Abstract
Piton de la Fournaise volcano, La Réunion, France, erupted between the 2 and 6 April 2020, one of a series of eruptive phases which occur typically two or three times per year. Here, we use back trajectory analysis of satellite data from the TROPOMI instrument to determine that gas emissions during the June 2020 eruption were of unusually high intensity and altitude, producing 34.9 ± 17.4 kt of SO(2) and plume heights up to 5 km a.s.l. The early stages of the eruption (2-4 April 2020) were characterised by relatively low SO(2) emission rates despite strong low frequency tremor (LFT); the latter phase followed an increase in intensity and explosivity in the early hours of 5 April 2020. This period included lava fountaining, significantly increased SO(2) emission rates, increased high frequency tremor (HFT) and decreased LFT. Using the PlumeTraj back trajectory analysis toolkit, we found the peak SO(2) emission rate was 284 ± 130 kg/s on the 6 April. The plume altitude peaked at ~ 5 km a.s.l. on 5 April, in the hours following a sudden increase in explosivity, producing one of the tallest eruption columns recorded at Piton de la Fournaise. PlumeTraj allowed us to discriminate each day's SO(2), which otherwise would have led to a mass overestimate due to the plumes remaining visible for more than 24 h. The eruption exhibited a remarkable decoupling and anti-correlation between the intensity of the LFT signal and that of the magma and gas emission rates. LFT intensity peaked during the first phase with low magma and SO(2) emissions, but quickly decreased during the second phase, replaced by unusually strong HFT. We conclude that the observation of strong HFT is associated with higher intensity of eruption, degassing, and greater height of neutral buoyancy of the plume, which may provide an alert to the presence of greater hazards produced by higher intensity eruptive activity. This might be particularly useful when direct visual observation is prevented by meteorological conditions. This eruption shows the importance of combining multiple data sets when monitoring volcanoes. Combining gas and seismic data sets allowed for a much more accurate assessment of the eruption than either could have done alone. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00445-023-01628-1.