 |
| Titel |
Completeness period analysis of SisFrance macroseismic database and interpretation in the light of historical context |
| VerfasserIn |
José Bonnet, Thibault Fradet, Paola Traversa, Christine Tuleau-Malot, Patricia Reynaud-Bouret, Thomas Laloe, Kevin Manchuel |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250096608
|
| Publikation (Nr.) |
 EGU/EGU2014-12118.pdf |
|
|
|
|
|
| Zusammenfassung |
| In metropolitan France the deformation rates are slow, implying low to moderate seismic
activity. Therefore, earthquakes observed during the instrumental period (since 1962), and
associated catalogs, cannot be representative of the seismic cycle for the French metropolitan
territory. In such context it is necessary, when performing seismic hazard studies, to
consider historical seismic data in order to extend the observation period and to
be more representative of the seismogenic behavior of geological structures. The
French macroseismic database SisFrance is jointly developed by EDF (Electricité de
France), BRGM (Bureau de Recherche Géologique et Minière) and IRSN (Institut de
Radioprotection et Sureté Nucléaire). It contains more than 6,000 events inventoried between
217 BC and 2007 and more than 100,000 macroseismic observations. SisFrance
is the reference macroseismic database for metropolitan France. The aim of this
study is to determine, over the whole catalog, the completeness periods for different
epicentral intensity (Iepc) classes≥IV. Two methods have been used: 1) the method of
Albarello et al. [2001], which has been adapted to best suit the French catalog,
and 2) a mathematical method based on change points estimation, proposed by
Muggeo et al. [2003], which has been adapted to the analysis of seismic datasets.
After a brief theoretical description, both methods are tested and validated using
synthetic catalogs, before being applied to the French catalog. The results show that
completeness periods estimated using these two methods are coherent with each
other for events with Iepc ≥IV (1876 using Albarello et al. [2001] method and
1872 using Muggeo et al. [2003] method) and events with Iepc ≥V (1852 using
Albarello et al. [2001] method and 1855 using Muggeo et al. [2003] method). Larger
differences in estimated completeness period appear when considering events with
Iepc ≥VI (around 30 years difference) and events with Iepc ≥VII (around 50 years
difference). These could be explained (1) by the differences in the way each method
approaches the data; Muggeo et al. [2003] method estimates all change points within data
series, whereas the method of Albarello et al. [2001] focus on the last one, and
(2) by a more limited number of data for these epicentral intensity classes (2056
events with Iepc ≥IV and 1252 events with Iepc ≥V vs. 486 events with Iepc ≥VI
and 199 events with Iepc ≥VII). Results obtained for epicentral intensity classes
greater than VIII are considered not reliable due to the short number of existing
data (around 30 events). The completeness periods determined in this study are
discussed in the light of their contemporary historical context, and in particular of the
evolution of the information available from historical archives since the 17th century. |
|
|
|
|
|
|