2017年中国滨海红树林生长区海岸带土地覆被样本数据集

本数据集为2017年中国滨海红树林生长区海岸带土地覆被样本数据，可为生态保育、湿地保护、区域可持续发展提供数据支撑。同时适用于辅助野外调查、遥感分类等情景。
b. 建设目的
样本能够驱动分类算法，是准确分类的先决条件。海岸地区位于陆地与海洋的过渡区域，土地覆被多样，需要更多类别的样本来刻画。但当前样本的共享稀少，使得相似研究独立重复进行费时费力的采样工作。因此，将红树林分类提取研究中获取的10个类别共计16444个样本公开。本样本集覆盖了沿海8个省区（海南、广西、广东、福建、浙江、台湾、澳门、香港）的低海拔海岸部分。它不仅能够用于支撑海岸土地覆被分类的模型训练，还能用于验证各分类结果。除了验证红树林分布图，还能用于潮滩提取、盐沼植被识别、农田制图等多种应用。 可广泛服务于农业、生态等方面和水文、湿地生态学等相关学科的科学研究工作者及学生。

This dataset is land cover sample data for coastal mangrove growth zones in China in 2017, which provides data support for ecological conservation, wetland protection, and regional sustainable development. It is also suitable for scenarios such as auxiliary field surveys and remote sensing classification.

b. Purpose of Establishment
Samples are capable of driving classification algorithms and serve as a prerequisite for accurate classification. Coastal regions, located in the transition zone between land and sea, feature diverse land covers and thus require samples of more categories for characterization. However, the current scarcity of publicly shared samples makes independent and repetitive sampling work in similar studies highly time-consuming and labor-intensive. Therefore, a total of 16,444 samples across 10 categories obtained from mangrove classification and extraction research are being made publicly available.

This sample set covers low-altitude coastal areas across 8 provincial-level administrative regions along China’s coast: Hainan, Guangxi, Guangdong, Fujian, Zhejiang, Taiwan, Macao, and Hong Kong. It can not only support model training for coastal land cover classification, but also be used to validate various classification results. In addition to validating mangrove distribution maps, it can be applied to multiple tasks such as tidal flat extraction, salt marsh vegetation identification, and farmland mapping. It can broadly serve scientific researchers and students in relevant disciplines including agriculture, ecology, hydrology, and wetland ecology.