Storm Amy observations with fibre-optic DAS data at the Svelvik CO₂ Field Lab, Norway: Implications for Monitoring and Networks

Distributed Acoustic Sensing (DAS) has become a powerful technique for high-resolution, continuous monitoring of near- and subsurface earth phenomena, with increasing applications in geohazards, seismology, and industry applications such as CO₂ storage monitoring. However, the sensitivity of DAS measurements to atmospheric forcing, particularly during extreme weather events, remains poorly understood. In this study, we investigate the response of a permanent, 1.2 km long straight fibre-optic array installed at the Svelvik CO₂ Field Laboratory (Norway), to intense wind conditions associated with the Amy Storm, which hit Norway from October 3-6, 2025. 

 

As part of efforts to understand passive methods to monitor CO2 migration in the subsurface, an Alcatel Submarine Networks (ASN) DAS system continuously recorded strain-rate data along a buried fibre that includes both near surface-installed sections and borehole down- and up-going segments reaching depths of approximately 100 m. The near-surface sections were installed inside protective pipes and were therefore not directly coupled to the surrounding ground. To characterise wind-induced seismic signatures, we analyse downsampled recordings using band-limited root-mean-square (RMS) amplitudes and spectral methods across three frequency ranges (0.1–1 Hz, 1–3 Hz, and 3–10 Hz) and time averages over 1 hr intervals. Time–frequency characteristics are examined using group-averaged spectrograms, and a Spectral Energy Index (SEI) is derived by integrating power spectral density within each frequency band. These seismic metrics are compared with near located meteorological observations, including mean wind speed, maximum mean wind speed, and maximum wind gusts. 

 

The results reveal a pronounced increase in DAS energy coincident with the maximum speed gusts of storm Amy, with the strongest responses observed at frequencies below 3 Hz. Correlation and lag analyses show that seismic energy variations are closely associated with periods of enhanced wind activity, particularly wind gusts, indicating a strong coupling between transient atmospheric forcing and ground vibrations. Importantly, the response differs significantly between surface and depth segments of the fibre. Surface-installed channels exhibit broadband amplitude increases correlated with direct wind–ground interaction, while depth channels display coherent low-frequency spectral patterns, suggesting excitation by wind-generated surface waves or distant secondary sources (e.g., waves from neighbouring fjord) rather than direct aerodynamic loading. 

 

These findings demonstrate that DAS arrays deployed at wells (abandoned or active) are sensitive to extreme meteorological forcing, which can imprint distinct and depth-dependent seismic signatures. Quantifying and distinguishing wind-induced signals is therefore critical for the robust interpretation of DAS data in long-term passive monitoring applications, particularly when subtle subsurface signals related to CO₂ injection, migration, or leakage must be detected in the presence of strong environmental noise. At the same time, this sensitivity highlights an additional benefit of such fibre-optic installations: DAS infrastructure deployed in future abandoned wells in the context of  Oil & Gas industry and their reutilization for CO2 capture and storage, can also provide valuable information for national seismic and environmental monitoring networks, extending their utility beyond site-specific applications.