To achieve the goals set forth for the JCSDA of accelerating the use of satellite observations in operational NWP, the GEOS-5 atmospheric data assimilation system (ADAS) shares a common atmospheric analysis system with NCEP. In addition to facilitating NASA's contributions to the Joint Center for Satellite Data Assimilation (JCSDA), the USWRP and the capabilities of the NWS, this strategy provides a validated, state-of-the-art system for NASA investigations. The GMAO contributes to NWP through the ingest of new data types in combination with the application of assimilation system tools to evaluate existing observation suites, identify new observing systems that have the greatest potential for adding useful information to the observing system and, ultimately, contribute to new mission planning.
The current GEOS-5 ADAS is a 3DVar implementation using the Gridpoint Statistical Interpolation (GSI) scheme originally developed at NCEP. It has been used for MERRA and to investigate avenues for improving the use of clear and cloud-cleared AIRS radiances. We have implemented the assimilation of MLS temperature and ozone retrievals, and we have undertaken a comprehensive assessment of the impact of different sensors on forecast skill using adjoint tools. This system has also formed the basis for the rapid development of an OSSE capability with a simulation of all the observations used in the current GEOS-5 production system.
The next generation system, which is currently being tested, is a 4DVar system. It is being developed in a continued partnership with NCEP. The ultimate goal is a hybrid ensemble 4DVar system with the background error covariances estimated from the ensembles. This hybrid system is being developed as a collaboration between GMAO, NCEP, and NOAA/ESRL.