Antarctic Sea Ice

Guest post by Dr Jonathan Day, Department of Meteorology, University of Reading

What is going on with the Antarctic sea ice?

March 2017 was an interesting month for sea ice. Both northern and southern hemispheres experienced record breaking low extents for the time of year. The extent of Arctic sea ice reached the maximum area of its seasonal cycle on March 7th coming in at 14.42 million km2. This was a fraction below the previous record, set in 2015 and is in line with what we expect to see in a warming climate. Meanwhile the other side of the planet Antarctic sea ice continues to confound expectations.

Increasing Antarctic sea ice

Over the last 38 years the area covered by sea ice in Antarctica has been increasing slightly in all seasons, leading to record high conditions reported in 2015. This is not what one would expect in a warming climate. However, this year has gone in completely the other direction and on March 3rd the all-time record minimum of 2.11 million km2 was announced, about 25% below normal. So what does this all mean and why was the sea ice increasing despite global warming?

Firstly, one year, even a record breaker, doesn’t tell us a lot more than we knew before. We know that the magnitude of year-to-year variability of sea ice in Antarctica is very high compared to the long term trend [Fig 1].

Figure 1. February monthly mean Antarctic sea ice extent from NSIDC.

There are a number of competing theories as to why the ice has been increasing and these can be split into two categories:

  1. Changes associated with human activities;
  2. Natural variability.

Human activities causing changes in Antarctic sea ice

In the first category, physically plausible mechanisms have been proposed that link human activities associated with the creation of the ozone hole1 and increased runoff from the Antarctic ice sheets2 (land ice) to increased sea ice. However, different studies have come to different conclusions regarding the magnitude of these effects.

Natural variability in Antarctic sea ice

The second category relates to climate variability from natural causes. For example we know that the major modes of climate variability such as the El Nino Southern Oscillation (ENSO) and the Southern Annular Mode (SAM) project strongly onto Antarctic sea ice variability. In addition, climate models and observations suggest that there may be modes of variability which act on multi-decadal timescales, although understanding of such modes is currently limited3.

Signal-to-noise ratio in sea ice changes

In order to detect the influence of climate change we need the signal caused by man-made changes to be large compared to natural variability. We can measure this ratio in climate model experiments and express it as a signal-to-noise ratio4. Climate models suggest that this ratio is small in the Southern Ocean compared to other parts of the world, therefore the signal of change may be drowned by the noise of variability [see the low values around Antarctica in Fig 2].

Figure 2. multi-model mean CMIP5 simulated change in air temperature over the 21st century divided by the simulated amplitude of natural variability – the signal-to-noise ratio (from Ed Hawkins).

Another line of evidence is that sea ice and temperature trends in the Southern Ocean changed sign in the 1970s for no apparent reason. The climate was generally warming from 1950-1978 and the cooling thereafter5 [Fig 3]. To me this is highly suggestive of natural multi-decadal variability, rather than a forced change6, but the jury is still out.

Figure 3. Southern Ocean SST and sea ice trends from HadSST, for the periods 1950-1978 (left) and 1979-2014 (right) and the zonal mean of both (middle) from Fan et al. (2014). Sea ice concentration is not available for the 1950-1978 period.

Is a signal starting to emerge?

Although one low year is not enough to tell if the sign of the trend is changing it is may be a sign that the climate change signal is starting to emerge from the noise of natural variability.