The Tornado Warning Mess -- National Service Replies

Since this blog was created in 2009, we've looked at dozens of storms across the United States. In numerous cases, I have praised the work of the National Weather Service (two of many examples, here and here), especially in the early years of the blog. More recently, we have expressed considerable alarm at the deterioration of the NWS's tornado warning program. 

In 2022, on two occasions, I made my concerns known to the then-director of the National Weather Service. The replies were of a patronizing "party-line" nature -- "everything is fine, no cause for concern." 

After the multiplicity of unwarned tornadoes in the last nine months of 2023, I decided to try again. This time, I sent the letter to Dr. Richard Spinrad, the Administrator of the National Oceaning and Atmospheric Administration (NOAA) which is the parent agency of the NWS. A copy of the letter is below.

I apologize for the edit marks, I can't seen to remove them.
They were not on the original.

In addition to the letter, I provided an appendix that included my missed tornado warning blog write-ups since April 1.

List of Problematic Tornado Warnings Since April 30, 2023

(in reverse chronological order)

·       https://www.mikesmithenterprisesblog.com/2023/12/no-advance-warning-for-fatal.html (fatal tornado)

·       https://www.mikesmithenterprisesblog.com/2023/12/it-happened-again-strong-tornado-on.html

·       https://www.mikesmithenterprisesblog.com/2023/09/dont-count-on-nws-to-warn-you-of.html

·       https://www.mikesmithenterprisesblog.com/2023/08/another-tragic-poorly-warned-tornado.html (fatal)

·       www.mikesmithenterprisesblog.com/2023/08/another-giant-tornado-warning-miss.html

·       www.mikesmithenterprisesblog.com/2023/08/tornado-warning-miss-3-this-weekend.html  

·       www.mikesmithenterprisesblog.com/2023/08/the-second-missouri-tornado-warning.html  

·       www.mikesmithenterprisesblog.com/2023/08/another-shocking-tornado-warning-miss.html  

·       www.mikesmithenterprisesblog.com/2023/07/strong-tornado-without-advance-warning.html  

·       www.mikesmithenterprisesblog.com/2023/07/another-bizarre-nws-tornado-warning.html

·       www.mikesmithenterprisesblog.com/2023/07/when-did-nws-decide-to-get-out-of.html

·       www.mikesmithenterprisesblog.com/2023/06/matador-another-fatal-tornado-without.html (fatal)

·       www.mikesmithenterprisesblog.com/2023/06/again-fatal-tornado-strikes-perryton.html  (fatal)

·       www.mikesmithenterprisesblog.com/2023/05/fatal-texas-tornado-struck-without.html  (fatal)

·       www.mikesmithenterprisesblog.com/2023/05/virginia-beach-ef-3-another.html  

·       www.mikesmithenterprisesblog.com/2023/04/there-is-exceptional-number-of.html

Since the letter was written on December 13, there have been six additional tornadoes with problematic warnings. And, there has been an additional write-up about a tornado that was unwarnable (by anyone) with the current state of meteorological knowledge and technology. I have been hoping that this increase in the rate of missed tornadoes is a fluke rather than an extension of the trend. 

Today, I received the reply from Ken Graham, the Director of the National Weather Service. In fairness to the NWS, I am reproducing the letter in full. 

I really appreciate the time taken by Ken, Dr. Spinrad and the others that had input into the reply. I have several comments about it.

• National Weather Service tornado warnings were excellent 15 years ago. They've decreased rapidly in accuracy and lead-time (advance notice) since 2011. While Ken acknowledges that decrease, no reason or plan for repair is given.
• False alarm rate: From 2016 to 2020 (latest figures available), the average false alarm rate (FAR) was 69.4 percent. In 2022, the FAR was 66%. That represents an improvement of 5%. I question the whether the public can perceive a difference that small. Plus, is the 5% decrease in FAR worth sacrificing 27 points of tornado warning accuracy? Or sacrificing 20% of the lead-time (advance notice) of the arrival of a tornado? My sense is "no," but others may disagree. 
• While mass casualties generally occur with EF-3 to EF-5 intensity tornadoes, deaths occur in all tornado intensities. If you were a loved one of the two women killed in the poorly warned Perryton Tornado (above), do you think you would care it was an EF-2 rather than, say, an EF-4? Here is an example of an F-1 tornado that killed 16 people. 
• Where are the statistics they reference? As we have discussed before, the page that used to have all of these stats is now behind a login and password. They are a government agency and, as such, their data should be widely available -- good or bad. 
• I take Ken at his word regarding the training NWS meteorologists receive before they "solo" on the tornado warning desk. However, that conflicts with first-person reports from three NWS meteorologists (one active, two retired) who have gotten in touch with me. 
• I appreciate the tornado and severe thunderstorm watches and other information referenced but none of those cause people to scurry to the basement or crawl into a bathtub to save their lives. That is why the tornado warnings are critically important. 
• I understand why they cannot comment on a bill before Congress. However, if my Natural Disaster Review Board (NDRB) becomes law, they will take over validation of NWS warnings and all of that data will be readily available. 

There is good news on one front. Earlier this week, I wrote about the Little Rock tornado warning false alarm just after midnight on January 12th. That tornado warning was caused by misinterpreting a type of radar issue caused "side lobes." Today, I learned the NWS had already scheduled seminars on January 24 and March 27 on side lobe interpretation. Hopefully, those will make this type of false warning less likely in the future. Well done, National Weather Service!

I am encouraged Ken Graham's reply because it is obvious they take this issue seriously. I've already had some post-letter communication with the NWS and it is my hope we can make progress with regard to tornado warnings.