Tuesday's Missed Tornado Warning in Kansas

Greenwood County tornado damage, KWCH TV News
It is not known which tornado produced this damage. 

We often discuss tornadoes that were improperly warned by the National Weather Service (NWS). 

We had a completely unwarned tornado in the Kansas Flint Hills region Tuesday morning but I don't fault the NWS for not anticipating it. 

They found the tornado path's (tornado #2, yellow path) when looking for a tornado that occurred earlier  that morning in Greenwood County. The tornado in question was an EF-2, which is rated "strong" intensity. 

As to why I don't fault the NWS, let's take a look at the radar data. 

The tornado touched down at 7:34am. The radar image immediately below was one minute before the start of the tornado and there is nothing remotely tornadic in the immediate area (circle).

The tornado has been on the ground for one minute at this point. Still, nothing. There is weak rotation to the SSW of the location but nothing that would say, "tornado."

At 7:37am, the tornado is still on the ground. I've removed the circle -- the location of the tornado -- and put arrowing showing the weak rotation. This rotation is so weak, I would not think "tornado" from looking at it. 

There is significant rotation here (tornado still on the ground) but, when compared to the previous frame, I would think, "I need to start watching this." But isn't nearly enough to issue a tornado warning. If we warned every time there was this level of rotation, the false alarms would be through the roof!

The tornado is still in progress and, if the green/red were touching, I'd be telling my Twitter followers ( @usweatherexpert ), "time to start monitoring this storm, you may want to get indoors and away from windows."

The tornado lifted one minute before this data. The green/red are still not touching in any significant way. I still would not issue a tornado warning and I agree with NWS Wichita's decision not to issue a tornado warning. 

But, the NWS is not totally blameless. They have refused to install a network of gap-filler radars even though such has been proposed multiple times (please see here and here). The NWS missed an EF-2

Baring, MO Tornado Damage. Weather.com

this past August that did considerable damage to the town of Baring, MO. Baring is 133 miles from the nearest radar in St. Louis. Baring is a typical rural town with an important difference: the BNSF Railway's "Transcon" (Chicago to Los Angeles) goes through the town -- one of the most important rail lines in the United States. Hazardous materials are occasionally transported. While not related to weather, we saw how serious freight derailments can be as a result of last year's in East Palestine, OH. In addition, Amtrak's Southwest Chief goes through the town.

BNSF Railway goes the extra mile for safety as it has an excellent commercial weather company issuing track-specific warnings. But, no one can issue consistently high-quality tornado or thunderstorm-generated wind warnings when the radar is 130+ miles away.

If the United States had a Natural Disaster Review Board, I have no doubt that this would be toward the top of its recommendations for the National Weather Service. When the original radar network was designed (late-1980's), we thought that seeing the middle of thunderstorms (which is what can be seen at long distances due to the earth's curvature) was adequate to alert meteorologists to the possibility of tornadoes. We now know that it is important to see lower in the storm and the radar network needs to be adjusted accordingly. 

A private sector company, Climavision, is installing a network of radars and is in the process of installing one in Humbolt, Kansas. That is ~30 miles from the path of Tuesday's tornado. I don't know much about their radars, but I am hoping they will prove useful in prevent these surprise tornadoes. 

NWS or private -- we need more radars!