Direct dating of active faults using luminescence: A case of study in New Zealand

The Alpine Fault in New Zealand is one of the world’s major active crustal-scale faults. It builds the boundary between the Pacific and the Australian plate, and branches into strike-slip faults known as the Marlborough fault system. The northeastern region of the southern island of New Zealand has a historical record of large, shallow earthquakes with magnitude (Mw > 6.5) since the Nineteenth Century. The most recent event, the 2016 Kaikōura earthquake, with a magnitude (Mw) of 7.8, is among the strongest. The severe impact on society and landscape explain the importance of a better understanding of the Quaternary activity of these active faults. Recent investigations in other tectonically-active settings worldwide indicate the potential feasibility of applying luminescence dating to unravel the timing and hence, re-occurrence of fault activity as a source for earthquakes. We aim to test the potential of luminescence dating to determine the relative activity of three active faults in New Zealand. To this end, we collected four dark-gray, fine to very fine grain-size samples classified as cataclasite and gouge from outcrops situated along the fault traces of the Alpine Fault, Hope Fault, and Hundalee Fault. Through sample processing, we obtained polymineral fine grains, ranging from 4 to 11 µm, to conduct post-infrared infrared stimulated luminescence (pIRIR225) dating. The method applied on faults is the signal-resetting event of the fault movement, and if the signal is not saturated in nature, this implies that frictional heating was enough to at least partially reset the system; for feldspar the closure temperature is 40-90 °C. The growth curves of the pIRIR225 signals reveal that the gouge samples extracted from the Hope fault and Hundalee fault approach saturation levels with equivalent doses around 850 Gy and 900 Gy, respectively. In contrast, the equivalent dose of cataclasite samples from the Alpine Fault was clearly below saturation ranging from 220 Gy to 410 Gy. All assessment criteria, including recycling ratio and recuperation rate, meet the rejection criteria for all samples, indicating a reliable signal to dose relationship. These results suggest there were events, which thermally eroded the pIRIR225 signal at the Alpine Fault. The comparison of the equivalent doses from the three faults also indicates that the method is applicable to evaluate the relative fault activity; the Alpine Fault is more active than the Hope and Hundelee faults. However, micro-structural analysis also indicated differences in brittle deformation mechanisms, differences that also potentially influence the variations between the pIRIR225 signals of individual samples. Observed features comprise, for example, grain fracturing, frictional sliding, pressure solution, and twinning. The micro-structural variation suggest differences in deformation, stress and pressure-temperature (P-T) conditions experienced by the studied cataclasite and gouge samples. This study presents the first findings of pIRIR225 dating on feldspar in active faults in New Zealand and points at the success of luminescence dating. However, we strongly emphasize the importance of combining luminescence analysis with microstructural and mineralogical data obtained through Scanning Electron Microscope (SEM)-Mineral Liberation Analysis (MLA) to better understand the P-T conditions and resulting degree of luminescence signal reset.