quality values, including visibility) also requires that Federal Land Managers
be informed of plans for new sources which
Section 126 further provides for States to
disallow a new source or modification which would'"interfere" or to require a
SIP revision where an existing source "interferes." This portion of the Act
as currently written would be most effectively considered in the context of
implementing a SNAAQS or Phase I.class I area visibility protection program.
could affect their Class I areas.
petition the Administrator to
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Effectiveness:
This approach is currently constrained in its effectiveness to protect visi-
bility by both legal and technical considerations. This .is demonstrated at
least in part by an Environmental Protection Agency decision in 1984 on inter-.
state air pollution petitions which were filed in 1980 and 1981 under the
provisions of Section 126. Based on the decision reached in consideration of
these petitions, it appears that States must have approved visibility protection
plans for regional haze before Section 110 and 126 can be used as vehicles for
addressing haze problems. Transboundary pollutants which may be responsible
for causing regional haze impacts appaently can only be controlled through the
petition mechanism of this time if these pollutants are "interfering" with
other requirements contained in State plans, such as compliance with PSD incre-
ments or NAAQS.
Even absent the legal question, there is still a lack of technical tools
and thorough understanding of the transformation and transport processes
responsible for regional haze. In order to effectively analyze and mitigate
regional haze impacts utilizing'the interstate air pollution control provisions
of the Clean Air Act, more research and development will be needed to address
these problems.
Simplicity:
This approach would not be simple to implement. Exercising the provisions
of Section lib and 126 with respect to the "interference" clauses requires the
affected State(s) to assume a significant burden of proof in demonstrating the
prohibited impacts coming from the accused State(s). Such an approach is, by;-
its nature, reacting to an existing problem and not preventative. It could also
be adversarial and might involve arbitration or court settlement to reach a
decision. The notification provision under Section 165 is a simple procedural
permitting activity which is ongoing. Some improvements to simplify this
section are offered in a subsequent section.
Cost:
This approach is likely, to be costly in terms of money,
resources regarding activities under Sections 110 and 126.
Enforceability:
time, and agency
It would be hard to enforce this approach pursuant to Section 110 and 126
due to potential legal recource and the technical uncertainties that currently
exist. It is also hard to track the' compliance with Federal Land Manager
notification procedures under Section 165.
Flexibility:.
The approach under Sections 110 and 126 as written appears to depend on
source/receptor specificity that is not likely to be achieved in the near
future. In that sense, it is not flexible for dealing with regional scale
transport.
-------
c) International Air Pollution
This approach is extremely flexible, perhaps at the expense of being
effective.
Compatibility:
I
10
Compatibility:
As applied to regional haze, this approach would not be compatible with I
existing air quality programs.
Acceptability: ' I
This approach is. partially acceptable in that it provides a mechanism for
affected parties to seek corrective action under Sections 110 and 126 and for I
Federal Land Managers to be apprised of impending permit actions. Its overall I
acceptibility is, however, adversely affected because of its inherent limitations.
I
Section 115 of the Act provides authority to call for States to control U.S. _
pollution that may endanger health and welfare in a foreign country, when that
country has reciprocal legislation with respect to the U.S. The complexity of "
source receptor relationships may limit the usefulness of this authority for
transboundry transport of haze between the U.S. annd Canada, but use of this
aauthority or other bilateral discussions regarding impacts of emissions from |
several smelters near the U.S. -Mexico border may be more tractable.
Effectiveness: . I
Partially effective as discussed above. The potential exists for greater
effectiveness given future advances in the understanding of source receptor
relationships and improvements predictive and analytical technology. I
Cost:
Low to moderate, depending upon the level of interest given to any
particular international air pollution problem.
Enforceability: . I
Difficult to enforce, givern the tenuous nature of international agreements
and jurisdictional aspects which would complicate any definitive action. g
Flexibility:
I
This approach should be compatible with existing air pollution control _
programs, but permits regulations of effects for which no other standards exist. I
Acceptability:
Provided there are benefits to all international parties engaging in problem |
resolution, this approach should be acceptable. It is likely that the affected
localities may find the apprach the least acceptable.
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
d) New Source Performance Standards (NSPS), Motor Vehicle Emission
Standards (MVESj
NSPS for new stationary sources and MVES for vehicles apply to new emission
sources and are generically called "technology related." Such emission standards
are selected on the basis of technology availability and cost-effectiveness and
apply nationwide. NSPS must be reviewed every four years and revised if superior
technology has become available. NSPS requirements are supplemented by BACT
determinations under PSD and Lowest Achievable Emission Reductions (LAER)
under new source review where NAAQS are violated.
In the long term, NSPS and MVES are the principal means presently used to
control regional air pollution, like haze. The retirement/replacement process
that ultimately reduces emissions operates over the long run. This may be
fairly rapid (10 years for vehicles), intermediate (20 years for smaller boilers
and many processes) or quite slow (up to 60 years for utility boilers). Given
that NSPS will eventually limit haze precursors, the key policy issues related
to the intermediate time domain are: Will NSPS/MVES operate with acceptable
speed, and if not, how could their operation be hastened or supplemented?
Effectiveness:
These approaches are effective over the long term as older, minimally
controlled sources are replaced by newer, better controlled sources. However-,
the degree of effectiveness will depend upon several factors such as the respective
rates of existing source retirements/replacements and new source growth, the
siting locations of new stationary sources, and the recognition and control of-
those emissions which are ultimately responsible for haze formation. The
effectiveness of this approach can be improved if emissions control require-
ments are considered with visibility protection as a goal. On the other hand,
the effectiveness may be limited by the inability of this approach to prevent
eventual increases in total visibilityrreducing emissions within a particular
region or area. In such situations, PSD (BACT) and LAER decisions can be used
to tailor requirements for specific areas.
Simplicity:
This approach is simple to implement in that both the NSPS and MVES programs
are ongoing and fromilar. Any new requirements could simply be added to the
existing regulations.
Cost:
In terms of program costs to the administering agencies, there should be
little additional costs since these are existing programs. Some additional
costs would be incurred by the regulated industry and the general public,
however, if,NSPS are developed for currently unregulated industry or pollutants
and if stricter MVES are adopted. Also, front-end costs would.be necessary to
determine the effectiveness of implementing new control requirements to achieve
public health and welfare goals.
-------
12
Enforceability:
This approach would be enforced through existing compliance mechanisms.
These include: 1) requiring stationary sources to install, operate, and
maintain control technology to minimize emissions of the affected pollutants;
2} requiring these same sources to install, operate, and maintain continuous
emissions monitoring systems and to report the monitoring data; 3) performing
compliance testing of these sources to verify industry reports of compliance to
the control agencies; and 4) implementation of vehicle inspection and maintenance
programs for compliance of any MVES.
Flexibility:'
This approach is flexible to the extent that it allows sources subject to
NSPS to employ any control technology they choose in order to meet the required
emission limits. Vehicle manufacturers may have fewer options under strict
MVES.
Compatibility:
this approach is compatible with existing regulatory programs.
Acceptability:
This approach would be acceptable to the implementing agencies due to the
familiarity with the programs and the simplicity of their implementation. It
is likely to be less acceptable to the affected industries due to stricter
regulation and associated increase in control costs.
2.1.3 Control of Haze Precursors to Meet Other Objectives
a) Primary National Ambient Air Quality Standards (PNAAQSl
Sections 109(a)(l)(B) and (b)(l) of the Act require the EPA to prescribe
national primary ambient air quality standards "the attainment and maintenance
of which ... are requisite to protect the public health." PNAAQS protect the
public health by limiting the ambient concentrations of SOg, N02, 03, particulate
matter and lead. Control programs designed to attain and maintain these ambient
concentration levels may also have an impact on reducing the frequency and
intensity of haze. If, in the future, a PNAAQS for fine particles (e.g., PM-
2.5) or sulfates is supported by health data, reductions in regional haze might
be substantial. A more stringent primary S02 standard could also result
in haze reductions.
Effectiveness
The past fifteen years have been marked by significant reductions in
emissions of particulate matter, S02, NOg, VOC and reductions i-n ambient 03
concentrations resulting from the continuing*effort at achieving and maintaining
the primary national ambient air quality standards. In the absence, of such
progress the frequency, intensity, and extent of regional haze would surely be
significanly greater than it is at present.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13
The existing set of PNAAQS do not effectively control important visibility
precursors and secondary aerosals such as fine particulate matter (PM-2.5) and
sulfates, nor address the long range transport of visibility precursors from
polluted urban areas to more pristine ones.
A primary NAAQS for a visibility important precursor, if justifiable on
health grounds, would be more effective at reducing haze problems in western
urban areas; the northwest; and the east. Control programs developed in response
to such PNAAQS would be effective because they would .utimately result in lower
ambient levels of visibility impairing precursors. A major issue of this approach
would be whether sufficient information is available to justify a health related
standard and when such standard setting could occur. EPA is currently in the
process of making revisions to the primary and secondary NAAQS for PM. During
March 1984, EPA proposed to replace the TSP PNAAQS with a standard based on
PM-10. With this revision process as of yet unfinished, it appears premature
to expect any initiative to create a health based PM-2.5 NAAQS within the next
five years. The review of the S02 PNAAQS is, however, not yet completed;-
Any standard established for ground-level concentrations may not protect
against all regional or layered hazes. An elevated haze layer could be readily
perceptible and yet ground-level fine particulate concentrations could be near
zero.
Simplicity:
This approach is simple and easy to understand and is identical or similar
to existing ambient air quality standards and proposed PM-10 standards. PNAAQS
for PM-2.5 and/or sulfates would require specialized monitoring and modeling
capabilities and implementation.
Cost:
Standard setting process, establishment and implementation of a monitoring
network and the development of a modeling capability may be fairly costly.
Enforceability:
The enforceability would depend on the effectiveness of the modeling
approaches and the extend and reliability of the monitoring data.
Flexibility:
The PNAAQS strategy is a relatively inflexible one because it cannot be
adapted to all problems, differences between geographic areas and vistas within
a given area. -
Compatibility: ' - " -
This approach would-be compatible with existing air quality standards and
proposed PM-10 fine particle standards." ' - '
-------
I
I
14
Acceptability:
Concerns about the adequacy of modeling and monitoring science or the |
justifiability of a health based standard for PM-2.5 or sulfates could affect
the acceptability of this approach. _
2.2.0 Modifications of Existing Authorities . "
2.2.1 Haze as Principal Objective
a) Codification of Integral Vistas
The Act, in 1977, established a national goal of protecting certain
national parks and other areas of national significance from visibility
impairment.resulting from industrial pollution. Section 169A of the Act
required EPA to promulgate regulations as guidelines to the States for
implementing a visibility protection program. The agency published these
regulations in December 1980. These regulations give the States the discretion
of extending the visibility protection to views perceived from within
certain national parks and other federal areas of specific landmarks or I
panoramas located outside the boundary of the area. These views are called
"integral vistas". Although it is the State that determines how much
protection (if any) to afford these integral vistas, it is the Federal Land
Manager that must identify and list them for the State's consideration.
The concept of integral vistas has been the subject of legal challenge.
This regulatory uncertainty could be eliminated by Congress including provisions |
for the listing and State-selective protection of integral vistas in amendments
to the Clean Air Act.
Effectiveness:
Congressional action to incorporate integral vistas into the Clean Air
Act would eliminate the uncertainty of this mechanism and encourage its
implementation. Unless authorized differently by Congress, the integral
vistas could be applied selectively to prevent or remedy visibility impairment
that was reasonably attributable to a source or group of sources. Under
this Phase I approach, regional haze would not be regulated. If regional
haze impairment of integral vistas were to be considered, this mechanism
would provide the means of States to balance the need for correcting or
preventing the vista haze problems with other factor such as economics and
energy considerations. Such protection would not be automatic and would
not be applicable to all lands external to class I areas, only those areas
included within the integral vistas. Certain important State lands not
designated as class I, where visibility nonetheless is an important value,
may not receive protection under this approach unless the State incorporates
special protection provisions in its own rulemaking. Given the greater
density of class I areas and potential integral vistas in the West, and
because regional haze tends to occur on large geographical scales, this
approach would be more effective at protecting the intervening spaces
between class I areas in the West rather than the East. It is questionable
this approach should be considered in remedying or preventing urban haze
caused by urban area emissions. Opting for the protection of an integral
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
15
vista, with the resulting economic and social costs of further reductions in
emissions in urban areas, may be a very difficult decision for a State to make
and enforce.
Simplicity: '
This integral vista approach is simple in principle. In practice, it may be
very difficult to regulate haze in this manner because of the difficulty States
may experience in performing the benefits/cost analysis needed to justify how
much (if any) protection is afforded integral vistas. Ambient monitoring as
well as source monitoring will be required to determine the origin of the haze.
Cost:
The cost of the design and implementation of this approach would be relatively
low. However, the monitoring and modeling support required by this approach
may be costly.
Enforceability:
The enforceability would depend on the effectiveness of the modeling
approaches, the extent and reliability of the monitoring data and the States
resolve to balance the need for industrial growth with that of resource protection.
Flexibility:
Class I areas with sensitive vistas could be addressed with this approach,
and states are given substantial flexibility. -
Compatibility:
This approach would be compatible with existing air quality regulations
and standards.
Acceptability:
Protection of integral vistas would probably be accepted eventually providing
that such protection is not mandated but rather part of a process that evaluates
alternative goals or objectives for the lands included in these vistas.
b) Modification of Section 169A
Section of 169A of-the Act could be strengthened or weakened with respect
to regional.haze. As noted above, the 1980.EPA regulations implementing
Section 169A distinguished between identifiable plumes and regional haze. The
decision to include regional haze type visibility impairment in the Section
169A visibility protection program was deferred until a later Phase II. Phase II
would be proposed and promulgated when improvements in monitoring techniques
provide more data on source-specific levels of visibility impairment, regional
scale models become refined, and our scientific knowledge about the;relation-
ships between emitted air pollutants and visibility impairment improves.
Section 169A could be modified by Congress to encourage the application of
the program to regional and urban haze; to exclude haze from Section 169A and
address it elsewhere; or could be left in its present form.
-------
16
Effectiveness:
As noted, left in its present form, 169A could be an effective approach
for addressing western visibility. Exclusion of haze from 169A and addressing
haze with acid deposition might be useful in the East, but of questionable
effectiveness in the West. A reauthorization by Congress could be a very
effective means for accelerating the development of programs for addressing the
haze problems,in the West. For example, Congress could establish a firm date
for the implementation of Phase II visibility protection. Such a statutory
deadline would create a high priority for haze research and would encourage
research monies to be directed in the areas of improving the science of regional
visibility modeling and monitoring. Other modifications that would be of use are
(1) increasing the subset of existing sources subject to BART, such as pre-1962
sources, (2) addressing the role of smoke management in preventing of remedying
haze problems caused by prescription burns and other land management practices,
(3) requiring regional haze to be analyzed in the new source review PSD process,
(4) specifically subjecting urban haze impairment to control and remedy via
this section, and (5) broading the geographical scope of this section to encompass
more than the federal class I areas.
Simplicity:
The concept of such an approach could be straightforward and understandable
to regulators, environmentalists, industry and the general public. Implementation,
however, must still address complex multi-source regional problems.
Cost:
The cost of design, implementation and enforcement be relatively high.
Determining the major, contributions to the haze would require expensive monitoring
and modeling. Enforcement would be expensive if smelters or urban areas were
the major contributions especially if such sources were located hundreds of
kilometers from the geographical areas with the haze problems.
Enforceability:
Enforceability would be difficult for urban area sources and older existing
sources that could not afford retrofit controls.
Flexibility:
The current Section 169A approach is very flexible one that provides a
consistent framework to evaluate different types of visibility impairment, in
different regions, and caused by a variety of sources. A modified approach
might reduce flexibility, but increase effectiveness.
Compatibility:
If this approach expands to cover class II areas, it would be inconsistant
with existing approaches for regulating air quality and visibility impairment.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
17
Acceptability:
Those urban areas or industries affected by this approach would probably
not find the controls required acceptable and challenge such rules in court.
Public comments from environmentalists and industry question both the desirablility
of expanding to class II areas and the need to reauthorize 169A to deal with
haze. ,
c) Modify the PSD Program
A separate task force is examining the PSD program and is discussing
modifications motivated by objectives other than haze. These should be examined
for haze implications. The PSD program could be modified to increase its
effectiveness at addressing regional haze problems.. Such modifications could
include:
1} Replacing the PSD increments with new ones based on PM-2.5 (instead of TSP).
2) Mandating short-term PSD increments for other pollutants that affect
visibility;
3} Allowing BACT determination to consider mitigating regional haze and impacts
on AQRVs in addition to economics, and energy impacts;
"i
4) Providing specific language enabling regional haze effects to be addressed
in class I and class II areas;
5) Requiring States and Federal Land Managers to set state wide, regional or
local visibility objectives or goals wich would be implemented through PSD;
6} Including review and control of "minor" air pollution sources which emit
pollutants that affect visibility.
Effectiveness:
The overall effectiveness in reducing or preventing regional haze would be
enhanced if the above modifications were made to the PSD-program.
Fine particles are commonly defined as particulate matter which is smaller
than 2.5 micrometers in diameter, and referred to as PM-2.5. Relatively high
concentrations of fine particles are primarily responsible for regional haze in
the West. Lowering the magnitude of the PSD particulate increments and redefining
them .to be based on PM-2.5 rather than total suspended participates would, focus
more control on the amounts of both those fine particulates that are directly
emitted and those that are formed in the atmosphere (e.g., sulfates from precursor
pollutants) (e.g., sulfur dioxide}. Exceedances of these PM-2.5 increments
would be addressed via the PSD'permit process or through a revision of the
State's SIP which could require reductions in emissions from existing sources.
Promulgating short-term'PSD increments for other pollutants that affect
visibility and haze would, like the PM-2.5 approach above, provide a direct
means for regulating haze producing emissions in areas where PSD is .applicable.
-------
I
18
Increment could be established for pollutants such as sulfates or optical air
quality indicators such as contrast.
Allowing the determinations of BACT to consider environmental impacts such
as regional haze and impacts on class I area air quality related values, could
increase the stringency of control technology requirements for PSD sources and
encourage additional reduction of haze producing pollutants. Presently, BACT |
represents the maximum degree of emission reduction achievable taking into
account energy, environmental and economic impacts and other costs.
Language could be added to the PSD statute that would enable regional haze
effects on class I and II areas to be addressed. Such added emphasis on regional
haze would encourage better analyses of haze to be performed during the PSD
permit review. I
A fifth modification of PSD would allow the State and/or Federal Land Managers
to set state-wide, regional or local visibility objectives or goals which would
be implemented through the PSD program. A visibility analysis for a proposed
new source would consider whether the new source impact is consistent with the
applicable objectives or goals with respect to haze, perceptible plumes or I
atmospheric discoloration. If the visibility analysis indicates an unacceptable I
impact relative to the objectives, the permit could be denied; the source could
relocate to more acceptable locations where meteorology,and terrain are more
favorable; the source could improve the emission controls; or the source could I
scale down the size of the project to reduce emissions.
Lastly, the effectiveness of the PSD program at addressing haze could be I
improved by broadening the coverage of PSD review to other source categories
which emit pollutants that affect visibility. Certain "minor" sources, general
urban development and smaller populations centers near class I areas that are
created in response to major industry could be reviewed under the PSD program. |
Simplicity:
These modifications represent major extensions to the existing PSD program
which has been in effect for over a decade. They would probably add to the
complexity of program. '
Cost:
The cost of design, implementation and enforcement of PSD will increase if I
these modifications are made. Establishing new increments, developing visibility
goals and objectives and the PSD review of additional sources are all resource
intensive activities. I
Enforceability:
These modifications would be as enforceable as the existing PSD provisions I
provided tht adequate modeling and monitoring tools .are available to characterize
the haze source/receptor relationships. .
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
19
Flexibility:
This approach would not be flexible enough to address the significant haze
problems associated with many urban nonattainment areas. The modifications are
better suited for managing visibility problems of class I and class II attainment
areas where the PSD provisions are applicable.
Compatibility:
Modifying the PSD program to better regulate regional haze would create a new
program that would be compatible with the existing PSD approach.
Acceptability:
Modifications such as establishment of new or revised increments, visibility
objectives, more stringent BACT determinations, would probably not be very well
accepted by industry and would be the subject of legal and technical challenge.
d) State Based Programs: Six Rocky MountainStates Cooperative Study
States would have the opportunity to develop a visibility protection
program, choose quantitative visibility parameters and standards, designate
specific areas or vistas requiring protection, and choose the most cost-effective
strategy in the absence of national program requirements. The States of Colorado,
Idaho, Montana, New Mexico, Utah and Wyoming cooperated on a study with the
National Park Service and U.S. Forest Service to examine such an approach. It
was the intent of these States to consider the feasibility and desirability of
a consistent region-wide approach to visibility protection.
Effectiveness:
This approach has the potential to be effective depending upon the degree to
which individual states develop and adopt visibility standards. Without specific
national requirements for adoption of state plans, however, it is doubtful that
all western states would adopt a program to control regional haze. Overall
effectiveness would be enhanced if there is comprehensive coverage of visibility
control programs.
Simplicity:
The concept of establishing and implementing a standard is easy to understand
but the establishment of an aesthetic visibility standard may not be a simple task.
Implementation of this standard may also be difficult due to the need for
acceptable modeling techniques and, monitoring methods.
Cost:
The cost of this approach would be less than that of a national program.
Much of the costs would be incurred during the planning and development stages
of standard setting. Control costs'may also be significant for both .existing
sources which must reduce emissions to make progress toward attaining a standard
and for new sources which may affect areas subject to a standard.
-------
20
Enforceability:
The enforcement of applicable standards will depend upon several factors.
These factors include: 1) the degree to which states cooperate on adopting a
consistant form of the standard; 2) the degree to which interstate cooperation
on visibility impacts occurs; and 3) the adequacy and acceptablity of modeling
techniques and monitoring methods. The lack of nationally applicable, federally
enforceable standards may also hamper strict enforcement, especially with
regard to interstate impacts.
Flexibility: .
This approach -is flexibile in that it allows for consideration of regional,
area and vista-specific goals in standard setting.
Compatibility:
Standard adoption and implementation should be compatible with existing
air quality protection programs.
Acceptability:
Acceptablity of this approach will be greater if ample opportunity is
allowed for public participation and review during the standard setting process.
Further development of acceptable models and monitoring methods will also
increase the acceptability of this approach.
2.2.2 Haze as Partial Objective
a) Interstate Air Pollution Control
"Regional" or "transported" air pollutants include fine PM, ozone (and
other oxidants), and acidifying agents (SOX, NOX). Regional air pollution
occurs over large (interstate) areas and is difficult to attribute to specific
sources except in a statistical sense. Modifications to Section 126 could
increase the flexibility to deal with interstate transport when it is not
possible to pinpoint impacts of specific sources. This could enhance the
implementation of a number of air quality management approaches in dealing
with regional transport, including NAAQS, 169A, PSD, International Air Pollution,
as well as modifications to these authorities.
Effectiveness:
The modifications under discussion here include:
1) retain 126 as is for identifiable sources and tight clusters of
sources,
2) replace "interfere with" and "prevent" in' 110(a)(Z)(E) .with
"substantially interfere with",
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
21
3) add a new subject addressing long range transport (>50 km) from
multiple sources
a) acknowledge problems with specific source,attribution, but empower
limits on emissions from a state or a large portion thereof,
b) provide authority to impose reasonable conditions on SIP acceptibility
to mitigate or prevent regional pollution,
c} provide a mechanism for prior consultation between EPA and states
involved, and
d) define or provide for defining "substantial interferences."
Such modifications could provide an effective tool for implementing regional
haze and other air quality management approaches by recognizing difficulties in
source-receptor relationships permitting reasonable actions, and providing
mechanisms for conflict resolution.
Simplicity:
As with any air quality management based approach, this would not be
simple to implement.
Cost:
The cost of this supplemental approach depends on the air quality
management approach it is used to implement.
Enforceability:
As with other air quality based approaches, enforceability is limited
by the availability of technical tools to support this kind of regulatory action.
Flexibility:
This approach is flexible in that it allows for "reasonable" measures and
prior states involvement. .
Compatibility:
This approach could affect compatibility between regional programs and
existing air quality programs.
Acceptability:
This approach is partially acceptable in that it provides an improved
mechanism for affected States to seek correcitve action for a problem over
which they would otherwise have no direct control. Its overall acceptability
depends on the program it is used to implement.
b} Variable Secondary National Ambient Air Quality Standards (SNAAQS)
Visibility is better in arid climates than in humid ones. A-small
addition of pollution-derived fine particlues produces a greater effect in an
area with naturally high visual ranges than in an area with lower visual ranges.
A single nationally applicable SNAAQS for haze that would maintain currently
good western visibility could be unnecessarily stringent for the eastern
-------
I
22
United States where the natural background of fine particles may exceed m
current loadings in the West. A similar imbalance could result if a SNAAQS
-for ozone or S02 were based on crop protection where sensitive crops were
region-specific rather than ubiquitous. A more effective approach could be I
revision of Section 109 (b)(2) of the Act to allow the welfare related Standards
to be region specific or even variable within a region. A second possible
NAAQS change would permit cost/benefit considerations to be considered in I
establishing SNAAQS.
Effectiveness:
The regional standard approach could be more effective than the single
nationally applicable SNAAQS. The regulatory goals of such a program could _
more realistically reflect the diverse nature of haze problems in the United I
States. The variable SNAAQS could be designed to protect currently good or '
excellent visibility in the southwest, for example, or prevent undesirable
visibility impairment, or improve existing visibility in those regions where it
is currently impaired. . |
The existence of regional specific SNAAQS could facilitate haze control
being incorporated in urban air quality management. This approach would be I
useful in addressing important haze issues in areas other than class I designated *
areas and in multistate areas. Use of costs to establish SNAAQS would put
great weight on uncertain benefits and costs estimates with the possibility of
standard levels becoming more stringent as lower cost controls were developed. I
To the extent that'national uniform standards are used, the "balance" could be
suboptional. Such an approach could, however, permit establishing practical
goals and more aggressive implementation.
Simplicity:
A variable SNAAQS approach would be more difficult and complicated to *
establish than setting of a single SNAAQS. Division of the nation into
potentially numerous regions would be particularly difficult.. In effect, the
standard setting process would have to be duplicated in each region. I
The development and articulation of regional and area visibility goals, _
objectives and numerical limits would demand the participation of industry I
Federal Land Management.agencies, State and federal agencies, and "environmental
groups. The federal government would have the responsibility for promulgating
these regional standards. States may argue that such a task is best accomplished
by the State rulemaking process, especially if the region in question is totally |
situated within one State. Interstate haze issues could be dealt with better
at the federal level. _
Cost: . «
Resources for setting variable SNAAQS would be greater in than in getting
a single standard. Once promulgated, the standards would require resources for |
implementation and enforcement that would be of the same order of magnitude as
other strategies analyzed. The cost of monitoring and modeling for the purposes H
of establishing the sources' contributions to haze and the degree of improvement I
that would result from implementing source emissions controls would be relatively
expensive and resource intensive. Use of cost/balancing approaches in standard
I
I
-------
'f
1
1
1
1
1
1
1
1
1
1
7
23
setting could lower costs of control, but it is not clear that it offers
substantial benefits in this regard over considering costs in implementation.
Enforceability:
The enforceability would depend on the adequacy and acceptability of the
modeling methodology and monitoring data.
Flexibility:
This approach is relatively flexible because of 'the ability to tailor the
standards on a regional or subregional basis. Protection of specific vistas
in class I areas is probably best handled with other mechanisms. Use of
national cost/benefit balancing could decrease flexibility of implementing
SNAAQS, but increase flexibility in standard setting.
Compatibility:
Variable SNAAQS would be compatible with the existing air quality management
programs for the existing primary and secondary NAAQS.
Acceptability:
Given sufficient opportunity for participation and review during the standard
setting process, those regulated may find the resulting standards acceptable.
The acceptability will depend strongly on the adequacy of the monitoring and
modeling tools. It may be argued, however, that this approach departs from the
traditional thrust of the Act, which encourages uniform national standards, and
discourages development of "havens" allowing poorer air quality. Use of
cost-benefit in standard setting would be controversial.
c) Episode Related Control
Haze, and other air pollution, varies substantially with weather and season.
Control of high pollution episodes could be more efficiently accomplished if
weather or seasonally variable emission control could be practiced. However,
Section 123 specifically prohibits SIP credit for weather variable controls.
The "reasonable progress, time, and/or measures" approach to SNAAQS attainment
is conducive to the use of efficient but imprecise measures such as weather
variable control.
Therefore, alteration of Section 123 may be useful for the control of haze
or other welfare related air pollution.
Effectiveness:
This approach may be partially effective, has limitations over the. long-term.
The traditional approach of controlling during periodic episodes would require
coordination and communication among' hundreds of sources and short response times.
Moreover, benefits would be limited to reducing haze on the worst days, not
making average conditions good or good conditions better.- With the possible
exception of health, associated benefits (acidic deposition, materials damage)
-------
I
I
would also be limited. Episodic control might have some application in the
West, but only if understanding of transport patterns improves. Possibly more
effective would be seasonal control approaches, with an emphasis on reducing
summertime emissions. |
Simplicity: . _
This approach would not be simple to implement. To make this alternative
viable, it would be necessary to estimate haze formation based on a number of
factors. These include variable meteorological conditions, transport and
dispersion parameters in areas of complex terrain, formation of secondary |
aerosols, emissions from single or multiple sources on a regional scale, and
the resulting perceptibility of the haze itself.
Cost:
Unless a fairly accurate and simplified methodology could be developed for
predicting haze episodes, this approach would be costly to operate, and not clearly I
less expensive than continuous control. Extensive monitoring and research
would be required to develop any useful predictive techniques initially and to m
verify the effectiveness of the program as it is implemented. The practicality I
of major fuel shifts or periodic curtailment of operation on a large scale needs
further assessment.
Enforceability: I
This approach would be difficult to enforce without the use of costly
continuous monitoring devices and/or frequent inspections of operation records. |
Flexibility:
This approach is flexible, permitting the possibility of improved visibility
that might not be reasonably affordable with fulltime control.
Compatibility:
I
This approach could have compatibility problems with the usual air pollution _
control procedures; such systems have however been used on individual sources I
(t*.a~ «ma1t»r«^ no an intorim hacic * H
I
Because of its non-traditional character and enforcement burden, this
approach might not be well received by state and local programs. Sources night H
or might not prefer such approaches depending on their costs and effects on I
reliability as compared to those for alternatives that might be required.
on an interim basis.
Acceptability:
d) Revised New Source Performance Standard Requirements
I
One possible revision to New Source Preformance Standards that has been
suggested concerns the percent reduction requirement of Section lll(a)(l)(A)(ii)
which in effect prevents the use of low sulfur coal as the sole means of achieving
NSPS for fossil fuel fired stationary sources. Removal of this provision has
been proposed by same, and if implemented, SOg emissions in the West would
I
-------
I
I
I
I
I
I
i
i
i
i
i
i
i
i
I
i
i
i
i
25
increase because flue gas desulfurization technology would not be routinely
applied to power plants using low sulfur coal. A related provision would
permit NSPS "bubbles", i.e., the reduction of emissions from existing sources
that offset new source growth (see "emissions trading" below).
Effectiveness:
The overall effectiveness in reducing or preventing visibility degradation,
would be limited, to the eastern U.S., where it could accelerate reductions in
the near term. Eliminating the percent removal requirement could, however,
increase future western $02 emissions. . . .
Simplicity:
Such an approach would be relatively simple and understandable to regulators,
affected industries, environmentalists and the public.
Cost:
The cost of development of revised NSPS under the current or modified
provision would be comparable. Control costs would decrease.
Enforceability:
Emission standards are generally easier to enforce than ambient air quality
standards, and nationally applicable standards facilitate enforcement.
Flexibility:
This is a relatively inflexible approach that, of itself, cannot effectively
consider regional, area or vista-specific differences.
Compatibility:
Revised NSPS would be compatible with the existing statutory .and regulatory
programs.
Acceptability:
Eliminating the percent removal requirement for the power plant NSPS would
be very controversial, and objected to by those whose livelihood depends, on
high sulfur coal, and by .those concerned about long-term emissions growth in the
West. - ' .
e) Requirements for Forest, Range, Agricultural
Practices
and Other Land Management
Congress could give generic guidance for developing regulations to control
the use of prescribed and agricultural burning, soil cultivation and construction
and maintenance-of unpaved roads under NAAQS, PSD, and-the visibility program.
Regulation of these "non-traditional" sources in a consistent, effective manner
has proven to be a difficult challenge under current authorities.
-------
26
Effectiveness:
This approach should be effective in controlling land management practices that
can contribute significantly to visibility impairment on a broad scale. Seasonal
impacts of these various activities could be minimized through substitute
activities or by specifying alternative methods.
Simplicity:
This approach would be complex to implement, but the mandate could
make it easier than at present. Some aspects may, however, not be readily
acceptable to land management agencies. The ease with which this approach is
carried out will depend upon the degree of authority delegated to the regulating
agencies.
Cost: . '
The cost of this approach is difficult to determine. It would vary by
region and depend upon feasibility of measures that could be imposed. Program
design and enforcement activities would require additional resources to implement.
Enforceability:
Successful enforcement of this approach also depends upon the authority
granted to the regulating agencies.
Flexibility:
This approach should be flexibile to the extent that different control
measures would be necessary for the various land management activities in
different regions. A combination of measures could be implemented to minimize
impacts of these activities on regional haze.
Compatibility:
Although control measures under this approach have not been utilized to a
large extent, historically, this approach should be compatible with existing
control programs.
Acceptability:
Efforts to control burning, and other temporary area sources have met with
resistance in the past; the number of "sources" of various sites make it likely
that some will resist regulation. Nevertheless, land managers and sources
recognize the need for consensus guidelines.
2.2.3 Modifications Based on Other Objectives
a) Emissions Trading
Another, option that could be considered in conjunction with the emissions and
technology based approaches discussed in this report is to amend the Clean Air
Act to permit emissions trading (bubbles) involving new as well as existing
sources. Such a measure would substantially expand currently allowable emissions
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
27
trades and may be well suited to regional pollution where source location is
relatively less, important. An emission trade would create an imaginary "bubble1.1
within which a source of air pollution could increase its emissions above the
legally allowable limits, if another source(s) within the bubble would agree to
reduce its emissions by an equal or greater amount. The effect of the combined
emission increases and decreases would be to create a net improvement in air
quality. A factor linking emission trading to haze is that power plant 'smelter
trades in some regions of the West might improve visibility because smelter
emissions are richer in directly emitted sulfates and other fine particles ,
that produce haze. It is best considered as a cost minimizing approach used I ^
in combination with an overall emissions reductions strategy e.g., age based I
control, or more stringent NSPS.
Effectiveness:
The effectiveness of such a strategy must be examined very carefully to
insure that any benefits realized by reducing haze on a regional scale would
not be offset by degradation of air quality on the local scale near the air
pollution sources.
EPA contracted with Systems Applications, Inc. (SAI) to perform a preliminary
analysis of the net regional air quality impacts resulting from a scenario in
which the S02 emissions from power plants were allowed to increase in a trade
that provided for decreases in emissions at smelters. In particular, increases
in emissions at 11 power plants and decreases in emissions at 7 copper smelters
in the southwestern United States were analyzed. The results of this modeling
analysis indicated that there would be areas of improved air quality near
copper smelters and degraded air quality near the power plants. However, for i
sulfate concentrations and visibility impacts, that are caused by fine particulate
sulfate, areas of relative improvement and degradation were less localized.
Visibility and sulfate air quality were predicted to improve slightly in most
areas in the southwest. In the small number of areas where degradation was
predicted, the precentage degradation was very small; Sulfate air quality and
visibility was predicted not to change significantly as a result of emission
trading. These results should be viewed with caution, however, due to
model based limitations.
The SAI report also indicated a potential problem resulting from this
proposed emission trading in that increased S02 emissions at many power plant
sites -may not be possible because of constraints imposed by PSD class I and II
increment consumption. These constraints would be most severe in areas of
complex terrain and areas in the vicinity of national parks and wilderness
areas. This report also assumed that the copper smelters would meet the SIP
emission limits. If the smelter'industry were unable to meet the SIP emission
limits without emission trading with electric utilities, much more significant
improvements in regional haze and air quality could result from an emission
trade.
Approved emission trades with new sources could circumvent the application and
installation of 8ACT on all new sources which has been a long-standing national
policy and requirement of the Clean Air Act.
The overall effectiveness in reducing or preventing visibility degradation
could be very limited because it would not prevent large increases in visibility
-------
28
reducing emissions within a particular area. This approach also does not have
any direct relationship to ambient air quality or visual air quality. Without
continuous monitoring and extensive regional modeling the effectiveness of
this approach would only be known after the fact and possibly too late to
effectively address visibility problems. Under this approach, area specific
visibility problems of specific class I areas and other visually sensitive
federal and non-federal areas and vistas would not be addressed. However, the
interstate nature of haze could be considered with such emission trades. By
itself, it would be insufficient to assure effective visibility protection.
Alternative approaches should be examined more carefully before steps are
taken toward developing such an initative.
Simplicity:
Emissions trading is a simple and understandable concept, but one that
would be complex.in its implementation. Extensive regional modeling analyses
would be required both to isolate those existing and new sources with a significant
potential to reduce haze by participating in a trade, and to calculate the
contribution to improvement or worsening of air and visual quality of each
source.
Cost:
The cost of this approach would be relatively high in dollars, time and
personnel on the part of the agency evaluating and implementing an emission
trend.
Enforceability:
The emission trade concept is enforceable only as long as the emission
limitations of the participating sources are enforceable.
Flexibility:
Emission trades would provide added flexibility in designing and modifying
other pollution control strategies.
Compatibility:
» " _
This approach would be compatible with the existing bubble policies and
regulations of the EPA and various State governments. Generally, the bubble
approach has been applied to existing sources located in nonattainment areas.
At present, the State of North Dakota is conducting a study of the feasibility
of an emission trading program to assist in the management of air quality class
I and II increment consumption.
Acceptability:
Emission trading for new sources would be
The smelter industry has indicated in its
little to be gained from such trading.
I
I
I
I
I
expected to be very controversial.
comments to the task force that it sees
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
29
b) Reasonable Effort to Attain SNAAQS
Section 110(a)(2)(A)(ii) of the Act specifies that a State plan implementing
a secondary National Ambient Air Quality Standard specify a "reasonable time"
at which such standard will be attained. A possible alternative would be to
require a "reasonable effort" at attaining the SNAAQS. The latter is more
amenable to regulatory interpretation via demonstrated, cost-effective control
methods. Such an amendment might be appropriate if SNAAQS were the vehicle by
which haze (or any other welfare effect) was to be addressed.
Effectiveness:
This amendment would facilitate the regulatory implementation of the
SNAAQS by encouraging technology/cost based procedures, e.g., RACT.
Simplicity:
The requirement for a "reasonable effort" appears to be a relatively simple
one to understand, but must be well defined in EPA guidelines.
Cost:
This measure would probably increase the cost-effectivenss of an approach
to control haze based on the SNAAQS by providing for a more efficient implementa-
tion. ' '
Enforceability:
Determining whether a reasonable effort has been made could be very difficult
which in turn, would make enforcement difficult.
Flexibi 1 ity:
As discussed above, the SNAAQS approach does not.provide a very flexiable
framework to attack diverse visibility problems. Incorporating the requirement
that a reasonable effort be made to attain the SNAAQS could increase the
approach's flexibility, if visibility aspects could be considered in the
determination of reasonable effort.
Compatibility:
This amendment would not be incompatible with.current approaches for
regulating air quality and visibility impairment, or other goals and
responsibilities of federal and State governments.
Acceptability:
The acceptability would be related the nature of guidance issued.
2.3.0 New Clean Air Act Authorities
2.3.1 Haze as Principal Objective
-------
30
a) Visibility Standards
Under this approach, a regional, area specific or local visibility standard
based on some quantitative measure of visibility (e.g., visual range, contrast,
etc.) would be promulgated by the Federal or State government. A deadline for
attaining the standard would be established and the States would be required to
develop plans for attaining the visibility standards by controlling existing
and new emissions of visibility-related pollutants.
Effectiveness:
This approach could be very effective in preventing and improving regional
visibility degradation. Provided that the attainment of any such standard
could be achieved through control of any and all sources contributing to
visibility-reducing emissions, this approach would be comprehensive in scope.
The key to the effectiveness of this approach is in the establishment and
enforcement of a standard-level which is acceptable to the public in any given
area of applicability.
Simplicity:
This approach would be implemented in the same fashion as other ambient
air quality standards, but accounting for non-pollutant influences on visibility
{fog, rain, illumination, clouds) would add complexity.
Cost:
The cost of this approach could be significant. Many states would have
to develop emission control strategies which may place increased control costs
on certain existing and new sources. Additional federal and State resoruces
may be necessary to develop and implement the standard.
Enforceability:
The enforceabi-lity would depend on the effectiveness of modeling approaches,
the extent and reliability of monitoring data, and authorities granted to
control interstate air pollution.
Flexibility:
Any new mandate for a standard to protect visibility should be flexibile.
It should allow the application of a standard in such a way that will meet the
goals of class I protection and provide for reasonable progress in other areas
of the nations.
Compatibility:
This approach would be compatible with existing approaches to protecting
the public welfare. . -
Acceptability: ;
This complexity of this approach may cause disagreement among air quality
control agencies, industry,'and the interested public.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
31
2.3.2 Haze as Partial Objective
a) SO? Rollback
The S02 rollback proposals are usually advocated in the name of "acid
rain" control. When the S02 rollback proposals are viewed as general air pollution
control, haze reduction is one of the potentially major benefits that can be
legitimately cited. Some examples are examined in detail in Appendix B.
Effectiveness:
Since regional visibility impairment has been associated with regional
sulfur oxides emissions, particularly in the east, this approach could be at
least partially effective at reducing existing impairment. If a cap on SOg
emissions is established as a provision of any required Toll backs, then this
approch might be effective in controlling any future increases of .502 from new
sources as well. The effectiveness could also be enhanced by making SOg rollback
provisions applicalbe to all sources regardless of emission potential. This
approach would not address the visibility impairment problems caused by other
pollutants such as primary fine'particles, nitrogen oxides, and volatile organics.
Simplicity:
This approach would be reltively simple to implement once equitable
scenarios are developed.
.Cost:
The cost of the design of this approach could be relatively low. However,
implementation and enforcement could be costly.
Enforceability:
This approach should be relatively easy to enforce if source specific $03
emissions levels are specified.
Flexibility:
Such approach can be flexibile, within the limits of the overall mandated
reductions. Control strategies could be designed to require emissions reductions
either on a national, regiona, or more localized scale or'by some other criteria
such as one which might consider source characteristics.
Compatibility:
This approach could have aspects of air quality management (SIP based emission
limits) compatible and technology based programs.
Acceptability:
Industry affected by this approach may not readily accept regulation based
solely on visibility concerns, but the other public health and welfare benefits
associated with S02 reductions may lend more support to this.strategy. A
congressional mandate offers strong advantages in this area.
-------
32
b) Age Based Control (existingpowerplants)
An alternative to SOg rollback, suggested by Congressmen Broyhill and
Madigan, would be to require existing powerplants to control at age X (or
retire) emissions to level y (e.g., x = 40 years, y -2 lb/106 BTU). The effect
would be similar to SOg rolTback, but would occur over a 20-30 year period
rather thanXal1 at once in the mid 1990's.
Effectiveness:
This approach would probably as effective than a general $03 rollback
strategy in the long-run. Its chief drawback is improvements would occur
10 to 20 years later than rollbacks in 1995-2000.
Simplicity:
This approach is simple and easy to understand.
Cost:
The costs are comparable to rollback but are incurred 10 to 20 years later
for many power plants significantly reducing present value of the costs.
The stream of costs would be similar to the stream of cost savings associated
with new existing source emissions trading. Preliminary analyses by DOE
indicates that a combination of those measures could break even or even yield
a net savings relative to current programs.
Enforceability:
Once the affected sources are identified and control schedules developed,
this strategy should be relatively easy to enforce.
Flexibility:
This approach would not be very flexibile. Its purpose is to address the
problem of long-lived power plants directly. When combined with emissions
trading, the overall approach is quite flexible.
Compatiblity:
This approach is supplementary to and compatible with current technology
based programs (NSPS, BACT, LAER).
Acceptability:
The approach would be more acceptable than roll back to industry but less
acceptable to those who would prefer to see continued visibility improvements
in the next 10 to 20 years. The acceptability of any further S02 control is,
however, contingent on a societal decision that such control is necessary and
beneficial.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
33
2.3.3 New Authorities for Haze Precursors to Meet Other Objectives
a) "Acid Rain"
"Acid rain" control is the subject of major new legislative proposals whose
side effect could be haze reduction. The primary focus of this legislation is
to reduce the impacts of acid deposition in the eastern United States with
little benefit for western areas of the nation. A comprehensive national acid
rain control program has the potential for significantly reducing total emissions
of pollutants that are primarily responsible for acid deposition and visibility
degradation, i.e., sulfur dioxide and nitrogen oxides.
If NOX is shown to be the principal cause of forest effects and spring
snowmelt-related acidification of aquatic areas, then SOg reduction could be
deemphasized and eastern haze reduction potential diminished.
Effectiveness:
This approach has the potential for greatly reducing the need for any separate
major visibility protection program for regional haze. As indicated in Appendix B,
very targeted acid deposition strategies may produce little haze related benefits.
The effectiveness in reducing western regional haze will depend upon the extent
and degree of required emissions reductions which impact haze in the West. The
need for further emissions reduction to protect visibility could be determined
after implementation of any acid rain control program, but this would delay the
timing of regional haze benefits.
Simplicity:
This approach can be simple or complex depending on the final program design
by Congress. Emissions reductions could be specified by State, region, or
nationwide and could apply across the board to all sources by categories, size,
or other source characteristics.
Cost:
The cost could be substantial. Control technology costs would be high as
well as control agency costs to implement and enforce the program.
Enforceability:
Once emission reduction plans and compliance schedules are developed,
enforcement of this approach should be reltively easy.
Flexibility:
The design of the program by Congress will determine, how much flexibility
is allowed in implementing this approach. A national program could prove to be
less flexible and adaptable to the needs of individual states- and the problems
of individual sources.
-------
34
Compatibility:
This approach should be compatible with existing air pollution control
programs.
Acceptability:
Uniform national requirements for acid rain control may be more acceptable
to all concerned, when compared to individual state or regional approaches that
would operate without Congressional backing.
3.0 MAJOR APPROACHES FOR FURTHER EXAMINATION
The available projections in Appendix B suggest that regional visibility
in both the East and West is likely to be stable for the next decade - the
East experiencing continued episodic regional haze and the West maintaining
relatively good current conditions. Smaller areas within these regions may,
however, experience visibility degradation depending upon local emissions and
meteorology. Based on its examination of strategies as they might be applied
to these somewhat distinct problems in these regions, the following findings
are offered.
3.1 Eastern Haze Strategies
Maintaining current regulatory programs will likely result in approximately
the present level of episodic regional haze in the East through about 2030,
with potential improvement thereafter. More rapid improvements would require
regional reductions of S02 emissions. The major approaches for accelerating
such reductions are ranked in order of effectiveness.
1. The analyses of SOg control strategies (Appendices B,C) suggest that
the benefits of improved visibility estimated for SOg roll backs are uncertain,
but might be substantial. Even at the higher end of the uncertainty range,
however, they are unlikely to equal or exceed control costs by themselves.
Some of the more targeted strategies designed to control acid-deposition produce
visibility improvements that are limited in extent and magnitude. If accelerated
reductions of S02 emissions are judged appropriate, the most effective long-term
regulatory approach would be to develop cost-effective regional strategies
designed to maximize all of the known and anticipated multiple environmental
benefits of SOg control, including haze reduction. Implementation of such a
strategy is best accomplished with new legislation.
2. Although less timely and more difficult to implement than a mandated
control strategy, if it is decided that accelerated reduction of eastern haze
is warranted, consideration should be given to developing a fine particle
SNAAQS. Outside groups may attempt to force action on such a standard in any
case, and in the absence of new legislation mandating regional SOg reduction,
a standard could guide long-term sulfur oxides and other control programs by
establishing an acceptable target. Some modifications to current Clean Air
Act authorities might improve the design and implementation of SNAAQS for
fine particles, especially if haze control is the major objective.
3. A supplemental (or alternative) long-term strategy for reducing SOg
is control of sources based after a defined source life-time. This could act"
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
35
to ensure and accelerate reductions projected from retirement (or control) of
higher emitting existing sources and their replacement with cleaner new
facilities.
3-2 Western Haze Strategies
Protection of western visibility may be effected well enough without
new legislation. Assuming implementation of current regulatory programs, the
western analyses suggests that delaying development of new regional haze
programs for five years while awaiting improved source-receptor information
would not result in unacceptable or irreversible degradation. As new research
results improve our ability to predict regional impacts of emission source
limitations, "phase II" regulations can be developed under the existing
section 169A. Protecting visibility in class I areas while implementing
existing NAAQS, PSD, and mobile source controls will provide substantial
protection against impairment from haze throughout the West. Interim
adjustments to existing 169A implementation procedures can facilitate the
transition to a phase II program. Alternatives for doing this are
discussed in the following sect'ion.
3.3 Issues Associated With Current Visibility Regulations
On December 2, 1980, at 45 FR 80084, EPA established requirements
for States to consider visibility in mandatory Class I Federal areas as
part of the State implementation plan (SIP) to meet Clean Air Act goals. '
At the time of promulgation of these requirements, technical limitations
prevented consideration of control strategies for regional haze. However,
the basic structure of the requirements did allow for documentation and '
consideration of regional haze impairment at the Class I areas. It also
anticipated future EPA action, called Phase II, which would address
control strategies for regional visibility impairment. Through this
report, EPA now is looking at control strategies which will address
regional haze, although the scope is'not limited to mandatory Federal
Class I areas.
It is important to note that the framework of Phase I visibility
regulations is not yet in place. Only in 1984 did EPA begin to seriously
encourage States to develop their own plans, and begin to act for those
States with deficient SIP's. Therefore, any consideration of control for
regional haze must also consider important issues yet unresolved from
Phase I.
3.3.1 Immediate Issues
The-.EPA is in the process of implementing changes to the SIP's to
incorporate rules promulgated.under Section 165 (new source review) and
Section 169A (visibility protection in mandatory Class I Federal areas).
Certain key issues should be resolved so as to provide a reasonable basis
for any future action on regional haze. The most immediate actions will
take place in June 1985, with promulgation of a Federal monitoring strategy
and new source review requirements. Certain issues EPA will address in
this rulemaking that may. have consequences on regional haze control are
discussed below.
-------
36
1. Documentation of Impairment
In addition to allowing Federal Land Managers (FLM's) to establish
integral vistas, the Phase I rules also allow FLM's to identify impairment
in the Class I areas, as well as any named integral vistas. This identi-
fication process is not_ only for the types of impairment that'can be
controlled under Phase I, but also covers regional haze type impairment.
The intent of the regulations is to assess all visiblity .impairment that
affects the Class I areas. Such identification by the FLM's is then
subject to further review as to whether the impairment is significant, and
then even further review to see if such significant impairment can be
considered reasonably attributable and therefore controlled under Phase
I. In implementation of Phase I and in development of strategies for
Phase II, EPA should stress the fact that a documented record of regional
haze impairments that exist at Class I areas is called for by the current
regulatory framework and should be part of the SIP's when the Phase I
rules are implemented.
2. Monitoring Strategy and the Proposed Network
Because EPA is now implementing the Phase I rules for most States, it
has proposed a national network for collection of visual air quality data
important to carrying out the review and permitting of sources near Class I
areas. This network will be used, in part, as a source of information on
background levels of visibility impairment for the purpose of determining a
potential source's contribution to that impairment. In collecting this
information, the monitors will also create a record of impairment resulting
from long range transport from distant sources or regional haze. This
information should be utilized in any national strategy for regional haze
control. Therefore, careful planning of collection and storage of the
information is needed to assure access to Phase II programs.
Better criteria for measurement of impairment and its relative levels
would also aid in assessing progress toward the national goal which is
required in the long-term strategy review under Phase I. The most direct
means of standardization is a EPA reference method. Work on a reference
method is scheduled for FY 86 and should be completed quickly to aid Phase I
implementation by the States and pave the way for Phase II considerations.
3.3.2 Long-Term Strategies Under Phase I . -
In the near future, EPA will be developing Federal plans to incorporate
consideration of integral vistas, review of existing source impairments, and
long-term.strategies for control of visibility impairment. Certain general
guidance should be issued, especially for review of. SIP's for a long-term .
strategy. For example: Los Angeles currently has a long-term strategy"
for ozone control which encourages residential development and other
growth 'into the desert to relieve ozone concentrations in the western
part of the valley. This clearly is in direct opposition to a.policy to
protect visibility in Class I areas of the desert that are immediately
east of Los Angeles. EPA should take firm stands on several issues
before completing development of this Federal promulgation or acting on
State submittals. These issues are:
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
1
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
37
1. Cumulative Impacts of New Sources
The current review of new sources for visibility impacts is essentially
tied to the prevention of significant deterioration program (PSD) and new
source review promgram (NSR) in nonattainment areas. While it is common
practice to add a permitted sources' emissions to background levels before
reviewing the next permit, neither the visibility rules nor the PSD/NSR
rules are explicit in dealing with situations when many sources are applying
for permits in the same general location at the same time. This is commonly
referred to as the "cumulative effect" of permit processing. In a situation
where several owner/operators apply for permits in one area (such as an oil
field development), each permit would be reviewed based on/that owner/operator's
contribution to background. Because the background information used to
judge impacts would reflect conditions as they exist now, the effect of all
sources once built would be underestimated. An explicit policy on permit
processing in this situation, either as part of PSD or visibility review,
should be set. This policy could continue the individual permit process or
require simultaneous permits applications for different source owner/
operators to be considered as one action. Combining the impacts of many
minor sources that if reviewed together would trigger a permit review
action should also be considered. This issue will affect all types of
visibility impairment, both plume and regional haze. However, it may be
more important in the case of regional haze, since there are almost
always multiple sources involved. As an example: a major mineral development
project involving many companies acting independently could create a regional .
problem, with none of the individual permits causing an adverse impairment
on its own.
2. Integral Vistas .
Integral vistas are views from a Class I area of panoramas or landmarks
that lie outside the statutory boundary-of the area. The EPA recognizes
that there are important views from Class I areas to lands surrounding
them. Also the establishment by Congress of certain Class I areas was
substantiated on spectacular views from these areas, which encompass great
distances. In consideration of the importance of such views, the provision
for protection of integral vistas was included in the existing regulations.
However, the existing rule allows the balanacing between protection of
integral vistas and economic and other social factors when issuing permits
or for developing strategies to meet the national goal. This is less stringent
than the requirement that a permit be'denied if an FLM proclaims adverse
impairment to visibility within a class I area.
Several petitions for reconsideration of the rules requested EPA to
remove integral vista protection from the rules. Although not responding
to these petitions, EPA is now implementing these rules by a court ordered
settlement agreement and will, at the appropriate time, incorporate consideration
for integral vistas named by the FLM's according .to the December 1980 rule.
However, EPA's delay in implementation of the rules, and in lack of response to
the petitions for reconsideration, have caused some problems. The December
1980 rules only give the FLM's unti.l December 31, 1985, to name any integral
vistas. As yet, only one set for one Class I area has been named. The FLM's
have, in addition to other internal considerations, been waiting for EPA's
response to the petitions and for court action before devoting resources to
naming integral vistas.
-------
38
Before consideration of any regional haze program, EPA should take
action on the petitions for reconsideration with respect to this issue.
Such action quickly would resolve the question of EPA support or nonsupport
for integral vistas and end FLM debate on establishing integal vistas. If
integral vistas are kept, they should become part of a regional haze control
strategy. If not, EPA need not consider such a specialized program.
3. Definitions and Guidance on Long Term Strategy Review
Phase I rules only trigger control strategies for impairment
attributable to one source or a group of sources: however, better guidance
on defining significant and adverse impairment for all types of visibility
degradation events needs to be developed. Although many subjective
components will determine actions taken under the Phase I program, some
review and compilation of different actions by EPA and the States in
handling impairment could become a case-by-case guideline. This would also
point to areas of needed research and regulatory investigation for areas
where visibility improvement or maintenance is not seen. This would be
useful also in a Phase II program for class I areas as envisioned, based on
the success of control decisions.
4. Prescribed Burning
In certain areas, particularly the Pacific Northwest, control of
wildfires and underbrush growth is accomplished through man-made fires.
These activities are sometimes the major cause of visual impairment in
mandatory Federal Class I areas. While these operations can cover large
expanses of land in and near Class I areas, controlling precribed burning
activities can be handled now as a "reasonably attributable" source issue.
During the development of the Phase I rules, discussions on factoring'
visibility impacts into decisions on when, where, and how often to use
presribed burning in and near Class I areas were begun. The Forest Service
continually updates its manuals and guidance on proper techniques for
prescribed buring. This guidance does consider visibility impacts when
deciding on burn days and amount of land to be burned. During implementation
of Phase I. rules, EPA and the States should specifically address prescribed
buring during development of the long-term strategies. In some cases,
existing Forest Service guidance will be sufficient to limit visibility
impacts to a reasonable level; in other cases, additional restrictions may
have to be considered. This issue is best resolved in a case-by-case
review during Phase I rulemaking.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
APPENDIX E. RESEARCH NEEDS
I
I
I
I
I
I
I
by: Al Galli, EPA/ORD
I Marc Pitchford, EPA/ORD
William Wilson, EPA/ORD
William Malm, NPS
I
'
I
I
I
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX E
TABLE OF CONTENTS
1.0 Human Perception
2.0 Monitoring Techniques
3.0 Characterization of Regional Visibility Conditions
4.0 Empirical Approaches to Source Assessment
5.0 Deterministic Modeling
Page
1
2
7
9
14
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX E. RESEARCH NEEDS
The first major examination of interagency visibility strategies and
information needs was presented in the 1979 EPA Report to Congress.
Progress has been made since that time, but significant opportunities
were lost by the reduction of EPA's visibility research program. The
Park Service, Department of Defense and industry groups have all mounted
visibility programs, but all cite the need to restore EPA to a leadership
role in visibility research. The reduction of uncertainties in a number
of key areas, particularly in source receptor relationships, perception
and values, is the highest priority for the development and implementation
of credible long-term strategies for regional haze. The task force
recognizes that visibility will not be the highest research priority, and
that EPA will not be able to fund the full extent of research needed.
The discussion in this Appendix focuses on general needs to be
addressed by the collective visibility research community. Continued and
expanded coordination is needed to agree upon priorities in these areas.
Near-term research priorities for EPA budget planning are documented in
the separate interim needs statement (March 1984} and have already been
factored into FY85 and FY86 budget planning.
1.0 HUMAN PERCEPTION
Studies of perception to date have yielded information that is pertinent
to understanding the response of the human eye-brain system to changes in
scenic quality resulting from changes in atmospheric aerosol concentration.
There are a number of needs that remain unanswered and in some cases
have not ever been addressed. Specifically:
1. Does a just perceptible change in image information content constitute
visibility impairment or is a just noticeable difference in judgments
of image quality the more pertinent approach?
2. What features of a scenic landscape that are perceptible are
sensitive to changes in air quality (color, texture, form etc)?
3. How much of a change in air quality constitutes visibility
impairment of a scenic vista, as a function of vista characteristics
(vista size, textural content, color etc.)?
4. How does the size, shape, edge sharpness, and color of layered haze
relate to visibility impairment (just-noticeable difference)?
5. How much of a change in optical characteristics of a visible
layered haze is noticeable, as a function of its size, shape and
edge sharpness?
-------
Introduction
There are several techniques in use at present for monitoring visibility
2.1 View Monitoring
I
I
6. How does the passage of time affect perception of differences in
visual air quality? ;
7. How does perception vary in the population?
Additional research into methods for quantifying information content and |
quality of landscape features will be required to address these research
questions. Of primary interest is the quantification of scenic textural _
content and scene color. Because conclusions derived from studies of I
visibility values are often highly sensitive to assumptions about perception,
strong linkages should be established among researchers in the fields of
perception and value.
2.0 MONITORING TECHNIQUES
I
and other related parameters. Some techniques have serious deficTencies which I
limit their utility. Even those without inherent flaws are presently limited I
by a lack of standardized guidance for their use and the use of data they
generate. This section discusses the desired characteristics and possible m-
directions for new technology plus the requirements for guidance in the use of
existing methods.
Monitoring of visibility and related parameters is performed to obtain I
three distinct types of information. The characteristics of a scene viewed at
a distance are monitored to document the scene-specific visibility. Optical
properties of the atmosphere are monitored for a scene-independent measure of
visual air quality. Aerosol characteristics are determined to associate |
atmospheric optical properties with the responsible pollutants. Monitoring
techniques development for each of these informational requirements is discussed _
separately below. Though most of the following discussion concerns monitoring I
regionally uniform visibility, monitoring techniques for plumes and layered
hazes are also covered.
I
View monitoring requires the development of well-tested equipment and
procedures to allow the brightness, color, and spatial detail of the viewed
scene to be accurately recorded and subsequently used. The photographic method "
is most commonly used for monitoring vistas. The availability of adequate
equipment, supplies and services to support this method make it the obvious
choice for qualitative documentation purposes. However, the present lack of
standardized procedures for its use in visibility monitoring make quantitative
analysis much less certain. Among the items which need to be addressed are the
choice of films and processing, camera exposure and lenses, methods to calibrate
or document system gamma functions, archiving, and quality assurance procedures.
Promising new techniques which may be applied to view monitoring include I
digital photography and video tape or disk archiving of images. The cost and
inherent difficulties associated with the use of film for long-term routine
monitoring make the non-film approaches very attractive. Prior to their use
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
for visibility monitoring, these new techniques must be evaluated and standard-
ized procedures for their use developed.
Some of the objectives of view monitoring are the collection of data which
can be used to,study perception, to evaluate visibility trends and to assess
the sensitivity of a view to changes in air quality. To accommodate these
objectives, equipment and procedures to extract quantitive measures of bright-
ness, color and spatial detail from photographs (film or digital medium) must
be developed. Several methods to accomplish this are presently in use. These
include film spot densitometry and several techniques to measure radiance of
projected images. Such techniques need to be evaluated, improved where possible
and standardized. New approaches, including methods to process digitally
recorded images, should also be developed and characterized.
Another objective of view monitoring is to make possible the determination
of frequency, persistence and Intensity of plumes and elevated hazes associated
with local sources. Time lapse photography has been used for this purpose.
Operations, quality assurance and information processing procedures are
required to promote comparability of data collected by different investigators.
2.2 Monitoring Atmospheric,Optical Properties
For scene-independent measurements of visual air quality, the goal is to.
develop techniques to routinely monitor the atmospheric extinction coefficient.
Measurement of the scattering phase function is also of interest where the
objective is to associate optical properties to view impacts. There are two
methods in common use for monitoring the atmospheric extinction coefficient. %
Both have deficiencies which limit their use, thus new techniques are being
investigated. Scattering phase function is not routinely measured for-visi-
bility research. .
The extinction monitoring methods in common use are long path measure-
ments of target/sky contrast by teleradiometer (or other radiance monitoring
method) and scattering coefficient measurements by integrating nephelometer.
The deficiencies of the contrast measurement technique primarily involve non-
uniform lighting conditions and inherent contrast {target/sky contrasts at the
target) uncertainty. For nephelometer measurements of scattering, the primary
deficiencies are inadvertent sample modification (aerosol size distribution
changes), and the need for separately measured or assumed absorption coefficient
information. With the possible exception of the nephelometer sample modifica-
tion, these deficiencies are inherent to the techniques. Until new techniques
are available for routine use, standardized guidance for the use of existing
methods must be developed to minimize the inherent deficiencies, quantify
the uncertainties, and promote some degree of uniformity in the quality of
data collected. Standardized measurement specification (precision, accuracy,
sensitivity, spectral characteristics, etc.) by instrument type should be
established. Guidance for instrument selection and siting and minimum acceptable
monitoring and quality assurance procedures should be developed. Determining
target/sky contrast from photographs is supported by many investigators as a
reasonable alternative to use of teleradiometers. This approach requires
additional characterization and comparability tests and procedures development
prior to its routine use. Also required for all methods is the development
of data handling and editing procedures, including methods used to estimate
-------
I
I
inherent contrasts. In addition, absorption coefficient measurement
techniques require additional development, testing and standardization.
Several new techniques are being developed and tested in hopes of provid-
ing methods without the deficiencies described above. The techniques being
investigated are transmission measurements and radiance measurements with
configurations beyond the commonly used target/sky contrast.- Both techniques |
will likely operate over an intermediate path length (1 to 10 km), though some
of the radiance measurement methods may be made over long paths if artificial _
targets are not required. Other desired instruments include a next generation I
integrating nephelometer (open air or other design to eliminate sample
modification), a continuous absorption coefficient monitor, and a field worthy
polar nephelometer or other device for continuous phase function monitoring.
Any new method must be fully characterized, field tested and have operation and |
quality control procedures developed prior to its routine use.
Monitoring optical characteristics of plumes and layered hazes is done
remotely, using radiance measurements through and outside of the plume and in
situ, typically with airborne measurements of scattering coefficients (nephelom-
eter) occasionally supplemented with absorption coefficient measurements. In
spite of the infrequent use of monitoring for this objective, guidance concern-
ing operation and quality control procedures is needed to promote comparability
among investigations.
2.3 Monitoring Aerosol Characteristics
To understand the causes of visibility impairment, it is necessary to I
monitor particle characteristics in conjuction with optical monitoring. The "
two approaches are to develop and employ techniques which provide optically
important aerosol information, and information which will aid in source
identification. The most common method employed to meet these goals is |
size-segregated aerosol sampling and subsequent sample analysis for physical
and chemical characteristics. There is a need to establish guidance for _
sampler and analysis selection, and minimum acceptable sampling, analysis, I
and data handling procedures based upon desired overall precision, accuracy
and sensitivity by particle characteristic (mass, species, etc., in each size
range). .
Improved sampling technology is required to minimize sampling artifacts
associated with volatile aerosol components (e.g., volatile organics and M
nitrates). If the volatile aerosol component contributions to visibility I
impairment are shown to be significant, the need for routine methods to sample
them will be critical. Other desired sampling characteristics include the
ability to sample in many size ranges with substrates and sample amounts consis- I
tent with full compositional analysis, and samplers that are field worthy for
remote, low-power, routine service.
More work should be done on the development and testing of non-traditional I
sampling methods of characterizing aerosols and their optical properties. One
such approach, which has resulted in several prototype instruments, involves
systematically modifying aerosol size distributions by physical or chemical
processes prior to measuring the scattering coefficient with an integrating
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
nephelometer. In this way, an association between chemical/physical properties
and the scattering of aerosols can be inferred. In the long term, such methods
may prove a cost-effective supplement or alternative to the sampling and analy-
sis approach. . J
Methods for aerosol characterization within plumes or elevated layers have
many of the same requirements as those discussed above. They have the additional
requirements usually associated with airborne monitoring including low power and
rapid response ot sample time. Instrument and procedures development for
aerosol plume characterization is required for comparability of data collected
-------
6
SUMMARY OF RESEARCH RECOMMENDATIONS
1. Current view monitoring techniques standardization and guidance
2. New view nitoring methods development, testing, and
standardization .
I
I
I
_
3. Equipment and procedures development to obtain quantitative
view data
4. View monitoring techniques development and standardization
for plumes _
5. Current extinction measurement techniques standardization
and guidance:
6. New extinction measurement techniques development testing |
and standardization
7. Fieldworthy measure of phase function development, testing,
and standardization
8. Optical monitoring of plumes-development, testing and B
standardization I
9. Current aerosol monitoring techniques-standardization and.
guidance jj
10. New aerosol monitoring techniques development, testing, and
standardization
11. Nontraditional aerosol characterization-techniques development
and testing
12. Aerosol monitoring of plumes development, testing and
standardization _
S ========= ======S=S======S======S==SS===========3=S======S==3==== =============
I
1
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.0 CHARACTERIZATION OF REGIONAL VISIBILITY CONDITIONS
The study of past and current regional visibility offers a wealth of
information concerning the spatial and temporal scales involved, the magnitude
and urgency of the problem and the major contributors to regional visibility
impairment. Some work of this kind has been done, though much remains to be
accomplished. The data available for interpretive analysis is marginal.
National Weather Service (NWS) visibility observations offer the only long term,
national-scale data. However, they are limited by their semiquantitative
nature and generally restricted range and resolution. Data from several recent
short to moderate term (1 to 8 year) regional monitoring programs-are available.
By design programs such as VIEW, SCENES, RAQS RESOLVE and PANORAMAS have provided
optical and aerosol data to determine the levels and 'causes of visibility
impairment within their respective study areas. Though the quality of these
data are generally superior to airport observations, their restricted spatial
and temporal coverage limits their utility for long-term continental scale
analysis. In spite of these difficulties, existing data should be systematically
investigated to extract the available information and as a means to identify
and design programs to fill data deficiencies.
Study of spatial and temporal distributions of regional visibility is a
logical starting point since it offers the chance of simplifying subsequent
efforts. Some work has been conducted to identify areas of similar visibility
fay time of year, primarily using NWS observations. Further analysis should be
conducted to identify areas of similarly varying visibility by time of year and
areas with similar historic trends. The main objective of this analysis is to
identify a set of visibility regions within which regional visibility is suffi-
ciently similar so that additional regional analysis and monitoring need only.
be done for one representative location per region.
Successful identification of visibility regions should be followed by
analysis to characterize each. This should include descriptive regional and
inter-region comparisons of current visibility levels, historic trends, and
diurnal and annual variations. Associations between visibility levels and
meteorological conditions should be investigated. The interaction between
regional and urban visibility is an important research topic. Studies are
needed to determine what fraction of urban impairment is regional; what pol-
lutants are of greatest importance within the urban setting; whether urban areas
can be categorized by source make up, size, or region and what are the urban
impacts on regional visibility.
To the degree possible, causes of visibility impairment for each region
should be identified. Visibility trends can be compared to historical changes
in pollution emission sources as a means to infer gross cause/effect relation-
ships. If sufficient aerosol data are available, extinction budget analysis
should-be performed. The two commonly used methods to perform extinction
budget analysis, Mie theory and regression analysis,' often provide significantly
different results. Regression analysis has tended to show a much larger contri-
bution by sulfur aerosol and a much smaller contribution by organic aerosol
than the theoretical approach. Possible causes of these discrepancies include
faulty aerosol data masking the statistical relationships between certain
aerosol components (e.g. organic) and extinction; and faulty assumptions
I
-------
SUMMARY OF RESEARCH RECOMMENDATIONS
I
I
concerning aerosol index of refraction, size or shape distributions for Mie
calculations. Specifically designed field and laboratory measurement programs I
should be conducted to resolve these questions. Error propagation analysis for I
both extinction budget analysis methods may prove a valuable approach to under-
standing the discrepancies between them.
The data analysis described above should test the limits of existing data
and identify deficiencies. In this way it provides the starting point for the
design of future monitoring programs to further characterize regional visibility. I
These could take the form of short term special studies,, routine long term
monitoring and/or an augmentation of existing programs.
While some of the 'specific spatial, temporal and characterization studies |
have been accomplished in both the Eastern and Western U.S., there has
not been a comprehensive assessment of regional haze in the East, beyond
interpretation of satellite and NWS data. The establishment of a visibility-
particle Eastern network to obtain a data base to support a standard to
address regional haze problems in the East is needed. This along with
continued support of western characterization research efforts in these
same subject areas is clearly needed. |
I
I
1. Visibility-particle eastern network
2. Visibility region identification, and characterization with I
existing and where necessary, supplemental data
3. Extinction analysis by region
4. Extinction budget analysis techniques research
5. Urban visibility characterization .
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4.0 EMPIRICAL APPROACHES TO SOURCE ASSESSMENT
41 Studiesof Specific Categories of Interest
4.1.1 Urge point sources .- smelters, power plants and industrial complexes
a. Contribution to attributable plumes (plume blight). Project
VISTTA has made a major contribution to understanding and modeling
visible plumes from smelters and power plants. The major source of
error in plume optics models lies in predicting the size, shape and
structure of the plume and the meteorological conditions which give
rise to cohesive, visible plumes. Other problem areas are: deter-
mining the sometimes variable emission rate of nitrogen oxides and
the amount and size distribution of primary particulates in the light
scattering range. Additional analyses of the existing VISTTA data
base would help define problem areas in EPA's plume model as well
as those of other developers and provide a guide to model improvement,
especially in complex terrain. Measurements of plume structure
by photography, correlation of plume structure with meteorological
parameters - especially turbulence at plume altitudes, and analysis
of model predictions of plume blight occurrence over the longer -
term such as periods of a year are needed. Lidar studies would
provide valuable details of plume structure to augment photographic
measurements.
b. Contribution to regional haze and layered haze. This area is
reasonably wellunderstood due to the results of studies such as
VISTTA, MISTT, Tennessee Plume Study, Cold Weather Plume Study and
PEPE. Additional analysis of the existing field data with emphasis
on visibility would be useful. Although power plants and smelters
have been reasonably well characterized for contributions to
visibility reduction, very little work has been done on industrial
complexes and similar large point sources. Some characterization of
emissions including size distribution of primary particulates and the
transformation of the emissions to particles through chemical reaction
is needed for industrial operations such as steel mills, refineries,
etc.
4.1.2 Large area sources - urban plumes, forest fires, natural emissions
and industrial complexes
a. Urban plumes. Project MISTT has documented the importance of
urban plumes for forming sulfate, ozone, and light scattering aerosols
as measured by bscat. The chemistry of sulfate is fairly well under-
stood. However, more work is needed to characterize the transforma-
tion, removal, response to changes in relative humidity, and optical
properties of other components-especially carbon soot, organic aero-
sol, ammonium nitrate and coarse particles. Project MI-STT did not
cover a wide enough variety of .meteorological conditions, seasons,
and geographic areas to define the frequency of occurrence of urban
plumes or the geographic range of influence of urban plumes. There-
fore, additional field studies of urban plumes are needed to define
the chemical processes occurring, the visibility reducing species
I
-------
10
other than sulfate, the meteorological.conditions which cause an
urban plume to form, the distance over which urban plumes will have
an impact, and how urban plumes disperse into the general regional
background.
b. Forest fires. Wild and prescribed forest fires make a significant
contribution to regional haze both in the eastern and western United
States. Little is known about the mass and size distribution of
primary aerosol or the subsequent formation of secondary aerosol in
forest fire emissions. Additional field studies are needed to charac-
terize the behavior of forest fire plumes, to characterize the chemi-
cal and physical properties of the emissions, and to relate the
visibility effect to the type and mass of the material burned.
c. Natural Emissions. Natural and biological processes lead to
emissions of reduced sulfur species, nitrogen oxides, ammonia, and a
variety of organic vapors. Ammonia emissions are very important
since the concentration of ammonia may determine whether nitrates
appear as nitric acid vapor or ammonium nitrate participate and
natural emissions are probably a major source of ammonia. However,
ammonia as well as other nitrogen and sulfur compounds are being
looked-at in the Acid Deposition Program.
Organic emissions from trees and plants include species such as
terpenes which are very effective in forming organic aerosol. In the
western United States the organic fraction of the aerosol collected
on filters is approximately the same order of magnitude as sulfate.
In the eastern United States the organic fraction is smaller but
still on the order of 1/3 of the sulfate. It is also possible that
some organic aerosol is present in the atmosphere but is not found
on filters because it evaporates between the time of collection and
the time at which the filter is weighed. Additional laboratory and
field studies are needed to determine the aerosol forming potential
of natural organic emissions, the rate of natural emissions, and
the composition of the organic aerosol.
4.1.3 Small area sources
a. Diesel exhaust. Diesel exhaust is a major source of atmospheric
soot~Further studies are needed to characterize the size of the
diesel soot particles and the quantity of soot emitted for various
types of diesel vehicles; driving modes; ambient temperature, pres-
sure, etc.; and of the aerosol forming potential of gaseous emissions
from diesels.
b. Road dust. Fugitive emissions from traffic, largely expected to
be particles from dirt roads or from dirt.tracked onto hard surface
roads, make an important contribution to the local aerosol concentra-
tion. Further studies are needed to determine what fraction of the
road dust is within the fine particle range and will thus contribute
to regional haze.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
11
c. Agricultural burning. In some areas of the country prescribed
burning of fields is used as a routine technique to remove the previous
year's growth and prepare the soil for planting. As is the case with
forest fires, more information is needed on the quantity and properties
of the smoke.
d. Mood smoke from home heating. A substantial effort is underway
to determine the chemical and biological properties of wood smoke
from home heating as a function of fuel type and completeness of
combustion. However, these studies do not include the measurements
necessary to determine the effects on visibility.. A small amount of
additional effort, funded through the visibility program, would
enable a visibility field and laboratory component to be added to
ongoing studies of wood smoke.
4.2 Source Identification and Impact Evaluation
4.2.1 Receptor-models
a. General. This technique depends on utilizing detailed composition
measurements of ambient aerosols to determine their sources and the
contribution of each source. When the composition of individual
sources is known, the chemical species balance technique can be
applied to a single sample. When the sources or their composition
are not known various types of statistical analyses, such as factor
analysis and multiple regression analysis, may be applied to infer
types of sources. This technique requires multiple samples. The
output is a group of chemical species which show the same pattern
of variation in concentration from sample to sample. These groups
of chemical species may frequently be identified with types of
sources. .
For either of these techniques it is'necessary to know the
composition of potential sources, not only the major components which
give rise to visibility-reducing aerosol but also those minor com-
ponents which may be used to identify specific sources. The statis-
tical analysis approach may identify previously unexpected soruces
whose identity will have to be confirmed by comparison of experimen-
tally determined composition with that predicted. Thus, for all of
the sources discussed earlier, composition measurements need to be,
available to develop fingerprints or signatures for the primary particles
and the secondary products, especially for organic aerosols. For car-
bon soot it seems possible that microscopy may be able to differentiate
' between various source types.
b. Sulfate. Source-receptor models yield the concentration of
sulfate but do not provide a convenient way to distinguish sulfate
from various types of sources or from individual sources of the same
type. This problem arises since sulfate is a secondary product
formed by conversion of S02 to sulfate. The conversion rate depends
on a variety of atmospheric conditions as well as reaction time.
The Acid Deposition Program is supporting some work which might lead
to the ability to differentiate between sulfate from coal and sulfate
-------
12
from oil.combustion sources. Some work is also underway to determine
if different power plants use coal with trace elements which differ
sufficiently to serve as tracers of opportunity. However, even if it
is possible to use source apportionment to apportion primary aerosol
from various coal-fired power plants the problem of conversion would
limit the ability to'apportion sulfate to the various power plants.
c. Organicaerosol. Analysis of the organic composition of source
and ambient aerosols should help in apportioning this component of
visibility-reducing aerosol. However, the same problems exist as in
the case of sulfate. If the primary aerosol particulate differs in
composition from the secondary aerosols it would be possible to
appportion it. It might also be possible to use the composition of
the secondary organic aerosol to identify types of sources even
though one could not differentiate within, source types. However,
more extensive studies of the organic composition of ambient and
source aerosol is needed and should provide a useful tool for iden-
tifying sources. One objective would be to identify stable organic
compounds or groups that could be used as tracers.
d. Soot. Carbon soot has a variety of sources - diesel combustion,
oil combustion, gasoline combustion, forest fires, wood smoke, indus-
trial sources, etc. There is some evidence that the shape and struc-
ture of soot particles, as determined by optical or electron micro-
scopy, may be used to differentiate the various sources of soot.
This approach would appear to be a profitable one for research. The
C12/C14 ratio can be used to differentiate fossil carbon from modern
carbon. Also specific gases, e.g., methyl chloride, have been
shown to be potential tracers for wood smoke.
e. Volatile aerosols. Recent experimental studies using heated
nephelometers indicate that a significant fraction of the light
scattering aerosol is volatile below 100°C. Extinction budget analy-
sis by Mie theory and by regression analysis give differing weights
.to the various components. Organic particles and NH4N03 particles
are known to be in equilibrium with vapor phase material. These
observations suggest that a portion of the light scattering particles
in the atmosphere, when collected on filters, may evaporate either
during the collection process or during storage before the filters
are weighed. This problem may be particularly important in pristine
western areas where non-volatile sulfate particles are a smaller
fraction of the total light scattering material.
Experimental studies in the field and in laboratory simulations
are required to determine how important this process may be and what
chemical species and sources are involved.
4.2.2 Artificial tracers .
The major problem in the use of artificial tracers is that there is no
convenient tracer which converts from gas to aerosol in the same manner as
S02. Within the Acid Deposition Program studies are investigating the use
of sulfur-35 (radioactive) or sulfur-34 (non-radioactive) for use as tra-
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13
cers which behave in the same manner as S02. If these were available
they would provide a very useful way of tracing individual sources of $02
and determining the amount of sulfate particulate which came from each one.
There are also possibilities of using transitional metal oxides added to
to coal combustion or stacks of coal-fired power plants as tracers.
These provide information on the transport and dispersion of power plant
plumes and the apportionment of the primary particulate but do not tell
us anything about S02 or other gaseous transformation rates.
4.2.3 Back trajectory methods
Source apportionment techniques provide information on the types of
sources. To determine the geographical location of the source it is
necessary to utilize back trajectories showing where the air came from.
Additional work is needed to provide state-of-the-art, user friendly
models with this capability. The models should allow flow in different
directions within at least four separate layers within the boundary
layer and additional independent flow in a ground removal layer and
above the boundary layer zone. The routinely available wind data
provides information at 6:00 and 18:00 GMT. To obtain accurate
trajectories it is necessary to use empirical factors or dynamical
models to interpolate the wind field between the twice daily radiosonde
data and account for the nighttime jet and the midday drag from
convective mixing. Back trajectory models could^developed that can t^
be run on a personal computer. The technique needs to be applied to
existing field data. New studies are needed to validate multi-day
trajectories completed by these models. ;'
4.2.4 Combinations
Combined applications of artificial tracers and tracers of opportu-
nity, back trajectories, and source receptor techniques based on chemical
composition or microscopy need to be developed and applied. New field
studies are needed to obtain the specific data in terms of emissions,
source and ambient composition and wind information needed to apply these
techniques.
In particular, methods described in Section 4.2.2 and 4.2.3 are
sometimes used to help investigators determine whether source profiles
derived statistically from factor analysis are reasonable. However,
derivation of profiles from factor analysis.is still very much a
subjective art. In order for more quantitative techniques (such
as chemical mass balance) to be used by program offices to apportion
fine particulate observations at rural/remote sites a protocol or
distinct set of rules needs to be derived regarding the application
of the methods in 4.2 to regional scale analyses. Such work is
needed to reduce the subjectivity and increase the reproducibility
of receptor .methods which are applied in regulatory programs to
address regional scale problems.
-------
14
5.0 DETERMINISTIC MODELING
5.1 General
a. Visibility is sensed instantaneously although the effect may
"persist" for a finite time (e.g., many hours or days for haze). It
is not the result of accumulated deposition or average concentration.
Therefore, visibility models should simulate least short time averages.
The concentration pattern of a visible plume changes rapidly. Instantaneous
optical properties are needed for plume optics models. On a larger
scale, such as for regional models, aerosol concentration changes
more slowly. However, one to three hour averages are needed.
Standards for fine particles, set to protect visibility, may be
given in terms of 24-hour or longer averages. However, models will
be needed to predict shorter term visibility in order to determine
improvements in visibility produced by a variety of possible control
strategies for fine particle concentrations and averaging times.
b. Regulatory analysis models are needed which can be run fre-
quently, cheaply, and quickly so that the effect of changing levels of
emissions and distribution of emission sources can be determined. In
the case of regional or mesoscale models it should be possible to
conveniently establish source-receptor relationships. It will often
be important to determine what sources contribute to visibility reduc-
tion in a given region or to determine the influence of a specific
source.
c. The performance of a model is best established by comparing the
model simulation with real world measurements.
5.2 Plume optics models
Several plume optics models exist and their results have been inter-
compared. The model which has had the most use and is most readily avail-
able to the scientific community is called PLUVUE. The major need is an
improved technique for determining the often very short-term plume
shape construction and the visual as well as pollutant concentration
pattern, especially in other than simple terrain situations.
5.3 Regional models
5.3.1 Requirements for regional visibility models
(1) Averaging time should permit resolution on the order of
one up to three to six hours.
(2) The model should have a removal layer where dry deposition
occurs, multiple layers in the boundary layer to a 1-1 ow for wind
sheer and at least one and preferably multi-layers in the trans-
port zone above the boundary layer.
(3) The model should be able to account for formation of sul-
fate which is the key aerosol species.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
15
(4) Advanced versions of the model should be able to handle
carbon soot, organic aerosols, nitrate, other fine mass and
coarse mass in that general order of importance.
(5) The model must allow the S02 transformation rate to vary
as a function of light intensity, cloud interaction, relative
humidity and aerosol content. Full-scale chemical transfor-
mation mechanisms for gas phase transformations provide no im-
provement over parameterization unless the full chemistry for
cloud droplet reactions and reactions in aerosol mechanisms are
included with the same degree of sophistication.
(6) There must be a convenient technique for determining source-
receptor relationships.
(7) There must be a mechanism for taking aerosol concentration
and converting it to optical parameters. This will require some
way of accounting .for the size distribution of the aerosol al-
though a complete aerosol dynamics capability may not be necessary
for the initial version of a regional model.
(8) The regional model should have a realistic treatment of sub-
grid scale phenomenon involving urban and power plant plumes.
(9) It should be possible to run the model easily and inexpen-
sively on readily available computers since multiple runs and
multi-year runs will be required to investigate regulatory
strategy options. *
(10) It would be desirable to have a user friendly version,
perhaps simplified for use on a personal computer.
5.3.2 Model availability
There are a number of regional models which have been intercom-
pared in a study resulting from the US and Canadian interest in acid
rain. From this intercomparispn there appears to be a number of
models which do reasonably well in predicting wet deposition of sul-
fate and presumably would also do a reasonable job on sulfate con-
centrations which are a prerequisite to predicting visibility
impairment. There are also several models which have been developed
specifically for visibility work. Most of these models fall into
two types, Lagrangian and Eulerian, which have there own specific
advantages and'disadvantages. The Lagrangian model provides for
convenient source-receptor relationships but cannot easily handle
non-linear processes. The Eulerian model is convenient for introducing
non-linear processes but does not provide, a convenient techniaue
for determining source-receptor relationships. It also requires
some form of parameterization, not yet well established, to account
for sub-grid processes. A third type of regional model developed
especially for visibility uses a Monte Carlo technique which combines
some.of the best features of both the Lagrangian and Eulerian models.
-------
16
Current model development work Is primarily in two areas. First
is the acid'rain model being developed at NCAR. This regional acid
deposition model, RADM, is expected to be a highly sophisticated
model including all of the processes which are known to be important
in determining wet and dry deposition of sulfate and nitrate on a
regional scale. However, it is not expected to be available for
general use since the computer requirements and complexities are
such that it is expected to be run for only a few episodes and only
at NCAR. NCAR will also develop an "engineering" model which will be a
mathematical model of the RADM. It is expected that this model will
have only three layers, including only one layer within the primary
transport layer, and will not provide a simple way to understand the
various processes that are occurring. It will also be designed for
the acid rain phenomenon for which monthly averages or seasonal
averages are adequate. Therefore, it is not anticipated that it will
serve the needs of visibility modeling, especially in regard to
averaging time, accuracy of trajectories, convenient source-receptor
relationships, and inexpensive running cost.
Another model development effort is the regional fine particle
model, an in-house ASRL project. This program will utilize the
regional oxidant model as a starting point and will add chemical
transformation modules for the conversion of S02 to sulfate and
eventually for the formation of nitrate and organic aerosol. Full-
scale chemical kinetic models will be used for the transformation
processes comparable in sophistication to those used in the regional
oxidant model. This model will also include a full-scale simulation
of aerosol dynamics in that the formation of condensible species and
the condensation and coagulation processes will be simulated. This
will be a very useful research model but will not satisfy the needs
of the visibility modeling program. It is currently envisioned that
it will only have three layers, a dry deposition layer, one layer
within the boundary layer and one layer above the boundary layer.
This will be adequate for longterra averages but not for modeling
episodes as is required of the visibility model. It also will be a
very sophisticated model which will require highly trained and skilled
personnel to run as well as large computers and long running times.
5..3.S Development of a regional visibility model
It had been initially hoped that either the acid deposition pro-
gram or the regional fine particle modeling program would provide
the basic model which could then be adapted for visibility. It
appears that the cost and complexity of using these models and of
adding a visibility module are considerable. Therefore, consid-
eration should be given to utilizing the existing Monte Carlo
visibility model and making some improvements to it as an alter-
native to trying to utilize either the acid rain or regional fine
particle modules. Desirable enhancements to the Monte Carlo model
include: 1) Add carbon aerosols. An emission inventory for carbon
soot will be required no matter what model is used and can be easily
added to the Monte Carlo Model. 2) Add a hydrocarbon/NOx ratio .
parameterization as a method for predicting ozone. This is a relatively
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
17
simple technique which will not greatly increase computational
cost and will provide an index of the oxidizing power of the
atmosphere which can be used especially for improving the cloud
conversion process and aerosol conversion process. 3) Add a dynamic
model to use for wind field interpolation. 4) Add a parameterization
for organic aerosols.
5.4 Special situations
There are a number of special situations which may require either
specific models or development or modification of existing models. These
include: (1) area source impact, (2) point source impacts, (3) fugitive
source impacts, (4) techniques for examining impact of multiple sources,
.(5) visibility impacts in complex terrain situations, (6) visibility
impacts during extended stagnations within a local or mesoscale area.
It seems likely that there is such a variety of special situations that
the models should be put together to satisfy the specific situation
or problem rather than trying to anticipate situations and devise
models for all of them. For most of these situations there will be
models which will predict the aerosol concentration under some type of
averaging situation. The visibility program should provide a technique
for converting aerosol concentrations to visibility parameters.
One situation, intermediate between urban models and regional models,.,.
occurs frequently enough to justify separate consideration. This is the *
case in which the urban (or other type) plume, because of stagnation or "
complex terrain situations may move back and forth across an area for 'f
several days building up pollution to serious levels. Such situations :'
occur in a number of western areas. Thus, while the area covered may
be mesoscale, the chemistry and aerosol dynamics in the model must be
capable of handling multi-day reaction sequences.
Simple optical models are probably adequate_for calculating,visibility
from aerosol concentration in the well-mixed haze frequently encountered
in the east. In western areas, however, elevated haze layers and other
non-uniform situations require a more complex accounting of optical
properties of the atmosphere. Optical models need'to be developed for
these situations.
5.5 Conservative screening techniques
It would be desirable for many of these situations, especially the
plume optics and the-regional models, and some of the area and point
source impacts, to develop conservative screening^techniques which could
be used to determine if more sophisticated modeling is needed.
5.6 Model Evaluation
This should involve verification of sub-modules or components of a
larger model as well as comparison of the ultimate model predictions
against measurements from various real world networks.
-------
18
5.6.1 Visibility
For visibility studies we are fortunate in having a remarkable
data base, that of the National Weather Services human observer
visibility measurements. These are available every hour from several
hundred stations within eastern North America. While the human obser-
ver visibility data do not have the resolution that might be obtained
from an instrumental measurement, it does provide adequate resolution
for model evaluation. While there is some criticism of the human
observer data as being subjective and perhaps not always accurate,
there are a number of reasons for believing that this is a very
useful and reliable data base. Analysis of this data base has indi-
cated that there are large geographical regions, each covering perhaps
several states, in which 5-10 National Weather Service Stations give
the same pattern of visibility changes. When one receives compar-
able data from a number of stations the confidence in the individual
data bases is increased. The NWS data also agree with measurements
from satellites which show the area and intensity of low visibility
episodes. Furthermore, comparisons of sulfate concentrations for
those periods and areas in which network data is available show a high
correlation with the National Weather Service human visibility data.
Further work in using the results of the NWS human observer network
and of satellite measurements of the hazy visibility blob are needed
to provide data bases for validation of regional models.
5.6.2 Transport
The transport and dispersion module of a model can frequently
be verified independently of the chemistry or optics component. For
regional scale models there is a need for data on the transport and
dispersion of tracers, especially over multi-day periods so that the
dispersion processes of the diurnal cycle are included. There have
been two studies in which a few releases of a perfluoro compound have
been made and tracked on the order of 500 to 1000 miles. Additional
studies of this type are probably needed although they are fairly'
expensive. Currently there are no plans for further studies of this
type but they will be necessary to validate the current hypotheses
regarding long-range transport (transport beyond several hundred
miles). Tracer studies are needed particularly in the pristine
West where long range transport of urban, forest fire, and smelter
plumes are thought to be especially important and where complex
terrain effects on transport cannot be predicted with confidence.
Another way to get at these processes,involves the use of a
fluorescent aerosol tracer with measurement by lidar. This has an
advantage over the flourbcarbon tracer in that one may map out the
. entire three-dimensional structure of the plume and follow it as it
moves and disperses. The fluorocabon tracer can only be measured as
an integrated sample at a point. .
In view of the limited extent and high cost of tracer studies,
it is desirable to seek additional means of obtaining data to check
transport modules. Analysis of NWS visibility distance data and
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
19
observation of hazy areas in satellite images indicate that during
certain episodes, especially in the summer and fall, hazy air masses
or blobs form and move around the country. These blobs have dimen-
sions of hundreds of miles, persist for several days and have surpris-
ingly sharp edges. A regional model, even with a crude sulfate
formation mechanism, should be able to give the outline of the blob and
its motion if the transport module is accurate. This combination of
NWS visibility data and satellite images provides a cost-effective
technique for obtaining a data base to evaluate the transport module
of regional models.
5.6.3 Chemistry
Chemistry modules can be tested from smog chamber studies or from
extensive measurements from platforms moving in power plants or urban
plumes. Full-scale predictions which include chemistry, transport
and dispersion can be evaluated against measurements from ambient
monitoring networks. Unfortunately the network data are not always
adequate and when such data are available they frequently do not have
adequate spatial or temporal coverage.
5.6.4 Error Sensitivity Analysis
Analysis of models for error sensitivity is also needed. Within
the Acid Rain Program protocols are being developed for model evalu- -
ation and verification. When these are developed they should be
applied to visibility models.
ft
5.6.5 Model Comparison
It should be possible to increase the reliability of both pre-
dictive and inferential (source-receptor) models by applying both
types of models to the same data base and comparing results.
5.7 Emissions
The Acid Deposition Program has a major effort underway to develop
emission inventories. Task Group B will provide anthropogenic emissions
of S02 and N02. Probably NH3 and ozone-forming hydrocarbons will be
included. Task Group A will provide information on natural emissions
of sulfur and nitrogen compounds. However, the Acid Deposition Program
has minimal interest in emissions of carbon soot or specific organic
compounds which are important aerosol precursors. Similarly, natural
organic emissions which may be significant aerosol precursors, especially
in the pristine west, are of little concern because they are not thought
to be important in forming ozone. Therefore,'the visibility program must
make provisions for obtaining inventories of these emissions as well as
those from forest fires and agricultural burning. Emission-s of wood
smoke from home heating and road dust are being examined in other pro-
grams. However, some augmentation will be required to insure that the
information needed for visibility models is obtained. For pristine
western areas coarse aerosols sometimes account for a significant amount
of extinction. More information is needed on both natural and man-made
emissions of coarse particles.
-------
20
5.8 Wind Field
Most existing air quality models use a wind field derived from
National Weather Service measurements of the upper air wind speed and
direction. These measurements provide a snapshot of the wind speed and
direction in the morning'and evening (0 and 12 GMT, 6 am and 6 pm EST).
Upper air measurements are made at over 60 stations. The data must be
interpolated in time and space to provide wind fields for models. These
interpolations are known to be inaccurate because the wind speed slows
down during the middle of the day due to turbulent mixing in the boun-
dary layer and speeds up at night due to lack of vertical momentum trans-
fer. Also the geographic coverage is inadequate for our modeling purposes.
This is an especially severe problem in the pristine West where fewer
stations are located and where complex terrain features produce local
and mesoscale effects on the wind field.
An attractive line of research is to use dynamical models to inter-
polate between NWS observations. A dynamical model is being used in the
NCAR regional acid rain model. Calculations have_shown that trajectories
obtained from dynamical wind field models are different from those ob-
tained from empirical wind field models. However, tracer studies will be
required to demonstrate which wind field model gives the best agreement
the NCAR dynamical model to eastern visi-
of a dynamical model for western visibility
significant improvement in the reliabilty
with reality. Application of
bility models and development
models would probably yield a
of visibility models.
Additional upper air measurements are also needed. In the pristine
West more stations are needed. At all sites additional measurements at
midnight and noon would be helpful.
Summary of Reserch Recommendations
A. Empirical Approaches to Source Assessment
1. Specific Source Categories - Developmental Improvement of Techniques
and Guidance
a. Large Point Sources
b. Large Area Sources
c. Small Area Sources
B. Empirical Approaches to Source Identification
1. Receptor Oriented Approaches - Development and Improvement of
Techniques and Guidance
a. Aerosol Components '
b. Back-trajectory Models
c. Data Bases for Model Validation
d. Combinations and Comparisons of Techniques .
and Models
'6. Extinction Budget Analysis
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2. Deterministic Modeling
and Guidance
21
- Development and Improvement of Techniques
a. Generalized Visibility Modules for Regional and
Other Aerosol Models
b. "Monte Carlo" regional model
c. Western Regional Model
d. Data Bases for Model Validation
6.0 VISUAL VALUES
6.1 Psychological Values
6.1.1 Natural Settings
A primary consideration in determining the adversity of an air
quality impact on a scenic vista is a determination of whether or not
the quality of the visitor experience and/or enjoyment of a class I area
is compromised. Assessing visitor experience and enjoyment raises
several interesting questions. For example, how can the quality of
the recreational experience be measured? Is good visual air quality
an important component of the recreational experience? If so, how
important is it? How might a given change in visual air quality
affect visitor satisfaction with the visual resource? How might the
impact on satisfaction with the visual resource affect the overall
quality of the recreational experience?
Employing recreation demand theory to assess the quality of a
recreational experience is one way to approach the problem. The
basic premise of demand theory is that individuals engage in behaviors
which will maximize their psychological well-being. If they feel
their psychological well-being is less than optimal and that a
recreational experience will restore their sense of well-being, they
will seek out a recreational experience and engage in behaviors which
will provide the needed restoration.
The recreational behaviors a person might engage in to reach a
higher state of psychological well-being and the relationship between
these behaviors and the environment can be best visualized with a
four level demand hierarchy. The first level is the demand to
participate in a desired activity. The second is to do. so in an
environment which offers both the type and quality of attributes
desired. The third is to experience short-term psychological
satisfactions which come from interacting, with desired environmental
attributes during activity participation. The fourth level is to
realize long-term benefits which might come from a satisfying
experience. "
Results of the research conducted so far have begun to provide a
better understanding concerning the importance of good visual air quality
to the recreational experience sought at National Park Service Class I
areas. The results have also begun to provide an understanding of
-------
22
4. Assess the long-term benefits of good visual air quality.
5. Develop, where possible, quantifiable relationships between
visibility "perception" indices and visitor demand for attributes
and visitor satisfaction with the quality of these attributes.
6.1.2 Urban Settings
I
relationships between good visual air quality and visitor satisfaction _
with the visual resource and the quality of the visitor experience I
and the level of visitor enjoyment. However, more complete answers
to questions such as those posed earlier are needed to prevent the
quality of the recreational experience from being negatively impacted
.by possible reductions in good visual air quality in class I areas. |
Specifically, additional work is needed with the following emphasis:
1. Further test the hypothesis concerning the naturalness cluster I
of attributes and determine a more complete picture of which
attributes are part of this naturalness.
2. Establish the importance of visibility-related attributes in |
various types of class I areas.
3. Develop relationships between good visual air quality and I
visitor satisfaction with the visual resource and the recreational
experience.
I
I
Studies of willingness to pay for improved visibility in urban
ares, urban perception surveys, and results of visitor studies-in non- -
urban settings suggest that improved visibility would increase psychological |
well being in urban areas. Despite increasing work on the effects of
illumination on behavior and mood, research on the effect of visibility on
psychological attributes in urban settings is almost non-existent. Approaches
to identifying and evaluating potential effects are needed.
6.2 Economic Assessment of Visibility Values
I
The visibility benefit research summarized in Chapter 4 points to
two principal limitations in the economic methodology for estimating «
dollar values of better visibility. -The more general problem transcends
the visibility application. As research in the visibility area has
repeatedly indicated, and as critics (e.g., the UARG sponsored reviews)
have pointed out, contingent survey approaches are poor substitutes for
market data, and may not provide unbiased answers. Thus, further work to
demonstrate the relationship between market based and contingent valuation
(CV) based methods is essential, even though such calibration, by definition,
cannot be performed on a norimarket good like visibility. Second, there
are a variety of unresolved questions from the visibility benefits research
itself that should be addressed. The results of current benefits analysis
are quite sensitive to the resolution of some of these questions.
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
23
6.2.1 The General Issue
CV and related analyses continue to be controversial, given the
potential for a wide variety of biases associated with surveys in general
and the contingent technique in particular [See Schulze et al. (1981) or
Randall et al. (1981)]. Bishop (1979) heightened concern with his finding
that in an experimental comparison between real cash payments and willingness
to pay (WTP) elicited from a survey, the average WTP was over 50 percent
higher than average realized cash payments. Though more recent unpublished
work by Bishop finds no statistical discrepancy in survey derived estimates,
the issue is not yet satisfactorily resolved. Thus, further work comparing
'market and contingent techniques on goods traded in well established
markets (thus, excluding visibility, for which market data are unavailable)
is recommended. It would be particularly useful if a framework similar
to the Bishop design were expanded to include comparisons among question
formats and the CR method.
6.2.2 Specific Issues on Visibility Valuation
0 Valuation in Multiple Good Setting: Recent studies have indicated
that respondents' valuation for a specific environmental .good
falls when competing environmental (or potentially nonenvironmental)
claims are included in the survey process [see Randall et al.
(1981) and Rae (1984)]. A split sample survey valuing visibility ;
in isolation and in competition with other goods is recommended
to resolve this uncertainty.
0 Specifying Visibility in Distributional Terms: Most of the
existing visibility benefits work, excepting Loehman (1981) and
Rae (1984), is based on respondents valuing an average change in
visibility in miles. Visibility, however, is a distribution of
daily conditions. It is quite conceivable an individual would
assign a low value to a modest average change in visual range,
but would value highly visibility improvements on particularly
poor days, especially in the summer or when recreation is
involved. It is not clear when these poor days are averaged
over a year, that the resulting aggregate valuation is the same
as when based on average visual range improvement. Further research
on the depiction of visibility conditions to respondents is
necessary to resolve this issue.
9 Relationship between visibility and recreational (or other)
activities: While research in this area has been undertaken by
Tolley et al. (1984) and by the National Park, Service, further
efforts to explore this complementarity are likely to. be productive.
0 Nonuser and Existence Value: Empirical research on nonuse values
for visibility has been problematic, as the results of- Randall
et al. (1981) clearly demonstrate. While this is a "high risk"
research area, it is of sufficient importance to try to narrow
the existing uncertainty.
-------
-------
I
I
I
I
I
APPENDIX F. SUMMARY OF PUBLIC COMMENTS
I
I
I
I
I
by: John Bachmann, EPA/OAR
Jeffery Smith, Consulting Attorney
I
I
I
I
I
I
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
APPENDIX F
TABLE OF CONTENTS
Summary of Written Comments
I. Goals, Criteria, and Definitions
II. Alternative Regulatory Strategies
III. Visibility Research
IV. Procedural Matters
Summary of Public Meetings of the Interagency Visibility
Task Force - Memorandum
Denver Meeting
Formal Presentations
Open Discussion
Goals for Western Visibility
Alternative Control Strategies
Research Needs
Washington Meeting
Goals
Alternative Control Strategies
Research Needs
Attendees
1
4
10
14
1
1
3
3
6
7
8
9
9
11
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
SUMMARY OF COMMENTS
This summary covers written comments EPA received in
response to its November 9, 1984 (49 Fed. Reg. 44770),
notice soliciting public comment on the issues and
alternatives being addressed by the Visibility Task Force
with regard to its charge of developing a long-range
strategy for addressing visibility impairment caused by
pollution-related regional haze. This summary does not
cover comments presented at the public hearings announced
by this same November 9 Federal Register notice.
This summary groups the comments into the three
general categories of issues the commenters addressed: (1)
the proper goals, evaluation criteria, and definitions of
visibility-related terms the Visibility Task Force should
use; (2) alternative regulatory strategies for addressing
regional haze; and (3) visibility research.
I. GOALS, CRITERIA, AND DEFINITIONS
Goaljs. The National Parks and Conservation
Association, the St. Cloud Environmental Council, and the
West Michigan Environmental Action Council urged EPA,
generally, to move forward firmly to control regional
haze. The West Michigan Environmental Action Council noted
also that such impairment can significantly disrupt the
enjoyment of the environment and adversely affect economies
through the loss of scenic vistas.
Several members of the oil industryChevron, the
American Petroleum Institute (API), and the Western Oil and
Gas Association (WOGA)--stated that visibility protection
for mandatory Class I areas should not be extended to other
areas without a further Congressional mandate. Chevron and
API noted that environmental groups participating at the
-------
Denver public meeting (December 5, 1984) agreed with this
position.
UARG stated .that a goal should be to ascertain with
much greater certainty than currently exists the trade-off
the public is willing to make between the cost of
electricity and the value of visibility. The public must
make this trade-off because both electricity and visibility
are "goods" used by everyone. That is, UARG explained, one
cannot view the regulation of visibility improvements as a
way to internalize the costs of electricity production
because those costs are already internalizedelectricity
users (the public) must bear all of the costs of
electricity production including the financial cost and
whatever visibility impairment that it causes.
Evaluation Criteria. API and WOGA agreed with each of
the Visibility Task Force's proposed criteria for
evaluating alternative strategies for visibility
regulation, but urged EPA to define precisely the nature of
each criterion. For example, these commenters stated, the
"cost-effectiveness criterion" should be expressed in terms
of the cost of the strategy per degree of visibility
improvement. API stated that this concept is not new,
noting that the San Francisco Bay Air Quality Management
District's criteria for setting control priorities to meet
ambient standards is $/ppm decrease in pollutant
concentration.
The Oregon Forest Industries Council stated that it
was important to tailor any long-range regional haze
strategy so that it would maintain and improve the economic
viability of the forest industry, which is just .emerging
from a difficult economic recession. This commenter also
stated that it was important to develop cost-benefit data
for each strategy.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The Colorado Mountain Club, on the other hand,
stressed the economic importance of natural resources. It
noted that Colorado's natural recreational
resources including high peaks, scenic areas, good
visibility, and clean airsupport a stable, state-wide
tourism and outdoor recreation industry of major economic
importance. A cohesive recreational management policy is
needed for these resources, the commenter stated, most of
which are on public lands managed by diverse agencies.
Definitions. Salt River Project expressed
considerable concern about the definition of "visibility"
and "visibility impairment." Specifically, Salt River
Project encouraged EPA to refine its concept of regional
haze based on the commenter*s analysis of National Park
Service photographs, which revealed three types of haze:
(1) uniform haze (characterized by a discontinuity in the
vertical distribution, and which corresponds most closely
to EPA's concept of regional haze); (2) ground-based
layered haze (which occurs predominantly in winter); and
(3) elevated, layered haze (which is somewhat synonymous
with visible plumes, but which has an origin that is not
necessarily tied to emission sources). Salt River Project
stated that EPA may also require a "no haze" haze-type,
which would occur when none of the three preceding
haze-types is perceptible.
In refining these definitions, Salt River Project
stated that EPA should establish a dialogue with the
scientific community to arrive at a consensus for the
definition of visibility arid visibility impairment. This
consensus agreement would then be submitted to an advisory
subcommittee of the EPA's Science Advisory Board.
-------
Also, in connection with these definitions, Salt River
Project urged EPA to adopt a concept of "prevailing
background" since it is practically impossible to establish
a true natural background. The prevailing background,
being variable in space and time, is therefore
site-specific and should be considered as a time-averaged
value at a given site, Salt River Project asserted.
Chevron and API stated that EPA should specify
procedures for the federal land managers to follow in their
adverse impact determinations. API stated also that EPA
should give quantitative parameters--which could be
area-specific--to the federal land managers to judge what
is an "adverse" impact requiring mitigation. API
stated that it agreed with the Environmental Defense Fund's
comment (at the Denver public meeting) that it is critical
to focus on the definition of good and bad visibility.
WOGA also stressed the need for an objective standard
against which to measure adverse impacts on visibility.
II. ALTERNATIVE REGULATORY STRATEGIES
Integrated Approach. Chevron, WOGA, and API agreed
with EPA that any regulatory program to protect visibility
from regional haze should be integrated with other control
programs that might affect visibility. WOGA noted in
particular that EPA should credit NSPSs with visibility
improvement as part of the integrated program concept.
Chevron and API explained that an integrated approach
(1) provides a mechanism for EPA to consider the cumulative
air quality benefits of the broad range of alternative
regulatory strategies, and (2) results in the most
cost-effective controls being implemented, as well as
helping form the basis for more equitabale regulatory
strategies.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Public Willingness to Pay. UARG^ stated that EPA
should use the public's willingness to pay for visibility
improvements to measure the adversity of any visibility
impairments. If the public is not willing to fund the
improvement, the status quo cannot be regarded as a state
of adversity. Also, UARG stated, if there is no adverse
effect on the public, no secondary ambient standard can be
justified.
In this regard, UARG stated that it has assessed the
errors and uncertainties in the draft report prepared for
EPA by SAI entitled "Visibility-and Other Air Quality
Benefits of Sulfur Dioxide Emission Controls in the Eastern
United States" (the SAI report). This assessment reveals
that even a 12 million ton per year S0£ reduction scenario
would produce benefits of less than one percent of the
multi-billion dollar price tag for controls. Since these
costs would be wholly out of proportion to the benefits, no
SC>2 controls can be justified, UARG concluded.
Integral Vistas. API and WOGA stated that EPA should
delete the concept of integral vistas since it is not
authorized by the Clean Air Act.
Forest industry representatives agreed: the Oregon
Forest Industries Council stated that protection of
integral vistas created the potential for giving decisional
authority about local land use matters to the federal
government, and the Industrial Forestry Association stated
that protection of integral vistas could curtail essential
slash burning even though the resulting smoke may have
little impact on. a Class I area itself.
-------
I
I
Modification of Clean Air Act Section 169A. API and
WOGA stated that EPA should not seek legislative
modification of Section 169A to explicitly |
covet protection of visibility from regional haze. API
stated that it would take time to obtain congressional I
action; WOGA noted that EPA should take a hard look at
Section 169A to see if protection of visibility from
regional haze is covered. *
Clean Air Act Section 126. API and WOGA stated that |
EPA should not seek any legislative change to Section 126
since that authority is presently receiving a thorough I
examination in connection with interstate ozone and acid
rain impacts. API noted further that there is inadequate
knowledge about visibility impairment for it to be a
driving force for legislative change to Section 126, and «
stated that Section 126 was not intended as a vehicle to I
regulate regional haze.
Episode Control. API and WOGA supported a legislative
amendment to Clean Air Act Section 123 that would allow I
seasonal variability of emission limitations.
Emission Trading. API and WOGA supported creating
trading rights between new and existing sources over a wide _
area as a cost-effective approach to solving many regional |
air pollution problems.
The Industrial Forestry Association also supported the
concept of emission trading, which it stated could be
applied to essential slash burning operations such that if '
future prescribed burning emissions' are below the natural «
pristine levels of smoke from wild fires, the future |
emissions could be traded for natural baseline levels.
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
On the .other hand, .the American Mining Congress (AMC)
stated that a flexible NSPS that would allow emissions
trading between copper smelters and electric utilities is
not economically or politically feasible. AMC noted that
(1) SAI's economic analysis regarding this concept failed
to take into account the large financial commitments
already made by several companies in the industry to
achieve state implementation plan (SIP) compliance, and (2)
choosing a lower and more realistic baseline would
necessarily decrease the calculated savings attributable to
this trading policy. A full analysis, AMC stated, would
show essentially no difference in capital and operating
costs at any given level of sulfur dioxide emissions for
smelters and powerplants.
Also, AMC stated, there has been no field measurement
to confirm the assertion that visibility would be improved
under an emissions trading policy by more vigorous controls
on smelters because of an assumed beneficial effect of NOX
on the oxidation rate of sulfur dioxide emissions from
powerplants. AMC also noted that smelters could not cause
the urban plume blight problem, which is caused by urban
nitrogen oxide emissions. In addition, there would be
political resistance to interstate trading. (AMC noted,
however, that its concerns do not apply to the currently-
allowed policy of emission trades.)
Prevention of Significant Deterioration (PSD) Program.
WOGA and API opposed the idea of requiring minor sources to
undergo visibility review as part of the PSD preconstruc-
tion review; EPA instead should publish de minimus criteria
for exempting small projects using, for example, a version
of the Level I screening-criteria in ElPA's "Work-book for
Estimating -Visibility Impairment."
-------
WOGA noted also that redefining the PSD participate
matter increments to realize regional haze benefits may
result in a multitude of different increments for the
numerous haze conditions of Class I areas, .thus further
complicating the PSD program. WOGA stated that it agreed,
however, that states and federal land managers should be
allowed to set state-wide, regional, or local visibility
objectives implemented through the PSD program if EPA
provides technical guidance such as objective visibility
standards and a validated reference method.
Non-Traditional Sources. API and WOGA agreed that if
visibility controls are warranted, the most cost-effective
controls should be implemented first whether they involve
traditional or non-traditional sources.
Primary Standard for Visibility. API stated that it
was premature to recommend a primary standard for
visibility to Congress because (1) many areas of the
country are still trying to attain the primary health
standard, and (2) the Visibility Task Force should
thoroughly evaluate all the visibility benefits associated
with controls for attaining the primary national ambient
air quality standards before suggesting visibility-specific
^controls. API also noted that such a course of action is
probably unnecessary since visibility will likely improve
in the future due to continued implementation of current
program control such as NSPS.
Secondary National Anbient Air Quality Standard
(StiAAQS) . WOGA stated that promulgation of a SNAAQS is not
a workable strategy to reduce regional haze because (1) EPA
would have to define many different SNAAQS to cover the
array of unique haze characteristics throughout Class I
areas, which would be too complicated and time-consuming;
and (2) instead, EPA should propose a specific, objective,
8
-------
and validated reference method to measure compliance with
such a standard.
Retiring Old Plants. API stated that it was concerned
that a program feature of retiring old plants could cover
refineries and chemical plants, as well as utilities, when
there are better and more cost-effective control approaches
for the former facilities.
Regional Strategies. The Oregon Department of
Environmental Quality stated that visibility strategies
must be regional in nature to reflect the "visual value
differences" among regions.
The Oregon Forest Industry Council agreed, stating in
particular that EPA should allow the present liaison among
Idaho, Oregon, and Washington to develop the most
cost-effective strategy for the Northwest to reflect the
uniqueness of the area.
Prescribed Burning. The Industrial Forestry
Association stated that the Visibility Task Force must
recognize the ecological, biological, and management
relations between fire and natural resources in any
visibility regional haze programs it proposes; otherwise,
the resulting regional haze visibility rules could (1)
increase the incidence and severity of wild fire, (2)
exacerbate the acute air quality degradation caused by
natural fires that may be more harmful to human health, (3)
contribute to a loss of annual habitat maintained by
natural and prescribed fires, (4) increase insect and
disease outbreaks, (5) create relatively greater
environmental degradation of other resources such as soil
productivity and water quality, and (6) unknowingly
influence professional resource management decisions.
-------
The Oregon Forest Industry Council noted that
prescribed forest residue burning control techniques have
improved greatly since the late 1970s.
The Industrial Forestry Association noted, as
discussed above, that emission trading could be used to
allow the continued practice of essential slash burning.
III. VISIBILITY RESEARCH
Current Research. AMC, Chevron, and API stated
generally that current research is inadequate to provide
sufficient data and analytical tools to show that a program
to protect visibility from regional haze is needed. Until
such data and tools are developed, these commenters stated,
EPA should delay adopting visibility-specific control
measures.
API and WOGA expressed specific concern with the
reference in EPA's November 1984 outline of alternative
regulatory strategies regarding the characterization of
regional haze trends as based on human observations at
airports from 1948 to the present. Such data, these
commenters noted, may not reflect accurate regional haze
trends because (1) the visual observers, their training,
their methods, and the visual range targets have changed
over the years; and (2) the observations of haze
characteristics were secondary to ensuring that airplane
landings and takeoffs could be done safely.
Similarly, API and WOGA expressed concern that the
same quality assurance problems may also exist in
projections of future regional visibility trends. Thus,
API and WOGA stated, the Visibility Task Force should
carefully qualify this data.
10
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In this regard, the Oregon Forest Industries Council
stated that suitable historical visibility trend data for
the Pacific Northwest should be obtained from forest
agencies, which possess data based on daily reports by
forest lookouts that include visibility estimates.
Salt River Project expressed concern that the planned
analyses of certain ongoing programs (VIEW, SCENES, and
RESOLVE) might be flawed because the meteorological data
that has been collected during those programs may not allow
detailed source-receptor-analyses. The reasons are, first,
that surface meteorological data are strongly influenced by
local drainage flows, and their use in inferring transport
directions is minimal. Second, the National Weather
Service rawinsonde network, which might provide better
indications of transport, could also provide misleading
results because the spacing between National Weather
Service rawinsonde stations of several hundreds of miles is
too large to depict haze transport.
The Utility Air Regulatory Group (UARG) expressedin
detailits concern regarding the draft report prepared for
EPA by SAI entitled "Visibility and Other Air Quality
Benefits of Sulfur Dioxide Emissions in the Eastern United
States." UARG noted two general problems with this
report.
First, UARG stated, the report mischaracterizes the
improvements in visibility that S02 controls can produce.
This results because the report overstates the ability of 8
and 12 million tons per year (mtpy) emission reduction
scenarios to benefit the public because, among other
things, SAI (1) actually modeled emission reductions
totaling 9.4 and 13.1 mtpy, respectively; (2) assumed that
502 emission reductions and changes in sulfate
concentrations are linearly related; (3) used an inflated
11
-------
value for the intrinsic contrast targets; and (4) did not
perform a comprehensive and reliable assessment of the
confidence that can be placed in its results.
The second principal flaw in the SAI report, according
to UARG, was that its estimates of the value of visibility
improvements are unreliable. In this regard, UARG
criticized SAI for examining reductions in electric utility
emissions and in regional haze as the vehicle to improve
visibility, when in fact both electricity and visibility
are "goods" used by everyone. Regulation of visibility
improvements cannot be viewed as a means to internalize the
cost of electricity production, UARG stated, because those
costs are already internalizedelectricity users (the
public) must bear all of the costs of electricity
production including the financial costs and whatever
visibility impairment that it causes.
Future Research. Because of the deficiencies in the
SAI report summarized in the preceding paragraphs, UARG
stated that the report cannot support a regulatory
program. If EPA decides to pursue the visibility issue, it
should initiate a research program to determine whether the
public is dissatisfied with current visibility, whether S02
controls can produce a visibility improvement that has a
significant value to the public, and how much visibility
improvement the public is willing to fund, UARG stated.
WOGA and API stated, generally, that EPA must commit
to developing and funding a comprehensive visibility
research program. API noted, in particular, the value of
»
EPA funding for the Visibility Research Forum.
12
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
API and WOGA also agreed generally with the research
projects the Visibility Task Force identified in its' April,
1984, report regarding the gathering and analysis of data
with respect to source-receptor relations and the
establishment of a. comprehensive monitoring and research
program. In particular, these commenters stated, research
is needed on parameters for aerosol-visibility relation-
ships, the sources and causes of visibility impairment in
the West, all high-quality visibility data, and development
of standard operating procedures and a reference method for
visibility monitoring techniques.
API also stated that EPA should give research on
volatile aerosols a higher priority since they are a
significant component of visibility impairment in the West
and little is presently known about them. In addition, API
stated, the Visibility Task Force should consider the
research and data being generated by the California Desert
Air Work Group in assessing directly the visibility impact
of the Clean Air Act non-attainment plan' programs of the
South Coast Air Basin and the San Joaquin Valley Air Basin
on visibility in the Southeastern California Desert.
The Oregon Department of Environmental Quality stated
that it was important for EPA to conduct research to (1)
establish a uniform scheme for regional haze measurement,
and (2) designate regional air basins (and require states
within those basins to adopt a regional-specific control
strategy).
Finally, Salt River Project proposed the following
priorities (from first, priority to last) for the research
programs the Visibility Task Force proposed for .FY 84: (1)
visibility parameters for aerosol-visibility relationships,
(2) visibility monitoring techniques, (3) mini-recession
S02/visibility analysis, (4) western visibility impairment,
and (5) aerosol absorption measurements.
13
-------
For FY 85, Salt River Project recommended the
following priorities (again from first to last): (1)
workshops/conference, (2) visibility parameters for
aerosol/visibility relationships, (3) visibility impairment
by volatile aerosols, (4) visibility monitoring techniques,
(5) western visibility impairment, (6) aerosol absorption
measurements, (7) source-receptor relationships, (8)
mini-recession S02 visibility analysis, (9) visibility
trends, and (10) regional visibility research network.
Finally, GAIA Associates stated that it hoped that EPA
will provide funding to the Visibility Research Forum to
t
support attendance by environmental groups at the next
year's program.
IV. PROCEDURAL MATTERS
Chevron, API, and AMC stated that EPA should expand
the role of visibility experts from industry and the public
in developing alternate regulatory strategies for
contolling regional haze. API and Chevron noted also that
EPA should continue to provide for public meetings and
public review and comment on developments (e.g., the
Western Visibility Benefits Study) as EPA continues to work
on programs to control regional haze.
The Oregon Forest Industries Council stated that the
Visibility Task Force should maintain close liaison with
the forest industry at national, regional, and state
levels. This is especially critical, the commenter stated,
when evaluating alternative strategies that will impact the
economic viability of the forest industry.
14
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The Oregon Department of Environmental Quality stated
that the Visibility Task Force should also maintain close
liaison with the joint effort by the states of Oregon,
Washington, and Idaho to study regional haze; this joint
study could yield important information on techniques for
measuring visibility, and the magnitude and sources of
regional haze in the Pacific Northwest.
Finally, GAIA Associates described in detail the
Visibilty Research Forum in order to supplement the
discussion of that group at the Denver public meeting.
GAIA stated that the Visibility Research Forum is organized
for the exchange' of information and ideas about current and
planned visibility research carried out in both the public
and private sectors of the United States. The Forum's
objective is to address the linkages among ongoing
visibility research, natural resource management, and
visibility and related air quality regulatory programs.
The Forum is interested in research efforts to develop
sound scientific information and planning tools appropriate
for regulatory and administrative processes, corporate
planning, and natural resource management. In addition,
the Forum provides a neutral setting for dialogue among
visibility researchers, managers, and planners in the
federal and state agencies and affected industry groups.
15
-------
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
January 3, 1985
MEMORANDUM
SUBJECT: Summary of Public Meetings of the Interagency Visibility Task
Force
I
I
I
I
I
11 \/» %>c yj *
FROM: John Bachmann *fa**j)
Task Force Chairmen
I TO: See Below /
I
I
I
I
I
I
I
I
I
I
I
I
The Interagency Visibility Task Force held two public meetings in
December to discuss the work of the task force and receive public suggestions
and comment. The meetings were announced in the November Federal Register
(attached) and both written and oral public participation was solicited.
Some 41 individuals and/or organizations requested the additional background
.materials mentioned in the notice. This memorandum summarizes the oral
presentations and discussions that occurred at each of the meetings, some
of which have implications for developing long-term strategy alternatives.
Further perspectives and insights from the public are expected in the
written submissions that are due by mid-January.
DENVER MEETING - DECEMBER 5, 1984
The Denver meeting was held in a conference room of the Customs House
and ran from 11:30 a.m. to 4:30 p.m. MST. Attendees (see attached) included
representatives of several task force agencies, a number of utilities,
mining and some other industries, two State Agencies, private contractors,
and environmental organizations. The discussants included a substantial
number of those present.
FORMAL PRESENTATIONS
John Leary - State of Colorado
The meeting opened with my brief overview of the background and mission
of the task force and order of discussion. Following this introduction,
John Leary of the Colorado Department of Health gave a presentation on 1) a
study of alternative control strategies on urban visibility in Denver
through the year 2000 and 2) the six Rocky Mountain State visibility strategy
study.
-------
The urban visibility analysis started with the current relative frequency
I
I
of four wintertime visual range (visibility) categories as represented by "
photographs and projected the change in frequency of these categories
through the year 2000 under alternative emissions scenarios. According to
the analysis, the major sources of wintertime urban haze in Denver are
mobile sources (gasoline and diesels), refineries and industry, power |
plants, woodstoves, and space heating. With no further controls on any of
these sources, the frequency of poorer than average visibility categories _
in Denver will shift from 50% to 80% of winter days where natural conditions . I
are a minor factor. Under this scenario, the total source contribution in *
2000 would be of 139% of current levels. If Federal regulations were to be
applied to heavy and light duty diesels, this figure would drop to 114% of
current, and with other strategies already planned by the State, the total |
would be 106% of current. Other approaches are being examined that might
result in improvements by 2000;
During the discussion of this analysis, John Leary and the other
Colorado State representatives appeared to agree that the major ways in
which the Federal government might assist the State with respect to urban
haze were 1) more stringent regulations of diesel particulate emissions and
2) additional research and/or resources to examine the causes, benefits,
and control of urban haze.
Leary then outlined the western states visibility study. Task force
members have already received the first summary document prepared for this _
study. The purpose is to develop a uniform approach to visibility standards
and to identify the key technical and policy issues. The project consists *
of three phases, two of which are completed:
I. Issues and Options;
II. Select Preferred Option; and «
III. ' Address specific issues with standard format chosen.
Colorado may use the results of this work in developing their visibility I
SIP for class I-area protection. Leary indicated that the State was looking
for visibility standards that were readily understandable and permitted
monitoring at a reasonable cost. The States are considering doing studies
that would, through photographs or other means, elicit public opinion on |
desirable or acceptable visibility levels.
In the ensuing discussion Ivar Tombach (AeroEnvironment) suggested I
that the states' goals of simplicity and understandability in visibility *
standards might be difficult to reconcile with a consistent measurement
technique. Tom Dodson (US Navy) emphasized that studies eliciting opinion
on desirable or acceptable visibility should also present information to I
the public on the costs of achieving such visibility levels.
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Anne Vlckery
4
Anne Vickery made a presentation on behalf of the Colorado Mountain
Club, a long standing organization now numbering some 8000 members. The
presentation consisted of a series of slides showing a number of Impressive
vistas seen from various mountains. The views included long paths that
extended from class I area to class II area, class II to class II and class
II to class I. The Mountain Club strongly feels that such vistas between
class I areas deserve the same protection as class I areas themselves; they
therefore support the Idea of "integral vistas,"-but might go beyond current
interpretations. The Club maintains ledgers at each of the 54 high peaks
(> 14000 ft), in Colorado that record on the order of 500-700 signatures
per year each. Ms. Vlckery pointed out that tourism accounts for,some $4
billion to the Colorado economy and that good visibility is an important
factor that attracts visitors. She provided a report prepared by the
Mountain Club on the importance of natural resources to the Colorado economy
(attached). She also submitted several reports done for the Forest Service
of visitors to wilderness areas that indicate Clean Air and good visibility
are among the highest ranked attributions of such areas.
In response to a question regarding the Mountain Club observations on
current visibility and perceived future threats to the areas of concern,
Ms. Vickery indicated that degredation of visibility has been noted in some
of the southern areas (Durango and Winimucha) and on the "Front Range." In
particular, Pike's Peak used to be visible from Long's Peak on a regular
basis; this is. rare view at the present time. She feels that future oil
shale development could present serious problems, but that if care were
taken in control and location, such development could proceed. In addition1,
she indicated that the southwestern and Mexican smelter control could be
important to Colorado.
OPEN DISCUSSION
Following a lunch break, the format shifted to an open discussion by
those present. To provide some structure, the time was divided among three
broad topics mentioned in the Federal Register Notice:
1. Desirable goals for regional visibility programs
2. Alternative strategies, and
3. Research needs
The discussion was focused on the western U.S. A substantial number of
attendees participated. The following summarizes some of the major points
made in each area.
Goals for Western Visibility
Bob Yunke (Environmental Defense Fund) opened the discussion in this
area. He pointed out that an apparent "schizophrenia" existed in EPA's
visibility protection programs. Many of the"PSD permitting decisions being
-------
impairment and resolve any remaining technical issues, relating to
indicators of visibility and visibility measurement.
I
I
made by Federal Land Managers that involve an analysis of visibility effects
now appear to include regional haze, even though the regulations themselves
were intended to focus on plume "reasonably attributable" impairment. An |
example was a decision concerning the Teddy Roosevelt class I area. Thus
the Federal Land Managers must make decisions that involve use of regional _
models or other approaches for assessing haze without official EPA guidance. I
In essence, a "back door" approach to addressing haze is developing. Yunke *
pointed out that this problem will emerge again in the upcoming resolution
of the two major remaining components in the EOF;settlement on Phase I
visibility regulations: 1} developing guidance for the BART assessment I
of sources - Yunke feels any regulation should deal more than just coherent
plumes and 2) the regional haze versus plume blight issue will most seriously
come into conflict in the context of developing long-range plans. He feels
that no real distinction has been made between plumes and haze in the
regulatory preamble and that the long-range plan must include haze. If no
federal guidelines exist on how to judge state implementation plans, it I
will be difficult for states to develop adequate procedures and regulations. "
He feels that the long-range .plans must address both area and point source
impacts, not only "SOj" reasonably attributable" impairment but also for
haze. . |
Given the above points, Yunke feels that two areas are paramount in _
developing goals for federal visibility-protection programs. I
(!) EPA should act quickly so that no conflict exists between long-
range plans developed for plume blight and regional haze, and I
(2) The Agency should finally specify the criteria for measuring
I
Ann Vickery felt that protection of class I areas alone from regional _
haze or plumes was not enough. .Many of the most popular areas are only
recently designated as parks or wilderness areas and are not included
under class I protection. In her opinion, these areas should be protected.
Many of them have more visitors than class I areas and the goal, at least
in Colorado, ought to be protecting existing visibility everywhere in the |
state. Several participants then discussed whether or not addressing
regional haze in class I areas in the West would also provide protection
for the spaces in between. The general consensus was that, with some possible
exceptions, protecting visibility in class I areas probably would go a long
way toward protecting visibility in the intervening spaces, because of the
density of Class I areas in the West and because regional haze tends to I
occur on large geographical scales.
Jim Blankenship (Forest Service) raised a Forest Services management
dilemna in this context. A number of wilderness areas have .been enlarged |
since the 1977 Act Amendments. These enlargements are class II areas,
while the rest of the area is designated as class I. Blankenship feels _
these class II portions should receive the same protection as the rest of
the area. The Forest Service apparently Intends to treat both in the same
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
way. Dennis Hadow (USFS) added that some new.wilderness areas are located
at distances over a 100 miles from class I areas and thus may not get
adequate protection from some sources of impairment without special treatment.
It was also pointed out that some states have agreed to allow impact reviews
in all cases, for example, Colorado and Wyoming. One commenter noted that
the need to protect areas between class I locations would seem to warrant a
cooperative effort between States and land managers rather than new Federal
regulations.
On the subject of goals, Rob Farber (Southern California Edison), said
his reading of the national goal established by Congress is to try to get
rid of all man-made pollution. In his words, this is a lofty goal - but
unattainable. He suggested a phased approach that would define the areas
in which we would want to maintain current visibility and areas in which we
would want to improve or allow some degregation. He indicated that we
should establish priorities with some areas e.g. the Grand Canyon, receiving
higher priority and some receiving Tower priority {e.g., some near Los
Angeles). The program could start with these kinds of goals, and develop a
track record before moving further. -Jim Blankenship indicated we should
not ignore the mandate and write off class I areas near Los Angeles because
detecting visibility would be difficult there. Steve Connolly (Jellinek,
Schwartz/1 Connolly, and Freshman ) said that in places like Los Angeles, we
are probably doing all we can do to meet health effects goals and that will
be as much as can be done to make progress towards improving visibility.
Nevertheless, as pointed out by Al Qb'yawayma (Salt River Project) and 7
others, some pollutants emitted in Los Angeles, e.g., $63 may affect visibilit
throughout broad regions and all health criteria have been met. Therefore
there may be some instances in which consideration could be given to urban
.controls beyond those required to meet health related standards.
Bob Yunke then addressed the question of whether or not the task force
report should make recommendations for changing the current legislation
regarding visibility protection. In his opinion, the point of reference for
such legislation is protection of visibility in parks and wilderness areas
where the Federal government has a legitimate interest. The goal of preservat
of visibility in important vistas in parks and other such areas gives
political validity to the program. To raise the prospect of the Federal
government involving itself in protection of visibility beyond these areas
to, for example urban areas or other spaces in between, would detract from
the validity of the program in the western U.S. Yunke would like to see
Congress consider applying class I visibility protection to new wilderness
areas but would not like to see the national goal otherwise expanded or cut
back.
In Yunke's opinion, the critical issue for the task force to address
is What is visibility impairment? The law leaves the determination of what
is adverse to the Federal Land Manager, but the, problem is that EPA has not
issued criteria for making the decision. He feels EPA shoul'd define quantifia
measures and guidelines for determining adverse impairment. One'approach
would be to set standards for each class I area. As Steve Connolly pointed
out, however, the legislative history shows that Congress wanted the decisions
on impairment to be subjective, not a hard test a source could pass or flunk
in an arbitrary way.
-------
I
I
Aaron Mann {Sierra Pacific Power) suggested that one focus for the
task force would be to provide the tools for decision making. These include
1) a definition of the units for measuring visibility and impairment in a |
"reference method" for visibility and 2) an assessment of the sources that
contribute to visibility impairment. _
Rob Farber asked whether one can control the growth patterns in the *
western U.S. under the current authorities, because in his opinion the
growth of urban sprawl is a major current and future problem with respect
to regional haze in the West. In particular, the growth of population and I
associated emissions in northern Arizonia near the "golden circle" areas
should be examined and programs should strike a reasonable balance between
point sources and urban area growth. Others suggested that secondary
impacts (area and other emissions sources that accompany major point sources)
should be addressed in the BACT reviews under the PSD program. Tom Dodson
indicated it would be difficult to prepare an air quality management plan
for each class or area. Anne Vickery appeared to agree and suggested that
she would hate to see each area have to develop individual plans. In her
view, it would be better to do it on an "as needed" basis with the threshold
question being are you where you want to be in respect to visibility in |
comparison to where you are now.
Alternative Control Strategies I
At this point in the discussion, the focus shifted to the area of
control strategy alternatives for the West and asked for comments both on
technical control strategies as well as regulatory alternatives. Rob
Farber asked whether the task force contractor report on emissions projections
in the West included emissions from off-road vehicles, which can cause a
semi-permanent fugitive dust problem, and emissions from natural vegetation. J
Farber indicated that coarse particles may account for about 20% of the
extinction budget in the southwest. George Roope (Utah International Inc.) _
suggested that a regional or detailed management plan could be helpful to I
industries deciding on appropriate new source locations. On the other '
hand, Steve Connolly indicated that if each class I area had a predetermined
target, it might be difficult to accommodate conflicting goals, useful
tradeoffs, and appeals on specific cases. Scott Archer (Bureau of Land |
Management) indicated it could be quite easy to address such issues in a
resource management plan. On the other hand, Jim Blankenship said that the
land managers could do all sorts of planning yet have little control in I
specific instances where there may be an adverse impact.
Ivar Tombach stated that industry needs to have a more stable regulatory
future and that requires well thought out plans that don't have to be
changed constantly. This would argue against the case-by-case approach or
at least in favor of well thought-out criteria and guidelines if a case-by-cas
approach is to be implemented. Paul Roberts (Chevron) felt that EPA should p
provide better guidance and criteria on ''adversity" and supported the
notion that predictability is very important. _
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Research Needs
At this point, the discussion moved to the area of research needs and
recommendations to be considered by the task force. Rob Farber opened the
discussion on the subject.of the relative role of EPA, the National Park
Service, Department of Defense, and industry with respect to visibility
research coordination. Farber suggested that even though EPA is not
the major source of visibility research, that it nevertheless should serve
a role in coordinating visibility research by the various groups in the
country. Most helpful would be some sort of coordinating group that
provided an overall guide with respect to who .is doing what now in the
way of visibility research and suggested time lines for various answers
that would be relevant to regulatory programs. This would serve to provide
better coordination among groups conducting visibility research and give a
better idea of where .various groups stand at this point.
Ivar Tombach felt that the first research need is to define visibility
in a way that is regulatable.. This involves (!) how wel.l we can see real
world things, and (2) the clarity of the atmosphere. He again stated the
need to fix an index against which standards can be set. Nels Larson (Salt
River Project) felt that all visibility monitoring equipment should be made
rugged enough to be used in remote areas.
Ron Farber stressed the issue of particle monitoring in light of the new
information that a substantial component of visibility impairment may be due t
volatile aerosols not collected by traditional monitors. He recommended that
care be used in analyzing any of the existing data bases as proposed in the-
interim research needs document. He also raised the issue of the role of
meteorology in normalizing visibility. In his evaluation of the data, a
7-year rolling average is needed to factor out meteorological variability.
Al Qoyawayma added that we need much better upper air and local meteorology
data in the West. The current radiosonde data are not nearly enough.
Discussion ensued with respect to whether better upper air data could be
gathered in a cost effective manner. Tom Dodson suggested that with enough
planning and cooperation among agencies improved meteorological could be
obtained using, data for example mini-sondes that measure 5 kilometer
winds. Scott Archer added that the BLM already is operating a 75 mile grid
of wind data in the west. Dodson also stressed that research into the
question of volatile aerosols was a very high priority.
Rob Farber stated that a good consensus existed among visibility
researchers on what should be done on a physical basis but felt that we were
weak on what we should do with the physical results. He felt that the work
on perception and values conducted by the Park Service and others should
be continued, but that many different ways to approach these issues
existed and could be used to link to the physical results later. Although
EPA has not conducted research on perception, it does maintain a program on
visibility values, at least in the economic sense, in OPPE. " Bill Malm felt
that part of EPA's overall coordination role would be to make a statement
of support for research on visibility perception and values that would
encourage the Park Service as well as other potential funding agencies to
continue work in this area.
-------
8
WASHINGTON MEETING - DECEMBER 10, 1984
The Washington meeting, held at EPA from 9:30 a.m. to 1:00 p.m., was
also relatively well attended by various groups and task force members but
many groups, particularly the industry representatives, seemed less inclined
to participate in the public discussion. Ironically, some participants
were unable to attend or were late due to a ground fog in the East that
curtailed or prevented aircraft operations.
I opened the meeting in the same way as in Denver, providing an overview
of the task force activities. Then, Doug Latimer of SAI made a presentation
on the modelling work and benefits analysis conducted under contract to OPPE
in support of the task force. Because the conduct and results of that work
are presented elsewhere, I will not summarize his presentation here.
Following the opening presentation, we proceeded to an open discussion
under subject headings similar to those used in the Denver meeting. In this
case, however, the discussion centered on the goals, strategies, and research
needs relating to visibility protection and related issues in the eastern U.S.
Goals
The discussion opened with David Hawkins (National Resources Denfense
Council). He stated that visibility goals can bft defined in the abstract
but that the next step, that is developing policy options, is a much more
difficult one. In this area, we should look at a wide range of practical
implications. He supports the idea of the policy analysis done to date and
feels it is important that EPA continue to have two focuses;
(1) a problem focus, for example, the problem of visibility impairment,
acid rain, and so on,
(2) a pollutant focus. This perspective is derived from the fact
that specific pollutants may be involved in multiple problems, for example
S02, which shows up frequently. A ranking of policy options for a single
pollutant should consider options that involve other pollutants and a need
exists to look at all the benefits of controlling a single pollutant.
Hawkins feels that the Agency is in pretty good shape on both focuses
with respect to sulfur oxides. Susan Buffone {National Parks and Conservatior
Association) raised a western related issue. She asked whether the task
force report would address international concerns, in particular, whether
the issue of the Mexican smelters would be addressed. She also noted that
two issues related to visibility were raised in the parti oil ate matter
proposal, (1) the question of deferring a secondary fine particle standard,
and (2) whether costs should be considered in setting secondary standards.
Betsy Agel (Clean Air Coalition) felt that the problems of visibility
go beyond class I areas to class II areas and places such as state parks.
She indicated that cost benefit analysis were not done to set these areas
aside and therefore questioned whether we could now set dollar values on
protecting visibility in such places.
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
-------
I (
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Dave Hawkins said that looking at the history of public agitation over
air pollution, people's concerns peak most greatly when they can see or
smell it. Therefore, in his opinion, in the goal for visibility protection
even in the east, should be the elimination of any man-made degregation
of visibility. He then provided a number of examples of the importance of
clear skies. He felt that even though visibility appears to be a soft
issue compared to health effects and materials damage, there are many clues
in the literature and everyday experience suggesting that aesthetic or soft
issues are quite important and could affect things that make a difference
in the GNP such as worker productivity.
Hawkins feels that the goal of eliminating man-made impairment should be
fairly aggressive and that an adverse effect on visibility is that which
degrades the clarity of skies and vistas. He feels that approach is implicit
in the Act now with respect to the goal and that the goal should not be
watered down with practical considerations of cost. Evaluating cost-effective
approaches and considering practical realities can and should be factored in
to efforts to pursue the goal..
Alternative Control Strategies
John Trijonis (Santa Fe Research) raised what he termed a philosophical
point with respect to developing control strategies. That is, air pollution
regulations are better the closer they are to the source of the problem.
He feels that visibility protection programs should regulate the source - -
not the effect - using as an example that regulations of automobile emissions
have been more effective than the air quality standards in reducing pollution
levels. Trijonis feels that it is harder to take the emissions approach
under current regulatory authorities. Bill Malm cautioned that if we only
look at the suspected sources, we may miss the real cause of the effect and
may have to switch programs at that point. Malm feels that if the standard
were set on the effect it would be easier to mid-course correct if new
information suggested that a major portion of the problem was not being
addressed. He feels that we should look at the effect. Trijonis agreed
that we should continue to look at the effect, but should base regulations
on emissions. Dave Hawkins stated that really both approaches were needed
and they complement each other's strengths and weaknesses.
On the question of whether to use existing or request new authorities
for dealing with regional haze in the East, Hawkins stated that the Clean
Air Coalition would prefer legislative approaches because the current
regulatory approaches are quite slow, and may well not work without Congress'
support in any case. Because a legislative recommendation might not be
accepted, he feels that the task force should include a regulatory approach
alternative or alternatives as well. Hawkins said that the reason legislation
is necessary is that the Clean Air Act in its current form is not working
with respect to visibility in the East and, therefore it should be fixed.
Research Needs
/
John Trijonis opened indicating that in his review of the interim researc
needs, he would place a low priority on the mini-recession study. He felt
that the change in emissions during the recession were probably too small
-------
I
to see any valid change in visibility. William Wilson responded that the
intent was to use models to filter the data and to use the information to
evaluate the models. Vicky Evans (Gaia) stated for the record that her
organization had been conducting a visibility research forum and had intended
to expand this effort to environmental groups. She suggested that the
group could use more EPA support. Dave Hawkins suggested that the noise |
program, despite little in the way of research dollars, was able to provide
something of a clearinghouse for noise research and coordinating noise _
research in the country in the past. He suggested that this might be a
useful model for EPA to use in coordinating visibility research.
There were some questions regarding the amounts of funding EPA
received for FY84. |
Molly Ross of the National Park Service stated that the Parks Services'
view was that the mandate for protection of visibility and air quality relatec
values extends to all of their areas, not just class I areas. A number of
smaller parks are designated .as class II. .
Dave Hawkins then made some recommendations that EPA consider research I
in psychological reactions to clean vs. dirty air, citing examples of the
effects of light on mood changes and studies of visitation rates to national
parks versus haziness. |
ADDRESSEES: . . -
Visibility Task Force Members
John O'Connor
Harvey Nozick
Barbara Wauchope
Barbara Bankoff I
I
I
I
I
I
I
I
I
-------
1
1
1
1
1
1
1
1
i
1
1
1
"
1
-
'
1
1
1
Public Mi
Task Force Member
Scott F. Archer
John Bachmann
Jim Blankenship
Jim Dicke
David Joseph
Bill Malm
Janet Metsa
Tom Pace
Marc Pitchford
Bruce Polkowsky
Butch Rachel
Sara Schneeberg
Vivian Thomson
Other Attendees
Paul Adams
Steven Arnold
Prem Bhardwaja
C.A. Bina
Bill Blankenship
John M. Clause
Steve Connolly
James B. Coyne
Attendees
Public Meeting of the Interagency Visibility Task Force
December 5, 1984
Denver, Colorado
US BLM/Colorado State Office
U.S. EPA/OAQPS
U.S. Forest Service
U.S. EPA/OAQPS
NPS
US NPS/Washington Office
U.S. EPA/OAQPS
U.S. EPA/OAQPS
U.S. EPA
U.S. EPA/OAQPS
EPA Region 8
U.S. EPA/06C
U.S. EPA, Washington
A. Coors Co.
Colorado Dept. of Health
Salt River Project
Basin Electric Power Corp.
New Mexico Environmental
Enforcement Division
Colo. Air Pollution Control Div.
Jellinek, Schwartz, Connolly &
Freshman, Washington, D.C.
Peabody Coal Co.
-------
John T. Dale
Tom Dodson
Victoria A. Evans
Rob Farber
Marcia Gelman
Rick Hardy
Dennis Kaddow
John Leary
Aaron Mann
C.V. Mathai
John McNeil 1
Brian Mitchell
Ron Ostop
Robert Pearson
Michael A. Poling
t»
Al Qoyawayma
Richard Raw!ings
Paul Roberts
George W. Roope
Joe Southerland
Mike league
Louis Thanukos
Ivar Tombach
Anne Vickery
Bob Yuhnke
EPA Region VIII
U.S. Navy
Gaia Associates
So. Calif. Edison
Kunton i Williams
Morrison Knudsen Co., Boise,"Id.
U.S. Forest Service
Colorado Dept. of Health
Sierra Pacific Power
Arizona Public Service Co., Pnoe
Colorado-UTE Elec. Assn.
NPS-Air Quality Denver
Colorado Springs Dept. of Utilit
Public Service Co. of Colorado
American Mining Congress
Salt River Project
New Mexico Environmental
Enforcement Division
Chevron Research
Utah International Inc.
Salt River Project
Hunton & Williams
Appl. Env. Cons.
AeroVironment,. Inc.
Colorado Mountain Club
Env. Defense Fund, Boulder, CO
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
Visibility Task Force Meeting
Washington, O.C.
December 10, 1984
1
1
1
1
I
-
1
1
1
1
Name
Task Force Members
John Bachmann
Jim Byrne
Jim Dicke
David Joseph
Bill Malm
Tom Pace
Dorothy Patton
Bruce Polkowsky
Sara Schneeberg
William Wilson
Sidney Worthington
Other Attendees
Betsy Agle
Richard Boubel
David Branand
Rob Brenner
Susan Buffone
Brendan Doyle
Victoria Evans
Robert B. Flagg
Marcia Gel man
Barbara Goldsmith
Judith Greenwald
Organization/Address
EPA, RTP, NC 27711
USDA-FS .
EPA, RTP, NC 27711
NPS, Denver, CO
NPS, Ft. Collins, CO
EPA, -RTP, NC 27711
EPA, OGC, Washington, DC
EPA, RTP, NC 27711
EPA, OGC, Washington, DC
EPA, RTP, NC 27711
EPA, OPPE, Washington, DC
Clean Air Coalition
DOD-Pentagon
NCA
EPA, OAR
NPCA
RM Dowd & Co.
Gaia Associates
Sausalito, CA
Mining & Reclamation Council
1575 Eye St. NW, Washington, DC
Hunton & Williams
ERT, Inc.
EPA, OPPE, Washington, DC
Phone
919-541-5531
703-235-8096
919-541-5681
919-541-5522
202-382-7625
919-541-554C
202-382-7635
919-541-2551
202-382-549C
202-653-1272
202-463-2636
202-382-7432
202-265-2717
202-737-5069
415-331-5932
202-789-0220
202-955-1540
617-369-8910
202-382-5490
-------
Name
Marcel L. Halberstadt
David Hawkins
Deborah K. Jessup
Don Leonard
Conrad Mackerron
Kim Mihelhaftz
David Parker
Debby Peck
Joseph W. Phillips
Tobey Prira
Dave Reeves
Molly, N. Ross
Lisa Shapiro
Mike league
Ivar Tombach
John Trijonis
Anthony Walters
John W. Wilson
Organization/ Address
Motor Vehicle Manufacturers Assoc.
300 New Center Bldg. Detroit, MI 48202
NRDC, 1350 NY Ave., Washington, DC
BNA
Western Regional Council
BNA
EPA, OFA
EPA, 'OFA
NPS, Washington, DC
TVA/.449 MPB-M, Muscle Shoals, MA
EPA, OAR, Washington, DC
Jell i nek, Schwartz, Connolly &
Freshman, Washington, DC
^NPS, Washington, DC
NPS, Washington, OC
H & W
AeroVironment Inc., Monovia, CA
Santa Research Corp.
. American Petroleum Institute
Stone & Webster Engineering Corp.
Boston, MA
Phone
313-872-4311
202-783-780(
202-452-441*
801-363-7997
202-452-440:
202-475-8797
202-475-8797
202-343-491J
617-386-203:
202-382-013(
202-783-338?
202-343-4911
202-343-491]
202-955-1527
818-357-9983
612-944-0602
617-539-289£
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
I
-------
APR | 6
------- |