Investigating the location of the ~1620 C.E. West Coast earthquake using coseismic landslide modelling.
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
Palæoseismic indicators of large earthquakes within the Southern Alps of New Zealand are widespread and the methods by which they are obtained and interpreted vary greatly. The palæo-earthquake record of the Alpine Fault is a primary goal of many research programmes, so attribution of these indicators to a particular fault aids in understanding the past to better prepare for the future. An earthquake event around 1620 C.E., hitherto attributed to the Alpine Fault, does not appear to directly reflect these various forms of evidence; neither does it match the currently understood recurrence interval of Alpine Fault ruptures.
One primary result of large earthquakes is coseismic landsliding, which has the potential for damage and destruction of both the natural and built environments. Within the Southern Alps, topography and climatic factors exacerbate erosion rates following such events. In addition to the immediate effects, longer-term issues like river aggradation and avulsion can further hinder recovery. On the West Coast, substantial aggradation has been directly attributed to the ~1620 C.E. earthquake, and is considerably higher than modelled aggradation from an Alpine Fault event. Based on the incongruence between the palæoseismic indicators, inferred earthquake properties, and other models, a 7.6 MW earthquake on a hypothetical fault within the western Southern Alps was investigated to see if it better matched these effects.
A Shakemap model was developed, producing distributed shaking intensities for this hypothetical earthquake. The results were incorporated into a fuzzy logic based coseismic landslide susceptibility model and, using Monte Carlo analysis, landslide volumes, denudation depths, and aggradation depths in some order 5, and all order 6 and higher river catchments of the South Island were produced.
The results show that the smaller catchments in the western Southern Alps close to the hypothetical fault are significantly more impacted by coseismic landsliding associated with its rupture than in the case of an Alpine Fault earthquake. However, when the more northern and southern indicators of this event are compared, the hypothetical earthquake does not produce sufficiently intense impacts to match the ~1620 C.E. earthquake palæoseismic indicators. This implies that either a longer fault producing a larger, more geographically extensive earthquake or a secondary earthquake within the same timeframe is necessary.