TY - JOUR T1 - Suspended sediment load estimation by using feed forward back propagation (FFBP) and radial basis function (RBF), (Songurchay river) TT - تخمین بار معلق رسوبی رودخانه سنگورچای با استفاده از دو مدل شبکه عصبی مصنوعی JF - geospace JO - geospace VL - 16 IS - 56 UR - http://geographical-space.iau-ahar.ac.ir/article-1-694-en.html Y1 - 2017 SP - 115 EP - 131 KW - Suspended sediment load KW - feed forward back propagation KW - radial basis function KW - Akaike Information System KW - Songurchay river N2 - Songurchay river basin is one of the Qaranghu river branches with high volume of sediments load. Due to this fact, feed forward back propagation model (FFBP) and radial basis function (RBF) was used in this study to estimate the suspended sediment load. Due to the non-linear nature of the suspended sediment load these using models as nonlinear models is inevitable to simulate these parameters. However, the input parameters for each model vary and in a single phase only discharge data were used and then beside the discharge data rainfall data were also used in each model. Then, to determine the efficiency of models, root mean square error (RMSE) and the error of determination (R2) was used and it was observed that the RBF model in the case of using two parameters, discharge and rainfall as an input parameters, has achieved better results with 0.9251 R2 and the root mean square error equivalent to 265 mg in liter. Finally, to assess the RMSE parameter the Akaike Information System (AIC) was used and it was observed that RBF models by having Akaike values equivalent to 1042 was more capabile. M3 ER -