Quantitative calibration of the lacustrine seismograph using sedimentary imprints of recent megathrust earthquakes in south-central Chile

Accurate seismic hazard assessment strongly relies on paleoseismic research, providing long timescales of past seismic shaking, hereby potentially also revealing earthquakes that were larger than the highest-magnitude earthquakes known from historical or instrumental records. In this respect, lakes often provide long and highly sensitive shaking records, and lacustrine paleoseismology has evolved into an invaluable methodology for reliable reconstructions of earthquake recurrence intervals around the world. A thorough understanding of the relation between the component(s) of strong ground motion (e.g., PGA, PGV, duration) and the resulting sedimentary signature is, however, still missing. As a result, characterization of the source parameters of paleo-earthquakes, such as magnitude and location, up to now relies solely on qualitative or semi-quantitative considerations. Previous studies have aimed at calibrating the lacustrine seismograph by attributing the occurrence or absence of coseismic imprints (e.g., turbidites) to intensities of seismic shaking. These intensity values are usually expressed on the macroseismic scale, as such information is more readily available compared to instrumental data for recent as well as historical earthquakes. These are, however, relatively subjective ground motion measures, unable to capture the various aspects of strong ground motion. In this study, we determine the relationship between quantitative ground motion measures on the bottom of a lake and the sedimentary shaking imprints identified therein. To achieve this, we focus on the sedimentary signature of instrumentally recorded megathrust earthquakes in south-central Chile. This includes the 1960 Mw 9.5 Valdivia earthquake and the 2010 Mw 8.8 Maule earthquake. A compilation of existing sedimentological data shows that coseismic deposits related to either of these events are identified in over 20 lakes. By linking the imprint characteristics for both earthquakes to the local ground motions, we bridge the gap between sedimentology and seismology, opening perspectives towards quantitative characterization of paleo-earthquakes.