&EPA
United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens GA30613
%i
Research and Development
EPA-600/S3-81-037 July 1981
Project Summary
Evaluation of Remote Sensing
Data for Input into
Hydrological Simulation
Program-FORTRAIM (HSPF)
D. D. Franz and S. M. Lieu
This report evaluates the feasibility
of using a remotely sensed database
as an input into the Hydrologic Simu-
lation Program - FORTRAN (HSPF).
Remotely sensed data from the
LANDSAT satellite and conventionally
obtained data were used to set up the
input parameters of HSPF. Simulations
were then run to compare the model
outputs from the two sets of input
data.
The remotely sensed data set per-
formed at least as well as the conven-
tional one when model results were
compared with observed data. Further,
the remotely sensed data set offered
savings estimated at 30 to 50 percent
in the cost of set-up and operation.
Areas of potential development of
remote sensing for modeling input are
suggested.
The Image Based Information Sys-
tem (IBIS) used to manipulate the
LANDSAT data was found to contrib-
ute significantly to the data collection
and handling process.
This Project Summary was devel-
oped by EPA's Environmental Research
Laboratory, Athens, GA, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
More and more frequently, planners
and scientists are turning to computer
modeling for assistance in understand-
ing the complexity of natural phenomena,
their interactions, and the impact of
human activities on the environment. To
successfully model the complex and
subtle components of the environment,
many types of accurate data describing
large geographical areas are needed.
These data are essential to the con-
struction, execution, testing, and ulti-
mately, to the improvement of any
model. Available data, however, is not
always adequate. Tremendous potential
for improving the ease and accuracy of
modeling efforts lies in the development
of satellites able to remotely acquire a
wide variety of environmental data over
large areas of the earth's surface.
Whether it .is feasible to use satellite
input into complex environmental models
is an area explored in this study.
Data from the LANDSAT satellite
were supplied by the National Aeronau-
tics and Space Administration (NASA).
Terrestrially based data were provided
by the Northern Virginia Planning District
Commission (NVPDC). These data were
used as input in simulation runs with
the Hydrological Simulation Program -
FORTRAN (HSPF). The questions ad-
dressed include the feasibility of using
remotely sensed data as model input,
difficulties in the implementation of this
-------
approach, and the modeling accuracy
and cost effectiveness compared with a
conventional database. The contribu-
tion of ancillary computer hardware/
software for handling LANDSAT infor-
mation to modeling, particularly the
Image Based Information System (IBIS),
is also assessed.
Conclusions
Conclusions are drawn for both the
results of the simulations themselves
and the process of use of remotely
sensed databases in HSPF modeling.
In terms of the simulations themselves,
the remotely sensed database performed
at least as well as the conventional
database. This is supported by the find-
ings below:
• Differences in land-use distribu-
tion (Broad Run of Occoquan Basin)
had little effect on the overall out-
come of HSPF modeling. The
changes in parameters resulting
from differing land use distributions
tended to balance out during
simulation.
• Both the conventional and the
remotely sensed parameter sets
performed well for certain simu-
lations. The results for flow rate,
water temperature, and dissolved
oxygen are comparable to observed
data.
• Neither parameter set seemed to
perform well in nutrient simula-
tions. Assessing the accuracy of
these results, however, was diffi-
cult because of inadequacies in the
observed data and the modeling
approach.
• Overall, the parameter set from the
remotely sensed database per-
formed at least as well as the
conventional set. Remote sensing
also offered some additional ad-
vantages.
• Refinement of information on urban
and water areas in the remotely
sensed land-use classification
would result in better segment
definition.
Conclusions concerning the use of a
remotely sensed database with IBIS are:
• The current use of remotely sensed
information is relatively indirect:
classes of land use are processed
by a transformation function to
generate modeling parameters.
With the availability of the appro-
priate transformation function(s),
the use of remotely sensed data-
bases for HSPF and other types of
modeling is feasible.
• Use of a remotely sensed database
offers significant savings in cost
(about 30 to 50 percent) in com-
parison with use of a conventional
database. Larger study areas should
provide even greater economic
advantages.
• A remotely sensed database offers
advantages of repeatability, relative
rapidity and ease of information
collection. It would be particularly
advantageous in large or inacces-
sible areas or in areas requiring
frequent database updates.
• IBIS made significant contributions
in verifying segment definition and
in computing basin wide average
parameters and other cross tabula-
tions, which greatly enhanced the
use of the remotely sensed data-
base. The image formatted data
plane concept of IBIS is vital to the
integration of various sources and
types of data, particularly on a large
scale.
Recommendations
Recommendations for further use of
remote sensing in hydrologic and water
quality simulations include:
• Additional development of the data
collection and transformation
processes. Because its lack is the
main impediment to more intensive
and extensive use of remote sens-
ing, further development and gei
eralization of transformatic
processes is needed.
• Refinement of the remotely senst
land-use classification of urba
areas. This would enhance the us
of satellite information in areas <
rapid urban development, wher
updates are required to accurate
assess changes in hydrology an
water quality.
• Use of remotely sensed database
for modeling agricultural or unde
veloped areas, particularly if thes
are large or inaccessible.
• Assessment of information the
can be obtained for modeling pui
poses from future satellite deve
opments. In long-range terms, thi
is a valuable avenue of researc
because the availability of nev
data may significantly affect o
even entirely modify the modelini
approach.
The utilization of IBIS occurred sub
sequent to the initiation of the curren
effort. A more comprehensive evalua
tion of the use of HSPF and remoti
sensing, particularly with the use o
IBIS, could be conducted on a signifi
cantly larger area. In conjunction, it i:
recommended that the digital terrair
data from the Defense Mapping Agenq
(DMA) be utilized to improve basin slop*
determinations.
D. D. Franz andS. M. Lieu are with Hydrocomp, Inc., Mountain View, CA 94040.
T. O. Barnwell is the EPA Project Officer (see below).
The complete report, entitled 'Evaluation of Remote Sensing Data for Input into
Hydrological Simulation Program-FORTRAN (HSPF)." (Order No. PB 81 -209 561;
Cost: $11.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, GA 30613
•> US. GOVERNMENT PRINTING OFFICE. 1M1 -757-OU/7ZZ4
-------
^ Postage and
United States Center for Environmental Research pees pajQ<
Environmental Protection Information Environmental
Agency Cincinnati OH 45268 Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS 0000329
AGENCY
------- |