Spanning the Arc: Margin Geometry and Topography Control Upper Plate Deformation in the Central Andean Subduction Zone

Subduction zone forearcs deform transiently and permanently due to the frictional coupling with the converging lower plate. Transient stresses are mostly the elastic response to the seismic cycle. Permanent deformation is evidenced by forearc topography, upper plate faulting, and earthquakes; its relation to the megathrust seismic cycle is debated. Here we study upper plate seismicity, interplate earthquake slip vectors, and the GNSS strain field in the northern Chile subduction zone to deduce the stress field and to separate elastic and permanent strain. We find that seismicity is distributed unevenly and that high seismicity rates concur both with a break in the forearc topography and tectonics of the Coastal Cordillera and the onset of a change in subduction obliqueness. Earthquakes in the South American crust under the sea and the Coastal Cordillera show a remarkably homogenous north-south, i.e., trench-parallel, compressional stress field. The trench-parallel compression above the plate coupling zone, almost perpendicular to plate convergence direction, may be explained by strain resulting from a change in subduction obliqueness due to the concave shape of the plate margin, which we demonstrate by investigating inter-plate earthquake slip vectors. From these, we derive a strain rate estimate (-510e-8 /a) and compare it to one derived from upper plate earthquakes (-810e-9 /a). We argue that the dominance of trench-parallel compressive stresses over trench-perpendicular ones is due to canceling of the latter by tensional gravitational stresses due to the topographic gradient between the Andes and the Nazca trench. Based on the distribution of the type of faulting we investigate the trench-perpendicular stress field with a force-balance model. The observed deep strike-slip earthquakes, expression of trench-perpendicular tension, require the deepest extent of the megathrust to be very weak.