APPLICATION OF VEGETATION INDECES FOR DOGITAL SOIL MAPPING BASED ON SENTINEL-2 SPACE IMAGES

Abstract

At present, the works devoted to the creation of digital soil maps using geographic information systems (GIS) and remote sensing (RS) data are relevant. In the work the analysis of vegetation indices (VI) for soil mapping was carried out, the maps of vegetation indices were created: Normalized Difference Vegetation Index (NDVI), Optimized Soil-Adjusted Vegetation Index (OSAVI), Soil-Adjusted Vegetation Index (SAVI), Tranformed Soil-Adjusted Vegetation Index (TSAVI), Enhanced Vegetation Index2 (EVI2) for the territory of ZAO Mirny farm of Kochenevsky District using Sentinel-2 A satellite image (May 10, 2021). ). As a result it was revealed that the vegetation indices OSAVI and EVI2 allow to establish spatial boundaries between the main types of soils of automorphic, half-hydromorphic and hydromorphic moisture regimes. Background. Multispectral space and aerial photos for thematic mapping of soil resources are of greater practical application. Sentinel-2 A space image with good spatial and spectrozonal resolution (10 m, 20 m and 60 m) and territorial coverage (290 km) was used in this article. This made it possible to calculate and analyze various vegetation indices for the purposes of digital soil mapping. Purpose. аnalysis of vegetation indices for digital soil mapping based on Sentinel-2 A images. Materials and research methods. The research was carried out on the territory of CJSC Mirniy, Kochenevsky District, Novosibirsk Region. The methods of digital processing of space images, mapping and geoinformation analysis with the use of Sentinel-2 A satellite image (May 10, 2021) were used. The method of equal intervals was used for comparative analysis of images. This allowed using GIS ArcGIS to make thematic maps of images with the allocation of gradations: very low, low, average, above average, high value. Results. Field soil surveys were carried out on the territory of CJSC Mirny farm in Kochenevsky district of Novosibirsk Region. Using SAGA geoinformation system the space image was atmospherically corrected and spatially referenced, NDVI, OSAVI, TSAVI, EVI2 raster maps were compiled. Geoinformation analysis of the large-scale 1:1000 soil map and raster EVI maps revealed that OSAVI allows to establish spatial boundaries between the main types of soils of automorphic, half-hydromorphic and hydromorphic moisture regimes. Very low values of VI are typical for the soils of hydromorphic humidification regime, formed near small lakes, along the banks of the Sharikh river. Very low values of UI have objects of hydrography, marsh peaty, meadow-marsh humus soils, marshy and peated marshes, marsh solonchaks, marsh solonchaks, formed in lowered areas of relief with depth of groundwater occurrence less than 0.5 m. Ploughed ordinary chernozems, deposited in the upper and middle part of the gentle slope, have low values of WP. These are soils of automorphous moisture regime with depth of groundwater occurrence more than 6 m. Average and above average values are characteristic of gray forest saltwort soils under woody vegetation, as well as meadow-chernozem soils under meadow vegetation in the lower part of the gentle slope with groundwater occurrence depth from 3 to 4 m. High WI values were obtained for meadow soils with dense grass cover. Wetting conditions of soils, location in the relief, and vegetation type significantly influence VI values. The obtained VV values can be used at the stage of training data preparation in the form of reference classes for basic soil types required for automatic image recognition.

Keywords

• digital soil mapping
• geographic (geoinformation) information systems
• space images
• vegetation indices
• agriculture