Strain partitioning and interplate friction in oblique subduction zones: Constraints provided by experimental modeling

Physical modeling of oblique subduction is performed to study the mechanism of strain partitioning. The model is two-layer and includes the elasto-pla stic lithosphere (the overriding and subducting plates) and the low-viscosi ty liquid asthenosphere. The subduction is driven by a push force from a pi ston and a pull force when the density contrast Delta(rho) between the subd ucting plate and the asthenosphere is positive. We vary both Delta rho and the interplate friction (frictional stresses). Slip partitioning is obtaine d only in the models with high interplate friction and only when the overri ding plate contains a weak zone. This zone in the models corresponds either to locally thinned lithosphere or to cut (fault). The horizontal, trench-n ormal component of the interplate friction force \F-fh\ can be comparable w ith the absolute value of the horizontal component of the nonhydrostatic in terplate pressure force \F-Ph\ in the subduction zone. F-fh is always negat ive (compression), while F-Ph can be either negative (compressional subduct ion regime) or positive (extensional regime). High friction, which promotes partitioning, can completely cancel the extensional (suction) force F-Ph B ack are tension and strike-slip faulting appear thus as conflicting process es, although they can coexist in the same subduction zone, depending on the relative values of relevant forces. It appears that high friction can exis t only in compressional subduction zones where partitioning should develop more easily. This conclusion is supported by the comparison of two oblique subduction zones, having similar geometry: the compressional southern Kuril e zone (strong partitioning) and extensional southern Ryukyu zone (no litho spheric-scale partitioning). Other factors controlling the strain partition ing are the length of the oblique subduction zone, the boundary conditions at the transverse limits of the forearc sliver, and of course, the obliquit y of subduction.