Characterization of volcanic stages using seismic features: Case of Tajogaite (2021) and Colima (2013−2022)

This study proposes a novel approach to volcano seismic monitoring by analyzing the time evolution of specific seismic parameters—Shannon Entropy (SE), Frequency Index (FI), and Kurtosis—with the aim of improving the forecasting, characterization, and definition of the end of eruptive episodes. We focus on two well-studied volcanoes: Volcán de Fuego de Colima (Mexico) and the Tajogaite eruption (La Palma, Spain) that occurred in 2021. For Volcán de Fuego de Colima, we analyze a 10-year seismic dataset (2013–2022) that spans periods of high volcanic activity—including dome growth, lava flows, and explosive events—as well as intervals of relative quiescence. Our analysis suggests that SE is a useful parameter for identifying the onset of intense eruptive phases, dome growth episodes, and transitions to dormant stages. When combined with FI and Kurtosis, these parameters provide insights into specific eruptive processes such as magma ascent, conduit pressurization, and fracture generation. SE decreases when seismic signals become more organized, which typically occurs just before eruption. FI highlights changes in dominant frequency bands associated with different types of activity, while Kurtosis detects impulsive seismic events. These features thus help constrain the interpretation and potential modeling of subsurface eruptive mechanisms . In the case of the 2021 Tajogaite eruption, our method forecasted the eruption at least nine hours in advance. The joint analysis of SE, FI, and Kurtosis also allowed us to determine the end of the eruptive process, marked by a clear change in SE that coincided with the last visual evidence of volcanic activity. Overall, this study highlights the potential of combining these seismic parameters for real-time volcano monitoring, eruption forecasting, and the characterization and modeling of eruptive processes, ultimately contributing to the improvement of early warning protocols in volcanic hazard management. • Shannon Entropy decay precedes eruptions, enabling reliable early warnings. • Kurtosis detects impulsive seismic events, swarms, and explosive phases. • Frequency Index tracks high-frequency VT fracturing and tremor dynamics. • Combined SE, FI, and Kurtosis enhance eruption forecasting and monitoring. • Rising SE after eruptions signals a return to volcanic quiescence.