Select simulations from the NCAR Community Climate System Model (CCSM) version 3.0 run for the IPCC 4th assessment report are being dynamically downscaled using the WRF model to 30km resolution to provide more detailed estimates of future changes in climate in regions of the United States. This study examines warm-season precipitation from the downscaled simulations for the 1990s from the Climate of the 20th Century (20c3m) scenario and compare it against the 2090s from the SRES A2 scenario. Changes in frequency distribution and other characteristics of precipitation will be explored with a focus on shifts in the extremes.

Warm season deep convection is the main component of precipitation in many parts of the U.S., yet climate model simulations of warm-season, continental convection and associated precipitation do not compare well with observations. Much of the uncertainty involved in the prediction of precipitation under climate change scenarios is rooted in model resolution and the parameterizations used at those scales. At their coarse resolution, climate models have problems with topography, eddy processes, and sub-grid parameterizations, giving them little predictive ability at regional scales. Dynamically downscaling climate model output helps address some of these issues and makes estimation of changes in regional precipitation more feasible. Since current climate models predict that extreme precipitation will change more than the mean, and any change in the hydrologic cycle would have a great impact in many sectors, especially at the extremes, it is important to provide plausible scenarios at resolutions finer than those currently available from GCMs to assess potential impacts from any changes.