ALBERT

Description: Volcanic explosions produce energy that propagates both in the subsurface as seismic waves and in the atmosphere as acoustic waves. We analyse thousands of explosions which occurred at different craters at Etna volcano (Italy) in 2018 and 2019. We recorded signals from infrasound sensors, geophones (GPH), broadband seismometers (BB) and Distributed Acoustic Sensing (DAS) with fibre optic cable. The instruments were deployed at Piano delle Concazze at about 2 to 2.5 km from the active craters, within (or onto) a ~300,000 m2 scoria layer deposited by recent volcanic eruptions. The DAS interrogator was setup inside the Pizzi Deneri Volcanic Observatory (~2800 m elevation). Infrasonic explosion records span over a large range of pressure amplitudes with the largest one reaching 130 Pa (peak to peak), with an energy of ca. 2.5x1011 J. In the DAS and the BB records, we find a 4-s long seismic “low frequency” signal (1-2 Hz) corresponding to the seismic waves, followed by a 2-s long “high-frequency” signal (16-21 Hz), induced by the infrasound pressure pulse. The infrasound sensors contain a 1-2 Hz infrasound pulse, but surprisingly no high frequency signal. At locations where the scoria layer is very thin or even non-existent, this high frequency signal is absent from both DAS strain-rate records and BB/GPH velocity seismograms. These observations suggest that the scoria layer is excited by the infrasound pressure pulse, leading to the resonance of lose material above more competent substratum. We relate the high frequency resonance to the layer thickness. Multichannel Analysis of Surface Wave from jumps performed along the fibre optic cable provide the structure of the subsurface, and confirm thicknesses derived from the explosion analysis. As not all captured explosions led to the observation of these high frequency resonance, we systematically analyze the amplitudes of the incident pressure wave versus the recorded strain and find a non-linear relationship between the two. This non-linear behaviour is likely to be found at other explosive volcanoes. Furthermore, our observations suggest it might also be triggered by other atmospheric pressure sources, like thunderstorms. This analysis can lead to a better understanding of acoustic-to-seismic ground coupling and near-surface rock response from natural, but also anthropogenic sources, such as fireworks and gas explosions.

Description: Distributed Dynamic Strain Sensing (DDSS), a.k.a. Distributed Acoustic Sensing (DAS), is becoming a popular tool for volcano monitoring. The sensing method relies on measuring the phase-shift of Rayleigh back-scattered light throughout the fibre due to strain variations in the fibre glass. This provides distributed strain-rate measurements at fine temporal and spatial sampling intervals. During 3 months in 2019, we recorded signals from thousands of mild volcanic explosions from Mt. Etna using a multi-instrument network deployed in an area at ca. 2.5 km distance from the active craters. Infrasound sensors were laying at the surface with a dense array of broadband seismometers (BB). Two types of fibres were also buried ca. 30 cm depth in the non-consolidated scoria from the area. First fibre was a 1.5 km long standard fibre, interrogated with an iDAS unit. The second fibre was a 0.5 km long engineered fibre, interrogated with a Carina unit. Relation between infrasound and DDSS data suggests a ground response of the loose scoria due to the acoustic pressure waves from explosions. Further analysis suggests a non-linear relationship between acoustic pressure and strain-rate data. However, signal saturation is encounter in some of the strain-rate data, which affects the interpretation of the non-linear relation. Therefore, we present an algorithm to correct the signal artefacts, allowing us to restore the true strain-rate signal and exceed the dynamic range limited by the initial DDSS recording parameters. The outcome includes strategies in the selection of acquisition parameters prior to DDSS campaigns to avoid signal saturation.