A semi-arid climate's use of exploratory data analysis (EDA) as a reliable non-parametric method for geochemical mapping.

In this study, exploratory data analysis (EDA) methods, specifically boxplots, were employed to examine the composition of stream sediments in the Collo area of Northeast Algeria. The region's diverse lithological formations, interactions between permanent and temporary watercourses, intermittent flash floods, rugged topography, and fluctuating climatic conditions contribute significantly to data variability. Utilizing the boxplot's capabilities, our analytical approach successfully identified outliers, recorded the actual distribution of data, and determined skewness without the need for prior data processing, often required in conventional statistical studies. The boxplot proved to be an effective tool, offering insights into the geographical distribution of elemental concentrations, including iron (Fe), zinc (Zn), copper (Cu), arsenic (As), chromium (Cr), and lead (Pb). Geochemical mapping, driven by a robust class selection mechanism developed from the boxplot, revealed strong geographical correlations. A notable northeastern anomaly, characterized by elevated concentrations of Pb, Zn, Cu, and As, aligned with known base-metal sulfide and arsenopyrite mineralization sites. Additionally, outlier data for Cr indicated proximity to a plagioclase-lherzolite intrusion and known chromite deposits. The geographical distribution of iron, corresponding with known magnetite and hematite resources, was effectively highlighted using robust class selection based on boxplot analysis. This comprehensive investigation underscores the value of EDA methods in identifying mineralization trends amid the complex variability of the Collo area.