Using Artificial Neural Networks for Modeling Suspended Sediment Concentration

For continuous monitoring of river water quality , this study assesses the potential of using artificial neural networks (ANNs) for modeling the event-based suspended sediments concentration (SSC) in Jiasian diversion weir in southern Taiwan. The hourly data collected include the water discharge, turbidity and SSC during the storm events. The feed forward backpropagation network (BP), generalized regression neural network (GRNN), and classical regression were employed to test their performances. The results showed that the performance of BP was slightly better than GRNN model. In addition, the classical regression performance was inferior to ANNs. Thus, using ANNs were more reliable than the classical method. Furthermore, the turbidity is a dominant variable over water discharge for SSC estimation in the weir. Key-Words: Artificial neural networks, suspended sediments concentration, turbidity, water discharge, modeling.