Unsupervised Coherent Noise Removal From Seismological Distributed Acoustic Sensing Data

• Published in: Journal of geophysical research. Machine learning and computation Vol. 1; no. 4
• Main Authors: Konietzny, Sebastian, Lai, Voon Hui, Miller, Meghan S., Townend, John, Harmeling, Stefan
• Format: Journal Article
• Language: English
• Published: Wiley 01.12.2024
• Subjects: coherent noise; distributed acoustic sensing; machine learning; seismology; traffic
• ISSN: 2993-5210; 2993-5210
• DOI: 10.1029/2024JH000356

Recent advances in sensing technologies, particularly Distributed Acoustic Sensing (DAS), have significantly improved the collection and analysis of seismological data. DAS is a powerful method for detecting vibrations from various sources, including earthquakes. However, the vast amount of data produced by DAS requires sophisticated analytical methods to differentiate between signals of interest and noise, such as traffic signals. We introduce an innovative approach by extending the Noise2Self framework to effectively remove unwanted, structured coherent noise from DAS data. By creating masks based on the characteristics of traffic signals, we isolate and preserve earthquake signals while maintaining the denoising performance of the original Noise2Self approach, which reduces noise without requiring clean reference data. To evaluate our method, we used synthetic data generated from seismic recordings of closely spaced seismometers and then applied our approach to data from a DAS array crossing the Alpine Fault near Haast, New Zealand. Our results demonstrate that our model successfully removes traffic noise and other non‐coherent noise while preserving seismic signals, leading to improvements in both Signal‐to‐Noise Ratio (SNR) and waveform coherence. Evaluations on real‐world DAS data further confirm the robustness of our method, positioning it as a valuable tool for analyzing large‐scale DAS data sets in various geoscientific contexts. This approach, which we refer to as “CoherentNoise2Self” to emphasize the extension of Noise2Self to coherent noise, advances the capabilities for near real‐time monitoring and analysis of seismic DAS data. Plain Language Summary Recent advancements in optical sensing, data storage, and computational methods for analyzing terabyte‐scale geophysical data have transformed the recording and interpretation of ground vibrations. Distributed acoustic sensing (DAS) uses laser pulses in unused optical fibers in commercial telecommunications networks to detect vibrations from earthquakes, landslides, and human activities. These vibrations affect the cable's scattering, allowing for high‐frequency recording over long distances. However, distinguishing signals of interest from coherent noise, like traffic, is challenging. In this study, we address removing coherent traffic noise from DAS earthquake recordings using an extended machine learning method called Noise2Self. Applied to DAS data from a highway crossing the Alpine Fault in New Zealand, our algorithm successfully removes traffic noise while preserving earthquake signals. Our approach can be generalized to other coherent noise sources in different contexts. Key Points We address the removal of coherent noise signals present in real‐world seismological Distributed Acoustic Sensing (DAS) data The self‐supervised Noise2Self framework is expanded to efficiently filter out coherent traffic signals from DAS earthquake records Both qualitative and quantitative testing with synthetic and real DAS data from the Alpine Fault show the effectiveness of our approach