Geostatistical modeling of stream sediment anomalies in the Biga Peninsula, NW Türkiye: A threshold-based approach for mineral exploration

Abstract

The Biga Peninsula (NW Türkiye), a key segment of the Alpine–Himalayan metallogenic belt, is characterized by extensive mineralization and geological complexity. This study integrates stream sediment geochemistry and geostatistics to model the spatial distributions of As, Cu, Pb, Zn, and Cd based on 2,369 stream sediment samples. The geochemical results indicate wide concentration ranges for the elements: As (from 3.00 to 839.00 ppm), Cu (from 3.67 to 1680.00 ppm), Pb (from 5.00 to 3989.62 ppm), Zn (from 8.00 to 17762.20 ppm), and Cd (from 0.10 to 210.20 ppm), with standard deviations ranging from 6.43 ppm (Cd) to 488.18 ppm (Zn). Descriptive statistics reveal non-normal distributions, with strong correlations among Zn-Cd (r=0.92), Pb-Zn (r=0.90), and Pb-Cd (r=0.84), while As and Cu appear to be geochemically independent. Cluster analysis confirms a strong geochemical affinity among Cd, Zn, and Pb, whereas As and Cu remain unrelated. Five statistical methods -arithmetic mean, median, interquartile range, percentile-based approach, and cumulative probability- were used to determine elemental threshold values, leading to the classification of anomalies as moderate, high, and very high. Variogram modeling reveals isotropic spatial patterns, enabling the use of ordinary kriging to estimate values at 15,312 grid points. Anomaly maps highlight a distinct Cd anomaly in the southeast, overlapping Zn-Pb-Cd zones, along with independent Cu-As patterns. Comparison with known mineral deposits confirms strong spatial correlations. High As concentrations were associated with the Sahinli, Kocatarla, and Balya deposits, whereas Cu anomalies aligned with the Koru, Kuşçayır, Handeresi, Bağırkaçdere, Tepeoba, Kocatarla, and Arapuçandere deposits. Pb- Zn-Cd anomalies correspond to mineralized zones such as Koru, Handeresi, Bağırkaçdere, and Balya. This study demonstrates a novel application of integrated geochemical and geostatistical modeling, employing a comparative, threshold-based approach to identify mineral exploration targets in the geologically complex Biga Peninsula.