Regional correlations for estimating seismic amplification. Implications for loss assessment in SW Iberia

Abstract

Most earthquake-related losses are enhanced by soil amplification phenomena, especially in areas where high vulnerability assets coexist. These mechanisms are frequently inferred from empirical formulas. These procedures are based on the average shear-wave velocity to 30 m depth (Vs30) as a proxy for amplification events. However, past analysis of spatial relationships between the Vs30 and nonlinear soil behaviour has frequently resulted in poor correlations, affecting all onward analysis. In this research, these relationships are further analysed using an Earthquake Loss Estimation Software. In this framework, the Algarve region and associated ground motion have been depicted based on a certified seismic catalogue. This region has a moderate seismic hazard but possesses a complex geology and dissimilar seismogenic sources. The source data for this research has been Vs30 data collected from geologic surveys made for the region in 2010 in the framework of the Study of Seismic Risk and Tsunamis in Algarve (ERSTA) coordinated by the Portuguese Civil Protection. Other data collected has been the Vs30 topography compiled by the USGS and geologic information available by the OneGeology collaborative project. The goal of this research is to identify inaccuracies that potentially occur in estimating the site effects by different approaches. In this framework three amplification factors based on different sources - Vs30 field-data, Vs30 using slope as proxy and Vs30 using geology as proxy – have been tested. The results have shown a good correlation between Vs30-field-data and geologic based Vs30 and a minor correlation between the former and slope based Vs30. However, regarding the seismic losses - building damage and human losses - the final results show similar values for all three amplification factors analysed.