The Incorrect Tornado Warnings Near Baltimore Monday Afternoon

A pair of obviously false tornado warnings were issued by the National Weather Service (NWS) yesterday afternoon near Baltimore. I'd like to explain what went wrong and how it might be repaired going forward.

The first issue Monday was the Sterling NWS radar was being operated at 4-minute intervals. Four minutes is too infrequent when tornadoes are an issue. It should have been operated at 80- or 120-second intervals. 

First Tornado Warning

Here is the earlier tornado warning at 4:14pm EDT:

The "radar indicated rotation" was allegedly over Randallstown (underlined). The radar displays Central Time.

This first warning is represented by the northern (upper on the image) red polygon. There is no radar echo at all over Randallstown (left image). A tornado is not possible when there is no echo in a situation like this. The line of thunderstorms had pushed a gust front ahead of the line. On the right, the Doppler wind display indicates the wind is blowing from west to east but there is no rotation in the area. 

No tornado warning should have been issued.

Second Tornado Warning

The second tornado warning (southern or lower polygon) was issued 4 minutes later at 4:18. 

The warning was issued because, supposedly, there was a tornado "over Ellicott City" (underlined) that was indicated by radar (below). 

As above, there is no radar echo at all over Ellicott City. The gust front hasn't even moved that far east (right image). Again, there is no rotation whatsoever on the Doppler wind display. The brighter reds, in this case, are likely birds or insects being carried by the gusty winds. 

The tornado warning should never have been issued. There was no tornadic rotation indicated in either warning for the duration of each warning; just some "false echoes" known as "side lobes" in places. 

The State of the NWS's Tornado Warning Program

It is difficult to understand why these warnings were issued, especially at the times they were issued. My educated guess is that it is a training issue at the NWS. They closed their in-person radar and warning training school in 2004. The zenith of their tornado warning quality was 2000 to 2010. Since 2011, their own statistics reveal the serious issues with tornado warning accuracy since that time. By my figuring, the quality has deteriorated by about 40%.

Former NWS meteorologist Frank Alsheimer says,

I completely agree. A first step would be to reopen the NWS's radar/storm warning in-house training school. Beyond that, we need a National Disaster Review Board modeled after the hugely successful National Transportation Safety Board. Here's how it would work. 

A disaster review board would have the respect and clout (I hope) that enable it to constructively work with the National Weather Service on this far-reaching and completely necessary restructuring of its tornado-severe thunderstorm-flash flood warning system. 

Finally, if appropriately constructed, the review board would take over grading NWS tornado and other warnings which would remove that major conflict of interest. 

These changes cannot be made soon enough. I'm hoping Congress and the Administration will work together to make these changes.