A preliminary tsunami hazard assessment of the Canadian Coastline

Abstract

The Canadian coastline is the longest of any country in the world, and is at risk from tsunamis generated in three oceans. The current state of knowledge precludes a complete probabilistic tsunami hazard assessment, which would require quantification of a wide range of possible scenarios for each tsunami source, coupled with modelling that incorporates fine-resolution bathymetry and onland topography to adequately assess potential runup at the coast. This preliminary assessment presents a first attempt to quantify the tsunami hazard on the Canadian Pacific, Atlantic and Arctic coastlines from local and far-field, earthquake and large landslide sources. For each source considered, we calculate the probability that tsunami runup at the coast will exceed 1.5 m (threshold for potential damage) and 3 m (significant damage potential), in a 50-year period. For each coastal region, we then combine the relative hazard from each source to calculate the overall probability that the coastline in question will experience tsunami runup exceeding 1.5 m (and 3 m) within a 50-year period, from any geological source. We also consider the maximum runup levels expected to occur within time periods of 100, 500, 1000, and 2500 years. Our assessment indicates that the overall tsunami hazard (runup ? 1.5 m) of the outer Pacific coastline (~40-80% probability of exceedance in 50 y) is an order of magnitude greater than that of the outer Atlantic coastline (~1-15%), which in turn is an order of magnitude greater than the Arctic coastline (< 1%). These probabilities are equivalent to an expected recurrence of runup exceeding 1.5 m of ~30-100 years for the outer Pacific coast, ~300-1700 for the Atlantic, and ~6500-17,000 years for the Arctic. For larger runup (? 3 m), the estimated Pacific hazard (~10-30% probability of exceedance in 50 y) is significantly larger than both the Atlantic (~1-5%) and the Arctic (< 1%). Equivalent recurrence intervals are ~150-500 years for the Pacific, ~650-4000 years for the Atlantic, and ~7000-20,000 years for the Arctic. On the outer Pacific coastline, the 1.5 m runup hazard is dominated by far-field subduction zone sources, whereas the more severe 3 m runup hazard is almost entirely contributed by local subduction zone sources. The Cascadia subduction zone presents the highest tsunami hazard to the Pacific coast, with the most extreme potential runup; potential thrust sources along the Explorer and Queen Charlotte margins contribute a significant proportion of the estimated tsunami hazard for the northern BC coastline. For the more sheltered inner Pacific coasts of Juan de Fuca and Georgia Straits, the hazard at both levels is contributed mostly by Cascadia subduction zone events. Tsunami hazard on the Atlantic coastline is dominated by far-field subduction zone sources, but this hazard is poorly constrained. Significant tsunami hazard is also provided by near-field continental slope failures similar to the 1929 Grand Banks event. Tsunami hazard on the Arctic coastline remains poorly constrained, but these regions are assumed to be sheltered from far-field tsunamis, so the hazard is provided by local sources. A hypothetical earthquake source beneath the Mackenzie delta requires further study. We discuss briefly but do not quantify the hazard of locally-damaging waves triggered by subaerial or submarine landslides, but we highlight susceptible areas. A probabilistic analysis of local landslide tsunamis would require (1) the identification of potential sources; (2) evidence for past tsunamigenic events to establish frequency-size relationships and/or slope stability analyses that incorporate expected earthquake shaking levels; (3) probabilistic tsunami modelling of a wide range of possible failures.