Section: New Results

Climate Modelling

Participant : J.P. Quadrat.

Motivated by the issue we have started to study the climate modeling literature to identify some domain where we can contribute. Our first interest was very simple model where the system is naturally stabilized by change of albedo when the temperature increases like the daisy world of Lovelock but they are too much simplistic. Our second interest was the global circulation model of the atmosphere and ocean to see if it is possible to design a toolbox in ScicosLab dedicated to climate simulation. After a while it appears clearly that it is almost impossible that this kind of models are able to make a prediction enough precise to be useful. Our third interest was the very rich literature of the sceptic about anthropic warming. In particular, we have looked more deeply to the radiation modeling of the atmosphere for which the influence of the atmospheric vapor and the clouds is largely debatable. At the end, it seems that there is still places for studies on classes of simplified model of the earth. For example the linearization of the dynamic of a simplified earth in thermal periodic regime could (by computation of eigenvectors) give information about the best places to measure the temperature to detect the abrupt change of the dynamics if there is one. This sensor optimization and detection problem could be an interesting research subject if the time life of the project METALAU was large enough.