Barrière, J., d'Oreye, N., Smets, B., Oth, A., Delhaye, L., Subira, J., Mashagiro, N., Derauw, D., Smittarello, D., Syavulisembo Muhindo, A. & Kervyn, F. 2022. ‘Intra‐crater eruption dynamics at Nyiragongo (DR Congo), 2002‐2021’. Geophysical Research: Solid Earth 127(4): e2021JB023858. DOI: 10.1109/IGARSS47720.2021.9553796. URL: https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2021JB023858 I.F. 3.848.
Article dans une revue scientifique / Article dans un périodique
Nyiragongo is one of the rare volcanoes on Earth hosting a lava lake. However, the understanding of its plumbing and lava lake systems remains limited, with, until recently, only sporadic or time-limited historical observations and measurements. Combining dense accurate lava lake and crater floor level measurements based on 1,703 satellite radar images and topographic reconstructions using photogrammetry, we obtain the first reliable picture and time evolution of intra-crater erupted lava volumes between the two last flank eruptions in January 2002 and May 2021. The filling of the crater by lava, initiated in 2002 and continued up to May 2021, is seen as an evidence of a long-term pressure build up of the magmatic system. This filling occurs through irregular pulsatory episodes of rising lava lake level, some of which overflow and solidify on the surrounding crater floor. Pauses of stable molten lava lake level and sudden numerous level drops also marked the summit's eruptive activity. The joint analysis with seismic records available since 2015 revealed that the largest lava lake drops are synchronous with seismic swarms associated with deep magma intrusions, generally preceded by an increase of extrusion rate within the crater. The appearance of a spatter cone in the summit crater in 2016, most likely superficially branched to the lava lake, was a clear marker of the change in eruption dynamics. This first long-term time series of Nyiragongo's crater topography between two hazardous flank eruptions might further help to better decipher Nyiragongo's past and future behavior using multi-parameter observations.