Developing a new index with time series Sentinel-2 for accurate tidal flats mapping in China

Tidal flats are essential habitats for a wide variety of plants and animals along the coast, contributing significantly to biodiversity preservation. Despite efforts to employ remote sensing technology for mapping these areas, misidentification remains a persistent issue. Within the context of tidal flat mapping, one of the primary challenges lies in distinguishing between tidal flats and turbid water bodies. This is primarily attributed to the high sediment content in turbid water bodies, which renders their spectral characteristics similar to those of tidal flats, consequently leading to confusion. To this end, this study developed a novel Tidal Flats-Water Difference Index (TWDI) with time series Sentinel-2 to accurately map tidal flats in China for the year 2020. The index demonstrates robust performance in distinguishing between tidal flats and water bodies, particularly in turbid water conditions. It reduces the likelihood of misclassifying water bodies as tidal flats, thus enhancing the accuracy of tidal flats mapping. Based on this index, we generated Tidal Flats Map of China at 10 m (TFMC) in 2020 based on the GEE. With validation from 10,234 samples, the TFMC map attained an overall accuracy (OA) and F1 score of 0.97. Based on our calculations, the total area of tidal flats in China was 9195.80 km², with Jiangsu Province accounting the largest area at the provincial level (2467.71 km²). This study is the first attempt to address the issue of misclassification of turbid water bodies as tidal flats. TWDI can serve as the spectral index in tidal flats mapping or as input features for training classification models of coastal wetland. The TFMC map can serve as foundational data to promote the protection of tidal flats and the sustainable development of coastal wetland ecosystems.