TY - JOUR
T1 - Recurrent Scattering Network Detects Metastable Behavior in Polyphonic Seismo-Volcanic Signals for Volcano Eruption Forecasting
AU - Bueno Rodriguez, Angel
AU - Balestriero, Randall
AU - De Angelis, Silvio
AU - Benitez, M. Carmen
AU - Zuccarello, Luciano
AU - Baraniuk, Richard
AU - Ibanez, Jesus M.
AU - De Hoop, Maarten V.
N1 - Funding Information:
This work was supported in part by KNOWAVES under Grant TECT2015-6872, in part by FEMALE under Grant PID2019-106260GB-I00, in part by the Department of Energy under Grant DE-SC0020345, in part by the Simons Foundation MATH+X Program, and in part by the Geo-Mathematical Imaging Group (Rice University).
Publisher Copyright:
© 1980-2012 IEEE.
PY - 2022
Y1 - 2022
N2 - We introduce an end-to-end (E2E) deep neural network architecture designed to perform seismo-volcanic monitoring focused on detecting change. Due to the complexity of volcanic processes, this requires a polyphonic detection, segmentation, and classification approach. Through evolving epistemic uncertainty, invoking a Bayesian network strategy, we detect change and demonstrate its significance as an indicator for possible forecasting of eruptions using data from the Bezymianny and Etna volcanoes. Specifically, we propose morphing the scattering transform from previous work into a novel E2E hybrid and recurrent learnable deep scattering network to adapt to multi-scale temporal dependencies from streaming data. The time-dependent scattering is in some sense physics informed, namely, through time-frequency representation (TFR) of the data. At the same time, with a carefully designed deep convolutional LSTM (ConvLSTM) architecture, we learn intra-event, temporal dynamics from the scattering coefficients or features. We verify the effectiveness of transfer learning switching between volcanoes. Our experimental results set a new norm for semi-supervised seismo-volcanic monitoring.
AB - We introduce an end-to-end (E2E) deep neural network architecture designed to perform seismo-volcanic monitoring focused on detecting change. Due to the complexity of volcanic processes, this requires a polyphonic detection, segmentation, and classification approach. Through evolving epistemic uncertainty, invoking a Bayesian network strategy, we detect change and demonstrate its significance as an indicator for possible forecasting of eruptions using data from the Bezymianny and Etna volcanoes. Specifically, we propose morphing the scattering transform from previous work into a novel E2E hybrid and recurrent learnable deep scattering network to adapt to multi-scale temporal dependencies from streaming data. The time-dependent scattering is in some sense physics informed, namely, through time-frequency representation (TFR) of the data. At the same time, with a carefully designed deep convolutional LSTM (ConvLSTM) architecture, we learn intra-event, temporal dynamics from the scattering coefficients or features. We verify the effectiveness of transfer learning switching between volcanoes. Our experimental results set a new norm for semi-supervised seismo-volcanic monitoring.
KW - Recurrent neural networks (RNNs)
KW - seismology
KW - uncertainty
KW - volcanoes
KW - wavelet transforms
UR - http://www.scopus.com/inward/record.url?scp=85121336134&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85121336134&partnerID=8YFLogxK
U2 - 10.1109/TGRS.2021.3134198
DO - 10.1109/TGRS.2021.3134198
M3 - Article
AN - SCOPUS:85121336134
SN - 0196-2892
VL - 60
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
ER -