Infrasound array criteria for automatic detection and front velocity estimation of snow avalanches: towards a real-time early-warning system

Abstract

Abstract. Avalanche risk management is strongly related to the ability to identify and timely report the occurrence of snow avalanches. Infrasound has been applied to avalanche research and monitoring for the last 20 years but it never turned into an operational tool for the ambiguity to identify clear signals related to avalanches. We present here a new method based on the analysis of infrasound signals recorded by a small aperture array in Ischgl (Austria), which overcome now this limit. The method is based on array derived wave parameters, such as back-azimuth and apparent velocity. The method defines threshold criteria for automatic avalanche identification considering avalanches as a moving source of infrasound. We validate efficiency of the automatic infrasound detection with continuous observations with Doppler Radar and we show how dynamics parameters such as the velocity of a snow avalanche in any given path around the array can be efficiently derived. Our results indicate that a proper infrasound array analysis allows a robust, real-time, remote detection of snow avalanches that could thus contribute significantly to avalanche forecast and risk management.

Snow Avalanche Detection and Source Constraints Made Using a Networked Array of Infrasound Sensors

“…As such, intensity calculated from TF and AF distributions represent a reasonable range of source intensities that could be more precisely calculated in future studies. Additional constraints, including observations from Doppler radars (e.g., Köhler et al., 2016; Marchetti et al., 2015) would be helpful for pinpointing the timing of flow surges and which portions of the flow are most energetic.…”

“…Without these filters, the avalanche infrasound is overwhelmed by nonnatural incident sound and avalanche signal analysis is not feasible. The additional bandpass filter (2–25 Hz) is designed to coincide with the expected band of infrasound associated with snow avalanches of similar size occurring in other regions (e.g., Havens et al., 2014; Johnson et al., 2018; Marchetti et al., 2015). The cumulative filter is represented in a Bode amplitude response plot (Figure 5).…”

“…The detection and study of rapid gravity‐driven mass movements using infrasound is a growing area of research and includes studies of hazardous flows like lahars (Johnson & Palma, 2015), debris flows (Kogelnig et al., 2014; Liu et al., 2015; Marchetti et al., 2019), rockfalls and ice‐falls (Johnson & Ronan, 2015; Preiswerk et al., 2016; Zimmer et al., 2012), pyroclastic density currents (Delle Donne et al., 2014), and iceberg calving (Richardson et al., 2010). Infrasound monitoring has also been proven as an effective tool for identifying the occurrences of snow avalanches and has been implemented as a monitoring strategy both remotely (at tens of kilometers) and proximally (within a few kilometers) using arrays of sensors (e.g., Adam et al., 1998; Arai et al., 2017; Bedard et al., 1988; Comey & Mendenhall, 2004; Marchetti et al., 2015; Mayer et al., 2020; Ulivieri et al., 2011). Infrasound remote sensing of avalanches may be complemented by seismic monitoring, which must be deployed relatively near the flow path trajectory owing to high seismic wave attenuation in the ground (e.g., Kogelnig et al., 2011; Marchetti et al., 2020).…”

Broadband Infrasound Signal of a Collapsing Hanging Glacier

“…Given the short computing time of infrasound array processing, with events possibly identified, and warning issued, within 2 min after their occurrence, and its efficiency to identify signals related to moving sources (Marchetti et al., 2015), this could be used as an additional system to radars and/or cameras to provide quantitative information, in near real‐time, on ice avalanches, that might reflect the general stability on an hanging glacier front.…”

Automatic detection of snow avalanches in continuous seismic data using hidden Markov models

“…In addition, further avalanche events may be used for training to improve model performance. Finally, incorporating localization parameters as new features in the HMM could open the door for further model improvement, as is done for the automatic detection of avalanches in continuous infrasound data (Marchetti et al, 2015;Thüring et al, 2015). These features can then either be implemented directly into the HMM or be used in additional post-processing steps.…”