ABSTRACT

The remote sensing technique often analyzes the thermal characteristics of any area. The present study focuses on the estimation of land surface temperature (LST) of Hyderabad city as a whole and for the most heated zones of the city and its relationship with modified normalized difference water index (MNDWI) and normalized difference built-up index (NDBI). Landsat-8 OLI/TIRS data of 1 February 2016 was used for the whole study. Several urban heat islands (UHIs) and urban hot spots (UHSs) were identified as the most heated zones throughout the city. The UHIs were developed along the outskirts of the city except for the central and eastern part, while the UHSs mainly existed along the western and south-eastern periphery within the UHIs. The result shows that LST has a strong negative correlation (−0.73) with MNDWI and a strong positive correlation (0.63) with NDBI for the whole Hyderabad city. But, these relationships become much weaker for UHIs. Even an inverse correlation was found for UHSs. It indicates that landscape heterogeneity reduces the reliability of the relationship between LST with MNDWI and NDBI.