Fatal Texas Tornado Struck Without Warning

Tragically, it has happened again. The National Weather Service has failed to provide a warning on a tornado that ended up killing two one and injuring at least ten in south Texas. This continues the trend of deteriorating NWS tornado quality experienced the past ten years. 

Update at 9:40pm. According to local officials, just one fatality occurred. They say that none of the injuries is life-threatening. That is very good news. 

They also said that the tornado began at 4:06am. I'm quite skeptical of that because of what the radar shows plus the NWS Storm Prediction Center's log said 4:01am -- which lines up with the radar. (see below). The time in Greenwich Mean Time (0901 = 4:01am CDT) is below at far left. 

The head of the local National Weather Service said there was "zero" warning. None of my write-up needs to be changed based on the NWS's comments at the news conference. 

There is no question that some tornadoes are not "warnable" in adequate time, but that was certainly not the case here. The signs were there more than 30 minutes in advance. 

In addition to (for meteorologists) an 0 to 1km EHI of 1 an hour before, a classic hook echo (ironically, discussed on this blog yesterday evening, scroll down) appeared a half hour in advance. 

While a supercell with a hook echo is not conclusive, it is certainly a sign that the storm should be closely watched. In addition, the radar which was collecting data at five minute intervals, should have been placed in 80 second "tornado" mode. The storm was moving north northwest. Laguna Heights is highlighted at upper left. 

At 3:51am, there was a hook echo with increasing rotation. The storm was still moving north northwest. Had I been on the warning desk, I would have seriously considered issuing a tornado warning at this point.

Unfortunately, the radar was still set at 5 minute intervals. By the next data period at 3:56, there was no question a tornado warning should have been issued. There was strong, and increasing, rotation along with a hook echo. 

The NWS finally issued a tornado warning at 4:04am, it was based on an obvious "tornado vortex signature" visible at 4:01am (below). That signature might have been seen earlier -- allowing for an earlier warning -- had the radar been set to tornado mode. Unfortunately, by 4:04, it was too late. Per the NWS Storm Prediction Center's tornado report log, the tornado occurred at 4:01. 

The tornado warning is below.

By 4:05am the tornado's circulation was weakening. 

Given the time of day, it is possible that people would have been seriously injured even with a timely tornado warning. However, with apps, weather radios, WEA or StormWarn it is possible to instantly warn people in time to take action. That was not possible in the past.

Just two weeks ago an unwarned EF-3 occurred in Virginia Beach. And, the day before, in Palm Beach Gardens. 

There is no question National Weather Service tornado warnings have seriously deteriorated in quality during the past decade. Based on my discussions with them, the senior management of the NWS is in denial. The only way to fix this is for Congress to establish a National Disaster Review Board that would be modeled after the hugely successful National Transportation Safety Board. 

Given the large number of these dangerous non-warning events, I recommend going to your congressperson's web site and emailing them to recommend the National Disaster Review Board be established as soon as possible. 

© 2023 Mike Smith Enterprises, LLC