selectively rained out, the rainfall tends to choose for the heavy isotope.
And so that means that the water vapor that's left in the atmosphere as you go
further and further north where it's colder and colder.
Tends to get lighter and lighter and
lighter in its proportion of oxygen-16 to 18.
Until finally, when it forms an ice sheet, the ice sheet is very light.
So by very light, I mean something like it has 5% less
oxygen-18 relative to 16 of what the ocean has.
Which doesn't sound like a huge change, but actually it is a very large change,
very easy to measure.
And the way geochemists describe this is to do it in per mille instead of percent.
So 5% is the same as 50 per mille.
And the fact that it's lighter, they put a minus sign there.
So the ice sheets have an isotopic signature of minus 50 per
mille relative to the ocean, which starts out at zero per mille.
And then when you take enough water out of the oceans to make this isotropically
light ice sheet, the water that's left behind is a little bit enriched.
So it turns out that the ocean can be about one or
two per mille heavier during a time when there's a lot of ice on the Earth.
[NOISE] So we can measure the oxygen isotopic composition
of the ice just by putting it in a mass spectrometer and
counting the different kinds of atoms.
But we can also figure out what the isotopic composition of the ocean was in
the past from sediment cores by looking at the oxygen in calcium carbonate.
So, here's an equilibrium reaction where we're changing this heavy oxygen for
the two light oxygens in this calcium carbonate here.
And it turns out that where, which side,
which chemical the oxygen isotope, the oxygen-18 wants to be bound to,
depends on the temperature at which the calcium carbonate forms.
So it's actually a bit of a trick, because measuring the proportion of oxygen-18
to 16 in calcium carbonate in sediments today.
There could be a component of signal that comes from the global ice volume
that can tell us about the shrinking and growing of ice sheets through the Ice Age.
But it can also be a function of the local temperature, because when