Land Surface Modeling: Using Sub-Basin Units Improves Hydrologic Simulations
Traditionally, land surface models used in Earth System Models (ESM) compute land surface processes on a structured grid that follows the globe’s latitudes and longitudes. Hydrologists, however, have more commonly employed watershed catchments, or sub-basins, as the computational units, because hydrological processes are strongly influenced by topography that in turn defines the sub-basin boundaries. A team of Department of Energy scientists at Pacific Northwest National Laboratory investigated the relative merits of grid- and sub-basin-based land surface modeling approaches for hydrologic simulations, with a focus on their ability to perform consistently across spatial resolutions in simulating runoff. They produced simulations in the grid- and sub-basin-based Community Land Model at 0.125, 0.25, 0.5, and 1 degree spatial resolutions over the U.S. Pacific Northwest. For the first time, the team examined the importance and relative advantages of using a different spatial structure for accurate modeling of hydrological processes (i.e., in the context of ESM applications). The significant scalability advantage for the sub-basin-based approach compared to the grid-based approach for runoff simulations affects other processes such as soil moisture, evaporative processes, and stream flow. The researchers found that the source of runoff scalability is related to the atmospheric forcing effects and land surface parameters of runoff generation, both of which are sensitive to surface elevation and more scalable in the sub-basin framework.
Tesfa, T. K., L. R. Leung, M. Huang, H. Li, N. Voisin, and M. S. Wigmosta. 2014. “Scalability of Grid- and Subbasin-Based Land Surface Modeling Frameworks for Hydrologic Simulations,” Journal of Geophysical Research: Atmospheres 19(6), 3166-84. DOI:10.1002/2013JD020493.