Abstract

In the last decade, the use of Distributed Acoustic Sensing (DAS) technology for observational seismology has flourished. DAS sensors can provide extremely dense and often unique datasets that enhance seismic catalogs. Nevertheless, DAS-only observations may lead to biased source characterization due to strong spatio-temporal variations in signal quality and limited azimuthal coverage. Conversely, sparse conventional three-component (3C) seismic networks may fail to detect microseismicity but often guarantee sufficient azimuthal coverage. In this context, integrating conventional 3C sensors with high-resolution fiber-optic data enables the full exploitation of the strengths of individual deployments. We propose a system designed for observational seismology that integrates DAS and seismometer networks for source location. Our approach employs efficient DAS data denoising strategies, smart spatial subsampling along the fiber optic cable, and a balanced integration of the seismic observations from the individual networks. Waveform stacking is adopted to estimate source location using an efficient grid search approach, enabling the method to operate in near real time. In particular, we developed a DAS channel selection strategy that accounts for both signal quality and spatial distribution, to balance the integration of single-component DAS observations with the typically higher-quality 3C data from conventional seismic networks. The method is validated using both synthetic datasets and real seismic sequences. Results demonstrate that combining the data redundancy of fiber-optic sensing with the azimuthal coverage of conventional networks enables a step change in the resolution of seismic sequences.