Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique
Inversion is a fundamental step in magnetotelluric (MT) data routine analysis to retrieve a subsurface geoelectrical model that can be used to inform geological interpretations. To reduce the effect of non-uniqueness and local minimum trapping problems and improve calculation speeds, a data-driven m...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-03-01
|
Series: | Minerals |
Subjects: | |
Online Access: | https://www.mdpi.com/2075-163X/13/4/461 |
_version_ | 1797604142096580608 |
---|---|
author | Xiaojun Liu James A. Craven Victoria Tschirhart |
author_facet | Xiaojun Liu James A. Craven Victoria Tschirhart |
author_sort | Xiaojun Liu |
collection | DOAJ |
description | Inversion is a fundamental step in magnetotelluric (MT) data routine analysis to retrieve a subsurface geoelectrical model that can be used to inform geological interpretations. To reduce the effect of non-uniqueness and local minimum trapping problems and improve calculation speeds, a data-driven mathematical method with a deep neural network was developed to estimate the subsurface resistivity. In this study, a deep learning (DL) inversion technique using a revised multi-head convolutional neural network (CNN) architecture was investigated for MT data analysis. We created synthetic datasets consisting of 100,000 random samples of resistivity layers to train the network’s parameters. The trained model was validated with independent noised datasets, and the predicted results displayed reasonable accuracy and reliability, which demonstrates the potential application of DL inversion for real-world MT data. The trained model was used to analyze MT data collected in the southwestern Athabasca Basin, Canada. The calculated results from the DL method displayed a detailed subsurface resistivity distribution compared to traditional iterative inversion. Since this approach can predict a resistivity model without multiple forward modeling operations after the CNN model is created, this framework is suitable to speed up the computation of multidimensional MT inversion for subsurface resistivity. |
first_indexed | 2024-03-11T04:42:10Z |
format | Article |
id | doaj.art-0b7afe9f208c434b93c3ec41aee85023 |
institution | Directory Open Access Journal |
issn | 2075-163X |
language | English |
last_indexed | 2024-03-11T04:42:10Z |
publishDate | 2023-03-01 |
publisher | MDPI AG |
record_format | Article |
series | Minerals |
spelling | doaj.art-0b7afe9f208c434b93c3ec41aee850232023-11-17T20:35:00ZengMDPI AGMinerals2075-163X2023-03-0113446110.3390/min13040461Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion TechniqueXiaojun Liu0James A. Craven1Victoria Tschirhart2Geological Survey of Canada, 3303-33rd Street NW, Calgary, AB T2L 2A7, CanadaGeological Survey of Canada, 601 Booth Street, Ottawa, ON K1A 0E8, CanadaGeological Survey of Canada, 601 Booth Street, Ottawa, ON K1A 0E8, CanadaInversion is a fundamental step in magnetotelluric (MT) data routine analysis to retrieve a subsurface geoelectrical model that can be used to inform geological interpretations. To reduce the effect of non-uniqueness and local minimum trapping problems and improve calculation speeds, a data-driven mathematical method with a deep neural network was developed to estimate the subsurface resistivity. In this study, a deep learning (DL) inversion technique using a revised multi-head convolutional neural network (CNN) architecture was investigated for MT data analysis. We created synthetic datasets consisting of 100,000 random samples of resistivity layers to train the network’s parameters. The trained model was validated with independent noised datasets, and the predicted results displayed reasonable accuracy and reliability, which demonstrates the potential application of DL inversion for real-world MT data. The trained model was used to analyze MT data collected in the southwestern Athabasca Basin, Canada. The calculated results from the DL method displayed a detailed subsurface resistivity distribution compared to traditional iterative inversion. Since this approach can predict a resistivity model without multiple forward modeling operations after the CNN model is created, this framework is suitable to speed up the computation of multidimensional MT inversion for subsurface resistivity.https://www.mdpi.com/2075-163X/13/4/461magnetotelluricdeep learningsubsurface resistivityCNN architectureinversion algorithm |
spellingShingle | Xiaojun Liu James A. Craven Victoria Tschirhart Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique Minerals magnetotelluric deep learning subsurface resistivity CNN architecture inversion algorithm |
title | Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique |
title_full | Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique |
title_fullStr | Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique |
title_full_unstemmed | Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique |
title_short | Retrieval of Subsurface Resistivity from Magnetotelluric Data Using a Deep-Learning-Based Inversion Technique |
title_sort | retrieval of subsurface resistivity from magnetotelluric data using a deep learning based inversion technique |
topic | magnetotelluric deep learning subsurface resistivity CNN architecture inversion algorithm |
url | https://www.mdpi.com/2075-163X/13/4/461 |
work_keys_str_mv | AT xiaojunliu retrievalofsubsurfaceresistivityfrommagnetotelluricdatausingadeeplearningbasedinversiontechnique AT jamesacraven retrievalofsubsurfaceresistivityfrommagnetotelluricdatausingadeeplearningbasedinversiontechnique AT victoriatschirhart retrievalofsubsurfaceresistivityfrommagnetotelluricdatausingadeeplearningbasedinversiontechnique |