Every year sea ice around Antarctica expands and contracts with the seasons and reaches its annual minimum extent at the end of summer. This year, Antarctic sea ice coverage in February was the lowest it has ever been since regular satellite observations began. This is remarkable – there have now been three record summer sea ice minima in the past seven years – with some suggesting Antarctic sea ice may have entered a new regime. This is the focus of our paper that we started writing in late January, as we watched sea ice approach its annual minimum with trepidation.

It’s a funny feeling, watching the subject of your research head towards another record. While the community was (and still are) working on understanding the record minimum sea ice in 2022, another record was unfolding in front of us. Climate models have long projected a decline in Antarctic sea ice, and yet, this didn’t make it any less alarming.

Our study is a brief rapid response to the remarkable sea ice behaviour this year, with two key findings. Firstly, when we look at the sea ice record, we detect three distinct periods: a neutral sea ice period from November 1978 to August 2007, a high sea ice period from September 2007 to August 2016, and a low sea ice period from September 2016 to the present. Of note – what we thought we knew about sea ice does not seem to be holding anymore in the current low sea ice state. And secondly, we find evidence that ocean warming is contributing to the current low sea ice state.

For context, let’s rewind a little bit.

The atmosphere is an important driver of Antarctic sea ice variability and change.

In the 2010s, Antarctic sea ice was increasing, counter to what we expected given increasing global temperatures. A lot of progress into understanding drivers of Antarctic sea ice was made and in particular, winds were found to be important in driving the increasing sea ice trend of the time.

Then, in the austral spring of 2016 Antarctic sea ice coverage dropped dramatically. Over two years, the Antarctic lost as much sea ice as the Arctic had lost in three decades. Once again, atmospheric conditions were largely identified as driving the decline in sea ice. And since 2016, Antarctic sea ice coverage has remained low. This year it’s crazy low.

 While the atmosphere remains an important driver of sea ice, and individual low sea ice events have been linked to anomalous atmospheric circulation events, an important consistency across the current low sea ice period is a warm subsurface ocean. Further, our analysis finds that subsurface ocean warming occurred prior to the 2016 sea ice decline, supporting the hypothesis of Meehl et al. (2019) that the warm subsurface ocean was an additional and important driver of the low sea ice in spring 2016 and the sustained low sea ice state since.

The characteristics of Antarctic sea ice persistence – memory in the sea ice system tightly linked with the ocean – have also changed. This is further evidence that the role played by the ocean has changed in this new sea ice state.

 Antarctic sea ice coverage since the summer minimum has been even more remarkable. Following the seasonal cycle, sea ice has expanded over autumn and winter, however it has done so at a much slower rate than normal. As a result, Antarctic sea ice coverage this winter has been much lower than normal for the entire austral winter. Much, much lower. Nothing like this has occurred during the observational period before.

We need to understand future sea ice variability and change because sea ice is important.

 Sea ice is important for ecosystems. Sea ice provides habitat for krill, a foundational species in Southern Ocean food webs, and low sea ice in late 2022 led to catastrophic breeding failure of emperor penguins. What will happen this breeding season?

 Sea ice is also important for global climate and sea levels. A new low sea ice regime will have important impacts including changed ocean stratification and circulation. This will affect the deep ocean ventilation and basal melting beneath ice shelves. Reduced sea ice will also lead to high coastal exposure with likely consequences for ice-shelf stability.

 We urgently need more research to understand the physical processes driving sea ice variability and change in this new, low sea ice state, and what this means for the future. We have an indication that the underlying processes controlling Antarctic sea ice coverage may have altered, and more work is needed to understand how both the atmosphere and ocean are driving sea ice changes.