DAS-N2N: machine learning distributed acoustic sensing (DAS) signal denoising without clean data

• Published in: Geophysical journal international Vol. 236; no. 2; pp. 1026 - 1041
• Main Authors: Lapins, S, Butcher, A, Kendall, J-M, Hudson, T S, Stork, A L, Werner, M J, Gunning, J, Brisbourne, A M
• Format: Journal Article
• Language: English
• Published: Oxford University Press 01.02.2024
• Subjects: Instrumental noise; Earthquake monitoring and test-ban treaty verification; Antarctica; Distributed acoustic sensing; Machine learning; Denoising
• ISSN: 0956-540X; 1365-246X
• DOI: 10.1093/gji/ggad460

SUMMARY This paper presents a weakly supervised machine learning method, which we call DAS-N2N, for suppressing strong random noise in distributed acoustic sensing (DAS) recordings. DAS-N2N requires no manually produced labels (i.e. pre-determined examples of clean event signals or sections of noise) for training and aims to map random noise processes to a chosen summary statistic, such as the distribution mean, median or mode, whilst retaining the true underlying signal. This is achieved by splicing (joining together) two fibres hosted within a single optical cable, recording two noisy copies of the same underlying signal corrupted by different independent realizations of random observational noise. A deep learning model can then be trained using only these two noisy copies of the data to produce a near fully denoised copy. Once the model is trained, only noisy data from a single fibre is required. Using a data set from a DAS array deployed on the surface of the Rutford Ice Stream in Antarctica, we demonstrate that DAS-N2N greatly suppresses incoherent noise and enhances the signal-to-noise ratios (SNR) of natural microseismic icequake events. We further show that this approach is inherently more efficient and effective than standard stop/pass band and white noise (e.g. Wiener) filtering routines, as well as a comparable self-supervised learning method based on masking individual DAS channels. Our preferred model for this task is lightweight, processing 30 s of data recorded at a sampling frequency of 1000 Hz over 985 channels (approximately 1 km of fibre) in <1 s. Due to the high noise levels in DAS recordings, efficient data-driven denoising methods, such as DAS-N2N, will prove essential to time-critical DAS earthquake detection, particularly in the case of microseismic monitoring.