Working at the Intersection of Research and Operations - Lessons learned and a future vision of community modeling from OAR’s Weather Program Office

Community-based numerical weather prediction has been the vision for NOAA for nearly a decade. Congress shares this vision and has instructed NOAA to prioritize open source development of its models in the Weather Research and Forecasting Innovation Act Reauthorization of 2018 and the LEGEND Act of 2023. Based on this guidance, NOAA has invested in simplifying its operational suite of models through the Unified Forecast System (UFS), as well as enabling the larger modeling community to work with the UFS through workshops, hackathons, online tutorials, portable and documented code, etc.

WPO has operated at the intersection of the research and operational modeling communities by virtue of where it sits as a program office in OAR, but also through management of Earth Prediction Innovation Center (EPIC). EPIC enables researchers to use and improve NOAA’s forecast models through advanced software engineering, community & user support, systems integration, and infrastructure development. WPO also supports community modeling through grants and cooperative agreements, including through a 2023 innovations for community modeling competition, a 2024 competition for fire weather modeling and climate testbeds, and a future 2025 competition that encompasses many of the programs in WPO including testbeds, air quality, and seasonal to subseasonal research. Furthermore, in 2024, WPO awarded nearly $7 million dollars for a data assimilation (DA) consortium of six universities using funds from the Inflation Reduction Act. This was achieved in collaboration with NOAA Laboratories and the NWS and NOAA-UKMET Transatlantic Data Science Academy. The DA Consortium stated goals are to support the training and development of the next generation of DA experts, advance DA science and innovation, and support a university team to work with NOAA to advance data assimilation for operational models.

Through this community modeling journey, we have learned many lessons and realized opportunities for improvement in the future. In many ways, NOAA is still figuring out how to be an effective contributor, enabler, and beneficiary of a community modeling construct for its operational models. These lessons learned, combined with the advent of AI data-driven modeling, necessitates NOAA and the community to take a step back and re-evaluate the future vision and implementation of community-based modeling. This presentation will discuss how WPO, sitting at the intersection of research and operations, can facilitate the collaborative development of a future vision of community-based modeling that achieves the objectives of the multiple stakeholders involved in the effort.