Application of JCSDA JEDI in Observing System Simulation Experiments (OSSE) for Improved UFS Forecasts

This study harnesses Joint Effort for Data assimilation Integration (JEDI)'s data assimilation (DA) system to explore its potential in Observing System Simulation Experiments (OSSE). Using the Unified Forecast System (UFS) 6-hour forecast data spanning from February 1-29, 2020, we have trained a static background error covariance deploying the System-Agnostic Background Error Representation (SABER) component within JEDI. This forms the foundation for a 3D-Var DA system equipped with the trained static background error covariance.

In preparing observation data, the Interface for Observation Data Access (IODA) component inside JEDI has been employed. Through this, meteorological data, initially in Binary Universal Form for Representation (BUFR), was efficiently converted into ioda-compatible netcdf format. The current set of assimilated observations encompasses data from conventional measurements, Advanced Technology Microwave Sounder (ATMS), Advanced Microwave Sounding Unit-A (AMSU-A), Global Precipitation Measurement Microwave Imager (GMI), Microwave Humidity Sounder (MHS), Special Sensor Microwave Imager / Sounder (SSMIS), and the Tomorrow Microwave Sounder (TMS).

A pivotal component of this study will be comparing UFS Forecasts initialized from analyses both without and with TMS data assimilation. Preliminary results suggest notable forecast differences with assimilated TMS data. Further assessment will provide insights into the potential advantages and challenges of integrating TMS data in the UFS forecast initiation process.

The outcomes of this research could have significant implications for improving weather prediction accuracy and highlight the potential of TMS measurements and JEDI's data assimilation system in optimizing meteorological forecasts.