R. Hartog et Sy. Schwartz, Subduction-induced strain in the upper mantle east of the Mendocino triplejunction, California, J GEO R-SOL, 105(B4), 2000, pp. 7909-7930
We observe splitting of teleseismic shear waves at five stations of the Ber
keley Digital Seismic Network located east of the Mendocino triple junction
in northeastern California that is dependent on the arrival direction of t
he seismic phases. The observed variations with back azimuth cannot be expl
ained with laterally varying anisotropy with a horizontal symmetry axis and
are attributed to the presence of fabric with an inclined symmetry axis. W
e assume that the anisotropy is caused by the preferred alignment of olivin
e crystals. A grid search over possible orientations of the olivine a axes
reveals that south of the Mendocino triple junction they dip to the east, w
hereas north of the triple junction they dip to the northeast. On the basis
of a comparison of the ray paths of our data to the spatial distribution o
f fast and slow P wave velocity anomalies in the upper mantle, we conclude
that the anisotropy is located within seismically slow regions and that the
directions are controlled by the geometry of a steeply dipping fast P wave
velocity anomaly. Assuming that the fast P velocity anomaly represents sub
ducted slab material, we conclude that the fabric beneath the stations nort
h of the triple junction is most likely caused by the differential motion b
etween this rigid, strong down going plate and the surrounding mantle. Sout
h of the triple junction the fabric may have developed while subduction of
the Farallon plate was still ongoing in this region (prior to 6 Ma). Howeve
r, we prefer to attribute the observations to more recent asthenospheric fl
ow associated with the opening of a slabless window beneath the North Ameri
can lithosphere. The flow is modulated by the presence of rigid lithosphere
to the north and east.