Sub3DNet1.0: a deep-learning model for regional-scale 3D subsurface structure mapping

Abstract

Abstract. This study introduces an efficient deep-learning model based on convolutional neural networks with joint autoencoder and adversarial structures for 3D subsurface mapping from 2D surface observations. The method was applied to delineate paleovalleys in an Australian desert landscape. The neural network was trained on a 6400 km2 domain by using a land surface topography as 2D input and an airborne electromagnetic (AEM)-derived probability map of paleovalley presence as 3D output. The trained neural network has a squared error <0.10 across 99 % of the training domain and produces a squared error <0.10 across 93 % of the validation domain, demonstrating that it is reliable in reconstructing 3D paleovalley patterns beyond the training area. Due to its generic structure, the neural network structure designed in this study and the training algorithm have broad application potential to construct 3D geological features (e.g., ore bodies, aquifer) from 2D land surface observations.

Expanding the Spatial Reach and Human Impacts of Critical Zone Science

“…Their model revealed complex relationships between surface and subsurface features and offered an automated, low‐cost method to generate 3‐D subsurface images that inherits the probability structure from 2‐D surface images. Machine‐learning methods have been shown to outperform traditional calibration approaches in estimating subsurface parameters because they can directly infer parameters from observations and better capture the highly nonlinear relationships with fewer realizations (Cromwell et al., 2021; Jiang et al., 2021). Another deep neural network model used widely available time series of streamflow data to estimate subsurface permeability, which can be further used to estimate flow paths and weathering fronts (Jiang et al., 2022; Tartakovsky et al., 2020).…”

Application of Multilayer Perceptron Neural Network in Geological Modeling of Categorical Variables: A Case Study in Peru

“…Recently, there has been a surge of interest in threedimensional geological modeling supported by machine learning techniques. Multiple approaches have been developed ranging from classical techniques, such as support vector machines (SVM) [24], decision trees (DT) [25], and random forests (RF) [26], to advanced neural structures such as depth-first neural networks (DFNN) [27], convolutional (CNN) [28], recurrent (RNN) [29], graph (GNN) [30], and generative adversarial (GAN) [31].…”