A methodology of integration of magnetometric and geochemical soil contamination measurements

•Integrated, surface and in-profile, measurement of soil magnetic susceptibility•Improved assessment of soil pollution using integrated magnetometric measurement•In small area: soil preparation before measurement at control points and regression•In large area: integration of surface and vertical measurements using cokriging

There is still a high interest in the improvement of soil magnetometry procedures that could increase its accuracy. As it was often discussed in the literature, the depth of penetration of the magnetometric probe that is used to measure magnetic susceptibility on the soil surface (e.g. MS2D) can be insufficient to acquire the magnetic signal of pollutants accumulated in the deeper-located soil horizons. The goal of this study was to propose practical methods of the integration of surface and vertical measurements of soil magnetic susceptibility, and thus eliminate, or at least, minimize the number of geochemical measurements. It was assumed that these methods must be cost- and time-effective, and easy to implement, because it is of vital importance for fast screening method. Results obtained in this paper clearly showed that the method of the integration of magnetometric measurements should depend mainly on a thickness of organic soil horizon and on a thickness of this soil layer where the magnetic enhancement was observed. As soon as the thickness of an organic soil horizon falls in the interval between 15% and 30% of the effective penetration range of a Bartington MS2D magnetometer the assessment of the potentially polluted area becomes problematic. In such situation, the extent of the polluted area was determined using values of soil magnetic susceptibility measured with a MS2D device that were integrated with the values of a width of magnetic enhancement and the depth where it started to be observed in a soil profile. Values of Pearson's correlation coefficient between soil magnetic susceptibility and the Pollution Load Index ranged between 0.63 and 0.77.In the case of a large area where the removal of soil horizon was problematic, it was sufficient to use cokriging in order to integrate the values of magnetic susceptibility measured on the soil surface with magnetometric parameters measured in the soil profile. However, it was observed that if the number of sample points is high, the extent of the potentially polluted area can be assessed using ordinary kriging and the values of magnetic susceptibility measured on the soil surface. Validation results showed that the values of Pearson's correlation coefficient between magnetic susceptibility that was estimated using kriging and the concentrations of Co, Ni, Cd, Pb were significant, and equaled to 0.77, 0.76, 0.65, and 0.74, respectively.