hide
Free keywords:
-
Abstract:
Earthquake focal mechanism solutions (FMS) form the basic
data input for many applications, e.g. stress tensor inversion
or ground-motion prediction equation estimation. In these
applications the FMS data is usually binned spatially or in
predetermined ranges of rake and dip based on expert elicitation.
However, due to the significant increase of FMS data
in the past decade an objective data-driven cluster analysis
is now possible. Here we present the method ACE (Angular
Classification with Expectation-Maximization) that identities
clusters of FMS without a priori information. The identified clusters can be used for the classification of the Style-of-
Faulting and as weights for FMS data binning in the aforementioned
applications. As an application example we use
ACE to identify FMS clusters according to their Style-of-
Faulting that are related to certain earthquake types (e.g.
subduction interface) in northern Chile, the Nazca Plate and
in Kyushu (Japan). We use the resulting clusters and weights
as a priori information for a stress tensor inversion for these
regions and show that uncertainties of the stress tensor estimates
are reduced significantly.
