Abstract
New Zealand’s Alpine Fault is a major crustal-scale structure that accommodates around 75% of relative Australian-Pacific plate motion. This active transform fault ruptures in a large earthquake every 291 ± 23 years, most recently in 1717. This means, the Alpine Fault is late in its seismic cycle and provides an opportunity to investigate earthquake nucleation processes. Herein I describe studies of the potential roles of graphite and porosity in weakening the Alpine Fault based on analyses of samples recovered from boreholes drilled during the two phases of the Deep Fault Drilling Project (DFDP).
In the Alpine Schist, highly crystalline graphite is mostly found on grain boundaries in quartzo-feldspathic domains. Mylonitisation within the Alpine Fault zone caused reorganization of this carbonaceous material, resulting in clustered (in protomylonites) and foliation aligned (in mylonites) graphite. In cataclasites derived from the mylonitised schists, graphite is more abundant and has two different habits: inherited mylonitic graphite and less mature patches of concentrated graphitic carbon. The occurrence of this structurally different graphite can be explained either by structural disorder of graphite due to deformation, or hydrothermal addition of carbonaceous material in the mid-shallow crust that then matured to a lesser degree.
To examine the possibility that mechanical processes resulted in modifications of graphite structure, highly crystalline graphite powder was sheared at normal stresses of 5 and 25 MPa with aseismic sliding velocities of 1 µm/s, 10 µm/s and 100 µm/s – insufficient to cause significant frictional heating. Raman spectrometry revealed that graphite aggregates became structurally disordered compared to the starting material as a function of increasing shear strain. SEM and TEM observations indicate this was accommodated by brittle processes. These findings compromise the validity of the calibrated graphite ‘thermometer’ in active tectonic settings.
The hypothesis of hydrothermal addition of graphite was examined by combining total organic carbon and stable carbon isotope analyses. Graphite enrichment in the cataclasites (≤ 2.74 wt.%) in comparison to the schist and mylonite rocks (≤ 0.52 wt.%) was clearly demonstrated and attributed to hydrothermal addition of graphite within strained areas. These data and interpretation suggest that: (1) mechanical disorder of the graphite structure possibly occurred; (2) the measured graphite content may be sufficient to weaken the fault during seismic slip.
Porosity analyses from synchrotron CT scans and TEM images of rocks from the fault core show grain boundary pores with total volumes of 0.016 to 0.029%. These porosity data suggest that: (1) porosity reduction due to gouge compaction and pressure solution processes took place; and (2) elevated pore fluid pressures were present to prevent pore collapse. It was concluded that porosity distribution of the fault core may affect the mechanical behaviour of the Alpine Fault due to pore fluid pressure evolution related to post-seismic healing processes and precipitation of weak mineral phases (i.e. graphite) within open pore spaces.
The findings in this thesis suggest that the linked processes of graphite enrichment in the cataclasite matrix and evolution of the state of porosity in rocks from the fault core may stimulate future seismic activity along the active Alpine Fault of New Zealand.