{"title":"Physics-Informed Convolutional Decoder (PICD): A Novel Approach for Direct Inversion of Heterogeneous Subsurface Flow","authors":"Nanzhe Wang, Xiang-Zhao Kong, Dongxiao Zhang","doi":"10.1029/2024GL108163","DOIUrl":null,"url":null,"abstract":"<p>We propose a physics-informed convolutional decoder (PICD) framework for inverse modeling of heterogenous groundwater flow. PICD stands out as a direct inversion method, eliminating the need for repeated forward model simulations. The framework combines data-driven and physics-driven approaches by integrating monitoring data and domain knowledge into the inversion process. PICD utilizes a convolutional decoder to effectively approximate the spatial distribution of hydraulic heads, while Karhunen–Loève expansion (KLE) is employed to parameterize hydraulic conductivities. During the training process, the stochastic vector in KLE and the parameters of the convolutional decoder are adjusted simultaneously to minimize the data-mismatch and the physical violation. The final optimized stochastic vectors correspond to the estimation of hydraulic conductivities, and the trained convolutional decoder can predict the evolution and distribution of hydraulic heads. Various scenarios of groundwater flow are examined and results demonstrate the framework's capability to accurately estimate heterogeneous hydraulic conductivities and to deliver satisfactory predictions of hydraulic heads, even with sparse measurements.</p>","PeriodicalId":12523,"journal":{"name":"Geophysical Research Letters","volume":null,"pages":null},"PeriodicalIF":4.6000,"publicationDate":"2024-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1029/2024GL108163","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Geophysical Research Letters","FirstCategoryId":"89","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1029/2024GL108163","RegionNum":1,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"GEOSCIENCES, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
We propose a physics-informed convolutional decoder (PICD) framework for inverse modeling of heterogenous groundwater flow. PICD stands out as a direct inversion method, eliminating the need for repeated forward model simulations. The framework combines data-driven and physics-driven approaches by integrating monitoring data and domain knowledge into the inversion process. PICD utilizes a convolutional decoder to effectively approximate the spatial distribution of hydraulic heads, while Karhunen–Loève expansion (KLE) is employed to parameterize hydraulic conductivities. During the training process, the stochastic vector in KLE and the parameters of the convolutional decoder are adjusted simultaneously to minimize the data-mismatch and the physical violation. The final optimized stochastic vectors correspond to the estimation of hydraulic conductivities, and the trained convolutional decoder can predict the evolution and distribution of hydraulic heads. Various scenarios of groundwater flow are examined and results demonstrate the framework's capability to accurately estimate heterogeneous hydraulic conductivities and to deliver satisfactory predictions of hydraulic heads, even with sparse measurements.
期刊介绍:
Geophysical Research Letters (GRL) publishes high-impact, innovative, and timely research on major scientific advances in all the major geoscience disciplines. Papers are communications-length articles and should have broad and immediate implications in their discipline or across the geosciences. GRLmaintains the fastest turn-around of all high-impact publications in the geosciences and works closely with authors to ensure broad visibility of top papers.