Abstract
My thesis documents the first-ever paleoseismic trench investigation of the Akatore Fault, which has long been considered the most active fault to exist near Dunedin City. Two trenches were excavated across the fault in order to investigate the late Quaternary activity (timing, magnitude and recurrence of large ground rupturing earthquakes).
Trenching investigations at Big Creek and Rocky Valley have concluded that there have been three ground-rupturing earthquakes in the Holocene. An antepenultimate event has been constrained between 10,400 ± 1,700 and 1,326 ± 22 cal. yr BP, while, the penultimate and most recent events have been constrained between 1,326 ± 22 and 776 ± 22 cal. yr BP. These events resulted in 5 m of dip slip, hence 1 - 2 m of surface displacement per event, which may to have produced earthquakes with moment magnitudes 6.8 - 7.4.
Further studies at Taieri Mouth provided information on the longer term behaviour of the Akatore Fault. We estimated only 2 – 3 m of scarp development since the 125 ka marine terrace was formed. Since the Big Creek trench results indicated similar displacements achieved over three Holocene earthquakes, it is plausible that the scarp development has happened by way of these same three Holocene events. This would imply that there has been no activity along the Akatore Fault for a long period prior to these Holocene events i.e. little to no movement between 125,000 – 10,000 cal. yr BP. Furthermore, the Holocene slip rate along the Akatore Fault is significantly greater than the long term slip rate. This suggests the fault does not act in a characteristic fashion. It has an episodic / irregular behaviour. Similar behaviours have been determined for other Otago faults, which is problematic for forecasting future earthquakes. If inception of uplift along the Akatore Fault occurred ~1 Ma, the implied long-term slip rate is such that the fault may not yet have slipped enough in these Holocene events to accommodate the accumulated slip over the previous ~110 ka. The Akatore Fault needs to become the focus of a time-dependent seismic hazard calculation for Dunedin.