THE 2001 MW 6.8 NISQUALLY INTRASLAB EARTHQUAKE AND ITS AFTERSHOCKS

We investigate the rupture history of the February 28, 2001 M 6.8 Nisqually earthquake using three-component strong-motion seismograms integrated to particle velocity and low-pass filtered at 0.25 Hz. These data are inverted for slip history assuming slip occurred along one of the two potential fault planes determined by the Harvard Centroid Moment Tensor. Complete theoretical Green’s functions are calculated for 1-D models. The 1-D velocity model for each station is the appropriate path average through our high-resolution 3-D tomography model. Seismograms from 20 stations within 100 km of the epicenter can generally be fit very well. The data are fit slightly better assuming rupture occurred along the nearly horizontal plane than along the nearly vertical plane. In either case the rupture propagated primarily to the north along the strike of the slab. In addition, the main shock and its 5 aftershocks (M 1.0 to 4.3) are relocated relative to each other by applying the double-difference method to absolute P and S picks as well as more accurate relative times determined using cross correlations of 100 pairs of P waves and 20 S-wave pairs. Each pair represents two earthquakes recorded at a common station. These relocated events do not clearly delineate the correct fault plane, but provide a slight preference for rupture on the near horizontal plane. Thus, it appears that this earthquake ruptured on the plane that is more nearly parallel to the slab dip, and that the fault is elongated along strike. The hypocenter is at the base of the 5-km thick zone of background seismicity. These observations are consistent with the hypothesis that the Nisqually earthquake nucleated near the Moho of the subducted oceanic lithosphere and ruptured up through the subducted oceanic crust. If so, all these earthquakes may result from dehydration embrittlement associated with the basalt to eclogite phase transition. Being elongated along strike and on a fault sub-parallel to the slab dip, this earthquake can just fit geometrically within a 5-km thick crust. A significantly larger earthquake (M8) would have to rupture through a significant thickness of the subducted mantle lithosphere as well.