Integrating theory and observations of the Pleistocene (iTOP)

The central motive of palaeoclimate dynamics theory is to use of both models and data to test hypotheses about the evolution of climate. Our ambition is to develop methodologies based on Bayesian inferrence theory and dynamical system analysis that will allow us to make an optimal and ideally unbiased use of all sources of information at our disposal : data, general circulation models, empirical principles.

Key challenges

What are the mechanisms of glacial-interglacial cycles ? When and why did they regime change ? Would they occur in absence of orbital forcing? What do we learn from recent marine, ice core and terrestrial records? Can we predict climate on long time scales? When will the next glacial inception occur? Do past and future greenhouse gas emissions have long-term consequences ? Is a glacial-inception overdue, as suggested by W. R. Ruddiman ? " ... a complete theory of the ice ages is still elusive" (M. Raymo and P. Huybers, Nature 2008)

The originality of the project is to build-up on the very powerful approach developed by B. Saltzman (in which both general circulation models and emperical principles merged and expressed in the form of reduced-order models) and attach to it a rigorous treatment of model error compatible with the modern understanding of Bayesian inferrence. In this framework, the model is validated and then calibrated with palaeoclimate data to provide informed predictions (ideally unbiased).

The project builds on a 30-year history of expertise developed at our Institute on the topics of

• astronomical forcing theory
• time-series analysis
• climate dynamics

However, we are also in the process of developing new areas of expertise in the framework of national and international collaborations :

• Bayesian inferrence and dynamical model calibration (in collaboration with Bristol University)
• Dynamical system analysis
• Complex system theory (dissipative systems and maximum entropy theory)

The project is funded by the European Research Council, in the framework of a independent research starting grant, by the Belgian National Fund of Scientific Research and the Université catholique de Louvain.

contact person : Michel Crucifix (michel.crucifix (at) uclouvain.be )

Job Offers

Three positions are open for enthusiastic individuals:

• A post-doctoral fellow with credentials in Bayesian statistics. He/she will be responsible for the development, validation and calibration of dynamical-statistical models. Calibration strategies envisaged in this project include particle filters, and the candidate will have the freedom to explore techniques at the frontier of modern applied statistics through active collaboration with our partner, the Institute of statistics of the University of Bristol.
• A post-graduate research fellow with training and interest in dynamical system theory and analysis. His/her mission is to develop and study low-order dynamical systems capturing ice age dynamics inferred from palaeoclimate data analysis.
• A post-graduate research fellow with training and interest in experiment design and analysis. The goal of his/her work will be to provide constraints on the Earth climate response to changes in orbital parameters, ice sheet topography and ocean state, on the basis of a set of well-designed computer experiments with general circulation models. The applicant will benefit IT support to actually run the experiments.

We are keen to develop an international and lively team spirit, with unified working practices and open-access mentality. Candidates must have good english spoken and oral skills and be ready to work with LaTeX and the R statistical environments. Candidates should be open to European mobility.

The “George Lemaître Center for Climate and Earth Research”, within the Université catholique de Louvain, is a particularly active environment with two currently running ERC projects. It is located in the lively and cosmopolitan town of Louvain-la-Neuve, 30 km South of Brussels, and is expected to move to fully renovated offices within 12 months.

Salaries are determined in accordance with the Belgian legal system. Positions are contractual.

Applications should be sent along with three scanned recommendation letters to itop@uclouvain.be by the 31st September 2009.

Related articles

How to cope with climate's complexity?

On the use of simple dynamical systems for climate predictions: A Bayesian prediction of the next glacial inception (DAMES Workshop, October 2008, Potsdam)

The first article available on the arXiv corresponds to the Heidelberg (2008) talk (see below) and will appear in European Reviews soon. It has bee commented in the Nature Physics by science writer Mark Buchanan : http://www.nature.com/nphys/journal/v4/n11/full/nphys1117.html (text available on request to me)

The second article is an extension co-authored with Jonathan Rougier after a talk at the DAMES 2008 workshop. It is accepted for publication in European Physics Journal.

Recent presentations :

Key challenges in Modeling Earth System Dynamics (QUEST ASM Meeting, March 2008, Winchester)

How to handle climate system's complexity ? (Heidelberg, April 2008)

Is Ruddiman's hypothesis undecidable (Vienna, April 2008)

Entropie : un concept utile pour l'étude du system climatique ? (Louvain-la-Neuve, May 2008)