A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model

Abstract Geophysical inversion plays a pivotal role in understanding the Earth's internal structure. Recently generative neural networks (GNNs), such as normalizing flows models (NFMs), have gained popularity for solving Bayesian inversion problems. However, the posterior probability density fu...

Full description

Saved in:
Bibliographic Details
Main Authors: Binbin Liao, Xiaodong Chen, Jianqiao Xu, Jiangcun Zhou, Heping Sun
Format: Article
Language:English
Published: Wiley 2025-03-01
Series:Journal of Geophysical Research: Machine Learning and Computation
Online Access:https://doi.org/10.1029/2024JH000479
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850206760958164992
author Binbin Liao
Xiaodong Chen
Jianqiao Xu
Jiangcun Zhou
Heping Sun
author_facet Binbin Liao
Xiaodong Chen
Jianqiao Xu
Jiangcun Zhou
Heping Sun
author_sort Binbin Liao
collection DOAJ
description Abstract Geophysical inversion plays a pivotal role in understanding the Earth's internal structure. Recently generative neural networks (GNNs), such as normalizing flows models (NFMs), have gained popularity for solving Bayesian inversion problems. However, the posterior probability density functions (PDFs) obtained by amortized GNN‐based methods often deviates from the target distribution. This discrepancy arises because traditional amortized methods use joint PDFs as the objective in loss functions, rather than the conditional PDFs of the observed data. To address this, we propose the Iterative Normalizing Flows Model (INFM), a novel approach that mitigates loss function bias by progressively narrowing the prior distribution's support set in each iteration, while ensuring that the posterior distribution accurately converges to the target distribution. Our experiment, validated on high‐dimensional Bayesian inversion tasks, shows that INFM significantly enhances inversion accuracy without increasing network complexity or computational cost. When applied to the Earth's 1‐D structure model inversion, our method revealed key insights, such as a lower core density compared to the Preliminary Reference Earth Model (PREM) model and the presence of anisotropy in both the mantle and core, consistent with previous studies. These findings suggest that the INFM method offer high computational efficiency and accuracy, making it well‐suited for large‐scale geophysical inversion problems.
format Article
id doaj-art-47f4e67dd7564a4dbeb8ed6ebcc54cb7
institution OA Journals
issn 2993-5210
language English
publishDate 2025-03-01
publisher Wiley
record_format Article
series Journal of Geophysical Research: Machine Learning and Computation
spelling doaj-art-47f4e67dd7564a4dbeb8ed6ebcc54cb72025-08-20T02:10:42ZengWileyJournal of Geophysical Research: Machine Learning and Computation2993-52102025-03-0121n/an/a10.1029/2024JH000479A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows ModelBinbin Liao0Xiaodong Chen1Jianqiao Xu2Jiangcun Zhou3Heping Sun4Innovation Academy for Precision Measurement Science and Technology Chinese Academy of Sciences Wuhan ChinaInnovation Academy for Precision Measurement Science and Technology Chinese Academy of Sciences Wuhan ChinaInnovation Academy for Precision Measurement Science and Technology Chinese Academy of Sciences Wuhan ChinaInnovation Academy for Precision Measurement Science and Technology Chinese Academy of Sciences Wuhan ChinaInnovation Academy for Precision Measurement Science and Technology Chinese Academy of Sciences Wuhan ChinaAbstract Geophysical inversion plays a pivotal role in understanding the Earth's internal structure. Recently generative neural networks (GNNs), such as normalizing flows models (NFMs), have gained popularity for solving Bayesian inversion problems. However, the posterior probability density functions (PDFs) obtained by amortized GNN‐based methods often deviates from the target distribution. This discrepancy arises because traditional amortized methods use joint PDFs as the objective in loss functions, rather than the conditional PDFs of the observed data. To address this, we propose the Iterative Normalizing Flows Model (INFM), a novel approach that mitigates loss function bias by progressively narrowing the prior distribution's support set in each iteration, while ensuring that the posterior distribution accurately converges to the target distribution. Our experiment, validated on high‐dimensional Bayesian inversion tasks, shows that INFM significantly enhances inversion accuracy without increasing network complexity or computational cost. When applied to the Earth's 1‐D structure model inversion, our method revealed key insights, such as a lower core density compared to the Preliminary Reference Earth Model (PREM) model and the presence of anisotropy in both the mantle and core, consistent with previous studies. These findings suggest that the INFM method offer high computational efficiency and accuracy, making it well‐suited for large‐scale geophysical inversion problems.https://doi.org/10.1029/2024JH000479
spellingShingle Binbin Liao
Xiaodong Chen
Jianqiao Xu
Jiangcun Zhou
Heping Sun
A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
Journal of Geophysical Research: Machine Learning and Computation
title A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
title_full A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
title_fullStr A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
title_full_unstemmed A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
title_short A Novel Bayesian Geophysical Inversion Method to Address Loss Function Bias: The Iterative Normalizing Flows Model
title_sort novel bayesian geophysical inversion method to address loss function bias the iterative normalizing flows model
url https://doi.org/10.1029/2024JH000479
work_keys_str_mv AT binbinliao anovelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT xiaodongchen anovelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT jianqiaoxu anovelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT jiangcunzhou anovelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT hepingsun anovelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT binbinliao novelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT xiaodongchen novelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT jianqiaoxu novelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT jiangcunzhou novelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel
AT hepingsun novelbayesiangeophysicalinversionmethodtoaddresslossfunctionbiastheiterativenormalizingflowsmodel