Tag Archives: ice age

Three Million Years of Climate Change

Using reconstructions of global temperature based on oxygen isotope ratio analyses of ocean sediment cores and polar glacial ice cores we can look back at the Earth’s climate for about 800,000 years in considerable detail. To go farther back in relatively high temporal detail, we have to rely on the ocean sediment core analyses that provide data back to five million years.

The three previous blog posts looked at the Earth’s climate over the last half million years using these same proxies for global temperature. During that period the Earth’s climate was dominated by five intense glacial periods each lasting about 60,000 to 100,000 years, alternating with much shorter interglacial warm periods lasting about 2,000 to 25,000 years. The Earth is currently in the most recent interglacial period that first reached near modern “normal” temperatures about 12,000 years ago. The glacial-interglacial cycle time was about 80,000 to 120,000 years during this period.

Looking back another half million years to a million years ago, the glacial periods were about the same duration but progressively weaker going backward in time as can be seen in the graph below.

Climate Reconstructions

Global temperature reconstructions for the last million years from oxygen isotope ratio analyses of ocean sediment cores (Bintanja), Antarctic ice cores (EPICA and Vostok), and a Greenland ice core (GISP2). Click on graph to enlarge.

The interglacial warm periods were similar in length to the most recent half million years, but were weaker and did not quite reach the modern “normal” temperature. Consequently, the average global temperature was about the same as for the most recent half million years, a little over 5 degrees Centigrade (C) below our modern “normal” temperature. Thus, over the last million years, the Earth has averaged a little over 5C colder than our current modern “normal” temperature. This estimate is based on adjusting the ocean sediment core reconstruction to match the Antarctic ice core reconstructions from EPICA and Vostok. The average is even lower in the unadjusted ocean sediment reconstruction as will be shown later.

Another difference is that the range in temperature during each cycle was only about 8C during the during the earlier half of the last million years, as compared to about 14C during the most recent half of the last million years.

Looking back all the way to three million years ago, the adjusted ocean sediment core reconstruction shows that global temperature progressively dropped from levels much closer to the modern “normal” around three million years ago to the much colder average of the last million years. Thus, the start of our current ice age was about three million years ago. The cycling between cold glacial periods and warm interglacial periods was present throughout the last three million years but each cycle was only about 40,000 years duration prior to one million years ago compared to 100,000 years duration in the most recent million years as can be seen in the graph below.

Climate Reconstructions

Global temperature reconstructions for the last 3 million years from oxygen isotope ratio analyses of ocean sediment cores (Bintanja), Antarctic ice cores (EPICA and Vostok), and a Greenland ice core (GISP2). Click on graph to enlarge.

Without adjustment, the ocean sediment core reconstruction shows even larger swings in amplitude of both warming and cooling with each cycle as seen in the graph below. The large difference between the adjusted and unadjusted global temperatures is an indication of the uncertainty involved in making these reconstruction estimates.

Climate Reconstruction

Global temperature reconstruction for the last 3 million years based on oxygen isotope ratio analyses of ocean sediment cores (Bintanja). Click on graph to enlarge.

Note that in the unadjusted ocean sediment core reconstruction the global average temperature is estimated to have been over 7C warmer than our modern “normal” for a period of several thousand years as recently as a little over 2.9 million years ago. With temperatures that warm there would have been much less permanent ice than today and sea levels would have been substantially higher.

One of the challenges to understanding the Earth’s climate is to determine what caused this gradual trend into our current ice age. Another is to understand what caused the 40,000 year cold to warm cycle to increase fairly abruptly around a million years ago to about 100,000 years. And yet another challenge is to determine what caused the increase in amplitude of the cycles about half a million years ago.

One of the leading hypotheses as to what started this ice age is that it may have resulted from the connection of North America to South America by the uplifting that created the Isthmus of Panama and blocked ocean currents from passing between the Atlantic and Pacific.

The 40,000 year cycling corresponds well with orbital/rotational mechanics of the Earth that induce changes of solar radiation at the poles. But what caused the shift to 100,000 year cycling and a greater amplification is more difficult to explain. Until our climate models can reproduce these past changes in climate, I have little confidence that they will be able to accurately predict the future climate. In the mean time, extrapolating past climate cycles is probably our best estimate of the future climate changes we can expect, as was attempted in the previous blog post linked below.

Interglacial Comparisons


Interglacial Comparisons

Most people don’t realize that the Earth is still in a long term ice age that started about three million years ago and has had many alternating cold glacial periods interspersed with warmer interglacial periods.  We are currently in an interglacial period where global temperatures have been near our modern “normal” for about 12,000 years now.  In addition to our current interglacial period, there have been four previous interglacial periods in the last 500,000 years.  Each one has been spaced about 100,000 years apart and lasted about 2,000 to 25,000 years with temperatures at or above our current modern “normal”.  These observations are based on the EPICA Antarctic ice core climate reconstruction using oxygen isotope ratios as a proxie for global temperature change.

The graph below shows the current and last four interglacial periods plotted together, normalized to the year where the estimated global temperature first reached the level of our modern “normal” climate. The approximate year where each interglacial episode first reached the modern “normal” temperature is shown in the legend.  Notice that all four previous interglacials had global temperatures reaching 2 to 4 degrees Centigrade higher than our current modern “normal” without any help from humans, based on this reconstruction.

Interglacial Period Comparison

This graph compares the current interglacial period with the previous four interglacials using the EPICA ice core climate reconstruction. Each interglacial period has been normalized to the time the global temperature departure first reached the current “normal” temperature. Click on graph to enlarge.

I find it amazing how abruptly and similarly each glacial period ended in about 5,000 years to start each following interglacial warm period.  In contrast, the duration of the interglacials has been much more variable. The most recent previous interglacial period that started about 130,000 years ago lasted about 14,000 years at temperatures at or above our current modern “normal”. The second previous interglacial lasted about as long as our current interglacial while the third previous was by far the shortest, lasting about 2,000 years, but arguably had somewhat of a double peak. However, the latter secondary peak did not quite reach the warmth of our current modern “normal”. The fourth previous interglacial which started about 418,000 years ago was by far the longest at about 25,000 years.

As repeatable as the glacial cycles have been over the last 500,000 years, I see little reason not to expect more of the same in the future. Using this interglacial comparison as a climate persistence forecast, we might expect about a 75% chance that the global average temperature will begin to drop dramatically sometime within the next few thousand years and about a 25% chance of staying warm for another 10,000 years or so … at most. Perhaps we need all the anthropogenic warming we can muster to stall or prevent the next glacial period?

Our understanding of what causes these glacial cycles, which are relatively recent on a geological scale, is still very limited although there are plenty of hypotheses. Our current climate models cannot predict them and therefore to me are somewhat useless. Until we can create climate models that can accurately track past glacial and interglacial periods I will not be too impressed and I certainly don’t believe our infant and untested climate models should be used to shape policy regarding “climate change”.

Update 2016 November

Below is a link to an interesting analysis of the causes of glacial cycles, along with conclusions made by the author, which seem reasonable to me.  The author hypothesizes that evidence suggests that the current interglacial period is likely to be only average in length and therefore should be ending soon, most likely sometime within the next two thousand years.

Nature Unbound I: The Glacial Cycle


1) Obliquity is the main factor driving the glacial-interglacial cycle. Precession, eccentricity and 65°N summer insolation play a secondary role. There is no 100 kyr cycle. Milankovitch Theory is incorrect.

2) The current pacing of interglacial periods is the consequence of the Earth being in a very cold state that prevents almost half of obliquity cycles from successfully emerging from glacial conditions. The rate for the past million years has been 72.7 kyr/interglacial, or 1.8 obliquity cycles between interglacials. This can be generally described as one interglacial every two obliquity cycles except when close to the 413 kyr eccentricity peaks, when interglacials take place at every obliquity cycle.

3) Glacial terminations require, in addition to rising obliquity, the existence of very strong feedback factors manifested as very low glacial maximum temperatures. High northern summer insolation at the second half of the rising obliquity period is a positive factor, and if high enough during eccentricity peaks can drive the termination process.

4) CO2 can only produce a minor effect in glacial terminations since the measured change in concentration (roughly a third of a doubling which represents half of the warming effect of a doubling) is too small to account for any important contribution to the large observed temperature changes.

5) Since the precession cycle has bottomed and the obliquity cycle is half way down we should expect the next glacial inception to take place within the next two millennia.