The aftershock series of the 2016 Petermann Ranges earthquake in Central Australia

Although located far from any active plate boundaries, Central Australia features significant seismic activity, with a total of four M>6 events in the last four decades. The most recent such event was the May 25^th, 2016 Petermann Ranges earthquake (Mw = 6.1), which occurred close to the border triangle between the Northern Territory, South Australia and Western Australia. The last tectonic reactivation of the region that hosted the earthquake occurred during the intraplate Petermann Orogeny that terminated about 540 Ma ago. It is commonly assumed that although recently inactive, this region still constitutes a lithospheric-scale zone of weakness, so that stresses imposed on the rigid Australian plate at its edges can localize and lead to seismicity here. Previous studies have shown that the Petermann earthquake occurred on a splay fault in the direct vicinity of the Woodroffe Thrust, one of the principal shear zones of the Petermann Orogeny. It occurred at a very shallow depth (<= 5 km) and had a thrust mechanism with a NW-SE oriented rupture plane. Due to its shallow depth, it created a surface rupture that was mapped over a length of about 20 km.

In the present study, we utilized a temporary deployment of 11 seismic stations that was installed in the aftermath of the Petermann earthquake to characterize its aftershock sequence. Since only a single permanent station was operating within a radius of 400 km around the rupture area before the deployments, we do not have much information about the earliest part (first two weeks) of the aftershock series.

We combine two different event catalogs, a handpicked one comprising 1231 events for the first 3.5 months of the aftershock sequence, and a semi-automatically derived one that contains a total of 4918 events. We derived an optimal 1D velocity model and station corrections from the handpicked catalog, and relocate all events with this model. In a second step, we apply a double-difference relocation to the entire dataset. Most of the relocated events occurred at depths between 2 and 4 km, and outline a tight NW-SE striking an NE-dipping plane that aligns well with the mapped trace of the surface rupture. As already indicated in previous studies, we also find scattered aftershock activity within the footwall of the fault.

We further applied a template-matching algorithm in order to further decrease the completeness magnitude of the catalog and to get the best possible estimate for event rate decay over time. Moreover, we present fault plane solutions for a few of the largest aftershocks.