Ground motion evaluation along the Chilean subduction zone: variability and possible causes

T he study and prediction of ground motions produced by moderate to large earthquakes is one of the key components for engineering seismology and seismic hazard assessments. In this context, Ground Motion Prediction Equations (GMPE) are models that predict the ground motion intensity at a given site, based on earthquake source, path and site effects parameters . These models can be used as a reference framework to evaluate observed ground motions in a specific region.

The Chile an section of the Nazca plate subduction under South America is one of the most seismically active regions in the world, where at least one significant earthquake with damaging ground motions occurs every 10 years. To monitor the seismicity, new digital seismic stations are continually being deployed, thus providing opportunities for strong ground motion studies.

In this work, we analyze ground motion variability along the Chilean subduction zone. We divide the country in to three regions (north, center and south) that are analyze d separately. We compile an interface earthquake catalog for each region, and infer site condition parameters for the newer stations deployed in Chile. Nationwide, we obtain more than 5400 strong motion observations from 375 recent 4.0 ≤ Mw ≤ 8 . 8 ea rthquakes, and compare th ese observations with an empirical GMPE developed with Chilean data.

We find a good overall fit of the GMPE in north and central Chile, with the best fit in central Chile. By analyzing between-event residuals, we find a clear increase of short-period ground-motion energy with depth in these two regions, with a sharper contrast in central Chile. T he events showing larger ground motions at short periods are located in areas of low interseismic coupling, similar to recent findings in Japan, suggesting a frictional related origin . Southern Chile shows less frequent seismicity and has not been considered for GMPE development. O bservations in this region are generally over-predicted by the GMPE. We propose that future GMPEs should include corrections for depth, and include data from Southern Chile, since recent studies have proposed a reawakening of significant seismicity since the 1960 Mw 9.5 Valdivia earthquake, starting with the 2016 Mw 7.6 Chiloé earthquake in the region.