The Effect of Revised Soil and Vegetation Properties on Simulated GEOS-5 Land Surface Hydrology

The climatology and temporal dynamics of simulated land surface state variables are largely affected by the values used for various land surface parameters, particularly those defining soil and vegetation properties. To improve simulations of the land surface hydrological cycle and its subsequent effect on near-surface atmospheric variables, some input parameters and process parameterizations have been upgraded in the Goddard Earth Observing System 5 (GEOS-5) Catchment land surface model. More specifically, (i) the soil texture for the surface and subsurface has been replaced by a composite of data from the National Geophysical Data Center (NGDC), the Harmonized World Soil Database (HWSD), and the State Soil Geographic (STATSGO) project, with the inclusion of information on the organic material, (ii) the hydraulic parameters are derived from these textures using more continuous pedotransfer functions, (iii) the minimal soil depth is increased, (iv) parameters describing transpiration and bare soil evaporation resistance are modified, and (v) the Leaf Area Index climatology is refined. We discuss the impact of these changes on relevant state variables (e.g., soil moisture) and hydrological fluxes (e.g., runoff), quantifying improvement using watershed-averaged and in situ ground measurements, Soil Moisture Ocean Salinity (SMOS) satellite soil moisture retrievals and runoff measured in large basins across the United States.