P. Martin Mai
King Abdullah University of Science and Technology
|The Source Inversion Validation (SIV) Project: Uncertainty Quantification and Quality Appraisal for Finite-Fault Earthquake Models|
|着任セミナー「建築物の地震災害軽減に向けて －これまでを振り返り、これからを 考える」|
|Seismic source characterization from global and specific earthquake analyses|
Institut des Sciences de la Terre, Grenoble
Lamont-Doherty Earth Observatory, Columbia University
|Investigating lava rheology using experiments, numerical models and field observations|
University of Colorado at Boulder
Insights on the rupture process of earthquakes can come from exhaustive analyses, where the number of earthquakes enables the extraction of generic behaviours, or from detailed studies of well-instrumented earthquakes, where specific or unusual characteristics can be determined.
An exhaustive view of the earthquake source can be based on the study of the moderate-to-large magnitude seismicity (above magnitude 5.5-6), well recorded by the global broadband networks (IRIS, Geoscope…) since the beginning of the 1990’s. This represents today a representative catalogue of several thousands of earthquakes, for which the behaviour of the earthquakes can be studied as a function of - for example - magnitude, depth or tectonic context. The detailed analysis of each event of this catalogue is likely out of reach, but some robust characteristics can be extracted from the moment rate functions (or source time functions – STFs). We will here briefly explain one of the techniques able to retrieve these STFs (SCARDEC method; Vallée et al., 2011), before showing how these STFs inform us about generic behaviours of the rupture process. Stress drop, strain drop or seismic energy are physical quantities that can be robustly estimated from STFs, and we will explore how the earthquake context affect these values. A special focus will be put on the influence of depth, which indicates that the strain drop does not change significantly with magnitude and depth (Vallée, 2013).
The latter approach finds its limits when we want to detect more than integral characteristics of the rupture process. In this case, other approaches (array analysis, empirical Green function techniques…) and geophysical data (strong motion records, geodesy) can be used in order to finely describe how rupture developed on the fault plane. These detailed analyses are able to detect specific properties of the earthquake process, as the existence of supershear velocities (e.g. Vallée and Dunham, 2012), the delayed rupture of some seismic asperities or the fact that simple earthquake recurrence models may fail in predicting earthquake repeat times (Vallée and Satriano, 2014).
Vallée, M., J. Charléty, A.M.G. Ferreira, B. Delouis, and J. Vergoz, SCARDEC : a new technique for the rapid determination of seismic moment magnitude, focal mechanism and source time functions for large earthquakes using body wave deconvolution, Geophys. J. Int., 184, 338-358, 2011.
Vallée, M., and E.M. Dunham, Observation of far-field Mach waves generated by the 2001 Kokoxili supershear earthquake, Geophys. Res. Lett., 39, L05311, 2012.
Vallée, M., Source time function properties indicate a strain drop independent of earthquake depth and magnitude, Nature Communications, doi: 10.1038/ncomms3606, 2013.
Vallée, M., and C. Satriano, Ten-year recurrence time between two major earthquakes affecting the same fault segment, Geophys. Res. Lett., 41, 2312-2318, 5, 2014.