Implications of GMPE's structure for multi-site seismic hazard

Previous research has discussed the implications of the structure of a classical ground motion prediction equation (GMPE) on single-site probabilistic seismic hazard analysis and disaggregation. Classical refers to a GMPE where local site conditions (or any other factor) are accounted for via constant (with respect to magnitude and source-to-site distance) terms added to the mean and that do not affect the distribution of the residuals. Herein, the implications of such a structure of the GMPE are briefly discussed with respect to multi-site hazard assessment that, typically, requires a large number of simulations of random fields of ground motion intensity measures. It is shown that this type of GMPEs enables to run the simulations only once, independently of the soil conditions (or any other factor modeled in a similar way) eventually assigned to each site, which can represent a significant computational advantage in the case of spatially distributed assets.