Structural heterogeneity of the midcrust adjacent to the central Alpine Fault, New Zealand: Inferences from seismic tomography and seismicity between Harihari and Ross

Determining the rates and distributions of microseismicity near major faults at different points in the seismic cycle is a crucial step toward understanding plate boundary seismogenesis. We analyze data from temporary seismic arrays spanning the central section of the Alpine Fault, New Zealand, using double-difference seismic tomography. This portion of the fault last ruptured in a large earthquake in 1717 AD and is now late in its typical 330 year cycle of Mw approximate to 8 earthquakes. Seismicity varies systematically with distance from the Alpine Fault: (1) directly beneath the fault trace, earthquakes are sparse and largely confined to the footwall at depths of 4-11 km; (2) at distances of 0-9 km southeast of the trace, seismicity is similarly sparse and shallower than 8 km; (3) at distances of 9-20 km southeast of the fault trace, earthquakes are much more prevalent and shallower than 7 km. Hypocenter lineations here are subparallel to faults mapped near the Main Divide of the Southern Alps, confirming that those faults are active. The region of enhanced seismicity is associated with the highest topography and a high-velocity tongue doming at 3-5 km depth. The low-seismicity zone adjacent to the Alpine Fault trace is associated with Vp and Vs values at midcrustal depths about 8 and 6% lower than further southeast. We interpret lateral variations in seismicity rate to reflect patterns of horizontal strain rate superimposed on heterogeneous crustal structure, and the variations in seismicity cutoff depth to be controlled by temperature and permeability structure variations.