Stream sediment geochemical anomaly identification and mapping in the Xinbo area, northern China: Using a catchment basin approach and the Apriori algorithm

Abstract

Stream sediment geochemical data are widely used for regional anomaly identification and mineral exploration mapping. However, conventional anomaly detection methods primarily rely on single-element anomaly extraction of ore-forming elements, an approach that often results in incomplete prospecting information and make it difficult to detect prospecting signals. This study reanalyzes and reinterprets 1:50000 geochemical reconnaissance survey data from the Xinbao area of North China (primarily stream sediment data, with supplementary soil geochemical data). By integrating the Catchment Basin Approach (CBA) and the Apriori algorithm, we (1) constructed a catchment basin model, (2) identified high-order correlations among elements and their concentration levels, and (3) established an association rule dataset of mineral exploration significance. Next, we extracted petrogenetically and metallogenically significant association rules for geochemical mapping and mineral exploration. We further compared these association rules with Principal Component Analysis (PCA) to validate the efficacy of the Apriori algorithm in delineating stream sediment anomalies. Within the target areas delineated based on mineralization element association rules, all known Pb, Zn, Ag and Au deposits previously identified through 1:10000 soil geochemical surveys were accurately detected. Our findings demonstrate that the combined CBA-Apriori approach significantly enhances the detection of geochemical anomalies, thereby improving the applicability of stream sediment data in mineral exploration. Also of note is that our approach served to identify two new target areas - CAVZ-1 (Pb–Zn–Au dominant) and CAVZ-2 (Ag–Au dominant) within the survey area.