An integrated simulation-optimization framework to optimize the reservoir operation adapted to climate change scenarios

摘要: Abstract In order to optimize the operation rule curve of Shahrchay reservoir in north-west Iran under climate change, a new integrated simulation-optimization framework different change scenarios was introduced current study. The proposed contains several steps: First, more accurate general circulation models (GCMs) were chosen among available GCMs and predictors screening based on decision tree model (M5) applied select most important enormous number potential large-scale variables GCMs. Then, artificial neural networks (ANNs) trained by observed temperature precipitation time series selected dominant downscale parameters for base period (1951–2000) over case area future monthly predictions implemented ANNs A1B, B1, RCP4.5, RCP8.5 assess changes between 2020 2060. Thereafter, estimated values imposed hybrid Wavelet-M5 simulate inflow (runoff) finally, genetic algorithm (GA) used curves system using predicted average precipitation, evaporation, considering water supply agricultural municipal targets minimization total squared deficiencies future. could result reliable performance one-step-ahead runoff prediction as obtained correlation coefficient (CC) training verifying data sets 98% 97.7%, respectively. results showed that long-term annual volume may be decreased 0.08% 2.27% with regard period, simulation present conditions indicated decrease availability. Also, optimal shape all employed this