Young Jean Choi - Forecast Research Lab, Meteorological Research Institute (METRI), Seoul, Korea

Soil moisture content is one of the most important land surface variables in land-atmosphere coupled model. McCumber and Pielke(l981) show that the effect of soil moisture on the surface heat fluxes is more important than that of surface albedo, soil texture, or soil type in the mesoscale numerical experiments. Using High Resolution Land Data Assimilation System (HRLDAS) to MMSILand Sufrace Model (LSM) coupled system, Chen (2004) shows that initialization of fine scale gradient of soil moisture is clearly important in the evolution of the convective initiation. The surface energy budget strongly affects both large and small scale circulation. Therefore in order to better simulate this effect, sophisticated pa.rameterization schemes of the biospheric and land surface processes are being incorporated into general circulation models, mesoscale model. In this study soil moisture initialization with uncoupled LSM was used. Observed rainfall was used to derive an off-line simulation of a LSM. The strength of this initialization method is the evolution of soil moisture does not suffer froin the model bias in surface forcing field. The stand alone LSM can adequately capture the evolution of soil moisture. Such initialization is obviously superior to imposing coarse resolution cliinatological soil water content over the domain. The purpose of this project is to develop a land data assimilation system that can produce soil moisture field for inesoscale NWP model.