Sebastián Barra , Marcos Moreno , Francisco Ortega-Culaciati , Roberto Benavente , Rodolfo Araya , Jonathan Bedford , Ignacia Calisto
{"title":"A supervised machine learning approach for estimating plate interface locking: Application to Central Chile","authors":"Sebastián Barra , Marcos Moreno , Francisco Ortega-Culaciati , Roberto Benavente , Rodolfo Araya , Jonathan Bedford , Ignacia Calisto","doi":"10.1016/j.pepi.2024.107207","DOIUrl":null,"url":null,"abstract":"<div><p>Estimating locking degree at faults is important for determining the spatial distribution of slip deficit at seismic gaps. Inverse methods of varying complexity are commonly used to estimate fault locking. Here we present an innovative approach to infer the degree of locking from surface GNSS velocities by means of supervised learning (SL) algorithms. We implemented six different SL regression methods and apply them in the Central Chile subduction. These methods were first trained on synthetic distributions of locking and then used to infer the locking from GNSS observations. We tested the performance of each algorithm and compared our results with a least squares inversion method. Our best results were obtained using the Ridge regression, which gives a root mean square error (RMSE) of 1.94 mm/yr compared to GNSS observations. The ML-based locking degree distribution is consistent with results from the EPIC Tikhonov regularized least squares inversion and previously published locking maps. Our study demonstrates the effectiveness of machine learning methods in estimating fault locking and slip, and provides flexible options for incorporating prior information to avoid slip instabilities based on the characteristics of the training set. Exploring uncertainties in the physical model during training could improve the robustness of locking estimates in future research efforts.</p></div>","PeriodicalId":54614,"journal":{"name":"Physics of the Earth and Planetary Interiors","volume":"352 ","pages":"Article 107207"},"PeriodicalIF":2.4000,"publicationDate":"2024-05-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Physics of the Earth and Planetary Interiors","FirstCategoryId":"89","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0031920124000657","RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"GEOCHEMISTRY & GEOPHYSICS","Score":null,"Total":0}
引用次数: 0
Abstract
Estimating locking degree at faults is important for determining the spatial distribution of slip deficit at seismic gaps. Inverse methods of varying complexity are commonly used to estimate fault locking. Here we present an innovative approach to infer the degree of locking from surface GNSS velocities by means of supervised learning (SL) algorithms. We implemented six different SL regression methods and apply them in the Central Chile subduction. These methods were first trained on synthetic distributions of locking and then used to infer the locking from GNSS observations. We tested the performance of each algorithm and compared our results with a least squares inversion method. Our best results were obtained using the Ridge regression, which gives a root mean square error (RMSE) of 1.94 mm/yr compared to GNSS observations. The ML-based locking degree distribution is consistent with results from the EPIC Tikhonov regularized least squares inversion and previously published locking maps. Our study demonstrates the effectiveness of machine learning methods in estimating fault locking and slip, and provides flexible options for incorporating prior information to avoid slip instabilities based on the characteristics of the training set. Exploring uncertainties in the physical model during training could improve the robustness of locking estimates in future research efforts.
期刊介绍:
Launched in 1968 to fill the need for an international journal in the field of planetary physics, geodesy and geophysics, Physics of the Earth and Planetary Interiors has now grown to become important reading matter for all geophysicists. It is the only journal to be entirely devoted to the physical and chemical processes of planetary interiors.
Original research papers, review articles, short communications and book reviews are all published on a regular basis; and from time to time special issues of the journal are devoted to the publication of the proceedings of symposia and congresses which the editors feel will be of particular interest to the reader.