Chance of a snowball

As the weather in New York oscillates between the warmth of spring and the bitter cold of a winter we'd all hoped was gone, I've been reminded of some of the peculiarities of a 'habitable' planet. Over the past few months I've tagged along with some interesting work on terrestrial planet climates. Recently we put out a couple of papers that discuss in some detail how the orbital arrangement of a world can alter climate and climate variation (Dressing et al., Spiegel et al.). The interesting thing about Earth-type planets in our models - rocky worlds with surface water and atmospheres - is that they have a tendency to go snowball. Going snowball isn't some odd mental state, it's when a planet drops below the freezing point of water from pole to pole. The Earth may have done this itself, about 700 million years ago, since there's evidence of glaciers near the equator during that time. Because frozen surface water reflects more radiation and keeps things cool then once you're snowballed you tend to stay that way. On the Earth it may have been geophysics that got us unfrozen, over millions of years more greenhouse gases got dumped into the atmosphere from volcanic activity until things heated up again. On different planets, where their orbits get strongly perturbed by other objects in the system - increasing and decreasing in ellipticity - this kind of thing may happen on a regular basis, but it's the orbital swings that push and pull a world in and out of a frozen state. I think it's fascinating, not least because if we eventually spot planets like this they may present a conundrum - either caught in the act of freezing up or thawing out. In this case we'll have to consider their orbital circumstances very carefully in order to know whether or not they are merely experiencing a bad case of seasonal adjustment....