Answers are grounded in this paper and cite their sources.
As a major ecological and environmental issue facing arid and semi-arid regions, the occurrence and evolution of desertification profoundly impact regional ecological security and land-use patterns. Conducting desertification monitoring based on remote sensing data is of great significance for providing objective ecological assessments and a scientific basis for ecological governance. To address the limitations of existing studies, which largely rely on a single spectral dataset, this study combines Sentinel-1 and Sentinel-2 remote sensing data to develop the three-dimensional desertification monitoring index (TDMI) that integrates vegetation condition (modified soil-adjusted vegetation index, MSAVI), surface reflectance (albedo), and surface texture (vertical-transmit–vertical-receive backscatter coefficient, VV) by using the spatial distance model (SDM). Subsequently, a gaussian mixture model (GMM) was employed to conduct monitoring of desertification in the Three-North Shelterbelt Forest Project area for the years 2020, 2022, and 2024. Combining field survey data with accuracy validation using high-resolution historical imagery from Google Earth, the results indicate that the TDMI demonstrated good performance across all years, with an average Overall Accuracy (OA) above 83%. This represents an improvement of 6.73 and 11.10 percentage points compared with the two-dimensional desertification index (DMI) and Fractional Vegetation Cover (FVC), respectively. In terms of spatial distribution characteristics, TDMI demonstrates stronger identification capabilities for the wind-sand desertification types in the study area. Generally, this study demonstrates that the synergistic application of multi-source remote sensing data can improve the accuracy and applicability of desertification monitoring.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: