Biostratigraphic constraints on megathrust earthquake deformation history in south central Chile

Abstract

A lack of comprehensive understanding of the seismic hazards associated with a subduction zone can lead to inadequate anticipation of earthquakes and tsunami magnitudes. Four hundred and fifty years of Chilean historical documents record the effects of numerous great earthquakes; however, with recurrence intervals between the largest megathrust earthquakes approaching 300 years, seismic hazard assessment requires longer chronologies. This thesis seeks to verify and extend historical records in south central Chile using a relative sea-level approach to palaeoseismology, developed in Alaska and the Pacific Northwest.

Quantitative, diatom-based approaches to relative sea-level reconstruction are successful in reconstructing the magnitude of coseismic deformation during recent, well documented Chilean earthquakes. Disparities between my estimates and independent data highlight the possibility of shaking-induced sediment consolidation of tidal marshes. Following this encouraging confirmation of the approach, I quantify land-level changes in longer sedimentary records from the centre of the 1960 rupture zone. Here, laterally extensive marsh soils abruptly overlain by low intertidal sediments attest to the occurrence of four megathrust earthquakes. Field sites preserve evidence of the 1960 and 1575 earthquakes and Bayesian age-depth modelling constrains the timing of two predecessors to 1270 to 1410 and 1050 to 1200. The sediments and biostratigraphy lack evidence for the historically documented 1737 and 1837 earthquakes. The distribution of documented effects of these ruptures and the new palaeoseismic data presented in this thesis suggests these earthquakes were smaller in magnitude and located in the southern portion of the 1960 rupture segment, as other authors have previously inferred.

Coastal sediments record relative sea-level changes reflecting both the earthquake deformation cycle and non-seismic processes. The 1000 year record of net relative sea-level rise implied by the new records presented here differs from the mid to late Holocene relative sea-level fall inferred from previous field studies and modelling approaches.