A cost effective method for generating extended streamflow predictions (ESP) is presented. ESP methods provide probabilistic streamflows during any designated time period. The method presented is based on a sensitivity analysis of the rainfall inputs and resulting streamflow of calibrated rainfall/runoff (R/R) models. The sensitivity approach assumes that knowledge of the variances and sensitivities (partial derivatives) of R/R model input parameters allows one to determine the variance of the output. Monthly rainfall is assumed Markovian. This assumption is used in selection of the input rainfall. The sensitivity approach is applied to the Occoquan watershed in northern Virginia for a period of severe drought. The streamflow variance is within 12% of the value obtained by the current National Weather Service (NWS) technique and required less than 10% of the computer effort required by the NWS technique.
No takes yet. Share an insight, caveat, or question.
Young et al. (1980) studied this question.