Assessing LULC changes and LST through NDVI and NDBI spatial indicators: a case of Bengaluru, India

Increased urbanization and growing needs in urban areas resulted in causing extensive changes to the land use land cover (LULC) of urban areas. It was observed from many previous studies that the LULC changes in urban areas result in short and long-term repercussions, including increased land surface temperature (LST) and urban heat island. However, the established relationship between the LULC changes and LST needs a deeper understanding. This study examines and explores the relationship between the LULC changes in Bengaluru, India, and the corresponding impact on the LST. LULC classification done using Landsat images revealed that over 50% of vegetation had been lost between 2003 and 2021, while the built-up area had nearly doubled during the same period. Subsequently, the city experienced an increase in the minimum temperature from 16 °C in 2003 to 21 °C in 2021, whereas the mean LST had increased from 26 °C in 2003 to 29 °C in 2021. Spatial aggregation of the LST is examined using Moran's I spatial autocorrelation, and the results suggested moderately clustered LST values in the study area. Spatial indicators like normalized differentiated vegetative index (NDVI) and normalized differentiated built index (NDBI) are calculated and compared against the LST using scatter plots. The relationship is identified using simple linear regression (SLR) and geographically weighted regression (GWR), considering the non-spatial and spatial attributes inherently present within the data. Results from GWR suggest a more defined relationship between LULC and LST, NDVI, and NDBI than SLR. While NDVI exhibited a negative relationship with LST, NDBI displayed a positive relationship with LST. The results from this study highlight the need for planned growth of urban areas with sustained growth in vegetated areas through urban planning and design strategies, which can help contain the increase in LST in future scenarios.