Recent Developments in Land Data Assimilation for Global NWP at NOAA.

This presentation will review the land data assimilation system planned for NOAA's upcoming global NWP system update, and also demonstrate the impact of the land data assimilation on the NWP forecast skill. Our global system update, called GFSv17, will include an upgraded snow analysis, and the introduction of our first soil moisture and soil temperature analysis. The snow analysis will be a JEDI-based 2DVar assimilation of station snow depth and remotely sensed snow cover, while the soil moisture and soil temperature analysis will be achieved by expanding the atmospheric control vector to include the soil states. This strongly coupled land/atmosphere data assimilation system will also benefit from the addition of the assimilation of screen-level temperature and humidity observations. The impact of the land data assimilation on the NWP forecast skill will be presented, based on a pair of winter- and summer-time forecast impact experiments.

Finally, the Hybrid 4DEnVar data assimilation system used in the atmosphere analysis includes a re-centering step, in which the ensemble mean analysis is re-centered on the deterministic analysis. Since the land is strongly non-linear, the adjustments made by re-centering the land states can have similar magnitude to the increments from the land DA system. Results from an additional set of forecast impact experiments will also then be presented, investigating the relative impact of recentering the ensemble land states.