ARIMA modeling for forecasting land surface temperature and determination of urban heat island using remote sensing techniques for Chennai city, India

Mounting human population and fast urban expansion has driven the ecosystem degradation within the past 3 decades by reducing permeable cultivable land surface for the construction and thus increasing land surface temperature (LST) and creating urban heat islands (UHI). It is well known that temporal assessment of land surface temperature and UHI are inevitable for city planning and ecosystem monitoring. Landsat-derived LST is widely used for urban temperature studies including the study of the urban heat island (UHI). In this study, remote sensing (RS) techniques have been used for the estimation and forecasting of LST and identification of UHI in one of the fast-growing cities of Tamil Nadu state, India, using Autoregressive Integrated Moving Average (ARIMA) model. Satellite data between the period 2008 and 2018 was used in the model study. Analysis of the images and model results show that there is a progressive increasing trend of LST in built-up areas. LST values obtained from model study exhibited a negative relationship with land use and land cover (LULC) for Chennai city and surrounding area. LST maps developed from the model study depicted growing UHI hotspots in the southeastern and western parts of the city where the development of the city is fast. The present study would help in forecasting the LST of a city and in identifying UHI hotspots for proper urban planning.