Lentas K., Ferreira A. M. G., Vallée Martin. (2013). Assessment of SCARDEC source parameters of global large (M-w >= 7.5) subduction earthquakes. Geophysical Journal International, 195 (3), p. 1989-2004. ISSN 0956-540X.
Titre du document
Assessment of SCARDEC source parameters of global large (M-w >= 7.5) subduction earthquakes
Geophysical Journal International, 2013,
195 (3), p. 1989-2004 ISSN 0956-540X
Rapid and accurate source characterizations of large, shallow subduction earthquakes are key for improved tsunami warning efforts. We assess the quality of source parameters of large magnitude (M-w >= 7.5) shallow subduction earthquakes of the past 20 yr determined using SCARDEC, a recent fully automated broad-band body-wave source inversion technique for the fast estimation of the moment magnitude, depth, focal mechanism and source time functions of global events. We find that SCARDEC source parameters agree well with those reported in the global centroid moment tensor (GCMT) catalogue, with only the fault dip angle showing a tendency for steeper SCARDEC dip values than GCMT. We investigate this discrepancy through independent validation tests of the source models by: (i) testing how well they explain data not used in their construction, notably low-frequency normal mode data; and, (ii) assessing the data fit using 3-D forward modelling tools more sophisticated than those used to build the source models; specifically, we use a spectral element method for a 3-D earth model. We find that SCARDEC source parameters explain normal mode data reasonably well compared to GCMT solutions. In addition, for the 3-D earth model used in our experiments, SCARDEC dip angles explain body-wave data similarly or slightly better than GCMT. Moreover, SCARDEC dip angles agree well with results from individual earthquake studies in the literature and with geophysical constraints for different subduction zones. Our results show that SCARDEC is a reliable technique for rapid determinations of source parameters of large (M-w >= 7.5) subduction earthquakes. Since the SCARDEC method provides realistic source time functions allowing the fast identification of long-duration tsunami earthquakes, it is complementary to existing methods routinely used for earthquake monitoring and suitable for ocean-wide tsunami warning purposes.
