This data set is led by researcher Wang Xin and associate researcher Zhang Rongwang from the South China Sea Institute of Oceanology,Chinese Academy of Sciences,co founded by the Oceanographic Research Center of the Chinese Academy of Sciences. It is mainly based on satellite remote sensing and reanalysis data,with observation data such as China's offshore buoys,offshore platforms and flux towers as training data,and is developed using technologies such as target depth neural network ensemble average model and improved COARE 3.5 algorithm,It effectively overcame such outstanding problems as underestimation of latent heat flux under medium atmospheric specific humidity and error of latent heat flux caused by sea surface cold skin effect. Compared with international mainstream air sea flux products,this product has significant regional advantages and performance index advantages in China's offshore.