Arguments
What causes Arctic amplification?
Posted on 2 May 2010 by John Cook
The warming trend in the Arctic is almost twice as large as the global average in recent decades. This is known as Arctic amplification. What's the cause? Changes in cloud cover, increases in atmospheric water vapour, more atmospheric heat transport from lower latitudes and declining sea ice have all been suggested as contributing factors. A new paper The central role of diminishing sea ice in recent Arctic temperature amplification (Screen & Simmonds 2010) (here's the full paper) examines this question. The title is a bit of a give-away - the decline in sea ice is the major driver of Arctic amplification.
The vertical profile of Arctic warming (eg - how much warming occurs at different altitudes) gives us insight into the underlying cause. If atmospheric heat transported from lower latitudes was the major driver, more warming would be expected at greater heights. On the other hand, if retreating snow and sea ice cover was the major cause, maximum warming would be expected at the surface. Figure 1 shows the simulated warming expected in each season if declining sea ice was the major cause of warming.
Figure 1: Temperature trends linked to changes in sea ice. Temperature trends over the 1989–2008 period averaged around circles of latitude for winter (a), spring (b), summer (c) and autumn (d). The trends are derived from projections of the temperature field on the sea ice time series.
Using higher resolution temperature data supplemented with updated satellite measurements, Screen 2010 analyse the observe warming trend in each season. What they find is maximum Arctic warming at the surface and that warming lessens with height in all seasons except summer. This vertical structure suggests that changes at the surface, such as decreases in sea ice and snow cover, are the primary causes of recent Arctic amplification.
Figure 2: Observation of temperature trends, 1989–2008. Temperature trends averaged around circles of latitude for winter (December–February; a), spring (March–May; b), summer (June–August; c) and autumn (September–November; d). Red shading indicates that the lower atmosphere has warmed faster than the atmospheric column as whole. Blue shading indicates that the lower atmosphere has warmed slower than the atmospheric column as a whole.
The surface warming is modest in summer because energy is used to melt remaining sea ice and warm the upper ocean. The majority of the winter warming is associated with changes in sea ice cover even though the sea ice declines at this time of the year are relatively small. During summer, the atmosphere loses heat to the ocean whereas during winter, the flux of heat is reversed. Reduced summer sea ice cover allows for greater warming of the upper ocean but atmospheric warming is modest. The excess heat stored in the upper ocean is subsequently released to the atmosphere during winter.
Another potential contributor to amplified warming that's investigated is changes in cloud cover. Spring is the only season that exhibits significant trends in Arctic average cloudiness and the trend is negative. However, decreased cloud cover is expected to cause surface cooling because clouds have a warming influence in spring. Thus no evidence is found of cloud cover changes contributing to recent near-surface Arctic warming.
Changes in atmospheric water vapour content may amplify Arctic warming. However, specific humidity trends are found only during summer and early autumn. The pronounced warming in winter and spring are not accompanied by increases in humidity. In fact, the evidence suggests part of the humidity increase is driven by enhanced surface moisture fluxes associated with sea ice reductions.
The empirical evidence from the past two decades reveals that declining sea ice cover and thickness have been great enough to enhance Arctic warming during most of the year. The emergence of strong ice–temperature positive feedbacks increases the likelihood of future rapid Arctic warming and sea ice decline
Then again the climate is so very complex that further study is always a good idea.
I don't get it either. Most of it seems counter-intuitive to my simple mind!
The paper may be useful for ruling out some possible causes but does it do anything more than show an apparent relationship for one decade between sea ice and temperature, where is the cause and effect?
Lupkes, C., T. Vihma, E. Jakobson, G. Konig-Langlo, and A. Tetzlaff (2010), Meteorological observations from ship cruises during summer to the central Arctic: A comparison with reanalysis data,Geophys. Res. Lett., doi:10.1029/2010GL042724, in press.
This paper in press at GRL questions the accuracy of the ERA reanalysis compared to some ship based observations.
Hansen mentions a "greater atmospheric stability" on the poles, which would also be a contributing factor. Can someone help me understand what this is?
This paper provides relatively little evidence for mechanism a) and never formally considers the other logical alternatives. The author's main evidence is that warming is greatest at the surface. Unfortunately, surface warming is a feature of the latest data (ERA-Interim) and wasn't present in earlier reanalyses (ERA-40). Clearly, our understanding of changing Arctic temperature with altitude is limited. The authors provide data that show that trends in cloud cover and humidity do not correlate with trends in temperature, suggesting these factors aren't responsible for temperature trends. No mechanism by which sea ice controls temperature is adequately discussed.
Is there any evidence that changes in Arctic sea ice are not likely to be the direct cause of changes in temperature? Absolutely! First, the temperature increase is smallest in the summer months, when sea ice has been diminishing at 5.9% per decade, and largest in the winter months, when sea ice has been diminishing at a rate of 2.6% per decade. Second, during the winter, changes in sea ice coverage are restricted to the small area of the Arctic Ocean north of Scandinavia (at about 75 degN) that remains ice free (due to the Gulf Stream). However, the greatest warming is observed thousands of kilometers away at 85-90 degN. If the author's hypothesis is correct, additional energy must be obtained from unfrozen ocean in this region and somehow must travel thousands of kilometers to the north without leaving the surface. Unfortunately, the author's data has been averaged over all longitudes, so there is no evidence to support or contradict this mechanism.
How do the authors explain the fact that changes in sea ice have minimal effect on surface temperatures in the summer? This is where the authors show the greatest confusion about cause-and-effect. They cite other sources saying that the trend towards warmer summer temperature is negated by energy loss from the air to the ice and/or ocean. This is mechanism b), not mechanism a). Before one can have increased energy transfer from the air to ice or ocean, one either needs warmer air to begin with, more ice, or a colder ocean. None of these phenomena exist: The author's data shows that: summer air temperature is only marginally warmer (much less than the +2 degC/decade observed in other seasons) and that sea ice is down. With less sea ice, a dramatic increase in albedo and increased humidity in the air, the ocean is unlikely to be colder.
What does all this say about "Polar Amplification"? In the introduction, the authors tell us that: "It is widely accepted that changes in the surface albedo associated with melting snow and ice enhance warming in the Arctic, but other processes may contribute." Changes in albedo are greatest and most important in the summer (when the sun is shining), but warming is greatest in the non-summer months. From the information present in this paper, it should be absolutely clear that popular hypotheses do not provide satisfying explanations for the current (limited) data.
A more satisfying explanation for "Arctic Amplification" might be constructed by remembering that the polar regions radiate far more energy into space than they receive from the sun (and none in winter). Therefore their temperature depends more on energy imported from other areas of the planet than anywhere else. Arctic temperature is least dependent on imported energy from other areas of the planet during the summer, when insolation is greatest. Temperature increase is least in the summer when the most dramatic changes in sea ice have been observed, so hypotheses a) and b) are relatively unattractive despite the strong statistical correlation between the two phenomena. Increased import of energy (Mechanism c.) provides the most satisfying explanation for the parallel reductions in sea ice and increases in temperature and water vapor. Unfortunately for alarmists, it suggests that a self-perpetuating cycle of positive feedbacks within the Arctic is not going to lead to climatic disaster.
"Summer, by contrast, has seen a recent increase
in the net surface heat flux (an increased net heat gain
by the ice-ocean column). This is understood in that ice melt(phase change) and heating of the ocean mixed layer have limited the increase in the surface temperature and hence the upward longwave flux." Remember it is the end of summer and autumn which have the most open water, also recall it takes lots of energy to melt the ice. Also note the water is not in general warmer than the air in the summer and would not heat it appreciably if it is not. The water is increasingly warmer than the air in the autumn, as autumn progresses. That is a key to the ocean being able to heat the air above it. In particular the air near the surface. Another point they make is:
"2) Consistent with an anomalous surface
heating source, development of the autumn warming pattern aligns with the observed reduction in September sea ice extent, and temperature anomalies strengthen from the lower troposphere to the surface; 3) Recent autumn warming is stronger in the Arctic than in lower latitudes; 4) Recent low level warming over the Arctic Ocean is less pronounced in winter when most open water areas have refrozen;"
The pattern does fit. Recall that this group used NCEP and JAR-25 data not just ERA-40. This suggest a very robust result and does support the joint effects of hypothesis a and b.
I guess the theory goes less sea ice means greater transfer of heat between the ocean and atmosphere. I don't understand how this is greatest in winter and at the highest latitudes. This is both the time and region where sea ice has seen least change. An insulating cover of ice must be present at this time. How does the energy get through the ice?
As Frank also mentioned there are questions about the temperature records. No buoy or ship data in winter at the highest latitudes and poor satellite coverage, where are the numbers coming from?
(It's a pity the supplementary data isn't working)
I'm also a little sceptical that there is a linear trend in winter sea ice over the period (1989-2008). The graph below (from Cryosphere Today) shows that winter sea ice recovered to it's early 20th century norm between 1997-2002. How is the winter sea ice a linear trend?
Thickness is important because the insulating effect of the ice depends on the thickness of the ice, just like the thickness of the insulation in your house determines its insulation rating (given as Rn in the US, where n is a number).
Heat can be transported through the ice in a couple of ways. One is by transport or percolation of water through the ice, but I don't know a lot about how important that is. Even with no water percolating through the ice, heat will be transported through the ice by conduction. The rate of heat flow for conductive heat transport in a uniform material is q = -k*dT/L, where k is a constant that depends on the material, dT is the temperature difference, and L is the distance between the hot and cold sides, in this case the thickness of the ice. The minus sign just means heat flows from hot to cold.
If you cut the ice thickness in half, you double the rate of heat flow given the same temperature difference. As the ice gets thinner, the heat flow will be dominated less by conduction, and will approach the heat flow out of open water (heat flow won't go infinite). The thickness of the ice is likely to be a controlling scale for other modes of heat transport as well, but somebody else needs to pick up beyond this because this exhausts my knowledge of heat transport through sea ice.
But the bottom line is that in autumn/early winter, there is a lot more open water than before, which then gets covered by thin ice, and the heat transport from ocean to the atmosphere is going to be a lot faster than it used to be throughout the entire winter. Compared to the Arctic atmosphere, the Arctic Ocean is a vast reservoir of heat. Yes, its very cold by ocean standards, but the value of T doesn't matter in the heat flow problem (except for radiation), only the temperature difference (dT). And in the winter dT is large and negative (atmo minus ocean).
there is a very good reason for the warm water effect having a larger effect in winter, it's air temperature. Relative to the winter-summer temperature difference, water temperature is almost constant and not far from air temperature in summer.
I'd also like to point out that the authors are smart enough to not confuse correlation with causation. Indeed their conclusions don't rely on the correlation but on the physics of the process. The correlation has been properly used to rule out other effects.
It always surprises me when people think that reputable scientists (authors and referees) could make such trivial mistakes. Betting on a trivial mistake in a published paper, good chances are that we're missing something.
a little sceptical that there is a linear trend
There are issues here indeed. Arctic sea ice extent reconstruction for the first half of 20th century at Cryosphere Today is higly problematic.
The following book is a must-read for all serious students of the topic (even if it is mentioned at WUWT).
Arctic Ice by Nikolai Nikolaevich Zubov, Moscow, 1943 (English translation 1963).
From the summary: “. . . the combined heating effect from the loss of sea ice and snow and the increased release of methane from permafrost in 2010 is roughly equal to releasing an additional 3 billion metric tons [my emphasis] of carbon dioxide into the atmosphere. For comparison, this equals 42 percent of current annual U.S.emissions of greenhouse gases.”
We are talking about an amplification of warming. Frank seems to be proposing some mechanism that moves more heat into the arctic region when the region is warmer, than it did when the region was relatively cooler. That is counter-intuitive for me.
The primary factor we are looking at is ice loss and regain. Frank's post is all about temperature, but in a phase change, there are large changes in energy with little or no change in measured temperature. When ice melts in the summer, there can be large changes in energy with very little change in temperature. This energy is released as the water freezes, and we see a positive temperature anomaly, in the fall/winter, compared to historical points when the water was already frozen in that place and time of year.
The average temperatures and trends of each hemisphere are very different with an obvious warming bias in the northern hemisphere and closer to no change in the southern hemisphere.
Warming due to CO2 is supposed to be global but if each hemisphere, and each polar region are examined individually, then obviously there are other factors to be considered, like long term cycles that will reverse over time.
Certain factors may be in play in the Arctic to produce amplification, but what is happening at the Antarctic that either enhances or offsets them that will be reflected in the global situation?
Figure 1 in Serreze shows that future Arctic Amplification is projected to be maximal near the summer solstice and minimal near the winter solstice. If these changes are driven by changes in sea ice - as Screen asserts - they should be maximal in the fall and minimal in the winter and spring - when changes in sea ice are occurring far away. If changing ice-albedo is the most critical factor in temperature change, that should be maximal when insolation is maximal.
Riccardo suggests that there is a bigger effect in the winter because the difference between ocean and air is greatest in winter. As the ice at 85 degN thickens, energy is transferred through the ice cover to the air. 334 J/kg*917 kg/m^3 = 3*10^8 J/m^2 per meter of ice frozen. If thinner sea ice means that an addition meter of sea ice more is melting and freezing every season (180 days), the heat flux amounts to +20 W/m^2 (+5 degK at 250 degK)the winter and -20 W/m^2 (-4 degK) in the summer. If Screen had data showing that an average of 1 additional meter of sea ice was melting and refreezing every year at 85 degN compared with two decades earlier, he would have a mechanism for a 2 degC/decade rise in temperature that depends on sea ice. Without mechanistic details, he is simply engaging in undisciplined speculation about possible reasons for correlation. (Notice that a discussion of mechanism directs attention away from changes in sea ice coverage to changes in the thickness of sea ice melted and refrozen each season. "Arctic Amplification" driven sea ice by Riccardo's mechanism turns out the be simply a transfer of heat from summer to winter, amplifying global warming in the winter and negating it in the summer.)
Figure 1 in Serreze shows that future Arctic Amplification is projected to be maximal near the summer solstice and minimal near the winter solstice. If these changes are driven by changes in sea ice - as Screen asserts - they should be maximal in the fall and minimal in the winter and spring - when changes in sea ice are occurring far away. If changing ice-albedo is the most critical factor in temperature change, that should be maximal when insolation is maximal.
"transfer of heat from summer to winter"
this happens all the time and will continue to happen untill there will be some ice around there. It's just heat capacity and latent heat at play, they both tend to stabilze temperature.
But these heat fluxes can be pushed one way or another by a change in ice volume and extention. I can't see any big surprise here. What is new in the paper is its relative amount with respect to the atmospheric flux.
But as always happen with cutting edge research, it might well not be the last word.
However understanding the physical mechanisms involved is another thing, and as in the article, scientists are just beginning to study some of those physical mechanisms.
As an example, despite most CFC's being released in the NH, the ozone hole is generally over the South Pole. This thought to be because the ozone layer over the Arctic stays warmer by about 10 degrees than the that over the Antarctic. That is what happens and is measured and accounted for, but is there any understanding of why it is so? Is it the surface climate changing conditions in the stratosphere, or vice versa?
What is not happening at one pole may be of as much importance as what is happening at the other, such being more likely to be revealed more by developing understanding of said physical mechanisms.
It seems to me that in many cases the cart is before the horse, with assumptions being made simply because data correlates more so than an understanding of why it correlates, or if even relevant.
But whether the writers really benefited from - such as - this work: Chylek, P., Folland, C. K. & Lesins, G. Dubey, M. K. & Wang, M. Arctic air temperature change amplification and the Atlantic multidecadal oscillation. Geophys. Res. Lett. 36, (2009); ?
"Analyzing temperature records of the Arctic meteorological stations we find that (a) the Arctic amplification (ratio of the Arctic to global temperature trends) is not a constant but varies in time on a multi-decadal time scale, (b) the Arctic warming from 1910-1940 proceeded at a significantly faster rate than the current 1970-2008 warming, and (c) the Arctic temperature changes are highly correlated with the Atlantic Multi-decadal Oscillation (AMO) suggesting the Atlantic Ocean THERMOHALINE CIRCULATION is linked to the Arctic temperature variability on a multi decadal time scale."
Recall it: Chylek et al. 2006 - Greenland warming of 1920–1930 and 1995–2005; write that instrumental measurements indicate a large and rapid heating of the coast of Greenland in the decade of 1920. The average annual temperature has risen when the 2 to 6 degrees C in less than 10 years. "Temperature increases in the two warming periods are of a similar magnitude, however, the rate of warming in 1920–1930 was about 50% higher than that in 1995–2005." This rapid increase in temperature, while CO2 emissions to the atmosphere was 9-fold lower than in 2003 (Marland et al., 2006) speak of a natural cause of such a powerful warming.
The same is here, only :http://ocean.am.gdynia.pl/p_k_p/pkp_19/Marsz-Stysz-pkp19.pdf
"... the STRONG CORRELATION between the sea surface temperature (SST) in the region of the Gulf Stream delta and anomalies in surface air temperature (SAT) in the Arctic over the period 1880-2007. SEA ICE MAY EITHER INCREASE OR LIMIT THE HEAT FLOW FROM THE OCEAN TO THE ATMOSPHERE."
"THE GENESIS OF THE 'GREAT WARMING OF THE ARCTIC' IN THE 1930S AND '40S IS THE SAME AS THAT OF THE PRESENT DAY."
Really worth reading the two works. Are long and ... full of very interesting calculations - in contrast to the Screen & Simmonds 2010, but ...
Marsz, also 2009 but by: Present warming - oceanic climate control; said:
"Changes in SST in the Sargasso Sea, explains about 70% of the variability of SAT anomalies in the NH in the period 1880-2008 and 68% of the variability of global SAT anomalies [...]. In times of growth of SST in the Atlantic and North Atlantic sea sector of the Arctic, associated with INTENSIFICATION of the INTENSITY of THERMOHALINE CIRCULATION [!], there is an increase in air temperature in the NH. This increase is particularly strong in the higher latitudes - the Arctic and temperate zone."
And I'll be back again also to FIG 11 - maps from this work: K.E. Trenberth, J. Fasullo, L. Smith, 2005: Trends and variability in column-integrated atmospheric water vapor. Climate Dynamics 24: 741–758; The largest increase in humidity over the past decades, we see it is in a place where the return of energy by the Gulf Stream ... Here the difference - in relation to the whole Arctic - is significant. The largest increase in humidity over the past decades in a place where we see the return of energy by the Gulf Stream ... Here the difference - in relation to the whole Arctic - is significant.
The balance of energy resulting from the local greenhouse effect caused by water vapor - "positives" are often strongly underestimated.
Conclusion. As Frank said - it's energy imports determines the current scale and pace of rising temperatures in the Arctic.
what the paper under discussion does is to put current beliefs into question by using a different reanalysis and a different point of view. Just reaffirming older methodologies without critical comparison does not bring us much further ahead.
OK. Agreed.
However, the work of Simmonds & Screen 2010 contains important conclusions. For example, such as the proposal:
"The findings reinforce suggestions that strong positive ice–temperature feedbacks have emerged in the Arctic, increasing the chances of FURTHER RAPID warming [?!] and sea ice loss, ..."
If this is: THC-AMO (mainly), the the conclusion is hasty.
Each methodology as the "older" and "younger", must be based on earlier findings - the results. Also, methods of measurement.
E.g.: Alekseev et al., 2008, Arctic Sea Ice Data Sets in the Context of Climate Change During the 20th Century.
- "September ice extent in the majority of the Siberian Arctic seas and in the Barents Sea reveal rapid shrinking during Arctic warmings in the 1920–1940s and 1990s. Significant correlation between surface air temperature and ice extent occurs in summer months with maximum in June under the influence of June maximum solar irradiation, and amplified by heat advection in the atmosphere and ice extent anomalies in the previous months. The relationship between variations of winter air temperature and ice extent is weaker because winter ice extent anomalies depend on air temperature anomalies as well as on the area occupied by a freshened upper layer. Good agreement between variations of the sum of summer air temperature in the marine Arctic and sea ice extent in September is found (correlation coefficient is 0.85)."
"Conclusion. As Frank said - it's energy imports determines the current scale and pace of rising temperatures in the Arctic. "
We are talking about temperature anomalies in the arctic that are greater than temperature anomalies globally. That is, there is an increase in the arctic that is above and beyond what is happening globally. What mechanism would amplify the amount of energy going into the arctic from the rest of the globe when the arctic is relatively warmer than it has been?
Your original concern was about differences in response of the two hemispheres to a deranged energy budget. The reasons for why this is true are well understood; glancing at a map of the globe shows us why we should not expect the two halves of the planet to behave identically. The takeaway is, we've got a good grasp of why the Southern Hemisphere diverges from the Northern Hemisphere w/regard to a deranged energy budget. The effects you mention are interesting but in terms of budget they're akin to worrying about plugging a nickel into a parking meter when the payment on your automobile is $500USD/month.
not sure I understand your point. The paper main focus is on the vertical structure of the temperature anomaly following the new ERA-interim reanalysis. The latter is in constrat with previous reanalysis datasets:
"We diverge considerably from ref.8 [Graversen et al., Nature 451, 53–56(2008).] in finding that the maximum Arctic warming is at the surface and that warming lessens with height in all seasons except summer.
This vertical structure suggests that changes at the surface, such as decreases in sea ice and snow cover, are
the primary causes of recent Arctic amplification."
So, they are not discussing the origin of the warming but the feedback that can be inferred from this new reanalysis.
The European Geosciences Union annual meeting currently underway in Vienna have put some of their summary "Press release" presentations on line as webcasts. There is an interesting presentation by Peter Wadhams entitled "Arctic sea ice is in terminal retreat" available from here. The question/answers session at the end is interesting.....
I lived in Anchorage, AK in 2008 and not only was the winter one of the coldest ever, but October 2008 was the 3rd coldest October in the same history frame. Fort Yukon had temperatures plummet to -79 degrees F for weeks at a time. The Yukon river was frozen from its outlet in Alaska all the way down through the Yukon territory of Canada. I drove along it for miles and have tons of pictures of it being completely frozen in June. It was frozen the whole summer the year before as well.
Also one of the things I see no one talking about is the fact that the fabled Northwest passage was completely open and navigable in spring, summer, and fall of 2007, but by winter of 2008 when a liberal ecologist (not an ad hominem that was his own listing) tried to take a kayak followed by two vessels of camera teams, journalists, and assistants tried to navigate the famed Northwest passage to show the world how the evils of ACGW was destroying our earth, the ice had already gathered in such a large vast amount he was turned back. In fact studies show that there was more ice growth in the Arctic in 2008 than in the last 20 years previous. The four famous polar bears talked about by Mr. Gore in “An Inconvenient Truth”, as dying because of Arctic warming. Were actually autopsied by experts and it was found they didn’t starve to death due to too much warming; they died from exposure to extreme cold during the worst wind storm known.(The Eskimo’s and other native tribes are excellent at recalling history of weather for many generations). The same windstorm pattern that blew from Alaska in the west had similar patterns all the way across to Hudson’s Bay in the east. This windstorm in the west also caused a major phenomenon which few people outside the skeptic realm ever talk about. Because there was such incredible wind for such an extensive period of time there was no snow to insulate the ice and make it take longer to build up like usual and the moisture that fell caused exponential amounts of ice to form.
Now for any of you who are close to being a skeptic but are unsure, look into the great Northwest passage looked for, for so long by so many… There are verified reports of Canadian vessels traveling east to west across the Inland Passage at least 4 times in the past 200 years. Those that never found it were not there at the right time. Just like the gentleman in the Kayak in 2008. Could it be that the Earth regulates itself, and has for eons and no matter what man does, it won’t allow him to step outside of its control of itself?
if you're looking for a discussion on a particular year or month you should try to find a weather related blog. Climate is different.
P.S. The place where I live has been called "the Heaven on Earth". I don't think we should take this ancient testimony litterally.
Are you kidding? 2008 wasn't even close to being that cold. In fact, it was above the average of the past 30 years*, and that 30 year period itself is way warmer than any other period in your 1775-present timeframe!
In fact, although 2008 was cooler than some other recent years, in the GISSTEMP record it was the 10th warmest year since the start of the record in 1880! Every year since 2001 is in the top 10.
Where on earth did you get the idea that "2008 was the third-coldest year since 1775"?
You need to check your sources, and preferably provide references for your claims.
* 2008 was warmer than the 1979-2009 average in the RSS, GISS, CRU, and NCDC records. It was slightly cooler than the average in the UAH record.
Does the particular study that we are discussing in this thread have anything to say about such a shift in the seasonal pattern?
Sigh. I'm trying to find a good explanation for NON-albedo driven Arctic amplification, but this post isn't it. In fact, I feel like this post summarily dismisses these other mechanisms based on a single paper, which is not appropriate: certainly, both the IPCC AR4 report or the draft AR5 report indicate that these other mechanisms may be responsible for as much as or more than half of the amplification. Work that might be relevant includes:
http://www.esrl.noaa.gov/psd/people/amy.solomon/pubs/Solomon_GRL_2006.pdf : Showing a link between basically thunderstorms and polar amplification
http://onlinelibrary.wiley.com/doi/10.1029/2005GL025244/pdf : surface albedo contributes but does not dominate amplification
http://yly-mac.gps.caltech.edu/AGU/AGU_2008/Zz_Others/Li_agu08/LangenAlexeev2007.pdf : Arctic amplification is observed in an idealized aquaplanet GCM
http://link.springer.com/content/pdf/10.1007%2Fs00382-009-0535-6.pdf : Perhaps the clearest explanation, using a locked-albedo model to explore how water vapor and cloud feedbacks as well as changes in heat transport and vertical stratification enhance Arctic warming.
-MMM
Another very recent paper:
http://www.cgd.ucar.edu/cas/cdeser/Docs/submitted.screen.arctic_forcing.mar12.pdf
This is perhaps closer to the original post - it finds that half of the Arctic surface warming is due to sea ice/sea temperature changes, with a sixth coming from remote transport due to sea surface temperature changes elsewhere... and it isn't clear where the remaining third comes from. The paper claims that changes in direct radiative forcing mainly matter for summer tropospheric changes.