PREDICTION OF RIVER DISCHARGE AND SURFACE-WATER QUALITY USING AN INTEGRATED GEOGRAPHICAL INFORMATION-SYSTEM APPROACH

A methodology is developed for the prediction of river discharge and s urface water quality (indexed by nitrogen loading) of a predominantly rural catchment using simple models in an integrated Geographical Info rmation System (GIS). River discharge is predicted using the Soil Cons ervation Service (SCS) runoff Curve Number model, and surface water qu ality by the export coefficient model. Main input variable to these mo dels is information on land-use along with ancillary information such as soils. Land-use is an important parameter that affects both dischar ge and water quality, and it can be derived from classification of rem otely sensed images. Unlike conventional models, the models employed h ere do not require large amounts of data on several hydro-meteorologic al variables. The models are applied to a rural catchment in eastern E ngland where major land-use changes have occurred in the recent past. Historical land-use data are derived from a variety of sources includi ng maps, aerial photographs and remotely sensed satellite images for v arious dates ranging from 1931 to 1989. A GIS is a valuable means to e nable large amounts of spatial data to be integrated, and to facilitat e data manipulation for the specific application of the models. Result s are validated using observed runoff and water quality records, and i t is shown that the model predictions are of acceptable accuracy. This study demonstrated an application of a GIS to employ simple models to predict river discharge and water quality.