Using localized microseismic noise sources to perform high-resolution seismic Imaging of seafloor using Distributed Acoustic Sensing

Abstract

In this study we use distributed acoustic sensing (DAS) on a 41-km-long submarine optical fibre (OF) cable located offshore Toulon, France. We record both the amplitude and frequency of seafloor strains induced by ocean surface gravity waves, as well as secondary microseisms. Combining the analysis of the two types of waves, we identify and localize local sources of secondary microseisms that manifest as Scholte waves generated by the reflection of oceanic gravity waves on the coastline. During the experiment, these local sources represent the most energetic contribution to the seismic noise recorded along the OF and by an onshore broad-band station located near the DAS interrogator. As a result, the characteristics of this noise are closely related to local wave conditions. One major challenge in performing seismic imaging using ambient seismic noise correlations using DAS data is that we cannot solve for the true seismic velocity because the noise wave field is dominated by local sources. To address this, we measure the incident angle of the dominant local noise sources, correct the apparent velocity using the incident angle retrieve from beamforming analysis and generate a 2D model. We then quantify the errors that arise from picking the dispersion curves of the most energetic velocities without correcting from the incident angle. Our results show that there are significant differences in velocities, with differences reaching up to several hundred meters per second. This highlights the importance of correcting these velocities before generating a tomography. Finally we evaluate an alternative strategy for a linear DAS fiber that cannot be use to localized the dominant noise source. We measure the dispersion curve of the slowest Scholte waves recorded and compare it to the corrected dispersion curves of the dominant source. Although this strategy suffers from limitation, it minimizes the error in the velocity model.