Without a Warning: The Role of Broadcast Meteorologists in Shaping Tornado Risk Perception during an Undetected Tornado

On 7 August 2023, an EF-2 tornado touched down in west Knox County, Tennessee. The tornado touched down in a west Knoxville suburb at 2:17 pm LDT and traveled northeast for 3.8 miles, dissipating at 2:23 pm LDT. The tornado was unique in several ways: it was the first August tornado on record in the county, the first significant tornado in nearly 60 years in the county, and was a significant tornado that went undetected. The only other confirmed August tornado in East Tennessee history was four days prior. That EF-1 tornado in Roane County (20 minutes west of the EF-2) was also undetected. In both cases, there were surprisingly no injuries or fatalities, which leads to the questions: did the public perceive that tornadoes were possible and take appropriate precautions? Did days of preparation by broadcast news and the National Weather Service prevent injury?

While there was no tornado warning, the Storm Prediction Center (SPC) had issued a tornado watch for the area from 11:20 am–7:00 pm LDT. The tornado developed quickly from a bowing structure of storms that were under severe thunderstorms issued by the Morristown NWS Weather Forecasting Office (WFO) with warning text that included a “tornado possible” tag, which is used when a severe thunderstorm is especially capable of producing a tornado but other hazards are still considered a greater threat. Thus, there were multiple ways the NWS encouraged the media to highlight the tornado threat and urged the public to prepare for tornadoes, even if no tornado warning was ultimately issued.

Our interdisciplinary research team is completing a rapid response project examining how broadcasters in Knoxville interpreted tornado risk on 7 August and presented it to the public, and how this influenced public risk perception and behavioral response during an undetected tornado. We are using a mixed-methods approach, including a focus group of broadcast meteorologists and a survey of the public. In this talk, we specifically focus on the risk interpretation and communication by broadcast meteorologists at one television station, WVLT, and how such television broadcasts may have affected perceptions and behavior of viewers. We aim to fill research gaps related to how television meteorologists use specific NWS products, including the tornado possible tag, to understand and communicate risk; and how television broadcasts affect risk perception and behavior in the absence of a warning.

At WVLT, wall-to-wall coverage began at approximately 12:44 p.m. LDT. The team of three TV meteorologists routinely showed the Tornado Watch, explained the differences between a watch and warning, showcased storm relative velocity products, the power outage threat, and efforted to ‘jump’ between the multiple storms. When the Farragut tornado was on the ground, the team was talking about another tornado warning in Greene County. That Greene warning did not verify, despite the claims of a social media video. With so many active warnings, the “tornado possible” tag could have been overlooked, especially in the new NWSChat format. TV meteorologists are curious to learn: how can different types of warnings bias us to cover separate storm cells? If they note a potential brief touchdown via radar that is not under a tornado warning by the NWS, how assertive can they be?

Public information seeking, perception of tornado possibility, and behavioral response are being assessed using an IRB-approved rapid-response survey of Knox County residents over the age of 18. Upon closing of the survey in mid-September 2023, we will assess where the public got their information and how that impacted their perceptions and protective action decision making. In this talk, we will focus on those survey respondents who used television broadcasts as an information source, and determine how the communication of risk by television meteorologists affected their risk perception and behavior. By tracing the risk information from the NWS through media dissemination and to the end-user, we provide end-user feedback to the NWS and media on their specific risk communication products.