Conference Dates

June 18-21, 2006


The destructive seismic events of January 13 (ML: 7.6) and February 13 (ML: 6.1) 2001 in El Salvador, with origin in the subduction area and the volcanic chain, respectively, provide an ideal scenario to analyse the factors that define the landslide hazard in this country. In this paper we analysed the events in terms of strong-motion and precedent climatic condition and their relation with the landslides induced; establishing a great opportunity to compare some hazard assessment methods as those proposed by Mora and Vahrson (1994) and Rodríguez (2001) which define the hazard in terms of the interaction between triggering agents and susceptibility conditions.

Historically both rainfalls and earthquakes have shown to be important triggers for landslides in El Salvador, and results show how the combination of these factors are also critical in defining trigger thresholds and in controlling failure mechanisms. It was found that the Mora and Vahrson method underestimates the landslides hazard; as was discussed by Bommer and Rodríguez (2002), this is mainly due to the rainfalls levels used by the method which were defined for the Costa Rica conditions, which are markedly different to those in El Salvador. The Rodríguez model describes in a better way the landslides hazard, however in some areas the hazard is overestimated due to the way as weighting factors for lithology shades the influence of topography, this suggests that assignment of weighting values in the model must be reviewed implementing a multivariate correspondence analysis instead of the bivariate model used so far.