Can Satellite-Derived Beach Images Resolve the Responses to Human Activities?

Satellite imagery is widely used to study beach changes. Studies based on satellite-derived results predict that by 2100, half of the world's beaches will disappear. Although previous studies have been productive and alarming, the reliability of these findings has not been validated. This study presents an initial comprehensive assessment of the reliability of satellite-derived beach shoreline changes and concludes that despite the use of state-of-the-art subpixel shoreline identification technologies and water-level correction methods based on beach profile shapes, the errors in satellite-derived shoreline positions remain significant. Thus, adequate satellite imagery is crucial for enhancing the reliability of satellite-derived beach changes. Furthermore, the historical water levels when the Landsat and Sentinel-2 images were taken exhibited significant systematic deviations and both long-term and short-term changes. Consequently, the beach changes obtained through the direct statistics obtained from satellite imagery based on big-data platforms such as the Google Earth Engine (GEE) are susceptible to systematic biases. The study also concluded that satellite-derived beach changes based on adequate satellite imagery have distinct, significant, and diverse responses to human activities. Given the current context of intensified human activities and notable inaccuracies in satellite-derived shoreline positions, fully utilizing satellite imagery to study beach changes can help establish stronger connections between beach changes and human activities. This contributes to a better understanding of beach evolution mechanisms and the need for beach restoration and protection, with multiple scientific, engineering, and social implications. Beaches around the globe are currently facing significant threats due to the combined influence of rapid sea-level rise and intensified human activities. Researchers have used satellite images to study beach changes, and the results are alarming. Some researchers even predict that by 2100, half of the world's beaches will have disappeared. However, the existing research has mainly evaluated the precision of satellite-derived shoreline position data rather than satellite-derived beach changes. Moreover, the identified beach changes are not clear enough to determine the links between beach changes and human activities. Based on all available Landsat and Sentinel-2 images, this study presents the first comprehensive assessment of the reliability of satellite-derived beach shoreline changes and demonstrates the distinct, remarkable, and diverse responses of a typical beach to human activities. The study revealed that the water levels when the satellite images were taken exhibited systematic deviations and both long-term and short-term changes. Consequently, water-level corrections to the shoreline positions that were obtained through direct statistical analyses of satellite imagery are essential for reliably studying both global and local beach changes. Finally, adequate satellite imagery is recommended for identifying high-precision beach changes. This paper presents an initial comprehensive assessment of the reliability of satellite-derived beach shoreline changes Beach changes obtained through the direct statistics obtained from satellite imagery on big-data platforms are prone to systematic biases Adequate satellite imagery is crucial for enhancing the reliability and revealing the impact of human activities