ESTIMATION OF SPATIAL DISTRIBUTION OF PROBABLE RAINFALL BY REGIONALIZATION

Abstract

Extreme environmental events such as rainstorms give severe consequences to human society. Magnitude evaluation of an extreme event for a given return period is an important step in planning of flood prevention facilities. The estimation of the spatial distribution of magnitude is difficult because extreme events are rare and the data record is often short. Regionalization resolves this problem by “trading space for time”; data from several sites are used in estimating event frequencies at any one site. Applying regional frequency analysis, regression analysis and kriging method to Miyagi prefecture, we have obtained the estimates and the accuracy for probable rainfall and discussed incorporation of orographic effects.