Estimation of Surface Water Quality Changes in Response to Land Use Change: Application of The Export Coefficient Model Using Remote Sensing and Geographical Information System

A trend of increasing solute (i.e. nitrogen and phosphorus) concentrations in surface water has been observed in the recent past. Solute concentrations have often exceeded the World Health Organization directives on the quality of drinking water. In predominantly rural watersheds of developed countries, increasing solute concentrations have been attributed to agricultural intensification (i.e. land use change) and increased fertilizer usage. This paper presents an approach based on remote sensing and Geographical Information System (GIS) to estimate and hindcast water quality changes using historical land use data for a watershed in eastern England. The water quality, as indexed by the nitrogen loading, was estimated using the export coefficient model. The model calculates solute loading at the outlet of a watershed using land use data, fertilizer application rates and export coefficients. Historical land use data were derived from a variety of sources including maps, aerial photographs and remotely-sensed Landsat and SPOT satellite images. Export coefficients indicate the proportion of the total nutrient lost to the surface water, and determine the amount of nutrient loading from various land use types. Export coefficients for various land use classes were compiled from literature. A GIS was necessary to store, manipulate and analyse large volumes of land use data, and to carry out land use change analysis. A number of special GIS routines were developed to carryout the model computations automatically. Results demonstrated that the model estimates nitrogen loading with an acceptable degree of accuracy. Analysis showed an overall increase of nitrogen loading from 1931–1984, after which it remained steady. Experiments were carried out to identify the separate effects of land use and fertilizer application on the nitrogen loading. The export coefficient model operating in a GIS has significant potential for the rapid estimation of surface water quality using land use data derived from remotely sensed satellite images.