Quantifying Temporal Changes in Sub-daily Precipitation Intensities Across Agricultural Regions of the U.S.

More frequent and intense precipitation events can result in damaging floods and have high societal impacts. While in the eastern U.S increases in daily precipitation intensities have been documented during the second half of the 20st century, in the southwest U.S. natural variability has made it difficult to identify trends in extreme precipitation. This study aims to quantify changes in precipitation intensities at different temporal scales across agricultural regions of the U.S. during the last 4-5 decades. The USDA-Agricultural Research Service (ARS) experimental watersheds, now largely part of the LTAR network, have a unique long-term and high-resolution precipitation measurement dataset that permit a detailed analysis of storm characteristics. We begin the multi-site evaluation at a semi-arid watershed in southern Arizona, the Walnut Gulch Experimental Watershed LTAR, and a humid watershed, the Mahantango Creek Watershed (Upper Chesapeake Bay LTAR), in eastern Pennsylvania. Changes in sub-daily precipitation intensities for durations ranging from 2 minutes to daily are analyzed with an over-the-threshold approach. A Generalized Pareto distribution function is used to compare precipitation intensities for different return periods. This work is part of ARS effort to quantify climate change impacts on watershed response across agro-eco regions the U.S.