Just 16 minutes before a gigantic twister first developed near Oklahoma City on Monday, the National Weather Service put out a tornado warning.
That doesn't sound like very much time to get out of the way. For many, it wasn't: At least 24 people died when the tornado ripped a mile-wide path through the city of Moore, Okla.
But those 16 minutes actually represent an enormous advance for weather science. Back in the 1980s, the average tornado lead time was a scant five minutes. Today, it's about 13 minutes.
What's more, meteorologists are now able to issue alerts and storm forecasts even earlier, thanks to powerful computers that allow them to run detailed weather simulations. The Oklahoma City area had been identified as an at-risk area days before the twister actually struck. And the National Weather Service's Rick Smith issued an eerily prescient forecast at 11:30 a.m. Monday, alerting people to the threat of tornadoes that very afternoon.
So could tornado forecasts keep improving? And how many lives would better warning systems actually save? A few points here:
Tornado detection is likely to improve in the years ahead. Scientists are developing a wide array of technologies to fine-tune their forecasts, from advanced radar systems to lightning mapping arrays that will help peer into the innards of storms. What's more, the National Oceanic and Atmospheric Administration is set to upgrade its computers that analyze weather data. That could all help lengthen the lead time for tornado warnings.
"Anything that gives us even a few additional minutes is extremely valuable," says John Trostel, director of the Severe Storms Research Center at Georgia Tech.
(That said, there may be some questions about the government's ability to conduct forecasts in the near term. The National Weather Service is facing an 8.2 percent cut from sequestration that could "trim already financially-depleted programs critical for maintaining and improving the NWS’ weather capabilities.")
But there's a limit to how accurate those predictions will ever be. "We will never have the degree of resolution we need in atmospheric variables and measures in order to achieve some perfect forecast," explained Greg Carbin of the National Weather Service Storm Prediction Center in a 2011 interview with Scientific American. "It's just not possible. …We might be able to get, say, an hour lead-time on a tornado."
While meteorologists have made enormous strides in predicting tornadoes once an individual storm has actually formed, that's still extremely difficult before the storm appears on the radar. (The National Weather Service operates 120 radar sites around the country.)
Here's Carbin's explanation for why it's so difficult: "There are so many mild adjustments, slight adjustments that can make a huge difference in whether you end up getting the formation of storms. The sensitivity the atmosphere has to ingredients in the formation of tornados and magnifying that slight change in something we can't even observe can have a dramatic impact on the forecast."
Better warnings are one thing — getting people to listen is another. Trostel points out that storm researchers are paying just as much attention to the sociological aspect of warnings. "That's an area people are working hard at," he says. "How do we get the warnings to people? How do we get them to pay attention? How do we get them to do the right things?"
This is trickier than it sounds. For instance, many people still rely on outdoor sirens, which are meant to be a warning system of last resort and are often difficult to hear in high winds. What's more, the weather service doesn't want to "over-warn" the public or produce too many false alarms — otherwise, there's a risk people will tune out.
Even a good warning system can't help if there's no place to hide. It appears that not many houses in Moore had basements or storm cellars. (Here's why.) "An F5 tornado" — the twister in Moore was likely a 5 on the Enhanced Fujita scale — "will wipe those houses right out," Trostel says.
The Oklahoma government has been offering $2,000 rebates to some residents to help defray the cost of building "safe rooms" in their homes, which research shows increase the likelihood of surviving a severe storm.
But even as the population of Moore has tripled since 1970, and even though Oklahoma has historically been at high risk for tornadoes, the local building codes don't require these rooms.
Then there are the elementary schools in Moore, which lay directly in the path of both a deadly 1999 tornado and Monday's twister. Those schools don't have underground shelters, either. (Many students at Plaza Towers elementary school were moved to a nearby church Monday, although the third-grade class is still unaccounted for.)
Over at Dot Earth, Andrew Revkin posts a note from Kevin Simmons, an economist at Austin College, who notes that there were two elementary schools in Moore a mile apart that suffered very different fates. "As of this writing," he notes, "there are no fatalities at Briarwood and many from Plaza Tower." Was that because of the way the storm shifted? Was it engineering? A difference in evacuation procedures? Or something else?
It's still way too early to say for sure. But it's a reminder that even our much-improved warning systems aren't always enough.