Fractal modeling for geochemical data of deep-sea surface sediments: A case study from Zhongsha Island, Southern China Sea

Abstract

This study applied principal component analysis (PCA), concentration-area (C-A) and spectrum-area (S-A) fractal models to identify geochemical element combination anomalies related to heavy metals from the data for 43 geochemical elements in surface sediments of the Zhongsha Islands. The relationships between these 43 elements were first examined by applying correlation and cluster analysis to the surface sediment sample dataset. The multifractal spectral functions of the 43 elements were then calculated and all elements were classified using the curve shape of the spectral function. The results of classification were consistent with that of cluster analysis, thereby verifying the reliability of fractal analysis of the study area. PCA was then used to identify element combination anomalies and to assess high-value areas of element contents, and the S-A model was used to distinguish between the anomaly and background signal. The results of this multiple-analysis approach show that the element combination anomaly is mainly located in the deep-sea plain area, with a few elevated values in the Zhongsha South Basin. This study showed that quantitative evaluation of anomalies of multi-element associations can effectively delineate a target area and that anomalies of element associations extracted by the S-A fractal method can provide a more accurate outline areas containing high contents of combined elements. This study can act as a theoretical reference for the environmental assessment of the Zhongsha sea area.