-------
I
I
f
I
I
I
I
I
I
Farther Evaluations of Selected Periods. The aforementioned
designations of a typical year, dry year, and wet year were based upon the
period 1966-1978 since this is the period covered by the NWS hourly rainfall
records acquired for the EPA Chesapeake Bay Program. The decision to
restrict the rainfall records to the period 1966-1978 was one based
primarily on costs in an effort to keep the budget for the acquisition and
analysis of NWS rainfall tapes from becoming prohibitive. Since NWS hourly-
rainfall records for the Chesapeake Bay Basin recording gages are currently
available for 17 additional years (i.e., 1949-1965), it was necessary to
screen observed April-October flow-duration curves for each year in the
period 1949-1978 to demonstrate that the selected years were the most
appropriate for model production runs.
For the typical year comparisons, the flow-duration curves for the
period 1966-1978 were found to adequately approximate the curves for the
full 30-year period (1949-1978) covered -by NWS hourly rainfall records,
indicating that selections of a typical year based on comparisons with the
shorter period were certainly appropriate. Further, 1974 exhibited as good
an agreement with the 1949-1978 period as did the only two years (1950 and
1961) in the earlier period which merited consideration as a typical year.
For the dry year comparisons, three earlier years (1954, 1957, and 1963)
merited consideration as a design dry year. However, although the total
April-October streamflow volumes for at least one (1963) of these earlier
years are less than April-October 1966 volumes, the similar and even higher
frequencies of extreme low flows during 1966 reinforces its selection for
the dry year production runs.
For the wet year comparisons, it was determined that none of the earlier
years produced April-October flow-duration curves which approached 1972 and
only one year (1960) approached 1975. Inspection of the flow-duration plots
indicated that 1975 is typically very similar to 1960 in terms of the
frequency of high flows and superior to 1960 in terms of the frequency of
low-to-moderate flows. Thus, the comparison of flow-duration curves for the
period 1949-1978 supports the selection of 1975 for the wet year production
runs.
Interpretation of Model Output
For each production run, the Basin Model can be operated in its entirety
and loading output for the April-October period (plus a one-month antecedent
period) will be input to the Bay system models to project receiving water
quality impacts. In addition, loadings delivered to the Bay system can be
tabulated on a seasonal, monthly, and daily basis for general assessments of
management strategies.
By subjecting the output time series to frequency analyses,
concentration-frequency relationships at the fall line can be derived for
each management strategy simulation. Concentration-frequency relationships
for typical year production runs may be assumed to represent the average
-196-
-------
impacts over a thirty-year period. In other words, the number of days (or
hours) with water quality criteria violations projected for the typical year
simulation may be assumed to represent the average number of violations per
year if the management strategy were applied over a thirty-year period
characterized by a good mix of wet years, dry years, and average years.
Likewise, concentration-frequency relationships for dry year and wet year
production runs define the duration of water quality criteria violations for
two periods which can be expected to occur relatively infrequently over a
thirty-year period. Since the selected wet year and dry year represent
extreme conditions, one would expect the frequency of any water quality
criteria violations during these two simulation periods to be greater than
under typical year simulations which are assumed to represent average algal
growing season conditions in the receiving waters.
By relating the simulated concentration-frequency relationships to
beneficial use criteria, water quality damages may be inferred from fall
line concentration simulations. Reductions in water quality damages
achieved by a particular management strategy can then be expressed in terms
of reductions in the frequency of undesirable concentrations. However, it
should be cautioned that since the models are idealized representations of
the prototype river systems, water quality output is best viewed in a
relative sense to provide insights into differences among management
strategies.
Framework for Model Production Runs
The first set of production runs to be carried out with the Basin Model
package were daily loading projections for the existing land use pattern and
the land use pattern projected for the Year 2000. As previously indicated
the existing land use pattern is based upon LANDSAT data interpretations.
The 2000 land use projection was derived by increasing the existing urban
land use in each sub-basin by the ratio of the 2000 population projection to
the 1980 population and reducing forest land to represent the consumption of
undeveloped land by urban development. Wastewater discharges and water
supply diversions for the Year 2000 were generally derived by multiplying
the 1980 values by the ratio of population values. The assumption that only
forest land will be consumed by urban development occurring between 1980 and
2000 is a conservative one designed to ensure a worst case projection of
Year 2000 water quality impacts. This assumption preserves the agricultural
land use distribution at 1980 levels while increasing urban development to
reflect population increases, thereby resulting in what is probably a
conservatively high estimate of nonpoint pollution loadings in the Year
2000. The comparison of existing and future land use impacts on the Bay's
estuarine system should provide considerable insight into the impacts of
urban development on Baywide water quality. These two model runs will also
establish the baseline receiving water quality conditions (i.e.,
concentration-frequency and loading-frequency relationships) for evaluations
of alternate management strategies.
-197-
-------
I
I
f
I
I
I
I
I
I
V
I
I
I
I
1
Included among the management strategy evaluations which were the
subject of model production runs are the following:
A. General Comparisons of Point Source and Nonpoint Source Loadings;
The contributions of point and nonpoint sources to fall line
loadings were determined by operating the Basin Model without any
point source loadings (i.e., with only the flow, temperature, and
dissolved oxygen associated with each wastewater discharge). The
difference between simulated fall line loadings "with" and
"without" point source inputs represents the fraction of the fall
line load that can be attributed to point sources.' Since these
production runs were intended to evaluate both point and nonpoint
sources, they were based on all three hydrologic conditions (i.e.,
typical year, dry year, and wet year).
B. General Comparisons of Water Quality Benefits Promised by Cropland
Best Management Practices (BMP's); Initial production runs
focussed on the relative loading contributions of cropland areas.
The Basin Model was operated without surface runoff loadings from
forest, pasture, and urban land uses, and no change in subsurface
flow loadings. The difference between fall line loadings for this
run and previous runs represented the contribution of non-cropland
land uses. In general, the non-cropland contributions were not
more than about one-third of the total fall line load of N and P
for average and wet years, indicating the significance of cropland
contributions. The principal cropland BMP which was studied
involved converting high tillage cropland areas to low tillage
cropland. The potential benefits of this cropland BMP were
determined by comparing Basin Model projections "with" and
"without" the low tillage BMP.
C. Assessments of Locational Differences in Nonpoint Pollution Loading
Delivecy to the Bay's Sstuarine System; As previously indicated,
it would be incorrect to assume that all nonpoint pollution
loadings released into the Basin's receiving waters are transported
to the Bay's estuaries. The Basin Model was used to develop
"pollutant delivery ratios" for different sections of the three
major river basins. By removing the nonpoint pollution loadings
produced by sub-basin clusters one cluster at a time (i.e.,
including only flow, temperature, and dissolved oxygen time series
for clusters deleted from the simulations) , the Basin Model was
used to estimate the fraction of sub-basin nonpoint pollution
loadings which is delivered to the Chesapeake Bay's estuarine
system. The difference between simulated annual loadings at the
mouth of each river basin for "with" and "without" sub-basin
cluster conditions may be assumed to be the total annual nonpoint
pollution loadings delivered to the Bay's estuarine system by each
sub-basin cluster. Pollutant delivery ratio estimates for
different sections of the Chesapeake Bay Basin can be used by
management agencies to identify those areas where nonpoint
-198-
-------
pollution controls promise the greatest benefit. For example, maps
showing pollutant delivery ratio contours could be developed for
use in comparing the Chesapeake Bay impacts of nonpoint pollution
loads from the various physiographic provinces in the Basin and to
determine whether nonpoint pollution controls should be
concentrated in downstream sections (e.g., Coastal Plain and
Piedmont provinces) of the Basin or extended into upland sections
(e.g., Appalachian Ridge & Valley and Appalachian Plateau
provinces) as well. Since these production runs are restricted to
nonpoint source impacts, they were based on typical year and wet
year meteorologic conditions only.
D. Assessments of Locational Differences in Point Source Pollution
Delivery to the Bay's Estuarine System; Using the same approach
outlined under C, loadings from point source clusters were removed
from the Basin Model one cluster at a time to develop estimates of
each cluster's contribution to the total fall line load. Estimates
of locational differences in the delivery of point source loadings
to the fall line were developed for different sections of the three
major river basins. Point source contributions were also defined
for each state traversed by the river basin. These production runs
were based upon all three hydrologic conditions (i.e., typical
year, dry year, and wet year).
E. Assessments of Point Source Management Strategies: The impacts of •
higher wastewater treatment levels on fall line loadings were
assessed with the Basin Model. The majority of the wastewater
management strategies involved more stringent effluent levels for
phosphorus. At least one management strategy involved more
stringent treatment levels for both nitrogen and phosphorus. These
production runs were typically restricted to dry year and average
year conditions.
-199-
-------
I
I
f
I
I
I
I
I
I
b
I
I
I
1
i
REFERENCES
1. Hydrocomp, Inc., "The Occoquan Basin Computer Model: Calibration,
Verification, and User's Manual," prepared for Northern Virginia
Planning District Commission, Falls Church, VA, May 1978.
2. Johanson, R.C., Imhoff, J.C., and Davis, H.H., "Users Manual for
Hydrological Simulation Program - FORTRAN (HSPF)," BPA-600/9-80-015,
U.S. Environmental Protection Agency, Environmental Research
Laboratory, Athens, GA, April 1980.
3. Linsley, R.K. and Crawford, N.H., "Continuous Simulation Models in
Urban Hydrology," Geophysical Research Letters, Vol. 1, No. 1,
May 1974, pp. 59-62.
4. Linsley, R.K. and Crawford, N.H., "Computation of a Synthetic
Streamflow Record on a Digital Computer," Journal of International
Association of Scientific Hydrology, No. 51, 1960, pp. 526-538.
5. Donigian, A.S. and Crawford, N.H., "Modeling Nonpoint Pollution from
the Land Surface," EPA-600/3-76-083, U.S. Environmental Protection
Agency, Environmental Research Laboratory, Athens, GA, July 1976.
6. Hartigan, J.P., Quasebarth, T.F., and Southerland, E., "Use of
Continuous Simulation Model Calibration Techniques to Develop Nonpoint
Pollution Loading Factors," Proceedings of Stormwater and Water Quality
Management Modeling Users Group Meeting; March 25-26, 1982,
EPA-600/9-82-015, U.S. Environmental Protection Agency, Environmental
Research Laboratory, Athens, GA, 1982, pp. 101-127.
7. Northern Virginia Planning District Commission and Virginia Polytechnic
Institute and State University, "Occoquan/Four Mile Run Nonpoint Source
Correlation Study," Final Report prepared for Metropolitan Washington
Council of Governments, Washington, D.C., July 1978.
8. Northern Virginia Planning District Commission, "Guidebook for
Screening Urban Nonpoint Pollution Management Strategies," prepared for
Metropolitan Washington Council of Governments, Washington, D.C.,
November 1979.
9. Hartigan, J.P., et al., "Calibration of Urban Nonpoint Pollution
Loading Models," Proceedings of ASCE Hydraulics Division Specialty
Conference on Verification of Mathematical and Physical Models in
Hydraulic Engineering, American Society of Civil Engineers, New York,
NY, August 1978, pp. 363-372.
10. Grizzard, T.J., Hartigan, J.P., and Randall, C.W., "The Development of
Water Quality Management Plans Using Data from Automatic Stormwater
Monitoring Networks," Proceedings of IAWPR International Workshop on
Water Quality Monitoring, held at Munich, Germany, June 22-25, 1981.
f
-200-
-------
11. U.S. Geological Survey, "Water Quality of the Three Major Tributaries
to the Chesapeake Bay, January 1979 - April 1981: Estimated Loads and
Examinations of Selected Water Quality Constituents," prepared for
USEPA Chesapeake Bay Program, November 1981.
12. Hartigan, J.P., et al. , "Areawide and Local Frameworks for Urban
Nonpoint Pollution Management in Northern Virginia," Stormwater
Management Alternatives, Tourbier, J.T. and Westmacott, R. , eds. ,
University of Delaware, DE, April 1980, pp. 211-245.
13. Biggers, D.J., Hartigan, J.P., and Bonuccelli, H.A., "Urban Best
Management Practices (BMP's) : Transition from Single-Purpose to
Multipurpose Stormwater Management," Proceedings of Seventh
International Symposium on Urban Storm Runoff, Report No. UKYBU121,
College of Engineering, University of Kentucky, Lexington, Kentucky,
July 1980, pp.. 249-274.
14. Northern Virginia Planning District Commission, "Follow-up Assessments
of Alternate AWT Operating Rules with Occoquan Basin Computer Model,"
prepared for Camp, Dresser, McKee, Inc., consultants to Virginia State
Water Control Board, Richmond, VA, February 1980.
15. Northern Virginia Planning District Commission, "Recommended Operating
Procedure for Nitrogen Treatment Facilities at UOSA AWT Plant,"
prepared for Upper Occoquan Sewage Authority, Fairfax County Water
Authority, and Occoquan Watershed Monitoring Laboratory, May 1981.
16. Northern Virginia Planning District Commission, "Modeling Study of
Nonpoint Pollution Loadings from Potomac Embayment Watersheds," Final
Report prepared for Virginia State Water Control Board, Richmond, VA,
March 1981.
17. CH2M~HiH' "Stormwater Management: A Comprehensive Study of the
Muddy Branch and Seneca Creek Watersheds," prepared for Montgomery
County (MD) Planning Board, Silver Spring, MD, April 1975.
18. Montgomery County Planning Board, "Conservation and Management Report
on the Rock Creek Watershed, Montgomery County," Silver Spring, MD,
1978.
19. Metropolitan Washington Council of Governments, "The Seneca Creek
Watershed Model: Documentation of the Hydrologic Calibration and
Verification," Washington, D.C., April 1982.
20. Northern Virginia Planning District Commission, "Technical Plan for a
Modeling Study of Nonpoint Pollution Loadings and Transport in the
Chesapeake Bay Basin," Interim Report prepared for USEPA Chesapeake Bay
Program, Annapolis, MD, February 1981.
-201-
-------
I
I
f
I
I
I
I
I
I
I
I
I
I
I
21. Roffe, K.A., "Computerized Mapping for Assessment of Environmental
Impacts," presented at Annual Conference of American Institute of
Planners, held at Kansas City, MO, October 8-12, 1977.
22. Arbruster, J.T., "Flow Routing in the Susquehanna River Basin: Part I
- Effects of Raystown Lake on the Low-Flow Frequency Characteristics of
the Juniata and Lower Susquehanna Rivers, Pennsylvania," U.S.
Geological Survey Water Resources Investigations 77-12, USGS,
Harrisburg, PA, April 1977.
23. Wischmeier, W.H. and Smith, D.D., "Predicting Rainfall Erosion Losses
from Cropland East of the Rocky Mountains," Agricultural Handbook
No. 282, Soil Conservation Service, U.S. Department of Agriculture,
Washington, D.C., May 1965.
24. Northern Virginia Planning District Commission, "Assessments of Water
Quality Management Strategies with the Occoquan Basin Computer Model,"
prepared for Occoquan Technical Review Committee, January 1979.
25. Hydrocomp, Inc., "Hydrocomp Simulation Programming: Water Quality
Simulation Operations Manual," Palo Alto, Cal., 1977.
26. Zison, S.W., et al., "Rates, Constants, and Kinetics Formulations in
Surface Water Quality Modeling," EPA-600/3-78-105, Environmental
Research Laboratory, U.S. Environmental Protection Agency, Athens, GA,
December 1978.
27. Northern Virginia Planning District Commission, "Water Quality'Modeling
Study of Goose Creek, Broad Run, and Sugarland Run Watersheds," Falls
Church, VA, June 1980.
28. Northern Virginia Planning District Commission, "Occoquan Basin
Computer Model: Summary of Calibration Results," Falls Church, VA,
January 1979.
29. Lumb, A.M. and James, L.D., "Runoff Files for Flood Hydrograph
Simulation," Journal of the Hydraulics Division, ASCE, Vol. 102,
No. HY10, October 1976, pp. 1515-1531.
30. Lumb, A.M., "UROS04: Urban Flood Simulation Model, Part 1:
Documentation and Users Manual," School of Civil Engineering, Georgia
Institute of Technology, Atlanta, GA, March 1975.
31. Ross, G.A., "The Stanford Watershed Model: The Correlation of
Parameter Values Selected by a Computerized Procedure with Measureable
Physical Characteristics of the Watershed," Research Report No. 35,
Water Resources Institute, University of Kentucky, Lexington, Ky., 1970.
32. Hydrocomp, Inc., Hydrocomp Simulation Programming: Hydrology
Simulation Operations Manual, Palo Alto, CA, January 1976.
-202-
-------
33. Winter, T.C., "Uncertainties in Estimating the Water Balances of
Lakes," Water Resources Bulletin, Vol. 17, No. 1, February 1981,
pp. 82-115.
34. Cavacas, A., et al., "Hydrologic Modeling for Studies of Pollutant
Loadings and Transport in Large River Basins," Proceedings of
Stormwater and Water Quality Management Modeling Users Group Meeting:
March 25-26, 1982, EPA-600/9-82-015, U.S. Environmental Protection
Agency, Environmental Research Laboratory, Athens, GA, 1982, pp. 69-89.
35. American Public Health Association, Standard Methods for the
Examination of Water and Wastewater, 14th Edition, APHA, Washington,
D.C., 1976.
36. U.S. Environmental Protection Agency, "Methods for Chemical Analysis of
Water and Wastes," EPA-625/6-74-003, National Environmental Research
Center, Washington, D.C., 1974.
37. Hollander, M. and Wolfe, D.A., Nonparametric Statistical Methods, John
Wiley and Sons, New York, NY, 1973.
38. Northern Virginia Planning District Commission, "Reverification of
Occoquan Computer Model for Post-AWT Conditions," prepared for NVPDC
Occoquan Basin Nonpoint Pollution Management Program, Annandale, VA,
November 1982.
39. Smullen, J.T. and Taft, J.L., "Nutrient and Sediment Sources to the
Water Column of the Chesapeake Bay," USEPA Chesapeake Bay Program
Synthesis Paper, U.S. Environmental Protection Agency, Annapolis, MD,
February 1982.
40. Northern Virginia Planning District Commission, "NURP Study Quarterly
Report: October - December 1981," prepared for Metropolitan Washington
Council of Governments, Washington, D.C., February 1982.
41. Wright, R.M. and McDonnell, A.J., "In-Stream Deoxygenation Rate
Prediction," Journal of Environmental Engineering Division, ASCE,
Vol. 105, No. EE2, April 1979, pp. 323-335.
42. Bansal, M.K., "Nitrification in Natural Streams," Journal of Water
Pollution Control Federation, Vol. 48, No. 10, October 1976,
pp. 2380-2393.
43. Roy F. Weston, Inc., "Susguehanna River Water Quality Study Pertaining
to the Safe Harbor and Holtwood Hydroelectric Projects: Interim
Report," prepared for Penna. Power and Light Co. and Safe Harbor Water
Power Corporation, February 1982.
44. Personal communication from Stuart Freudberg, Metropolitan Washington
Council of Governments, June 22, 1982.
-203-
-------
I
I
l'
I
I
I
I
45. Huber, W., "Discussion Remarks," Proceedings of Seminar on Design Storm
Concept, Ecole Polytechnique de Montreal, Montreal, Quebec, 1979.
46. Southerland, E., "A Continuous Simulation Modeling Approach to Nonpoint
Pollution Management," Dissertation presented to Virginia Polytechnic
Institute and State University, Blacksburg, VA, in December 1981, in
partial fulfillment of the requirements for the degree Doctor of
Philosophy in Environmental Sciences and Engineering.
-204-
-------
-------
I
I
f
I
I
I
I
I
I
APPENDIX A
SUBBASIN MAPS
WITH
PRECIPITATION GAGE LOCATIONS
-------
Figure Al
186980
O
073570
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-l
-------
Figure A2
180470
0
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
)
MAP
NOT TO SCALE
A-2
-------
I
I
Figure A3
f
I
I
I
I
I
I
LEGEND
O HOURLY RAINGAGE
DAILY RAINGAGE
SEGMENT NUMBER
xxxxxx GAGE NUMBER
A-3
MAP
NOT TO SCALE
-------
Figure A4
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
I
MAP
NOT TO SCALE
A-4
-------
Figure AS
079595
o
LEGEND
O HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
f
MAP
NOT TO SCALE
A-5
-------
Figure A6
363699
Q.
364166
O
12
361(778
o
361589
O
189030
o
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
X SEGMENT NUMBER
xxxxxx GBGE NUMBER
1
MAP
NOT TO SCALE
A-6
-------
Figure A7
464763
O
O
A
LEGEND
HOURLY -RAINGAGE
DAILY RAINGAGE
X SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-7
-------
Figure AS
O
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-8
-------
I
I
f
I
I
I
I
I
I
I
I
I
I
I
Figure 9
441322
446533
A
o
LEGEND
HOURLY RA1NGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
\
MAP
NOT TO SCALE
A-9
-------
Figure 10
18^030
O
O
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
)
MAP
NOT TO SCALE
A-10
-------
I
I
f
I
I
I
I
I
I
Figure 11
I
I
I
I
442044
O
448046
465739
O
445120
O
O
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-11
-------
Figure 12
O
O
A
.LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
I
MAP
NOT TO SCALE
A-I;
-------
I
I
f
I
I
I
I
I
I
Figure 13
I
I
I
I
I
f
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
f
MAP
NOT TO SCALE
A-13
-------
Figure 14
188065
O
3*3298
O
362721
O
20
467730
O
184030
O
367965
O
0
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-14
-------
I
I
f
I
I
I
I
I
I
Figure 15
I
I
1
I
I
362298
O
O
A
l- — —
LEG
HOURLY RAINGAGE
DAILY RAINGAGE
END
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
MAP
NOT TO SCALE
A-15
-------
Figure 16
366916
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
X SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
I
MAP
NOT TO SCALE
A-16
-------
I
I
f
I
I
I
I
I
I
Figure 17
365817
o
LEGEND
HOURLY RAINGAGE
A DAILY RAINGAGE
X SEGMENT NUMBER
xxxxxx GAGE NUMBER
t
MAP
NOT TO SCALE
A-17
-------
Figure 18
A-18
-------
Figure 19
3622^5
O
36^001
O
O
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
»
1
1
MAP
NOT TO SCALE
A-19
-------
Figure 20
369408
O
O
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
t
1
MAP
NOT TO SCALE
A-20
-------
Figure 21
368905
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
1
MAP
NOT TO SCALE
A-21
-------
Figure 22
o
A
LEGEND
HOURLY RAINGAGE
DAILY RAINGAGE
)( SEGMENT NUMBER
xxxxxx GAGE NUMBER
1
I
MAP
NOT TO SCALE
A-22
-------
1
Figure 23
1
m
LEGEND
HOURLY RAINGAGE
A DAILY RAINGAGE
SEGMENT NUMBER
xxxxxx GAGE NUMBER
f
MAP
NOT TO SCALE
A-23
-------
APPENDIX B
LANDSAT LAND USE DATA ACQUISITION AND ANALYSIS
LANDSAT OVERVIEW
Landsat is a satellite data acquisition system dedicated to earth
resource investigation. The three Landsat satellites are in circular, near
polar orbits at an altitude of 570 miles (920 km). The orbit brings an
individual satellite over the same point on the earth, at the same time of
day, every 18 days. The satellites pass over the equator from north to
south at approximately 9:30 a.m.
The primary sensor on Landsat is the Multispectral Scanner (MSS). The
MSS collects data for each Landsat scene over an area 185 x 185 km (13,000
square miles). The data is collected in the following spectral bands:
Band 4 0.5-0.6 micrometers
Band 5 0.6-0.7 micrometers
Band 6 '0.7-0.8 micrometers
Band 7 0.8-1.1 micrometers
which correspond to the colors blue-green, green-yellow, red, and
near-infrared, respectively. The instantaneous field of view (INFOV) or
resolution of the MSS is 79 meters x 56 meters. The data is collected in
each of the four bands in west to east strips called scan lines in a Landsat
scene. Each scan line is divided into picture elements (pixels) that are 56
meters in length. Hereafter, when referring to the location of individual
pixels, a scan line will be called a line, and the elements in a scan line
will be called samples. Therefore, each Landsat scene is comprised of over
7,500,000 pixels (3440 scan lines x 3300 PIXELS )
( LANDSAT SCENE SCAN LINE)
B-l
-------
Each pixel has an area of 1.1 acres (79 meters x 56 meters); and there are 4
spectral reflectance values for each pixel.
Each land cover type (soil, vegetation, water, etc.) has a distinct
spectral reflectance pattern called its spectral signature. Using the
spectral signatures for each land cover, one can assign or "classify" each
pixel to a land cover type from the pixel's spectral reflectance. For more
information on Landsat, consult "Landsat Data Users Handbook," U.S.
Geological Survey, 1200 South Eads Street, Arlington, VA or "Manual of
Remote Sensing" by the American Society of Phtogrammetry, 105 N. Virginia
Avenue, Falls Church, VA.
B-2
-------
I
I
I
OBJECTIVES
The objectives of this project were to:
o Produce a Level I land cover classification from Landsat data of
the Chesapeake Bay watershed.
o Within the agricultural land cover, determine tillage practices.
o Tabulate the land cover statistics by river subbasin.
GENERAL PROCEDURES
The land cover analysis was performed on the Eastern Regional Remote
Sensing Application Center's (ERRSAC) Hewlett-Packard 3000 computer. The
land cover dataset was developed using the Interactive Digital Image
Manipulation System (IDIMS) and Geographic Entry System (GES) software
packages.
An IDIMS file containing a whole Landsat scene was created for each of
the twelve scenes covering the Chesapeake Bay drainage basin. For the eight
Landsat scenes with dates prior to 1979, the known Landsat distortions were
removed using IDIMS automated preprocessing programs (keskewing and rotating
to true north). The three 1979 scenes were ordered in a geometrically
correct format.
Each Landsat scene was divided into subscenes with similar physiographic
and land cover characteristics. Land cover signatures were developed for
each subscene from subsampled areas by an unsupervised classification
program. The signatures from the unsupervised classification were used to
produce a land cover map of the subscene. Figure Bl shows the subscene
location and Table Bl the names of the images and Land Cover Spectral
Signatures statfile for each of the Landsat scenes.
B-3
-------
High altitude color infrared, high altitude black and white, and
orthophotquads of ground truth sites were used to verify the land cover
dataset. Thematic line printer maps (line printer maps with grouped
spectral classes) were produced for each of the ground truth sites, and the
verification work was done at the Northern Virginia Planning District
Commission (NVPDC) office." If the land cover maps were not"adequate, a
supervised approach was used to augment the land cover signatures.
The digitized subbasin boundaries were registered and superimposed on
the final land cover map. Pixel counts for each land cover type were
tabulated by subbasin.
To establish the validity of the land cover statistics, the land cover
statistics were compared with data from: NVPDC, Maryland State Planning,
Susquehanna River Basin Commission, Soil Conservation Service Rockingham
County, and Piedmont Planning District. In addition, eight sites of
approximately four square miles each were randomly selected and their land
cover statistics compared with land cover statistics interpreted from
1:222,000 scale color infrared photography.
B-4
-------
Figure Bl
LANDSAT SUBSCENE LOCATION
CHESAPEAKE BAY
BASIN
B-5
-------
w
kJ.
M
fa
1
H
H
fa
CO
<
3
CJ
X
a
*
Q
Z
3
EH
05
0
CJ
X
J
Q
Z
<
3
EH
§
O
X
05
E-i
05
O
CJ
s
•-3
>i
fa
co
<;
3
u
05
Q
1
EH
05
O
CJ
X
05
Q
Z
«S
J
EH
05
O
CJ
1 i
i
X
05
i
' S
, 1
Q
Z
5
EH
05
O
CJ
o
ro
1
vO
M^
i — <
X
3
<
05
O
CO
s
>n
>H
fa
CO
H
fa
co
<
J
u
X
J
*
z
1
<
05
O
CO
3
•
z
0
EH
Z
g
CJ
CO
i
X X
3 ' J
i
! . i
z
z
s
a
co
H
fa
CO
<;
j
o
j
3
EH
05
O
cu
CO
<
H
3
M
3
u3
3
EH
05
O
a.
CO
2
<
i— t
J
M
3
05
3
J
J
H
3
S
1^3
>H
fa
' CO
<
3
u
05
3
EH
05
o
Cu
CO
<
M
, J
h4
M
3
05
3
EH
05
O
Oi
CO
s
<
H
iJ
J
H
2
J
J
J
a
M
2
S
1-3
M
>H
fa
CO
rt
J
O
j
kH
EH
05
O
AJ
CO
<:
H
J
J
f— I
3
J
J
EH
05
O
PJ
CO
s
<
I-H
J
J
M
3
05
J
J
i-J
I-H
3
3
t-3
>H
fa
CO
<
l-d
O
05
J
EH
05
0
a.
CO
hH
J
•4
M
2
K
J
>
EH
05
O
a,
CO
s
<
M
J
J
H
2
j |
05 . 05
: 3 J
i i
i
p-3 J
! 3 1 iJ
1 '
' . i
CH
05
o
cu
CO
M
J
J
>— <
3
a\
r~
X
r-
in
-— j
n
I
r^
r— i
a
3
05
05
H
fa
CO
<
J
CJ
J
3
•
U
05
3
ca
CO
I-H
05
05
<
J
3
o
05
3
m
CO
M
05
05
<
X
J
3
05
05
<:
X
s
1— J
>H
fa
CO
$
CJ
05
O
05
3
03
CO
1— 1
05
05
<
X
05
3
CJ)
05
3
CO
CO
M
05
05
<
J
J
05
05
H
fa
CO
rij
o
J
J
t
o
05
3
CQ
CO
M
K
05
H
fa
CO
<
I-H
u
05
J
CJ)
cs
o
ca
CO
M
05
O5
<
05
J
O
05
3
(23
CO
IH
05
O5
<
X
05 05
3 J
I
|
l _ .,
| J iJ
^ ^ i
t !
u
05
3
CD
CO
M
05
K
S
i-L,
CTi
r-
X
[^
in
-------
•
1 &
1 3
fa
rH
< M CC
ft < z co a
L O c^
" i a a a
» 3 2 a 2
n> < a X
" 2 U M 2
»j CO Q< M
i.
13
It
b
^^
18 3
in §
1 i
13 S
JMALTO
Jx
fa
CO
5
CJ
J
D
LTOONA .
i-J
o
2:
|
<
OS
'
JMALTO
>H
fa
CO
3
CJ
os
D
LTOONA .
OS
£3
Z
0
g
J
e£
JMALTO
>H
fa
CO
3
CJ
J
J
LTOONA .
^
1-3
i
|
<
OS
JMALTO
^~*
t-l
CO
5
CJ
QS
>_3
LTOONA .
QS
J
2
O
<
as os
O hJ
D i-I i
^
Cu
cn
»-j
CJ
a:
'-j
^H
ALISBUR
CO
X
1-3
SH
QS
D
CQ
CO
M
CO
X X
133 H^
I
'
as
£3
CQ
CO
M
H
s
EH
Z
o
OS
fa
QS
JMFRONI
fa
CO
! )
CJ
OS
t
<
^4
O
OS
0
OS
fa
QS
,
J
<
>H
0
QS
H
2
O
OS
fa
JMFRON]
fa
CO
! 1
CJ
^
<
^H
0
QS
B
O
QS
fa
^
.
<
^H
O
OS
E-i
Z
O
OS
fa
OS
J
JMFRON]
fa
CO
l_]
CJ
as
.
<
Jx
0
QS
2
O
OS
fa
OS
<
J
rt
>H
O
OS
E-<
§
OS
fa
j
t
os os
IO i i-4
— _ -
i
J ' iJ
3 J
<
g
OS
2
O
QS
fa
en
r-
X
in
ro
ro
1
rH
-------
H
fa
1
ff
E-i
cn
Q fa
H 2
M H
fa CJ M
H cn s
cn i <
cn CQ 2
rt| O
h4 CO
O
TABLE Bl (contd
RAW DATA
SUB-SCENE
NAME
2
2 O
0 H
M EH
£-• 3
CJ 03
a M
cn os
1 6-1
CQ cn
•— ^ |_^
co a
y a cn
< 2 N 05
< 2 CO 6-
Q Cd
i a J s
3 z a
< a x 2
o5 CJ i i i i
'/5 Oj
Q
05
4,
H
o,
O
fa
05
O
2
S
LASFY
CJ
J
O
•
x;
J
o
o
2
NORFOLK. UL
O5
pi
fa
O5
0
2
LASFY
O
•
OS
o
*
^
(_]
0
0
2
NORFOLK. UR
, ^J
j
fa
05
O
2
LASFY2
O
X
(
**s
ij
o
o
2
NORFOLK . LH J
05 X
3 J
.
1
-
3
y|
J
O
fa
05
0
2
P^
LO
X
r~
Lfl
(^
1
U1
1—1
Z3
CJ
H
05
2.
CLASFY ,
1
(_^
3
Q
2
0
2
X
CJ
H
G5
"
RICHMOND. UL
PS
D
X
O
M
05
CLASFY
•
05
*
Q
2
O
&
U
H
05
RICHMOND. UR
^
X
CJ
H
QS
5^
CLASFY
•
f^
i— 3
.
Q
2
0
X
CJ
H
K
RICHMOND. LL
OS
1 ~\
CJ
M
as
CLASFYl
•
05
•
Q
2
O
2
CJ
H
OS
!
RICHMOND. LR
05 05
D J
—
1
3 Jj
Q
2
g
*T*
CJ
M
cn
X
f-v.
in
'X
f>1
1
kO
B-8
X
CJ
2
| -i
•^
.CLASFYl
w
J
.
O
OS
CQ
ffi
g
.
LYNCHBURG. LH
Q^
O
2
( 1
s
.CLASFY j
X
05
CJJ
OS
3
CQ
X
£H
j
LYNCHBURG. RH
on
- -
5
O
05
3
CQ
X
CJ
2
r~
LTJ
X
r-
in
J
•
m
I
r~
a
1
05
2
LASFY |
CJ
X
os
w
^
0
2
O
ROANOKE.RH
X
OS
1
M
^
O
!2
<
0
CTi
p~
X
r*»
u">
OS
•3-
m
I
as
i — i
-------
I
I
1
I
I
I
I
I
I
I
I*
I
I
I
I
CRITERIA FOR SELECTING LANDSAT SCENES
The criteria for Landsat scene selection were determined by the
requirements of the Chesapeake Bay fluvial model. The land cover dataset
needed to be a USGS Level I, present use, land cover classification. Within
the agricultural class, minimum and conventional tillage as well as pasture
land covers needed to be identified. A study was undertaken to determine
the feasibility of developing two land cover datasets for the Chesapeake'Bay
Basin: a more detailed land cover dataset developed from a multitemporal
analysis for areas with an estimated river transport time to the Bay of less
than 5 days; and a less detailed dataset developed from a single date for
all other areas. The multitemporal was to be done at ERRSAC by NVPDC
personnel and the single data work by Kennedy Space Center (KSC). After
meeting with Kennedy Space Center personnel, it was mutually agreed that KSC
did not have the resources to participate in the project. Therefore, it was
planned that a 'single land cover database for the entire Chesapeake Bay
Basin would be developed at ERRSAC by NVPDC and that it would not use the
multitemporal approach. After discussions with District Conservationists
and Agriculture Extension Agents from Fauguier, Fairfax, Loudoun, and Prince
William Counties in Virginia and Agricultural Engineers from the University
of Maryland, it was decided that only Landsat scenes from 1977 to the
present would adequately reflect present agricultural practices. To
differentiate between minimum and conventional tillage practices, the
Landsat scenes' date had to be after spring planting but before the crop
cover reached 20%. To minimize the amount of undefined land cover, only
Landsat scenes with a cloud cover of 10% or less were used.
B-9
-------
Table B2 lists the Landsat scenes in possession of the Chesapeake Bay
Program. "Used in study" denotes the scenes used in the development of the
land cover database. "Quality" denotes the image quality on a scale of 1 to
8 for bands 4, 5, 6, and 7, respectively.
There are six areas—three in Virginia/ one in Pennsylvania and two in
New York—where there is not total coverage of Chesapeake Bay Basin by the
Landsat scenes. In Virginia's George Washington National Forest near Back
Creek Mountain, there is an area of approximately 250 square miles that is
not covered. This area .is over 95% forested, and it is not covered because
the Front Royal and Roanoke Landsat scenes do "riot overlap. There are two
small portions of the drainage basin- south of Norfolk, Virginia that are not
covered by the Norfolk Landsat scene. They are about 15 square miles each.
One is west of the Dismal Swamp and south of Route 58; the other is east of
Dismal Swamp and south of Great Bridge, Virginia. The missing area west of
Dismal Swamp is about 80% forest, and the area east of the Dismal Swamp is
predominately marsh. On the West Branch Susquehanna River south of
Curwensville, Pa., there is an area of approximately 300 square miles which
is 85% forested that is not covered by the Altoona Landsat scene. In New
York State there'are two areas which fall outside of the Williamsport
Landsat scene. The first is an area of approximately 150 square miles that
is west of Kenka Lake and north of Avoca; the second is 25 square miles in
area and is east of Kenka Lake and north of Weston. Both areas are
predominately agriculture. For each of the missing areas it was assumed
that the ratio of land cover types within the missing areas was the same as
that for the entire subbasin.
B-10
-------
TABLE B2
EPA/GBP AVAILABLE LANDSAT CCT'S
I
TAPE
NUMBERS
21
22
23
24
25
26
27
28
29 '
JO
31
32
33
34
35
36
37
PATH ROW
15-33
15-34
16-30
16-31
16-32
16-33
16-34
17-31
17-32
17-33
17-34
18-34
SCENE I.D.
30422-15025
30422-15031
28321-44815
28141-44935
21192-14425
28681-44715
21192-14432
28501-44855
21192-14434
21193-14482
30082-15125
28151-45545
28151-45605
21229-14505
30082-15131
21589-15062
30101-15193
DATE
5-1-79
5-1-79
5-3-77
4-15-77
4-28-78
6-8-77
4-28-78
5-21-77
4-28-78
4-29-78
5-26-78
4-16-77
4-16-77
6-4-78
5-26-78
5-30-7-9
6-14-78
CLOUD COVER
10%
0%
10%
10% '
10%
10%
10%
10%
0%
10%
0%
0%
0%
10%
10%
10%
10%
QUALITY
8888
8888
8888
8888
5888
8888
8888
8888
8888
8858
8888
8888
5888
8888
5888
8888
8588
USED
IN STUDi
*
*
*
*
*
*
*
*
*
*
*
B-ll
a
-------
LAND COVER SIGNATURE DEVELOPMENT
An unsupervised classifier was the first step in developing land cover
signatures for each subscene. Only spectral data from every other line and
sample pixel (within a subscene) were used as input to the unsupervised
classifier. The unsupervised classifier, using an iterative clustering
algorithm, grouped the spectral data into 50 unique spectral classes. The
statistics from the 50 classes (mean, variance, and covariance in the 4
spectral bands) were used as the land cover signatures and stored in a
statfile. The entire subscene (every line and sample) was then classified
by a "maximum likelihood" classifier using the signatures generated by the
unsupervised classifier.
Thematic line printer maps were output for the various ground truth
sites within each subscene. The line printer map in conjunction with the
ground truth were used to assign each of the 50 spectral classes to one of
the following land cover classes: undefined (clouds and shadows), forest,
winter agriculture, conventional tillage, minimum tillage, pasture, water,
marsh, urban, strip mine, and idle land.
If all the land cover classes were not adequately represented by the 50
spectral classes developed by the unsupervised classifier, a supervised
approach was used to develop the missing land cover signature. The
supervised signature development was done at ERRSAC and the typical
procedure was:
1. The Landsat data corresponding to the ground truth site was
displayed on a color cathode ray tube (CRT).
2. Training sites (locatable areas on a Landsat scene with distinct
spectral features) for the missing land cover classes were selected
B-12
-------
I
I
and their statistics calculated and appended to the existing
statfile.
• 3. The subscene was reclassified using the updated statfile.
4. The resulting classified subscene was examined on the CRT and
checked with ground truth.
5. If the new classification was not adequate, steps 2-4 were repeated.
When a satisfactory land cover classification was developed for all 36
• subscenes, the data was overlaid with the geographic data and tabulated.
I
I
I
I
I
I
I
I
I
f
B-13
-------
GEOGRAPHIC ENTRY OVERVIEW
In order to facilitate the registration of ground data to Landsat
imagery, the Geographic Entry System (GES) was used. GES is an interactive
system capable of entering, editing, and storing geographic information
taken directly from a map surface. GES has no analytical capabilities. Its
only purpose is to allow geographic information to be put into a data
structure that is accessible to other programs such as IDIMS.
The GES program uses three storable entities as it acquires geographical
information. They are points, lines, and polygons. Discrete points consist
of a coordinate and a point name. Such points may be used as ground control
points for use in generating registration transformations between ground
data and digital imagery. Discrete lines consist of a series of
coordinates. Discrete lines are useful for digitizing roads, pipelines,
trails, and to a limited extent, contour lines. Polygons consist of lines
that intersect at nodes called junction points. Polygons may be used to
digitize any geographic boundary.
The GES program can also construct a latitude/longitude or UTM base
grid. The grid facility is very useful for cell by cell stratification of
data.
All digitized information in GES is done with reference to a
"geoblock." A geoblock is a rectangular subsection of the earth's surface
whose edges are oriented N-S and E-W. The user defines the position and
dimensions of the geoblock, and then the GES program creates an internal
representation of the geoblock and superimposes a grid over it. The grid's
inData Stratification and the Geographic Entry System: A User's Manual,"
ESL Inc. Technical Memo. ESL-TM 991, 1978.
B-14
-------
I
I
JB individual cells are called internal units. As a result, ground resolution
varies with the geoblocks's size. When the size of the geoblock increases,
• the ground resolution, as defined by the internal units, decreases. GES
I
•
•
•
I
I
I
I
I
I
also recognizes three additional coordinate systems: latitude/longitude,
UTM, and digitizer coordinates.
The data digitized in GES is stored in six different Multiprogramming
Executive (MPE) files. They are: the geoblock directory, the overlay, the
polygon, the line, the junction point, and the discrete point files. These
six files exist for each geoblock, and the files are unique for that
geoblock. Table B3 lists the geoblock name, MPE files and the information
stored in them for the Chesapeake Bay Program.
B-15
-------
TABLE B3
GES DATA STRUCTURE
GEOBLOCK NAME: WHITE
MPE FILES
ARCPBSBB
CPCPBSBB
JPCPBSBB
LSCPBSBB
OVCPBSBB
TXCPBSBB
WHITE
INFORMATION STORED
POLYGONS
DISCRETE POINTS
JUNCTION POINTS
LINES
OVERLAY
GEOBLOCK DIRECTORY
B-16
-------
I
I
• GEOGRAPHIC DATA
All geographic data was digitized from nineteen 1:250,000 scale, USGS
• topographic maps. The following information was digitized from these base
maps: river subbasin boundaries, state boundaries, populated area
• boundaries, contiguous boundaries of the Landsat images, and the control
M points used to transform the geographic data to the Landsat images. Table
B4 lists the GES overlay number, its name, and the geographic data it
I contains. The following information on the GES overlays: the overlay
number, the overlay name, the GES class, and the description of that class,
• is contained in a subsequent section entitled "Tape Catalog."
The river subbasins were divided into two overlays. Overlay 1
I " containing fifty-six subbasins covers the basins above the fall line, and
overlay 2 containing twenty-nine subfaasins covers the basins below the fall
lines.
9 The first step in digitizing the subbasin boundaries on these two
overlays was locating the USGS gaging stations on the maps. This was done
9 using information the USGS provided on the gage's latitude and longitude
coordinates and the description of the gage's location. From this point,
the boundaries of the subbasins were hand drawn on the 1:250,000 scale
_ maps. The subbasins were then digitized as polygons which are stored in GES
• overlays 1, "SUBBASIN," and 3, "COASTAL."
fl State boundaries in the basin for Delaware, Maryland, New York,
Pennsylvania, Virginia and West Virginia were digitized from the base maps.
§| The boundaries were digitized as polygons on overlay 5, "STATE." This
information was not needed for the hydrologic tabulations, but was digitized
1
H for possible future use.
B-17
-------
Populated areas below forty degrees north latitude which have relatively
short travel times to the Bay, as highlighted on the base maps, plus
populated areas on the Harrisonburg base map were digitized. The USGS
criteria for defining populated areas on the USGS 1:250,000 scale maps is
that any area large enough on this cartographic product with detectable
housing or construction was a populated area. The purpose for digitizing
these populated areas was to obtain data on the distribution of urban,
suburban, and rural areas in the subbasins. The populated areas were
digitized as polygons on GES overlay 6, "URBAN."
Due to the fact that adjacent Landsat images overlap between 20 and 30
percent, there is a problem of double counting the areas within the
overlap. Double counting the area within the overlap can be prevented by
merging the images together or dividing the images along a common boundary
where the images overlap. Rather than merging the twelve Landsat images
together, a costly and time-consuming task which in this case required too
much data storage, it was decided that the most efficient method for
correcting the problem of double counting the area of overlap would be by
using contiguous boundaries. To do this, an accurate approximation of the
area covered by each image was needed. Based on 9 x 9 inch hard copy prints
of the images, the boundaries of each image were drawn on the base maps.
Noting the overlap between the areas covered by the images, contiguous
boundaries for the images were then drawn in the areas of overlap. Also, at
the edge of the basin, the image boundaries were extended past the basin
boundary. This insured the maximum coverage of the basin by the Landsat
images. The outer boundaries drawn on the map contained the total basin.
B-18
-------
I
f
I
I
I
I
I
I
fc
I
I
I
I
The contiguous boundaries and the outer boundaries were then digitized as
polygons on GES Overlay 2, "Images."
In order to generate a GES-to-Landsat transformation/ landmarks that
were identifiable on both the IDIMS digital display and the GES base maps
were selected as control points. The control points were marked and labeled
on the base maps and their line and sample values from the Landsat image
were noted and saved in a file. A total of 570 points, more or less
uniformly distributed throughout the basin, were chosen. The control points
on the base maps were then digitized on the point overlay, overlay 4,
"CONTROL."
B-19
-------
TABLE B4
GEOGRAPHIC ENTRY SYSTEM
OVERLAYS
OVERLAY HO. NAME DESCRIPTION
1 Subbasin Non-Tidal River Subbasins
2 Images Contiguous Boundaries of the
Landsat Images
3 Coastal Tidal River Subbasins
4 Control Control Points
5 State State Boundary Lines
6 Urban Populated Areas Highlighted
on the USGS, 1:250,000
Scale Maps
B-20
-------
I
I
I
I
I
I
I
I
I
I
I
1
REGISTRATION OF GEOGRAPHIC DATA
The general procedure for registering the geographic data to the Landsat
scenes was as follows:
1. The Landsat scenes were displayed on a CRT and significant landmarks
(ground control points) were located. The line and sample coordinates of
the landmarks along with their identifying labels were stored in a
mensuration file (a unique file for each Landsat scene). The location of
the landmarks along with their identifying labels were marked on the
1:250/000 scale USGS topographic maps.
2. The ground control points were digitized off the 1:250,000 scale
maps using GES software. The identifying labels and their coordinates (in
internal units) were saved in overlay 4.
3. Using IDIMS software, a third order polynomial equation was
developed (a unique equation for each Landsat scene) which converts GES
internal unit coordinates to IDMS line and sample coordinates. The
residuals from the transformation equations were all within ±3 pixels.
4. Using the transformation equation, the geographic data was
transferred into an IDIMS image format such that they had the same number of
lines and samples as their corresponding Landsat scenes. For each Landsat
scene, an image was created containing the Landsat scene boundaries (overlay
2). For Landsat scenes where the data was digitized, an image was created
containing the urban boundaries (overlay 6).
[NOTE: The Landsat data was not transformed to a ground coordinate system
(Latitude/Longitude, Universal Transverse Mercator, State Plane) because of
increased computer time and storage limitations. The GES files will be
B-21
-------
available for future users to do the transformations. See TECHNICAL
MEMORANDUM "A Procedure for merging Landsat data with other geographic data
using the Geographic Entry System (G.E.S.) and the Interactive Digital Image
Manipulation Systems (IDIMS)" by Wayne A. Hallada, Computer Science
Corporation, April 22, 1981)].
B-22
-------
I
I
l'
I
I
I
I
I
I
TABULATION OF LANDCOVER DATA
The landcover data was tabulated by river subbasin for each of the
thirty-six subscenes. The general procedure for tabulating the land cover
data was:
I
1. The classified subscene was overlayed by the image containing the
• , Landsat scene boundaries. Only the portion of the subscene inside
the Landsat scene boundary was used in the tabulation.
| 2. The classified subscene was then overlayed by the image containing
urban boundaries. The subscene areas inside the urban boundaries
• were changed to either high, medium, or low density urban depending
• on their spectral characteristics.
3. The subscene was then overlayed by the image containing the river
subbasin boundaries and the land cover types tabulated by subbasin.
B-23
-------
LANDSAT ACCURACY ANALYSIS
The verification of the Landsat land cover data set followed two
different procedures. The first procedure was to randomly select areas
within the Chesapeake Bay drainage basin and compare the Landsat land cover
statistics with statistics from photointerpreting color infrared
photography. The second was to compare land cover statistics from Landsat
with land cover statistics from various agencies in the Chesapeake Bay basin.
The general procedure for comparing the Landsat land cover statistics
with the photointerpreted land cover statistics was:
1. The classified Landsat image was displayed on a CRT. An IDIMS
function was run which randomly selects blocks whose area was
selected by the operator from within the Landsat scene. The land
cover statistics of the block were tabulated.
2. The positions of the blocks were located on the color infrared
photography, a mylar overlay was placed on the photography and land
cover types drawn and labeled. The blocks were labeled for
geographic features near their location.
3. The mylar was then transferred to a digital planimeter and the area
of each land cover calculated. These were compared with the values
tabulated from the classified Landsat data.
The only requirement on the location of the randomly selected areas was
that it could be identified on the color infrared photographs (which was the
best ground truth available) on loan from Maryland State Planning. There
was at least one randomly selected site in each of the Chesapeake Bay basin
physiographic provinces. The Appalachian Plateau, Appalachian Ridge and
B-24
-------
I
I
• Valley, and the Blue Ridge each have one site. The Piedmont has two sites,
and the Coastal Plain has three randomly selected areas. The randomly
• selected areas fell across three different Landsat scenes (Front Royal,
Washington, Salisbury). Each of the sites was 50 x 50 pixels. For the
g Front Royal and Washington scenes, that corresponds to 4.3 square miles.
_ For the Salisbury scene the area is 3.1 square miles.
• Because the randomly selected areas were located on three different
• Landsat scenes, on different dates, and represent all 5 geographic
provinces, it was decided that, when aggregated, their statistics would
• adequately represent the accuracy of the land cover over the entire basin.
There was some evidence to suggest that the randomly selected site near
I Loch Raven Reservoir should not be included in the accuracy analysis. The
site, 10 miles north of the Baltimore beltway, was in a state of transition
between agriculture land use and residential land use. There were
•indications on the aerial photography that many of the agricultural fields
had been left idle. In one field a road network was in place for a
fl subdivision development. If the fields had been left idle, their spectral
characteristics would not be representative of the land cover type and
H therefore should not be included in the analysis. The photointerpreted land
covers for Loch Raven Reservoir in winter agriculture and pasture were 2.2%
9 and 41.44%, respectively. There was no other site that had such large
M discrepancies between these two land cover types. There were no indications
that the classification errors were the result of physiographic influences.
PI Indeed, another site, Liberty Reservoir which is only 15 miles from Loch
Raven Reservoir, has a land cover classification that correlated well with
m
t>q
fei the photointerpreted land cover.
B-25
-------
Table B5 contains the Landsat land cover accuracy results. Lines A-H
are the 8 randomly selected sites and their tabulation of land cover types.
Line I is the sum of all 8 sites and Line J is the sum of 7 sites with line
E (Loch Raven Reservoir) excluded. Line K lists the percent error for the
lines I and J. Columns 1-5, 9, and 10 are the land cover types tabulated
from the randomly selected sites. The value in column 1 (FOREST) includes
Idle land. Columns 6-8 are different aggregations of the agriculture land
covers. From inspection of line K, it can be seen that only columns 2, 5,
and 9 (winter agriculture, pasture, and urban) are significantly effected by
the deletion of Loch Raven Reservoir from the accuracy analysis. The change
in percent error in winter agriculture from 50.3% to 3.43% and in pasture
from -11.45% to +15.91% are reasonable. After visual inspection of the
classified data and comparison of the classified data to other land cover
data sets, it was determined that the percent error tabulated without Loch
Raven Reservoir was more representative. The percent error in the urban
classification is probably not representative of the error in urban
classification basinwide. Most of the error in the urban classification is
caused by Catoctin Mountain and Cumberland sites. In both cases, the urban
areas were located in the shadows of mountains, which is not representative
of urban areas basinwide. Visual inspection of the classified data and
comparison of the classified data with other land use data sets also
indicates that the urban area classification errors are less than the
accuracy analysis indicates.
In addition to the randomly selected sites, the Landsat land cover data
set was compared with land use data sets from various state and federal
agencies. The following river basins were used in this comparison:
B-26
-------
I
I
• Potomac, James, York, Rappahannock, Patuxent, Gunpowder, Occoquan,
Appomattox, and Susquehanna. The Landsat land cover data set was compared
• with the Environmental Protection Agency (EPA) Chesapeake Bay Program
Historic Land Use data base. The EPA data base was compiled from
• Agricultural Census data (1978) and Timber Survey data (1975). Tables B6 -
• B9 show the comparisons for the Potomac, James, York, and Rappahannock river
basins. For the Patuxent and Gunpowder river basins, the Landsat data set
• was compared with data from the Maryland Department of State Planning's
Maryland Automated Geographic Information (MAGI) System. The MAGI's 1978
B land use data set was developed from high-altitude color infrared
photography with a minimum mapping area of 10 acres. Table BIO shows the
• comparison of the Landsat land cover data with MAGI land use for the
g Patuxent and Gunpowder river basins. For the Occoquan River Basin, the
^^r Landsat land cover data was compared with the land cover statistics
• planimetered from a 1979 1:48,000 scale land use map developed by the
Northern Virginia Planning District Commission. Table Bll shows the
• comparison for the Occoquan River Basin. In the Appomattox River Basin, the
Landsat land cover statistics were compared with land cover statistics from
B the Pidemont Planning District Commission and Forest Statistics for the
« South Piedmont of Virginia. This comparison is found in Table B12. Table
™ B13 shows a comparison of the percent forest developed by the Landsat data
Bbase with percent forest figures developed by the Susquehanna River Basin
Commission.
B-27
-------
0
0 g
rH «Q
3
z
<
a.
cn a
~
rj in
< •*•
*•/
CO J +
< cn
^ ^
gd
j
J
M
f« -.
T
p- j j.
< cn
E* +
p fN
B
2 en
* 5
W J ~
5 VO gd
> O M
j SH
2
<
U
LD XI 2
03 0 p
3 Ul P
« 1 s
i-i o 2
03 y
,-r* ^
<; **•
EI r-,
"* > -J
co ^ £ 3
a o M
!
s.i
fl Kt k-
S H
«
£o
CN 2 <
M
2t
CH
W
- §
o
u.
i ^* CN
1 CN .H
: ° d
' Sg
J d
*-< T
VO •"*
r- r-
rH 10
*r rn
TT ry
-H ^i
vo o
T ^"
O CO
IN IN
CO CD
CN CN
O CN
CO -<
•H o
in T
CJ» O
O rH
en en
10 O
vO O
m in
»H rH
en 10
cn en
en in
CN CN
Liberty % Photo
Reservoir % LANDSAT
Area=2782 Acres
i i
i i
i i
en vo
rH rH
r* o
IO Tf
VO IN
rH O
rH CN
O IN
vo in
^ p*
O CO
vo •«•
^T p*
ID CN
O P-
r- IN
rH
1 T
1 O
d
T ID
en in
•H en
10 CN
IN cn
en en
cn o
rH VO
rt cn
co p>
Catoctin % Photo
Mountain % LANDSAT
Area=2782 Acres
*T
1 O
1 <
1 O
O CN
en en
•H
CO CN
10 M
in en
•H CN
p* ^"
•H 10
^ •*?
CN
i in
i
i ^
.H CO
in T
^ CO
•H >H
p- IN
*H *H
-4 d
VO
1 p-
1
1 CN
IO
1 P-
1 •
1 »H
IN CN
O in
d in
p* P*
Cumberland % Photo
\ LANDSAT
Area=2782 Acres
CD O
CN 0
-H d
*H -H
P- T
•H CN
tH iH
CN TT
in 10
en vo
P- O
in o
O T
0
1 0
i .
1 en
in ^ .
a\ 10
IN IN
r- o
in IN
d d
0
1 CO
1 •
1 CN
0
1 O
1
1 •-*
en IN
O r*
^r IN
CO CO
Deep Creek % Photo
Lake % lANDSAT
Area=2782 Acres
i i
i i
i i
in *»
en co
cn o
t-i
CN *T
in Ti-
en d
in in
O IN
o in
CN c-H
FH m
en o
•
in TT
Swan Creek * Photo
% LANDSAT
Area=2782 Acres
cn CN
in rH
en en
^ *r
in vo
cn n
d d
Tf VO
*-» iH
in cn
m IN
VO VO
*H p»
cn vd
CN IN
CO O
.H O
VO TT
CN CN
CO O
cn TT
in IN
co vo
cn p*
IN CN
"4 IN
r-1 iH
ui en
CN IN
r- co
O CO
-i d
CO VO
en en
d >
IN CN
Tolcljester % Photo
Beech % LANDSAT
Area=2007 Acres
i
rH
CN 1
. 1
0 1
CO
1 0
1 t
r o
a* IN
P» e>
•^ rH
en cn
.H CN
p- CO
co eo
0 O
CO CN
-H IN
^ T
cn IN
vo in
p» en
o o
o o
cn vo
en ^
IN T
IN O
cn in
IN CN
CO vO
CO -H
CO P^
«"H rH
O CN
in eo
Lang ford % Photo
% LAUDSAT
Area=2007 Acres
i
0 T
-J CO
vo m
p* in
x cn
^r ^
in co
cn co
cn .4
en T
in IN
vO IN
in cn
IN IN
in CN
CO T
c en
CN CN
O VO
en vo
T IN
rH rH
0 0
eo co
T in
O CO
cn cn
in m
rH rH
in T
Cn T
TT p*
CO en
0 en
cn o
TT in
TOTAL % Photo !
% LANDSAT
Area=20, 706 Acres
m m
o r-
p- vO
0 p.
c m
T O
rH p.
cc in
r^ o
en
+ +
IN .n
cn -H
en *r
rH
"*" *
p» en
cn vo
d d
rH
•t- +
in rH
*r cn
rH '^1
rH rH
1 *-
en IN
co in
o co
IN ^H
+ +
O cn
in IN
d d
+ I
en
en *r
d en
in
•f -t-
in o
in -H
CN CN
•f -t-
TOTAL* % Error
LANDSAT % Error**
Area=20,706 Acres
1>
01
Z)
•s
3
t:
• J) CD
01 JJ
=> C1J
u
•O 2^
IP
SM
w o
Q JJ
Z O
<£ s:
3 a
n n
-a -a
t) 01
> -w
t. u
HI cj
01 a
J X
O a
0)
W Ul
a •-*
i §
Vj
u
O wl
O 0)
rH CC,
X C
— D
TJ >
HI eg
JJ U
0 J<
CD O
a-o Q
(Observed-Ex
* %ERROR= Expecte
** Tabulated without L
Q UJ l_t_ ten —• - — , ^>
B-28
-------
I
I
f
I
I
I
I
I
I
LAND COVER
PASTURE
CROP
FOREST
OTHER
TABLE B6
LAND COVER COMPARISON POTOMAC RIVER BASIN
POTOMAC ABOVE FALL-LINE POTOMAC BELOW FALL-LINE
LANDSAT
18.42%
16.35%
61.00%
4.23%
EPA/CPB* LANDSAT
18.32% 15.90%
16.53% 15.23%
57.00% 54.47%
8.15% 14.40%
EPA/CPB*
7.64%
14.59%
51.70%
26.07%
* Environmental Protection Agency Chesapeake Bay Program Historic Land Use Database
I
I
I
I
B-29
-------
LAND COVER
PASTURE
CROP
FOREST
OTHER
TABLE 37
LAND COVER COMPARISON JAMES RIVER BASIN
JAMES ABOVE FALL-LINE
LANDSAT EPA/CPB*
14.62% 14.49%
12.19% 7.16%
65.23% 74.21%
7.96% 4.14%
JAMES BELOW FALL-LINE
LANDSAT EPA/CPB*
10.81% 2.75%
14.64% 12.78%
62.68% 61.52%
11.87% 22.95%
* Environmental Protection Agency Chesapeake Bay Program Historic Land Use Database
B-30
-------
TABLE E3
LAND COVER COMPARISON YORK RIVER BASIN
YORK ABOVE FALL-LINE
LAND COVER LANDSAT EPA/CPB*
PASTURE 12.88 % 9.11 %
CROP 13.90% 11.00%
FOREST 72.80% 69.70%
OTHER 0.42% 10.19%
YORK BELOW FALL-LINE
LANDSAT EPA/CPB*
11.84% 3.25%
19.26% 12.33%
68.82% 71.59%
0.05% 12.83%
* Environmental Protection Agency Chesapeake Bay Program Historic Land Use Database
B-31
-------
TABLE B9
LAND COVER COMPARISON RAPPAHANNOCK RIVER BASIN
LAND COVER
PASTURE
CROP
FOREST
OTHER
RAPPAHANNOCK ABOVE FALL-LINE
LANDSAT
26.21%
12.08%
61.42%
0.29% •
EPA/CPB*
23.59%
15.32%
55.23%
5.86%
RAPPAHANNOCK BELOW FALL-LINE
LANDSAT
15.1%
21.64%
62.20%
1.06%
EPA/CPB"
3.22%
19.94%
65.96%
10.88%
* Environmental Protection Agency Chesapeake Bay Program Historic Land Use Database
B-32
-------
I
•
_
I
I
I
TABLE BIO
LAND COVER COMPARISON PATUXENT AND GUNPOWDER RIVER BASINS
Forest
Agriculture
Urban
PATUXENT
LANDSAT MAGI*
53%
41%
6%
48%
38%
9%
GUNPOWDER
LANDSAT
47%
44%
MAGI*
41%
48%
I
_
•
I
I
_
*From Maryland Automated Geographic Information (MAGI) , 1978 Dept. of
State Planning
B-33
-------
TABLE Bll
LAND COVER COMPARISON OCCOQUAN RIVER BASIN
LANDSAT NVPDC*
Forest 62% 59%
Tilled Agriculture 14% 11%
Pasture ' 20% 22%
Urban 4 % 8 %
*Planimetered from a 1:48,000 scale Land Use Map NVPDC 1979
B-34
-------
I
I
l'
I
I
I
I
I
I
I
I
I
I
TABLE B12
LAND COVER COMPARISON APPOMATOX RIVER BASIN, VIRGINIA
PIEDMONT PLANNING FOREST
LANDSAT* DISTRICT COMMISSION** PUBLICATION**_*
Forest 69% 65% 72%
Pasture 18% 20%
Agriculture . 12% 13%
From Subbasin in Appomatix River Basin
Average of Amelia and Prince Edward Counties
Average of Amelia and Prince Edward Counties from "Forest Statistics
for the South Piedmont of Virginia" 1976
B-35
-------
TABLE B13
LAND COVER'COMPARISON SUSQUEHANNA RIVER BASIN
ZONE
1
2
3
4
5
6
LANDSAT
% FOREST
57
53
60
77
72
44
SRBC*
% FOREST
57
54
55
80
66
40
AREA SQUARE
MILES
4771
2596
3700
6900
3404
5288
TOTAL
61
60
26600
ZONE
1
2
3
4
5
6
NAME
Susquehanna River, Upstream From Athens, Pa.
Chemung River
Susquehann River, Sayre, Pa. to Sunbury, Pa.
West Branch Susquehanna River, Source to Mouth
Juniata River
Susquehann, Sunbury, Pa. to Mouth (excluding
Juniata River)
* From "Assessment of the Water Quality of Streams in the Susquehanna River
Basin," Susquehanna River Basin Comm., January 1976
B-36
-------
I
I
l'
I
I
I
I
I
I
I
I
I
I
I
t
I
LANDSAT IMAGE INFORMATION
CHESAPEAKE BAY BASIN
The following sections contain information on the location of the
Landsat imagery, the location of the subscene on the imagery, the land cover
classification scheme, and the tape number where each Landsat image is
stored. Table B14 provides the information needed to locate and access the
classified images for each Landsat scene. It lists the Landsat image's
path/row location, the name of the file containing the classified
information by the subscene name, the location of the subscene on the images
by the starting line, starting sample, number of lines and number of
samples, the number of classes in each file, and the tape number where the
information is stored. Table B15 lists the identification numbers for he
land cover types. Tables B16 through B27, which use the land cover
identification numbers, relate the land cover types to the subscene class
numbers. These tables provide all the information needed to access the land
cover data for the Chesepeake Bay Basin.
B-37
-------
TABLE B14
LOCATION ON ORIGINAL IMAGE
PATH ROW
16-30
16-31
17-31
16-32
17-32
15-33
16-33
17-33
15-34
16-34
17-34
18-34
SUBSCENE NAME
CORTLAND . LH . CLASFY
CORTLAND . RH . CLASFY
SCRANTON.UH. CLASFY .
S CRANTON . LH . CLASFY
WILLIAMSPORT . UL .
1 CLASFY1
WILLIAMSPORT . UR.
CLASFY
WILLIAMSPORT . LL .
CLASFY1
WILLIAMSPORT . LR.
CLASFY
HARRISBURG - UL . CLASFY2
HARRISBURG . UR. CLASFY
HARRISBURG . LL . CLASFY1
HARRISBURG . LR. CLASFY
ALTOONA . UL . CLASFY
ALTOONA . UR . CLASFY
.ALTOONA . LL . CLASFY
ALTOONA . LR. CLASFY
SALISBURY . UH . CLASFY
SALISBURY . LH . CLASFY
WASH. UL. CLASFY
WASH. UR. CLASFY
WASH. LL. CLASFY
WASH. LR. CLASFY
FRONT . ROYAL . UL . CLASFY
FRONT . ROYAL . UR. CLASFY
FRONT . ROYAL . LL . CLASFYl
FRONT . ROYAL . LR . CLASFY
NO RFOLK . UL . CLAS F Y
NORFOLK. UR. CLASFY
NORFOLK . LH . CLAS FY 2
RICHMOND . UL . CLASFY
RICHMOND . UR. CLASFY
RICHMOND . LL . CLASFY
RI CHMOND . LR . CLAS F Y 1
LYNCHBURG . LH . CLASFYl
LYNCHBURG . RH . CLASFY
ROANOKE . RH . CLASFY
STARTING
LINE
1100
1100
1
1171
1
1
1171
1171
1
1
1171
1171
1
1
1171
1171
1
1492
1
1
1171
1171
1
1
1171
1171
1
1
2101
1
1
1171
1171
1
1
1
B-38
STARTING
SAMPLE
1
1501
1
1
1
1717
1
1717^'
1
1719
1
1719
1
1683
1
1683
441
441
1
1721
1
1721
1
1721
1
1721
1
1433
1
1
1723
1
1723
1
1799
1000
NUMBER
OF
LINES
1241
1241
1170
1170
1170
1170
1170
1170
1170
1170
1170
1170
1170
1170
1170
1170
1491
1492
1170
1170
1170
1170
1170
1170
1170
1170
2100
2100
883
1170
1170
1170
1170
2983
2983
1350
NUMBER
OF
SAMPLES
1500
1500
3343
3343
1716
1717
1716
1717
1718
1719
1718
1719
1682
1683
1682
1683
1936
1936
1720
1721
1720
1721
1720
1721
1720
1721
1432
1000
2432
1722
1722
1722
1722
1798
1798
2373
NUMBER
OF
CLASSES
48
50 '
50
50
49
50
50
50
51
51
52
50
50
50
50
52
55
50
50
50
46
50
50
50
51
50
50
50
51
48
45
50
46
50
50
50
TAPE
NO.
1
1
2
2
3
3
3
3
4
4
4
4
.
5
5
5
6
6
/
7
7
7
8
8
8
8
9
9
9
10
10
10
10 ,
^
11
11
12
-------
1
1
^^
f
1
1
1
1
1
1
•
NUMBER
1
2
3
4
5
6
7
8
9
10
11
I
I
I
I
I
TABLE B15
IDENTIFICATION NUMBERS FOR LAND COVER TYPES
LAND COVER TYPE
Unclassified
Forest
Winter Agriculture
Conventional Tillage
Minimum Tillage
Pasture
Water
Marsh
Urban
Strip Mine
Idle
B-39
-------
?A7H-?.CW 16-30
TABLE B16
LAND COVER TYPES 3Y SU3SCZNS CLASSES
— ^-s-.*-. r - -v
£.
3
4
5
o
™
3
?
10
— —
12
13
14
15
15
* -*
i ~
13
20
21
^ ^
23
24
25
25
27
23
29
30
31
32
33
34
J ^
35
37
33
3?
1 n
41
42
43
t-T
T C
45
LAND COVER
CORTLAND . LH . CLASFY
1
6
2
5
3
4
2
D
3
6
2
2
6
6
6
7
6
5
2
9
2
5
2
2
6
o
2
2
6
6
2
1
6
2
5
2
3
2
2
3
6
2
6
6
2
1
TYPE ^^
CORTLAND . RH . CLASFY
1
o
2
5
6
4
6
6
2
6
6
2
6
6
2
2
3
6
9
5
2
5
6 4ft
2 ^
D
6
2
2
2
2
2
7
. 6
6
2
5
2
4
2
2
o
6
2
2
6
5
4" 62
4o
49
2
.
5 j 5
Z —
52 B-40
-------
I
TABLE B17
^ND COVER TY?ES BY SUSSCZMZ CLASSES
?ATH-=.CW 16-31
,
,
2
•3
4
5
0
7
3
Ci
10
11
J_
13
14
15
15
LAND COVER
SCRANTON . UH . CLASFY
A
2
3
2
5
2
^
2
3
2
2
2
6
2
2
7
TYPE
SCRANTON . LH . CLASFY
4
2
6
2
4
2
5
2
3
2
6
2
6
2
2
2
17 63
13
19
20
21
22
23
24
25
25
27
23
29
30
"« 1
32
33
34
3:
3c
37
33
39
.in
• T
_i T
43
44
4:
45
4~
4-=
49
-'w
2
6
2
5
2
D
2
3
2
2
2
5
9
5
7
4
2
6
2
5
2
5
2
3
2
2
2
6
7
o
2
D
2
6
6
2
5
2
5
2
3
2
2
2
5
9
2
7
4
2
3
2
5
5
2
3
2
9
2
6
2
2
3
2
o
2
; _
=2 B-41
-------
TABLE B18
LAND COVER TYPES BY SU3SCENS CLASSES
CLASS NO.
1
^
3
•?
0
5
/
-*
3
~2
' Q
• 1
1 I
±2
14
15
15
— ;
15
^ ^
20
21
22
23
24
23
25
27
23
29
30
2 i
3 2
33
34
3-
•^ •-
JC
2 7
33
39
40
41
42
43
44
i £
•s —
46
47
43
i (-
-» ^
0 w
WILLIAiMSPORT.
UL.CLASFvi
4
1
2
1
3
1
2
1
5
•
2
1
5
1
2
6
1
7
1
6
1
2
1
5
1
2
1
2
1
7
1
4
1
2
1
6
1
2
1
5
1
2
1
5
1
2
1
0
1
-t
WILLIAiMSPORT.
UR.CLASFY
1
o
5
2
4
2
2
2
3
2
6
2
6
2
2
2
6
2
5
2
6
2
2
2
6
2
2
2
6
2
9
7
6
2
5
9
5 '
2
2
3
6
•?
6
2
2
6
5
5
->
6
WILLIAiMSPORT.
LL.CLASFY1
4
2
6
2
6
2
2
3
2
2
2
2
6
2
2
2
6
2
6
2
5
2
2
2
6
2
2
2
2
2
2
7
4
2
6
5
2
2
3
2
2
6
5
5
s
2
5
2
O
2
•
WILLIAiMSPORT .
LR.CLASFY
4
2
5
2
3
2
5
2
4
2
2
2
5
2
7
2
6
2
6
2
6
2
5 4|
2 ™
5
2
2
2
6
5
2
7
4
2
5
2
3
5
2
5 ,
5
9
2
6
5
5
2
2 ^
3 V
2
fl B-42
52
-------
I
16-32
TABLE B19
U\D COVER TYPES BY 3U3SCENE CLASS!
CLASS :ic.
<
^
3
4
3
C
7
3
o
10
' i
12
13
14
15
15
; "7
15
19
20
21
22
23
24
25
25
27
23
29
30
31
32
33
34
35
3 c
37
•« f^
j a
39
TU
41
42
"1 J
44
45
45
• —
4o
-» ^
50
~ •
52
HARRISBURG.
UL.CLASFY2
4
2
5
2
3
2
5
2
5
2
D
2
b
2
2
7
5
2
0
2
3
9
6
2
6
2
5
2
6
2
5
7
4
2
5
3
5
6
2
6
2
5
3
2
5
7
5
2
D
2
10
HARRISBURG .
UR . CLASFY
4
2
5
2
3
2
5
2
5
2
5
2
6
2
2
7
3
2
5
2
3
9
6
2
6
2
5
2
o
2
5
7
4
2
^
3
5
6
2
6
2
5
3
2
5
7
5
2
5
2
10
HARRISBURG.
LL.CLA3FY1
5
3
2
4
2
5
2
5
5
5
2
5
2
6
2
1
0
3
2
4
2
6
2
6
5
3
2
5
2
2
7
4
2
3
2
5
9
6
2
6
5
6
9
6
2
5
2
4
2
1
2
9
HARRISBURG.
LR. CLASFY
4
o
4
2
3
5
6
7
6
2
5
2
3
2
5
7
6
2
c;
9
3
2
2
7
6
2
5
8
6
2
2
7
4
2
4
2
3
5
5
7
6
2
5
7
6
2
5
7
5
9
53 B-43
-------
= Y---?.C-v 17-32
TABLE B20
iND COVER TYPES 3Y Su3SCENE CLASSES
--^cc -;c_
7
2
3
4
0
5
^
3
~t
10
^. _
]_2
13
i i
15
15
i /
15
19
20
21
22
23
24
25
25
27
23
29
30
2 i
32
33
34
35
35
37
33
39
40
4 1
42
43
f 1
**-*
45
-* c
T .
43
49
— ^
ALTOONA .
UL.CLASFY
3
4
4
•5
6
2
2
2
5
2
2
2
6
2
2
9
6
4
6
5
2
2
2
2
2
2
2
2
2
2
2
7
3
5
6
7
9
2
2
9
6
2
2
2
2
2
7
3
5
7
ALTOONA .
UR.CLASFY
3
4
2
5
3
6
2
9
6
6
2
2
6
6
2
4
3
4
2
5
2
5
2
7
2
5
2
2
2
2
2
7
3
5
2
2
2
6
2
4
6
6
•j
2
6
6
7
-,
5
2
ALTOONA .
LL.CLASFY
3
4
5
4
3
6
5
2
f.
2
2
2
6
2
2
2
3
6
2
5
2
2
6
2
3
2
2
2
2
2
2
7
3
6
6
5
2
5
6
2
2
2
2
2
2
3
6
2
9
2
ALTOCNA . ^^
LR.CLASFY
3
4
6
2
6
6
6
2
V
4
2
2
5
5
6
9
3
5
2
2
6
5
2
2 A
3
5
2
2
2
5
2
7
3
4
6
2
5
6
2
5
6
5
2
2
2
5
2
2
2 J|
5 ™
51 9
32
B-44
3
z 3
-------
I
TABLE B21
LAND COVER TYPES 3Y SUBSCENE CLASSES
15-3
_, . _,. ,.n
L
2
3
SALISBURY.
UH.CLASFY
4
5
3
SALISBURY.
LH.CLASFY
4
9
3
4 77
5
5
7
3
9
10
— . J.
12
13
14
15
15
17
13
•19
20
21
22
23
24
25
25
27
23
29
30
2 i
32
33
34
35
35
J /
33
39
40
4 1
42
43
TT
45
45
-,"
43
49
3
2
2
7
5
2
2
7
5
7
2
7
d
9
3
7
5
2
5
7
5
2
2
7
5
2
2
7
4
2
3
7
6
2
2
5
2
2
5
7
2
. 4
2
2
6
6
2
2
7
5
2
2
7
4
2
2
"7
4
7
3
7
4
7
5
5
2
2
7
5
7
2
4
9
3
7
5
7
2
5
2
2
4
7
2
4
2
2
7
6
7
50 2 2
— -» 2
52
5
52 2 B-45
-------
?Ar-:-?.cw 16-33
TABLE B22
LAND COVER TY?ES SY SU3SCSNE CLASSES
1
-1
3
4
o
6
1
3
n
10
• T
12
13
14
15
15
: —
i ^
19
20
21
? 7
23
24
25
25
27
28
29
30
3 i_
32
• 33
J -I
3 5
35
37
33
39
40
f T
•^ —
42
43
« 1
Tt
-9 3
45
47
4c
49
-^
WASH . UL .
CLASFY
4
b
3
2
0
o
6
o
-
q
o
9
5
9
6
7
4
2
' 3
2
6
5
3
2
5
2
3
2
6
2
6
7
4
5
1
2
6
6
9
5
9
3
2
5
7
• 6
7
5
2
3
WASH . UR .
CLASFY
4
2
S
7
3
y
2
7
6
5
5
7
6
2
2
4
2
6
7
3
7
2
5
2
5
7
6
7
2
4
2
5
7
3
9
2
5
5
3
3
2
2
4
2
6
T
7
7
5
2
WASH . LL .
. CLASFY
3
2
4
2
4
2
2
7
6
2
2
7
s
2
2
6
2
4
7
6
5
2
6
2
5
7
5
2
2
3
2
5
2
6
2
2
7
3
2
2
7
fi
3
2
5
3
•
WASH . LR .
CLASFV
4
5
5
7
3
7
2
7
3
2
6
6
7
2
6
2
5
5
7
2
6
2
2 A
2
7
2
4
5
5
3
7
2
6
2
2
3
7
2
5
2
s
1
7
6
7
S
7
7 ^1
J ^
5
= 1
s:
5: B-46
-------
I
I
l'
I
I
I
I
I
I
TABLE B23
LAND COVER TYPES BY SUBSCZN'Z
PATH-ROW 17-33
CLASSES
-LA =3 .1C
i
2
3
t
-»
5
5
"7
3
£)
10
•; i
L2
13
14
15
15
;_7
13
^_a
20
21
1 "*
4* «
23
24
25
26
27
23
29
30
2 T
32
33
34
35
35
37
33
39
i n
t w
• 1
"t _
t -)
"t —
t 1
-Ij
• t
•9-1
"! Zi
4c
T /
-»O
45
c ^
FRONT. ROYAL.
UL.CLASFY
6
2
o
2
4
2
2
2
5
2
2
2
6
2
2
2
3
2
5
2
6
2
2
2
3
2
2
2
2
9
7
3
2
2
2
5
2
2
2
6
0
6
9
1
2
2
5
2
^
6
FRONT. ROYAL.
UR . CLASFY
6
2
4
2
6
2
6
2
6
5
5
2
3
2
3
7
3
2
4
2
6
?
9
2
. 3
5
5
6
2
2
2
7
3
2
4
2
6
2
6
2
6
5
5
2
6
2
5
->
3
4
FRONT. ROYAL.
LL.CLASFY2
3
2
4
2
4
2
D
2
6
2
6
2
6
2
2
2
6
2
5
2
6
2
5
2
6
2
2
2
6
2
5
2
3
2
5
2
5
2
2
3
6
3
2
3
5
2
6
2
3
2
FRONT. ROYAL.
LR. CLASFY
3
2
4
2
4
2
6
2
^i
2
o
2
6
2
6
7
3
2
5
2
6
2
2
2
3
2
5
2
o
2
2
7
3
2
3
6
5
b
2
o
-
6
o
3
2
-i
->
3
t
3
51 7
52 B-47
-------
PATH-ROW 15-34
TABLE 324
COVER TYPES 3Y SUBSCZNZ CLASSES
CLASS ::c.
-
7
3
4
5
c
~
3
n
NORFOLK .
UL.CLASFY
4
2
3
7
6
7
2
7
5
NORFOLK.
UR.CLASFY
4
7
5
7
3
7
2
7
5
NORFOLK . ^^
LH.CLASFY2
4
2
3
7
o
7
2
7
D
10 878
1 '
1 ^
13
T t
15
.L C
• •"
13
19
20
^ j.
22
25
24
25
26
27
23
29
30
31
32
33
34
35
3c
37
33
39
40
_il
42
43
-t-5
,! ^
45
4"
43
2
7
5
7
6
7
4
2
2
7
5
7
2
5
7
2
6
7
2
7
4
8
3
6
7
2
5
7
2
5
7
2
4
2
2
6
2
5
2
7
6
"7
8
4
7
2
3
2
5
7
2
2
8
4
5
3
2
b
7
2
2
7
5
7
7
2
8
5
7
2
2
8
4
5
3
2
2
7
5
7
6
7
4
2
2
7
6
8
2 A
5 ™
8
2
7
5
7
2
4
2
2
7
6
7
2
5
2
2
7
6
2
4
2
2
6
7 ^
49 7 5 o ^
^ 0
2
7
5
51 9
52 B-48
-------
TABLE B25
P.ND COVER TYPES BY SU35CZNE CLASSES
PAO:-?.CW 16-34
CLASS :;c.
T_
2
T
!
-S
C
5
7
3
3
10
•L —
12
13
14
15
16
17
_c
1 9
20
21
22
23
24
25
26
27
23
7Q
30
3 1
32
33
34
35
36
37
33
33
40
1 1
-i *.
42
43
44
45
4c
4"
4o
49
RICHMOND.
UL.CLASFY
4
6
3
2
6
3
9
5
6
3
2
3
2
2
7
4
2
3
9
6
2
5
2
5
2
2
2
6
2
2
7
4
2
3
2
6
5
2
2
5
2
2
5
2
7
5
2
3
RICHMOND.
UR.CLASFY
4
6
<•
2
5
5
2
4
2
3
7
3
2
2
4
6
2
7
6
5
2
7
5
2
2
7
5
2
2
4
6
2
5
6
2
4
3
7
5
2
2
4
2
2
7
RICHMOND .
LL.CLASFY
1
5
3
2
6
5
2
2
4
6
6
2
5
2
2
9
1
o
2
2
6
9
2
2
4
2
2
2
6
2
2
7
1
5
3
2
5
5
2
6
6
2
2
5
2
2
2
4
5
RICHMOND.
LR.CLASFY1
4
6
2
2
3
3
2
4
2
3
7
5
2
2
4
6
2
7
6
9
2
7
5
2
2
7
5
2
2
4
6
2
5
6
2
4
3
7
5
2
2
4
2
2
7
9
5: 2
31
;- B-ay
: j
-------
TABLE B26
CL--.53 *IC .
i
^
3
-I
5
0
T
3
3
10
• i
j. ^
13
14
15
la
^ /
15
19
20
21
22
23
24
25
25
27
23
29
30
J ^
32
33
34
35
35
37
33
39
4C
41
-1 T
43
-t-T
-» ^
45
-T I
4o
-» ^
5C
^ ^
c _
^At-iL) LUViK TYPES 3Y 5UBSCZME CLASSE
LYNCHBURG.
RH CLASFY
1
2
3
5
5
5
2
5
4
2
2
2
6
2
2
5
2
3
6
6
6
2
2
4
6
2
2
6
2
6
7
2
3
5
5
5
2
2
4
2
2
2
6
2
2
2
1
2
2
2
B-50
3
LYNCHBURG. ^^
LH.v_LASFY2
3
2
2
6
4
4
5
5
9
2
2
2
2
2
2
5
2
6
2
5
6
5
6 A
2 W
2
2
2
2
2
2
2
7
-5
2
2
6
6
5
6
2
2
2
2
2
2
2
2
2 -
6 4
2 ^
-------
?ATH-=CW 13-34
TABLE B27
COVER TYPES BY SU3SCZME CLASSES
CL.
4
1.53 NO.
•
t.
3
I
T
5
c
1
3
a
uO
, ^
_2
i3
.4
.5
.5
. f
.5
.9
20
/I
Z2
23
24
25
25
27
2S
29
50
; 1
if — •
j2
)3
,4
! ^
55
7
;3
9
0
_
2
3
4
5
8
~
o
y
C
—
-
ROANOKE . RH . CLASFY
3
6
2
2
2
2
4
2
2
4
2
5
6
6
2
9
2
6
2
2
2
2
5
2
2
2
2
2
5
2
7
3
6
2
2
2
2
5
2
2
6
2
2
6
6
2
2
6
2
2
B-51
-------
TAPE CATALOG
All the Landsat land cover data and ancillary information is stored on
twenty 9-track tapes. Tapes 1-12 contain the classified data, the Landsat
edge boundary, the subbasin boundaries, and where available, the urban
boundaries. Tapes 13-18 contain the Landsat scenes preprocessed by IDIMS
software. Tapes 19 and 20 contain the mensuration files, statfiles, and the
GES files.
The classified data (Tapes 1-12) is stored by subscene without the
spectral classes being aggregated into land cover types. If the classified
data is to be overlayed with the geographic data'(river subbasins, urban
areas, Landsat edge), the following procedure should be followed:
1. Each of the classified scenes should be loaded off the tape.
2. The spectral classes should be aggregated into land cover types
using Tables B15 - B27.
3. The subscenes should be reunited or mosaiced together. They will
now overlay the geographic, pixel for pixel.
Tables B28 - B30 list the data contained on each tape. Each file on the
tape contains all the data for a subscene. Within each file are a given
number of records. Each record is one scan-line or line of data in a
subscene. Each record is a given number of bytes in length, and each byte
is one pixel. If there is an odd number of pixels in a line, an extra pixel
of intensity zero is added to make the number of pixels even.
The GES files for the Chesapeake Bay program which are stored on Tape 19
are found in Table B31. The GES files can only be used with GES and IDIMS
software.
B-52
-------
Tape 20 contains the files for mensuration and transformation. The
mensuration files for each Landsat image are listed in Table B32. (The
mensuration file for the Salisbury Image was lost because of tape errors.)
The transformation file contains the third degree polynomial transformation
equations for each Landsat scene and are prefixed by a six letter code
listed in Table B33. The six letter code is used to identify and access the
transformation file for each Landsat image. The mensuration and
transformation files can only be used with GES and IDIMS software.
Tape 20 also contains the land cover spectral signature files
(statfiles). A list of the statfiles are found'in Table B34. The format of
the statfiles is as follows:
1. There is 1 record for each spectral signature of class in the
statfile. See Table B14 for the number of classes in each statfile.
2. Each record contains 168 words, each word is 16 bits. Figure B2
represents the makeup of the record in a statfile.
The Chesapeake Bay Basin Landsat data has been transferred to the
University of Maryland Remote Sensing Systems Library (RSSL) for permanent
storage. The RSSL will make copies of the tapes available at cost to
interested parties. Contact the Department of Civil Engineering, University
of Maryland, College Park, Maryland, for information.
B-53
-------
TABLE B28
TAPE NUMBER/
DENSITY
1
(6250JBPI
2
(6250)BPI
3
(6250)BPI
4
(6250)BPI
5
(6250)BPI
6
(6250)BPI
7
(6250)BPI
SUBSCENE NAME
CORTLAND . LH . CLASFY
CORTLAND . LH . CLASFY
CORTLAND. EDGE
CORTLAND . SUBBASIN
SCRANTON . UH . CLASFY
SCRANTON . LH . CLASFY
SCRANTON. EDGE
SCRANTON . SUBBAS IN
WILLIAMSPORT . UL . CLASFY1
WILLIAMSPORT. UR. CLASFY
WILLIAMSPORT. LL . CLASFY1
WILLIAMSPORT . LR. CLASFY
WILLIAMSPORT . EDGE
WILLIAMSPORT . SUBBAS IN
HARRISBURG . UL . CLASFY2
HARRISBURG . UR. CLASFY
HARRISBURG . LL . CLASFY1
HARRISBURG . LR. CLASFY
HARRISBURG. EDGE
HARRISBURG . SUBBAS IN
HARRISBURG . URBAN
ALTOONA . UL . CLASFY
ALTOONA . UR . CLAS F Y
ALTOONA . LL . CLASFY
ALTOONA . LR . CLASFY
ALTOONA. EDGE
ALTOONA . SUBBAS IN
ALTOONA. URBAN
SALISBURY . UH . CLASFY
SALISBURY . LH . CLASFY
SALISBURY. EDGE
SALISBURY . SUBBASIN
SALISBURY . URBAN
WASH. UL. CLASFY
WASH. UR. CLASFY
WASH . LL . CLASFY
WASH. LR. CLASFY
WASH. EDGE
WASH. SUBBAS IN
WASH. URBAN
FILE
NO.
1
2
3
4
1
2
3
4
1
2
3
4
' 5
6
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
1
2
3
4
5
6
7
NO. OF
RECORDS
1241
1241
1241
1241
1170
1170
2340
2340
1170
1170
1170
1170
2340
2340
1170
1170
1170
1170
2340
2340
2340
1170
1170
1170
1170
2340
2340
2340
1491
1492
2983
2983
2983
1170
1170
1170
1170
2340
2340
2340
LENGTH OF
RECORD (BYTES)
1500
1500
3000
3000
3344
3344
3344
3344
1716
1718
1716
1718
3434
3434
1718
1720
1718
1720
3438
3438
3438 •
1682
1684
1682
1684
3366
3366
3366
1936
1936
2376
2376
2376
1720
1722
1720
1722
3442
3442
3442
BYTES PER
PIXEL
1
1
1
1
1
1
1
1 •
1.
1
1
1
1
1
1
1
1
1
1
i m
i •
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
i
B-54
-------
TABLE B29
rf
i
i
i
i
F TAPE NUMBER/
DENSITY SUBSCSNE NAME
8 FRONT . ROYAL . UL . CLASFY
(6250)BPI FRONT. ROYAL. UR. CLASFY
FRONT . ROYAL . LL . CLASFY1
FRONT . ROYAL . LR . CLASFY
FRONT . ROYAL . EDGE
FRONT . ROYAL . SUBBAS IN
FRONT . ROYAL ..URBAN
9 NORFOLK. UL. CLASFY
(6250) BPI NORFOLK. UR. CLASFY
NORFOLK . LH . CLASFY
NORFOLK. EDGE
NORFOLK. SUBBAS IN
NORFOLK. URBAN
10 RICHMOND . UL . CLASFY
(6250) BPI RI CHMOND. UR. CLASFY
RI CHMOND . LL . CLASFY
RICHMOND . LR. CLASFY2
RICHMOND. EDGE
RI CHMOND . S UBB AS IN
RI CHMOND . URBAN
| 11 LYNCHBURG. LH.CLASFY1
™ (6250) BPI LYHCHBURG . RH . CLASFY
LYNCHBURG. EDGE
LYNCHBURG . SUBBASIN
LYNCHBURG . URBAN
1 2 ROANOKE . RH . CLAS F Y
(6250) BPI ROANOKE. EDGE
ROANOKE . SUBBAS IN
ROANOKE . URBAN
13 ROANOKE (RAW DATA BAND 4)
(1600)BPI ROANOKE(RAW DATA BAND 5)
ROANOKE (RAW DATA BAND 6)
ROANOKE (RAW DATA BAND 7)
14 CORTLAND(RAW DATA BAND 4)
(1600) BPI CORTLAND(RAW DATA BAND 5)
CORTLAND(RAW DATA BAND 6)
CORTLAND(RAW DATA BAND 7)
FILE
NO.
1
2
3
4
5
6
7
1
2
3
4
5
6
1
2
3
4
5
6
7
1
2
3
4
5
1
2
3
4
1
2
3
4
1
2
3
4
NO. OF
RECORDS
1170
1170
1170
1170
2340
2340
2340
2100
2100
883
2983
2983
2983
1170
1170
1170
1170
2340
2340
2340
2983
2983
2983
2983
2983
1350
1350
1350
1350
2340
2340
2340
2340
2340
2340
2340
2340
LENGTH OF
RECORD (BYTES)
1720
1722
1720
1722
3442
3442
3442
1432
1000
2432
2432
2432
2432
1722
1722
1722
1722
3444
3444
3444
1798
1798
3596
3596
3596
2373
2373
2373
2373
3372
3372
3372
3372
3358
3358
3358
3358
BYTES PER
PIXEL
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
B-55
-------
TABLE B30
TAPE NUMBER/ FILE
DENSITY SUBSCENE NAME NO.
15
(6250)BPI
16
(625COBPI
17
(6250)BPI
18
(6250JBPI
19
(6250)
20
(6250)
FRONT. ROYAL (RAW DATA BAND 4)
FRONT. ROYAL (RAW DATA BAND 5)
FRONT. ROYAL (RAW DATA BAND 6)
FRONT. ROYAL (RAW DATA BAND 7)
HARRISBURG(RAW DATA BAND 4)
HARRISBURG(RAW DATA BAND 5)
HARRISBURG(RAW DATA BAND 6)
HARRISBURG(RAW DATA BAND 7)
CORTLAND.SB(RAW DATA BAND 4)
CORTLAND.SB(RAW DATA BAND 5)
CORTLAND.SB (RAW DATA BAND 6)
CORTLAND.SB (RAW DATA BAND 7)
RICHMOND (RAW DATA BAND 4)
RICHMOND (RAW DATA BAND 5)
RICHMOND (RAW DATA BAND 6)
RICHMOND (RAW DATA BAND 7)
NORFOLK. SB (RAW DATA BAND 4)
NORFOLK. SB (RAW DATA BAND 5)
NORFOLK. SB (RAW DATA BAND 6)
NORFOLK. SB (RAW DATA BAND 7)
ROANOKE.RH(RAW DATA BAND 4)
ROANOKE.RH(RAW DATA BAND 5)
ROANOKE.RH(RAW DATA BAND 6)
ROANOKE.RH(RAW DATA BAND 7)
SCRANTON (RAW DATA BAND 4)
SCRANTON (RAW DATA BAND 5)
SCRANTON (RAW DATA BAND 6)
SCRANTON (RAW DATA BAND 7)
WILLIAMSPORT (RAW DATA BAND4)
WILLIAMSPORT(RAW DATA BANDS)
WILLIAMSPORT (RAW DATA BAND6)
WILLIAMSPORT (RAW DATA BAND7)
WASH (FILE!) (RAW DATA BAND 4)
WASH (FILE!) (RAW DATA BAND 5)
WASH (FILED (RAW DATA BAND 6)
WASH (FILED (RAW DATA BAND 7)
ALTOONA (RAW DATA BAND 4)
ALTOONA (RAW DATA BAND 5)
ALTOONA (RAW DATA BAND 6)
ALTOONA (RAW DATA BAND 7)
GIS FILES
MENSURATION
ANO
STAT FILES
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
8
9
10
11
12
1
2
3
4
NO. OF
RECORDS
2340
2340
2340
2340
2340
2340
2340
2340
1241
1241
1241
1241
2340
234.0
2340
2340
2983
2983
2983-
2983
1350
1350
1350
1350
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
2340
LENGTH OF
RECORD (BYTES)
3442
3442
3442
3442
3438
3438
3438
'3438
3000
3000
3000
3000
3444
3444
3444
3444
2432
2432
2432
2432
2373
2373
2373
2373
3434
3434
3434
3434
3434
3434
3434
3434
3442
3442
3442
3442
3366
3366
3366
3366
BYTES PER
PIXEL
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1 ^
1 t
1 ^
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
M
1
B-56
-------
TABLE B-31
GES MPE FILES
FILE NO.
FILE NAME
DATA DESCRIPTION
1 ARCPBSBB. WHITE. CIS
4 CPCPBSBB. WHITE. CIS
6 JPCPBSBB. WHITE. CIS
9 LSCPBSBB. WHITE. CIS
11 OVCPBSBB. WHITE. CIS
13 TXCPBSBB. WHITE. CIS
14 WHITE. WHITE. CIS
Polygons
Control Points
Junction points
Line segments
Overlay
Text
Geoblock
B-57
-------
TABLE B-32
MENSURATION FILES
FILE NO.
3
4
5
6
7
8
9
10
11
13
14
FILE NAME
JMWMALTO . USERFILE . IDIMS
JMWMCORT . USERFILE . IDIMS
JMKSEROEL USERFILE . IDIMS
JMWMHARR . USERFILE . IDIMS
JMWMLYNC . USERFILE . IDIMS
JMWMNORF .USERFILE . IDIMS
JMWMRICH . USERFILE . IDIMS
JMWMROAN .USERFILE . IDIMS
JMWMSCRA . USERFILE . IDIMS
JMWMWASH . USERFILE . IDIMS
JMWMWILL. USERFILE. IDIMS
PATH/ROW
17-32
16-30
17-33
16-32
17-34
15-34
16-34
18-34
16-31
16-33
17-31
SCENE NAME
ALTOONA
CORTLAND
FRONT ROYAL
HARRISBURG
LYNCHBURG
NORFOLK
RICHMOND
ROANOKE
SCRANTON
WASHINGTON
WILLIAMSPORT
B-58
-------
TABLE B-33
TRANSFORMATION FILES
FILE NO.
FILE NAME
12
JMWMTRNF.USERFILE.IDIMS
CODE
ALTOON
NEWCOR
FRONTR
HARRIS
LYNCHB
NORFOL
RICHMO
NEWROA
SALISB
SCRANT "
WASHIN
WILLIA
_PATH-ROW
17-32
16-30
17-33
16-32
17-34
15-34
16-34
18-34
15-33
16-31
16-33
17-31
SCENE NAME
ALTOON
CORTLAND
FRONT ROYAL
HARRISBURG
LYNCHBURG
NORFOLK
RICHMOND
ROANOKE
SALISBURY
SCRANTON
WASHINGTON
WILLIAMSPORT
B-59
-------
FILE NO.
16
17
18
19
20
21
22
23
24
' 25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
TABLE B-34
STATS FILES
FILE NAME
JMALTOLL, STATS . ID IMS
JMALTOLR .STATS . ID IMS
JMALTOUL . STATS . IDIMS
JMALTOUR. STATS . IDIMS
JMCORTLH .STATS . IDIMS
JMCORTRH. STATS . IDIMS
JMFRONLL .STATS . IDIMS
JMFRONLR . STATS . IDIMS
JMFRONUL . STATS . IDIMS
JMFRONUR . STATS . IDIMS
JMHARRLL . STATS - IDIMS
JMHARRLR. STATS . IDIMS
JMHARRUL . STATS . IDIMS
JMHARRUR . STATS . IDIMS
JMLYNCLH . STATS . IDIMS
JMLYNCRH . S TATS . ID IMS
JMNORFLH . STATS . IDIMS
JMNORFUL. STATS. ID IMS
JMNORFUR. STATS . IDIMS
JMRICHLL. STATS . IDIMS
JMRICHLR . STATS . IDIMS
JMRICHUL. STATS . IDIMS
JMRICHUR .STATS . IDIMS
JMROANRH .STATS . IDIMS
JMSALILH . STATS . IDIMS
JMSALIUH . STATS . IDIMS
JMSCRAUH . STATS . IDIMS
JMSCRAUH . STATS . IDIMS
JMWASHLL. STATS . IDIMS
JMWASHLR. STATS . IDIMS
JMWASHUL . STATS . IDIMS
JM WAS HUR. STATS . IDIMS
JMWILLLL. STATS . IDIMS
JMWILLLR .STATS . IDIMS
JMWILLUL .STATS . IDIMS
JMWILLUR . STATS . IDIMS
PATH/ROW
17-32
17-32
17-32
17-32
16-30
16-30
17-33
17-33
17-33
17-33
16-32
16-32
16-32
16-32
17-34
17-34
15-34
15-34
15-34
16-34
16-34
16-34
16-34
18-34
15-33
15-33
16-31
16-31
16-33
16-33
16-33
16-33
17-31
17-31
17-31
17-31
SCENE NAME
ALTOONA. LL
ALTCONA.LR
ALTOONA. UL
ALTOONA . UR
CORTLAND . LH
CORTLAND.RH
FRONT . ROYAL . LL
FRONT. ROYAL. LR
FRONT. ROYAL. UL
FRONT.ROYAL.UR
HARRIS BURG. LL
HARRIS BURG. LR
HARRISBURG.UL
HARRISBURG.UR
LYNCHBURG . LH
LYNCHBURG . RH
NORFOLK. LH
NORFOLK. UL
NORFOLK. UR
RICHMOND . LL
RICHMOND. LR
RICHMOND. UL
RICHMOND. UR
ROANOKE.RH
SALISBURY. LH |
SALISBURY. UH
SCRANTON.LH
SCRANTON.UH
WASHINGTON.LL
WASHINGTON . LR
WASHING TON. UL
WASHINGTON. UR
WILLIAMSPORT . LL
WILLIAMSPORT . LR
WILLIAMSPORT . UL
WILLIAMSPORT .UR
B-60
-------
i
i
i
i
o
o
to
uo
IT)
LO
o
H
(XL
O
o
o
LU
CJ
i
i
i
t
co
H
CQ
CN
II
O
cr>
co
co
o
2:
cu
cu a;
O Q-r
<*- i— _a
3
CO
4-> T3
O O)
6uLuie.li #
•AOO'^C
spueg
d OJ M
-a c
to JZ (O
rO ra
10
CU
U
a
to
(tJ
>
o
(/I
It}
01
•o
s_
o
CO
in + +
-o
S- VO CM
O •—»
S "S -a
c o c
f- 3 10
o
o
a>
C
cu
+
o
•*->
C
O
B-61
-------
GEOGRAPHIC ENTRY SYSTEM
This section contains a list of the base maps and the GES overlay
information needed to access the digitized geographical information. Table
B35 lists the names of the 1:250,000 scale, USGS maps used for the GES base
maps. Tables B36 through B41 list the GES classes and the geographic
information contained in the classes for each overlay.
B-62
-------
TABLE B35
GEOGRAPHIC ENTRY SYSTEM BASE MAPS
Scale: 1:250,000
MAP NAME
Wilmington
Washington
Baltimore
Cumberland
Salisbury
Newark
Binghamton
Elmira
Harrisburg
Pittsburgh
Scranton
Warren
Williamsport
Charlottesville
Eastville
Norfolk
. Richmond
Roanoke
Bluefield
Clarkesburg
STATE
Del.
D. C.
Md.
Md.
Md.
N. J.
N. Y.
N. Y.
Pa.
Pa.
Pa.
Pa.
Pa.
Va.
Va.
Va.
Va.
Va.
W. Va.
W. Va.
B-63
-------
TABLE B36
GES OVERLAY NO. 1 -SUBBASIN
CLASS
88
89
87
81,
74,
76,
79,
60,
55
72,
61,
66,
32,
30
54,
13,
33,
58,
38,
35,
23
12,
1, 2
7, 8
11
10
83
75
85
80
78, 84
73, 77
82
67
51, 52, 53
56, 57
29, 31, 34
39
59
40
36, 37
22
, 3, 5, 6
, 9
SUBBASIN 10
20
30
40
50
60
70
80
'90
100
110
120, 130, 140, 150
160
170
180
190
200
210
220
230
240
250, 260
270
280
290
300, 310, 320
B-64
-------
TABLE B37
GES OVERLAY NO. 2 IMAGE
CLASS DESCRIPTION OF CLASS
1 ROANOKE LANDSAT IMAGE
2 LYNCHBURG LANDSAT IMAGE
3 RICHMOND LANDSAT IMAGE
4 NORFOLK LANDSAT IMAGE
5 FRONT ROYAL LANDSAT IMAGE
6 ^ WASHINGTON LANDSAT IMAGE
7 SALISBURY LANDSAT IMAGE
8 ALTOONA LANDSAT IMAGE
9 . HARRISBURG LANDSAT IMAGE
10 WILLIAMSPORT LANDSAT IMAGE
11 SCRANTON LANDSAT IMAGE
12 CORTLAND LANDSAT IMAGE
B-65
-------
TABLE B38
GES OVERLAY NO. 3 COASTAL
CLASS DESCRIPTION OF CLASS
69 LAND SEGMENT CLASS 1
42 LAND SEGMENT CLASS 5
65 LAND SEGMENT CLASS 6
19 LAND SEGMENT CLASS 8
17 LAND SEGMENT CLASS 9
68 LAND SEGMENT CLASS 11
47 ANACOSTIA
10 APPOMATTOX
64 BALTIMORE HARBOR
71 BOHEMIA
48 ' " CHESTER
14 CHICKAHOMINY
44 CHOPTANK
25 ELIZABETH
20 GREAT WICOMICO
62 GUNPOWDER
15 JAMES
24 NANSEMOND
46 NANTICOKE
91 OCCOQUAN
63 PATAPSCO
41 PATUXENT
28 POCOMOKE
21 POTOMAC
18 RAPPAHANNOCK
90 SEVERN
49 WICOMICO
43 WYE
16 YORK
B-66
-------
I
TABLE B39
GES OVERLAY NO. 4 CONTROL
CLASS
DESCRIPTION OF CLASS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
BLUEFIELD MAP: PT 7, 8, 144, 145
ROANOKE MAP: PT 1-6, 10-49
RICHMOND MAP: PT 50-119, 131, 132
NORFOLK MAP: PT 120-130
EASTVILLE MAP: PT 133-143
CHARLOTTESVILLE MAP: PT150,_151, 326, 330, 332-351
WASHINGTON MAP: PT 152-169, 180-241, 250, 251
SALISBURY MAP: PT 278-281, 290-297
WILMINGTON MAP: PT 276, 277, 288, 289, 431, 440
BALTIMORE MAP: PT 242-249, 252-275, 283-287, .
299-303, 398, 400, 402, 405-408, 433-439
CUMBERLAND MAP: PT 304-325, 327, 329, 331, 395, 404
NEWARK MAP: PT 419-421, 430
HARRISBURG MAP: PT 358-368, 378-388, 396, 397, 399,
401, 413-415, 417, 418, 422-429
PITTSBURGH MAP: PT 354, 357, 370-374, 377, 389-394, 403
SCRANTON MAP: PT 491-496, 507-514, 528-531
WILLIAMSPORT MAP: PT 410-412, 416, 482, 484, 489,
497-506, 515-527, 532, 53.3, 548, 549,
551, 555-561, 565, 566, 568-570
WARREN MAP: PT 353, 355, 356, 552-554, 562-564, 567
BINGHAMTON MAP: PT 445-459, 465-473, 477, 478, 490
ELMIRA MAP: PT 441-444, 460-464, 474, 476, 479-481,
483, 485-488, 540-547, 550
B-67
-------
TABLE B40
GES OVERLAY NO. 5 STATE
CLASS DESCRIPTION OF CLASS
1 VIRGINIA
2 MARYLAND
3 DELAWARE
4 WEST VIRGINIA
5 ARLINGTON COUNTY, VIRGINIA
6 ALEXANDRIA CITY, VIRGINIA
7 FAIRFAX COUNTY, VIRGINIA
8 ' PRINCE WILLIAM COUNTY, VIRGINIA
9 LOUDOUN COUNTY, VIRGINIA
10 PENNSYLVANIA
11 NEW YORK
B-68
-------
TABLE B41
GES OVERLAY NO. 6 URBAN
CLASS DESCRIPTION OF CLASS
1 SALISBURY MAP
EASTVILLE MAP
NORFOLK MAP
2 RICHMOND MAP
3 ROANOKE MAP
5 BLUEFIELD MAP
6 " WASHINGTON MAP
8 CHARLOTTESVILLE MAP
9 WILMINGTON MAP
BALTIMORE MAP
10 CUMBERLAND MAP
11 HARRISBURG MAP
i
I
I
I
B-69
-------
GROUND TRUTH
This section contains lists of aerial photography (Table B42) and USGS
orthophoto quads (Table B43) used as ground truth for the Landsat study of
the Chesapeake Bay basin.
I
I
I
B-70
-------
H
it
i
t
t
I
I
I
I
I
CM
CQ
id
CO
S '
Z
\ — 1
CO
*3*
CQ
>•
CQ
g
U Cu
/3
Q3
~ 8
H ^
•^ o
23 ~
=H a,
Cl. ^
2 ' <
Z) M
o cs
Ct3 Cd
O <
Cd
^
<£
O*
&J
J
O
03
0
o 6
X 2
ft,
< Q
CO •
< H
Z
^
CJ
Cd
O
03 Cd
a, Q
O
* CJ
CO
CO
Cd
04
£4
EH
V
§
O
t4
J
f^
0
CO
w
£-<
m r^ [^» r^ fr"] fT1 fr1 O O O O O
ininini£>voinu>>>>>>>
03Q3030303(Xa:Q3
x.x.x,\x.x,x.\
rj^n^iHi-irin
oooooooo
"o O OO 0 O O OX,x.x,x.x.x,x.\
CJCJCJCJOaCJCJCQCacQCQCQCQcaCQ
U U4. r)
(l)CUQJCl)(l)(DCi^Cl](l)ClJCl)OJCl)DrDCU
n, -lJ-l 3 S-lrH^tt)
0)0 OnODiJfdrH
•-HJ-I SiJDUXiCDiJrH
—1 CJ cnwOJfJCW'OXl -H
H i-i SE^ eUSO£tOt/)>
>fT3?fl) ^^tSUlS-i^rflO^^^iC
(A O 0 ~> -H -H -H O -r-l -H 5 jH C UH
£ _C3 (0 C tl rH rH i-J f^ 4_) rH r*H ^4 ^ Vl H™*
(^ -i-l OJfi30HH[T3(L)!l2 'H O O 0 fO 'H
^rHZI — ^SSCJSCdSSSEHES
in
CM
, — 1
1>
•g
"3
Q
^
X.
CQ
^
n.
frj
a,
cn
X
o
o
o
CO
p.*
rH
CM
r-l
o
r-i
vjQ
r"
CO
o
CO
3
it!
cu
S1
0
^Q
CO
•r-4
Jj
^J
f^J
32
xf
0*i
(•**_
in
xT
01
f-%.
LO
H
p^
LO
rH
(*!-)
f-x.
LO
a:
o
0
o
I]
M
Q)
p
fG
O4
0^
X
0
o
0
CO
CM
rH
rH
o
rH
ro
""-
5!
(fj*
2
t/}
,
ch
4J
Jj
Q
Q
CO
CN
CNJ
i—4
^r
CN
CN
r-l
xT
rH
(T>
o
CN
LO
r^
LO
03
X
M
0
0
o
Oj
d*
(T3
CM
cn
X
o
o
o
cn
r— 1
rH
o
o
p-
s
^
CO
^
c
•H
"c
(Tf
r— 1
i J
r^
3
Q
s
in
OO
rH
T
m
ro
rH
"^
rH
en
o
CM
m
[^
in
03
X.
o
Q
o"
LJ
M
OJ
p
(^
CM
cn
X
0
o
o
cn
•H
rH
o
in
o
^
y
^
<
f-C>
z
(0
*^»
n
rH
.2
"-W
J^
Q)
jj
cn
c^
r^
o
!_T) ^J*
CO Lf
1— 1
^r
LO
CO
r-l
<^p
rH
a^
o
CN
O^
*
m cc
r^ [jj
in |>
03
X.
^
0
0 X.
U CQ
t, (.
M >H
03 Q)
d, O4
crj fO
&4 CU
cn cn
X X
O 0
o o
0 0
r^ 0
rH
rH rH
com
0 rH
0 0
r- r-
s
<: co
CO O
*£ CO
Z D
n3 ro
?* w>
o
O i-l
d) CQ
J^
CU Cfl
M £
<^s o
3 E-1
.
co *vj* co CN O ^
i_n i — i cr* 0*1 LO co
rH
rH
O a. ffl w 03 03
ffi rJ 2J 2 2H 2
H O W W tJ W
^> ^ [> I> i> £>
\ \ \ X. X. \.
CD CQ CQ CQ CQ CQ
(, tl It t, (j (,
Pi Q_i Clj PH ^i PH
T3 (T3 fd ^> ,> £> ^> ^>
OJ
cn
rH
•H JH
> 3
_Q C
o; wo
rH ^3 E J-*
ft3 *H fH C7^
p^-l CO O "H i
GJ QJ i-4 C rH -rH
C JyJ Q eg --H .>
•H n3 > o 'r~' ^
^ J M E-t 2 M
r-
i—
a.
J
Q
^
\
CQ
(,
Q-i
ft3
a.
cn
X
o
o
o
o
CO
r-H
CM
CO
0
r-
CO
O
CO
D
ftj
£>
CU
c
Q
4J
C/}
^
nj
i-H
U
0"^
CN
rH
Oi
rJ
Q
£>
-x^
CQ
I.
Qw
tO
&4
CTl
X
cn
o
o
o
0
rH
CN
PO
O
r-
co
o
CO
D
n3
[>
rH
rH
n3
U
QJ
j^
Q
4-J
CO
CO
< — 1
•^fi
^
o
£>
x^
CQ
M
d)
CU
CO
rH
CO
o
o
o
CO
[-^
rH
CN
ro
O
^
CO
CO
D
03
^>
C
n3
C
(^
^c
(J
-J
CQ
cc
CN
•K
O
Q
^
X.
CQ
!H
0)
Qj
CO
i- H
X
00
o
o
o
CO
rH
CM
C*")
O
^
CO
CO
ID
T3
£>
>i
rH
rH
>
rrj
rH
Q
C
M
5
cc
O*1
J.
-•j
Q
^
\
CQ
^
0)
n
WH
n3
p_t
00
rH
CO
o
o
o
CO
ps.
rH
CN
OO
O
r~
CO
CU
CO
o
n3
£>
0)
rH
( — j
•H
^
UJ
,_^
,-rj
Q
Q
O
rH
X
^
Q
^
x^
CQ
!H
CU
Q,
CO
rH
CO
o
o
o
co
rH
CN
on
O
^
CO
CO
.r>
03
^>
cu
•y*
0
c
fQ
Q
cc
CO
CN
f — 1
•ic
itD
Q
^
\_
CQ
rH
a
n
*-!-(
Qj
CO
r- H
CO
o
o
o
CO
rH
CM
OO
0
"^
CO
co
o
tQ
[>
(T.
0
CTi
£
>p^
5^
ty
t-i
1 1
1 1
1 |
1 1
1 1
I |
! |
| |
1 1
O3 ££
X. X,
0 0
rH i— I
O 0
0 CJ
•
CO CO
z z
03 Q3
H EH
en cn
X X
cn cn
o o
o o
0 0
CM CN
CM CM
rH r— |
rH rH
CO CN
CM rH
o o
r-- r-
2 ^
<1 f£
S! -<3
CO CO
^ ^
2 Z
'O 13
2 ^*
1 (0
W -U
2 EH
B-71
-------
TABLE B43
GROUND TRUTH: CHESAPEAKE BAY BASIN
ORTHOPHOTO QUADS
SCALE = 1:24,000
MAP NAME
STATE
MAP NAME
STATE
Barton
Binghamton West
Cortland
Endicott
Guilford
Holmesville
New Berlin North
New Berlin South
Norwich
Oxford
Pitcher
• Seeley Creek
Sherburne
Spafford
Waver ly
Altoona
Beech Creek
Berwick
Black Moshannon
Crooked Creek
Harveys Lake
Howard
Julian
Kingston
Knoxville
Lopez
Mifflinetown
Millersburg
Phillipsburg
Ransom
Schellsburg
State College
Tamaqua
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
NY.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
• Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Pa.
Beach
Bentonville
Bon Air
Bowers Hill
Charlottesville, East
Charlottesville, West
Harrisonburg
Hopewell
Lake Drumond -
Luray
McKenney
Midlothian
New Kent
Newport News North
Petersburg
Prince George
Strasburg
Walkers
Warfield
Waynesboro East
Waynesboro West
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
•Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
Va.
B-72
-------
*r.
••>?-\'»;f
-., t _,;.ijVA
-------
1 • i-'sft* ''i
m 3 o
<- o =•
O Q. ?f
W ai CD
cn ~o ID
-^ o ^<
-------