THE SLAB WINDOW, THE PLATE BOUNDARY, AND THE FAULTS: CONTROLS ON THE FORMATION AND EVOLUTION OF THE SAN ANDREAS

Since the 1970's it has been recognized that the Pacific - North America (San Andreas) plate boundary develops in response to migrating triple junctions. The San Andreas plate boundary in northern California localizes within a slab window that forms in the wake of the Mendocino triple junction (MTJ) passage. Two key details of the plate boundary formation remain enigmatic - (i) when and where does the lithospheric-scale plate boundary shear zone form, and (ii) how do upper-crustal faults connect to this lithospheric-scale plate boundary at depth? To address these questions, we combine geodetic observations, patterns of seismicity, and newly available high-resolution seismic tomography that defines the 3-D plate boundary structure, with constraints from plate kinematics. Geodetic data in northern California indicate that PAC-NAM plate motions are accommodated on a localized shear zone within < 30 km south of the southern edge of the subducted Gorda slab, implying the plate boundary structure forms < 1 million years after MTJ passage. This plate boundary shear zone lies approximately along the same small circle as the San Andreas fault in central California, and underlies the Hayward-Rodgers Creek-Maacama fault systems in northern California. The shear zone is offset to the east from the western edge of the North American crust, and its development appears to be controlled by the eastern edge of the Pioneer Fragment, a segment of the Pacific plate that extends eastward beneath the North American margin, immediately south of the MTJ. The plate boundary structure - including the slab window and Pioneer Fragment - is imaged in the tomography and indicates that the slab window that forms immediately south of the southern edge of the Gorda slab is narrower than previously inferred. The resulting plate boundary geometry throughout northern California after MTJ passage is similar to the 3-D structure imaged in the BASIX experiment in the San Francisco Bay area. One key implication of this is that the San Andreas Fault through northern California, that last ruptured in 1906, is not a lithospheric cutting structure, but rather likely connects to the main plate boundary to its east along a low-angle detachment, whose seismogenic potential is unclear.