Long-term seismicity in regions of present day low seismic activity: the example of western Europe

In western Europe, the knowledge of long-term seismicity is based on reliab le historical seismicity and covers a time period of less than 700 years. D espite the fact that the seismic activity is considered as low in the regio n extending from the Lower Rhine Embayment to England, historical informati on collected recently suggests the occurrence of three earthquakes with mag nitude around 6.0 or greater. These events are a source of information for the engineer or the scientist involved in mitigation against large earthqua kes. We provide information relevant to this aspect for the Belgian earthqu ake of September 18, 1692. The severity of the damage described in original sources indicates that its epicentral intensity could be IX (EMS-98 scale) and that the area with intensity VII and greater than VII has at least a m ean radius of 45 km. Following relationships between average macroseismic r adii and magnitude for earthquakes in stable continental regions, its magni tude M-s is estimated as between 6.0 and 6.5. To extend in time our knowled ge of the seismic activity, we conducted paleoseismic investigations in the Roer Graben to address the question of the possible occurrence of large ea rthquakes with coseismic surface ruptures. Our study along the Feldbiss fau lt (the western border of the graben) demonstrates its recent activity and provides numerous lines of evidence of Holocene and Late Pleistocene large earthquakes. It suggests that along the 10 km long Bree fault scarp, the re turn period for earthquakes with magnitude from 6.2 to 6.7 ranges from 10,0 00 to 20,000 years during the last 50,000 years. Considering as possible th e occurrence of similar earthquakes along all the Quaternary faults in the Lower Rhine Embayment, a large earthquake could occur there each 500-1000 y ears. These results are important in two ways. (i) The evidence that large earthquakes occur in western Europe in the very recent past which is not on ly attested by historical sources, but also suggested by paleoseismic inves tigations in the Roer Graben. (ii) The existence of a scientific basis to b etter evaluate the long-term seismicity in this part of Europe (maximal mag nitude and return period) in the framework of seismic hazard assessment. (C ) 2001 Elsevier Science Ltd. All rights reserved.