Volcanic eruptions are spectacular manifestations of natural forces that dynamically shape our planet. Their impacts spread from the geosphere to the hydrosphere and the atmosphere with the potential to have severe consequences at global scale. Within the volcanological community, forecasting volcanic eruptions remains a primary goal in volcanic hazards and risk mitigation. Over the past decades, the quantity and resolution of observations and the quality of monitoring resources have steadily increased, providing a wealth of data on the underlying physical processes that drive volcanic eruptions. Novel technological advancements have significantly broadened both the spatial coverage and frequency bandwidth of geochemical and geophysical observations at active volcanoes. By integrating multi-parametric data from both ground and space, scientists now gain an unprecedented vision of the surface manifestations of mass transport beneath volcanoes as well as the internal structure from static and functional imaging techniques. This enables the detection and tracking of subtle signals of volcanic unrest prior to eruption, even at remote or inaccessible volcanoes. These advancements have been accompanied by new models and processing techniques including with artificial intelligence and machine learning, leading to innovative paradigms for the interpretation and inversion of observational data (geophysical, geochemical, geological). Within this context, this session aims to convene a multidisciplinary audience for discussing the most recent innovations in monitoring tools and to present observations, methods, and models that enhance our understanding of volcanic processes fostering our capabilities in volcanic early warnings and risk reduction.