Archive for October 9th, 2011

Analysis of San Antonio Tornado

Published on Sunday, 09 October 2011 20:57

David Reimer

0 Comments

Good evening,

For those who are not aware, a brief and weak tornado occurred in the western portions of San Antonio earlier this morning. The storm that produced the tornado was not under a Severe Thunderstorm or Tornado Warning at the time it produced the tornado. I’ve seen several folks slamming the National Weather Service on the lack of warning and I wanted to provide a brief explanation on how rotating thunderstorms are detected on radar and how we’re able to determine if a thunderstorm is capable of producing a tornado.

Before I begin analyzing the storm itself, it’s important to understand the environment that was present across the San Antonio area. For that, we need to look at two parameters specifically.

[caption id="attachment_4371" align="aligncenter" width="550" caption="Storm Relative Helicity (spin) in the lowest 1 KM of the atmosphere"]width=”550″ height=”412″ />[/caption]

[caption id="attachment_4372" align="aligncenter" width="550" caption="Instability (CAPE) in the lowest 3 KM of the atmosphere"]width=”550″ height=”412″ />[/caption]

To briefly summarize the overall atmospheric setup, a large area of rain was occurring across most of West Texas, roughly along and west of I-35. Instability values were meager but wind shear was sufficient enough to cause some organization in thunderstorms. That allowed one or two storms to become marginally severe with brief periods of small hail and some gusty downburst winds. However, it takes more then borderline conditions in most cases to produce tornadoes, even brief/weak ones. I want to direct your attention to the two graphics above. The first graphic shows Storm Relative Helicity, or the amount of spin a storm could utalize, in the lowest 1 kilometer of the atmosphere. In most fall/winter/early spring setups, these values are normally quite high. Generally, any number over 100 m/s supports the possibility of organized low level rotation, assuming all other ingredients are in place. The 1 KM SRH (Storm Relative Helicity) in the San Antonio area was in excess of 200 m/s, thus supporting the possibility of low level rotation. The second graphic shows the amount of instability within the lowest 3 kilometers of the atmosphere. When looking for the possibility of organized low level rotation and tornadoes in fall and winter setups, this is a particular algorithm I like to use to gauge the tornado threat on a particular day. This algorithm shows something of interest that may explain how a brief tornado was able to develop this morning in the San Antonio area. Notice that for the most part, values across the state are quite low, with the exception of deep south Texas and the Interstate 10 corridor. That’s how far north the instability pool was located at 1 AM. Now go back to the first graphic that shows the amount of spin in the lowest 1 kilometer of the atmosphere. For the most part, the highest amounts are located north of the low level instability pool, with the exception of the San Antonio area where just enough instability and low level wind shear were present  to combine and produce marginally favorable conditions for organized low level rotation. The overall atmospheric setup wasn’t favorable for an organized severe weather event with long-track supercells and significant tornadoes. However, conditions were marginally supportive of brief, weak low level rotation. One storm was able to organize just enough to produce a brief tornado. At the time of this post, the National Weather Service in Austin/San Antonio had found a damage track 1.5 miles long with a very narrow width of 50 yards wide. Based on storm motions yesterday, this tornado was likely on the ground for only 60 to 100 seconds and mainly resulted in a few trees being uprooted, which then caused property damage. This tornado will likely be rated EF0 to EF1 on the Enhanced Fujita Scale.

Now that I’ve briefly explained the atmospheric setup, let’s dive into the storm itself. What I’m about to show you are three radar scans taken between 1:14 AM and 1:23 AM CDT on October 9th. In addition to being able to see reflectivity(rain, hail, tornado debris) inside a storm, Doppler Radar is able to determine which way and how quickly rain is moving inside a thunderstorm. This is how we’re able to identify a possible tornado and issue a Tornado Warning based off that. The following images are displaying Base Velocity data, meaning that they are displaying raw velocity data without taking into account the storm’s motion or speed. Because the storm was so close to the NWS San Antonio Radar, that doesn’t really matter in this case. I’ve laid the data out in a 4-tilt arrangement, meaning the lowest tilt is on the top-left box, increasing in height as you go right on the image. The highest tilt is on the bottom right of the image and I’ve put the altitudes of the radar beam in each box to make it easier to identify.

[caption id="attachment_4376" align="aligncenter" width="550" caption="4/27/2011 5:20 PM - 243 Knot Couplet on the Tuscaloosa/Birmingham, AL Tornado"]width=”550″ height=”302″ />[/caption]

For reference to the discussion below, here is what a significant tornado looks like on velocity data on radar. This comes from the EF-4/Borderline EF-5 Tuscaloosa, Alabama Tornado back on April 27, 2011.

[caption id="attachment_4373" align="aligncenter" width="550" caption="San Antonio Radar - Base Velocity at 1:14 AM CDT"]width=”550″ height=”302″ />[/caption]

Now that you’ve seen what a velocity couplet looks like on a significant tornado, lets take a look at what the storm that produced the brief tornado looked like this morning. This radar data was taken at 1:14 AM, just a few moments before the tornado occurred. Do you see anything close to what the April 27 image looked like? Without getting to technical, the only type of organized rotation I see is at 8,000 feet and that’s still quite weak, around 30-40 knots of rotation. Below that, there is no organized rotation.

[caption id="attachment_4375" align="aligncenter" width="550" caption="San Antonio Radar - Base Velocity at 1:18 AM CDT"]width=”550″ height=”302″ />[/caption]

Fast forward about four minutes. There still isn’t a defined couplet on the lowest tilt of the radar, but there is signs of weak, but organized rotation at 4,400 feet and especially at 6,000 feet. Seeing weak rotation signatures like this are not out of the ordinary, and had the rotation been as organized on the lowest tilt, there may have been a Tornado Warning issued, but take a look for yourself. There’s maybe 15 knots of broad rotation at 2,800 feet. with around 55 to 60 MPH at 6,000 feet. This radar image would definitely make you want to pay attention of the possibility of the rotation lowering, but this image itself doesn’t support a tornado warning.

[caption id="attachment_4374" align="aligncenter" width="550" caption="San Antonio Radar - Base Velocity at 1:23 AM CDT"]width=”550″ height=”302″ />[/caption]

Fast forward another four to five minutes and any sort or organized rotation is gone. The lowest tilt shows no rotation, with the next 3 tilts showing very weak, disorganized rotation. So, you may be asking yourself why the radar did not show any sort of organized rotation on the lowest tilt of the radar. Well, there’s a very good answer to that. As I said above, the tornado was on the ground for one to two minutes at the most. In these cases, mesocyclones (the storm’s rotation that can lower to the ground as a tornado) quickly ramped up and ramped right back down. The big issue with these brief circulations are that the National Weather Service radar normally takes four to six minutes per scan. The reason being is that it normally scans 13-17 different tilts so you can get a picture of the entire storm. This is necessary so you can see the entire storm and see higher up, where hail develops along with rotation before lowering to the ground. In this case, the rotation simply developed and dissipated between the radar scans. You may be asking yourself why does this not happen with longer track tornadoes. Simple, longer track (significant) tornadoes normally have a stronger mesocyclone, which is normally developing 10-20 minutes before the tornado touches down. This tornado was also heavily wrapped in rain and was weak. Such brief tornadoes occurring are difficult to warn for because they normally only last for mere seconds. In addition, they are normally quite weak and comparable with straight line winds. This was a freak event that provided little lead time to the forecasters at the National Weather Service office in Austin/San Antonio. It’s likely by the time they had written up the warning and issued it, the tornado would have already come and gone. This is a good, but rare example showing that tornadoes can and DO occur with little to no warning in some cases.

I really hope this helped explain the process of detecting tornadoes to you. If you have any questions, shoot us a note on our Facebook Page or message us on Twitter (@TxStormChasers)! We’d be happy to answer any of your questions. Have a good night!