Consistency between a strain rate model and the ESHM20 earthquake rate forecast in Europe: insights for seismic hazard

Abstract. This work aims at investigating the consistency between a strain rate model and a long-term earthquake forecast at the European scale. We take advantage of the release of geodetic strain rate models by Piña-Valdés et al. (2022) and the release of the European Seismic Hazard Model 2020 (ESHM20) by Danciu et al. (2024) to compare geodetic and seismic moment rates across Europe. Seismic moments are inferred from the magnitude–frequency distributions that constitute the ESHM20 source model. We explore the full ESHM20 source model logic tree to account for epistemic uncertainties. On the geodesy side, we use the strain rates to calculate the geodetic moment for each area source zone of the hazard model, considering associated epistemic uncertainties. We show that the parameters contributing the most to the overall uncertainty in the geodetic moment rate are the distance weighting scheme used in the spatial inversion, the equation used to convert surface strain to a scalar moment rate, and the effective seismic thickness. We compare the distributions of geodetic and seismic moment rates at different geographical scales. In highly seismic activity zones, such as the Apennines in Italy, Greece, the Balkans, and the Betics in Spain, primary compatibility between seismic and geodetic moment rates is evident. Discrepancies emerge in low- to moderate-seismic-activity zones, particularly in areas affected by the Scandinavian glacial isostatic adjustment, where geodetic moment rates exceed seismic moment rates significantly. We show that considering broader zones enhances the match between geodetic and seismic moment rate distributions. In zones where ESHM20 magnitude–frequency distributions are well-constrained (established on more than 30 complete events), the distributions of seismic and geodetic moments usually overlap significantly, suggesting the potential for integrating geodetic data into hazard models, even in regions with low deformation.