Restoration of a group in late Pleistocene northern Spain, by Mauricio Antón

Why do Ice Ages occur? A New Paradigm Shift on a Prehistoric Problem


Why do ice ages occur?  Surprisingly, even after many decades of paleoclimatic research we simply do not know for sure. Most scientists will agree that ice age cycles have something to do with precession: the slow wobble of the axis of the Earth.

The ancient Egyptians and Greeks knew of precession and called it the Great Year, because it gives warm and cool seasons over its approximate 23,000-year cycle. But there is a problem with invoking the Great Year as the regulator of ice ages, because we should really get an interglacial warming every 23,000 years or so. And we don't—they only happen every fourth or fifth Great Year.

Warming and Cooling

But why should the global climate give a selective response to orbital warming and cooling (called 'forcing' in the climate trade)? This is one of the great unknowns of modern science. Many suggestions have been made, from interstellar dust blocking sunlight to the weight of the ice sheets depressing the lithosphere and warming the ice. And yet all of these theories share one thing in common: they stretch credulity.

The Athabasca Glacier is one of a handful that makes up the Columbia Icefield in Jasper, Alberta, Canada. An Ice Age relic, the ice is 10,000 to 15,000 years old.

The Athabasca Glacier is one of a handful that makes up the Columbia Icefield in Jasper, Alberta, Canada. An Ice Age relic, the ice is 10,000 to 15,000 years old. ( CC BY-NC-SA 2.0 )

The only thing that is certain, is that the science is not settled in this area of ancient climate research.

The Perennial Conundrum of Paleoclimatology

But a new theory by independent researcher Ralph Ellis may have unlocked this perennial conundrum of paleoclimatology. And considering the myriad implications that flow from this theory, its mechanism is nevertheless very simple to understand.

What we need is a selective feedback system that can act either in concert with, or in opposition to, the warming and cooling provided by the ~23,000-year Great Year. At present it is claimed that this feedback agent is CO 2, because CO 2 is a warming agent and its concentrations do indeed rise and fall with the ice ages.

But there is a big problem with this assertion, because when CO 2 concentrations reach a minimum in the depths of the ice age, the world warms. And when CO 2 concentrations reach a maximum during the warm interglacial period, the world cools. And yet this is the very opposite of what should happen if CO 2 was the most powerful warming feedback agent.

Artist’s depiction of Ice Age on Earth

Artist’s depiction of Ice Age on Earth ( CC BY-SA 4.0 )

The Achilles Heel of Albedo

This is where the new theory by Ralph Ellis shifts the paradigm, and the primary character in this new climatic drama is albedo - the reflectivity of the Earth.

Obviously polar ice sheets are highly reflective, they have a high albedo, and reflect up to 90 percent of the incident sunlight during the all-important northern summer. This allows the polar ice sheets to slowly grow, year after year, and slowly reflect more and more incident sunlight in the process. And this ice-albedo reflection mechanism is so strong, it can even resist the next Great Summer when northern sunlight is once again at a maximum, which is why some Great Summers produce no global warming whatsoever.

A graph of sunlight strength at 65ºN (blue) vs Antarctic temperatures (red). Each red peak represents an interglacial warming, which occur roughly every 100,000 years. The blue peaks represent Great Summers in the northern hemisphere. Note that not every Great Summer produces an interglacial global warming.  (Sources:  Laskar 2004 orbital cycles,  Epica3 2007 temperature data.)

But if albedo is so strong that it can shrug off the increased sunlight of a Great Summer, then how does the climate system generate an interglacial warming? The answer is that ice-sheet albedo has a very prominent Achilles heel: dust.

If dust gets on ice sheets their albedo is reduced considerably and they can melt very quickly. And surprising as it may seem this is exactly what happens, because every interglacial warming period is preceded by about 10,000 years of intense dust storms. And we know that this dust settles on polar ice-sheets, because the ice-core data still contains that dust. And in the Greenland ice-sheet, the majority of that dust appears to have originated from the Gobi Desert. But rather than being a shifting-sand desert, most of the Gobi is currently pastoral steppe grasslands, where the herds of the Mongolian nomads graze. So how and why does this large steppe plateau become a true sand desert?

It has been claimed that desertification at the ice age maximum was caused by a cooling and drying of the climate. But there are many strands of evidence that conflict with this argument, and point towards a cooling of just 3ºc in tropical and extra-tropical regions, combined with only a slight reduction in precipitation .

The answer to this problem is CO 2 concentrations, which reduced during the ice ages as the seas cooled and absorbed more atmospheric CO 2. They eventually reached as low as 190 ppm, which is dangerously low for much of the world's plant-life. Most plants suffer severe stress at 190 ppm CO 2 and die at 150 ppm, because CO 2 is a primary plant-food. And the fact that vegetation was suffering from CO 2 starvation during the depths of the ice age was confirmed recently by a study of ancient trees preserved in the La Brea tar pits in southern California.

Great Summer and Great Winter

And now we have the entire ice age forcing and feedback mechanism, laid out and plain for all to see. It begins when a Great Summer turns into a Great Winter, which reduces the sun-strength in the northern hemisphere and allows ice sheets to grow. This is a slow process that takes tens of thousands of years, and appears destined to turn the world into a complete snowball.

Let’s hope this isn’t the case anytime soon. Artist’s depiction of modern Ice Age.

Let’s hope this isn’t the case anytime soon. Artist’s depiction of modern Ice Age. ( CC BY-NC 2.0 )

However, the Achilles heel of the high albedo polar ice sheets come into play— the dust. As the ice sheets grow and the seas cool, CO 2 also reduces. The concentration finally reaches the critical 190 ppm level where world flora begins to die, especially at higher altitude, and the Gobi steppe-lands turn into a true sand desert. This turns northern China into the equivalent of 1930s Dust Bowl America, and the ensuing dust storms dump thousands of tons of dust onto the northern ice sheets each year.

A dust storm approaches Stratford, Texas, in 1935.

A dust storm approaches Stratford, Texas, in 1935. ( Public Domain )

And so when the next Great Summer comes along, the dusty polar ice sheets can warm and melt and the next interglacial is born. So CO 2 can indeed cause global warming, but its effect is much more pronounced at low rather than high concentrations.

A summary graph of all the factors that play a role in glacial modulation.

A summary graph of all the factors that play a role in glacial modulation.

Ice sheet extent (light blue) grows, forcing the temperature to fall (red). And CO 2 concentrations (yellow) fall alongside temperatures, as CO 2 is absorbed by the cooler oceans. When CO 2 concentrations reach a minimum of 190 ppm there is desertification and large dust storms (purple). And when the next Great Summer comes along (blue sine wave), the dusty-ice sheets can melt and the world warms into another interglacial (red peak).   (Sources:  Laskar 2004  orbital cycles,  Epica3 2007 temperature data)

© Ralph Ellis, Sept 2015

Featured image: Restoration of a group in late Pleistocene northern Spain, by Mauricio Antón ( CC BY 2.5 )

By Ralph Ellis


Ralph Ellis, 2016. “Modulation of Ice Ages via Precession and Dust-Albedo Feedbacks”. [Online] Available at:

PMIP  (2015) “Paleoclimate Modeling Intercomparison Project (PMIP)” [Online] Available at:

The average of eleven precipitation models:

The average of eleven temperature models:

Ward, J.  et al   (2004) “Carbon starvation in glacial trees recovered from the La Brea tar pits in southern California.” [Online] Available at:    doi: 10.1073/pnas.0408315102

Laskar, J.  Joutel, F.  Boudin, F.  et al  (2004) “The Institut de mecanique celeste et de calcul des ephemerides.” [Online] Available at:

Epica3 data  (2007, 2008, 2015):

For Epica3 dating methodology please see:

Parrenin, F.  Barnola, J. Beer, J. Blunier, T.  Castellano, E.  Chappellaz, J.  Dreyfus, G.  Fischer, H. Fujita, S., 2007. “The EDC3 chronology for the Epica Dome C ice core”. [Online] Available at:


190 ppm is not the necessary trigger. but the gun when fired, the reverse gear engaged, yet not until climate glass (or delicacy) has gone just beyond overfill or underempty.

I suspect also that methane and ammonia concentrations can rise and fall as well, thus attributing aid in the off balances. Additionally, similarly other Whole Planets, such as Pluto show the same signs and results, just with different gases emissions being the culprit. Mars probably use to have them both, but now just dust storms are left.. Are there enough animals along geological timeline to add affect to legal climate enigma ? I think there may seriously be, a lack of scientists willing to work on all those drilled glacial cores we have. Yet also, it may be beyond suspect that there is nothing man can do to counteract the earths way of balancing. global warming is a hoax; climate change is only the earth, at work :)

Perhaps animal methane can be added catalyst, once methane is formed or released, what counteracts methane ? or is there just simple dissipation ? amonia and other elements too.

Colin Berry's picture

Wiki gives a nett lifetime for methane – atmosphere mainly, a little in the soil – of 8.4 years, which is pretty quick if thinking in terms of planetary glaciation cycles.

Mechanism of removal?  Again, according to wiki (let the user beware) it gets attacked by OH free radicals, the latter deriving from singlet oxygen, being turned first into formaldehyde,  HCHO, and then, after reaction with more OH, into CO2 and water.

The dust theory is quite a promising one, given that it's testable in principle by looking at ancient ice cores. Now if they could just find powdered Gobi desert yak dung ...



>>Now if they could just find powdered Gobi desert yak dung ...

Perhaps they will. ;-). But this does indeed seem to be Gobi dust, as it has been isotopically identified as comming from that region. See the full paper, in the references section.



Register to become part of our active community, get updates, receive a monthly newsletter, and enjoy the benefits and rewards of our member point system OR just post your comment below as a Guest.

Next article