TRACE

TRACE is an interdisciplinary scientific research project and AI lab, providing a shared computational environment for seismology and earthquake science.

Through a guided web interface and computational tools, it supports AI model development, real-time earthquake monitoring, early warning, research, and education.

Project Scope

TRACE brings together AI research, seismic data, physical modeling, and computational tools for earthquake science, real-time monitoring, and early warning.

AI Model Development

We develop, train, adapt, and evaluate AI models for analyzing seismic data and addressing research and operational challenges in earthquake science.

AI Training Lab

We are building an open research environment that combines a guided web interface with computational tools for training, testing, and comparing AI models.

Regional Data Adaptation

Users can incorporate regional datasets, velocity structures, physics-based synthetic data, and observed noise to create training datasets that better represent local conditions.

Research and Education

TRACE supports professional researchers and operational institutions while helping students and early-career researchers learn how AI models are applied to problems in seismology and earthquake science.

Research Topics

Seismology

We combine seismic observations, regional Earth models, and seismological knowledge to study earthquakes and support reliable monitoring, characterization, and cataloging.

Signal Processing

We analyze continuous seismic records using filtering, phase analysis, waveform methods, and noise characterization to identify earthquake signals in complex seismic observations.

Artificial Intelligence

We develop, train, and evaluate AI models for seismic analysis, real-time earthquake monitoring, early warning, and other problems in earthquake science.

Applied Mathematics

We develop mathematical and statistical methods for physical modeling, uncertainty quantification, inverse problems, and reliable AI-based seismic analysis.

TRACE Lab

We provide a guided web interface and computational tools to generate training datasets and to train, adapt, and compare modern AI models.

Data Analysis and Visualization

We develop tools for exploring seismic data, examining model behavior, comparing predictions, evaluating earthquake catalogs, and communicating results.