Assessment of the Dynamics of Land Cover at the Local Level to achieve Land Degradation Neutrality (for Desertified Lands of the Dry-Steppe Zavolzh’e)

Abstract

The land structure in dry-steppe Zavolzh’e is complicated and includes irrigated and rain-fed croplands and pastures in natural steppe areas and on abandoned plow lands, which results in significantly incorrect assessment of the Land Degradation Neutrality (LDN) based on standard approaches elaborated to analyze the land status at the global and regional levels. The problem of interpreting the results of the assessment of the dynamics of land cover as one of the main global indicators of LDN proposed by the UN Convention to Combat Desertification is analyzed in this work. A method based on the analysis of seasonal series of Landsat satellite images, using Kohonen self-organizing neural networks implemented in the Scanex Image Processor software package was developed and applied for the study area. It is shown that this approach makes significantly greater the possibilities of assessing the LDN as a result of the detailed analysis and delineation of a larger number of cartographic areas and when using the transition assessment matrix based on the extended classification of the land cover. The application of the Change Detection module of the SAGA GIS software package enables us not only to identify the nature of restoration and degradation changes, but also to compare the trends for land categories. The combination of the proposed approaches in the form of a functional algorithm may be recommended for monitoring and evaluation of the achieving LDN at the local level.