Estimation of land surface temperature in agricultural lands using Sentinel 2 images: A case study for sunflower fields

Abstract

AbstractLand surface temperature (LST) is usually calculated based on the thermal bands of satellite images. The present study aimed to estimate LST with reasonable accuracy for on‐time agricultural management by integrating Sentinel 2 with Landsat 8 or MODIS images. For this purpose, LST was first calculated based on Landsat 8 and MODIS images in the Satar Plain located in the western of Kermanshah province, Iran. Then, three methods for LST estimation using Sentinel 2 images were presented and evaluated. In the first method, a relationship was established between LST and elevation and the normalized difference vegetation index. In the second method, LST was determined based on the relationship between Sentinel 2 spectral bands and Landsat‐based LST in sunflower fields and agricultural lands. In the third method, the LST product of the MODIS was downscaled using spectral bands of Sentinel 2. Finally, the results of the three methods were compared with Landsat‐based LST. The results showed that the second method (the relationship between LST and spectral bands of Sentinel 2) with coefficient of determination and root mean square error values of 0.9 and 1.48°C in sunflower fields and 0.625 and 3.85°C in agricultural lands, respectively, has better accuracy than the other two methods.