Extraction of Surface Soil Geochemical Characteristics of Element Concertation by Principal Component Analysis

Abstract

Understanding the geochemical characteristics of various soils is significant for revealing the mechanisms occurring under natural conditions, assessing the environmental risks and managing the land use. However these various soils have complex forming mechanisms. This study examined geochemical characteristics of surface soils using statistical analysis. Principal component analysis (PCA), a very commonly used multivariate technique that can extract hidden structures and patterns from high-dimensional data, was applied to 633 classified soil samples. Soil samples collected in the Miyagi and Ibaraki prefectures were analyzed for major elements (Si, Ti, Al, Fe, Mg, Mn, Ca, Na and K) and heavy metals (Cr, Cu, Zn, As and Pb). Major elements showed most distinctive relationships with an inverse correlation between Si and Ti, Fe and Al. Heavy metals did not exhibit clear correlations with each other. However, background concentrations could be estimated using the frequency distribution and log-normal distribution curves. The background concentration of Cr, Cu, Zn, As and Pb were 50 ～150 mg/kg, 75 ～150 mg/kg, 200 ～300 mg/kg, 25 ～50 mg/kg and 30 ～60 mg/kg, respectively. Results of PCA clarified four common factors controlling major elements and heavy metals in the soil samples. They are ‘Physicochemical formation', ‘Soil components',‘ Vegetation effect' and ‘Adsorption and desorption'. Soils are formed by leaching reactions from rain and groundwater and mixed organic matter from different vegetation types. Therefore, major elements and heavy metals in soil are affected by these processes, which are common to most soil types.

Keywords

• soil classification
• principal component analysis
• geochemical characteristics