NWS Joplin Service Assessment is Out

[Mike Bennett]

There seems to be a lot of emphasis, at least in this discussion, on warning sirens. In the area where I live, (northern Oklahoma) the public are told to not rely on sirens. The siren warning system is primarily a system to warn the public who are outdoors at the time. Depending on how close you live to a siren, you may not be able to hear them if you're indoors. I live relatively close to a siren (probably 1/4 to 1/3 of a mile) and you can barely hear it if the TV is on or it is already raining or thundering outside so my wife says. I'm never there because I am always out chasing. I have however been home during siren tests on a clear day and it doesn't take a lot of ambient noise in the house to cancel it out.

 

[rdale]

I won't dispute what you are saying. However, to me, this can be addressed in the context of the current paradigm by changing bad practices not the introduction of more warning types.

Amen...

There seems to be a lot of emphasis, at least in this discussion, on warning sirens.

Because people who live in areas with sirens rely heavily on them. Plus they were poorly utilized in Joplin (sounding for non-warned storms just to be safe, sounding outside the polygon, etc.) which hurt things.

 

[J Tyler]

One problem we have in my home town is, they took out the old Thunderbolts, and replaced them with sirens that sound just like firetrucks, or police cars, or ambulances, or any other siren. The old Civil Defence Thunderbolts had such a distinctive wail to them that you KNEW it was the end of the world when you heard them. The Thunderbolts don't sound like ANYTHING else. When I hear the weekly test now, I always think it is a firetruck.

 

[Mike Smith]

As the NWS SA notes, the local EM sounded the sirens for SVR and TOR in JLN for a warning that included any part of Jasper Co.

This is not the only case of gross overuse of sirens. In St. Louis Co. they sound the all sirens when there is a TOR for any part of St. Louis Co. and if there is a TOR in an adjacent county. Here is a news story on the topic: www.kmov.com/news/local/Could-a-mas...-sirens-do-more-harm-than-good-122759099.html The EM says, at 1:03 in the story, "you never know what path a storm is going to take."

It is difficult to imagine that so many EM's are not aware of polygon warnings. EM's are, to one extent or another, political creatures. I believe they are focused on "better safe than sorry" without realizing that they themselves are creating situation where they will be "sorry" about the results: People not taking shelter due to siren fatigue.

 

[cdcollura]

Good day all,

The Joplin storm was something to be reckoned with to say the least.

The tornado-genesis phase of this supercell took mere seconds ... I was on that storm, and saw the backside of the meso wrap up (then rain wrapped) phenominally fast. This storm had a LOT going for it in terms of dynamics.

Even more disturbing was the time from T-Genesis to EF4+ intensity, like 4 minutes or so. These dynamics are very low-level, and may be missed between the volume scans (6 minutes) and / or too low in the storm to even be picked up at all. This is where better technology in detecting such things becomes paramount.

Extremely rapid tornado-genesis like this is something that definitely deserves attention and farther study.

On an even more disturbing note ... Complacency (?) ... How many folks ignored the warnings and possibly looked out at a rain-wrapped tornado sky and ASSUMED there was "Nothing In There"?

There is a pretty amazing video taken only a mile AHEAD of the Joplin Tornado and it appears more like an HP storm (the violent tornado is completely hidden). See below...

http://youtu.be/CburjPYmSdo

I wonder how many Joplin Residents (now either dead or disabled) thought it was "just rain" ;-(

 

[MClarkson]

A legitimate question is, "Should the NWS be concerned with capturing every perceived tornado with a tornado warning?" my answer is no. Focus on the tornadoes that have the potential to persist longer than the time itntakes to craft the warning.

Agreed. Not only are warnings being issued for every TVS, when we all know some are way more threatening than others, but the warning polygons are more than a little wider than they need to be... tornadoes rarely change direction as much as some of these warnings indicate. Bottom line, you have to be willing to accept a few false negatives or you are going to have a massive amount of false positives.

I think severe warnings for the flanks of tornadic storms and for weak TVS is good enough, and save tornado warnings for the core of strong storms. Make it a warning that actually means something. I don't know how much it would have helped in this case, as this storm rapidly intensified just before a population center, but in general saving tornado warnings fore exceptional circumstances seems like the most accurate method.

 

[Clarence Bennett]

As the NWS SA notes, the local EM sounded the sirens for SVR and TOR in JLN for a warning that included any part of Jasper Co.

This is not the only case of gross overuse of sirens. In St. Louis Co. they sound the all sirens when there is a TOR for any part of St. Louis Co. and if there is a TOR in an adjacent county. Here is a news story on the topic: www.kmov.com/news/local/Could-a-mas...-sirens-do-more-harm-than-good-122759099.html The EM says, at 1:03 in the story, "you never know what path a storm is going to take."

It is difficult to imagine that so many EM's are not aware of polygon warnings. EM's are, to one extent or another, political creatures. I believe they are focused on "better safe than sorry" without realizing that they themselves are creating situation where they will be "sorry" about the results: People not taking shelter due to siren fatigue.

Mike is absolutely correct in this last paragraph, but how do we bridge the gap between NWS and EM? It is this way in middle Tennessee and is maddening. I have discussed this before in another thread. The public has been conditioned over a long period of time because of the behavior of EM's and their handling of sirens. The average person may not know that the person making the decision to sound a tornado siren has nothing to do with the person issuing the storm warning, so who do they believe? This problem is much more complicated due to politics on both sides...in a perfect world, the sirens would be tied into the warning system and if a polygon encompasses an area where sirens are present, they are sounded as soon as the warning hits the system, cutting out the human element on the EM side.

 

[Rich Long]

Clarence I agree whole heartedly. We are trying to get sirens in La Vergne but there is a good public outcry right now not to do it because they have seen how they have been ignored other places and view it as a waste of money. I just did a post about it on my blog (www.box100weather.wordpress.com) but the policy has to change in their usage and public education has to be very aggressive. Remember the public views sirens as outdated in some areas. I just don't see how to fight that perception effectively until the false alarm rate goes down.

 

[Frank Kienast]

On a slightly different but related subject, it seems to me that there has been some improvement in not "calling wolf" when it comes to hurricane forecasts. I remember hurricane Gilbert 25 years ago. I kept watching the zone forecasts for southern Texas, and the predicted conditions were always consistent with a direct hit, despite the fact that the most likely track was well south into Mexico. I remember thinking how no one was likely to pay attention to the next warning given how unrealistic those local forecasts had been. I remember reading similar forecasts for Norfolk VA back in the 1990s.

By contrast, the forecast for Norfolk with hurricane Irene was much more realistic. Headlines read "Tropical storm conditions expected with hurricane conditions possible". The wording for winds was something like "Winds increasing to 60 to 85 miles per hour. Winds MAY gust to over 105 miles per hour". That's a lot better than blanket predictions of 120 mile per hour winds like I used to read with similar storms. It lets people know the danger is there, but not necessarily imminent.

 

[rdale]

It's still a problem... Even Irene never came close to having winds observed that were anywhere close to being advertised. I haven't seen any evidence that it was even remotely close to being a hurricane when it hit NYC.

 

[MClarkson]

Irene was the worst forecast from the media point of view that I can remember in my life. I didn't think it was possible to overhype a weak storm so much. The models handled the storm really well too... it was purely a function of hype and headlines.

 

[Elaine Spencer]

Getting back to our topic ... check out Andrew Revkin's "Dot Earth" blog in the New York Times for the Sept. 21 entry titled "Unheeded Tornado Warnings Just One Lesson From Joplin" (can't copy the link so just google the title)

This paragraph really grabbed my attention:

"The fatality total in Joplin in one way is not out of line with other EF-5 tornadoes, when you consider fatalities per building damaged or destroyed. In EF-5 tornadoes between 1997 and 2010, a total of 8,425 buildings were damaged or destroyed, and 110 fatalities occurred in buildings in these tornadoes. In Joplin an estimated 7,500 housing units and 500 businesses were damaged or destroyed, and according to the Storm Prediction Center, 133 fatalities occurred in buildings."

In other words... the Joplin tornado killed more people and destroyed almost as many buildings (about 8,000) as all F-5/EF-5 tornadoes in the U.S. combined during the previous 13 calendar years! (There were 6 such tornadoes during the period cited, including Jarrell, Bridge Creek/Moore and Greensburg.)

Since these figures exclude the other EF-5 tornadoes recorded in 2011 (there were 5 others, four from the Dixie Outbreak and one 2 days later in El Reno OK) I wonder if the "fatalities per building" figure in Joplin would still hold up if those were included. I suspect it would.

 

[rdale]

Until 2011 there were only 2 F5's in the last 10 years and neither impacted the population base of Joplin's size, so I'm not sure how valid that comparison is?

 

[Dean Baron]

"The fatality total in Joplin in one way is not out of line with other EF-5 tornadoes, when you consider fatalities per building damaged or destroyed. In EF-5 tornadoes between 1997 and 2010, a total of 8,425 buildings were damaged or destroyed, and 110 fatalities occurred in buildings in these tornadoes. In Joplin an estimated 7,500 housing units and 500 businesses were damaged or destroyed, and according to the Storm Prediction Center, 133 fatalities occurred in buildings."

In other words... the Joplin tornado killed more people and destroyed almost as many buildings (about 8,000) as all F-5/EF-5 tornadoes in the U.S. combined during the previous 13 calendar years!

I understand where he's coming from and I see his point, but Joplin was obviously a completely different situation than anything experienced in the previous 13 years (excluding Moore, OK). I don't know if that comparison can directly be made. I dont know how to explain it. I guess the best way is to say an EF5 ripping apart the heart of a city with a population of 70,000 is different than a few EF5 hitting/missing a few towns over the period of 13 years. Obviously Moore, Greensburg, and Parkersburg were significant and devastating, and I dont mean to take away from that. Maybe a better way to put it is I bet the death toll would have been significantly higher higher in the Greensburg or Parkersburg tornadoes had the respective city's population been that of Joplin's.

 

[hazelmaryjackson]

The problem remains how to warn the public so that they take notice. All these years everyone has been saying, if we give people more warning they will seek shelter. Not always they won't. OH, who has never been in a tornado, says that he would probably respond to a warning by going outside to see if he could see it. If he could not, he would assume it was an over reaction. Sad but true and I suppose typical of a great many people, even in regions where tornadoes are not uncommon, like Missouri. (There is also a great reluctance to spend public money on storm shelters - see the response to the local newspaper when a Joplin resident proposed a storm shelter only a few weeks earlier because she felt their town was at risk.) What I found interesting about Joplin is how some residents clearly and deliberately ignored all the warnings, it wasn't a case of not hearing them they ignored them. There are also disputed accounts about how much warning notice was actually issued inside the hospital, it does not seem to have been 20 minutes. Again they seem to have been relying on seeing it first.

How do we change this mindset?