Don’t say we didn’t warn you. Anyone, especially in the northeast and Midwest, who is surprised by the arctic express that moved in over the weekend and is still gripping most of us today wasn’t paying attention last month when climate scientist Judah Cohen issued his detailed winter forecast, the subject of my column in the December 17 issue of NEWSWEEK.

Unlike the official forecasters at the National Weather Service and elsewhere, Cohen, director of seasonal forecasting for Atmospheric and Environmental Research, Inc., doesn’t base his prognostications on crude indicators such as the status of El Nino. His research shows that something much subtler—the amount of snow cover in Siberia around the end of October—is the best crystal ball ever discovered when it comes to predicting winter weather in the U.S.

The connection, as I explained in the column, has several steps. But when Cohen ran the data through his forecasting model he came up with a prediction for cold for about the first three weeks in December, followed by milder weather especially during the first half of January (for which my heating bill thanks you) and a return to cold around Martin Luther King, Jr.’s birthday. Yesterday’s Giant-Packers playoff game was one of the coldest in NFL history. As Cohen writes me, “To predict swings in the weather almost to the day two months in advance should be impossible based on accepted climate theory. After all, chaos theory was invented to explain climate dynamics.”

Dumb luck? Doubtful. Will the rest of Cohen’s winter forecast, as well as those in coming years, prove as accurate? His track record is impressive, beating the official forecasts numerous times, but only time will tell. Will government forecasters acknowledge that Siberian snow cover can be a useful forecasting tool? When I spoke to them last month they were politely dismissive. But it’s tough to argue with success after success.