Methodological Advances in Volcanology: The Role of Artificial Intelligence in Volcano Monitoring, Modelling, and Hazard Assessment

This chapter explores the opportunities and challenges of using Artificial Intelligence (AI) and Machine Learning (ML) in volcanology. It starts by introducing the basic concepts of AI and ML. Then, it discusses the current and potential applications of AI and ML in volcanology, including recent advances in petrology, geophysics, remote sensing, and ground monitoring. We highlight that AI and ML can potentially have a transformative effect in understanding volcanic systems, from deciphering the architecture of magma feeding systems and pre-eruptive processes to eruption forecasting. However, the success of AI in volcanology relies heavily on having access to extensive, cohesive, and high-quality data sources for both training and testing models. Data scarcity, noise, and interpretability remain key challenges. Furthermore, many volcanoes lack comprehensive and multi parametric monitoring networks, limiting AI’s full potential. Education also needs to evolve, with universities offering curricula that include AI and ML skills to prepare future researchers for an AI-aware future. Understanding the limitations of and pitfalls associated with these tools, will be important.