Hydrologic processes are affected by the spatial variability of soils, topography, land use and cover, climate, and humaninduced changes and management. Because contaminants are transported by runoff to surface waters and by infiltration and deep percolation to groundwater, hydrologic processes are often at the core of water quality and quantity concerns. Water quality and quantity affected by non-point sources of contaiminants is dependent on spatially distributed attributes of a catchment or watershed.
The integration or linkage of the spatial data handling capabilities of a GIS with a distributed process, hydrologic model offers the advantages associated with utilizing the full information content of the spatially distributed data to analyse the hydrologic processes. Distributed process models use computational elements smaller than the scale of the process or the system being modelled. The information content of the model output is greater than a lumped model though obtained at greater cost. A CIS is capable of manipulating both the input and output parameters required by distributed process models.
This paper presents a review of water quality and quantity modelling and GIS applications in water resources. An application of the GIs, ARC/INFO and the finite element solution to the kinematic wave equations is presented. In this application, ARC/INFO is used to process the spatially variable terrain in a small watershed using a Triangular Irregular Network on TIN. The TIN facets are used to provide land surface slope in a finite element solution of overland flow which is a fundamental subprocess affecting non-point source, water quality and quantity.
KEY WORDS Geographic Information Systems Water quality and quantity Non-point source Finite Element Method
Catchment runoff Modelling of the spatial and temporal variability in the system. Because of this, values of the model parameters are often not known or easily represented in the conceptual model. The 'unknowability' of this variability described by Beven (1986). presents an impediment to the progress of hydrologic science. Distributed process models act directly on the spatially distributed data to model hydrologic transport processes affecting water quality. The advantages of distributed process models versus lumped models is that the effect of spatially variable data may be calculated and analysed throughout the watershed or hydrologic unit. A disadvantage of distributed process models are the large amounts of data to be compiled and analysed. Further, model parameters or the distribution throughout the watershed may not be known, resulting in the model's capability exceeding data availability.