Lacustrine Paleoseismometers Reveal Spatial and Temporal Patterns of Rupture during the Last Ten Large Earthquakes on the Alpine Fault, New Zealand.

Abstract:

The rarity of long, well-dated paleoseismic records from sites along plate boundary transform faults is a major constraint on the development and evaluation of conceptual models of fault rupture behaviour. This is the case for the 800 km long, high slip rate (27±5 mm yr^-1), dextral strike-slip Alpine Fault at the boundary between the Pacific and Australian plates in southern New Zealand. We use lacustrine paleoseismology to evaluate the hypothesis that the Alpine Fault exhibits self-similar behaviour, that is, the fault always produces earthquakes at or near a maximum magnitude of M_w8. The hypothesis is tested using reconstructions of high intensity shaking from five lakes situated along 240 km of the Alpine Fault’s Central section. Sedimentological investigation of lake cores shows that high intensity shaking events are recorded in the lake sediments as turbidites formed by subaqueous slumping. These turbidites are overlain by terrigenous sediment from co- and post-seismic landsliding on hillslopes in the lakes’ mountainous catchments. Chronologies derived from Bayesian modelling of AMS ¹⁴C dates on terrestrial leaf macrofossils precisely constrain the timing of earthquakes at the lake sites, facilitating along-strike correlation. Shaking events correlate between the sites and with known ruptures of the Alpine Fault, confirming the seismic origin of the deposits and allowing thresholds of shaking intensity for deposit formation to be determined using isoseismal modelling. Modelled shaking intensities for the last two Alpine Fault earthquakes show that subaqueous slumping occurs when shaking intensities exceed Modified Mercalli scale (MM) VI-VII, and that increased fluvial sediment fluxes from earthquake-induced landslides occur when shaking intensities exceed MM IX. The record of synchronous MM IX shaking at the lake sites provides first order constraint on the rupture length of the last ten earthquakes on the central Alpine Fault. Rupture scenarios for these earthquakes are augmented by correlating event timing with long earthquake records from the South Westland section and geomorphic reconstructions of the slip distribution for the most recent ruptures to explore the best-fit model of fault behaviour and to test the hypothesis that the Alpine Fault always ruptures in great (M_w8) earthquakes.