Description:

The Warn-on-Forecast (WoF) project fosters multidisciplinary research and development, and experimental application of new concepts in both testbed and real-world environments. The relocatable, rapid-data-assimilation, rapid-forecast-update, storm-scale ensemble system (WoFS) that embodies the project breaks new ground, and is one of the early maturing technologies that will support the Forecasting a Continuum of Environmental Threats (FACETs) paradigm for severe weather on the watch-to-warning time scale. The WoF concept of issuing warnings via ensemble-based prediction of individual thunderstorms represents a significant step forward. Another area WoFS has helped pioneer is the development of new and unique probabilistic products for forecasters to use to determine both the probability and intensity of convective weather hazards. Emphasis on probabilistic forecasts of specific hazards which are updated multiple times per hour makes WoFS well suited to assist the cultural and paradigm shift of severe storm science toward FACETs. Developing and operating WoFS as well as using it effectively in fast-paced severe weather environments will require frequent collaboration between researchers, forecasters (both government and private), developers, and social scientists.

Research-to-operations-to-research (R2O2R) loops already exist within the WoF project. The web-based product viewer has iteratively incorporated direct forecaster feedback obtained via the Hazardous Weather Testbed Spring Forecasting Experiment since 2017, yielding more efficient visualizations and user-controlled options. The number of National Weather Service (NWS) Forecast Offices presenting examples of real-world application of WoFS at conferences has grown each year. NWS national center and local office meteorologists participated in an experiment in 2021 to document visionary ideas and challenges for using WoFS in a collaborative forecast process; the results will move toward publication and conference presentation this year. A Google chat room internal to NWS that hosts real-time discussion of WoFS output between developers and forecasters has expanded to the lower 48 states this year. And demonstrations of what is possible through a new cloud-based, cb-WoFS, which can run multiple targeted domains, can help the NWS develop plans for how to run the relocatable WoFS domain for operations to help predict multiple threats.

WoFS developers at the National Severe Storms Laboratory will work with the NWS to draft a transition plan by fall of 2022. Many groups were already working on several items that will be needed for a successful transition, such as tailoring WoFS to the Unified Forecast System architecture, developing calibrated, post-processed probabilities of severe weather hazards from machine-learning models, and documenting WoFS strengths and weaknesses in contexts that speak to the variety of event-types for which this ensemble will be applied around the lower 48 states. The transition planning process will also likely reveal NWS needs in areas such as computing, verification, and training that will spur further collaborations between research groups. Warn-on-Forecast presents a wealth of ongoing R2O2R efforts that can focus the conversation on FACETs for near-term severe weather.