Wednesday, March 07, 2012

Warning Fatigue Has No Easy Solution

All I can say is, "Enjoy the 60s!" There's plenty more where that's coming from in the next couple weeks. Incredible how some of the temperature departures look going forward. 20-30, even 40 degrees above normal in parts of the Midwest. More on that in other posts.

In the meantime, today I wanted to touch on something that is and will continue to be an issue in the wake of recent tornado outbreaks: Warning Fatigue. Tom did a good job summarizing things and explaining the benefits of a weather radio the other day. No, major severe weather outbreaks don't often happen in the Delaware Valley, but you never, ever know. Weather Radios are great, especially in those severe weather prone parts of the country. They save lives, especially at night. But the question many have posed, especially in the wake of the Joplin tornado was whether or not people are getting hit too hard with warnings. The NWS has even decided to begin experimenting with "tiered" warnings in parts of Missouri and Kansas in an effort to try and distinguish the big ones from the small ones and make certain warnings stand out over others. I respect their wanting to experiment with this and think it's a step in the right direction.

With that being said...I don't particularly think it's the answer. What is the answer then? I can't answer that question. And that's the problem with the overwarning debate. Everyone is willing to scream and shout at the NWS for overwarning, but the issue is that no real solution exists. Technology simply doesn't exist for us to be able to distinguish certain things. Yes, you can usually tell when a massive tornado exists (but sometimes even that occurs too quickly to "see" via radar). What is next to impossible to distinguish are the lower end to moderate tornadoes, especially those that spin up on squall lines (QLCS's). And those tornadoes are capable of being just as deadly or destructive as anything.
Radar Capture of Tornado
Credit: Brad Panovich, Chief Meteorologist
WCNC-TV, Charlotte, NC

A recent example: In Charlotte this past weekend, the NWS missed an EF-2 tornado that tracked over 3 miles that was 200 yards width at maximum. The EF-2 twister managed to suck a child out of his bedroom and drop him 350 yards away on a highway embankment. The tornado formed in a QLCS scenario, was difficult to see on radar, according to the NWS, and ended up injuring four people and causing a fair amount of significant damage. Now, you can see the image at left, captured from WCNC-TV's doppler radar. Brad Panovich, the chief meteorologist there (and a great Twitter follow), did a fantastic blog entry explaining why there was a legitimate reason to "see" the tornado and why there probably should have been a warning. The NWS has come out in defense of their warnings (or in this case, lack of warning).

NWS Defends Storm Warnings
Why NWS Missed Tornado Warning
NWS Issues Explanation

I can buy that the NWS is capable of missing a tornado. It's a tough scenario and sometimes things are just missed. But if the reasoning is that the Doppler Radar system was unable to see something in a particular location that a TV news Doppler was able to see almost clearly (as seen above), then we have a serious problem. Warning capabilities are only as good as the equipment and knowledge we have available to detect meteorological threats. I bring up this example because I want you to realize that it is still very easy to miss a serious event. And if the NWS organization in the Charlotte area made more sense, this may not have happened. As an aside, instead of discussing cutting the NWS budget, perhaps we should find a way to re-appropriate funds from elsewhere to this vital agency? Simply put, in its current state, the NWS cannot afford budget cuts. If we've learned anything in the last two years we need improved ways of detection to distinguish threats from false alarms. The NWS is a very well working, efficient government agency. They do absolutely amazing work and the cost-benefit is probably moreso than any other government agency. There are issues that need to be looked at to make them better (such as the Charlotte Problem). But I digress, as this is all another debate for another post.

Here's the problem: We have a lot of warnings...clearly too many. But the NWS cannot just stop issuing tornado warnings for "minor" events. We cannot always distinguish low-grade events from potentially damaging ones in our current system and with our current knowledge. The NWS may be guilty of overwarning in certain scenarios, but it is because they have a mission to fulfill. The NWS' mission is to protect life and property. They will not just stop warning. Weather changes rapidly and is still extremely unpredictable. What doesn't look like a big deal in one radar scan can quickly evolve to a major concern in the next scan...and then back to minor in the third scan.

We have a lot of avenues to deliver these warnings: TV, radio, social media, weather radios, sirens, etc....perhaps also, clearly too many. We have no way in the present warning system to truly distinguish bad from potentially bad from really bad. Tiered warnings are a possible solution....but then you may just muddy the picture. I worked in television for 4+ years. Many average viewers STILL cannot distinguish simply between a watch and a warning. We've done well in improving how we communicate those details, but still not everyone understands it. I don't know what the answer is to fixing this problem...and I don't know if anyone out there has a rational solution. I do know that this needs to continue to get publicity and be discussed in forums such as this.