Those are what we study.

This is a piece of ice – about 20,000 years old – from Antarctica.

And bubbles trap air from 20,000 years ago, so we can find out what air was like back then.

(We) Can figure out if carbon dioxide has gone up or down.

And what we've learned from that is carbon dioxide is higher now than it's been for at least the last million years, probably the last 20 million years, but that's less certain.

So it's really quite a dramatic thing that we humans have done to the carbon dioxide.

Hey smart people.

Joe here.

Earth's atmosphere and climate have changed in a big way, and they are continuing to change.

There's no doubt about that, and we've known it for decades.

But Earth's climate has always changed throughout its history.

So how do we know this time is different?

We know because at places like the Scripps Institution of Oceanography in southern California, we have freezers full of ancient ice that let us look into the past, thousands.

Even millions of years, and measure exactly what Earth's atmosphere, and its climate, were like throughout deep history.

I recently stopped by to visit Dr. Jeffrey Severinghaus, who studies ice cores.

He's part of a team working to find the oldest ice on Earth.

Each of these little blocks of frozen water can tell us something about our planet's past, long before we existed – and where it's heading, now that we do.

And inside these tiny bubbles in this ice are old bubbles of air that existed on this planet as old as that ice is.

Yeah.

That's the atmosphere of the planet, trapped in those little bubbles.

What happens in the polar regions is it's too cold to melt.

So when snow falls it doesn't melt, it just piles up and piles up, and eventually turns into ice under its own weight.

But if you think about what snow is like, if you have a snowflake you have air in between the snowflake.

As snow becomes more and more dense, it tends to squeeze out the air between snowflakes, but it turns out it doesn't squeeze out all the air.

As more layers of snow fall and condense, those tiny voids are literally frozen in time, layer upon layer.

And, there are a lot of layers.

Some ice cores have annual layers just like trees do, you know how you can count tree rings?

So some graduate student sits there and counts 50,000 annual layers.

Of course it has to be a graduate student!

What a lot of work.

But to study ancient ice, first you have to find ancient ice.

Where are you doing this research?

Where are you collecting these ice cores?

This is from a place called Taylor Glacier in Antarctica.

Taylor Glacier is a 54 kilometer stretch of ice and rock.

People like Dr. Severinghaus can read it like a book–full of stories about our ancient climate.

Taylor Glacier is special because it's one of the few places on Earth where the ancient ice has risen to the surface.

So, you only have to drill 5-10 meters to get the ice.

Which is much easier than drilling a deep ice core which is 3,000 meters and costs 50 million dollars.

It's basically a cylinder that has little tiny teeth on the bottom.

And when you rotate the barrel it carves out the ice on the edges and leaves behind an ice core in the middle.

Once the core is pulled up, it's packed up and sent off, carrying a slice of history inside it.

It's a slow process, it takes like a month for the ship to get here.

Whether you're standing in the middle of the Amazon rainforest or at the North Pole, you're breathing roughly the same air.

Our atmosphere is pretty much the same everywhere.

Which means that a tiny air bubble from that one spot is enough to paint a picture of what the entire planet's atmosphere looked like so many years ago.

This is the freezer.

We won't be in there long, so don't worry about the cold.

So this is what a typical ice core sample looks like.

Now you'll notice that there's no bubbles.

That's because when you get down below 600-700 meters, the pressure is so high that the air turns into something called a clathrate which is an ice-like substance.

Clathrates are crystals, where instead of bubbles, the molecules are trapped in a cage made by the bonds between frozen water molecules.

There's still gas in there.

There's still gas molecules but they're not in a gas phase.

Man the patterns are so cool, you must randomly see such cool ice phenomena.

It's cold in here!

This cold!

Funny how that works.

Okay, but how do you get the ancient air out of the ice to measure it?

I mean, without contaminating it with… all this air around us?

So this is how we actually extract the ancient air, if you will.

We take a piece of ice and put it in a vacuum flask, and pump out all of the modern air, the air we're breathing right now, using a vacuum line.

This is a vacuum pump here.

So we make a seal, we pump all the modern air, and close this valve, and then you only have an ice cube and a little bit of water vapor, but no air.

Then we melt the ice, and the melting of the ice releases those little air bubbles of ancient air.

So because you already let out the "now air," the only gasses that are coming out are the ones that are trapped inside the ice.

Right, and then once we've done that, we can purify the gas a little bit by freezing the water.

So they pump out all the modern air, melt the ice to let the ancient atmosphere vaporize, re-freeze the water, and pump that ancient atmosphere out so it can be measured.

This is a liquid helium tank, it's cold enough - it's at 4 kelvin, 4 degrees about absolute zero.

It's cold enough that all the air actually condenses and turns into ice - air ice.

Every gas, will freeze.

Every gas except helium.

So then we take it over here.

This is the analysis part of it.

This tube is actually a bottle, a long skinny bottle that's capable of dipping itself into the liquid helium.

You wouldn't want to be getting your own hands too close to 4 kelvin.

No.

The frozen air gets put into this, a mass spectrometer, which basically measures the masses of really tiny things.

We measure the chemical composition of the atmosphere using isotopes: they're like different flavors of atomic elements.

Isotopes, those flavors of elements, have unique masses, and the mixture of them in the air bubbles can tell us all kinds of things about ancient earth.

We use the isotopes of nitrogen to tell ancient temperature at the time the snow was falling.

Ordinary nitrogen has a mass of 14, but the rare isotope nitrogen 15 has a mass of 15.

It turns out that relative proportions of N15 and N14 are sensitive to temperature.

So, whatever the temperature is at a particular time, it's creating different mixes of different flavors of gasses in the atmosphere, like a fingerprint for temperature.

That's right, and that's trapped in air bubbles for posterity.

So the sample here starts out waiting its turn and when its turn comes the valve opens and goes into this little tiny tube, which leads into the mass spectrometer, here.

And it gets accelerated by a 3,000 volt electrical gradient, which makes the ions go really fast.

And then they hit this magnet and they're forced to make a 90-degree right turn.

And in doing so, heavy things like N15 try to go straight, and lighter things like N14 get bent more.

It's like being in a car.

You can't turn as fast in a big heavy car.

So they swing out, and then the detector is seeing what swung out farther.

So, you're getting resolution of things that differ by a single neutron when they're flying through that curve?

That's pretty cool.

The same idea can be used to find out more than just temperature.

Labs all over the world use elements trapped in air, trapped in ice cores, to paint a map from our distant past to today.

Oxygen isotopes can tell us how oceans changed, mineral dust tells us about how the atmosphere moved around, there are chemical clues about early volcanoes.

But maybe most importantly, we can trace changing levels of carbon dioxide.

So the climate has changed before, how do we know that this time it's us.

The way we know, is just like we talked about with nitrogen, the carbon in carbon dioxide also has two flavors.

There's carbon 12, which is ordinary carbon, and then a very rare form of carbon, carbon 13.

So, that's how we know it's human caused.

The atmosphere, as it goes up in CO2 concentration, the carbon 13 of the atmosphere is taking a nosedive.

And that's not what would happen if it was natural CO2.

Because fossil fuel CO2 is very depleted in carbon 13.

This comes from the fact that plants prefer to eat CO2 made of carbon-12.

And when we burn fossil fuels made from those ancient plants, the fraction of carbon-12 in the atmosphere goes up while carbon-13 goes down.

We've only been measuring carbon dioxide in the atmosphere since 1957, but using the data from ice cores, we can trace levels back way farther.

And this is what we see: CO2 was pretty flat for most of the past 1,000 years.

All around 280 ppm.

Now we're going to add in the carbon 13 abundance, this gold line.

And you can see that was also pretty constant for most of the last thousand years.

But then around 1850, right when carbon dioxide concentration started to rise, the carbon 13 abundance started taking a nosedive.

And this kind of unambiguously tells you that humans did it.

That's why I call it the smoking gun of human causation.

There are lots of other ways we know, but this is the simplest.

We're moving into uncharted territory.

The last time something like this shows up in the ice record is around 55 million years ago, when a volcano popped up under an oil field and cooked basically everything.

It sent all the carbon dioxide into the atmosphere.

So, the carbon dioxide shot up, we think it nearly quadrupled, and the climate warmed by 6 degrees.

The most important thing is right away to solve this global warming problem.

We don't have much time left.

We got to put aside all of our political differences.

We can do this, I know we can.

We can. But will we?

I hope so. Stay curious.