Computational Study of Foreshocks in the Burridge-Knopoff Earthquake Model Using Machine Learning
Description:
The study of mechanical models of earthquake faults are important to understand the different behaviors observed in real earthquakes. The model we consider for this work was introduced by Burridge and Knopoff [Bull. Seismol. Soc. Am. 57, 341 (1967)], which, consists of blocks connected by linear springs in contact with a moving rough surface. It was implemented a numerical simulation of the model founding a wide variety in the events size which have a power law distribution. As the next step, it was made a data base of artificial earthquakes with the purpose of training an artificial neural networks (ANN) model that are able to predict the events generated by the simulation. The ANN models have an acceptable prediction to calculate the event magnitude, but it was found that they have difficulty to predict the time when they occurred.
Session: Numerical Modeling in Seismology: Developments and Applications [Poster]
Type: Poster
Date: 4/20/2023
Presentation Time: 08:00 AM (local time)
Presenting Author: Jesus G. Ortega
Student Presenter: Yes
Invited Presentation:
Authors
Jesus Ortega
Presenting Author
Corresponding Author
jesus.ortega3@upr.edu
Puerto Rico Seismic Network, University of Puerto Rico, Mayagüez
Rafael Ramos
rafaela.ramos@upr.edu
University of Puerto Rico, Mayaguez
Computational Study of Foreshocks in the Burridge-Knopoff Earthquake Model Using Machine Learning
Category
Numerical Modeling in Seismology: Developments and Applications