Suspended Sediment Estimation for Rivers using Artificial Neural Networks and Sediment Rating Curves

The methods available for sediment concentration and flux estimation are largely empirical, with sediment rating curves being the most widely applied. In this study, a comparison is made between artificial neural networks (ANNs) and sediment rating curves for two rivers with very similar catchment areas and characteristics in the north of England. Data from one river are used to estimate sediment concentrations and flux in the other for both estimation techniques. A more traditional, split-sample approach is also used, in which part of the available data from a site is used to develop a predictive relationship, which is then tested with the remaining data from the same site. The results of the two estimation techniques and the two approaches for the derivation of predictive capability are compared and discussed. The potential advantages of ANNs in sediment concentration and flux estimation are highlighted. In particular, an ANN approach can give information about the structure of events (e.g., hysteresis in the sediment concentration water discharge relationship, and the effect of antecedent conditions) which is impossible to achieve with sediment rating curves.