98th JLG: European ground motion models from the past to the future

Last updated: 18 November 2016

Thanks to all participants for a highly productive meeting!

The 98th “Journées Luxembourgeoises de Géodynamique” (JLG) took place on 14-16 November 2016 at the Alvisse Parc Hotel in Luxembourg City. The 98th edition focused on the subject of European ground motion models from the past to the future: Lessons learned and new trends and was held as a strongly discussion-based workshop. The 98th JLG were organised by ECGS and Section 2.6 of the German Research Center for Geosciences (GFZ) in Potsdam, Germany.

JLG98 Group Photo

Journées Luxembourgeoises de Géodynamique

  • President: Manfred Bonatz
  • Scientific Secretary: Adrien Oth
  • Organizing Committee: Adrien Oth (ECGS), Dino Bindi (GFZ), Fabrice Cotton (GFZ), Corine Galassi (ECGS), Gilles Celli (ECGS/Mnhn)

Workshop Description

Ground-motion prediction in Europe is a difficult challenge. Indeed, Europe is a complex tectonic puzzle. To image the tectonic map of Europe, we first of all need to match the subduction pieces for the Calabrian and Hellenic arc with those for the oceanic crust in the Mediterranean Sea and for shallow crustal active regions all around the Mediterranean Sea. Moving north towards continental Europe, the map develops matching the pieces for stable continental regions with those for the Baltic shield. In Europe we also have active volcanos and intermediate-depth earthquakes occurring far from plate boundaries. Such a complex tectonic pattern poses several challenges for the seismic hazard assessment that require suitable models for describing the seismic sources and the attenuation effects. From the ground motion point of view, the large spatial variability of the deformation processes reflects in a large variability of the seismic activity rates, making the capturing of ground motion variability in low seismicity areas particularly challenging.

Most existing ground-motion models do not take into account such regional specificities of earthquake or wave-propagation properties. For example, ground-motion models used in engineering seismology are calibrated on global databases that are usually created by mixing data from different regions. Seismological data recorded by local networks show, however, that regional variations of faults, earthquakes and crustal properties are significant. These observations pose significant problems as they show that regionalization of ground-motion models is required. There is little guidance in the literature on how this should be done.

The aim of the 2016 edition of the Journées Luxembourgeoises de Géodynamique is therefore to bring together the experiences on seismic hazard assessment recently gained in the framework of regional and national projects across Europe. The meeting will be laid out as a discussion-driven workshop framed in two sections:

  • The first section will be devoted to the lessons learned thus far. Approaches followed by different projects for sampling regional variations of ground motion at different scales, from a single country to the whole of Europe, will be presented and discussed, both for low and high seismicity regions. Ongoing activities, like those currently developing in the framework of EPOS, are of interest for this section as well.
  • In the second part, new developments and trends in ground motion characterization and hazard assessment will be discussed. Examples are procedures for site-specific hazard assessment; exploration of the potentiality of merging weak and strong motion data; data-driven ground motion modeling; new needs from the earthquake engineering community.


Monday morning

  • 11:00 – 12:00 Registration
  • 12:00 – 12:30 Welcome

12:30 – 14:00 Lunch

Monday afternoon: Lessons learned & needs

  • 14:00 – 15:45 Keynote Presentations
    • 14:00 – 14:45 Talk 1: Predicting and Using Ground Motions in Regions with Sparse Data for Magnitudes and Distances of Engineering Interest: Recent Developments in the US and Greece D. Boore (30 min + 15 min discussion)
    • 14:45 – 15:15 Talk 2: Ground motion modelling and PSHA: lessons learned from recent European projects. L. Danciu (20 min + 10 min discussion)
    • 15:15 – 15:45 Talk 3: What are the engineers requirements from ground motion prediction? S. Akkar (20 min + 10 min discussion)
  • 15:45 – 16:45 Coffee break and poster session
  • 16:45 – 18:00 Discussion and brainstorming
    • Do we really need more data to improve ground motion prediction? If yes, what type of data?
    • Is there still a gap between the seismological and earthquake engineering communities?
    • Have we been surprised by ground motions in recent earthquakes?

Tuesday morning: Selection, epistemic uncertainties and regionalisation

  • 08:30 – 10:15 Keynote Presentations
    • 08:30 – 09:15 Talk 1: GMPE selection and alternative methods for developing ground motion models (e.g. backbone approach). Which criteria? Are we capturing the right level of epistemic uncertainty? J. Douglas (30 min + 15 min discussion)
    • 09:15 – 09:45 Talk 2: Application of GMPEs in PSHA across the Globe: Current considerations, good practice and emerging issues. G. Weatherill (20 min + 10 min discussion)
    • 09:45 – 10:15 Talk 3: Comparison of GMPEs to observations. What does it really tell us? S. Mak (20 min + 10 min discussion)
  • 10:15 – 11:15 Coffee break and poster session
  • 11:15 – 12:30 Discussion and brainstorming
    • Is there a better approach than logic trees?
    • Back-bone models or selections of multiple GMPEs?

12:30 – 13:30 Lunch

Tuesday afternoon: Physics based ground motion prediction

  • 13:30 – 15:15 Keynote Presentations
    • 13:30 – 14:15 Talk 1: Development of physics-based approaches to assess the hazard in large urban areas. R. Paolucci (30 min + 15 min discussion)
    • 14:15 – 14:45 Talk 2: Lessons learned from the use and development of stochastic models. B. Edwards (20 min + 10 min discussion)
    • 14:45 – 15:15 Talk 3: Engineering use of physics-based models. L. Dalguer (20 min + 10 min discussion)
  • 15:15 – 16:15 Coffee break and poster session
  • 16:15 – 17:30 Discussion and brainstorming
    • Are physics-based model mature enough to be used for engineering applications? For which regions, scenarios and structural period ranges?
    • Are empirical and physics-based models able to capture realistic near-fault effects?
    • What is the actual level of epistemic and aleatory uncertainties in physics based ground motion predictions?

18:00 Social evening & meeting dinner

Wednesday morning: Site specific ground motion prediction

  • 08:30 – 10:15 Keynote Presentations
    • 08:30 – 09:15 Talk 1: Strategies to take into account site-specific conditions and their uncertainties P.Y. Bard, C. Aristizabal, C. Beauval, E. Chaljub, C. Cornou, B. Derras, B. Guillier, F. Hollender, A. Laurendeau, E. Maufroy, V. Perron, J. Régnier (30 min + 15 min discussion)
    • 09:15 – 09:45 Talk 2: European station site characterization and status of the EPOS ground motion databases. L. Luzi, (20 min + 10 min discussion)
    • 09:45 – 10:15 Talk 3: Site specific and partially ergodic GMPEs. Implication for the development of an EPOS GMPE service. F. Cotton, D. Bindi, S. Bora, S.R. Kotha, Y.S. Chen (20 min + 10 min discussion)
  • 10:15 – 11:15 Coffee break and poster session
  • 11:15 – 12:30 Discussion and brainstorming
    • What is beyond vs30?
    • Strengths and weaknesses of host-to-target adjustments
    • Do building codes correctly estimate non-linear site effects?
    • Do building codes correctly estimate topographic effects?

12:30 – 13:30 Lunch

Wednesday afternoon: Final discussion, 13:30 – 15:00

  • Where do we expect to be in 25 years?
  • Which kind of service/coordination (infrastructures, databases, projects) are needed to improve ground motion prediction in Europe?

For More Information

Please contact:
Ms Corine GALASSI, ECGS secretary
19 rue Josy Welter
L-7256 Walferdange
Grand Duchy of Luxembourg
Tel: +352 33.14.87-1
Fax: +352 33.14.87-88
Email: corine.galassi@ecgs.lu