Impact of data uncertainties on french PSHA results using synthetic catalogues.

C. Beauval (1*), D. Bertil (2), F. Bonilla (1) and O. Scotti (1) European Geophysical Society XXVII General Assembly - Nice, 21-26 avril 2002.

Seismicity in France is moderate but the occurrence of several historical earthquakes with magnitudes greater than 5.5 show that seismic hazard assessment must be performed. As a first step towards the evaluation of probabilistic seismic hazard, parameters describing seismicity such as seismic rates and b-values must be estimated. This work uses Monte-Carlo technique to study the influence of the uncertainties inherent to our French catalogue on the evaluation of seismic parameters. The uncertainties we are dealing with concern the estimation of magnitude and of periods of completeness. The initial catalogue is perturbed to generate synthetic catalogues. This catalogue contains around 2000 earthquakes with magnitudes above 3.5: 1200 instrumental ones and 800 macroseismic (before 1962). Uncertainty on the macroseismic magnitude can be very high and the periods of completeness are not very well constrained for magnitudes above 5.5-6.0. The resulting distributions on the seismic parameters will be explored by Monte-Carlo technique to produce coefficients of variation of hazard results and thus estimate the reliability of hazard assessment. (1) Institut de Radioprotection et Sûreté Nucléaire, * étudiant en contrat de thèse à l'IRSN (2) Cenergys