GEOLOGISTS had their work cut out at the meeting of the AGU last week when
they presented new evidence in favour of the controversial 鈥渟nowball Earth鈥�
theory. They were up against climate researchers who say they simply can鈥檛 make
their best computer models produce the thick ice sheets which the geologists
claim once shrouded the Earth.
The controversy was ignited two years ago by Paul Hoffman, a geologist at
Harvard University, who was trying to account for evidence that suggested
glaciers reached the equator between 750 and 600 million years ago
(快猫短视频, 6 November 1999, p 28).
He proposed that the atmosphere lost
much of its carbon dioxide, causing it to cool and ice to start spreading. As
the ice advanced, it reflected more and more sunlight away from Earth, leading
to a runaway cooling effect and thick ice over the whole planet.
To explain how the freeze ended, Joe Kirschvink of the California Institute
of Technology had earlier shown that volcanoes could add enough greenhouse gases
to the atmosphere over millions of years to produce an intense warming effect.
This would melt the global ice sheet and cause a brief exceptionally hot spell
after the thaw.
Advertisement
At the AGU meeting, Kirschvink presented new evidence in favour of the
theory: a huge manganese ore deposit in South Africa which, he says, formed
during an early snowball episode 2.4 to 2.3 billion years ago. Underneath the
global ice sheet, minerals such as manganese from hydrothermal vents would build
up, creating a mixture 鈥渟imilar to the nutrient media typically used for growing
肠测补苍辞产补肠迟别谤颈补鈥�.
These bacteria would have been unable to grow while thick ice blocked out the
sunlight. But as soon as the ice melted, cyanobacteria would have bloomed in the
mineral-rich waters, Kirschvink says, releasing huge amounts of oxygen. Some of
this escaped into the atmosphere, but the rest reacted with soluble manganese
and iron compounds, precipitating them on the ocean floor. Kirschvink says a
similar mechanism created deposits during the later snowball episodes.
But other researchers still doubt whether the oceans could have frozen as
solidly as Hoffman and Kirschvink propose. Biologist Bruce Runnegar of the
University of California at Los Angeles prefers a 鈥渟lushball鈥� model, in which
the tropics did not freeze completely. He says that cells which require oxygen
would struggle to survive a hard freeze. Thick ice would halt photosynthesis,
block oxygen supplies, and the post-thaw heat would have been deadly. 鈥淎ll this
would have been very troublesome for life,鈥� he says. A slushball 鈥渨ould have
been much easier to survive鈥�.
Climate modellers say that in the later proposed snowball episodes, it would
be difficult to produce a hard snowball, even though solar radiation was about 6
per cent lower than it is today. The stumbling block is that ocean currents
transport heat, says Greg Jenkins, a meteorologist at Pennsylvania State
University. Richard Peltier of the University of Toronto found that in a model
created by the US National Center for Atmospheric Research, these ocean currents
鈥渟trongly inhibit global glaciation鈥�.
Three other researchers using different ocean-coupled models also reported
that the equatorial oceans remained open, even when the models included our best
guess at the actual geography of the time. 鈥淭he greater the complexity of the
ocean model, the harder it was to make a snowball,鈥� says Mark Chandler of NASA鈥檚
Goddard Institute for Space Studies in New York. Hoffman suggested that the ice
might grow if the models were run for longer, but Peltier said the ocean reaches
equilibrium very rapidly and the ice front stops advancing.
Given the meagre geological evidence in support of the snowball Earth, and
the limitations of climate models, hell may freeze over before the issue is
settled.