Earthquake Early Warning Systems (EEWS) are critical tools for regions prone to seismic activity. However, their widespread adoption is hampered by the high cost of traditional systems, particularly in low-income areas. Recently, researchers have proposed low-cost alternatives, such as smartphone-based EEWSs, despite the reliability challenges of smartphones. This work presents a statistical methodology for estimating key earthquake parameters using smartphone data. Borrowing from survival data analysis, a Bayesian cure model is proposed that treats smartphones as patients in a clinical trial, with earthquake detection as the censoring event. Incorporating spatial and temporal data, a mixture of parametric densities is developed to represent detectable earthquake waves. The model is fitted using an adaptive Markov chain Monte Carlo algorithm. A real-world case study demonstrates the robustness of the model and provides insights into smartphone-based earthquake monitoring.
Finazzi, F., Aiello, L., Paci, L., A Bayesian cure model for earthquake parameter estimation using crowdsourced smartphone data, in Methodological and Applied Statistics and Demography III, (Bari, 17-20 June 2024), Springer, Cham 2024:<<ITALIAN STATISTICAL SOCIETY SERIES ON ADVANCES IN STATISTICS>>, 665-670. [10.1007/978-3-031-64431-3] [https://hdl.handle.net/10807/310738]
A Bayesian cure model for earthquake parameter estimation using crowdsourced smartphone data
Paci, Lucia
2024
Abstract
Earthquake Early Warning Systems (EEWS) are critical tools for regions prone to seismic activity. However, their widespread adoption is hampered by the high cost of traditional systems, particularly in low-income areas. Recently, researchers have proposed low-cost alternatives, such as smartphone-based EEWSs, despite the reliability challenges of smartphones. This work presents a statistical methodology for estimating key earthquake parameters using smartphone data. Borrowing from survival data analysis, a Bayesian cure model is proposed that treats smartphones as patients in a clinical trial, with earthquake detection as the censoring event. Incorporating spatial and temporal data, a mixture of parametric densities is developed to represent detectable earthquake waves. The model is fitted using an adaptive Markov chain Monte Carlo algorithm. A real-world case study demonstrates the robustness of the model and provides insights into smartphone-based earthquake monitoring.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.