Daily reservoir inflow forecasting using weather forecast downscaling and rainfall-runoff modeling: Application to Urmia Lake basin, Iran

This study develops the first daily runoff forecast system for Bukan reservoir in Urmia Lake basin (ULB), Iran, a region suffering from water shortages and competing demands. A weather downscaling model is developed large-scale raw forecasts of ECMWF NCEP to small-scale spatial resolutions. Various methods are compared, including deterministic Artificial Intelligence (AI) techniques Bayesian Belief Network (BBN). Downscaled precipitation temperature then fed into rainfall-runoff that accounts snow soil moisture dynamics sub-basins upstream reservoir. The multi-objective Particle Swarm Optimization (MOPSO) method used estimate hydrological parameters by maximizing simulation accuracy observed river flow (NSEQ) logarithm (NSELogQ) each sub-basin. Results show BBN greater than various AI tested. Calibration results indicate no significant trade-off between fitting high low flows, with an average NSEQ NSELogQ 0.43 0.63 calibration period, 0.54 0.57 validation period. entire forecasting was evaluated using inflow observations years 2020 2021, resulting NSE 0.66 can be policymakers operators optimize allocation agricultural environmental demands ULB.