Why is the Arctic warming so quickly?
The Arctic is warming at a rate twice the global average . It is the most sensitive place to climate change, its loss of sea ice is perhaps the first climate tipping point we can no longer return from, and its utter transformation should convince policy-makers of the power and delicacy of the climate system. Nonetheless, the science explaining this trend of Arctic amplification is comparatively youthful. Whilst the idea that Arctic amplification is a natural part of the climate system has become accepted, paleoclimatology revealing the Arctic to be an amplifying chamber for warming and cooling stretching back to the Pliocene , attributing relative significance to local versus large-scale climate mechanisms is much harder. So, why is the Arctic warming so fast?
Local feedbacks from sea ice
At a local scale, the feedbacks caused by a loss of sea ice drive Arctic amplification. Sea ice acts as a barrier that insulates a comparatively warm ocean from the much colder atmosphere. As the ice melts during the short Arctic summer, the heat stored beneath it radiates into the atmosphere, trapped by a cloud cover that forms from the increased rate of evaporation from now-exposed open water. Due to the difference in albedo between sea ice and water, the surface now absorbs more thermal energy, in turn radiating it back to the atmosphere. These constant fluxes naturally amplify anthropogenic forcing, which keeps the sea warm for a longer period of time, delays the formation of winter sea ice, and makes it thinner in spring, prone to melting earlier .
Lindsay and Schweiger (2015)  report a shocking 65% reduction in ice thickness (-2.34m) since 1975 in the central Arctic Basin where sufficient records exist, and they suggest an average decline in sea ice thickness of -0.58m ± 0.07m decade-1 (2000-2012). Thinner sea ice allows more longwave radiation to reach the atmosphere from the sea, enhancing the localized greenhouse effect as clouds form earlier . Only at the poles do clouds act as a positive rather than negative feedback mechanism, their insulating properties more significant than their reflective properties due to the low levels of insolation high latitudes receive. Cloud formation is also favoured by the lowering of air pressure close to the surface, up to ~850hPa in the areas of strongest warming . Sea ice thinning is therefore thought to have amplified Arctic warming by of up to 37%.
As the Arctic becomes warmer for a longer season, the treeline is creeping northward , accompanied by an increasingly productive, though less diverse, tundra biome  . A process termed shrubification means that the tundra is slowly converting into straggly boreal forest, lowering its albedo and amplifying longwave radiation absorption, and reradiation to the atmosphere, subject to local variability . As the greenest part of the season is June, peak plant productivity and peak transpiration (which humidifies the air promoting cloud formation) occur before peak of sea ice melt, which amplifies temperatures and is shifting peak warming in the Arctic earlier in the season, Chae et al. (2014)  modelling a shift from November to as early as August. At very localized scales, plant productivity is even more closely coupled to sea ice extent: as ice extent declines, the sea breeze bringing cold air to the coast is warmer, which increases plant productivity with associated feedback .
Nonetheless, “Arctic amplification is also found in models without changes in snow and ice cover” . Why?
Using a column of marine sediment from Western Svalbard stretching back 2000 years, Spielhagen et al. (2011)  found an “unprecedented” level of warming Atlantic water, suggesting the Fram Strait bordering the archipelago, the only deep-water inflow to the Arctic Ocean, has warmed by some 1.9K. A warmer Arctic Ocean during the winter is hence another heat flux which thins sea ice, and although studies highlight the importance of meridional heat transport, such as a 2016 study, which found the decreasing temperature gradient between the Arctic and the tropics to be the major factor in winter amplification in an ice-free Arctic Ocean , the dominant consensus is that more local temperature feedback mechanisms are dominant.
Planck radiation and lapse-rate feedback
Pithan and Mauritsen (2014)  argue persuasively that two of the negative temperature feedback mechanisms apparent elsewhere on Earth are more limited in the Arctic, highlighting that mechanisms, like Planck radiation, considered so fundamental to the greenhouse effect, are rarely given due consideration. Given that increasing the temperature of a surface increases the amount of radiation it re-emits by a fourth power (R = εσT4, where ε and σ are constants), the Arctic’s freezing temperatures mean that it retains a far larger amount of the thermal heat it absorbed than at the tropics. Secondly, they highlight lapse-rate feedback, although the importance they subscribe to this, up to 50% of total amplification – is qualified in other studies, Graversen (2014)  for example placing it at ~15%. With well-stratified, unmixed air in the Arctic, little convective transfer is present, preventing the upper troposphere warming and thereby radiating heat to space. Compared to lower latitudes, this retention of thermal energy close to the surface is much higher, consequently amplifying surface albedo change, but in itself a dominant cause of amplified temperatures beneath the cloud cover.
Pithan and Mauritsen (2014) present an informative chart of different models and the contributions of each process in different model simulations which informed their study.
It can be clearly seen, across the literature, that surface albedo change is a major, if not the key, factor driving Arctic amplification, a robust conclusion from many years of research and observation. Nonetheless, this is modulated and affected by the interlinking parts of the climate system – clouds, tundra, ocean flux, lapse rate – meaning that the greenhouse effect is concentrated much closer to the surface, everything amplified in vast feedback cycles that are likely to make the Arctic nearly ice-free by 2040.