Declining Antarctic sea ice extents were a cornerstone of climate models - unless they began increasing. It may be that both are just natural fluctuation according to a new paper which shows that the negative phase of the Interdecadal Pacific Oscillation (IPO), which is characterized by cooler-than-average sea surface temperatures in the tropical eastern Pacific, has created favorable conditions for additional Antarctic sea ice growth since 2000.

Obviously that could mean that sea ice may begin to shrink as the IPO switches to a positive phase. Climate models have done a poor job of accounting for nature, they have tended to take a trend and made it linear into the future. Nature is not that predictable. 

Expanding ice

The sea ice surrounding Antarctica has been slowly increasing in area since the satellite record began in 1979. But the rate of increase rose nearly five fold between 2000 and 2014, following the IPO transition to a negative phase in 1999. The new study finds that when the IPO changes phase, from positive to negative or vice versa, it touches off a chain reaction of climate impacts that may ultimately affect sea ice formation at the bottom of the world.

When the IPO transitions to a negative phase, the sea surface temperatures in the tropical eastern Pacific become somewhat cooler than average when measured over a decade or two. These sea surface temperatures, in turn, change tropical precipitation, which drives large-scale changes to the winds that extend all the way down to Antarctica. The ultimate impact is a deepening of a low-pressure system off the coast of Antarctica known as the Amundsen Sea Low. Winds generated on the western flank of this system blow sea ice northward, away from Antarctica, helping to enlarge the extent of sea ice coverage.

Compared to the Arctic, global warming might cause only weak Antarctic sea ice loss, which is why the IPO can have such a striking effect in the Antarctic. There is no comparable natural variability in the Arctic that competes with global warming, the authors believe.

Sifting through simulations

To test if these IPO-related impacts were sufficient to cause the growth in sea ice extent observed between 2000 and 2014, the scientists first examined 262 climate simulations created by different modeling groups from around the world.

When all of those simulations are averaged, the natural variability cancels itself out. For example, simulations with a positive IPO offset those with a negative IPO. What remains is the expected impact of human-caused climate change: a decline in Antarctic sea ice extent.

But for this study, the scientists were not interested in the average. Instead, they wanted to find individual members that correctly characterized the natural variability between 2000-2014, including the negative phase of the IPO. The team discovered 10 simulations that met the criteria, and all of them showed an increase in Antarctic sea ice extent across all seasons.

When is the next change? Scientists speculate that in 2014, the IPO began to change from negative to positive and so climate models would start to look more accurate. An upcoming period of warmer eastern Pacific Ocean surface temperatures on average, though year-to-year temperatures may go up or down, depending on El Niño/La Niña conditions. Accordingly, the trend of increasing Antarctic sea ice extent may also change in response.