GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 1.2 GNGTS 2024 Fig. 1 – Maps showing the positon of Panarea Island in Southeastern Tyrrhenian Sea (red marker in lef panel) and the positon of the investgated vent (cyan marker in right panel). To carry out the estmaton of the gas volume emited by the hydrothermal vent we implemented a spectral method (Leighton and White 2011, Roche et al. 2022), based on a specially developed inverse modelling algorithm. The adopted approach, founded on the assumpton that the acoustc signature of a single bubble event evolves over tme as a sinusoid that exponentally decays (Leighton et al. 2011, Walton et al. 2005), is based on the formulaton of a forward model for the sound radiated by the bubbles plume, then the path is backward analysed to obtain an estmaton of the fow emission rate. High-resoluton audio frames were recorded by using an autonomous smart hydrophone, able to collect and store digitsed audio frames in the frequency band [1 - 12.800] Hz. The hydrophone was deployed in mooring confguraton in the proximity of the investgated vent, acquiring data for ~10 hours at the maximum sampling frequency. Hence the collected audio frames were treated with preliminary analyses, aimed to characterise the spectral features of the vent, thus identfying the acoustc signature of the source and the frequency range connected to signals atributable to the bubbling events. The analyses of the Power Spectral Density (PSD) and Pressure Power Spectrum show the presence of diferent persistent energetc frequency peaks over the environmental background noise which are compatble with the dynamic of the hydrothermal feld (see Fig. 2). Among these, the most energetc ones are likely due to the acoustc signal radiated by a huge, resonant bubble plume, consistently confrmed by the coupling of the estmated radius with direct observatons.