Multispectral images of agricultural landscapes as a tool for detecting heterogeneity of soil fertility

The aim of the work was to identify the causes of soil fertility heterogeneity and assess the state of spring wheat crops in the agrolandscape of the steppe zone of the Omsk region using the analysis of multispectral images. The studies were carried out in 2021 in the Novovarshavsky district of the Omsk region on the production array of the field. The object of the study was the soil cover, represented by meadow-chernozem soil, and the sowing of spring wheat of the OmGAU-100 variety, as an indicator of soil fertility. To study the soil and vegetation cover, we used multispectral images obtained by surveying the earth's surface with the Sentinel-2 L2A satellite. Satellite data were processed in the ENVI and QGIS software packages using the Sentinel Нab, Land Viewer, and Vega PRO Internet platforms of the Space Research Institute of the Russian Academy of Sciences. In four areas, differing in the density of wheat standing and the spectral characteristics of the surface, wells were laid to a depth of 100 cm and soil samples were taken every 10 cm. In them, the reaction of the environment, the content of humus and easily soluble salts were determined. An analysis of the values of the NDVI, EVI, NDMI, Agriculture indices indicated the spatial variability of wheat biomass. It was established that the reasons for the heterogeneity of crops were microrelief, uneven distribution of moisture over the surface and soil properties. In the control plot, the meadow-chernozem soil was characterized by the presence of readily soluble salts from a depth of 50 cm and was classified as medium solonchakous slightly saline. In areas with sparse vegetation, the soil was highly saline solonchak. The use of vegetation indices to identify the heterogeneity of soil fertility in the early phases of wheat development was complicated by the influence of the microrelief, waterlogging of microdepressions, and the development of weeds. The relationship between the values of vegetation indices and the value of the above-ground mass of wheat, as an indicator of fertility, was more clearly manifested by the end of its vegetation, starting from the earing phase. The results of the research indicate the possibility of using multispectral images to identify the heterogeneity of soil fertility in the steppe agricultural landscapes of the Omsk Irtysh region.