T52A-03:
Lacustrine Turbidites as a Tool for Quantitative Paleoseismology: Evidence from 17 Chilean Lakes

Friday, 19 December 2014: 10:50 AM
Maarten E Van Daele1, Jasper Moernaut2, Lindsey Doom1, Evelien Boes1, Katrien Heirman3, Willem Vandoorne1, Karen Fontijn4, Sabine Schmidt5, Michael Strasser6, Mario Pino2, Roberto Urrutia7 and Marc A O De Batist1, (1)Ghent University, Renard Centre of Marine Geology, Ghent, Belgium, (2)Universidad Austral de Chile, Escuela de Geología, Instituto de Ciencias Ambientales & Evolutivas, Valdivia, Chile, (3)Geological Survey of Denmark and Greenland, Department of Geophysics, Copenhagen, Denmark, (4)University of Oxford, Oxford, United Kingdom, (5)EPOC Environnements et Paléoenvironnements Océaniques et Continentaux, Centre National de la Recherche Scientifique, Talence Cedex, France, (6)ETH Swiss Federal Institute of Technology Zurich, Zurich, Switzerland, (7)Universidad de Concepcion, Concepcion, Chile
Abstract:
One of the remaining challenges in paleoseismological research is quantification of ancient-earthquake magnitudes, and rupture locations and extent. Such reconstructions require multiple continuous paleoseismic records with excellent temporal and spatial resolution and stable threshold conditions, which can be offered by turbidite records in lakes. Recent, well-documented, great megathrust earthquakes provide an ideal opportunity to calibrate the turbidite record in different types of lakes, and determine which settings are suitable for quantitative paleoseismology. We study the sedimentary impact of the 1960 and 2010 Chilean megathrust earthquakes using 107 short sediment cores in 17 lakes. A combination of image analysis, magnetic susceptibility and grain-size analysis allowed to distinguish two types of lacustrine turbidites (i.e. LT1s and LT2s) from background sediments and other seismically-induced event deposits. LT1s have a composition similar to the hemipelagic background sediments and result from shallow mass wasting on sublacustrine slopes covered by hemipelagic sediments. LT2s have a composition different from the background sediments and can result from delta collapses, floods etc. Comparison of the frequency of all seismically-induced event deposits with local seismic intensities shows that turbidites, and especially LT1s, are most reliable for quantitative paleoseismology. LT1s are almost exclusively triggered by earthquake shaking and they most sensitively record varying seismic intensities. Over all studied lakes, the number of locations with LT1s linearly increases with increasing seismic intensity, starting with no LT1s at intensities just below VI and reaching 100% when intensities are higher than VII½. In 3 lakes that have been studied in higher detail, we found that also a turbidite’s spatial extent and thickness within a certain basin are a function of the local seismic intensity. Hence, these records allow determining the seismic intensity of ancient earthquakes at each lake. Consequently, our multilake turbidite record aids in pinpointing magnitudes, rupture locations, and extent of past subduction earthquakes in South-Central Chile. The consistent results in 17 lakes indicate that many more lacustrine basins worldwide have similar potential.