SeismicX-Cont Mini Subset

This mini subset is a lightweight, runnable subset of the SeismicX-Cont continuous seismic monitoring benchmark. It is designed for quick testing, tutorials, software validation, and demonstration of the SeismicX-Cont workflow without requiring users to download the full 387 GiB archive.

The complete SeismicX-Cont dataset is available at:
https://www.modelscope.cn/datasets/cangyeone/SeismicX-Cont

Contents

The mini subset contains two one-hour continuous waveform windows:

- 2019-07-06 04:00–05:00 UTC: a dense Ridgecrest aftershock-period example.
- 2021-11-14 16:00–17:00 UTC: a comparatively quiet regional-background example.

The package includes:

• hourly HDF5 waveform files;
• mini annotation JSON;
• SQLite waveform index;
• station metadata;
• instrument-response JSON for optional response removal or response simulation;
• example picker outputs;
• REAL association/event-detection example outputs;
• scripts for waveform querying, dataloader use, picker inference, picker evaluation, consensus candidates, REAL association, and event-level catalog comparison;
• a quickstart notebook and checksum manifest.

Purpose

This subset is intended to let users reproduce the main SeismicX-Cont access and validation workflow on a small local example. It supports:

• HDF5 waveform loading;
• SQLite station-time waveform queries;
• PyTorch dataloader tests;
• optional instrument-response correction and response simulation;
• continuous phase-picking examples;
• coverage-aware picker evaluation;
• optional consensus-candidate construction;
• association-based earthquake detection;
• event-level comparison against the catalog.

The mini subset is not intended for large-scale model training or for replacing the complete SeismicX-Cont benchmark.

Relationship to the Full Dataset

The files in this release are derived from the full SeismicX-Cont benchmark and preserve the same data organization and script interfaces where possible. Users should cite the complete SeismicX-Cont release and the original waveform/data sources when using this mini subset in publications or tutorials.

License

The SeismicX-Cont derived metadata and scripts are released under the licenses described in the repository. Waveform data remain subject to the data usage policies of the original data centers. Users should consult and cite the original SCEDC, NCEDC, and CEED sources where applicable.