Description: The 27 February, 2010 Maule earthquake (Mw=8.8) ruptured ~400 km of the Nazca-South America plate boundary and caused hundreds of fatalities and billions of dollars in material losses. Here we present constraints on the fore-arc structure and subduction zone of the rupture area derived from seismic refraction and wide-angle data. The results show a wedge shaped body ~40 km wide with typical sedimentary velocities interpreted as a frontal accretionary prism (FAP). Landward of the imaged FAP, the velocity model shows an abrupt velocity-contrast, suggesting a lithological change which is interpreted as the contact between the FAP and the paleo accretionary prism (backstop). The backstop location is coincident with the seaward limit of the aftershocks, defining the updip limit of the co-seismic rupture and seismogenic zone. Furthermore, the seaward limit of the aftershocks coincides with the location of the shelf break in the entire earthquake rupture area (33°S–38.5°S), which is interpreted as the location of the backstop along the margin. Published seismic profiles at the northern and southern limit of the rupture area also show the presence of a strong horizontal velocity gradient seismic backstop at a distance of ~30 km from the deformation front. The seismic wide-angle reflections from the top of the subducting oceanic crust constrain the location of the plate boundary offshore, dipping at ~10°. The projection of the epicenter of the Maule earthquake onto our derived interplate boundary yielded a hypocenter around 20 km depth, this implies that this earthquake nucleated somewhere in the middle of the seismogenic zone, neither at its updip nor at its downdip limit.

Description: The south-central Chilean subduction zone has witnessed some of the largest earthquakes in history, making this region particularly important for understanding plate coupling. Here we present the results of a local earthquake tomography study from a temporary local seismic network in the Villarrica region between 39 and 40°S, where the largest coseismic displacement of the 1960 Valdivia earthquake occurred. A lowvelocity anomaly and high Vp/Vs values occur under the coastal region, indicating mantle serpentinisation and/or underthrusting of forearc material. Further east, a high-velocity anomaly is observed, interpreted as “normal” high-velocity mantle. Under the active volcanic arc a low-velocity anomaly together with high Vp/Vs ratios (1.8 and higher) likely images fluid ascent beneath the volcanoes. Close to the subducting Valdivia Fracture Zone, the coastal low-velocity anomaly extends further inland, where it interrupts and shifts the high-velocity anomalies associated with “normal” fast mantle velocities. This may indicate enhanced fluid presence along this part of the margin, probably caused by a stronger hydration of the incoming plate along the Valdivia Fracture Zone. This is consistent with geochemical fluid proxies (U/Th, Pb/Ce, Ba/Nb) in young volcanic rocks displaying peak values along the volcanic front at Llaima and Villarrica Volcanoes, and with recent GPS measurements, which suggested a local reduction in plate coupling in this region. The shift in the high-velocity anomaly underlying the central part may be caused by a north to south decrease in plate age and hydration across the Valdivia Fracture Zone, and may explain why a Central Valley is absent in this segment of the margin. The low La/Yb ratios in the volcanic rocks from Villarrica and Llaima suggest that the high slab-derived fluid flux causes elevated degrees of melting beneath these volcanoes, providing an explanation as to why these are amongst the most active volcanoes in South America.