Sorry about that White Christmas, Philadelphia. It’s not looking good; temperatures are going to be at least in the 40s. But don’t despair, snow-lovers; December is going to end up snowier than average.
Looking further ahead and afield: Come March, warm weather will boost shorts sales 28% in Chicago over last year, but the shorts business will be off a cool 15% in Tokyo.
Oh, and good news for local travelers: Flight delays at Philadelphia International Airport will be down 42% next year, compared with 2019, thanks to considerably less precipitation.
That’s all according to the voluble Bill “Captain” Kirk (yes, he’s a Star Trek junkie), cofounder and CEO of Weather Trends International, a Lehigh Valley-based company that sells long-range — as in 11 months in advance — forecasts to retailers and investors.
In fact, the company issues those long-range forecasts for every square mile of the planet. Once a customer purchases an 11-month outlook, the company doesn’t change or update it.
Home Depot Canada has been a customer since 2017, a spokesperson says, and Kirk claims a long list of clients that have used his services, including retail giants such as Target and Walmart, investment houses that include JP Morgan, and Coca-Cola. Kirk put the privately held company’s annual revenue at $5 million to $10 million.
Reliable year-ahead forecasts would be invaluable to retailers. “Once merchandise has been put out for the season,” said Sucharita Kodali, an analyst and executive at Forrester research, “the cake is baked.”
Has Weather Trends found the silver bullet? Scientists who have wrestled with the vicissitudes of the atmosphere aren’t buying. They view the long-range-forecast problem as a mechanical rabbit the hounds may never catch.
Kirk recalled that Paul Knight, former chair of the Pennsylvania State University meteorology department and the erstwhile state climatologist, once called his forecasts “dartboard” science. Gary Szatkowski, former head of the local National Weather Service Office in Mount Holly, mused that if those 11-month-ahead forecasts were as good as the site claims, why wouldn’t Kirk just go off and make a killing in the futures markets.
Kirk dismisses the skepticism.
“That’s the problem with meteorology," he says. "We can’t get out of our own way.”
The energetic Air Force veteran, who studied meteorology at Rutgers University and speaks with a rapidity that suggests he is driven by an overcharged battery, counters that they don’t know what he knows.
He says his methods eschew the standard physics and computer modeling used for conventional short-term forecasting, acknowledging that he was no fan of the calculus and assorted, unsightly meteorological math of his undergraduate education.
Kirk said the formula is about two-thirds statistical analysis and the rest an analytical blend that includes weather “cycles,” primarily slow-occurring shifts in ocean temperatures and air-pressure patterns, including the El Niño/Southern Oscillation and the Pacific Decadal Oscillation, not to mention 417 trillion bits of data.
For example, he cites March 2016, which was quite warm in the Northeast. Temperatures in Philadelphia were 7-plus degrees above normal. Business would tend to assume March 2017 would be a repeat, he reasons, but based on statistical analysis, after a warm March in the Northeast the next one will be colder seven out of eight times.
He says his temperature forecasts can be calibrated to retail sales: A one-degree difference makes a 15% change in air-conditioner sales and a 7% difference in the sunscreen business.
Why doesn’t everyone try this? He said it’s a matter of “training.” Traditional forecasting is mired in “the physics-based meteorological model,” but not much help beyond the short term.
In embracing statistics, Kirk is onto something, says Kyle Imhoff, a Penn State researcher who is the current state climatologist.
“It is certainly a scientifically sound approach to longer-range forecasting that goes beyond what our numerical weather prediction models in the short to medium range are currently designed for,” he said.
He points out that other private companies, academic researchers, and the government’s Climate Prediction Center all are using statistical analysis in long-term forecasting. The center “relies strongly” on climatological data, said Szatkowski.
Kirk says his company has found a better way.
Kirk says an auditor, whom he did not identify, showed that his extended temperature outlooks are more accurate than other 7- to 14-day forecasts — an error margin of 2 to 6 degrees, compared with 4 to 8 degrees.
However, those one- to two-week outlooks remain a challenge, something easily documented by looking at how they have performed, and researchers believe that for now 10 days is probably the limit for reliable forecasting.
The problem? Computer models have to rely on the observations that are fed into them. The world is imperfectly observed.
“The atmosphere is a chaotic, nonlinear system,” says Imhoff. “Any errors in attempting to capture the current state of the atmosphere will grow significantly over time."
A study by two climatologists concluded that for temperatures, the Weather Trends’ year-ahead forecast outperformed the climate center’s three-month forecasts, which are updated monthly, four out of five times.
A caveat: Lead author Laurence Kalkstein, a University of Miami emeritus professor and respected climatologist known for his pioneering work in heat-related mortality, is also a member of the Weather Trends advisory board, but said his affiliation did not diminish the validity of the study.
It should be noted that the government’s forecasts are cautiously confined to probabilities, which the climate center says are dictated by the limits of science.
Don’t rule it out yet. The outcomes of Weather Trends’ snow forecasts for 2018-19 were a real wintry mix: off in Philadelphia, Boston, and Casper, Wyo., on target in Harrisburg, St. Louis, and Des Moines.