Palaeoclimate and Climate Modelling with the New CSIRO Coupled Atmosphere-Ocean - Land Surface - Sea Ice Scheme of General Circulation Models

Principal Investigator

John Chappell

Biogeography & Geomorphology


The CSIRO coupled ocean-atmosphere-sea ice climate model was used to simulate some of the salient features of the glacial-interglacial climatic cycle (the last 130 000 years). These were the first ever simulations with such a complex climate model. This work has far reaching implications as it is directly relevant to Quaternary palaeoclimatology and highly relevant to model validationB issues.

Four critical time-slices were simulated; the pre-Industrial climate; the mid-Holocene climatic optimum (ca. 6 000 years BP); the early glacial transition (ca. 116 000 years BP); and the last Interglacial (ca. 126 000 years BP). Two of these simulations were integrated in time for 160 model years on the VP2200 supercomputer and the results were archived at the mass storage facility for further analysis.


Joseph Syktus

Division of Atmospheric Research



r51 - VPP

What are the results to date and the future of the work?

This work demonstrated for the first time, using a complex climate model, that the main features of the proposed orbital theory of climate change are simulated successfully. This demonstrates that our understanding of the physical mechanisms driving climatic changes in the past is correct and that climate models are able to simulate a range of the climatic conditions, some of them very different from present day climatic conditions.

The model results show that two different climatic states can eventuate, depending on the seasonal distribution of insolation, which is governed by the orbital elements (i.e. obliquity, eccentricity and the precession of the equinoxes). The orbital forcing which is typical of a "cold summer orbit" leads to accumulation of snow and ice on the continental fringes in the Northern Hemisphere, which in time leads to the growth of massive ice sheets more than 3 km high. The climatic state of such a regime is glacial with weak monsoonal circulation in the subtropics (i.e. African-Asian monsoonal system). On the other hand, conditions typical of a "warm summer orbit" lead to excessive heating of the continents in summer, which in turn melts all snow and ice accumulated during the winter. The monsoonal circulation is strong and typical of that observed during the interglacial times.

Some of the results from this work were presented at the Western Pacific Geophysical meeting of the American Geophysical Union in Brisbane, July 1996, and a paper was submitted to the Journal of Climate. In the near future we plan to produce a paper describing the ocean circulation characteristics of these two climatic states.

What computational techniques are used?

The atmospheric part of the climate model used spectral techniques and the ocean part used finite difference equations, which were integrated in time to produce a climatic solution close to equilibrium under given climatic forcing. One year of the coupled model integration required some four hours of CPU time on the VP2200 supercomputer and it was 72% vector efficient. The model was integrated for 160 years in synchronous ocean-atmosphere coupling mode. The three bottom layers (out of twelve in total) in the ocean model used acceleration techniques in order to bring the bottom ocean circulation to equilibrium more quickly. In a normal non-accelerated integration, the deep ocean circulation would require well over 1000 years to equilibrate.


Syktus, J., Chappell, J., Orbital forcing of the interglacial and early glacial climates: A CSIRO coupled GCMs study. Journal of Climate (in press), 1997.

Chappell, J., Syktus, J., Palaeoclimatic modelling: a western Pacific perspective. In: Giambelluca, Thomas W. and Henderson-Sellers, Ann, editors. Climate change: developing southern hemisphere perspectives. Chichester: Wiley; pp. 175-193, 1996. Research and developments in climate and climatology.

Chappell, J., Syktus, J., Orbitally-forced monsoonal circulation changes during the last glacial cycle: interhemispheric connections. Supplement to EOS, 77, 22, 1996.

Syktus, J., Chappell, J., Orbitally-forced ice-age initiation using the coupled CSIRO GCM's. Supplement to EOS, 77, 22, 1996.

Masson, V., Syktus, J., Watterson, I., Mid-Holocene: Impact of ocean model on simulated climate at 6,000 years BP. Supplement to EOS, 77, 22, 1996.

Chappell, J., Syktus, J., Palaeoclimate modelling and its relation to palaeodata. In: CLIMANZ IV: Quaternary Climates of Australia and New Zealand, 26-28 February 1996, ANU, Canberra (abs), 1996.

- Appendix A