Relationships Between Storm Mode, Tornado Warning Outcomes, and Societal Vulnerability

This study provides a cross-disciplinary perspective in severe weather research by examining the relationships between tornadic storm mode, tornado warning outcomes, and social vulnerability. These three variables were analyzed within eight United States (U.S.) regions and 79 National Weather Service (NWS) county warning areas (CWAs). Specifically, radar data, tornado warnings, and tornado reports were gathered from 2007 to 2017 to assess spatiotemporal patterns in storm mode types, false alarm warnings, lead time warnings, and missed events. Tornadic storms were classified into five storm mode categories (i.e., discrete, linear, multicell, tropical, or disorganized) using machine learning methods. Results show that regions and CWAs have varying types and degrees of storm modes, warning outcomes, and vulnerabilities, suggesting the need for localized resilience solutions. The Southeast U.S. has the highest tornado-society impact potential given its elevated social vulnerability, more frequent linear storm modes, greater odds of winter (December, January, February) storms, and greater nocturnal tornado risk. This research provides a comprehensive understanding of tornado disaster risk, especially at the local scale. NWS weather forecast offices (WFOs), emergency managers, and policymakers can use the information provided in this research to target vulnerable groups for community preparedness, mitigation, and resilience-building strategies for tornado hazards.