Coherence-based Amplification of Rayleigh Waves from Urban Anthropogenic Noise recorded with Distributed Acoustic Sensing

At a local scale, e.g. in urban settlements, seismic subsurface characterization requires implementing experiments at high spatial resolution. Distributed acoustic sensing (DAS) provides the opportunity of using pre-existing fiber optic cables as dense receiver arrays, thus potentially reducing the effort for active seismic surveying in urban areas. Due to their small footprint, passive experiments appear particularly appealing. However, extracting coherent signals in an urban environment, i.e. in the presence of anthropogenic activity in the receivers' vicinity, remains a challenge.

 

In this study, we present results from combining the well known technique of Multichannel Analysis of Surface Waves (MASW) with the coherency-based enhancement of wavefields. The investigation is based on a DAS dataset acquired along a major road in Berlin, Germany. We analyse a 4.5 km long straight subsegment of a dark fiber that was sampled at 8 m intervals with 1000 Hz over a period of 15 days. After temporal decimation and the interferometric analysis, clear causal and a-causal branches of Rayleigh-surface waves emerge in the virtual shot gathers.

 

In the further processing, we employ coherence-based enhancement of wavefields to amplify the Signal to Noise Ratio of the virtual shot gathers. Compared to the traditional workflow of ambient-noise tomography the modified one yields improved dispersion curves particularly in the low-frequency part of the signal. This leads to an increased investigation depth along with lower uncertainties in the inversion result. The final velocity model reaches depths down to 300 m. We show that the application of coherence-based enhancement of the virtual shot gathers in the MASW-workflow may significantly relax the necessity of collecting long baselines for passive tomography in urban environments.