Acoustic monitoring of proglacial discharge at Qaanaaq Glacier, Northwest Greenland

Glaciers around the world have experienced substantial mass loss due to global warming (Hugonnet et al., 2021). In Greenland, meltwater runoff is one of the major contributors to mass loss from the Greenland Ice Sheet and surrounding glaciers (Mouginot et al., 2019). This meltwater increases the discharge of proglacial rivers and poses a growing flood hazard to local communities (Kondo et al., 2021). Therefore, there is an urgent need to develop passive, robust, and low-maintenance methods for monitoring proglacial discharge under rapidly changing channel conditions.

Recent studies have shown a strong correlation between proglacial river discharge and fluvial sound (Podolskiy et al., 2023). Fluvial sound is mainly generated by air-bubble entrainment and collapse within turbulent flow features such as rapids and waterfalls, and its amplitude and spectral characteristics systematically respond to changes in discharge (Bolghasi et al., 2017). Passive acoustic monitoring therefore enables non-invasive and cost-effective discharge observation by simply recording the self-generated sound of a river, yet its applicability and limitations remain insufficiently understood.

In this study, we investigate the potential of passive acoustic monitoring to track proglacial discharge at Qaanaaq Glacier in northwestern Greenland (77°28’ N, 69°14’ W). During the summer of 2024, we deployed four passive acoustic sensors along the proglacial river and continuously recorded fluvial sound. Acoustic power in the 94–375 Hz frequency band showed a strong correlation with river discharge (R ≈ 0.90). Cross-correlation analysis between two sensors separated by 1,850 m revealed highly correlated acoustic signals (R = 0.90) with repeatable time lags of up to approximately one hour, although data gaps occurred during very low- and high-discharge conditions when the acoustic time lag became poorly resolved. This limitation suggests that larger sensor separations or array-based deployments may be required to robustly resolve time lags under variable flow conditions.

In addition to fluvial sound, the acoustic sensors recorded traffic-related noise from a bridge crossing the river. More than 200 traffic events were detected, providing supplementary information relevant to local flood risk and infrastructure usage. The usage of bridge reached maximum around 13 to 16 local time of Qaanaaq (LT), whereas discharge reached maximum around 18 to 23 LT. The peak in bridge usage occurred during the rising phase of discharge, highlighting the importance of early-stage flood awareness for local communities.

These results demonstrate that passive acoustic monitoring offers a low-cost, non-invasive tool that can complement conventional methods for monitoring proglacial river discharge, particularly in dynamically evolving glacial river systems. In addition, acoustic observations can provide complementary information on human activity near rivers, which is relevant for local flood-risk awareness and infrastructure management.