multi-source fusion seasonal coastline dataset of China at 10m resolution (2019-2025)

This dataset provides a high-frequency (seasonal) and high-resolution (10m) tracking of China's coastline dynamics from 2019 to 2025. It is developed using the Google Earth Engine (GEE) platform by integrating Sentinel-1 Synthetic Aperture Radar (SAR) and Sentinel-2 multispectral imagery. Methodology: Robust Compositing: Provincial-level coastlines were extracted from 50th percentile (median) composites of seasonal time-series data (MNDWI and SAR VV). This statistical strategy effectively suppresses transient noise such as wave foam, cloud shadows, and ship interference. Adaptive Extraction: A Tiled OTSU (block-based adaptive thresholding) algorithm was employed to handle spatial heterogeneity in water-land reflectance across the vast Chinese coastline, ensuring geometric continuity. Tidal Normalization: For four representative port sites (Binhai, Weihai, Kanmen, and Chongwu), the dataset provides both original median waterlines and tidally corrected coastlines. The latter were generated by coupling the FES2022 global tide model with a local slope-inversion algorithm, normalizing instantaneous waterlines to the Mean Sea Level (MSL) datum.本数据集提供了2019年至2025年中国海岸线动态的高频（季节性）、高分辨率（10米）追踪数据。该数据集基于谷歌地球引擎（Google Earth Engine，GEE）平台开发，整合了哨兵-1（Sentinel-1）合成孔径雷达（Synthetic Aperture Radar，SAR）与哨兵-2（Sentinel-2）多光谱影像。 ### 研究方法： 1. 稳健合成：从季节性时序数据（归一化差异水体指数MNDWI与SAR VV波段）的50百分位（中位数）合成结果中提取省级海岸线。该统计策略可有效抑制海浪泡沫、云阴影及船舶干扰等瞬态噪声。 2. 自适应提取：采用分块大津（Tiled OTSU，基于分块的自适应阈值）算法，以处理中国广阔海岸线范围内水陆反射率的空间异质性，保障海岸线的几何连续性。 3. 潮汐归一化：针对滨海、威海、坎门、崇武四个典型港口站点，本数据集同时提供原始中位数水边线与经潮汐校正的海岸线。后者通过耦合FES2022全球潮汐模型与局地坡度反演算法生成，将瞬时水边线归一化至平均海平面（Mean Sea Level，MSL）基准面。