Comparison of four algorithms to retrieve land surface temperature using Landsat 8 satellite

Abstract

Land Surface Temperature (LST) plays an important role in energy exchange between the land surface and the atmosphere. LST is a key variable in many applications, such as land surface modeling. Many satellite-based algorithms have been proposed to retrieve LST, such as Split-Window (SW), dual-angle, and single-channel algorithms. In this study, four satellite-based LST retrieval algorithms, including two SW algorithms (Juan C. Jimenez-Munoz and Offer Rozenstein SW algorithms) and two mono-window algorithms (Juan C. Jimenez-Munoz and Qin Zhihao mono-window algorithms), were compared with Landsat-8 satellite data over the region around Wuxi City. The accuracy of the four algorithms was evaluated against the ground measurements from 16 floating stations over Lake Tai. The results showed that the performance of the two SW algorithms, which have an average error of 0.7 K, was better than that of the SW algorithms, which have an average error of 1.3-1.4 K, when compared with ground measurements. The sensitivity analysis of these algorithms showed that the Juan C. Jimenez-Munoz SW algorithm was the least sensitive to key input parameters (emissivity and water vapor), whereas the Offer Rozenstein SW algorithm and the Qin Zhihao mono-window algorithm showed high sensitivity to input parameters. The limitations of these four LST retrieving algorithms were also discussed. (Less)