The Role of the CLM2 in Seasonal Dynamical Downscaling for Crop Model Application

An advanced land model (the National Center for Atmospheric Research Community Land Model, NCAR CLM2) is coupled to the Florida State University (FSU) regional spectral model to improve seasonal surface climate outlooks in a very high spatial and temporal resolution and examine its potential for crop yield estimation. The regional model is placed over the southeast United States and run at 20 km resolution, roughly resolving the county level. Warm-season (March-September) simulations from the regional model coupled to the CLM2 are compared to those from the model with a simple land surface scheme (i.e., the original FSU model). In this comparison, two convective schemes are also used to evaluate their roles in simulating seasonal climate, primarily for rainfall. It is shown that the inclusion of the CLM2 produces consistently better seasonal climate scenarios of surface maximum and minimum temperatures, precipitation, and shortwave radiation, hence provides superior inputs into a site-based crop model to determine crop yields. The FSU regional model with the CLM2 exhibits some capability in the simulation of peanut (Arachis hypogaea L.) yields, depending upon the convective scheme employed and the site selected.