Enhancing Deep Learning based RMT Data Inversion using Gaussian Random Field

TL;DR

This paper introduces a Gaussian Random Field-based data generation method to improve the generalization of deep learning models for Radio Magnetotelluric data inversion, especially on out-of-distribution and noisy data.

Contribution

The paper proposes a novel DL inversion scheme using Gaussian Random Fields to enhance model generalization to unseen data distributions.

Findings

GRF dataset improves generalization over homogeneous background datasets.

The network accurately recovers structures in synthetic and real field data.

The scheme outperforms traditional gradient-based methods in noisy conditions.

Abstract

Deep learning (DL) methods have emerged as a powerful tool for the inversion of geophysical data. When applied to field data, these models often struggle without additional fine-tuning of the network. This is because they are built on the assumption that the statistical patterns in the training and test datasets are the same. To address this, we propose a DL-based inversion scheme for Radio Magnetotelluric data where the subsurface resistivity models are generated using Gaussian Random Fields (GRF). The network's generalization ability was tested with an out-of-distribution (OOD) dataset comprising a homogeneous background and various rectangular-shaped anomalous bodies. After end-to-end training with the GRF dataset, the pre-trained network successfully identified anomalies in the OOD dataset. Synthetic experiments confirmed that the GRF dataset enhances generalization compared to a homogeneous background OOD dataset. The network accurately recovered structures in a checkerboard resistivity model, and demonstrated robustness to noise, outperforming traditional gradient-based methods. Finally, the developed scheme is tested using exemplary field data from a waste site near Roorkee, India. The proposed scheme enhances generalization in a data-driven supervised learning framework, suggesting a promising direction for OOD generalization in DL methods.