Hourly SO2 emissions and plume dispersion simulated by inverse modelling using TROPOMI, OMPS, IASI, and ground-based LIDAR observations: case studies of the 2021 Etna and 2018 Ambrym eruptions

Volcanic sulphur dioxide (SO₂), a precursor of sulphate aerosols, can have a deleterious impact on the atmosphere, ecosystems, and air quality at multiple scales. Knowledge of highly variable volcanic SO₂ emissions, i.e., mass flux rates and injection heights, would not only aid comprehension of such atmospheric implications but would also provide information on subterranean volcanological processes of magma transport. Furthermore, volcanic SO₂, which frequently co-exists in volcanic plumes with ash and sulphate aerosols, can pose a threat to aviation as ash and acidic aerosols are alarming to aircraft. Therefore, comprehensive knowledge of volcanic SO₂ emissions is essential for a thorough evaluation of near-source volcanic hazards and large-scale atmospheric impacts.

Hyper-spectral nadir-viewing UV and infrared satellite instruments record global SO2 mass-loadings on a daily or bi-daily basis. Geostationary sensors, on the other hand, deliver high temporal information on SO₂ emissions but with much lower sensitivity. Consequently, there are still gaps in our knowledge of volcanic SO₂ emissions and SO₂ to sulphate oxidation rates, notably inside tropospheric plumes, and hence volcanic sulphur-rich compound feedback on the atmosphere.

TROPOMI, a hyperspectral UV sensor with increased spatial and spectral resolution than that of the pre-existing UV (OMPS) sensor in the same orbit, was launched in 2017. We discuss how an inverse modelling approach that assimilates TROPOMI SO₂ column amounts (CA) improves the retrieval of hourly SO₂ emissions when compared to the assimilation of OMPS data acquired at approximately the same time and the new SO2 products (both SO₂ CA and layer heights) from IASI. The purpose of using IASI data is to assess the impact of assimilating SO2 data available bi-daily into inverse modelling with additional information on SO₂ layer height. The inverse modelling is performed utilizing a time series of daily or bi-daily SO₂ CA snapshots from the TROPOMI, OMPS, and IASI satellite instruments, respectively. Contrary to OMPS, which has 50x50 km² of spatial resolution, and IASI, which has a 12 km circular footprint, TROPOMI has an extraordinary spatial resolution of 5.5x3.5 km² (7x3.5 km² before August 2019). We find that because of their sensitivity to low-level SO₂ fluxes and thin SO₂ plumes, and the numerous SO₂-rich pixels defining dense parcels, TROPOMI observations enable better evaluation of SO₂ degassing during paroxysmal eruption phases, offering better-resolved SO₂ emissions by inverse modelling. However, if meteorological clouds hide the volcanic SO₂ plumes, the results can be inconsistent, especially if the clouds are near the source. So the additional data, the SO₂ height product from IASI observations, is used to reconcile and offer more robust SO₂ emissions. As a second step, we perform inverse modelling using both the SO₂ CA and layer heights from IASI. This research investigates the Mount Etna eruption in February 2021, the SO2 plume reaching France, and the 2018 Ambrym eruption, which was the top world-ranking SO₂ emitter. In the context of Etna eruption, we use ground-based OHP LIDAR aerosol height measurements to explore the presence of sulphate aerosols and their height in the SO₂ plume.