Discharge modeling in compound channels with non-prismatic floodplains using GMDH and MARS models

Abstract In this study, modeling of discharge was performed in compound open channels with non-prismatic floodplains (CCNPF) using soft computation models including multivariate adaptive regression splines (MARS) and group method data handling (GMDH), then their results were compared the multilayer perceptron neural networks (MLPNN). addition to total discharge, separation between floodplain main channel modeled predicted. The parameters relative roughness coefficient, area flow cross-section, hydraulic radius, bed slope, width water surface, depth, convergence or divergence angle, longitudinal distance as inputs, considered output. demonstrated that statistical indices MARS, GMDH, MLPNN testing stage are R2 = 0.962(RMSE 0.003), 0.930(RMSE 0.004), 0.933(RMSE 0.004) respectively. Examination error shows all developed have appropriate accuracy estimate CCNPF. structure GMDH MARS parameters: roughness, area, aspect ratio, angle greatest impact on estimation discharge.