Simulating the Climate of the Last Glacial Cycle


Great uncertainty remains as to the degree of natural variability in the Earth's climate, and as to the changes that might take place during the coming centuries. This project will use a global climate system model to simulate past changes – those that have taken place since the height of the last Ice Age. These simulations will enable us to investigate natural variability within the climate system, and the processes that drive long-term climate change. By assessing the ability of the model to simulate past changes, we will also be able to anticipate how reliable it might be at simulating future changes, and identify areas where it might be improved.

In order to adequately simulate the range of processes and interactions that occur within the climate system, models require the computing power and data storage capacity of supercomputers such as the APAC National Facility.


Principal Investigator

William Budd
IASOS
University of Tasmania

Project

e56

Co-Investigators

Steven Phipps
ACE CRC
University of Tasmania


Tas Van Ommen
Xingren Wu
Glaciology
Australian Antarctic Division

RFCD Codes

260602


Significant Achievements, Anticipated Outcomes and Future Work

The CSIRO climate system model has been ported to the APAC National Facility, with extensive work also carried out on optimisation and parallelisation of the code. The result has been a model which can be used to complete multi-millennial climate simulations in a matter of months. Initial control experiments identified some deficiencies within the model physics, which required investigation before any further simulations could be carried out. Working with researchers at CSIRO, the problems were identified and fixed, leading to a number of improvements being made to the model. As a direct result of this project, CSIRO are therefore already producing more realistic projections of future climate change.

The atmospheric and sea ice components of the model, coupled to a simple one-layer ocean model, have been used to produce initial simulations of both past and future changes in the Earth's climate. During 2004 this work will be expanded, with the full climate system model being used to simulate not only the conditions that prevailed 6,000 and 21,000 years ago, but also those that will prevail if the amount of carbon dioxide in the atmosphere is allowed to increase to three times its pre-industrial value (see Figure). At the same time, the role of flux adjustments will be studied. Although these artificial adjustments to the simulated fluxes of heat and freshwater between the atmosphere and ocean are generally required in order to ensure the stability of coupled climate models over long periods of time, they may also influence the response of the model to changes in external conditions.

The value of the past climate simulations will be considerably enhanced through their inclusion in Phase 2 of the Paleoclimate Modelling Intercomparison Project. This international project will compare the results of coupled climate models from around the world, studying both the role of feedbacks within the climate system and the ability of state-of-the-art climate models to reproduce climate states radically different from those of today.

Figure 1. Simulated increase in surface air temperature which arises when the amount of carbon dioxide in the atmosphere is increased to three times its pre-industrial level [units: degrees Celsius].

 

Computational Techniques Used

The project uses the CSIRO climate system model. This performs a forward-in-time integration, using a semi-implicit leapfrog time scheme, of the differential equations describing the motions of the atmosphere and oceans. A finite-difference method is used to calculate derivatives, and Fast Fourier Transforms are used to convert spectral fields to and from values on a latitude/longitude grid. Parallelisation is achieved using OpenMP.

 

Publications, Awards and External Funding

External Funding and Awards

None.

Publications

None.