← Back to Sessions

Clusty, A Versatile Waveform-Based Network Similarity Clustering Toolbox

Session: Waveform Cross-Correlation-Based Methods in Observational Seismology [Poster]

Type: Poster

Date: 4/21/2021

Presentation Time: 03:45 PM Pacific

Description: 

Clusty is an open-source toolbox for earthquake clustering based on event waveform similarity developed to identify and characterize active faults and fault networks. It is highly versatile with possible applications in complex tectonic regions but also in studies of acoustic emissions from small-scale experiments. The toolbox includes several techniques to compute network-based event similarities based on cross-correlations of waveforms recorded by a network of seismic stations. The density-based clustering allows the identification of events pertaining to a common fault over long segments. The clustering procedure is transparent thanks to several graphical outputs visualizing the homogeneity within each cluster and the differences between distinct clusters. Here, we present results of two case-studies over two different scales:

(1) The aftershock sequence of the Mw 6.9 2018 Zakynthos (Greece) Earthquake. The western termination of the Hellenic Subduction System is known for its complex tectonics. Our clustering results confirm the activation of multiple faults of various kinematics and complex geometric configurations. Based on our results we can identify the causative fault plane of representative focal mechanisms and assign them to individual active faults.

(2) A decameter-scale injection experiment conducted at Äspö Hard Rock Laboratory, Sweden. Here, we apply the clustering toolbox to acoustic emissions detected during high-pressure water injections into granitic rock to gain insight into the growth of hydraulic fractures.

Even in applications where the mechanisms of single events might be unknown, as in the second case study, the waveform-based clustering allows the identification of event clusters that cannot be discriminated based on locations or origin times alone.

Presenting Author: Peter Niemz

Student Presenter: Yes

Authors