Estimation of water temperature exceedence probabilities using a deterministic-stochastic modeling approach

A continuous simulation approach is proposed for estimating water temperature exceedence probabilities. A deterministic thermo-hydrodynamic model is used to simulate flows and water temperatures based on historical and synthetic records. Representative hydrologic and meteorologic records are obtained from nearby continuous gages. Synthetic records of input water temperatures and other inputs are generated using stochastic models calibrated with short sample records. The approach is used to estimate water temperature exceedence probabilities for the Upper Illinois Waterway based on a 46-year simulated record. Sampling resampling methods (bootstrapping) are used to estimate the uncertainty in exceedence probability estimates.