Understanding the Efficacy of U-Net & Vision Transformer for Groundwater Numerical Modelling

Maria Luisa Taccari, Oded Ovadia, He Wang, Xiaohui Chen, Adar Kahana, Peter Jimack

Keywords: Groundwater Modelling, Time-Dependent Forward Modelling, U-Net, Vision Transformer

TL;DR: This paper compares U-Net, U-Net with ViT, and FNO for groundwater modeling. U-Net models outperform FNO in accuracy and efficiency in sparse data scenarios, highlighting their real-world application potential.

Abstract: This paper presents a comprehensive comparison of various machine learning models, namely U-Net, U-Net integrated with Vision Transformers (ViT), and Fourier Neural Operator (FNO), for time-dependent forward modelling in groundwater systems. Through testing on synthetic datasets, it is demonstrated that U-Net and U-Net + ViT models outperform FNO in accuracy and efficiency, especially in sparse data scenarios. These findings underscore the potential of U-Net-based models for groundwater modelling in real-world applications where data scarcity is prevalent.

Submission Number: 31