Generalised storage-yield-reliability modelling: Independent validation of the Vogel–Stedinger (V–S) model using a Monte Carlo simulation approach

The generalised storage-yield-reliability model by Vogel and Stedinger (V–S) is widely used in the literature but there has never been a complete validation of the model with measured river runoff data. This study investigated the problem using three different river runoff data records within a Monte Carlo simulation framework. The V–S model is predicated on the assumption that the probability distribution of reservoir capacity can be described by the 3-parameter log-normal density function. The results showed that the V–S model over-predicted reservoir capacity quantiles for most yield (or demand) ratios but especially so for very high (>0.8) yield ratios. Further investigations revealed that the poor performance of the V–S model was caused by the huge errors associated with its expression for two of the three parameters of the 3-parameter log-normal distribution, i.e. the variance and the lower limit, although the mean of the distribution was relatively well predicted.