Landscape Characterization for the Identification of Environmental Gradients in the Tropics

Identification of spatial and temporal gradients in the environment is a first step in modeling environmental processes at all scales. Understanding these environmental gradients helps us improve modeling efforts by clarifying the relevant suites of input data and highlighting ideal regions in which to perform experiments in modeling applications. Landscape classification is one method for identifying gradients in time and space. Delineating patterns in precipitation, topography, soil chemistry, solar radiation, and similar variables has proven to be useful for understanding ecological processes and improving model performance. The work presented here utilizes multivariate spatiotemporal clustering analysis for quantitative landscape characterization in the tropics. Modeled climatic variables output from the Community Land Model (CLM) are used with additional observationally-based data layers to better understand the range of tropical ecosystem function and pinpoint poorly understood or under-sampled environments. Results offer methods for the upscaling of available data in environmental modeling and highlight the application of these data in identifying regions appropriate for modeling experimentation.