EPA/600/R-01/004
April 2002
Imaging Spectroscopy for
Determining Rangeland Stressors
to Western Watersheds
Pollution Prevention and
New Technology
by
David Williams
Principal Investigator
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Environmental Photographic Interpretation Center
12201 Sunrise Valley Drive
555 National Center, Suite 2D-115
Reston, Virginia 20192
(703) 648-4798
and
William Kepner
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Environmental Sciences Division
944 East Harmon Avenue
Las Vegas, Nevada 89119
(702) 798-2193
176LEB02.RPT ~ 5/6/02
-------
TECHNICAL REPORT DATA
1. Report No.
EPA/600/R-01/004
2.
3. Recipient's Accession No.
4. Title and Subtitle
Imaging Spectroscopy for determining Rangeland Stressors to Western Watersheds
5. Report Date
6. Performing Organization Code
7. Author(s)
Williams, D.J., Kepner, W.G.
8. Performing Organization Report
No. (TIP#)
NERL-LV-ESD 01-004
9. Performing Organization Name & Address
U.S. EPA, Office of Research & Development, National Exposure Research Laboratory
Environmental Sciences Division, Environmental Chemistry Branch
944 E. Harmon Ave.
Las Vegas, NV 89119
10. Program Element Number
11. Contract/Grant No.
12. Sponsoring Agency Name & Address
U.S. EPA, Office of Research & Development, National Exposure Research Laboratory
Environmental Sciences Division, Environmental Chemistry Branch
944 E. Harmon Ave.
Las Vegas, NV 89119
13. Type of Report & Period Covered
14. Sponsoring Agency Code
EPA 600/07
15. Supplemental Notes or Complete Citation!
Williams, D.J., Kepner, W.G. Imaging Spectroscopy for determining Rangeland Stressors to Western
Watersheds EPA/600/R-01/004 Las Vegas, NV:U.S. Environmental Protection Agency
16. Abstract The Environmental Protection Agency is developing rangeland ecological indicators in twelve western states using advanced remote
sensing techniques. Fine spectral resolution (hyperspectral) sensors, or imaging spectrometers, can detect the subtle spectral features that make
vegetation and soil discrimination possible. This study will use hyperspectral remote sensing data, such as NASA's Airborne Visible-Infra-Red
Imaging Spectrometer (AVIRIS), a system capable of 5 to 20 meter spatial resolution. Airborne and satellite remote sensing will provide vegetation
mapping at the species level, soil types and characteristics, and landscape information such as erosional features. Vegetation community structure,
spatial distribution, and health can then be determined and combined with climatic data to classify rangeland condition and identify disturbed
regions.
Accurate determination of rangeland vegetation and soils is required to establish reliable landscape indicators. Rangelands in the West encompass a
range of ecological conditions or states from healthy to at risk to degraded. This gradient of conditions can be quantitatively determined and used to
develop landscape indicators. Vegetation communities differ over the gradient of rangeland conditions. Soil attributes such as organic matter
content, salinity, moisture, mineralogy, and physical condition influence and are influenced by vegetation cover. The water quality of the watershed
is directly impacted by these rangeland variables. Imaging spectroscopy can detect these variables and allows for landscape scale assessment and
monitoring of stressors to water resources in the West.
Potential research with the Bureau of Land Management, US Department of Agriculture, and U.S. Geological Survey will correlate remote sensing
data with ground measurements. The long-term goal of this work is to develop a methodology using current technologies for use with the
forthcoming hyperspectral satellite platforms scheduled for operational service within the next 2 to 3 years.
17. Key Words and Document Analysis
a. Descriptors: imaging spectroscopy
b. Identifiers/Open-ended Terms
c. COSATI
18. Distribution Statement
RELEASE TO THE PUBLIC
19. Security Class (This Report)
UNCLASSIFIED
21. No. of Pages
39
20. Security Class (This Page)
UNCLASSIFIED
22. Price
EPA 2220-1
-------
Table of Contents
Page
Section 1 Core Task Summary 1
Section 2 Core Task Narrative 2
Goals/Objectives 2
Background/Literature Review 2
Statement of Problem 2
Problem 3
Grazing pressure/intensity 3
Climate change 3
Invasive species 4
Watershed impacts 4
Development of Ecologial Indicators to Access Rangeland Condition 5
Need 5
Indicators 6
Use of Remote Sensing to Develop Rangeland Ecological Indicators 6
Remote sensing systems 7
Section 3 Imaging Spectroscopy 8
Spectral reflectance of vegetation 8
Discrimination of plant species using imaging spectroscopy 8
Section 4 Imaging Radar 11
Development of Landscape Ecological Indicators Using Fine Spectral and Spatial
Resolution Sensors 11
Determination of ecological indicators 11
Soil mapping and soil assessment 12
Soil classification 12
Soil mineralogy 12
Soil organic matter 12
Relating soil parameters to vegetation 13
Salinization 13
Soil crusting 13
Biological soil crusts 13
Stability 13
Ancillary Data 14
Climate 14
Ground measurements 14
Grazing plot intensity, duration and season of use records and other land use data 14
Topography 14
iii
-------
Hypothesis 14
Conceptual Basis for Hypothesis 15
Technical Approach 16
Methodology 16
General approach 16
Subtask 1 16
Subtask 2 17
Data Acquisition and Planning 17
Data Analysis and Interpretation 17
Section 5 Potential for Reducing Uncertainty in Exposure Assessments and Importance to the
Milestones and Program Area Goals 19
Section 6 Quality Assurance Statement 21
Anticipated Results/Milestones 21
Results and reporting 22
Milestones 22
Deliverables 22
References 23
Proposed Staffing 30
Description of Facilities 31
Appendix A Description of Potential Study Sites 32
Pilot Study Sites 32
Follow-on Study Sites 32
Appendix B Personnel 34
David James Williams 35
William G. Kepner 37
iv
-------
List of Tables
Table Page
1 Criteria and Indicators of Rangeland Health (Adapted from National Research Council,
1994) 6
2 Indicator Development Adapted from the Rangeland Health Evaluation Matrix (National
Research Council, 1994) 15
List of Figures
Figure Page
1 Rill and gully formation due to soil erosion (National Research Council, 1994) 4
2 The electromagnetic spectrum. Scale in micrometers (|im) 7
3 Spectral reflectance of common rangeland vegetation and soils. Data courtesy of
USGS Speclab 9
4 Comparison of vegetation species by band center analysis 9
-------
Section 1
Core Task Summary
The Environmental Protection Agency is developing rangeland ecological indicators in twelve western
states using advanced remote sensing techniques. Fine spectral resolution (hyperspectral) sensors, or
imaging spectrometers, can detect the subtle spectral features that make vegetation and soil
discrimination possible. This study will use hyperspectral remote sensing data, such as NASA's
Airborne Visible-Infra-Red Imaging Spectrometer (AVIRIS), a system capable of 5 to 20 meter spatial
resolution. Airborne and satellite remote sensing will provide vegetation mapping at the species level,
soil types and characteristics, and landscape information such as erosional features. Vegetation
community structure, spatial distribution, and health can then be determined and combined with climatic
data to classify rangeland condition and identify disturbed regions.
Accurate determination of rangeland vegetation and soils is required to establish reliable landscape
indicators. Rangelands in the West encompass a range of ecological conditions or states from healthy to
at risk to degraded. This gradient of conditions can be quantitatively determined and used to develop
landscape indicators. Vegetation communities differ over the gradient of rangeland conditions. Soil
attributes such as organic matter content, salinity, moisture, mineralogy, and physical condition influence
and are influenced by vegetation cover. The water quality of the watershed is directly impacted by these
rangeland variables. Imaging spectroscopy can detect these variables and allows for landscape scale
assessment and monitoring of stressors to water resources in the West.
Potential research with the Bureau of Land Management, US Department of Agriculture, and U.S.
Geological Survey will correlate remote sensing data with ground measurements. The long-term goal of
this work is to develop a methodology using current technologies for use with the forthcoming
hyperspectral satellite platforms scheduled for operational service within the next 2 to 3 years.
1
-------
Section 2
Core Task Narrative
Goals/Objectives
The long-term goal of this research is to institute the use of imaging spectroscopic and radar methods for
relating rangeland landscape pattern and biophysical variables to watershed condition. The objectives of
this project are to:
• Identify and map at a fine spatial resolution biophysical variables such as vegetation
communities at the species level and soil types
• Determine the extremes of rangeland condition using remote sensing and correlate
these with ground measurements. Identify and evaluate the gradients that exist
between the extremes of rangeland conditions
• Establish landscape ecological indicators that predict the condition of rangelands in
the West
• Refine landscape pattern metrics that relate to functioning and degraded watershed
conditions
Background/Literature Review
Statement of Problem
Rangelands are valuable natural resources that provide forage for livestock as well as wildlife habitat and
recreational opportunities. The nation's rangelands encompass more than 310 million hectares, primarily
in the western states. Rangelands regulate the quantity and quality of water for the surrounding
watersheds. Rangeland vegetation has the ability to intercept runoff from rain events, and increase water
infiltration to the soil allowing for aquifer recharge. Rangeland soils contribute to water quality by
immobilizing and transforming nutrients and contaminants and act as a groundwater filter. Rangeland
degradation can be caused by overgrazing, drought, improper recreational use, and other anthropogenic
and natural stresses. Information on the condition, or state of the nation's rangelands is important for the
protection of these resources.
The demands placed upon rangelands are escalating as the populations of the western states increase.
Development of land for commercial and residential use can remove areas of rangeland for these
purposes, simultaneously increasing the utilization of the remaining rangelands. Rangeland function can
be defined as the ability of the rangeland ecosystem to provide commodities such forage and habitat,
regulate water quantity and quality, and provide recreational opportunities. Improper rangeland
2
-------
management practices can cause severe degradation that can lead to irreversible changes of the
ecosystem. Degradation of these areas can negatively impact the surrounding watersheds through soil
erosion and desertification.
There is an urgent need for resource managers to know the state of the nation's rangelands, knowledge
that is hampered by lack of reliable and continuous data especially over large regional or watershed
scales. Development of data collection systems and ecological assessment methods are required to
evaluate and monitor these resources (National Research Council, 1994).
Problem
Livestock overgrazing, climate change, and certain types of land use can lead to a decline in rangeland
condition. The recovery process of rangelands to severe degradation may be long, or might never occur
(West et al., 1984). Arid and semi-arid ecosystems are especially prone to degradation due to climatic
conditions and major population shifts to the western US. The time scales and vegetal dynamics
associated with the recovery process is not fully understood for these ecosystems (Anderson and Holte,
1981).
Grazing pressure/intensity: Rangeland ecosystems vary due to the climatic and geologic conditions of
the region. Each rangeland ecosystem is a complex combination of vegetation species and soil types.
The carrying capacity for each rangeland is unique because of this ecological variability. Grazing
pressure or intensity is a way to determine if livestock utilization of a rangeland is proportional to the
area's carrying capacity. Grazing pressure is defined as the hypothetical Animal Unit Months
(AUM), the amount of forage required to support a cow and calf for one month according to the
area's carrying capacity (Mouat et al., 1997). High grazing pressure, as well as several other factors,
may lead to loss of desirable or palatable forage species, encroachment of non-palatable species, and
topsoil erosion due to overall groundcover loss (Sedgwick and Knopf, 1991).
Long term ecological studies on rangeland vegetation response to grazing practices have shown the
changes occurring are not continuous, reversible or consistent (Westoby et al., 1989). Previous
ecological theory held that plant communities strive toward a climax condition. Changes in
vegetation species and abundance due to disturbance follow a linear path from the climax state to
degraded states. This ecological succession model can be applied to rangeland management
decisions. Vegetation response due to grazing pressure could then be managed by varying the
stocking rate for a particular rangeland site. However, a wealth of evidence from rangeland studies
now indicates that the successional theory does not hold, especially for arid and semi-arid ecosystems
(Westoby et al., 1989).
Vegetation dynamics are now described by a phase transition model that allows for discrete
transitions between a set of states (Westoby et al., 1989). The state and transition theory proposes
that 1) plant communities changes may not be reversible to a parent climax state; 2) plant
communities may not be in equilibrium with the current climate; and 3) random (stochastic) events
may influence changes in vegetation dynamics.
Climate change: Climate is a major controlling factor for vegetation in the western rangelands (Olsen
et al., 1985). Drought is the greatest determinant of vegetation change, and trends in vegetative cover
can be directly related to climatic fluctuations. Within trends of vegetation change, grazing has a
large impact on vegetation resilience and recovery. Native rangeland vegetation has developed under
widely varying climatic conditions, and it takes a severe, long-term drought to destroy plant
3
-------
communities in these ecosystems (Cook and Sims, 1975). Rangeland resource managers must be
sensitive to climatic fluctuations as well as grazing pressure and stocking rates to maintain rangeland
productivity and environmental quality effectively.
Invasive species: Livestock often graze on forage preferentially according to their nutritional needs.
This selective grazing tends to reduce relatively homogenous vegetation units into a landscape
mosaic of heavily grazed and lightly grazed patches of vegetation (Bakker et al., 1983). Over grazed
units are usually characterized by invasive or exotic plant species of little forage value to wildlife and
livestock (Hatch and Tainton, 1991). These patches tend to increase in area and number as this type
of grazing continues (Fuls, 1992). Overall species diversity is low in the disturbed areas, a situation
that may lead to exotic species invasion (Pimm, 1991). However, recent work by Stohlgren et al.,
(1999) has shown that the richness (or dearth) of species diversity is not always a predictor of exotic
invasibility. The invasibility of certain landscapes depends upon variables such as availability of
resources, spatial scale, species-specific responses to grazing and other disturbances, and vegetation
and biome types. Landscape scale inventorying and monitoring of invasive species along with the
study of the effects of invasive species across gradients of resource availability is required to deal
with this problem (Stohlgren et al., 1999).
The vegetation communities forming the landscape mosaic are determined by soil types, grazing
pressure, and climate. Community composition can be used to identify disturbance and estimate site
degradation. Weixelman et al. (1997) identified plant community compositions that corresponded to
range condition of the site. A gradient was determined ranging from a grass dominated state being
the most desirable, to a grass/forb/shrub state of low productivity.
Overgrazing changes the soil moisture
capacity due to the reduction of basal
cover (Figure 1). Less vegetative soil
cover results in increased
evaporation. Absence of plants
increases the runoff potential and also
decreases the permeability of the soil
from lack of plant rooting and soil
compaction. More drought tolerant
plants such as shrubs displace grass
species and the area begins to convert
from a grassland to a woodland, and
from a stable soil environment to an
erosional landscape (Freidel, 1991).
The conversion of grassland to a
shrub dominated system increases the
spatial heterogeneity of water and
nutrients in the soil (Fuls, 1992).
Such changes are often irreversible in
arid and semi-arid ecosystems.
Watershed impacts: Watershed
function (soil moisture storage, water
table stability, percolation) and water
quality is impacted by rangeland
m
mm
assies
Figure 1. Rill and gully formation due to soil erosion (National
Research Council, 1994).
4
-------
topsoil loss and erosion. Streams in arid and semi-arid environments produce the highest sediment
yields over the runoff season (Langbein and Schumm, 1958). Infrequent rainfall creates sparse
vegetation distributions. This natural lack of vegetation allows for greater runoff and sediment yield
in grassland dominated ecosystems than in forested regions. These environments usually have
intense rainfall events, which increases the movement of soil particles off the soil surface (Gordon et
al., 1992). Soil stability directly influences the condition of rangeland watersheds (Branson et al.,
1981). The stability of the soil is determined by the vegetative composition and grazing intensity of
the site, and climatic and landscape variables (Fuls, 1992). Denuding of vegetation by overgrazing
exposes the soil to increased solar radiation and erosion by wind and water. Erosion removes the
surface horizon of the soil, along with most of the soil organic matter and nutrients. Without the
plant canopy, the high incident solar radiation increases the temperature of the soil. Soil
denitrification occurs as the ability of nitrogen oxidizing bacteria declines at soil surface
temperatures above 40°C (Van Wambeke, 1992). Loss of the surface horizon and organic matter
causes a crusting of the soil surface, preventing water infiltration and seed germination (Brady,
1990).
Streambank erosion is a natural and continuous process in stream evolution. The banks of streams
can become unstable if the annual amounts of sediment and organic debris are reduced. Periods of
greatly increased sediment input can alter stream morphology and instream biotic ecology. Streams
in semi-arid regions have greater variation in sediment loads than streams in humid environments
(Nordin, 1985). Excessive erosion produces sediment in quantities that can severely impact the
environmental quality of the receiving waters. Sediment particles contain nutrients that may cause
eutrophication of downstream water bodies. The eutrophication may be less of a problem waters
receiving sediments from nutrient poor soils of the arid and semi-arid western US. The
sedimentation of streams and lakes results in increased turbidity, streambed blanketing, habitat
destruction, and other deleterious impacts (Pierzynski et al., 1994). High turbidity reduces the
photosynthetic ability of aquatic plants, disturbing the energy flow of the system. Voluminous
sediments irritate fish gills, blanket spawning beds, and reduce visibility (Rabeni and Smale, 1995).
Long term accumulation of sediment can foul water supply intakes and fill in navigable waterways.
Development of Ecological Indicators to Assess Rangeland Condition
Need: The National Research Council in 1994 determined that all attempts at national-level assessment
revealed that significant degradation is occurring on a majority of rangelands in the United States.
The council also determined that despite the evidence of historical and current rangeland
degradation, inadequate data on current rangeland conditions exist. To add to the problem, there is
no consensus between managing agencies on how to assess and monitor the status and health of these
resources (National Research Council, 1994). Further, the ability to measure rangeland health
reliably and efficiently, and in a systematic manner, over the western U.S. is currently lacking.
Extensive monitoring of rangeland condition is inhibited by several factors; debates over the methods
employed for rangeland assessment, the cost of large scale measurements and research, and a lack of
proven technology for variable quantification.
A coordinated federal effort is required to develop, test, and standardize indicators and methods for
inventorying and monitoring rangeland health (National Research Council, 1994). This effort should
be coordinated with the federal environmental and agricultural programs to:
1. Develop a set of indicators for a minimum dataset for inventorying and monitoring rangeland
health.
5
-------
2. Standardize methods of measuring indicators and categorizing rangelands as healthy, at risk,
or unhealthy
3. Validate the indicators and methods selected by field testing
4. Quantify the correlation between measures of rangeland health and condition
Indicators: Determination of rangeland condition is a complex task, a mixture of judgement and
quantification that requires an interdisciplinary approach to solve. The ecological processes
controlling degradation and recovery are numerous and not fully understood (Teuller, 1973, Glenn-
Lewin and van der Maarel, 1992). Ecological indicators relating to rangeland condition can be
measured and used to understand the status or health of the ecosystem. Proper selection and
measurement of indicators can tell the manager whether an ecosystem is healthy, at risk of
deterioration, or unhealthy. Defining rangeland condition this way requires the determination of
boundaries between healthy, at risk, and unhealthy. Selected ecological indicators are used to assess
the state of a rangeland, and long term monitoring will tell if the system is deteriorating to a lower
state, or recovering to a higher state. By understanding the relationship between the ecological
indicators and the state of the ecosystem, thresholds can be developed to define the boundary
between at risk and unhealthy conditions. These boundaries are significant since they represent a
possible irreversible shift to a state of lowest productivity and environmental quality (National
Research Council, 1994).
Indicators recognized by the National Research Council (1994) are established around three criteria;
1) soil stability and watershed function, 2) nutrient cycling and energy flow, and 3) recovery
mechanisms. These criteria encompass a host of indicators relating to the overall condition of a
given rangeland (Table 1). All three criteria are interconnected, and one by itself cannot be used to
determine an area's ecological status.
Table 1. Criteria and Indicators of Rangeland Health (Adapted from National Research
Council, 1994)
Criteria
Indicators
Soil Stability and Watershed Function
A-horizon present
Rills and gullies
Pedestaling
Scour or sheet erosion
Sedimentation of dunes
nutrient cycling and energy flow
Distribution of plants
Litter distribution
Rooting depth
Photosynthetic period
recovery mechanisms
Plant age-class distribution
Plant vigor
Germination and presence of microsites
Use of Remote Sensing to Develop Rangeland Ecological Indicators
Satellite and airborne remote sensing has been a powerful tool for the analysis of earth features. Remote
sensing data has been used to solve complex environmental problems, investigate the impacts for global
6
-------
change, identify archeological sites, and solve geopolitical-political problems (Lillesand and Kiefer,
1994). Ecological and environmental applications often involve point or plot measurements taken in the
field. Many resources vary in space and time, and sampling these variations is difficult (Lyon et al.,
1992). Remote sensing affords the user the opportunity to sample multiple dates and extrapolate point
measurements collected in the field over large spatial scales (Lyon and McCarthy, 1995, Vincent, 1997).
There have been many uses of remote sensing to inventory and assess rangelands (Tueller, 1992).
Landsat TM imagery has been used to estimate rangeland plant productivity, monitor change, and
identify disease and pest outbreaks (Haas, 1992). Airborne multispectral video has been successfully
employed to distinguish plant communities and species and detect drought stress, grazing intensity, and
burned areas (Everitt and Escobar, 1992).
The use of remote sensing data as input to an ecological model is an important next step for rangeland
assessment and monitoring. Landscape ecological models require spatially-intensive data (Lobo et al.,
1998). Remote sensing provides this information, often at the fine spatial resolution and large areal
extents needed for most landscape studies (Frohn, 1997).
Green
Visible
Remote sensing systems: Remote sensing systems operate by observing the interactions of
electromagnetic (EM) radiation with materials on the surface of the earth (Figure 2) (Vincent, 1997).
These systems work by either
passively collecting light energy
reflected or emitted off the earth's
surface, or by actively
transmitting EM radiation and
collecting the reflected or scattered
energy from the source. Both
active and passive
systems record
different information
regarding the physical
and chemical pro-
perties of a material.
These systems can
complement each
other greatly by y-Rays
simply gathering
more information on a
material than one system
could do alone.
-6 -5
10 10
104 103
io2
io"1
a
10 10 10
4 5
10 10
6 7 8
10 10 10
J 1
X-RaysJ
^.Thermal
Infrared
t
Liyii
L-
TV/Radio
Microwave
.Near & Mid
Infrared
Ultraviolet -1
Figure 2. The electromagnetic spectrum. Scale in micrometers (|jm).
7
-------
Section 3
Imaging Spectroscopy
Materials such as minerals, vegetation, and man-made substances have characteristic reflectance spectra
caused by the absorption and emission of impinging EM energy. Multispectral systems are not designed
to sample a large portion of the EM spectrum at a fine detail, and may miss important information
contained in the non-sampled bands. Diagnostic absorption features that characterize materials often
occur over a small portion of the spectrum (Clark, 1999). The large, noncontinuous bands of sensors
such as Landsat, SPOT, ASTER, and others are not able to detect these subtle features.
Hyperspectral sensors are designed to sample, or image a large segment to the EM spectrum, usually
visible to near Infra-Red (NIR) at a fine spectral resolution (Green et al., 1990). These systems are
termed imaging spectrometers since they collect data in ways similar to laboratory reflectance
spectrometers. Imaging spectrometers can see absorption bands not detectable by other systems, making
them the preferred sensor for accurate material identification (Clark, 1997).
Spectral reflectance of vegetation: Vegetation has a very characteristic spectral reflectance, or
signature caused by chlorophyll absorption of a segment of the EM spectrum known as
photosynthetic active radiation (PAR). The two chlorophyll types, a and b, absorb light energy at
0.43 - 0.66 |im and 0.45 - 0.64 nm respectively (Schanda, 1986). Leaf chemistry also produces
spectral features that can vary according to plant species and health. Absorption features associated
with foliar nitrogen, lignin, and cellulose concentration can be diagnostic (Martin et al., 1998).
Water is a strong absorbent of light and these absorption bands can be related to leaf moisture content
(Goetz, 1983).
Discrimination of plant species using imaging spectroscopy: Vegetation species can be
characterized by plant and leaf structure and foliar chemical composition. These components affect
the absorption of light in various ways (Figure 3). The physical properties of the plant canopy, such
as leaf size (leaf area index), orientation (leaf angle distribution), and woody stem abundance relates
to overall canopy spectral reflectance (Ramsey and Jenson, 1995). These properties, together with
leaf chemistry, create a spectral "fingerprint" which is used to identify plant species and physiology
(Martin et al., 1998).
The diagnostic spectral features produced by the physical and chemical properties of vegetation
occur over discrete wavelengths. These features can be resolved by removal of the continuum in the
spectra (Clark and Roush, 1984). Two components make up the diffuse reflectance spectrum; the
continuum and the actual absorption features. The continuum can be thought of as the background.
Features of interest are superimposed onto this background absorption (Kokaly and Clark, 1999).
The diffuse reflectance spectrum of a sample generally has an increasing slope due to a number of
effects such as scattering, the additive effect of optical constants, and Beers Law (Clark et al., 1990,
8
-------
Clark, 1981, Morris et al., 1982). Removal of the continuum isolates the absorption band center and
allows for these features to be easily compared with other diffuse reflectance spectra (Figure 4).
Imaging spectroscopy can detect these subtle features, which allows for vegetation species mapping,
determination of plant health, and by collecting this type of data over a long period of time, detection
of environmental change (Kokaly and Clark, 1999).
Visible to Near IR Diffuse Reflectance
0)
o
c
<9
o
0)
©
a:
o
u>
.q
<
0.6
0.4
0.2
1 I 1 1 1
/oA\'
. 1
A j '
II /
/ / /f
/ f / '
/y \ <
(/j ^
•
\ t
v
//
1 . . . 1 , , , 1
Sage_Brush
Cheatgrass
Pinon_Pine
Grassland Soil
0.5 1.0 1.5 2.0
Wavelength
2.5
Figure 3. Spectral reflectance of common rangeland vegetation and
soils. Data courtesy of USGS Speclab.
Continuum-removed Spectra
Sage_Brush
Cheatgrass
Pinon Pine
-- 0.67
i i .
0.40
0.80
0.50 0.60 0.70
Wavelength
Figure 4. Comparison of vegetation species by band center analysis.
9
-------
Remote sensing of vegetation in arid and semi-arid regions has proven to be challenging due to the
effect of soil-vegetation mixtures within pixels (Huete et al., 1985). The low vegetation cover of
these ecosystems together with the high spectral variability of the soils creates problems when trying
to employ standard image analysis techniques such as vegetation indices (Elvidge and Chen, 1995).
The spectral signature of these mixed pixels is so complex as to require the use of data of high
spectral dimensionality to accurately determine the components occurring within the mixtures (Okin
et al., 1998). Spectral unmixing techniques can be applied to hyperspectral data to analyze multiple
component pixels. Methods such as linear spectral mixture analysis use spectral end-members,
common components occurring within the scene that account for most of the image's spectral
variability. The assumption is that most of the pixels in the scene contain some proportion of these
end-members (Mustard and Sunshine, 1999). The mixed spectrum in a pixel is a linear combination
of the "pure" end-member spectra, weighted for their fractional abundance within the pixel (Kruze,
1999). Mixture modeling has been successfully used to map sparse vegetation, soils, and minerals in
arid and semi-arid environments (Drake et al., 1999; Okin et al., 1998; Roberts et al., 1993).
10
-------
Section 4
Imaging Radar
Advanced radar systems have great promise for use in ecological studies (Kasischke, 1997). Radar
systems interact with the earth's surface differently than sensors that work in the visible and NIR spectral
region. Radar systems actively illuminate a scene with microwave energy and collect the reflected and
scattered return of this energy (Lillesand and Kiefer, 1994). Microwave radiation is scattered according
to the sensor's wavelength of operation. Objects whose height is greater than Va cos 0 (where 0 is the
angle between the sensor and the target) of the operating wavelength are diffuse reflectors to radar and
are able to scatter energy back to the sensor. The information collected by radar systems relate to earth
surface texture and dielectric variations caused by the interaction of microwave energy with conductive
and resistive materials (Ulaby et al., 1982).
Traditional radar systems transmit and receive in one wavelength only. Advanced systems, such as
AIRSAR can work in 3 frequencies, and thus image more of the EM spectrum than single frequency
sensors. Microwave radiation is transmitted in a linear polarized wave. Most systems transmit and
receive in a single specific polarization, either vertical or horizontal. AIRSAR works in a quad
polarization mode by transmitting and receiving in both vertical and horizontal polarizations (Van Zyl et
al., 1992). More information regarding surface characteristics can be collected in this mode.
Mathematical models can be employed to analyze multifrequency polarimetric radar data to estimate a
materials physical properties (Van Zyl, 1992).
Development of Landscape Ecological Indicators
Using Fine Spectral and Spatial Resolution Sensors
The National Research Council (1994) recommended the development of landscape indicators for the
evaluation of rangeland condition. These indicators are to be adopted by agencies responsible for the
management of public and private rangelands to create a uniform assessment and monitoring system.
Remote sensing is aptly suited for the identification of the components related to rangeland indicators
(Haas et al., 1983). Specifically, the use of imaging spectrometry and radar can be used to accurately and
reliably identify these components and, through the use of GIS, develop the rangeland ecological
indicators necessary for efficient and robust assessments (Kasischke et al., 1997; Smith et al., 1995).
Determination of ecological indicators: Vegetation community structure, ecological change, and
landscape pattern can be determined using various landscape metrics (Washington-Allen, 1999). The
use of hyperspectral data for the derivation of landscape metrics will an innovative use of this
technology. The vegetation community structure for a rangeland can be related to the area's
disturbances, such as grazing, drought, fire (Fuls, 1992). Invasion of weedy species changes the
community structure of a system and can indicate site degradation (Mouat et al., 1997). The ratio of
perennial versus annual species, or native versus introduced species as a percent of the total
11
-------
vegetative coverage of an area is an important indicator of rangeland condition (Mouat et al., 1997).
The fragmentation, or patchiness of an area caused by the introduction of weedy plants can be
determined using landscape metrics (Frohn, 1997). The fine spatial resolution of the AVIRIS low
altitude deployment can provide vegetation mapping at a resolution desirable for landscape metric
analysis.
Soil mapping and soil assessment: Soils form a continuum across the landscape and do not occur as
discrete homogenous units (Odeh et al., 1992). Remote sensing can provide various levels of
information concerning soil properties depending on spatial scale and spectral resolution of the
system (Wessman, 1990). For many decades the USDA Natural Resource Conservation Agency's
Soil Survey has used aerial photography to aid in mapping soils (Soil Survey Staff, 1951). However,
the use of multispectral image classification to map soils is problematic due to the complexity of the
soil mixtures (Huete, 1986). Hyperspectral data can be used to evaluate these complex signatures.
The relative contributions of different soil physical and chemical properties can be analyzed using
mixture modeling (Boardman et al., 1995). Ahn, (1999) used linear unmixing analysis of
hyperspectral imagery together with geostatistics and fuzzy clustering to create boundary maps
showing delineations between soil class memberships. These soil pattern maps can be integrated
using a GIS with other spatial ancillary data to create true soil maps. A comprehensive collection of
soil spectra has been created by Stoner and Baumgardner (1980, 1981).
Soil classification: Soils are three-dimensional components of the landscape (Buol et al., 1989). Soil
classification is the grouping of soils according to a set objective. This grouping relies on the soil's
morphology, or the specific characteristics of a soil's horizons (Soil Survey Staff, 1975). Most soils,
except those which are newly formed, are made up of distinct horizons that extend from the surface
to often greater than 3 meters below the surface (Brady, 1994). Remote sensing is capable of
imaging only the surface horizon of a soil, but useful information can be derived from this
component of the soil profile. In healthy soils, the greatest amount of biological activity occurs in the
surface, or A horizon of the soil profile. In the arid and semi-arid western regions of the country, the
A-horizon is often very low in organic matter or is absent due to erosion or intense weathering. Soil
groups, or orders (see Buol et al., 1989 for descriptions of soil orders) can be identified in part by the
presence or absence of a unique surface horizon called an epipedon. Soil properties such mineralogy,
structure, and organic matter content are elements of a soil's surface horizon or epipedon. These
components can be quantified using remote sensing. In areas with an absent A-horizon, the surface
presentation of a high clay B-horizon might be an indicator of the reduction or elimination of the
overlying A-horizon.
Soil mineralogy: Soils are a matrix of clay minerals and other geologic and organic materials. Some of
the most common soil minerals are silicates, carbonates, and iron oxides (Brady, 1994). A great deal
of work has been done to characterize the reflectance spectra of minerals (Hunt, 1971a,b). The
matrix can be a simple linear mixture or a complex combination (Clark, 1999). The complexity of
material such as soils requires an imaging spectroscopy approach to properly characterize these
materials (Kruse et al., 1993).
Soil organic matter: Soil organic matter (OM), a complex carbon-rich material characterized by several
organic acids, is important to soil biological productivity and health (Sposito, 1989). Wessman
(1991) discusses the effect of soil organic matter on the spectral response of soils. Organic materials
have diagnostic absorption bands caused by the presence of specific functional groups (Weyer,
1985). Soils having high organic matter content appear dark due to these absorptions, and
spectrometers are sensitive to these features (Bigham, 1995). The almost negligible concentrations
12
-------
of OM in arid and semi-arid soils still may be detectable using imaging spectroscopy due to the
significant influence OM has on a soil's spectral response (Stoner and Baumgardner, 1981).
Relating soil parameters to vegetation: The soils of an area has an influence on the occurring
vegetation. Soils parameters such as OM, texture, cation exchange capacity favor one type of
vegetation over another (Buol et al., 1989). Vegetation, in turn, influences the always evolving
formation of a soil. In areas where vegetation has been rapidly removed, the soils of that area can
change rapidly as well due to wind and water erosion, reduction in soil OM, and decreased biological
activity. The type of soil can then be linked to the types of resident vegetation. By accurately
mapping soil types predictions of vegetation occurrence can be done. The information is valuable
for areas that have vegetation in abundances, as a proportion of a pixel relative to the soil
background, too small to be accurately detected using remote sensing. In areas where the vegetation
obscures the soil, information on the linkage between soil and vegetation coverage may be useful for
estimation of the underlying soils.
Salinization: Soil with high salinity can be toxic to vegetation and is an indicator of soil degradation
(Logan, 1992). Areas receiving irrigation water are prone to salinization due to the high electrolytic
concentration of groundwater. Through evaporation, the salts present in the irrigation water are
deposited in the soil where it accumulates. The presence of salt minerals have an influence on the
spectra of the soil (Csillag et al., 1993). In addition, the unique dielectric properties of salts are
detectable using imaging radar (Kierein-Young, 1997).
Soil crusting: The absence of vegetative cover increases the oxidation of soil organic matter and leads
to topsoil loss. Crusting occurs when the topmost layer of soil, when dry, becomes harder than the
layer beneath it. Soil crust formation is caused by the loss of fine soil particles under the impact of
raindrops (van Wambeke, 1992). The surface layer becomes aggregated and impenetrable to
moisture and plant seedlings. Re-vegetation of the area becomes difficult without intervention. This
condition can lead to severe erosion and site degradation. The textural variation between crust
formations and the surrounding soils may be detectable by imaging radar and spectroscopy.
Biological soil crusts: Cryptobiotic soil crusts provide improved surface stability, water infiltration, and
fertility in arid ecosystems. These fragile biological crusts are composed of cyanobacteria, lichens
and mosses and are easily disturbed by human, animal and mechanized trampling (Belnap and
Gardner, 1993). Disturbance of these biotic crusts can cause accelerated soil erosion and
degradation. The characteristically long recovery time of arid ecosystems make disruption to these
crusts a near irreversible process. These soil crusts have a characteristic dark color that make them
identifiable using remote sensing instruments. Mapping of microbiotic soils has been accomplished
for an area along the Colorado Plateau using imaging spectroscopy (Kokaly et al., 1994).
Stability: Erosional features indicate the lack of soil stability. Rills and gullies are conduits for sediment
transport, while scouring or sheet erosion indicate large scale soil loss. Sedimentation and dune
formation indicate mass soil movement and site degradation. The drainage patterns formed by rills
and gullies can be interpreted from remote sensing imagery. Radar systems can detect these patterns
by topographical differences and the soil dielectric variations caused by the increased soil moisture
present in these depressions (Ulaby et al., 1982).
13
-------
Ancillary Data
The inclusion of non-remote sensing data, or supporting datasets, improve landscape ecological analysis
and landcover mapping (Mouat et al., 1997). Ancillary data can be used to assist in the image
classification procedure and are used as inputs to spatial models (Williams et al., 1999; Vogelmann,
1998). Riparian habitat was characterized for a watershed in Arizona using a combination of satellite
imagery with water level data, and temperature and precipitation maps (Lee and Marsh, 1995).
Climate: Climate determines the characteristic vegetation communities for an ecosystem (Bailey, 1983).
Vegetation change cannot be understood without knowledge of the climatic context in which a
community evolved. The perceived anthropogenic disturbance may in fact be of natural causes, an
example being the effect of drought on grass dominated regions. Classification of range condition is
outlined by the National Research Council (1994) by mapping and correlating soil and climatic
information with a sites vegetation community. A vegetation species occurring out of its climatic
context may be an indicator of site degradation. Introduced, or invasive weedy species are such
candidates.
Ground measurements: Field measurements are essential to ecological indicator development
(Ringrose et al., 1994). Range research stations provide long-term measurements of important
ecological variables which can be used with remote sensing data (Washington-Allen et al., 1999).
Field based spectral measurements of rangeland vegetation cover is essential for determination of the
vegetative components using overhead imagery (Bork et al., 1999a). Ground spectral measurements
are required for several steps in the image processing procedure, especially when imaging
spectroscopy data is calibrated (Clark, et al., 1999). Field based measurements are also used to
establish a gradient of conditions to test ecological indicator and remote sensor sensitivity.
Grazing plot intensity, duration and season of use records and other land use data: Information
on grazing area intensity of use, duration of use, and season of use are important in determining
grazing pressure. This type of data can be used similar to census data to aid in regional assessments.
Experimental rangelands and watershed stations often have long-term stocking rate, livestock type,
and season of use records that can be used for site specific and regional studies.
Topography: The topography of the landscape influences the vegetation community structure (Bailey,
1983). Mouat (1997) used slope length along with soil morphology and erodibility as components of
an erosion hazard indicator. Slope, aspect, and elevation are important attributes of soil formation as
well (Buol, 1992).
Hypothesis
The observations described above lead to the following hypotheses:
1. Arid and semi-arid shrub, forb and graminoid vegetation have distinct and separate chemical and
morphological features that can be detected using imaging spectroscopy and radar. These features
allow for accurate vegetation mapping at the species level.
2. Landscape erosional features such as gullies and scouring can be detected using spectral and
microwave remote sensing, especially systems having fine spatial resolutions.
14
-------
3. Soil properties such as mineralogy, organic matter content, texture, surface horizon determination,
and other features can be determined using imaging spectroscopy and radar. Basic soil mapping is
possible.
4. Ancillary spatial and non-spatial data can be used with remote sensing data to reduce the uncertainty
as well as enhance image classification and provide modifier variables in model development.
Degraded, disturbed and functioning systems can be identified by modeling the interactions between
these variables.
5. Rangeland ecological indicators can be determined using remote sensing allowing for large scale
assessments of resource conditions.
Conceptual Basis for Hypothesis
The following is a listing of the rangeland indicators that are predicted to be collected using remote
sensing. The model to be developed will use this evaluation matrix as its structure. The model's
equations will use various indicators as variables and weighting factors (Table 2).
Table 2. Indicator Development Adapted from the Rangeland Health Evaluation Matrix (National
Research Council, 1994)
Indicator
Component to be Measured
Sensor
Soil Stability and Watershed
Function
IS = imaging spectroscopy
IR = imaging radar
Soil A horizon
Soil organic matter
IS imagery to determine soil
organic matter content and
mineralogy
Texture
IR to estimate soil texture
Mineralogy
IR to determine erosional features
Rills and gullies
Landscape drainage patterns
IS for drainage pattern
characterization
Scouring or sheet erosion
Elimination of soil surface
horizon
IS to identify presence/absence of
soil horizination
Sedimentation
Soil accumulation, dune
formation
IS/IR to identify sediment deposits
Distribution of Nutrient
Cycling and Energy Flow
Distribution of plants
Landscape fragmentation using
pattern metrics
IS/IR to map plant species
Litter distribution and
incorporation
Above ground detrital biomass
IS for leaf litter estimation
Recovery Systems
Age-class distribution
Plant species identification and
age estimation
IS/IR to determine plant species
and height for age estimation
Plant vigor
Determine photosynthetic
efficiency
IS to determine plant chlorophyll
absorption
Germination
Soil crust formation and areas of
soil translocation
IS/IR to identify soil organic matter
depletion and crusting
15
-------
Technical Approach
Methodology
General approach: The complex nature of rangeland ecological research requires an interdisciplinary
approach. Multiple agencies involved in research and management of rangelands such as USGS,
BLM, and USDA will be asked to enlist in this effort. Accessing the considerable knowledge of
these groups will help ensure a better understanding of the nature of the ecological interactions
occurring in these ecosystems and identify gradients in ecological condition. Cooperation with
pertinent field research stations and scientific personnel will be important to this study. The research
will consist of field data collection, laboratory analysis of collected samples, and remote sensing data
collection of hyperspectral and polarimetric radar imagery (Bork et al., 1999b; Smith et al., 1995).
Ancillary data such as climatic information, animal stocking rates, and spatial data (DEM's, DLG's)
will be merged with the remote sensing data products. Modeling of data will be done in a GIS
framework in order to develop the landscape indicators required for rangeland assessments and
monitoring.
This research will be conducted in a two-phase approach. A pilot study will be done on several areas
that have recent and historical imagery. Sites may be part of or adjacent to field research stations or
part of a research study such as the US EPA's Environmental Monitoring and Assessment Program
(EMAP). The mission of EMAP is to provide documentation on the current condition of the
Nation's ecological resources, why that condition exists, and predict future ecological conditions
(EPA, 1990). This will allow for the collection of good representative data and will provide
assistance in the data analysis effort. Sites to be included in the study will encompass a range of
ecological conditions or rangeland states from healthy to at risk to degraded. Landscape indicators
will then be developed across this gradient of rangeland conditions using remote sensing data and
existing data and knowledge that will be obtained from other rangeland management agencies.
The second phase will be an extrapolation of the pilot study to a larger area of the western United
States. Subsets of the imagery collected during normal operations by government and commercial
data vendors will be added to the study. This imagery will be analyzed according to the methods
developed in the pilot study to evaluate the effectiveness of the model over a wide range of
ecological conditions. It is hoped that near the end of this phase there will be an operational space
borne hyperspectral satellite. This will provide the final segment of this studies objectives, to use
advanced remote sensing systems for large scale assessments of this nation's rangelands.
Subtask 1: The pilot study will be done on two long-term rangeland research sites: the Walnut Gulch
Experimental Watershed, Tombstone, Arizona, and the Sevilleta National Wildlife Refuge, Socorro,
New Mexico. Field data collection will consist of reflectance measurements of soils and vegetation
using EPIC's ASD FR field spectrometer. Soil samples will be collected and analyzed for
mineralogy (using x-ray diffraction [XRD], infra-red spectroscopy [FTIR], and chemical procedures),
texture, organic carbon, major cations and anions, moisture content and other parameters such as
dielectric constants and particle size using published methodologies (Klute, 1986). Additional
laboratory soil and vegetation reflectance measurements will be obtained using a suite of
spectrometers available at the USGS Spectroscopy labs in Reston, Virginia and Denver, Colorado. If
a soil survey of the site has not previously been accomplished, then one will be conducted to
determine the soil orders and other landscape information occurring at the site. Vegetation species
identification will be accomplished by range scientists. Plant chlorophyll measurements will also be
taken. Topographic, climatic, census, and other ancillary data will be compiled for each site. The
16
-------
remote sensing datasets will be correlated with ground measurements obtained at the study sites.
Model development will be accomplished in this phase using spatial and nonspatial statistical
methods.
Remote sensing data for the study sites will be obtained using current and historical sources. The
HyVista HyMap system will be the primary system and the NASA/JPL AVIRIS and AIRSAR
sensors will be used if available. Other aerial photography, airborne video such as Digital
MultiSpectral Video (DSMV), and medium and high resolution multispectral imagery (Landsat,
SPOT, and IKONOS) will be obtained as necessary. An accuracy assessment of the remote sensing
derived products will be accomplished using the methodologies outlined in Congalton and Green
(1998).
Subtask 2: This phase of the study will test the rangeland assessment model's accuracy and seek to
develop additional model parameters using two-three other long-term rangeland research sites in the
southwestern United States, the Muleshoe Ranch Cooperative Management Area, Willcox,
Arizona, the Jornada Experimental Range, Las Cruces, New Mexico, and the Desert Experimental
Range in Pine Valley, Utah. Other sites may be added if resources permit. These sites are the
Amenderas Ranch, Truth or Consequences, New Mexico and Vermejo Park Ranch near Raton, New
Mexico and the Desert Ranch in Utah. Selected study sites in this phase will be characterized using
the methods developed in subtask 1. Rigorous model calibration and testing using these additional
study sites will occur in this phase.
The rangeland ecological indicators developed in subtask 1 and 2 will be applied over a gradient of
conditions in many additional areas as remote sensing data for these regions becomes available. The
landscape indicator model for these areas will rely on remote sensing data primarily. Collection of
field data will be used for solely for accuracy assessment purposes. This final iteration of the model
will set the stage for an operational methodology for rangeland assessments using the forthcoming
hyperspectral satellites due near the end of this research study.
Data Acquisition and Planning
Hyperspectral data will acquired for the pilot study using the HyMap system obtained under a group
shoot scenario where the mobilization costs are shared by multiply federal, and university entities.
NASA AVIRIS data collection will be obtained if resources permit. The pilot study sites are long term
ecological research (LTER) stations and the federal agencies and universities in charge of the site have
obtained AVIRIS data. Collaborative arrangements have been set up with these partners for data sharing
and cooperative research. Planning for hyperspectral data acquisition, whether AVIRIS or HyMap must
be done several years in advance. Strategic plans are in place to coordinate airborne and ground based
data collections for the pilot and follow-on study areas.
Data Analysis and Interpretation
Imaging spectroscopy data will be analyzed using techniques outlined by Clark (1997), Boardman (1995)
and others (see reference list). Surface reflectance calibration of the imaging spectrometer data will be
accomplished using the steps outlined in Clark (1999). Briefly, a radiative transfer algorithm is applied
to the data to remove atmospheric absorptions and Rayleigh and aerosol scattering from the data. Next,
ground spectral measurements using a portable field spectrometer of areas imaged by the sensor is used
to correct the overflight data. Highly reflective and spectrally uniform areas occurring within the
flightlines of the sensor are used as ground calibration sites. The spectra of these sites are collected by
17
-------
the field spectrometer and are compared to the imagery and corrections to the image data are made
through the use of multipliers to achieve accurate reflectance signatures of materials in the image.
Materials of interest such as soils, minerals, and vegetation will be identified and mapped by comparing
the imaging spectrometer reflectance data to in situ ground measurements, standards contained in
spectral libraries, and field samples analyzed using laboratory spectrometers. Multiple endmember
analysis techniques will be applied to the imagery to detect and map whole pixel and sub-pixel
occurrences of target materials (Okin et al., 1998). The steps in the image processing and analysis are as
follows (Kruse, 1999); 1) imaging spectrometer data reduction using a minimum noise fraction (MNF)
transformation; 2) spatial data reduction using a pixel purity index (PPI); 3) endmember extraction using
n-dimensional visualization; 4) spectral identification of target materials by comparison of the image
derived reflectance data to in situ measurements and known standards; 5) mapping of materials using the
mixture tuned matched filtering (MTMF) method (Boardman, 1998; Kruze, 1996).
Analysis of soil and vegetation interactions will be accomplished using spectral and GIS modeling. The
soil/vegetation spectra will be analyzed by testing linear and non-linear models and accuracy assessing
the results. Probabilistic maps of vegetation will be generated using soil information derived from
remotely sensed data, field data, and historical ecological information. Pixel in the hyperspectral
imagery having mixed spectra will be compared to the vegetation probability maps to assist in the image
classification process. This will allow for the identification of possible vegetation species occurring in
the pixel. Once the species is estimated, its spectral contribution to the whole pixel spectra can be
determined. The vegetation spectra can then be de-convolved from the whole pixel signature and the
abundance of vegetation in the pixel can be calculated (Huguenin et al., 1997).
Software tools such as ENVI, Grams32, Tetracorder, SpecPR will be used to process imagery and map
target materials. Landscape metrics will be developed using GIS tools such as ESRI ARC/INFO and
ArcView Spatial Analyst, and Erdas Imagine. Modeling of data to develop landscape indicators will be
accomplished using SAS and Systat statistical software packages.
Radar imagery will be processed and analyzed using an approach developed by van Zyl (1989). Radar
backscatter modeling will be accomplished using the methods outlined in Taylor et al. (1996),
Kierein-Young (1993) and Kierein-Young and Kruse (1997). Radar fusion with hyperspectral imagery
will use the method described in Kierein-Young (1997).
18
-------
Section 5
Potential for Reducing Uncertainty in Exposure Assessments
and Importance to the Milestones and Program Area Goals
This research reduces uncertainty in conducting assessments of rangeland ecological conditions by
coordinating research efforts with agencies responsible for rangeland management with EPA scientific
and field personnel. The influence of healthy, at risk, or degraded rangelands on the water quality of the
rangeland watersheds can be more accurately determined. This approach allows remotely sensed
measurements to be compared against field samples measured with state-of-the-art analytical methods
that are, by law, designed to withstand legal scrutiny. Further, when spectra of known chemicals or
vegetation species are obtained, they will be compared against existing library spectra as a method of
reducing uncertainty in assessments. This is central to milestones of this research effort. By
coordinating closely with site-specific operations, especially in situ sample analysis, this research can
enjoy a level of quality control and reduction of uncertainty that driven by operational requirements.
The EPA, to better account for the success of its actions, has developed a cascading set of goals,
objectives, subobjectives, milestones, measures, tasks, and products in compliance with the Government
Performance Results Act (GPRA). There are currently ten longer-term goals for the EPA under the
GPRA. Goal 8, "Provide sound science to improve the understanding of environmental risk, and develop
and implement innovative approaches for current and future environmental problems" serves as the
foundation, or core of the ORD's Ecological Research Program. The specific objective associated with
ORD's ecological research under this "Sound Science" goal is to provide the scientific understanding to
measure, model, maintain, or restore, at multiple scales the integrity and sustainability of ecosystems
now, and in the future.
In addition, the ORD's "Ecological Research Strategy" identifies major objectives, sub-objectives and
products associated with its core research program areas of:
• Ecosystem monitoring research
• Ecological processes and modeling research
• Ecological risk assessment research
• Ecosystem risk management restoration research
A shorter-term accounting of success is accomplished by establishing, and monitoring the response to the
annual performance goals (APGs) and measures (APMs) under GPRA and progress toward completion
of any additional critical research products identified in the ORD's "Ecological Research Strategy" and
its subsequent updates. These goals and measures provide the "why" and the "what" of our research
tasks and projects. This document, as a technical research plan addresses not only the "why" and the
19
-------
"what," but also the "how" - the approach to providing products that satisfy the specific performance
goals associated with this activity. Those specific annual performance goals and measures are:
• Review Article: Rangeland ecological indicator development using advanced remote sensing
technologies
• Field Data Collection, Number of Sites Visited
• Hyperspectral Remote Sensing Imagery Acquired
• Reports, Papers, Journal Articles
20
-------
Section 6
Quality Assurance Statement
This research will be conducted in accordance with the Quality Management Plan for the Environmental
Sciences Division (ESD), National Exposure Research Laboratory, Las Vegas. A Quality Assurance
(QA) Project Plan will be prepared prior to the initiation of field activities and data processing. The QA
Project Plan will document: (1) questions to be answered or decisions to be made based upon study data;
(2) The nature, number and quality of data points needed to achieve a selected level of confidence in
those decisions; (3) the experimental design and methods necessary to meet those data objectives (EPA
1999).
Also, the Environmental Photographic Interpretation Center (EPIC) has a Master Quality Assurance
Project Plan in place and has developed a full set of 53 Standard Operating Procedures (SOP) for many
aspects of photographic and digital imagery acquisition, scanning, processing, analysis, and graphics
(Lockheed 1999). Specific quality assurance measures for this research include:
1) The use of the manufacturers calibration systems and reports for field spectrometer data
collection (Beal, 1997).
2) The Use of a Spectralon Standard white reference for field spectrometer data collection (ASD
1997).
3) The averaging of multiple spectral measurements for all field data collection (Clark et al., 1999).
4) Comparison of collected spectra with that of known spectra from existing spectral libraries as
developed by USGS, NASA and others.
5) The use of accepted calibration and atmospheric removal algorithms for all hyperspectral remote
sensing data (Clark et al., 1999).
6) The comparison of field and remote sensing spectra with in situ samples collected and analyzed
by conventional analytical chemistry methods.
7) The publication of all significant results in peer-reviewed scientific journals.
Anticipated Results/Milestones
A standardized, effective, and accurate method for determining the state of this nation's rangelands is
needed. This research will provide rangeland managers with the tools they need to assess and monitor
these resources. The long-term goal of this work is to develop a methodology using current technologies
for use with the forthcoming hyperspectral satellites due in the next 2 to 3 years.
21
-------
This research will enhance other ESD research efforts that deal with the development of landscape
indicators and those that deal with ecosystem assessments in the western EMAP regions. Specifically,
the results of this work are directly related to GPRA goal 8.1.1, Development of Landscape Indicators for
Use in Regional Risk Assessments and its following subtasks: H - EMAP Western Landscape Pilot and J
- Quantification of Landscape Indicators/Watershed Conditions Relationships in a Semiarid Watershed.
Results and reporting: The proposed working relationship with rangeland management agencies will
allow for the dissemination of results and model improvements. All results will be reported through
a series of internal EPA reports, symposium poster presentations and proceedings, and journal
articles in peer-reviewed scientific journals.
Milestones
1. Research Plan: A fully developed and peer-reviewed research plan will be developed and in
place by 12/01/00.
2. Literature Review: A literature review of the use of imaging spectroscopy and radar for
rangeland assessments will be completed by 3/01/01.
3. Field Data: Selected sites will be visited and spectral data Spectral data will be collected with
EPIC's ASD FR Spectrometer. Ongoing.
4. Remote Sensing Data Collection: Working with NASA and commercial vendors, hyperspectral
and radar imagery will be acquired over many of the study areas. Ongoing.
5. Research Partnerships: Developing cooperative partnerships with Federal Agencies responsible
for rangeland management and research. Also, developing cooperative partnerships with Federal
Agencies and commercial remote sensing providers for the acquisition and analysis of
hyperspectral data for environmental issues. Ongoing.
6. Research Results: Developed on a site by site basis, research results will be reported as internal
reports, symposium papers and peer-reviewed journal articles. As needed.
Deliverables
1. Preliminary guide to the development of landscape indicators of rangeland condition utilizing
remote sensing, primarily imaging spectroscopy or hyperspectral data, and ancillary data.
2. GIS model for determining rangeland condition using indicators derived from research results.
This model would run on ArcView and let users choose data layers such as soil types, vegetation
species, areas of soil erosion. Users may also assign weights to the layers for statistical analysis.
Developed indicators may also be chosen and operated upon using statistical weights. These
indicators include; vegetation patch metrics, exotic vs. native vegetation ratios, percent
grass/forb/shrub species per unit area (area units to be determined).
3. Regional assessments using the GIS model. Rangeland scientists involved in the project will be
accomplishing the analysis and deriving the results. The regions used in the assessments first be
the areas in the subtasks. Additional regions will be accomplished as more hyperspectral data
becomes available.
22
-------
References
Ahn, C-W., M.F. Baumgardner, and L.L. Biehl. 1999. Delineation of soil variability using geostatistics
and fuzzy clustering analyses of hyperspectral data. Soil Sci. Soc. Am. J. 63:142-150.
Anderson, J.E., and K.E. Holte. 1981. Vegetation change over 25 years without grazing on sagebrush-
dominated rangeland in southeastern Idaho. J. Range Manage., 34(l):25-29.
Bailey, R. G. 1983. Delineation of ecosystem regions. Environ. Manage. 7: 365-373.
Bakker, J.P., J. de Leeuw, and S.E. von Wieren. 1983. Micro-patterns in grassland vegetation created and
sustained by sheep. Vegetatio, 55:153-161.
Belnap, J., and J.S. Gardener. 1993. Soil microstructure in soils of the Colorado Plateau: the role of the
cyanobacterium Microcoleus vaginatus. Great Basin Naturalist, 53:40-47.
Bigham, J.M., D.F. Post, F.S. Jones, and M.E. Konen. 1993. Relationship between color and organic
carbon in soils in western Ohio. Agron. Abst. 85:294.
Boardman, J.W., Kruse, F. A., and Green, R. O., 1995, Mapping target signatures via partial unmixing of
AVIRIS data: in Summaries, Fifth JPL Airborne Earth Science Workshop, JPL Publication 95-1,
v. 1., p. 23 - 26. U.S. Gov. Print. Office, Washington, DC.
Boardman, Joseph, 1998, Leveraging the high dimensionality of AVIRIS data for improved sub-pixel
target unmixing and rejection of false positives: mixture tuned matched filtering, in: Summaries of
the Seventh Annual JPL Airborne Geoscience Workshop, Pasadena, CA.
Bork, E. W., N.E. West, and K.P. Price. 1999. Calibration of broad and narrow-band spectral variables
for rangeland cover component quantification. Int. J. Remote Sens. 20(18):3641-3662.
Bork, E. W., N.E. West, K.P. Price and J.W. Walker. 1999. Rangeland cover component using broad
(TM) and narrow-band (1.4 NM) spectrometry. J. Range Manage. 52:249-257.
Brady, N.C. 1990. The nature and properties of soils. Macmillan, New York. USA.
Branson, F.A., G.F. Gifford, K.G. Renard, and R.F. Hadley. 1981. In: E.H. Reid (ed), Rangeland
hydrology. Kendall/Hunt, Dubuque, LA.
Buol, S.W, F.D. Hole, and R.J. McCracken. 1989. Soil genesis and classification. Iowa State Univ. Press.
Ames, Iowa, USA.
23
-------
Clark, R.N. 1981. The spectral reflectance of water-mineral mixtures at low temperatures. J. Geophys.
Res. 86:3074-3086.
Clark, R.N., and T.L. Roush. 1984. Reflectance spectroscopy: quantitative analysis techniques for remote
sensing applications. J. Geophys. Res. 89:6329-6340.
Clark, R.N., T.V.V. King, M. Klejwa, G.A. Swayze, and N. Vergo. 1990. High spectral resolution
reflectance spectroscopy of minerals. J. Geophys. Res. 95:12653-12680.
Clark, R.N., A.J. Gallagher, and G.A. Swayze. 1990. Material absorption band depth mapping of imaging
spectrometer data using a complete band shape least-squares fit with library reference spectra.
Proceed, of the Second Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) Workshop. JPL
Publication 90-54.
Clark, R.N., G.A. Swayze, A. Gallagher, T.V.V. King, and W.M. Calvin. 1993. The U.S. Geological
Survey, Digital Spectral Library: Version 1: 0.2 to 3.0 mm. U.S. Geological Survey, Open File
Report, 93-592.
Clark, RN. 1999. Spectroscopy of rocks and minerals, and principles of spectroscopy. In: A. N. Rencz
(ed), Remote Sensing for the Earth Sciences. American Society for Photogrammetry and Remote
Sensing, John Wiley and Sons Inc. New York, New York, USA.
Clark, R.N., G.A. Swayze, T.V. King, K.E. Livo, R.F. Kokaly, J.B. Dalton, J.S. Vance, B.W. Rockwell,
and R.R. McDougal. 1999. Surface reflectance calibration of terrestrial imaging spectroscopy data: a
tutorial using AVIRIS. US Geological Survey Open File Report (in press).
Cook, C.W. and P.L Sims. 1975. Drought and its relationship to dynamics of primary productivity and
production of grazing animals. In: Proceedings, Evaluation and Mapping of Tropical African
Rangelands. Addis Ababa, Ethiopia.
Congalton, R.G. and K. Green. 1998. Assessing the accuracy of remotely sensed data. CRC Press, Lewis
Publishers, Boca Raton, Florida, USA.
Csillag, F., L. Pasztor, and L. L. Biehl, 1993. Spectral band selection for the characterization of salinity
status of soils. Remote Sens. Environ. 43:231-242.
Drake, N.A., S. Mackin, and J.J. Settle. 1999. Mapping vegetation, soils, and geology in semi-arid
shrublands using spectral matching and mixture modeling of SWIR AVIRIS imagery. Remote Sens.
Environ., 68:12-25.
Elvidge, C.D. and Z. Chen. 1995. Comparison of broad band and narrow band red and near infrared
vegetation indices. Remote Sens. Environ. 54:38-48.
Everitt, J.H. and D.E. Escobar. 1992. Airborne video systems for real time assessment of rangelands.
Geocarto, 7(1): 19-26.
Friedel, M.H. 1991. Range condition assessment and concept of thresholds: a viewpoint. J. Range
Manage., 44:422-426.
24
-------
Frohn, R.C. 1997. Remote sensing for landscape ecology: new metric indicators for monitoring,
modeling, and assessment of ecosystems. CRC Press, Lewis Publishers, Boca Raton, Florida, USA.
Fuls, E.R. 1992. Ecosystem modification created by patch-overgrazing in semi-arid grassland. J. Arid
Environ., 23:59-69.
Glenn-Lewin, D.C, and E. van der Maarel. 1992. Patterns and processes of vegetation dynamics. In: D.C
Glenn-Lewin, R.K. Peet, and T.T. Veblen (eds), Plant Succession: Theory and Prediction. Chapman
and Hall, New York, NY.
Goetz, A.F.H., G. Vane, J.E. Solomon, and B.N. Brock. 1985. Imaging spectrometry for earth remote
sensing. Science, 228:1147-1153.
Gordon, N.D., T.A. McMahon, and B.L. Finlayson. 1992. Steam hydrology: an introduction for
ecologists. John Wiley and Sons, New York, NY.
Green, R.O., J.E. Conel, V. Carrere, C.J. Bruegge, J.S. Margolis, M. Rast, and G. Hoover. 1990.
Determination of the in-flight spectral and radiometric characteristics of the Airborne Visible-Infra-
Red Imaging Spectrometer (AVIRIS). In: Proceedings of the Second Airborne Visible-Infra-Red
Imaging Spectrometer (AVIRIS) Workshop. JPL Publication 90-54.
Haas, R. H. 1992. Landsat thematic mapper products for rangeland assessment. Geocarto, 7(l):27-33.
Haas, R.H., J.A. Newcomer, and E.H. Horvath. 1983. A multitemporal approach for classification and
mapping rangeland vegetation. U.S Geol. Surv. Open-File Rep. 83-877. U.S. Gov. Print. Office,
Washington, DC., USA.
Hatch, G.P., and N.M. Tainton. 1991. A preliminary investigation of area selective grazing in the
Southern Tall Grassveld of Natal. J. Grassland Soc. Southern Africa, 7:238-242. Huete, A.R. 1986.
Separation of soil-plant spectral mixtures by factor analysis. Remote Sens. Environ., 19:237-251.
Huete, A.R., R.D. Jackson, and D.F. Post. 1985. Spectral response of a plant canopy with different soil
backgrounds. Remote Sens. Environ. 17:37-55.
Huguenin, R.L., M.A. Karaska, D. Van Blaricom, J.R. Jensen. 1997. Subpixel classification of bald
cypress and tupelo gum trees in thematic mapper imagery. Photogrammetric Eng. Remote Sens.
63(6):717-725.
Hunt, G.R., J.W. Salisbury, and C.J. Lenhoff. 1971a. Visible and near-infrared spectra of minerals and
rocks IE. Oxides and hydroxides. Mod. Geol., 2:195-205.
Hunt, G.R., J.W. Salisbury, and C.J. Lenhoff. 1971b. Visible and near-infrared spectra of minerals and
rocks III. Sulfides and Sulfates. Mod. Geol., 3:1-4.
Kasischke, E. S., J.M. Melack, and M.C. Dobson. 1997. The use of imaging radars for ecological
applications- a review. Remote Sens. Environ. 59:141-156.
25
-------
Kierein-Young, K. S., 1997. Integration of optical and radar data to characterize mineralogy and
morphology of surfaces in Death Valley, California: International Journal of Remote Sensing, v. 18,
no. 7, p. 1517-1541.
Kierein - Young, K. S. and Kruse, F. A., 1992. Extraction of Quantitative Surface Characteristics from
AIRSAR data for Death Valley, California: Proceedings of the Fourth AIRSAR Workshop, JPL, pp.
46-48. U.S. Gov. Print. Office, Washington, DC.
Klute. A. 1986. Methods of Soil Analysis, Part 1- Physical and Mineralogical Methods. Soil Science
Society of America Book, No 5. American Society of Agronomy, Madison, Wisconsin, USA.
Kokaly, R.F., and R.N. Clark. 1999. Spectroscopic determination of leaf biochemistry using band-depth
analysis of absorption features and stepwise multiple linear regression. Remote Sens. Environ.
67:267-287.
Kokaly, R.F., R.N. Clark, and G.A. Swayze. 1994. Vegetation and cryptobiotic soils mapping in arid
regions. In: N. Milton and J. Lancaster (eds), Spectral Analysis Workshop: The Use of Vegetation as
an Indicator of Environmental Contamination, Desert Research Institute, Reno, Nevada, November
9-10, 1994.
Kruse, F.A., 1999. Visible and infrared: sensors and case studies. In: A. N. Rencz (ed), Remote Sensing
for the Earth Sciences. American Society for Photogrammetry and Remote Sensing, John Wiley and
Sons Inc. New York, New York, USA.
Kruse, F. A., 1999. Mapping Hot Spring Deposits with AVIRIS at Steamboat Springs, Nevada: in
Proceedings of the 8th JPL Airborne Earth Science Workshop: Jet Propulsion Laboratory Publication
(in press).
Kruse, F. A., 1996. Geologic Mapping Using Combined Analysis of Airborne Visible/Infrared Imaging
Spectrometer (AVIRIS) and SIR-C/X-SAR Data: in Proceedings, International Symposium on
Optical Science, Engineering, and Instrumentation, SPIE Proceedings, August 1996, Denver, v.
2819, p. 24-35.
Kruse, F. A., Lefkoff, A. B., and Dietz, J. B., 1993, Expert System-Based Mineral Mapping in northern
Death Valley, California/Nevada using the Airborne Visible/Infrared Imaging Spectrometer
(AVIRIS): Remote Sensing of Environment, Special issue on AVIRIS, May-June 1993, v. 44, p.
309-336.
Langbein, W.B., and S.A. Schumm. 1958. Yield of sediment in relation to mean annual precipitation.
Am. Geophys. Union Trans., 39:1076-1084.
Lee, C., and S. Marsh. 1995. The use of archival Landsat MSS and ancillary data in a GIS environment to
map historical change in an urban riparian habitat. In: J.G Lyon and J. McCarthy (eds), Wetland and
Environmental Applications of GIS. CRC Press, Lewis Publishers, Boca Raton, Florida, USA.
Lillesand, T.M., and R.W. Kiefer. 1994. Remote sensing and image interpretation. John Wiley and Sons,
New York, NY.
26
-------
Lobo, A., K. Moloney, O. Chic, and N. Chiarello. 1998. Analysis of fine-scaled spatial pattern of a
grassland from remotely-sensed imagery and field collected data. Landscape Ecol. 13:111-131.
Logan, T.J. 1990. Chemical degradation of soil. In: R. Lai and B.A. Stewart (eds), Soil Degradation.
Springer-Verlag, New York, NY.
Lyon, J.G., R.S. Lunetta, and D.C. Williams. 1992. Airborne multispectral scanner data for evaluating
bottom sediment types and water depths of the St. Mary's River, Michigan. In: J.G. Lyon and J.
McCarthy (eds), Wetland and Environmental Applications of GIS. CRC Press, Lewis Publishers,
Boca Raton, Florida, USA.
Lyon, J.G. and J. McCarthy. 1995. Introduction to wetland applications. In: J.G. Lyon and J. McCarthy
(eds), Wetland and Environmental Applications of GIS. CRC Press, Lewis Publishers, Boca Raton,
Florida, USA.
Martin, M.E., S.D. Newman, J.D. Aber, and R.G. Congalton. 1998. Determining forest species
composition using high spectral resolution remote sensing data. Remote Sens. Environ. 65:249-254.
Morris, R.V., S.C. Neely, and W.W. Mendell. 1982. Application of Kubelka-Munk theory of diffuse
reflectance to geological problems: the role of scattering. Geophys. Res. Lett. 9:113-116.
Mouat, D., J. Lancaster, T. Wade, J. Wickham, C. Fox, W. Kepner, and T. Ball. 1997. Desertification
evaluated using an integrated environmental assessment model. Environ. Monitor. Assess.,
48:139-156.
Mustard, J.F., J.M. Sunshine. 1999. Spectral analysis for earth science investigation. In: A. N. Rencz
(ed), Remote Sensing for the Earth Sciences. American Society for Photogrammetiy and Remote
Sensing, John Wiley and Sons Inc. New York, New York, USA.
National Research Council. 1994. Rangeland health: new methods to classify, inventory, and monitor
rangelands. National Academy Press, Washington D.C., USA.
Nordin, C.F., Jr. 1985. The sediment loads of rivers. In: J.C. Rodda (Ed), Facets of hydrology.
John Wiley, Chichester.
Odeh, I.O.A, A.B. McBratney, and D.J. Chittleborough. 1992. Fuzzy-c-means and kriging for mapping
soil as a continuous system. Soil Sci. Soc. Am. J. 56:1848-1854.
Okin, G.S., W.J. Okin, D.A. Roberts, B. Murray. 1998. Multiple endmember spectral mixture analysis:
application to an arid/semi-arid landscape. Proceedings of the 8th Annual JPL Airborne Geoscience
Workshop, Pasadena, California, USA.
Olsen, K.C., R.S. White, and B.W. Sindela. 1985. Response of vegetation of the northern Great Plains to
precipitation amount and grazing intensity. J. Range Manage., 38(4):357-361.
Pierzynski, G.M., J.T. Sims, and G.F. Vance. 1994. Soils and environmental quality. CRC Press, Lewis
Publishers, Boca Raton, Florida, USA.
Pimm, S.L. 1991. The balance of nature? University of Chicago Press, Chicago, Illinois, USA.
27
-------
Rabeni, C.F., and M.A. Smale. 1995. Effects of siltation on stream fishes and the potential mitigating
role of the buffering riparian zone. Hydrobiologia 303:211-219.
Ramsey, E.W., and J.R. Jensen. 1995. Modeling mangrove canopy reflectance using a light interaction
model and an optimization technique. In: J.G. Lyon and J. McCarthy (eds), Wetland and
Environmental Applications of GIS. CRC Press, Lewis Publishers, Boca Raton, Florida, USA.
Ringrose, S., W. Matheson, C J.S.S. Matala, and T. O'Neill. 1994. Vegetation spectral reflectance along
a north-south vegetation gradient in northern Australia. J. Biogeogr. 21:33-47.
Roberts, D.A, M.O. Smith., and J.B. Adams. 1993. Green vegetation, nonphotosynthetic vegetation and
soils in AVIRIS data. Remote Sens. Environ. 44:255-269.
Schanda, E. 1986. Physical fundamentals of remote sensing. Springer-Verlag, Berlin. Germany.
Sedgwick, J.A. and F.L. Knopf. 1991. Prescribed grazing as a secondaiy impact in a western riparian
floodplain. J. Range Manage., 36(5):28-34.
Smith, A.M., D.J. Major, R.L. McNeil, W.D. Willms, B. Brisco, and R.J. Brown. 1995. Complementarity
of radar and visible-infrared sensors in assessing rangeland condition. Remote Sens. Environ.
52:173-180.
Soil Survey Staff. 1951. Soil survey manual. U.S. Department of Agriculture Handbook 18. U.S. Gov.
Printing Office, Washington DC., USA.
Soil Survey Staff. 1951. Soil taxonomy - a basic system of soil classification for making and interpreting
soil survey. Soil Conservation Service. U.S. Department of Agriculture Handbook no. 436. U.S.
Gov. Printing Office, Washington DC., USA.
Sposito, G. 1989. The chemistry of soils. Oxford University Press. New York, NY.
Stohlgen, T.J., D. Binkley, G.W. Chong, M.A. Kalkhan, L.D. Schell, K.A. Bull, Y. Otsuki, G. Newman,
M. Bashkin, and Y. Son. 1999. Exotic plant species invade hot spots of native plant diversity. Ecol.
Mono. 69(1)25-46.
Stoner, E.R, and M.F. Baumgardener. 1980. Physicochemical, site, and bidirectional reflectance factor
characteristics of uniformly moist soils. LARS Tech. Rep. 111679. Purdue University, West
Lafayette, Indiana, USA.
Stoner, E.R, and M.F. Baumgardener. 1981. Characteristic variations in reflectance of surface soils. Soil
Sci. Soc Am. J. 45:1161-1165.
Taylor, G.R., A.H. Mah, F.A. Kruse, K.S. Kierein-Young, R.D. Hewson, and B.A. Bennett. 1996. The
extraction of soil dielectric properties from AIRSAR data. Int. J. Remote Sens. 17(3):501-512.
Tueller, P.T. 1973. Secondary succession, disclimax, and range condition standards in desert shrub
vegetation. In: Proceedings of the Third Workshop of U.S/Australia Rangelands Panel. Tucson,
Arizona, USA.
28
-------
Ulaby, F.T., R.K. Moore, and A.K. Fung. 1982. Microwave remote sensing; active and passive, volume
II. Addison-Wesley Publishing, Reading, Massachusetts, USA.
Van Wambeke, A.. 1992. Soils of the tropics: properties and appraisal. McGraw-Hill, New York, USA.
Van Zyl, J. J., C.F. Burnette, and T.G. Farr. 1991. Inference of surface power spectra from inversion of
multifrequency polarimetric radar data. Geophy. Res. L. 18(9): 1787-1790.
Van Zyl, J.J., R. Carande, Y. Lou, T. Miller, and K. Wheeler. 1992. The NASA/JPL three-frequency
polarimetric AIRSAR system. Proceedings of IGARSS 1992.
Vincent, R.K. 1997. Fundamentals of geologic and environmental remote sensing. Prentice-Hall, Upper
Saddle River, New Jersey, USA.
Vogelmann, J.E., T. Sohl, and S.M. Howard. 1998. Regional characterization of land cover using
multiple sources of data. Photogrammetric Eng. Remote Sens. 64:45-47.
Washington-Allen, R.A., N.E. West, R.D. Ramsey, and C.A. Hunsaker. 1999. Characterization of the
ecological integrity of commercially grazed rangelands using remote-sensing-based ecological
indicators. In: People and Rangelands (Eds D. Eldridge and D. Freudenberger), pp 778-780. VI
International Rangeland Congress Inc., Townsville, Australia.
Weixelman, D.A., D.C. Zamudio, K.A. Zamudio, and R.J. Tausch. 1997. Classifying ecological types
and evaluating site degradation. J. Range Manage., 50:316-332.
Werth, L.F., and E.A. Work. 1992. Applications of large-sale aerial photography for rangeland
monitoring. Geocarto, 7:11-18.
Wessman, C.A., 1991. Remote sensing of soil processes. Agric. Ecosystems Environ., 34:479-493.
West, N.E., F.D. Provenza, P.S. Johnson, and M.K. Owens. 1984. Vegetation change after 13 years of
livestock exclusion on sagebrush semidesert in west central Utah. J. Range Manage., 37(3):262-269.
Westoby, M., B.H. Walker and I. Noy-Meir. 1989. Opportunistic range management for rangelands not at
equilibrium. J. Range Management 42:266-274.
Weyer, L.G. 1985. Near-infrared spectroscopy of organic substances. Appl. Spectrosc. Rev., 21:1-43.
29
-------
Proposed Staffing
Percent Time
Name
Duties
FY99 FY00 FY01
David J. Williams
Investigator/Physical Scientist
50% 50% 50%
William G. Kepner
Co-Investigator/Landscape Ecologist
10% 10% 10%
30
-------
Description of Facilities
All data potentially acquired under this proposal will be analyzed by EPA image processing scientists at
the image processing laboratory of the Environmental Photographic Interpretation Center (EPIC) in
Reston, Virginia. The EPA/EPIC laboratory is currently outfitted with three Unix workstations (two
SUN Ultra 30, one SGI Indigo 2) and one Windows NT workstation. Included is an ASD Full-Range
field Spectrometer for field data collection. Software in the EPIC IP lab includes two copies of Arc/Info
(GIS), five copies of ARCVIEW, three copies of ENVI, one copy each of ERDAS Imagine and PCI
(image processing). Spectral processing software includes GRAMS32, and ATREM. Also, under
Interagency Agreement with USGS, EPA scientists have access to the USGS reflectance Spectroscopy
lab and their Specpr and Tetracorder spectral analysis software.
31
-------
Appendix A
Description of Potential Study Sites
Descriptive text courtesy of the USFS, USDA, USGS-BRD, and Utah State University.
Pilot Study Sites
Sevilleta National Wildlife Refuge, Socorro, New Mexico. The University of New Mexico's Sevilleta
Long-Term Ecological Research Program (LTER) in the central Rio Grande Basin is one of the U.S.
National Science Foundation's LTER Network sites. The Sevilleta LTER Program is located
primarily on the Sevilleta National Wildlife Refuge in Socorro County, NM. Varied study sites
include a wide range of ecosystem types, including Chihuahuan Desert, Great Plains Grassland,
Great Basin Shrub-Steppe, Pinon-Juniper Woodland, Bosque Riparian Forests and Wetlands,
Ponderosa Pine Forests, Mixed-Conifer Montane Forests, and Sub-alpine Forests and Meadows.
Walnut Gulch Experimental Watershed, Tombstone, Arizona. Located in SE Arizona surrounding
the historic city of Tombstone, the 150 sq. km. Walnut Gulch Experimental Watershed was
established in the early 1950's to study the role of watershed treatments on downstream water yield.
The site was deemed typical of the black grama grass-brush dominated areas of southern New
Mexico and Arizona. Beginning in 1954, a network of recording precipitation gages was established.
Since then, the network of recording precipitation gages has grown to the nearly 100 gages currently
maintained by SWRC staff. Complementary to the precipitation gages, 25 flumes and weirs have
been established within the Walnut Gulch Experimental Watershed defining subwatersheds ranging
in size from 112 sq. km. to 0.3 sq. km.
Follow-on Study Sites
Jornada Experimental Range, Las Cruces, New Mexico. The Jornada Experimental Range is a
193,394 acre (78,325 hectare) research facility established in 1912. The Range is located in the
northern portion (the Trans Pecos region) of the Chihuahuan Desert, the largest desert in North
America. The mission of the Range is to develop new knowledge of ecosystem processes as a basis
for management and remediation of desert rangelands. AVIRIS data has been collected for the range
in 1998 and a current overflight for the area and supporting ground measurements is planned in
October, 1999.
Desert Experimental Range. The Desert Experimental Range is in Pine Valley approximately 70 km
(43 miles) west of Milford, Utah. It is geographically and floristically representative of
approximately 200,000 km2 (77,220 miles2) of the Great Basin, an arid region of the Western United
States comprising a series of north-and south-aligned ranges and closed basins. AVIRIS data has
been collected for regions adjacent to the range.
32
-------
Great Basin Experimental Range, Ephraim, Utah. The Great Basin Experimental Range is on the
south portion of the Ephraim or Cottonwood Creek drainage on the west front of the Wasatch Plateau
about 8 km (5 miles) east of Ephraim, Utah, on the Manti-LaSal National Forest. AVIRIS data has
been collected for adjacent areas in 1998.
33
-------
Appendix B
Personnel
34
-------
David James Williams
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Environmental Photographic Interpretation Center
12201 Sunrise Valley Drive
555 National Center, Reston, VA, 20192
Telephone: (703) 648-4798
Facsimile: (703) 648-4290
E-mail: vvilliams.davidi@epa.gov
or dxwilliams@usgs.gov
Education
Ph.D. (In progress) Computational Sciences and Informatics, Earth Observing and Remote Sensing,
George Mason University
M.S. Environmental Science: Soil Physical Chemistry, 1998. The Ohio State University
B.S. (cum laude) Natural Resources, 1996. The Ohio State University
Research Interests
Remote sensing with an emphasis on hyperspectral data analysis or imaging spectroscopy and
computational techniques; evaluation of current and future remote sensing systems; reflectance
spectroscopy of organic and inorganic contaminates in soils and sediments; chemistry and mineralogy of
sediments from mining and industrial effluents; spatial analysis of environmental data including the
development of landscape ecological indicators; GIS and spatial statistics. Scientific background in
analytical and physical chemistry, geochemistry, hydrogeology, and pedology.
Professional Memberships
American Society for Photogrammetry and Remote Sensing
IEEE Geoscience and Remote Sensing Society
Honorary Societies: Phi Kappa Phi, Gamma Sigma Delta
Employment
1998 - Present:
1997- 1998:
1996- 1998:
1993 - 1996:
35
Research Physical Scientist, USEPA National Exposure Research Lab, Environmental
Photographic Interpretation Center, Reston, Virginia.
Image Analyst, Ohio Environmental Protection Agency, Division of Surface Water,
Columbus, Ohio.
Graduate Fellow, Environmental Science Graduate Program, The Ohio State Univ.
Columbus, Ohio.
Research Assistant, Soil Characterization Lab, School of Natural Resources, The Ohio
State Univ. Columbus, Ohio.
-------
Scholarships and Awards
1999: USEPA Office of Research and Development Honor Award: Exceptional ORD Technical
Assistance to the Regions or Program Offices
1996: U.S. Department of Defense Environmental Restoration Fellowship
1995: Ohio State University, School of Natural Resources Undergraduate Research Scholarship
Selected Publications
Williams, D.J., J.M. Bigham, and S.J. Traina. 2000. Assessing mine drainage water quality form the
color and spectral reflectance of chemical precipitates. Applied Geochemistry (submitted).
Williams, D.J., and S., Norton. 2000. Determining impervious surfaces in satellite imagery using digital
orthophotography. Proc. of the Am. Soc. for Photogrammetry and Remote Sens. Annual Conference,
Washington, DC.
Williams, D.J., D.A. White, and A. Engelmann. 1999. Riparian characterization using sub-pixel analysis
in an ecological risk framework. Proc. of the Am. Soc. for Photogrammetry and Remote Sens.
Annual Conference, Portland, OR. May 21st 1999.
Williams, D.J., and W.G. Kepner. 1999. Imaging spectroscopy for determining rangeland stressors to
western watersheds. Abstracts with program, EMAP Symposium on Western Ecological Systems:
Status, Issues, and New Approaches, San Francisco, CA., April 6-8, 1999. p. 152.
36
-------
William G. Kepner
U.S. Environmental Protection Agency
National Exposure Research Laboratory
Environmental Sciences Division
944 East Harmon Avenue
Las Vegas, NV 89119
Telephone: (702)798-2193
Facsimile: (702) 798-2692
E-mail: Kepner.William@epa.gov
Education
M.S. Zoology, Arizona State Univ., 1982
B.S. Biology, Univ. of Arizona, 1975
A.A. Biology, Phoenix College, 1973
Professional Registration
Certified Wildlife Biologist, The Wildlife Society, 1983
Certified Fisheries Scientist, American Fisheries Society, 1982
Certificate in Business Management, UNLV College of Business and Economics, Department of
Management in cooperation with the American Management Association, 1995
Experience
1990-Present:
1984-1990:
1978-1984:
1977-1978:
1997:
Research Ecologist: U.S. Environmental Protection Agency, National Exposure
Research Laboratory, Las Vegas, Nevada
Environmental Contaminant Specialist: U.S. Fish and Wildlife Service, Phoenix,
Arizona
Wildlife Biologist: U.S. Bureau of Land Management, Phoenix, Arizona
Research Assistant: Lower Colorado River Basin Research Laboratory, Arizona State
University, Tempe, Arizona
Hydrologist: U.S. Forest Service, Apache-Sitgreaves National Forests, Springerville,
Arizona
37
-------
Selected Publications
Mouat, D.A., J. Lancaster, T. Wade, J. Wickham, C. Fox, W.G. Kepner, and T. Ball. 1997.
Desertification Evaluated Using an Integrated Environmental Assessment Model. Environmental
Monitoring and Assessment 48:139-156.
Riitters, K.H., J.D. Wickham, K.B. Jones, W.G. Kepner, and D.J. Chaloud. 1996. A Landscape Atlas of
the Chesapeake Bay Watershed. Tennessee Valley Authority, Norris, TN. 29 pp.
Kepner, W.G., K.B. Jones, D.J. Chaloud, J.D. Wickham, K.H. Riitters, and R.V. O'Neill. 1995.
Mid-Atlantic Landscape Indicators Project Plan. EPA/620/R-95/003. U.S. Environmental Protection
Agency, Office of Research and Development, Washington, D.C. 38 pp.
Wade, T.G., J.D. Wickham, and W.G. Kepner. 1995. Using GIS and a Graphical User Interface to Model
Land Degradation. Geo Info Systems. Pp. 38-42.
Kepner, W.G. 1995. Environmental Monitoring and Assessment Program: A Landscape Approach to
Environmental Assessment: Application to Neotropical Migratory Bird Issues. Partners in Flight,
4(2). National Fish and Wildlife Foundation, Washington, D.C.
Breckenridge, R.P., W.G. Kepner, and D.A. Mouat. 1995. A Procedure for Selecting Indicators for
Monitoring Condition of Rangeland Health. International Journal of Environmental Monitoring and
Assessment, 36:45-60.
Kepner, W.G. et al. 1994. Environmental Monitoring and Assessment Program, Arid Ecosystems 1992
Pilot Report. EPA/620/R-94/015. U.S. Environmental Protection Agency, Office of Research and
Development, Washington, D.C. 116 pp.
Kepner, W.G. et al. 1991. Arid Ecosystems Strategic Monitoring Plan. EPA/600/4-91/018. U.S.
Environmental Protection Agency, Office of Research and Development, Washington, D.C. 299 pp.
Radtke, D.B., W.G. Kepner, and R.J. Effertz. 1988. Reconnaissance investigation of water quality,
bottom sediment, and biota associated with irrigation drainage in the Lower Colorado River Valley,
Arizona, California and Nevada, 1986-87. U.S. Geological Survey Water-Resources Investigations
Report 88-4002. Pp. 1-77. Denver, CO.
38
------- |