The real difficulty with the current tornado warning system

[Andy Wehrle]

It has been my observation that the city of Stoughton sounds its tornado sirens whenever the NWS-Milwaukee/Sullivan issues a tornado warning containing a given path location within approximately the southeastern 1/4 of Dane County, as shown below (not known for fact that this is community policy, just a general observation of mine):

This has occurred five times since 2003:

May 10, 2003: No confirmed touchdown

May 30, 2003: No confirmed touchdown in SE 1/4, though earlier produced brief F0 (no property damage) touchdown near Verona.

May 23, 2004: Brief F0 (no property damage) touchdown near Albion (east of Stoughton near Jefferson County border).

June 23, 2004: No confirmed touchdown in SE 1/4, though earlier produced F1 damage on southwest side of Madison.

August 18, 2005: 1 dead, many injured, millions of dollars in damage, long-tracked F3 with 1/4 mile wide path.

The problem: That is one out of five events that actually got noticed by anyone outside of the "meteorlogical community". As far as the general public is concerned those first four warnings were all "false alarms".

What is the siren vs. NWS-warning policy in your community, either officially or observed in practice? How many of these warnings get chalked up as "false alarms" (that is, nobody outside of the damage surveyors and maybe 1 or 2 unlucky farmers is aware that a tornado actually happened)?

 

[Jeff Snyder]

Just as a note, I believe the current False Alarm Rate for tornado warning is about 75%. The implementation and deployment of the WSR-88D / NEXRAD system helped PoD quite a bit, but it hasn't improved FAR nearly as much. So, from if the county's policy is to blow the siren for all tornado warnings, then the 4-out-of-5 false warnings isn't too off par from what the average is.

I grew up in suburb south of St. Paul, MN... The county in which I grew up (Dakota Co.) was the only one in the area that blew the siren for EVERY severe thunderstorm and tornado warning. A marginal severe storm with 0.75" hail? Blow the sirens. A storm blowing over lawn furniture with 60mph winds? Blow the sirens. I'd be interested to see the percentage of people in Dakota county who ignore ths sirens, especially in comparison with surrounding counties which only blow the sirens for tornado warnings and particularly severe storms (winds >= 75mph). I've always found my county's policy annoying, since sirens should invoke a quick response (turning on the TV or radio to find out the specific threat, taking shelter in a basement, etc). By blowing the sirens for every marginally-severe storm, many of which pose a rather insignificant threat to life or property (or at least a threat that anyone do anything about -- it's good for farmers to know, but there's very little they can do about it), I can't help to think that makes the complacency issue that much worse.

I'm not sure Dakota county still has that policy, but it's always bugged me.

 

[Bill Tabor]

Jeff, perhaps they do it because there is a big concern for Farmer Brown out on his tractor to realize something is approaching and they don't want him to get whacked by large hail or blown by strong winds. In other words maybe the fact that some of these people are out in the fields and out of direct communication with other mediums helps motivate to blow the sirens more. Does that justify the actions and lack of consistency with other areas? Not sure - you'd probably have to take a vote in the community based on their level of concern.

Also thinking about FAR 75% for tornadoes seems like a lot. But if you think about it the bigger the threat to life and property, along with requirement for lead time to take shelter, combined with the unknown of complete accuracy of a tornado actually being present somewhat justifies higher FAR's as being acceptable. Of course this is implicit. I'm just saying that if for every four times you are warned one of those times will kill you then it still may be respectable and reasonable to get warned at that level if it saves your life or that of your family / friends. That said, this point of view doesn't diminish at all the desire, and need to reduce the FAR further.

 

[Damon Poole]

Actually, NWS offices are not allowed by policy to sound or operate outdoor tornado sirens. The communities themselves are solely responsible for making the choice to activate. Thus, NWS can only issue warnings and advise whether or not to activate. LZK makes this very statement each year during Severe Weather Awareness Week. NWS can activate the Emergency Alert System via NWR-SAME, thus this is their primary direct means to issue warnings.

 

[Joe Nield]

While we're on the subject of sirens, please please please help us get this fact out to the public.

Tornado sirens are NOT designed to be heard indoors. While you may or may not be able to hear the siren from your living room, don't count on this as your only warning source. Two dozen people died in Evansville, Indiana last November partially due to this problem. We've been hitting this hard in preparedness work and interviews since then, but we need to hammer the concept home.

 

[GPhillips]

I grew up in suburb south of St. Paul, MN... The county in which I grew up (Dakota Co.) was the only one in the area that blew the siren for EVERY severe thunderstorm and tornado warning. A marginal severe storm with 0.75" hail? Blow the sirens. A storm blowing over lawn furniture with 60mph winds?[/b]

Annoying wasn't it. I lived in an apartment complex in Eagan right next to a siren and that thing would wake me up in the middle of the night for nothing...usually.

I'm sure the policy came about as a CYA from a county official who had been burned, or was paranoid, and it has been continued on by people who are afraid to change the policy and then have something bad happen.

 

[John MacKay]

In Manhattan, the sirens go off for any tornado warning in the county (especially the southern half, but they will let it go for the northern half).

They generally do a good job here, sounding the tornado sirens twice on May 8, 2003 (even though the afternoon one was out of the county by the time the sirens went off since the storm had rotation as it left he county), once in August of 2004 (I think), and once last Thursday.

Even though there was no tornado warning for Riley County last Thursday, spotters reported a wall cloud in Geary County, and with their fast movement, it was probably the safe way to go, especially at rush hour when everyone out could hear them. Then again, people would get confused being in a car in a busy street with no tornado warning.

I know in my hometown, a wall cloud or tornado warning for a storm close by gives them the incentive to sound the sirens.

 

[Chris Geelhart]

While we're on the subject of sirens, please please please help us get this fact out to the public.

Tornado sirens are NOT designed to be heard indoors. While you may or may not be able to hear the siren from your living room, don't count on this as your only warning source. Two dozen people died in Evansville, Indiana last November partially due to this problem. We've been hitting this hard in preparedness work and interviews since then, but we need to hammer the concept home.
[/b]

I've been interviewed twice in the last week by Springfield, IL media. SPI is having a problem in getting all of its sirens functioning at one time as of late. Usually it's only about 5-10 (out of 57, I think) that malfunction at a time, but a couple were damaged by the initial tornadoes of March 12. Apparently it is nearly impossible to buy NWR receivers in SPI at this time, as basically everyone is sold out. I'm not sure if people are buying the receivers because of the siren situation or because of increased publicity in NWR since March 12. (I noticed on the ILX webpage stats, that the SPI NWR page only had about 75 hits in February, and nearly 850 in March.)

A personal observation on the sirens, if I have the windows closed at my house, I can only barely hear the siren from my bedroom (siren is NE of the house, bedroom is on the SW corner). I am certain I would not be able to wake up from the noise in that situation. I know that during our second tornado threat that night (circa 2 AM), the local sirens never sounded because there was no power.

Chris G.

Actually, NWS offices are not allowed by policy to sound or operate outdoor tornado sirens. The communities themselves are solely responsible for making the choice to activate.[/b]

I remember in the early 90's, when I worked at Sioux Falls SD, there was a device at the office that could be used to sound the sirens, but the NWS staff couldn't use it. In severe weather situations, a city official would be at the office to activate the sirens. I think there might have been a key that was used to turn on the box, which we didn't have.

On the original question (when are sirens activated in your area), it seems that here in central IL, it varies by location. Some sound it for all TOR's, some only for specific towns named in the TOR, some only if one of their spotters sees something (even if we have their town mentioned in the TOR).

Chris G.

 

[Nick Smith]

The policy for Hamilton County, Ohio (including the city of Cincinnati) is to sound a steady tone for five minutes, for tornado warnings and severe thunderstorms warnings that coincide with an existing tornado watch. Some of the northern Kentucky counties have a similar policy (Boone and Kenton). From June of 1986 until the end of 1991, the policy in Hamilton County was to sound the sirens for ALL severe thunderstorm warnings. This came after extensive wind damage in northern Kentucky and unwarned tornadoes in the surrounding areas on March 10, 1986 (severe thunderstorm warnings were in effect, as well as a tornado watch). There tended to be many false alarms with this policy and warnings for marginal severe weather (0.75" hail). The public did become complacent (8 severe thunderstorm warnings in 1991 with marginal severe weather) and the county decided to change the policy for 1992 (as noted above). It was influenced by the Civil Defense director, Meteorologist in Charge of the Cincinnati WSO and several representatives from the county. This policy may have been challeged a few times since, but it still remains in place today.

Since 1992, the system has been activated 17 times for severe thunderstorm warnings that coincided with a tornado watch (one time for severe t-storm warning without a tornado watch), for which 16 of those instances verified. The system has been activated 6 times for tornado warnings, for which one of those verified. There was also two activations where no offical warning was in effect.

The Cincinnati WSO did have controls for the Hamilton Co system (and may have for others) up until shortly before the office closed in 1995. Sometimes the sirens would come on before the warning hit NOAA weather radio and TV.

This year alone, the system has been activated four times for severe thunderstorm warnings. The first was on April 2nd and again on April 7th, when the sirens were sounded countywide three times in less than one hour. Severe thunderstorm warnings were in effect for different parts of the county at different times on the 7th. The county has the capability to sound each siren individually, but still chooses to sound them all.

 

[rdale]

"for which 16 of those instances verified"

You mean that

"for which 16 of those instances verified"

You mean that 16 times out of 17 that a SVR was issued, they had damage the equated tornado damage and worthy of the sirens? Or they had 3/4" hail and some 2-3" tree limbs down?

 

[Brandon Clement]

The NWS out of Jackson goes overboard with their warnings. I can be sitting at the computer and my weather radio goes off with a tornado warning in the eastern most parishes of La. I will pull the radar up and the cell that is being warned managed to make it all the way across Northern La without being warned but as soon as it makes its way into JAN's area the horns blow. I would be very suprised if the warning average is anywhere greater than 10%. I'm not sure if the reason they do this is because so many events in Ms happen at night combined with a large percentage of mobile homes so they take the extra precautions or if they are just horn happy.

Last March we had a huge squall line moving our way when we went to bed at 1AM. I knew the weather radio would be going off early in the morning and was anticipating a bunch of severe warnings. When the radio went off it was a tornado warning for my county and I just shrugged it off as another lame warning and went back to sleep. After an hour and about a dozen more warnings I got up and when I checked the radar and the squall line was gone. Instead we had sups firing one after the other and they were producing.

A tornado warning in Ks and a tornado warning in Ms are two completely different things. In Ks their is usually a few dozen spotters on any storm capable of producing and it is visible for many miles. In Ms there are usually no chasers and the storms tend to be at night in very bad terrain and many times they are wrapped. In Ks there is usually a tornado warning when a storm is about to produce or first puts a tornado down. In Ms a tornado warning is used for any storm showing shear at the low levels regardless of whether the storm is actually producing or even close to it. Many times their is a tornado warning for storms in the middle of a line. Lead time is very important, especially for people in mobile homes but when the false alarms happen so frequently is the person in the mobile home going to actually leave and go to a strong building for every warning? How many times does a person wake up to a warning, get their family together, leave in a huge rush and go to a building and see little more than rain before they ignore the next warning?

 

[Nick Smith]

"for which 16 of those instances verified"

You mean that

"for which 16 of those instances verified"

You mean that 16 times out of 17 that a SVR was issued, they had damage the equated tornado damage and worthy of the sirens? Or they had 3/4" hail and some 2-3" tree limbs down?
[/b]

I meant that there was either wind damage, measured wind gusts greater than 58 mph or hail 0.75" in diameter. There was few instances of measured winds greater than 75 mph or hail over 2" in diameter (which could be considered "significant severe").

There was one instance that I can think of where the sirens were not sounded (because the watch/warning system did not meet criteria), but wind damage over a relatively large area was equivilant to an F1 tornado.

 

[rdale]

"The NWS out of Jackson goes overboard with their warnings."

A quick check showed FAR of .81, national average runs about .75

 

[Damon Poole]

Here's an interesting story about how some tornado sirens being replaced, from KARK-TV in Little Rock. Independence County Arkansas is replacing its sirens with a volunteer group of callers to be equipped with NWR recievers.

http://www.kark.com/news/default.asp?mode=shownews&id=6149

 

[Alex Lamers]

"The NWS out of Jackson goes overboard with their warnings."

A quick check showed FAR of .81, national average runs about .75
[/b]

Had no idea that the national average was so high. Where are you getting these statistics?