DEFINING HISTORICAL EARTHQUAKE RUPTURE PARAMETERS AND PROPOSED SLIP DISTRIBUTIONS THROUGH TSUNAMI MODELING IN SOUTH-CENTRAL CHILE

South-central Chile has experienced at least 17 tsunamigenic earthquakes from 1570 to 1960 between 46°S and 31°S. Because most of these events are pre-instrumental, information regarding the distribution of high slip is unknown. We aim to uncover these previously undocumented slip distributions by using historic and geologic tsunami records as a benchmark to compare against tsunami simulations that are derived from finite-slip earthquake source models. We ran 424 tsunamis simulations to generate waveform outputs (using GeoClaw) at 47 coastal sites in south-central Chile. Inputted earthquake scenarios ranged from M_w 8.6 to M_w 9.4 with stochastic slip distributions along the Chilean Subduction Zone. We compared the waveforms of all tsunami simulations against observations within the historical record using Akaike Information Criteria (AIC) statistical models. With AIC, we propose a suite of probable earthquake scenarios that best describe these 17 pre-instrumental earthquakes, e.g. 1960, 1835, and 1730 events. The AIC statistical modeling application identifies the most-likely earthquake parameters by matching simulated wave height outputs with recordings in the written and geologic record. Our AIC analyses conclude that the 1960 event experienced a M_w 9.2-M_w 9.4 rupture with high concentration of slip (~30 m) at ~ 38-40ºS; the 1835 event experienced either a M_w 9.2_- 9.4 rupture with max slip at ~ 39ºS or a M_w 8.6 rupture with maximum slip at ~ 36ºS; and the 1730 event experienced a M_w 9.2 rupture with maximum, shallow slip at ~ 36ºS. These results agree with the previously estimated magnitudes of these events within the literature. We produced high resolution maps at three coastal sites with geologic evidence to simulate the inundation dynamics of different tsunami scenarios. We also evaluated the relationship between seafloor deformation and total water displacement. These supplemental techniques allowed for further refinement of potential rupture scenarios. Defining the slip distribution of historical earthquakes may give clues to the temporal and spatial variability of locking zones. This information may be useful for predicting future near-field tsunami wave heights for particularly vulnerable coastal regions.