Use of Satellite Skin Temperatures to Improve Surface Evapotranspiration Performance in WRF

The physical atmosphere plays a critical role in air quality modeling performance. It is the purpose of paper to evaluate and improve the performance of a land surface model (Pleim –Xiu) used in the meteorological model (WRF) by the use of satellite skin temperatures to better specify surface evapotranspiration (ET). While considerable work has been done by the national community to develop improved land use classifications, land use classes themselves are not directly used in models. Rather, physical parameters such as heat capacity, thermal resistance, roughness, surface moisture availability, albedo etc. associated with a land use class are actually used in the land surface model. Many of the land use class associated parameters such as surface moisture availability are dynamic and ill-observed depending on antecedent precipitation and evaporation, soil transport, the phenological state of the vegetation, irrigation applications etc. Other parameters such as heat capacity, thermal resistance or deep soil temperature are not only difficult to observe they are often model heuristics unknowable a priori. This project will use satellite data skin temperature data to retrieve or adjust these critical land surface parameters.

Morning skin temperatures observed from geostationary satellites are used to adjust soil moisture in WRF within the Pleim-Xiu boundary layer scheme and alter surface ET. The Pleim-Xiu scheme has a two-stream (vegetation and bare soil) surface layer model. Adjustments in soil moisture are made based on differences between model skin temperatures and satellite observed skin temperatures. In addition to the adjustments in soil moisture, evening differences in model and satellite observed are used to adjust the thermal resistance of the surface.

Results are provided for the Discovery AQ special observation program in the Texas area for September 2013. Model performance with and without the satellite assimilation is provided. Performance statistics are provided over the domain using both satellite skin temperatures and NWS 2-m temperature and winds as performance metrics. Special data comparisons will be made for flux sites, wind profiler sites and aircraft measurements.