Milankovitch Cycles and the Pleistocene Ice Ages

by Bruce Scofield

Milankovitch Cycles are astronomical cycles of the Earth-Sun relationship. Cycles of this nature were suggested in a paper by Sir John Herschel in 1830, Joseph Alphonse Adhemar in a book in 1842 (Revolutions of the Sea), and more importantly, by James Croll in 1875 (Climate and Time). Milutin Milankovitch (1879-1958) developed these ideas over the course of his lifetime publishing Mathematical Theory of Heat Phenomena Produced by Solar Radiation in 1920 and Canon of Insolation of the Earth and its Application to the Problem of Ice Ages in 1941. (Insolation = INcident SOLar RADiation.) Milankovitch showed that three key orbital cycles modulate the amount of insolation the Earth receives over long periods of time.

Astronomical cycles

Precession: The cycle of the gradual change in the direction of the Earth’s axis. This change is caused by gravitational torques exerted by the Sun and Moon on the bulge of the Earth at its equator. When the Earth is at perihelion (today around January 3rd) the northern hemisphere is tiltled toward the sun. During summer (July 4th) the northern hemisphere tilts away from the sun. This moderates both seasons. At the present time the Earth receives about 3.5% more solar radiation at perihelion than at aphelion.

It follows that when aphelion occurs near the winter solstice, there will be less radiation from the sun and winters may be cooler.

The anomalistic year is the length of the year measured from perihelion to perihelion. It is about 25 minutes longer than the tropical year, a figure that amounts to a full day every 58 years. The position of perihelion and aphelion shift over a 21,000-year cycle (21,700 is average, but ranges between 19 ky to 23.7 ky) at a rate of about 1 day every 58 years. Currently, there is a 3% difference in distance between aphelion and perihelion.

The fact that there is more land mass in the northern portion of the northern hemisphere than there is the the southern hemisphere accounts for a stronger response to lower insolation there.

Obliquity: The Earth’s axis is currently tilted from the vertical to the plane of its orbit by 23.45 degrees, but this is not a constant. This tilt varies from a miniumum of about 21.8 degrees to a maximum of 24.4 degrees within a cycle of approximately 41,000 years. It is the tilt of the axis that produces the season and it follows that when the axis is tilted less the contrast between the season will be less. The affects of this cycle are felt equally in both hemispheres. Changes in obliquity have only a minor effect on solar radiation at low latitudes, but this increases in the higher latitudes. Changes in obliquity affect insolation in both north and south hemispheres equally.

Eccentricity: The orbit of the Earth around the sun varies from nearly circular to eliptical over a period of approximately 100,000 years (95,800 on average over the past 5 million years with a range from 95ky to 123ky). When the orbit is most eliptical (about 6 million miles farther from the sun at aphelion), and when winter in the northern hemisphere occurs at aphelion, one would expect long cold winters and short hot summers. The reverse would be true of the southern hemisphere. At maximum eccentricity insolation can vary by as much as 30%, affecting the intensities of the seasons with opposite effects in each hemisphere. This cycle modulates the precession cycle. There is a second eccentricity cycle of about 400,000 years (413ky) that also appears to be reflected in long-term climate cycles.

Milankovitch theorized that the total summer radiation that is received in the northern latitudes (near 65 degrees north – where ice sheets have formed previously) is the key factor in the development of an ice age. When summer in the northern hemisphere coincides with aphelion, the obliquity is at minimum, and eccentricity is high, conditions are most conducive to glaciation. The triggering of an ice age is thought to occur during times of cool summers which allow an accumulation of ice and snow from year to year in the higher latitudes. Eventually this builds into an ice sheet that reflects solar radiation back into space, which further cools the Earth (positive feedback). The amount of CO2 in the atmosphere declines as the ice sheet grows, yet another cooling factor.

The Earth’s axis and its orbit are perturbed by the gravitational forces of the other planets. These perturbations produce cycles in insolation. When the solar system is viewed dowward, asssuming the plane of the Earth’s orbit as fixed, the orbits of the other planets (their shapes) are regarded as “g” frequencies. Differences in the planetary orbits on the vertical level (parallel to the plane of the Earth’s orbit) are called “s” frequencies. Climatic frequencies are produced by the combinations of the individual g and s frequencies and the precessional constant p. The limits of precision allow the g and s frequencies to drift over long periods of time making it impossible to produce isolation curves for more than about 20 million years into the past or future. At 200 million years there can be a 40% error. The 404 ky cycle appears to be very stable and geological records (Newark basin and other Triassic-Jurassic basin cyclical records) may serve as a means of focusing the celestial mechanics of orbital variables in the remote past.

The Pleistocene Ice Ages

The actual beginning of the Pleistocene was first suggested by Lyell in 1893 – the point in the geological record in which 90-95% of the fossils found were of organisms alive at present. Forbes later suggested that this era be dated by evidence of a cold climate. In 1948 it was suggested that the first appearance of cold-water species found in a specific southern Italian sediment marked the beginning. Later it was found that this point coincided with the Olduvai Normal Event at 1.8 mya.

Traditionally, four ices ages during this period were documented in North American and six or seven in Europe. In the 1960’s 100,000 year cycles emerged from data derived from Czechoslovakian brickyards (sedimentary deposits) and Caribbean deep sea core V12-122. It is now thought that during the Pleistocene (1.8 mya) there were at least 20 alterations of warm/cold cycles.

The climatological record for the termination of the previous ice age about 12,000 years ago is accounted for by the Milankovitch model – an increase in northern hemisphere summer insolation.

The Climatological Record

In the 1950’s Milankovitch’s theory was rejected by most geologists due to the advent of radiocarbon dating and the results it produced in regard to the the Pleistocene. It was discovered that there had been warm climate intervals as recent as 25,000 years ago and other ice fluctuations iver 80,000 years that were at variance with Milankovitch cycles. In the 1960’s the analysis of deep sea cores (Ericson, Emiliani, etc.) produced evidence that sea level fluctuations appeared to occur in a 21,000 year cycle.

Dating methods of deep sea cores include the following:

  1. The ratio of Oxygen isotope 18 to Oxygen Isotope 16 as found in the calcium carbonate shells of foraminifera. This data reflects changes in the volume of the ice sheets.
  2. Radio carbon dating tell the age of a sample, but is less accurate over 40,000 years.
  3. Protactinium and Ionium are radioactive elements in the ocean floor mud and ooze remaining from the decay of Uranium. This data has revealed eight cold climate periods alternating with warm periiods from 700,000 years ago.
  4. Magnetic stratigraphy analyzes the record of magnetic reversals and has produced a well-dated history of climate. Some extinctions concur with this data and some specific bench marks in Earth history are mapped in this manner. The Brunhes-Matuyama boundary occurred 700,000 years ago and the Olduvai Normal Event at 1.8 million years ago.

In 1971 the CLIMAP project began to reconstruct the history of the North Atlantic and the North Pacific during the Brunhes epoch. The cores obtained were analyzed by Shakleton who found a dominant 100,000 year cycle, but the evidence of the other cycles was unclear. Hays analyzed core RC11-120 (for radiolaria and isotopic data) from the Indian Ocean in which there was a high deposition rate which could show better details. This showed that the southern hemisphere was changing in synch with the northern hemisphere. The addition of core E-49-18 extended the range of data to 450,000 years with enough detail to show cycles as short as 10,000 years. In these evidence of all Milankovitch cycles were found and the results were published in 1976.

Using sediment cores from the Indian Ocean, Hays et al. (1976) showed that there was a fairly consistent phase relationship between insolation, sea surface temperature and ice volume. They also found that the 100ky cycle was far more prominent that had been expected. Of the three cycles, eccentricity was believed to have the least effect on climate. The 100ky cycle has been consistent for the previous 700ky, but before that time climate cycles of about 40ky (obliquity) appear to dominate. There have been two approaches to this problem: One seeks an explanation in orbital forcing by the 19ky and 23ky precessional frequencies which affect the terminations of glacial episodes. The second proposes that ice-volume fluctuations are modulated, not driven, by orbital forcing in a highly complex system.

Uranium-thorium based data from “Devils Hole” in Nevada suggests that the termination of the ice ages (Termination II) was not consistent with the Milankovitch model. (Karner, 2000) Also, coral terraces from New Guinea suggested that sea levels were at a high point (glacial melting) as early as 142,000 years ago. Imbrie et al. set Termination II at 127,000 (+ or – 6ky). Other sea-level records are also inconsistent with the Milankovitch model and this raises the possibility that climate is too complicated to be predicted by a single parameter. It has also been shown that while the oxygen isotope record of the Vostok ice core matches the precession cycle, the temperature data does not. (J.R. Petit et al, Nature 399, 429 (1999).

Cores taken from the Antartic (Cape Roberts Project – 1998-1999) show ice sheet advances between 24.1 and 23.7 mya as occurring in cycles of 100,000 and 40,000 years.


Berger, A., J. Imbrie, J. Hays, G. Kukla, and B. Saltzman. Milankovitch and Climate: Understanding the Response to Astronomical Forcing. (1984). Dordrecht: D. Reidel Publishing Co.

Bradley, Raymond S. (1999) Paleoclimatology: Reconstructing Climates of the Quaternary. New York: Harcourt Academic Press.

Hays, J.D., J. Imbrie, N.J. Shackleton (1976). Variations in the Earth’s Orbit: Pacemaker of the Ice Ages. Science, Vol. 194, Number 4270, pp 1121-1132.

Imbrie, John, and Katherine Plamer Imbrie. (1979). Ice Ages: Solving the Mystery. Cambridge, MA: Harvard University Press.

Karner, Daniel B. and Richard A. Muller. A Causality Problem for Milankovitch. (2000). Science, Vol. 288, pp. 2143-2144.

Kerr, Richard. A. (1999). Why the Ice Ages Don’t Keep Time. Science, Vol. 285, pp. 503-504.

Climate, Astronomical Forcing, and Chaos. Milankovitch Theory and Climate.