Roles of atmospheric and land surface data in dynamic regional downscaling

[1] In studies dealing with the impact of land use changes on atmospheric processes, a key methodological step is the validation of simulated current conditions. However, regions lacking detailed atmospheric and land use data provide limited information with which to accurately generate control simulations. In this situation, the difference between baseline control simulations and different land use change simulations can be quite different owing to the quality of the atmospheric and land use data sets. Using multiple simulations at the Monteverde cloud forest region of Costa Rica as an example, we show that when a regional climate model is used to study the effect of land use change, it can produce distinctly different results at regional scales, depending on the amount of data available to run the climate simulations. We show that for the specific case of land use change impact studies, the simulation results are very sensitive to the prescribed atmospheric information (e.g., lateral boundary conditions) compared to the land use (surface boundary) information.