Quantile trends of subhourly extreme rainfall: Marmara Region, Turkey

Abstract

Global climate change will probably cause intensification of the hydrologic cycle, which can lead to alterations in extreme precipitation properties. In this study, we investigated the trend of 5-, 10-, 15-, and 30-min annual maximum rainfall series at 12 stations in the Marmara Region, Turkey, using quantile regression. The data ranges were from 46 to 71 years long. Five quantiles were used to examine the extreme rainfall series, and their quantile regression parameters were calculated. The results show that quantile regression is a powerful tool to compute trends with a more inferential context, which was validated with the notable differences between the trends at chosen quantiles and the classical ordinary least squares method. Concerning the problem of the analysis of climate trends, the quantile regression method seems to provide a perspective from a more detailed understanding of processes in the climate system in terms of characteristics of climate variability and extremity.