Modelling the spatial distribution of drought effects on land degradation in the Blue Nile Region, Sudan

Land degradation (LD) has a significant impact on Sudan's economy, society, and environment, driven by both climatic variations and human activities. Quantifying this degradation and its link to land use change (LUC) in Sudan's semiarid regions is crucial but often overlooked. The study provides precise spatiotemporal data to evaluate and model the impact of drought on land degradation in Sudan's Blue Nile Region. Four cloud-free Landsat images (1993, 2003, 2013, and 2023) were used and downloaded from the United States Geological Survey (USGS) repository. The study integrated vegetation indices (VIs) (NDVI, SARVI, SAVI, VHI) and soil indices (SIs) (BSI, TGSI) to assess LD. Change detection matrices estimated spatiotemporal land use and degradation shifts. Correlation and modeling (Kriging) determined relationships and causes of LD. Model validation was through the use of the coefficient of determination (R2), , Kappa coefficient, and principal component analysis. SIs and VIs effectively detected LD. Very severe and severe degradation increased by 15.8% and 23.3%. Conversely, non and non-light-degraded areas decreased to 27% and 16.2%, respectively. Moderately degraded areas increased by 2.5% and 1.7%. The study revealed positive correlations between (TGSI, NDVI, VHI), with R2 of 0.99, 0.98, respectively, and negative correlations between (BSI, NDVI, SAVI) with R2 of -0.92 and -0.89. The Kriging model showed reasonable performance with an R2 of 0.52, Kappa coefficients of 72%, and PC1 and PC2 capturing 78% of the variance. This work provides a robust, low-cost approach for predicting soil degradation via vegetation, especially valuable for semiarid regions. The integrated methodology and validation offer a reliable tool for environmental monitoring and management.