Pytheas: An open-source software solution for local shear-wave splitting studies

Abstract

Shear-wave splitting analysis of local earthquakes is a challenge tackled by a long history of published software. However, there is currently no complete solution that integrates manual and (partially or fully) automatic methods. The Pytheas software aims to achieve this by offering a collection of popular analysis techniques accompanied by an intuitive Graphical User Interface. The software offers its own implementation of the following methods: (a) Visual inspection of polarigrams and hodograms (manual), (b) Rotation-Correlation (semi-automatic), (c) Eigenvalue (semi-automatic) and (d) Minimum Energy (semi-automatic). Moreover, it automates the last three operations by using Cluster Analysis. The software also utilizes databases to effectively manage results storing. It is compatible with standard file formats for event and station metadata (QuakeML and StationXML), as well as waveform data (such as SAC and miniSEED). Pytheas also introduces a novel algorithm for automatically assigning quality grades to measurements. To showcase the utility of the software, we analyzed a set of 263 events recorded in the Western Gulf of Corinth (Greece). The workflow included the establishment of a control group through manual analysis, the fully automatic processing of the catalogue and the verification of the results through visual inspection of a sample. Automatic and manual measurements show good agreement, validating the potential of Cluster Analysis. The presented software aims to provide a friendly experience to the end-user in all stages of its usage (installation, operation and database management) and trivialize the time-consuming part of a local shear-wave splitting study, leaving more time for interpretation and analysis of results.