Intermediate deep earthquakes from flat to normal sections in Chile-Argentina Nazca plate

SPAGNOTTO SILVANA LIZ; ENRIQUE TRIEP; SILVINA NACIF; RENZO FURLANI

Foz do Iguazzu

Congreso; 2010 Meeting of the Americas; 2010

AGU

We report teleseismic waveform modeling focal mechanisms and slip distributions for 7 earthquakes, depth>100km, 31°-34°S. All mechanisms are normal, 4 pure dip-slips, 1 with strike-slip component, and 2 almost pure strike-slips. Generally, at least the strike of one nodal plane closely correlates with corresponding constant depth contours lines of the slab. Two pure normal-dip earthquakes are not too far of the flat slab western end and have similar mechanisms: one nodal plane dipping 35° southeast and the other 56° northeast. It is difficult to choose one of them as the fault plane, because the combination of the slab negative buoyancy, the deformation suffered nearby where the slab had been not horizontal and the possible inheritance of faults from the outer ridge. The earthquake in the normal section of the slab, just below active volcanoes, has almost pure dip-slip mechanism with a plane in the down-dip direction. The 3 earthquakes with strike slip component are associated to bends and/or contortions of the plate. Two additional local P-wave first motion mechanisms, Md ~4.5 and ~3.9, at depths ~217 and ~199km, are located within the pronounced slab bend at the center of the transition zone. Mechanisms are strike-slip with thrust component, the P axis are consistent to the bending. The normal dip mechanism with a little strike-slip at ~33.6°S, ~68.9°W, 160km depth, near the boundary between the transition and normal zones, behaves differently to all the others. The nodal plane strikes are near perpendicular to contour lines. The nodal plane dipping to the southeast could be evidence of some breaking between the two zones. A similar CMT mechanism, up-dip and towards northeast, could reinforce that interpretation, which also agree with the proposed geological processes that suggest slab steepening south of ~33.5°S. We conclude that slab pull and/or slab lateral distortion determine the occurrence of the analyzed earthquakes except probably those in the western end of the flat slab. We chose the depth of the earthquakes from the inversion rms variation, tried different rupture velocities, V and initial source time functions, STF. V varies from 1.9-3.6 km/sec. We constrained the slip distribution by mechanism and V obtained in the inversion, and then choose one nodal planes. The recovered STF shows time duration t~13sec. Some STF show a foreshock and/or an aftershock. When the STF has two picks we tried again the inversion with two shocks, obtaining t ~22sec. We looked for the ruptured area length along strike, L, and along the down-dip, W. In the flat slab the two western earthquakes have W aproximate of 18km and the one to the east of it Waproximate10km. So, if the thickened oceanic crust could reach these values then the ruptures should be inside of it. The earthquake in normal section has W aproximate of 16km. Here the crust has ~6-7km. The other 4 earthquakes have 25≤W≤29km, 3 of them in the down-dip slop. Thus, in these cases the rupture goes well into the subducting lithosphere mantle.