Climate Model Parameter Optimization

While computational models of global climate show remarkable qualitative agreement, they also differ widely in their sensitivity to projected increases in greenhouse gases. There is enough variability among the various model predictions to make it worthwhile to try to pin down and reduce the sources of uncertainty. That is the task undertaken by Charles S. Jackson and the Climate Research Group he co-directs with Robert Scott, which includes researchers at the UT Institute for Geophysics (UTIG) and in the Department of Geological Sciences as well as collaborators at several other universities.

Jackson determines climate uncertainties by applying a Bayesian statistical method called stochastic inversion. This is the same method used by seismologists at UTIG and elsewhere to determine the composition of the Earth's crust and mantle from seismic data. Astronomers also use the method to determine the detailed composition and structure of the sun and other stars. Treating the results of climate models as initial data, Jackson can calculate the degree of uncertainty that can be associated with individual parameters used in all the model calculations.

For a detailed description of this computational project and others in which Jackson is also involved and working on at TACC, see "The Unbearable Uncertainties of Climate Change."