Studying for Master of Science
Area of study:
Modelling spatio-temporal patterns and magnitude of non-volcanic tremor using a hidden Markov model framework
Supervisor: Ting Wang—
Title: Modelling spatio-temporal patterns and magnitude of non-volcanic tremor using a hidden Markov model framework
Supervisor: Ting Wang
Non-volcanic tremor is a sequence of weak seismic activity. Features of the migration and recurrence of tremor activity are related to larger destructive earthquakes. Thus, further understanding of tremor patterns can inform future hazard assessment.
Wang et al. (2017) developed a hidden Markov model (HMM) for sparse time series with many null events to investigate the spatio-temporal migration of tremor clusters by combining a binary variable and a continuous variable in the HMM framework. Here, we propose to extend this model to include event sizes. This requires development of the theory of a mixture distribution of discrete and continuous variables, and then use of this type of distribution in a HMM framework again allowing for many null events. We will look for further information on the patterns of events occurring in sub-systems of the Tokai, Kii and Shikoku regions of Japan.