Q&A: NWS's science officer on how technology is changing forecasting, warning systems
It seems to the layperson as if we've had an unprecedented number of destructive weather events recently. Is that perception, or is it reality?
I think it's both. We're living in a much different society than we did 10 years ago. Turn on your cable and how many 24-7 news channels are there now versus 10 years ago? There's a whole lot more, and they're all looking for eye-grabbing, attention-getting stories to draw you in. Now, you've got storm chasers and thrill seekers, a healthy mix of both of those out there. They can film a tornado on their camera, put it on their laptop, put it up on Facebook, put it up on Twitter and the media can grab it. It's instantaneous weather information that we've never seen before, and it's led to a perception that all these catastrophic weather events are occurring now and they never did in the past.
But I want to step back and say this is a really active period we've been in for the country in the last three to four months. It is abnormal. We've issued at this office and at Weather Service offices across the country a very high rate of severe thunderstorm warnings and tornado warnings, and there have been a number of destructive tornadoes. How has technology changed your ability to predict the weather or warn the public of dangerous situations?
Back in the mid to late 1990s, the network Doppler radars were installed across the entire country. They have probably been the biggest technological improvement with their capabilities to see hazardous weather. And then also in the late '90s and 2000s we revamped our entire computer system to AWIPS -- Advanced Weather Interactive Processing System. And that piece of software allows us to integrate radar, satellite, computer model data, and observations all on the screen at the same time -- which allows us to verify what the computer models are trying to tell us.
We're upgrading that radar system to dual polarization, which is going to be much better at discriminating "hydro meteors" -- small raindrops, large raindrops, hailstones, non-meteorological targets like dust, bugs – basically anything that's flying around in the air that can scatter back radar energy. For example, are we seeing hailstones in that thunderstorm or just very large raindrops? We've also seen here locally, and at some other offices across the country, implementation and use of FAA terminal Doppler weather radars. And these are radars that sit near major airports across the country. They were deployed by the FAA near airports to be able to see wind shears and downbursts and detect those in real time. We now get live access to those radars and we're able to see parts of thunderstorms like downburst signatures, or very small rotational couplets that may be indicative of a tornado.
Another area where technology is really evolving is the ability to use very high resolution, short-term computer models to predict thunderstorm initiation and longevity and thunderstorm mode. Most of the big, longer-term computer models used to forecast day-to-day weather run on a horizontal resolution of anywhere from 20 kilometers to 60 kilometers. There are computer models out there now that run at horizontal resolution of three to four kilometers, which allows the computer model to simulate (real atmospheric conditions). That's gone a long way in helping us to say, for example, what's the most probable area for thunderstorm initiation and development and is it going to evolve into a squall line with wind damage, or is it going to be discrete super cells with the potential for damaging tornadoes?You spoke about societal changes. I'm wondering if there are things like shifts in population density that have changed how you work with communities.
Definitely. The Weather Service has a Storm Ready program. I don't oversee this area in our office, but there's a series of things that a community has to have in place to be declared storm ready. And it's a way for us to get the word out about storm safety and it allows the community to say 'hey, we're a storm ready community so that if a storm strikes, here are the things we have in place as a form of community response.' What we're doing locally with emergency managers is we provide briefings to them -- it's called decision support -- and we're interacting with our emergency managers in all of our counties and providing them information before significant weather events so they can go to their communities and they can go to their local governments and implement a disaster plan or disaster response plan before the weather actually hits. That's something that's actually taken off in the last couple of years.Have these advances helped to reduce deaths and injuries?
Unfortunately we never know about the lives we saved, we only learn about the lives we've lost. It's an interesting point that with weather disasters you can give people all the warning in the world and provide them with as much detail about actions they can take, but it all comes down to how that person takes the information and what they choose to do with it. There's the segment of the population that goes into a panic mode and runs down to the basement and stays there for six hours. There's the other side of the spectrum where, when there's a tornado warning, they run out with a video camera and even contemplating getting in their car and going out to try to find it. There's a huge social science issue here about how people take warning information and what they do with it. Is there research going on along those lines?
Very much so. I think we're finally understanding there's a lot more to this than just the meteorology of why tornados form or why thunderstorms form. We feel like we understand a lot of that pretty well, but now how do we get people to act? What are some capabilities that aren't possible now but that we may see 10 years from now?
There are a number of things out there that right now, admittedly, seem kind of pie in the sky. I was telling you about these real high-resolution computer models that can run at three to four kilometers. There are ideas out there that with computing power and resources we'll be able to run computer models down to one kilometer resolution, and not just run one model but 30 different computer models, run them all at one kilometer resolution. These storms haven't developed yet, but you're almost issuing a warning type product . . . instead of telling people in an area 15 to 30 minutes in advance of a thunderstorm, you're giving them an hour to two hours of lead time. I think radar technology is going to continue to improve. And I would say we'll see more integration of social sciences and social media into our forecasting product. To be able to take all the wonderful things that come with social media on Facebook and Twitter and turn that information around and get it out to the public in a much more efficient way than we do right now -- I think we're just on the cusp of that.