Data quality oriented procedure, for detailed mapping of heavy metals in urban topsoil as an approach to human health risk assessment

Abstract

Urban soils' health is important to the community because of the soils' potential use for recreational activities. A data quality-oriented approach to sampling design is proposed for performing soil representative surveys that gives support to defensible and statistically-based decisions. Krowoderski park in Cracow (Poland) was selected as a study case to investigate heavy metals (HMs) accumulation and to assess human risk exposure according to simulated scenarios. Statistical power was computed for optimizing the number of samples to compare HMs concentration against legal upper tolerance levels (LUTL). The samples' location was iteratively designed as random spatial distribution throughout the study area, followed by K Ripley's test for selecting the best sampling scheme and avoiding points of clustering or dispersion at several ranges. The total content of Cd, Cu, Pb, Zn, coarse size particles fraction and fine size particles texture, bulk density, pH, total C and S were measured in topsoil at each location using composite sampling. The hydraulic properties were estimated using pedotransfer functions. Statistical analysis of topsoil data shows low correlation between heavy metals, whereas high correlation was found between total S with Cu and Pb as well as total C with Cu and Pb. The concentration of all the HMs analysed was found to be under LUTL in all locations in the park, except for one point that is an outlier for Pb, although the values of several indexes for pooled HMs categorize the park as medium to highly polluted. Spatial autocorrelation was explored for every heavy metal and for elaborated pollution indexes, then maps were drawn using geostatistics. A human health risk assessment (HHRA) was computed for several simulated scenarios finding that risk exists for children from Pb through high ingestion of soil particles.