← Back to Sessions

Chirped-pulse DAS for Ambient Noise Tomography

Description: 

Subsurface imaging is a powerful tool for terrain studies, revealing essential information about geological processes and material composition of the soil. Ambient noise interferometry (ANI) has emerged as a reliable technique to extract subsurface information without the need for active seismic sources. As high-resolution ambient noise tomography requires an extensive and dense detector array, distributed acoustic sensing (DAS) results in a cost-effective solution to seismic noise measurements. It provides high spatial resolution (~10m) and large distance measurement capabilities (10s of km) with a simple and non-invasive deployment. While DAS is a powerful tool for seismic measurement, it presents several inherent disadvantages such as non-linear measurements, non-uniform SNR and the presence of random fading points along the measured fiber. Chirped-Pulse DAS technology overcomes all these challenges, providing a linear and quantitative measurement.

For this study, a dark fiber on the trackside of a high-speed railway in Spain was monitored for a full 24-hour period. Raw data from the interrogator unit was then analyzed with a full ANI algorithm as follows. Firstly, data was filtered both in the frequency and spatial domain, restricting the range of interest to common frequencies and wavelengths for surface seismic waves. After that, FK filtering was applied, which allowed for a dispersion velocity filtering, enabling for low velocity signals such as trains to be discarded. Temporal normalization and spectral whitening were then applied to the dataset, so all noise sources contribute to the final data equally. Cross-correlations between signal from adjacent spatial points were then computed, obtaining a full map where signals propagating through the surface can be observed. Phase-weight stacking was employed to reduce the total number of needed files to obtain a clear correlation map. Once the 24-hour period was analyzed and stacked, slant-stacking was employed to compute a dispersion velocity map, showing at least two different dispersion relationships, revealing different layers of the surface structure.

Session: Fiber-optic Sensing Applications in Seismology [Poster]

Type: Poster

Date: 4/15/2025

Presentation Time: 08:00 AM (local time)

Presenting Author: Diego

Student Presenter: No

Invited Presentation: 

Poster Number: 55

Authors