-------

-------
      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 cloud1—I.  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 Report—Visibility 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 aerosol•concentrations
    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.-oral•communication  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
    
    4•1  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 industry—Chevron, 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 internalized—electricity
    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 air—support 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) expressed—in
    
    detail—its 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 internalized—electricity 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
    

    -------