Seismic hazard

The seismic hazard assessment corresponds to the estimation of the seismic movements expected in a particular region or site, their levels and their return periods. Past earthquakes and their effects are analyzed in order to answer questions such as: « What is the ground acceleration with a 10% chance of being exceeded in the next 50 years in a region of interest? ».

The 3 main areas of research and development in the field of seismic hazard assessment are the following:

1. Quantification of the seismic potential of seismic sources

The main hypothesis governing seismic hazard assessment is that past earthquakes may recur in the region where they have already occurred. It is therefore very important to focus on describing the past earthquakes. To characterize the most recent sources, the studies are based on instrumental data, i.e. earthquakes characterized by seismograms (since 1962 in France). The characterization of events that occurred during the historical period is based on the SisFrance database, which lists all known earthquakes in the metropolitan territory (the oldest date from the 6th century) and identifies, for each one, the effects felt by the population as well as the damage caused to the buildings.

Sismicité historique (points oranges) et instrumentale (points rouges) de la France en magnitude locale ML et zonage sismotectonique du LDG (en trait noir).
Historical (orange dots) and instrumental (red dots) seismicity of France in local magnitude (ML) and seismotectonic zoning of the laboratory LDG (black line).

In France, there are few strong earthquakes and they rarely reoccur in the same place on an instrumental and historical scale. We could not therefore expect to have an exhaustive picture of the potential seismicity of a region using only instrumental and historical seismicity data. The search for neotectonic indications, paleoearthquakes and the use of deformation measurements are therefore essential complements for the analysis of past seismicity.

Champ de vitesse GPS du Sud Est de la France relativement à l’Eurasie stable (Nocquet, 2013), sismicité instrumentale de magnitude supérieure à 3 (LDG) et grands systèmes de failles (en bleu, failles normales, en vert failles décrochantes, en noir : chevauchements).
GPS speed field of South-East of France relative to stable Eurasia (Nocquet, 2013), instrumental seismicity of magnitude higher than 3 (LDG) and large fault systems (in blue: normal faults, in green: strike-slip fault, in black: thrust faults).

The identification of the faults responsible for the past strong earthquakes, completed by the study of structures capable of carrying the future large earthquakes, are necessary prerequisites for a better definition of the seismic hazard. These studies, because they focus on understanding the functioning of faults on several scales of time (from the second to the last millions of years) and space (from the local geological outcrop to the tectonic plate that supports it), involve several disciplines. The Neotectonic and paleoseismological studies allow to characterize the faults activity by estimating their sliding velocities on a geological time scale and the incremental displacements generated by repeated surface breaks of the strong earthquakes they have generated. The study of historical seismicity, operational seismology and tectonic geodesy complete this knowledge through the characterization of seismic sources and the current mechanical behavior of faults. The comparison of all these observations and measurements allows to estimate the seismic potential of the faults through the determination of the magnitudes and return times of the earthquakes they produce.

2. Prediction of seismic movement

The acquisition of strong seismic movement data and numerical modelling provide a better understanding of the variability of near-field seismic movement, i.e. in the most vulnerable areas.

Principales méthodes de prédiction du mouvement sismique
Main methods for predicting seismic movement

In France, the seismicity is moderate and the establishment of Ground Motion Prediction Equations (GMPEs) is not straightforward as no major earthquake record at close distance exists. The prediction is generally made using links built from databases of other parts of the world with comparable seismotectonic characteristics and compatible with the limited data registered on our territory. This approach is not always applicable due to the fairly quick variation of the attenuation depending on the seismotectonic context, even within a country of the size of France. Ground movement simulation can provide missing estimates for strong earthquakes and produce realistic accelerograms useful for sizing structures.

3. Study of site effects

The term “site effect” refers to the amplification or attenuation of seismic movement at the surface caused by the local characteristics of the site.

Simulation de la propagation des ondes sismiques, sur des distances plurikilométriques.
Simulation of seismic wave propagation over multi-kilometer distances.

These 3 areas of research and development focus on quantifying uncertainties (both in terms of basic data and methods), a prerequisite for probabilistic approaches (assessment such as PSHA – Probabilistic Seismic Hazard Assessment) as an input to Probabilistic Safety Studies.

Probabilistic assessment of seismic hazard

Most of the regulations currently in force in the world, whether for “normal” risk or for industrial or nuclear facilities, explicitly require a probabilistic approach in the assessment of seismic hazard (International Atomic Energy Agency, large dams, normal risk in France, European regulation, Eurocode 8…). The SEISM Institute is working on this probabilistic methodology and has operational tools to assess seismic hazard at a particular site.

The probabilistic method requires a sufficiently detailed description of the seismotectonic model, sometimes difficult to achieve based on knowledge of seismicity and tectonic, especially in contexts of moderate seismicity such as the intraplate domain. There are non-zoning methods that rely solely on the observed seismicity, its spatial correlations and its recurrence rate. These methods assume that future events are likely to occur in the vicinity of those of the past. These principles, although integrated into a probabilistic calculation, are very close to the deterministic approach. Their limit therefore remains that, in less active areas such as metropolitan France, activity is not high enough and the seismic cycle is not short enough for the seismicity sample observed to represent a reliable hypothesis of the seismic potential. In these contexts, methods based on a more complete interpretation of seismotectonics are generally preferred. The counterpart is that all uncertainties due to lack of knowledge and data must be taken into account in order to reflect the most plausible level of confidence. This requires a mix of methodical, wise and consistent expertise at all stages of modelling and calculation. However, it is still difficult to identify and discriminate between the different uncertainty factors in the definition of sources, movement prediction and site effects. Modelling these uncertainties is a key issue in probabilistic assessment.

Exemple de résultat d’étude d’aléa sismique : carte du pic d’accélération (en m/s²) atteint ou dépassé en 1975 ans en France.
Example of seismic hazard study result: map of the peak acceleration (in m/s²) reached or exceeded in 1975 years in France.

The ground movements that the installations could undergo during their lifetime resulting from the hazard assessments are then used for seismic design of the structures.

Link to seismic engineering

A final area of research specific to the interface between seismic hazard and structural vulnerability emerges through the study of “relevant indicators”, i.e. which parameters of seismic movement are relevant to characterize the damage of a given structure or equipment? The use of the response or floor spectrum alone or of the values such as the maximum acceleration peak is not sufficient to describe the potential for damage from a seismic signal.