Sudden Southern Ocean sea ice decline

[tsemmler05]

In the 2030s of the ssp585d simulation a sudden decline of Antarctic sea ice occurs, after it had been relatively stable up to around 2025. In this issue plots can be collected.

[tsemmler05]

This is the Southern Ocean sea ice area as monthly and annual means and anomalies along with global surface air temperature and El Nino surface air temperature and index analysed by Dae-Won.

Kim-DW-HR_230802.pdf

[tsemmler05]

It seems like that south of 70S mainly a sudden surface warming is taking place while the lower ocean layers are not that much affected. However, if considering the whole Southern Ocean basin, a gradual warming can be seen also in deeper ocean layers up to 500 m.

South of 65 S:

South of 40 S:

[tsemmler05]

To see if the simulations gets to a different trajectory when stabilizing the forcing at 2020 conditions, the simulation c2020d is branched off from ssp585d simulation at the end of 2019 with constant greenhouse gas and aerosol forcing. The Antarctic sea ice is rapidly melting in both cases - even a bit faster in c2020d than in ssp585d:

The years 2020-2034 are shown - sorry, not sure why the years are not displayed on the x-axis.

[tsemmler05]

This is now the Hovmoeller diagram of the potential temperature for two simulations south of 65 S, once the standard ssp585d simulation and once the constant 2020 forcing simulation c2020d branched off at the end of 2019 from the SSP585d simulation. The warming is even a bit stronger in c2020d.

[tsemmler05]

This is the same Hovmoeller diagram but south of 40 S:

Here, a similar warming can be seen in both simulations. It seems like the slightly different forcing does not bring the model towards a different trajectory.

[tsemmler05]

One could say that c2020d warms slightly less and slightly later than ssp585d in 100 to 500 m depth when averaged over the large area south of 40S but slightly faster and slightly more at the surface when averaged over the smaller area south of 65S.

[qiangclimate]

hi Tido, would it be informative to have a look at the spatial changes in 2D plots? Is the quick change in sea ice confined to some sectors?

[tsemmler05]

It is especially in the Weddell Sea and east of it, in the c2020d even more than in the ssp585d simulation - analysis by Dae-Won.

ssp585&c2020d.pdf

[tsemmler05]

While one clearly sees the reduced forcing in the global mean temperature which increases less in c2020d compared to ssp585d as expected, the Weddell Sea ice and the ice further east does something unexpected, maybe related to convection?

[mzapponi]

this is the ocean potential temperature change (scenario minus control run) for both the experiments. For the gloabl ocean (left) the temperature increase is indeed weaker but looking only at the southern ocean (SO) and at the area south of 65 degrees (65S) it clearly shows an earlier warming in c2020d.

[mzapponi]

Sea ice extent in the SH in September looks really interesting! An in general the two simulations c2000d and c2020d seem to follow a quite similar path

[mzapponi]

and here is the updated one, is interesting to see how the Arctic Ocean sea ice extent seems to be not much affected by the not increasing forcing and continues decreasing. While in the Southern Ocean one can really see the difference.

[mzapponi]

The ocean potential temperature change (with respect to the control run) shows quite a difference indeed between the ssp585 (top row) and the c2020 (middle row) simulations. The bottom row refers to the c2000d simulation.

[mzapponi]

Looking at the mean meridional ocean heat transport and surface energy balance for two time interval before and after the sea ice extent minimum of c2020d (and comparing it to the same years of ssp585d) there is more difference in the atmosphere rather than in the ocean heat transport.