Automation of Distributed Acoustic Sensing Passive Seismic Processing

Summary As the need for continuous passive seismic monitoring grows, we continue to seek sensors that can provide a large aperture and are highly repeatable. Distributed acoustic sensing provides a unique opportunity to deploy optical fibres as a complementary distributed sensing instrument for seismic monitoring that is repeatable once permanently deployed (i.e., trenched or cemented) and with minimal maintenance effort when installed correctly. Additional cost savings can be achieved when repurposing the existing network of telecommunication optical fibres. However, the data rates, both temporally and spatially, can be challenging for near-real-time seismic monitoring solutions. We propose using a machine learning-enabled seismic phase picker (i.e., PhaseNet) and an automated pick corrections procedure in preparation for seismic event location. We demonstrate our workflow on synthetic data from a modified SEAM II Barrett unconventional model.