Statistical analysis of pollution in stormwater infiltration basins

Abstract

In order to characterize soil pollution in stormwater infiltration basins, four basins of the Lyon area were selected and sampled. Altogether, 66 soil samples were collected. Each sample was analyzed for pH, cationic exchange capacity, particle size distribution, total organic carbon, total nitrogen, total phosphorus, metals, total hydrocarbons and polycyclic aromatic hydrocarbons (PAHs). Data was explored in a classic way (soil profiles) and using multivariate analysis techniques. Two methods were applied: Principal Component Analysis (PCA) and Cluster Analysis (CA). Main results are presented in this paper. Pollutant concentrations decrease rapidly with depth while pH and grain size increase. Sustainable concentrations are reached at a 50-cm depth, even after 21 years of operation. Multivariate analysis shows how pollution affects each sampling depth.