Predicting Fault Slip via Transfer Learning

TL;DR

This paper presents a transfer learning approach that uses numerical simulations to train a model predicting fault-slip, which is then fine-tuned with limited laboratory data to improve earthquake fault predictions.

Contribution

The study introduces a novel transfer learning method combining numerical simulations and laboratory data to predict fault-slip behavior, addressing data scarcity in geophysical applications.

Findings

Transfer learning improves fault-slip prediction accuracy.

Simulation-trained models generalize well to laboratory data.

Fine-tuning with minimal data enhances model performance.

Abstract

Data-driven machine-learning for predicting instantaneous and future fault-slip in laboratory experiments has recently progressed markedly due to large training data sets. In Earth however, earthquake interevent times range from 10's-100's of years and geophysical data typically exist for only a portion of an earthquake cycle. Sparse data presents a serious challenge to training machine learning models. Here we describe a transfer learning approach using numerical simulations to train a convolutional encoder-decoder that predicts fault-slip behavior in laboratory experiments. The model learns a mapping between acoustic emission histories and fault-slip from numerical simulations, and generalizes to produce accurate results using laboratory data. Notably slip-predictions markedly improve using the simulation-data trained-model and training the latent space using a portion of a single laboratory earthquake-cycle. The transfer learning results elucidate the potential of using models trained on numerical simulations and fine-tuned with small geophysical data sets for potential applications to faults in Earth.