Data-driven Mineral Prospectivity Mapping: Unlocking Critical Mineral Resources using Artificial intelligence techniques in Jharkhand and its Surroundings

Unearthing mineral ore deposits involves a complex and resource-intensive endeavor that typically integrates the diversity of geological, geochemical, geophysical, and remote sensing data. This study can set a framework for a preliminary structure in mineral exploration through the application of machine learning and deep learning (ML/DL) techniques which is one of the most demanding approaches now a days in artificial intelligence (AI) world. Leveraging neural networks, convolutional techniques, and spectral analysis methods, our proposed efficient and time-saving approach seeks to uncover meaningful insights from large and intricate geospatial datasets. The workflow begins with the collection and preprocessing of diverse datasets, including Lithological unit, Tectonic component, Multispectral imagery, Geophysical anomaly, Geochemical composition, and point-based sample evidence of Copper and Graphite commodity (critical minerals in India) in Jharkhand and its surroundings. These data layers further stack and cross-correlate through ensemble supervised ML/DL algorithms and are taking through rigorous model sampling and training process to recognize trends indicative of mineralization, enabling automated identification and classification of mineralogical features for critical mineral deposits. Results from the application of this advanced technique in our study with some statistics such as (Area under the Receiver Operating Characteristics Curve (AUC-ROC), F1-score, Precision, Recall) > 0.85 which would be able to showcase the model's ability to identify prospective areas and generate insightful geologic depositional environments with an accuracy over 80%. This also validate with ground truth data and comparison with traditional exploration methods which demonstrate the effectiveness of the proposed approach. In conclusion, this approach surpasses traditional methods by incorporating temporal aspects and cost-effective analysis, revolutionizing the identification and prioritization of evolving patterns and trends in mineral occurrences. Mineral Prospectivity Mapping (MPM) is employed to predict the likelihood of mineral deposits, providing exploration teams with valuable information for targeted and efficient resource allocation.Keywords: Mineralization, Iron Ore deposits, Mineral Prospectivity Mapping, Machine learning, Multispectral Imagery.Graphical Abstract: