The assimilation of layer precipitable water and the impacts on weather forecasting in a regional NWP model

The quality of a humidity analysis directly impacts severe storm analysis and forecasts. With high temporal and spatial resolution, the humidity information from geostationary satellite can improve regional/storm scale initialization through data assimilation. The Advanced Baseline Imager (ABI) (Schmit et al. 2005) from GOES-R can provide atmospheric water vapor with three water vapor absorption spectral bands during both day and night. The Advanced Himawari Imager (AHI) on Himawari-8 used operationally has almost the same features (spectral, spatial and temporal) as ABI in infrared bands. The GOES-R ABI LAP (Legacy Atmospheric Profile) algorithm can be applied on Himawari and provide the atmospheric water vapor information. The layer precipitable water (LPW) products at three sigma layers (0.3-0.7, 0.7-0.9, and 0.9-1) can be retrieved using the ABI for GOES Sounder data and AHI for Himawari Sounder data. The Weather Research and Forecasting (WRF) is used as n the forecast model and the DTC-GSI 3D-var is used as assimilation system. A forward operator is developed and integrated into GSI for assimilating LPW. Focuses are on how to better assimilate the high temporal moisture information, including cycling assimilation scheme, bias adjustment, observation error setting, use of information in cloudy region, radiance assimilation versus retrieval assimilation, etc. To verify the impacts of assimilating LPW data from GOES Sounder, the forecasted accumulated precipitation, radar reflectivity, and temperature and moisture profiles are compared with the real measurements. The 24 hour, 6 hour and 1 hour accumulated precipitation from forecasts are verified against the NCEP Stage IV analysis precipitation. The temperature and moisture profiles are verified with the radiosondes profiles. Both the frequency bias and the equitable threat score (ETS) are calculated to show the impacts of LPW and radiances on storm precipitation forecasts over CONUS. Typhoon Soudelor (2015) is selected to show the impacts of assimilating LPW data from Himawari, the improvement of typhoon track and intensity is discussed in this study.