Modelling point and non-point source water pollution in large drainage basins

The paper describes the features and capabilities of two model systems, computation tools, developed at VITUKI, that were especially designed for handling the water pollution issues of large river (lake) basins, providing also a decision support tool for the evaluation of overall management strategies. These model systems are SENSMOD and NOSNUL. Recent versions of SENSMOD are designed for GIS based application, with relative abundance of data on land uses, point source discharges and non-point land application/deposition rates of pollutants. The model describes the hydrological and pollutant load/concentration conditions over a complex water system for a larger period of time (e.g. annual or multiannual runoff and load conditions) by trying to ''explain'', with a self-calibration algorithm, the observed quantitative and qualitative state of the water system, in their relation to natural and manmade inputs (specific runoff, point source discharge, non-point source application rates). It considers runoff and mass transport routes over land and in-stream. The main output of the model is the spatially distributed identification and quantification of pollutant loads originating from non-point (unidentified) sources. NOSNUL relies on the same principle as SENSMOD, but assumes the existence of a very limited data base of very large drainage basins. It calculates water and mass balances of larger sub-drainage basins on the basis of hydrographic, water quality and point source monitoring data. Next it attempts to explain ''missing'' load components as those of non-paint sources by relating these missing load fractions to land-uses of the basin, using literature based unit area loading rates and a single delivery rate coefficient, to be determined by calibration against measurement data for each sub-basin. The use of the models are illustrated by references to case studies: 1. the Rhine River Basin study where SENSMOD was applied for the identification of non-point source contributions of selected chemicals (cadmium, phosphorus, nitrogen and BOD), 2. A Lake Balaton case study when SENSMOD was adapted to a digital terrain model of the Zala River Basin, using a detailed inventory of fertilizer application and accumulation on agricultural lots for identifying the fate of phosphorus within the river basin. 3. Application of NOSNUL for the Hungarian part of the Danube basin (with an outlook for application to the entire Danube basin) for identifying sources of plant nutrients.