Remote Sensing Analysis of Urban Heat Island Dynamics Using NDVI, NDBI, and LST in Constantine, Algeria (2005–2024)

The intensification of the Urban Heat Island (UHI) effect represents a major environmental challenge in rapidly expanding semi-arid cities, where vegetation loss and the growth of impervious surfaces significantly alter the urban thermal environment. This study analyses the spatiotemporal evolution of vegetation cover, built-up surface characteristics, and land surface temperature (LST) in Constantine, Algeria, over the period 2005–2024, using multi-temporal Landsat imagery integrated within a Geographic Information System (GIS) framework. Biophysical indices, namely the Normalized Difference Vegetation Index (NDVI) and the Normalized Difference Built-up Index (NDBI), were derived to characterise land cover dynamics and examined in relation to LST retrieved from thermal bands through radiometric calibration and emissivity-based correction methods. The results indicate pronounced land surface transformations, marked by a substantial decline in vegetated areas and a significant increase in built-up–related spectral responses over the study period. Regression analyses reveal a consistent negative relationship between NDVI and LST, confirming the cooling role of vegetation, with the strongest association observed in 2015 (R² ≈ 0.26), followed by a notable weakening in 2024 (R² ≈ 0.08). Conversely, NDBI exhibits a positive relationship with LST, highlighting the warming influence of built-up surface intensity, with regression slopes increasing from approximately 9.8°C in 2005 to 29°C in 2024, despite moderate explanatory power (R² = 0.29–0.42). Spatial analysis of UHI patterns further demonstrates the growing dominance of high-LST zones (24–30°C) in 2024 compared with cooler thermal conditions in 2005.