Empirical Tsunami Hazard Assessment of Near-Field Plate-Boundary and Crustal Fault Sources Demonstrated for the Pacific Coast of Canada

Abstract:

We demonstrate an empirical method for preliminary tsunami hazard assessment of near-field fault sources that lack long historic records, paleoseismic/paleotsunami data and/or adequate tsunami modelling studies. Along the Pacific coast of Canada, the North America plate boundary is characterized by varying degrees of convergence with adjacent oceanic plates and microplates. The 1700 M~9 Cascadia earthquake ruptured at least the full extent of Juan de Fuca plate subduction as far north as central Vancouver Island; paleoseismic data show that similar events have occurred approximately every 500 years throughout the Holocene, accompanied by large tsunamis. Further north along the margin, the paleoseismic and paleotsunami histories of the Explorer, Winona, and Haida Gwaii segments of the margin are unknown. The Explorer plate is subducting beneath Vancouver Island at about half the rate of the Juan de Fuca plate; this locked segment may rupture independently or it may slip concurrently with the rest of the Cascadia subduction zone system to the south. The tsunamigenic potential of the Winona segment off northern Vancouver Island is poorly understood. The occurrence of the 2012 M7.8 thrust earthquake off southern Haida Gwaii confirmed the tsunamigenic nature of partitioned convergent slip on this dominantly transform margin segment. Parts of the coastline face additional tsunami hazard from submarine crustal faults. For potentially tsunamigenic faults with unknown history, we use (1) geophysical data to constrain fault rupture area, (2) empirical relations to estimate earthquake magnitude from the rupture area, (3) plate motion models and geodetic data to constrain convergence and thrust earthquake recurrence rates, and (4) empirical relations to estimate near-field tsunami runup at coastal sites, given distance from the rupture. The success of this approach is demonstrated by general agreement between expected and observed earthquake magnitude and near-field tsunami runup in the 2012 M7.8 Haida Gwaii event. However, this example also highlights that sites of extreme runup cannot be predicted without detailed tsunami modelling of a range of rupture scenarios. We also compare tsunami runup estimated using our approach with runup modelled for scenario ruptures of the Cascadia subduction zone.