The characterization of volcano state is not a simple task due the complexity of physics processes underway. Understanding their evolution prior to and during eruptions is a critical point for identifying transitions in volcanic state. Recent developments in the field of Machine Learning (ML) have proven to be very useful and efficient for automatic discrimination, decision, prediction, clustering and information extraction in many fields, including volcanology. In Romano et al. (2022) the use of ML algorithms led to classify strain VLP families related with changes in volcano dynamics prior of paroxysmal eruptions: algorithms have been able to discriminate little differences in VLPs shape and to find a correspondence among a higher number of families and volcanic phenomenologies. For paroxysmal events occurring outside any long-lasting eruption, the initial success of our approach, although applied only to the few available examples, could permit us to anticipate the time of alert to several days, instead of few minutes, by detecting medium-term strain anomalies: this could be crucial for risk mitigation for inhabitants and tourists. The neural network method used in previous analysis has been extended to a wider (2007-2022) period to verify that families found in the previous narrower time interval were still present. We tried, then, to associate families with volcanic activity, confirming the conceptual model previously introduced (Mattia et al., 2021 and Romano et al., 2022), capable of explaining the changes found. Our innovative analysis of dynamic strain, systematically conducted on several years of available data, may be used to provide an early-warning system also on other open conduit active volcanoes. Valuable information is embedded in the data used in the current work, which could be used not only for scientific purposes but also by civil protection for monitoring reasons. Such a variety of possible usage needs the setting of principles and legal arrangements to be implemented in order to ensure that data will be properly and ethically managed and in turn can be used and accessed by the scientific community.
Dynamic strain anomalies detection at Stromboli from 2007 eruptive phase using machine learning
Romano, PierdomenicoMembro del Collaboration Group
;Di Lieto, BellinaMembro del Collaboration Group
;Scarpetta, SilviaMembro del Collaboration Group
;Messuti, GiovanniMembro del Collaboration Group
;Scarpa, RobertoMembro del Collaboration Group
2023-01-01
Abstract
The characterization of volcano state is not a simple task due the complexity of physics processes underway. Understanding their evolution prior to and during eruptions is a critical point for identifying transitions in volcanic state. Recent developments in the field of Machine Learning (ML) have proven to be very useful and efficient for automatic discrimination, decision, prediction, clustering and information extraction in many fields, including volcanology. In Romano et al. (2022) the use of ML algorithms led to classify strain VLP families related with changes in volcano dynamics prior of paroxysmal eruptions: algorithms have been able to discriminate little differences in VLPs shape and to find a correspondence among a higher number of families and volcanic phenomenologies. For paroxysmal events occurring outside any long-lasting eruption, the initial success of our approach, although applied only to the few available examples, could permit us to anticipate the time of alert to several days, instead of few minutes, by detecting medium-term strain anomalies: this could be crucial for risk mitigation for inhabitants and tourists. The neural network method used in previous analysis has been extended to a wider (2007-2022) period to verify that families found in the previous narrower time interval were still present. We tried, then, to associate families with volcanic activity, confirming the conceptual model previously introduced (Mattia et al., 2021 and Romano et al., 2022), capable of explaining the changes found. Our innovative analysis of dynamic strain, systematically conducted on several years of available data, may be used to provide an early-warning system also on other open conduit active volcanoes. Valuable information is embedded in the data used in the current work, which could be used not only for scientific purposes but also by civil protection for monitoring reasons. Such a variety of possible usage needs the setting of principles and legal arrangements to be implemented in order to ensure that data will be properly and ethically managed and in turn can be used and accessed by the scientific community.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.