Central-Western Italy Seismic Waveforms for Rapid Ground Motion Prediction

Dario Jozinović of Roma Tre University published this dataset of seismic waveforms from central-western Italy. It was used in a 2021 paper exploring transfer learning with convolutional neural networks for rapid prediction of earthquake ground shaking. The dataset includes 915 events from the 2016 central Italy sequence and 266 events from an area near Pisa.

Use Cases

• Training CNN models for rapid peak ground motion prediction based on raw multistation waveforms.
• Applying transfer learning techniques to improve model performance in regions with insufficient training data.
• Investigating the impact of incorporating station position data as an additional neural network layer on prediction accuracy.
• Benchmarking early warning times for specific stations, such as station PII, using the described methodology.

Strengths

• Dataset is directly linked to a peer-reviewed research paper published on arXiv in 2021.
• Includes a specific, larger dataset of 915 events from a dense station network in central Italy.
• Associated code for the paper's methodology is publicly available on GitHub.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the primary dataset is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Roma Tre University

Collection Method

Seismic recordings from stations in central-western Italy, used for academic research.

Time Range

Includes data from the 2016 central Italy earthquake sequence.

Geography

Central-western Italy, specifically including an area near Pisa.