Projects
Fiber-optic Seismology: Monitoring and data processing system for Distributed Acoustic Sensing (DAS) arrays
PI: Zhongwen Zhan (Division of Geological and Planetary Sciences)
SASE: Donnie Pinkston, Instructor
From earthquake detection and monitoring to subsurface imaging, distributed acoustic sensing (DAS) is proven to be a viable and powerful tool. The effectiveness of DAS is due to its ability to record seismic data using existing telecommunication fiber cables at an unprecedented spatial resolution for tens of kilometers. For this reason, DAS could also be an integral component of the next generation of earthquake early warning tools, especially, when deployed on submarine cable infrastructure. A long-term goal of the Zhan’s group research is to convert thousands of kilometers of cables in California into seismic networks to form the next-generation Southern California Seismic Network.
One of the major challenges associated with DAS is the significant amount of data generated even by single DAS experiments, which can easily reach tens to hundreds of TBs per year. With the envisioned state-wide DAS network, we are expecting 10s of TB of data per day. Additionally, given the novelty that DAS represents in seismology, conventional analysis tools are unsuitable for effectively and efficiently processing this vast amount of recorded DAS data. Some of the applications of DAS, such as earthquake early warning, need real-time data streaming and processing capabilities.
Two hours of the signal from the Ridgecrest DAS array for a magnitude 5.2 earthquake and a series of aftershocks near Bakersfield, CA.
Initial efforts in collaboration with the Schmidt Academy are focused on building a monitoring infrastructure for existing DAS arrays, with the ability to scale up as more systems come online. At the same time, techniques are being explored to reduce the volumes of data that must be brought into central systems by pushing analysis and computation out to edge nodes in the network, and using lossless or lossy compression techniques where appropriate. The ultimate goal is an integrated, accelerated, and user-friendly cloud-based platform to read, visualize, process, and analyze DAS continuous waveforms, event data, and data products. This platform would represent a significant advancement compared to any existing solution and serve as the software foundation toward the long-term goal of a large-scale DAS network, a key component of the next-generation seismic network.