Zuo Jian, Zhang Li, Xiao Jingfeng, Chen Bowei, Zhang Bo, Hu Yingwen, Mamun M M Abdullah Al, Wang Yang, Li Kaixin
International Research Center of Big Data for Sustainable Development Goals, Beijing, 100094, China.
Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, 100094, China.
Sci Data. 2025 Jan 22;12(1):129. doi: 10.1038/s41597-025-04430-0.
The coastline reflects coastal environmental processes and dynamic changes, serving as a fundamental parameter for coast. Although several global coastline datasets have been developed, they mainly focus on coastal morphology, the typology of coastlines are still lacking. We produced a Global CoastLine Dataset (GCL_FCS30) with a detailed classification system. The coastline extraction employed a combined algorithm incorporating the Modified Normalized Difference Water Index and an adaptive threshold segmentation method. The coastline classification was performed a hybrid transect classifier that integrates a random forest algorithm with stable training samples derived from multi-source geophysical data. The GCL_FCS30 offers significant advantages in capturing artificial coastlines, reflecting strong alignment with location validation data. The GCL_FCS30 classification was found to achieve an overall accuracy and Kappa coefficient over 85% and 0.75. Each coastline category accurately covered the majority of the area represented in third-party data and exhibited a high degree of spatial relevance. Therefore, the GCL_FCS30 is the first global coastline category dataset covering the high latitudes in a continuous and smooth line vector format.
海岸线反映了海岸环境过程和动态变化，是海岸的一个基本参数。尽管已经开发了几个全球海岸线数据集，但它们主要关注海岸地貌，仍然缺乏海岸线类型学。我们制作了一个具有详细分类系统的全球海岸线数据集（GCL_FCS30）。海岸线提取采用了一种结合了改进的归一化差异水体指数和自适应阈值分割方法的组合算法。海岸线分类采用了一种混合断面分类器，该分类器将随机森林算法与从多源地球物理数据中获取的稳定训练样本相结合。GCL_FCS30在捕捉人工海岸线方面具有显著优势，与位置验证数据高度吻合。研究发现，GCL_FCS30分类的总体准确率和kappa系数超过85%和0.75。每个海岸线类别准确覆盖了第三方数据中代表的大部分区域，并表现出高度的空间相关性。因此，GCL_FCS30是第一个以连续平滑线矢量格式覆盖高纬度地区的全球海岸线类别数据集。
