Probabilistic Seismic Hazard Assessment for the Southern part of the Netherlands

Abstract

Abstract Due to the necessity of a national Annex for the Eurocode 8, safety requirements of engineering structures and public interest in the effect of induced earthquakes, an update of the seismic hazard assessment is required for the Netherlands. A first version of a new seismic hazard model for tectonic seismicity is prepared, using two separate probabilistic seismic hazard methodologies. The input parameters (zonation, magnitude-frequency relation, ground motion prediction equation) are determined from the most recent earthquake catalogue and from the literature. In the first method the analytical equations of the Probabilistic Seismic Hazard Analysis (PSHA) method of Cornell (1968) are numerically approached in the program EQRISK (McGuire, 1976). This method is developed for larger tectonic earthquakes, and turns out to be inappropriate to apply to low-seismicity areas, such as the Netherlands. An alternative approach of PSHA, based on a Monte Carlo simulation, is used in the new national Eurocode 8 Annex in the United Kingdom. The M3C hazard software presented by Musson (2009) is used to model the seismic hazard for the Netherlands. Since little data are available for the Netherlands, the seismic hazard analysis admits to a wide range of interpretations and uncertainties. The sensitivity of the hazard estimate to the input parameters is investigated. It turns out that the hazard estimate is mainly sensitive to the definition of source zonation and the choice of ground motion prediction equation. Finally, the revised seismic hazard map presented in this study is compared to alternative studies, which are performed for surrounding countries. The results of Leynaud et al. (2000) agree fairly well with those presented in this study. Probably due to the use of another zonation model, the seismic hazard maps of Meskouris (2005) and Grünthal et al. (2007) is more different, although in the Roer Valley region similar peak ground accelerations are expected. The results of this study are quite different from those of a former study (De Crook, 1996). This is probably due to the smaller amount of data used for the former study and the use of another source definition. The final proposed seismic hazard map contains regions for which (taking the uncertainty into account) the critical value mentioned in the Eurocode 8 will most probably not be exceeded in 475 years. For the remaining areas (i.e. Roer Valley, Hainaut zone, the areas surrounding Liège and Hautes Fagnes and the Neuwied basin), the peak ground acceleration values are just below, or above the critical value. It is up to the user to decide whether seismic hazard should be taken into account. For further research it is recommended to quantify the uncertainty of the hazard estimate.