9:15 AM - 9:30 AM
[SSS11-02] Refining the ETAS model: non-stationary background rate, hypocentral depth, source rupture geometry, and focal mechanism
Keywords:ETAS model, focal mechanism, rupture geometry, background rate
The space-time ETAS model has been widely used as a standard model to analyze local, regional, or even global seismicity. The assumptions of this model include:
1. The magnitudes are identically independently distributed according to the Gutenberg-Richter magnitude-frequency relationship and are independent from other components such as occurrence times and locations.
2. The background seismicity is stationary in time but nonhomogeneous in space of epicenter locations.
3. Each event, no matter whether it is a background event or is triggered by other shocks, triggers its own offspring independently according to some probability rules: the occurrence times of triggered shocks follows the Omori-Utsu formula, and the epicenter locations of the triggered events follow an inverse power law in space.
Many features of seismicity that are not included in the ETAS model have been revealed by analysis of seismic data from different regions, indicating that the ETAS model should be extended to incorporate more information of seismicity. In this talk, the author summarizes the current developments on the extension of the ETAS model, including:
(1) non-stationary background rate, such as long-term trend and seasonality, (2) earthquake depth, (3) geometry of earthquake rupture, (4) earthquake focal mechanisms.
In the implementations of the above extensions, the stochastic declustering and stochastic reconstruction have been used as basic tools to estimate the non-parametric parts.
1. The magnitudes are identically independently distributed according to the Gutenberg-Richter magnitude-frequency relationship and are independent from other components such as occurrence times and locations.
2. The background seismicity is stationary in time but nonhomogeneous in space of epicenter locations.
3. Each event, no matter whether it is a background event or is triggered by other shocks, triggers its own offspring independently according to some probability rules: the occurrence times of triggered shocks follows the Omori-Utsu formula, and the epicenter locations of the triggered events follow an inverse power law in space.
Many features of seismicity that are not included in the ETAS model have been revealed by analysis of seismic data from different regions, indicating that the ETAS model should be extended to incorporate more information of seismicity. In this talk, the author summarizes the current developments on the extension of the ETAS model, including:
(1) non-stationary background rate, such as long-term trend and seasonality, (2) earthquake depth, (3) geometry of earthquake rupture, (4) earthquake focal mechanisms.
In the implementations of the above extensions, the stochastic declustering and stochastic reconstruction have been used as basic tools to estimate the non-parametric parts.