Preprints
https://doi.org/10.5194/gmd-2022-117
https://doi.org/10.5194/gmd-2022-117
Submitted as: development and technical paper
05 May 2022
Submitted as: development and technical paper | 05 May 2022
Status: this preprint is currently under review for the journal GMD.

DeepISMNet: Three-Dimensional Implicit Structural Modeling with Convolutional Neural Network

Zhengfa Bi1, Xinming Wu1, Zhaoliang Li2, Dekuan Chang3, and Xueshan Yong3 Zhengfa Bi et al.
  • 1School of Earth and Space Sciences, University of Science and Technology of China, Hefei, Anhui, P.R.China
  • 2China Aero Geophysical Survey and Remote Sensing Center for Natural Resources, Beijing, P.R.China
  • 3Research Institute of Petroleum Exploration & Development-NorthWest(NWGI), PetroChina, Gansu, Lanzhou, P.R.China

Abstract. Implicit structural modeling using sparse and unevenly distributed data is essential for various scientific and societal purposes ranging from natural source exploration to geological hazard forecasts. Most advanced implicit approaches formulate structural modeling as least-squares minimization or spatial interpolation problem and solve partial differential equations (PDEs) for a scalar field that optimally fits all the input data under smooth regularization assumption. However, the PDEs in these methods might be insufficient to model highly complex structures in practice and may fail to reasonably fit a global structure trend when the known data are too sparse. In addition, solving the PDEs with iterative optimization solvers could be computationally expensive in 3-D. In this study, we propose an efficient deep learning method using a convolution neural network to predict a scalar field from sparse structural data associated with multiple distinct stratigraphic layers and faults. Our deep learning architecture is beneficial for the flexible incorporation of empirical geological knowledge by training with numerous and realistic structural models that are automatically generated from a data simulation workflow. It also presents an impressive characteristic of integrating various types of structural constraints by optimally minimizing a hybrid loss function to compare predicted and reference structural models, opening new opportunities for further improving geological modeling. Moreover, the deep neural network, after training, is highly efficient to predict implicit structural models in practical applications. The capacity of our approach for modeling highly deformed geological structures is verified by using both synthetic and real-world datasets, where the produced models are geologically reasonable and structurally consistent with the inputs.

Zhengfa Bi et al.

Status: open (until 30 Jun 2022)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Zhengfa Bi et al.

Zhengfa Bi et al.

Viewed

Total article views: 186 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
149 32 5 186 2 1
  • HTML: 149
  • PDF: 32
  • XML: 5
  • Total: 186
  • BibTeX: 2
  • EndNote: 1
Views and downloads (calculated since 05 May 2022)
Cumulative views and downloads (calculated since 05 May 2022)

Viewed (geographical distribution)

Total article views: 167 (including HTML, PDF, and XML) Thereof 167 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 20 May 2022
Download
Short summary
We present a learning-based implicit structural modeling method to predict a geologically valid and structurally consistent model from spare structural data that are associated with incomplete horizons and faults. Trained with numerous and realistic synthetic data, our network is able to build geologically reasonable models in field examples without the need of solving large systems of equations, opening new opportunities for improving the efficiency and accuracy of geologic modeling.