Survival modelling of smartphone trigger data in crowdsourced seismic monitoring: with applications to the 2023 Pazarcik and 2019 Ridgecrest earthquakes

Abstract Crowdsourced smartphone-based earthquake early warning systems have recently emerged as reliable alternatives to more expensive solutions based on scientific instruments. For example, during the deadly 2023 Pazarcik event in Turkey, the system implemented by the Earthquake Network citizen science initiative provided up to 58 s of warning to people exposed to life-threatening ground shaking. We develop a statistical methodology based on a survival mixture cure model that provides full Bayesian inference on epicentre, depth, and origin time, and we design a tempering Markov chain Monte Carlo algorithm to account for the multi-modality of the posterior distribution. The methodology is applied to data collected by the Earthquake Network during three seismic events, including the 2023 Pazarcik and 2019 Ridgecrest earthquakes.