The Structure of the Chilean Subduction Zone from Seismic Imaging and Tomography at 34.5°S

The south-central Chilean seismogenic zone has produced some of the largest megathrust earthquakes ever recorded, including the 2010 Mw 8.8 Maule event. To understand the rupture behaviour of this earthquake in the region of maximum coseismic slip at 34.5°S, we analyse the tectonic and elastic structure of the margin using 2D wide-angle seismic (WAS) data, and spatially coincident 2D multichannel (MCS) data acquired with a 15-km-long streamer.

To improve the seismic velocity model relative to previous results along the same WAS line, we jointly invert travel times from WAS and MCS data using a combined refraction–reflection tomographic approach and statistical uncertainty analysis. In addition, we apply downward continuation to the MCS shot gathers to increase the number of usable MCS travel times and to improve ray coverage with refracted arrivals from the shallow part of the velocity model. This approach enhances coverage and reduces tomographic velocity uncertainties, and improves constraints on the position of the interplate reflector from the megathrust.

The resulting 2D P-wave velocity (Vp) model includes the velocity structure of a 50-60 km-wide accretionary prism, and a sharp velocity transition into crystalline basement landward. We convert the velocity structure of the upper plate into density and S-wave seismic velocity to then calculate rigidity (Shear modulus), and infer dynamic rupture parameters such as slip and rupture velocity. Comparison of the expected slip distribution from our results with existing kinematic slip models shows significant discrepancies, particularly beneath the accretionary prism, where the time-migrated 127-km-long seismic profile reveals intense internal deformation and increasing thrust faulting and folding towards the trench. We discuss potential upper-plate coseismic deformation processes to explain such discrepancy.