The Joint Effort for Data assimilation Integration (JEDI) effort will be the backbone for performing data assimilation across the UFS applications, and will be critical for realizing a unified data assimilation system for any coupled application of the UFS. In order to meet the goals for improved data assimilation, investment in continued JEDI development, transition, and testing needs to be expanded. Transitioning to JEDI will also accelerate innovations and science development across a whole range of data assimilation issues. The UFS R2O project ensures continued development in JEDI infrastructure, executing the transition plan from the current operational data assimilation infrastructure for atmospheric data assimilation, the Gridpoint Statistical Interpolation (GSI). An additional part of the infrastructure for realizing improved initial conditions and concomitant forecasts is related to the processing and utilization of observations. This project has significant emphasis on continued investment in the improved use of observations for realizing improved medium range forecast skill. This includes improved use of currently available observations as well as preparations for a new generation of observations that will be deployed within the next several years. Additional investment is needed in the pre-processing, quality control, and front-end handling of observations for use in assimilation. This project ensures alignment of the elements of the JEDI project (observation database and unified forward operator) that should be targeted for the re-engineering of the currently operational “observation pre-processing”.
It is also important to consider that the system that will be utilized for UFS applications will fundamentally be coupled to other non-atmospheric components, requiring the initialization of states for the ocean, sea ice, waves, land, and more. Therefore this project not only includes the development of data assimilation for marine and land components of the UFS, but also includes the development of a scientifically validated coupled data assimilation (CDA) capability. In particular, there is a significant effort underway to build out a JEDI-based marine assimilation system for a MOM6 (ocean)-CICE5/6 (sea-ice) application for the purposes of monitoring (e.g. to replace the legacy Global Ocean Data Assimilation System - GODAS). Additionally, this project will emphasize the usage of JEDI for building an initial capability for the initialization of soil moisture and snow within the land component of the UFS. These components will be used to develop a baseline weakly coupled data assimilation (WCDA) capability for the UFS. Since coupled reforecasts are a requirement for GEFS v13, a major component of this effort involves the production of a coupled UFS reanalysis and reforecast (R&R) product ahead of the GEFS v13 operational implementation.
In the immediate short term, the main forecast system targets for the efforts contained in this project are GFS v17, GEFS v13, and the SFS v1 thereafter. More specific sub-project tasks within the first two years include transition from GSI to JEDI for the atmosphere, replacement of legacy GODAS systems with JEDI-based SOCA (sea-ice, ocean and coupled assimilation) application including a fully completed ocean/sea-ice reanalysis, initial capability for soil moisture and snow assimilation, leveraging of IODA/UFO components for observation pre-processing re-engineering, and algorithm inter-comparison results from global NWP testing. As this project involves general infrastructure development activities for data assimilation, follow-on operational targets will include things such as the Hurricane Analysis and Forecast System (HAFS), Rapid Refresh Forecast System (RRFS), among others.