Magnitude problems in historical earthquake catalogues and their impact on seismic hazard assessment

• Published in: Geophysical journal international 2011-12, Vol.187 (3), p.1687-1698
• Main Authors: Rong, Yufang, Mahdyiar, Mehrdad, Shen-Tu, Bingming, Shabestari, Khosrow
• Format: Article
• Language: English
• Subjects: Bias; Catalogues; Earthquake construction; Europe; Hazard assessment; Hazards; Regional; Seismic engineering; Seismic phenomena; Seismicity and tectonics; Statistical seismology
• ISSN: 0956-540X; 1365-246X
• DOI: 10.1111/j.1365-246X.2011.05226.x

Summary A reliable historical earthquake catalogue is a highly critical component in any regional seismic hazard analysis. Although many of the various catalogues used for constructing national or regional seismic hazard maps in Europe cover much of the same territory and use standardized parameters, some discrepancies have been revealed between different catalogues. Switzerland′s ECOS catalogue, compiled by the Swiss Seismological Service (SED), and Italy′s CPTI04 catalogue, compiled by the CPTI Working Group, spatially overlap each other to a large extent and employ a uniform moment magnitude scale (M w). However, a careful review of these two catalogues has revealed significant inconsistencies in the moment magnitude (M w): (1) The moment magnitudes used for the same earthquakes were found to differ, which could potentially have a substantial impact on any regional seismic hazard assessment that is based on these catalogues. This type of inconsistency often arises when different regression relationships are used to convert other magnitude scales or event intensity into M w, which can cause many of the event magnitudes in one catalogue to be systematically biased higher or lower with respect to those in another catalogue. In this case, the magnitudes of small to medium historical earthquakes in Italy′s CPTI04 catalogue are systematically higher than those in Switzerland′s ECOS catalogue. (2) An abnormally high frequency of large magnitude events was observed for some of the time periods, particularly if most of the available data was on intensity. This has been observed in both the CPTI04 and ECOS catalogues for the time period prior to 1975 and is possibly due to a bias during the intensity-to-magnitude conversion. (3) A systematic bias in magnitude resulted in biased estimations for the a and b values of the Gutenberg-Richter magnitude-frequency relationships. All of these issues can lead to skewed seismic hazard results, or inconsistencies in the seismic hazard from different studies. By constructing seismic hazard models based on the ECOS and CPTI04 catalogues, we can evaluate the impact of these inconsistencies on the regional seismic hazard. The results show that the seismic hazard from the two models can disagree by as much as a factor of three at some locations, which demonstrates the importance of developing consistent and unbiased earthquake catalogues across neighbouring countries.