The ﬂow assessment in a river is of vital interest in hydraulic engineering for ﬂood warning and evacuation measures. To operate water structures more efﬁciently, models that forecast river discharge are desired to be of high precision and certain degree of accuracy. Therefore, in this study, two artificial intelligence models, namely kernel extreme learning machine (KELM) and multivariate adaptive regression splines (MARS), were applied for the monthly river flow (MRF) modeling. For this aim, Mississippi river with three consecutive hydrometric stations was selected as case study. Using the previous MRF values during the period of 1950–2019, several models were developed and tested under two scenarios (i.e. modeling based on station's own data or previous station's data). Wavelet transform (WT) and ensemble empirical mode decomposition (EEMD) as data processing approaches were used for enhancing modeling capability. Obtained results indicated that the integrated models resulted in more accurate outcomes. Data processing enhanced the model's capability up to 25%. It was observed that the previous station's data could be applied successfully for MRF modeling when the station's own data were not available. The best-applied model dependability was assessed via uncertainty analysis, and an allowable degree of uncertainty was found in MRF modeling.
Kernel extreme learning machine (KELM) and multivariate adaptive regression splines (MARS) approaches were used for MRF modeling in three successive hydrometric stations.
The WT and EEMD as pre-processing methods were used for improving the model's efficiency.
Monte Carlo uncertainty analysis was applied to investigate the dependability of the applied models.