Land surface models (LSMs) must have reliable forcing, validation and surface attribute data as the foundation for effective model development and improvement. Eddy covariance flux tower data are considered the benchmarking data for LSMs. However, these datasets lack site-observed attribute data, limiting their use as benchmarking data. Here we collected their vegetation, soil information, and wind speed measurement height from literature, regional networks, and Biological, Ancillary, Disturbance, and Metadata (BADM) files. Then through global data product complement and plant functional types (PFTs) classification, we obtained the final flux tower attribute dataset, presented in NetCDF format with necessary descriptions and reference sources.