NH2.3 | Technologies for Forecasting Volcanic Hazards
EDI
Technologies for Forecasting Volcanic Hazards
Co-organized by GMPV9, co-sponsored by AGU
Convener: Ciro Del Negro | Co-conveners: Alessio Alexiadis, Eleonora Amato, Federica Torrisi

When a volcano erupts, providing information on hazardous volcanic phenomena, their effects, and the eruption's duration is crucial. However, eruptions are complex phenomena governed by interactions of many processes, which are often nonlinear and stochastic. Numerous uncertainties in the involved parameters make precise predictions of specific events in time and space usually unattainable; that is, volcanic eruptions can be intrinsically unpredictable. Despite these limitations, significant progress has been made in forecasting volcanic hazards and, in specific circumstances, in making predictions. Improvements in forecasting are closely related to the wealth of data from enhanced monitoring techniques, such as satellite observations, and tremendous advances in computing power. This has led to the increased use of data-driven approaches, including artificial intelligence (AI) techniques, to address volcanic hazards. Machine learning, a type of AI in which computers learn from data, is gaining importance in volcanology not only for monitoring purposes (i.e., in real time) but also for hazard analysis (e.g., modeling tools). Looking to the future, AI models can be combined with physical constraints to bridge the gap between data-driven methods and physical modeling, thereby increasing the interpretability of AI predictions. This offers an alternative approach to dealing with the strongly nonlinear and time-dependent character of volcanic phenomena. Several hybrid strategies, utilizing growing computational resources, are currently being developed to achieve greater flexibility and full synergy between numerical physics-based simulations, machine learning, and data-driven approaches. This multidisciplinary session invites contributions focusing on enhancing traditional ground-based volcano monitoring systems through technological innovation in satellite remote sensing and computational methods, integrating deep-learning, data-driven approaches, and physics-based simulations, to better understand and forecast volcanic hazards.

When a volcano erupts, providing information on hazardous volcanic phenomena, their effects, and the eruption's duration is crucial. However, eruptions are complex phenomena governed by interactions of many processes, which are often nonlinear and stochastic. Numerous uncertainties in the involved parameters make precise predictions of specific events in time and space usually unattainable; that is, volcanic eruptions can be intrinsically unpredictable. Despite these limitations, significant progress has been made in forecasting volcanic hazards and, in specific circumstances, in making predictions. Improvements in forecasting are closely related to the wealth of data from enhanced monitoring techniques, such as satellite observations, and tremendous advances in computing power. This has led to the increased use of data-driven approaches, including artificial intelligence (AI) techniques, to address volcanic hazards. Machine learning, a type of AI in which computers learn from data, is gaining importance in volcanology not only for monitoring purposes (i.e., in real time) but also for hazard analysis (e.g., modeling tools). Looking to the future, AI models can be combined with physical constraints to bridge the gap between data-driven methods and physical modeling, thereby increasing the interpretability of AI predictions. This offers an alternative approach to dealing with the strongly nonlinear and time-dependent character of volcanic phenomena. Several hybrid strategies, utilizing growing computational resources, are currently being developed to achieve greater flexibility and full synergy between numerical physics-based simulations, machine learning, and data-driven approaches. This multidisciplinary session invites contributions focusing on enhancing traditional ground-based volcano monitoring systems through technological innovation in satellite remote sensing and computational methods, integrating deep-learning, data-driven approaches, and physics-based simulations, to better understand and forecast volcanic hazards.