Warn on Forecast: From vision to reality!
This presentation will discuss the history and science advancements associated with NSSL's Warn on Forecast system (WoFS). Developed over the last 14 years, WoFS is an experimental relocatable, rapidly updating ensemble analysis and forecast system operating on a relocatable regional domain, providing new probabilistic storm-scale forecasts every thirty minutes to forecasters from SPC, WPC, and other NWS offices.
WoFS assimilates radar, satellite, and mesonet/surface observations every fifteen minutes over a period of 12-18 hours to predict high-impact weather including tornadoes, severe winds, and flash flooding.
WoFS origins can be traced back to the convective storm research by Ogura,
Wilhelmson, Klemp, Rotunno, and Weisman during the 1980's. This led to Lilly's seminal 1990 paper, "Numerical prediction of thunderstorms - has its time come" which spurred the founding of the Center for the Analysis and Prediction of Storms (CAPS) at the University of Oklahoma. In the early 2000's the application of Everson's 1994 ensemble Kalman filter method for geophysical flows to the convective-storm analyses problem by Zhang and Snyder is perhaps the critical advancement that helped create WoFS; this removed the need for an adjoint model in data assimilation.
Concurrently, the viability of convection-allowing models for next-day prediction of severe weather was demonstrated yearly in the NSSL/SPC's Spring Forecasting Experiment. In 2009 NOAA provided new funding to NSSL to embark on a research project to determine whether a numerical weather prediction system could extend average tornado-warning lead times to the public beyond the fifteen-minute limit from radar-based warnings. Over the next seven years, NSSL scientists researched probabilistic storm-scale prediction and subsequently developed an WRF-based EnKF system with a relocatable regional domain which could be executed rapidly enough to provide new guidance twice per hour.
Since 2017 WoFS has been extensively evaluated and used by forecasters in real time during the spring and summer (and nowwinter) severe weather seasons. WoFS is now scheduled to be independently run by the NWS in a new demonstration project starting in 2025.
I will present the new science associated with WoFS, some unique decisions associated with its development, and NSSL's plans for a next-generation WoFS and storm-scale prediction during the next few years.