Continental-Scale Convection-Permitting Regional Climate Modeling

Abstract:

Convection-permitting regional climate models (CPCMs) have proven to be useful for down scaling large-scale climate information to regional and local scales. They add value to the representation of impact relevant parameters such as near surface temperature, precipitation, and the representation of extremes by improving local scale processes such as soil atmosphere interactions, snowpack dynamics, or the representation of deep convection. Due to their high computational costs most CPCM simulations have been restricted to small domains on the order of a few 100 km. On such small domains CPCMs might not reach their full potential because they are restricted by the lateral boundary forcing and may not be able to spin up properly.

In this study we investigate the ability of a continental scale CPCM to simulate climate conditions in the Contiguous United States within the period October 2000 to December 2010. We downscale ERA-Interim reanalysis data to a horizontal grid spacing of 4 km with the Weather Research and Forecasting (WRF) Model that allows an explicit treatment of deep convection. The model performance is analyzed in different synoptic-scale weather regimes, which enables a process-oriented evaluation. The significance of model biases in simulated precipitation and temperature is investigated by including observational uncertainties in the analysis. Significant biases are further investigated and possible error sources are discussed. The goal of this study is to provide a benchmark on the state-of-the-art convection-permitting regional climate modeling and to give guidance for future model development.