Thursday, March 1, 2012

Interesting Look at Rotation on Tuesday Night / Wednesday Morning's Storms...


The above image was provided by the NOAA Environmental Visualization Laboratory (EVL). It shows the output of a computer algorithm which attempts to identify where rotation in a storm (as indicated on area radars) may be strong enough to produce a tornado on the ground.  The algorithm uses a combination of wind and hail data to arrive at its conclusions.  This particular image shows the output of that data for the storms of Tuesday evening (February 28th) and early Wednesday morning (February 29th) of this week.

I have taken the same image and added the locations of the 4 cities that were hardest hit by tornadoes that night:  Branson, MO, Buffalo, MO, Harveyville, KS and Harrisburg, IL (click to enlarge):


A scale isn't provided, however the lightest shadings indicate the weakest rotational/tornado signature, and the darkest indicate the strongest signatures.

As you can see, the storm responsible for producing the Branson area tornado actually began rotating as it moved through Tulsa, and continued doing so all the way across the southern edge of Missouri and into extreme southern Illinois and extreme southwest Kentucky!

The storm responsible for the EF-4 tornado in Harrisburg, IL began showing strong rotation over southcentral Missouri (to the Northeast of the Branson storm).  

Both the Buffalo, MO and Harveyville, KS storms also showed relatively strong signs of rotation on radar well before they actually produced tornadoes on the ground (yet they too did not produce a tornado for the entirety of the rotational cycle).

Herein lies the challenge to the meteorological community:  lots of severe storms rotate, but not every rotating storm produces a tornado on the ground.  To further complicate matters, we have yet to establish any type of rule of thumb that says something like "a storm rotating at such and such velocity will generally produce a tornado 90 minutes after it starts rotating..."

This problem is by no means new, yet it is one that needs to be solved quickly in order to improve the current tornado warning system.  Until then, you are likely to continue to feel "overwarned" from time to time as meteorologists struggle with when to issue a tornado warning based off of radar with no ground truth data available to support the decision.

One step in the right direction is the ongoing upgrade of existing NWS radars to dual polarization technology.  You can see my post here for more details on what that technology promises to bring.

The Springfield, MO radar (which is the site nearest Branson) had the new upgrade in place prior to Tuesday's storms.  The output from that radar clearly showed a strong signature of tornado debris as the storm moved through Branson on Tuesday night:


Branson is located near the center of the image.  The dark blue and light green pixels indicate the presence of the debris associated with the tornado that was ongoing at the time the image was taken.

If you've been reading the blog for very long, you know that it's not unheard of for a debris ball to show up on radar (even before the upgrade).  What is unusual about the above image taken by the new technology on Tuesday night is that the Branson tornado was "only" rated EF-2 intensity.  In the past, with the old technology, a debris ball would not typically show up with a tornado that "weak" (most of the tornado debris ball signatures we saw last year with the old technology were EF-4 intensity or higher).

The ability to potentially detect the debris signatures of weaker and/or shorter lived tornadoes is a very promising advancement, indeed!

Several of the radars in the areas to be affected by severe weather tomorrow have recently been upgraded to the new technology as well.  I plan to monitor that new data tomorrow evening and will be sure to post any interesting findings...


If you enjoy reading 'The Original Weather Blog', please be sure to "like" our facebook page!

No comments: