Paleoclimatology ... the study of Earth's climate history

Enter Text


The Earth is about 4.5 billion years old.

From astrophysical theory we know that, when it was young, in the infancy of the solar system, the sun must have been about 30% less luminous than it is now; and as it ages, in the very distant future, our star will eventually become so bright it will consume all its planets. The very remote past is barely known to us - although we know that the original atmosphere must have been different, and that the planet has had episodes of extreme cold hundreds of millions of years ago. But the record of the last 70 million years or so is much better known - in fact remarkably so. The most recent couple of million is more accessible still, and has been unravelled in amazing detail.

The four diagrams below, taken from the excellent little book by Robert Henson The Rough Guide to Climate Change (highly recommended), give you a sort of sketchy overview of this very long history. Each one magnifies a section of the one before, focussing down on the "now" in the centre.

The first one, on the left, shows the whole of Earth’s history on the left of the vertical line, and its future on the right. The brown line represents the Earth’s surface temperature. In this immense span about all you can see is that Earth’s been cooling since its birth (in spite of the brightening sun), and its been cold a couple of times before. Billions of years in the future, the sun will get so hot all its planets will be consumed, then it will die.

Now look at number 2, which magnifies a small bit near the centre.

This second one, on the right, shows steady cooling for the last 10 million years, with an oscillating period of repeated freezing and warming for 2.5-3 million before the present - the Pleistocene epoch of recent ice ages. Evidently, the behaviour of the climate system changed in some fundamental way at the onset of this epoch.

Number 3 focusses on the last 50,000 years - roughly the time aborigines are known to have occupied the Australian continent. It shows the last half of the most recent glacial cycle (the Wisconsinan), the warming that ended it, beginning about 18,000 years ago, and the ‘holocene’ - the 10,000 years of stable warmth in which human civilization developed. You can now see that on a scale that matters to us - hundreds or thousands of years, the climate record is quite irregular - full of ups & downs - but also that the civilized era has been unusually stable.

In the fourth chart, you can see what’s been happening in historical times and on the scale of human life-times. It looks as if a more or less steady cooling trend has been broken by rather sudden warming about a century ago. The two lines on the right of the vertical represent two possible futures - one that would follow if we go right on adding enormous quantities of gases to the atmosphere; and another if we quit doing this.

Let's turn now to some detailed climate history - the last 65 million years, known as the Cenozoic era. How can we possibly know much about the climate so long ago? you might be wondering. The answer lies in some very clever techniques for getting temperature data from mud on the sea floor. For more about this, see below.

This remarkable graph is a landmark in paleoclimatology. James Zachos and his colleagues published it in 2001.

It was made by putting together a suite of records of deep sea temperature proxies derived from analyses of ocean sediment cores taken from all the world's ocean basins.

It shows in amazing detail  how the Earth first warmed, then cooled over this long period.

Many of its details have stimulated lots of further research and many insights into how the climate system works over tens of millions of years.

You can read the original paper here:

In case you are wondering how the temperature record was made, and what the various ups & downs are all about, this short movie explains a few things about the the Zachos graph.

The world in a grain of sand

If you haven’t got a clue what the scale label on the right edge means, here’s a short explanation. 99.8% of all the oxygen atoms on Earth have 8 neutrons in their nucleus; two in a thousand have 10. This stuff is called oxygen 18, the heavy isotope, or 18O for short. Because its a bit heavier than the other atoms, it behaves a bit differently, and it can be measured in any sample of anything at all that contains oxygen in a gadget called a mass spectrometer. Now consider sea water. Water is a compound of oxygen & hydrogen. Foraminifera living in the sea use the oxygen in water to make their calcium carbonate shells - but the chemical reaction prefers light oxygen, so shells contain less 18O than sea water. Here’s the nice part: this effect depends precisely on the temperature. So by measuring the ratio of 18O & 16O in these little shells the size of a grain of sand you can work out how hot it was when the shell was made - and the shells can last for tens of millions of years under the sea floor.

But there’s more. The two isotopes also do different things at the sea surface, where water molecules containing the heavier ones are less likely than the lighter to jump out of the water into the air (evaporate). If water vapour falls back into the sea either in rain over the ocean, or in river discharge, in the long run it doesn’t make any difference - but if it travels all the way to the poles and precipitates as ice, the proportion of heavy isotope in the ocean eventually rises. So the forams can tell us how much ice was in the world when they lived, as well.

If this all sounds complicated (and almost miraculous), it is - but believe me it’s the backbone of recovering our picture of ancient climates and it’s been thoroughly worked out. The theory was the Nobel prize winning work of two University of Chicago scientists, Harold Urey and Cesare Emiliani in the 1950s; the practical technique was firmly established by Emiliani and Nicholas Shackleton by the 1970s

How do we learn about climate history?

Many very clever techniques are used. None of them is perfect, but their limitations are well understood, and together they provide incredible detail.

          Physical and chemical geology. Ice leaves marks on rocks that can be dated; it also leaves sometimes very characteristic deposits that lend themselves to analysis and dating. Other chemical and physical signatures are due to heating. Fossil plants and animals, their parts and residues can be identified and dated, revealing a pattern of the distribution of living things.

          Radio-chemical dating. This is the key to the magic of revealing the past. An astonishing variety of very ingenious methods are used for measuring tiny amounts of rare isotopes and then figuring out what they mean for past temperatures and many other climate features.

          Ice cores. Deep cores have been drilled in many places over the past 40 years, most notably in Greenland and Antarctica. From Greenland we learned in extraordinary detail just how irregular the last ice age (back to 120,000 years) was; from Antarctica we’ve got an incredible record of all the ice age cycles for 800,000 years. These cores tell not only about temperature, but atmospheric composition, global sea level, windiness and dryness, the distribution of plants & more.

          Ocean and fresh water sediments. In the oceans, tiny shelled organisms (foraminifera) enter the sediment deposited on the sea floor. In places, they remain undisturbed for very long times, and when recovered in cores, can be used to infer a great deal about conditions in ocean and land at the time they lived. In bodies of fresh water, other organisms, especially pollen can be recovered after long periods and analysed.

          Tree rings. Under the right conditions, the pattern of tree rings, both living and dead, can be used to create a climate record of several millennia. Ancient corals and stalagmites, which also exhibit layered growth, can be analysed using different techniques.

This incomplete list might give you an idea just how searching the study of ancient climate has become. It’s important to appreciate that these are not brand new technologies, but by now well tried; their inherent limitations (as to error & uncertainty) are pretty well known. As you’d expect, precise knowledge drops off fairly sharply & becomes more speculative the further one looks back in time - but we nonetheless know an amazing amount & can speak with confidence about many aspects of Earth’s climate history. This is really the only way we can say much about the future - which has become pretty important.

How careful study of 18O abundance can tell us about the amount of ice that was on the planet at some remote time in the past.

An example of an ice core study: the first detailed analysis of the long Vostok core from central Antarctica, published in 1999.

The core yields excellent records of Antarctic air temperature, as well as the atmospheric composition, for four whole glacial cycles (ice ages).

The close correlation between CO2, methane and temperature, and the very fast rise in the gases in the last century are clearly visible (the time scale for this short period has been greatly magnified so you can see it).

A well known example of a temperature proxy study covering the last 1,800 years: Michael Mann's 2008 record, using methods suitable for this interval - tree rings, coral, stalagmites.

The spread of the individual components of Mann's ensemble are shown, with error estimates. The red line at the right is the instrumental record of the last 120 years.


You get some idea of the colossal impact of the ice-ages from these maps, showing the extent of the great continental ice-sheets only 18,000 years ago.

So much water was frozen on land that the sea was 120m lower, and the all the continental shelves were dry land.

4-5℃ temperature change is all it took to do this.

This is the guy who did most to figure it out - Milutin Milankovic, who published the results of his laborious calculations in 1920. It took another 50 years or so before most of the details were worked out, and even now, there are things we don't understand. But we understand it well enough to be very confident about the following story:

Essentially, ice-ages are caused by cold summers in the northern hemisphere. In the south, the Antarctic circle is almost fully occupied by a big continent, so when cold summers occur there (which they do) there's nowhere for the Antarctic ice sheet to expand onto. But in the north, close to the Arctic circle, there is plenty of land to host an expanding ice sheet ... and expand they do. 20,000 years ago, for example, at its maximum extant, the one on North America (the Laurentide) was bigger than Antarctica and up to three miles thick.

The puzzle is even greater now we know exactly what the cycles were like in amazing detail. Here is the record of temperature & CO2 from the deepest Antarctic ice core, the EPICA Dome C, with the cycles & sub-cycles numbered above, and the terminations named below.

Milankovic, of course, had no idea of this, and yet, his idea turned out to be basically correct.

But what could make summers cold, over and over for tens of thousands, or even millions of years?

The answer is something surprising. It isn't any terrestrial factor at all. The ultimate cause of ice-ages lies elsewhere in the solar system. For reasons no one understands, Earth's axis of rotation tilts 23.5° away from perpendicular to the orbital plane. That's the cause of the seasons. But this angle changes slowly, from a minimum of 22.1° to a maximum of 24.5°, and back again, over 41,000 years. Clearly, the more it leans, the greater the contrast between summer & winter on Earth, so this has consequences.

Second, the axis also wobbles. It describes a circle at the poles, turning right round every 21,000 years. this has the effect of rotating the dates of the seasons slowly through the calendar at about 25 minutes a year. Right now, we have northern mid-summer on June 21st; southern on December 21st; but in 10,500 years, these will be reversed. The significance of this depends on a third regular celestial cycle: the eccentricity of Earth's orbit.

The orbit slowly changes from elliptical to almost circular, taking about 105,000 years to complete a cycle. The cycles are not quite regular (specially the second and third) because they are affected by the gravity of the other planets and the Moon - but this is now pretty well known. Today, the maximum Sun-Earth distance is 152,000,000 km (reached on July 3rd), and the minimum 147,000,000 km (January 3rd). This means sunshine in southern summers is 7% stronger than northern ones. But when the orbit is most elliptical, this difference is 23%.

Now for the interesting bit. Imagine a circumstance in which three things hold: the axial tilt is near minimum; the axial wobble has northern mid-summer at the same time as Earth is farthest from the Sun in its orbit; and the orbit is most elliptical. This makes northern summers cold. Snow on Baffin Island and Labrador stays all summer, and the snowfield expands year after year.

The expanding white reflects a lot more sun (increased albedo). That makes the air colder. the nearby Arctic ocean surface gets colder, and dissolves more CO2, reducing the greenhouse effect, making it colder still. After a century or two, the falling CO2 and the albedo effect and some other feedbacks drive the onset of the ice-age by themselves. It gets colder for 80,000 years, then warms in 10,000 or so, then after 5-10,000, the whole thing happens again - not exactly the same each time, because the intersection of three cycles isn't quite the same.

The question of why Earth became susceptible to these cycles of freezing & thawing when it did ... look again at the Zachos graph and you can see the present era of ice-ages began about 5.2 million years ago - the beginning of the Pliocene ... this question has no clear answer now. All we can say with reasonable confidence is that some set of geophysical conditions (including the presence of polar ice itself) must be responsible. For instance, the first glaciation in the north seems to have coincided with the closure of the Panama seaway connecting the Atlantic & Pacific, when the isthmus separated the two great oceans. We know this had a big effect on the transport of heat around the oceans, and various suggestions are being examined to see if they can explain what actually happened.

On the next page you can examine a couple of instances where important questions have been answered by studying climate history.

The best known and most recent big episodes of natural climate change are the recurring glaciations we call ice-ages.

They have fascinated researchers ever since they were deduced from geological evidence 150 years ago. What can have caused such enormous climate disturbances, over and over? ... and why did they start at some point in Earth's history? ... and are they finished, or should we expect another one?

In view of our present problem, the really useful thing about studying climate history is what it can tell you about the responses of the climate system to forced change.

Scientists use this insight to try to answer questions about our future.

If you find the subject of paleoclimatology interesting, there's a good introduction at the NASA Earth Observatory site here




Studying how the climate on Earth has changed in the past has become our best way to understand what's happening to it now.

The study is not very old, but it is the source of some of the most interesting and important discoveries in science today.