New Zealand scientists are looking to the planet's past to solve a question vital to its future: how warm would ocean waters have to get to melt Antarctica's marine ice sheets?
The frozen continent below us locks down an equivalent 60m of potential global sea level rise – and there's still much uncertainty around how and when climate change might push it to tipping point.
But as our understanding widens, the picture grows more serious.
One recent study co-authored by a Kiwi glaciologist suggested the threshold could be crossed somewhere near another 1C of global warming, or over coming decades.
In a new project, supported with a $960,000 grant from the Marsden Fund, a group of researchers have set out to tackle some of the most enduring questions around how Antarctica will respond to a warming world.
The Victoria University-led study builds on a key takeaway of the New Zealand-designed, international ANDRILL drilling expedition, which was carried out on the ice a decade ago.
That offered the first clear evidence that the vast West Antarctic Ice Sheet had collapsed between one and three million years ago, in climates that were similar to what's expected by the end of this century.
More recent drilling by the International Ocean Drilling Program found that part of the much larger East Antarctic Ice Sheet collapsed as recently as 120,000 years ago.
The parts lost were marine-based ice sheets sitting on the Earth's surface below sea level, which have the potential to raise sea levels by 20m.
Importantly, the loss of these sheets was thought to be due to oceanic warming melting the ice at its marine margin, rather than "top down" melting by the atmosphere.
The new study's leader, Associate Professor Rob McKay of Victoria University's Antarctic Research Centre, said the focus this time wouldn't be on the sheets themselves, but what happened in the oceans surrounding them.
"Key questions we aim to answer are: how warm did the oceans directly offshore of Antarctica get before we lost the marine ice sheet?" he said.
"And did shifting oceanic currents push warm waters that currently exist further north in the Southern Ocean closer to the Antarctic ice sheets, greatly accelerating their retreat?"
McKay and his team also want to know what happened when a large volume of freshwater was released from the Antarctic Ice Sheet to the Southern Ocean, and how marine plankton might respond to that biologically.
"The most obvious impact is sea level, but a fresher ocean could lead to changes in sea ice extent, nutrient delivery to marine plankton, and affect the way oceanic heat is transported from Antarctica to the rest of the planet - as a large component ocean circulation near the edge of the Antarctic continent is driven by changes in the density of surface ocean waters," he said.
They're also interested in how a smaller ice sheet might influence the wind fields in the Southern Ocean.
A smaller ice sheet could result in the primary wind field moving southward towards the continent, which in turn may push warmer waters further south, causing a further acceleration of retreat.
Joining McKay is a team of 30 international researchers who collected a series of sediment samples from the outer Ross Sea last summer.
"These sediments carry important indicators of the physical strength of ocean currents that may deliver warmer waters southwards towards Antarctica, and glacial meltwater being washed into the ocean."
Other researchers will look at microfossil of plankton and chemical tracers that will reconstruct the temperature and salinity of past waters offshore of Antarctica, as well the biological response of these changing oceanic and climatic conditions.
"We expect to be able to identify thresholds of oceanic warming for past ice sheet retreat, as well as understanding the physical drivers for how waters may migrate southward towards Antarctica and contributing to past ice sheet collapse event.
"We also want to understand the processes that led to polar amplification of warming - whereby the poles warm up much more that global average temperature during past warm climates."
Sea ice, ice shelves and marine-based ice sheets, he explained, all helped to keep the ocean cold.
If these components of the cryosphere were lost or reduced due to human-induced climate change from greenhouse gas emissions, the overall warming response of the planet would be greatly amplified compared to that from the greenhouse gas input alone.
McKay noted a big part of the study was linking in with computer-simulated ice sheet models.
"Existing models indicate the area we drilled was the most sensitive to changes in oceanic heat changes in the geological past," he said.
"The data will help validate if those models are behaving realistically based on naturally-induced climate change example in the past, and from this we can assess if these same models are likely to be reliable indicators of future changes in this modern example of human induced climate change.
"If we see part of the natural system from our geological record behaving differently to these models, it may also provide indication of what processes need to be better represented in these models - and whether these could act to amplify or dampen future ice sheet melt."