Sensitivity of Land Surface Simulations to the Distribution and Treatment of Vegetation Properties in GSWP-2

The Second Global Soil Wetness Project (GSWP-2) is an initiative to produce and evaluate 10-year simulations by a broad range of land surface models under controlled conditions. An essential component of GSWP-2 involves the production of a suite of sensitivity studies by each participating land surface scheme (LSS) where forcing data or boundary conditions are altered to test the response of the models to uncertainties in those parameters. In this study, two sensitivity experiments from several LSSs over the 10-year period have been analyzed. The first experiment investigates the sensitivity to the global specification of vegetation type where the University of Maryland (UMD) vegetation classification is used in place of the default IGBP-derived vegetation categories. A second study examines the sensitivity to choice of mean seasonal cycle versus time-varying vegetation parameters (LAI, FPAR, greenness fraction, roughness length, and displacement height). The impact of alternative surface vegetation data sets on model simulations of surface fluxes and state variables has been assessed. These studies will help us better understand the role of land cover change and interannual variations on the land surface in an uncoupled mode (without atmosphere feedback) and the potential importance of near-real-time global vegetation observations for weather and climate prediction and monitoring.