This study uses the dry season NDVI of 1982, to 2000 derived from AVHRR to identify regions of substantial deforestation in Central America. High-resolution nested grid simulations over the identified regions of deforestation using the Colorado State University Regional Atmospheric Modeling System (RAMS) are used to examine cloud formation under three scenarios: 1. region completely forested; 2. current land use conditions; and 3. deforested conditions. In the deforestation scenario the locations of the proposed Mesoamerican Biological Corridor are kept forested with pasture conditions outside the protected areas. These simulations will help identify locations where high deforestation could make the proposed corridor climatically unstable. The study also uses MODIS and GOES satellite data to understand how land use currently influences cloud formation and their properties. Preliminary satellite data analyses show that for the dry season month of March cumulus cloud frequency is nearly one and a half times higher over forests compared to grasslands irrespective of elevation. Cumulus clouds over forests are also more likely to be drizzling than over grasslands. Model simulations show that with pastures surrounding protected forests and corridors, several locations of the corridor will have suppressed cloud formation. Further simulations are being conducted and results from statistical analysis of a set of simulations and comparison against satellite observations will be presented.