This paper reviews recent progress in data-science-based approaches in volcanology. Firstly, the concept of machine-learning, the basis of recent advances in data science, is described. Recent advances in data science driven by machine-learning techniques have led to a major change to a data-driven approach, whereby models are automatically determined from data, objectively and reproducibly. An overview of the representative statistical analysis methods used in machine-learning approaches is then presented. Next, examples of volcanological studies based on data science approaches are reviewed for each research field. Data science approaches have been applied to almost the entire field of volcanology, leading to significant advances in data processing, analytical accuracy, and modeling of high-dimensional data and nonlinear relationships. Data-science-based approaches have been widely applied, especially in seismology, geochemistry, and geothermobarometry. These approaches are also being applied in geodetic and remote sensing studies, image processing, and analytical data processing, and are being utilized for real-time monitoring and short-term prediction of volcanic eruptions based on time-series data. Finally, an overview of the current status and future perspectives of data-science-based volcanology is presented. In future research, it will be essential to conduct machine-learning analyses that can confirm the interpretations of the outputs and their consistency with geophysical and geological evidence. To gain novel insights into volcanic phenomena using data-science-based approaches, further development and establishment of “data-driven volcanology” is essential.

Volcanic disaster prevention involves protecting human lives and livelihoods from various disasters caused by volcanic eruptions. Disasters caused by volcanic eruptions include direct damage caused by eruption phenomena (ballistic ejecta, ash fall, pyroclastic flows, and lava flows) and sediment-related disaster caused by debris flows and mudflows that occur when erupted pyroclastic material is washed away by rainfall or melting snow. Sabo works (erosion control) are the act of protecting human lives and property from debris flows, landslides, slope failures, and other sediment-related disasters. The work in volcanic areas is called "volcanic sabo works ". Their purpose is to prevent and mitigate sediment-related disaster caused by debris flows, lava flows and pyroclastic flows etc. through structural measures and non-structural measures. Structural measures include sabo dams and sand pockets to capture debris flows and volcanic mudflows, as well as to control their direction. Non-structural measures include monitoring and detection of sediment movement for evacuation, volcanic hazard mapping, provision of disaster prevention information (sediment-related disaster emergency information and real-time hazard map).

The foundation of volcanic disaster prevention is the volcanological research findings on eruption history and transitions in eruptive activity, the accumulation and analysis of data from monitoring and observation of volcanic activity. Based on these research results, eruption scenarios and sediment movement scenarios are developed, and the introduction of eruption alert levels, evacuation plans, and volcanic erosion control plans are considered. This process involves volcanologists, officials of the Japan Meteorological Agency, erosion control departments, and local governments. Therefore, we believe that in addition to cooperation among various institutions, a cooperative framework with volcano experts is essential. In the future, the content and level of precision required for volcanic disaster prevention are expected to increase, making collaboration between volcano researchers and related organizations more important than ever before.

Observations of tephra deposit loads and atmospheric tephra concentrations are crucial for understanding the processes characterizing ash transport: source, advection, diffusion, and deposition. Over the past decade, research on small-scale eruptions has led to the emerging use of optical disdrometers for high temporal resolution and real-time tephra-fall observations and radar for monitoring volcanic plume height and tephra concentrations in the atmosphere. These observations have already begun contributing to the advancement of research on other tephra transport processes, such as tephra-fall modelling using wind fields that account for volcanic topography and understanding the finger phenomenon in the sedimentation process. Other computational research has also progressed in the practical application of a bending plume in a crosswind, real-time estimation of the erupted mass, and probabilistic prediction of tephra fallout. Disdrometer and radar observations are expected to make even greater contributions over the next decade.

This review focuses on zircon chronology using LA-ICP-MS, which is thought to have made particularly remarkable progress in relation to volcanology over the past decade. LA-ICP-MS allows for in-situ analysis, therefore a lot of information can be extracted from a single zircon grain. In particular, double dating using the U-Th-Pb and the (U-Th)/He methods can elucidate with high precision the time scales in the magma process from zircon crystallization to eruption. Zircon petrochronology, which reveals the composition and age from a single zircon grain, has also made great progress. In the future, chronologies using LA-ICP-MS other than zircon will also attract attention, such as the U-Th method for ilmenite and the Rb-Sr and K-Ca methods using a reaction cell.

Several types of textures occur in volcanic rocks: textures as an assemblage of crystals and glass (e.g., intersertal texture, hyalopilitic texture, and diktytaxitic texture), crystal textures (e.g., sieve texture, zoning texture, and skeletal texture), vesicular textures, and glass textures (e.g., eutaxitic texture and liquid immiscibility). Recent textural studies of volcanic rocks have enabled us to understand the time scales of eruptive activity (e.g., residence time, decompression rate, and cooling rate) and the conditions under which textures are generated. In this review paper, I aim to provide a broad overview of the progress in each textural study of volcanic rocks over the past decade. Recent studies on textures in volcanic rocks have investigated the microlite number density (MND), crystal size distribution (CSD), nanocrystals (nanolite and ultrananolite), crystals on sample surface, crystal shapes (morphology/habit/tracht), broken crystals, diffusion chronometry (zoning in phenocrysts, melt embayment, and chlorine contents in glass), bubble number density (BND), and liquid immiscibility. In particular, diffusion chronometry has developed remarkably and several review papers have been published. Among them, the cold storage model of the magma reservoir shown by diffusion chronometry studies and the determination of the decompression rate by melt embayment diffusion chronometry have garnered attention. Theoretical studies on MND, CSD, and BND have progressed in terms of heterogeneous nucleation. Understanding the changes in crystal shape in magma reservoirs and during the magma ascent process has improved. The crystallization processes in the lava dome, as represented by the diktytaxitic texture, have been revealed. Experimental studies on crystallization in flow fields and crystal nucleation and growth processes are currently in progress. Although we are currently examining various assumptions and identifying problems in various textural studies, textural studies of volcanic rocks can be used to elucidate volcanic eruption mechanisms and monitor active volcanoes in the near future.

The study of large-scale explosive eruptions that form collapse calderas has progressed dramatically in the last decade. Several caldera systems have been studied using geological and petrological approaches to investigate the development of long-term magma reservoirs as a preparatory process for large-scale eruptions. The concept of caldera cycle is useful for understanding the long-term activity of caldera volcanoes with repetition of large-scale caldera-forming eruptions. Several calderas have been studied for their eruptive phenomena that immediately preceded a major caldera-forming eruption. However, the development of the magmatic system and its relationship with precursory eruptive activities is still poorly understood. The eruptive processes of several caldera systems in Japan, such as those at Kutcharo, Shikotsu, Toya, Aso, Aira and Kikai, have been intensely studied, using geological and petrological approaches. These studies reveal the detailed sequence of caldera-forming, large-scale, explosive eruptions. In addition, the eruption history of smaller calderas, such as Ikeda and Nigorigawa, has also been studied, highlighting the similarity and contrast against the larger calderas. Theoretical studies on the collapse mechanisms of calderas that cause these large-scale eruptions have also made significant progress in the past decade. Furthermore, geophysical surveys have been conducted to gather information on the deep structures and activity of caldera volcanoes such as Aira, Aso and Kutcaharo. However, due to the lack of real-world observations of large-scale explosive eruptions, our knowledge of such eruptions relies primarily on the geological and petrological analysis of past eruptions. In the next decade, we should aim to integrate these geological and petrological data with geophysical and geochemical observations to improve our understanding of the current activity of caldera volcanoes and evaluate their future activity.

Volcanoes on other worlds outside our Earth provide a good opportunity to compare with terrestrial ones and deepen our understanding of fundamental volcanology. The Japanese volcanology community, however, has stood back from the progress of volcanology on the other planetary bodies in the recent 20 years. This paper aims to rekindle scientific enthusiasm for planetary volcanology among the volcanology community in Japan. To prepare for approaching opportunities such as human space exploration on the other volcanic world, we should now create an environment for discussion on planetary volcanology in Japan.