Scholarlink

Gold is recognized as a major economic mineral resource within the Egyptian Nubian Shield. With advancements in data collection and machine learning algorithms (MLAs), the geoscience community is increasingly turning to remote sensing and MLAs as preliminary exploration tools to narrow down potential exploration zones. This chapter presents the application of several satellite datasets, including multispectral (e.g., Sentinel-2 and ASTER) and hyperspectral (e.g., PRISMA) imagery, to enhance the detection of gold-bearing rocks through four key stages: (1) precise lithological identification using multispectral imagery, digital elevation models (DEMs), and radar data; (2) accurate identification of source and/or host rocks (e.g., ultramafics and related rocks); (3) detection of hydrothermal alteration zones; and (4) localization and proposed distribution maps for gold-bearing zones. Various MLAs, including Artificial Neural Networks (ANN), Extreme Gradient Boosting (XGB), Random Forest (RF), and Support Vector Machines (SVM), were employed to achieve these objectives. Additionally, their accuracies were assessed, and the advantages and limitations of each method were highlighted to provide insights for future research within the geological community.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning Algorithms for Gold-Bearing Alteration Mapping in the Egyptian Nubian Shield Utilizing Remote Sensing Datasets

  • 章节

  • 发表日期：

    2025年

查看全文
  • Ali Shebl,
  • Árpád Csámer

摘要

Gold is recognized as a major economic mineral resource within the Egyptian Nubian Shield. With advancements in data collection and machine learning algorithms (MLAs), the geoscience community is increasingly turning to remote sensing and MLAs as preliminary exploration tools to narrow down potential exploration zones. This chapter presents the application of several satellite datasets, including multispectral (e.g., Sentinel-2 and ASTER) and hyperspectral (e.g., PRISMA) imagery, to enhance the detection of gold-bearing rocks through four key stages: (1) precise lithological identification using multispectral imagery, digital elevation models (DEMs), and radar data; (2) accurate identification of source and/or host rocks (e.g., ultramafics and related rocks); (3) detection of hydrothermal alteration zones; and (4) localization and proposed distribution maps for gold-bearing zones. Various MLAs, including Artificial Neural Networks (ANN), Extreme Gradient Boosting (XGB), Random Forest (RF), and Support Vector Machines (SVM), were employed to achieve these objectives. Additionally, their accuracies were assessed, and the advantages and limitations of each method were highlighted to provide insights for future research within the geological community.