Quantifying the dominant modes of subgrid variability in land surface modeling using the mosaic approach

The Earth's land surface characteristics, such as vegetation, soils and topography, vary spatially and across a range of spatial scales. This heterogeneity has significant impacts on terrestrial energy and water budgets and on non-linear processes occurring at the land-atmosphere interface. One of the commonly used approaches to represent spatial heterogeneity in land surface models is to use grid cell aggregation methods by modeling the land surface as a mosaic consisting of homogenous elements. Each of these elements is created based on the distribution of land surface parameters within a grid cell. Most reported mosaic-tiling approaches, however, are focused on using a single factor such as vegetation as the basis of tiling. In this study, we present the results of using a mosaic approach where different modes of variability introduced by vegetation, soils and topography are simultaneously considered. The effectiveness of the tiling approaches is evaluated by comparing them to the explicit modeling of the land surface at a fine spatial resolution (e.g., ~1 km). The impacts of subgrid tiling approaches on both terrestrial energy and water cycle variables will be presented.