Submarine Distributed Acoustic Sensing: Covariance-Based Detection of Volcanic Tremor at La Palma Eruption, 2021 (Canary Island)

This study explores the use of Distributed Acoustic Sensing (DAS) data recorded along a submarine fiber-optic cable at La Palma (Canary Islands) during the 2021 Tajogaite eruption, with the aim of improving the detection of volcanic tremor sources. The cable, originally installed for telecommunications, was instrumented with a High-Fidelity DAS (HDAS) system by CanaLink on October 19, 2021. The array is located about 10 km from the Cumbre Vieja crater, providing continuous measurements with a spatial sampling of 10 m along 50 km. In this work, we focus on the first 30 km of the cable, corresponding to around 3,000 potential channels. From these, 12 representative channels were selected based on signal quality and spectral characteristics.The methodology follows the covariance-matrix analysis originally proposed by Seydoux (2016) and Soubestre (2018). This approach enables the detection and characterization of coherent seismic signals by analyzing the eigenstructure of frequency-dependent covariance matrices. In particular, the dominant eigenvalue is used as an indicator of coherent sources such as volcanic tremor, providing information about their presence and spectral characteristics. The analysis of three months of continuous data shows a persistent and coherent spectral band centered around 1 Hz, observed consistently in both DAS data and conventional seismic records. Compared to traditional seismic stations, the DAS measurements tend to exhibit a narrower and more stable spectral response, which helps to better constrain the dominant tremor frequency. Overall, these results suggest that submarine DAS systems may offer a useful complement in areas where conventional instrumentation is limited.