What causes the low Antarctic sea ice extent since 2016?

[Steven]

I couldn't find a thread about this subject, so I'm starting a new one.

Antarctic sea ice extent had a slight upward trend from 1979 to 2015, followed by a precipitous decline in 2016 to a new state of low sea ice extent that has persisted since then.

An interesting recent paper about this subject:

https://www.nature.com/articles/s43247-023-00961-9#MOESM1

The study gives some evidence that subsurface ocean heat may be the primary reason for the reduced sea ice since 2016.

Also very interesting is the suggestion that the warming since 2016 is the second phase in a two-phase process triggered by the strengthened westerly winds around Antarctica in the last few decades:

Ferreira et al. proposed that the surface Southern Ocean response to strengthened westerly winds occurred over two distinct timescales: a rapid cooling due to enhanced northwards Ekman transport and vertical mixing, and a long-term warming due to upwelling of warmer subsurface warmers. In this context, it is noteworthy that the current observed ocean warming [since 2016] follows a period of surface cooling [between 2007 and 2014], consistent with the first short-term cooling phase of the two-timescale response, and that both the initial cooling and subsequent near-surface ocean warming have occurred under a backdrop of prolonged Ekman upwelling.

Some more extracts from the paper:

The change points [in 2007 and 2016] identified do not align with changes in large scale climate modes known to impact sea ice variability, such as the Southern Annular Mode, the El Niño Southern Oscillation, the Interdecadal Pacific Oscillation, or the Atlantic Multidecadal Oscillation.   Furthermore, climate modes have strong regional impacts, while the current low sea ice state is characterised by circumpolar low sea ice anomalies.

...

The atmosphere is an important driver of Antarctic sea ice variability and change.   Individual extreme sea ice events have been linked to anomalous atmospheric circulation events .... However, the phase of the Southern Annular Mode cannot account for the low Antarctic sea ice over recent years: the positive phase of the Southern Annular Mode has historically been associated with colder sea surface temperatures and increased sea ice extent, and in line with this, the 2017 record summer sea ice minimum occurred at a time of anomalously warm surface temperatures in the Southern Ocean and an anomalously negative Southern Annular Mode.   In contrast, the 2022 and 2023 record sea ice minima occurred when surface temperatures were close to the climatological mean and the Southern Annular Mode was anomalously positive, while the subsurface ocean was anomalously warm.

...

Using a gridded Argo dataset, we present the circumpolar average ocean temperature anomaly between 50–65°S (Fig. 1c).   The shift from cold subsurface temperatures to warm subsurface temperatures (below 100 m; Fig. 1c) occurs approximately one year before the second change point [in 2016] identified in the sea ice extent time series (Fig. 1a), strongly suggesting that the current low sea ice extent state is due, at least partly, to subsurface ocean warming.

...

While there is evidence that climate modes contributed to low sea ice coverage in recent individual years, when averaged over the seven years of the current period, a circumpolar decline in sea ice is seen, coincident with a circumpolar subsurface ocean warming.   Importantly, subsurface ocean warming occurs priors to the reduction in Antarctic sea ice, with strong spatial agreement between warm anomalies at 100–200 m depth in 2015 (340–20°E and 190–250°E), and the subsequent strong negative sea ice anomalies in 2016 (10–90°E and 190–290°E), accounting for the eastward advection of anomalies over time  (Fig. 3).

[gerontocrat]

I found one paper that associates strengthening poleward-shifting circumpolar southern hemisphere Westerly winds with AGW by looking at 20th Century data and well as satellite data.

However, I can't find anything about any connection of this with sub-surface ocean temperatures and/or Antarctic sea ice extent and area.

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2021GL095999

Strengthening Southern Hemisphere Westerlies and Amundsen Sea Low Deepening Over the 20th Century Revealed by Proxy-Data Assimilation

Abstract
Winds and pressure over the Southern Ocean are critical to many aspects of the climate system, but the brevity of climate data in this region makes it challenging to interpret recent changes. Here, we reconstruct 20th century sea level pressure and zonal surface wind anomalies over the Southern Ocean, using data assimilation with a global paleoclimate proxy database and four climate-model priors. The reconstructions agree well with instrumental reanalysis products, especially in the circumpolar westerly and Pacific regions. We observe significant strengthening in the midlatitude Pacific westerlies, associated with a deepening Amundsen Sea Low, throughout the 20th century in all four reconstructions. When the prior includes anthropogenic forcing, we observe poleward-shifting circumpolar westerlies throughout the 20th century. Our results highlight the combined roles of natural variability and anthropogenic forcing, and the zonally asymmetric character of atmospheric circulation changes at high southern latitudes, with implications for Antarctic ice sheet change.

Key Points
- Paleoclimate proxy-data assimilation skillfully reconstructs sea level pressure and surface winds in the high latitude Southern Hemisphere

- Reconstructions show 20th century Amundsen Sea Low deepening and strengthening circumpolar westerlies, particularly in the Pacific sector

- Twentieth century poleward-shifting circumpolar westerlies are observed only when anthropogenic forcing is included in the climate-model prior

Plain Language Summary
Wind and pressure patterns over the Southern Ocean modulate ocean carbon uptake and may influence the stability of the Antarctic ice sheet. Observations suggest that westerly winds around Antarctica are strengthening and shifting closer to Antarctica. However, modern instrumental observations (i.e., data from satellites, radiosondes, and weather stations) in this region begin only in 1979, making it difficult to place these changes in the context of natural climate variability over longer timescales. Here, we use paleoclimate proxy records (e.g., oxygen-isotope records from ice cores), combined with simulations from climate models, to produce maps of surface pressure and winds over the Southern Ocean spanning the 20th century. Our results compare well with available independent data sets. We find that strengthening westerly winds in the Pacific sector of the Southern Ocean has occurred through most of the 20th century and is associated with a deepening of the Amundsen Sea Low. Our results suggest that this is accompanied by poleward-shifting westerly winds around the rest of Antarctica, attributable to anthropogenic forcing (i.e., increased greenhouse gases).

[kassy]

But what time scales and effects should we look at?
If it is the ocean which seems most likely to me (see AABW decline paper) then that would probably change winds or at least influence them locally. And while local changes happen there are some global ones.

[oren]

The discussion is about Antarctic sea ice, I've now modified the thread subject to reflect this.

[Steven]

I found one paper that associates strengthening poleward-shifting circumpolar southern hemisphere Westerly winds with AGW by looking at 20th Century data and well as satellite data.

However, I can't find anything about any connection of this with sub-surface ocean temperatures and/or Antarctic sea ice extent and area.

The ring of westerly winds is at the northern edge of the ring of cyclones around Antarctica.  There is upwelling of subsurface water at the center of each cyclone (Ekman upwelling). 

Strengthened westerly winds over the past decades cause strengthened upwelling.


As for the effect on sea ice:

Initially the strengthened Ekman upwelling/transport would have a cooling effect on the surface for several years. 

But a study suggested that eventually the upwelling would drag deeper warmer water from below the cold surface layer upward toward the surface.  That would lead to a long term warming of the surface layer.

But that is still speculative.  Another study suggests that the deeper warmer water wouldn't be upwelled, since the upwelling would be disrupted by eddies and hence only cold water from the mixed surface layer would be upwelled.


TLDR: strengthened westerly winds around Antarctica may have caused upwelling of deeper warmer water, causing the warming of the topmost ocean layer around Antarctica since 2015/16, but that is still unsure.  Maybe other factors such as the 2015/16 super El Nino played a role but the warmth has persisted since then.