Incorporating Deep Learning Into Hydrogeological Modeling: Advancements, Challenges, and Future Directions
Journal of Geophysical Research: Machine Learning and Computation · 2025-06 · 2 (2)
DOI: 10.1029/2025jh000703 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Geophysical Research: Machine Learning and Computation, online erschienen am 20.06.2025. Verfasst von Zhenxue Dai und 17 weiteren Autor:innen. Die Arbeit wurde bisher 20-mal zitiert (Crossref) und nennt 315 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Shandong Province.
Abstract (Originalsprache)
Hydrogeological modeling is essential for managing groundwater systems, especially in the transport and remediation of contaminants. Traditional modeling methods face challenges due to the increasing complexity and volume of data. Deep learning (DL) has emerged as a promising tool, offering significant improvements in accuracy and efficiency for tasks such as time series prediction, spatial data analysis, and inverse modeling. Although recent applications of DL in hydrogeology have shown potential, many models are still in the testing phase due to limited hydrogeological data, the “black box” nature of DL models hindering interpretability, and the substantial computational resources needed for training. Furthermore, the lack of standardized evaluation benchmarks makes it difficult to compare the performance of different DL models in hydrogeological contexts. To advance DL‐based hydrogeological modeling, future research should focus on enhancing data availability through data fusion and public databases, improving model interpretability using physics‐informed and explainable DL techniques, and developing more efficient algorithms for training large‐scale models. Additionally, exploring new computational paradigms, such as quantum computing, could provide revolutionary solutions for handling the computational challenges associated with training complex models. Establishing standardized benchmarks will also be key for assessing the practical utility of DL models and facilitating their generalization in real‐world scenarios. By addressing these challenges and leveraging DL alongside emerging computational technologies, hydrogeological modeling can be significantly advanced, improving the management and remediation of groundwater systems impacted by contaminants.
Förderung
- National Natural Science Foundation of China
- Natural Science Foundation of Shandong Province
Zitieren
APA
Dai, Z., Zhan, C., Yin, H., Chen, J., Xu, L., Xia, Y., Yang, S., Chen, W., Cao, M., Du, Z., Zhang, X., Yan, B., Ma, Y., Wang, H., Moeini, F., Soltanian, M. R., Thanh, H. V., & Carroll, K. C. (2025). Incorporating Deep Learning Into Hydrogeological Modeling: Advancements, Challenges, and Future Directions. Journal of Geophysical Research: Machine Learning and Computation, 2(2). https://doi.org/10.1029/2025jh000703
Vancouver
Dai Z, Zhan C, Yin H, Chen J, Xu L, Xia Y, et al. Incorporating Deep Learning Into Hydrogeological Modeling: Advancements, Challenges, and Future Directions. Journal of Geophysical Research: Machine Learning and Computation. 2025;2(2). doi:10.1029/2025jh000703
BibTeX
@article{Dai2025,
author = {Dai, Zhenxue and Zhan, Chuanjun and Yin, Huichao and Chen, Junjun and Xu, Lulu and Xia, Yuzhou and Yang, Songlin and Chen, Wei and Cao, Mingxu and Du, Zhengyang and Zhang, Xiaoying and Yan, Bicheng and Ma, Yue and Wang, Hao and Moeini, Farzad and Soltanian, Mohamad Reza and Thanh, Hung Vo and Carroll, Kenneth C.},
title = {Incorporating Deep Learning Into Hydrogeological Modeling: Advancements, Challenges, and Future Directions},
journal = {Journal of Geophysical Research: Machine Learning and Computation},
year = {2025},
volume = {2},
number = {2},
doi = {10.1029/2025jh000703}
}
EUVIMED