Abstract Slip rate, locking width and ground-motion prediction equation (GMPE) selection are important in seismic hazard analysis because they are used to estimate earthquake recurrence, to limit the maximum magnitude in an earthquake source and to estimate earthquake ground shaking. In this study, we examine the sensitivity of probabilistic seismic hazard analysis (PSHA) to fault slip rates, fault locking width and the selection of GMPEs for the Aceh Fault Segment, Indonesia. The hazard level differences vary considerably owing to changes in these three parameters. Therefore, careful consideration is needed in applying PSHA in areas of high fault parameter uncertainty.

Abstract Subduction zones are highly productive seismic sources both at the interface and within the slab, and have produced many of the highest-impact earthquakes throughout history. Several current probabilistic seismic hazard analyses model interface sources as faults with rates constrained by only seismicity or tectonics singularly, and intraslab sources as gridded points, sometimes smoothing rates according to former occurrences. Here, we propose a subduction source modelling methodology that improves on current standards for both source types. We use cross-sections of earthquake catalogues, focal mechanisms and geophysical models to define the locked subduction interface and slab volume, and to classify seismicity. We divide these geometries along-strike to impose segment boundaries. For the interface, we combine tectonic and seismicity components to assign occurrence properties, considering fault area, convergence rate and seismic coupling. We use ruptures that are confined to the slab volume to demarcate intraslab sources, enhancing their utility with ground motion prediction equations that use site-to-rupture distance. We demonstrate the proposed approach on the Nazca subduction zone of South America, validating the source model parameters against observed earthquakes and palaeoseismic information when available, and showing that the resulting magnitude–frequency distributions better fit the observed occurrences than traditional approaches.