← Back to Sessions

Virtual Earthquake Analysis of Future Alpine Fault Earthquakes and Ground-Shaking Using the Southern Alps Long Skinny Array (SALSA)

Description: 

A challenge in seismic hazard forecasting is to evaluate the ground motions likely to be produced in an earthquake without knowing which of many geologically- and geophysically-plausible scenarios will eventuate. Time is of particular essence in the case of the central Alpine Fault, New Zealand, which is recognized as being late in the typical interseismic phase of its 249±58 earthquake cycle, having last ruptured 306 years ago. The likelihood of a M_w>7.5 Central Section earthquake occurring in the next 50 years ~75% (29–99%; 95% confidence interval), and the likelihood of this earthquake also rupturing the South Westland Section in a M_w~8 event is ~82% (Howarth et al., 2021).

Previous models of Alpine Fault earthquakes have addressed only idealized scenarios that ignore recognized along-strike and down-dip variations in fault geometry, present-day microseismicity, or thermal regime that likely influence coseismic rupture. To investigate the effects of these and other factors such as the direction and speed of rupture on ground motions, we are undertaking virtual earthquake analysis (Denolle et al., 2014, 2018) using data from the Southern Alps Long Skinny Array (SALSA). SALSA consists of 56 seismometers (42 temporary and 14 permanent, 40 broadband and 16 short-period) deployed along a ~450 km length of the Alpine Fault: 45 of the seismometers are installed within ~3 km of the fault trace, primarily on the hanging-wall, at 10 km nominal spacing. SALSA noise recordings will be combined with data from seismometers throughout southern New Zealand to synthesize Green’s functions representing the farfield response to incremental slip anywhere on the fault. By convolving these with kinematic rupture models incorporating empirical constraints on fault-zone structure and properties, we can compute ground motions at locations of interest in response to large numbers (millions) of plausible ruptures. This enables us to efficiently explore the range of ground motions consistent with measurements made along the length of the Alpine Fault, and to quantify the epistemic uncertainties in existing ground motion estimates.

Session: Future Directions in Physics-based Ground-motion Modeling in Preparation for the Fall 2023 Meeting

Type: Oral

Date: 4/18/2023

Presentation Time: 04:45 PM (local time)

Presenting Author: John Townend

Student Presenter: No

Invited Presentation: 

Authors