Uncertainty Analysis of Reservoir Operation Based on Stochastic Optimization Approach Using the Generalized Likelihood Uncertainty Estimation Method

This study evaluated the reservoir operation for effective water allocation under uncertainty using linear programming (the water evaluation and planning model: WEAP) and evolutionary optimization (the particle swarm optimization algorithm: PSOA). The stochastic autoregressive integrated moving average (ARIMA) and seasonal autoregressive integrated moving average (SARIMA) models were used to forecast monthly inflow into the AmirKabir dam reservoir (Iran) as well as its net evaporation (2018-2025), respectively. For this purpose, the best-fitted model was selected based on the minimum Akaike criterion and the autocorrelation function (ACF) test. The resulting uncertainty of monthly release for the WEAP and PSOA were analyzed using the generalized likelihood uncertainty estimation method (GLUE). The results showed that the WEAP model proved to be better in meeting water demands during low inflows, whereas the PSOA had a higher certainty in meeting demands during high inflows periods. Also, the WEAP model had a high uncertainty in January-April compared to May-December, whereas the PSO algorithm had a high uncertainty in all months. This evaluation of water allocation considering uncertainties of fluctuating water supply and net evaporation helps us answer questions about optimal allocating of different demand sites: how is water shortage affects economic, social, environmental aspects, and the relationship between systems sustainability and water shortage.