Two years of DAS acquisitions at GEUS

Distributed Acoustic Sensing is a transformative technology rapidly advancing in seismology. It enables recording high-resolution seismic data in remote areas and is especially beneficial for seafloor sensing, where standard seismic instruments are complex to operate in real time.

Over the past two years, the Geological Survey of Denmark and Greenland (GEUS) has conducted continuous and campaign-based Distributed Acoustic Sensing (DAS) acquisitions on terrestrial and submarine fiber-optic cables, exploring the potential of DAS as a complement to the national seismic monitoring infrastructure. Using two ASN C-band interrogators deployed in both permanent and mobile configurations, GEUS has collected tens of TB of DAS data, with applications ranging from earthquake monitoring and onshore active seismic experiments to the detection of anthropogenic and marine activities.

This exploration constitutes a steep learning curve for GEUS in field logistics (how to access dark fibers, how to trench our own fibers, ...), as well as in data acquisition, processing, and archiving. Here, we present lessons learned along the way from fieldwork operations and observations of regional and teleseismic earthquakes, quarry blasts, submarine explosions, controlled naval experiments, active seismic tests, and so on. 

Regional earthquakes recorded on submarine cables demonstrate DAS's sensitivity to S-waves even when P-wave signal-to-noise ratios are low. In contrast, teleseismic events (e.g., the M6.5 Jan Mayen earthquake) reveal complex wavefields that are strongly modulated by bathymetry and sedimentary structures. Comparative analyses across multiple cables highlight significant variations in signal quality and frequency content linked to cable type, installation conditions, and seafloor environment. Automatic phase picking using PhaseNet shows promising results, particularly when combining multiple pre-trained models, but also exposes key limitations related to strain-rate measurements, coupling variability, and absolute timing errors caused by the lack of GPS synchronization. Anthropogenic signals, including quarry blasts, anchor drops, vessel traffic, and rare submarine passages offshore, and active and passive seismic acquisition onshore, illustrate both the detection capabilities and the sensitivity to acquisition parameters.

Overall, this two-year dataset demonstrates that DAS can significantly enhance seismic and environmental monitoring, while also identifying critical technical challenges, data volume management, timing accuracy, and site-dependent coupling that must be addressed before DAS can be fully integrated into our operational seismological workflows.