Use of a Regionally-Refined Model for Simulations of North American Climate

Tuesday, 16 December 2014
Stephen A Klein1, Larry K Berg2, James S Boyle1, Scott E Giangrande3, Wuyin Lin3, Richard B Neale4, Yun Qian5, Laura Riihimaki2, Qi Tang1 and Mark Taylor6, (1)Lawrence Livermore National Laboratory, Livermore, CA, United States, (2)Pacific Northwest National Lab, Richland, WA, United States, (3)Brookhaven National Laboratory, Upton, NY, United States, (4)NCAR, Boulder, CO, United States, (5)Pacific Northwest Natl Lab, Richland, WA, United States, (6)Sandia National Laboratories, Albuquerque, United States
While global climate models with horizontal resolutions approaching 10 kilometers are becoming more common, their computational expense is too great to permit routine testing with long simulations. Lower resolution global models with regional refinement are a more affordable modeling framework that is potentially useful for understanding the behavior of a high-resolution global model in the region with refinement. In this work, we test this notion by comparing simulations of the Community Atmosphere Model using the Spectral-Element dynamical core with 1/8° latitude-longitude regional refinement over North America to simulations of globally uniform 1/8° resolution model. The focus is on characteristics of warm season precipitation. Comparison to observations collected at the Department of Energy’s Atmospheric Radiation Measurement program site in Oklahoma is facilitated by simulations of the regionally-refined model that are nudged to analysis data over the coarse-resolution portion of the grid. This work is performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.