Source Apportionment of Soil Heavy Metals Based on Spatial Co-Location Patterns

Abstract

Existing source apportionment methods for soil heavy metals have failed to establish a credible relationship through global models because the high levels of spatial heterogeneity in soil heavy metals. Bivariate local Moran’s I (B-LISA) provides a local geographic perspective to explore the spatial co-location patterns of driving factors and receptors. The accumulation of heavy metals in soil originates from the high aggregation of driving factors. A factor is considered to be a driving factor if the high aggregation of the factor is spatially co-located with a high accumulation of heavy metals in the soil. Based on this, we propose the use of DHH index to quantify the effect of the potential driving factors on soil heavy metals. The results showed that among the nine factors, the Non-metallic Mineral Product (NMP) industry was the main factor in soil Cd pollution followed by the Chemical (CHM) industry and Non-Metallic Mining and Dressing (NMD). Our analysis revealed that soil Cd contamination was mainly distributed in the economic development zone(YSH town) and the historical coal mining area(TJS town). These two areas show different localized pollution patterns. One is the YSH with multiple industries dominated by NMP, CHM and non-ferrous metals (NMF) with complex interactions. The other is TJS with a single dominant influence of mining industry. This study propose a technical approach to precisely locate and quantify the effects of pollution and source and suggest that source deterrence should be considered based on the spatial co-location patterns of sources and pollution.

Keywords: Soil heavy metal, source apportionment, Bivariate local Moran's I, Spatial co-location pattern, Spatial interaction