CISL Seminar: The Joint Effort for Data assimilation Integration (JEDI)

Abstract

The long term objective of the Joint Effort for Data assimilation Integration (JEDI) is to provide a unified data assimilation framework for research and operational use, for all components of the Earth system, and for different applications. The objectives are of reducing or avoiding redundant work between government agencies and within the community, and increasing efficiency of research and of the transition from development teams to operations.

In this project, the concept of forecast model is well separated from the observation handling with clear interfaces. The data assimilation algorithms are themselves separated from the model space and observation space components. This results in a data assimilation system that is generic with respect to the model it is interfaced with, including atmospheric models (UFS, MPAS, LFRic, NEPTUNE...), other components of the Earth system, coupled models, or AI/ML models. With this separation of concerns, it is easier for specialists of each area to contribute without unexpected consequences in other parts of the system, and to work in parallel on each component. For example, new observation types can be added easily using the generic infrastructure and be made available to all the partner agencies in a very short amount of time and with no duplication of effort. The abstraction of code away from the specific implementation details allows for a very flexible system and fast prototyping of innovative data assimilation science.

After several years of development the code is getting mature and is planned for operational implementation by the JCSDA partner agencies (NOAA, NASA, US Navy, US Air Force, UKMO). In this talk, an overview of the system and the current status of the implementation and testing for several applications will be presented and we will discuss directions for upcoming developments.