SO2 flux measurements from ground, air and space before, during and after a lava fountain on Mt Etna

Quantifying volcanic sulfur dioxide (SO2) emissions is essential for understanding magmatic processes and improving eruption forecasting. We present a four-day, multi-platform investigation of SO₂ emissions at Mt Etna, Italy, spanning 15–18th July 2024 and capturing the build-up, climax, and recovery from a major paroxysmal lava-fountaining eruption at the Voragine crater on 15th July. Spectra were acquired using ultraviolet spectrometers mounted on a car and a custom vertical take-off and landing uncrewed aerial system (UAS). This was complemented with spectra from the permanent FLAME scanning network, and by satellite-derived emission rates from TROPOMI imagery analysed using the PlumeTraj analysis toolkit. 

All ground-based and airborne spectra were analysed using the iFit intensity-fitting algorithm, enabling consistent SO₂ slant column density retrievals and correction of light dilution effects using the dual-waveband approach. Wind speeds used in flux calculations were derived from Pitot tube measurements from the UAS when available. Across the integrated dataset, SO2 emission rates increased from steady background levels of ~6 kg s⁻¹ on the morning of 15th July to >40 kg s⁻¹ several hours before the onset of lava fountaining, at which point daylight-dependent ground and airborne measurements ended. Analysis of the TROPOMI imagery provides an average SO₂ emission rate of 270 ± 80 kg s⁻¹ during the 6-hour fountaining phase, corresponding to a total emitted mass of 5.7 ± 1.7 kt of SO2. 

Volcanic tremor amplitude rose concurrently with the pre-eruptive increase in SO₂ flux, showing strong correlation prior to fountaining (Spearman rank correlation, ρ = 0.85), but alters during the eruption, likely reflecting a shift in the dominant tremor source. Following the eruption, all platforms recorded greatly reduced quiescent emissions on 16–17th July (<2 kg s⁻¹), before partial recovery by 18th July (<5 kg/s). 

Each platform contributed complementary strengths: UAS measurements provided high signal-to-noise ratios and light-dilution quantification; car traverses most easily captured complete plume cross-sections; scanners resolved short-term degassing variability; and satellite observations quantified eruptive emissions inaccessible to the other methods. Together, these results demonstrate that coordinated, multi-platform SO₂ monitoring is essential for resolving rapid degassing dynamics across an eruptive cycle and for enhancing eruption-forecasting at persistently active volcanoes.