Developing and Testing Watch-Scale Forecast Products in the FACETs Paradigm

One of the challenges facing the FACETs continuum is bridging the gap between long lead time convective outlook probabilities and short lead time warning-scale probabilities. Previous work has shown that a majority of convective outlook day events occur in just 4-hours of the 24-hour day. This lead to the development and testing of a forecast product aimed at identifying when that 4-hour period would occur. This product, called isochrones, was verified and evaluated during the 2016 and 2017 Spring Forecasting Experiments in Norman, OK. Forecasters generally identified the start time of an event accurately, but often progressed events too slowly, forecasting areas downstream of initiation to be later than what observations showed. However, forecasters showed improvement in the second year of the experiment, likely due to better training and identifying the temporal bias from the year before.

The major concern expressed by forecasters in the 2016 and 2017 experiments was the utility of this product, especially with outside partners. Confusion over what the graphics were conveying and how long it would take for users to interpret them raised questions about how useful this product would be even if forecasters produced a perfectly accurate product. Due to these challenges, researchers re-invented a timing product related to isochrones, but made up of areas instead of contours. Forecasters and emergency managers in the 2018 HWT tested the Potential Severe Timing areas with promising results and feedback from both sides.