EVER AGAIN - Snowball Earth - page 105

The chill margin takes on the local magnetic field the moment it cools. If it then shatters and the
shards point in random directions, their magnetic arrows should be just as random. But if the magnetic
field of these rocks has been overwritten after they formed, cooled and shattered, the fields in pillows,
shards and everything else will all point neatly in one overall direction. Joe knew that if he measured
the fields of the chill fragments, he would be able to tell if the magnetic birth certificate was original.
Here at the Ongeluk, we can still see the small cylindrical holes where Joe took his samples. Joe
calls them “palaeomaggot holes”, but he made them himself with a rock drill almost a decade ago. He
eventually analyzed the results years later, at the prompting of a particularly insistent graduate student.
And he then discovered that the volcanics, and the ice rocks that bracketed them, contained a field that
was almost flat. What's more, the shards of shattered chill margin had fields that pointed in every dir-
ection. This meant the measurement was genuine, and that 2.4 billion years ago, ice lay within a few
degrees of the equator. These ice rocks, in other words, are just like Paul's. They are the remnants of
another, earlier Snowball. 1
S O NOW we know that Snowballs have happened twice. At least one occurred a little over 2 billion
years ago, and then a series of perhaps four engulfed the Earth between 750 and 590 million years ago.
There have apparently been no others. What, then, did these Snowball periods have in common, and
what made them different from every other time period in Earth's long history? Was there anything
unusual about them that could have triggered the ice onslaught?
Perhaps. There are intriguing magnetic hints that both of these time periods had a peculiar contin-
ental alignment. As the world's tectonic plates drift over its surface, the continents sometimes bunch
and sometimes scatter. When they spread out, they can end up anywhere. But on a few rare occa-
sions, they can find themselves in a band around the Earth's equator. And this might be exactly what
happened during the Snowball periods.
Though magnetic measurements are difficult, and many of the sites have had their magnetic
memories rewritten in the intervening time, decent data exist from about half the continents that were
around during the later Snowballs. And every one of these lay near the equator. So, too, did the half-
dozen sets of ice rocks that have now been measured. For the earlier Snowball, the task is harder and
the measurements are fewer. But still, all of them point to low-latitude continents.
If the continents truly were arranged around the equator during these two Snowball periods, that
could be just what the ice needed. One reason is that the tropics soak up most of the heat that arrives on
Earth from the sun. Because land is more reflective than ocean, putting all available land in the tropics
could reflect more of the incoming sunlight, and help the planet to cool. Joe Kirschvink suggested this
in his short paper back in 1992.
Dan Schrag, the ideas man, has come up with another reason why equatorial continents could be
the key. When continents spread out to the far north and south, he says, they act as an important brake
on overenthusiastic polar ice caps.
Ice naturally wants to spread: white ice reflects sunlight, which causes cooling, which breeds more
ice—and if this were left unchecked, Earth would spend its entire life as a Snowball. Fortunately for
us, high-latitude continents stop that from happening by helping to warm the Earth back up again
whenever polar ice becomes rampant.
Normally, rocks do the opposite. They help prevent the Earth from overheating by soaking up
the greenhouse gases like carbon dioxide that are pumped out by volcanoes. But if polar ice starts to