Please use this identifier to cite or link to this item:
Authors: Liechti, D.* 
Ruettener, E.* 
Eugster, S.* 
Streit, R.* 
Title: The impact of a and b value uncertainty on loss estimation in the reinsurance industry
Issue Date: Feb-2000
Series/Report no.: 1/43 (2000)
Keywords: Completeness
seismic risk
average annual loss
Subject Classification04. Solid Earth::04.06. Seismology::04.06.99. General or miscellaneous 
04. Solid Earth::04.06. Seismology::04.06.11. Seismic risk 
Abstract: In the reinsurance industry different probabilistic models are currently used for seismic risk analysis. A credible loss estimation of the insured values depends on seismic hazard analysis and on the vulnerability functions of the given structures. Besides attenuation and local soil amplification, the earthquake occurrence model (often represented by the Gutenberg and Richter relation) is a key element in the analysis. However, earthquake catalogues are usually incomplete, the time of observation is too short and the data themselves contain errors. Therefore, a and b values can only be estimated with uncertainties. The knowledge of their variation provides a valuable input for earthquake risk analysis, because they allow the probability distribution of expected losses (expressed by Average Annual Loss (AAL)) to be modelled. The variations of a and b have a direct effect on the estimated exceeding probability and consequently on the calculated loss level. This effect is best illustrated by exceeding probability versus loss level and AAL versus magnitude graphs. The sensitivity of average annual losses due to different a to b ratios and magnitudes is obvious. The estimation of the variation of a and b and the quantification of the sensitivity of calculated losses are fundamental for optimal earthquake risk management. Ignoring these uncertainties means that risk management decisions neglect possible variations of the earthquake loss estimations.
Appears in Collections:Annals of Geophysics

Files in This Item:
File Description SizeFormat
03.pdf2.72 MBAdobe PDFView/Open
Show full item record

Page view(s) 50

checked on Feb 1, 2023

Download(s) 10

checked on Feb 1, 2023

Google ScholarTM