KEYNOTE: THE OPENING OF THE GULF OF CALIFORNIA: TEMPORAL VS SPATIAL VARIATIONS IN RIFT KINEMATICS

Plate boundaries commonly initiate along thermal or mechanical weaknesses that are oblique to the direction of relative plate motion. This inherent obliquity requires the strain of plate margin shearing to be three dimensional with finite changes in length occurring on all three principal strain axes. Faults with similar orientations and slip directions can only accommodate two-dimensional strain and therefore most plate margins are composed of a geometrically complex network of faults. Faults are commonly classified into sets with similar orientations and kinematics and as a rule these sets are heterogeneously distributed throughout the deformation belt. Such spatial partitioning of strain into discrete faults produces many potential pitfalls in the interpretation of the kinematic history of shearing across plate margins.

The Gulf of California is a crustal scale dextral shear zone that is extending perpendicular to its margins and this transtensional shearing produces strongly constrictional 3D strain. Kinematic interpretations of faulting in the Gulf of California extensional province have been a source of controversy for more than a decade and many workers are beginning to question the original hypothesis of temporally discrete phases of rifting with different directions of relative plate motion, which is based entirely on inversions of fault kinematic data that were interpreted to show changes in paleostress.

Kinematic characterization of modern seismicity and Quaternary faults demonstrates how transtensional shearing is accommodated by a network of faults that violate many of the basic assumptions paleostress inversions. Single earthquakes commonly activate faults a continuous and broad spectrum of orientations that defy simple classification schemes based on end member slip models. Ruptures from a single earthquake commonly propagate through a series of faults that intersect at high angles and conventional analysis of cross-cutting relationships cannot be used to infer relative timing between fault sets. As faults rotate with progressive shear, the direction of shear stress on the plane changes and thus faults with multiple striae are expected in a single kinematic regime and do not constitute strong evidence for changes in the orientation of paleostress.