Seeing the Big Picture on Climate Change and Variability

Can using only high-end climate models to study climate change mean losing sight of the forest for the trees? Yes, according to ICTP scientists Fred Kucharski and Riccardo Farneti of the Centre's Earth System Physics section. To avoid the situation where the climate community is too lost in details to see the big picture, Kucharski and Farneti emphasize the importance of using "intermediate complexity models".

"The best models are as complex as nature itself," says Kucharski, adding, "So it is important to use simpler models along with the state-of-the art models to better understand climate variability and change."

In their recently published paper titled " On the need of intermediate complexity general circulation models", which appeared in the Bulletin of the American Meteorological Society (Volume 94 Issue 1), Kucharski and Farneti stress the importance of intermediate complexity models, using SPEEDY (Simplified Parameterizations, primitivE-Equation DYnamics) as an example.

SPEEDY is an atmospheric general circulation model (AGCM) that has been developed at ICTP.  "While this kind of a model may not be used to study every aspect of climate change, it can be used to study the fundamental dynamics that affect climate systems," says Farneti. "Rather than making climate projections, the model tries to answer why climatic variations occur."

"Intermediate complexity models can run with lower computational power," says Kucharski.  "Moreover, many of them are freely available for download and use."

The memory requirement to run SPEEDY, for example, is about 35 MB, and it can run on older desktop computers with low hardware specifications. "This means the models can be used by researchers from developing countries and by graduate students who have access to limited resources," adds Farneti.

"If you look at the mean climate, the results obtained using SPEEDY are as good as those from the best models," says Kucharski. Both Kucharski and Farneti say that the aim is to keep things simple and provide a modelling system that can run on inexpensive computational equipment and at the same time allow efficient investigations for a broad range of research questions arising in the field of Earth-system science.