Forecasts may soon put real-life spin on storms

The day after Tom Hultquist arrived in the Twin Cities in 2007 to work at the National Weather Service (NWS), a bit of snow turned his 30-minute commute into a three-hour slog.

Now he and other forecasters are looking for ways to combine winter storm forecasts with a more precise sense of the disruption they might mean for the metro area. Their aim is to develop a Winter Impact Index, so that winter storm forecasts might reflect the practical differences between 2 inches of snow at 8 p.m. (an easy brushoff) and 2 inches at 4 p.m. (metrowide gridlock).

"We want to correlate what we forecast to what happens to people," said Hultquist, chief science officer in the Chanhassen office of the NWS.

In pursuit of their index, Hultquist and forecaster Lisa Schmit have begun boring through voluminous traffic data gathered by Minnesota Department of Transportation pavement sensors and cross-checking it with weather data.

Today the agency generally issues winter weather advisories, watches and warnings that are nearly nuance-free.

Warnings are issued for expected snowfalls of 6 inches in 12 hours or 8 inches in 24 hours. Advisories are for expected snow of 3 to 5 inches in 12 hours. Over the past two winters, 40 percent of the 35 storm events prompted no alerts at all.

A Winter Impact Index could consider the time of day and the day of the week, on top of the expected snowfall. That might yield warnings that would give rush-hour commuters an estimate of how much longer their drive could be.

The NWS issued a warning -- its highest alert -- for that storm, and indeed, 7.4 inches of snow fell over the two days. On the second day, the accident rate decreased markedly, particularly in the morning rush hour, even as the temperature dropped and winds picked up.

Compare that with the scene a month later, Jan. 7, 2010. The snowfall amounted to only 2.2 inches, and the NWS didn't issue any alert. But the metro area saw 752 accidents that day, nearly two-thirds as many as in the two-day, 7.4-inch snow.

Did the alert make the difference? Or maybe people adjusted to conditions on the second day of the December storm?

"Driving is going to be tough in a winter storm. A lot of people already know that," Hultquist said. "But not everybody. We're trying to see if there's anything else we can do to get the response we want."

Custom-made forecasts?

More fine-toothed winter weather warnings would also indicate an expanded role for the Weather Service in what Hulquist called "decision support." That means forecasts better tailored to the needs of highway departments, emergency managers and other public agencies, similar to what private forecasting firms now provide to sports teams, farmers, trucking companies, investors and others.

The Nov. 20-21 ice storm illustrated the need for precise forecasting, even on weekends. Freezing rain hit the metro on a Saturday night, about five hours before predicted. The result was 376 accidents in the metro area. By the time MnDOT salt-truck drivers were called in to work, highways were already choked with crashes and slow-moving traffic, so they were late getting out on the roads. But Hultquist said Weather Service forecasting tools keep getting better, and that better practical interpretation should follow.

MnDOT spokesman Kevin Gutknecht endorsed that idea.

"We use a lot of different tools to try to figure out what the weather is going to do," he said. "If there's a way to get more accurate and timely weather reports, it would help us do a better job of clearing the roads."