v>EPA
United States
Environmental Protection
Agency
Environmental Monitoring
and Support Laboratory
PO. Box 15027
Las Vegas NV 89114
EPA-600/7-79 080
April 1979
Research and Development
Computer Processing
of Multispectral Scanner
Data Over Coal
Strip Mines
Interagency
Energy-Environment
Research
and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY—ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA'S mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the Pro-
gram is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses of the transport of energy-related
pollutants and their health and ecological effects: assessments of, and development of,
control technologies for energy systems; and integrated assessments of a wide range
of energy-related environmental issues.
This document is available to the public through the National Technical Information
Service. Springfield, Virginia 22161
-------
EPA-600/7-79-080
March 1979
COMPUTER PROCESSING OF MULTISPECTRAL SCANNER
DATA OVER COAL STRIP MINES
by
Charles E. Tanner
Lockheed Electronics Company, Inc,
Las Vegas, Nevada 89114
Contract No. 68-03-2636
Project Officer
G. J. D'Alessio
Office of Energy, Minerals, and Industry
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D. C. 20460
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring
and Support Laboratory-Las Vegas, Nevada, U.S. Environmental Pro-
tection Agency, and approved for publication. Approval does not
signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or rec-
ommendation for use.
11
-------
FOREWORD
Protection of the environment requires effective regulatory
actions that are based on sound technical and scientific informa-
tion. This information must include the quantitative description
and linking of pollutant sources, transport mechanisms, inter-
actions, and resulting effects on man and his environment. Be-
cause of the complexities involved, assessment of specific pollu-
tants in the environment requires a total systems approach that
transcends the media of air, water, and land. The Environmental
Monitoring and Support Laboratory-Las Vegas contributes to the
formation and enhancement of a sound, integrated monitoring data
base through multidisciplinary, multimedia programs designed to:
• develop and optimize systems and strategies for
monitoring pollutants and their impact on the
environment
• demonstrate new monitoring systems and technologies
by applying them to fulfill special monitoring needs
of the Agency's operating programs
This report presents the results of the Phase II operations
of the Western Energy Overhead Monitoring Project conducted by the
Environmental Monitoring and Support Laboratory-Las Vegas, Nevada.
It describes and outlines procedures used to produce Level I land-
cover classification maps of selected coal strip mines. These
data are then used to assess reclamation efforts and monitor
changes on active strip mines in^ the Western^ United States .
George B. Morgan
Director
Environmental Monitoring and
Support Laboratory
Las Vegas, Nevada
1X1
-------
ABSTRACT
There is little doubt that remote sensing techniques can be
effectively applied to the task of monitoring coal strip mine
progress and reclamation work. Aircraft multispectral scanner
data acquired over six coal strip mines in the States of Wyoming,
Montana, Colorado, and Arizona were processed on the Data Analysis
System (DAS) using a clustering approach to automatic pattern
recognition. The classification results demonstrated that a
Level I hierarchy of vegetation features, manmade features, and
disturbed areas could be easily obtained with a minimum amount
of time. Aside from satisfying a Level I hierarchy, the results
may be used as input to other classification approaches to pattern
recognition, or they may be incorporated into a data base for
planning or for conducting temporal analyses studies.
This report was submitted in partial fulfillment of Contract
No. 68-03-2636 by Lockheed Electronics Company, Incorporated,
under the sponsorship of the U.S. Environmental Protection Agency.
IV
-------
CONTENTS
Page
Foreword iii
Abstract iv
Figures vi
Tables viii
Scientific Names of Vegetation ix
Acknowledgement x
Introduction 1
Conclusions and Recommendations 4
Data Acquisition 5
Multispectral Scanner 5
Metric Mapping Camera 9
Data Processing 11
Data Analysis 14
Results and Discussion 20
Colstrip Coal Strip Mine 20
Decker Coal Strip Mine 22
Dave Johnston Coal Strip Mine 27
Wyodak Coal Strip Mine 35
Black Mesa Coal Strip Mine 35
Nucla Coal Strip Mine 40
Summary 48
References 49
v
-------
FIGURES
Number Page
1 The location of active coal strip mines analyzed in
Phase II operations 3
2 Twin-engine Aero Commander used by EPA in support of
ongoing investigations and surveillance 6
3 An 11-channel multispectral scanner used by EPA in
ongoing investigations 7
4 Multispectral scanner imaging characteristics
(simplified) .... 8
5 Configuration of the EMSL-LV Data Analysis System. . . 12
6 Aircraft multispectral scanner data reduction flow
(simplified) 13
7 Simplified data processing flow for aircraft multi-
spectral scanner data over coal strip mines 18
8 The locations of the Colstrip and Decker Coal Strip
Mines within the State of Montana 21
9 Land-cover classification map of the Colstrip Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data 23
10 Aerial color-infrared photography of the Colstrip
Coal Mine acquired by NASA U-2 aircraft 24
11 Aerial color-infrared photography of the Decker Coal
Strip Mine acquired by NASA U-2 aircraft 26
12 Land-cover classification map of the Decker Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data 28
13 The location of the Dave Johnston and Wyodak Coal
Strip Mines in the State of Wyoming 30
14 Land-cover classification map of the Dave Johnston
Coal Strip Mine generated from aircraft-acquired
multispectral scanner data 31
vi
-------
FIGURES
Number Page
15 Aerial color-infrared photography of the Dave
Johnston Coal Strip Mine acquired by NASA U-2
aircraft 32
16 Aerial color-infrared photography of the Dave
Johnston Coal Strip Mine acquired by NASA U-2
aircraft 33
17 Land-cover classification map of the Wyodak Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data 36
18 Aerial Color-infrared photography of the Wyodak Coal
Strip Mine acquired by NASA U-2 aircraft 37
19 The location of the Black Mesa Coal Strip Mine with-
in the State of Arizona. 39
20 Land-cover classification map of the Black Mesa Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data 41
21 Aerial color-infrared photography of the Black Mesa
Coal Strip Mine acquired by NASA U-2 aircraft. ... 42
22 The location of the Nucla Coal Strip Mine in the
State of Colorado 44
23 Aerial color-infrared photography of the Nucla Coal
Strip Mine acquired by NASA U-2 aircraft 45
24 Land-cover classification map of the Nucla Coal
Strip Mine generated from aircraft acquired multi-
spectral scanner data 46
Vll
-------
TABLES
Number Page
1 Multispectral Scanner Wavelength Bands 9
2 Land-cover Classification Hierarchy for Coal Strip
Mine Monitoring 15
3 Modified Land-cover Classification Hierarchy with
Coding System 17
4 Acreage Statistics from the Entire Scene of the
Colstrip Coal Strip Mine 25
5 Acreage Statistics from the Entire Scene of the
Decker Coal Strip Mine 29
6 Acreage Statistics from the Entire Scene of the
Dave Johnston Coal Strip Mine 34
7 Acreage Statistics from the Entire Scene of the
Wyodak Coal Strip Mine 38
8 Acreage Statistics from the Entire Scene of the
Black Mesa Coal Strip Mine 43
9 Acreage Statistics from the Entire Scene of the
Nucla Coal Strip Mine 47
Vlll
-------
SCIENTIFIC NAMES OF VEGETATION IDENTIFIED
IN PHASE II OPERATIONS
Common Name
Alkali sacatone
Blue grama
Greasewood
Green needlegrass
Indian Ricegrass
Junegrass
Juniper
Mormon tea
Needle-and-threadgrasses
Pinyon pine
Rabbit brush
Sand dropseed
Sagebrush
Serviceberry
Snakeweed
Wheatgrass
Winterfat
Genus and Species
Sporobolus airvoides
Bouteloua gracilis
Sarcobatus vermiculatus
Stipa viridula
Oryzopsis hymenoides
Koeleria cristata
Juniperus sp.
Ephedra sp.
Stipa comata
Pinus edulis
Shrysothamnus sp.
Sporobolus cryptandrus
Artemisia sp.
Amelanchier sp.
Gutierrezia sp.
Agropyron sp.
Eurotia lanata
IX
-------
ACKNOWLEDGMENTS
The scale and scope of this report were made possible
through the cooperation of the U.S. Environmental Protection
Agency and the National Aeronautics and Space Administration/
Earth Resources Laboratory (NASA/ERL) in Slidell, Louisiana.
Sincere thanks are extended to mine personnel at the Decker,
Nucla, Dave Johnston, Colstrip, and Black Mesa Coal Strip Mines
for providing assistance during ground-truth operations.
x
-------
INTRODUCTION
The extraction and transportation of coal can be and some-
times is unattractive, but a certain price in environmental
damages usually must be paid to obtain the coal required for our
standard of living. In the West, coal is extracted by surface
mining techniques in which the overburden is removed to expose
and allow the extraction of the underlying coal seam. Surface
mining is divided into two general types: area mining and con-
tour mining. Area mining is practiced in relatively flat to
gently rolling terrain. Contour mining is practiced where de-
posits occur in hilly or mountainous country.1
It has been demonstrated that surface mining can be done
responsibly without permanent aesthetic damage to the land and
water provided the mining company plans ahead and follows estab-
lished or recommended guidelines. Technology exists for the
effective reclamation of mined lands, and many mining companies
are taking full advantage of such technology as evidenced by the
reclamation efforts in the West.
The first national attempt to regulate surface mining activ-
ities was made by the 95th Congress of the United States when it
passed the Surface Mining Control and Reclamation Act of 1977.
The purpose of this act is to provide for the cooperation between
the Secretary of the Interior and the States with respect to the
regulation of surface coal mining operations and the purchase and
reclamation of abandoned mines.
In anticipation of the passage of a comprehensive strip
mining bill and because of its Congressional charter, the
Environmental Protection Agency (EPA) entered into a five-year
interagency project with the National Aeronautics and Space
Administration (NASA). The purpose of the project is to transfer
hardware and software technology for processing remotely sensed
digital data from aircraft or satellite platforms to the monitor-
ing of coal strip mine operations in the Western United States.
This project was divided into three phases. Phase I
(as reported by Anderson, et al.) was an 18-month task during
which time NASA, at its Earth Resources Laboratory (ERL) in
Slidell, Louisiana, defined a data collection system for instal-
lation in an EPA aircraft located at the Environmental Monitoring
and Support Laboratory in Las Vegas, Nevada (EMSL-LV). The Earth
-------
Resources Laboratory was also to process Landsat and aircraft-
acquired multispectral scanner data of selected strip mines in
the West using basic pattern recognition techniques refined and/or
developed at NASA/ERL. Finally, NASA was to build a Data Analysis
System and train EPA personnel in its use and maintenance.2
Phase II of the joint project was undertaken by the Environ-
mental Monitoring and Support Laboratory Las Vegas. EMSL-LV used
the system developed at ERL to monitor selected coal strip mines
in 6 Western States and at the same time to develop its own pro-
cedure for performing "quick turnaround" land-cover classification
maps and acreage statistics for coal strip mines.
Phase III will be the testing phase of the system and
techniques in an operational mode, with parallel software develop-
ment and assistance being provided by NASA/ERL.
Phase II of the joint project was begun in January 1977.
During this phase, EMSL-LV used an 11-channel multispectral
scanner (MSS) and the Data Analysis System (DAS) computer and its
software to monitor selected coal strip mines in the West and
Northern Great Plains area. Figure 1 illustrates the general
location of the mines within their respective states.
Also, during this time, NASA/ERL (using a similar system)
is investigating specific problem areas related to coal strip
mines, mine-mouth power plants, and oil shale and geothermal
energy sites which require additional research.2
This report discusses the results of the phase II operations
conducted by EMSL-LV and briefly outlines the techniques employed
or developed to obtain the results.
-------
WYODAK
DAVE JOHNSTON «
Figure 1.
The location of active coal strip mines
analyzed in phase II operations.
-------
CONCLUSIONS AND RECOMMENDATIONS
The results of a sequential clustering approach to pattern
recognition provided useful Level I land-cover classification
maps of active coal strip mines. The map that was produced by
the clustering program can also be upgraded to Level II or Level
III if the necessary ancillary data (vegetation-type maps, ground-
truth data, etc.) are available. As an added feature, the Level I
classification map can be used as input to other classification
approaches to pattern recognition. These data would facilitate
the selection of homogeneous training samples that are required
when using the supervised classification approach.
Based on the results of this portion of the EMSL-LV phase II
operations it is recommended that:
• Modifications to the clustering algorithm be per-
formed to reduce run time and to accept statistical
input parameters as well as a nonrectangular or
square field input,
• other available software programs similar to the
clustering algorithm available on the EMSL-LV inter-
active computer be evaluated to determine their
utility in classifying coal strip mine areas, and
• use of the sequential clustering algorithm be con-
tinued as the "first cut" at land-cover classifica-
tion of coal strip mine areas.
-------
DATA ACQUISITION
Aircraft multispectral scanner data were collected over the
selected coal strip mines in late June, 1977. Data were used be-
cause their spectral and spatial resolution would permit computa-
tion of very detailed statistics for major land-cover categories.
The Environmental Monitoring and Support Laboratory-Las Vegas
subleases a light twin-engine Aero Commander (Figure 2) that is
used to conduct special surveys and provide support to ongoing
investigations. This aircraft is capable of attaining a maximum
altitude of 7,315 meters (24,000 feet) above mean sea level (MSL).
Instrumentation aboard the aircraft consists of a metric camera
and an 11-channel multispectral scanner. The scanner and the
metric camera are used for monitoring coal strip mines and are
discussed in the following paragraphs.
MULTISPECTRAL SCANNER
The multispectral scanner (Figure 3) is an airborne system
capable of collecting data at altitudes ranging from 500 feet
(152.4 meters) to 20,000 feet (6,096m) above ground level. This
system is designed to collect and record radiant energy data in
the ultraviolet through the thermal infrared portions of the elec-
tromagnetic spectrum (see Table 1 for bandwidths).
The scanner has a rotating mirror that scans across the
ground scene perpendicular to the line of flight (Figure 4).
Radiant energy from the ground surface is reflected through an
aperture onto a beam splitter, which diverts the visible radiation
(0.38 - l.lOym) to a 10-channel spectrometer and the thermal
infrared radiation (8 - 14um) to a solid-state detector. Electronic
signals from the 11 detectors are digitized and recorded on magnetic
tape in a high-density format. The scan rate is synchronized to the
aircraft ground speed and altitude, resulting in scan-line contigu-
ity at nadir, thereby avoiding over- or under-scan of the ground
scene. The scanner is equipped with internal visible and thermal
reference sources, which provide information for calibration of
the data. The aircraft sensor tape is processed on the ground-based
Data Analysis System (DAS) to display, analyze, and create images
of the surveyed scene. These processes will be discussed in greater
detail in another section of this report.
-------
Figure 2. Twin-engine Aero Commander used by EPA in support of
ongoing investigations and surveillance.
-------
Figure 3.
An 11-channel multispectral scanner used by EPA
in ongoing investigations.
-------
SCAN LINE
Figure 4.
Multispectral scanner imaging characteristics
(simplified).
8
-------
TABLE 1. MSS WAVELENGTH BANDS
Channel
1
2
3
4
5
6
7
8
9
10
11
Wavelength
0
0
0
0
0
0
0
0
0
0
8
.38 -
.42 -
.45 -
.50 -
.55 -
.60 -
.65 -
.70 -
.80 -
.92 -
.00 -
Band (vim)
0.42
0.45
0.50
0.55
0.60
0.65
0.70
0.79
0.89
1.10
14.00
Color/Spectrum
Near
Blue
Blue
Green
Green
Red
Red
Near
Near
Near
Ultraviolet
Infrared
Infrared
Infrared
Thermal Infrared
METRIC MAPPING CAMERA
The camera that is part of the instrumentation aboard the air-
craft is a 6-inch focal length (15.24-centimeter) metric camera.
Its primary use is a ground-coverage documenting camera. The film/
filter combinations and exposure techniques are chosen to obtain
photographic imagery in the desired spectra.
The following lists the camera specifications:
• Type
9-inch (22.86cm) square image utilizing a 6-inch
(15.24cm) F/5.6 Universal Aviogon lens. Unit has
no forward-motion compensation.
• Lens
Focal Length, 6-inch (15.24cm) nominal, with
Waterhouse stops; apertures of F/5.6, 6.8, 8,
11, 16, 22, and 32. View angle 74 degrees.
-------
Shutter
Continuous variable rotary, speed from 1/100
to 1/700 second.
Cycle Interval
One cycle each 3.5 seconds, maximum rate.
10
-------
DATA PROCESSING
Data processing begins after the scanner data have been re-
corded on a high density digital tape (HDDT) and returned to the
laboratory. The aircraft-acquired multispectral scanner data
must be converted into another format because conventional com-
puters, such as the Data Analysis System, cannot read the data
format produced by an aircraft multispectral scanner. Therefore,
a special device has been designed to convert aircraft multi-
spectral scanner data into the format expected by conventional
computers. This reformatting procedure has been termed decommuta-
tion and is made in the HDDT or pulse-coded-modulation (PCM) front-
end hardware (Figure 5). Following the decommutation operation,
standard data processing techniques are applied to all other func-
tions (Figure 6) in order to produce a classified image and to
generate the required acreage statistics.3
11
-------
OPERATOR'S TERMINAL
AND CARD READER
9-TRACK MAGNETIC
TAPE DRIVES
COLOR FILM
RECORDER
PLAYBACK SYSTEM AND
CENTRAL COMPUTER
INTERACTIVE DISPLAY SYSTEM
Figure 5. Configuration of the EMSL-LV Data Analysis System,
12
-------
DATA ACQUISITION
DECOMMUTATION
PREPROCESSING
AND DATA
TRANSFORMATION
PRE-
| PROCESSED'
DATA
TAPE
INTERACTIVE
DISPLAY
SYSTEM
I
INTERACTIVE
DISPLAY
SYSTEM
Figure 6.
Aircraft multispectral scanner data reduction
flow (simplified).
13
-------
DATA ANALYSIS
Before processing and analysis of the data could be initiated,
a land-cover hierarchy similar to Table 2 (photointerpretation
hierarchy) had to be developed. This hierarchy (Table 3) is a
categorical ranking of the natural or manmade ground features that
are of primary interest to the analyst.
In the Phase II operations, it was felt that a Level I/II
hierarchy would prove to be more economical and faster for pro-
ducing the results needed to monitor coal strip mine operations.
The Level II hierarchy was implemented whenever the needed ancil-
lary data or photography was available.
Data analysis was begun immediately after the hierarchy was
approved. Processing of the aircraft-acquired multispectral scan-
ner data was performed using EMSL-LV developed techniques and hard-
ware for the automatic recognition of spectral patterns.
Analysis of the MSS data proceeded as detailed in Figure 7.
A computer-compatible tape that had been preprocessed to remove
scanner anomalies was viewed on the Data Analysis System to locate
the start/stop picture element (pixel) and the start/stop scan-
line numbers for the area of interest. Also during this procedure,
a quality check of the data was performed. Following the comple-
tion of the aforementioned procedures, the analyst had to decide
which classification approach or combination of approaches to use.
It was decided to use a combination approach and to classify each
data set using a sequential clustering algorithm. The classified
image generated by the sequential clustering algorithm would then
be used to facilitate the selection of homogeneous training samples
for use in other classification approaches to pattern recognition.
The clustering program is an algorithm that groups pixels with
similar statistical characteristics. The major assumption in the
clustering program is that the data have some homogeneous patterns.
Success at running this algorithm requires that the analyst know
something about the data, e.g., spectral ranges of the data, approx-
imate number of classes that can be generated, the categories that
may possibly cause confusion, the best channels to use for class
discrimination, and the number of channels with which to work.
14
-------
TABLE 2. LAND-COVER CLASSIFICATION HIERARCHY FOR COAL
STRIP MINE MONITORING
00 - Active Mine Features
01 - Advance cut
02 - Topsoil stockpile including embankment
03 - Stripping bench, high wall, open cut, pit, and related
features
04 - Exposed coal seam
05 - Raw spoil bank and/or sideslope
06 - Coal storage pile
07 - Recontoured spoil
08 - Haul road including cuts, fills, turnouts, etc.
09 - Miscellaneous disturbed areas within active mine complex
10 - Barren Land
14 - Shorelines, river banks
15 - Badlands (barren silts and clays, related metamorphic
rocks)
18 - Manmade barrens
20 - Water Resources
21 - Settling ponds
22 - Pit ponds
23 - Undifferentiated ponds, lakes, and reservoirs
24 - Water courses, including canals
30 - Natural Vegetation
31 - Herbaceous types
32 - Shrub/scrub types
33 - Savanna-like types
34 - Forest and woodland types
15
-------
TABLE 2. LAND-COVER CLASSIFICATION HIERARCHY FOR COAL
STRIP MINE MONITORING (Continued)
40 - Cultural Vegetation (plantations and seedings)
Numerators
41 - Grass/forb seedings
42 - Tree/grass or tree/scrub plantations
45 - Seeding trails and test plots
Denominators
A - Vegetative cover 80 - 100%
B - Vegetative cover 60 - 80%
C - Vegetative cover 30 - 60%
D - Vegetative cover 5 - 30%
E - New seeding, vegetative cover less than 5%
1 - Production level better than native level
2 - Production level about equal with native level
3 - Production level less than native level
50 - Agricultural Production
51 - Field crops
56 - Fallow land
60 - Urban/Industrial
61 - Residential
62 - Commercial and services
63 - Institutional
64 - Industrial
65 - Transportation, communications and utilities
66 - Resource extraction
16
-------
TABLE 3. MODIFIED LAND-COVER CLASSIFICATION HIERARCHY WITH
CODING SYSTEM
LEVEL I
LAND-COVER
Exposed coal
seam
Spoil pile
Recontoured
area
Haul road
Rehabilitated
area
Natural vege-
tation
Agricultural
land
Barren land
Urban
Residential
CODE*
00/04
00/05
00/07
00/08
20
30
50
10
60
60/61
LEVEL II pn p LEVEL III
LAND-COVER LAND -COVER
Dense vegeta-
tion
Medium vege-
tation
Light density
vegetation
Aquatic vegeta- 30/31 Dense vegeta-
tion tion
Grassland 30/33 Medium density
vegetation
Forest/wood- 30/34 Light density
land vegetation
Planted land 50/51
Fallow land 50/56
CODE
40/A
40/B
40/C
40/A
40/B
40/C
*Land-cover classification codes taken
Roach Hierarchy.1*
17
from the modified Anderson-
-------
PRE-
/PROCESSEI
COMPUTER
I COMPATIBLE
.TAPE
INTERACTIVE
DISPLAY
SYSTEM
PATTERN
RECOGNITION
SUPERVISED CLASSIFICATION
APPROVAL
•TRAINING FIELD SELECTION
•STATISTICS COMPUTATIONS
•CHANNEL SELECTION
•CLASSIFICATION
UNSUPERVISED CLASSIFICATION
APPROACH
•DETERMINE SPECTRAL RANGE OF DATA
•DETERMINE STATISTICAL CONSTRAINTS
•SET CRITERIA FOR CLASSES
• CHANNEL SELECTION
4 BEST CHANNELS
CORRELATION TABLE
MEANS
STANDARD DEVIATIONS
ACREAGE
COMPILATION
THEME
INVENTORY
HARDCOPY
OUTPUT
ACREAGE
COMPILATION
THEME
INVENTORY
Figure 7. Simplified data processing flow for aircraft multi-
spectral scanner data over coal strip mines.
18
-------
The data flow for the sequential clustering program proceeds
as follows:
• Establishing the initial population (land-cover
class),
• Developing statistical profile tables for classes,
• Determining if the picture element belongs to
established populations, and
• Establishing additional populations.
Each pixel of the digital data is compared to the existing
statistical profiles to determine which population it best fits.
When all the pixels have been assigned to one of the active popu-
lations or to zero population (pixels exceeding specified count
ranges), the run terminates and the results are saved on a 9-track
(800 BPI) magnetic tape. Statistics that were generated during
the run and stored in the computer are dumped to the line printer
device for printing. These hard copies are used in color-coding
the images and grouping similar land-cover areas.5
After the images are color-coded, a special software program
is used to convert the classified digital imagery to a format that
is acceptable to the film recorder. The image is filmed in color
on a 9- by 9-inch (22.86cm) film format. Contact prints are then
produced for use in presentations or reports such as this. The
following section presents the land-cover classification images
and land-cover acreage statistics.
19
-------
RESULTS AND DISCUSSION
The aircraft MSS data assessment of the areas affected by
strip mining was performed in a single phase analysis through the
use of the DAS computer system. The use of the MSS computer-com-
patible tapes with the DAS provided reliable data in the detection
and inventory of strip mines. Results from the computer output
were correlated with photography acquired by the EPA lease plane
and were found to have excellent size and positional agreement.
The clustering algorithm proved to be very reliable in de-
tecting and distinguishing various earth surface conditions and
land-cover situations (mine-related and nonmine-related) in the
study areas. The results generated by the clustering algorithm
will provide needed information for future planning and data for
incorporation into the EPA data base. The data base will be an
important tool in the temporal analysis of MSS data. The data
reduction and processing techniques developed at this laboratory
will furnish additional capabilities and provide a speedy means of
strip mine assessment. Also, the acreage compilation algorithm
generates reliable statistical data that will be essential in
determining the progress of planting and reclamation work of coal
strip mines. Presently, no comparative analyses have been per-
formed to determine the correlation between the computer algorithm
for calculating acreages and the manual digitizer. However, the
dot-count method and the computer algorithm were evaluated and the
results differed by 2 percent. The real advantage of the automated
procedure is speed. Dot counting is too tedious and time consuming
to be utilized in a production atmosphere.
The following sections are devoted to a discussion on the
classification results of each of the strip mine studied in this
phase of the project.
COLSTRIP COAL STRIP MINE
The Colstrip Coal Strip Mine is located in the south-central
portion of Bighorn County, Montana (Figure 8). Colstrip is oper-
ated by Western Energy Company and Montana Power. Most of the
coal that is produced from this mine is shipped north via a 35-mile
rail spur to the Burlington-Northern main line and then to Billings,
Montana; St. Paul, Minnesota; and Chicago, Illinois, for steam
electric use. Estimated yearly consumption for the electrical gen-
erating plants is 4,336,000 tons. The remaining tonnage (500,000
tons) is used to fuel the power plant located in the nearby city
of Colstrip, Montana.
20
-------
' ? . * 1. A t 1
'' ' M •> - I' -L-
, . pf . ..t-a.« |
'-* r'il.fo S
- J.*.' 1
r-v : •> 'T»"«'o «
i* A "'' ^-4-<->.
~~— Ck-r; "i-
- ^ ' i
T C O L C ,'
1
\um
-4 1 c
"1 f
«.!«' — r"J
r~
p -4 •
,„ !__ "H°" __!' '. • ', ._i._
i-, ,1 ' r~~~ i r^"'^'!
| «; ---rp -'.""•'
'
V. 1__ jj~ T.
L *-t- — , 1, U
V r*"1- • % ! • °< , ri
J"\, T^ I.^IVI -V
X-1 < '
j < -'_
H
( 5
' GRANITE 1 .' (J"
_.r--^-/ *.,,
' ^ ^ ' ^"•'f"Vl«B&*. f-v" '
*• '""", \ /--/''"V'.|^ILV*'t *OW,'
S /^" "y*''
r
1" '^
0
L€NA V, ^
1 — J
'•.
LfKMbWAIl"^'
1 ?
UMIM«U i
•L>/
fTV
»-' H ,
'__
r
—
i
H 0 U T t »
.(i •"""'»"
">»^
f
j
'~! JUDITH
x-.| N
A Q H C R P
Wh'it S ion ! Y
i
- is«
:. •
f
i
«CI«na
H...J I,
• LA
I 1,
I— j
u i
i
j
. • j
NE H ,
1
> PHI
i
1 '
S"-^~^'~^' ii
\ .
F E R O i
~>
N f
1
i » _j ,
r ^
f -*
T *A
-j._._i-4- ,.
i •
i
,H
•Mil j
L L 1 P S -
1
i
|
r~
\ O A R F
1
\
N
N '
j O.pEL. [
[1 _J[^
VALLEY
JWMfM
/fcrj
j'*
\ "'
\
i
1 E L 0
i
A
i
i
Colstrip Coal Strip Mine
i
Sr.4-1' !^ ...
• 'y ^ 8«Vm«n J lirmpton [
i. < ' J"' * " K 1
, -.0 .0* ! , i
i -...-., •,/.
«i -1 ,
*
"4
j „(._ s .-% __
1
— I
I •• -r
" ..,„»., v° i
/ y\
^ .._!-, f
v L— _.
' T 1 r.l,,Tititl| I ^*
,— 'J.. r-* i '
Decker Coal Strip Mine
r '«« i
.. f . •°d"« , .
. i /
0 i
.*
S^; ^
! c (
i
i R 0 0 •
j
! Ml Po...
-hl^MI-
J
,-j"
! i
CONE |
Ci'tlt •
i DA
;
1
T • A 1 » 1 1 '— ^
~"! ii..
-i
;C°"T"
a
i
1 i
1
-j .U
\!
i 1
i
3 W D E R Rl'v E R
.;,o7;
--i
V I i T
- — ^"~>,^
ICH L AN D
Mfl
»»ON \
1
fc— .. j
»- tf
r~ ' ,
'"" .1
i •
^ r A L L o
j
1 I..,1...
CARTE
Figure 8. The locations of the Colstrip and Decker Coal Strip Mines
within the State of Montana.
-------
The mine has two coal seams (the Rosebud and the McKay) but
only the Rosebud is being mined actively now. The Rosebud seam
is approximately 27 feet (8.23 meters) in thickness. Reaching
this seam requires the removal of 30 (9.14 meters) to 160 feet
(48.77 meters) of overburden, averaging around 90 feet (27.43
meters) in thickness.6
Vegetation is of the ponderosa pine forest type. This vege-
tation type occurs mainly in eastern Montana and northeastern
Wyoming on uplands, ridges, and north slopes that have shallow loam
soils. Prominent species are ponderosa pine, snowberry, bluegrasses,
fescues, and June grass. These species are only fairly suitable but
have good availability for rehabilitation.
The various vegetation types of the area that were separated
by sequential clustering algorithm had to be grouped into three den-
sity classes to facilitate color-coding of the classes and to pro-
duce a reasonable facsimile of a vegetation map. Note how well the
various density classes on the MSS classification image (Figure 9)
compare with the aerial photograph (Figure 10). Also, compare
the positional accuracy of the coal stock piles and the extent and
location of the various water bodies, easily defined by the clus-
tering algorithm, to the photograph. Table 4 provides statistical
summaries by class for the entire MSS classification scene.
DECKER COAL STRIP MINE
The Decker Coal Strip Mine is located in Big Horn County,
Montana, approximately 25 miles (15.54km) north of Sheridan,
Wyoming, and adjacent to the Tongue River Reservoir (Figure 8).
Decker is operated by Peter Kiewit Sons, Company, and Pacific
Power and Light Company. Coal from this mine is shipped, by rail,
to Commonwealth Edison in Havanna, Illinois, for use in generating
electricity.
At the present time, the 52 foot (15.85 meter) Dietz No. 1
coal seam is being actively mined. The depth of an active pit
varies from 30 to 150 feet (9.14 to 45.72 meters) and requires
the removal of approximately 70 to 150 feet (21.34 to 45.72 meters)
of sandstone and shale.6
Vegetation of this area is the grassland-sagebrush type,
which occurs on open grassland of medium and short grass with
scattered sagebrush on silty clay-loam soils in southeastern
Montana and northeastern Wyoming. Dominant species are as follows:
western wheatgrass, crested wheatgrass, blue grama, green needle
grass, June grass, winterfat, and Indian rice grass.
Once again the various vegetation types were grouped to facil-
itate color-coding and to create a less cluttered image. As can
be seen from the aerial photograph (Figure 11), a considerable
amount of reclamation work has already begun. This is by far the
22
-------
Figure 9.
Land-cover classification map of the Colstrip Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data.
23
-------
Figure 10.
Aerial color-infrared photography of the Colstrip
Coal Mine acquired by NASA U-2 aircraft.
24
-------
TABLE 4. ACREAGE STATISTICS FROM THE ENTIRE SCENE OF THE COLSTRIP
COAL STRIP MINE
COLOR
CLASS
PIXEL
COUNT
PERCENT ACREAGE
SQUARE
MILES
Light Density 87070
Vegetation
Medium Density 218973
Vegetation
Soil 30101
High Density 71248
Vegetation
Exposed Soil 28890
Ponderosa Pine 11533
Water 8224
Coal 8514
Coal Covered 392
Soil
18.72 1269.00
47.09 3193.00
6.47 439.00
15.32 1039.00
6.21
2.48
1.77
1.83
0.08
421.00
168.00
120.00
124.00
6. 00
1.98
4.99
0.69
1.62
0.66
0.26
0.19
0.19
0.01
25
-------
Figure 11.
Aerial color-infrared photography of the Decker Coal
Strip Mine acquired by NASA U-2 aircraft.
26
-------
best example of the reclaiming procedure in progress. Note how
the mine is proceeding in a westerly direction (as indicated by
the location of spoil piles in relation to the coal seam and rec-
lamation work, e.g./ recontouring of spoil piles, seeding, fallow,
etc.). The reclaimed areas are planted with approximately 8 to 16
vegetation types that occur naturally in the area. This is the
reason that grouping the vegetation types is mandatory; otherwise,
the image would be too speckled to be of use. Although the re-
claimed areas are comprised of the same vegetation species as the
surrounding land, the DAS has the capability to allow the areas
to be assigned a specific, distinguishable color for ease of dis-
crimination (Figure 12). Table 5 is a statistical summary by
classes for the entire MSS classified area.
DAVE JOHNSTON COAL STRIP MINE
The Dave Johnston Coal Strip Mine is located in the east-
central portion of the State of Wyoming in Converse County (Figure
13) . Dave Johnston Mine is operated by Pacific Power and Light
Company and is termed a "captive mine." It is given this name
because it ships coal by rail to the Dave Johnston Power Plant
located in nearby Glenrock, Wyoming.
The mine has two coal seams that are being mined at the pre-
sent time. The youngest, the Badger seam, is approximately 16
feet (4.88 meters) in thickness, and the oldest, the School seam,
is approximately 37 feet (11.28 meters) in thickness. The average
depth of the active pit is about 140 feet (42.67 meters) with a
maximum depth of approximately 180 feet (54.86 meters).6
Vegetation is of the short-grass prairie type. This type
occurs on dry prairies in shallow soils in southeastern Montana
and northeastern Wyoming. Dominant species are grama, wheatgrasses,
and various needlegrasses. The species that characterize this type
have moderately poor suitability and fair availability for rehabil-
itation.
Both the MSS classified image (Figure 14) and the aerial
photography (Figures 15 and 16) verify that an extensive amount
of reclamation work has been performed on this westwardly advancing
coal strip mine. The coal seam is well defined in the MSS class-
ified image as is the lush vegetation located in the drainage pat-
terns of the area. Most of the transportation network of the mine
is well defined; moreover, the road that parallels the western
side of the mine also delineates the extent of the active mining
area as evidenced by the difference in vegetation densities when
moving from west to east on the classified image. Table 6 presents
the statistical results from the sequential clustering algorithm
classification for the entire MSS scene.
27
-------
Figure 12.
Land-cover classification map of the Decker Coal
Strip Mine generated from aircraft-acquired multi-
spectral scanner data.
28
-------
TABLE 5. ACREAGE STATISTICS FROM THE ENTIRE SCENE OF THE
DECKER COAL STRIP MINE
COLOR
CLASS
PIXEL
COUNT
PERCENT ACREAGE
SQUARE
MILES
Light Density 16636
Vegetation
Soil 30317
Soil 30728
Water 41331
Natural Vege- 75877
tation
Reclaimed Area 9323
Trees 5177
Agricultural 996
Land
Water 1759
Coal 1272
Coal Covered 481
Soil
7.78
14.17
14.37
19.32
35.47
4.36
2.42
0.47
0.82
0.59
0.22
243.00
442.00
448.00
603.00
1106.00
136.00
75.00
7.00
26.00
19.00
6.89
0.38
0.69
0.70
0.94
1.73
0.21
0.12
0.01
0.04
0.03
0.03
29
-------
I Wyodak Coal Strip Mine
o ii
Dave Johnston Coal Strip Mine
CHEYENNE
Figure 13.
The location of the Dave Johnston and Wyodak Coal
Strip Mines in the State of Wyoming.
30
-------
Figure 14.
Land-cover classification map of the Dave Johnston
Coal Strip Mine generated from aircraft-acquired
multispectral scanner data.
31
-------
Figure 15. Aerial color-infrared photography of the northern
portion of the Dave Johnston Coal Strip Mine ac-
quired by NASA U-2 aircraft.
32
-------
Figure 16.
Aerial color-infrared photography of the southern
portion of the Dave Johnston Coal Strip Mine ac-
quired by NASA U-2 aircraft.
33
-------
TABLE 6. ACREAGE STATISTICS FROM THE ENTIRE SCENE OF THE
DAVE JOHNSTON COAL STRIP MINE
COLOR
CLASS
PIXEL
COUNT
PERCENT ACREAGE
SQUARE
MILES
High Density
Vegetation
Light Density
Vegetation
Native Grasses
Soil
Soil
Medium Density
Vegetation
126842
81555
29.40
18.90
1849.00
1189.00
0.89
1.86
76265
64598
47066
30914
3692
295
17.68
14.97
10.91
7.17
0.86
0.07
1112.00
942.00
686.00
451.00
54.00
4.00
1.74
1.47
1.07
0.70
0.08
0.01
34
-------
WYODAK COAL STRIP MINE
The Wyodak Coal Strip Mine is located in Campbell County,
Wyoming, approximately 20 miles (32.18km) east of Gillette,
Wyoming (Figure 13). Wyodak is operated by Wyodak Resources Devel-
opment Corporation and produces in excess of 62,000 tons per month.
This coal is shipped, via rail, to electrical generating stations
in the cities of Osage and Wyodak, Wyoming, and Lead and Rapid
City, South Dakota.
Wyodak has two coal seams, the Anderson and the Canyon, which
are both approximately 40 feet (12.19 meters) thick. The average
overburden thickness is approximately 30 feet (9.14 meters) with
the minimum thickness of 15 feet (4.57 meters) and the maximum
thickness reaching 40 feet (12.19 meters). The depth of the active
pit ranges from 110 feet to 160 feet (33.53 meters to 48.77 meters)
with 85 feet (25.91 meters) the minimum depth.
Vegetation is of the short-grass prairie type. This vegeta-
tion type occupies dry prairies in shallow soils in southeastern
Montana and northeastern Wyoming. Dominant plant species are blue
grama, western wheatgrass, and various needlegrasses.
The land-cover classification (Figure 17) of the Wyodak Coal
Strip Mine was performed without any problems. The land-cover
classes of natural vegetation were grouped for the same reason as
cited earlier. Donkey Creek, which meanders through this area,
was readily resolved and classified as were the active pit area
and the haul roads leading to the tipple. The housing subdivision
located north of the western half of the mine was also spectrally
distinct, resulting in an urban category for the mine scene. The
accompanying aerial photograph (Figure 18) is provided for compara-
tive purposes. Table 7 summarizes the statistical analysis per-
formed on the data.
BLACK MESA COAL STRIP MINE
The Black Mesa Coal Strip Mine is located in the north-
northeastern corner of the State of Arizona, in Navajo and Apache
Counties (Figure 19). It is operated by Peabody Coal Company
headquartered in St. Louis, Missouri. Black Mesa is in the
500,000-and-over tonnage class and strips in excess of 20,000 tons
per day. Total tonnage mined in 1974 (last year of available
statistics) was 3,933,493 tons. Coal from this mine is transported
via pipeline to power plants outside the local area. Average
elevation on the lease area is approximately 6,500 feet (1,981
meters), placing it in the pinyon pine/juniper vegetation zone.
Dominant vegetation types within the lease area are: pinyon pine,
juniper, alkali sacatone, greasewood, snakeweed, sagebrush, blue
grama, winterfat, and rabbit brush.
35
-------
Figure 17.
Land-cover classification map of the Wyodak Coal Strip
Mine generated from aircraft-acquired multispectral
scanner data.
36
-------
Figure 18,
Aerial color-infrared photography of the Wyodak
Coal Strip Mine acquired by NASA U-2 aircraft.
37
-------
TABLE 7. ACREAGE STATISTICS FROM THE ENTIRE SCENE OF THE
WYODAK COAL STRIP MINE
COLOR CLASS
: Bar, Soil
3 Light Density
PIXEL
COUNT
63646
95893
PERCENT
24.95
37.58
ACREAGE
928.00
1398.00
SQUARE
MILES
1.45
2.18
Vegetation
4 High Density
Vegetation
5 Natural Vege-
tation
6 Medium Density
Vegetation
7 Agricultural
Land
8 Disturbed Area
9 Water
10 Industrial
11 Coal Covered
Soil
48944
4044
34469
325
781
5369
614
81
55
835
85
19.18
1.59
13.51
0.13
714.00
59.00
503.00
5.00
1.12
0.09
0.79
0.01
0.31
2.10
0.24
0.03
0.02
0.32
0.03
11.
78.
9.
1.
1.
12.
1.
00
00
00
00
00
00
00
0.02
0.12
0.01
0.00
0.00
0.02
0.00
38
-------
.
•
u
i
:
Black Mesa Coal Strip Mine
E)
i
Y
A
y~'J<
"
* i
\ i
i
i '-•
M A i M
j
Kl^Kl:
1
j
j
ix-
C O C O
~"v
~\
i
L
AVAR
R n
.-•v
PHOENJ^I
A R 1 C .O
i
P
j
1
i
P
,.
f
NINO
*
H..SI.II
j
A ,
I
\
\ 38
r ,A
r *
P A
1 N
M
i
' S f
O
(
1 -
t
1 >
1
~ ' "
| HoltHOok
2
J N '
1
[
G 1 L A
\ 1
\ ^
\ ; |
.
'-!
. . J
i «
i •
A !
cc
•" *" " 1
^N T A CRUZ '
Ul
I
o
1
<
» SlJM.so
.
k.
i sJ~~
-vn -
U
1
2
LJ
J CllltD"! U
W-H A M\ «
SitiOfd (^
j
C H 1 S E
Figure 19.
The location of the Black Mesa Coal Strip Mine
within the State of Arizona.
39
-------
By comparing the classification image in Figure 20 with the
aerial photograph in Figure 21 it is easy to verify the validity of
the classified image. The most prominent vegetation type, pinyon
pine (dark green on the MSS image), is mapped fairly accurately.
It was noted that the percent of ground cover for most vegetation
species around the mine rarely exceeded 50 percent as evidenced
by the white and gray colors that represent exposed soil classes
in Figure 21. Note also how the spoil piles located on the west-
ern edge of the mine were resolved and classified. Again, Table 8
provides the needed statistical summarization of the classification
results by individual classes from the entire scene.
NUCLA COAL STRIP MINE
The Nucla Coal Strip Mine is located in Montrose County,
Colorado, approximately 100 miles southeast of Grand Junction,
Colorado (Figure 22). Nucla is operated by Peabody Coal Company
with headquarters located in St. Louis, Missouri. Nucla is a
small strip mining operation and is in the 100,OOO-to-199,999
tonnage class.
Vegetation of the area is of the pinyon pine/juniper type.
Blue grama is the most abundant and frequent understory species.
Common species include several gramas, western wheatgrass, galleta,
sand dropseed, and June grass. Associated shrubs are often rabbit
brush, big sagebrush, snakeweed, and serviceberry.
Figure 23 illustrates that the Nucla Coal Strip Mine is a
relatively small operation bordered on the north, east, and south
sides by agricultural lands and bordered on the west by its dom-
inant vegetation type — the pinyon pine/juniper association.
This major concentration of pinyon pine/juniper vegetation, as
well as small pockets within the scene and other tree species, were
easily separated from other vegetation types (Figure 24). The
agricultural lands that comprise approximately 50 percent of the
scene are planted in alfalfa, which is used as silage by the
farmers. Both figures verify that little has been done to reclaim
the previously mined areas. Table 9 was prepared to provide the
reader with statistical information concerning the result of the
classification.
40
-------
Figure 20.
Land-cover classification map of the Black Mesa
Coal Strip Mine generated from aircraft-acquired
multispectral scanner data.
41
-------
Figure 21.
Aerial color-infrared photography of the Black Mesa
Coal Strip Mine acquired by NASA U-2 aircraft.
42
-------
TABLE 8. ACREAGE STATISTICS FROM THE ENTIRE SCENE OF THE
BLACK MESA COAL STRIP MINE
COLOR
CLASS
PIXEL
COUNT
PERCENT ACREAGE
SQUARE
MILES
0 Recontoured
Area
1 Bare Soil
2 Light Density
Vegetation
3 Medium Density
Vegetation
4 High Density
Vegetation
6 Natural Vege-
tation
10 Coal Covered
Soil
12 Bare Soil
13 Bare Soil
15 Trees
11049
60871
146445
72465
45553
64232
27960
15259
3730
5339
39728
1062
2.23
12.33
29.66
14.56
9.23
13.01
5.66
3.09
0.76
1.08
8.05
0.22
161.00
887.00
2135.00
10.56
664.00
927.00
408.00
222.00
54.00
78.00
580.00
15.00
0.25
1.39
3.34
1.65
1.04
1.46
0.64
0.35
0.08
0.13
0.90
0.02
43
-------
O F F! A
' 1 O B L
•••«. j
j
J
Me«w
N C 0
["
SB
1
rH
J 4
* • \^
i
G R
.
\ LARIMER |
FortCo.lrs° j
• 1
., ^ V
r
1 I E.|*
G A R F I E L D
GkniKKXl Springs '
* N °
| SEDOWCK
I L O 0 » N
PHILLIPS
-} ^, 10 1^ :
E * G L E r-L^T £- —i f :P"—-;—
KIT CARSON
m
r~
-*7 Mnnlrrr
Moiiifcn. I
MONTR^SE ('
•C l_!
£J*_*LL2_Mi>LL _*"">_._.."""c'"
DOLORES
«ONTE2UMA ,-
term '*-
rr
a <
GUN
• kr City
|
1
| M|
-rl-
1CH U
«•* ,,--
-it- -
loal Strip
N 1 S O N *"\
L S*'idj
Gunniion v^
S A G U A c" H
ERAL ^ t>l l?o.rt
j RIO GRANDE t
._.; — :_.^
LETA'ICONEJ
0 "RT
Mine 4-«
"- •
»/ FREMONT
_•
X /yC'>1
\ . r
\ *eslti>H* ,
E \
Vx
v-- *,
-J II
.' H U E R
ALAMOSA '
«"•*»• / \
\ > -
\ COST1LLA Y
0 S I 5,^
j
C
"*"lt"~
- i
gColOTada Spnngi
"j E L. PASO j
"I
i 1'
1
,"•** j
PUEBLO
|
,''^-^.J
F A N O /
/
1
r'C' 'A s A ^
D! U-
i
j ;
LINCOLN
1 '
K
CROWLCV \
- " L'" i
UJ«nu° BEN
OT E RO
1
, - — -' f
1 M A S
;
|
i
H E Y E N N E
'
O; W A
S,
1
„,., 1 l l
T ' p R O * E R
j
!
L — - --
B A C A
1
-
Figure 22.
The location of the Nucla Coal Strip Mine in the
State of Colorado.
44
-------
Figure 23.
Aerial color-infrared photography of the Nucla Coal
Strip Mine acquired by NASA U-2 aircraft.
45
-------
•«
'•^ ' •"*•" ?>
'#>'-.: t'i^Ofe^
..»/ •„;.«• •" -V'"-, - r: »<*
• «
Figure 24.
Lan.d cover classification map of the Nucla Coal
Strip mine generated from aircraft acquired
multispectral scanner data.
46
-------
TABLE 9. ACREAGE STATISTICS DERIVED FROM THE ENTIRE SCENE OF
THE NUCLA COAL STRIP MINE
COLOR
CLASS
PIXEL
COUNT
PERCENT ACREAGE
SQUARE
MILES
0 Recontoured 1509
Area
1 Trees 51366
2 Light Density 49465
Vegetation
3 High Density 25984
Vegetation
4 Natural Vege- 28610
tation
5 Medium Density 34119
Vegetation
6 Agricultural 13487
Land
7 Agricultural 29689
Land
10 Coal 138
12 Agricultural 30620
Land
15 Water 152
0.58
19.38
18.65
9.80
10.08
12.87
5.20
11.22
22.00
749.00
721.00
379.00
417.00
497.00
197.00
433.00
0.03
1.17
1.13
0.59
0.65
0.78
0.31
0.68
0.05 2.00 0.00
11.55 446.00 0.70
0.05
2.00
0.00
47
-------
SUMMARY
In the summer of 1977, aircraft multispectral scanner data
and color-infrared photography were collected over six coal strip
mines in five western States (Wyoming, Montana, Colorado, Arizona
and New Mexico) at an altitude of 12,000 feet (3,660 meters).
These data were used by image analysts of EPA's Environmental
Monitoring and Support Laboratory in Las Vegas, Nevada, to develop
an interactive computer processing procedure for producing "quick
turnaround" land-cover classification maps and class acreage
statistics for coal strip mines. These classified data sets will
serve as the initial input to a geographically referenced data
base which will provide a cost-effective means of performing change
detection analyses of coal strip mine reclamation activities.
The procedure developed at the EPA laboratory uses a sequen-
tial clustering algorithm as an approach to pattern recognition.
Basically the clustering program groups pixels with similar statis-
tical characteristics. Each pixel of the digital data is compared
to the existing statistical profiles to determine which population
it best fits. This clustering procedure is faster and requires
less computer time than the training field (supervised) approach
to pattern recognition. Also, the classified data set produced by
the clustering algorithm can be used to facilitate the selection
of homogeneous training fields if an analyst chooses to use the
supervised approach to pattern recognition. Overall, this pro-
cedure will provide state and federal agencies with an effective
and efficient means of monitoring reclamation activities and mine
progress and direction.
Finally, the aircraft acquired multispectral scanner data at
an altitude of approximately 12,000 feet (3,660 meters) appears to
possess sufficient spatial resolution (element size) and spectral
resolution (land width and location) for generating land-cover
classifications of western coal strip mines using the sequential
clustering technique.
48
-------
REFERENCES
1. Fogiel, Max. Modern Energy Technology. Research and Educa-
tion Association, New York, New York, 1975. pp. 947-979.
2. Anderson, James E., and C. E. Tanner. Remote Monitoring of
Coal Strip Mine Rehabilitation. National Technical Information
Service, Springfield, Virginia, 1978. pp. 1-2.
3. Whitley, S. W. Low-Cost Analysis Systems for Processing
Multispectral Scanner Data. NASA TR R-467, NASA, Earth
Resources Laboratory, Slidell, Louisiana, 1976. pp. 9-11.
4. U.S. Geological Survey. A Land Use and Land Cover Classifica-
tion System for Use with Remote Sensor Data. Geological Sur-
vey Professional Paper 964, Government Printing Office,
Washington, D.C., 1967.
5. Pooley, John. Unsupervised Sequential Cluster Program.
NASA, Earth Resource Laboratory, Slidell, Louisiana, 1976.
6. Nielsen, George F. Keystone Coal,Industry Manual. McGraw-
Hill Publications, New York, New York, 1975. pp. 732-960.
49
•&U.5. GOVERNMENT PRINTING OFFICE: 1979—684-558,
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-080
I. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
COMPUTER PROCESSING RESULTS OF SCANNER DATA
OVER SELECTED COAL STRIP MINES
5. REPORT DATE
March 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Charles E. Tanner
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Lockheed Electronics Company, Inc.
Remote Sensing Laboratory
Las Vegas, Nevada 89114
10. PROGRAM ELEMENT NO.
INE 625C
11. CONTRACT/GRANT NO.
EPA 68-03-2636
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency—Las Vegas, NV
Office of Research and Development
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
13. TYPE OF REPORT AND PERIOD COVERED
Final (01-01-78/6-30-78)
14. SPONSORING AGENCY CODE
EPA/600/07
15. SUPPLEMENTARY NOTES
G. J. D'Alessio, Project Officer, Western Energy/Environmental Monitoring Study,
U.S. Environmental Protection Agency, Washington, D.C. 20460
16. ABSTRACT
Aircraft multispectral scanner data over six coal strip mines in the States of
Wyoming, Montana, Colorado, and Arizona were processed on the data analysis mini-
computer system using a clustering approach to automatic pattern recognition. The
classification results demonstrated that a Level I hierarchy of vegetation, manmade
features, and disturbed areas is easily obtained with a minimum expenditure of time.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Aerial photography
Land use
Monitoring
Photographic reconnaissance
Photointerpretation
Sterophotography
Multispectral scanner
Ground observations
9B
14E
43F
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
60
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
A04
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |