Abstract

The selection of ground motion models, and the representation of their epistemic uncertainty in the form of a logic tree, is one of the fundamental components of probabilistic seismic hazard and risk analysis. A new ground motion model (GMM) logic tree has been developed for the 2020 European seismic hazard model, which develops upon recently compiled ground motion data sets in Europe. In contrast to previous European seismic hazard models, the new ground model logic tree is built around the scaled backbone concept. Epistemic uncertainties are represented as calibrations to a reference model and aim to characterise the potential distributions of median ground motions resulting from variability in source scaling and attenuation. These scaled backbone logic trees are developed and presented for shallow crustal seismic sources in Europe. Using the new European strong motion flatfile, and capitalising on recent perspectives in ground motion modelling in the scientific literature, a general and transferable procedure is presented for the construction of a backbone model and the regionalisation of epistemic uncertainty. This innovative approach forms a general framework for revising and updating the GMM logic tree at national and European scale as new strong motion data emerge in the future.

Highlights

  • The effective mitigation of seismic risk within a region begins with a quantitative assessment of the shaking hazard posed by earthquakes

  • Given the wealth of new ground motion data for Europe, the core backbone model for shallow crustal seismicity is that of Kotha et al (2020), which is fit to data from the Engineering Strong Motion flatfile of Lanzano et al (2019)

  • Whilst the primary focus of the construction of the ground motion model (GMM) logic tree is on the representation of epistemic uncertainty, as a means of understanding the potential implications of the new backbone model for Probabilistic seismic hazard assessment (PSHA) it is important to compare the aleatory uncertainty in the new model with others derived previously

Read more

Summary

Introduction

The effective mitigation of seismic risk within a region begins with a quantitative assessment of the shaking hazard posed by earthquakes. Completed at the end of 2013, the ESHM13 was a generational leap-forward in seismic hazard mapping in the region It introduced explicit modelling of active faults into PSHA in Europe, a comprehensive logic tree capturing epistemic (or model-to-model) uncertainty, and a hybrid expert- and data-driven procedure for the selection and weighting of ground motion models. For the characterisation of strong motion, a new and updated selection of ground motion models (GMMs) is required, alongside a corresponding representation of the epistemic uncertainty. This paper describes the new developments of the ESHM20 ground motion model logic tree for Europe in order to characterise the seismic hazard on Eurocode 8 class A rock, with a measured 30 m average shearwave velocity (V S30 ) of 800 m/s. Though tightly connected to the ground motion logic tree of the ESHM20, the European site response model itself requires such a sufficient level of detail that this too will be addressed in a subsequent paper

Epistemic uncertainty: insights from new data and new perspectives
The shallow crustal backbone model
Region‐specific components of variability
Within‐model statistical uncertainty
The “default” shallow crustal logic tree
Regional calibration
A pseudo‐volcanic GMM?
Constraining the c3 distribution per cluster
Aleatory uncertainty
The complete shallow crustal logic tree
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call