Thoughts on This Week's Tornadoes

It has been a bad two days for severe thunderstorms and tornadoes as they have occurred from Texas to Kansas to Ohio. Per news reports, three people were killed by tornadoes in Ohio yesterday. I want to pass along some thoughts to our readers. 

Tornado Forecasting

Tuesday afternoon, I posted the forecast below. The base forecast came from the National Weather Service's Storm Prediction Center (SPC). However, I was concerned there was a risk of tornadoes farther west in Kansas. So, I added the dashed lines. 

There were two EF-2 "strong" (strong defined as EF-2 or -3 by NWS) tornadoes in Kansas as indicated by the red lines (per survey by the National Weather Service's Topeka office). 

In the lead-up to the storms, it looked like that area would be affected most adversely. There is a relative new short-term rotation forecasting tool created by NOAA that looked that it was accurate a couple of hours before the storms began. So, I posted it here. 

The pink = where rotating thunderstorms were most likely -- with the darker areas the higher probabilities. 

This is how it came out. 

The radar of the tornado near Alma showed a classic "hook echo" that may be used in radar meteorology classes in the future. 

And, if you look at all of the reports large hail (up to softball size in places), damaging winds and tornadoes, it did quite well. Red = tornado reports; blue = damaging winds; green = large hail. 

The problem is that this system is so computationally demanding that it can only cover limited areas at a time. Yesterday, when it would have been extremely valuable, it was set to Arkansas (where the SPC thought the worst storms would be) rather than Indiana and Ohio where the worst tornadoes actually occurred. They also test it only on days when the threat is dynamic rather than subtle. The latter is often the more difficult to forecast. It would be beneficial if NOAA could find the computational resources to run it over the U.S. east of the Rockies daily at full resolution so we could see how good it actually is. 

This Blog's Original Forecasts and Warnings

Per the statistics, we've had thousands of new readers this week. I wish to emphasize that the content is based on hours and hours of work and that, as above, many or most of the forecasts and storm warnings are original. 

For example, there was a tornado in Frisco, Texas, yesterday evening. 

Per the NWS, there is video of the tornado on social media. 

At 6:26pm, I forecast the possibility of a tornado with the thunderstorm moving toward Frisco with large hail also forecast to occur. 

At 6:26pm, the above was posted on Twitter. You'll note there is no official tornado warning (red polygon).  

The official tornado warning came out twenty minutes later. 

Please allow me to point out that this NWS warning (based on the tentative time the tornado touched down of 7:05pm) provided 9 minutes of advance warning. 

The NWS issued an adequate tornado warning based on what we currently know about that tornado; that is not my point. My point this blog provides original content. 

To all of our new readers, welcome!! If you have friends interested in weather or who need "heads up" weather information, please send them our way.