MONITORING SOIL COMPACTION AND DEGRADATION WITH SATELLITE DATA: A REVIEW OF CURRENT APPROACHES AND PERSPECTIVES

Abstract

This review article presents a comprehensive analysis of current remote sensing approaches for assessing soil compaction and surface physical properties, with a particular focus on dark-grey podzolized soils typical of the Western Forest-Steppe of Ukraine. The study emphasizes the use of multispectral (Sentinel-2), radar (Sentinel-1), and thermal (MODIS, Landsat) data to identify indirect indicators of soil structural degradation. Vegetation and bare-soil indices such as NDVI, Brightness Index, NDWI, and SAR backscatter are evaluated as proxies for compaction-related stress on crops and soil hydrology. Additionally, the article highlights recent advances in Digital Soil Mapping (DSM) and machine learning applications (e.g., Random Forest, Cubist, CNN), which enable accurate spatial modeling of compaction risks and topsoil texture variations. Case studies from Ukraine and Central Europe are reviewed, including the application of the Two-Dimensional Bare Soil Separation (TDBSS) algorithm, which combines SAVI and NDRGI indices for precise detection of bare soil from Sentinel-2 images. Field-validated results demonstrate that integration of remotely sensed indices with in-situ penetrometer measurements can achieve mapping accuracies exceeding 80%. Limitations of current methods–such as atmospheric interference, shallow sensor penetration, and the need for local calibration—are discussed in detail. The authors conclude that hybrid systems combining satellite imagery, field observations, and AI-based soil models represent a promising pathway for implementing nationalscale soil monitoring frameworks. Future research directions include the development of open-access soil data platforms and operational systems for early detection of physical soil degradation under climate change and intensive farming.