Deep Learning-Based Earthquake Magnitude Estimation Using Seismic Waveform Images

Marat Nurtas, Ayazhan Kumarkhanova, Takhmina Nessipbay

2025

Closed access Metadata (CSL-JSON)
DOI 10.1109/sist61657.2025.11139211
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

In recent decades, researchers have increasingly applied deep learning in seismology, especially for estimating earthquake magnitudes from seismic data analysis. This study introduces deep learning model for earthquake magnitude estimation from seismic wave images. A high-quality Stanford Earthquake Dataset (STEAD), which includes seismic wave data for earthquakes with magnitude M ≥ 3, is used to train the model. Compared to traditional methods, the prediction accuracy is improved using a hybrid convolutional neural networks (CNN) model with bi-directional LSTM (BiLSTM). The results obtained as mean square error (MSE) = 0.19 and mean absolute error (MAE) = 0.34 indicate that the proposed model can predict earthquake magnitude more accurately compared to existing methods.