Elements of Seismic Risk Evaluation at the City of Lourdes, France
Abstract
The pilgrim city of Lourdes, in the French Pyrenees, is located in a complex area 12 km north of the North Pyrenean Fault, which represents the boundary between Eurasian and Iberian plates. It has been partly destroyed in 1660 and 1750 by two events with maximum intensity IX and VIII respectively. In order to evaluate the damages which could now result from similar events, estimated to ML=6.0, at least three informations are necessary: 1- where are located the active faults, and what are the depths and focal mechanisms of the events on these faults; 2- what are the site effects at the various places in Lourdes; 3- how to extrapolate the ground responses from small events to large events. A detailed study of the seismicity with focal mechanisms determinations has first been performed using permanent and temporary networks. It reveals that the active fault is probably not the North Pyrenean Fault: the activity near Lourdes is mostly concentrated along several segments approximately perpendicular to this fault. In order to evaluate the site responses inside the city, ten velocimetric stations have been set up on the various geological structures. 32 events at distances 5 to 65 km from Lourdes with backazimuths 130 to 320 degrees and local magnitude 1.4 to 3.2 have been recorded in 7 months. The classical spectral ratio method, with one reference station on the bedrock, has been applied. It reveals amplification by a factor up to ten at frequencies 5 to 20 Hz for stations on the sediments, compared to the stations on the bedrock. Topographic effects are also clearly apparent, but neither the hypocenter locations nor the magnitudes seem to be important parameters for the spectral ratio. 1D and 2D modeling using the available geological informations allow to partly explain the observations. The main difficulty is to predict the ground response to a large earthquake. We have attempted to establish empirical relationships between ground acceleration and magnitude, using a permanent accelerometric station set up inside the city, and nearby velocimetric stations with longer history, which recorded a few M=5 events. However, some models will be necessary to evaluate the ground motion at larger magnitudes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2001
- Bibcode:
- 2001AGUFM.S52C0655D
- Keywords:
-
- 7212 Earthquake ground motions and engineering;
- 7223 Seismic hazard assessment and prediction;
- 7230 Seismicity and seismotectonics