Abstract This study contains two major tasks. The first task is concerned with the identification of representative significant storm patterns in Wyoming. Significant storms refer to those events with the total depth equal to or greater than that of a IO-year storm. The method employed was cluster analysis. Contingency tests were performed to examine the dependency of the occurrence of storm patterns on the climatic region, storm duration, and seasonality. Parametric models were used to fit storm patterns. The second task is the development of procedures to stochasticaBy generate storm patterns. Three methods based on the multivariate Johnson system are proposed. An additional issue addressed under this task is the selection of alternative normal transformations with favorable small sample qualities. Four such procedures are proposed and their performance is examined.
Water Resources Publications List
Water Resources Data System Library | Water Resources Data System Homepage