LAUSR

We analyze shallow (0–20 km) microseismicity adjacent to the Alpine Fault in New Zealand, where there is oblique convergence of the Australian and Pacific plates. Focal mechanisms for 155 earthquakes (June 2012 - October 2013) are inverted to determine the orientation of the stress field. This yields a principal horizontal axis of compression, SHmax = 114° ± 10°, which cannot be explained in terms of the sum of stress from tectonic loading due to plate convergence, indicated by GPS observations, and gravitational stresses. The azimuth of slip vectors for individual focal mechanisms cluster perpendicular and parallel to the plate convergence vector. These faults, however, strike at ~45° to SHmax from the stress inversion, suggesting a very low coefficient of friction. The earthquake slip directions may be kinematically controlled, accommodating the plate convergence on a limited set of fractures, similar to the segmentation for neotectonic faulting along the Alpine Fault, which is partitioned into strike-slip and thrust segments at a 1–10 km scale. We suggest two possible controls on our calculated SHmax azimuths. Firstly, there may be a slight clockwise bias in the estimates of SHmax from earthquakes; slip may be occurring on a more limited range of fractures than assumed by the stress inversion method, although this effect is likely to be relatively small (±5°). More importantly, the components of the stress field may be relieved at different time scales during big earthquakes, resulting in a residual stress field that varies significantly (±15°) on timescales of several large earthquakes.