Land surface models are used to represent terrestrial processes that shape global climate; examples of these processes include evaporation, plant water use, and photosynthesis. While much progress has been made to improve and refine these models, some hydrological processes are not well captured, which hinders our ability to understand land-atmosphere interactions and ultimately to predict impacts of climate change on water resources. The inadequate representation of evapotranspiration may partly explain why global climate models do not match observed precipitation patterns. Multiple factors contribute to this problem. In moist tropical regions, high humidity, leaf wetness, and cloud cover combine to suppress forest water use and possibly reduce forest growth in ways that are poorly understood. Mountainous areas pose additional difficulties, as standard modeling and measurement techniques are not readily applied in rough terrain.
The overall goal of this project is to improve the modeling of fluxes of water vapor and carbon dioxide to and from tropical forests. This goal will be achieved through a combined program of field-based measurements in a mountainous tropical forest in Costa Rica and regional scale modeling of land surface fluxes in the Neotropic ecozone of South America, Central America, and the Caribbean. Specifically, the project will: 1) collect targeted hydrological and meteorological measurements along in-canopy and above-canopy profiles at locations throughout a mountainous forest watershed at the Texas A&M Soltis Center; 2) develop a new conceptual framework for modeling wet canopy processes based on the new dataset; 3) appropriately revise the Community Land Model (CLM) to improve its estimates of evapotranspiration in tropical forests; and 4) model tropical forests and their interactions with rainfall using the improved CLM coupled with the Weather Research and Forecasting (WRF) model capable of resolving processes in mountainous forests.