Transient Electromagnetic Machine Learning Inversion Based on Pseudo Wave Field Data

Chen, Jian; Zhang, Yang; Lin, Tingting

doi:10.1109/tgrs.2022.3187021

Cited by 13 publications

(3 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…DL methods can learn both prior parameter relationships and structural similarity of the multiphysics data sets. Therefore, the accuracy of EM imaging may be greatly improved (Chen et al, 2022;Guo et al, 2022), and the uncertainty of imaging could be reduced (Um et al, 2022). These aspects are crucial for DL decision-making inversion and need further study (Oh and Byun, 2021).…”

Section: Potential and Challengesmentioning

confidence: 99%

Electromagnetic imaging and deep learning for transition to renewable energies: a technology review

Castillo-Reyes,

Hu,

Wang

et al. 2023

Front. Earth Sci.

View full text Add to dashboard Cite

Electromagnetic imaging is a technique that has been employed and perfected to investigate the Earth subsurface over the past three decades. Besides the traditional geophysical surveys (e.g., hydrocarbon exploration, geological mapping), several new applications have appeared (e.g., characterization of geothermal energy reservoirs, capture and storage of carbon dioxide, water prospecting, and monitoring of hazardous-waste deposits). The development of new numerical schemes, algorithms, and easy access to supercomputers have supported innovation throughout the geo-electromagnetic community. In particular, deep learning solutions have taken electromagnetic imaging technology to a different level. These emerging deep learning tools have significantly contributed to data processing for enhanced electromagnetic imaging of the Earth. Herein, we review innovative electromagnetic imaging technologies and deep learning solutions and their role in better understanding useful resources for the energy transition path. To better understand this landscape, we describe the physics behind electromagnetic imaging, current trends in its numerical modeling, development of computational tools (traditional approaches and emerging deep learning schemes), and discuss some key applications for the energy transition. We focus on the need to explore all the alternatives of technologies and expertise transfer to propel the energy landscape forward. We hope this review may be useful for the entire geo-electromagnetic community and inspire and drive the further development of innovative electromagnetic imaging technologies to power a safer future based on energy sources.

show abstract

Section: Potential and Challengesmentioning

confidence: 99%

Electromagnetic imaging and deep learning for transition to renewable energies: a technology review

Castillo-Reyes,

Hu,

Wang

et al. 2023

Front. Earth Sci.

View full text Add to dashboard Cite

show abstract

“…This method is an artificial source secondary field method, which has advantages such as high data signal-tonoise ratio, no interference from the primary field, less affected by topographic relief in high-resistivity surrounding rock areas, strong ability to penetrate high-resistivity overburden, and large detection depth. In recent years, it has been favored by karst researchers and applied to karst detection [20][21][22]. Since TDEM observes the electromotive force data in the time domain, it is necessary to obtain the resistivity in the depth domain through inversion and then interpret related geological problems.…”

Section: H Zmentioning

confidence: 99%

Fine Detection and Analysis of Hidden Karst in Wellsite with Quasi-Three-Dimensional TDEM Based on Lateral Constraint

Wu,

Liao,

Chen

et al. 2023

Geofluids

View full text Add to dashboard Cite

Considering that karst caves, underground rivers, and dissolution fractures in shallow carbonate formation in the Sichuan Basin are extremely developed, leakage, failure and plugging difficulties are easy to occur in the drilling process. The TDEM was used to carry out the exploration of hidden karst geological bodies in well QM2, and the quasi-three-dimensional inversion based on lateral constrain was used to invert the TDEM data. Three NW trending anomalous bands were identified in the lower Triassic Jialingjiang Formation within the range of drilling, consisting of seven relatively low-resistivity anomalous zones. Under the guidance of TDEM quasi-three-dimensional inversion resistivity data, the low-resistivity karst development area is avoided, and the specific drilling location of well QM2 is determined. No karst cave and underground river were drilled in the later drilling process of well QM2, as well as no instability phenomenon occurred. It indicates that the TDEM detection results are consistent with the actual drilling, and the quasi-three-dimensional TDEM inversion interpretation data based on lateral constraints is reliable and can accurately detect the buried karst in the wellsite.

show abstract

“…This method is an artificial source secondary field method, which has advantages such as high data signal-to-noise ratio, no interference from primary field, less affected by topographic relief in high-resistivity surrounding rock areas, strong ability to penetrate high-resistivity overburden, and large detection depth. In recent years, it has been favored by karst researchers and applied to karst detection [19][20][21]. Since TDEM observes the electromotive force data in time domain, it is necessary to obtain the resistivity in depth domain through inversion, and then interpret related geological problems.…”

Section: Introductionmentioning

confidence: 99%

Fine detection and analysis of hidden Karst in wellsite with quasi-three-dimensional TDEM based on lateral constraint

Liao

Chen

et al. 2023

Preprint

View full text Add to dashboard Cite

Considering that karst caves, underground rivers and dissolution fractures in shallow carbonate formation in Sichuan Basin are extremely developed, leakage, failure and plugging difficulties are easy to occur in the drilling process. The TDEM was used to carry out the exploration of hidden karst geological bodies in well QM2, and the quasi-three-dimensional inversion based on lateral constrain was used to invert the TDEM data. Three NW trending anomalous bands were identified in the lower Triassic Jialingjiang formation within the range of drilling, consisting of seven relatively low resistivity anomalous zones. Under the guidance of TDEM quasi-three-dimensional inversion resistivity data, the low-resistivity karst development area is avoided and the specific drilling location of well QM2 is determined. No karst cave and underground river were drilled in the later drilling process of well QM2, as well as no instability phenomenon occurred. It indicates that the TDEM detection results are consistent with the actual drilling, and the quasi-three-dimensional TDEM inversion interpretation data based on lateral constraints is reliable, and can accurately detect the buried karst in the well site.

show abstract

Transient Electromagnetic Machine Learning Inversion Based on Pseudo Wave Field Data

Cited by 13 publications

References 34 publications

Electromagnetic imaging and deep learning for transition to renewable energies: a technology review

Electromagnetic imaging and deep learning for transition to renewable energies: a technology review

Fine Detection and Analysis of Hidden Karst in Wellsite with Quasi-Three-Dimensional TDEM Based on Lateral Constraint

Fine detection and analysis of hidden Karst in wellsite with quasi-three-dimensional TDEM based on lateral constraint

Contact Info

Product

Resources

About