A New Model for the Seismogenic Behavior of Subducted Seamounts Based on Multi-Channel Seismic Reflection and GPS Data Collected in Central Ecuador.
Thursday, 18 December 2014: 10:35 AM
The relationship between subducted seamounts and earthquakes has remained controversial. Although seamounts are expected to subduct aseismically, they have also been considered to generate large earthquakes. Based on a remarkable case study in Central Ecuador, we show that a subducted seamount can lock the shallow megathrust along its trailing flank preparing for a possible shallow (<20 km), large magnitude (Mw ~7.0) tsunamogenic earthquake, while its leading flank keeps partially creeping along with frequent earthquake swarms and slow slip events (SSE). The erosive Ecuador convergent margin, which basement consists of high velocity (Vp=5 km/s) mafic rocks, is underthrust eastward at 4.7 cm/yr by the rugged Carnegie Ridge. As modeled by global positioning system (GPS) measurements acquired as close as 35 km from the trench axis at La Plata Island, the Central Ecuador margin figures a creeping subduction segment with the exception of a 50 km-diameter locked patch centered over the uplifted La Plata Island region. The 3D geometry of the plate-interface megathrust obtained from 2D-PreStack-Depth-Migration of a grid of multi-channel seismic reflection data collected near La Plata Island reveals a collection of closely spaced peaks that belong to a broad (55 X ~50 km) low-drag shape subducted seamount. The clear spatial correlation between the seamount and the highly coupled zone denotes the seamount as the main cause for both the locked patch and the island uplift. The absence of a seismically imaged subduction channel, the highly jagged seamount-trailing flank and the stiffness of the oceanic margin are found to be the principal long-term characteristics associated with shallow locking of the megathrust. Moreover, the combination of our structural interpretation and inter-seismic coupling map with 14-years of relocated seismicity, and the 2010 SSE and its associated microseismicity allow to propose a new model for the seismogenic behavior of subducting seamounts.