Using Geostatistical Gaussian Simulation for Designing and Interpreting Soil Surface Magnetic Susceptibility Measurements

This paper presents a new approach to the assessment of the uncertainty of using geostatistical Gaussian simulation in soil magnetometry. In the study area, numerous measurements of soil magnetic susceptibility were made, and spatial distributions of soil magnetic susceptibility were simulated. The parameters of variograms of soil magnetic susceptibility measured in the study area were determined and compared with those of simulated soil magnetic susceptibility. Regardless of the measurement scheme used, reproducibility of the original semivariograms of soil magnetic susceptibility was satisfactorily achieved when applying simulated values. A nugget effect, a sill, and a range of correlations of variograms of simulated values of soil magnetic susceptibility were similar to those of measured values. When the input data for the geostatistical simulation were averaged, the measured values of soil magnetic susceptibility and simulated spatial distributions were characterized by slightly lower standard deviations in comparison with the result of simulations based on the non-averaged, measured ones. At the same time, however, local variability of soil magnetic susceptibility was reproduced less. The accuracy of the calculations of point parameters and spatial distributions—based on the averaged values of soil magnetic susceptibility—were satisfactory, but when using geostatistical methods, it is recommended to use non-averaged magnetic susceptibility measurements.