Earth's climate shifts between short periods of warm and long, long periods of frigid cold.

Based on past pans, there's reason to think that the current warm period might be nearly done.

Is the Ice Age coming back, or will human activity swing us wildly in the opposite direction?

We live in an ice age.

Our geological period is the Quaternary, and is characterized by a massive glaciation-- vast ice sheets stretching from the Arctic all the way down to the Missouri River through Siberia, much of Europe, and spreading out from all major mountain ranges.

OK, sure.

Right now, we're in a brief interglacial phase-- a relatively summery stretch in which the glaciers have retreated.

These interglacial periods are short lived.

The Quaternary Ice Age has lasted 2.5 million years so far.

It's 10,000 to 15,000-year warm patches are separated by glacial periods that last several times as long as.

The current respite is called the Holocene era, and it began around 11,000 years ago.

Temperatures rose, glaciers, and woolly mammoths migrated north, and humans thrived.

This new era of warmth and plenty saw the rise of agriculture, writing, cities, and technology.

All of our recorded, even our remembered history, is of the Holocene.

You might forgive us for imagining that these relatively summery millennia are normal for this planet.

That is not the case.

The current interglacial is already long.

Does this mean that the glaciers are overdue?

Is winter coming?

To answer these questions, we need to understand what triggers the march of the glaciers and why they eventually retreat.

In fact, we know the broad answer to this, even if the details are under debate.

Earth's motion around the sun changes, and with it, the intensity and distribution of sunlight.

It was Serbian scientist Milutin Milankovitch who realized that the gravitational tug of Jupiter and Saturn would lead to three periodic shifts that might explain the enormous climatic swings of the Quaternary period.

These are the Milankovitch cycles.

Let me summarize.

One-- the elongation or the eccentricity of Earth's elliptical orbit shifts from almost completely circular to somewhat more elliptical in 100,000-years cycle.

At the absolute maximum eccentricity, Earth's most distant point from the sun-- the Aphelion-- is about 30% further than the closest point, the Perihelion.

One hemisphere will experience summer at Aphelion and winter at Perihelion and milder seasons all around.

That's the north at the moment.

The Southern Hemisphere is closer to the sun in summer and further in the winter, so more extreme seasons.

However, the difference in sunlight intensity due to this difference in distance from the sun is much less than the simple difference due to the seasons themselves.

So this shouldn't be a huge effect.

Two, the pointing of Earth's axis precesses.

It rotates 360 degrees over approximately 26,000 years.

In addition, the long axis of Earth's elliptical orbit also precesses.

Together, these two effects define where in the orbit the seasons occur.

They combine to produce a 21,000-year cycle called the precession of the equinoxes.

So eventually, the north's mild Perihelion winter will turn into a cold Aphelion winter.

And 3- Earth's tilt changes.

Our spin axis is now tilted at 23 1/2 degrees relative to the axis of our orbit.

This obliquity oscillates between 22.1 and 24.5 degrees over 41,000 years.

High obliquity means more extreme seasons.

But it's low obliquity that ultimately leads to a colder global climate climate.

Because then the highest latitudes, where glaciation begins, never get much sun.

Now, Milankovitch predicted that obliquity would drive climate variations, because it governs the strength of the seasons.

But how can we test this?

Paleoclimatology.

We can reconstruct our planet's climate history by digging holes.

First, glacial ice cores.

The most famous is the nearly four-kilometer-deep hole drilled in the Vostok Glacier in Antarctica.

This glacier was built up by millennia of snowfall.

Each year's layer carries bubbles of the Earth's atmosphere from that time.

Isotope ratios and greenhouse gas content in those bubbles traces global climate over the past 420,000 years.

Second-- oceanic sediment cores reveal the changes in ocean floor sea life, whose composition also depends sensitively on ocean temperatures and salinity, and so also on global climate and ice volume.

Ocean cores get us a climate record back tens of millions of years.

If you look back to the early Quaternary-- earlier than, say, a million years ago-- it seems Milankovitch was right.

Temperature goes up and down on the roughly 40,000-year time scale of changing obliquity.

But then, around 800,000 to 900,000 years ago, something changed.

As Earth reached the depth of the current ice age, the cycle shifted.

Now the warm periods come only once every 100,000 years.

They seem to follow the change in eccentricity, not obliquity.

Every time Earth's orbit becomes more circular, the planet warms and the glaciers go away.

As eccentricity increases again, the glaciers return.

This is totally weird, because eccentricity should produce a much smaller effect than obliquity.

So what changed?

It's not entirely clear.

But it may be that we're now so deep in the ice age that it takes all of the Milankovitch cycles together to cause the glaciers to retreat.

Eccentricity and obliquity and precession must line up perfectly.

The eccentricity cycle is the longest, and so the shifts correspond to its period.

OK.

So we're now in a warm interlude in the depth of an ice age.

You might be wondering, when are the glaciers going to rush down from the north, bringing polar bears, white walkers, Tontons?

One thing is for sure-- the glaciers will come from the north.

The vast oceans of the Southern Hemisphere provide a powerful buffer against changes in temperature.

Ice struggles to build up on water.

But even now, northern winters see ice and snow cover the land all the way down to the continental US, Europe, and China.

In summer, it retreats completely.

But if the climate were a little bit cooler, then summer may not be warm enough to melt all of the winter snow.

Then it would build up year after year, slowly creeping south.

Now, by themselves, shifts in Earth's orbit aren't enough to radically change climate.

But they are enough to trigger positive feedback cycles.

As ice cover increases, Earth starts to reflect more incoming sunlight.

Its albedo increases.

More ice means less absorbed sunlight, lowering global temperature and allowing even more ice to grow.

The glaciation initiated by the Milankovitch cycles accelerates.

A second feedback cycle is equally important.

Cooler oceans are better at absorbing carbon dioxide from the atmosphere, and so the Earth's natural greenhouse effect is diminished.

There is an unfortunate combination of orbital properties that kickstarts this process.

First, low obliquity means less overall sun at high latitudes where the glaciers start.

Second, high eccentricity means one hemisphere experiences a bad winter at Aphelion, further from the sun.

Earth also moves slower at Aphelion, and so those long, cold winters are not counteracted by the short, warmer summers.

And third, the procession of the equinoxes sends the glacier-prone Northern Hemisphere into a bitter Aphelion winter while the eccentricity is high.

So when does this happen next?

Well, right now, obliquity is decreasing, and it will bottom out in around 12,000 years.

It's currently winter at Perihelion in the Northern Hemisphere, but it'll persist completely to the bad situation in 10,000 years.

So over 10,000 to 12,000 years, all of that points to cooling.

What about the 100,00-year eccentricity cycle that seems to define the overall cycle?

Well, actually, we're just coming out of a peak in eccentricity.

That should've been bad.

And perhaps it would have meant that the upcoming cooling trend would bring the glaciers with it.

However, we may have dodged a bullet.

See, the recent eccentricity maximum was a sad little pig, and our orbit remains pretty circular.

See, as well as the 100,000-year cycle, there's a longer 400,000-year cycle on top of that.

Roughly, every fourth eccentricity peak is very low.

That just happened.

And the next peak will be weak, also.

We got lucky.

We're in a long, stable, low-eccentricity phase.

Because of this, climate models predict that we have another 25,000 to 50,000 years of interglacial period left And that's only if you ignore anthropogenic climate change.

Human influence on the climate messes with the whole equation.

With CO2 now at 400 parts per million, it's higher than at any point in the Quaternary period.

It's been predicted that this may extent the current interglacial for 100,000 years.

So we've probably at least offset the next glaciation, although it wasn't coming any time soon, anyway.

The real question is have we ended the entire Quaternary ice age?

Also possible.

However, the recent increase in greenhouse gases is so large and so sudden that there's no precedent anywhere in the climate record.

This makes modeling our influence a huge challenge.

But don't mistake that for a lack of certainty.

Our influence is certainly enormous.

There is another climate extreme that's much less fun than a long, mild interglacial.

That's a sweltering greenhouse climate, like the one that dominated the Mesozoic when the dinosaurs roamed, or Venus.

See you next week for more cold, hard facts on Space Time.