Correlation Between Built-Up and Land Surface Temperature Using Sentinel-2 and Landsat-8 Images through Semi-Supervised Deep Learning Model for Efficient Land Surface Monitoring
Satellite images are a powerful tool of earth observation that are providing a different view of earth surface monitoring. Currently, more than 200 satellites are available in orbit, and some of them are operational which are producing different kinds of images including optical, SAR, and thermal images with different spatial, spectral, and temporal resolutions. Many researchers have developed various frameworks, models, and systems that are using machine learning and neural network techniques for Earth's surface monitoring such as agricultural monitoring, urban area monitoring, flood monitoring, soil monitoring, etc. Satellite image contains heterogeneous characteristics of big data so machine learning and neural network techniques are less capable to produce an efficient result. Therefore, to overcome this problem, a semi-supervised deep learning model is proposed that is capable to classify the images into four land cover classes such as urban, barren lands, water bodies, and agricultural fields, and also capable to measure the land surface temperature (LST). In this paper, Sentinel-2 and Landsat-8 optical satellite images of the summer season from 2017 to 2021 have been used to estimate the urban area and land surface temperature of Lucknow city situated in Uttar Pradesh, India. The Sentinel-2 image is a high-resolution satellite image so it has been used in urban monitoring and the Landsat-8 image have two Thermal-Infrared bands that are used to measure the temperature of the land surface. Further, estimated urban area and LST have been used to find out the correlation between them, and are also used to conduct the trend analysis for the 10 years (2022 to 2031) using the linear trend analysis technique. The LST has been increasing in the taken period, according to trend analyses. It is also observed that LST is heavily influenced by the growth and development of built-up areas as well as the alteration of the thermal characteristics of the urban core and barren lands. In the future, this type of study may be very useful to construct the plan urban areas, which may be very helpful to control the temperature rise.