Vast uncertainty remains over the causes of melting Arctic sea ice and when it may disappear altogether during the summer, which would have consequences for oil explorers, shipping firms and the fight against climate change.
The answer will depend on the balance of natural and manmade causes.
Those causes include warmer air and seas as a result of greenhouse gas emissions, variations in atmospheric circulation, and a faster southward ice drift down the east coast of Greenland.
A major study said last month that science had underestimated the impact of soot, also called black carbon, adding another complication to an already difficult task of making a forecast.
The Arctic region accounts for 15 percent of the world's undiscovered oil and 30 percent of its undiscovered gas, the U.S. Geological Survey estimates.
Scientists are still unsure about the exact causes of sea ice melt, and predictions of when the North Pole could have an ice-free summer vary from 2015 to 2080 or later, with forecasts centring around 2040.
Oil producers including Russia's Rosneft, Norway's Statoil, U.S.-based Exxon Mobil and Anglo-Dutch Shell are preparing to drill in areas of melting sea ice, despite the technological difficulties and costs. .
SEA ICE RETREAT
Last September sea ice reached its lowest level in the satellite record, which dates back to 1979, a development that has implications for local native communities and wildlife, local coastal erosion and possibly northern hemisphere weather.
Observations of the extent of Arctic sea ice made by ships in earlier decades suggest last year's record would at least stretch back to a cold period in the mid-nineteenth century known as the Little Ice Age.
Data show a clear trajectory of ice losses in recent decades.
An acceleration last year past a previous record in 2007 reflects a self-perpetuating process as well as thinning over many decades.
Under the albedo effect, dry snow reflects more than 80 percent of solar radiation; bare ice 65 percent; and open water just 5 percent.
And increasing expanse of open water each summer warms up faster than ice-covered sea, meaning new ice will be thinner and more vulnerable the following year.
There are various underlying causes of the melt.
Rising greenhouse gas emissions drive up air and sea temperatures.
For example, a discussion paper published online last month, "North Atlantic warming and declining volume of arctic sea ice", highlighted the role of water from a warmer Atlantic entering the Arctic Ocean through the Fram Strait between eastern Greenland and Svalbard, an island due north of Norway.
A similar inflow of warmer Pacific water through the Bering Strait has long been identified as an important process that causes the thinning of ice in the central Arctic.
Last month's study estimated a particularly strong impact from warm Atlantic water north of Svalbard, where recent accelerated melting is clearly visible.
"The Atlantic water temperature has been increasing since the mid-1960s," said four scientists from U.S., Russian and Norwegian institutions.
"Our lower-end conservative estimates indicate that the recent Atlantic water warming episode could have contributed up to 150-200 cubic kilometres of sea ice melt per year, which would constitute about 20 percent of the total 900 cubic kilometres per year negative trend in sea ice volume since 2004."
Another human-derived impact is from soot emissions as a result of forest fires; household and industrial burning of wood and coal; and diesel transport fuel.
Such soot, also called black carbon or particulate matter, contributes more to global warming than previously thought according to a major study published last month - "Bounding the role of black carbon in the climate system: A scientific assessment".
Black carbon absorbs and scatters sunlight both in the atmosphere and when deposited on snow and ice, where it darkens the surface.
But there are complications. For example, local warming of the Arctic atmosphere could reduce poleward winds from the south, which may result in a cooling effect.
The net impact from the combined effects of black carbon is to warm the Arctic, experts who contributed to the recent report concluded in an email correspondence.
There are also weather effects that may be natural, partly natural or entirely due to greenhouse gas emissions.
A particular weather pattern contributed to a big melt six years ago by sending air towards the central Arctic, according to researchers from the Netherlands, Sweden and Germany.
"We find that in summer 2007 there was an anomalous atmospheric flow of warm and humid air into the region that suffered severe melt," the researchers said in the paper, "Warm winds from the Pacific caused extensive Arctic sea-ice melt in summer 2007".
In a longer trend over the past 50 years, stronger winds have caused a faster southward drift of ice through the Fram Strait along the east coast of Greenland, according to researchers in Norway.
The winds are a result of air pressure changes and more intense storms in the Nordic Seas north of Norway. The ice flow is additionally channelled along the Greenland coast and rides on the underlying, southward East Greenland Current.
"The dramatic loss of Arctic sea ice in the last decades is thus only partly caused by increased long-wave radiation related to ongoing atmospheric CO2 increase," the authors reported in their 2011 paper, "Recent wind driven high sea ice export in the Fram Strait contributes to Arctic sea ice decline".
"The ice export has likely been an effective contributor to Arctic ice loss since the 1960's."
Given that scientists have not yet pinned down the causes, it is impossible to predict further sea ice decline.
The outlook depends on how much of a role has been played by natural variability, which could reverse.
The opposite scenario is one of a growing, cumulative impact from rising greenhouse gas emissions, which combined with a few freak summers, ice export and the albedo effect could finish off summer sea ice rather quickly.
"Simulations ... show that after the ice thins to a more vulnerable state in response to rising greenhouse gas concentrations, a reinforcing kick from natural variability may trigger an initial, abrupt ice loss," several seasoned Arctic ice watchers wrote in the journal, "Eos Transactions American Geophysical Union" in 2008.
"While natural variability may instead stabilize the ice cover for the next few years, the long-term outlook is disturbing. Our view is that a seasonally ice-free Arctic Ocean might be realized as early as 2030."