The Warning Process Must Get Better

| June 8, 2011 @ 9:03 pm | 191 Replies

Now that I have had a chance to catch by breath after the historic April 27 tornado outbreak across Alabama, time to jot down a few important thoughts…

*I firmly believe apathy and complacency due to a high false alarm ratio over the years led to inaction in many cases that could have cost lives.

The FAR (false alarm ratio) for many NWS offices when it comes to tornado warnings is in the 80-90 percent category. I say this is simply not acceptable. Sure, the POD is excellent (probability of detection), but if most of the warnings are bad, then what good is a high POD?

I ask the NWS to consider stopping the use of tornado warnings when trying to catch small spin-ups within a squall line (or QLCS). These tornadoes rarely last more than a few minutes, and are next to impossible to detect in advance. And, in most cases, the greatest damage from a QLCS is from widespread damaging straight line winds, not tornadoes.

These kind of warnings force us to go on the air for 40-45 minutes, often after tornado signature has vanished from the radar. Sirens sound, the NOAA Weather Alarm goes off, severe weather apps on smart phones alert users. Getting these kind of warnings over and over and over again totally create an ocean of people that won’t be paying attention when a real tornado emergency is in progress.

I heard it over and over as people described their April 27 experience. “I hear those sirens all the time, and nothing ever happens”. The cry wolf syndrome is very real, and very dangerous.

*Too many people believe they should hear a siren before a tornado strikes.

I think the time has come to take them down. Sirens are not efficient, reach a limited number of people, and can’t be heard in most homes, schools, and businesses. And, in most counties, the sirens don’t sound only in the warned polygon, they sound county wide. In some cases, this means you are hearing a siren when the actual tornado threat is over 40 miles away.

Sirens were born during the Cold War with the Soviet Union in the 1950s and 1960s… their time has come and gone. If the sirens are taken down, then you KNOW you won’t hear one next time there is a tornado threat.

Most southerners still have the “siren mentality”, and that no doubt killed people April 27.

*NOAA Weather Radio must be upgraded to the polygon warning system soon, or it will become obsolete.

Sure, it is the best thing we have now, and I still promote it heavily. But, why hasn’t NOAA upgraded their system so the receiver manufacturers can produce models with GPS included so they sound only when the receiver is a in a warning polygon? If something doesn’t change soon, the private sector will be the ones that push the warning process into the new technological era.

The best model I see now for the future warning device is the WeatherRadio app from the iMap weather guys. If you are in a polygon, you get the notification on your smart phone. If you are not, nothing happens. it works beautifully.

*TV stations must stream their long form tornado coverage in a way that is accessible to all portable devices, not just some of them. And, make the stream easy to find either via a web page or app.

Why in the world are so many TV stations streaming only in Flash or Windows Media format? Yeah, maybe I am an Apple fan boy, but there are 200 million iOS devices in the world, and if you don’t offer an HTML5/H.264 stream that plays with iPhones, iPads, and iPods, you are telling those 200 million people you are not interested in serving them or providing emergency weather information to them.

We use uStream for this very reason… it is universal, and can be viewed on ANY smart phone, Android, iPhone, whatever. I can’t tell you how many people sent me notes letting me know that they were watching us in their tornado safe place via uStream. And, thousands had no commercial power after the morning storms April 27, and the smart phone/uStream method was the only way their could see our live coverage. A lifeline for so many.

*Social media is not a time waster or a novelty, it is a lifeline during severe weather, and must be used by TV meteorologists.

Lives were saved April 27 by pushing tornado information to the masses through Facebook and Twitter. Seems like many old school news directors think this stuff is for high schoolers. How wrong is that; these social media services are mainstream and reaches across all demographics.

And, you just can’t throw up a Twitter or Facebook account and expect to be successful. It takes years of conversation and interaction with followers to grow your numbers and reach critical mass.

Broadcasting is now a conversation. The people that follow you on Facebook and Twitter aren’t idiots… they are our friends that can offer a treasure trove of information during active weather and any kind of breaking news event. They follow you, you follow them. Most media people just don’t get it.

*When there is a genuine tornado emergency, TV stations must have the guts to blow off regular programming and go with wall to wall weather coverage. No matter what the regularly scheduled program happens to be.

I was appalled to see a TV station in a top ten market opting to air the season finale with “Dancing With The Stars” when a tornado warning was up for the two major counties in the metro area this spring. No guts, no glory. If you can handle a little email from the haters, you don’t need to be in this business. People have a long memory, and if you aren’t there for them, they will go somewhere else.

*TV meteorologists need to get on the key wall and look their audience in the eye.

Too many of us are now hovering around computers and radar screens, off camera. I have noticed some stations go for almost an hour or so before you even get a glimpse of who is speaking. I understand the need to be looking at a computer screen, but get it out in front of the key wall so you can go face to face with your audience. There is no substitute for this; it is a key element of the communication process during a tornado emergency.

I could go on and on… but one more thing. Despite my long tenure here, I still learn something after every severe weather event. Some things worked, others didn’t. Learn from your mistakes and don’t do them again… and always strive to get better.

The media landscape is changing like a meteor streaking through the might. We must adapt to the changes, and provide the weather information the audience wants and needs on their time schedule and at their convenience. You see, the power has shifted from the TV newsrooms to the people, which is an exciting opportunity for us to personally serve the masses. You take care of the customer, which is your viewer, and they will be faithful.

I doubt if I see another April 27 on my watch, but there will be more severe weather days, hurricanes, snow storms, and blizzards. We must strive to be better at what we do while learning to have a servant’s heart!


Powered by Facebook Comments


Category: Hodgepodge

About the Author ()

James Spann is one of the most recognized and trusted television meteorologists in the industry. He holds the AMS CCM designation and television seals from the AMS and NWA. He is a past winner of the Broadcast Meteorologist of the Year from both professional organizations.

Comments (191)

Trackback URL | Comments RSS Feed

  1. Deb says:

    As someone mentioned earlier, this outbreak was predicted well in advance. That being said, I know that most people probably don’t stay on top of the weather as much as I do…I am terrified of storms and have been since I was a child.

    I watched James on UStream that day…I told my boss I was staying home from work (we have a basement at home) and I know most people don’t have that luxury. In the path of an EF-5, I don’t know that it would have helped me but we were fortunate in our area of southern middle TN and didn’t get hit like Alabama did.

    I understand what James means about “siren mentality”, but people just need to be educated that they won’t always be able to hear a siren. Where I used to live, I could hear one….even when I was inside my home. I don’t think they should be taken down…rather that people need to realize that when the signs are as potent as they were for this event, you just need to find a way to be weather alert…whatever that may be. In fact, whenever we are under a Severe Thunderstorm Warning, you couldn’t pay me to be in the shower or anywhere except watching to see what happens. My husband says not everybody puts their lives on hold that way, but I do…and make no apologies for it. Weather is serious business….and I, too, although at first shocked at the huge numbers of lives lost…realize it could have been a far greater number.

    My prayers go out to all those who lost family members, friends, pets, and property that day. I hope never to see another severe weather event like this in my lifetime.

  2. James M says:

    What this blog implies is that the following, taken directly from the Birmingham NWS, should not have received a tornado warning because it was part of a linear system:

    “National Weather Service meteorologists surveyed damage across central Walker County. It was determined that the damage was the result of a tornado. This tornado spun up along a shear axis in rural southwestern Walker County east of Hwy 69 and south of County Road 6. It rapidly intensified and caused major damage to a home, leaving a foundation swept clean on Horseshoe Bend. The tornado continued north-northeast through the Richardson subdivision and across Pleasantville Road, destroying at least two single-wide manufactured homes. In the Aldridge community, several homes were damaged, large trees were uprooted, and another single-wide manufactured home was tossed and obliterated. In Cordova, the tornado caused significant roof and parapet damage to brick buildings in the downtown area. The tornado crossed Old US Hwy 78 just east of the Mulberry Fork, snapping and uprooting trees, and continued northeast just south of County Road 22 before dissipating less than a mile north of Hwy 78. The damage path was 19 miles long and approximately 300 yards wide at its widest point.”

    I don’t understand how anybody in the meteorology field could argue against this damage needing a tornado warning. We HAVE to keep trying with QLCS tornadoes because of what they can do.

    As for false alarms, nobody is under tornado warnings more than about 10 days/year. If somebody cannot take a couple of hours to take cover once every 2-3 weeks (on average), then they are going to find any excuse to not pay attention to the weather. We absolutely cannot deprive good, focused Americans valuable and potentially life saving information just because others are too lazy to be inconvenienced. It’s not like the NWS issues these tornado warnings for fun, to ruin people’s day.

  3. boy says:

    Why not have a QLCS warning.You know we get damage as often from a QLCS event than a supercell.If you can work into peoples mentality the dangers posed by a QLCS event then maybe one extra life could be saved.I know almost all damage here in pell city over the years have came from QLCS events so imagine if a warning related to a severe tstorm or tornado was put up for all QLCS events say 15-30 minutes before reaching your area.

  4. Paulette Williams says:

    James, I think you would really be surprised at the number of people who do not have a Weather Alert Radio. I am continually amazed at the number as I talk to people. In North Alabama it would not have mattered any way because it went out early in the day. Don’t know the problem but it did not work. My family relied on local meterologists as long as we had power and then back to the age old battery powered
    AM/FM radio. I agree that something needs to be done about warning but really don’t know the solution. Personal responsibility is a must in all cases. If the meterologist tells you it is going to rain, don’t you make sure you have an umbrella? Same thing for severe weather, in most cases we are told days in advance, everyone needs to take it seriously.

  5. Bryce says:

    I also wonder how the TV show Storm Chasers has affected people’s perception of what is real and what is fake. I’ve been storm chasing for 11 years, and was in Joplin when it was hit a few weeks ago. I have never seen destruction as bad as that. I do my best to alert the NWS offices as well as local media as to what is really going on out there, hopefully saving lives in the process. Some people are just out for the thrill of it. Yes, there is a thrill, but there is also the serious nature of the situation. What you see on TV, in the matter of the Storm Chasers Discover Channel show is not what it’s like out in the field. I just hope and pray that people don’t go and stand out on their front porch to see what is coming because they want to see something like they have on TV. I really feel this is what happened in Joplin. The tornado was wrapped in rain until it was too late to know it was coming. Yes, they had warning and the sirens went off, but until the Tornado Emergency came out and the talk of a violent tornado moving into Joplin proper started, I’m not sure I would have taken shelter either.

  6. JoMo says:

    Bryce. I live 2 1/2 blocks away from the Joplin tornado path and I could hear the actual tornado while I was inside and had the TV on. It was a very loud rumble that at first I thought was rolling thunder. I took shelter in a closet and I may not be here now if the tornado had been a few blocks closer. Joplin has very few basements due to the high water table and rocky soil. We also suffer from ‘warning fatigue’ where we are warned but nothing happens, so most people do look for a visual cue before taking shelter. The tornado touched down just west of Joplin so the news media did not know it had touched down until it had made its way into Joplin.

  7. Trish says:

    I think James is the best meteorologist there is. I think the sirens can be helpful but people need to realize that they shouldn’t totally rely on them. I am like Deb. I am very weather aware. I take any warning serious and I don’t mind an inconvenience as long as I know that it could save my life or another life. I have heard people say there is no reason to be scared of storms well I think there is because you never know if it is going to be a really bad one or not. I do understand what James is talking about with the sirens though I know in the county I use to live in when a warning was issued the sirens wouldn’t go off until the storm was right on top of you. I would love to have the Weather Alert Radios issue the warning with the polygons. I really am with James on the TV stations should go wall to wall coverage when it is a tornado warning and not worry about making some viewers mad. Safety is a lot more important than any TV show. I would not watch that station at all. James I would just like to say thanks for all you do for all of us. Keep up the good work and just know that you are loved and appreciated.

  8. ObsGuy says:

    I would like to provide some insight. I humbly submit that I can vouch for what I’m saying, and I’ll leave it at that. No two radar operators have the same mentality. No two NWS offices have the same approach. And, unless you actually sit behind the wheel, and know what you are talking about with respect to warning decision making, be careful what you preach, and what you ask for.

    This event was a very bad example to site with respect to lives lost, and NWS performance. This debate needs to happen after an event similar to what James is concerned with. I think it is in poor taste and the timing is not good. There will always be people out there saying, AWW they always sound the sirens or issue warnings and nothing happens. Similar things were said in Joplin, MO too. This event is NOT unique in that regard. In the meantime, James has a good point. There are on occasion (maybe more often as of late), events where offices are issuing a large number of large TOR polygons (covering many counties at once), trying to catch a bunch of leading edge “spin ups”, as A LINE OF SEVERE THUNDERSTORMS CAPABLE OF PRODUCING RAIN WRAPPED TORNADOES”. Well, often times, these isolated spin ups are not producing wind much stronger than the overall line of severe storms. However, on occasion, within QLCS systems, you can in fact get an EF2 or EF3 tornado. In fact, one event I am familiar with, a QLCS with very strong rapid shear development thundered an F3 down suddenly, and killed somebody.

    It’s really not rocket science. The data we look at is good enough (usually), to determine whether or not the circulation and signature is strong enough to be a potentially significant tornado or not. Our greatest challenge is when these QLCS systems are farther from the radar, and you are detecting mid level rotation, vs. the needed near ground shear signatures we rely on, to catch these. None of this is explained in his article. Radar fundamentals, principles and limitations are just a small sampling of the overall warning decision making process. Lead time, county configurations, confidence in the data we are looking at, all fall into the mix.

    So, there is no easy answer to this. To blanket say, the NWS should stop issuing on QLCS signatures and small spin ups is weak on its own. It’s not that simple. What I would like to see within the warning decision making process, is greater discretion. Do not “blanket” warn, because there may be a spin up. Thousands of square miles being warned for, for a 100 yard wide, 1/4 mile long 80 mph wind damage event (in the hopes of catching it), does tend to water down the Tornado Warning. Not to mention, you are triggering large corporations, hospitals, entities, and other agencies to implement their tornado plans of action, which cost time, and money. You can see how there is a very FINE line here. They need to do this, but not time after time after time, because offices hope to catch the “spin ups”.

    I do agree with James that the Tornado Warning needs to mean something. People need to understand its importance, and that it could be a matter of protecting their own life, and the lives of those they care for.

    So, my intention here is not to attack James. This debate is good. Very good. The passions expressed within the feed are valuable, and well intended by all. But some additional insight, might offer up a little bit more to the story. That way, some things discussed here are not endorsed as fact, or assumed, when in reality they are not, or should not be.

  9. K.L. Smith says:

    One of the best posts concerning the “siren mentality” ever. Everyone here should post James’ post to Facebook, Twitter, and any other Social Networking site, print it out and post it on the ‘fridge. Tornado Sirens are outdated and dangerous warning systems. Those of you over 30 can remember that more times than not, a Tornado Watch or Warning broadcast on television or radio during the 70’s, 80’s and early 90’s were hardly ever accurate. I can remember many times hearing a Watch/Warning issued for my county and never hearing thunder or seeing rain or lightning. As technology hurdles toward the age of “instant everything”, and our thirst for knowledge and information is being quenched with weather apps for smart phones, text message alerts, home phone alerts, and live streaming of severe weather coverage, the Tornado Siren is as outdated as the telegraph or typewriter. Imagine someone coming to your door and screaming “FIRE!”. Your first instinct is to flee and get to a safe place. Now imagine someone doing this over and over again over a period of time. Yet you never smell smoke, never feel the flames, and never experience any damage to your home. It is only human nature to begin to believe that your house is not, nor will it ever be on fire, and you eventually believe that the person screaming fire is ignorant, stupid, and has no idea what he is doing. Such is the case with meteorology. In the rush to provide information to the public, weather services and TV news outlets have become the ineffective, intrusive neighbor, always pounding on your door with news of impending disaster while we ignore them, and complain about not getting to see our favorite show. Please take time to educate your family to NEVER rely on Tornado Sirens during severe weather. And always trust James Spann and ABC 33/40 for your news, sports and weather coverage. They have proven to be the most accurate and most reliable team broadcasting in Alabama, and in my opinion, the entire Unites States.

  10. James,
    Good article. I have a question for you and would like your input and thoughts on this.
    After reviewing all of the deaths, in all the various locations of Alabama, I noticed that a very large percentage of the deaths were to senior citizens. (age 60+) Granted some of this may be caused by being a more frail person not able to handle injury as well as younger persons. Also, possible age related illnesses / conditions (heart) may be a contributing factor also.
    But, with all of this, the percentage still seems considerably higher for the 60+ age group compared to those in all of the other ages combined.
    Just as an added note to this, this trend seems to follow in other areas of the country also.
    What are your thoughts on this?
    Thanks as always James.

  11. Jason says:

    James, I have been watching you forecast the weather for longer than I can remember and your about the only one I watch in times of bad weather but I don’t agree with everything you say here. You are pushing NOAA weather radio which is fine and dandy, until it just don’t work like on April 27th which their was NO excuse for. On one hand I understand what you are saying about the siren mentality but on the other hand I believed it could have saved some lives on April 27th as well. You are the weather man but what you say is just your opinion, doesn’t mean it’s right. Nobody probably is exactly right but until the system has some major improvements we don’t need to get rid of anything in place right now, that’s my opinion. You might have a different point of view had one of these monsters been staring you down.

    Ok, let the bashing begin.

  12. Matt Smith says:

    James, I do not agree that sirens should be taken down because they are the most basic of warnings. What you are advocating are very complex systems that can be prone to failure in a true emergency. What needs to happen though is the siren tech needs to improve, just like you advocate for the NOAA weather radio tech to improve. I.E only sirens in the polygon going off and not the other ones. Also, municipalities need to stop sounding the sirens for non-emergency situations or for storms that will not hit them. Where I live they are sounded to mark the noon hour. Also they have been sounded by the city just because a storm was in the area.

  13. Ryne says:

    All of Walker County was without power just as the tornado event was getting underway on April 27th. No Power, No Cell Service, No Internet. Luckily, we had enough batteries to turn on the radio and find weather coverage. If we didn’t have that, our only warning would have come from the tornado sirens. Some people don’t understand that technology can fail. Technology did fail on April 27th in my area. I’m all for improving the warning system, but taking away the tornado sirens, which could be somebody’s only warning, is not the answer.

  14. Wx_Enthusiast says:

    Before I dive into my criticisms and questions, I want to first say that although I’m not from Alabama and have never lived there, I did have the pleasure of watching James Spann’s live coverage on the internet during the April 27th tornado outbreak. I found it to be extremely good coverage and left with the impression that James is a very, very good communicator of life saving weather information when the crap hits the fan such as in a major tornado outbreak. I believe that he should be commended for those efforts.

    Now, for the criticisms. James M and ObsGuy have already summed up my position on the issue of QLCS tornado warnings, but I’ll add a few more of my own insights. To put it bluntly, wanting the NWS to stop issuing tornado warnings for QLCS tornadoes is foolish. A good authoritative source in the meteorological community for QLCS tornadoes is the Trapp et. al. 2005 paper that appeared in the Weather and Forecasting Journal. A link to it is available at and I encourage anyone who wants detailed information about the climatology of QLCS tornadoes from a peer-reviewed article to read through it. Essentially, all 3828 tornadoes that occurred in the 1998-2000 period in the U.S. were examined, and it was found that 693 (or about 18%) of those were from QLCS’s. Alabama had a slightly higher percentage of its tornadoes in the QLCS category. Also, while it is true based on the study that QLCS tornadoes tend to be weaker overall than tornadoes spawned from supercells, there are still many QLCS tornadoes that can be on the strong side. Based on Figure 3 in the paper, about 73 QLCS tornadoes were F2 or higher (about 59 F2, 12 F3, and 2 F4). If the NWS does what James Spann seemingly wants them to do, then that’s 73 significant tornadoes in that 3 year period that would not have a tornado warning. That’s not good. Instead of taking such an extremist position on this issue, I think a better position to take is to acknowledge that while the NWS False-Alarm-Ratio (FAR) is fairly high (something I’ll discuss in the next paragraph), the NWS still needs to continue to try to issue accurate tornado warnings for these types of events since there are certainly a non-trivial amount of cases where QLCS’s do produce significant tornadoes. There are also a large amount of F1 tornadoes produced by QLCS’s according to the study, and certainly an F1 tornado is nothing to sneeze at either. In fact, ALL tornadoes should be taken seriously. I also personally believe that even F0 and F1 tornadic winds should be treated differently than straight-line winds of similar magnitude. I suspect that the vertical component of the tornadic winds can potentially cause more structural damage and thus more threat to human life than straight-line winds of the same magnitude. I can’t find any peer-reviewed articles that discuss this issue offhand and I’m no structural engineer, but it seems to me that it’s easier to lift objects and destabilize structures if you can get wind up underneath them (like a car or a roof of a house for example).

    Now to address the FAR issue. From a raw number’s perspective, yes it sure seems like the NWS FAR is pretty high. However, what does this really mean for the individual person living in neighborhood X? In the case of JoMo from Joplin who commented earlier, there is certainly a concern about “warning fatigue”, or the notion that a lot of tornado warnings are issued and nothing happens. I want to ask an open question about this first and then present data from 3 cases this year. My question for the rest of the public is simply this, how many false alarms per year are too many for you? In other words, what do people think the magic number of false alarms is such that once that number is reached, any tornado warning after that begins to lose credibility? I believe that this is the kind of question that needs to be asked, and I certainly don’t know the answer to it or know of any research that has indicated what this magic number might be. With that question in mind, here is the unofficial data that I’ve compiled for three counties (Jasper County, MO, Jefferson County, AL, and Tuscaloosa County, AL). Each of these counties has seen a terrible tornado disaster this year (Joplin, Tuscaloosa, and near Birmingham). I used the publically available Iowa State Unofficial NWS Storm Based Warning Verification site at in order to calculate these stats. Using this site, I counted up the total number of tornado warnings issued for each county from 2005 to the present day as well as how many tornado warnings “verified” with a tornado local storm report, how many did not have a tornado report, and how many tornadoes occurred with no tornado warning in those three counties. I also then calculated how many tornado warnings on average each county receives per year by simply dividing the total number of tornado warnings for each county by 6.5 years (2005-present day). Here are the results for each county. Keep in mind these numbers are unofficial and the official numbers are probably slightly different due to updated surveys and information, but these are probably good ballpark figures.

    Jasper County, MO: total TOR warnings=39, verified TOR warnings=11, non-verified TOR warnings=28, tornado reports=12, warned events=11, POD=0.92 (11/12), FAR=0.72 (28/39), average number of TOR warnings/year=6

    Jefferson County, AL: total TOR warnings=62, verified TOR warnings=20, non-verified TOR warnings=42, tornado reports=27, warned events=20, POD=0.74 (20/27), FAR=0.68 (42/62), average number of TOR warnings/year=9 to 10

    Tuscaloosa County, AL: total TOR warnings=90, verified TOR warnings=28, non-verified TOR warnings=62, tornado reports=35, warned events=28, POD=0.80 (28/35), FAR=0.69 (62/90), average number of TOR warnings/year=13 to 14

    Here are some things to think about when examining these numbers. First of all, these are all county-based numbers. We have been in the polygon era for a few years now, and in reality, a person in neighborhood X is likely to be under a tornado warning less times per year than these numbers indicate, and perhaps significantly less. I do think the county based numbers still carry some validity though because as has been pointed out already by James Spann, the NOAA wx radios and many siren based systems are still county based in their technologies. That means that even if the NWS polygon does not include neighborhood X, person X in that neighborhood might still believe that he or she is under the warning since his/her NOAA weather radio will still go off and/or the sirens nearby still might sound. In other words, the advantage of the NWS polygon has just been negated in that particular scenario by creating a needless false alarm even though that wasn’t the intention of the office issuing the polygon. In addition, I know for a fact that at least one county in Georgia will sound their sirens for a severe thunderstorm warning IF a tornado watch is also in affect. There’s no telling what kind of harm that is doing in the false alarm arena by potentially making people think a tornado is coming when in fact no such warning is in effect. In light of JoMo’s comment about warning fatigue, the Jasper County numbers are particularly interesting. The numbers indicate that out of the 6 tornado warnings/year issued, roughly 2 of those warnings had a tornado within the polygon and 4 did not. Are 4 tornado false alarms/year too many when about 2 times a year, a tornado is relatively close by? Or, perhaps something else is going on that’s beyond the numbers? I don’t know these answers, but I do think that this is the kind of information that the meteorological community needs to find out. In reality, if the NOAA Wx Radios and the siren technology were upgraded to polygon technology, then the number of warnings that neighborhood X felt they were under each year may be significantly reduced further from the numbers I presented.

  15. Charles says:

    I have enjoyed this discussion very much. Is the elephant in the room the fact that some tornado warnings come during Dancing With the Stars and are very unpopular? I don’t watch that show but might make an exception if James were on it.

  16. JoMo says:

    Interesting stats Wx_Enthusiast.

    I think that people say ‘nothing happens’ when a weather event doesn’t personally have any effect on them. This adds to the warning fatigue even if a tornado occurred nearby.

    The last tornado we had in Joplin prior to the EF-5 was in the 90’s I believe and it was technically a gustnado. The last strong tornado was an F3 in 1971.

    We have had several close calls recently:
    May 10th 2008 tornado (Picher, OK tornado) which went just south of Joplin. May 4th 2003 tornado which went just north of Joplin and hit Carl Junction and I’m probably forgetting some others, but these storms did not hit Joplin proper, so in the eyes of the public ‘nothing happened’ where they were.

    I’m not sure how you would fix that.

  17. Nadine says:

    Many stations do go “wall to wall” and have ‘weather alert days’ when bad weather is approaching. People are being stupid and not taking cover when they need to. Or they want to go out and get a picture of the tornado that will most certainly get them$1,000 and their name in the media. The lead time has now gotten long enough that people get warned and expect the event to happen right now, but then ‘forget about it’ so that when it does come, their short attention span (thank you TV) has caused them to forget it. What about the people who are nowhere near the warned area hearing of the warning and driving specifically into the area just because they might get to see an actual tornado and they are actually putting themselves in harms’ way. Also, what about the cable companies not breaking into the cable shows or HBO, Showtime, etc not interrupting heir programming? Or people on their computers not being given a warning unless someone takes them by the hand and tells them to look outside or look at radar. People need to step up and do the proper action themselves for the right situation. Quit putting the blame elsewhere.

  18. Nadine says:

    JoMo – your comment is also true about the hurricane season. This past season was the third busiest on record, but since none hit the United States, almost everyone says it was a slow year. People are not going to believe it will happen to them until it finally does, even if it wipes out the other side of the street, and then its’ too late!

  19. Dot says:

    James, First thank you and thanks to your staff.
    I believe that sirens should be replaced and even some added in areas that “don’t hear”. How anyone can not hear a tornado siren is beyond my comprehension. They are deafening!EVERY household should also have a weather alert radio, EVERY household, as well as disaster relief supplies if but enough to get you through the first 24 hours.
    County wide alerts should be given to alert all people and then drop it to site specific. F1-F5.. I, for one, want to know where it is, probable heading and when it’s gone, they all are capable of death, injury and destruction. You should be the determining factor as to how long YOU interrupt programming. After all it is YOUR business to keep YOUR viewing area in the know. A tv program can go to “reruns”, the weather is in real time.
    Grants should be given to EACH state to study weather phenomena, not solely nationally. I care about national issues but am more concerned with my region and state, especially where the weather is concerned.
    Lastly, my family & friends always chided me about my seeming obsession with the weather. The weather radio was annoying and unplugged, the channel changed on the tv, or some off the wall comment made. The weather radio went off tonight, and it is not Wednesday noon. Everyone stopped in their tracks and listened, decided, turned the tv on to 33/40 and then returned to their activity. It was not unplugged and was left on to repeat till it went back to regular programming and then turned back to alert. One very difficult lesson learned, thank God.
    Yes, my family was affected with loss of material things, not loss of life, thank the Lord. Our hometown, Pleasant Grove.

  20. ~Eddie~ in Warrior says:

    I feel that the NWS should have what James Spann has said for many years now. First a “Tornado Watch” is issued, then if with these indications that show up or radar that “could” be a tornado then the NWS should had a “Tornado Alert” and last the “Tornado Warning” for when there is a tornado on the ground and has been seen by someone.

  21. Steve says:

    Bryce, I know what you mean about going and looking for the tornado. I now live in Tennessee, but I grew up in the OKC metro. The joke my wife and I tell our Tennessee friends is that the tornado siren is the signal to go out on your porch and watch it go by.

    The choice to use uStream by ABC 33/40 is absolutely the right thing to do. I work in broadcasting, and in my line of work I have to convert assets to more than a dozen different codecs and file formats for use on various devices. Having a one-stop output is an essential way of getting past the old ways of the broadcast ecosystem.

    Another great resource is the radio spectrum. In Knoxville, WBIR-TV simulcast on B97.5 FM for the entirety of the April 27th event. When my power went out, and my cell phone went dead, I went into the garage and listened to the broadcast (continuous, for over 8 hours, BTW) on my car radio. Not ideal, but at least I had enough warning to get to my safe place. In all of the rush by the federal government to auction off the broadcast spectrum for cell use, I think they might want to think about reserving a small portion of that bandwidth to allow for a weather frequency that will simulcast weather coverage in areas prone to severe outbreaks.

  22. JCoop says:

    On April 27th I awoke at approximately 5:30AM because, in the midst of a fairly intense thunderstorm, I heard the tornado siren going off. I have literally slept through bombings in the past; I do not just wake up for nothing. My wife, who normally awakens at the sound of someone shutting their car door a block away, did not hear it.

    So, while I agree that too many false alarms will ultimately create apathy among the vast majority of the populace, I strongly disagree that the siren system should be elimintate. It must be continued.

    For one thing, for people working (or playing) outdoors – and those of us who simply don’t spend our lives glued to the television or handcuffed to our Blackberries – the sirens are essential warning devices. We have nothing else.

    For another, there is often power to the siren system when homes in the area have lost power. In which case, we still have the sirens when the televisions are no longer working.

    Third, eliminating the sirens does nothing to prevent the interruption of television programming or alternate warnings broadcast on television or radio which – by your own words – are very often false and very often ignored. I fail to see how taking down the siren system improves anything.

    Finally, if I’m not mistaken, the Civil Defense siren system was not erected solely for weather warnings. It serves other purposes and can be used to warn citizens of other emergencies.

    So, while almost everyone would agree that NWS and the weather forecasting community should continually strive to improve the accuracy of their detection, reporting, and warning, it would be foolish in the real world to tear down an entire warning system which has served citizens well for decades.

  23. David says:

    I am not from the Midwest – I can tell you that here, in Maryland, when the sirens go off, people understand it is time to take appropriate shelter. (Frankly, I am kind of surprised about people who don’t take shelter during a severe t-storm. I’m not referring to a basement, but I am referring to going to a safe place). I personally have a Midland Weather Radio, and I feel safe sleeping at night knowing that the radio next to me will warn me if there is severe weather headed my way.

    That being said, I am for tornado warnings in storms that “are capable of producing a tornado”. A couple of months ago, following the April 27 outbreak in the Midwest, parts of Maryland were tornado warned multiple times due to the rotation in the storms, but no tornado was ever formed.

    In my opinion, the purpose of a tornado warning is to warn people to seek shelter BEFORE a tornado forms. If the atmospheric conditions exist for a tornado to form, that’s good enough for me to have a tornado warning issued. What’s the point of issuing a warning if the tornado is already on the ground a couple of blocks from your location?

    I agree, through, on the complacency issue, especially in areas more prone to tornadoes than Maryland. As an emergency manager, complacency is one of the problems we face when trying to have people listen, understand and take action on our warnings.

  24. Keith says:

    James, I have seen the situation here in NW Ohio you mentioned. November 2002 an F4 tornado hit Van Wert Ohio, it then crossed into portions of Paulding, Putnam, Defiance and Henry counties dissipating 6 miles from Napoleon. The sirens were going off here in Napoleon, and people were out walking their dogs, doing yardwork etc.

    Sirens are a two edged sword, they can be a great tool in emergency notification, but they can also be a hindrance. Last year two tornadoes hit NW Ohio. In the case of the one in Fulton County some of the sirens out in the rural area could only be heard in some cases less than a mile away.

    As an emergency management professional, I look at these cases to determine how we locally can better prepare and warn the public.

    One of the most important things is public education. NOAA radios are not that expensive ($40.00). I know many people who rely only on the sirens or broadcast media for warnings. If the power or other infrastructure (cable/phone/internet) goes out these people will have no way of hearing the warnings that are issued.

    How many people spend $40.00 a month on junk food, movie/game rentals, alcohol and/or tobacco products? Surely they can spend that money on something that could very well save their life or the lives of their loved ones.

    Sure warnings are issued based off of radar detection that many times do not actually produce a tornado. But if people do actually take heed to the warnings they will be safer than if they don’t and one actually occurs.

  25. Matt says:

    I agree with Mr. Spann. Here in Fishers, IN we’ve had several tornado warnings this spring where there was a line of strong storms rolling through but no tornado on the ground. I used to think that a tornado warning meant that there was a tornado on the ground and you should take cover immediately. Lately it seems tornado warning has become synonomous with severe thunderstorm warning. It doesn’t have the same shock value.

  26. John says:

    I agree with Eddie. The warning system needs to be more granular to allow for the unknown factors. Defining a “Warning” as a sighted tornado and adding an additional category for possible radar-indicated tornadoes seems like an ideal way to do it.

  27. mercy says:

    I have to disagree with you on most of what you have written here. I have lived most of my life in Oklahoma , I can tell you that yes there are some improvements that need to be made on the warning systems. But I can tell you that the ones we have in place here. Sirens , weather radio , tv , reg. radio , smart phone apps , facebook , twitter , and even the land lines that normally issues Silver and Amber Alerts SAVE LIVES! It’s not the warning system. it’s the people who choose not to get out of their mobile homes , get off the road before it’s to late. It’s the people who choose not to get somewhere safe before the sirens go off. In 2009 , my town was hit by an F4 it killed 8 people. 7 were from a trailer park. 1 was on the highway. The numbers could have been higher. We had 30mins to get some place safe because of our local news station told us it was coming. 3-5 min. warning when the sirens went off.I personally think you need to come to Oklahoma and get with Gary England and other meteorologists and learn more from them.

  28. Robert says:

    Well James, you said you wanted comment and you’re getting it – thanks in large part to Matt Drudge. 🙂

  29. Kevin says:

    On the storm sirens I say keep them since digital TV signals seem to only want to work part of the time.

    This may sound odd to some people but my parents and many older people do not have cable TV. Which wasn’t a issue in the past, but with the switch over the digital TV you can’t even get local ABC, FOX,and NBC stations for the Atlanta area on a clear day living 30 miles from Atlanta on a hill half the time.

    On the day/night of the outbreak it came down to me on the internet streaming from various top notch weather people across the region to let them know how things were going. As is it was only dumb luck the tornado that blew threw our county was 15 miles or so north of us.

    If we had lost internet/lost the radio and the storm hadn’t unexpectedly tracked farther north they along with many others may have been killed or injured due to no or little warning.

    So yeah keep the sirens, just use them better.

  30. bobby says:

    It doesn’t matter how much warning you have if you do not have a safe place to go to.

  31. Michael Scott says:

    The problem with warning on QLCS tornadoes is by the time you detect them and get the warning out, a high percentage will have either done their damage and fallen apart or never made it to ground in the first place. This is a subject that needs hard data, i.e. of warnings on QLCS tornadoes what percentage were either too late to save lives or had fallen apart by the time the warnings hit the air waves? If that percentage is high, you are probably indeed doing a disservice by desensitizing the population to these warnings, making them more likely to ignore warnings to supercell tornadoes which are generally both more powerful and more persistent and hence more likely to take lives.

    There is no perfect solution here. The best you can do is gather as much hard data as possible and then try to make rational decisions that minimize the loss of life. It’s possible that by raising the threshold on QLCS warnings that you might see a slight rise in the number of QLCS tornado deaths while seeing an overall drop in total tornado deaths. I don’t know that would be the outcome but it is certainly possible.

  32. Lindsey says:

    While there are certainly some points made here that aren’t going to be easy to resolve, I can definitely say I agree with the Cry Wolf Syndrome at work here. In the 4 or 5 weeks prior to April 27, we had at least one tornado warning in Jefferson County each week. Usually you get one or two a season, but when the warnings seem to come every other day — and nothing comes of it — then after a while you start to ignore the warnings as just another nuisance. In these cases, yes it IS annoying when regular programming is scratched… there’s only so much to be said about the non-event in the hour or so it takes to expire.

    On April 27th, the alarm sirens went off for the first time at about 5:50am, and I remember grumbling and whining that it woke me up 45 minutes before my alarm. Did I turn on a weather radio or the news to check what was happening? Of course not, I pulled a pillow over my head because surely it was another false alarm, or it was affecting the extreme opposite side of the county and had nothing to do with me. If I had checked, I would’ve known that severe weather was coming right at me in Cahaba Heights. I’m lucky that I live about 1 mile from the worst damage there and wasn’t affected at all, but it easily could’ve been a different story. Apathy is definitely at play in cases like this.

    I don’t know what the solution is. According to most stats, if James’s suggestion about being more selective with warnings was to be taken then there would’ve been no siren or warning at all for those early morning storms. But then again, all of those false alarms in the previous weeks would not have been triggered either, and maybe people would have paid more attention when the alarms actually did go off that afternoon. As for the sirens, they absolutely should not be removed. I only turn on a weather radio AFTER I hear a siren go off, and while I know that’s not the ideal, proper way to deal with severe weather I’m pretty sure it’s the way most people do.

    I do believe the ABC 33/40 weather team’s non-stop coverage that day saved a lot of lives — seeing that monstrous tornado in Tuscaloosa, and being told the exact path it was taking into Birmingham, surely knocked some sense into quite a few apathetic viewers, and they knew this was for real and action needed to be taken. With monsters like that, there’s only so much protection hiding in a bathroom with a bike helmet will give you as it hovers right overhead. I can only imagine how much worse the casualty count might have been.

  33. Alan says:

    Why can’t there be three categories:
    1. Conditions are favorable
    2. Rotation has been detected by radar
    3. Tornado on the ground

    Sirens would only be sounded for #3. This way everyone would know that when they hear the siren there is a real and imminent danger.

  34. survivor in Pleasant Grove says:

    I live in Pleasant Grove and everyone knows the devastation there. Our power went off after the %:45 a.m. storm and I made the decision to stay home from work that day due to the prediction of worse storms that afternoon. I did not want my teenagers alone in that kind of danger.

    My power came on around noon. We were watching James Spann when the tornado formed in Tuscaloosa and I knew we were in trouble. The siren sounded in Jefferson and I had my kids downstairs in the basement bedroom. I was looking outside thru the garage and saw debris falling. About that time, I walked onto the driveway and saw the tornado–I knew that is the sound I had been hearing. I got the kids into the basement closet just as it hit–we were lucky—just roof damage and trees down. My sister, grandmother and cousin lost everything–and they were less than 1/2 mile from me. Just 3 blocks away houses were gone. Friends were killed. It was horrible. Without that siren and tv coverage, it couldve been worse. I keep thinking what if I had not heard James tell of the threat that afternoon and went on to work. I would not have been able to get home to my kids!

    That being said, it is offensive and tacky, in my opinion and that of many others affected by this tornado, to see these storms named! To see commercials that are made of clips of James Spann’s coverage and images of devastation in my hometown, with a name such as “April’s Fury” (not even sure if that was it,but it was something to that effect)along with a voice saying “we were there giving the best up to date coverage” flashed on the screen turned my stomach. I have not watched ABC 33/40 since. The images we see and have seen since that day are hard to live with and deal with in light of what this tornado took from us is bad enough, but to turn on the television and have it flashed before us in what seems an effort to get ratings is appalling and downright tacky. I knew all of the people who were killed in Pleasant Grove. How do you think their families feel when they see this?

    Keep your severe weather coverage. I feel it did/does save lives, but PLEASE have consideration for the victims of this and other horrific weather events and don’t use it to get higher ratings! We know that April 27 was a historic tragedy, but it doesn’t need a name. We also know that James Spann was on air with coverage and we do not need the constant “we were there” commercial with clips of our suffering and loss.

  35. James Spann says:

    Hi Survivor…

    Thanks for the kind comments about our severe weather coverage. In defense of our news department, they have not branded this event anything. “April’s Fury” is used by another local TV station:

  36. Survivor in Pleasant Grove says:

    My apologies Mr. Spann, on mistaking that your station named the storm.

  37. Danny says:

    Are sirens out of date? Yes and no. In many rural and urban communities sirens are woefully out of date, in a state of constant disrepair and too expensive for the community to maintain. In such cases they do more harm than good if one never really knows if they will work or not. Can they be used more effectively? Absolutely. The signaling technology exists to activate sirens in specific areas as to not “overwarn.” The larger problem is that many emergency managers (who operate the sirens) do not take advantage of that capability. They would PREFER to warn for the entire populus…”just in case.” This leads to the “cry wolf syndromne” and accompanying apathy that James is absoultely correct in pointing out.

    Inevitably the first sound bite on the news is the survivor who says “there was no warning….” which we know is basically not true unless they were paying no attention to anything. Most native Alabamians don’t have to be warned when a storm is approaching. Simply looking at the sky or hearing the wind come up is a good indicator depending on the time of year. Someone asked earlier, “how can one NOT hear a siren?” Easy. The power can be knocked out; the distance you are from the siren; high wind and foliage on the trees in the spring can attenuate sound, not to mention very heavy rain or hail. That’s before you add the fact that most siren systems, were a product of the 50s and 60s cold war and installed before every home had central AC, R-14 insulation and double pane windows. We build houses today designed to keep elements (including sound) OUT of our homes. While sirens can effectively serve the intended function of OUTDOOR warning, they should NOT be anyone primary means of warning.

    The digital S.A.M.E. signal used by NOAA Weather Radio and the Emergency Alert System (EAS) has had the capacity of providing location specific warning abilites since it was adopted for use in the mid 90s. However, it is vastly underutalized. The county FIPS codes can be parsed down into 9 additional locations within EACH COUNTY. EAS/NWR receiver devices were designed to be addressable, down to 1/9th of a defined area. This ability is rarely used or adopted by any warning authority. That defined area can be anything. A specific community, section of the county, a building….whatever the warning managers choose to define it as….even a specific siren if desired.

    If building codes can REQUIRE a smoke detector, why can that same requirement include an EAS chip and receiver that would activate the same smoke detector for an emergency? There are available EAS receivers that will operate bed shakers, digital signs and blinking lights. They will turn themselves on and off automatically and only when there is an IMMEDIATE threat to a specific populus. The need to blanket an ENTIRE area for a targeted event is not necessary. The whole point in providing a warning is to spur a physical response to that person who is warned. If nothing but heightened awareness all the way to physically taking shelter, we do have better means of doing this today without the overwarning which results in people ignoring the warning altogether. That’s why and how people get killed.

    Detection technologies have vastly improved. All I hear James saying (and I agree) is that the means of effectively communicating the warnings that are generated should be improving as well…but in many cases they are not. The technologies capable of doing this have been around for quite some time. It is time they are adopted and used correctly to their full potential.

  38. As a former on air weatherman / reporter in the Memphis market during the 90’s, I commend your position on this. Bravo! You have nailed it perfectly! One of the reasons I left the business in 2004 was because of much nonsense that had penetrated the business. Thank you, thank you, thank you! For saying what’s needed to be said for a very long time.

  39. Andy says:

    Not sure if anyone has said this yet, but here goes anyway. Why are we still warning marginally severe storms? How many SVRs are put out each year for weak squall lines that might gust up to 60 mph, or for pulse storms that might manage a penny-size hailstone? Talk about warning fatigue! What’s the worst a 60 mph gust can do? It excites us weather nerds, but the general public does not care if there’s a slight chance they may lose a twig off the tree out back. Eliminating those warnings might also allow us to stop issuing TORs for dubious QLCS events.

    If we restrict SVRs to what the SPC calls “significant severe” events, they would become much rarer and people in the path might actually seek shelter. SVRs would then be sufficient for QLCSs that might cause a brief spinup. Then TORs can be limited to cases where a true tornado, not a leading edge gustnado, is likely to form and remain on the ground for more than a few hundred yards. Would there be exceptions, like the F3 spinup mentioned above? Probably, but if the redesigned SVR is issued, people should be aware the potential for real danger exists.

    What about the marginally severe storms that would no longer receive warnings? No worries–for those weather-savvy enough (or in love with their cars enough) to care about 1″ hail, special weather statements would suffice. Also, maybe special weather statements would stop being such a joke then: I read one not long ago for “a line of gusty showers.” Gusty showers? Seriously?

    On another note, public awareness of a given day’s potential to turn exceptionally dangerous is at a shocking low. Joplin residents never saw it coming–but the SPC had them in a moderate risk and a hatched tornado area all afternoon. That means a 10%+ probability of EF-2 or stronger tornadoes. If you’re in that area, you MUST know that if a warning is issued, you need to be below ground because it could be an absolute monster.

    Public service announcement campaigns across the Plains, Midwest, and Dixie Alley might help a little. Ads encouraging residents to not seek visual confirmation of warnings would be appropriate, as would ads reminding people to get to the best shelter around (underground if possible) instead of settling for a closet in a one-story vinyl siding house. Also, that so many people died in mobile homes in Alabama blows my mind, so I guess we need even more reminders that trailers can’t handle twisters.

  40. Wx_Enthusiast says:

    Andy, I’m going to disagree with a lot of what you said about warning criteria. First, a slight clarification. Warnings for penny size hail (0.75″) were discontinued in January 2010 so there haven’t been any warnings for that since the end of 2009 basically. Warnings are now for quarter size hail (1.00″) or larger. This was based on sound research conducted by Tim Marshall and other engineers as to when hail damage starts to occur with the findings available at and . SPC defines “significant hail” as 2 inch diameter or larger. Back in 1998, I experienced golf ball sized hail (1.75 inch) that the insurance company deemed did enough damage for a whole new roof to be put on the house. Not having a severe thunderstorm warning for that as you suggest would be silly in my opinion. It would have probably injured someone if they had been outside.

    Now, as for the wind criteria, “significant wind” is defined by SPC as hurricane force or greater. However, in the area where I live (Florida panhandle and southwest Georgia), weaker trees can start coming down with winds as low as 40-50 mph and several trees will come down in 60 mph winds. The city of Tallahassee is particularly vulnerable to strong winds given its number of tall, skinny trees. Winds to 60 mph in Tallahassee will cause big problems with several roads blocked by downed trees and numerous power outages from trees falling on power lines. There is definitely a legitimate threat to life and property with 60 mph winds in Tallahassee, and I am sure the same is true in other areas along the gulf coast and east coast as well. In light of the above vulnerabilities, special weather statements for winds of 40-50 mph are not a joke. They are simply telling people that winds of 40-50 mph are likely in the area. They’re not telling people to go run and hide in their bathroom. It is just passing along information that some people will find useful to know.

    On the QLCS tornado issue, I laid out my case for why it’s a bad idea to stop warning for those in my long post above, but again I will reiterate that stopping tornado warnings for QLCS tornadoes would mean missing about 24 EF2 or higher tornadoes (defined as significant tornadoes by SPC) nationwide per year. That is not acceptable in my opinion. That number is based on the peer-reviewed research I posted above. I do agree that the false alarm ratio for tornado warnings is high. However, a draconian solution such as stopping tornado warnings for QLCS events is not the answer. More research and advances in technology are needed to bring the false alarm ratio down significantly over the long term without sacrificing the high probability of detection. I will take a high POD/high FAR combination any day over a low POD/low FAR combination, although that’s just my personal preference.

    I do agree with you on the issue of public awareness. I think that it is generally too low in a lot of cases, and people will claim “warning fatigue” as one of the reasons. However, while the warning fatigue phenomenon may be real, I do not believe the primary reason is from too many warnings by the NWS. I believe instead it is a perception problem combined with technology that is still county-based. I tried to provide evidence for my belief in my long post above when I responded to JoMo’s comment about warning fatigue. The county where Joplin, MO is located has only averaged about 6 tornado warnings per year over the last several years according to the stats that I could find. Is that really too many tornado warnings? Tuscaloosa and Jefferson counties in Alabama have averaged more per year over the last several years, but I believe they may be a bit larger in size as well and even their numbers are not excessive from a raw numbers standpoint per year in my opinion. The solution that will make the largest and most immediate impact on the warning fatigue issue in my opinion is to upgrade all of the currently existing county based dissemination technologies to polygon based. As an example of what this will accomplish, I will use the Birmingham forecast area as an example. Stats from this link show that for 2011 so far, the polygon warnings have reduced the size of the warned areas compared to county based warnings by 61.3% in the Birmingham forecast area. That means that if all dissemination based technology used the actual polygon instead of the whole counties, then there would have been a 61.3% reduction in the size of the areas warned this year for both severe thunderstorm warnings and tornado warnings combined. In my opinion, that would have an immediate and significant impact on the warning fatigue issue. Far less people would be warned unnecessarily, and it also means that in reality, the average number of tornado warnings for any given neighborhood per year is probably significantly less than the already relatively small (in my opinion) county based warning numbers that I provided in the earlier post.

  41. Just wish to say your article is as astounding. The clarity to your put up is simply spectacular and that i can assume you are an expert in this subject. Well along with your permission let me to grab your feed to stay up to date with impending post. Thanks 1,000,000 and please continue the enjoyable work.

  42. Just want to say your article is as astounding. The clarity on your put up is just spectacular and that i can think you are a professional in this subject. Fine with your permission allow me to snatch your RSS feed to keep updated with approaching post. Thank you one million and please keep up the rewarding work.

Leave a Reply

You must be logged in to post a comment.