Search

Introduction: Volcano seismology has successfully predicted several eruptions and includes many reliable methods that have been adopted extensively by volcanic observatories; however, there are several problems that still lack solutions. Meanwhile, the overwhelming success of data-driven models to solve predictive complex real-world problems positions them as an effective addition to the monitoring systems deployed in volcanological observatories. Methods: By applying signal processing techniques on seismic records, we extracted four different seismic features, which usually change their trend when the system is approaching an eruptive episode. We built a temporal matrix with these parameters then defined a label for each temporal moment according to the real state of the volcanic activity (Unrest, Pre-Eruptive, Eruptive). To solve the remaining problem developing early warning systems that are transferable between volcanoes, we applied our methodology to databases associated with different volcanic systems, including data from both explosive and effusive episodes, recorded at several volcanic scenarios with open and closed conduits: Mt. Etna, Bezymianny, Volcán de Colima, Mount St. Helens and Augustine. Results and Discussion: This work proposes the use of Neural Networks to classify the volcanic state of alert based on the behaviour of these features, providing a probability of having an eruption. This approach offers a Machine Learning tool for probabilistic short-term volcanic eruption forecasting, transferable to different volcanic systems. This innovative method classifies the state of volcanic hazard in near real-time and estimates a probability of the occurrence of an eruption, resulting in a period from at least hours to several days to forecast an eruption. [ABSTRACT FROM AUTHOR]

An increase in volcanic thermal emissions can indicate subsurface and surface processes that precede, or coincide with, volcanic eruptions. Space-borne infrared sensors can detect hotspots -- defined here as localized volcanic thermal emissions -- in near-real-time. However, automatic hotspot detection systems are needed to efficiently analyze the large quantities of data produced. While hotspots have been automatically detected for over 20 years with simple thresholding algorithms, new computer vision technologies, such as convolutional neural networks (CNNs), can enable improved detection capabilities. Here we introduce HotLINK: the Hotspot Learning and Identification Network, a CNN trained to detect hotspots with a dataset of -3,800 satellite-based, Visible Infrared Imaging Radiometer Suite (VIIRS) images from Mount Veniaminof and Mount Cleveland volcanoes, Alaska. We find that our model achieves an accuracy of 96% (Fl-score 0.92) when evaluated on -1,700 unseen images from the same volcanoes, and 95% (Fl-score 0.67) when evaluated on -3,000 images from six additional Alaska volcanoes (Augustine Volcano, Bogoslof Island, Okmok Caldera, Pavlof Volcano, Redoubt Volcano, Shishaldin Volcano). In comparison with an existing threshold-based hotspot detection algorithm, MIROVA (Coppola et al., Geological Society, London, Special Publications, 2016, 426, 181-205), our model detects 22% more hotspots and produces l2% fewer false positives. Additional testing on -700 labeled Moderate Resolution Imaging Spectroradiometer (MODIS) images from Mount Veniaminof demonstrates that our model is applicable to this sensor's data as well, achieving an accuracy of 98% (F1-score 0.95). We apply HotLINK to 10 years of VIIRS data and 22 years of MODIS data for the eight aforementioned Alaska volcanoes and calculate the radiative power of detected hotspots. From these time series we find that HotLINK accurately characterizes background and eruptive periods, similar to MIROVA, but also detects more subtle warming signals, potentially related to volcanic unrest. We identify three advantages to our model over its predecessors: 1) the ability to detect more subtle volcanic hotspots and produce fewer false positives, especially in daytime images; 2) probabilistic predictions provide a measure of detection confidence; and 3) its transferability, i.e., the successful application to multiple sensors and multiple volcanoes without the need for threshold tuning, suggesting the potential for global application. [ABSTRACT FROM AUTHOR]