Abstract:
The Earth’s land surface is observed by various satellite-based sensors, but satellite observations are intermittent in space and time, mostly only available at a coarse resolution and prone to error. Land surface models simulate the land surface based on physical laws that can be solved continuously in time and space. However, these models are a great simplification of nature and are thus imperfect. By assimilating satellite-based observations into large-scale land surface models, improved land surface estimates can be expected. The land data assimilation system of NASA’s Global Modeling and Assimilation Office has been exercised with ensemble Kalman filter/smoother and rule-based techniques to assimilate various types of satellite observations for a large-scale estimation of soil moisture, temperature and snow. A short overview of this research will be presented, followed by an in depth discussion of the preparation for the operational Soil Moisture Active Passive (SMAP, to be launched in November 2014) level 4 root zone soil moisture product (L4_SM). This L4_SM product will assimilate L-band microwave radiances at 40 degree incidence angle from SMAP to update surface and root-zone soil moisture, as well as surface temperature, in the Catchment land surface model of the Goddard Earth Observing System Model (GEOS). To prepare for SMAP, the system is currently tested with multi-angular L-band microwave observations from the Soil Moisture Ocean Salinity (SMOS) mission, which was launched in 2009.