PROGRESS IN UNDERSTANDING THE GREAT EARTHQUAKE THREAT IN THE PACIFIC NORTHWEST

Since the pioneering work of Brian Atwater, Tom Heaton, and Hiroo Kanamori in the 1980s, much progress has been made in understanding the seismotectonic framework of the Cascadia subduction zone. Stratigraphic studies in coastal marshes and documentation of turbidite sequences on the northeast Pacific Ocean floor indicate that there have been 20 great subduction earthquakes of the entire Cascadia margin over the past 10,000 years. However, there remain important questions that require attention. First, the deep-sea turbidite record suggests that the aforementioned 20 earthquakes have resulted from rupture of most of the 1100 km length of the plate boundary and, accordingly, are M9+ events. This rupture behavior is unique – all historic great earthquakes at other subduction zones around the Pacific Ocean have involved segmented rupture – and also unexpected because of (1) the presence of microplates at both tends of the subduction zone, (2) the ~30^° change in the trend of the plate boundary north of Washington state, and (3) clustering in times of the past 20 events, with return periods ranging from about 100 years or less to more than 1000 years. Second, differences between the coastal marsh and turbidite records have not yet been reconciled, although they may be more apparent that real given the inevitable uncertainties in the records. Third, the tectonic implications of the northern (Explorer) and southern (Gorda) microplates are not fully understood. Both are thought to be subducting beneath North American crust and to be capable of independent rupture during great Cascadia events. The two microplates, however, are adjacent to triple junctions and are experiencing complex internal deformation due to stresses exerted by adjacent transform faults. Third, the degree of coupling of the North American and Juan de Fuca plates remains uncertain, with important implications for seismic risk in the region – active crustal faults that might slip during a subduction earthquake occur in or near the large cities of the region. Plate coupling also carries implications for the aftershock behaviour. Lastly, the inboard extent and depth of the locked portion of the plate boundary are not yet firmly established.