A stochastic ground motion accelerogram model for Northwest Europe

• A fully non-stationary stochastic accelerogram model for NW Europe is proposed.
• One nonlinear functional form of predictive equations for such a model is presented.
• Model inputs are three variables: moment magnitude, epicentral distance and soil type.
• Variability of intensity, frequency and time span of NW European records is modelled.
• Model is validated against attenuation relations for UK, Europe and stable regions.

This article presents a stochastic ground-motion accelerogram model for Northwest Europe which simulates accelerograms compatible with seismic scenarios defined by earthquake magnitudes 4