In the March 2010 issue of Nature Geoscience, Yin and Berger analyze the warming of interglacial (e.g. present-day) intervals across the Mid-Brunhes Event (MBE) at approximately 430,000 years ago. Interglacials prior to that time were cooler, and exhibited lower sea levels, larger ice sheets, and lower atmospheric CO2, unlike the present-day interglacial or any after the MBE. In an attempt to explain why this change occurred, Yin and Berger modeled interglacial peaks over the past 800,000 years as a function of greenhouse gases and Milankovitch orbital forcing.

The Milankovitch hypothesis proposes that the 100,000 year periodicity of glacial and interglacial intervals present in the geological record is a function of orbital forcing. Changes in the Earth’s axial tilt (obliquity), rotational wobble (precession), and distance from the Sun (eccentricity) combine to affect the latitudinal distribution of incoming solar radiation across the Northern and Southern Hemispheres (i.e. boreal and austral), which is responsible for long-term climatic variation. Axial tilt and precession are the primary controls on insolation (41,000- and 25,700-year periodicity, respectively), but are amplified and modulated by longer-scale changes in orbital eccentricity (100,000 years and, to a lesser degree, 413,000 years). With this in mind, it’s clear that the discrepancies between pre- and post-MBE interglacial intervals need to be explained.

To answer the question, Yin and Berger chose standard insolation values for each successive interglacial peak across the past 800,000 years. Phase congruency between precession and tilt created the insolation standard, and interglacial peaks (i.e. the warmest part of the 100,000 year cycle) were chosen based on the marine δ18O record, otherwise named the ” δ18O minima”. Recall that the marine record exhibits lower δ18O during warm periods while glacial ice records higher δ18O due to evaporation fractionation.

What does this mean? The following figure provides a quick idea of what’s going on:


Fig.1. Obliquity, precession, and δ18O modeled over the past two interglacials from 135,000 to present day. The δ18O black bar represents the peak warming of each interglacial (δ18O minima), while the precession black bar represents when the Earth was closest to the Sun in the Northern Hemisphere summer (perihelion) and the obliquity black bar indicates the maximum tilt of the Earth (~24.5 degrees). Full article text shows 800,000-year record. Adapted from Yin and Berger (2010).

As indicated, the in-phase obliquity and precession curves indicate peak insolation, and pre-date the warmest part of the interglacial interval by about 5,000 years, which is more or less consistent with what is expected. Deglaciation is a long and protracted process with many starts, stops, and feedbacks, so it is typical for there to be a lag between maximum insolation and peak warmth. In a general sense, the Milankovitch hypothesis would only be in trouble if maximum insolation post-dated the warmest part of interglaciation.

In a snapshot discussion, Yin and Berger reveal that complications arise for specific interglacial intervals. And would we expect anything less from Mother Nature? However, averaging the pre- and post-MBE interglacials, the authors identified a clear interglacial warming after 430,000 years ago.

It appears the primary culprit is increased atmospheric CO2 post-MBE present in the geological record, in conjunction with increased winter insolation in the Northern Hemisphere. The authors cite a 60%-30% split (final 10%??), and conclude that a) boreal winters are generally warmer during interglacials after 430,000 years ago than they were pre-MBE, and that b) increased winter warming exerts a stronger control on climate than increased summer warming.

All told, the paper is a good look into some interesting aspects of the late Quaternary icehouse climate. If you have access to Nature Geoscience, check it out.


Yin, Q.Z., and Berger, A., 2010. Insolation and CO2 contribution to the interglacial climate before and after the Mid-Brunhes Event. Nature Geoscience, vol. 3, pg. 243-246.