Meteorology: Tornado or Side Lobes?

This past Friday, we had a great example of how a meteorologist can tell the difference between side-lobes (a false echo that mimics a tornado's circulation on radar) and one indicating a tornado is forming or in progress. I'm going to walk you through it so young meteorologists, in a similar case, won't make the mistake of mistaking side lobes for a tornado. 

This case was in north central Texas on February 2nd. I'm using the Abilene/Sweetwater WSR-88D and the software is RadarScope. When I draw on one panel of the screen, it shows up on the other in the same place, so the measurements are about as exact as any in meteorology. 

The Thunderstorm Cluster, 4:24pm

Above is a cluster of thunderstorms with the two storms with arrows starting to transition to supercells. We already have the northern storm (just south of Aspermont)  moving north northeast. In a situation like this, the storm will likely be "a hailer" -- meaning it is likely to produce hail, possibly large, -- but a tornado is unlikely. 

The southeasternmost storm, just north of Hamlin, is moving to the "right" (straight northeast) of the others. While large hail and, maybe, damaging winds may occur, it is this type of storm which is more likely to produce a tornado. Meteorologists call this type of storm a "right-mover." Before we had Doppler wind data, the storm would be labeled suspicious because of its movement. 

4:44pm, Beginning of a Hook Echo

Twenty minutes later, we have the very first sign of a hook echo forming with a "weak echo region" (little or no reflectivity ahead of the hook and, aloft, a bounded weak echo region [not shown here]). The possibility of a hook ratchets up the danger. However, at this point, there was little low-level rotation. 

4:53pm, Side Lobes Appear

The "hookish" feature is wrapping around the mid-level rotation (left, circled). At right is the Doppler wind data. There is decent, abrupt rotation but the center (circled) is not inside of, or coincident to, the developing hook. 

At this point, I began preparing to notify my Twitter readers as to the threat. 

4:57pm, Hook and Tornado Likely Forming

The original "left-mover" is north northeast of Aspermont with its hail core (lighter colors) just to the right of the U.S. 83 shield. The right-mover has a full-fledged hook and, before Doppler wind data, I would have issued a tornado warning for my clients at this point. But, knowing from research that the ideal amount of warning is 13-15 minutes, I held off. The National Weather Service (NWS) has a severe thunderstorm warning (yellow polygons) for both storms. 

4:57pm, Wind Velocity Data Added

The reflectivity data (left) is the same as the image immediately above. The rotation is more or less stationary while the right-mover's hook is about to be coincident with it. I've put a blue line at the center of rotation at right, which the software carries over to the left image. The center of rotation is just outside of the hook. 

5pm -- Time to Issue a Tornado Warning

Classic hook at left. Classic tornado "couplet" at right. Again, there is a blue line at the center of rotation and the center is within the circulation indirectly denoted by the hook. 

There were two tornadoes, the first of which touched down at 5:13pm -- thirteen minutes' lead-time. Storm chaser Brian Scholl has images.

I put out my version of a severe thunderstorm warning (NWS, as shown above, already had one) and, after hitting "send," was quickly typing my version of a tornado warning when the NWS sent one a few seconds later. I prefer to use NWS warnings so there is no confusion on the part of the public, so I used their warnings until the situation calmed down.

The National Weather Service did a perfect job with this storm providing ideal warning on both tornadoes. 

To summarize,

• Right mover,
• Hook echo, 
• Rotation with a hook.

That is really all you need to know to issue a tornado warning. At this point in the state-of-the-art, making it more complex doesn't really improve accuracy. And, please keep in mind there is such a thing as too much lead time. 

If you have a question, just post it in the comments.