Tracking Killer Tornadoes

Transcript

FLORA LICHTMAN, HOST:

The tornado that struck Oklahoma this week was classified as an EF5, highest ranking on the Enhanced Fujita scale. And if you've seen the pictures from Oklahoma, you have a sense of how powerful it was. And this storm visited towns that have been struck in the past. Moore, Oklahoma, had been hit by another EF5 in 1999. In fact, if you look at a national map of tornado probability, you'll see a kind of bull's eye radiating out right from around Moore. But why is that? What makes some parts of the country so susceptible to tornadoes? Basically, why does Tornado Alley exist at all?

Joining me now to talk about how tornadoes form and how our understanding of tornadoes is linked to our ability to predict them is Marshall Shepherd. He's the president of the American Meteorological Society and the director of the Atmospheric Sciences Program at the University of Georgia. Welcome back to the show, Dr. Shepherd.

MARSHALL SHEPHERD: Oh, always happy to join you.

LICHTMAN: So tell us why. Why is it that some parts of the country are so twister prone?

SHEPHERD: Yeah. Well, the United States itself, we have this dubious honor of sitting right in the collision zone between really cold air masses coming down from Canada and moist, warm air that comes up from the Gulf of Mexico. And so the well-known Tornado Alley is a convergence zone, couple that with the position of jet stream and other factors that we as meteorologists look for, a dry, mid-layer or wind sheer, and you have the perfect ingredients for tornadoes.

Most tornadoes affect the United States east of the Rocky Mountains. And in fact, there is another alley we call Dixie Alley because if you look at a map of climatological occurrence of tornadoes, there are areas in the South as well that seemed to get quite a few tornadoes. But certainly this region, as you were leading into the story, was an area that we would've expected tornadoes for May 20th based on climatology.

LICHTMAN: Well, let's talk about the forecasting because one of the amazing parts of sort of science of this story was that hours before, the National Weather Service locally issued a warning that there might be a tornado. And it was pretty spot-on, it sounded like.

SHEPHERD: Yeah. You know, even days before. I mean, with this region was under the gun, we had outlooks noting this particular area would be prone to these types of storms on that particular date. Even that day, as you noted, some of our high resolution models indicated the possibility of tornadic storms. I know the National Weather Service put out some public communications, even warning, I think, somewhere around 11, 11:30, noontime hour that there was a possibility of storms later that afternoon. And even special warnings for schools.

So I want to give out a shout out to the colleagues and members of AMS there working at the National Weather Service and in the broadcast community because they were on top of it from the beginning. And I want to also emphasize that the Weather Service put out warnings 16 minutes before the storm even touched down. The average warning time these days is around 13 to 15 minutes, and that's up from five minutes in 1990. So although to many in the public that doesn't sound like a lot, that's actually above average so - and even more time before the storm reached Moore.

LICHTMAN: Talking with Marshall Shepherd, president of the American Meteorological Society. I'm Flora Lichtman, and this is SCIENCE FRIDAY from NPR. So what accounts for this change, this bettering of our prediction time?

SHEPHERD: Well, there are so many things to look for. I mean, if you talk - and I was talking to Harold Brooks of the NSSL, and he talked about the increase in being able to detect storms. There are things like growth and increase of people observing the storms. You know, so the probability of detection has gone up.

But one of the other things that I should mention is over the last several decades, we've seen our radar network go from essentially a pretty basic system that can only detect rainfall to a system that could actually detect movement of the raindrops in the cloud. That's the Doppler radar capability so we can actually see circulations, the so-called mesocyclone in these super cells. And that's very important. And often times now, storms are warned based on a Doppler indicated signature.

When I was a kid, someone actually had to see the tornado on the ground. Or a spotter had to note it. Now, the other thing I should note is that more recently, the National Weather Service has upgraded its radar network to something called dual polarization radar. And that's just a fancy term that allows us to see various aspects of the tornadic storm and the rainfall in ways that we couldn't see before because it can actually see the storm from double perspectives, from a horizontal and vertical perspective in terms of the - what we call polarization signature of the radar.

LICHTMAN: I read that one of the things the radar is looking for is flying debris. Is that right?

SHEPHERD: Yeah. Yeah. And one of the things that the dual polarization radar can do with a couple of the signatures or products, is it actually can help to see that debris ball. You can somewhat detect that if it's a really well-defined debris ball in the reflectivity, the sort of old school way of looking at the radar data because as debris is being lofted in the tornado, debris actually has a radar cross-section. You can see it in the same way that the radar, quote, unquote, "sees the raindrops."

But using correlation coefficient and some of the other advanced products, we can actually get a bit more information on that debris ball so we can, you know, there are times it's not clear whether that circulation in the Doppler signature is actually on the ground as a tornado. Using the debris ball and the correlation coefficient and other products together, we really can improve our ability, perhaps, to detect the tornado on the ground. And that certainly leads to improved warning.

LICHTMAN: Hmm. You know, whenever there's a big storm, it feels like climate change gets thrown into the mix. Is there any evidence of a link between tornado frequency or strength and global warming?

SHEPHERD: Yeah. I mean, you know, I've been all over places talking about this because clearly as a scientist who understands the peer review literature and the president of the AMS, who recently released a 2012 climate change statement, I know that climate change is an issue and there are things that we need to be concerned about with human-caused climate change and the natural climate change.

But on tornados, the peer review literature is just not very mature on this right now. It's much more mature on the link between climate change and drought or heat waves or even the intensity of rainfall, and even more so as well on hurricanes in terms of the intensity of hurricanes, perhaps, increasing in the future. And if I were to cascade along that spectrum, scientific literature relating individual tornadic occurrence in climate change is very immature at this point

I think we have to be careful on that jump, although I know there's a tendency to do it. There is great deal of research ongoing right now at places like Purdue University or even by one of our doctoral students, Victo Gensini, here at the University of Georgia, looking at climate change and the type of environment that might emerge that could support, you know, tornadic storms, wind shear profiles that Jeff Trapp and others at Purdue have been doing this work as well. So that's sort of where we are, and I think we just have to be very careful on sort of causality for individual tornadic storms.

LICHTMAN: And that's where we have to leave it. Thank you for joining us today, Dr. Shepherd.

SHEPHERD: Thank you for having me, Flora.

Marshall Shepherd is the president of the American Meteorological Society and director of the Atmospheric Sciences Program at the University of Georgia. Transcript provided by NPR, Copyright NPR.