Data to be used as constraints for the rotor inflow benchmarks (stable cases) as defined in https://rotor-inflow-benchmark.readthedocs.io/UVW*.csv : high-frequency (20 Hz) time series of three-dimensional wind velocity vector over 24 hours centered on the periods of interest, obtained from meteorological tower sonic anemometers NL*.nc: low-frequency (1 Hz) time series of longitudinal wind velocity over each period, obtained from nacelle-mounted scanning lidar after quality control, temporal upsampling, interpolation to a regular grid and the use of a machine learning model to fill in gaps introduced by blade passage in front of the lidar LES*.nc: low-frequency (1 Hz) time series of longitudinal wind velocity over each period, obtained from large-eddy simulations performed to match the experimental field measurements UVW_les*.csv: high-frequency (25 Hz) time series of three-dimensional wind velocity vector at two heights for the 704 seconds of the large-eddy simulations optional-lidar-ppi-scans.zip: This contains the 1 Hz horizontal lidar scans that are the basis for the NL*.nc files. We include this optional dataset to explore the question: how different are reconstructed inflows when using the PPI scans versus the heavily processed inflow plane data as a constraint?