Empirical site classification of CSN network using strong-motion records

Abstract

The National Seismological Center of the University of Chile (CSN, Centro Sismologico Nacional) has been operating more than 400 seismic stations throughout the country. The data collected from this network, from March 2012 to August 2017, includes more than 4000 Chilean strong-motion records, from more than 1000 events (magnitudes ranging from 4.0 up to 8.3). In this study, we use this data set and classify 118 stations from this network, using the horizontal-to-vertical response spectral ratio (HVRSR) of strong-motion records. This classification considers not only the predominant period obtained from the average HVRSR but also the peak amplitude from this curve. The results indicate no correlation between the two parameters, despite the common practice of combining all curves with similar predominant periods. Even more, we believe that relevant information of the site's impedance contrast between the soil and bedrock is lost in the process of averaging HVRSR curves from different stations.