Photo by Kevin A. Roberts
We’re in one of Saint Louis University’s creaky old buildings, in an office once occupied by a theologian, and thunder’s rumbling in the distance.
On Bob Pasken’s high-res monitor, it’s caught in a swirl of glowing red and orange. With a few keystrokes, he can consult more than 100 sensors and drones and tell me when the heavens will open and just how apocalyptic this particular storm will be.
It all started as a prediction of sneezes.
Pasken, a meteorology professor, agreed to forecast high pollen counts for a local consortium. Bureaucracy snagged the project, but by that time Pasken had calculated not just the days of high pollen but also when and where it would peak.
So when a compressed-gas facility exploded on Chouteau in 2005, he wondered whether he could track where asbestos had scattered. He figured it out—but nobody knew he had the data.
In July 2006, he predicted a storm with 70-mph winds that would knock out power—but only the National Weather Service can issue severe storm warnings. About 500,000 people were out of power for up to 10 days in 100-degree heat.
"They stalled Game 6 in the World Series because of Quantum rain predictions," says Pasken.
Five months later, half an inch of ice coated surfaces all over the metro area, zapping power from hundreds of thousands of Ameren customers. Pasken and a colleague set up an appointment. By 2008, SLU was providing Ameren Missouri with forecasts from a system that they called QuantumWeather. It meshed Ameren’s vital data—where the power lines and transformers are, where the trees are, how healthy they are, when they were last trimmed—with a neighborhood-by-neighborhood weather model. The approach was so novel, they patented it.
Ameren was especially worried about freezing rain, heavy wet snow, and the high winds of spring and summer thunderstorms. “Leaves on a tree act like a sail and cause more problems,” Pasken says. “They push the tree back and forth—and into the power lines. If the wind blows just right, a power line gallops, going up and down a few feet.”
The first month that QuantumWeather was up, a test came. The National Weather Service was predicting that Hurricane Ike would pass north of I-44, Pasken says. “We said, ‘No, it’s going to be south.’ Ameren had to make a decision. They sent everybody to Cape Girardeau, and that’s where all the damage occurred.”
In 2011, Pasken says, “we were forecasting for the Cardinals. They stalled Game 6 in the World Series because of Quantum rain predictions. The Cards rested, won Game 6, and went on to win the series. Not that I’m taking credit…”
He savors his own wins, because a few times he’s felt like they blew it. “In 2015, we said it was going to be in Washington County’s northwest corner, and it actually happened in Potosi,” which is in the middle. “For us, busted forecast—we missed it by 20 miles. But Ameren said, ‘Nah, it’s perfect.’ They just want to know how to allocate resources.”
Why aren’t forecasts ever rock solid? “Because I don’t have a perfect description of what’s happening now, I can’t make a perfect forecast,” he says. “I only have sensors. One here, one here… What’s happening between those two points? The National Weather Service says, ‘Today, it will be 68.’ Well, 68 where? They forecast for a particular location. A summer thunderstorm will cause a rain shower—well, which thunderstorm? The one over Brentwood, or the one over Waterloo?” Snow’s more widespread, but it’s still easy to miscalculate. “People will say, ‘Yeah, well, there’s 2 inches of partly cloudy in my driveway waiting to be shoveled.’”
To smooth the uncertainties, QuantumWeather makes an “ensemble” forecast, changing variables slightly—maybe the temp’s a little higher, the wind a little slower—then looking at the outcomes. The narrower the window, the higher their confidence.
“Over the decades, the error bar has gotten smaller,” Pasken insists. “Radar’s better.” The sensors keep improving, too. QuantumWeather now has three balloon-borne sensors and data from drones. “They can fly anywhere—following FAA regulations, of course,” says Pasken. “They’re reusable—unless you fly them into a wall.”
I chuckle. “No,” he says. “I did that Tuesday.”
I ask why TV weather’s hyped to the point of hysteria, and he chuckles. “Ameren called me once and said, ‘Are we really going to get three-quarters of an inch of ice on our power lines?’ I said, ‘What the heck are you talking about?’ [TV forecasters’] budgets and jobs all depend on being able to get somebody to watch TV. My mother-in-law watched those forecasts, and not even I could convince her not to worry.”
Not even a man who’s flown straight into the eye of a hurricane, taught undergrads how to drop sensors into tropical storms, traced the role of the Saharan sands in tropical cyclone formation.
What weather event’s been the scariest for him?
Scary?
“OK, most dramatic.”
“In ’06, we were flying out of Africa on a NASA DC-8, and it got hit by lightning that fused parts of the airplane together, like, flaps and rudders. And in St. Thomas, we sucked a cowbird into one of the engines on takeoff, and it set the engine on fire.”
He grins, thrilled to have weathered it all.
