TY - JOUR
T1 - Southeastern U.S. Residents’ Perceptions and Responses to Evolving Probabilistic Tornado Forecasts and Warnings
AU - Wirz, Christopher D.
AU - Demuth, Julie L.
AU - Prestley, Robert
AU - Morss, Rebecca
AU - Joslyn, Susan L.
AU - Qin, Chao
AU - Savelli, Sonia
N1 - Publisher Copyright:
© 2025 American Meteorological Society.
PY - 2025/11
Y1 - 2025/11
N2 - In this study, we examine how individuals perceive and respond to probabilistic tornado information that evolves across different space and time scales. We conducted 36 structured in-person interviews in July of 2024 with residents of the southeastern United States to examine (i) probabilistic tornado forecasts that evolve from a 2-h “Warn-on-Forecast” System (“WoFS”) lead time to a probabilistic tornado warning and (ii) evolving tornado warnings that update on the order of minutes, with both deterministic and probabilistic conditions. We found notable variability in how people assess evolving tornado forecast information across information types, formats, and locations. Despite the novelty of the WoFS graphic, respondents generally understood what was being communicated. Many participants drew comparisons with a tornado watch and indicated that they would respond in similar ways. As the presented tornado forecast information evolved from the probabilistic WoFS graphic to a probabilistic tornado warning, participants also evolved from saying they would not immediately worry and would attend to future forecasts to indicating they would more seriously consider the risk to themselves. This evolution seemed to be a function of the change in spatial and temporal scales. We also observed some different interpretations of risk based on the locations respondents were in for the evolving warnings and based on whether the warning was probabilistic or deterministic. Overall, our findings demonstrate how each type and format of tornado forecast information has strengths and weaknesses that result in trade-offs among different ways of communicating.
AB - In this study, we examine how individuals perceive and respond to probabilistic tornado information that evolves across different space and time scales. We conducted 36 structured in-person interviews in July of 2024 with residents of the southeastern United States to examine (i) probabilistic tornado forecasts that evolve from a 2-h “Warn-on-Forecast” System (“WoFS”) lead time to a probabilistic tornado warning and (ii) evolving tornado warnings that update on the order of minutes, with both deterministic and probabilistic conditions. We found notable variability in how people assess evolving tornado forecast information across information types, formats, and locations. Despite the novelty of the WoFS graphic, respondents generally understood what was being communicated. Many participants drew comparisons with a tornado watch and indicated that they would respond in similar ways. As the presented tornado forecast information evolved from the probabilistic WoFS graphic to a probabilistic tornado warning, participants also evolved from saying they would not immediately worry and would attend to future forecasts to indicating they would more seriously consider the risk to themselves. This evolution seemed to be a function of the change in spatial and temporal scales. We also observed some different interpretations of risk based on the locations respondents were in for the evolving warnings and based on whether the warning was probabilistic or deterministic. Overall, our findings demonstrate how each type and format of tornado forecast information has strengths and weaknesses that result in trade-offs among different ways of communicating.
KW - Communications/decision making
KW - Probability forecasts/models/distribution
KW - Social Science
KW - Tornadoes
UR - https://www.scopus.com/pages/publications/105022428699
U2 - 10.1175/WAF-D-25-0086.1
DO - 10.1175/WAF-D-25-0086.1
M3 - Article
AN - SCOPUS:105022428699
SN - 0882-8156
VL - 40
SP - 2391
EP - 2407
JO - Weather and Forecasting
JF - Weather and Forecasting
IS - 11
ER -