MetWatch Forecasters’ First Evaluations of NSSL’s Experimental Warn-on-Forecast System for Ensembles (NEWS-e)

Operational convection-allowing models provide day-1 forecasts (≥18 hrs) useful for the production of rainfall products by operational forecasters. Owing to advancements in high-resolution modeling and compute resource availability, the frequency of forecast runs has been increasing. The HRRR, for example, currently runs on an hourly cycle. However, the spin-up time for convection to develop in HRRR and the latency of these forecast runs diminishes the real-time usability of the first 1–2-hour forecasts. Given the dynamic nature of extreme rainfall events, more frequently-cycled, short-term forecasts have the potential to improve the forecast guidance available to users during rapidly evolving extreme rainfall and flash flood events.

With this need in mind, researchers at the National Severe Storms Lab have developed the NSSL Experimental Warn-on-Forecast System for ensembles (NEWS-e). NEWS-e is a prototype Warn-on-Forecast (WoF) system that provides real-time, short-term probabilistic forecast guidance for a variety of weather hazards, including extreme rainfall, within the watch-to-warning timeframe (0–6 hr) over a preselected subdomain. Unique to this system is frequent data assimilation cycling (every 15 min) and frequent forecast launches (e.g., every 30 min).

An important step in the development process is forecaster-in-the-loop studies. One such study took place in collaboration with MetWatch Desk forecasters at the Weather Prediction Center on days during May, June, and July 2018. In this study, NEWS-e probabilistic forecast guidance was accessible to MetWatch forecasters via a website daily during the afternoon and evening hours. Forecasters who used NEWS-e during operations were asked to complete a co-created, web-based survey available on the same website. The survey was designed to elucidate information such as: 1) the most frequently used forecast parameters, 2) how and in what cases NEWS-e contributed to forecasters’ decision making process, 3) forecasters’ perspectives on what constitutes a forecast rated as “poor”, “fair”, “good”, or “excellent” in terms of location, timing, intensity, and coverage, and 4) what improvements forecasters’ recommend to improve the usefulness of NEWS-e. This presentation will report findings from the MetWatch forecasters’ first evaluation of NEWS-e.