ASEISMIC SLIP ON THE DEEP CASCADIA SUBDUCTION INTERFACE: CHARACTERIZATION AND IMPACTS

Detailed studies of data from continuous GPS sites along the northern Cascadia margin suggest that "silent slip" events on the deeper subduction interface may be more common than first suspected. An examination of similarities between observed events is a first step in understanding the processes involved. These similarities include the following aspects: 1) all surface displacements caused by slips are in a direction opposite to the long-term interseismic deformation; 2) maximum surface displacements (~ 6 mm to date) occur at sites located in the inner margin, attenuating rapidly to the east but more gradually to the west; 3) slip events appear to be centered around southeastern Vancouver Is. and the eastern Olympic Plateau, at times propagating northwest along the plate interface, but (apparently) not occurring to the south; 4) the duration of slip at any given site ranges from 5 to 15 days. These spatial and temporal aspects of transient surface displacement can be modeled by up to 3 cm of slip on the deep subducting slab interface, that propagates both updip and along strike at an equivalent rupture velocity of about 5 km/day, pinching out to zero as it approaches the updip seismogenic zone. It is not clear what limits the magnitude of the slips and their propagation velocity, but it is expected that a megathrust earthquake may begin with one such silent slip. The GPS time series also clearly show that in the period between slips, deformation velocities due to the locked seismogenic zone are augmented substantially at sites that show the largest slip displacements. This is consistent with transient coupling across the deeper interface, having stronger coupling which impedes plate-rate slip for a period of a year or two, followed by aseismic slip which releases this accumulated stress locally. With a denser continuous GPS network, it may become possible to define a "frequency-magnitude" relationship for these slip events, which will be key to evaluating the role of aseismic slip in the temporal redistribution of stress and the possible triggering of megathrust earthquakes.