Atmospheric and oceanic impacts of Antarctic glaciation across the Eocene–Oligocene transition

Composite satellite image of what the Earth may have looked like prior to Antarctic
glaciation during the late Eocene (image by Alan Kennedy).

The Eocene-Oligocene Transition occurred approx. 34 million years ago and was one of the biggest climatic shifts since the end of the Cretaceous (with the extinction of the dinosaurs). The Earth dramatically cooled and the Antarctic ice sheet first formed, but the cause and nature of the cooling remain uncertain. Using a climate model, HadCM3L, we looked at the effect of ice sheet growth and palaeogeographical change (i.e. continental reconfiguration as Australia separated from Antarctica) on the Earth’s steady-state climate. We utilised four simulations: a late Eocene palaeogeography with and without an ice sheet and an early Oligocene palaeogeography with and without an ice sheet.

The formation of the Antarctic ice sheet causes a similar atmospheric response for both palaeogeographies: cooling of the air over Antarctica, intensification of the polar atmospheric cell and increased winds over the Southern Ocean. The sea surface temperature response to the growth of ice is very different, however, between the two palaeogeographies. For the Eocene palaeogeography there is a 6°C warming in the South Pacific sector of the Southern Ocean in response to ice growth, but very little change (or even a slight cooling) for the Oligocene palaeogeography. Why, under the same forcing (the appearance of the ice sheet), do these different palaeogeographies have such different sea surface temperature responses?

The stronger winds over the Southern Ocean force more-saline water from the southern Indian Ocean into the less-saline southern Pacific Ocean. This is particularly important for the Eocene simulations, where the narrow gap between Australia and Antarctica limits flow from the Indian to the Pacific Ocean. As salinity in the southern Pacific Ocean increases the water becomes denser and sinks, releasing heat. This accounts for the increase in sea surface temperature in the Eocene simulations. In the Oligocene simulations, flow is already much greater between the Indian and Pacific Oceans, and so there is no marked increase in density, sinking or sea surface temperature following glaciation. There is only a mild cooling due to the presence of the large, cold ice sheet.

Whether in reality the dominant ocean response to glaciation was warming or cooling may have impacted the growth of the ice sheet at this major transition in the Earth’s history. However, more importantly, this research highlights that sensitivity to subtle changes in palaeogeography can potentially have very large effects on the modelled climatic response to an event such as Antarctic glaciation. This could be very important for understanding palaeoclimate records and interpreting climate model results.

This research, carried out by Alan Kennedy, Dr Alex Farnsworth and Prof Dan Lunt of the Cabot Institute and University of Bristol with others, is featured in a special issue of the Philosophical Transactions of the Royal Society A. The full special issue on the theme of ‘Feedbacks on climate in the Earth System’ and the paper can be accessed here.

Special issue cover (image from Royal Society).

Citation: Kennedy A.T., Farnsworth A., Lunt D.J., Lear C.H., & Markwick P.J. (2015) Atmospheric and oceanic impacts of Antarctic glaciation across the Eocene–Oligocene transition. Phil. Trans. R. Soc. A, 373, 20140419, doi:10.1098/rsta.2014.0419.
This blog is written by Alan Kennedy from the School of Geographical Sciences at the University of Bristol.  This blog post was edited from Alan’s blog post at Ezekial Boom.

Alan Kennedy


Oligocene discussion day

On the 16th of May, the University of Bristol held a half-day meeting devoted to the discussion of the Oligocene epoch (34 to 23 million years ago [Ma]). The Oligocene is a period of relative climate stability following the establishment of permanent ice sheets on Antarctica (34Ma). By the early Miocene (23Ma), atmospheric CO2 was low enough to allow the development of northern hemispheric ice sheets1. As a result, the Oligocene may have been the only time in the Cenozoic era (65-0Ma) during which a unipolar glaciation could exist.

Despite this, the Oligocene has received little attention from the Cenozoic palaeoclimate community. The aim of this event was to promote awareness of the Oligocene and encourage future research within this field.

Ellen Thomas, currently in Bristol on sabbatical from Yale, and David Armstrong-McKay, from the National Oceanography Centre (NOC), began the morning session with a series of talks devoted to the late Eocene and early Oligocene. Ellen discussed the Eocene-Oligocene transition (34Ma) from both a modern2 and historical3 perspective while David outlined the competing hypothesis put forward to explain the event4.   Dierderik Liebrand, also from the NOC, followed this with a talk on late Oligocene and early Miocene (24-19Ma) cyclostratigraphy5.  Following lunch, Bridget Wade gave an hour-long seminar on the Eocene-Oligocene boundary (34Ma)6 and the middle Oligocene (24-30Ma)7. Bridget’s talk doubled as a departmental seminar in the School of Geography.

Figure 1: A compilation of benthic foraminifera oxygen isotope values. During the Oligocene, this reflects a combination of ice volume and temperature7

The event was hosted by Gordon Inglis, a PhD student in the School of Chemistry, and was funded by Professor Rich Pancost (Global Change) and Professor Paul Valdes (School of Geography).


For more information, please consult the following references:

  1. Zachos, et al. (2008) An early Cenozoic perspective on greenhouse warming and carbon-cycle dynamics: Nature, v. 451, p. 279-283
  2. Liu, Z. et al (2009) Global cooling during the Eocene-Oligocene transition: Science, v. 323, p. 1187-1190
  3. Kennett and Shackleton (1976) Oxygen isotopic evidence for the development of the psychrosphere 38 Myr ago: Nature, v. 260, p. 513-515
  4. Merico, A, et al. (2008) Eocene/Oligocene ocean de-acidifiation linked to Antarctic glaciation by sea level fall: Nature, v. 452, p. 979-982
  5. Liebrand, D., et al. (2011) Antarctic ice sheets and oceanographic response to eccentricity forcing during the early Miocene: Climate of the Past, v. 7, p. 869-880
  6. Wade, B., et al (2011) Multiproxy record of abrupt sea-surface cooling across the Eocene-Oligocene transition in the Gulf of Mexico: Geology, v. 40, p. 159-162
  7. Wade, B. And Palike, H., (2004) Oligocene climate dynamics: Palaeoceanography, v. 19, PA4019