Update of Land Surface Category in the NCEP Data Assimilation System and Its Impact on Satellite Data Assimilation

Satellite observed brightness temperatures (Tb) in various spectral channels are assimilated through the JCSDA Community Radiative Transfer Model (CRTM) on the NCEP Gridpoint Statistical Interpolation (GSI). Over land, CRTM simulated satellite brightness temperatures at surface sensitive channels usually have large errors, either in infrared or in microwave, because of inaccurate surface emissivity calculation in the CRTM, resulting in rejection of many satellite data in the GSI.

Surface characteristics, which is a function of soil type and vegetation type, plays an important role in quantify the land emissivity. In this study, vegetation category is updated from the 1-degree 13-type SiB vegetation classification to the MODIS 1-km 20-type IGPB vegetation classification, and soil category is updated from the 1-degree 9-type Zolber soil dataset to the 1-km 19-type State Soil Geographic Database (STATSGO) soil dataset. A series of experiments were carried out to investigate impact on brightness temperature simulation in the GSI with the update of land surface category. Results from this investigation will be presented.