Enhancing NCEP-NAM Weather Forecasts via Assimilating Real-time Satellite-based Soil Moisture and Green Vegetation Fraction (Invited Presentation)

Accurate forecasts of temperature and precipitation from numerical weather prediction (NWP) models rely on the quality of the initialization of land surface state variables (e.g. soil moisture(SM)) and the representativeness of parameters that describe the current land surface (e.g. green vegetation fraction (GVF)). Real time satellite-based land surface products are capable of providing spatially continuous observations of surface parameters while accurately capturing the dynamics of surface conditions. Studies have shown the unique value of satellite-based SM retrievals and vegetation cover information and the feasibility of assimilating SM and vegetation dynamics products into the land surface models (LSMs) to improve the land-atmosphere water and energy exchange simulations. While most studies have focused on the assimilation of land surface data products into uncoupled LSMs, the potential impact of SM and GVF assimilation through coupled NWP-LSM modeling system may be more realistic. This study aims at assessing the impact of assimilating real-time satellite based SM retrievals and GVF on the weather forecasts of the NCEP North American Mesoscale Forecast System (NAM) NAM model. Specifically, we examine the assimilation of satellite based SM products from two independent sources. One is a thermal infrared (TIR) based SM product retrieved from operational GOES satellite using the Atmosphere Land Exchange Inverse (ALEXI) model, and the other is microwave (MW) based surface SM product generated from the Climate Change Initiative (CCI) of EUMETSAT. The impact of satellite based SM retrievals on model simulations is assessed by comparing forecasts variables with in-situ observations. On top of that, how the variations within these two SM data sets (MW vs. ALEXI) impact the weather forecasts differently is analyzed from spatial and temporal aspects. On another aspect, the current NCEP Noah LSM within NAM uses only a multiyear climatology of GVF although land-atmosphere interactions are well known to be sensitive to realistic vegetation status. Fang et al. (2014) compared Noah LSM SM estimates using either the multiyear climatology or real time GVF and found the later could improve Noah LSM performance. The impact of real-time GVF on the weather forecasts from NWP models is further investigated in this study. The effectiveness of assimilating real time satellite based GVF and SM observations is evaluated against both in situ soil moisture measurements and the standard NWS-NCEP NWP evaluation metrics. Results of this evaluation will be presented and path of transitioning this LIS-NAM SM and GVF assimilation system into NCEP operations will be discussed.