Improving Basins/HSPF Predictions of Nitrogen Export to Improve
TMDL Accuracy Using NASA Imagery
Investigators: Philip A. Townsend (PI, U. Wisconsin),
Angelica L. Gutierrez-Magness (U. Maryland),
Keith N. Eshleman (U. Maryland),
Brenden E. McNeil (West Virginia U.)
|
Abstract
|
|
Mathematical models are a critical element in predicting the effects of human activities
and natural processes on the quality of water in lakes and streams. Water quality
decision support systems such as BASINS (EPA software based on the hydrological simulation
model known as HSPF) are presently being used to simulate pollutant loads from large
complex multi-use watersheds in order to partition the total load of various constituents
into more management relevant components, thus improving the efficacy of nutrient management
efforts. Such partitioning is an important aspect of establishing and addressing
total maximum daily loads (TMDLs) of various pollutants -- a task presently mandated
by USEPA under the Clean Water Act regulations. Some of the more useful applications
of BASINS in this regard include
(1) categorizing non-point source loads as either controllable or non-controllable
to facilitate establishing meaningful tributary nutrient reduction goals
for Chesapeake Bay subwatersheds; and
(2) estimating pollutant loads by major land use category (e.g., forest, agriculture, urban, etc.).
Clearly, remote sensing imagery can play a significant role in
(1) partitioning pollutant loads among different land cover types;
(2) aiding in the detection of changes in nutrient retention due to land cover conversions
(i.e., the permanent change from one land cover type to another); and
(3) incorporating landscape dynamics (i.e., seasonal changes in vegetation density or properties)
into simulations of nutrient loads over time.
Remote sensing imagery might also play a useful role in addressing the problem of
non-point source pollution in situations in which nutrient loads show appreciable
interannual variation (as well as spatial variation), despite the fact that the
cover type remains unchanged (and is spatially uniform).
Previous research has shown that data from MODIS and Landsat can greatly increase the
predictability of stream nutrient concentrations in watersheds that have a significant
forest component.
The primary objectives of the proposed project are to:
(1) develop an approach to directly incorporate information derived from MODIS into BASINS
to improve the calibration of the water quality module of this decision support system (DSS);
and (2) improve estimates of annual non-point source loads of nitrogen (N) from forested lands
to surface waters. Measures of forest condition will be integrated with measures of landscape
condition to further improve predictions of annual N loads.
The study will be implemented in three regions with considerable water quality data available
for model calibration and validation: the Chesapeake Bay Watershed, the Adirondacks, and Wisconsin.
End-users include: (1) EPA, the agency responsible for oversight of the Clean Water Act
and maintenance of the BASINS modeling framework, as well as (2) the Maryland Department of
the Environment and (3) the Chesapeake Bay Program, both of whom are interested in improving
water quality predictions to assist regional, state, and local agencies in performing
water quality studies and establishing TMDLs in the Chesapeake Bay region.
The primary activities of the project will be:
(1) derivation of annual image-derived metrics of landscape condition;
(2) incorporation of the derived metrics into the EPA/BASINS software to demonstrate
its use for predicting surface water N concentrations and loads; and
(3) interacting with end-users to incorporate the improvements into the decision-making process.
|
|
|
NASA Products
MODIS, Landsat
Project Partners
EPA, Maryland Dept. of the Environment, Chesapeake Bay Program
Decision Support Tools
BASINS (EPA software based on the hydrological simulation model known as HSPF)
Reports
