Evaluating Parameters of the Distributions of Extreme Storms with Several Events per Year Taken into Account

Abstract

Applied problems of the analysis of data series of flood-forming storm precipitation (rain intensity over short time intervals), containing several events per year are discussed. The use of data containing several events per year has been shown to be justified for reliable determination of statistical characteristics of time series at a short observation period. The statistics of time series containing one or more events per year have been shown to correlate well with the frequency of the observed phenomenon in the Ural region. Recommendations have been developed for recalculating the statistics of the series containing several events per year into statistics for a single event per year; a brief comparative analysis of the methods used in Russia in this field is given.