A Comparative Analysis of Land Surface Retrieval Methods Using Landsat 7 and 8 Data to Study Urban Heat Island Effect in Madurai

In this paper the urban heat islands (UHI) effect in Madurai region has been identified by retrieving the land surface temperature (LST) distribution. The aim of this paper is to implement an algorithm to measure land surface temperature in Madurai region and it’s surrounding from Landsat thermal imagery. The Land surface temperature has been estimated by using Single Channel (SC) algorithm and Split-Window (SW) algorithm. These two algorithms can be implemented using Landsat 7 and Landsat 8 satellite data. The various methods adopted for retrieving the algorithm has been addressed in the present study. The results show that for Single Channel (SC) algorithm the error is approximately 1K-2K and in Split-Window (SW) algorithm the error is reduced because as SW algorithm uses two Thermal Infrared (TIR) bands for land surface temperature retrieval. In SW algorithm error is less than 1K. The results show that the LST generated using the SW algorithm was more reliable and accurate. From the final output it is revealed that barren lands, uncultivable land and urban areas experienced high LST and the areas with high vegetation cover and water body experiencing low LST. The results from both the algorithms show a variance of 5-6°C between urban areas, barren lands and vegetation covers thus indicating the presence of UHI in Madurai city.