University of Texas at Dallas Earthquake Dataset

A global earthquake dataset constructed from high-quality source and receiver metadata, including associated seismic phase picks across diverse station geometries.

Documentation

Full documentation for UTDQuake is available here:

You will see:

• QuickStart guide to get you up and running
• Detailed API reference
• Tutorials and example workflows

Dataset

The dataset is available on Hugging Face: UTDQuake

Installation (utdquake)

pip install utdquake

Why this dataset matters?

Curated datasets of earthquake events and phase picks are essential for modern seismology, especially in the AI era. While waveform datasets have advanced earthquake detection, multistation picks provide complementary information crucial for phase association and earthquake location.

This dataset offers structured event catalogs, station metadata, and phase picks across networks, supporting reproducible research and the development of data-driven seismological methods.

What’s inside?

For details on the contents and schema of each dataset, please refer to the Hugging Face dataset viewer.

To get started, see the Quick Start section below, or click “Use this dataset” on the Hugging Face dataset page for example loading code.

Quick start

Basic Access

import utdquake as utdq

# dataset overview 
dataset = utdq.Dataset()
print(dataset)

# network level
network_data = dataset.networks
print(network_data)

dataset.plot_overview(savepath="utdquake.png")

Network Data

# load network 
network = dataset.get_network(name="tx")
print(network)

# events
events = network.events
print(events)

# stations
stations = network.stations
print(stations)

# picks
picks = network.picks
print(picks)

Event Bank

Check ObsPlus EventBank for more details.

# get event bank
ebank = network.bank # 

# Example: Filter by event_id
ev_ids = events["event_id"].iloc[:5].tolist()
cat = ebank.get_events(event_id=ev_ids)
print(cat)

# Example 2: Other filter (check obsplus.EventBank for more details)
cat2 = ebank.get_events(minmagnitude=4.3)
print(cat2)

Plot

# get Obspy Event
network = dataset.get_network(name="tx")
network.plot_overview(savepath="overview.png")
network.plot_uncertainty_boxplots(savepath="uncertainty_boxplots.png")
network.plot_station_location_uncertainty(savepath="station_location_uncertainty.png")
network.plot_stats(savepath="stats.png")
network.plot_pick_histograms(savepath="histograms.png")
network.plot_pick_stats(savepath="pick_stats.png")

Thanks

Thanks to the UT Dallas HPC team for providing the computational resources for this dataset.

We also thank the seismology and AI communities for their work in earthquake research, and Hugging Face for hosting and sharing open datasets.

We welcome feedback and contributions!

Authors

Emmanuel Castillo University of Texas at Dallas
Nadine Ushakov University of Texas at Dallas
Marine Denolle University of Washington