Published online 10 December 1998 | Nature | doi:10.1038/news981210-1


At full tilt

The severity of the ice ages of the past two million years was eclipsed by glaciations in the distant past. The period between 820 and 550 million years ago saw intermittent ice ages in which even the tropics were covered by glaciers. According to the 'Snowball Earth' hypothesis, the Earth was completely glaciated, from the poles down to the Equator. Perhaps even the oceans froze, down to a depth of a kilometre or more. The planet was only saved from a permanent freeze by a massive injection of carbon dioxide from volcanoes, a 'greenhouse gas' that melted the ice. The problem with this idea is that life, already well-established, might easily have been driven to extinction on a Snowball Earth.

One way of explaining the coexistence of tropical ice-sheets and the continuation of life is to imagine a system of climate turned on its head, with the tropics colder than the poles. That way, glaciers could encircle the tropics, while life continued to flourish elsewhere, even at ice-free poles. This idea is perhaps even more bizarre than the Snowball Earth idea.

Putting scepticism aside, how could such a reversal of roles have happened? Geologist George E. Williams of the University of Adelaide in Australia has advanced one mechanism, in which the Earth's axial tilt - the degree to which the Earth's axis of rotation is tilted away from the vertical, where the vertical is itself at right angles to the plane of the Earth's orbit around the Sun - was much greater than it is today.

At present, the axial tilt, or 'obliquity', is just over 23 degrees from the vertical. This ensures that the Equator receives more sunshine, averaged over the year, than the poles, partly explaining why the tropics have palms and the poles have polar bears and penguins. Williams suggested that a larger obliquity, greater than about 54 degrees, would result in the poles receiving more sunshine than the Equator - leaving room, in principle, for tropical glaciation and ice-free poles. The problem of the Snowball Earth, then, could be resolved if evidence could be found that the Earth once had a much greater obliquity than it does today.

This idea is controversial, and raises further problems. One is that there is good evidence from other sources that the Earth's axial tilt 430 million years ago was similar to that of today. If so, then the axial tilt must have shifted by around thirty degrees in an interval of 100 million years or less.

How can obliquity change so rapidly? Darren E. Williams of Penn State Erie, Pennsylvania and colleagues have come up with a possible solution, which they describe in a report in Nature. Their answer relies on the fact that obliquity is not fixed: it oscillates very slightly, and this oscillation influences climate. Indeed, the succession of ice ages in the past two million years is related to a 41,000-year cycle of fluctuations in obliquity, in which the axial tilt varies by a dgeree or so either way.

But climate may also, in turn, have an effect on obliquity. When, in ice ages, large volumes of water are sucked out of the oceans to form ice caps, the consequent alteration in the distribution of mass also affects axial tilt, as the Earth 'shifts itself' to compensate. Williams and colleagues suggest that the interaction between obliquity and mass distribution drove itself in a process of self-amplification called a 'positive feedback' loop: A shift of mass led to a change of tilt, which produced to a further shift of mass, another change of tilt, and so on. The researchers calculate that the process was sufficient to shift the Earth's axis of rotation through the required angle.

The researchers find support for their view from an unexpected quarter - the Moon. Earth's only satellite has an orbit inclined by about five degrees to the plane of the Earth's orbit around the Sun. This inclination is hard to explain. However, there is an interaction between the lunar orbit and the Earth's obliquity. For example, obliquity is currently increasing slightly, as a result of tidal interactions with the Moon. Past changes in obliquity might have left an imprint on the Moon's orbit. Using the principle that momentum in the Earth-Moon system will have been conserved, and asuming that the Moon's orbit orginally had no inclination, the researchers calculate that a shift in obliquity of around 25 degrees would be compensated by the inclination of the lunar orbit by about five degrees, as observed.

This research will not be greeted with approbation everywhere, and runs counter to other evidence that obliquity underwent no massive shifts. For example, the alignment growth patterns in fossil stromatolites (colonial algae) dating back more than 800 million years suggest that obliquity was more or less the same as today. This only regenerates the Snowball Earth problem - but there may be good reason for thinking that the Snowball Earth was not such a problem after all. Rocks laid down during the intervals concerned have an unusual geochemistry, best explained by supposing that the oceans and the atmosphere were sealed off from one another for five to ten million years. A kilometres-thick sheet of ice would have done the trick nicely - but how would have life fared on a world totally frozen for this length of time?

This confused picture is complicated further by other sudden alterations that the Earth is postulated to have undergone at around the same time. Reporting in Science last year, Joseph L. Kirschvink of Caltech, Pasadena, California and colleagues presented evidence for a massive and sudden reorganization of the Earth's crust and mantle between around 535 and 520 million years ago. This event, a response to an unstable distribution of mass, led to a 90-degree rotation of most, if not all the continental plates in just 15 million years. This dramatic series of events was not, however, related to any changes in obliquity that might have occurred at the same time.

If that wasn't enough, the period between 600 and 540 million years ago saw an explosive burst of evolution that produced a plethora of animal forms. Kirschvink and colleagues suggest a link between this evolutionary burst and the tectonic events that occurred at about the same time.

Clearly, a lot more work needs to be done before a universally agreed picture of the Earth's turbulent past begins to emerge. It is true that the Earth's biosphere underwent dramatic changes between 800 and 500 million years ago, including alterations in the chemistry of the oceans and the atmosphere, all of which could have produced a response in the living world, and which would have, in turn, have influenced - and have been influenced by - the geophysical events of an extraordinary interval in the Earth's history.