Abstract

Locating sources of volcano-seismic event is very relevant to monitor and comprehend volcanic processes. Ordinary estimation of source seismic events is based on phase picking. The most accurate procedure of phase selection is the visual inspection of the records by experts, who employ local characteristics for phase detection and comparison with observed signals from other stations. This activity is highly time demanding, which in turn is a strong motivation to automatize the epicenter estimation process. However, automatic phase picking in volcano signals is highly inaccurate because of the short distances between the event epicenters and the seismograph stations. In this paper, an end-to-end based LSTM (Long-Short Term Memory) scheme is proposed to address the problem of volcano event localization without any a priori model relating phase picking with localization estimation. LSTM was chosen due to its capability to capture the dynamics of time varying signals, and to remove or add information within the memory cell state and model long-term dependencies. A brief insight into LSTM is also discussed here to justify the use of this neural network. The results presented in this paper show that the LSTM based architecture provided a success rate, i.e., an error smaller than 1.0 km, equal to 48.5%, which in turn is dramatically superior to the one delivered by automatic phase picking. Moreover, the proposed end-to-end LSTM based method gave a success rate (18%) higher than CNN (Convolutional Neural Network). The results presented suggest that the approach proposed here for automatic volcano event epicenter localization can be applied to other geophysics problems.