RoLaSTIM: a novel method for lakeshore type identification and utilization rate assessment

Lakeshores are both vital components of lake ecosystems and non-renewable resources that support human development. Understanding their utilization is essential for assessing human–lake interactions, yet accurate lakeshore type identification remains challenging because of the complex water–land interfaces and diverse natural and anthropogenic contexts of lakes. To address this, we developed RoLaSTIM (Robust Lake Shore Type Identification Method), an operational approach for classifying lakeshores as natural or artificial based on spatial analysis and land cover conditions. RoLaSTIM was applied to 12 typical global lakes, and their annual Shore Utilization Rates (SURs) were analysed from 2000 to 2022. RoLaSTIM achieved high classification accuracy (area-adjusted overall accuracy: 98.93 ± 0.02%, weighted F1 score: 98.94%), benefiting from a two-level buffer strategy and corresponding physical rules that minimize misclassification. Key factors influencing accuracy included the lake SUR, artificial lakeshore length, land cover complexity, and satellite image resolution. Moreover, we identified four distinct SUR evolution trends: increasing, stable, inverted (‘∩’-shaped), and fluctuating (‘M’-shaped). RoLaSTIM is suitable for robust and consistent lakeshore classification at large scales, supporting efficient SUR assessment and long-term monitoring. This method fills a critical gap in global lakeshore research and offers valuable insights for lakeshore management, conservation, planning, and sustainable development.