RESERVOIR OPTIMIZATION USING SAMPLING SDP WITH ENSEMBLE STREAMFLOW PREDICTION (ESP) FORECASTS

Abstract

The National Weather Service (NWS) produces ensemble streamflow prediction (ESP) forecasts. These forecasts are used as the basis of a Sampling Stochastic Dynamic Programming (SSDP) model to optimize reservoir operations. The SSDP optimization algorithm, which is driven by individual streamflow scenarios rather than a Markov description of streamflow probabilities, allows the ESP forecast traces to be employed directly, taking full advantage of the description of streamflow variability, and temporal and spatial correlations captured within the traces. Frequently-updated ESP forecasts in a real-time SSDP reservoir system optimization model (and a simpler two-stage decision model) provide more efficient operating decisions than a sophisticated SSDP model employing historical time series coupled with snowmelt-season volume forecasts. Both models were driven by an appropriately weighted and representative subset of the original forecast and streamflow samples.