Areas that have a high incidence of these storms share similar geographic and weather-related features that will be reviewed later. Worldwide, tornadoes can occur on any day and at any time. Springtime and the hours between 4:00 PM and 7:00 PM are the most prevalent times, with a peak time of 5:00 PM.3
There are a number of variables that determine the extent of physical damage, injury, and death. These include storm strength, population demographics in the storms track, time on the ground, structure design, and advanced warning times. The annual number of tornadoes worldwide is unclear. The United States has the largest number, with averages around 1,000 per year, while Canada comes second with approximately 80 annually.4 Although Bangladesh has fewer storms, it has the highest total mortality. This is due to the high population density, poor quality of construction, and other factors.4 The most deadly tornadoes by numbers of fatalities and continent are listed in Figure 37.2.5
Current State of the Art
In the framework of disaster management, the subcategories of mitigation, preparedness, response, and recovery will be considered in the context of tornado threats.
Mitigation: Activities that need to be performed before a tornado event include a broad range of categories. In order to reduce injury, death, and adverse health events, advance notification of potential storm risk and warning of imminent threats are paramount. Governmental meteorological agencies typically direct these activities. “Lead time,” discussed in more detail later in the chapter, is the time interval from issuing a tornado warning to tornado impact. Some U.S. severe storm experts believe the National Weather Service (NWS) has reached the optimum warning lead time with current technology. They promote the concept that risk will be reduced by public education directed at appropriate protective actions once a warning has been issued, rather than increasing the lead time.6 Countries with elevated risk for tornadoes have variable degrees of sophistication in detection and warning capabilities.
Building construction practices, which are also inconsistent worldwide, may have an impact on injury and death. Retrofitting homes with safe rooms or shelters has been recommended, as well as maintaining adequate shelters for public gathering places and hospitals.
Preparedness: Routine testing of warning methods (e.g., sirens, public broadcasts, cell phone notifications, other media) should be conducted regularly to check for and address system inadequacies. Testing also provides information to the general population regarding warning methods. Individuals must engage in personal preparedness to understand the protective measures necessary when public health officials issue a warning.
Response: Health systems, public safety, and other response agencies should have action plans in place, practiced and updated as recommended by after-action reviews. In addition, individual behavior can have a major effect on health outcome. The psychological/sociological aspect of an individual’s response to warning will be reviewed later in the chapter.
Recovery: Although the impact phase of a tornado has obvious health risks, the post-impact phase has its own set of challenges. During clean-up, the risk of injury from sharp objects, trips and falls, being struck by falling debris, and electrocution from power lines mistakenly thought to be without electricity are potential health risks.
Both event and post-event wound contamination and death from opportunistic infectious agents are concerns that require close monitoring in susceptible populations.7
Tornado Science
Tornado Rating Scales
The Enhanced Fujita scale (EF scale) rates the strength of tornadoes in the United States based on the damage they cause. Implemented in place of the Fujita scale introduced in 1971 by Ted Fujita, it became operational in 2007. Canada began using this scale in 2013.8 Great Britain uses a wind speed scale termed TORRO for tornado rating.9
The EF scale is based on U.S.-specific construction practices and its application in other countries may be inexact for rating tornado strength.10 Figure 37.3 illustrates structural damage as related to wind speed.
Forecasting Thunderstorms
Meteorological science is constantly improving forecasting ability for severe thunderstorms, from which tornadoes are spawned. The Storm Prediction Center (SPC) of the U.S. National Oceanic and Atmospheric Administration (NOAA) produces severe weather outlook forecasts.11 These incorporate modeling of upcoming atmospheric conditions and compare them to historic weather events. Using probability predictions, the SPC produces data for the current day with projections for the next 8 days.11
Worldwide locations with tornado occurrences share similar atmospheric features. Jet stream–driven cold dry air mass travels above a warm moist air mass that creates vertical turbulence and thunderstorm development. In the United States, topography associated with frequent storms is in the Central Midwest and Southeast, as cool dry air mass traveling from the Rocky Mountain range interacts with warm air from the Gulf of Mexico.4
In Bangladesh, the cool air mass originates from the Himalayas and the warm air mass from the Bay of Bengal. This results in most storms occurring in a relatively small area of central, south central, and southeast areas regions of the country. March, June, and July are the high-occurrence months.12
Southern Ontario and Quebec and the Canadian Prairie Region are the high-occurrence areas in Canada, with June, July, and August being the most likely months of occurrence. In this region of the world, the Rocky Mountain cold front interacts with warm fronts to create thunderstorms.4
Forecasting Tornadoes
While predicting thunderstorm probability over the upcoming week is fairly accurate, it is more challenging to predict tornadoes, which are spawned from thunderstorms. Mesocyclone formations within storms are areas of rotation. If the rotating column of air descends and reaches the ground, it is classified as a tornado. Tornadoes can develop within minutes, sometimes leaving a short period of time for detection and warning. Doppler radar, which detects wind circulation within a storm, suggests potential tornado formation. Storm spotters on the ground in contact with the radar operators assist with confirmation by direct visualization.
Tornado Warning
NWS issues tornado warnings that are disseminated to weather alert radios. NWS partners with various media and communication services that rebroadcast warnings by television, radio, cell phone messaging, and social media. As described previously, the time from issue of warning to tornado impact is termed “lead time.” At the time of this writing, NOAA reports an average lead time of 13 minutes in the United States.13
American meteorologist Harold Brooks of NOAA and the National Severe Storms Laboratory (NSSL) explains that the lead time has remained unchanged for 25 years and that it has averaged 18.5 minutes. He believes the discrepancies in times reported are due to the inclusion in more recent reports of “no warning” tornado events (lead time = 0) that skew the data.14
Intuitively, longer lead times would allow more time to take protective action and reduce injury and death. However, some weather experts question that premise. They theorize that long lead times indicate that the tornado is large in size and thus was detected sooner and has greater destructive energy. If this were the case, taking refuge inside a structure would increase risk of serious injury or death if that shelter is demolished.6 A regression model analysis of lead time versus injury and death in more than 18,000 tornado events from 1986–2002 revealed that, when compared to no warning, lead times up to 15 minutes reduced injury and death, while lead times beyond 15 minutes were associated with increased morbidity and mortality.6,15
An explanation for this seemingly paradoxical finding is that people erroneously think they have ample time before storm arrival for various activities like getting in a vehicle to outrace the storm. In addition, some people in harm’s way have been observed to go outside to watch the storm, mistakenly thinking they would have enough time to get to shelter. Considering that nearly 75% of warnings are false alarms, after 15 minutes, people might assume this is the case and leave the shelter even though they are in a true event.6
In Bangladesh, a central government agency, the Bangladesh Meteorological Department (BMD), monitors adverse weather events. BMD uses a sophisticated weather observation and data collection system located centrally and in strategic outlying locations to develop and issue warnings via electronic media.16 Table 37.1 compares the three countries by population and media capabilities illustrating that Bangladesh’s population is mostly rural and without widespread electronic media coverage. Therefore timely advanced warning may be sparse in many areas of the country.17