Open Mind

What caused the ice pack to recede during the ice age?

The present ice age began 40 million years ago with the growth of an ice sheet in Antarctica, but intensified around 3 million years ago with the spread of ice sheets in the Northern Hemisphere. The entire period is an ice age, but within it, times with lots of ice are called glacial periods, those with only a little (like today) are called interglacials.

From about 3 million years ago to about 800 thousand years ago, the ice sheets advanced and retreated in a roughly 41,000-year cycle. But around and about 800,000 years ago (a time referred to as the mid-Pleistocene transition), the pattern changed. Glaciations intensified, with ice advancing much further, and the cyclic nature changed from a reasonably consistent, roughly 41,000-year cycle, to a much more irregular, roughly 100,000-year cycle. This is visible in this graph, which shows isotope ratios of deep-sea sediment core microfossils, which is a reasonably good indicator of global ice volume. “Up” corresponds to less ice (generally warmer teperatures), “down” to more ice (data from Lisiecki, L.E., & Raymo, M.E. 2005, A Pliocene-Pleistocene stack of 57 globally distributed benthic d18O records, Paleoceanography, vol. 20, pg. 1003):

The roughly 41,000-year cyclic nature of the earlier time period, and the roughly 100,000-year cyclic nature of the last 800,000 years, are quite evident. Detailed analysis reveals that the data also show another cycle (weaker than the other two but undeniable) of about 23,000 years. This particular data are just the latest version of what became clear about 40 years ago, that regular cycles are clearly present in glaciations/deglaciations, with periods of about 100,000 years, 41,000 years, and 23,000 years. What could cause such cyclic behavior?

Milankovitch Cycles

The answer had already been suggested, even before the cyclic nature of glaciations was observed. Nearly 100 years ago a Serbian mathematician, Milutin Milankovitch, suggested that ice advances and retreats would be driven by changes in the energy we receive from the sun. These changes aren’t due to changes in the sun itself, but to the changing position and orientation of earth — changes in earth’s orbit, and in the tilt of earth’s axis.

Obliquity

Most of us are aware that it’s the tilt of earth axis, relative to its orbit, that causes the seasons. When, in the course of its yearly trip around the sun, the northern hemisphere is tilted toward the sun, we in the north get the bulk of the incoming solar energy, while the southern hemisphere gets the left-overs. Hence we have summer in the north and winter in the south. Six months later, earth is on the other side of its orbit, so the northern hemisphere is tilted away from the sun while the southern is tilted toward. That’s when we have winter in the north and summer in the south. The tilt angle of earth’s axis is called its obliquity.

What most people don’t know is that the tilt of earth’s axis is not constant. Right now it’s about 23.5^o, but it varies between about 22^o and 24.5^o. When the tilt is greater, more sunshine falls near the polar regions during summer and less during winter, so the poles have more extreme seasons. More heat in summer means more ice-melt, while more cold in winter means less humidity and therefore less snowfall. Both factors tend to reduce the size of ice sheets — melting them away during summer, starving them of fresh snowfall during winter. Also, with greater tilt the polar regions receive a greater share of total solar energy when averaged over the whole year. So, greater tilt means more total solar energy at extreme latitudes and more extreme seasons, both of which make it possible to reduce an ice sheet.

Precession

Another factor pointed out by Milankovitch is precession. The direction in which earth’s axis points changes; right now our axis points toward the star Polaris (our “north star”), but the axis actually moves very slowly; several thousand years ago, when the pyramids were built, the axis pointed roughly at the star Thuban, which at that time was the “north star.”

Now consider the fact that earth’s orbit is not a perfect circle; it’s an ellipse. So, at some points on our orbit we’re closer to the sun than average, at other points we’re further than average. When we’re closer to the sun, we’re closer to the “heater” so we get more solar energy. The amount of difference between closest approach to the sun (perihelion) and furthest (aphelion) is measured by a quantity called the eccentricity.

Suppose that we get to closest approach when the north pole is pointed toward the sun: the peak of summer in the northern hemisphere. Then during summer, the north receives more solar energy simply because we’re closer to the sun than average. During northern hemisphere winter, the north receives less solar energy because six months later, we’ll be further from the sun. Again, the result is more extreme seasons, allowing for more ice-melt in summer and less fresh snowfall in winter — a recipe for reducing the size of ice sheets.

One of the big differences between obliquity and precession changes is that obliquity (tilt) affects both hemispheres in the same way at the same time, while precession affects the hemispheres oppositely. But the dominant factor in ice mass is the northern hemisphere; there’s so little land in the southern hemisphere that it’s much harder for large ice sheets to form.

Eccentricity

We mentioned that precession affects the seasonal distribution of solar energy because earth’s orbit is not perfectly circular; if it were, then precession would have no effect. It turns out that the eccentricity (non-circularity) of earth’s orbit is also variable. When eccentricity is high, precession has a much stronger affect; when it’s low (when our orbit is nearly circular) it has almost no effect. So, eccentricity also modulates the distribution of incoming solar energy. In addition, eccentricity changes the total solar energy received by the entire earth throughout the year — but only slightly.

All these astronomical factors — eccentricity, obliquity, and precession — that can affect the growth and decay of ice sheets, also show cyclic changes. Eccentricity varies mostly on a 100,000-year cycle, obliquity on a 41,000-year cycle, and the precession cycle is about 23,000 years. Do those numbers sound familiar? Yes! They’re precisely the cycles we observe in the growth and decay of ice sheets.

Not the Whole Story

But Milankovitch cycles are not the whole story in the advance and retreat of ice sheets. The fact is, that the changes in earth’s energy budget due to Milankovitch cycles are not strong enough to account for the change in global average temperature between a glacial period and an interglacial — generally about 5 or 6^oC. To have this effect, the slight influence of Milankovitch cycles must be amplified by feedback mechanisms

One of the important (and rather obvious!) is ice-albedo feedback. Albedo is the reflectivity of a surface. Ice and snow are highly reflective, sending about 90% of incoming solar energy bouncing right back to space, so it never enters the climate system at all. When ice sheets retreat, however, ice and snow are replaced by open land or sea, which are not very reflective; only about 20% of the incoming solar energy is reflected, the rest is absorbed and hence becomes available to the climate system. So, more warming implies less ice/snow implies more solar energy absorbed by earth implies more warming: ice albedo feedback.

Another important factor is greenhouse gas feedback. When ice sheet retreat causes slight global warming, the oceans warm along with everything else. There’s a lot more carbon dioxide dissolved in the oceans than in the atmosphere, but the amount depends on temperature; warmer water holds less CO₂ while colder water holds more. So, slight warming implies less CO₂ solubility in the oceans implies more greenhouse gases in the atmosphere implies even more warming: greenhouse gas feedback.

That this is real is evident from studies of past CO₂ concentrations measured in air bubbles trapped in ice cores. The most famous such measurements are from the Vostok ice core (this graph shows the last 350,000 years, the blue line indicates temperature, the green line atmospheric CO₂):

The combined effect of ice-albedo feedback and greenhouse-gas feedback are enough to amplify the Milankovitch cycles, causing regular episodes of global warming and cooling throughout the ice ages.

During the last several million years (at least!) CO₂ concentration has, as a result, fluctuated between about 180 ppmv (parts per million by volume) and 280 ppmv. The red line in the above graph shows what CO₂ concentration has done recently, because of human burning of fossil fuels. The present concentration is about 380 ppmv, some 30% higher than we’ve experienced in the last 700,000 years at least, and probably a lot longer.

Greenhouse gases (like CO₂) certainly do absorb infrared radiation, interfering with the earth’s natural cooling mechanism. Less efficient cooling means temperature will rise: global warming. Just as certainly, the levels of CO₂ and other greenhouse gases are much higher than earth has seen in a very long time. Just as certainly, the reason is: us.

The current warming is not just the end of another glacial period; we’re already in an interglacial. It’s also worth noting that the deglaciation which takes us from a glacial period to an interglacial is fast — on geologic timescales — but slow as molasses compared to modern global warming. The total temperature change of 5 or 6^oC from glacial to interglacial typically takes 5000 years or more; the rate of warming is generally less than 0.1^oC/century. That’s what happens during a deglaciation, and it’s rapid enough to be a major stress on ecosystems. The rate of global warming right now is 1.8^oC/century — nearly twenty times faster! The stress on ecosystems is already being observed. It’s only going to get worse.