photo by Erin Nekervis
If you are somewhere east of the Mississippi River this weekend, shivering as the harshest Arctic outbreak of the season threatens to break records in many places, it may interest you to know that your misery is associated with record warmth at the very top of the atmosphere, at the very top of the world.
Of course, this may not interest you in the least unless you share my lifelong fascination with how the atmosphere works. But since climate is a hot topic in the world of politics and economics, it is worth being aware that the weather we experience is the product of an extremely complex delivery system whose details are still, in significant part, beyond the realm of human knowledge. But we are getting closer to figuring it out.
Since early last year, I have been following the Arctic Oscillation blog written by Judah Cohen, a Ph.D. who works at the Atmospheric and Environmental Research division of New Jersey-based Verisk Analytics. The AO, as meteorologists call it, is closely related to (some would even say just another manifestation of) the North Atlantic Oscillation, which is a major influence on winter weather in eastern North America and western Europe. Grossly oversimplifying, when the AO/NAO is in its positive state, much of the high Arctic is colder than average, especially the region from Greenland to the North Pole. When the AO/NAO is negative, the Arctic is warmer than normal and a mound of relatively warm and dry air tends to park itself over Greenland. This disrupts the circumpolar wind flow and causes it to buckle southward, transporting frigid air from the Arctic to the mid-latitude continents where a lot of us live. This southward-displaced gyre of cold air is the famous “polar vortex,” which sometimes may actually split into two or more vortices at various locations around the northern hemisphere.
As you would expect, the AO/NAO tends to oscillate. It typically switches from positive to negative every few weeks. But sometimes it maintains a given state more or less continuously for months at a time, changing only briefly and mildly to the opposite condition before reverting to the norm it seems to have chosen for itself in a given season. Climatologists have long known that if they could predict the favored state of the AO/NAO for an upcoming winter, it would take them a long way toward forecasting the severity of the season.
Think how much more efficiently we could plan our lives if we could count on reliable long-range seasonal forecasts. It would have huge impacts on everyone from energy traders, to the people who manage snow blower inventories at retail chains, to highway departments planning their budgets for overtime and road salt. But unfortunately, up to this point computer models have had limited skill at forecasting the AO/NAO more than a couple of weeks in advance.
About a decade ago, researchers discovered that there seems to be a correlation between how rapidly the snow cover advances southward across and beyond Siberia in the autumn and how much time the AO/NAO spends in its negative state in the ensuing winter. Faster snow cover advance seemed to presage a more negative AO/NAO state, and thus harsher winters in eastern North America and much of Europe. But how and why does that happen, and what other factors also influence the outcome?
In his blog, Cohen has been studying the connection between Eurasian snow cover advance in the autumn (specifically in October) and the occurrence in midwinter of a “sudden stratospheric warming” event over the pole. The hypothesis is that extensive autumn snow cover can initiate a chain of events that, typically in December or early January, forces energy to rise through the polar atmosphere until it warms the stratosphere, disrupting or displacing the polar vortex and allowing cold air to escape southward. This energy then ricochets downward to the surface, warming the high Arctic and forcing the AO/NAO into its negative state, tending to lock that cold air into place over our shivering heads.
Cohen has written for months that the conditions seemed to be in place for this to occur this season. But there were other variables at play, notably the most intense El Nino phenomenon to be observed in the past two decades. El Nino is typically associated with warmer than normal winter weather in the north-central and northeastern United States. How would El Nino and the potential for sudden stratospheric warming affect one another?
The jury is still out on this point, but here is what we know so far. In December the AO/NAO stayed positive, and much of the eastern U.S. and adjacent Canada experienced record warmth. It was also mild in much of northwestern Europe. These are typical El Nino patterns. The energy transfer over the polar region seemed to be delayed.
But by late December, Cohen noted in his blog that the first significant pulse of energy was occurring. It was followed by additional transfers in January. The eastern U.S. turned colder, and the combination of this cold air and a classic coastal storm (favored by El Nino) brought a record blizzard to the mid-Atlantic states. This week the polar energy pulse reached the top of the atmosphere, where record warm conditions were measured in the stratosphere. And the polar vortex has pushed south into Quebec, dragging frigid air with it. Cohen is now forecasting that we are likely to see a negative AO/NAO over most of the remainder of the winter, which is very good news if you operate or frequent a New England ski area.
These phenomena have effects all over the globe. December’s warmth in the American East was matched by cold and welcome snowfall in the drought-stricken West, but right now the storms are blocked away from California by a ridge of warm and dry air. That warm air extends north to Alaska, which had a cold and snowy autumn. In east Asia, the early Siberian snow has fostered a cold regime for most of the winter. Hong Kong residents who climbed the city’s highest peak in late January encountered an ice storm that is exceptionally rare in a place that sits at a more southerly latitude than Havana.
Regular readers know that I consider myself a climate change skeptic. Not that I question human influence over the atmosphere; because we have changed its physical and chemical composition, it is almost a given that we have changed how it responds. But I strongly question our current ability to accurately model its future behavior over years and decades, when we are still trying to decode how its various components will behave just months from now. A changing sea level, for example, does not depend only on how much ice might melt in Greenland; it is also affected by how much snow is deposited each year in Antarctica, and by how much moisture is retained in a warmer atmosphere and by changing vegetation patterns on land. Storm tracks and continental temperatures do not depend solely on total atmospheric heat; they are greatly affected by variations in local cloud cover and by broad atmospheric patterns such as the AO/NAO. Every climate model necessarily makes assumptions about these things. These assumptions are likely to require adjustment, and maybe substantial adjustment, as we learn more about how the world really works.
We are slowly getting better. The Arctic Oscillation blog pretty much called this weekend’s cold snap months ago, and if the AO/NAO goes negative as predicted, it will have been a nice piece of meteorological work.