DIGITAL.CSIC. Repositorio Institucional del CSIC
Datos de investigación
Set de datos (Dataset)
DASPack: Controlled Data Compression for Distributed Acoustic Sensing [Dataset]
The study introduces DASPack, an open-source, high-performance compression tool specifically designed to address the growing storage challenges of Distributed Acoustic Sensing (DAS) data, which already reach hundreds of terabytes in large-scale deployments. Its objective is to provide both lossless and controlled lossy compression, allowing users to balance storage efficiency and signal fidelity, while remaining scalable and practical for real-world applications. DASPack combines wavelet transforms, linear predictive coding, and entropy coding to achieve up to 3× file size reduction in lossless mode and between 6× and 10× in lossy mode, depending on the acceptable level of signal degradation. The tool delivers high throughput, with single-thread speeds of 100–200 MB/s and up to 750 MB/s using eight threads, making it suitable for real-time deployment under high data rates. Its performance was validated on 15 datasets from diverse acquisition environments, demonstrating robustness, speed, and broad applicability. Overall, DASPack establishes a practical foundation for sustainable, long-term management of DAS data in large-scale monitoring networks, European Union NextGenerationEU/PRTR Program under projects PSI (ref. PLEC2021-007875) and TREMORS (ref. CPP2021-008869), and the CSIC contract SAFE with Telxius Cable España, S. L. (ref. 20233069), This work contributes to the Institut de Ciències del Mar "Severo Ochoa Centre of Excellence" accreditation CEX2024-001494-S funded by AEI 10.13039/501100011033 of the Spanish Ministry of Science and Innovation, The dataset files follow the naming structure: <FDSN network>.<Experiment>.<YYYYMMDD-HHMMSS>.h5 where: <FDSN network> – The network code assigned to the DAS experiment. <Experiment> – A single-letter identifier for the experiment name/location: P – La Palma C – Castor S – SAFE <YYYYMMDD-HHMMSS> – The timestamp indicating the beginning time (UTC). .h5 – The data (nanostrain) is stored in HDF5 format. The Athens dataset is in ZIP format, which includes 20-h5 files containing 30-second-long optical phase changes, and can be converted to nanostrain-rate using the scaling factor 116.0 / 8192.0 * fs / 10.0 The timestamp indicates the beginning time (UTC). The naming structure is: <Experiment>_<Sampling rate>_<YYYYMMDD-HHMMSS.SSSSSS>.h5, Peer reviewed