SYNTHETIC HYDROLOGY AND WATER-SUPPLY RELIABILITY

Synthetic hydrology is a tool that may be used in evaluating the consequences of water supply management decisions. Given that there are many proposed generating methods for synthetic hydrology, it is necessary to evaluate the performance of various synthetic streamflow generators (SSG's). Six different single‐site, monthly generators are considered here. The comparison demonstrates that preservation of statistical moments (mean, standard deviation, and lag correlation coefficients) may be a misleading criterion for judging the ability of an SSG to provide plausible estimates of water supply system performance. It is shown that it is operationally superior to base the specification of marginal distributions of monthly streamflows on transformed values of historical streamflow data rather than on the data itself. Two transformations are considered, the log transform and a newly developed normalizing transform. A new type of autoregressive‐moving average model is shown to be operationally superior to an autoregressive model. This paper is not subject to U.S. copyright. Published in 1979 by the American Geophysical Union.