David Walsh

Victoria University of Wellington

Bayesian methods of earthquake focal mechanism estimation

This is joint work with Richard Arnold and John Townend, both VUW

Earthquakes are important signatures of the stress in the earth’s crust. In order to interpret the constraints earthquakes place on the stress field it is important to summarise — in a statistically valid way — the information that observations of earthquakes contain. The focal mechanism of an earthquake describes the geometry of the fault on which the earthquake occurred using three parameters: the strike, dip and rake. We are investigating a Bayesian method of calculating a posterior probability distribution for these focal mechanism parameters. We approximate such a distribution using a Matrix-Fisher distribution, and here we apply these methods to phase data from the Raukumara Peninsula, New Zealand. Our approach has the advantage of enabling us to incorporate observational errors and give robust posterior uncertainties for the estimated focal mechanism parameters.