Investigating Rupture Cascades on the Alpine Fault and Marlborough Fault System Using Lacustrine Paleoseismology

Rupture cascade describes a pattern of earthquake occurrence whereby a high-magnitude earthquake on one fault hastens the occurrence of failure on nearby faults, potentially resulting in sequences of large earthquakes on timescales of years to decades. Earthquake simulations and paleoseismic data suggest that rupture cascades may be a prevalent behaviour of the interconnected Alpine Fault and Marlborough Fault System (A-MFS), and that large earthquakes on the Alpine Fault may trigger rupture cascades on the Marlborough Fault System (MFS). However, to date there has not been a robust test of the temporal sequencing of large earthquakes on the A-MFS. Although major earthquakes on the central Alpine Fault are precisely dated, paleoearthquake ages on the MFS span centuries, precluding identification of earthquakes that occurred within years to decades of each other. More precise paleoearthquake chronologies are required from the MFS to determine if rupture cascades triggered by the Alpine Fault have occurred in the past. Lake sediments may provide suitable paleoseismic records because they can provide long, continuous records of earthquakes that have decadal temporal precision.

This thesis applied a lacustrine paleoseismology approach to Lake Tennyson, on the western MFS, to test the hypothesis that large earthquakes on the central Alpine Fault can trigger rupture cascades on the MFS. A paleoseismic record was developed using sedimentology, seismic survey data, and geochemical and paleo-ecological analyses, and dated with decadal precision using numeric age control (210Pb and 14C) in conjunction with Bayesian age modelling. The ages of eight earthquakes, spanning the last ~2000 years, were then compared with lacustrine paleoseismic records from the central Alpine Fault and two other lake sites on the western MFS to investigate spatio-temporal patterns of rupture on the A-MFS. Results showed that at least two of the last five central Alpine Fault earthquakes were followed within a maximum of 60 years by an earthquake on a western Marlborough fault, supporting the hypothesis that the Alpine Fault rupture can trigger rupture cascades in the MFS. The results from this research demonstrate the usefulness of lacustrine paleoseismology for investigating spatio-temporal patterns of rupture across the A-MFS, and future research should use more lake sites from the wider MFS, in conjunction with quantitative fault source modelling, to better understand the behaviour of this complex fault system.