Out of Sight: The Science and Economics of Visibility Impairment

Abt Associates Inc.
# 4800 Montgomery Lane #

Bethesda, MD 20814-5341 # www.abtassoc.com
Out of Sight:
The Science and
Economics of
Visibility Impairment
Out of Sight:
The Science and Economics of
Visibility Impairment
August 2000
Prepared for
Clean Air Task Force
Boston, MA
Project Manager:
Dr. L. Bruce Hill
Prepared by
Abt Associates Inc.
4800 Montgomery Lane
Bethesda, MD 20814-5341
www.abtassoc.com
Abt Associates’ Environmental Research Area provides multi-disciplinary scientific research and
environmental policy analysis to the EPA, the U.S. Agency for International Development, the Inter-American
Development Bank, the World Bank, and directly to foreign, state and local governments. Abt Associates has
extensive experience in estimating the potential public health improvements and economic costs and benefits
from improving ambient air quality. The Environmental Research Area conducted extensive health analysis
for the U.S. EPA in support of the 1997 revisions to both the ozone and the particulate matter National
Ambient Air Quality Standards. They also prepared the health and economic analyses for EPA’s 1997 Report
to Congress The Benefits and Costs of the Clean Air Act: 1970 to 1990, and conducted similar policy, health
and economic analyses for EPA of regulations on the electric generating industry, automobile exhaust, and
potential policies for climate change mitigation strategies. Abt Associate’s Environmental Research Area
conducts public health analysis projects worldwide, including air pollution health assessment projects with the
environmental and health ministries in Argentina, Brazil, Canada, Chile, Korea, Russia, Thailand, the Ukraine
and the World Health Organization.
Mr. Kenneth Davidson specializes in the analysis of air quality policy. He has a master's degree in resource
economics and policy from Duke University's Nicholas School of the Environment, and as a student worked
with the Innovative Strategies and Economics Group at the U.S. EPA's Office of Air Quality Planning and
Standards.
Dr. Leland Deck specializes in economic and risk analysis of environmental policies. His research projects
include estimating the risks and economic value of health and welfare benefits from reducing air pollution, the
costs of alternative pollution prevention technologies, and designing effective and enforceable economic
incentive programs as a part of an overall strategy for controlling pollution from stationary and mobile sources.
In addition to his own research projects, Dr. Deck manages Abt Associates’ Environmental Economics
Practice, and is a Vice President of Abt Associates.
Dr. Don McCubbin has eleven years of experience in the analysis of environmental issues, with a special
emphasis on the adverse effects of criteria air pollutants.
Dr. Ellen Post has fourteen years of experience in the scientific, economic, and policy analysis of
environmental issues, with particular emphasis on (1) criteria air pollution risk assessment and economic
benefit analysis, and (2) methods of assessing uncertainty surrounding individual estimates. She is one of the
primary analysts conducting a particulate matter air pollution risk assessment for EPA’s Office of Air Quality
Planning and Standards, and has been a key economist in ongoing work analyzing the economic benefits
associated with risk reductions from a number of air quality regulations, including the implementation of
proposed particulate matter and ozone standards in the United States.
Table of Contents
EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-1
1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. VISIBILITY IMPAIRMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.1 Causes of Visibility Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2 Observing and Measuring Visibility Impairment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 People Can Tell When Views Are Impaired . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.2 How Visibility Impairment is Measured . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3. TRENDS IN VISIBILITY IMPAIRMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.1 Recent Visibility Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.1.1 Visibility at Parks Where the Vista Is Integral to the Experience . . . . . . . . . . . 14
3.1.2 Visibility at Smaller Parks and Wilderness Areas . . . . . . . . . . . . . . . . . . . . . . 22
3.1.3 Visibility in Urban Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4. LEGISLATIVE AND REGULATORY HISTORY OF VISIBILITY IMPAIRMENT . . . . . . . 26
5. ECONOMICS: VISIBILITY IMPAIRMENT AND PARK VISITATION . . . . . . . . . . . . . . . 31

5.1 An Undisturbed Environment, Including Clean, Clear Air and Good Visibility, is Very
Important to Park Visitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.2 Visitors are Willing to Alter Their Length of Stay Based on Visual Air Quality Conditions at
National Parks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3 What is at Stake to the Economy if Visitation Rates Decline? . . . . . . . . . . . . . . . . . . . 34
5.4 Local Economic Benefits from Visibility Improvements . . . . . . . . . . . . . . . . . . . . . . . . 36
6. ECONOMICS: THE NON-MARKET VALUE OF VISIBILITY . . . . . . . . . . . . . . . . . . . . . . 41

6.1 The Economic Valuation of Visibility Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.2 The Value of Visibility Improvements at National Parks: Evidence from Studies . . . . . 43
6.3 The Value of Visibility Improvements in Residential Areas: Evidence from Studies . . . 46
6.4 Applying the Information from Studies to Assess the Visibility Benefits of Reducing Air
Pollution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
7. POWER PLANT EMISSION REDUCTIONS AND ASSOCIATED VISIBILITY BENEFITS

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
8. CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
9. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
APPENDIX A DETAILED TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1
APPENDIX B METHOD TO ESTIMATE VISIBILITY BENEFITS . . . . . . . . . . . . . . . . . . . . . . . . B-1

B.1 Basic Utility Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-2
B.2 Measure of Visibility: Environmental “Goods” Versus “Bads” . . . . . . . . . . . . . . . . . . B-3
B.3 Estimating the Parameters for Visibility at Class I Areas: the (’s and *’s . . . . . . . . . . B-5
B.3.1 Estimating Region-Specific Recreational Visibility Parameters for the Region
Covered in the Chestnut and Rowe Study (Regions 1, 2, and 3) . . . . . . . . . . B-8
B.3.2 Inferring Region-Specific Recreational Visibility Parameters for Regions Not
Covered in the Chestnut and Rowe Study (Regions 4, 5, and 6) . . . . . . . . . . B-8
B.3.3 Estimating Park- and Wilderness Area-Specific Parameters . . . . . . . . . . . . . B-10
B.3.4 Derivation of Region-specific WTP for National Parks and Wilderness Areas
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-10
B.3.5 Derivation of park- and wilderness area-specific WTPs, given region-specific WTPs
for national parks and wilderness areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-11
B.3.6 Derivation of park- and wilderness area-specific parameters, given park- and
wilderness area-specific WTPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-12
B.4 Estimating the Parameter for Visibility in Residential Areas: 2 . . . . . . . . . . . . . . . . B-13
B.5 Putting it All Together: the Household Utility and WTP Functions . . . . . . . . . . . . . B-13
List of Exhibits
Exhibit 1-1 Denver on a Clear Day and on a Hazy Day . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Exhibit 2-1a Visible Plume from Local Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Exhibit 2-1b Layered Haze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Exhibit 2-1c Regional Haze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Exhibit 2-2 National Emissions of Nitrogen Oxides and Sulfur Dioxide by Source in 1998 . . . . . . . . . . . 4
Exhibit 2-3a Contribution to Visibility Impairment in the Eastern U.S. . . . . . . . . . . . . . . . . . . . . . . . . . 5
Exhibit 2-3b Contribution to Visibility Impairment in the Mid-West . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Exhibit 2-3c Contribution to Visibility Impairment in the West . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Exhibit 2-4 Scattering and Absorption of Image-Forming Light from the Observer’s Sight Path . . . . . . . 7
Exhibit 2-5 Aerosol Size and Light Scattering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Exhibit 2-6 Types of Particulate Matter and Impact on Image-Forming Light . . . . . . . . . . . . . . . . . . . . 8
Exhibit 2-7 Relationship Between Perceived Visual Air Quality and the Amount of Particulate Matter in the
Air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Exhibit 2-8 Inverse Relationship Between Visual Range and the Deciview Index . . . . . . . . . . . . . . . . . . 11
Exhibit 3-1 Airport Visual Data: Trend in 75th Percentile Light Extinction Coefficient for July-September
(measured in km-1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Exhibit 3-2 Visibility Trends at Acadia National Park, Maine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Exhibit 3-3 Visibility Trends at Grand Canyon National Park, Arizona . . . . . . . . . . . . . . . . . . . . . . . . 17
Exhibit 3-4 Visibility Trends at Great Smoky Mountains National Park, Tennessee . . . . . . . . . . . . . . . 18
Exhibit 3-5 Visibility Trends at Shenandoah National Park, Virginia . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Exhibit 3-6 Visibility Trends at Yosemite National Park, California . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Exhibit 3-7a Extreme Visibility Days at Acadia National Park . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Exhibit 3-7b Extreme Visibility Days at Grand Canyon National Park . . . . . . . . . . . . . . . . . . . . . . . . . 21
Exhibit 3-7c Extreme Visibility Days at Great Smoky Mountains National Park . . . . . . . . . . . . . . . . . 21
Exhibit 3-7d Extreme Visibility Days at Shenandoah National Park . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Exhibit 3-7e Extreme Visibility Days at Yosemite National Park . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Exhibit 3-8 Visibility Trends at San Gorgonio Wilderness Area, California . . . . . . . . . . . . . . . . . . . . . 23
Exhibit 3-9 Visibility Trends at Chassahowitzka Wilderness Area, Florida . . . . . . . . . . . . . . . . . . . . . 23
Exhibit 3-10 Visibility Trends in Washington, D.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Exhibit 4-1 Map of Mandatory Class I Areas with Visibility Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Exhibit 5-1 Visitor Rated Importance of Park Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Exhibit 5-2 Total Recreation Visits to U.S. National Parks, 1951-1998 . . . . . . . . . . . . . . . . . . . . . . . . 35
Exhibit 5-3 Sales Increase Near Park Due to an Increase in Park Visitation . . . . . . . . . . . . . . . . . . . . . 39
Exhibit 5-4 Tax Revenue Increase Near Park Due to an Increase in Park Visitation . . . . . . . . . . . . . . . 39
Exhibit 5-5 Local Job Increase Near Park Due to an Increase in Park Visitation . . . . . . . . . . . . . . . . . . 40
Exhibit 6-1 Economic Valuation Studies for Recreational and Residential Visibility . . . . . . . . . . . . . . . 44
Exhibit 6-2 Visibility Benefits from Different Policy Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Exhibit 7-1 Residential (Urban) Visibility Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Exhibit 7-2 Recreational Visibility Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Exhibit 7-3 State-Level Recreational Visibility Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Exhibit 7-4 WinHaze Split-Images at Great Smoky Mountains National Park for Status Quo, No-EGU, and
Partial-EGU Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Exhibit A-1 Percentage Contribution of Constituents to Visibility Impairment on Good, Mid-Range, and Poor
Visibility Days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1
Exhibit A-2 Annual Slope Estimates with Probabilities for Rejection for the Average of the Worst, Median,
and Best Visibility Days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-3
Exhibit A-3 Visibility Benefits from Different Policy Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-4
Exhibit B-1 Available Information on WTP for Visibility Improvements in National Parks . . . . . . . . B-6
Exhibit B-2 Summary of Region-Specific Recreational Visibility Parameters to be Estimated in Household
Utility Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-7
1. INTRODUCTION
Visibility impairment is a basic form of air pollution, one that people can see and recognize without
special instruments. It is also one of the scientifically best-understood air quality-related impacts of fossil fuel
combustion. Despite this common knowledge, the full costs of impaired visibility are not well understood by
policymakers and the public. Some people are not aware that visibility is impaired at all, incorrectly believing
that the milky-white haze that blankets parts of the country is somehow a natural phenomenon associated with
humidity, especially on hot summer days. In fact, visibility impairment is a major problem in the United States,
with both aesthetic and economic consequences.
A variety of sources contribute to the air pollutant

emissions that lead to visibility impairment, including power
plants, motor vehicles, wildfires and industrial processes like
smelting. The largest source in many areas is power plant
emissions. This report is meant to give the educated reader a
basic understanding of the nature and science of visibility
impairment and to provide an overview of the economic costs
of visibility due to sources such as power plants. The report
is not, however, meant to go into great depth in every area,
although there is a large reference section that one can refer to
in order to get more detail on particular topic discussed here.
The report begins with a general overview of the basics behind the nature and science of visibility.
Included in this section is a discussion on what causes visibility impairment, how visibility is impaired, how
humans perceive visibility impairment, and how it is measured. The second section presents both historical
trends on national visibility and examples of visibility degradation at specific places. The legislative history
in specific regard to visibility regulation is then presented, along with a discussion on other air pollution
policies that have had an impact on visual air quality. The economics of visibility follows, and is presented
in two separate sections. The first discusses the economics of visibility in terms of its impact on the direct
consumption of visibility as a resource, or, in other words, how visibility impacts visitation and tourism
behavior. The next section presents the economics of visibility in terms of non-direct consumption, or how
people value improvements in visibility in areas where they may or may not be experiencing it directly. Finally,
an applied example of the valuation of visibility improvements is provided, specifically calculating the visibility
benefits associated with reductions in power plant emissions.
Exhibit 1-1 Denver on a Clear Day and on a Hazy Day
Abt Associates Inc. 1 August 2000

2. VISIBILITY IMPAIRMENT
Visibility impairment comes in a variety of forms: intrusive plumes from local smokestacks, a dirty
low-lying inversion layer, a milky or brown regional haze blanketing the view in all directions. Each of these
forms of visibility impairment is a function of the nature and source of emissions and the prevailing
meteorological conditions (Malm, 1999, p. 20). With stable atmospheric conditions and large, local emission
sources, plumes and layered hazes are likely to occur (Exhibits 2-1a and 2-1b). Regional haze occurs under
meteorological conditions favorable for regional transport (Exhibit 2-1c).
Plumes, layered haze, and regional haze differ from the clouds and fog that we might see on a rainy
day, and instead are manmade impediments to visibility that federal, state, and local governments are actively
trying to reduce. For regulatory purposes, the Environmental Protection Agency distinguishes between
visibility impairment that is caused by one or a small group of sources, such as a the plume from a smoke
stack, and visibility impairment that is caused by emissions over a wide geographic region. The distinction is
made because emissions over a wide region are more diffuse and less easy to attribute to specific sources and,
thus, more difficult to identify and control.
When the view is obscured by pollution, it especially affects people’s enjoyment and sense of
“wilderness” experience. Many visitors to our nation’s parks and wilderness areas are unable to see the
spectacular vistas they had expected, because a veil of white or brown haze hangs in the air blurring the view.
Because this reduction can significantly reduce people’s enjoyment of the views, and it may reduce the
likelihood that they come back to visit, this can have a significant local economic impact. As we discuss in
subsequent chapters, there is also evidence that people value visibility at parks even when they are at home,
whether they visited the area or not.
Exhibit 2-1a Visible Plume from Local Stacks

Source: Malm (1999, Figure 1-5) and NPS-CIRA (2000b).

Exhibit 2-1b Layered Haze
Exhibit 2-1c Regional Haze

2.1 Causes of Visibility Impairment
Most visibility impairment is caused by human-induced particulate air pollution, the same pollution
linked to premature death (Pope et al., 1995) and acid rain (NAPAP, 1990). Because of its very small particle
size, this pollution is often carried by the wind hundreds of miles from where it originated. The large coal-fired
electric utilities in the Ohio valley, some with stacks approaching 1000 feet tall, contribute the largest share
to visibility problems over a wide area of the Eastern and Midwestern U.S. Other contributors include motor
vehicles, industrial fuel burning, manufacturing operations, and natural sources such as windblown dust,
volatile organic compounds from plants, and soot from wildfires.

These emission sources contribute “primary” particles, which are particulates emitted directly into the
air, and “secondary” particles, which form from gases often carried many miles from their source. Both
primary emissions and secondary formation of particles contribute to visibility impairment. Primary particles
come from a variety of sources including diesel and wood combustion, and dust from industrial activities and
natural sources. Secondary particles form in the atmosphere from gases emitted from power plants, cars, and
a number of other sources, and are the most important in forming haze. Sulfate is, in most areas, the most
important secondary particulate, and forms from sulfur dioxide and ammonia emissions. Other secondary
particulates include nitrates, forming from nitrogen oxide and ammonia emissions, and organic carbon particles
from condensed hydrocarbon emissions.
Some particles, like sulfates and nitrates become more effective during humid conditions as they absorb
atmospheric moisture and grow (Day et al., 2000, p. 716; Sisler, 1996, p. 4-7; U.S. EPA, 1996, p. 8-44).
Sulfates and nitrates can more than triple in size as relative humidity increases, thus making visibility worse
during periods of high humidity, such as the humid summer months in the East (National Research Council,
1993, p. 103). However, humidity alone does not cause visibility impairment.
The impact of particles can be measured many miles away. California and Mexico both make
substantial contributions to sulfate particles in the Grand Canyon (Eatough et al., 2000, Figure 9; Malm, 1999,
p. 51). The U.S., east of the Mississippi, and Canada are both affected by emissions from power plants.
Electric utilities are perhaps the single largest contributor to poor visibility. Nationwide in 1998, electric
utilities contributed 67 percent of sulfur dioxide emissions and 25 percent of nitrogen oxide emissions (Exhibit
2-2 based on U.S. EPA, 2000, Tables A-2 and A-4). Coal-powered electric utilities dominate these emissions,
contributing 94 percent of sulfur dioxide and 88 percent of nitrogen oxide emissions from electric utilities.
Exhibit 2-2 National Emissions of Nitrogen Oxides and Sulfur Dioxide by Source in 1998
Nitrogen Oxides Sulfur Dioxide
70%
60%
50%
40%
30%
20%
10%
0%
Electric Utilities Industry Transport Other
Sulfur dioxide gas is especially important because it contributes to the formation of sulfates, which
often dominate other causes of visibility impairment, particularly in the Eastern U.S. Exhibits 2-3a, 2-3b, and
2-3c present the average contribution to visibility impairment of different particulate matter constituents at a
variety of mainly rural monitoring sites through out the U.S., and Exhibit A-1 presents the park-level data
underlying these regional averages. The exhibits present the contribution on a good, medium and bad visibility
days. The sites presented in the exhibits are typically located in national parks, with the exception of
Washington D.C.

In the eastern and mid-western U.S., sulfates account for the majority of visibility impairment. In the
West, sulfates and organic carbon play about equal roles. Throughout the U.S., nitrates typically account for
less than 10 percent of visibility impairment in most locations, with a notable exception of San Gorgonio,
located in southern California, where nitrates can contribute over 30 percent of visibility impairment (Exhibit
A-1). Perhaps most significantly, in almost all locations, sulfates are responsible for a greater percentage of
visibility impairment on bad visibility days than on good visibility days. In Great Smoky Mountains National
Park, sulfates account for 46 percent of impairment on a good day, 63 percent on a median day, and 76 percent
on a bad visibility day. In Acadia, sulfates account for 37 percent, 48 percent, and 69 percent, respectively
(Exhibit A-1). In most of the East, sulfur dioxide emissions, largely from electric utilities, account for two
thirds to three quarters of the visibility impairment on haziest days.
Exhibit 2-3a Contribution to Visibility Impairment in the Eastern U.S.

Source: NPS-CIRA (2000a).
80% Sulfate
Contribution to Visibility
70%
Impairment (%)
60%
Nitrate
50%
40%
Organic Carbon
30%
20%
10% Light Absorbing
Carbon
0%
Good Day Median Day Poor Day
Coarse Matter and
Eastern U.S. Fine Soil
Exhibit 2-3b Contribution to Visibility Impairment in the Mid-West

80% Sulfate
Contribution to Visibility
70%
Impairment (%)
60%
Nitrate
50%
40%
Organic Carbon
30%
20%
10% Light Absorbing
Carbon
0%
Coarse Matter and
Mid-Western U.S. Fine Soil

Exhibit 2-3c Contribution to Visibility Impairment in the West
80% Sulfate
Contribution to Visibility 70%
Impairment (%) 60%
Nitrate
50%
40%
Organic Carbon
30%
20%
10% Light Absorbing
Carbon
0%
Coarse Matter and
Western U.S. Fine Soil
2.2 Observing and Measuring Visibility Impairment
To better understand visibility impairment it is useful to consider the nature of light and how the human
eye functions. The human eye recognizes a spectrum of colors from blues with a wavelength of about 0.4
microns to reds with a wavelength of 0.7 microns (Malm, 1999, p. 3). We perceive a rose to be red because
it absorbs all of the wavelengths in the visible spectrum except for those around 0.7 microns, and we perceive
a blue butterfly because it absorbs most of the visible spectrum except for those wavelengths around 0.4
microns. The range of color we see are simply light with different wavelengths reflected back to us and
captured by our eyes. The same process of differential absorption and reflection of light occurs with visibility
impairment.
When we observe a low-lying brownish layered haze caused by nitrogen dioxide emissions, it appears
brown because nitrogen-dioxide absorbs blue light and reflects back to us the remainder of the spectrum
(Malm, 1999, p.4). When we see a dark plume coming from a smokestack, it appears black because the carbon
soot and other emissions absorb all of the visible spectrum. Conversely, a white plume of water vapor coming
from a power plant’s cooling tower appears white because it absorbs none of the incoming light and simply
scatters and reflects back the full spectrum to the eye.
When we are on a mountain top enjoying a view of the landscape or just walking down the street and
looking at a distant object, a number of processes interfere with the light that is reflected from objects that we
are trying to see (Exhibit 2-4). Our ability to see a distant mountain depends on “transmission radiance” and
“air light” (U.S. EPA, 1996, p. 8-23). Transmission radiance refers to the light reflected from the mountain
and the subsequent interaction of this light in the atmosphere. As this light is absorbed and scattered by gases
and particles in the atmosphere, our ability to see the mountain is reduced. Scattering by particles is usually
the most important source of interference. Another source is air light, which has a variety of effects and refers
to the light from sources other than the object of interest that are scattered towards us and affect what we are
trying to see. Air light scattered from behind the mountain provides backlighting and makes the mountain
standout, while air light scattered from particles and gases between the mountain and our eye obscures our
vision (National Research Council, 1993, p. 82).

Exhibit 2-4 Scattering and Absorption of Image-Forming Light from the Observer’s Sight Path
Transmission radiance usually dominates visibility impairment, and the scattering of photons is the
most important process in transmission radiance. Although gaseous pollutants, such as nitrogen dioxide,
contribute to visibility impairment, they usually play a small role, and instead scattering due to particulates
dominate the visibility impairment. However, not all particles are equally important in scattering light (Exhibit
2-5). Particles about the same size as the visible spectrum are the most efficient at scattering light (Malm,
1999, p.8).

Exhibit 2-5 Aerosol Size and Light Scattering
These fine particles, ultra-efficient at scattering light, include: sulfates, nitrates, organics, and soil
(Malm, 1999, p. 26). Elemental carbon is another fine particle that contributes to visibility impairment by
efficiently absorbing light because of its black color (Exhibit 2-6). In contrast to white sulfur, which absorbs
little light and instead scatters it effectively, elemental carbon absorbs light, much like a blacktop absorbs heat
on a hot summer day. However, the contribution of absorption by elemental carbon is generally less than 10
percent of the loss in transmission radiance. Sulfates often dominate, particularly in the East, and can
contribute 80 percent or more of the loss in transmission radiance (Exhibit 2-3a).
Exhibit 2-6 Types of Particulate Matter and Impact on Image-Forming Light


2.2.1 People Can Tell When Views Are Impaired
It is well known how visibility is impaired by air pollution, and how the human eye perceives these
changes. However, do humans regard visibility impairment in a consistent and equal fashion? That is, when
visibility is impaired, do humans perceive that visibility has changed equally? This is an important question
to answer because, if impaired visibility is not perceived equally between people from region to region, it would
be nearly impossible to measure and regulate.
In the first perception and judgement studies, conducted by the NPS (NPS, 1988), people were asked
to judge the visual air quality in several slides depicting vistas under different visibility conditions using a scale
of one to ten, one being worst and ten being best. The one to ten judgement is called perceived visual air
quality (PVAQ), and it reflects people’s perceptions and judgements concerning the visual air quality depicted
in the slide. Studying the differences among the slides and their average PVAQ has helped researchers
determine what factors are most important to human observers in their judgements of visual air quality.
These studies first addressed the question of whether individuals judged poor visibility in the slides
similarly to actual views under the same air quality conditions. It was found that the PVAQ judgements were
comparable and that the use of slides in studies concerning perceived visual air quality was a valid method for
comparing perceptions of visual air quality. They also found that, regardless of the demographic
characteristics of the individual (age, sex, education), PVAQ judgements made by different people were
consistent with each other. This suggested that for a given slide, people generally judged air quality in a similar
fashion.
Exhibit 2-7 Relationship Between Perceived Visual Air Quality
The analysis of the PVAQ
and the Amount of Particulate Matter in the Air
judgments revealed that increases in
air pollution are more noticeable and
objectionable to the human observer 9
Poor ------------------------------------------------- Good
when the air is relatively clean.

Perceived Visual Air Quality (PVAQ)
Exhibit 2-7, based on NPS (1988, 8

Figure 3-1), demonstrates the
relationship between the PVAQ 7
judgement and ambient particulate
levels. The curved line indicates that 6
a one-unit increase in particulates will
result in a much larger decrease in an
5
individual’s PVAQ at the lower
particulate levels than at the higher
4
particulate levels.
These studies were also able

3
to test whether the visual air quality
perceived by individuals is more 2
sensitive to changes in air pollution at
some vistas than at others. The 1
results indicate that people find 1 2 3 4 5 6 7 8 9 10
increases in air pollution more
3
objectionable in vistas with features Particulate Concentration (ug/m )
that are more highly colored and
textured.

These first few perception and judgement studies set the foundation for subsequent research on the
effects of visual air quality on the visitor experience. These studies suggested that visitors do have preferences
concerning visibility conditions and that a variety of circumstances, such as what is actually being viewed and
how good the air quality was prior to the pollution, influence changes in perceived visual air quality.
2.2.2 How Visibility Impairment is Measured
It is important to determine that people can actually perceive changes in visibility condition, which is
why the perceived visual air quality (PVAQ) index was developed. However, this index is not an actual
measure of visibility conditions as they exist from place to place. To conduct meaningful analyses of how
visibility changes due to the presence of pollution in the atmosphere, there must be some standardized approach
to measuring visual air quality. Visibility conditions are, therefore, commonly expressed in terms of three
mathematically related metrics: standard visual range (SVR), light extinction (ext), and deciviews (dv).
Standard visual range is the metric best known by the general public. It is the maximum distance at which one
can identify a black object against the horizon, and is typically described in kilometers (or miles). Higher visual
range estimates mean better visibility. While the theoretical maximum is 391 kilometers on a perfectly clear
day, this is never achieved due to the natural scattering of light by gases in the atmosphere, so-called “Rayleigh
scattering” (U.S. EPA, 1996, p. 8-12). While standard visual range is a simple measure that can be easily used
to characterize visual conditions, it is somewhat imprecise and cannot be used to effectively determine the
relative importance of the contributors to reduced visibility. It is also useless in cloudy conditions near
monitors.
Light extinction is a somewhat better alternative than visual range because it allows one to express
more objectively the relative contribution of a PM constituent to overall visibility impairment. Light extinction
is the sum of the light scattering and light absorption by particles and gases in the atmosphere, and is measured
in inverse megameters (Mm-1), relating how much light is extinguished per megameter. Higher extinction
values mean worse visibility. This is the inverse of visual range, where higher visual range estimates suggest
better visibility (U.S. EPA, 1996, p. 8-56). For example, in the Great Smoky Mountains, a relatively clear day
has an extinction of 47 Mm-1 and a visual range of 82 kilometers, and a hazy day has an extinction of about
211 Mm-1 and a visual range of 19 kilometers. Both extinction and visual range are similar in that they are not
proportional to human perception (Malm, 1999, p. 35). In other words, a one unit change in either light
extinction or visual range is perceived differently, depending on the starting point. For example, a five mile
change in visual range can be either very apparent or not perceptible, depending on whether the starting point
is a clear day or a hazy one.
A third measure of visibility is the deciview index, which EPA selected as the standard metric for
tracking progress in EPA's regional haze program, largely because it provides a linear scale for perceived visual
changes over a wide range of conditions.1 On a particle-free, pristine day, the deciview index has a value of
zero (SVR=391 km). On a relatively clear day in the Great Smoky Mountains the deciview index might be
about 16 (SVR=79 km) and on a relatively hazy day the deciview index might be about 31 (SVR=201 km).
For each 10 percent increase in light-extinction, the deciview index goes up by one. So, higher deciview values
mean worse visibility (Exhibit 2-8). This logarithmic scaling is analogous to the decibel scale used for the
perception of sound (U.S. EPA, 1996, p. 8-57). Under many scenic conditions, a change of one deciview is
considered to be just perceptible by the average person. However, it is important to understand that the same
amount of pollution can have dramatically different effects on visibility depending on existing conditions. Most
1
The Federal Register (July 1, 1999, vol. 64, no. 126, p. 35,725) discusses the choice of the deciview index for EPA's
regional haze program.

importantly, visibility in cleaner environments is more sensitive to increases in particle concentrations than
visibility in more polluted areas.
Exhibit 2-8 Inverse Relationship Between Visual Range and the Deciview Index
400
350
300
Visual Range (km)
250
Grand Canyon Good Day
200
150
Shenandoah Good Day
100
Shenandoah Median Day
50
Shenandoah Bad Day
0
0 5 10 15 20 25 30 35 40 45
Deciview Index
Visibility impairment is roughly proportional to the product of ambient particle levels and viewing
distance (National Research Council, 1993, Figure 4-3). As particle levels increase, we must move closer to
an object to see it as well as before. This phenomenon is particularly a problem in pristine areas, where long-
range transport of pollution may increase naturally low particulate levels and significantly reduce viewing
distance. For example, in pristine areas of the Southwest, where visibility is exceptionally good, small
increases in sulfate concentrations can lead to readily apparent reductions in visibility (National Research
Council, 1993, p. 106).
This principle is illustrated in Exhibits 2-9, which characterize a range of visibility conditions at
Shenandoah National Park. Generated by the WinHaze computer program (Air Resource Specialists Inc.,
1998), the two top scenes in Exhibit 2-9 are of a clear day at Shenandoah in 1998 with a visual range of 94
miles, and a day slightly worse than the median, with a visual range of 40 miles. The two bottom scenes are
of relatively hazy days with a visual range of 13 and 11 miles. In both the top and bottom sequences, the
difference in visual range is the result of an additional five µg/m 3 of sulfates in the atmosphere. This illustrates
that the perceived change in visibility due to an additional five µg/m 3 of sulfates to an already degraded
atmosphere is less noticeable than adding it to a pristine atmosphere. Thus, to achieve a given level of
perceived visibility improvement, a larger reduction in fine particle concentrations is needed in more polluted
areas. Conversely, a small amount of pollution in a clean area can dramatically decrease visibility.

Exhibit 2-9 Shenandoah National Park: WinHaze Photos Showing Effect of a Five µg/m3 Increment
of Sulfate on a Clear Day and a Hazy Day
(a) Clear Day – visual range 94 kilometers (b) Clear Day + 5 µg/m 3 sulfate – visual range 40 kilometers
(c) Hazy Day – visual range 18 kilometers (d) Hazy Day + 5 µg/m 3 sulfate – visual range 14 kilometers
100
Clear Day
80
Visual Range (km)
60
40 3
Clear Day + 5 ug/m sulfate
20
Hazy Day
3
Hazy Day + 5 ug/m sulfate
0
13 18 23 28 33 38 43
Deciview Index

3. TRENDS IN VISIBILITY IMPAIRMENT
Historical airport visibility data shows that visibility declined significantly in the 1940s up through the
1960s in the Western U.S. In the East, this decline continued up until about 1980. Based on EPA (1998,
Figure 6-4), Exhibit 3-1 depicts 75th percentile light extinction at airports across the U.S. from 1970 to 1990.
Since then, while there have been some areas of improvement, significant problems remain.
A variety of ways to record atmospheric

visibility have been used to provide an idea of
how visibility impairment has changed over time.
Human eye observations of visual range have
been recorded at airport weather stations for most
of the 20th century and have only recently been
phased out in favor of more quantitative
measures. One of the best sources of recent data
is the Interagency Monitoring of Protected Visual
Environments (IMPROVE) program, which was
established in 1987 to provide a variety of
visibility measurements including detailed
measurements of particulate constituents (U.S.
EPA, 1996, Table 8-3).
The airport data are the most extensive,

as they cover hundreds of stations across the
United States, and go back to the early 1900s.
While these data are somewhat limited because of
variations in observers and inconsistent reporting
procedures, they are nevertheless useful in
developing historical trends. After analyzing
these data, Malm (1999, pp. 39-41) reported that
in the East visibility has generally worsened
between the late 1940s and early 1980s, especially
during the summer in the Southeast. This decline
in visibility is closely matched with an increase in
sulfur emissions (Malm, 1999, Figure 6.16b). In
the Rocky Mountains southwest, the trends are
mixed from 1948-1976. While in California
visibility declined from the late 1940s to 1966,
and has since generally improved. However,
Malm noted that while the overall average is
Exhibit 3-1 Airport Visual Data: Trend in 75th improving in California, the number of very good
Percentile Light Extinction Coefficient for July- or “superior” visibility days at a couple of pristine
September (measured in km-1) monitoring sites has gradually declined.

3.1 Recent Visibility Trends
In this section, we consider more closely how visibility has changed in specific areas throughout the
country; in areas where the vista is an integral characteristic of the place itself, at smaller parks and wilderness
areas, and in urban settings with scenic, “postcard” skylines. We have comprehensive visibility data since
1987, when the IMPROVE network of monitors was established at national parks throughout the nation.
3.1.1 Visibility at Parks Where the Vista Is Integral to the Experience
Pristine national parks and wilderness areas are clearly some of the areas where visibility is extremely
important. To provide a sample of visibility at national parks where the vista is integral to the experience, we
chose five well-known parks from throughout the country. They are Acadia National Park in Maine, Grand
Canyon National Park in Arizona, Great Smoky Mountain National Park in Tennessee and North Carolina,
Shenandoah National Park in Virginia, and Yosemite National Park in California.
The National Park Service (NPS-CIRA, 2000a) sorted each park's daily visibility measurements from
low to high for each year, and placed them into three groups for analysis: good visibility days are the lowest
20 percent of daily measurements, mid-range days are the middle 40-60 percent, and poor visibility days are
those days above the 80th percentile of the ordered data. We then plotted the annual average visibility
measurement for each of these categories. To put the visibility category ranges observed at each park into
perspective, we present actual photographs that represent the deciview levels for each national park for the
good and poor visibility categories.2 Exhibits 3-2 through 3-6 present the trends graph and companion photos
for each park. These trends were also examined by Sisler and Malm (2000), who conducted a statistical
analysis of the slopes of the park-specific visibility trend lines. Exhibit A-2 contains the slope of the visibility
trend line for each park, and many additional parks, over the last decade and identifies whether or not the trend
identified by the slope is statistically significant.
The “good,” “midrange,” and “poor” categories that we considered represent averages for the visibility
levels within each group, and do not capture the full range of visibility levels at these parks. In Exhibit 3-7 we
present photographs to capture the range of conditions for each park, from pristine to extremely poor visibility
days.
Over the last decade there were no major changes to visibility levels at the parks examined in this
analysis. Both improvements and declines in visibility were generally very slight between 1988 and 1998,
though there was a significant amount of variation in visibility in the range of years. The points plotted on the
graphs do not lead to smooth trend lines in one direction or another for any of the parks. It appears that some
of the air quality controls already in effect may be preventing significant additional deterioration to visibility
at these national parks.
At Acadia National Park, visibility improved slightly from 1988 to 1998 in each of the three visibility
categories, however, only the visibility trends in the good visibility and median visibility categories were
statistically significant (Exhibit A-2). Like Acadia, visibility improved slightly at Grand Canyon National Park
in each of the three categories over the last decade. However, only the trend over good days was even
marginally significant. Between 1988 and 1998, visibility at Great Smoky Mountains National Park worsened
2
Air Resources Specialists, Inc. (2000c) did not report the deciview levels for the pictures of Half Dome, Yosemite. Using
a guide of roughly 4 deciviews for a good day and 16 deciviews for a poor day, we chose Half Dome pictures that compared
reasonably well with WinHaze photos of comparable deciview levels.

on the poorest days, and improved slightly on mid-range days and good days. None of the trends, though, were
statistically significant. Shenandoah National Park experienced declines in visual air quality on the poorest
visibility days and improvements in visibility on the mid-range and good days. Only improvements on the mid-
range days were found to be marginally significant. Finally, at Yosemite National Park, visibility improved
on the good and mid-range days, and worsened on the poor visibility days. None of the Yosemite trends,
however, were found to be significant.

Exhibit 3-2 Visibility Trends at Acadia National Park, Maine
( Photos: Air Resource Specialists Inc., 2000a, Img0004.pcd and Img0009.pcd; IMPROVE data: NPS-CIRA, 2000a)
Acadia National Park

Measurements of haze (in deciviews) and its effect on visibility
27
25
23
21
Deciviews
19
17
15
13
11
9
88 89 90 91 92 93 94 95 96 97 98
Year
Good Visibility Days Mid-Range Poor Visibility Days
Good Visibility Day (10 deciviews) Poor Visibility Day (23 deciviews)

Exhibit 3-3 Visibility Trends at Grand Canyon National Park, Arizona
( Photos: Air Resource Specialists Inc., 1997, Img0015.pcd and Img0023.pcd; IMPROVE data: NPS-CIRA, 2000a)
Grand Canyon National Park

14
12
Deciviews
10
4
89 90 91 92 93 94 95 96 97 98
Year

Exhibit 3-4 Visibility Trends at Great Smoky Mountains National Park, Tennessee
( Photos: Air Resource Specialists Inc., 2000b, Img0008.pcd and Img0013.pcd; IMPROVE data: NPS-CIRA, 2000a)
Great Smoky Mountains National Park

31
29
27
25
Deciviews
23
21
19
17
15
13
88 89 90 91 92 93 94 95 96 97 98
Year

Exhibit 3-5 Visibility Trends at Shenandoah National Park, Virginia
( Photos: Air Resource Specialists Inc., 1999, Img0082.pcd and Img0085.pcd; IMPROVE data: NPS-CIRA, 2000a)
Shenandoah National Park

32
30
28
26
Deciviews
24
22
20
18
16
14
88 89 90 91 92 93 94 95 96 97 98
Year

Exhibit 3-6 Visibility Trends at Yosemite National Park, California
( Photos: Air Resource Specialists Inc., 2000c, Img0002.pcd and Img0004.pcd; IMPROVE data: NPS-CIRA, 2000a)
Yosemite National Park

21
19
17
15
Deciviews
13
11
9
7
5
3
88 89 90 91 92 93 94 95 96 97 98
Year
Good Visibility Day Poor Visibility Day

Exhibit 3-7a Extreme Visibility Days at Acadia National Park
(Air Resource Specialists Inc., 2000a, Img0001.pcd and Img0012.pcd)
Excellent Visibility Day (3 deciviews) Bad Visibility Day (33 deciviews)
Exhibit 3-7b Extreme Visibility Days at Grand Canyon National Park

(Air Resource Specialists Inc., 1997, Img0010.pcd and Img0024.pcd)
Exhibit 3-7c Extreme Visibility Days at Great Smoky Mountains National Park
(Air Resource Specialists Inc., 2000b, Img0016.pcd and Img0026.pcd)

Exhibit 3-7d Extreme Visibility Days at Shenandoah National Park
(Air Resource Specialists Inc., 1999, Img0001.pcd and Img0011.pcd)
Exhibit 3-7e Extreme Visibility Days at Yosemite National Park

(Air Resource Specialists Inc., 2000c, Img0001.pcd and Img0005.pcd)
Excellent Visibility Day Bad Visibility Day
3.1.2 Visibility at Smaller Parks and Wilderness Areas
We may only think of major national parks when we think of the impacts of visibility impairment, but
poor visual air quality impacts smaller parks and wilderness areas, as well. Though the data may be sparse
for these parks, more data are added annually as the IMPROVE system includes new areas into the monitoring
system (NPS-CIRA, 2000a). We considered visibility trends at two lesser known areas, San Gorgonio
Wilderness Area in California and Chassahowitzka in Florida to demonstrate that visibility trends observed
at larger, well known parks can also be observed at smaller parks and wilderness areas.
At the San Gorgonio Wilderness Area, visibility on poor days improved between 1988 and 1998
(Exhibit 3-8), due primarily to a reduction in ambient ammonium nitrate (NPS-CIRA, 2000a). This trend was
found to be statistically significant (Exhibit A-2). Mid-range days and good days also experienced
improvements in visibility, though they were slight and not statistically significant. At the Chassahowitzka
wilderness area, where visibility data has only been collected since 1993, we again see the visibility day

categories remained relatively constant over the six year time period (Exhibit 3-9). The significance of the
trends for each of the visibility day categories, however, were not calculated, so no inference on the statistical
significance of the trends can be made. There appears, however, to be a slight increase in 1998 deciview levels
compared to those in 1993. The absence of large swings in visual air quality echo the trends seen at the larger
parks – that visibility has not changed dramatically over the last decade and that improvements can still be
made.
Exhibit 3-8 Visibility Trends at San Gorgonio Wilderness Area, California

(IMPROVE data: NPS-CIRA, 2000a)
San Gorgonio Wilderness Area

26
24
22
20
Deciviews
18
16
14
12
10
8
6
4
88 89 90 91 92 93 94 95 96 97 98
Year
Exhibit 3-9 Visibility Trends at Chassahowitzka Wilderness Area, Florida

(IMPROVE data: NPS-CIRA, 2000a)
Chassahowitzka Wilderness Area

29
27
25
Deciviews
23
21
19
17
93 94 95 96 97 98
Year

3.1.3 Visibility in Urban Settings
Urban settings are also impacted by trends in visibility. The IMPROVE network has acknowledged
this fact by placing a monitor in Washington, D.C., a tourist destination where the vistas of the historical sights
and monuments play an integral role to the attraction of the city. Though data and photos are only examined
here for Washington, D.C., the importance of visibility in other urban areas with a “postcard skyline,” like New
York City, San Francisco, or Boston, should not be overlooked. Exhibit 3-10 shows that visibility levels for
all visibility day categories have improved slightly between 1988 and 1998, though the trends were not found
to be significant (NPS-CIRA, 2000a).3
3
Savig (2000) did not report the deciview levels for a series of five pictures of Washington, D.C., ranging from excellent to
very poor visibility. To estimate good and poor visibility levels, we present the second and fourth pictures of the series.

Exhibit 3-10 Visibility Trends in Washington, D.C.
( IMPROVE data: NPS-CIRA, 2000a; Savig, 2000, naca_2.jpg and naca_4.jpg)
Washington, DC
33
31
29
Deciviews
27
25
23
21
19
17
89 90 91 92 93 94 95 96 97 98
Year
Good Visibility Day Poor Visibility Day

4. LEGISLATIVE AND REGULATORY HISTORY OF VISIBILITY IMPAIRMENT
Over the past 25 years there have been a number of legislative and regulatory initiatives designed to
improve visibility, such as the Clean Air Act Amendments of 1977 and the 1999 Regional Haze Rule. In
addition, there are a number of other legislative and regulatory initiatives that should reduce visibility
impairment in addition to the primary goal of reducing ambient pollutant levels. For example, the Title IV acid
rain provision of the Clean Air Act is reducing SO2 emissions, which will reduce the visibility impairment
related to sulfate aerosols. Similarly, tighter tailpipe emission standards for gasoline and diesel-powered
vehicles, along with programs for cleaner-burning fuels, will further reduce visibility impairment in some areas.
The Clean Air Act Amendments of 1977 included the first significant federal legislation that
specifically addresses visibility impairment. The 1977 Amendments established a national goal of “prevention
of any future, and the remedying of any existing, impairment of visibility in Class I Areas which impairment
comes from manmade pollution.” It included two programs aimed at visibility impairment in Class I Areas
(pristine wilderness locations “of great scenic importance”). The first is the prevention of significant
deterioration (PSD) program outlined in Sections 160-169, which is aimed at reducing ambient levels of criteria
pollutants. The PSD program requires that new or modified major emitting facilities must not adversely affect
nearby Class I Areas. The second is the Section 169 regional haze program, where Congress set a national
goal of preventing and remedying visibility impairment in pristine areas of the U.S.
The pristine areas identified in the legislation come from a group of so-called “mandatory Class I
Areas”, which are selected national monuments, wilderness, wildlife refuge and memorial areas and parks
larger than 5,000 acres, national parks over 6,000 acres, and all international parks in existence on the day
President Carter signed the 1977 Amendments into law. The legislation required EPA to identify from the 158
mandatory Class I Areas those in which visibility is an important value. In November 1979, EPA complied
and identified 156 areas including one international park in the Virgin Islands (Scott and Stonefield, 1990,
Table 1). Three federal agencies have primary responsibility for most of these areas: Forest Service (98
wilderness areas), National Park Service (36 national), and Fish & Wildlife Service (21 wilderness areas).
Exhibit 4-1 maps the 156 parks and wilderness areas chosen by EPA. Adding new Class I Areas would require
an Act of Congress.
In addition to identifying areas where visibility is an important value, the 1977 Amendments require
EPA to promulgate regulations to assure reasonable progress in meeting the goal of preventing and remedying
visibility impairment. Impairment causes were categorized as either “reasonably attributable to an individual
source or small group of sources”, or as regional haze, which “emanate(s) from a variety of ...regionally
distributed sources”. As a first major step in 1980, EPA established regulations to address visibility
impairment in Class I Areas that could be reasonably attributed to major stationary air pollution sources (U.S.
EPA, 1996, p. 8-2). At that time, EPA deferred regulatory action on regional haze until they had better
scientific tools, and instead focused on more local problems.
Responsibility for identifying and regulating specific sources that impair visibility involves three
different parties: the federal agency managing a Class I Area , the state where the source is located, and the
EPA. If a Federal Land Manager establishes that visibility at the Class I Area is impaired, the EPA must make
a “reasonably attributable” decision linking the impairment with a specific source. The State then conducts
a case-by-case review to determine what is the best available retrofit technology (BART) to control the source’s
emissions. The BART determination for a specific source depends on a number of factors, including the
control technologies that are available, the cost installing and operating the available technologies, all
environmental impacts of compliance, pollution control equipment already existing at the source, the remaining
useful life of the source, and the degree of improvement in visibility that may result from the use of BART.

Aside from the possibility that large point sources chose not to locate near Class I Areas, the 1980
regulations developed by EPA had limited impact on existing point sources. Identifying specific sources that
impair visibility has proven to be difficult, in large part because visibility is a regional problem with
contributions from many sources over a wide area (Latimer, 1990, p. 51). EPA has initiated separate studies
of the Navajo Generating Station (NGS) and the Mojave Power project. To date only the NGS, a large power
plant in Page, Arizona, has added BART to its facility. Extensive atmospheric studies demonstrated that the
NGS, located on the Colorado River at the north-eastern edge of the Grand Canyon National Park, was a major
cause of haze within the Canyon during certain wintertime conditions. At other times of the year the NGS is
only one of many sources that contribute to the major haze problem within the Canyon. In 1991 the NGS
BART review resulted in a negotiated settlement to reduce sulfur emissions from NGS by 90 percent, with
additional reductions in NOx fine particle emissions. Those emissions controls are now operating at the NGS,
helping to improve visibility not only at the Grand Canyon but throughout the “golden circle” of National Parks
and Monuments on the Colorado Plateau
Negotiated agreements to improve Class I visibility have also been reached involving two other major
western power plants: the Mojave and Centralia Power Plants. A 1996 negotiated settlement avoided a formal
and potentially lengthy BART proceeding, while successfully leading to a reduction in emission producing haze
at Mt. Rainier National Park in the State of Washington. The Centralia Power Plant is located 50 miles
southwest of Mt. Rainier, and is the largest remaining point source of sulfur emissions in the western United
States. Emitting over 69,000 tons of sulfur annually, Centralia is estimated to cause a third of the visibility-
impairing sulfur concentrations at Mr. Rainier NP, and a quarter of the haze-induced bad visibility days. The
agreement reached by the State of Washington, local regulatory agencies, the National Park Service, the EPA,
and PacifiCorp (one of the plant’s owners) will reduce sulfur emissions by 90 percent by 2002 (National Park
Service, 1997).
Another negotiated agreement reached in 1999 involves the Mojave Power Plant, located 75 miles west
of the Grand Canyon in Laughlin, Nevada. Like NGS, Mojave has also been linked to haze in the Grand
Canyon. The settlement will reduce Mojave’s sulfur emissions by over 85 percent. Construction planning is
beginning, and Mojave will meet the new emission limits by the end of 2005.
Progress on controlling regional haze began with the 1990 Clean Air Act Amendments, which added
Section 169B authorizing EPA to conduct research on regional haze (National Research Council, 1993, p. 61).
The research involves an expansion of visibility-related monitoring, assessment of sources of visibility-
impairing pollution, adaption of air quality models to measure visibility, studies on the chemistry and physics
of visibility, and an assessment of visibility levels in Class I Areas every five years. The 1990 Amendments
also allowed EPA to establish “visibility transport regions” for any Class I Area whose visibility is impaired
by the interstate transport of air pollution, and required establishing a transport region for the Grand Canyon
National Park. EPA later expanded this to include 15 other Class I parks and wilderness areas on the Colorado
Plateau (U.S. EPA, 1996, p. 8-2). The Grand Canyon Transport Commission must assess current and
projected emission sources and suggest corrective action, and in turn EPA must develop regulations that result
in reasonable progress toward reducing visibility impairment. The regulations to protect the Colorado Plateau
must also be coordinated with other federal regional haze and PM programs.
In response to the problem of regional haze, the National Academy of Science established a committee
to address regional haze in national parks and wilderness areas. The committee considered the state of
knowledge on a variety of issues including determining individual source contributions to visibility impairment,
factors that affect haze, improvements in air quality models, and emission controls. In 1993, the committee
published an influential report of their findings (National Research Council, 1993). The report stated that the
current state of science is adequate and control technologies are available to improve and protect visibility.

However, a regional approach is necessary to control visibility impairment, and an approach that focuses on
individual emission sources is “doomed to failure” (p. 7).
In part propelled by this report, on July 31, 1997, EPA published proposed amendments to its 1980
regulations that would require control of regional haze. The“Regional Haze Rule” calls for state and federal
agencies to work together to improve visibility, and establishes goals for each Class I Area that are designed
to improve visibility on the worst days and prevent degradation of visibility on the best days. Each state must
address its contribution to visibility problems in national parks and wilderness areas both within and outside
its borders, and to develop long-term strategies aimed at returning visibility to natural conditions. The first
State plans for regional haze are due by 2008.
The Grand Canyon Visibility Transport Commission (GCVTC) is the only visibility transport
commission established to date. The GCVTC consists of the governors of eight western states, leaders of five
Native American Tribes, and five federal agencies. They conducted an extensive scientific and policy analysis
project designed to identify an effective combination of policy recommendations to protect and improve
visibility at 16 National Parks and Wilderness Areas that make up the “Golden Circle” of parks on the
Colorado Plateau. The GCVTC emphasized the importance of active participation by a broad range of stake
holders, and participants ranged from individual firms to environmental organizations. In 1996 the GCVTC
completed their recommendations, designed to improve visibility on the Colorado Plateau by limiting emissions,
protecting “clean air corridors”, increasing the monitoring, and integrating visibility considerations into forest
fire management policies.
Although the GCVTC is the only visibility transport commission, other multi-state organizations are
involved with regional visibility as part of integrated regional air quality planning activities covering all
portions of the continental United States. These organizations are sponsoring a wide range of activities to
understand the unique causes, effects, and policy alternatives for improving all aspects of air pollution
including visibility within their region. Currently, regional organizations include:
! The Western Region Air Partnership, or WRAP [http://www.wrapair.org] is a successor

organization to the GCVTC, with a goal of implementing the GCVTC recommendations. It
includes 12 states: Arizona, California, Colorado, Idaho, Montana, New Mexico, North
Dakota, Oregon, South Dakota, Utah, Washington and Wyoming.
! The Western States Air Resources Council, WESTAR [http://www.westar.org] is an

organization of air agencies from 15 western states: Alaska, Arizona California, Colorado,
Hawaii, Idaho, Montana, Nevada, New Mexico, North Dakota, Oregon, South Dakota, Utah,
Washington, Wyoming.
! The Central States Air Resource Agencies, CesSARA [http://www.censara.org] includes air
agencies from nine states: Nebraska, Kansas, Oklahoma, Texas, Minnesota, Iowa, Missouri,
Arkansas and Louisiana.
! The Lake Michigan Air Directors Consortium, LADCO [http://www.ladco.org] includes the
four states: Illinois, Indiana, Michigan, and Wisconsin.
! The Mid-Atlantic Regional Air Management Association, MARAMA

[http://www.marama.org] includes nine states: Delaware, the District of Columbia, Maryland,
New Jersey, North Carolina, Pennsylvania, Virginia, West Virginia and the District of
Columbia.

! The Northeast States for Coordinated Air Use Management, NESCAUM
[http://www.nescaum.org] includes eight states: Maine, New Hampshire, Vermont,
Massachusetts, Connecticut, Rhode Island, New York and New Jersey.
! The Southeastern States Air Resources Managers, SESARM [http://www.metro4.org]

includes 8 states: Florida, Georgia, Alabama, Mississippi, Tennessee, Kentucky, North
Carolina and South Carolina..
! The Ozone Transport Commission, OTC [http://www.sso.org/otc] includes 14 states:

Connecticut, Delaware, the District of Columbia, Maine, Maryland, Massachusetts, New
Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont, and Virginia .
! The Southern Appalachian Mountains Initiative, SAMI [http://www.saminet.org] focuses on

the mountainous regions of 8 states: Alabama, Georgia, Kentucky, North Carolina, South
Carolina, Tennessee, Virginia, and West Virginia
The EPA maintains a website [http://www.epa.gov/oar/vis/] with links to each of the regional visibility
planning organizations, as well as other federal agencies and programs involved with regional visibility issues.

Exhibit 4-1 Map of Mandatory Class I Areas with Visibility Value
(Source: http://www.epa.gov/oar/vis/)

5. ECONOMICS: VISIBILITY IMPAIRMENT AND PARK VISITATION
Most people would probably say they want good visibility both in the communities where they live,
and in the national parks and scenic areas they visit for relaxation and recreation – and the evidence suggests
that it is not just talk. It is one thing to say you want something; it is another to actually be willing to pay for
it. In fact, that is how economists measure the economic value of something – by how much people are willing
to pay for it. If it is a good that can be bought in a store, the market price reflects its value. Visibility,
however, is a commodity that can not be bought or sold; in economic terms, visibility is considered a non-
market good. This makes it difficult to measure what it is worth to people.
Another fundamental challenge to valuing visibility is that a change in visibility must be perceptible
to people if they are going to place some value on that change. However, it is not so obvious how small
changes in visibility conditions should be treated when trying to gauge how people feel about visibility
impairment. Some changes, especially when measured in seasonal or annual averages, may not exceed
perception thresholds. For instance, a change in visibility may be less than one deciview, though one deciview
is approximately described as a just noticeable difference in visual air quality. When visibility changes are this
low, it is often suggested that they have no value. This conclusion, however, is wrong for two reasons. First,
whether or not a change is interpreted as perceptible may depend on the averaging time used to measure the
change. Visibility changes may be largest during certain times of the year on certain days. When these changes
are averaged over a season or the entire year, however, the overall change may appear to be quite small and
incorrectly treated as having no value. Second, while a single change may not be perceptible, the cumulative
effect of all visibility changes over an extended time period may be perceptible. Yet, specific policy decisions
that affect visibility are most always evaluated individually and incrementally. A policy scenario may only
create a small, perhaps even imperceptible, change in visibility. The change should still be valued, however,
because this change contributes to overall visibility improvements and may make a very large difference over
time.
Despite these obstacles – that visibility is a non-market good and visibility often changes in increments
the public may or may not notice – economists are devising ways of measuring the value of changes in
visibility. One approach that has been used by economists measures what improved visibility is worth to
people based upon what they say their willingness to pay for better visibility is, both at home and in natural
settings. This approach does not depend on a person’s actual behavior, but on what they say their behavior
would be with better visibility. Chapter 6 discusses this approach in detail.
Another approach is to measure the economic impacts of people’s behavior based upon improvements
or declines in a particular resource. There are a number of behavioral changes people may make when
confronted with poor visual air quality. For instance, people may base their decision to purchase a house, at
least in part, on the quality of the surrounding views and visibility, which can impact the housing market. If
the decision to relocate to an entirely new area is also influenced by visual air quality, the economic impacts
could reach beyond just the housing market and impact the entire local economy. These types of economic
impacts have the potential to be quite substantial if visibility is extremely poor. However, to isolate that
portion of the impact due to poor visibility from all of the additional factors that go into purchasing a house
or relocating to another area is extremely difficult to do.
Another type of behavioral change that is much easier to measure, and easier to attribute to visibility,
is the affect poor air quality has on a person’s decision to recreate. If, because of poor visibility, a person
decides to shorten their visit to a particular park, or go to another recreation site altogether, the economic
impacts will be felt in the way of lost revenue at the recreation site (park fees, lodging, concessions) and lost
revenue in the communities close to the site (gas, food, lodging, concessions, additional tourist attractions).

While visibility degradation is likely to affect most types of outdoor recreation to one degree or another, the
impacts of impaired visibility are perhaps most pronounced at national parks.
5.1 An Undisturbed Environment, Including Clean, Clear Air and Good Visibility, is Very
Important to Park Visitors
It has consistently been shown that American vacationers believe that one of the leading attributes of
a desirable travel destination is beautiful scenery and clean, clear air. In fact, the National Park Service (NPS)
has conducted a number of studies that have examined the importance of clean air as a park feature. The
results of these studies were summarized by the NPS (1988).
A 1983 NPS study confirmed that visitors are able to perceive different degrees of visibility
impairment. Visitors were asked if they had noticed haze at the parks, and if so, whether they thought it was
slightly, moderately, very, or extremely hazy. After comparing their responses to actual visibility measures
taken on the same day, they found that when the range of view was lower, visitors were more aware of haze
and were more likely to say it was very to extremely hazy. This same study also found that visitors to the
Grand Canyon and Mesa Verde National Parks who said the view was hazy enjoyed the park less than those
visitors who said they were not aware of haze or were aware of only slight to moderate haze. This meant that
not only did park visitors notice haze, but when they considered the view to be relatively hazy, it detracted from
their enjoyment of the park.
Another series of NPS studies conducted during the summers of 1983, 1984, and 1985, found that
people visit parks first to experience a natural setting and second to enjoy specific unique features associated
with various parks. Surveys were given to visitors at Grand Canyon, Mesa Verde, Mount Rainier, Great
Smoky Mountains, and Everglades National Parks. The surveys listed a number of park features and asked
visitors how important each one was to their recreational experience. Some of the listed features were the same
at all the parks and some were specific to each park. For example, “clean, clear air” and “interpretive
signs/information” were listed for all the parks while “viewing canyon rims” was listed for the Grand Canyon,
“ruins on mesa tops” was listed for Mesa Verde, and “views of chimney tops” (natural landscape features in
the area) was listed for Great Smoky Mountains.
The survey results revealed that it is very important to visitors that parks be natural and free of
pollution; in other words, as undisturbed by humans as possible. In fact, the survey consistently showed the
importance of clean air to the recreational experience, with “clean, clear air” one of the top four features at
every park. At the Grand Canyon, over 80 percent of the respondents rated “clean, clear air” as very important
or extremely important to their recreational experience.
Additional analysis of these surveys was conducted to determine if, when asked what the most
important features were at these parks, visitors identified features that belonged to a common recreational
theme. For example, features such as “clean, clear air” and “park cleanliness” were grouped into a
“naturalness” category. Based on work by the National Park Service (1988), Exhibit 5-1 shows the groups
of park features and their relative importance at three sample parks: Grand Canyon, Mesa Verde, and Great
Smoky Mountain. The most significant finding was that the group of “naturalness” features (which included
“clean, clear air”) was rated the most important at each park. The second most important set of park features
was associated with each park’s unique qualities. For example, while “naturalness” was the most important
group of features at both Grand Canyon and Mesa Verde, “viewing scenic vistas” at Grand Canyon and
“information/park history” at Mesa Verde ranked as the second most important group of features at those
parks, respectively.

Exhibit 5-1 Visitor Rated Importance of Park Features
Not at All Slightly Moderately Very Extremely

Important Important Important Important Important
Naturalness
Viewing
Grand
Information
Canyon
Activity Related
Visual Obscurement
Naturalness
Information
Mesa
Viewing
Verde
Activity Related
Visual Obscurement
Naturalness
Backcountry
Information
Great Smoky
Flora-Fauna
Mountains
Viewing
Management
Backcountry Reservations
5.2 Visitors are Willing to Alter Their Length of Stay Based on Visual Air Quality Conditions at
National Parks
The above findings suggest that if visibility as a park resource was allowed to deteriorate, visitor
enjoyment of the parks would decline. But how would a change in visual air quality affect visitation patterns?
When confronted with poor visibility at their destination, it is likely that travelers will do one of two things:
shorten their stay at a national park or go elsewhere. In either case, we assume that people will allocate the
time available to them in such a way as to maximize their enjoyment. There have been a number of studies
conducted to evaluate whether visitors would be willing to spend more time traveling to alternative viewing sites

in order to obtain better visual air quality during their park visit. Results from these studies provide evidence
that changes in visual air quality affects visitor use patterns.
Summarized in NPS (1988), a 1983 NPS study conducted at Grand Canyon National Park asked
visitors to rank possible alternative combinations of travel time to vistas and visibility conditions through the
use of photographs. These rankings revealed that the average change in the amount of time a visitor was
willing to spend traveling to a vista for a 10 Mm-1 change in visibility was between 15 minutes and 4 hours.
Studies conducted outside the scope of the NPS also support the finding that as perceived visual air
quality gets worse, visitation patterns to national parks is altered. One study found that if visibility at a vista
in either the Grand Canyon or Mesa Verde national parks changed from “average” to “poor,” 61 percent of
the survey participants said they would spend less time at the vista, while 80 percent said they would spend less
time total at the park (MacFarland et al., 1983). The average stated reduction in park visitation was about 13
hours, which is quite significant when compared to the average park visit of 14 hours.
In a 1985 study conducted at Grand Canyon National Park (Bell et al., 1985), visitors were asked to
participate in a simulation where they could choose between four hypothetical activities: viewing three vista
points (depicted on photographs) and touring an archaeological site. The driving time, scenic beauty, and
visibility of each vista, as well as the waiting time of the archeological tour, were varied between participants.
Analysis of the time a visitor would spend on each activity showed that visitors placed a high value on small
changes in visual air quality. For example, for one of the vista points, visitors revealed their willingness to
increase their driving time an average of about one-half hour for an improvement in the visual air quality scale
used in the survey.
Other studies, in slightly different recreational settings to national parks, have also shown that a
positive relationship between visibility conditions and actual visitation exists. A recent study examining the
impact that reducing the quality of views had on travel to the Blue Ridge Parkway in North Carolina found that
if 25 percent of the views in the park were at least partially obstructed, visits to the park would decrease from
2 to 1 at the median (Kask, 1998). Though the nature of the view obstruction examined in this study was not
visual air quality but actual physical obstructions like new homes, logging, and telephone poles, it still
demonstrates that visitors value natural, unobstructed views. In wilderness settings, it has been shown that
users of the Colorado wilderness ranked “clean, fresh air” as the most important of 73 physical resource
attributes (Brown et al., 1977). Similarly, “viewing scenery” was found to be the most important of a list of
20 wilderness experiences (Walsh and Loomis, 1982). These findings add to the evidence that visitation to
national parks, and participation in outdoor recreation in general, is clearly impacted when visitors are aware
that their ability to view scenic vistas is impaired due to poor visual air quality.
5.3 What is at Stake to the Economy if Visitation Rates Decline?
Tourism to national parks has grown tremendously, increasing from approximately 30 million visits
in 1951 to 287 million in 1998 (see Exhibit 5-2 based on National Park Service, 2000). The park system has
also expanded, with 376 parks comprising a total of 83 million acres and employing approximately 21,000
permanent and seasonal employees. There are also over 600 concessionaires in operation throughout the
national park system providing visitors with lodging, transportation, food, shops and recreational services. In
1997, the NPS collected $122.2 million in recreation fee revenue (entrance fees, season passes, parking fees)
and sales from concessions were estimated to be $650 million (U.S. Department of the Interior, 1988). One
study, conducted by the Department of the Interior, measured the state and national economic impacts
associated with all travel related expenditures (Peacock et al., 1998). The study estimated that travel related
expenditures by visitors to National Parks generated an average of $14.55 billion (1996$) in annual output and

approximately 210,000 jobs. By including visitation to all lands managed by the Department of the Interior
(Bureau of Land Management, Fish and Wildlife Service, Bureau of Reclamation, and National Park Service)
annual economic impacts associated with travel related expenditures increased to an average of $35 billion
(1996$) and generated approximately 500,000 jobs.
Exhibit 5-2 Total Recreation Visits to U.S. National Parks, 1951-1998
350
300
Number of Visits (millions)
250
200
150
100
50
0
1951
1954
1957
1960
1963
1966
1969
1972
1975
1978
1981
1984
1987
1990
1993
1996
Year
The majority of revenue associated with visitation to national parks and other federally-managed lands
is generated outside of the park boundaries. Recall that revenue from in-park concessions and park recreational
fees is only about $800 million, while total estimated revenue associated with park visitation is estimated to
be $14.55 billion. This out-of-park revenue results in three categories of benefits to local communities. First,
out-of-park expenditures are a primary source of income to local area businesses and individuals that provide
goods and services to park visitors traveling from out-of-town. Next, sales to park visitors increase local area
tax revenue that result from expenditures by non-local park visitors. Third, expenditures by non-local park
visitors create new jobs in the local area.
One study that measured the potential economic impact of decreased visitation to a recreational area
was conducted by Kask (1998), which looked at how obstructed visibility affected the number of visits to the
Blue Ridge Parkway in North Carolina. Using total visitation revenue statistics compiled by Brothers et al.
(Brothers and Chen, 1997) for the North Carolina portion of the Blue Ridge Parkway, Kask estimated that with
a reduction in annual visitation from 2 visits to 1, the potential monetary impact was, on average, $641 million
in direct expenditures and $51 million in lost tax revenues.
It is easy to see that if visitor patterns were to change due to impaired visibility (if visitors were
spending less time in a given area), revenue generation within the local economy would also decrease. And,
if visitors altered their travel plans altogether and decided to forego a park visit due to poor visual air quality,
not only would the local economy lose out on these visitors' expenditures, but the national park and the
concessionaires within the park would lose, as well. Dollars spent in and around the park would be spent
elsewhere on other goods and services. On the other hand, if visibility improved at national parks, research
suggests that it is not unreasonable to expect that the number of visitors to national parks would increase and/or
the time spent at the park would increase, as well.

5.4 Local Economic Benefits from Visibility Improvements
In 1990, the NPS developed an economic model to generate quick and inexpensive estimates of the
economic impacts of national parks on local communities (National Park Service, 1998). This model, called
the Money Generation Model (MGM), focuses primarily on the economic impacts associated with National
Park tourism expenditures. Three types of economic benefits are estimated:
! Sales Benefits: Income to local area businesses or individuals for goods and services that are
provided as a result of expenditures by non-local park visitors;
! Tax Benefits: Increases in local area tax revenues that result from expenditures by non-local
park visitors;
! Job Benefits: New jobs created in the local area as a result of expenditures by non-local park
visitors.
To illustrate the potential economic impact increased visitation due to improved visibility has on
communities that neighbor national parks, we have run the model for seven National Parks: Acadia NP in
Maine, Big Bend NP in Texas, Grand Canyon NP in Arizona, Great Smoky Mountains NP in Tennessee and
North Carolina, Indiana Dunes National Lakeshore in Indiana, Mt. Rainier NP in Washington, and Shenandoah
NP in Virginia. These parks were chosen for a number of reasons. First, visibility is an important feature at
each of these parks. Visitors travel to these parks not only to recreate, but also to see the views. For instance,
the Blue Ridge Parkway, which spans the length of Shenandoah NP, is often driven for the scenic vistas alone.
Second, these parks span a wide range of the nation and are impacted by a number of different pollution
sources, including power plants. In fact, Great Smoky Mountains and Shenandoah National Parks are the two
most air polluted parks in the nation, due in large part to emissions from electricity generating stations located
upwind of each park. Third, the parks afford visitors easy access to destination vistas and are located in areas
with many alternative entertainment options outside of the parks themselves. If a park visitor decides to shorten
their stay due to poor visibility and take their tourist dollars elsewhere (to communities beyond those that
immediately neighbor the national park), there are plenty of choices that are within easy driving distance.
To run the model, a number of inputs and assumptions are necessary. These include:
! Recreation Visitor Days. A visitor day represents the time visitors spend in the park,
measured in twelve hour increments. One visitor day equals twelve hours spent in the park
by one or a combination of visitors. Visitor days are calculated by multiplying the number
of recreation visits to a particular park (measured as the number of people that enter the park)
by the average length of stay at the park and dividing this product by twelve. Visitor days are
reported for each park in the NPS statistical abstract.
! Non-Local Percent of Park Use. Sales benefits are based only on spending by non-local
park users rather than spending by people who live around the park. The park attracts non-
local people to the area, and their spending is income for local businesses that supply a wide
variety of goods and services.

! Average Daily Expenditures. In addition to lodging and meal costs, visitors spend money on
a variety of things such as fuel, souvenirs, admissions and tolls, rentals, retail sales and other
recreational pursuits.
! Sales Multiplier. A sales multiplier attempts to estimate the secondary effects of visitor
spending. For example, when tourists purchase meals at a local restaurant the restaurant
owner then has to purchase additional inputs such as groceries and utilities to continue and
perhaps expand their operations. Their spending creates additional spending, and so on. A
sales multiplier can be thought of as a way to measure the number of times a tourist dollar
changes hands within a local community. Sales multipliers for parks in this analysis ranged
from 1.1 to 2.4.
! Taxable Income Ratio. This is the taxable portion of salaries and business profits. Studies
indicate this multiplier has a range of 0.2 to 0.6. This analysis used an average taxable
income ratio of 0.3.
! Jobs per Million Dollars. Studies indicate jobs vary with the volume of sales. For the
tourism industry the number of jobs has been found to range from 10 to 50 with an average
of 30 jobs created per million dollars in total sales. In this analysis, job rates varied from 15
to 47 jobs per million dollars.
Other variables can be included in the model, such as Federal Government expenditures on park-related
items (employee salaries, supplies, services, construction) and State park-related expenditures (like park access
roads). These expenditures, however, are not directly dependent on the fluctuation of visitor patterns (though
a case could be made that an indirect relationship exists between park-related Federal and State expenditures
and park visitation rates -- i.e. parks with a higher number of visitors require greater expenditures on
infrastructure), and tend not to fluctuate by great amounts from year to year. We therefore exclude these
expenditures from the model and estimate economic benefits of the parks considering only money spent by non-
local tourists. Parameters used in the MGM for this analysis were taken directly from those reported by the
NPS (Wade, 2000).
Using the above input variables, each category of economic benefit due to non-local visitation at
national parks is calculated as follows. Direct sales benefits equal the product of non-local percent of park use
multiplied by the number of recreation visitor days multiplied by the average daily expenditures of park
visitors. Sales tax revenue is equal to the product of total direct sales multiplied by the combined state and
local sales tax rate. Income tax revenue is equal to the product of total direct sales multiplied by the taxable
income ratio multiplied by the combined state and local income tax rate. Job benefits resulting from tourism
expenditures are equal to the product of total direct sales multiplied by the average number of jobs per million
dollars.
The impact of increased visitation due to an improvement in air quality on each of these benefits
categories is calculated for the case-study parks in the following manner. First, the MGM is run using input
data that reflect the status quo. Next, the MGM is run based on an assumption of increased visitation (an
increase in a visitor’s length of stay) due to improved visibility assumed to take place at each park. Economic
impacts can then calculated as the difference between the status quo economic benefits and the economic
benefits calculated when improved visibility is assumed to affect visitation. We have run two scenarios per
park that vary the impact visibility impairment has on park visitation to present a range of possible results.
Results are presented in Exhibit 5-3, 5-4, and 5-5. Exhibit 5-3 presents the per park impact on sales revenues
with an increase in visitation, Exhibit 5-4 presents the per park impact on tax revenues with an increase in
visitation, and Exhibit 5-5 presents the per-park impact visitation increases have on job creation.

We chose a conservative range of increased visitation rates to reflect the uncertainty of this exercise.
We assume that in the first scenario, visibility improves enough to increase visitation at the park by 10 percent.
In the second scenario, we assume that visibility improves enough to increase visitation at the park by 25
percent. It is important to point out that changes in visitation patterns can be translated into changes in length
of stay. The Money Generation Model uses recreation visitor days when calculating the economic impacts of
visitation. As noted above, a recreation visitor day is measured as twelve hours spent in a park. It is calculated
as the number of people that enter a given park multiplied by a park visitor’s average length of stay divided
by twelve. Using this equation, if visitation to a park decreases, a corresponding reduction in length of stay
can also be calculated.
Recall that one study (MacFarland et al., 1983) found that the average stated reduction in park
visitation due to poor visibility at Grand Canyon and Mesa Verde national parks was 13 hours (an 84%
reduction in visitation). Another study (NPS, 1988) found that visitors to the Grand Canyon would be willing
to travel 15 minutes to 4 hours just to find alternative viewing sites where the visual air quality was better (a
2% to 26% reduction in visitation), while another (Kask, 1998) found that if 25% of the natural views on the
Blue Ridge Parkway in North Carolina were obstructed, the number of visits would be cut in half (a 50%
reduction in visitation). For parks included in this example, the average length of stay for visitors ranged from
2 ½ hours at Indiana Dunes to 15 ½ hours at Grand Canyon, and 30 hours at Big Bend. If we assume that a
given park has an average length of stay of 8 hours, a 10 percent increase in visitor length of stay equals about
50 minutes, and a 25 percent increase in length of stay equals about two hours. Even with the conservative
assumption that impaired visual air quality will lessen a visitor’s length of stay by 10 percent, one can see that
the potential impacts presented in Exhibits 5-3 through 5-5 are substantial.
It should be noted that the MGM model was designed to be a very simple and low cost way to estimate
economic benefits of parks for local economies. Therefore, the results are only as good as the assumptions the
model makes. Assumptions were made about taxable income ratios, indirect sales multipliers, job rates, etc.,
in order to simplify the economic benefit calculations. Because the accuracy of the results depends on the
quality of the input data, every attempt was made by the NPS to use values representative of each local area
that was examined. However, the benefits presented in this section should only be interpreted as an indicator
of the potential economic benefits improved visibility will have on the local economies surrounding national
parks, and should not be interpreted as actual benefits. The most important result output from the MGM model
is that for even small changes in visitor activity at each of the parks, there are potentially large positive benefits
to be gained.

Exhibit 5-3 Sales Increase Near Park Due to an Increase in Park Visitation
10% Increase in Visitation 25% Increase in Visitation
300
250
Sales Benefit (millions 1999$)
200
150
100
50
0
Acadia Big Bend Grand Great Smoky Indiana Mt. Rainier Shenandoah
Canyon Dunes
National Park
Exhibit 5-4 Tax Revenue Increase Near Park Due to an Increase in Park
Visitation
25
Tax Revenue Benefit (millions 1999$)
20
15
10
0
Canyon Dunes
National Park

Exhibit 5-5 Local Job Increase Near Park Due to an Increase in Park
Visitation
7000
6000
5000
Jobs Created
4000
3000
2000
1000
0
Canyon Dunes
National Park

6. ECONOMICS: THE NON-MARKET VALUE OF VISIBILITY
We saw in the previous section that one way to measure the worth of visibility is to estimate the impact
poor visibility has on visitation rates to recreational destinations and then estimate the economic impact these
visitation reductions have on the neighboring communities. This method is an attempt to convert visibility into
a good (tourism) whose value can then be measured on the marketplace. There is value in visibility
improvements, however, beyond the impact good visual air quality has on revenue generation in communities
close to national parks. Because this value is not reflected in a market, however, measuring it presents a
challenge. Economists have taken on the challenge to devise a way to measure the non-market value of
improvements in visibility by measuring how much people are willing to pay for it.
In tackling the problem of how to measure the economic value of preserving or improving visibility,
economists have made two important distinctions. The first distinction is between “residential visibility” – the
visibility in and around one’s community -- and “recreational visibility” – the visibility at places to which
people go for natural scenic beauty and recreation. Economists make this distinction primarily because the
values of these two types of visibility are likely to be different from each other. Even within the category of
“recreational visibility,” values for visibility improvements may vary substantially from one place to another.
People may care much more, for example, about good visibility at well known places like the Grand Canyon
and Shenandoah National Park than they do about good visibility at lesser known places. Another reason
economists distinguish between residential and recreational visibility is that, as discussed below, there are, at
least potentially, different techniques for measuring the values of the two types of visibility. Most public
policies that are designed to affect air pollution on a broad geographic basis will impact both residential and
recreational visibility for many people. Assessing the visibility-related economic value of such a policy will
therefore require valuing improvements in both types of visibility.
Economists also distinguish between “use” and “non-use” value. It is easy to see that people will value
things they can use or directly experience. Goods and services that people purchase and use obviously have
value. If they did not, people would be unwilling to purchase them at any price. Even though people do not
“use” an improvement in visibility in the same way they would use a market product or service, they derive
pleasure or satisfaction from directly experiencing it. How much a person is willing to pay for a visibility
improvement he would directly experience is his “use value” -- what it is worth to him to be able to enjoy that
improvement in visibility.
In addition, however, people often want, and are willing to pay for, improvements in visibility that they
themselves will not directly experience. They may value such improvements so that other people can enjoy
them, or so that future generations can enjoy them. Sometimes, just the knowledge that something exists has
value to people, regardless of whether anyone ever directly uses or experiences it. Economists call these types
of value “non-use” value, and the evidence suggests that they are indeed real.
Like most environmental quality improvements, improvements in visibility can be enjoyed by many
people. One person’s enjoyment of better visibility in no way diminishes or “uses up” another person’s
enjoyment of it. Because of this, the value of an improvement in visibility is the sum of all individuals’ values
– both use and non-use values -- for it. This is important, because, while any individual’s value for a particular
improvement in visibility may be relatively small, there may be many people who would derive some value.
The total value of the improvement can therefore be quite substantial.

6.1 The Economic Valuation of Visibility Improvements
We have thus far categorized different types of non-market value for visibility improvements. We
distinguished between the value of visibility improvements in residential areas (“residential visibility”) and in
places to which people go for natural scenic beauty and creation (“recreational visibility”); and we
distinguished “use” value from “non-use” value. The remainder of this section discusses the techniques that
have been developed for estimating these values. We begin with a discussion of the main techniques that have
been developed to value environmental quality improvements in general. We then focus specifically on the
methods and results of (1) the studies that have estimated the value of visibility improvements at national parks
and (2) the studies that have estimated the value of visibility improvements in residential areas. Finally, using
an example, we show how the specific information obtained from relevant studies may be applied to assess the
visibility benefits of reducing air pollution
The problems in measuring what people would be willing to pay for improvements in visibility are
formidable. Some researchers have looked at property values and examined how they change as various
characteristics that are likely to affect property values change. One of these characteristics is the quality of
the air in the surrounding community. It has been hypothesized that people are willing to pay something to live
in a location with better air quality, all else equal, and the evidence so far supports this hypothesis. The
problem with using this “hedonic housing market” approach to valuing residential visibility, however, is that
it is extremely difficult to isolate the value of better visibility in particular from the value of better air quality
in general. These studies tell us that people are willing to pay something extra to live in a house in a
neighborhood with cleaner air, but they do not tell us how much of that extra amount is the value people put
on the better visibility that results from the cleaner air versus, say, the value they put on the health-related
aspects of cleaner air. Because of this, while hedonic housing market studies have been used to value air
quality improvements, it is difficult to infer from them values for residential visibility improvements in
particular.
The most common method for measuring the value people place on environmental quality
improvements, such as both residential and recreational visibility improvements, is known as the contingent
valuation (CV) method. Although there are various different forms of the CV method, they all essentially rely
on the researcher asking the study subjects how much they would be willing to pay for a specified
environmental quality improvement.
CV studies of the value of visibility improvements typically present study subjects with a series of
photographs from actual historical records or model-generated WinHaze images of the same location at
different visibility levels and ask what their household would be willing to pay each year to have average
visibility improve from a specified level of lower visibility to a specified level of higher visibility. In a
residential visibility valuation study, the pictures would be of a particular urban area – for example, a city
skyline – shown with different levels of urban smog. In a recreational visibility study, the picture would be of
an important vista in a national park or scenic area of interest in the study. Researchers try to measure both
use and non-use value by including subjects from various locations, both close to and far from the national park
in question and by querying subjects about whether they ever have or ever intend to visit it.
The CV method has several often-noted advantages and disadvantages. The most notable disadvantage
is that what people say they would pay for a hypothetical improvement in a hypothetical situation may not be
what they would actually pay if the situation were real and they actually were required to pay. There are
several possible reasons for this. Subjects in a study may not fully understand what it is they are paying for.
They may not take the task entirely seriously if there is no consequence for false answers, or they may give
“strategic” answers if they believe that their answers might in some way ultimately affect policy. Or they may

simply find it very difficult to assess how much they would be willing to pay for something without being in
a situation in which they would actually have to pay it.
Sometimes it is difficult to separate out the value of one “good” from the value of other different but
related “goods.” In visibility valuation studies, some subjects may not be able to completely separate out the
value they place on better visibility from the value they place on other improvements, such as reduced health
risks, that they would also enjoy if air pollution were cleaned up. It is also possible that subjects’ responses
may to some extent reflect a general desire to “clean up the environment,” rather than the amount they would
be willing to pay specifically for better visibility.
Finally, people’s willingness to pay for visibility improvements may depend on what other
environmental quality improvements they may be considering paying for as well. Because people do not have
unlimited budgets, their willingness to pay for something should depend, to some extent, on how much money
they have at their disposal. There is some evidence from studies that people’s willingness to pay for an
environmental quality improvement does depend on whether it is the only one they are being asked to pay for
or one of several.
These issues and problems with CV studies have been the focus of substantial attention among
economists and others. Tremendous work has gone into refining CV survey designs to address the potential
pitfalls. While CV studies are still controversial because of their hypothetical nature, there has been much
progress in designing studies to minimize the pitfalls and to maximize the likelihood that subjects clearly
understand what it is they are being asked to value and that their answers will reflect their true willingness to
pay.
Despite the problems with CV studies, they have some important advantages. One important
advantage of the CV method is that it is attempting to measure the correct value (individuals’ willingness to
pay). Another is that it can be used to measure the value of anything. Perhaps most important, unlike any
market-based method, it can be used to measure non-use values. This is an important advantage because in
some cases non-use values may be substantial – perhaps more substantial, in aggregate, than use values. This
appears to be the case, for instance, for recreational visibility, as discussed below.
Contingent valuation is one form of a broader category of valuation methods collectively referred to
as stated preference methods. Although CV is the type of stated preference method most widely applied to
environmental quality valuation problems, another stated preference method, conjoint analysis, has very
recently been adapted to environmental quality valuation from the field of market research, where it has been
used for some time. Like CV, conjoint analysis relies on asking people to state their preferences, and so shares
some of the drawbacks of all methods based on hypothetical scenarios. Unlike CV, however, conjoint analysis
is typically designed in such a way that the value of a (market or nonmarket) good can be decomposed into the
values of its various attributes. A well designed conjoint analysis survey might therefore be able, at least in
theory, to distinguish the values of different “attributes” of an improvement in air quality, such as (1) the
consequent improvement in visibility and (2) the consequent reduction in health risks. Although there are to
date no published applications of conjoint analysis to valuing visibility improvements in particular, an
unpublished Master’s Thesis (Porras, 1999) and a related paper (Hill et al., 2000) both attempt to apply
conjoint analysis to value visibility in New Hampshire’s White Mountain National Forest.

6.2 The Value of Visibility Improvements at National Parks: Evidence from Studies
A number of studies have examined the value of visibility improvement (Exhibit 6-1). One type of
study that values improvements in visibility at national parks may be characterized as an “on-site use value
study.” In this type of study, researchers ask visitors to national parks what they would be willing to pay in
additional park entrance fees to have specified visibility improvements during their visits to those parks. This
type of study captures only direct use value – what people would be willing to pay for visibility improvements
that they can enjoy while visiting the park. These types of studies have been conducted at several locations
around the United States. MacFarland et al. (1983) elicited on-site visibility values from visitors at Grand
Canyon National Park and at Mesa Verde National Park; Rae (Rae, 1983) surveyed visitors at Great Smoky
Mountain National Park and at Mesa Verde National Park; and Rowe et al. (1980) studied the Navajo
Reservoir, near Farmington, New Mexico. This last study provides the only estimates of on-site use value for
visibility at a recreational area that is not a national park.
The on-site use value studies have generally estimated values on the order of a few dollars to several
dollars a day per household for noticeable improvements in visibility (ranging from about twenty miles increase
in visual range to well over one hundred miles increase). Because most households do not spend very many
days a year at national parks, however, these studies suggest that the total value of people being able to directly
enjoy visibility improvements at national parks is likely to be small relative to the value of improvements in
visibility at and around the communities where people live and spend much more of their time.
Exhibit 6-1 Economic Valuation Studies for Recreational and Residential Visibility
Study Location Valuation Method Mean WTP (1999$)
Recreational - On-Site
Rowe et al. (1980) Navajo Reservoir, NM Contingent Valuation $10.94 per visitor party day
MacFarland et al. (1983) Grand Canyon NP, AZ and Contingent Valuation GC - $3.82 per visitor party day
Mesa Verde NP, CO MV - $3.24 per visitor party day
Rae (1983) Mesa Verde NP, CO and Great Contingent Valuation MV - $11.48 per visitor party day
Smoky Mountain NP, TN GS - $5.81 per visitor party day
Recreational - Off-Site
Schulze et al. (1983) Grand Canyon NP, AZ Contingent Valuation $7.48 per month per household
Chestnut and Rowe National Parks in the Contingent Valuation CA - $73.93 per year per household
(1990b) Southwest, California, and the for parks in-region and $50.56 for
Southeast parks out-of-region
SW - $62.46 per year per household
for parks in-region and $50.56 for
parks out-of-region
SE - $76.06 per year per household
for parks in-region and $46.31 for
parks out-of-region
Residential
Tolley et al. (1986) Chicago, Atlanta, Boston, Contingent Valuation $10.75 - $68.52 per household per
Mobile, Washington, DC, year
Miami, Cincinnati
McClelland et al. (1991) Atlanta, Chicago Contingent Valuation $49.46 per household per year

Another type of study draws respondents from the general public, rather than from among visitors at
national parks. This “off-site” type of study investigates a broader cross-section of people. In this type of
study, respondents are identified in each of several specified locations, some of which are near the national
park(s) in question, and some of which are not. Subjects are asked whether they have ever visited the park in
question, and whether they ever intend to. Even people who say they have never visited a particular national
park and never plan to are asked what they would be willing to pay for specified improvements in visibility at
the park. This type of visibility valuation study allows researchers greater latitude in investigating both how
much value people place on being able to directly enjoy the good visibility at national parks and how much
value they place on just knowing it is there – for others, for themselves, if they want to visit at some time in
the future, or just in general. Studies using this type of approach have estimated both use and non-use values,
as well as the sum of the two – the total value of preserving or improving visibility at national parks.
Two studies of this second type have estimated the value of preserving or improving visibility at
national parks in different regions of the United States. The National Parks Visibility Values Study (Chestnut
and Rowe, 1990a; Chestnut et al., 1990b) focused on national parks in California, the Southwest, and the
Southeast. A second study (Schulze et al., 1983) covered national parks in the Southwest, including the Grand
Canyon. These studies provide compelling evidence that people want good visibility preserved at national
parks – even if they do not live anywhere near the parks and even if they have no plans to enjoy that good
visibility themselves. People’s values for visibility improvements at national parks in these studies far exceeded
the values for comparable improvements in visibility at national parks in the “on-site use value” studies.
Although people who had visited or planned to visit the parks were generally willing to pay more, even those
who reported that they had never visited the park(s) in question and never planned to generally said they were
willing to pay something to preserve the visibility in these parks. These results suggest that a substantial
component of the value of good visibility at national parks derives from knowing that it is there for others to
enjoy, knowing that it is there to enjoy if one ever goes to visit the park, and just knowing that it is there,
regardless – what economists call “non-use” value.
The studies suggest, moreover, that it is not only people who live close to national parks who care
about preserving the good visibility in them. Although households relatively close to the national parks in the
studies were generally willing to pay more than those far away, even “out of region” households were willing
to pay positive amounts per year to preserve the visibility at national parks. This is important because there
are a lot of households in the United States, and even small amounts per household add up to a large value.
For example, based on the responses from in-region and out-of-region respondents in the National Parks
Visibility Values Study (Chestnut et al., 1990a; Chestnut et al., 1990b), Chestnut and Dennis (Chestnut and
Dennis, 1997, p. 400) calculated that the average out-of-region household was willing to pay about $10
(1999$) a year for a twenty percent improvement in visibility at national parks in the Southeast. With over
110 million households in the United States, even if all households were “out of region,” that adds up to a value
of over $1 billion.
It is difficult to compare the results of these studies because different studies asked people to value
different amounts of visibility improvement at different national parks. We cannot say, for example, that
households were willing to pay from x dollars a year to y dollars a year for a twenty percent improvement in
visibility at national parks across the country because not all studies asked people to value a twenty percent
improvement in visibility. A hypothetical example using values from the National Parks Visibility Values
Study (Chestnut et al., 1990a; Chestnut et al., 1990b), however, gives us an idea of what these studies found.
Because the average household income in the sample of households in the National Parks Visibility
Values Study was higher than the national average ($41,000 for the sample versus a national average of
$32,000 in 1987), the study authors adjusted the average WTPs for changes in visual range in their sample to
better reflect what the average household in the United States would be willing to pay for these changes. Using

the estimated income elasticity of WTP (the percent change in WTP corresponding to a one percent increase
in income) across the three regions in their study (0.9), the average household WTP in the United States at the
time of the study was about 80 percent of the average household WTP in their sample. Using these income-
adjusted WTPs, households in the Southeast were, on average, willing to pay $84 (1999$) a year for a two
hundred percent visibility improvement (from 25 to 75 kilometers) in national parks in the Southeast.
Households in regions other than the Southeast were willing to pay, on average, $36 (1999$) a year for a one
hundred percent visibility improvement (from 25 to 50 kilometers) in national parks in the Southeast. All other
willingness to pay values from that study (for different visibility improvements in national parks in different
regions of the country) fell somewhere in between. If the “out-of-region” households in this study are
representative of all households in the United States, then a one hundred percent visibility improvement (from
25 to 50 kilometers) in national parks in the Southeast would be worth over $3.6 billion.
6.3 The Value of Visibility Improvements in Residential Areas: Evidence from Studies
People spend a lot more time in their communities than they do visiting parks and wilderness areas,
and studies that have estimated the value of good visibility in and around people’s residences suggest that it
is even more important than good visibility at national parks. Residential visibility valuation studies have been
carried out in several cities, including Atlanta, Chicago, Boston, Mobile, Washington, D.C., Cincinnati, Miami,
Los Angeles and San Francisco. Although different studies considered different visibility changes, the average
per household values placed on improvements in residential visibility were generally substantially higher than
those for comparable improvements in visibility in the national park visibility valuation studies. For example,
Tolley et al.(1986), who conducted residential visibility studies in several cities, found that the average
household in Chicago was willing to pay $389 (1999$) a year for a one hundred percent increase in visibility
(from 9 miles to 18 miles); in Washington, D.C. the average household value for a sixty-seven percent
improvement (from 15 to 25 miles) was $411 (1999$).
It is difficult to compare the results of different residential visibility valuation studies for the same
reason it was difficult to compare the results of different recreational visibility valuation studies – different
studies considered different visibility improvements. It may also be the case that people in different urban areas
value visibility differently -- even the same visibility improvement may have different values in different cities.
(This is also true for recreational visibility values -- they may vary from one national park to another, or people
in one region may value visibility in a given park differently from people in another region.) These studies,
however, collectively provide strong evidence that people value good visibility in their communities. The
smallest annual average household value reported was $22 (1999$) for a small visibility improvement of less
than fourteen percent (from 17.6 to 20 miles) in both Atlanta and Chicago (McClelland et al., 1991). Although
the per-household value is relatively small, there are many households in Atlanta and Chicago. In 1996, there
were an estimated 3,541,000 and 7,734,000 people in Atlanta and Chicago, respectively (Statistical Abstract
of the United States, 1999 Table 43). Assuming an average 2.68 persons per household, that would be over
$92 million.
The largest annual average household value reported in these studies was $523 (1999$) for a visibility
improvement of 133 percent (from 15 to 35 miles) in Washington, D.C. (Tolley et al., 1986). With 4,563,000
people in Washington, D.C. in 1996, (Statistical Abstract of the United States, 1999, Table 43), that sums to
over $890 million.

6.4 Applying the Information from Studies to Assess the Visibility Benefits of Reducing Air Pollution
The value of better visibility depends in part, of course, on how much better it will be. When
policymakers try to assess the value of improving visibility at national parks and in residential areas through
an air pollution policy or regulation, they must first estimate how much visibility will improve as a result of
that regulation or policy. Of course, any policy is likely to affect visibility in many residential areas and in
various national parks and wilderness areas, and the estimated changes in visibility are not necessarily going
to be exactly the same at all locations nor will they necessarily be the same as those considered in any of the
studies that have valued recreational or residential visibility improvements.
As noted above, each recreational or residential visibility valuation study of necessity considers specific
visibility changes at specific national parks or urban areas. Any particular willingness to pay value is for a
particular change in visibility at a particular location. To use the information from these studies to estimate
the value of improved visibility that would result from a policy or regulation, we must be able to generalize
from the specific visibility improvements and values reported in the studies. What can be said about visibility
changes that were not specifically valued in the studies? Or about the value of visibility changes at national
parks or urban areas that were not specifically included in the studies?
For example, the National Parks Visibility Values Study (Chestnut et al., 1990a; Chestnut et al.,
1990b) found that the average household in the Southeast would be willing to pay $68 (1999$) a year for a one
hundred percent increase in visibility, from 25 kilometers to 50 kilometers, in national parks in that part of the
country, and $84 (1999$) a year for a two hundred percent increase from 25 kilometers to 75 kilometers. What
does that imply about what the average Southeastern household would be willing to pay for a twenty percent
improvement in national parks in the Southeast? What about a fifty percent improvement? Or a three hundred
percent improvement?
McClelland et al. (1991) found that the average household in Atlanta would be willing to pay $325
(1999$) a year for an eighty-three percent improvement (from 12 to 22 miles) in residential visibility, and $486
(1999$) for a one hundred and sixty-seven percent improvement (from 12 to 32 miles). What would they be
willing to pay in Atlanta for a twenty percent improvement in residential visibility? Or a two hundred percent
improvement?
To estimate the value of the many different visibility improvements that would result in different
locations from an implemented air pollution policy or regulation, the evidence from these studies can be used
to estimate a general relationship between the amount of improvement in visibility and the average value
households place on that improvement. One plausible relationship that has been fitted to the available
recreational visibility values assumes that what matters is the percent change and not the absolute change in
visibility. This relationship says that household WTP for a change in visual range from v1 to v2 is a coefficient
times the natural logarithm of (v2/v1). The coefficient is estimated using the reported WTPs and the
corresponding (v1, v2) pairs of visual range in studies which value visibility improvements. For example,
using WTP estimates for visual range improvements in national parks in the Southeast from the National Parks
Visibility Values Study, Chestnut and Dennis (1997) reported estimated coefficients of 85 and 50 (based on
1994$) for in-region and out-of-region households, respectively. Using these fitted relationships, household
WTP for any percent improvement in visibility at national parks in the Southeast can be estimated. For
example, a ten percent improvement in visibility (i.e., v2/v1=1.1) would be worth about $8 a year per
household in 1994$ (=85*ln(1.1)) for households near the park, which is about $9 in 1999$; Using the out-of-
region coefficient of 50, a ten percent improvement in visibility would be worth about $6 (1999$) a year per
household, for those households far from the park. A twenty percent improvement in visibility would be worth
about 17$ and $10 to near and far households, respectively, and a one hundred percent improvement would
be worth about $66 and $39 a year. Using a general relationship based on the information in the recreational

visibility valuation studies, the average household value for any visibility improvement in a national park can
be estimated.
Similarly, a general relationship between visibility changes in residential areas and household values
for those changes can be estimated, based on the residential visibility valuation studies. If we assume the
relationship to be the same in all urban areas, we would use all the reported information from all the studies
to estimate a single relationship. If, alternatively, we allow that the relationship may vary from one urban area
to another, then we could estimate the relationship that best fits the information we have for each city. Chestnut
and Dennis (1997) did the latter, using the same general form of relationship that was used for national parks
(in which the percent change in visibility, rather than the absolute change, is what matters). Using these city-
specific functions, they calculated what a twenty percent improvement in visibility would be worth in each city.
There was a broad range. For example, a twenty percent improvement in residential visibility was worth as
little as $22 a year per household (in Cincinnati) and as much as $272 a year per household (in San Francisco)
.
While none of these per-household values is extremely large, there are many households in the United
States. If the values derived from the collective evidence of these studies are reasonably representative of
households throughout the United States, then with over 110 million households in the United States, we could
roughly estimate that a ten percent improvement in visibility in our nation’s national parks would be worth well
over $660 million a year, a twenty percent improvement would be worth well over $1 billion a year, and a one
hundred percent improvement would be worth well over $4 billion a year. A twenty percent improvement in
visibility in all residential areas in the United States could be worth anywhere from over $2 billion (if all
households were like the sample in Cincinnati) to almost $30 billion (if all households were like the sample in
San Francisco).
In practice, the value of visibility improvements due to a particular policy scenario has been a regular
component of regulatory impact analyses for years. In each analysis, the policy-specific relationship between
the amount of improvement in visibility and the average value households place on that improvement has been
estimated, resulting in a range of benefit results. Exhibit 6-2 displays various recreational and residential
visibility benefit totals that have been estimated for significant air quality policies over the last decade. While
the results show a number of cases where there are substantial benefits, the results are not easily comparable
since the geographic coverage and policy emphasis have varied from analysis to analysis. For example, the
Section 812 Retrospective and Prospective Analyses measured visibility changes over the entire nation, while
the Section 126 and NOx SIP Call Analyses only estimated visibility changes over a portion of the nation. In
another instance, the Haze Rule was specifically designed to improve visibility and has relatively large visibility
benefits, while the Tier 2 Rule is not focused on visibility improvement and the benefits are relatively small.
Exhibit A-3 presents these visibility benefits in tabular format with an explanation of the policy scenario
examined in each analysis.
Valuation methods have also changed over time. Many earlier estimates of visibility benefits, such
as the NOx SIP call (Abt Associates Inc., 1998), were calculated using visibility changes measured in terms
of “visual range.” The percentage change in visual range was multiplied by a given valuation coefficient to
estimate annual household willingness to pay. More recent estimates of visibility benefits, such as in the Tier
II analysis (Abt Associates Inc., 1999), use what economists refer to as utility functions. A utility function
is a mathematical model of a person’s (or a household’s) rational economic behavior. Utility functions capture
the basic characteristics of how people (or households) make economic tradeoffs between the possible goods
they might consume (buy), given that they have only a finite amount of money to spend in total. The model
expresses how much “utility,” or satisfaction a person gets from the set of goods he chooses to consume, and
is used to calculate what combination of goods will give him the greatest satisfaction, given the prices of the
goods and the person’s budget. Non-market “goods” like visibility improvements can be included among the

goods considered using a utility function. In this case, there is no “price” of the good, and the amount that the
person (household) would be willing to pay for the non-market good, given the total budget, the prices of the
market goods, and how much satisfaction is gotten from each, can be calculated. A utility function approach
therefore balances the tradeoff between people’s (households’) desire for improvements in visibility and their
desire for the other goods on which they normally spend their money. The appeal of this approach is that it is
designed to avoid benefits estimates that suggest people would spend an unrealistic amount of their income for
improved visibility.
As research continues, our knowledge about how much people value visibility improvements of
different amounts in different urban areas and at different national parks and wilderness areas will increase.
At any point in the research process, however, analysts must rely on the information that is available at that
time, and draw tentative conclusions about the general relationship between the amount of improvement in
visibility and the average value households place on that improvement, as discussed above. Analysts can then
use this general relationship to estimate the value of different visibility improvements at different urban areas
and in different national parks to help the policymaker estimate the value of the improved visibility that would
result from an air pollution policy or regulation.
Exhibit 6-2 Visibility Benefits from Different Policy Analyses
7000
6000
Benefits (millions 1999$)
5000
4000
3000
2000
1000
0
812 812 Haze Rule Ozone/PM Tier 2 NAPAP/Title Navajo NOx SIP Call Section 126
Prospective Retrospective NAAQS IV
[---------------------- National ----------------------] [----------------- Regional -----------------]
Policy Analysis
Recreational Benefits Residential Benefits

7. POWER PLANT EMISSION REDUCTIONS AND ASSOCIATED VISIBILITY BENEFITS
Legislators and regulators are considering a variety of methods to decrease sulfur dioxide (SO2) and
nitrogen oxide (NOx) emissions from electric generating units (EGUs) in the United States, because EGU
emissions contribute significantly to visibility impairment, poor health, acid rain, and other adverse effects.
In this hypothetical example, we consider the overall impact of EGU emissions. We estimate what visibility
impairment would be in 2007 with the status quo policy, and what it would be if all power plant emissions were
eliminated, or, in other words, if power plants were “shut off” and society generated electricity without
pollution emissions. The difference between these two scenarios equals the cumulative impact all power plant
emissions have on visibility impairment in the United States. Once this visibility impact is quantified, we then
estimate the monetary value for two broad categories of visibility benefits: (1) changes in “residential” visibility
– i.e., the visibility in and around the locations where people live; and (2) changes in “recreational” visibility
at Class I Areas – i.e., visibility at Class I national parks and wilderness areas. The total value of impaired
residential and recreational visibility quantified in this analysis can be thought of as the cost of foregone
visibility due to all EGU emissions.4
In order to estimate the monetary impact of EGU emissions, we need to define the baseline and control
scenarios for 2007. The status quo, or baseline scenario, estimates the visibility levels that would occur in
2007 with a continuation of current policy, and the control scenario models the case where power plant
emissions are eliminated. Developing the baseline scenario is the most difficult, because no one can predict
with perfect certainty what will occur in the future. In developing the baseline, we make the following
assumptions about the future regulatory climate:
! We assume implementation of the NOx SIP call rule, which requires certain States to reduce NOx
emissions that contribute to nonattainment of ozone standards in downwind States.
! We assume full implementation of Title IV of the 1990 Clean Air Act Amendments. In two
phases, Title IV will reduce SO2 emissions to 10 million tons below 1980 levels and reduce NOx
emissions to two million tons below 1980 levels (Clean Air Act, Title IV, Section 401).5 Phase I,
begun in 1995, affected mostly coal-burning utilities in the East and Midwest, and resulted in
substantial emissions reductions. Phase II, begun in 2000, expands the regulation’s coverage to over
2,000 EGUs nationwide to meet the overall goal of reducing SO2 and NOx emissions by 2010.
! We assume partial implementation of the Tier II rule, which reduces emissions from light-duty
vehicles and light-duty trucks and reduces sulfur in gasoline. The Clean Air Act Amendments of 1990
set so-called Tier I emission standards, beginning with the 1994 model year. The Clean Air Act
Amendments also require EPA to study whether further reductions in emissions from these vehicles
should be required. These are the Tier II standards, which would not take effect before the 2004 model
year. A phase-in would occur between 2004 and 2009, and gradually lead to nearly a full fleet of Tier
II compliant vehicles in 2030. Since hypothetical example is for the year 2007, we assume the partial
implementation that would occur by this date.
4
The total value of impaired recreational and residential visibility does not include the economic impacts of reductions in
tourism to national parks and other wilderness areas. This is because the money generated or lost due to increases or decreases in
visual air quality is a transfer; if not spent on goods and services in and around the park, the money will be spent on alternative goods
and services. There is no net effect to the U.S. economy.
5
The full text of the 1990 Clean Air Act Amendments is available at: http://www.epa.gov/oar/caa/contents.html#titleiv

! We assume no explicit adoption of emissions reductions from a climate change treaty, and we
assume no additional emission reductions due to the PM ambient air quality standard.
In developing the control scenario we use the same projected 2007 emission levels used in the baseline
scenario, and simply eliminated the contribution from power plants. So the difference between the two
scenarios is just the elimination of power plant emissions.
Given our estimate of emissions for the baseline and control scenarios, our first step is translating these
emissions into estimated visibility levels. To estimate visibility in 2007 for the baseline and control scenarios,
we use the CRDM Source-Receptor Matrix model (Pechan-Avanti Group, 1999). The Source-Receptor Matrix
estimates annual average pollution levels for the ambient pollutants that impair visibility, such as ammonium
sulfate, ammonium nitrate, elemental carbon and other aerosols.
The second step is to translate the annual average level of ambient pollutants into estimates of light
extinction. We quantify light extinction with four terms that represent the scattering and absorption of light
by gases and particles (Sisler, 1996, p. 5-1):
bext = bRay + bsp + bag + babs ,
where:
bext = total light extinction coefficient (Mm-1),
bRay = light extinction coefficient due to natural Rayleigh scatter (Mm-1),
bsp = light extinction coefficient due to scattering by particles (Mm-1),
bag = light extinction coefficient due to absorption by gases (Mm-1), and
babs = light extinction coefficient due to absorption by particles (Mm-1).
The light extinction coefficient is calculated by multiplying the concentration of an aerosol species by
its light-extinction efficiency, and summing over all species.
The term bRay refers to the natural Rayleigh scatter from air molecules, mainly nitrogen and oxygen.
Depending on altitude, this term has a value of 9 to 12 Mm-1 (inverse megameters). We use a value of 10 in
this analysis (Sisler, 1996, p. 5-1).
The term bsp can be broken into the various species of fine and coarse particles that scatter light.
Because fine particles are much more efficient at light scattering than coarse particles, several fine particle
species are specified, whereas coarse particles are kept as one category. Fine particles with significant light-
extinction efficiencies include sulfates, nitrates, organic carbon, elemental carbon (soot), and soil.
A complicating factor for sulfates, nitrates, and some organic compounds is that these aerosols are
hygroscopic, i.e., they absorb water, which greatly enhances their light-scattering abilities. The amount of
water absorbed is a function of the relative humidity. A relationship between the relative humidity and
scattering efficiency for ammonium sulfate aerosols has been developed, and is also applied to ammonium
nitrate aerosols (Sisler, 1996, p. 5-2).
The term bag represents absorption due to gases; NO2 is the only major light-absorbing gas in the lower
atmosphere. This component is assumed to be negligible since concentrations of NO2 are expected to be
negligible in rural areas (Sisler and Malm, 1994) which is generally applicable for Class I Areas. However,
this may be a poor assumption for locations close to significant NOx emission sources, such as power plants
or urban areas.

The final term of the light-extinction coefficient equation, babs, represents absorption of light by
elemental carbon. This term represents approximately 30 percent of the non-Rayleigh extinction budget (Sisler,
1996).
We convert the light extinction estimates for the baseline and control scenario into more familiar
estimates of visual range, using relationships expressed in Pitchford and Malm (1994, p. 1051). We then enter
these visual range estimates into a utility function, which balances the tradeoff between a household’s income
and the amount they value improvements in visibility. Our approach is based on the preference calibration
method developed by Smith et al. (1999), and we discuss it in more detail in Appendix B.
The results of the analysis estimate that status-quo EGU emissions in the year 2007 will cost the
United States $3.4 billion (1999$) in foregone residential visibility benefits and $4.3 billion (1999$) in foregone
recreational visibility benefits. In total, EGU emissions will cost the United States $7.7 billion in foregone total
visibility benefits. Of course, the full extent of these foregone benefits will never be realized because EGU
emissions will never be completely eliminated. These foregone visibility benefit totals, however, reflect the
potential for substantial visibility-related benefits if EGUs were to reduce emissions beyond current levels.
Exhibit 7-1 presents results at the state level for foregone residential visibility benefits. Exhibits 7-2
and 7-3 present foregone recreational visibility benefits at the park and state level, respectively. State level
recreational visibility totals are simply the sum of foregone benefits estimated for parks that fall within a given
state. Therefore, some states may not have economic impacts attributed to them, even though the population
within the state contributed to a park’s recreational visibility valuation.
To give a rough idea of the impact EGU emissions have on a National Park with significant visibility
problems, we used WinHaze to generate photographs of a poor visibility day and the worst visibility day for
the status-quo and no-EGU scenarios in Great Smoky Mountains National Park (Exhibit 7-4). We do not have
the distribution of visibility estimates for 2007 since the S-R Matrix only models the mean visibility level.
However, we do have 1996 relative humidity data and speciated particulate matter data from NPS-CIRA
(2000a). We used these data to reconstruct daily extinction levels at Great Smoky Mountains National Park.
To estimate the 2007 status-quo poor visibility day, we averaged extinction levels for the haziest 20
percent of days in 1996, and then corrected for the slightly lower average visibility levels in 2007 relative to
1996, by multiplying this estimate by the ratio of the mean S-R matrix visibility estimates for 2007 and 1996.6
We then considered the contribution of EGUs to sulfate levels, and chose 67 percent, based on the fact that
EGUs contribute roughly 67 percent of national SO2 emissions (Exhibit 2-2). To estimate the 2007 no-EGU
poor visibility day, we subtracted 67 percent of the average sulfate contribution on poor days (173 Mm-1) in
1996 from the total extinction levels on poor days (215 Mm-1) in 1996, and then multiplied this estimate by the
ratio of the modeled mean visibility estimates for 2007 and 1996 (Exhibit 7-4a).
To estimate visibility levels on the worst day for the 2007 status-quo and no-EGU, we followed an
analogous procedure: we subtracted 67 percent of the sulfate contribution on the worst day (289 Mm-1) in 1996
from the total extinction on the worst day (289 Mm-1) in 1996, and multiplied this by the ratio of the modeled
mean visibility estimates for 2007 and 1996 (Exhibit 7-4b). We also considered a partial-EGU scenario, where
we reduced EGU emissions by 75 percent, rather than eliminating them completely. This translates into
subtracting 50 percent of the sulfate contribution on poor and worst days (Exhibit 7-4c and 7-4d).
6
Great Smoky Mountains National Park lies in Blount County, TN (FIPS=47009). The mean 1996 extinction level is
95.25 Mm-1 and the mean 2007 extinction level is somewhat lower at 88.78 Mm-1. We define a “poor” day as the average extinction
values from the 80th percentile to the worst day (100th percentile).

Exhibit 7-1 Residential (Urban) Visibility Benefits
State Monetary Benefits State Monetary Benefits

(millions 1999$) (millions 1999$)
Alabama $74 Nebraska $149
Arizona $35 Nevada $5
Arkansas $8 New Hampshire $19
California $61 New Jersey $25
Colorado $14 New Mexico $139
Connecticut $56 New York $4
Delaware $9 North Carolina $4
District of Columbia $17 North Dakota $270
Florida $208 Ohio $183
Georgia $123 Oklahoma $33
Idaho $46 Oregon $8
Illinois $3 Pennsylvania $230
Indiana $190 Rhode Island $16
Iowa $114 South Carolina $66
Kansas $28 South Dakota $8
Kentucky $71 Tennessee $95
Louisiana $37 Texas $170
Maine $93 Utah $6
Maryland $124 Vermont $132
Massachusetts $17 Virginia $10
Michigan $167 Washington $27
Minnesota $47 West Virginia $81
Mississippi $93 Wisconsin $29
Missouri $38 Wyoming $2
Montana $3 TOTAL $3,385

Exhibit 7-2 Recreational Visibility Benefits
Park Name State Park Benefits Park Name State Park Benefits
Typea (millions Type (millions
1999$) 1999$)
Sipsey AL W $2.1 Hercules-Glades MO W $2.3
Caney Creek AR W $1.4 Anaconda-Pintlar MT W $0.7
Upper Buffalo AR W $0.2 Bob Marshall MT W $3.1
Chiricahua AZ W $1.2 Cabinet Mountains MT W $0.5
Galiuro AZ W $0.2 Gates of the Mountain MT W $0.1
Grand Canyon AZ NP $278.9 Glacier MT NP $64.5
Mazatzal AZ W $0.3 Mission Mountain MT W $0.2
Mount Baldy AZ W $0.3 Scapegoat MT W $0.5
Petrified Forest AZ NP $7.3 Joyce Kilmer-Slickrock NC W $3.6
Pine Mountain AZ W $0.1 Linville Gorge NC W $1.2
Saguaro AZ W $2.9 Shining Rock NC W $2.0
Sierra Ancha AZ W $0.1 Theodore Roosevelt ND NP $16.8
Superstition AZ W $1.9 Great Gulf NH W $4.1
Sycamore Canyon AZ W $0.3 Presidential Range-Dry River NH W $1.9
Agua Tibia CA W $0.0 Bandelier NM W $5.8
Caribou CA W $0.1 Carlsbad Caverns NM NP $8.5
Cucamonga CA W $0.0 Gila NM W $0.1
Death Valley CA NP $9.9 Pecos NM W $0.2
Desolation CA W $1.9 San Pedro Parks NM W $0.8
Dome Land CA W $0.0 Wheeler Peak NM W $0.1
Emigrant CA W $0.5 White Mountain NM W $5.1
Hoover CA W $0.6 Jarbidge NV W $0.1
John Muir CA W $0.1 Crater Lake OR NP $4.3
Joshua Tree CA W $9.7 Diamond Peak OR W $0.1
Kaiser CA W $3.5 Eagle Cap OR W $2.6
Kings Canyon CA NP $23.7 Kalmiopsis OR W $0.1
Lassen Volcanic CA NP $5.1 Mount Hood OR W $1.1
Lava Beds CA W $0.6 Mount Jefferson OR W $0.4
Marble Mountain CA W $0.9 Mount Washington OR W $0.0
Mokelumne CA W $0.3 Mountain Lakes OR W $0.0
Pinnacles CA W $0.9 Strawberry Mountain OR W $0.8
Point Reyes CA W $17.1 Three Sisters OR W $0.6
Redwood CA NP $4.7 Badlands SD W $67.0
San Gabriel CA W $0.3 Wind Cave SD NP $28.0
San Gorgino CA W $1.7 Cohotta TN W $9.1
San Jacinto CA W $0.2 Great Smoky Mountains TN NP $1,830.1
San Rafael CA W $0.1 Big Bend TX NP $42.0

Exhibit 7-2 Recreational Visibility Benefits (cont.)
Park Name State Park Benefits Park Name State Park Benefits
Typea (millions Type (millions
1999$) 1999$)
Sequoia CA NP $59.4 Guadalupe Mountains TX NP $2.4
Southarner CA W $0.1 Arches UT NP $16.8
Thousand Lakes CA W $0.0 Bryce Canyon UT NP $36.7
Ventana CA W $0.6 Canyonlands UT NP $18.2
Yolla-Bolly-Middle-Eel CA W $0.1 Capitol Reef UT NP $4.6
Yosemite CA NP $76.6 Zion UT NP $66.0
Black Canyon of the Gun CO NP $5.5 James River Face VA W $0.6
Eagles Nest CO W $2.8 Shenandoah VA NP $127.7
Flat Tops CO W $3.0 Lye Brook VT W $0.6
Great Sand Dunes CO W $9.1 Alpine Lakes WA W $5.1
La Garita CO W $0.5 Glacier Peak WA W $3.6
Maroon Bells-Snowmass CO W $3.3 Goat Rocks WA W $1.0
Mesa Verde CO NP $23.3 Mount Adams WA W $2.1
Mount Zirkel CO W $1.6 Mount Rainier WA NP $148.4
Rawah CO W $0.4 North Cascades WA NP $1.8
Rocky Mountain CO NP $94.3 Olympic WA NP $99.5
Weminuche CO W $5.8 Pasayten WA W $2.2
West Elk CO W $3.0 Dolly Sods WV W $3.0
Everglades FL NP $58.6 Otter Creek WV W $1.4
Craters of the Moon ID W $2.2 Bridger WY W $4.3
Sawtooth ID W $1.2 Fitzpatrick WY W $0.8
Selway-Bitterroot ID W $2.3 Grand Teton WY NP $60.9
Mammoth Cave KY NP $147.8 North Absaroka WY W $0.2
Acadia ME NP $327.8 Teton WY W $2.3
Isle Royale MI NP $20.6 Washakie WY W $1.0
Boundary Waters Canoe MN W $81.8 Yellowstone WY NP $208.2
Voyageurs MN NP $15.4 TOTAL $4,253.9
a
Benefits are calculated for two park types, National Parks (NP) and Wilderness Areas (W).

Exhibit 7-3 State-Level Recreational Visibility Benefits
State Benefits (millions 1999$) State Benefits (millions 1999$)
Alabama $2.1 North Dakota $16.8
Arkansas $1.6 New Hampshire $6.0
Arizona $293.3 New Mexico $20.6
California $218.7 Nevada $0.1
Colorado $152.6 Oregon $10.0
Florida $58.6 South Dakota $95.0
Idaho $5.8 Tennessee $1,839.2
Kentucky $147.8 Texas $44.4
Maine $327.8 Utah $142.3
Michigan $20.6 Virginia $128.3
Minnesota $97.2 Vermont $0.6
Missouri $2.3 Washington $263.8
Montana $69.4 West Virginia $4.4
North Carolina $6.8 Wyoming $277.7

Exhibit 7-4 WinHaze Split-Images at Great Smoky Mountains National Park for Status Quo, No-EGU, and Partial-EGU Scenarios
(a) Poor Visibility Day: (b) Worst Visibility Day:

Status Quo (VR=20km) and No-EGU (VR=43km) Status Quo (VR=13km) and No-EGU (VR=31km)
(c) Poor Visibility Day: (d) Worst Visibility Day:

Status Quo (VR=20km) and Partial-EGU (VR=33km) Status Quo (VR=13km) and Partial-EGU (VR=23 km)

8. CONCLUSION
Power plants make a substantial contribution to visibility impairment, which has substantial economic
and aesthetic consequences. People value natural, unobstructed vistas around their home and, most especially,
at parks and wilderness areas. In fact, people believe that the natural features of parks and wilderness areas,
especially “clean, clear air,” are the most important features of these areas (NPS 1988). People are therefore
willing to pay substantial sums for improved visibility. And when visibility is poor at parks and wilderness
areas, they are less likely to visit. In turn, this can cause a significant drop in tourism-related economic activity
around parks.
Local economies benefit from increased tourism when visibility is improved. We estimated that if
visibility improvements increased visitation to one of the parks examined in this report by 25 percent, the
potential annual benefit to a park and its local communities could range from $13 million and 390 new jobs
at Indiana Dunes National Lakeshore to $320 million and 4,188 new jobs at Great Smoky Mountain National
Park. National parks and concessionaires also stand to benefit from an increase in visitation. An 25 percent
increase in visitors could yield approximately $30 million in increased fee collection and $160 million in
additional concession sales.
The benefits associated with the non-market value of visibility, people’s willingness to pay for better
visibility, can be even more substantial. A number of policy analyses examining the potential costs and benefits
of air pollution abatement have estimated the non-market value of both recreational and residential visibility
improvements to visibility “consumers.” Some policies target air pollution controls that will have a greater
impact on visual air quality, so the estimated benefits range substantially. Nevertheless, even for those policies
where visibility improvement is only a secondary benefit, the magnitude of the benefits can still be substantial.
We found that for national policy analyses, recreational visibility benefits ranged from $385 million to $5,983
million and residential visibility benefits ranged from $149 million to $4,311 million. Total visibility benefits,
the sum of residential and recreational visibility benefits, ranged from $963 million to $9,889 million for the
same policy analyses.
To demonstrate the extent to which power plant emissions impact our ability to enjoy visibility, we
modeled the change in visibility due to an hypothetical elimination of power plant emissions in 2007 and valued
that change in visibility applying the same methodology used in the most recent federally-mandated policy
analyses. The analysis estimated that if power plant emissions were completely eliminated, residential visibility
benefits for such a change in visibility would be $3.4 billion and recreational visibility benefits would be $4.3
billion. In total, power plant emissions were estimated to cost the United States approximately $8 billion in
foregone visibility benefits. Of course, all emissions from power plants will not be eliminated by the year 2007.
However, even if a fraction of the current emissions from power plants were controlled, the monetary benefits
would still be substantial.
The technology exists to reduce emissions from power plants. For instance, the Department of
Energy’s Clean Coal Technology Demonstration Program has had many technological successes in controlling
pollution emissions (U.S. Department of Energy, 1999; U.S. EPA, 1997). They currently have 40 projects in
18 states, with over half the projects having already reached successful completion. For plants already in
operation, effective, low-cost technologies are available to retrofit industrial processes and reduce the amount
of visibility-impairing pollutants present in their emissions. For the construction of new power plants, a
number of new technologies exist that effectively remove pollutants, like SO2, from their emissions. SO2
emission control technologies, both to retrofit existing power plants and for the construction of new plants,
should move into the utility marketplace and provide cost-effective regulatory compliance. Coupling emission
control technologies with strengthened legislation and oversight of the utility industry, the United States can
Abt Associates Inc. 58

move towards significant improvements in visibility throughout the nation, and the aesthetic and local
economic benefits to be gained with such improvements could be substantial.

9. REFERENCES
Abt Associates Inc. 1998. Selected Health and Welfare Benefits Methods for the NOx SIP Call RIA.
Prepared for U.S. EPA, Office of Air Quality Planning and Standards, under contract no. 68-D-98-
001. Research Triangle Park, NC. September.
Abt Associates Inc. 1999. Tier 2 Proposed Rule: Air Quality Estimation, Selected Health and Welfare
Benefits Methods, and Benefit Analysis Results. Prepared for U.S. EPA, Office of Air Quality
Planning and Standards. Research Triangle Park, NC. April.
Air Resource Specialists Inc. 1997. Grand Canyon National Park Arizona: IMPROVE Monitoring Network
Permanent Photographic Archive 1983-1996, Slide Spectrum. September. Prepared for the
IMPROVE program, NPS Contract #CX-1270-96-006. Fort Collins, CO.
Air Resource Specialists Inc. 1998. WinHaze. Dated May 29. http://www.air-resource.com.
Air Resource Specialists Inc. 1999. Shenandoah National Park Virginia: IMPROVE Monitoring Network
Permanent Photographic Archive 1991-1995, Slide Spectrum. Prepared for the IMPROVE program,
NPS Contract #CX-1270-96-006. Fort Collins, CO.
Air Resource Specialists Inc. 2000a. Acadia National Park Maine: IMPROVE Monitoring Network
Permanent Photographic Archive 1983-1995, Slide Spectrum. Prepared for the IMPROVE program,
NPS Contract #CX-1270-96-006. Fort Collins, CO.
Air Resource Specialists Inc. 2000b. Great Smoky Mountains National Park Tennessee: IMPROVE
Monitoring Network Permanent Photographic Archive 1984-1995, Slide Spectrum. Prepared for the
IMPROVE program, NPS Contract #CX-1270-96-006. Fort Collins, CO.
Air Resource Specialists Inc. 2000c. Yosemite National Park California: IMPROVE Monitoring Network
Permanent Photographic Archive 1984-1995, Slide Spectrum, Volume I. Prepared for the IMPROVE
program, NPS Contract #CX-1270-96-006. Fort Collins, CO.
Bell, P., W. Malm and G. McGlothin. 1985. A Utility Model for Assessing the Impact of Impaired Visibility
on Visitor Enjoyment of the Grand Canyon. In: Assessment of Visibility Impairment on Visitor
Enjoyment and Utilization of Park Resources. Cooperative Institute for Research in the Atmosphere,
Colorado State University: Fort Collins, Colorado.
Brothers, G. and R.J.C. Chen. 1997. 1995-1996 Economic Impact of Travel to the Blue Ridge Parkway:
Virginia and North Carolina. The Coalition for the Blue Ridge Parkway and th National Park Service.
Asheville, NC and Roanoke, VA.
Brown, P., G. Haas and M. Manfredo. 1977. Identifying Resource Attributes Providing Opportunities for
Dispersed Recreation: Final Report. Rocky Mountain Forest and Range Experiment Station. Fort
Collins, CO.
Chestnut, L.G. 1997. Draft Memorandum: Methodology for Estimating Values for Changes in Visibility at
National Parks. April 15.

Chestnut, L.G. and R.L. Dennis. 1997. Economic benefits of improvements in visibility: Acid Rain Provisions
of the 1990 Clean Air Act Amendments. Journal of the Air & Waste Management Association. Vol.
47(3): 395-402.
Chestnut, L.G. and R.D. Rowe. 1990a. New National Park Visibility Value Estimates. Mathai, C.V., Ed.
Visibility and Fine Particles: An Air and Waste Management Association / EPA Specialty Conference
at Estes Park, CO, October 1989. Air and Waste Management Association. Pittsburgh, PA.
Chestnut, L.G. and R.D. Rowe. 1990b. Preservation Values for Visibility Protection at the National Parks:
Draft Final Report. Prepared for U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Economic Analysis Branch. Research Triangle, NC. February, 16.
Day, D.E., W.C. Malm and S.M. Kreidenweis. 2000. Aerosol light scattering measurements as a function
of relative humidity. J Air Waste Manag Assoc. Vol. 50(5): 710-716.
Eatough, D.J., R.J. Farber and J.G. Watson. 2000. Second generation chemical mass balance source
apportionment of sulfur oxides and sulfate at the Grand Canyon during the project MOHAVE summer
intensive. J Air Waste Manag Assoc. Vol. 50(5): 759-774.
Hill, B.L., W. Harper, J.M. Halstead, T.H. Stevens, I.T. Porras and K.D. Kimball. 2000. Visitor Perceptions
and Valuation of Visibility in the Great Gulf Wilderness, New Hampshire. In: Proceedings:
Wilderness Science in a Time of Change. D.N. Cole and S.F. McCool, Ed. U.S. Department of
Agriculture, Forest Service, Rocky Mountain Research Station: Ogden, UT. Vol: RMRS-P-000
Kask, S. 1998. Valuing Scenic Beauty: A Pilot Study on the Blue Ridge Parkway (Draft). Western Carolina
University. Cullowhee, NC.
Latimer, D.A. 1990. A Conceptual Plan for a National Policy to Foster Long-Term Visibility Improvement.
Mathai, C.V., Ed. Visibility and Fine Particles: An Air and Waste Management Association / EPA
Specialty Conference at Estes Park, CO, October 1989. Air and Wast Management Association.
Estes Park, CO. p. 50-60.
MacFarland, K.K., W. Malm and J. Molenar. 1983. An Examination of Methodologies and Social Indicators
for Assessing the Value of Visibility. In: Managing Air Quality and Scenic Resources at National
Parks and Wilderness Areas. Chestnut, R.D.R.a.L.G., Ed. Westview Press: Boulder, Colorado.
Malm, W.C. 1999. Introduction to Visibility. Cooperative Institute for Research in the Atmosphere (CIRA).
Fort Collins, CO. Cooperative Agreement. May.
Malm, W.C., J. Sisler, D. Huffman, R. Eldred and T. Cahill. 1994. Spatial and Seasonal Trends in Particle
Concentration and Optical Extinction in the United States. Journal of Geophysical Research. Vol.
99(D1): 1247-1370.
McClelland, G., W. Schulze, D. Waldman, J. Irwin, D. Schenk, T. Stewart, L. Deck and M. Thayer. 1991.
Valuing Eastern Visibility: A Field Test of the Contingent Valuation Method. Prepared for U.S.
Environmental Protection Agency, Office of Policy, Planning and Evaluation. June.
NAPAP. 1990. Acidic Deposition: State of Science and Technology, Report 18: Response of Vegetation to
Atmospheric Deposition and Air Pollution. National Acid Precipitation Assessment Program, Office
of Director. Washington, DC.

National Park Service. 1988. Air Quality in the National Parks. National Park Service, Air Quality Division.
Denver, Colorado. Natural Resources Programs, Natural Resources Report 88-1.
National Park Service. 1997. Natural Resource Year in Review -1996. U.S. Department of the Interior,
National Park Service. Publication D-1182. May.
National Park Service. 1998. The Money Generation Model. Executable dated March 4. National Park
Service, Public Use Statistics Office. http://www.nps.gov/planning/mgm.
National Park Service. 2000. Recreation Visits 1951-1998. May. National Park Service, Public Use
Statistics Office. http://www2.nature.nps.gov/stats/.
National Park Service (NPS) and the Cooperative Institute for Research in the Atmosphere (CIRA). 2000a.
IMPROVE Data Available at FTP Site. Downloaded May 28. National Park Service, Air Resources
Division - Cooperative Institute for Research in the Atmosphere, Colorado State University. Fort
Collins, CO. http://alta_vista.cira.colostate.edu/.
National Park Service (NPS) and the Cooperative Institute for Research in the Atmosphere (CIRA). 2000b.
Visibility-Related Figures Available at FTP Site. Downloaded June 25. National Park Service, Air
Resources Division - Cooperative Institute for Research in the Atmosphere, Colorado State University.
Fort Collins, CO. http://alta_vista.cira.colostate.edu/.
National Research Council. 1993. Protecting visibility in national parks and wilderness areas. National
Academy Press: Washington, DC.
Peacock, B., C. Killingsworth and B. Simon. 1998. State and National Economic Impacts Associated with
Travel Related Expenditures by Recreational Visitors to Lands Managed by the U.S. Department of
the Interior. U.S. Department of the Interior. January 16.
Pechan-Avanti Group. 1999. Emissions and Air Quality Impacts of Final Motor Vehicle Tier 2 and Fuel
Sulfur Standards: Draft Report. Prepared for U.S. EPA, Office of Air Quality Planning and
Standards, Innovative Strategies and Economics Group. Springfield, VA. Pechan Report No.
99.10.001/9004.223. October.
Pitchford, M.L. and W.C. Malm. 1994. Development and Applications of a Standard Visual Index.
Atmospheric Environment. Vol. 28(5): 1049-1054.
Pope, C.A., M.J. Thun, M.M. Namboodiri, D.W. Dockery, J.S. Evans, F.E. Speizer and C.W. Heath. 1995.
Particulate air pollution as a predictor of mortality in a prospective study of U.S. adults. Am J Respir
Crit Care Med. Vol. 151(3): 669-674.
Porras, I.T. 1999. Off-Site Value of Visibility: New Hampshire's White Mountains Case Study. Unpublished
Master's Thesis, University of Massachusetts.
Rae, D.A. 1983. The Value to Visitors of Improving Visibility at Mesa Verde and Great Smoky National
Parks. In: Managing Air Quality and Scenic Resources at National Parks and Wilderness Areas.
Chestnut, R.D.R.a.L.G., Ed. Westview Press: Boulder, Colorado.
Rowe, R.D., R. d'Arge and D. Brookshire. 1980. An experiment on the economic value of visibility. J.
Environmental Economics and Management. Vol. (March): 1-19.

Savig, K. 2000. Air Resource Specialists, Inc., Fort Collins, Colorado. Email with photos sent to Kenneth
Davidson. June 7.
Schulze, W.D., D.S. Brookshire and E.G. Waltham. 1983. The benefits of preserving visibility in the national
parklands of the Southwest. Natural Resources Journal. Vol. 23: 149-173.
Scott, D.W. and D.H. Stonefield. 1990. The Environmental Protection Agency's Visibility Protection
Program: Ten Years of Regulatory Development. Mathai, C.V., Ed. Visibility and Fine Particles: An
Air and Waste Management Association / EPA Specialty Conference at Estes Park, CO, October
1989. Air and Wast Management Association. Estes Park, CO. p. 2-10.
Sisler, J.F. 1996. Spatial and Seasonal Patterns and Long Term Variability of the Composition of the Haze
in the United States: An Analysis of Data from the IMPROVE Network. Colorado State University,
Cooperative Institute for Research in the Atmosphere. Fort Collins, CO. July.
Sisler, J.F. and W. Malm. 1994. The Relative Importance of Soluble Aerosols to Spatial and Seasonal Trends
of Impaired Visibility in the United States. Atmospheric Environment. Vol. 28(5): 851-862.
Sisler, J.F. and W.C. Malm. 2000. Interpretation of trends of PM2.5 and reconstructed visibility from the
IMPROVE network. J Air Waste Manag Assoc. Vol. 50(5): 775-789.
Smith, V.K., G. Van Houten and S. Pattanayak. 1999. Benefits Transfer as Preference Calibration.
Resources for the Future. Washington, DC. Working Paper 99-36. May.
Tolley, G.S. and et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks. Prepared for
U.S. Environmental Protection Agency. January.
U.S. Department of Energy. 1999. Clean Coal Technology: The Investment Pays Off. U.S. Department of
Energy. November.
U.S. Department of the Interior. 1988. Recreational Fee Demonstration Program, Volume I – Overview and
Summary. Progress Report to Congress. January 31.
U.S. EPA. 1996. Air Quality Criteria for Particulate Matter. Volume II. U.S. EPA, Office of Research and
Development. Washington, DC. EPA-/600/P-95/001bF. April.
U.S. EPA. 1997. Profile of the Fossil Fuel Electric Power Generation Industry. Office of Compliance, Office
of Enforcement and Compliance Assurance, U.S. Environmental Protection Agency. Washington, DC.
EPA/310-R-97-007. September.
U.S. EPA. 1998. National Air Pollutant Emissions Trends, 1997. Office of Air Quality Planning and
Standards. Research Triangle Park, NC. EPA-454/R-98-016. December.
U.S. EPA. 2000. National Air Pollutant Emissions Trends, 1900-1998. Office of Air Quality Planning and
Standards. Research Triangle Park, NC. EPA-454/R-00-002. March.
Wade, T. 2000. Park Service, Public Use Statistics Office. Personal communication with Ken Davidson of
Abt Associates, Regarding Money Generation Model Input Assumptions. May 10.

Walsh, R.G. and J. Loomis. 1982. Wilderness Resource Economics: Recreation Use and Preservation Values.
Department of Economics, Colorado State University. Fort Collins, CO.

Exhibit A-1 Percentage Contribution of Constituents to Visibility Impairment on Good, Mid-Range, and Poor Visibility Days
Sulfate Nitrate Organic Carbon Light Absorbing Coarse Matter and Fine
Carbon Soil
Area Region State Good Mid Poor Good Mid Poor Good Mid Poor Good Mid Poor Good Mid Poor
Sipsey Wilderness East AL 57 67 76 13 5 2 9 12 11 5 6 4 4 3 2
Washington East DC 41 50 68 10 13 2 14 13 14 13 11 9 5 3 2
Chassahowitzka East FL 53 66 66 6 5 3 12 10 13 7 5 5 6 6 6
Okefenokee East GA 62 64 74 5 4 2 12 13 12 5 5 4 4 4 4
Mammoth Cave East KY 56 67 78 11 7 2 9 10 10 5 5 3 4 3 3
Acadia East ME 37 48 69 4 6 4 12 12 10 4 5 4 8 6 4
Shining Rock East NC 41 67 79 4 4 1 10 10 10 5 4 3 5 3 3
Brigantine East NJ 45 52 73 7 9 2 10 10 11 6 6 5 11 10 4
Cape Romain East SC 59 67 76 5 6 2 10 10 10 5 4 4 7 4 4
Great Smoky East TN 44 63 76 5 3 1 15 13 11 7 6 4 8 3 3
Shenandoah East VA 46 61 81 12 9 1 9 10 9 6 5 3 3 3 2
Lye Brook East VT 34 48 74 7 13 4 12 12 9 4 5 4 6 3 2
Dolly Sods East WV 49 57 84 7 9 1 12 13 7 6 6 3 4 3 1
Upper Buffalo Midwest AR 37 58 69 15 9 3 11 12 13 6 5 4 5 4 4
Boundary Waters Midwest MN 29 35 42 4 15 18 12 15 20 3 5 5 6 6 4
Badlands Midwest SD 20 36 44 6 10 9 9 12 18 4 5 5 11 7 9
Big Bend Midwest TX 26 34 38 2 3 2 13 15 26 6 4 5 13 17 15
Abt Associates Inc. A-1 August 2000

Exhibit A-1 Percentage Contribution of Particle Constituents to Visibility Impairment On Good, Median, and Bad Visibility Days (cont.)
Sulfate Nitrate Organic Carbon Light Absorbing Coarse Matter and Fine
Carbon Soil
Area Region State Good Mid Poor Good Mid Poor Good Mid Poor Good Mid Poor Good Mid Poor
Chiricahua West AZ 17 28 36 3 3 2 14 17 21 6 6 6 12 12 13
Grand Canyon West AZ 15 21 28 3 6 4 7 15 19 4 7 7 18 13 16
Redwood West CA 39 48 54 9 14 18 7 6 7 2 1 2 6 6 6
Yosemite West CA 9 18 29 3 11 7 12 23 27 3 7 7 6 7 9
San Gorgonio West CA 11 17 20 16 36 37 14 15 16 9 7 9 8 7 6
Rocky Mountain West CO 10 23 25 2 5 6 14 18 24 5 6 7 8 8 10
Glacier West MT 21 36 15 5 5 4 17 20 39 7 7 11 7 10 17
Crater Lake West OR 7 15 20 1 3 3 10 17 30 8 7 8 10 10 13
Bryce Canyon West UT 14 21 34 4 9 4 10 14 17 4 4 5 9 10 9
Mount Rainier West WA 26 49 58 3 8 6 11 15 16 4 6 7 8 4 3
Yellowstone West WY 12 18 18 4 4 3 8 23 31 3 4 6 3 6 10
Source: National Park Service (2000a). Note that a “good” visibility day is defined as the average of the lowest 20 percent of visibility days over the course of the year. A “mid” day is
the average of the middle 20 percent of days between the 40th and 60th percentiles, and a “poor” day is average of the worst 20 percent of days.

Exhibit A-2 Annual Slope Estimates with Probabilities for Rejection for the Average of the Worst,
Median, and Best Visibility Days
Worst 20% Median 20% Best 20%

Site Annual Probability Annual Probability Annual Probability
Slope Slope Slope
Acadia NP -0.07 0.24 -0.15 0.02 -0.24 0.01
Badlands NP 0.07 0.06 -0.11 0.01 -0.10 0.13
Bandelier NM -0.03 0.38 -0.18 0.00 -0.17 0.06
Big Bend NP 0.07 0.13 -0.02 0.18 -0.09 0.13
Bryce Canyon NP -0.02 0.38 -0.11 0.09 -0.13 0.01
Bridger W -0.12 0.31 -0.09 0.13 -0.11 0.09
Canyonlands NP -0.16 0.06 -0.12 0.04 -0.15 0.01
Chiricahua NM 0.00 0.54 -0.06 0.06 -0.01 0.46
Crater Lake NP 0.07 0.31 -0.04 0.46 -0.17 0.01
Denali NP -0.14 0.09 -0.16 0.09 -0.16 0.04
Glacier NP -0.09 0.09 0.12 0.02 -0.11 0.24
Grand Canyon NP -0.07 0.24 -0.11 0.13 -0.10 0.06
Great Sand Dunes NM -0.18 0.18 -0.20 0.02 -0.27 0.01
Great Smoky Mountains NP 0.09 0.18 -0.06 0.38 -0.06 0.31
Guadalupe Mountains NP -0.09 0.13 -0.06 0.24 -0.08 0.24
Lassen Volcanic NP -0.04 0.38 -0.15 0.01 -0.17 0.02
Mesa Verde NP 0.00 0.54 -0.06 0.31 -0.15 0.06
Mount Rainier NP -0.16 0.06 -0.03 0.46 -0.18 0.13
Petrified Forest NP -0.15 0.06 -0.18 0.02 -0.18 0.01
Pinnacles NM -0.21 0.01 -0.21 0.00 -0.21 0.04
Point Reyes NS -0.22 0.09 -0.21 0.02 -0.07 0.18
Redwood NP -0.13 0.06 -0.13 0.01 -0.19 0.01
Rocky Mountain NP -0.03 0.31 -0.17 0.01 -0.14 0.04
San Gorgonio W -0.19 0.02 -0.14 0.13 -0.03 0.46
Shenandoah NP 0.01 0.54 -0.13 0.09 0.00 0.54
Tonto NM -0.04 0.31 -0.08 0.06 -0.14 0.02
Washington -0.01 0.46 -0.04 0.46 -0.13 0.24
Weminuche W -0.13 0.02 -0.12 0.01 -0.01 0.46
Yellowstone NP -0.17 0.31 -0.12 0.13 -0.33 0.00
Yosemite NP 0.07 0.38 -0.07 0.24 -0.06 0.13
Source: (Sisler and Malm, 2000, Tables 5-7).

Exhibit A-3 Visibility Benefits from Different Policy Analyses
Benefits (millions 1999$)

Analysisa National Year of Analysis
Coverage Recreational Residential
NAPAP Partial 1994 1,451 2,559
Navajo Partial 1991 268 n/ab
Ozone/PM NAAQS National 1997 5,983 3,906
812 Retrospective National 1997 n/a 4,311
NOx SIP Call Partial 1998 13 75
Haze Rule National 1999 1,816 149
812 Prospective National 1999 3,697 n/a
Section 126 Partial 1999 43 31
Tier 2 National 1999 385 578

a
The following describes the policy scenarios considered in each of the analyses:
National Acid Precipitation Assessment Program (NAPAP): Benefits due to improvements in 2010 visibility conditions in the
Eastern U.S., relative to what would have been expected to occur without Title IV.
Navajo: Benefits due to a 90% reduction in sulfur emissions from the Navajo Generating Station near the Grand Canyon in
1995.
812 Retrospective: Visibility benefits associated with the Clean Air Act between 1970 and 1990.
Ozone/PM National Ambient Air Quality Standards: Benefits associated with changes from the PM NAAQS 10/65 incremental
to the current PM10 standard assuming partial attainment.
NOx SIP Call: 0.15 trading alternative incremental to baseline 2007 using a threshold of background.
Haze Rule: 1.0 dv/10 years incremental to baseline 2015, fugitive dust controls included, threshold of background, wilderness
area benefits not additional to total regional values.
812 Prospective: Visibility benefits due to criteria pollutants in 2010 post-1990 CAA Amendments.
Section 126: Visibiltiy benefits associated with changes from the 2007 representative year scenario.
Tier 2: Benefits due to changes from the final Tier 2 2030 control scenario.
b
n/a: benefits were not calculated for this visibility category in the analysis.

APPENDIX B METHOD TO ESTIMATE VISIBILITY BENEFITS
Economic benefits may result from two broad categories of visibility changes: (1) changes in
“residential” visibility – i.e., the visibility in and around the locations where people live; and (2) changes in
“recreational” visibility at Class I Areas – i.e., visibility at Class I national parks and wilderness areas.7 In this
analysis, only recreational benefits are included in the primary presentation of benefits; residential benefits are
presented as an alternative calculation of visibility benefits.
Within the category of recreational visibility, further distinctions have been made. There is evidence
(Chestnut et al., 1990b) that an individual’s WTP for improvements in visibility at a Class I Area is influenced
by whether it is in the region in which the individual lives, or whether it is somewhere else. In general people
appear to be willing to pay more for visibility improvements at parks and wilderness areas that are “in-region”
than at those that are “out-of-region.” This is plausible, because people are more likely to visit, be familiar
with, and care about parks and wilderness areas in their own part of the country.
To value estimated visibility changes, we are using an approach consistent with economic theory.
Below we discuss an application of the Constant Elasticity of Substitution (CES) utility function approach8
to value both residential visibility improvements and visibility improvements at Class I Areas in the United
States. This approach is based on the preference calibration method developed by Smith et al. (1999). The
presentation of this methodology is organized as follows. The basic utility model is presented in Section 0.1.
In Section 0.2 we discuss the measurement of visibility, and the mapping from environmental “bads” to
environmental “goods.” In Sections 0.3 and 0.4 we summarize the information that is available to estimate the
parameters of the model corresponding to visibility at in-region and out-of-region Class I Areas, and visibility
in residential areas, respectively, and we describe the methods used to estimate these parameters. Section 0.5
synthesizes the results.
B.1 Basic Utility Model
We begin with a CES utility function in which a household derives utility from
(1) “all consumption goods,” X,

(2) visibility in the residential area in which the household is located (“residential visibility”),9
(3) visibility at Class I Areas in the same region as the household (“in-region recreational visibility”),
and
(4) visibility at Class I Areas outside the household’s region (“out-of-region recreational visibility”).
7
Hereafter referred to as Class I Areas, which are defined as areas of the country such as national parks, national wilderness
areas, and national monuments that have been set aside under Section 162(a) of the Clean Air Act to receive the most stringent degree
of air quality protection. Class I federal lands fall under the jurisdiction of three federal agencies, the National Park Service, the Fish
and Wildlife Service, and the Forest Service.
8
The Constant Elasticity of Substitution utility function has been chosen for use in this analysis due to its flexibility when
illustrating the degree of substitutability present in various economic relationships (in this case, the tradeoff between income and
improvements in visibility).
9
We remind the reader that, although residential and recreational visibility benefits estimation is discussed simultaneously in
this section, benefits are calculated and presented separately for each visibility category.
Abt Associates Inc. B-1 August 2000

There are a total of six regions being considered, so there are 5 regions for which any household is out-of-
region. The utility function of a household in the nth residential area and the ith region of the country is:
Ni Nj
U ni = ( X + θ Z + ∑ γ ik Q +
ρ ρ
n
ρ
ik ∑∑δ jk Q ρjk )1/ ρ ,
k=1 j≠ i k=1
θ > 0 , γ ik > 0 , ∀ i , k , δ jk > 0 , ∀ j , k , ρ ≤ 1.
where
Zn = the level of visibility in the nth residential area;
Qik = the level of visibility at the kth in-region park (i.e., the kth park in the ith region);
Qjk = the level of visibility at the kth park in the jth region ( for which the household is out-of-region),
j…i;
Ni = the number of Class I Areas in the ith region;
Nj = the number of Class I Areas in the jth region (for which the household is out-of-region), j…i;
and
2, the (’s and *’s are parameters of the utility function corresponding to the visibility levels at residential areas,
and at in-region and out-of-region Class I Areas, respectively. In particular, the (ik's are the parameters
corresponding to visibility at in-region Class I Areas; the *1’s are the parameters corresponding to visibility
at Class I Areas in region 1 (California), if i…1; the *2’s are the parameters corresponding to visibility at Class
I Areas in region 2 (Colorado Plateau), if i…2, and so forth. Because the model assumes that the relationship
between residential visibility and utility is the same everywhere, there is only one 2. The parameter D in this
CES utility function is an important determinant of the slope of the marginal WTP curve associated with any
of the environmental quality variables. When D=1, the marginal WTP curve is horizontal. When D<1, it is
downward sloping.
The household’s budget constraint is:
m − p⋅ X ≤ 0 ,
where m is income, and p is the price of X. Without loss of generality, set p = 1. The only choice variable is
X. The household maximizes its utility by choosing X=m. The indirect utility function for a household in the
nth residential area and the ith region is therefore
Ni Nj
Vni ( m, Z n ,Q ; θ ,γ ,δ , ρ ) = ( m ρ + θ Z nρ + ∑ γ ik Qikρ + ∑ ∑δ jk Q ρjk )1 / ρ ,

k =1 j≠ i k = 1
where Q denotes the vector of vectors, Q1, Q2, Q3, Q4, Q5, and Q6, and the unsubscripted ( and * denote vectors
as well.
Given estimates of D, 2, the (’s and the *’s, the household’s utility function and the corresponding
WTP functions are fully specified. The household’s WTP for any set of changes in the levels of visibility at
in-region Class I Areas, out-of-region Class I Areas, and the household’s residential area can be shown to be:

Ni Nj
WTPni ( ∆ Z , ∆ Q ) = m − [ m + θ ( Z − Z ) + ∑ γ ik ( Q
ρ ρ
0n
ρ
1n
ρ
0 ik
ρ
− Q )+
1 ik ∑∑δ jk ( Q0ρ jk − Q1ρjk ) ]1 / ρ .
k =1 j ≠ i k =1
The household’s WTP for a single visibility improvement will depend on its order in the series of visibility
improvements the household is valuing. If it is the first visibility improvement to be valued, the household’s
WTP for it follows directly from the previous equation. For example, the household’s WTP for an
improvement in visibility at the first in-region park, from Qi1 = Q0i1 to Qi1 = Q1i1, is
WTP( ∆ Qi1 ) = m − [mρ + γ i1 (Q0ρi1 − Q1ρi1 )]1/ ρ ,
if this is the first (or only) visibility change the household values.
B.2 Measure of Visibility: Environmental “Goods” Versus “Bads”
In the above model, Q and Z are environmental “goods.” As the level of visibility increases, utility
increases. The utility function and the corresponding WTP function both have reasonable properties. The first
derivative of the indirect utility function with respect to Q (or Z) is positive; the second derivative is negative.
WTP for a change from Q0 to a higher (improved) level of visibility, Q1, is therefore a concave function of Q1,
with decreasing marginal WTP.
The measure of visibility that is currently preferred by air quality scientists is the deciview, which
increases as visibility decreases. Deciview, in effect, is a measure of the lack of visibility. As deciviews
increase, visibility, and therefore utility, decreases. The deciview, then, is a measure of an environmental
“bad.” There are many examples of environmental “bads” – all types of pollution are environmental “bads.”
Utility decreases, for example, as the concentration of particulate matter in the atmosphere increases.
One way to value decreases in environmental bads is to consider the “goods” with which they are
associated, and to incorporate those goods into the utility function. In particular, if B denotes an environmental
“bad,” such that:
∂V
<0 ,
∂B
and the environmental “good,” Q, is a function of B,
Q = F ( B) ,
then the environmental “bad” can be related to utility via the corresponding environmental “good”:10
V = V (m, Q ) = V (m, F ( B )) .
The relationship between Q and B, F(B), is an empirical relationship that must be estimated.
10
There may be more than one “good” related to a given environmental “bad.” To simplify the discussion, however, we
assume only a single “good.”

There is a potential problem with this approach, however. If the function relating B and Q is not the
same everywhere (i.e., if for a given value of B, the value of Q depends on other factors as well), then there
can be more than one value of the environmental good corresponding to any given value of the environmental
bad, and it is not clear which value to use. This has been identified as a problem with translating deciviews
(an environmental “bad”) into visual range (an environmental “good”). It has been noted that, for a given
deciview value, there can be many different visual ranges, depending on the other factors that affect visual
range – such as light angle and altitude. We note here, however, that this problem is not unique to visibility,
but is a general problem when trying to translate environmental “bads” into “goods.”11
In order to translate deciviews (a “bad”) into visual range (a “good”), we use a relationship derived
by Malm and Pitchford (1994) in which
391
DV = 10* ln( ),
VR
where DV denotes deciview and VR denotes visual range (in kilometers). Solving for VR as a function of DV
yields
VR= 391* e−0.1DV .
This conversion is based on specific assumptions characterizing the “average” conditions of those factors,
such as light angle, that affect visual range. To the extent that specific locations depart from the average
conditions, the relationship will be an imperfect approximation.12
B.3 Estimating the Parameters for Visibility at Class I Areas: the (’s and *’s
As noted in Section 0.2, if we consider a particular visibility change as the first or the only
visibility change valued by the household, the household’s WTP for that change in visibility can be
calculated, given income (m), the “shape” parameter, D, and the corresponding recreational visibility
parameter. For example, a Southeast household’s WTP for a change in visibility at in-region parks
(collectively) from Q1 = Q01 to Q1 = Q11 is:
WTP(DQ1 ) = m − [mr + g1 (Q01

r
− Q11
r
)]1/ r
if this is the first (or only) visibility change the household values.
11
Another example of an environmental “bad” is particulate matter air pollution (PM). The relationship between survival
probability (Q) and the ambient PM level is generally taken to be of the form
Q = 1 − α eβPM .
where " denotes the mortality rate (or level) when there is no ambient PM (i.e., when PM=0). However, " is implicitly a function of
all the factors other than PM that affect mortality. As these factors change (e.g., from one location to another), " will change (just as
visual range changes as light angle changes). It is therefore possible to have many values of Q corresponding to a given value of PM,
as the values of " vary.
12
Ideally, we would want the location- , time-, and meteorological condition-specific relationships between deciviews and
visual range, which could be applied as appropriate. This is probably not feasible, however.

Alternatively, if we have estimates of m as well as WTP1in and WTP1out of in-region and out-of-
region households, respectively, for a given change in visibility from Q01 to Q11 in Southeast parks, we can
solve for (1 and *1 as a function of our estimates of m, WTP1in and WTP1out, for any given value of D.
Generalizing, we can derive the values of ( and * for the jth region as follows:
( m − WTPjin )ρ − mρ
γj =
( Q0ρ j − Q1ρj )
and
( m − WTPj out )ρ − mρ
δj = .
( Q0ρ j − Q1ρj )
Chestnut and Rowe (1990b) and Chestnut (1997) estimated WTP (per household) for specific
visibility changes at national parks in three regions of the United States – both for households that are in-
region (in the same region as the park) and for households that are out-of-region. The Chestnut and Rowe
study asked study subjects what they would be willing to pay for each of three visibility improvements in
the national parks in a given region. Study subjects were shown a map of the region, with dots indicating
the locations of the parks in question. The WTP questions referred to the three visibility improvements in
all the parks collectively; the survey did not ask subjects’ WTP for these improvements in specific parks
individually. Responses were categorized according to whether the respondents lived in the same region as
the parks in question (“in-region” respondents) or in a different region (“out-of-region” respondents). The
areas for which in-region and out-of-region WTP estimates are available from Chestnut and Rowe (1990b),
and the sources of benefits transfer-based estimates that we employ in the absence of estimates, are
summarized in Exhibit B-1. In all cases, WTP refers to WTP per household.

Exhibit B-1 Available Information on WTP for Visibility Improvements in National Parks
Region of Park Region of Household
In-Regiona Out-of-Regionb
1. California WTP estimate from study WTP estimate from study
2. Colorado Plateau WTP estimate from study WTP estimate from study
3. Southeast United States WTP estimate from study WTP estimate from study
4. Northwest United States (based on benefits transfer from California)
5. Northern Rockies (based on benefits transfer from Colorado Plateau)
6. Rest of United States (based on benefits transfer from Southeast U.S.)

a
In-region” WTP is WTP for a visibility improvement in a park in the same region as that in which the household is located. For
example, in-region WTP in the “Southeast” row is the estimate of the average Southeast household’s WTP for a visibility
improvement in a Southeast park.
b
Out-of-region” WTP is WTP for a visibility improvement in a park that is not in the same region in which the household is
located. For example, out-of-region WTP in the “Southeast” row is the estimate of WTP for a visibility improvement in a park in
the Southeast by a household outside of the Southeast.
In the primary calculation of visibility benefits for this analysis, only visibility changes at parks
within visibility regions for which a WTP estimate was available from Chestnut and Rowe (1990b) are
considered (for both in- and out-of-region benefits). Primary estimates will not include visibility benefits
calculated by transferring WTP values to visibility changes at parks not included in the Chestnut and Rowe
study. Transferred benefits at parks located outside of the Chestnut and Rowe visibility regions will,
however, be included as an alternative calculation.
The values of the parameters in a household’s utility function will depend on where the household
is located. The region-specific parameters associated with visibility at Class I Areas (that is, all parameters
except the residential visibility parameter) are arrayed in Exhibit B-2. The parameters in columns 1-3 can
be directly estimated using WTP estimates from Chestnut and Rowe (1990b) (the columns labeled “Region
1,” “Region 2,” and “Region 3").

Exhibit B-2 Summary of Region-Specific Recreational Visibility Parameters to be Estimated in
Household Utility Functions
Region of Region of Park

Household
Region 1 Region 2 Region 3 Region 4 Region 5 Region 6
Region 1 (1a *2 *3 *4 *5 *6
Region 2 *1 (2 *3 *4 *5 *6
Region 3 *1 *2 (3 *4 *5 *6
Region 4 *1 *2 *3 (4 *5 *6
Region 5 *1 *2 *3 *4 (5 *6
Region 6 *1 *2 *3 *4 *5 (6
a
The parameters arrayed in this table are region-specific rather than park-specific or wilderness area-specific. For example, *1 is
the parameter associated with visibility at “ Class I Areas in region 1" for a household in any region other than region 1. The
benefits analysis must derive Class I Area-specific parameters – e.g., *1k, for the kth Class I Area in the first region.
For the three regions covered in Chestnut and Rowe (1990b) (California, the Colorado Plateau, and
the Southeast United States), we can directly use the in-region WTP estimates from the study to estimate
the parameters in the utility functions corresponding to visibility at in-region parks ((1); similarly, we can
directly use the out-of-region WTP estimates from the study to estimate the parameters for out-of-region
parks (*1). For the other three regions not covered in the study, however, we must rely on benefits transfer
to estimate the necessary parameters.
While Chestnut and Rowe (1990b) provide useful information on households’ WTP for visibility
improvements in national parks, there are several significant gaps remaining between the information
provided in that study and the information necessary for the benefits analysis. First, as noted above, the
WTP responses were not park-specific, but only region-specific. Because visibility improvements vary
from one park in a region to another, the benefits analysis must value park-specific visibility changes.
Second, not all Class I Areas in each of the three regions considered in the study were included on the maps
shown to study subjects. Because the focus of the study was primarily national parks, most Class I
wilderness areas were not included. Third, only three regions of the United States were included, leaving
the three remaining regions without direct WTP estimates.
In addition, Chestnut and Rowe (1990b) elicited WTP responses for three different visibility
changes, rather than a single change. In theory, if the CES utility function accurately describes household
preferences, and if all households in a region have the same preference structure, then households’ three
WTP responses corresponding to the three different visibility changes should all produce the same value of
the associated recreational visibility parameter, given a value of D and an income, m. In practice, of
course, this is not the case.
In addressing these issues, we take a three-phase approach:
(1) We estimate region-specific parameters for the region in the modeled domain covered by
Chestnut and Rowe (1990b) (California, the Colorado Plateau, and the Southeast) – (1, (2, and (3 and *1,

*2, and *3. (2) We infer region-specific parameters for those regions not covered by the Chestnut and Rowe
study (the Northwest United States, the Northern Rockies, and the rest of the U.S.) – (4, (5, and (6 and *4,
*5, and *6. (3) We derive park- and wilderness area-specific parameters within each region ((1k and *1k, for
k=1, ..., N1; (2k and *2k, for k=1, ..., N2; and so forth).
The question that must be addressed in the first phase is how to estimate a single region-specific in-
region parameter and a single region-specific out-of-region parameter for each of the three regions covered
in Chestnut and Rowe (1990b) from study respondents’ WTPs for three different visibility changes in each
region. All parks in a region are treated collectively as if they were a single “regional park” in this first
phase. In the second phase, we infer region-specific recreational visibility parameters for regions not
covered in the Chestnut and Rowe study (the Northwest United States, the Northern Rockies, and the rest
of the U.S.). As in the first phase, we ignore the necessity to derive park-specific parameters at this phase.
Finally, in the third phase, we derive park- and wilderness area-specific parameters for each region.
B.3.1 Estimating Region-Specific Recreational Visibility Parameters for the Region Covered in the
Chestnut and Rowe Study (Regions 1, 2, and 3)
Given a value of D and estimates of m and in-region and out-of-region WTPs for a change from Q0
to Q1 in a given region, the in-region parameter, (, and the out-of-region parameter, *, for that region can
be solved for. Chestnut and Rowe (1990b), however, considered not just one, but three visibility changes
in each region, each of which results in a different calibrated ( and a different calibrated *, even though in
theory all the (’s should be the same and similarly, all the *’s should be the same. For each region,
however, we must have only a single ( and a single *.
Denoting γ$ j as our estimate of ( for the jth region, based on all three visibility changes, we chose
γ$ j to best predict the three WTPs observed in the study for the three visibility improvements in the jth
region. First, we calculated γ$ ji , i=1, 2, 3, corresponding to each of the three visibility improvements
considered in the study. Then, using a grid search method beginning at the average of the three γ$ ji ’s , we
chose γ$ j to minimize the sum of the squared differences between the WTPs we predict using γ$ j and the
three region-specific WTPs observed in the study. That is, we selected γ$ j to minimize:
∑ ( WTP ( γ$
ij j ) − WTPij )2
i= 1
where WTPij and WTPij( γ$ j ) are the observed and the predicted WTPs for a change in visibility in the jth
region from Q0 = Q0i to Q1= Q1i, i=1, ..., 3. An analogous procedure was used to select an optimal *, for
each of the three regions in the Chestnut and Rowe study.
B.3.2 Inferring Region-Specific Recreational Visibility Parameters for Regions Not Covered in the
Chestnut and Rowe Study (Regions 4, 5, and 6)
One possible approach to estimating region-specific parameters for regions not covered by
Chestnut and Rowe (1990b) ((4, (5, and (6 and *4, *5, and *6) is to simply assume that households’ utility
functions are the same everywhere, and that the environmental goods being valued are the same – e.g., that

a change in visibility at national parks in California is the same environmental good to a Californian as a
change in visibility at national parks in Minnesota is to a Minnesotan.
For example, to estimate *4 in the utility function of a California household, corresponding to

visibility at national parks in the Northwest United States, we might assume that out-of-region WTP for a
given visibility change at national parks in the Northwest United States is the same as out-of-region WTP
for the same visibility change at national parks in California (income held constant). Suppose, for
example, that we have an estimated mean WTP of out-of-region households for a visibility change from Q01
to Q11 at national parks in Califonia (region 1), denoted WTP1out. Suppose the mean income of the out-of-
region subjects in the study was m. We might assume that, for the same change in visibility at national
parks in the Northwest United States, WTP4out = WTP1out among out-of-region individuals with income m.
We could then derive the value of *4, given a value of D as follows:
( m − WTP4out )ρ − m ρ
δ4 = ρ
Q04 − Q14ρ
where Q04 = Q01 and Q14 = Q11, (i.e., where it is the same visibility change in parks in region 4 that was
valued at parks in the region 1).
This benefits transfer method assumes that (1) all households have the same preference structures
and (2) what is being valued in the Northwest United States (by a California household) is the same as
what is being valued in the California (by all out-of-region households). While we cannot know the extent
to which the first assumption approximates reality, the second assumption is clearly problematic. National
parks in one region are likely to differ from national parks in another region in both quality and quantity
(i.e., number of parks).
One statistic which is likely to reflect both the quality and quantity of national parks in a region is
the average annual visitation rate to the parks in that region. A reasonable way to gauge the extent to
which out-of-region people would be willing to pay for visibility changes in parks in the Northwest United
States versus in California might be to compare visitation rates in the two regions.13 Suppose, for example,
that twice as many visitor-days are spent in California parks per year as in parks in the Northwest United
States per year. This could be an indication that the parks in California are in some way more desirable
than those in the Northwest United States and/or that there are more of them -- i.e., that the environmental
goods being valued in the two regions (“visibility at national parks”) are not the same.
A preferable way to estimate *4, then, might be to assume the following relationship:
WTP4out n4
=
WTP1out n1
(income held constant), where n1 = the average annual number of visitor-days to California parks and n4 =
the average annual number of visitor-days to parks in the Northwest United States. This implies that
13
We acknowledge that reliance on visitation rates does not get at nonuse value.

n4
WTP4out = * WTP1out
n1
for the same change in visibility in region 4 parks among out-of-region individuals with income m. If, for
example, n1 = 2n4, WTP4out would be half of WTP1out. The interpretation would be the following:
California national parks have twice as many visitor-days per year as national parks in the Northwest
United States; therefore they must be twice as desirable/plentiful; therefore, out-of-region people would be
willing to pay twice as much for visibility changes in California parks as in parks in the Northwest United
States; therefore a Californian would be willing to pay only half as much for a visibility change in national
parks in the Northwest United States as an out-of-region individual would be willing to pay for the same
visibility change in national parks in California. This adjustment, then, is based on the premise that the
environmental goods being valued (by people out-of-region) are not the same in all regions.
The parameter *4 is estimated as shown above, using this adjusted WTP4out. The same procedure
is used to estimate *5 and *6. We estimate (4, (5, and (6 in an analogous way, using the in-region WTP
estimates from the transfer regions, e.g.,
n4
WTP4in = *WTP1in .
n1
B.3.3 Estimating Park- and Wilderness Area-Specific Parameters
As noted above, Chestnut and Rowe (1990b) estimated WTP for a region’s national parks
collectively, rather than providing park-specific WTP estimates. The (’s and *’s are therefore the
parameters that would be in household utility functions if there were only a single park in each region, or if
the many parks in a region were effectively indistinguishable from one another. Also noted above is the
fact that the Chestnut and Rowe study did not include all Class I Areas in the regions it covered, focusing
primarily on national parks rather than wilderness areas. Most Class I wilderness areas were not
represented on the maps shown to study subjects. In California, for example, there are 31 Class I Areas,
including six national parks and 25 wilderness areas. The Chestnut and Rowe study map of California
included only 10 of these Class I Areas, including all six of the national parks. It is unclear whether
subjects had in mind “all parks and wilderness areas” when they offered their WTPs for visibility
improvements, or whether they had in mind the specific number of (mostly) parks that were shown on the
maps. The derivation of park- and wilderness area-specific parameters depends on this.

Abt Associates Inc. # 4800 Montgomery Lane # Bethesda, MD 20814-5341
www.abtassoc.com # ph: 301-913-0500 # fax: 301-652-3618

Out of Sight: The Science and Economics of Visibility Impairment

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Out of Sight: The Science and Economics of Visibility Impairment

Uploaded by

Copyright:

Available Formats

Abt Associates Inc.

# 4800 Montgomery Lane #

EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ES-1

3. TRENDS IN VISIBILITY IMPAIRMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

4. LEGISLATIVE AND REGULATORY HISTORY OF VISIBILITY IMPAIRMENT . . . . . . . 26

5. ECONOMICS: VISIBILITY IMPAIRMENT AND PARK VISITATION . . . . . . . . . . . . . . . 31

6. ECONOMICS: THE NON-MARKET VALUE OF VISIBILITY . . . . . . . . . . . . . . . . . . . . . . 41

7. POWER PLANT EMISSION REDUCTIONS AND ASSOCIATED VISIBILITY BENEFITS

APPENDIX A DETAILED TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1

APPENDIX B METHOD TO ESTIMATE VISIBILITY BENEFITS . . . . . . . . . . . . . . . . . . . . . . . . B-1

Exhibit 1-1 Denver on a Clear Day and on a Hazy Day . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

A variety of sources contribute to the air pollutant

Exhibit 1-1 Denver on a Clear Day and on a Hazy Day

Abt Associates Inc. 1 August 2000

Exhibit 2-1a Visible Plume from Local Stacks

Abt Associates Inc. 2 August 2000

Exhibit 2-1c Regional Haze

2.1 Causes of Visibility Impairment

Abt Associates Inc. 3 August 2000

Abt Associates Inc. 4 August 2000

Exhibit 2-3a Contribution to Visibility Impairment in the Eastern U.S.

Exhibit 2-3b Contribution to Visibility Impairment in the Mid-West

Abt Associates Inc. 5 August 2000

2.2 Observing and Measuring Visibility Impairment

Abt Associates Inc. 6 August 2000

Abt Associates Inc. 7 August 2000

Exhibit 2-6 Types of Particulate Matter and Impact on Image-Forming Light

Abt Associates Inc. 8 August 2000

when the air is relatively clean.

Exhibit 2-7, based on NPS (1988, 8

These studies were also able

Abt Associates Inc. 9 August 2000

2.2.2 How Visibility Impairment is Measured

Abt Associates Inc. 10 August 2000

Abt Associates Inc. 11 August 2000

Abt Associates Inc. 12 August 2000

A variety of ways to record atmospheric

The airport data are the most extensive,

Abt Associates Inc. 13 August 2000

3.1.1 Visibility at Parks Where the Vista Is Integral to the Experience

Abt Associates Inc. 14 August 2000

Abt Associates Inc. 15 August 2000

Acadia National Park

Good Visibility Days Mid-Range Poor Visibility Days

Abt Associates Inc. 16 August 2000

Grand Canyon National Park

Good Visibility Days Mid-Range Poor Visibility Days

Good Visibility Day (5 deciviews) Poor Visibility Day (13 deciviews)

Abt Associates Inc. 17 August 2000

Great Smoky Mountains National Park

Good Visibility Days Mid-Range Poor Visibility Days

Abt Associates Inc. 18 August 2000

Shenandoah National Park

Good Visibility Days Mid-Range Poor Visibility Days

Abt Associates Inc. 19 August 2000

Yosemite National Park

Good Visibility Days Mid-Range Poor Visibility Days

Good Visibility Day Poor Visibility Day

Abt Associates Inc. 20 August 2000

Excellent Visibility Day (3 deciviews) Bad Visibility Day (33 deciviews)

Exhibit 3-7b Extreme Visibility Days at Grand Canyon National Park

Excellent Visibility Day (0 deciviews) Bad Visibility Day (22 deciviews)

Excellent Visibility Day (4 deciviews) Bad Visibility Day (37 deciviews)

Abt Associates Inc. 21 August 2000