Measuring the onset of locking in the Peru–Chile trench with GPS and acoustic measurements

Abstract

The subduction zone off the west coast of South America marks the convergence of the oceanic Nazca plate and the continental South America plate. Nazca-South America convergence over the past 23 million years has created the 6-km-deep Peru-Chile trench, 150 km offshore. High pressure between the plates creates a locked zone, leading to deformation of the overriding plate. The surface area of this locked zone is thought to control the magnitude of co-seismic release and is limited by pressure, temperature, sediment type and fluid content. Here we present seafloor deformation data from the submerged South America plate obtained from a combination of Global Positioning System (GPS) receivers and acoustic transponders. We estimate that the measured horizontal surface motion perpendicular to the trench is consistent with a model having no slip along the thrust fault between 2 and 40 km depth. A tsunami in 1996, 200 km north of our site, was interpreted as being the result of an anomalously shallow interplate earthquake. Seismic coupling at shallow depths, such as we observe, may explain why co-seismic events in the Peruvian subduction zone create large tsunamis.

Scenario Source Models and Strong Ground Motion for Future Mega‐earthquakes: Application to Lima, Central Peru

“…Black arrows are velocity vectors of the Nazca plate at the convergence margin with the South American plate (Chlieh et al, 2011). 87 sites surveyed from the period 1993-2003 in Peru (Chlieh et al, 2011). These data include interseismic seafloor deformations measurements obtained offshore Lima city from a deployment of seafloor acoustic transponders and inland GPS receivers (Gagnon et al, 2005). The ISC model is based on the assumption that convergence between the Nazca and South American plates is dominantly accommodated along the megathrust including 4 mm=yr of back-arc shortening absorbed at the subandean fold and thrust belt regions (Chlieh et al, 2011).…”

Seafloor Geodetic Approaches to Subduction Thrust Earthquakes

“…An alternative interpretation of the result of Gagnon et al (2005) is that the shallow part is not locked, but looks apparently locked due to the interplate locking in a deeper part of the plate boundary. Whether this difference can be observed by geodetic observations on the seafloor, or not, is an important problem.…”

“…One of the critical reasons why the giant earthquake had not been suspected is the lack of information on the locking in the shallow part of the plate boundary. While GPSA observations by Gagnon et al (2005) in the Peru-Chile subduction zone suggested that the shallow part of the plate boundary is locked, few repeating earthquakes had been observed on the shallow plate boundary off Miyagi (Uchida and Matsuzawa, 2011). A low degree of locking was estimated there, and GPSA observations in the area had been delayed.…”

“…Therefore, we usually make GPSA measurements for one day at a site to reduce the effect of sound speed variation. Gagnon et al (2005) estimated that the final positions of two arrays changed by less than 1 cm after 80 hours of continuous data collection in the subduction zone off Peru. A key problem, in a practical sense, however, remains to be solved.…”

“…It shows a clear contrast to the episodic motion across the 1-km-wide axial valley observed by acoustic ranging (Chadwell et al, 1999;Chadwick and Stapp, 2002). Gagnon et al (2005) presented important results of GPSA observations at the PeruChile Trench. They showed that the measured horizontal surface motion perpendicular to the trench is consistent with the model that has no slip along the thrust fault between 2 and 40 km depth and suggested interplate locking at shallow depths of the plate boundary.…”