First author

Climate is never static, but it does veer between relatively steady states and periods of drastic fluctuation. During the past 3 million years, climate has become increasingly variable. About 1 million years ago, the magnitude of glacial–interglacial oscillations increased significantly, with two of the most severe glaciations of the 3-million-year period occurring in the past 150,000 years. These nonlinear events are generally attributed to subtle cyclical changes in Earth's orbit or variations in atmospheric carbon dioxide concentrations. Thomas Crowley, a geoscientist at the University of Edinburgh, UK, and William Hyde, a physicist at the University of Toronto in Canada, suggest that this picture is incomplete. On page 226, they show how models of past climate transitions bolster their suggestion that the more variable recent climate may be an indication of an imminent — in geological terms — shift to a new steady state comprising stable mid-latitude ice sheets. But, Crowley tells Nature, anthropogenic increases in CO2 make any such transitions unlikely.

How did your climate modelling experience lead you to this hypothesis?

Twenty years ago, I noticed enhanced variability around climate transitions in very simple climate models. This stuck in my mind, and one day a few years ago I started tinkering with data to look for similar trends in ice-sheet fluctuations during the past half a million years. Since then, we've tested more realistic models, including ones for ice-sheet dynamics, to see whether they could simulate that enhanced variability. They did.

What causes the variability?

In some ways, it's analogous to variability in the financial markets. Conditions tend to be more volatile during times of rapid change. As the climate system approaches a more stable state, such as an ice-sheet expansion, the system varies even more. With respect to the current climate, we're suggesting that the system may be oscillating because it is approaching a stable state. If Earth's climate had not been interfered with by humans, our model suggests the transition could have happened within the next 100,000 years.

So humans might inadvertently have prevented the next ice age?

It's possible. At the moment, global warming is occurring at a much faster rate than is any transition likely to cause a climate switch in the opposite direction. The amount of CO2 in the atmosphere is already more than enough to prevent Earth from reaching a state with considerable ice cover. Some might argue that we should continue to pollute to prevent an ice age, but this is not a valid argument.