Precipitation extremes derived from temporally aggregated time series and the efficiency of their correction

Abstract

ABSTRACT Extreme precipitation intensities derived from temporally aggregated time series can be considerably underestimated. Therefore, some form of correction is appropriate prior to their usage e.g. for the derivation of idf curves. The correction is usually performed in a multiplicative manner, using the coefficient obtained as a mean ratio of real to aggregated extremes. In this paper a novel correction approach is derived, allowing an individual treatment of each aggregated extreme. The precipitation time series from Prague (central Europe) are used for the assessment of newly introduced method. The comparison with the standard approach shows that the new method reaches better results as it reduces an undesirable overestimation of the corrected extremes. Moreover, the effect of data aggregation on extreme precipitation intensities is evaluated as well as the effect on quantiles of the generalized extreme value (GEV) distribution. The results are compared to those published for other climatic conditions.