Snow Avalanche Characteristics Along Milford Road, Aotearoa New Zealand, Using Distributed Acoustic Sensing

Abstract We repurposed an unlit (“dark”) telecommunication fiber cable running along the Milford Road in Fiordland, Aotearoa New Zealand, into a distributed acoustic sensing (DAS) array to monitor snow avalanches during the 2024 austral winter and spring season. Using DAS data, we constrained the timings of 16 avalanches and analyzed two triggered and two natural events in detail. We present the first seismic evidence for the plunging dynamics of snow avalanches along Milford Road. We also document impulsive signals associated with the avalanche-stopping phase and a precursory signal in wet slab avalanches that we attribute to the collapse of water-saturated structures within the snowpack caused by liquid water percolation before the main avalanche. In addition, we also examine the interplay between avalanche occurrence and local meteorological conditions, showing that naturally occurring avalanches are strongly clustered in time during periods of high precipitation rates. However, not all high precipitation periods release avalanches, which suggests that other parameters, such as snowpack stability, play an important role. Our results demonstrate the potential of dark fiber DAS networks as a low-cost, continuous seismic monitoring system for snow avalanche characterization in remote alpine environments. However, avalanche detection in this study relied on independently constrained event timings, and traffic noise remains a significant source of false detections.