Quan, Yu ， Zheng, Yufeng ， Yang, Yuanyuan ， Zhang, Xuanze ， Hong, Er
2026-02-01 REMOTE SENSING LETTERS 2026 17(卷), 2(期), (198-209页)
This study proposed a novel remote sensing approach for desertification assessment, the Normalized Desertification Difference Index (NDDI), based on the band combination of the Wide Field of View (WFV) sensor aboard the GF-6 satellite, using the Kubuqi Desert as the study area. Validation results indicate that the desertification classification derived from the NDDI achieves an average accuracy of 0.81 and a Kappa coefficient of 0.76. The analysis of land use transitions related to desertification in the Kubuqi Desert from 2019 to 2023 revealed that the overall desertification level remained stable, with interannual variations largely corresponding to changes in climatic conditions. Furthermore, NDDI demonstrated improved performance in distinguishing non-vegetative land features such as water bodies and photovoltaic power stations, leveraging the multi-spectral capabilities of GF-6 data. By reducing dependency on mid-infrared bands, NDDI simplifies the computational process and enhances its applicability for large-scale monitoring,Deserts in different regions all exhibit good classification performance.