Understanding the extent of future climate change is highly dependent on the use of climate models to assess the degree and rate of change. More recently, dynamic downscaling has been employed with the use of regional climate models (RCMs) as an alternative to statistical downscaling. However, while RCMs provide a much finer resolution, they have still been shown to exhibit bias within their simulations. In order to understand the biases and uncertainties that exist within the RCMs, simulations of current climate must first be evaluated using past climatic data. In this study, a comparative analysis examining mean surface-air temperature and daily accumulated precipitation differences between RCM projections from the North American Regional Climate Change Assessment Program (NARCCAP) and reanalysis data from the North American Regional Reanalysis (NARR) is performed for the contiguous United States (CONUS). NCEP-driven RCM runs from 1981-2000 were validated to discern the biases that exist solely due to the RCM. To evaluate RCM simulations at a regional level, the domain is then further divided into 5 sub-regions: Northeast (NE), Southeast (SE), Northwest (NW), Southwest (SW) and the Great Lakes Region (GLR). The Great Lakes Region was chosen in addition to the four main sub-regions as future climate change is expected to impact future water resources and agricultural production within this region. Results indicate that RCM performance varies from region to region and season to season.
Read
