Spatiotemporal assessment of environmental change in Kushtia, Bangladesh (1998–2023) using remote sensing-based environmental indices

Abstract

Bangladesh’s rapidly evolving environment, influenced by land use and climate dynamics, requires effective monitoring systems. This study examines spatiotemporal environmental changes in Kushtia from 1998 to 2023, utilizing NDVI, NDWI, NDMI, and LST derived from multi-spectral satellite imagery. A PCA-based weighted overlay method integrates these indices into a composite index to evaluate each year's overall environmental condition. Satellite images were processed in ENVI 5.3 and analyzed in ArcGIS 10.8, while PCA determined the weights of the indices. The weighted indices were overlaid in ArcGIS Pro 3.01, and image differencing was applied to detect environmental changes over time. The results represent a decline in the highest NDVI value from 0.72 in 1998 to 0.47 in 2023, reflecting an 18.23 % decrease in vegetated area coverage, while water availability (NDWI) and moisture content (NDMI) declined by 57.64 % and 47.01 %, respectively. Daulatpur experienced the highest reductions, with NDWI decreasing by 72.89 % and NDMI decreasing by 83.25 %, followed by Kumarkhali, which experienced a 51.16 % decline in NDWI. LST remained stable at 32.84 °C in 2023, with weak correlations to vegetation and water indices (R² < 0.13), indicating non-climatic drivers. PCA results suggest that PC1 (55.6 %) and PC2 (38.58 %) collectively account for 94.2 % of the environmental variation, with NDWI (weight: 0.312) and NDVI (weight: 0.298) as the dominant factors. PCA-weighted analysis reveals that degraded areas have shown improvement, while high-performing zones have experienced a decline. Environmental change was more significant from 1998 to 2010 than from 2010 to 2023, with notable degradation and recovery in Mirpur and Daulatpur. The findings emphasize the importance of adequate water governance, afforestation, and adaptive policies in addressing vegetation health and water stress.