Quantifying and Visualizing Severe Thunderstorm Motion Uncertainty for Improved Decision Support

Abstract Forecasts and observations of thunderstorm motion are key components of severe thunderstorm communication and decision support for preparedness and protective behaviors, as all National Weather Service (NWS) Storm Prediction Center (SPC) convective watches and NWS warnings contain information about storm translational speed and direction of motion. The goal of this research was to quantify uncertainty in thunderstorm motion using the information contained in NWS watch and warning products. Analyzing severe thunderstorm and tornado watches and warnings from 2008 to 2020, we matched and compared forecasted storm motions given in NWS warnings with those given in SPC watches. Translational speed deviations between watches and warnings exhibited symmetry, with about one-third of warning speeds falling within ±2.5 m s −1 of forecasts. Directional deviations were also symmetrical, with one-third of warning directions within ±9° of forecasts. However, when stratified to account for translational speed, the comparisons revealed greater directional uncertainty for slower-moving storms. Temporal analysis revealed faster and rightward translations in warnings during the latter half of watch durations. Using the distributions of speed and directional deviations, we developed new visualizations to represent both mean storm motion vectors and associated uncertainty sectors, aiding the interpretation of motion uncertainties. Quantifiable storm motion uncertainty also can inform estimated ranges of storm arrival times, potentially enhancing NWS communication and decision support with key partners and a range of publics for safer outcomes during severe weather events. Significance Statement Severe thunderstorms are a frequent phenomenon in the United States, threatening lives and property. The goal of this research was to expand on deterministic forecasts of thunderstorm motion by quantifying uncertainty for translational speed and direction of movement. Using the results of the uncertainty analyses, we visualized a forecast motion vector together with a motion sector that showed how much variability in motion is likely based on past forecasts in NWS watch and warning products. The results demonstrated how the motion uncertainty sectors varied according to the forecasted speed, with greater directional uncertainty at slower speeds and reduced directional uncertainty at faster speeds. Findings from this study can improve forecast communication and decision support for stakeholders who use NWS convective watch and warning products.