Assimilation of Screen-Level Temperature and Humidity Observations into NOAA's Global NWP System

In contrast to all other major NWP centers, NOAA does not currently assimilate screen-level Temperature (T) or humidity (q) observations over land into our global NWP system, although these observations are assimilated into our regional forecast system. There are over 6000 stations collecting these observations globally, and they represent a significant source of observed information that we are not currently utilizing. This presentation will then review recent work at NOAA towards assimilating screen-level T and q over land into NOAA’s future NWP system (using a prototype of the GFSv17 model, which includes the new Noah-MP land model). In the examples mentioned above, the screen-level observations are used to update the model soil moisture and soil temperature, using relatively simple data assimilation methods. By contrast, the experiments conducted here use an Ensemble Kalman Filter (EnKF), which is considered the gold standard for land data assimilation. Using the EnKF also allows us to use the same data assimilation system to update the model’s land and atmospheric components, so that we can easily test the impact of assimilating the observations into each component. Based on these experiments, an initial approach to assimilating screen-level T and q into the land and/or atmosphere of NOAA’s future global NWP system will be outlined. Finally, NOAA’s atmospheric assimilation is a 4D Hybrid Ensemble-Variational Data Assimilation (4D EnVar), and the opportunities and challenges of expanding the EnKF-based screen-level assimilation to use the full 4DEnVar will also be briefly reviewed.