As of a few days ago in October, snowfall for the month is a bit above average over a portion of Siberia. What does that have to do with the price of tea in China? What MIGHT it have to do with what kind of winter we can expect in eastern North America?
A couple of decades ago, MIT researcher Dr. Judah Cohen started his statistical research on a potential relationship between snowfall amounts during October in Siberia and the pattern taken by the polar jet stream during the winter over and near eastern North America.
This Siberian snowfall has been heavier more often in recent decades. That increased frequency is thought to be linked to a warming climate and reduced ice cover in the arctic, which had been predicted in climate models in the 1980s and which has come to fruition. With more open water late in the summer, there is more evaporation from more parts of far northern waters, leading to greater volume of water vapor and more snowfall in October.
Cohen found a correlation between heavier than average October Siberian snowfall along with the rate of the advance of that snow cover and a tendency for what’s called the Arctic oscillation to go into its negative/cold phase more often than average during the winter. The Arctic oscillation in its cold phase often favors the polar jet stream taking a dive over eastern North America. That dive allows polar air masses to be delivered more often to eastern North America, especially the northeast and middle Atlantic states.
Cohen’s hypothesis initially wasn’t paid much attention by the National Weather Service Climate Prediction Center, which is in charge of seasonal outlooks from National Oceanic and Atmospheric Administration. But during the first decade of this century, Cohen made winter outlooks a number of years that verified far better than those from the Climate Prediction Center, based largely on his growing understanding of the relationship between excessive October Siberian snow and a more prevalent cold phase of the Arctic oscillation.
In the fall of 2009, Cohen successfully predicted a much colder and snowier winter for much of the east, particularly the east coast. His prediction came after observing much above average snowfall in Siberia during October 2009, and a rapid advance of that snow cover during October. The rate of advance of snow cover may be just as important as the actual volume of snowfall.
When this snow cover advances rapidly and builds up beyond average for the month, a deep and cold dome of high pressure builds over Siberia and its very cold surface. This mountainous ridge tends to “pump up” the polar jet to the north while the famed polar vortex weakens. The pumped-up ridge over Alaska and western North America, along with a weakened polar vortex, allows more frequent incursions of polar air to take a dive into the Great Lakes, the Northeast and the middle Atlantic states. The thermal contrast between these waves of cold air and the much warmer air offshore by the Gulf Stream can fuel the development of deep storm systems, commonly called nor’easters.
The winter of 2009-10 brought multiple East Coast storms, leading to what was called “Snowmageddon” for Baltimore, Washington and Philadelphia. Baltimore actually received more snow that winter than Buffalo, which was probably unprecedented.
However, reliance on Cohen’s research is anything but a foolproof tool in making more accurate winter outlooks every year. He has had great success some years and much less in a few other years. The interaction between his studied variable — Siberian snowfall — and other oscillations around the globe is poorly understood.
Last year, for example, the October Siberian snowfall and its rate of advance taken alone would have suggested a colder winter in eastern North America. But most of us knew of a much stronger signal that would likely overwhelm that Siberian variable: the exceptionally strong El Niño that was developing. Last winter’s El Niño was the strongest forcing mechanism leading to our exceptionally mild winter, and it actually made for higher than usual confidence in winter outlooks for most meteorologists.
What do we see so far this fall?
Siberian snowfall started to advance rapidly earlier this month, but it largely stalled as we got closer to mid-month. Dr. Cohen is expecting snowfall rates to increase in the far northern hemisphere again later this month with the help of a more negative Arctic oscillation. However, I’m not seeing much evidence of that in the extended range European model or another extended range U.S. model.
In the absence of a strong El Niño, there IS some evidence the Arctic oscillation will be able to take on a negative/cold phase more frequently later this year — even by mid-November — or once we get into December-January-February-March.
In other words, winter is not looking so benign as it did last year at this time. Cohen favors the negative/cold Arctic oscillation being dominant this winter; though, of course, he recognizes large uncertainty this early in the season. There are signs in the extended range of the European model more frequent incursions of arctic air could begin to arrive in mid-November.
By the way, a negative Arctic oscillation does not necessarily mean a very snowy winter for Western New York. It is more connected with colder temperatures for Western New York. In 2009-10, Buffalo snowfall was somewhat below average, despite colder temperatures.
Another related oscillation, the North Atlantic oscillation, may have a greater correlation with our snowfall in its cold phase. In strict terms, neither the Arctic oscillation nor the North Atlantic oscillation can be forecast very well beyond two weeks in advance. However, there may be a growing skill in predicting their tendency to be more negative or positive/warm phase more often during the course of a winter.
And all these variables, folks, is why if I had a farm I wouldn’t bet it on winter outlooks verifying very well, with the probable exception of strong El Niño years.
Don’t look for any help from the animal kingdom, either. There is zero evidence in studies actually done that animals can anticipate seasonal tendencies. Some animals can respond to changes in barometric pressure or humidity related to approaching short term storms — like many arthritic people can — but not seasons.
If you’d like to learn more about these oscillations, see this excellent page from North Carolina State University’s State Climate Office.
Story topics: By Don Paul