IMPACT OF USGS VEGETATION MAP ON GCM SIMULATIONS OVER THE UNITED-STATES

A global atmospheric general circulation model (GCM) coupled with a bi osphere model (SSiB, simplified simple biosphere model) was used to st udy the impact of vegetation on simulations over the United States dur ing summer. Ensembles of 90-d integrations were performed from early J une initial conditions with different vegetation maps and different bi ophysical characteristics. Monthly and seasonal mean differences among these ensembles were analyzed. Incorporation of a new SiB vegetation map produced from the latest available data by the U.S. Geological Sur vey's EROS (Earth Resources Observation Systems) Data Center has a sig nificant impact on monthly and seasonal simulations of evaporation, su rface air temperature, and precipitation over some regions of the Unit ed States. The impact is greater over the western half of the United S tates than over the eastern half, where moisture convergence prays a s tronger role in the hydrological cycle. Systematic errors in the model simulations appear to be related to the use of a single crop vegetati on type in SiB (simple biosphere model). Replacing the crops over the United States with broadleaf deciduous trees reduces the systematic er rors. It appears that the strong seasonality of the SiB crop vegetatio n type makes it an unsuitable representative for crops in general. The importance of vegetation specification in monthly and seasonal predic tions is emphasized.