Investigating trends of hydrochemical time series of small catchments by artificial neural networks

The short-term variation of discharge and solute concentration of the runof f of small catchments generally reflects the interplay of a variety of diff erent processes. This makes the investigation of anthropogenic impacts on t he catchment's runoff often rather difficult. On the other hand, short-term dynamics at the output boundary provide information about the system. This information can be used, in principle at least, to assess its long-term be haviour more precisely. In this paper examples of time series of sulphate a nd nitrate in the runoff of two small forested catchments are presented. To minimise the danger of over-parametrisation, the objective was to find st very simple empirical model to map a substantial portion of the observed va riance (daily values). Here artificial neural networks were applied. They y ield an efficiency of more than 0.7 for the solutes investigated, based on discharge depth and air temperature as input variables only. As a next step , the invariance of these relationships was investigated. In the case of su lphate, a significant trend is observed. However, it differs considerably f or different subregions of the regression plane. Thus the neural network ap proach reveals a much more detailed insight into temporal shifts of the dyn amics than an overall trend analysis. (C) 2000 Elsevier Science Ltd. All ri ghts reserved.