On May 31, 2013 an exceptionally large and
powerful tornado formed near El Reno, Oklahoma. Up to 2.6 miles in diameter, this tornado
produced winds near 300 mph, ranking with some of the most intense tornadoes in history. This tornado exhibited an unusual penchant for changing both speed and direction:
Forward speeds ranged from nearly stationary to over 50 mph while the direction of movement
spanned over 360 degrees, looping over Interstate 40. This made safe observing at close range
almost impossible. The high-precipitation character of the parent
thunderstorm made viewing very difficult for storm spotters and chasers. All of these factors
combined to produce an incredibly dangerous situation in which many storm observers were
forced to flee for their lives. Unfortunately, not everyone made it out. Among the eight victims this storm claimed
were well-known storm chasers Tim Samaras, Paul Samaras, and Carl Young - as well as
at least one other storm observer. In view of this tragedy, we deem it important
to document the lessons that this storm has pressed upon the chasing and spotting community.
We hope that this documentation will reduce the risk of another such tragedy. Friday, May 31st Severe weather has been in the forecast for days. But where? And would a tornado threat
exist? It is early morning. Forecasters are convinced
that a potent combination of atmospheric ingredients will set up over central Oklahoma. Tornadoes
look likely - in an area that has already seen more than its fair share of wicked weather. On May 20th, a rare EF-5 tornado ravaged Moore,
Oklahoma, a suburb of Oklahoma City. This tornado took the lives of over 20 people,
and left hundreds injured. Similar to the 20th, a large, slow-moving,
upper-tropospheric trough is positioned over the Plains.
The southern periphery of this trough features an intense mid-upper tropospheric jet, which
has initiated strong southerly low-level flow. Very rich moisture is moving toward the Plains.
In the middle troposphere, strong westerly winds are transporting a cool, dry layer of
air over the warm, moist layer to the east. The overlap of these two air masses is creating
a zone of conditionally unstable air over the Southern Plains and the Upper Midwest. Conceptually, conditional instability is similar to the hot air balloon. Hot air is less dense
than cold air at the same pressure level, so the balloon rises. This force is known
as buoyancy. Gravity acts in the opposite direction. If buoyancy is greater than gravity,
an upward force is created Instability can be measured from weather balloon
data. These balloons rise through the troposphere, creating a vertical profile of temperature,
humidity and winds. During the early evening, a weather balloon
is launched from Norman, Oklahoma, revealing a vertical profile of the troposphere. The
environmental temperature with height is in red. The parcel temperature, analogous to
the hot air balloon, is the dotted line. Notice that the parcel temperature is much greater
than the environmental temperature. This means that air originating from the ground will
rise with great force once the environmental temperature is cool enough. The total difference
in the two temperature traces is known as the Convective Available Potential Energy,
or CAPE. The area between the two temperature traces determines the value of CAPE.
All of this energy is contained by a warm, dry layer of air known as the "the cap".
This layer of air originates in the Desert Southwest. It prevents thunderstorm development
through most of the day, generally until peak heating. When it breaks, it breaks explosively
- similar to the effect of removing a lid from a boiling pot of water. The combination of rich, Gulf moisture with
cool, dry air above has led to the development of extreme instability. CAPE values reach
as high as 6000 j/kg in SW OK. A surface low pressure has formed over southwestern
Oklahoma, backing the flow in central Oklahoma from south to southeast. In addition to strengthening
convergence in southwestern Oklahoma that may lead to the initiation of thunderstorms,
this is also increasing the turning with height in the troposphere, leading to stronger vertical
wind shear. In the middle troposphere, winds are from
the west and southwest - generally above 40 kts over the Southern Plains. Combined with the backed surface flow, this
has led to the development of deep layer shear more than 40 kts. Given the extreme instability,
conditions have become more than sufficient for rotating thunderstorms. To illustrate,
we turn to the paddle wheel. If water at the top a paddle wheel is flowing quickly, and water at the bottom is flowing
slowly, the paddle wheel will tend to rotate. This effect also occurs in the atmosphere,
since air is a fluid. Here is an image from a developing thunderstorm on a day when tornadoes
occurred nearby. The winds at the top of this towering cumulus are much stronger than at
the bottom. This causes the developing storm to lean downshear. Unseen by the eye are numerous
horizontal circulations. The formation of supercell thunderstorms -
from which tornadoes develop - begins with these horizontal circulations, created by
differences in wind speed and direction with height. When a developing thunderstorm updraft encounters
a circulation, the circulation is tilted upward by the updraft into the shape of a horseshoe.
This creates circulations spinning opposite directions. The northern circulation spins
clockwise, and the southern spins counterclockwise. The wind pattern often favors strengthening
of the southern circulation, while the northern circulation tends to dissipate. The southern
updraft continues to strengthen as an adjacent downdraft forms -created (in part) by precipitation.
This precipitation is wrapped around the updraft, creating the characteristic hook echo. A supercell
is born. The stage is set for intense supercells. A
stationary boundary - generated by previous convection is positioned over Interstate 40
in central Oklahoma. It is adding fuel to the proverbial fire: it is simultaneously
enhancing the low-level shear while creating a focus for convective initiation at its intersection
with the dryline. Visible satellite imagery shows partly cloudy
skies in central Oklahoma, warming the air near the stationary boundary, providing power
for explosive updrafts. It is 4 p.m. The atmosphere is primed for
the development of intense thunderstorms. Radar shows a number of attempts at thunderstorm
initiation west of Oklahoma City, but as of yet, nothing has developed. However, on the ground, cumulus clouds are seen towering high just west of Oklahoma City,
indicating that thunderstorms will initiate soon. Storm chasers are beginning to
converge in the small town of El Reno, as the location of key boundaries strongly hints
that it will be ground zero. Around 4:30, the capping inversion - the
layer of warm air that often suppresses storm development in the Plains - breaks in a northeast
to southwest line about 50 miles west of Oklahoma City. Several storms rapidly form, reaching heights well above 50,000 feet. Still, considerable
uncertainty remains concerning which storm, or storms, will dominate. This storm, approximately 15 miles west-southwest
of Calumet, quickly develops mid-level rotation. Radial velocity shows a weak - but classic
- rotational signature, about 10 miles west of El Reno. At the ground, storm observers note a rapidly
rotating wall cloud just south of the interstate. The National Weather Service issues a tornado
warning. Unconfirmed reports of a large tornado are hitting the TV airways. Just as this storm is approaching maturity,
another storm develops to its south, merges and begins to rotate. Shortly after 6 p.m., a vigorous circulation
develops rapidly 10 miles southwest of El Reno. From the ground this circulation is seen as
another large, rapidly rotating wall cloud. Within two minutes of its development, a ground
circulation develops. The El Reno tornado has begun. The tornado immediately shows strong multiple-vortex
structure. The Tempest Tour group is located southeast
of the tornado - a traditionally safe spot - if a tornado is moving northeast. The tornado does not initially appear to deviate
from the expected motion - to the east. It also doesnít appear very close: an illusion
created by the low-hanging cloud base. Subvortices appear as well. But instead of moving left
to right, they appear to move closer to the photographer. The tour leader begins to
sense danger, and signals his tour members to come back to the vehicle. The tornado does
not appear very large, but the entire circulation is almost a half-mile across. The tourists
frantically make their way back to the vehicle as the tornado rapidly closes on their location.
The tourists enter their vehicle, and are barely able to escape - the widening tornado
can be seen from a rear-view camera. It is about 6:06 p.m. The Tempest Tour group
is making their escape just south of the tornado. Meanwhile, Brandon Sullivan and Brett Wright
are stopped along Chiles Road, just south of Reno Street, filming the developing tornado
to their west. The tornado appears to be a half-mile away,
which should leave enough time for a safe escape. What they don't realize is that the
tornado has expanded to 3/4 of a mile wide, and the forward speed has increased to 35
mph. The edge of the tornadic wind field is close - and closing. Realizing the danger,
Sullivan and Wright begin to pack up their gear. Sullivan and Wright head south. They realize
the tornado is much closer and they make a frantic escape. They are overtaken by the
outer edge of the tornado. They survive without injury. Meanwhile, Dave Demko and Heidi Farrar
are observing the tornado a few miles to the north. Demko and Farrar are in the notch - the space
in between the large hail to the north and the tornado to the south. The storm is high
precipitation in character, and so it is difficult to see the tornado - especially from the
north. Not knowing the exact size and movement of the tornado, they begin to worry about
where to go. They agree to bail west. While Demko and Farrar are observing the widening
tornado from the north, Skip Talbot and Jenn Brindley are observing the approaching tornado
from the east. The time is 6:13 p.m. From their location,
the tornado is difficult to see because of the rain. Talbot notices very fast-moving
rain curtains - very close - and decides it's time to bail east. From Talbot and Brindley's location, the
condensation funnel is faintly visible The visible funnel is approximately 0.3 miles
wide. The tornado width - based on mobile radar data - is about 1.6 miles in diameter, noted by the
red circle. Thus, the area occupied by the full tornadic windfield is more than 10 times
larger than its condensation funnel. In this modified version of Talbot's video,
you can see how large the tornadic windfield is compared to the condensation funnel. Now realizing the imminent threat, Talbot
and Brindley retreat east on 15th Street, the tornado keeping pace at a blistering 45
mph. While they are fleeing, Ray Bohac and his
crew are following the tornado on Reno, just a few miles northwest of Talbot and Brindley. They watch as the large tornado intensifies
in front of them. It is 6:15. The tornado is difficult to see, enshrouded by rain and
debris. Suddenly, two more tornadoes appear: satellites. These tornadoes are 1/4 to 1/2 mile
from the edge of the main tornado. These tornadoes spin about the main funnel in a counter-clockwise
orbit. It is 6:18. A Weather Channel crew, led by
Mike Bettes, is racing south down HW 81. Meanwhile, Richard Henderson is trying to beat the tornado
to the east, his progress delayed by a chaser traffic jam. Near the intersection of 15th Street with
Highway 81, the tornado appears as a wall of condensation. Mikey Gribble shoots video to the west along
15th Street as the Weather Channel crew is frantically attempting to get past the tornado
a mile to his west. They do not make it. While the Weather Channel is attempting to
outrace the tornado, Richard Henderson continues east on Reno, hoping to beat the tornado to
the east. Hindered by the blinding rain and powerful winds, Henderson stops. With brutal
force, the tornado overtakes him. He does not survive. The lead car of the Weather Channel crew is
hit by a sub-vortex within the larger tornado, and their vehicle is rolled almost 200 yards.
The car is badly damaged, and all airbags are deployed. Amazingly, all of the team's
crew members have survived with non-serious injuries. As the Weather Channel crew is getting hit,
Dan Robinson is heading east on Reuter Road, just west of Highway 81. He briefly considers
heading south, but realizes there isn't enough time. The tornado is moving fast - much faster
than expected. Robinson decides to continue heading east on Reuter. Unknown to Robinson at the time, a white Chevy
Cobalt is following him. In it are Tim Samaras, Paul Samaras, and Carl Young. As Robinson crosses HW 81, he approaches Alfadale
Road and realizes that something is wrong - terribly wrong. The time is 6:20. Rain curtains - associated
with the tornadic circulation - begin to envelop his vehicle. The winds become much
stronger, too Though the main condensation funnel is still
a mile away, Robinson is overtaken by the invisible edge of the tornado. Robinson's
4-cylinder vehicle struggles to go eastward against strong east and northeasterly winds. The wind slows down Robinson's attempt to escape the
approaching tornado. Visibility is very low, as dust and precipitation
bands periodically hide the road in front of him. The tornado is just to Robinson's
south. To make matters worse, the traction control on his Toyota Yaris engages, reducing power to his wheels. Robinson struggles
to stay on the road. Not long after passing Radio Road, Robinson clears the future path
of the core flow - with winds now approaching 300 mph. Less than a half mile behind Robinson, a powerful
subvortex is swinging northwest toward Reuter. In the direct path, the TWISTEX crew is riding
out the storm alongside the road near a creek. The vortex briefly stalls over their Chevy
Cobalt. It tumbles over 5 times. They do not survive. The tornado continues to head north. The forward
speed of the tornado slows down to less than 10 mph. The tornado becomes increasingly wrapped
in rain. Meanwhile, at 6:41, Skip Talbot spots another
tornado approximately 5 miles to the southeast of the main tornado. This tornado, though, is spinning clockwise,
or anticyclonically. Though only a couple hundred yards in diameter this tornado is
powerful, with peak winds approaching 150 mph. Finally, at 6:43 pm, the tornado dissipates
near the intersection of I-40 and Banner Road. Several lessons have re-emerged from this
tragic event. It is our hope that these will lower the probability of another chasing or
spotting tragedy. First, tornado motion is always unpredictable
- even for big tornadoes. They don't move in straight lines or at constant speed. And
it's often difficult to tell where a tornado is moving.
The El Reno tornado changed directions over 360 degrees. Thus, if you were close, there
was no safe spot, regardless of what direction the tornado had been moving.
Additionally, the range of speeds in the El Reno tornado was enormous - from nearly stationary
to over 55 mph. Near Highway 81, the tornado doubled its speed - from 25 to 50 mph in
5 minutes. If you can't see the tornado - as was the
case with Dan Robinson and the TWISTEX chase team north of the tornado - you may be in
mortal danger. Sudden turns can and do happen. Tornadoes can expand rapidly. From 6:05 to
6:10 - when Brandon Sullivan and Brett Wright stopped to shoot video along Chiles
Road - the width of the tornado increased from 6/10 of a mile to 1.2 miles wide! Making
a close approach to a tornado can be very dangerous - and potentially deadly. A tornado is often larger than its condensation funnel - in some cases, much larger. Skip
Talbot's view of the tornado at 6:13 demonstrates this quite well. The tornado appeared to be
1/3 of a mile wide, but the tornadic wind field was well over a mile wide.
In the case of the TWISTEX group, it is quite likely that they thought they had more time
to escape the tornado than they actually had, since the outer edges of the tornado were
not visible. However, the easterly winds inside the uncondensed tornadic circulation were
powerful enough to hinder their escape on Reuter, resulting in tragedy. When big tornadoes occur, they are often accompanied by other tornadoes. These additional tornadoes
present a big problem for those trying to observe storms safely.
The first type is the satellite tornado. At 6:15 p.m., multiple satellite tornadoes were
observed on the west and south side of the El Reno tornado. These tornadoes generally
occur within a mile of the main tornado in any direction. Close observers are particularly
vulnerable to this type of tornado. The second type is the anticyclonic tornado. This type
of tornado spins in the opposite direction of the main tornado. While the El Reno tornado
was wrapped in rain near I-40, a powerful anticyclonic tornado - with winds up to 150
mph - developed to the southeast of the main tornado. Typically, these tornadoes form to
the right of the hook echo, a fair distance away from the main cyclonic tornado. The final
type of danger comes from new tornadoes forming in new circulations within the parent thunderstorm.
They generally form downstream of the existing tornado. The "notch" of a high precipitation supercell is extremely dangerous. It is why "core punching" - approaching the tornado from the rain and
hail - is so perilous. This is the area of the storm immediately
to the left of the tornado, and just to the right of the large hail. Those in the notch are in danger of the sharp
left turns that tornadoes often make when they are dissipating. If the tornado can be seen, successful escapes
can be made. However, heavy rain may hide the tornado. In that case, radar updates may
be the only way of knowing where the tornado is located.
But in the case of the El Reno storm, the tornado moved 2 miles north in less than 5
minutes - less than the interval of a WSR-88D volume scan. Additionally, the radar cannot
pinpoint the exact location of the tornado with certainty. So you should not depend on radar to know
where the tornado is. And even if the position of a tornado is known,
strong inflow winds may hinder a quick escape. This almost certainly was the case for the
TWISTEX team on Reuter Road. As mentioned previously, new tornadoes are
always a danger in the notch And, of course, there's the lesser threat
of very large, glass-breaking hail in the core of the storm. In the path of an approaching tornado, a quick escape may not be possible. A lack of good
road options, poor road conditions, or even traffic may hinder a safe escape. And in the
case of the El Reno tornado, numerous traffic jams were reported. It appears that these
traffic jams may have resulted in the deaths of at least 3 people in 3 different cars. Based on these lessons, we suggest that storm
spotters and chasers place more distance between themselves and the tornado - especially on
days when parameters are particularly volatile. When the instability and shear combination
is high, the storm evolution may occur more quickly, decreasing the margin for error.
Moreover, given the rarity of the ingredients that produced the El Reno tornado, storm behavior
may differ greatly from more "normal" tornado days. For example, on June 8th, 1995, a very
large tornado accelerated to nearly 60 mph near Allison, TX, before slowing down to nearly
stationary. You may have seen videos of people who escaped death and serious injury when their vehicles
were hit by the El Reno tornado. But it's critical to remember that in most of those
cases, the vehicles in most of were not impacted by the strongest winds in the tornado. It
is possible that this has resulted in a false sense of security within the storm observing
community. The most powerful winds in a tornado are located in sub-vortices, which are smaller tornadoes
within the larger circulation. These are the vortices responsible for leveling one house,
but leaving a house next door unscathed. Given the small area they occupy, the probability
of being hit is actually rather low. However, as the number of close encounters increase,
the odds increase that more chasers will encounter these deadly winds. This is especially true
now given the growing trend of "extreme" chasing. Remember that no footage, report, or data is worth your life. Of the 8 deaths in the
tornado, at least 4 were chasers. The number of chasers that were killed, injured, or narrowly
escaped was far larger than any other documented tornado. There will always be more storms.
