Evaluating the role of vegetation dynamics and complex topography on land surface fluxes using a tree-shrub-grass competition model in a semiarid landscape

The composition and structure of semiarid ecosystems are governed by the availability of water and energy and their variation in space and time. By modulating energy and water distribution, geomorphic structure of landscapes creates niches of ecosystem productivity, which often lead to differences in vegetation functional types and species composition. In central New Mexico, ecosystem patterns clearly reflect such controls. In desert elevations, the north facing slopes typically host tree-grass coexistence while south facing slopes are primarily shrub dominated. In this study we present an ecohydrological model of tree-shrub-grass competition driven by rainfall and solar radiation and some preliminary model results. The model is first used to examine the observed vegetation patterns, their dynamic response to climate fluctuations, and associated latent heat fluxes in two basins in central New Mexico at a fine spatial resolution representative of individual tree canopies. Second, model outputs of latent heat at fine spatial scales are aggregated to larger scales typically used in regional land surface models, and the impacts of vegetation dynamics in predicting land surface fluxes are discussed in relation to spatial and temporal scales of model predictions. Our preliminary model results underscore the importance of rainfall variability and suggest the need for a dynamic representation of vegetation in land surface models for climate change predictions over decadal time scales in semiarid regions.