Assessment of Soft Computing Techniques for the Prediction of Suspended Sediment Loads in Rivers

A key goal of sediment management is the quantification suspended load (SSL) in rivers. This research focused on a comparison different means estimation includes rating curves (SRC) and soft computing techniques, i.e., local linear regression (LLR), artificial neural networks (ANN) wavelet-cum-ANN (WANN) method. Then, techniques were applied to predict daily SSL at Pirna Magdeburg Stations Elbe River Germany. By comparing results all best models, it can be concluded that (LLR, ANN WANN) better predicted than SRC due fact former employed non-linear for data series reconstruction. The WANN models overall performer. testing phase showed mean R2 0.92 PBIAS −0.59%. Additionally, they able capture peaks with greater accuracy. They more successful as captured dynamic features time-variant load, while other methods used simple raw data. Thus, could an efficient technique simulate time because extract embedded signal.