This article presents a seismic-hazard disaggregation methodology based on the probabilistic approach introduced by Mortgat and Shah (1979). The disaggregation technique we propose computes the probability of exceedence of a disaggregated ground motion, gmd, characterized by a probability of at least one exceedence, P (GM > gmd), in a time interval, t, for given values of magnitude, m, distance, r, a measure of the deviation of the ground motion from its predicted value, ϵ, and the number of occurrences, n.
The original points of the procedure presented here are: (1) disaggregation is made in terms of probabilities with significant advantages over the most common approach based on occurrences, and (2) the method proposed shows whether the contributions to the disaggregated hazard value are given by single or multiple earthquake occurrences. To test the approach three source models have been considered: a single point source, gridded seismicity, and a vertical fault. The three examples show the basic capabilities of the procedure and demonstrate some interesting features.