Analysis of anomaly characteristics in mineralized segments using core drilling and geophysical logging

Reliable recognition of mineralized intervals from borehole logging data remains difficult because weak mineralization, thin layers, borehole disturbance, and lithological changes may produce similar curve responses. This study develops a multi-parameter coupled anomaly identification framework using density, resistivity, natural gamma, acoustic travel time, magnetic susceptibility, and polarizability logs acquired by the Kezuan Geophysical Logging Method. The framework integrates depth resampling, normalization, composite anomaly indexing, boundary-response calculation, anomaly-pattern matching, and lithology-constrained interval merging. Validation was performed using core records from Kezuan Well 1 and additional boreholes. In Well 1, 118 anomaly intervals were identified, of which 93 matched core-confirmed mineralized intervals, giving an accuracy of 78.8%, a false-negative rate of 20.3%, and a false-positive rate of 15.3%. The accuracy exceeded the single-parameter threshold method by 28.8 percentage points and the biparameter intersection method by 14.2 percentage points. External borehole validation produced accuracies of 75.4%–77.2%. The novelty of the framework lies in converting multi-parameter logging responses into boundary-constrained and geologically interpretable mineralized intervals rather than merely stacking abnormal curves. The results indicate that the method improves anomaly recognition under complex borehole conditions, although very thin and weakly mineralized zones remain difficult to detect.