Satellite-Driven Land Surface Temperature (LST) Using Landsat 5, 7 (TM/ETM+ SLC) and Landsat 8 (OLI/TIRS) Data and Its Association with Built-Up and Green Cover Over Urban Delhi, India

Abstract

One of the most important impacts of urbanisation in Indian cities is the conversion of green belts and agriculture land into the built-up area in the periphery (Chadchan and Shankar, Int J Sustainable Built Environ 1:36–49, 2012; Pandey and Seto, J Environ Manag 148:53–66, 2015). With these physical changes, i.e. decrease in green cover and increase in built-up, the land surface temperature (LST) is bound to increase. The green area is a basic need of any city because it is a must for a healthy life and also maintains the aesthetic and ecological beauty in the urban areas (Low et al. 2007). The present study aims to analyse the association between built-up, green cover and land surface temperature for which district-level analysis of the normalised differential built-up index (NDBI), normalised differential vegetation index (NDVI) and land surface temperature (LST) has been done over the urban area of Delhi. In this study, Landsat 7 (ETM+ SLC) for 2003, Landsat 5 (TM) for 2010, and Landsat 8 (OLI/TIRS) for 2017 have been used together with Survey of India (SOI) toposheet of Delhi at 1:25,000. Indices like NDBI, NDVI and LST are calculated for 2003, 2010 and 2017 using the spectral radiance model (SRM), the mono-window algorithm (MWA) and the split window algorithm (SWA). Thereafter, district-wise NDBI, NDVI and LST are extracted by using clip tools in ArcGIS 10.5 software. To analyse the relationship between built-up and green cover with LST, correlation is done in SPSS software and a scatter diagram is made to assess the correlation amongst the variables. The further surface temperature profile is created to know which part of the Delhi has the highest and lowest temperatures on a particular surface. The study shows that NDVI and LST are negatively correlated with each other as vegetation has a cooling effect on the land surface temperature whereas NDBI and LST are positively correlated with each other. The studies show a change in the distribution of vegetation cover and gradually increase in the built-up land which results in the increase in land surface temperature to about 3.31 °C in the last 14 years. The result shows that MWA give the most accurate result in this study since RMSE of MWA is the lowest (0.71 °C) amongst the three algorithms used in the study. Temporal analysis of land surface temperature by all the three algorithms shows the increase in land surface temperature of Delhi between 2003 and 2017.