Heavier Damages without Site Effects and Site Effects with Lighter Damages: Boumerdes City (Algeria) after the May 2003 Earthquake

Abstract

The ambient vibration H/V method (Nakamura 1989) is now widely used to determine the fundamental soil frequency and has proven to be a good tool for site effect evaluation ( e.g. , Bonnefoy-Claudet, Cornou et al. 2006). More particularly, it has been used for the zoning of a number of cities around the world, such as Basel, Switzerland (Fah et al. 1997); Quito, Ecuador (Gueguen et al. 2000); Barcelona, Spain (Alfaro et al. 2001); Caracas, Venezuela (Duval et al. 2001); Almeria, Spain (Navarro et al. 2001); and Thessaloniki, Greece (Panou et al. 2005). This method has also been used to find a correlation between damages from an earthquake and the frequency of ambient vibration H/V peaks ( e.g. Gueguen et al. 1998; Panou et al. 2005; Teves Costa et al. 2007; Cara et al. 2008). Other studies show that this correlation is not always clear or even does not exist (Fallahi 2003; Gonzales et al. 2004; Chatelain, Guillier, and Parvez 2008; Theodulis et al. 2008). The 21 May 2003 Algerian ( Mw = 6.8) earthquake produced severe damages in the city of Boumerdes, among others (see Figure 1 for locations). These damages were not uniformly distributed throughout the city, with a destruction level varying from 0 up to 5%, except in two neighborhoods where it reached 30% (Hamane et al. 2007). These observations led us to try to link the destruction pattern to variations of local seismic wave amplification. One of the most widely used methods for studying this type of site effect is based on recording ambient vibrations (noise) in both the vertical and horizontal directions and computing their spectral ratio (a comprehensive review of ambient vibration studies may be found in Bonnefoy-Claudet, Bard, and Cotton 2006). The physical meanings of the ambient …