Distributed Acoustic Sensing of debris-flow activity in the Öschibach torrent (Swiss Alps)

Debris-flow and debris-flood activity in the Öschibach torrent (Switzerland), driven by sediment supply from the unstable rock slope Spitze Stei, poses a significant hazard to the village of Kandersteg. While long-term monitoring exists, the dynamic linkage between sediment supply and torrential activity remains poorly constrained due to the spatial restrictions of conventional sensors.

In summer 2025, we addressed this limitation by deploying a dense seismic array on the rock slope and interrogating an existing ~4 km-long dark fiber optic cable along the stream using Distributed Acoustic Sensing (DAS). The DAS system provided strain-rate measurements at meter-scale resolution (inter-channel spacing of ~ 5 m with an effective gauge length of ~10 m) with a sampling frequency of ~600 Hz, along  ~850 geolocated channels. Torrential events were identified using water-level thresholds combined with moving-average and minimum duration filtering, to generate a catalog of candidate events. DAS data reveal increased high-frequency energy (20–30 Hz) in channels near the torrent during these events and coherent signals allow estimation of apparent of the propagating seismic sources. In addition, the fiber recorded other coherent signals, including rockfalls and local to teleseismic earthquakes.

To move beyond detection, we applied Matched Field Processing (MFP) to estimate locations of developing debris flows using the frequency-dependent phase information of the DAS data. We performed synthetic tests to evaluate the ability of DAS to distinguish between sources in two adjacent stream branches from which debris flows may originate. These tests demonstrate that the cable geometry can resolve sources even in closely spaced initiation areas. We further apply this approach to recorded DAS data to characterize torrential event. By looping the MFP framework through time, we aim to track the velocity and evolution of individual debris-flow surges.

Preliminary results show that our multisensor approach – combining rock-slope seismic arrays with fiber-optic DAS allows for the association between rockfall activity from Spitze Stei and debris-flow dynamics within the torrent. This work highlights the potential of DAS and array-based processing for spatially dense monitoring, warning and source location in steep Alpine catchments.