Tracking seismicity in an underfunded institution: The case of La Soufrière St Vincent volcanic eruption 2020–2021

• Published in: Journal of volcanology and geothermal research 2024-02, Vol.446, p.107990, Article 107990
• Main Authors: Contreras-Arratia, Rodrigo, Chouet, Bernard, Caudron, Corentin, Nath, Nisha, Balchan, Amit, Madoo, Farrah, Ramsingh-Reddock, Hannah
• Format: Article
• Language: English
• Subjects: Caribbean volcano; Fluid-filled crack model; La Soufriere; Seismic interferometry; Volcano monitoring; Volcano seismology
• ISSN: 0377-0273; 1872-6097
• DOI: 10.1016/j.jvolgeores.2023.107990

An unusual volcanic transition was recorded at La Soufrière Volcano in the Lesser Antilles during the 2020–2021 eruptive episode. Initial activity was an effusive phase that lasted three months and turned explosive on 9 April 2021. Although the transition was recognised when it occurred and people were timely evacuated thanks to the detection of subtle geophysical patterns, this work attempts to extract more detailed information from the seismic dataset to improve the monitoring of future events. Due to network limitations and the low amplitudes of the recorded signals, the elaboration of a conceptual model from earthquake locations turned out to be challenging. Accordingly, new analysis methods were developed, including: seismic interferometry using cross-correlation functions to locate the sources of tremor and explosions and a fluid-filled crack model to quantify the bulk modulus of magma stored in the system and translate it to gas content. Both are now applied in quasi-real time for the interpretation of future unrest. The VT locations feature a NNW-SSE alignment across the eruption site with focal depths shallower than 4 km. The banded tremor located at shallow levels (∼1 km depth) and the explosions featured two distinct behaviours, deepening of sources as the eruption progressed, as well as the activation of near-surface sources associated with the eruptive jet. A fluid-filled crack model points to an injection of gas into the system on 9 April, which amounts to approximately 76% gas-volume fraction. We propose a volatile injection equivalent to a 7% gas-volume fraction entering the system on 12 March; however, it was not energetic enough to produce changes in the extrusion rate. Seismological analyses support the conceptual model proposed by Joseph et al. (2022), where a dome sealed the system with a minimal amount of gas exsolved (0.7% gas-volume fraction). For the onset of the explosive phase, the system accumulated enough gases to produce the dome collapse and cleared the vent, decreased the internal pressure and triggered fragmentation.