SOURCE PARAMETER SELECTION FOR PROBABILISTIC TSUNAMI HAZARD ANALYSIS IN EASTERN CANADA

Unlike probabilistic seismic hazard analysis (PSHA), there is not a well-established methodology for probabilistic tsunami hazard analysis (PTHA). The PTHA methodology is similar to the widely used PSHA methodology for ground motion, and incorporates aleatory and epistemic uncertainty in calculating the probability of exceeding runup and drawdown values produced by tsunamigenic sources considered in the source model. Evaluating tsunami hazard becomes more difficult in locations such as the eastern coastline of Canada with low tsunami recurrence rates and few historical examples. In this study, we evaluated sources and assigned rates for local and far field earthquake and landslide sources for a site on the Bay of Fundy in New Brunswick, Canada. These sources included local faults, the Puerto Rico trench, fault sources linked to the 1755 Lisbon earthquake in the Azores-Gibraltar plate boundary region, and landslides on the continental slope and the Canary Islands. Fault source and Canary Islands landslide recurrence and size is based on literature review and compilation of available data. Two methods are evaluated for continental slope landslide recurrence. One relates the occurrence and size of mapped landslides to the frequency of landslides observed in seismic stratigraphy. The second method estimates the frequency of continental slope failures based on the earthquake catalog for the region, assuming that landslides are triggered by local earthquakes. The earthquake catalog based method had significantly lower recurrence rates of landslides than the than the method based on mapped occurrence. These results affirm that in regions with low tsunami recurrence, source selection methodology can have substantial effects on calculated hazard.