PyLST: a remote sensing application for retrieving land surface temperature (LST) from Landsat data

• Published in: Environmental earth sciences Vol. 83; no. 12; p. 373
• Main Authors: Parvar, Zahra, Salmanmahiny, Abdolrassoul
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.06.2024; Springer Nature B.V
• Subjects: Algorithms; Biogeosciences; computer software; Earth and Environmental Science; Earth Sciences; Environmental Science and Engineering; equations; Geochemistry; Geology; Hydrology/Water Resources; Land cover; Land surface temperature; Land use; land use and land cover maps; Landsat; Landsat satellites; Original Article; Radiative transfer; Remote sensing; Root-mean-square errors; Satellite imagery; Surface temperature; Terrestrial Pollution; user interface; Radiative transfer equation (RTE) method; Single Channel Algorithm (SCA); Split Window Algorithm (SWA); Mono Window Algorithm (MWA); Landsat
• ISSN: 1866-6280; 1866-6299
• DOI: 10.1007/s12665-024-11644-9

Understanding land surface temperature (LST) dynamics is crucial for assessing the impacts of changes in land use and land cover (LULC) through remote sensing. However, the complexity and time-intensive nature of existing LST extraction algorithms pose challenges for many users. In response, this study introduces an open-access Python-based user interface tailored for extracting LST from Landsat images (Landsat 5, 7, 8 and 9) using multiple algorithms, including the Mono-Window Algorithm (MWA), radiative transfer equation (RTE) method, Single Channel Algorithm (SCA), and Split Window Algorithm (SWA). The primary problem addressed by this research is the accessibility and usability of LST extraction methods for researchers and practitioners. By developing a user-friendly interface that facilitates algorithm comparison and selection, the software aims to streamline the process of LST retrieval and analysis. To evaluate the efficacy of the implemented algorithms, 24 Landsat images, spanning different seasons (six images per Landsat mission), were utilized. Results indicate that while all methods yielded acceptable outcomes, the RTE method demonstrated slightly superior accuracy for Landsat 5 and 7, with lower root mean square error (RMSE) values. Conversely, for Landsat 8 and 9, the SWA exhibited the best performance, achieving an RMSE of 2.1 °C.