Impact Study of AMSR2 Soil Moisture Product in the NCEP Global Forecast System

A global soil moisture environmental data record (EDR) is generated by NOAA NESDIS from the advance microwave scanning radiometer (AMSR2) on board of JAXA's GCOM-W1 satellite launched in May 2012. AMSR2 has very similar characteristics to the AMSR-E on NASA EOS Aqua satellite except that AMSR2 has two subbands of the C-band microwave and larger satellite antenna. The two subbands provide the possibility of mitigating the radio frequency interference (RFI) on C-band observations. Larger antenna improves the spatial resolutions of observations slightly. Consequently the soil moisture data product from AMSR2 may have better quality than that from the previous AMSR-E. In this paper, the potential impact of the AMSR2 soil moisture data product on numerical weather forecast is assessed using the NCEP Global Forecast System (GFS) in which the Ensemble Kalman Filter (EnKF) data assimilation algorithm has been implemented. The AMSR2 soil moisture retrievals are first evaluated against in situ soil moistrue measurements and their biases from the Noah land surface model simulations in GFS are corrected before assimilating them into the model. A sequential soil moisture data assimilation experiment with full cycle runs of the EnKF-GFS and NCEP Gridpoint Statistical Interpolation (GSI) analysis system were performed. Results from this investigation were analyzed through the verification of GFS forecasts against surface and sounding observations of temperature, humidity and wind as well as precipitation forecasts.