“Siberia is kind of the refrigerator for the Northern Hemisphere,” Cohen says.
More than luck?
Cohen’s work has been greeted with interest in the forecasting community, although some researchers question whether he has come across a relationship that has real predictive value or has just gotten lucky during the past few years.
“He definitely has at least three of the years in that series where he did nail it in a way in which the forecast models did not,” says Lisa Goddard, a research scientist at the International Research Institute for Climate and Society, based in Palisades, N.Y. Noting that the years he correctly forecast were extreme outliers, she says the question remains how strong the snow cover signal needs to be to serve as a good predictive tool.
Cohen has been issuing winter forecasts since the winter of 1999-2000, and although there have been hits as well as misses, his forecasts have outperformed many other models. For example, his group’s forecasts were largely on target during the winters of 2002-2003, 2009-2010 and 2010-2011, each of which brought memorable snowstorms to the Washington area. In contrast, he did not accurately predict conditions during the winter of 2005-2006.
Cohen said his forecast for last year got the entire hemisphere correct, which challenges the notion that the Arctic Oscillation can’t be accurately predicted.
“I don’t think it’s just luck,” he said.
The winter outlook issued last month by the National Oceanic and Atmospheric Administration’s Climate Prediction Center gave the Mid-Atlantic “equal chances” of experiencing below-average, average or above-average temperatures and precipitation during this coming winter.
To most people, that forecast translates to, essentially, “we don’t know.”
The reason NOAA cited for the uncertainty was the lack of reliable tools to predict the behavior of the Arctic Oscillation, which the agency referred to as the “wild card” in the winter forecast.
Mike Halpert, deputy director of the Climate Prediction Center, said government forecasters are examining Cohen’s work to see if his methods should be incorporated into their forecasting process.
However, he said it’s probably naive to think that Siberian snow cover alone controls most of the variability in the Arctic Oscillation. “[I] have to admit that the correlations are eye-catching, but the period of record is very short,” Halpert said via e-mail. “The results warrant more study and we plan to take a close look at these results and see if these correlations hold up going forward. The cynic in me does not expect them to hold up over a longer period of record.”
A recent paper in Nature Geoscience found that low solar activity is correlated with periods of negative oscillation, whereas higher-activity years tend to favor the milder phase of the pattern. Solar activity has been low for the past several years.
Looking ahead
For the upcoming winter, Cohen is forecasting a milder than average winter in the Washington area, with near normal snowfall. He said he hasn’t “thrown in the towel” on a heavy snowstorm or two, despite his mild forecast. (The Post’s Capital Weather Gang blog is forecasting roughly average temperatures and snowfall for this winter, which means about 12 inches at Reagan National Airport.)
Siberian snow cover advanced at almost exactly the normal rate during most of October, with the exception of a dramatic expansion at the end of the month, which has Cohen a little nervous about this year’s forecast. Without a clear Siberian snow signal, he relied more on other factors to make his forecast, including the likelihood of continued La Nina conditions in the Pacific.
His message to snow lovers: Don’t get your hopes up for a repeat of 2009-2010, but don’t despair, either.
Freedman is a freelance contributor to the Capital Weather Gang blog.
Loading...
Comments