On Thursday night, conditions perfect for power disaster

Just the right temperature – although “wrong temperature” might be a better term – little wind and no snow to speak of: That’s the recipe for a winter power disaster in New Hampshire.

This was the finding of a 2004 study by Plymouth State University researchers, and the recent ice storm showed it still holds true.

“The difference between 33 and 34 and 32 (degrees) is all the difference in the world,” said Eric Hoffman, an associate professor of meteorology and chair of the Atmospheric Science and Chemistry Department at PSU.

“The reason why there was so much damage (this time) probably has to do with the fact that the temps stayed colder than were predicted. That’s a very difficult forecast to make in that situation.”

Hoffman and two students conducted a study in 2004 and 2005, in conjunction with Public Service of New Hampshire, looking at eight years of data about weather and major power outages. They were seeking to quantify the weather factors that contribute to major problems.

“The one surprising result was we found that the big, heavy snowstorms, the two-footers, were not on our list” of disasters, he said. “We investigated that a little bit and found they are typically colder – lower 20s statewide – and the winds stronger.”

Lower temperatures mean precipitation is frozen before it hits power lines and trees, making it more likely to bounce off than cling, while higher winds blow down any precipitation that does stick.

The other big factor in making winter storms dangerous, he said, is that the low-pressure area travels south of the state, putting us on the colder side.

Hoffman also speculated that one of the surprising points of the recent story – that it wiped out so many trees without encasing cars and roads in thick ice – was because it came so early in the season. The lack of snow on the ground and a fair amount of latent heat remaining in the earth may have kept ice from forming on or near the ground, he said.

The research findings were put into a Web-based tool for helping PSNH “better identify when significant weather was coming” – to “put the research into practice,” he said. He believes it’s still being used in some form.

The unexpected amount of damage in this storm, however, shows it and other prediction tools remain imperfect.

“There’s always going to be some amount of uncertainty. . . . Reducing that level of uncertainty is certainly possible, and scientists are working on that as we speak, and as we reduce those amounts of uncertainty, then being able to insert that kind of information into decision making becomes more valuable,” Hoffman said.

More observers and observation stations, better theoretical knowledge of meteorology and likely continued advances in computing power to crunch data will improve things, he said.

It has improved things already, he noted: “The uncertainty certainly has gone down over the years.

“People actually make decisions – ‘We’re going to close school tomorrow’ – based on forecasts. That never happened 20 years ago because forecasts weren’t good enough to make decisions a day ahead of time.”