ETAS= Epidemic Type Aftershock Sequence
Utilizzato anche per foreshocks, ma in terremoti continentali non predice pe rniente bene:
http://pubs.usgs.gov/of/2005/1131/UJNR_ThursFri/Jordan_1/UJNR_Jordan_EPR.pdfUna trattazione dettagliata del modello, ceh conclude:
Taking these results all to-
gether, this suggests that the physics of aftershocks is
sufficient to explain the properties of foreshocks, and
that there is no essential physical difference between
foreshocks, aftershocks and mainshocks.
abstract:
Foreshock probability in Southern California explained by clustering models
Zhuang, J.; Jackson, D. D.
American Geophysical Union, Fall Meeting 2006, abstract #S13A-0213
The foreshock is one of the most popular issues in the researches on seismicity and earthquake prediction, and even of public attentions. Up to now, there are still many problems on foreshocks under discussions. The first problem in foreshock studies is how to define foreshocks from other shocks. The second problem is whether a foreshock is a mainshock whose aftershocks happen to be large. The third problem is how to make use of foreshocks in prediction. To make a good understanding to the above problems, we need a good model of earthquake clustering. In the ETAS model, it is assumed that the seismicity can be divided into a background component and a clustering component, and that each event, no matter it is from the background or it is directly triggered by another event, triggers its own offspring according to some general rules. In this study, we firstly give a brief description of necessary concepts and methodologies associated with the ETAS model in this study, such as triggering ability, thinning operation, stochastic declustering. We secondly concentrate on the discrepancies of the ETAS model and earthquake data, by which we proposed an improved version of the model. Then, different from conventional studies, we re-define foreshocks as events that are from the background and that have at least one larger descendant in all the generation. Making use of the ETAS model and the updated version, we investigate the features of foreshocks in the Southern California region. it is found that the ETAS model tends to produce more foreshocks (about 20% in the background events) than in the real data (about 10% in the background events). Such discrepancies could be explained by a more complicated clustering model, where the differences in clustering characteristics between background events and clustering events are considered.
Keywords: 7209 Earthquake dynamics (1242), 7223 Earthquake interaction, forecasting, and prediction (1217, 1242), 7260 Theory, 7290 Computational seismology
