Interdisciplinary

Assessment of agricultural region dynamics using object contour properties from Sentinel-2 images

AI Insight

This study introduces a Contour Density Indicator (CDI) derived from Sentinel-2 satellite imagery to monitor structural changes in agricultural landscapes in southern Ukraine between 2020 and 2024, a period encompassing active military operations. The CDI measures the total extent and contrast of linear landscape features such as field boundaries, roads, and irrigation infrastructure. Results show a systematic decline in CDI modal values from 2022 to 2023, with reductions of up to 55% in the 30-km battle-line zone relative to 2020 baseline values, while Ukrainian-controlled territories showed partial recovery by 2024.


The methodology provides a scalable, remote sensing-based tool for detecting agricultural degradation caused by conflict, depopulation, or reduced cultivation intensity, with direct applications for land management decisions in both wartime and postwar reconstruction contexts.


by Serhii Nikulin, Kateryna Sergieieva, Oleksandr Kovrov, Anton Kozhevnykov, Olga Korobko

Economic activity in agricultural regions is influenced by various natural and anthropogenic factors. Their impact on spatial and temporal changes can be assessed by studying satellite images acquired at different times. The methodology for detecting changes is continually improving with the use of new image processing techniques. Changes resulting from destructive events can be used to assess their effectiveness. Agriculture in southern Ukraine has been catastrophically affected by military actions, which have led to the degradation of the agricultural landscape structure. One of the manifestations of this process is the destruction and fragmentation of linear contours of agricultural landscapes such as cropland boundaries, field roads, irrigation infrastructure, forest stands, etc. The study assesses the potential of using structural features of object contours, such as their total extent and contrast, in satellite image bands to monitor alterations in the condition of agricultural landscapes. A methodology is proposed to analyze the dynamics of cultivated land development or degradation by analyzing a developed Contour Density Indicator (CDI). The CDI was applied to assess changes in the configuration of an agricultural region in southern Ukraine from 2020 to 2024 using Sentinel-2 images. This period includes 2.5 years of military operations of varying intensity. Analysis of Sentinel-2 images demonstrated a systematic decline in CDI modal values from 2022 to 2023, with the strongest decrease observed within the 30-km-wide battle-line zone, where CDI mode dropped by up to 55% relative to 2020. In contrast, Ukrainian-controlled areas showed a partial recovery of CDI values in 2024, while the buffer zone continued to exhibit persistent loss of contour density, indicating sustained degradation of the agricultural landscape. The results enabled the identification of zones with signs of agricultural degradation resulting from depopulation caused by military activity and the subsequent decline in cultivation intensity. The methodology can be used for operational monitoring of agricultural regions and for making informed management decisions regarding land use in both wartime and peacetime.

Source: Assessment of agricultural region dynamics using object contour properties from Sentinel-2 images