T120 Three-Dimensional Geological Mapping to Support Societal Development II

Wednesday, 28 October 2020
10:00 AM - 12:00 PM
Meeting Rooms

MACHINE LEARNING METHODS FOR 3D GEOLOGICAL MODEL CONSTRUCTION

Paper number: T120-7
Program: Technical Sessions
Time: 11:10 AM – 11:25 AM

Authors

Abstract

Although machine learning methods have been widely used for a myriad of geoscience applications, their use for 3D geological modelling has been limited. However, due to rapid advancement of machine learning techniques, dramatic increases in computational resources, and easy to use machine learning frameworks (e.g. Pytorch, Tensorflow) there has been increasing interest in applying these methods. Machine learning methodologies can be used to address inherent limitations of extensively used numerical methods for 3D geological modelling, namely their inability to incorporate all relevant available data and knowledge. These limitations can lead to the production of geologically invalid models, particularly in complex settings. Building robust and representative 3D geological models of a region then requires incorporating all relevant available data and knowledge. Furthermore, with an ever-increasing volume of multidisciplinary data arising from open data and technological advancements in data acquisition, flexible and scalable 3D modelling methods are required. This is especially important for initiatives requiring large-scale 3D geological modelling, such as digital twin earth system models and national 3D geological models. To overcome these challenges, machine learning methods can provide a promising scalable end-to-end framework in which all relevant available data and knowledge is incorporated.

This presentation will briefly review current machine learning methodologies used for the purpose of 3D geological modelling, including examples and a summary of current research gaps. In addition, preliminary 3D geological modelling results will be presented from a developed Graph Neural Network (GNN) architecture constrained by data and knowledge.

Geological Society of America Abstracts with Programs. Vol. 52, No. 6

doi: 10.1130/abs/2020AM-355922

© Copyright 2020 The Geological Society of America (GSA), all rights reserved. Permission is hereby granted to the author(s) of this abstract to reproduce and distribute it freely, for noncommercial purposes. Permission is hereby granted to any individual scientist to download a single copy of this electronic file and reproduce up to 20 paper copies for noncommercial purposes advancing science and education, including classroom use, providing all reproductions include the complete content shown here, including the author information. All other forms of reproduction and/or transmittal are prohibited without written permission from GSA Copyright Permissions.