June 2003

Pinpoint Weather

Cheap computer power and high-tech observation systems mean precise forecasts, offering consumers personalized reports and saving weather-sensitive businesses millions.

By David H. Freedman

Television meteorologist Paul Douglas remembers the day back in 1997 when he had the inspiration that led to the launch of his company. He had predicted on-air a rainstorm moving through Minnesota's Twin Cities, only to be confronted off-air by a flood of e-mails from local viewers wanting to know how the storm might affect their plans for the day. "It was so frustrating," he recalls. "What time will it start raining in my town?' I'm driving north; will I beat the rain?' My wedding is this afternoon; will it be rained out?'"

Douglas was helpless to answer such queries. His forecast was based largely on information from the National Weather Service, which predicts the conditions for a 12-kilometer by 12-kilometer area and whose predictions have nothing to say about how the weather varies within that area. And even if Douglas could make such localized forecasts, there was no way he could disseminate the personalized information effectively. And then it hit him: maybe he could give everyone a customized weather report. Two years later, in 1999, he founded Minneapolis-based Digital Cyclone, which predicts weather events over six-kilometer-by-six-kilometer areas and offers the information over mobile phones.

Digital Cyclone is just one of several companies taking weather forecasting to new levels of usefulness and precision. The Weather Channel in Atlanta, GA, for one, now provides eight-kilometer-resolution maps and an alert service for desktop computers. And AccuWeather in State College, PA, generates one-kilometer-resolution weather maps that are available on personal digital assistants and Internet-enabled phones. Fed by the availability of vast reservoirs of cheap computer power, new mathematical techniques for fine-tuning weather models, and high-tech observation systems, these firms are exploiting the much neglected facet of weather forecasting Douglas calls the "short game": that is, advising people about the particular weather conditions in their individual neighborhoods or towns, not just their regions.

The move to such higher-resolution forecasts could translate into huge savings. According to John Dutton, dean emeritus of Pennsylvania State University's College of Earth and Mineral Sciences, over $3 trillion of the nation's annual economy is affected by weather events. Farmers, construction workers, snow removal crews, energy maintenance workers, railroad dispatchers, and truck drivers depend on accurate and precise forecasts to effectively manage their time and resources. An unexpected cool air mass on a summer day could stick a power company with millions of dollars of unused electricity. A change in wind speed could modify a farmer's choice to spray fertilizers, which can disperse and even ruin neighboring crops in winds above 11 kilometers per hour. A minor temperature difference can determine whether a snow removal crew will lay down sand or salt. Salt is generally only effective above -7 C, and a wrong decision to use it can not only be ineffective at reducing ice but also waste thousands of dollars of taxpayers' money.

"Whether it's for employees or soccer moms, there's an insatiable demand for weather information," says Douglas. "A lot of incredibly useful information is already available, but there's a real opportunity to filter it, make it more timely and detailed and accurate, and provide it in more useful form."