Comparison of PSH results with historical macroseismic observations at different scales. Part 1: methodology

Abstract

The need for verifying and/or validating the results of probabilistic seismic hazard studies has become evident, despite the absence of generally accepted methods for judging such results. This paper proposes a methodology for comparing the results of PSH with historical macroseismic observations, at different scale levels. The selected metric of the comparison is the mean damage, i.e. the average annually expected damage. Hence, a procedure for converting macroseismic observations and PGA levels, for which PSH estimates are provided, into mean damage values is first presented. The procedure is based on the macroseismic method for converting intensities into mean damage values, whereas it takes advantage of fragility curves (representative of the seismic vulnerability of the building stock at the time of historical observations) to transform the rates of occurrence of PGA values into the equivalent quantity in terms of mean damage. A methodology for site-specific comparison is first outlined. To overcome some of the limitations of the comparison at a single site, two procedures for aggregating several sites by sampling in space are then presented, with the aim of enlarging the available macroseismic dataset. Finally, a procedure for the comparison at the regional level is discussed. Applications of the proposed methodologies will be discussed in a companion paper (Rosti and Rota in Bull Earthq Eng, 2017).