 |
| Titel |
Retrospective performance investigation of the predictive models of statistical seismology for the May 12, 2008, Wenchuan earthquake: PI, RI, AMR, and quiescence |
| VerfasserIn |
C. Jiang, Z. Wu |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250020839
|
|
|
|
|
|
| Zusammenfassung |
| To explore whether there were some indications of the increase of strong earthquake
probabilities before the May 12, 2008, Wenchuan earthquake, we conducted retrospective
forecast test of some of the predictive models of statistical seismology. Earthquake catalogue
down to ML3.0 from 01/01/1977 to 15/06/2008 in the Sichuan-Yunnan region is used. For
the investigation of the Pattern Informatics (PI) algorithm and the Relative Intensity (RI)
algorithm, we consider a 15-year long ‘sliding time window’, with ‘anomaly training time
window’ and ‘forecast time window’ being both 5 years. Grid size is taken as 0.2 degrees.
Pre-shock PI ‘hotspots’ can be seen along the northern Longmenshan fault which is
responsible for the Wenchuan earthquake. On the other hand, when considering bigger grid
size and higher cutoff magnitude, such ‘hotspots’ disappeared and there was almost no
indication of the approaching of this great event. In the perspective of the newly proposed
migration pattern, however, there seem to be clues to the preparation of the earthquake.
To systematically analyze the precursory moment release, considering the special
characteristics of this great inland thrust event, we took four upgraded approaches. We
propose a BIC criterion as a development of the curvature parameter of Bowman et al.
(1998) to identify the statistically significant acceleration or quiescence behavior
as compared to linear increase. We use an ‘eclipse method’ as a development of
the ‘interfering event consideration’ to eliminate the interference from the nearby
seismically active fault zones. We consider the distribution of m-coefficient in the
(T,R,MC) space, to explore the variation of moment release behavior with temporal
window length T and spatial window radius R centered at the nucleation point, and
cutoff magnitude MC of the catalogue in use. We use not only circular windows
but also ‘crack-like’ windows to investigate the overall behavior of the moment
release associated with the Wenchuan earthquake. It is observed that moment release
before the Wenchuan earthquake did show accelerating moment release (AMR) and
quiescence behavior for different spatio-temporal ranges, respectively, with robustness to
some extent against the changing of parameters. However, for this earthquake, to
constrain the failure time in a forward forecasting mode is shown to be difficult. |
|
|
|
|
|
|