Earth Prediction Innovation Center

Executive Summary

The Earth Prediction Innovation Center (EPIC) will enable the most accurate and reliable operational numerical forecast model in the world (The Vision). To be the catalyst for community research and modeling advances that continually inform and accelerate advances in our nation’s operational forecast modeling systems (The Mission), EPIC is partnering with the community for the benefit of the nation (The Mantra).

Goal: Accelerate scientific research and modeling contributions through continuous and sustained community engagement to produce the most accurate and reliable operational modeling systems in the world.

1.0 Why is EPIC needed? As our Nation pursues comprehensive and accurate Earth system models, advances in research and modeling must be rapidly incorporated into operational modeling systems. Where it was once reasonable to provide forecasts that were accurate to 2 or 3 days out, it is now necessary to provide forecasts that are accurate out to 21 days or even out to months and seasons. This opportunity to provide the types of accurate forecasts that underpin decisions across sectors depends upon making massive advances in science and modeling that can be continually incorporated into operational modeling systems.

Historically, this has been accomplished through a mix of research and modeling at NOAA labs, cooperative institutes, academia, and in the private sector that filters through the organization to operational use. Constraints to that transition into operations have included the complexity of introducing the required innovations into the existing operational model suite; consistently applying resources from research/modeling and operations to get the job done; balancing research/modeling interests with operational priorities; and the daily operational pressures to produce forecasts crowding out resources for innovation. Over the last decades, each of these pressures have been eased by hosting testbeds, increasing co-development between research/modeling and operations, and clarifying research priorities. However, NOAA has significant opportunities to improve in each of these areas by transitioning to the EPIC strategic approach.

The EPIC strategic approach leverages these realities with an extramural, distributed center with some individuals co-located (e.g., as visiting scientists, modelers, or software engineers) with the National Centers for Environmental Prediction (NCEP). It will be accountable for continuous connections between research and operations to communicate, and align operational needs, innovation possibilities, and the programs that fund research and development; ongoing engagement between modeling and operations; and access to an integrated development environment that is platform-agnostic and co-locates the high-performance computing environment, storage and delivery of data with the latest codes, utilities, documentation, and user support. EPIC will accomplish this transformation by offering a cloud development environment; code repository; observations and tools; and community support and engagement that will, in turn:

● Phase One (near term, 1-2 years): Improve the Unified Forecast System (UFS), here defined as the UFS community that shares science components and software infrastructure for weather forecasting, by incorporating contributions from the external research and modeling community.

● Phase Two (mid term, 3-5 years): Extend infrastructure and user support for UFS to full earth-system coupling such as convective-allowing models (e.g., High Resolution Rapid Refresh) and fully coupled forecast systems (e.g., the Seasonal Forecast System, National Water Model, and Ocean Forecast Systems). Also expand the breadth of earth-system modeling, consistent with NOAA’s mission.