Convolutional neural networks for mineral prospecting through alteration mapping with remote sensing data

TL;DR

The paper addresses the challenge of mapping mineral alteration zones for regional exploration using remote sensing. It proposes a CNN-based framework applied to Landsat 8/9 and ASTER data, with two training-data strategies (ground truth and PCA-derived) to produce alteration maps. Results indicate CNNs provide improved accuracy and spatial coherence over traditional classifiers, particularly when trained on high-quality ground-truth data, with Landsat 9 and ASTER showing complementary strengths for different alteration types. The work demonstrates the practical potential of CNN-driven alteration mapping and provides open-source tools for broader application in geological remote sensing.

Abstract

Traditional geological mapping, based on field observations and rock sample analysis, is inefficient for continuous spatial mapping of features like alteration zones. Deep learning models, such as convolutional neural networks (CNNs), have revolutionised remote sensing data analysis by automatically extracting features for classification and regression tasks. CNNs can detect specific mineralogical changes linked to mineralisation by identifying subtle features in remote sensing data. This study uses CNNs with Landsat 8, Landsat 9, and ASTER data to map alteration zones north of Broken Hill, New South Wales, Australia. The model is trained using ground truth data and an automated approach with selective principal component analysis (PCA). We compare CNNs with traditional machine learning models, including k-nearest neighbours, support vector machines, and multilayer perceptron. Results show that ground truth-based training yields more reliable maps, with CNNs slightly outperforming conventional models in capturing spatial patterns. Landsat 9 outperforms Landsat 8 in mapping iron oxide areas using ground truth-trained CNNs, while ASTER data provides the most accurate argillic and propylitic alteration maps. This highlights CNNs' effectiveness in improving geological mapping precision, especially for identifying subtle mineralisation-related alterations.

Figures (13)

• Figure 1: Simplified geological map of the study area located in the far west of NSW, Australia. The coordinate system is WGS 84 / UTM zone 54S.
• Figure 2: Training samples generated using ground truth data for Landsat 8, Landsat 9, and ASTER data.
• Figure 3: Training samples generated by applying selective PCA shirmard2020integration to Landsat 8, Landsat 9, and ASTER data.
• Figure 4: Machine learning framework for mapping alteration zones utilising remote sensing data. The framework features KNN, SVM, MLP, and CNNs for the classification module to create alteration maps.
• Figure 5: Model architectures featuring a) MLP classifier for the Landsat data, b) MLP classifier for the ASTER data, and c) CNN classifier for mapping target alteration zones using both the Landsat and ASTER data.