Forecasters Look at Ways to Improve Alabama Weather Warnings

When it comes to weather in Alabama, no matter what time of year, there’s just one thing to do: “Prepare for it to be bad.”

That’s according to Kevin Laws, the science and operations officer at the National Weather Service in Birmingham and one of many meteorologists and researchers in Alabama taking a hard look at severe weather warnings, false alarms and how to improve the warning process.

They have cut their false alarm rate nearly in half since Alabama’s deadly tornado outbreak of April 27, 2011.

But reducing the numbers is only half the battle, it turns out. How do you change public perception?

“It’s that ‘cry wolf’ syndrome,” said David Nadler, the warning coordination meteorologist at the National Weather Service office in Huntsville. “We can reduce our statistics all we want, but invariably that … doesn’t necessarily alter public perception.”

There’s been some progress on that front since 2011, when a historic outbreak killed more than 250 people in Alabama alone.

“The warning process has improved so much that there is less opportunity for desensitization,” said Dr. Laura Myers, deputy director and research scientist at The Center for Advanced Public Safety at the University of Alabama. “Warnings are more accurate now, so people have begun to have more faith in them.”

But there’s still more to do.

“I think we’ve got a long way to go,” said ABC 33/40 chief meteorologist James Spann. “What happened out here 3 1/2 years ago (in 2011), it opened our eyes. We’re all physical science people, we’re not experts in human behavior, we’re not experts in communication … that day the physical science side of things worked almost flawlessly. The warnings were so good – and death toll was so high.”

With the experience from 2011, extensive research and the addition of dual-pol radar, meteorologists are more skilled than ever at detecting tornadoes – something to consider when a warning is issued for your area.

“Historically, people are used to having a false alarm rate that’s relatively high,” Laws said. “How do we re-educate folks … that hey, you might want to take these warnings a little bit more seriously? Because now you have a better chance of having a tornado than you don’t.

“Before it (the false alarm ratio) was 80 percent, now you’re only 40. So the chances that you are going to be impacted now are a lot higher if you’re in the polygon. That’s the key.”

The polygon is important. But do most people understand how it works?

“I don’t guess we’ve done a good job of explaining that because over and over we hear of people that are going to a safe place and doing their tornado precautions when they’re not even close to a polygon,” Spann said. “We’ve got to be better at telling people who’s in and who’s out. So that’s another big problem that gives the illusion of a false alarm. Most people who do all that stuff and they’re not in the polygon and nothing happens, they think `well, that was a false alarm.’ It wasn’t a false alarm – you weren’t in the warning.”

In the past, severe weather warnings were issued for entire counties. The polygon warning system was first introduced in October 2007. Its intent is to warn only the locations and people in the direct path of severe weather instead of the entire county, thus cutting down on false alarms.

Now meteorologists are working on refining the polygon. Laws calls it “polygonology.”

One thing being considered is clipping polygons at county borders while extending the length of time a warning is in effect, Laws said.

“So in the old days we’d issue (a warning) for 45 minutes. In the new days, what I’m saying to do is we issue this now for 60 minutes,” he said. “We don’t extend the polygon out to 60, we just extend the time out to 60.”

He explained that the trouble with shorter timeframes is that warnings sometimes have to be reissued before a storm clears a county. By extending the timeframe, “I can ensure that this polygon will be completely gone,” he said

That added time also allows meteorologists to focus more on the evolution of the storm, better track its progress and target warnings more effectively.

That’s what Laws calls the “art” side of polygons. Then there’s the science.

“The science side is a lot harder, though, because the science requires us to gather data about this, and what we’ve done here is, we’ve gone back and looked at every event we’ve had over the last decade. And we’ve parsed every single event in every way known to man,” he said. “We looked at staffing. We looked at the environments tornadoes are in. We looked at warning performance. So some events we did really well; some events we didn’t do so well. … A couple of things fell out of that.”

One was that he wanted to start fully using the wealth of data collected by the weather service, and not rely solely on the radar screen when it comes to issuing warnings.

“We’re learning more and more about tornadic environments. So how favorable is this environment vs. ones we’ve seen in the past? That needs to be incorporated into the warning decisions,” he said. “If you have an environment that’s not very favorable for tornadoes, then we may want to hold back on some warnings. Even if the radar signature is suggesting otherwise we may want to give it a scan or two and see what happens.”

The sheer number of variables is daunting. Laws called it “tornado alphabet soup.”

“What we’ve developed here is a system where we are trying to better incorporate those variables – environmental variables – instability, wind shear, cloud base height … we’ve looked at 60, 70 parameters,” he said. “It’s not a yes or no answer. It’s a sliding scale. So you have to look at everything in perspective. You can’t just look at one thing by itself.”

The science continues to evolve. Now the technology has to catch up.

“We’re not there yet,” Laws said. “We haven’t caught up to the technology because it’s expensive … So what you have is part of the populace who have the ability to know where the storm is and know where the polygon is, but you have another part that relies on county-based warnings still. So we’re in this in-between period.”

Weather radios and outdoor sirens are county-based, which means that they will go off countywide – even if the storm is nowhere near your location.

“So me as a scientist, as a warning forecaster, what do I do? I’ve got to warn for the lowest common denominator, which is county-based,” Laws said. “So on some level I still have to do what I have to do as far as who is going to receive those. We’re trying to save everybody. And we know there are people out there who are totally relying on weather radio and/or the sirens.”

Spann has been very vocal about the dangers of reliance on outdoor sirens, which can’t always be heard when you are inside.

“There are days I feel like I beat my brains in, and you go out and you speak, and you get an email that says, ‘I’d like to know who to talk to about getting more sirens in my neighborhood’,” he said. “I used to be nice and I’d give them some name with the Emergency Management Agency, but now I say ‘Are you crazy?’ We believe that people have to have a layered approach to methods, because anything can fail.”

When a severe weather warning is issued, “The most common reaction is the need for secondary confirmation,” Myers said. “They want a second opinion, either another credible source or to see it for themselves. With this knowledge, we now recommend that people get their warnings from multiple sources – at least two – so they can have secondary confirmation as quickly as possible.”

Nadler said they are also learning more about how people perceive and react to the wording in warnings.

For example: the use of “tornado emergency” when a tornado is on the ground.

In the past that wording was buried in the warning, he said. They advocated to move it to the headline of a warning to convey the potential danger of an approaching storm. They used that method for the first time on April 28 and it proved beneficial, he said.

Also effective is using social media and photos to add more layers of awareness, Nadler said. “People will respond to that,” he said.

You may get the warning. But then what will you do?

“The message is there if you want to hear it,” Laws said. “That’s the thing. Do you want to hear it? And I’ll suggest that if you live in this state you probably do hear it on some level and you probably are more engaged than most people. These are things that can save your life.”