Molecular fossils and computer modelling have enabled scientists to build the first catalogue of Antarctic ocean temperatures over the past 45 million years, offering new insights into future sea level changes.

The team, led by scientists from Te Herenga Waka—Victoria University of Wellington, GNS Science, and Birmingham University (UK), says the results provide further evidence we may be nearing a ‘tipping point’ where ocean warming caused by atmospheric carbon dioxide (CO₂) will cause a major rise in sea levels as ice sheets melt.

The results are published today in Nature Geoscience.

“The record we’ve produced shows there is a very clear and direct response of Antarctic temperature to changing atmospheric CO₂ throughout geological time,” says Dr Bella Duncan, the study’s lead author.

“Our findings indicate we’ve crossed a threshold where increased CO₂ means oceans will warm to a level where there is major ice loss at the marine margins of Antarctic ice sheets, resulting in global sea-level rise over the coming decades and centuries,” she says.

In the study, researchers used molecular fossils from core samples taken during ocean drilling projects. These fossil remains are single lipid (insoluble in water) molecules produced by archaea—single-celled organisms that are similar to bacteria.

The archaea adjust the composition of their outer membrane lipids in response to changing sea temperatures. By studying these changes, scientists can draw conclusions about the ancient sea temperature at the time a particular sample died.

The researchers then used ‘machine learning’—computer systems employing algorithms and statistical models to analyse patterns in data.

The result of this work was the first record of changing Antarctic sea temperatures throughout much of the Cenozoic period, covering the past 45 million years. This record meant the team was able to identify with much more precision the historic temperatures that have caused ice sheets to grow and shrink in the past.

Dr Duncan says there is a clear link between CO₂, sea-surface temperatures, and the amount of ice on Antarctica throughout the past 45 million years.

She says one surprising finding was that ocean cooling did not always correspond with increases in Antarctic ice. This was observed in a one-million-year period of ocean cooling 24 to 25 million years ago.

“We showed this paradox is likely related to a gradual subsidence event where a once mountainous West Antarctic started to lower below sea level. This enabled a direct oceanic connection with the ice sheets, and because oceans melt ice sheets far more efficiently than the atmosphere, less ice was able to persist,” says Dr Duncan.

“Once West Antarctic lowered below sea level, colder oceans were needed to grow the marine ice sheet. In our computer simulations, this only occurred when CO₂ lowered below 400 parts per million (ppm). This indicates marine ice sheets can’t persist when atmospheric CO₂ exceeds 400ppm, a threshold we passed in 2013.”

The future loss of ice sheets and the retreat of glaciers in the Antarctic is critically important as melting ice in the region could cause sea levels to rise by several metres, Dr Duncan says.

“Ice in Antarctica is currently responding to warming seas with the loss of some ice shelves leading to accelerated retreat of one of the largest glaciers in the region, the Thwaites Glacier. The key challenge now is to slow emissions to prevent further acceleration of this ice loss and limit the impact of sea level rise as much as possible.”

The research team now plans to use the techniques in this study to reconstruct the climatic evolution of Antarctica and explore the implications for future warming and sea-level rise.

The team’s work was funded and facilitated by the International Ocean Drilling Programme, Antarctica New Zealand, the Royal Society Te Apārangi Marsden Fund, Natural Environment Research Council (UK), Scientific Committee on Antarctic Research, and a US National Science Foundation award. In-kind support was provided by the University of Birmingham, Yale University, and the Royal Netherlands Institute for Sea Research.