Reconstructing Land Surface Temperature for Cloud-Covered Regions: A Review of Methods

Understanding surface thermal conditions is essential for studying ecosystem responses, hydrological processes, and climate-driven environmental change. Land Surface Temperature (LST) is a key parameter for studying a wide range of environmental and climatic processes. Although remote sensing technologies enable the acquisition of LST data on a global scale, satellite observations are frequently obscured by cloud cover, resulting in spatial and temporal data gaps that limit their usability. This work provides a comprehensive review of existing methods developed to reconstruct gap-free LST data. First, state-of-the-art methods are systematically reviewed and classified into rule-based and data driven methods. For each method, the underlying principles are explained, alongside details on implementation, validation, and the selected case study, providing critical insights into their performance and generalizability. Second, the growing potential of advanced methods—particularly deep learning frameworks—is explored, highlighting their adaptability for the task of LST gap-filling. The findings indicate that despite the development of sophisticated models, significant challenges remain, especially regarding the generalizability of the methods in varying climatic and urban environments. In addition, inconsistencies are identified in the evaluation metrics, limiting the direct comparability of the methods. This review can help guide future research efforts by highlighting current limitations and identifying promising directions for the development of more robust and adaptable LST gap-filling approaches.