Bayesian Inference on the Magnitude of the Largest Expected Earthquake

Date: 4/24/2019

Time: 06:00 PM

Room: Grand Ballroom

Majority of earthquakes occur unexpectedly and can trigger subsequent sequences of events that can culminate in more powerful earthquakes. This self-exciting nature of seismicity generates complex clustering of earthquakes in space and in time. Therefore, the problem of constraining the magnitude of the largest expected earthquake during a future time interval is of critical importance in mitigating earthquake hazard. We address this problem by developing a methodology to compute the probabilities for such extreme earthquakes to be above certain magnitudes. We combine the Bayesian analysis with extreme value statistics to compute the Bayesian predictive distribution for the magnitude of the largest event to exceed a certain value in the near future. In the analysis, we assume that the earthquake occurrence rate can be modelled by the ETAS process, where each earthquake is capable of triggering subsequent events. To model the uncertainties of the model parameters, we employ the Markov Chain Monte Carlo method to sample the posterior distribution of the model parameters and use the generated chain of the parameters to simulate forward in time an ensemble of the ETAS processes. To illustrate our approach, we analyzed one recent prominent sequence, the 2016 Kumamoto, Japan, earthquake sequence, where we were able to compute the probabilities of having the largest expected events above certain magnitudes to occur during several stages of the sequence. As a main result of this work, we developed and tested an inference procedure to estimate the probabilities of having largest expected events during an earthquake sequence governed by the ETAS process. The suggested approach can be implemented in current or future operational earthquake forecasting schemes, where the constraints on the magnitudes of future large earthquakes are taken into account.

Presenting Author: Robert Shcherbakov

Authors