Nonstationary Regional Frequency Analysis of Maximum 24-Hour Precipitation in West of Iran

During last century, anthropogenic forcing and climate change have influenced the stationarity of the many extremes. Therefore, using of stationary pre-assumption-based Frequency Analysis (FA) techniques for estimating the risk of extremes may not be reliable anymore. In present study we introduced a nonstationary FA method to estimate risk of 24-hour maximum precipitation variable, as one of the major factors for generating floods in West of Iran. Nonstationarity assessments showed that among 53 investigated stations, only 4 stations (i.e., Khorramabad and Poldokhtar stations with increasing linear trends and Izeh and Ramhormuz stations with decreasing linear trends) have significant nonstationarity. Three hydrological homogeneous regions were characterized based on statistical properties of 24-hour maximum precipitation variable at study stations. Results revealed that stations in the Region_II, mostly including the stations in Fars, Chaharmahal and Bakhtiari, Bushehr and Kohgiluyeh and Boyerahmad Provinces, have higher risks for more severe 24-hour maximum precipitation compared to stations in other regions. Also, results showed that if nonstationarity in Khorramabad and Poldokhtar stations is ignored, 100-year quantiles of 24-hour maximum precipitation will underestimate around 10.1% and 13.8% in those stations.