The ice age that never was

03 September 2008
From New Scientist Print Edition.
Hazel Muir

Ice age enigma

THE romantic notion that early humans lived in harmony with their environment has taken quite a
battering lately. Modern humans may have started eliminating other species right from the start: our
ancestors stand accused of wiping out megafauna - from giant flightless birds in Australia to
mammoths in Asia and the ground sloth of North America - as they spread across the planet.
Even so, by around 6000 years ago there were only about 12 million people on Earth - less than a
quarter of the current population of Great Britain. That's a far cry from today's 6.6 billion, many of us
guzzling fossil fuels, churning out greenhouse gases and messing with our planet's climate like there's
no tomorrow. So it may seem far-fetched to suggest that humans have been causing global warming
ever since our ancestors started burning and cutting forests to make way for fields at least 7000 years
ago.
Yet that's the view of retired climate scientist William Ruddiman, formerly of the University of Virginia,
Charlottesville. Ancient farmers were pumping climate-warming carbon dioxide and methane into the
atmosphere long before recorded history began, he says. Far from causing catastrophe, however,
early farmers halted the planet's descent into another ice age and kept Earth warm and stable for
thousands of years.
Hugely controversial
Could a few primitive farmers really have changed the climate of the entire globe? If you find this hard
to believe, you're not the only one. Ruddiman's idea has been hugely controversial ever since he
proposed it in 2003. "Most new ideas, especially controversial ones, die out pretty fast. It doesn't take
science long to weed them out," he says. Yet five years on, his idea is still not dead. On the contrary,
he says the latest evidence strengthens his case. "It has become clear that natural explanations for the
rise in greenhouse gases over the past few thousand years are the ones that are not measuring up,
and we can reject them," he claims.
There is no doubt that the soaring levels of carbon dioxide and other greenhouse gases we see in the
atmosphere today - causing a 0.7 C rise in average global temperature during the 20th century - are
the result of human activities. In the late 1990s, however, Ruddiman started to suspect that our
contribution to the global greenhouse began to become significant long before the industrial age
began. This was when an ice core drilled at the Vostok station in Antarctica revealed how atmospheric
CO2 and methane levels have changed over the past 400,000 years. Bubbles trapped in the ice
provide a record of the ancient atmosphere during the past three interglacials.
What we see is a regular pattern of rises and falls with a period of about 100,000 years, coinciding with
the coming and going of ice ages. There is a good explanation for these cycles: periodic changes in

the planet's orbit and axis of rotation alter the amount of sunlight reaching the Earth (see "The ice
ages"). We are now in one of the relatively brief, warm interglacial periods that follow an ice age.
Regular patterns
Within this larger pattern there are regular peaks in methane every 22,000 years that coincide with the
times when the Earth's orbit makes summers in the northern hemisphere warmest. This makes sense,
because warm northern summers drive strong tropical monsoons in southern Asia that both encourage
the growth of vegetation and cause flooding during which vegetation rotting in oxygen-poor water will
emit methane. Around the Arctic, hot summers thaw wetlands for longer, again promoting both
vegetation growth and methane emission.
In recent times, however, this regular pattern has changed. The last methane peak occurred around
11,000 years ago, at about 700 parts per billion, after which levels began to fall. But instead of
continuing to fall to what Ruddiman says should have been a minimum of about 450 ppb today, the
atmospheric methane began to climb again 5000 years ago. "It just went the wrong way," Ruddiman
says.
Intrigued, he turned to records for CO2. No one understands exactly why the level of atmospheric CO 2
varies naturally - a complex combination of factors is involved, from the growth of vegetation to
volcanic activity. The Vostok core, however, shows that CO2 levels correlate closely with temperature,
peaking at 280 to 300 parts per million as an ice age ends, then falling to lower values. (Today the
level is 384 ppm and rising ever faster.)
This pattern was repeated until about 7000 years ago, when CO 2 bucked a downward trend and
started to rise again. Just before the industrial era began it was already around 40 ppm higher than
would be expected based on the last three previous interglacial periods.
Canadian ice sheet
The effect of the rising levels of these greenhouse gases on Earth's climate would have been
significant. The extra 250 ppb of methane and the extra 40 ppm of CO 2 would have kept the average
global temperature nearly 0.7 C warmer than it would otherwise have been, equalling the warming
during the 20th century due to industrial emissions.
Working with climate modellers Stephen Vavrus and John Kutzbach of the University of WisconsinMadison, Ruddiman has shown that if the levels of these gases had continued to fall rather than rising
when they did, ice sheets would now cover swathes of northern Canada and Siberia. The world would
be heading into another ice age.
So why did both methane and CO2 rise over the past few thousand years? In other words, why has this
interglacial been different from previous ones. Could humans be to blame?
Forest clearing
Agriculture emerged around the eastern Mediterranean some 11,000 years ago, then shortly
afterwards in China, and several thousand years later in the Americas. Farming can release
greenhouse gases in various ways: clearing forests liberates lots of stored carbon as the wood rots or
is burned, for instance, while flooded rice paddies release methane just as wetlands do.
To find out more about early farming, Ruddiman began to dig around in studies of agricultural history.
These revealed that there was a sharp rise in rice cultivation in Asia around 5000 years ago, with the
practice spreading across China and south-east Asia. Here at least was a possible source for the
unexpected methane rise.

In Europe, people had begun clearing forests to grow cereal crops such as barley and wheat 7500
years ago (see graphic). There is no firm figure for the total extent of this forest clearance, but
Ruddiman says it could have been vast. One pointer to this is the Domesday Book, which documents
the 11th-century census of England ordered by William the Conqueror. "There were 1.5 million people
and they had cleared 85 per cent of the forest," Ruddiman says.
When he published his theory in 2003 (Climatic Change, vol 61, p 261), there was no shortage of
criticism. The most devastating came from Fortunat Joos at the University of Bern, Switzerland, who
pointed out that there was simply not enough forest on Earth to account for the anomalous rise in CO 2.
Cutting down every tree on the planet wouldn't explain the CO 2 rise seen in ice cores over the past
7000 years.
Amplified
Ruddiman has conceded that Joos was right, and now argues that deforestation and rice growing
released enough CO2 and methane to cause a little warming, which then triggered feedback
mechanisms that released more CO2, causing further warming and so on, amplifying the influence of
early farmers.
That such feedback mechanisms exist is not in doubt. Climate scientists all agree that the magnitude
of the temperature changes as the Earth has gone in and out of ice ages cannot be explained by
orbital changes alone. The issue is whether early farmers could have released enough CO 2 and
methane to trigger feedback effects large enough to explain the total observed rise in CO 2.
With the help of Vavrus and Kutzbach, Ruddiman is using climate models to explore whether this is
plausible. The models suggest that the early human influence might have been great enough to keep
the oceans warm compared with previous interglacials. This could have boosted CO 2 levels in two
ways. "Carbon dioxide is less soluble if the oceans are warmer," says Ruddiman. "It's a bit like a soda
drink bubbling off its gases on a warm summer's day."
Warmer oceans
Warmer southern oceans would also mean less sea ice around Antarctica, increasing CO 2 levels by
boosting gas exchange between the atmosphere and seawater. "Both the observational data and our
model suggest these two factors are players," Ruddiman says. "Whether they fill the entire gap in the
CO2 budget we can't say, but they look very promising."
The feedback idea also overcomes another objection to the "early anthropogenic hypothesis". During
photosynthesis, plants take up a slightly larger proportion of the carbon-12 in the atmosphere than of
the heavier isotope, carbon-13. This means that if all the CO 2 came from deforestation - which
reverses the process - there should have been a gradual rise in the proportion of carbon-12 in the
atmosphere relative to carbon-13. Yet the ice cores show no such thing, a fact pointed out in a 2006
article by renowned climate scientist Wally Broecker of Columbia University's Lamont-Doherty Earth
Observatory in Palisades, New York.
If a feedback mechanism was at work, however, 75 per cent of the extra CO 2 would have come from
the oceans, and the relatively small rise in the proportion of carbon-12 would be consistent with the
only isotope record published so far, Ruddiman says.
What's more, since Ruddiman first suggested his hypothesis, ice core records of atmospheric methane
and CO2 levels have been extended back to 800,000 years ago, covering another five interglacial
periods, three similar to the present one. So we now have data for six comparable interglacials, all of
which suggest the rise in the gases over the past few thousand years is unprecedented, he claims.
"Natural explanations for the rise have been tested six times and they failed six times - 12 times if you
allow for the two gases."

Insane argument
Compelling? Emphatically not, says Broecker, who is Ruddiman's most fervent critic. "It's an insane
argument," he says. "It's total and utter nonsense."
"If Ruddiman wanted to argue that over the last thousand years there had been an effect, that would
be harder to argue against because maybe there was," adds Broecker. "But if you look at that record
during the last thousand years, CO2 sort of flattens out. Just at the time you'd think that man's effect
would be the strongest, you see the least effect."
Ruddiman argues that this is because deforestation had levelled out. "There's good evidence that most
of Eurasia was deforested by the time of Christ," he says.
Broecker reckons there's a natural explanation for the CO 2 rise. The deep oceans store lots of carbon
as solid calcium carbonate, and deep-sea sediments record a drop in carbonate concentrations that
could account for the rise in atmospheric CO2.
Previous interglacials
But if the CO2 rise before the industrial age is natural, why did it not happen in previous interglacial
periods? It did, Broeker claims, during the interglacial period 400,000 years ago (see graphic). He
compares this interglacial with the present one by aligning them from the point at which the previous
ice age ended, which shows CO2 naturally staying high, above 270 ppm, for about 28,000 years.
Ruddiman aligns the interglacials on the basis of the point in the orbital cycle at which incoming solar
radiation is at a minimum, signalling the end of the interglacial. This paints a very different picture.
Other climate experts say it's not possible to match the two cycles exactly because the three orbital
parameters (see graphic) were different, so any choice is somewhat arbitrary.
Another of Broecker's complaints is that Ruddiman has handed global-warming deniers some cheap
ammunition: if ancient farmers headed off a looming ice age, isn't that a good thing? "It confuses the
situation and we don't need that," says Broecker. "People use it in some curious ways to speak against
the effects of global warming," he adds. "I get very emotional about it, yes, because I think it's very bad
science."
A lot of progress
Some climate-change deniers did pounce on Ruddiman's hypothesis when it was first publicised, but
largely dropped it when it was pointed out that they were effectively conceding that humans really do
cause climate change. Many researchers, by contrast, remain intrigued by Ruddiman's idea. "We know
that humans have had a lot of influence," says Richard Alley, a geologist at Penn State University in
University Park, Pennsylvania. "Bill's idea is that you see it in the atmosphere. He may be right."
Alley is impressed by a paper that Ruddiman and his colleagues published this year in Quaternary
Science Reviews (vol 27, p 1291) which shows that the number of new rice-growing sites appearing in
China from 6000 to 4000 years ago was 10 times that in previous millennia. "I would not say
Ruddiman wins, it's over, let's go home - but I think that he's making a lot of progress with the
methane."
Geochemist Ed Brook at Oregon State University in Corvallis isn't so sure. He says the methane rise
over the past 5000 years could have arisen from wetlands in the southern hemisphere. "There are
some nice records from caves in South America that show that region was getting wetter."
Circumstantial

On CO2, things are even more uncertain. "It's a tough one, because the evidence is primarily
circumstantial," Brook says. "I'm just as willing to think that there are some natural mechanisms that
we don't understand." Subtle changes in the Earth's orbit might drive feedback that makes the
atmospheric chemistry very different from one interglacial period to the next.
Despite the doubts, Brook welcomes Ruddiman's research. "He has questioned the conventional
wisdom in a way that I think is productive," he says. "It makes us really stop and think."
Future results should help resolve matters. Measurements of carbon and hydrogen isotope ratios in
methane from ice cores could provide subtle clues to the methane's origin, while comparisons of
methane levels in cores from Antarctica and Greenland might reveal where the extra methane came
from geographically. The existing measurements are not conclusive but suggest the extra methane
came from low latitudes, which is consistent both with Ruddiman's hypothesis and with natural
methane from tropical South America.
Ruddiman, meanwhile, plans to continue working on climate models and researching the history of
human land use. "If we don't understand something as basic as why greenhouse gases have
increased in the past several thousand years, then there's a huge gap in our knowledge of the
climate," he says.
For now, it is still an open question: did ancient farmers kick off global warming or not? The answer is
in a sense academic, because it has no bearing on the far more rapid warming caused by
industrialised societies. If it's true, though, it could be seen as a warning. If a few million people
wielding stone axes averted an ice age, just imagine what the legacy of their 7 billion oil-guzzling
descendants will be.
A chilling tale
Did ancient plagues and pestilence that killed millions of people also alter the planet's climate?
Slight dips in the level of carbon dioxide in the atmosphere over the past 2000 years are usually put
down to natural causes, such as lower emissions from volcanoes. But when human populations were
decimated by disease, large areas of farmland would have been abandoned to nature. As forests
reclaimed the land, huge quantities of CO2 would have been sucked out of the atmosphere.
Ruddiman has shown that the timing of one of the larger dips in CO 2 matches a series of plagues that
peaked around AD 540. Another coincides with the "black death" of the 14th century. In both cases
Europe's population may have fallen by a third or more. Worse still was the effect of European settlers
bringing smallpox and other diseases to the Americas, causing populations to fall by as much as 90
per cent. This coincides with a relatively cool period known as the "little ice age".
Historians have suggested that the little ice age caused famine, disease and depopulation - but was
disease the trigger for the little ice age, rather than the upshot?
The ice ages
The coming and going of ice ages is driven by periodic changes in Earth's orbit and rotation axis,
known as Milankovich cycles, caused by the pull of other bodies in the solar system.
The tilt of the Earth's axis varies in a cycle lasting about 41,000 years. In the more "upright" position,
both poles have colder winters, making the Earth more prone to glaciations.
Another periodic change is precession, in which the pointing direction of the Earth's axis rotates, going
full circle in 22,000 years. Because the Earth's orbit is slightly elliptical, or eccentric, this alters the

season in which the Earth is farthest from the sun. The eccentricity of Earth's orbit also varies every
100,000 years, though it's not clear how this is linked to changes in climate.
The rapid swings between ice ages and interglacials began in earnest only around 2.5 million years
ago - before this the Earth had been much warmer. A long-term cooling trend in the past 55 million
years could be due to the rise of the Tibetan plateau increasing chemical weathering of rocks, which
removes CO2 from the atmosphere.
Related Articles

Huge sea level rises are coming - unless we act now

http://environment.newscientist.com/article/mg19526141.600
25 July 2007
Climate change: A guide for the perplexed
http://environment.newscientist.com/article/dn11462
16 May 2007
Early farmers warmed Earth's climate
http://environment.newscientist.com/article/dn4464
11 December 2003
Did Tibet cool the world?
http://environment.newscientist.com/article/mg13918804.400
3 July 1993

Weblinks

William Ruddiman, University of Virginia

http://www.evsc.virginia.edu/faculty/people/ruddiman.shtml
Richard Alley, Penn State University
http://www.geosc.psu.edu/people/faculty/personalpages/ralley
Ed Brook, Oregon State University
http://geo.oregonstate.edu/people/faculty/brooke.htm

From issue 2672 of New Scientist magazine, 03 September 2008, page 32-36