Integrating remote sensing indices and multivariate statistics for land degradation assessment in the semi-arid Chambal region

Abstract This study presents a spatiotemporal assessment of land degradation (1992, 2000, 2008, 2016, and 2024) in the Chambal region, one of the world’s most geomorphologically complex landscapes. Pearson correlation analysis and Principal Component Analysis (PCA) were employed to develop a data-driven Land Degradation Index. Eight remote sensing indices (BSI, NDVI, NDMI, MNDWI, NDBI, SAVI, SSI, and TGSI) were initially computed. Six non-redundant remote sensing indices (BSI, SAVI, NDMI, MNDWI, SSI, and TGSI) were integrated using PCA to develop the Land Degradation Index. The first principal component explained 81% of the total variance in 1992 and 87% in 2024, indicating a dominant land-degradation gradient. The PCA-derived Land Degradation Index was classified into five risk categories and analysed using zonal statistics across major land-use–land-cover classes. Results show that “High” and “Very High” degradation zones declined from 29.37% (1992) to 21.75% (2024), while “Low” and “Very Low” categories expanded from 41.66% to 52.65%. Ravines exhibited a 25.85% reduction in high-risk exposure, whereas forest areas showed an 11.27% increase in high-risk classes, indicating emerging ecological vulnerability. This framework supports SDG 15 by enabling effective monitoring of land degradation and promoting informed land-management interventions.