Advances in Coupling Regional Circulation/Land Surface Models and Dynamic Crop Models

Regional circulation models (RCM) coupled to land surface models (LSM), have a static representation of agricultural surfaces. This type of surface, which is around 30% of the total area in the southeast USA, is just characterized by monthly tables containing a set of optic and topographic characteristics without making differences between crops and its different management (i.e. a field of carrots has the same energy and water fluxes than a field of sugar cane). Agriculture is responsible for the largest man-made inter-seasonal and inter-annual variability land use change. The regional aggregated effect of this man-made variability has an un-explored impact on the regional climate, which can be analyzed only if a dynamic approach of agricultural lands is incorporated to the RCM/LSM models. In this research we coupled the Florida State University/Center for Ocean-Atmospheric Prediction Studies (FSU/COAPS) Regional Spectral Model already coupled to the National Center for Atmospheric Research (NCAR) community Land Model (CLM2) with the dynamic crop models from the Crop System Models – Decision Support System for Agrotechnology Transfer (CSM-DSSAT) family models. The objectives were (i) to better represent the effects of land use inter-annual variability in RCM/LSM models, (ii) to increase RCM performance replacing the current empirical agricultural parameterization by a new dynamical one, (iii) to analyze the bidirectional relationships between regional climate and agriculture, and (iv) to analyze scenarios of land use change. After coupling the models, a series of simulation experiments were conducted in order to measure improvements in the model retrospective forecasts' skill. The model was also used to simulate projections of land-cover changes resulting from transforming forest and agricultural lands to land development for housing in the SE-USA. Significant differences were found between the simulated regional climate patterns under different land development scenarios, differences that would potentially influence population requirements of water and energy.