Assessment of the hydrological and coupled soft computing models, based on different satellite precipitation datasets, to simulate streamflow and sediment load in a mountainous catchment

Abstract This study evaluated the performance and hydrologic utility of four different satellite precipitation datasets (SPDs), including GPM (IMERG_F), PERSIANN_CDR, CHIRPS, CMORPH, to predict daily streamflow SL using SWAT hydrological model as well coupled soft computing models (SCMs) such artificial neural networks (SWAT-ANNs), random forests (SWAT-RFs), support vector regression (SWAT-SVR), in mountainous Upper Jhelum River Basin (UJRB), Pakistan. SCMs were developed outputs un-calibrated improve predictions. Overall, shows highest for entire simulation with R2 PBIAS varying from 0.71 0.96 −13.1 0.01%, respectively. For best GPM-based models, SWAT-RF showed a superior ability simulate 0.96, compared SWAT-ANN (R2 = 0.90), SWAT-SVR 0.87), SWAT-CUP 0.71). Similarly, presented capability an 0.71, 0.66), 0.52), 0.42). Hence, based on SPDs could be effective technique simulating parameters, particularly complex terrain where gauge network density is low or uneven.