در مورد زمستان سرد و برفی امسال و ارتباطش با گرم شدن زمین تا به حال چند مطلبی درج کرده ام. به نظرم دوستان علاقمند از خواندن این دو مطلب روی سایت نیویورک تایمز هم لذت ببرند. اولی و دومی! البته دقت کنید که دومی در واقع نسخه ی کامل و به روز شده ی اولی است. من برای سهولت دسترسی دوستان مطلب جدید را عینا اینجا کپی می کنم.

 ماجرا از آنجا شروع می شود که یکی از کارشناسان هواشناسی در ایالات متحده ادعا می کند که ارتباط میان گرم شدن زمین، برف در سیبری و بروز هوای سرد در نیمکره ی شمالی را شناسایی کرده است و به همراه یک تیم تحقیقاتی مطالعاتی را هم در همین حوزه انجام داده اند. در بیان دلایل وقوع این قضیه شاید پاراگراف زیر راهگشا باشد:

Researchers have validated a new weather prediction model that uses autumn snowfall to predict winter cold in the United States and Europe. When snowfall is high in Siberia, the resultant cold air enhances atmospheric disturbances, which propagate into the upper level of the atmosphere, or stratosphere, warming the polar vortex. When the polar vortex warms, the jet stream is pushed south leading to colder winters across the eastern United states and Europe. Conversely, under these conditions the Arctic will have a warmer than average winter.

Putting a Siberian Snow Connection to the Test

Is wintry winter weather more predictable than we thought? Judah Cohen thinks he’s proved that the answer is yes. He’s the commercial climate analyst who wrote an op-ed article for The Times charting a connection between global warming, snow in Siberia and outbreaks of cold weather in parts of the Northern Hemisphere and whose work was discussed here yesterday.

He has submitted a comment defending, and elaborating on, his work. It’s posted below as a “Your Dot” contribution, along with Cohen’s answers to several additional questions I sent Monday night.

Cohen points to past successful predictions of  winter temperatures in North America and  Europe as evidence that autumn patterns of snow cover in Siberia have a lot of relevance to people thousands of miles away. He has a prediction for the remainder of this winter, offering a fresh test of his model (I’ll check back in March):

Cohen, the director of seasonal forecasting at Atmospheric and Environmental Research, has been collaborating with scientists at the Massachusetts Institute of Technology with support from the National Science Foundation, which seems quite happy with his work. Here’s a description from the foundation’s Web site:

A forecast model developed by AER scientist Judah Cohen has consistently achieved on-target forecasts for most major cities in the industrialized countries.

“Weather impacts peoples’ lives and the global economy on a daily basis,” says Jay Fein, program director in N.S.F.’s climate dynamics program. “Improving our ability to predict severe events such as the cold weather in the eastern U.S. this past winter, and the heavy snow during the prior winter, has obvious benefits. The success of Cohen’s real-time forecasts offers a way to improve our ability to anticipate important climate events.”

In 2007, the science foundation produced a graphic (by Nicolle Rager Fuller) explaining his model for how Siberian snow can distort air flow in ways that have big consequences for the eastern United States and Europe:

 Click here for a larger version. Here’s the science foundation caption:

Researchers have validated a new weather prediction model that uses autumn snowfall to predict winter cold in the United States and Europe. When snowfall is high in Siberia, the resultant cold air enhances atmospheric disturbances, which propagate into the upper level of the atmosphere, or stratosphere, warming the polar vortex. When the polar vortex warms, the jet stream is pushed south leading to colder winters across the eastern United states and Europe. Conversely, under these conditions the Arctic will have a warmer than average winter.

Here’s Cohen’s comment following the initial Dot Earth post on his work and the reactions of other climate scientists:

Many of the ideas … in my Op Ed article were taken from two of my most recently published peer reviewed papers (Both can be downloaded from http://web.mit.edu/jlcohen/www/papers.html).

In Cohen et al. 2009, we showed a statistically significant link between increasing snow cover extent in Siberia and colder Northern Hemisphere temperatures, so there is more to my article than thinking out loud. Furthermore this relationship follows the conceptual model that we published earlier on how Siberian snow cover influences winter Northern Hemisphere temperatures.

What is not mentioned in my article but which gives my ideas much more credibility, is that I make real-time forecasts based on these ideas. The best way to validate a scientific hypothesis is to make a successful prediction. The forecast model uses Siberian snow cover as one of its main predictors and I have been posting those forecasts to the National Science Foundation website: http://www.nsf.gov…

You can see this winter’s forecast and examples of previous forecasts on the site. Those forecasts have consistently outperformed other publicly available forecasts. Finally, I was invited to a prediction workshop in October where I presented a forecast for December-February 2010/11.

The forecast was for a cold winter in Northern Europe and the Eastern U.S. (a similar forecast was provided to my clients but that is not public and therefore not verifiable.) I will not declare success one month into a three month forecast, but I think that your readers will agree that since the severe weather so far this winter is consistent with my forecast, [this] makes the ideas that I expressed more compelling than if I had not made a forecast.

Finally, here are three questions I sent to Cohen Monday night, with his replies:

What got you focused on this particular puzzle piece, Siberian snow, a decade or so ago?

I was hired by Prof. Dara Entekhabi in the civil and environmental engineering department at M.I.T. to study the possible influence of Northern Hemisphere (NH) snow cover extent on the Pacific/North American (PNA) pattern using the NASA/GISS global circulation model. The response to the model to varied snow cover did not resemble the PNA pattern but instead was much closer an atmospheric pattern associated with the NAO (North Atlantic Oscillation) or AO/NAM (Arctic Oscillation or Northern Annular Mode). But when I started to publish/present my idea that snow cover could force the atmosphere, it was met with tremendous skepticism.

The community [studying] seasonal and longer climate variability was (and still is) dominated by the ocean community and especially a decade ago [ the El Niño Southern Oscillation] ENSO was the only boundary forcing that anyone was paying attention to. I was so confident (naively, I may add) in my modeling results that I decided to show that I could find the same signal in the observations. I did find a statistically significant relationship between the winter NAO/AO and NH winter snow cover extent, which was great but still had the chicken and egg problem. But then when I computed different lead and lags the strongest relationship was between fall snow cover and the winter AO. Then I divided the NH into North America and Eurasia and just about all of the NH signal could be explained by Eurasia. And ever since then I have been trying to understand the physical link between fall Eurasian snow cover and the winter AO and applying that relationship to improved seasonal forecasts.

Some scientists and environmental campaigners have been asserting that  it’s delayed freezing of sea ice that is the most important influence jogging winter Northern Hemisphere patterns. Given the complexities of NAM/NAO, ENSO, sea ice, and Siberian snow, is it possible to know which are chickens and which eggs – or irrelevant?

I think this is an interesting idea that I am looking into myself (Dave Robinson alluded to this in his comment). Simple statistical analysis shows that the relationship between sea ice variability and the winter climate is not clear and is much weaker than between snow cover and winter climate. Therefore it will be very difficult to isolate a sea ice signal in the winter climate using observations alone. The reason why I believe that snow cover has the cleanest signal with the winter AO, from the stratosphere to the earth’s surface, more so than any other boundary forcing is because its variability lies between 40-80 degrees North.

That happens to coincide with the region where the largest waves in the atmosphere (called Rossby waves) propagate, so therefore snow cover variability can have a direct impact and those waves, which dominate the large-scale atmospheric circulation. Sea ice variability occurs near the Pole, which is too far north to have a direct impact on these waves but it could have an indirect impact on another phenomenon, such as snow cover or a regional atmospheric patter (which could then impact the winter AO).

I know your main focus is seasonal forecasting, but do you have your own best sense of how the “character” of Northern Hemisphere winters is likely to change with continued greenhouse gas accumulation?

I have tried to stay focused on seasonal forecasting and not get distracted by global warming. The only comment I would say is, currently it is still cold enough in Siberia in the fall for precipitation to fall as snow. But if at some point temperatures warmed sufficiently that snow would fall as rain instead, then I think the lack of snow cover across Siberia in the fall could amplify winter warming. There are modeling studies looking at when this may happen, including the paper Dave Robinson cited, and it would be the latter half of this century. Also snow cover and the winter AO go through natural decadal cycles. If both went through a natural reversal in the upcoming years, I believe that we would experience much milder winters in the Eastern U.S. and Europe.