Lake-sediment based paleoseismology: Limitations and perspectives from the Swiss Alps

In regions with moderate seismicity and large intervals between strong earthquakes, paleoseismological archives that exceed the historical and instrumental timescale are needed to establish reliable estimates of earthquake recurrence for long return periods. In several regions, lake sediments have shown to be suitable for paleoseismological studies by causally linking characteristic sedimentological features to historic earthquakes. Studies on single lakes, however, do neither allow determining the paleoepicentre nor the paleomagnitude for the potential paleoearthquakes. Here we compile, using shaking-induced mass movements and micro deformations (summarized as Sedimentary Event Deposits SEDs), the sedimentary paleoseismic record of 11 lakes from Switzerland over the last 10,000 years. The large dating uncertainty attributed to such deposits (up to 250 years) does not allow us to conclusively test for one large earthquake hypothesis when comparing the different lake records and therefore represents one of the major limitations of this approach. Instead, using a new approach of exploring the normalized frequency of occurrence averaged over a larger area, the compiled dataset reveals striking periods of enhanced occurrence of SEDs in the studied lakes during several phases of the past 10,000 years, centered at 9700, 6500 and during the last 4000 cal yr BP. Moreover, we use a calibrated intensity attenuation relation in order to model scenarios of possible epicentral areas and ranges of magnitudes of paleoearthquakes. We differentiate two cases: (i) a 'single-earthquake' scenario if SEDs occur simultaneously in various studied lakes, or (ii) a 'multi-earthquake' scenario if SEDs in the studied lakes cluster within a time interval. The modelled scenarios allow us to propose maximally possible magnitudes of large paleoearthquakes, constituting an important input for seismic hazard assessment in the Swiss Alps.