On May 31, 2013, El Reno,
Oklahoma experienced the widest tornado in recorded history (Ukusurri et al,
111). Believed to have reached its
maximum extent at 2.6 miles, it is one of the best-documented cases of a mass
tornado evacuation in history (Hatzis & Klockow-McClain, 722). As most tornado responses don’t involve an
evacuation, this one was mostly due to its sheer size. It was also compounded by a rough 2 weeks of
severe weather for the city, which had started with the disastrous Moore
tornado that killed 24 people on May 20th that year.
What
generated this monster tornado was the combination of a classic southwest dryline,
a slow-moving front across the Great Plains, high surface dew points, unstable
lapse rates, and high vertical wind shear (ibid., 722). The supercell that formed from this
combination tracked east over central Oklahoma.
Tornadogenesis was just south of El Reno, but the path alarmingly went
in the direction of metropolitan Oklahoma City.
Fortunately, it did not reach the city, which is about 30 miles east. It primarily passed over open country, killing
8 people (all in vehicles), including 3 storm chasers (ibid. 722) before stalling
over I40 and dissipating.
The
official rating from the National Weather Service was EF3, but Doppler radar detected
velocities that exceeded the wind threshold of an EF5 (ibid. 722). Based on the size, intensity, and damage inflicted
on things it did intercept, including a brutally mangled car, many believe the rating
should have been upgraded to EF5.
The
El Reno tornado is known for having many traits that aren’t typically seen even
among EF5 tornadoes. When the tornado
was visible, several sub vortices were detected by storm chasers, spinning
around the main vortex like a top. It
also took a deadly turn just as it was doubling in size, causing many chasers
to get caught in its path. Because the
tornado was rain-wrapped, it was difficult for many to see. The deceptively wide base made it appear to
be merely a rainstorm. As it stalled on
I40, poor visibility caused several people to drive right into it, including
the driver of a semi-truck. These anomalies
make El Reno a highly unique tornado, and one of the most fascinating to study
in recent history.
Reports
of the tornado’s intensity led to a large-scale evacuation of metropolitan
Oklahoma City, which had been reeling from the disastrous Moore tornado only 11
days prior. The evacuation created major
traffic jams that would have caused fatalities in the hundreds if the tornado
had reached Oklahoma City (ibid. 721). The
traffic jams increased potential for a violent tornado hitting gridlocked
traffic, especially as the tornado struck during the afternoon commute. As vehicle fatalities account for 10-20% of
all tornado fatalities (ibid. 722), a response that didn’t involve thousands of
people stuck in their vehicles should have been executed by officials during this
event. Telling people who are already on
high alert from 2 weeks of severe weather that they need to evacuate if they
cannot get below ground can also lead to mass hysteria, putting more lives at
risk (ibid. 733). Any emergency communication
network must clearly decide where a tornado is heading and prioritize the
safety of individuals in its path without alerting a whole city.
Two solutions to the communication problem are in the use
of social media to monitor tornadogenesis, and the crowdsourcing of information
by storm chasers and researchers. In the
case of crowdsourcing, storm chaser video can be collected and fixed precisely
in time and location (Seimon et al, 2070).
Geographic Information Systems (GIS) can then be used to georeference
storm chaser video (ibid. 2079) as it is happening. Though this may be difficult to achieve in
real time, it is possible. While this
provides a benefit to future researchers, it also presents an opportunity for emergency
alert systems to track the speed and direction of tornados. It would help respondents see any surprising
developments or abrupt changes in direction on camera rather than relying on
radar or ground reporting alone, which are vulnerable to delays. In the case of El Reno, the tornado was
crowdsourced extensively, but only after the event happened and not by
emergency alert systems. Even if this
approach proves impractical in emergency settings, the added benefit of
crowdsourcing a tornado is that researchers can reconstruct it to find behavior
that led to any accidents and fatalities (ibid. 2079), thus helping emergency
planners predict problem areas in future events.
Another solution is the use of social media to geolocate
disaster spots. Twitter (now X) has
already been used as a source of information for pinpointing disasters or
social emergencies (Ukkusuri et al, 110).
Posts with hashtags can provide unique and valuable information toward
ground responses, information sharing, and can also help with crowdsourcing. Crucially, it accelerates the speed of
information by the sharing nature of threatening situations (ibid. 110). The information can help public and emergency
management authorities improve the understanding of on-the-ground realities
during emergency events like tornadogenesis (ibid. 110). As some posts contain geolocation data, it is
useful in identifying local hotspots of activity (ibid. 111). However, posts that do not have this
information would require a bit of data mining, which can be slow in real time.
If these communication methods had been used during the
El Reno event, it would have prevented the major traffic jams that put many
lives at risk. A whole city simply does
not have enough time to evacuate from a tornado that just formed 30 miles
away. A small section of the city could,
but even this wasn’t necessary for this tornado. With improvements in crowdsourcing and data
collection on social media, an evacuation for the El Reno tornado wouldn’t have
been necessary, as respondents would have seen it hooking away from the city and
slowing down near the interstate. A
real-time GIS generated map can provide all the functions of a spatially
motivated evacuation plan, provided the emergency team has enough data from crowdsourcing
and social media.
Sources:
Seimon,
A., Allen, J. T., Seimon, T. A., Talbot, S. J., & Hoadley, D. K. (2016). Crowdsourcing The El Reno 2013 Tornado: A New Approach
for Collation and Display of Storm Chaser Imagery for Scientific Applications. Bulletin
of the American Meteorological Society, 97(11), 2069-2084. https://doi.org/10.1175/BAMS-D-15-00174.1
Hatzis,
J. J., & Klockow-McClain, K. E. (2022). A
Spatiotemporal Perspective on the 31 May 2013 tornado evacuation in the
Oklahoma City Metropolitan Area. Weather, Climate, and Society, 14(3),
721-735. https://doi.org/10.1175/WCAS-D-21-0106.1
Ukkusuri,
S. V., Zhan, X., Sadri, A. M., & Ye, Q. (2014). Use of Social Media Data to
Explore Crisis Informatics: Study of 2013 Oklahoma Tornado. Transportation
Research Record, 2459(1), 110-118. https://doi.org/10.3141/2459-13