It was mentioned in recent discussion that Cohen et al. (2012) found recent winter cooling in much of the boreal (northern but not necessarily Arctic) northern hemisphere. In fact, here’s their key graph for that particular question:
Colors show the wintertime temperature trend from 1988 through 2010, with red indicating warming while blue indicates cooling, based on the HadCRUT3 gridded temperature data set. They note cooling in northern Europe, Russia, and eastern North America.
I decided to take a look at another data set for part of the area they studied, the data from NCDC (National Climate Data Center) for the 344 climate divisions covering the contiguous U.S. If we estimate the trend (by linear regression) for the winter season (Dec-Jan-Feb) only, for all U.S. climate divisions since 1988 (which is a little bit longer than the time studied in Cohen et al., since these data go through Sep.2012 while they ended with Dec.2010), it looks like this (red is warming, blue is cooling, larger dots mean faster rates):
Sure enough, the broad swath of blue in the eastern U.S. shows cooling just as reported by Cohen et al.
A natural question is whether the cooling is statistically significant. Sure, the trend estimate is negative but is that a genuine trend or could it be random fluctuation? Let’s plot the same graph again, but use open circles for the estimates that are not statistically significant and closed circles for those that are:
Clearly the vast majority of the trends are not significant, and those that are, are few enough in number that it could be the result of randomness. By the way, I did not correct for autocorrelation but it’s much smaller for regional than for global temperature, and it’s my blog so I’m allowed to get a little lazy from time to time.
That doesn’t mean that this region didn’t cool off. But it does mean that we don’t have the evidence to call it a trend. It could be — the variance of regional temperature is quite a bit larger than that for global, so it’s harder to confirm a trend significantly. But the evidence of trend isn’t really there.
In any case, the pattern for winter (DJF) differs (whether significantly or not) from that for the entire year:
The bulk of the country shows warming since 1988, with a pocket of cooling in the far Pacific northwest. If we do the “non-significant open, significant closed” version,we see that for many of the regions the trend is significant, so yes the country as a whole has warmed since 1988 with a pocket of cooling in the far Pacific northwest:
An interesting behavior emerges if we analyze each month separately (this is an animated GIF, you may have to click the graph to see the animated version):
The eastern cooling pattern doesn’t show in December, but it does during January and February. So, it’s really more a Jan-Feb cooling than the entire winter. But again, those monthly trends aren’t really significant — here are the not-significant-open, significant-closed versions for Jan and Feb:
Despite the fact that these patterns may not be genuine trends, they are at least genuine fluctuations. Cohen et al. attempt to understand them by exploring their relationship to other recent (since 1988) climate observations.
Their proposed theory is that the root cause is warmer summers and autumns. This increases Eurasian snow cover during Autumn (especially October) due to greater atmospheric water vapor content. This in turn promotes the cold phase of the Arctic Oscillation (AO) in winter, leading to colder winter temperatures over much of the northern hemisphere.
Thus, the regional NH wintertime temperature cooling is directly tied to the declining trend in the wintertime AO, which is forced partly by increases in Eurasian snow cover during the autumn, driven in response to warmer surface temperatures at high latitudes prior to and concurrent with the autumnal advance in snow cover.
They point out that a reason climate models may have predicted otherwise, is that they don’t model snow cover variability very well:
A warmer, more moisture-laden Arctic atmosphere in the autumn contributes to an increase in Eurasian snow cover during that season. This change in snow cover dynamically forces negative AO conditions the following winter. We deduce that one main reason for models failing to capture the observed wintertime cooling is probably their poor representation of snow cover variability and the associated dynamical relationships with atmospheric circulation trends (Hardiman et al 2008, Jeong et al 2011). Incorporation of the snow cover–AO relationship into seasonal forecasts is shown to greatly improve their abilities, and hence long-term climate solutions from coupled climate models may also bene?t from improved snow–AO relationships.
It’s speculative of course, as Cohen et al. themselves point out. As we’ve noted, the reported trends don’t pass statistical significance so it remains to be seen whether the observed phenomenon itself is even meaningful, let alone their purported explanation. But they do find impressive correlations between autumn snow and winter AO, and between winter AO and northern hemisphere temperatures, in just the right regions.
Those interested in more details should read their paper. As for whether or not they’re fundamentally correct … time will tell.
There’s spatial correlations to worry about as well. This can go two ways: make it easier to detect trends, because the trends are spatially correlated; or make it harder since the errors (noise/weather) are spatially correlated.
Interesting that there are only four significant dots for DJF, since you’d expect 9–25 by chance (if they were independent).
Well, there isn’t a near Pacific Northwest and a far Pacific Northwest. Roughly, the Pacific Northwest extends east to the continental divide and south into far northern California.
Anyway, having lived in the inland Pacific Northwest for 42 years now, I rather doubt any temperature trend whatsoever. Subjective, I know.
The exeption might be that increased aerosols from east Asia are moderating summer daytime highs. Overnight lows also seem moderated but the averages feel trendless.
The key issue here is that the seasonal winter cooling in the continental NH mid-latitudes is apparently the difference between the difference between 0.15 deg C per decade warming over the last several decades in the global temperature anomalies, and the 0.20 deg C per decade warming. Judah Cohen says the other three NH seasons hit much higher decadal warming rates, consistent with the average of climate models, but the NH winter is not warming like the models predicted.
So the apparent global warming rate below the warming trend projected by models, dwelled on by climate change skeptics, and especially lukewarmers, appears due to this anomalous winter cooling of the NH continents in the mid-latitudes.
Eventually of course, even the winter mid-latitudes will begin heating. Since the climate models are actually based on energy balances, eventually the increasing thermal energy retained by the planet, will begin manifesting itself everywhere.
“A warmer, more moisture-laden Arctic atmosphere in the autumn contributes to an increase in Eurasian snow cover during that season. This change in snow cover dynamically forces negative AO conditions the following winter. ”
That conclusion is a bit of a stretch…. Admitting one fall ultimately changes the next season by morphing the winter AO to be consistently at a negative phase is too much. It completely disregards sea ice conditions which onto itself affects the AO. For instance thinner over all ice allows more cyclones to penetrate the High Arctic. More snow cover in Siberia
makes it colder for Asia at a specific time period, this could only affect the adjoining regions which are rather large and well spread out.
I would like to point out that the boxes in Cohen et al 2012 contain more than one station. Hence, you expect that more of them would be statistically significant if the trend exist.
[Response: I believe they used, not station data, but the grid averages from the HadCRUT3 gridded data set.]
Yes, this is I am saying. Your analysis is station based. Their is grid based. Signal to noise ratio is likely to be better is more station are include in each bin.
[Response: Mine is not station based. It’s based on climate-division averages rather than grid-box averages.]
I must admit myself convinced the Judah Cohen has a point and is probably correct. This is far from the only paper he has published. I’ve blogged on his work previously. IMO he makes a convincing case for Siberian linkage to the AO in causing the winter of 2009/10, here. He also has produced a Snow Advance Index, based on snowline advance over northern Eurasia, which seems to have some predictive power for the following autumn, discussed here. You’ll find links to the source papers on those pages.
This is clearly not the only factor, wider factors such as ENSO/PNA have a role. However your comment suggests to me that you’ve not read much of Cohen’s work.
If you look at the map plot from Cohen you can just about see a quadropole pattern of cold Northern Europe, warm to the south of that, cold eastern US, warm to the north of that.
This is the classic quadropole pattern of the negative North Atlantic Oscillation, and it seems to me that it mostly reflects the record -NAO of December 2010, and other recent winters.
It would be interesting to see what the pattern looks like for different years. Leading up to the late 1990s, there was a trend of an increase in the NAO index, so one would expect to see the opposite of this quadropole pattern.
The key question is whether this circulation change is anything other than normal natural variability on top of the global warming trend, or whether it is a circulation feedback as a result of global warming. It has been speculated that the decline in Arctic sea-ice is the cause of this circulation change, but we’re some way from settling that convincingly in my opinion.
One recalls that there was also a lot of speculation that the change to positive NAO a few decades ago was also caused by global warming, but that now does not seem to be the case (I think there’s a recent paper on the stratosphere model of the Met Office that looks at this).
Tamino, could you make the open-circle/closed-circle version of the full-year animated gif? I think it might be interesting.
OT, but American Facebook users whose appetite for symbolic actions is not totally jaded may wish to participate in this action:
I’m a touch out of date but some years ago, due to stratospheric cooling a positive AO was expected – i.e. an intensification of the polar vortex due to stratospheric cooling from the enhanced greenhouse effect leading to +ve AO. Throwing in the rapid decline of sea ice and possible linkages through to increased autumn snowfall advance is a complicating factor that wouldn’t have been picked up in the models.
As things stand there are two mechanisms that have been proposed for cold outbreaks from the Arctic, and both seem to have recent events to back them. The first is Cohen’s proposed impact of rapid advance of snowline in Eurasia causing a low index AO during the following months. That seems to have the winter of 2009/10 in support of it. The second is the Petoukhov/Semenov pattern, as discussed by RealClimate here. That pattern occurred in early February causing a cold outbreak in Europe that was asociated with low sea ice extent in Barents/Kara, as I discussed here.
Recently Overland et al published “The Recent Shift in Early Summer Arctic Atmospheric Circulation”, discussed here. This relates to a shift in northern hemisphere atmospheric circulation that has happened since 2007. It’s the same pattern I had found while looking at my perception of the UK having cool wet, low pressure dominated, summers since 2007, see here. Overland et al look to changes in May snow cover as an explanation, I had been thinking of this new pattern as due to loss of Arctic sea ice since 2007. However I’m becoming more persuaded by snow cover, the problem is Eurasian snow cover looks to me like the best bet, and I haven’t a clue how that would teleconnect to Greenland! Whatever the details, this shift has set up a dipole anomaly in summers since 2007, which would have to play a role in the increased melt and low area/extents of Arctic sea ice since 2007.
But as this year has shown, atmospheric impacts are not the only issue now because the ice is so critically thin it almost doesn’t matter what the atmosphere does.
Compelling animated GIF; is there one for the planet too?
For that US animation — October, unusually cool from the Rockies to the Pacific — is that the fall’s early cold fronts showing up?
Chris, Cohen’s theory is pretty standard, That is why its potent . Totally agree with snow cover influence, I just don’t think Siberia alone can change the nature of winter. I view the AO, this strange creature, holistically with all the other regions and oscillations. By AO i mean the Arctic oscillation which I don’t particularly like, especially as a tool to explain weather events in greater detail.
What is at issue is winter, how strong can it get, how fierce it will become. We have just recently learned that thin sea ice have shaken winter to its weakest expanse as exemplified by the “March summer”
of 2012. Siberia was rich with snow despite this famous North American heat wave :
It is rather important when the Arctic does not become a conduit for a Siberian blast.
I’m not sure what Siberian snow in March has to do with the March heatwave, the AO linkage proposed by Cohen relates to Autumn advance of the snowline in Eurasia, specifically in October and November, although anomalous advance later on with warming may have an effect as well. But it is the rate of advance that Cohen is concerned with.
I’m not aware of any good reason to see the March heatwaves in the US as being directly (first cause) linked to sea ice loss. NCEP/NCAR shows that aside from the US surface air temperature anomalies were strongest over the low sea ice anomaly of Barents/Kara. Image. Geopotential height anomalies show that the pattern was due to a ridge anomaly over the US countered by a trough to the north. Image. In this respect the situation seemed more similar to the claims by Stu Ostro that there has been an increase in unusually deep lows and unusually high troughs that corresponds with general Arctic warming, not specifically with the loss of Arctic sea ice. Indeed Ostro implies that the March ‘Sci-Fi’ US heatwave was the sort of event he’s been citing, see here, as fitting in with what he’s been seeing – see his lecture pdf here , warning – very large PDF.
So if the US Sci-Fi heatwave of March 2012 can be linked to the Arctic, and I suspect it can, it is more properly associated with Arctic amplification, rather than directly associated with sea ice loss. I agree that ultimately we can’t seperate the two issues – they are inter-related. But for clarity I think it important that we seperate the loss of sea ice with its enhanced latent and sensible heat fluxes (most prominent in Autumn) from the wider Arctic warming, so that we can be more clear about the proimate causes and the mechanisms involved.
Chris, I could have shown snow in November or December:
The Siberian snow cover appeared expansive. I have another version
for 2010-2011. I documented it on my website. 2010-11 culminated into a very warm Arctic troposphere at start of winter, in large part from a persistent high over Greenland in December. The very high troposphere caused a very cold stratosphere helping create a huge and stable Polar Stratospheric Vortex, this in turn reduced stratospheric ozone , an important ozone hole was present in the spring.
2011-12 was different, from an even thinner Arctic Ocean sea ice base.
At winter start Low pressures criss crossed the Arctic Basin almost like there was only open water, especially where the ice was thinnest. At the beginning the troposphere was equally high from heat injected by these cyclones, again the the stratosphere started quite cold, but ozone in the Polar region was stronger and the vortex collapsed way earlier than preceding year. Especially at Arctic sunrise. The thinner ice literally cut off the Siberian to Alaska corridor, cold Arctic “blasts” from Canada were rare. Siberian cold spread elsewhere.
This year starts much like 2011-2012. I don’t use the AO because:
It confuses the hell out of me, its too large in scope since:
The Arctic had a low smack in the middle of the North Pole of late. And a persistent cyclone over the Arctic Basin mainly driving all circulation all around it.
Siberian snow is one factor to consider, but look at it this way, admit Cohen is right, as a result persistent High lies of Siberia, this drives the Asian Jet stream way low, however over the Pacific the stream heads NE, like the Gulf stream in the Atlantic, driving Pacific Lows to warm the Arctic further along with the Atlantic ones.
Breaking the super cold Arctic Anticyclone corridor even further, just like the March image you displayed…. Cohen was right but not Holistic enough.
Cohen has restricted himself to the causal pathways and resultant impacts that are strongly evidenced. Hence activity in the Pacific sector is not addressed. Yes it is quite possible that the blocking siberian high could change the phase or wavenumber of the Rossby waves thus causing an impact in the Pacific sector. However, as far as I can recall such wider impacts are not mentioned in his research. This would be a wise move on Cohen’s part even if he has noted such wider impacts on the Polar Jet, what is apparent from his work is that he has persued this matter for over a decade. It is only recently that he has developed a sufficiently strong case that people like the UK Met Office are paying attention – from what I can work out they’ve adapted their model to include the process he outlines. Being holisitic can often come across as hand-waving and general, not something that tends to win scientists over.
The point about the Siberian snow linkage is that it seems to offer opportunity to predict the worst type of winters, such as 2009/10. Yes other factors play a role. But being able to predict the worst class of winters is a distinct advantage. Furthermore with regards the Petoukhov/Semenov pattern, although low sea ice in Barents/Kara doesn’t predict when this pattern will cause a cold outbreak in Europe, if sea ice is low in those regions and a resultant high develops then there is good reason to be prepared. It’s all a matter of cutting the risk problem into manageable chunks.
With regards Tamino’s lead post. Here are the trends from GISS Maps. I’m not going to ask Tamino to do the analysis for Eurasia – I’ve analysed PIOMAS gridded data so should be able to figure out how to do the analysis myself. But it does strike me that the trends over the US are marginal, whereas those over Eurasia are much stronger, so I suspect would be more significant.