Abstract: The Yellow Sea enteromorpha green tide occured year after year since 2007. In order to trace the source and development process of green tide, various remote sensing identification methods have been carried out. In this study, Sentinel-2 MSI multispectral data were used to identify floating green tide. Firstly, field-measured spectrum of the floating green tide and the suspended sediment water was analyzed. And then three remote sensing index methods, NDVI, FAI, VB-FAH, were applied into floating green tide identification. Finally, comparison between remote sensing identification and manual extraction was proceeded. Compared with NDVI and FAI, VB-FAH index based on the reflectance difference between near-infrared and green bands, which can reduce the influence of high suspended sediment water on near-infrared and red channel, showed a better identification effect. This study can provide technical support for remote sensing identification of floating green tide in high suspended sediment water.