The Emergency Management and Storm Warning Mess in St. Louis Friday, May 16, 2025

[Mike Smith]

You wouldn't believe all of the things that went wrong in STL Friday. The reason we didn't have another JLN is because of the highly knowledgeable TV meteorologists rather than the beginners on weekend duty on May 22, 2011.

Issues With the Forecasts and Warnings of the 2025 EF-3 St. Louis Tornado

It is time to face facts: perhaps we should no longer consider the government (NWS + FEMA) to be the most essential agencies for storm warnings, at least in major markets staffed with well-qualified meteorologists. I didn't include in my report that TV meteorology cars staffed with mets or qualified storm chasers) or helicopters in other markets, were reporting on the minute-by-minute location of the tornado -- something the NWS didn't come close to doing.

As usual, the WFO wasn't operating the radar in SAILS tornado mode (80-second). I don't understand why this is such a big issue. The FAA radar was more valuable to the well-trained.

Finally, people should be fired in STL emergency management! You'll see why when you read the report.

 

[Matthew E Herbert]

LSX put out a warning before it touched down. It was upgraded to a PDS as it moved into STL City proper. The real problem was the emergency manager that wasn't at her post on a day when a moderate risk was issued and a tornado watch was in effect. On a busy Friday in a big metro when people are outdoors is when the outdoor warning sirens were needed most

 

[Mike Smith]

It was upgraded to a PDS as it moved into STL City proper.

Matthew, I can't find a message that upgrades it to PDS until it was past St. Louis and over Venice, Illinois. Can you give me a reference to the earlier message to which you are referring? Thank you.

 

[John Farley]

The latest from the Post-Dispatch. You can't make this stuff up. Horrible that people died because of it.

Tornado siren button was broken in St. Louis fire department. Unclear if it was pushed.

Mayor Cara Spencer said the city learned in a test of the system on Tuesday that the button, one of two that turns on the sirens, wasn't properly functioning.

www.stltoday.com

 

[John Farley]

Matthew, I can't find a message that upgrades it to PDS until it was past St. Louis and over Venice, Illinois. Can you give me a reference to the earlier message to which you are referring? Thank you.

Here it is, Mike:

IEM :: SVS from NWS LSX

SVS issued by NWS LSX at 16 May 2025 19:45 UTC

mesonet.agron.iastate.edu

 

[Mike Smith]

John,

Thanks so much!

The problem is that the statement contains factually incorrect information. The original warning and this statement says the tornado was moving "east" when it was moving northeast. We know that in JLN people died because they believed the tornado was going to miss the city based on the NWS three times saying the tornado was moving northeast (out of SE KS) when it was moving straight east at JLN.

If you were listening to KMOX Radio and hearing the tornado was in Richmond Heights "moving east," you would not think you were in danger in North City or North County.

These aren't the "quick spinups" the NWS too often blames for warning misses. This was a supercell thunderstorm with consistent rotation and direction of movement all the way southwest to Des Peres.

Finally, saying the tornado is over "St. Louis" was too vague. "North St. Louis near O'Fallon Park" is a far better description.

And, finally, with regard to the 2:50pm statement, the tornado was just NW of Granite City -- well north of the "Venice" location in that statement.

I simply don't understand. These errors have nothing to do with being shorthanded.

Mike

 

[John Farley]

I agree with most of what you say here, Mike. The one thing I would note is that their list of places impacted DID include most of the places that were impacted:

"Locations impacted include...
St. Louis, Florissant, University City, Granite City, O'Fallon,
East St. Louis, Hazelwood, Collinsville, Edwardsville, Webster
Groves, Ferguson, Fairview Heights, Overland, Clayton, Jennings,
Glen Carbon, Bellefontaine Neighbors, Troy, Berkeley and Richmond
Heights."

That is, however, an overly broad list (CYA?) that also includes a lot of places that were NOT impacted. And some that had been impacted but were no longer in danger by then.

I do think it is a social science question whether people pay relatively more attention in warnings to the direction of movement or to the list of places impacted. Some of both, I would be pretty sure, but my guess would be more so the list of places. There may be some studies on this, have not looked at the relevant literature recently.

 

[Mike Smith]

I do think it is a social science question whether people pay relatively more attention in warnings to the direction of movement or to the list of places impacted.

In Wichita, the list of locations is not used. I saw my first TOR in KC this past Monday night (we moved in September) and they didn't use the list of locations.

EAS isn't used in ICT, OKC or KC.

The NWS is off base in the way words and publishes tornado warning texts. There was considerable criticism of the TOR messages during JLN (including one nearly 300 words in length!) and the NWS responded by saying, "We're hiring social scientists to fix them!" Millions $$ later, tornado warning messages are as bad as ever.

 

[Mike Smith]

This video, which shows the tornado in progress (

) also transmits the warning with the time-of-day at lower right. It proves there was no lead-time with this tornado warning, unfortunately.