Machine Learning-Based Classification of Geological Structures from Magnetic Anomaly Data: Case study of Northern Nigeria Basement Complex

Abstract

The geological terrain of Northern Nigeria features a complex landscape of mineral resources that calls for systematic exploration. This study employs a machine learning framework on geomagnetic data to improve the detection of subsurface geological (mineralized) structures. By combining analytic signal processing with machine learning classifiers (Random Forest (RF) and Gradient Boosting (GB)), we analyze magnetic anomalies to predict subsurface geological features with a classification accuracy of 95.5%. In the first case study focused on the northern Nigeria basement complex, the results identify mineral-rich zones at various depths, from near-surface (280 m) to deep crustal levels (> 2000 m). Key prospective areas include Het, Kagoro, and Durbi, which are notable for deposits of monazite, tantalite, columbite, tourmaline, beryl, and kaolin. The study achieves a Pearson correlation coefficient of 0.956 between predicted and observed structures, demonstrating the effectiveness of this approach in subsurface exploration and geo-resource assessment. The methodology not only confirms known geological features but also uncovers previously unrecognized mineral-rich structures, supporting a more data-driven strategy for resource evaluation in geologically complex regions.