From Climate to Cloud: Advancing Fog Detection Through Satellite Imagery

The broad spatiotemporal coverage provided by satellite remote sensing is fundamental for monitoring fog events, a phenomenon that impacts transportation, agriculture, and ecosystem functioning. Despite advances in remote sensing technology, significant knowledge gaps remain regarding the application of these techniques to fog detection, especially over terrestrial ecosystems. This scoping review synthesizes the trends in methods used for fog detection by analyzing 38 papers retrieved from Scopus and Web of Science. Only studies that utilized satellite imagery to analyze the spatiotemporal dynamics of fog were included. Articles that employed non-satellite methodologies or focused on processes other than the detection, formation, or identification of fog events were excluded. In addition to a term co-occurrence analysis of abstracts using VOSviewer, this study examines key parameters of the detection methods—including sensor type, spectral bands, temporal resolution, and algorithmic approaches (e.g., threshold methods and deep learning techniques)—to evaluate their evolution and current limitations. Our results reveal that while approximately 53% of studies rely on geostationary satellite data (95% CI: 36.7–68.5%), favored for their high temporal resolution, the remaining 47% employ polar-orbiting sensors (95% CI: 31.5–63.2%) that offer superior spatial resolution. Notably, most research has concentrated on maritime fog detection, with few studies extending these techniques to complex terrestrial environments. The review highlights critical gaps in current approaches and proposes an integrated framework that combines traditional brightness temperature difference methods with emerging machine learning techniques, which could advance fog detection in diverse settings.