Assessment of regime shifts and risk of extreme precipitation in urban regions in Illinois, USA

Intensification of extreme precipitation, especially in urban regions, is mostly responsible for causing flash floods leading to significant damages of civil infrastructure and loss of human lives throughout the world. Understanding the pattern and risk of these extreme events could be immensely beneficial in the planning and managing built infrastructure under climate change. Although many studies have been conducted addressing extreme precipitation events, there is still a lack of understanding of these events and associated risk at the regional scale in a non-stationary environment.  The objectives of this study were to (a) identify the regime shifts in extreme precipitation (R95) using Fisher Information in Illinois, USA; (b) determine the trend of R95 applying Modified Mann-Kendall method; (c) quantify the risk associated with R95 applying Extreme Value Theory (EVT). Daily precipitation data from 1950-2010 was collected for 78 urban gauge stations from the United States Department of Agriculture –Agriculture Research Service (USDA-ARS) database, including Cooperative Observer Network (COOP), Weather-Bureau-Army-Navy (WBAN), and Midwestern Regional Climate Center (MRCC). The future precipitation data was collected for regional climate models (RCMs) under two representative concentration pathway (RCP) scenarios (i.e., RCP 4.5 & RCP 8.5) for 2021-2100 from the NA-CORDEX data archive. All the future data was bias-corrected using gauge station data applying the quantile mapping technique. Initial results showed a significant bias in the RCM dataset. Further, most of the regime shits in R95 were identified between 1971-1985 in the urban regions.