Integrating Recommendations for Shelter and Refuge Adequacy with a Tornado Watch Scale

Abstract

The 2011 U.S. tornado season resulted in 553 deaths, ranking behind only 1925 (794 fatalities) in the number of tornado-related fatalities in the United States (NWS Storm Prediction Center, 2012). Of the 553 fatalities, 6 were the result of “Weak” (EF0-EF1) tornadoes, 103 were the result of “Strong” (EF2-EF3) tornadoes, and 444 were the result of “Violent” (EF4-EF5) tornadoes (Storm Prediction Center, 2012). A potential latent factor in this unusually high number of fatalities was appropriate shelter/refuge, which has been poorly represented in the literature.

While interviewing Tuscaloosa residents following the April 27, 2011 Tuscaloosa, Alabama EF4 tornado, it was found that many residents were unsure of where to shelter in the event of another tornado because the areas they understood to be safe, recommended shelter (interior room, lowest level of a sturdy building) were destroyed during the tornado. This need for a clear definition of tornado refuge adequacy led to an effort to develop a way of determining what types of structures would provide adequate shelter during a tornado of a given intensity.

Data on the structural strength of most typical tornado refuge locations in regard to wind resistance can be found in the Enhanced Fujita (EF) Scale (McDonald, 2004). The EF scale provides a Lower Bound (LB), Expected (EXP), and Upper Bound (UB) wind speed corresponding to different levels of damage for 28 Damage Indicators (DI) that may possibly be used as refuge locations to account for structures that fall below, meet, or exceed code requirements, respectively. Using the LB wind speed estimate data for these 28 DIs, along with locations such as FEMA 320 or FEMA 361 storm shelters, partially-underground or fully-underground basements, and vehicles, a Tornado Safety Rubric (TSR) was developed to show the expected adequacy of each, based on the safety of someone sheltering inside during tornadoes of different intensities.

The TSR was then incorporated into a hierarchical Tornado Watch Scale based on the likelihood of weak, strong, or violent tornadoes. The TSR contains categories from Level 0 to Level 5 for use in communicating the expected intensity of tornadoes for an area on a given day. Because this is to be used as a safety recommendation, the levels correspond to a generous estimate of the maximum expected tornado intensity and include shelter/refuge categories of “Adequate”, “Questionable”, or “Inadequate” which encompass a full range of shelter/refuge locations. Audio recordings similar to those used in current NOAA weather radio communications were then developed to correspond to each TSR intensity level. These audio recordings representing the TSR and recordings representing a current NWS Tornado Watch and NWS PDS Tornado Watch were then used in interviews of Alabama residents to determine their comprehension of the information contained in each communication, what actions they would take when presented with each communication, where their refuge location would be if presented with a tornado warning while under each type of tornado watch, how they perceived the risk associated with each statement, and their preference between the current NWS Tornado Watch communication system, and the proposed multi-level Tornado Watch Scale.

Preliminary results from interview participants indicate a strong preference for the proposed Tornado Watch Scale compared to the current NWS Tornado Watch communications. Most participants report the reason for this preference is the additional shelter/refuge recommendation information contained within the TWS communication. Initial results also show the TWS stimulates participants to choose more adequate shelter when presented with more information on the potential intensity of an event and the appropriate actions to be taken during such an event.

A recommendation is made for the Tornado Watch Scale to be considered as an experimental product to be used operationally by the SPC, local NWS offices, or broadcast meteorologists. Convective outlooks issued by the SPC currently provide an accurate estimate of potential tornado intensity, but the use of these products is uncommon by the general public. It is hoped that this research will be adopted for use in conjunction with those products and disseminated to the public as a type of revised tornado watch communication tool meant to provide the best potential forecast estimate of maximum expected tornado intensity combined with adequate shelter/refuge recommendations.