CMIP5 GCM Alaskan Extreme Precipitation Events and Their Physical Processes Analyzed Using Self-Organizing Maps

We analyze physical processes leading to extreme daily precipitation events in Alaska applying Self-Organizing Maps (SOMs) to NCDC precipitation, ERA-Interim reanalysis and CMIP5 GCMs. We focus on widespread extreme events, defined as the top 0.01% of daily precipitation occurring on at least 20 grid points on the same day. The SOMs technique allows for the identification of large-scale circulations conducive to extreme events. Preliminary results show that the models do not replicate the highest precipitation intensities contained in observations. However, the models' widespread extremes do replicate the observed physical behavior in that extreme events are generally orographically forced in conjunction with a slow moving 500 hPa trough west of Alaska.