Interesting Weather Information

Saturday, June 22, 2013

Occluding Mesocyclones - Evidence from the El Reno , OK Tornado - Part II Radar

Updated 06.22.2013 20:45z

In Part I of this post I covered the basic meteorological setup of the El Reno, OK EF5 tornado of May 31, 2013. It was the widest tornado ever measured in the U.S. with a maximum width of 2.6 miles (4.2 km).The former record holder was the Wilber - Hallum, NE tornado of  May 22, 2004 which was 2.5 miles (4.02 km) wide.

In Part II I take a look at radar data, more of it than you would normally see on a TV weather program. I hoped to post this a couple days after the first post but this is so much going on with this tornado that I had to go back and review a number of journal articles to clarify this complicated tornado.

Incredible Storm Chasing Video

If you think the only dangerous part of a tornado is the funnel,  KFOR Meteorologist Emily Sutton probably has some advice for you. Take a look at what she and her crew got into while chasing the El Reno tornado.

Here is the video link:

In an email about her close call Emily wrote "I'm blessed to be alive".  After staring at the radar data for hours and hours, I think that is an understatement.  Emily and Kevin Josefy got about as close as a chase team can get and live to tell the story.  Her encounter was on the same road and a very short distance from where professional storm chasers Tim Samaras, his son Paul Samaras and Carl Young were killed and not too far from where the tornado ripped the hood from the armored tornado attack vehicle called The Dominator.

Meteorologist Emily Sutton, KFOR TV, Oklahoma City.

 In Emily's opinion she got into an intense rear flank downdraft (RFD) and the air sinking behind the tornado was screeching towards and into the tornado circulation.

RFDs are known to at the very least help intensify mesocyclone tornadoes and there is mounting evidence that they may be an essential element of tornado formation by supplying air to the surface beneath the cloud base and updraft.  RFD air wraps around and confines funnel rotation, concentrates the spin and helps bring rotation to the surface. As the updraft stretches the rotating air vertically the funnel narrows and the rotation rate increases. Think ice skater.

Top: Meteorologist Reed Timmer and  and
an early version of The Dominator. Below:
Current version of The Dominator .
Courtesy: Discovery Channel
When discussing the storm with Reed Timmer meterologist, storm chaser, tornado attack vehicle designer and Discovery Channel star who also chases for KFOR TV, Reed told Emily it was probably an inflow jet that smashed the back windshield of her chase SUV as her team were in a backwards mad dash to avoid the El Reno tornado.

An inflow jet is just what it sounds like it is.  Because of the extremely low pressure in the tornado air rushes in from all directions. When tornado-bound air is confined to a narrow channel, for example by the gust fronts of the  RFD  or the forward flank downdraft (FFD) of a newly forming upwind supercell  (in this case to the west-southwest) air rushes into the vortex.

Inflow jets can be fed by RFD air and driven to impressive velocities. Inflow jets can also race parallel to the RFD gust front independent of the RFD itself. In fact inflow jets can be found almost anywhere around a tornado as air screaming into the incredible pressure falls of a tornado vortex interact with terrain, other inflow jets and neighboring storms.

Doppler radar images from the El Reno - Yukon tornado tell an incredible story. And it looks like both Emily Sutton and Reed Timmer are right about the presence of RFD air and inflow jets. Ultimately data indicates Emily was just on the edge of the tornado circulation when it was at it widest and strongest. Luckily far enough to sustain only a broken back windshield.

Radar data shows the El Reno - Yukon tornado was complicated and there is ample evidence of a strong RFD which wrapped around the tornado vortex as a strong inflow jet.  In addition it is possible that a newly formed supercell may have helped channel and intensive the RFD inflow.

Supercell Fundamentals

The diagrams below show what you will find in most tornadic supercell environments. The specific geometric arrangement varies from storm to storm and not every feature that you see in these generalized schematics may be present in each storm. You have to think in 3D to get a full understanding of what is going on.

A supercell is a thunderstorm with, by definition, a tilted, rotating updraft. Because the updraft is tilted most of the rain falls outside the updraft. In a thunderstorm if the updraft is vertical rain falling  through the updraft works against the rising air. A thunderstorm with a vertical updraft has a short lifetime while a supercell can last 12 hours and travel hundreds of miles.

The tilt is the result of vertical wind shear, that is, wind speeds increase with increasing elevation so the top of the updraft is farther downwind (farther in the direction of storm motion) than the updraft base.

Supercell rotation is largely a matter of the storm importing rotation from the environment, although there are mechanisms internal to the supercell that can create rotation also.  In general meteorologists use the term vorticity to measure rotation and for qualitative discussions vorticity and rotation can be considered to be the same thing.

For more on rotation imported into thunderstorms use this link to a post about the March 2, 2012 tornadoes south of Cincinnati, OH:

Heavy rain ahead of the updraft core creates the Forward Flank Downdraft or FFD. The cool rush of air you feel just before the heavy rain and the arrival of the thunderstorm is the Forward Flank Downdraft.

There are two mechanisms that account for the FFD Downdraft:
  1. Each raindrop of the billions upon billions of drops pushes a bit of air towards the surface as it falls. The combined total can create and impressive wind gust when the descending air reaches the ground.
  2. Some of the raindrops evaporate on their journey to the surface cooling the descending air. Cooler air is more dense and adds to the velocity of the downdraft.
The abrupt change from warm muggy air to the gusty, cool air marks the leading edge of the FFD which we call the Forward Flank Gust Front. Just like the fronts on a weather map it is the boundary between contrasting air masses only smaller. The FFD Gust Front is also called an outflow boundary.  As the cool air undercuts the less dense warm and humid air it helps to increase the lift on the air flowing into the thunderstorm. The FFD Gust Front can:
  1. Cause the supercell to become more intense due to the increased lift.
  2. Help generate rotation when the air is lifted along the gust front.
  3. Cause new cells or a squall line to form along the gust front. 
The Rear Flank Downdraft or RFD is still a bit of a mystery with several candidates as possible mechanisms for formation of the RFD. The RFD is also the cause of several features that may contribute to tornado formation. The RFD is dry, sinking air behind the storm (relative to storm movement). It may be seen as a clear slot behind a tornado. It is likely the cause of the hook echo as the sinking dry air spirals into the mesocyclone. Radar sees the rain curtain as a spiral of high reflectivity and the air originating in the RFD a spiral lower or no reflectivity.

The leading edge of the RFD outflow is the RFD Gust Front and new cells that form along it are called the flanking line.

Possible causes of the RFD (all may operate together):
  1. Stagnation of mid-level air as it flows around the back side of the storm's updraft chimney. Pressure increases aloft forcing air to sink at the rear of the storm.
  2. Cooling of air flowing out of the storm aloft by evaporation or melting hail.
  3. Cooling of drier air moving into the storm from the rear also by evaporation or melting hail
  4. Dynamic effects that are beyond the scope of this post. Just think of it as what goes up must come down - an action/reaction pair.
Possible effects of the RFD:
  1. RFD outflow supplies air that wraps around the developing tornado vortex and confines the air to a narrower column helping to increase spin and velocity of the updraft. The updraft stretches the spinning column vertically and as the column narrows the spin rate increases.
  2. RFD air may transport rotation from aloft to the surface that aids in spinup of the tornado. Evidence for this is descending reflectivity cores (DRCs), areas of heavy rain that sink towards the base of the supercell rotating updraft.
  3. RFD air helps increase low level spin and may be crucial in bringing mesocyclone rotation to the surface or alternately in creating surface rotation that grows upward via the updraft joining the mesocyclone.
  4. RFD air sinks and is often warm and can arrive at the updraft as lower density air that is easily lifted.
  5.  If there is insufficient air flowing into the updraft of a supercell a tornado cannot form.The RFD probably supplies air that helps maintain a tornado. In other words if the updraft can remove more air than can be pulled into the storm there will not be a large, long-track tornado. Meteorologists refer to the swirl ratio of the storm. In simple terms it is the ratio of inflow volume  to updraft volume. As the swirl ration grows (> 1) the inflow air cannot be evacuated and the funnel becomes fatter eventually breaking dow to a multiple vortex tornado.
  6. New low-level mesocyclones may form on the RFD Gust Front and as the old mesocyclone occludes, that is it is absorbed by the supercell the new mesocyclone may generate a new tornado. These are called cyclic supercells and are the source of tornado families.

If you are not yet completely comfortable with stormglish ( aka chase speak), refer to these diagrams for reference during the following discussions.

A typical supercell  and surrounding environment from above.  The inflow jet shown is a "classic" situation. Because of the strength of tornado circulations inflow jets can develop from any direction.  The inflow jet along the RFD gust front is most common. the RFD flow itself can accelerate into the tornado circulation also forming an inflow jet and interaction with terrain can confine inflowing air enough to cause a jet from any direction. For more investigate the "Venturi Effect".

A typical supercell and surrounding environment from the side.

Radar Loops 
Reflectivity, Radial Velocity and Normalized Rotation
May 31, 2013 - KTLX Radar (Norman, OK NWS Office)

If these animated gifs are not looping, either refresh your browser or left click on the image to display the larger version in the blogger image viewer.

Reflectivity Display. Animation of the El Reno tornado hook echo from  KTLX the Norman, OK NWS Radar. The area enclosed by the yellow path is the damage area from the tornado as surveyed by NWS meteorologists in Norman. The small yellow square south-southeast of El Reno, is the approximate location of meteorologist Emily Sutton and her crew as the tornado crossed U.S. Route 81 the road from El Reno to Union City, and the point where the tornado turned left  (i.e. headed to the northeast).  The yellow arrow shows the direction in which Sutton and crew backed up to get away from the tornado. How close were they to disaster? Look at the entire series of animations below for a better idea. Notice how the hook seems to aquire reversed curvature (anticyclonic curvature) at 23:19:11z due to strong inflow from the RFD/inflow jet from the west-northwest.

KTLX reflectivity radar loop, 2246z 31May -0037z  01 Jun 2013 (5:46 PM - 8:37 PM CDT). The down-pointing triangle is where the MDA (Mesocyclone Detection Algorithm) of GRLevel II Analyst computes enough rotation for a tornado. The animation below is a closer view. Later in this post using a series of stills I will point out some of the features you can see in these animations and discuss what those features mean.
Same as above but at the second tilt elevation. In both this view and the view above a well defined RFD can be seen developing as a low reflectivity notch on the west side of the hook. 
Two of 7 tornadoes in the Grand Island, NE area on June 3, 1980. the right funnel is a mesocyclone funnel and the debris on the left is from an anticyclonic (?) gustnado. See diagram below.

Ted Fujita's diagram of a cyclonic/anticyclonic tornado pair during the Grand Island, NE tornadoes of  3Jun1980 is similar to the radar view of the El Reno - Yukon tornado  hook echo as it intensified and turned to the northeast with the end of the hook developing anticyclonic curvature. In the Grand Island event a small anticyclonic tornado developed. There was no such development in the El Reno - Yukon tornado. It is possible that the anticyclonic tornado developed on the RFD gust front as it occluded (curled into the center of the circulation), making the anticyclonic tornado a "gustnado".  In the El Reno - Yukon case it looks like simultaneous strengthening of the RFD gust front and the FFD (forward flank downdraft) of the second cell created an intense inflow jet and the shear induced the anticyclonic curvature and rotation. See the diagram below.

Base velocity display. Red is air movement away from the radar (+ in polar coordinates), green is towards the radar (- in polar coordinates). A doppler radar can only measure movement directly along a radial so the display does not show true wind velocity or direction but the component of the wind along a radial. Because of this any wind that crosses a radial will have a doppler measured velocity slower than the true wind. 

El - Reno - Yukon tornado. Background: base reflectivity, antenna angle 0.5° at 23:19:14z as the tornado was crossing U.S. Route 81, reaching maximum size and intensity and turning to the northeast. Red and Green outlines: location of  GRLevel 2 Analyst derived positive (red) and negative (green) rotation. The red outline is for NROT approximately >= +1.0, the green outline is for NROT approximately <= -0.90. The area of anticyclonic (negative) rotation is shear induced as indicated by translucent white arrows. Inflow jet is shown. It is probably a combination of RFD inflow and the influence of the developing second supercell.  The greatest inbound radial velocity is located at the small square (-140 kts/72.1 m/s).
Normalized Rotation ( Range +5 to -5, positive is cyclonic, negative anticyclonic, valuesbeyond +/-2.5 are extreme).  This product is unique to GRLevel 2 Analyst and was developed by Gibson Ridge Software. It differs from rotation algorithms used in NWS radars.  Because of the polar coordinate geometry native to radar systems data bins become larger as distance increases from the radar. A target of given size will occupy a decreasing proportion of the volume as distance increases.  This means that velocity measurements slow as range increases which changes rotation calculations.  In addition position within a volume becomes increasingly important with increasing distance from the radar due to increasing volume of data bins.

The Gibson Ridge normalized rotation algorithm takes all this into account and from this example preforms well with both placement of the center of rotation and the change of intensity when compared to ground surveys. The strange behavior of the TVS symbol in the last couple frames is likely due to the multi-vortex nature and great width of the vortex/mesocyclone as it decayed.

Same as above but with a graph of positive (cyclonic) and negative (anticyclonic) normalized rotation at an antenna elevation of 0.5°. Left click for a larger version. On the graph any value  greater than 2.5 or less than -2.5 (the yellow lines) is considered to be extreme. 

Same as above but the map is for an antenna elevation angle of 0.9°. The left graph is for the 0.9° tilt and the right graph is the next tilt up 1.4°. Note the map scale is different from the map above to acomodate two graphs. 
Emily Sutton and Kevin Josefy were close,  too close for comfort. Essentially they were within the outer fringe of the tornado's circulation when it was the widest and strongest. 

Contrast what Emily and Kevin did  when faced with life a threatening situation to try and stay safe with  what the Weather Channel crew did.  Which crew acted the smart way?

Radar Cross Section
Along a west (left) to east (right) line from
98.36° West Longitude to 97.65° West Longitude
Through the Center of Normalized Rotation @ Antenna Tile of 0.5°
Vertical reflectivity  cross section, west to east (left >> right) through the center of normalized rotation.  The BWER  is squeezed as the mesocyclone occludes and moves northeast into the main body of the supercell. The RFD flow, wraps completely around the mesocyclone and  is also squeezed as the next supercell grows and starts to merge with the supecell that caused the El Reno - Yukon tornado. In the last couple frames that have the RFD labeled  there is also evidence that the RFD is undercutting the vortex.

High angle view of the supercell's 40 dBz reflectivity from the southwest beginning 23:37:48z and ending 23:37:37z 31May 2013. Notice the following: 1. Areas of high reflectivity merging with the mesocyclone, 2. the well developed RFD (low or no reflectivity) as the hook reaches  US Route 81 and the anticyclonic curvature develops and 3. the next cell trying to spin up.

Medium angle view of 40 dBz reflectivity from the south-southeast beginning 23:56:14z and ending 23:37:37z. Careful observation reveals what are most likely descending reflectivity cores before the large curtain of rain wrapping around and into the circulation develops. Before the anticyclonic curvature develops the RFD is evident.

Medium angle view of 50 dBz reflectivity from the south-southeast beginning 22:37:49z and ending 23:37:37. What appear to be numerous descending reflectivity cores (DRCs) before and during the strengthening of the tornado descend and spiral into the mesocyclone circulation. The same thing happens with the development of the second supercell.

TDWR TOKC view of the evolution of the tornado, 1-minute time resolution. Antenna elevation angle = 0.5°

Wednesday, June 5, 2013

Occluding Mesocyclones - Evidence from the El Reno, OK Tornado, Part I - The Meteorological Set Up

This is the first of two posts on the El Reno - Yukon, OK tornado of May 31, 2013.
Part II: Radar Loops and 3D - Just What is Going On Here?
These posts are semi-technical and should be easy to understand for weather enthusiasts.


Left click on any image for a larger view.

It's official! The El Reno, OK tornado is the widest tornado on record to touch down the United States with a width of 2.6 miles and it is another EF5, the 60th in the United States since 1950 and number 8 for Oklahoma making it the state with the greatest number of F5 or EF5 tornadoes with a total of 8 since 1950.

Courtesy: NWS, Norman, Ok.

The damage path of the El Reno tornado as seen by the MODIS (Moderate Resolution Imaging
Spectroradiometer) Satellite. Modified from the original by Steve Horstmeyer.
Courtesy: Space Science and Engineering Center, University of Madison, Wisconsin.

The upgrade from EF3 to EF5 came about when the National Weather Service consulted with both Howard Bluestein and Josh Wurman.  They are on my list of Severe Storms Rock Stars (my post of March 9, 2013). Their Doppler on Wheels indicated extreme wind speeds and Dr. Bluestein's crew clocked winds at 296 mph while Dr. Wurman's crew, from a different vantage point,  measured winds from 246 to 258 mph.

Along with the damage path surveyed by the National Weather Service, Dr. Bluestein's crew measured the width of the tornado at 2.6 miles using their high resolution Doppler on Wheels.

A 2.6 mile wide tornado is BIG. Take a look at what would happen if one would move north through Downtown Cincinnati. Yellow shows what a 2.6 mile wide path would cover and everything from Lower Price Hill to Mount Adams would be affected.

A 2.6 mile wide path across Downtown Cincinnati.

The Satellite View

The El Reno tornado was a monster and satellite images and loops tell an interesting story. On these loops you can see the storms grow in a band from northwestern Illinois to southern Oklahoma.  As a former colleague liked to say, "... it looks like they are growing out of the ground". On the wide view notice the small cumulus clouds streaming northward from the Gulf of Mexico as warm, moist tropical air streams into the supercells.

Satellite Loops -If not looping <ctrl> re-load

Courtesy: NOAA

They look like flat round pancakes ( I have heard some meteorologists say they look like cow pies). As the storms bump into the tropopause at the base of the stratosphere the cloud cover spreads laterally blown by high altitude winds. Unless there is a big push the cloud cannot penetrate into the stratosphere. 

The big push comes in the rotating updraft. You can see the "over shooting tops" on the close up loop as rough, textured looking areas. As the air rushes upwards it forces its way into the stratosphere. The overshooting top has also been called the updraft dome. This term is rarely used today but if you look at older literature you may run into it.

Courtesy: Space Science and Engineering Center, University of Wisconsin - Madison.

The Role of the Jet Stream

If not looping <ctrl> re-load

The animation just below, courtesy of the College of Dupage ( shows why the storms were growing where they did and why they were oriented in a line from northeast to south west.

The severe storms were being powered by what you have heard TV meteorologists call an "upper level disturbance", UAD for short.  A UAD is a bundle of energy embedded in the jet stream. The bundle travels with the jet flow and air accelerates into the disturbance and slows coming out of the disturbance.

You can track the UAD moving into the central Great Plains from the west coast in the animation below. The map for 00z 01 June 2013 (8PM EDT, 7PM CDT the evening of May 31 in the U.S.) is enlarged below to show the UAD. 

A UAD  may go by other names too:  velocity max, vorticity max, speed max, impulse, jet streak and short wave. All of these are used by TV meteorologists to describe a bundle of energy moving in the jet stream.

Upper level disturbances in the jet stream before, during and after the El Reno, OK tornado.
Courtesy: College of DuPage,

The map below shows the UAD at 00z (7 PM CDT) Friday May 31, 2013. Yellow is the core of the disturbance with the highest wind speeds. The highest velocity is 80 knots (92 mph). From the center of the yellow oval the wind speed decreases outwards.

The weather map at the 500 mb (or 500 hPa) level at 00z 31May2013.  The height of this pressure level varies but a good target height is 18,000' above sea level and at thye base of what we call "jet stream level". The yellow core of the UAD or velocity max stretches from northwest Illinois to northeast Kansas. Courtesy: College of DuPage,

A UAD is like a wind tunnel in the jet stream - air accelerates into the wind tunnel through the entrance region and slows as it leaves the wind tunnel through the exit region.  Looking downwind, the wind is at your back, we can divide the UAD into quadrants. Air rises from the lower atmosphere to jet stream level beneath the Left Front Quadrant and beneath the Right Rear Quadrant. The reasons this happens in these quadrants are well known but beyond the scope of this post. Rising air means falling surface pressure and those towering, churning thermodynamic machines we call supercells are just low pressure cells that get stronger when a UAD provides lift.

The same map as above but annotated. Courtesy: College of DuPage,

On the map above the double headed arrow marks the middle of the jet streak half way from entrance to exit region.  The axis of the jet is indicated by the yellow arrows, longer arrows represent higher wind speeds. When air enters the UAD it accelerates to the middle of the UAD then slows as it heads from the middle towards the exit region.  For dynamic reasons air rises (the same as surface pressure dropping) in the Left Front and Right rear Quadrants.  The tornadic thunderstorms in Oklahoma were powered that night by lift provided by jet streak's right rear quadrant.

The Surface Map 

The surface map at 7PM CDT om May 31, 2013 was complex with a quasi-stationary front across Oklahoma from southwest to northeast.  South of that a dryline separating the dry desert air from the warm, moist tropical air off the Gulf of Mexico.
The map date/time information says 21Z which is universal time coordinates (UTC) and it is based on Greenwich, England. This map 
is for 4PM CDT 31 May 2013 based on the U.S. Central Time Zone. The cool outflow from the thunderstorms was plotted as 
outflow boundaries.

Regional Surface Maps
Surface observations at 23z May 31, 2013. Station data use the standard plotting model. The heavy line is the 60° isodrosotherm (dew point line) and yellow represents where dew point temperatures are 60° or greater, orange shows areas with dew point temperatures of 70° or higher. Created with Digital Atmosphere,

Same color scheme as above with streamlines added to show the flow of tropical moisture into the developing supercell complex.
A Surface Pressure overlay on the map with the same dew point color scheme as in the maps above. A stationary front to the north was increasing surface convergence. The high south of the front is a thunderstorm induced "bubble" high pressure system. The dryline separates desert air from very warm and moist tropical air that originated over the Gulf of Mexico. The trof (meteorological version of trough) from the low at the north end of the dryline is probably a result of the analysis scheme and the  the developing thunderstorms over southern Oklahoma you can see in the image below.

23z (6PM CDT) KTLX Radar (Oklahoma City). Note the hook echo east northeast of the radar site.

To sum the situation up:
1. A quasi-stationary front with warm moist air to the southeast over most of Oklahoma provided surface convergence and therefore lift.
2. The right rear quadrant of a moderately strong upper level disturbance was moving over the area also providing lift.
3. Instability was high, at Norman, OK the lifted index was  -7.69 and CAPE was 3260.19 at 00Z (7PM CDT). The atmosphere was primed to explode.

Next Part II - Interesting Radar Loops and Radar in 3D - Just What is Going On Here?

Tuesday, June 4, 2013

The El Reno, Oklahoma Tornado and a Tale of Many Motives

It was bound to happen. Fate is rarely cheated of the shocking headlines that seemingly happen way more today than they did a few years ago. May 31, 2013 is an example.  It is a day most of us who have spent our adult lives in the weather business will never forget.  In the headlines a Weather Channel meteorologist guides his crew right into an EF5 tornado, the widest ever recorded and  luckily they survive. They accomplished their goal - more viewers.  In the same storm a well respected research storm chaser, his son and another colleague lost their lives when their storm chase SUV became EF5 debris then was crushed, twisted and mangled beyond recognition.

Research storm chaser? I call Tim Samaras that to distinguish media storm chasers in search of viewers and  thrill seeking chasers in search of an adrenaline rush from serious scientists like Tim Samaras.  To be sure, I am not judging any of them. Consenting adults who recognize the danger and accept the risk are free to put themselves in harms way in a free society. I hope all of them always return home safely.

Tim Samaras did something no one else had done.  He placed instrument packages in the direct path of the EF4, Manchester, SD tornado on June 24, 2003. After a decade of trying, that day he scored a direct hit. The results of that dangerous activity are shown below in two graphs.

The top graph shows the 100mb pressure drop as the tornado passed over the instruments. The bottom graph shows a smaller pressure drop in the same tornado at a separate location.

It doesn't look like much but it took Tim Samaras 10 years to score a bulls eye with many close calls along the way. We know much more about tornadoes because he did.