Probabilistic Forecasts of Streamflow Scenarios Using ESP Approach (Case study: Halil River)

Stream flow forecast is a fundamental tool that could be utilized for appropriate water resource management. The enhancement of accuracy as well as the increase of time horizon for stream flow forecasts is essential especially for agricultural sector which is the most vulnerable sector during water shortages. In this regard, the application of stochastic approaches like Ensemble Streamflow Prediction (ESP) procedure for long-term forecast with respect to streamflow uncertainty can be proposed. The ESP procedure produces streamflow forecasts in the form of multiple hydrographs, each a possible realization of seasonal streamflow (Day, 1985). One example of ensemble streamflow forecasts is the National Weather Service’s ESP procedure. Faber and Stedinger (2001) successfully combined reservoir operation models with updated information from ESP of the National Weather Service (Faber and Stedinger, 2001). Using ESP forecasts, Eum et al. (2011) also developed a procedure to calculate optimal water release curtailments during droughts using a future value function derived with a sampling stochastic dynamic programming model.
The main objective of this paper is to present a probabilistic approach and forecast the inflows to Jiroft dam reservoir. In this regards, using ESP approach as well as Artificial Neural Networks (ANNs) the 1- to 12- month ahead probabilistic scenarios of Halil river were forecasted