Active Tectonics and Geomorphology of the central South Island, New Zealand: Earthquake Hazards of Reverse Faults

摘要

Oblique continental collision between the Pacific and Australian Plates in the central South Island of New Zealand (between c. 44 and 46°S) results in distributed reverse faulting. Only a few of these faults have been studied in detail, highlighting a major knowledge deficit in the earthquake behaviour, magnitude potential and contribution to seismic hazard for many faults in this part of the orogen. Three reverse faults are investigated in detail in this thesis: the Moonlight Fault Zone (MFZ), the Fox Peak Fault and the Forest Creek Fault. Geochronologic approaches, including Schmidt hammer exposure-age dating, radiocarbon dating, and optically stimulated luminescence dating, are combined with paleoseismic trenching, fault surface trace mapping, analysis of GPS and LiDAR survey data, and numerical modelling to characterise the rupture behaviour of these faults. A new Schmidt hammer chronofunction based on over 7000 clast analyses is developed that relates rebound value (R-value) to age for river terraces. The rapid, inexpensive, non-destructive, and statistically valid nature of this technique makes it widely applicable for age dating here and globally. I use Schmidt hammer exposure-age dating along with other geochronologic and surveying methods to show that stranded post-last glacial lake shorelines of Lake Wakatipu are undeformed and at a uniform elevation across the MFZ. This indicates an absence of uplift across the MFZ since c. 13 ka and suggests that this fault may be inactive or subject to long periods of interseismic quiescence despite its location in the active orogen. This result also challenges the long-held hypothesis that lake shorelines throughout central NZ are tilted due to isostatic rebound. Three segments of the Fox Peak Fault are identified through field mapping and surveying. Slip rates at over 50 locations along the 36.5 km total length of the fault (c. 1.5 mm yr⁻¹ maximum) co-vary with the bounding range topography and exhibit large gradients near intersecting NW-striking faults. Four paleoseismic trenches were excavated to determine if these segment boundaries represent barriers to earthquake rupture propagation. Evidence of 3-4 earthquakes since c. 16 ka on the two end segments with overlapping age uncertainties indicates that the recurrence interval of the fault is 2000-3000 years. The most recent event (MRE) occurred at c. 2.5 ka. Large single event displacement to length ratios on these segments and a single event scarp on the central segment indicate that while the segment boundaries control on-fault slip gradients, they are not likely to impede through-going ruptures in an earthquake. This is a relatively recent development from the long-term tectonic geomorphology, which is suggestive of range growth on separate faults.