SOIL EROSION INDICATION BY MAGNETIC METHODS IN KHARKIV REGION

The new land regulation system in Ukraine, as well as European integration strategy, require to update the methodological approaches of farming land monitoring, increasing the accuracy of research and reducing the cost of the studies. Among such effective tools is soil magnetism application to study the erosion processes. The introduction of magnetic methods at the technology level is associated with methodological problems: density of the sampling network, the reliability of the statistics. The object of the research was soil erosion processes at the territory of Rusko-Tishkivska land council (Kharkiv region). Field, statistical, laboratory, and cartographic methods were used. The parameters of magnetic susceptibility (MS) and humus content (G) were obtained for soil samples from three genetic horizons of the sections (the depth is 0–10, 10–20 and 25–40 cm). A sequence of indicators was formed, the numerical values of MS 1, MS 2, MS 3, and G 1, G 2, G 3 were systematized. The humus content index is characterized by a greater degree of variation than the magnetic susceptibility. The sample data has a more dense distribution than the humus content in all horizons of the soil. Comparison of the studied indicators with respect to the magnitude of the relative sampling error established that more reliable results are observed for magnetic susceptibility. The maximum sampling errors for all horizons of the soil are lower than for the humus content. The article widely considers the statistical tests to check the homogeneity of the collection of selected field data. According to the Cramer–Welch T criterion, the homogeneity of the mathematical expectation is satisfied for all samples except MS 3. The correlation analysis of the samples showed that the small coefficients of rank correlation of Spearman were found in the pair of G 1 and MC 1, and in the pair G 1 and MC 3. The corellation is 0,697 and 0,641 respectively. Magnetic susceptibility and humus content show a high degree of correlation, which can be traced for different genetic horizons.