Antarctic Geothermal Heat Flow : Investigations by Geophysical and Statistical Analyses

Antarctica’s contribution to past and future sea level changes is highly uncertain given the poor understanding of ice sheet dynamics in which solid Earth interactions play an important role. Geothermal heat flow (GHF) is one of the least constrained solid Earth components but has a significant influence on the visco-elastic behavior of the lithosphere and the thermal state at the base of the ice sheet.

Estimations of Antarctic GHF seldom come from direct measurements and rely on indirect methods based on geophysical observations. Forward models using constraints on lithospheric isotherms and assumptions on uniform thermal parameters exhibit large differences, both in amplitude and spatial distribution of the calculated heat flow. The consistency of such models is explored using a Bayesian inversion approach in an effort to reconcile different modeled lithospheric structures. Further, a machine learning approach is adopted that enables a statistical derivation of GHF incorporating multiple global geophysical data sets and in situ heat flow measurements. Lastly, a novel joint inversion approach is applied to magnetic and gravity data to invert for the crustal structure of the Wilkes Land region in East Antarctica and South Australia. This improves the understanding of the subglacial geology and small-scale GHF contributions.

The methods and results presented in this thesis are relevant for the thermal modeling of ice sheets and the lithosphere, especially with regard to understanding the coupling between ice and solid Earth.