ATMOSPHERIC SCIENCE LETTERS Atmos. Sci. Let. 0: 000–000 (2013) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/asl.440 Projections of heavy rainfall over the central united states based on CMIP5 models Gabriele Villarini, 1 * Enrico Scoccimarro 2,3 and Silvio Gualdi 2,3 1 IIHR-Hydroscience & Engineering, The University of Iowa, Iowa City, IA, USA 2 Centro euro-Mediterraneo sui Cambiamenti Climatici •, Bologna, Italy AQ1 3 Istituto Nazionale di Geofisica e Vulcanologia, Bologna, Italy *Correspondence to: G. Villarini, IIHR-Hydroscience & Engineering, The University of Iowa, Iowa City, IA, USA. E-mail: gabriele-villarini@ uiowa.edu Received: 18 January 2013 Revised: 7 May 2013 Accepted: 8 May 2013 Abstract The central United States is a region for which observational studies have indicated an increase in heavy rainfall. This study uses projections of daily rainfall from 20 state-of-the- art global climate models and one scenario [representative concentration pathway (RCP) 8.5] to examine projected changes in extreme rainfall. Analyses are performed focusing on trends in the 90th and 99th percentiles of the daily rainfall distributions for two periods (2006–2045 and 2046–2085). These results indicate a large increase in extreme rainfall mostly over the northern part of the region, with a much less clear signal over the Great Plains and states along the Gulf of Mexico. Copyright  2013 Royal Meteorological Society Keywords: extreme rainfall; central united states; CMIP5; trends 1. Introduction The central United States is a vital area for the US agri- culture, producing more than half of the total harvested area of principal crops in the US (US Department of Agriculture, 2011). It is also highly susceptible to catastrophic flooding, with the 2008, 2010 and 2011 being the latest events of a long list, causing billions of dollars in economic damage and numerous fatalities (http://www.ncdc.noaa.gov/oa/reports/billionz.html). Because of these large societal and economic impacts, it is of paramount importance to assess poten- tial changes in the upper tail of the rainfall dis- tribution over this area. Several studies based on observational records found increasing trends over the central United States(Karl and Knight, 1998, Kunkel et al ., 1999, Groisman et al ., 2004, 2012, Villarini et al ., 2011a, 2013). Recently, Villarini et al . (2013) found a large increase in the number of rainfall days exceeding the 95th percentile of the rainfall distribu- tion over the Upper Mississippi River Basin, and a much weaker signal in the Lower Mississippi River Basin. They also discussed these results in light of trends in temperature and increasing irrigation over the Great Plains. Over the central United States, Gro- isman et al . (2012) found a redistribution in the intense rain, with decreasing trends in the rain days within the 12.7 – 25.4 mm day -1 and an increase in the number of events above 76.2 mm day -1 . They also examined different possible mechanisms that could explain the observed increasing trends in intense rainfall. These results were based on analyses of the observational records. Are these trends bound to continue in the future? Here we address this question by using daily rainfall projections from 20 state-of-the-art global cli- mate models (GCMs) produced under the Fifth Cou- pled Model Intercomparison Project (CMIP5; Taylor et al ., 2012; Table I). Results will be based on the representative concentration pathways (RCPs) reflect- ing an 8.5 W m -2 radiative forcing change at the end of the 21st century (RCP 8.5) due to high greenhouse gas concentration level. 2. Data and methodology We use daily rainfall data from 20 CMIP5 GCMs (Table I) and focus on the 90th and 99th percentiles of the rainfall distribution to examine projected changes in the upper tail of the rainfall distribution. These models have different resolutions, ranging from 0.8 to 3.8 latitude degrees. All the analyses are performed at the 1.0-latitude degree and regridding is performed according to a bicubic interpolation. The study area is the central United States, defined as the region between 26 ◦ N and 50 ◦ N, and 83 ◦ W and 104 ◦ W. We focus on the ensemble-mean trend for the 90th and 99th percentiles of the rainfall distribution over two periods (2006–2045 and 2046–2085). More specifically, at each grid point we compute the values of these two percentiles for each year. We then compute the slope of the trend line for each of these times series and for each of the 20 models. Finally, we take the average of these 20 slopes as our representative slope values. As discussed in Tebaldi et al . (2006) and Orlowski and Seneviratne (2012), there is a generally weak agreement among models when working with extreme precipitation, with large inter-model variability. Resorting to the ensemble 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 Copyright  2013 Royal Meteorological Society