Spatial Interpolation Methods and Pollution Assessment of Heavy Metals of Soil in Typical Areas

Abstract

It is important to choose the best spatial interpolation method to reflect spatial distribution features and evaluate soil heavy metal pollution. The spatial distribution of arsenic (As) and cadmium (Cd) concentrations in top soil samples from Hubei Province were studied by four frequently-used spatial interpolation methods, including inverse distance weighted (IDW), radial basis function (RBF), local polynomial interpolation (LPI) and ordinary kriging (OK). The interpolation precision and effect of the spatial distribution of the four methods were compared with the results of cross validation and spatial distribution, and the pollution was assessed by the geoaccumulative index (Igeo) and indicator kriging (IK). The results showed that the four interpolation methods had small prediction errors, but that the interpolation effects were quite different. Among them, LPI had the most serious smoothing effect, followed by OK. The IDW and RBF best retained the extreme value information for element concentrations, and interpolation results were more detailed-and so to accurately understand the distribution of soil heavy metals, IDW or RBF methods were recommended. Taking the arithmetic mean of heavy metal concentrations in deep soil of Hubei Province as the background value, the evaluation result of geo-cumulative index pollution allowed exceedance percentages for As and Cd accounted for 5.5% and 99.0% respectively. The soils of the study areas were heavily contaminated with Cd. The pollution evaluation result from IK showed that high probability contaminated areas, with moderate-heavy contamination levels, were mainly located in the central part of the study area. The authors concluded that development of agriculture in the research area should include attention to Cd pollution and that soils there required the effective treatment and restoration of Cd levels.