Study of the seismo-acoustic energy radiation by debris flows

This doctoral thesis is focused on the study of the seismo-acoustic energy radiation by debris flows. Occurring within steep mountain catchments as sudden floods carrying large amounts of boulders and solid debris, debris flows represent a major hazard in worldwide mountain environments. The debris-flow monitoring is traditionally performed following a wide range of approaches, involving rainfall forecast and measurements and catchment observations. In the last 20 years, the use of seismo-acoustic signals for the study and monitoring of debris flows gained attention worldwide. Indeed, a debris flow radiates elastic energy both in the atmosphere, in the form of infrasound, and in the ground in the form of seismic waves. Seismic waves are believed to be generated by solid particle collisions, turbulent structures and friction with riverbed and banks, while infrasound is believed to be generated by flow waves that develop at the free surface of the debris flow. However, the radiation processes of both wavefields are not yet fully understood and open questions remain in linking the generated geophysical signal to the debris-flow dynamics and parameters. To investigate the seismo-acoustic radiation processes within debris flows, this thesis presents the analysis of the infrasonic and seismic signals generated by the debris-flow activity in the Illgraben catchment (Switzerland, Canton Valais) between 2017 and 2019, when 18 events were observed. Each event was characterized in terms of hydraulic and physical parameters (front velocity, flow depth, flow density). The infrasonic and seismic signals were analysed both in the time domain, performing the root mean square amplitude (RMSA) analysis, and in the frequency domain, computing the signal spectra with the Fourier analysis. These analyses allowed to characterize the signals in terms of amplitude envelope, maximum infrasonic and seismic RMSA, frequency content and peak frequency. Despite an excellent match resulted between the recorded infrasonic and seismic maximum amplitudes and despite the RMSA analysis showed that the Illgraben debris-flow events are characterized by a distinctive succession of the seismo-acoustic source processes, which tend to be repeated during each event, the infrasonic and seismic signals show clearly different frequency contents, indicating that the two wavefields are generated by different decoupled processes simultaneously acting at the flow surface and at the riverbed respectively. The relation between seismo-acoustic signals was further investigated by applying the cross-correlation analysis. The strong cross-correlation resulted between infrasonic and seismic primary signal components indicates that the two wavefield are closely related and suggest that, despite being decoupled, they are equally modulated in amplitude. In addition, results showed that the seismo-acoustic cross-correlation can be used to roughly locate the debris-flow events along the Illgraben channel. Furthermore, to deeply investigate the infrasound source mechanism within debris flows, the array processing was applied to the recorded infrasonic data. The analysis revealed that the infrasound by debris flows is dominated by coherent signal components generated in fixed position along the channel, in particular in correspondence of check dams and other significant topographic irregularities of the Illgraben channel. Furthermore, the infrasonic array processing permits to identify the position of the infrasonic source, thus allowing to track the motion of the debris flow along the entire Illgraben channel. To investigate if and how the hydraulic parameters influence and control the seismo-acoustic energy radiation, the amplitude and frequency content of the infrasonic and seismic signals were then compared to the measured front velocity, depth and peak discharge. Results show a positive correlation with both infrasonic and seismic maximum RMSA, suggesting that seismo-acoustic amplitudes are controlled by these flow parameters. The comparison between seismo-acoustic peak frequencies and flow parameters instead revealed that, unlike seismic signals, characterized by a constant peak frequency regardless of the magnitude of the flow, infrasound peak frequency decreases with increasing flow velocity, depth and discharge. Based on presented results and on previous models and experiments, a simplified conceptual source mechanism is proposed for the seismo-acoustic energy radiation by debris flows, according to which the infrasonic and seismic waves are generated by different source processes acting simultaneously at the ground and at the surface of the debris flow respectively, which, despite being intimately related to each other and equally modulated in amplitude, are clearly decoupled. In addition, results highlighted a strong influence of the flow parameters on the generated seismo-acoustic signals. For seismic signals, presented results agree with previous models and observations of seismic energy radiation by rivers and debris flows, indicating that seismic waves are generated by solid particle collisions and friction with the riverbed and banks and by fluid dynamic structures. For infrasound, non-stationary turbulence-induced waves and oscillations that develop at the free surface of the flow are thought to be the most likely source mechanism. The formation of such surface waves is enhanced wherever the flow encounters channel irregularities, such as significant topographic steps, like check dams, and steep bends, which result as locations of preferential infrasound radiation, consistently with the results obtained from the infrasonic array processing. Moreover, large flow depth and/or velocity is expected to generate higher and larger waves at the free surface of the flow. The development and the motion of these flow surface waves pushes the atmosphere and thus radiates infrasound. This motion of the flow surface can be modelled as a series of vertical pistons generating infrasound at the frequency of the piston motion, which is controlled by flow parameters and channel geometry. This model is also in agreements with the empirical relationships resulted between the infrasonic features and flow parameters. Finally, presented results highlight how the infrasonic and seismic recordings could be used for monitoring and warning purposes, not only for event detection, but also for the real time estimation of the event parameters.