The data set consists of Network Common Data Format (NETCDF) files (https: //www.unidata.ucar.edu/software/netcdf/). The NETCDF is self explanatory and upon data download the command “ncdump -h filename” can be used to obtain full information on the data. The data set consists of seven main directories. Each of them separately contains one of the source cases investigated in the paper by Cassiani et al (2023, submitted) and the names of the directories are self explanatory. Inside the main directories several sub-directories contains separate time sections of the complete simulation. Typically a time section consists of 15s or 30s of simulation. Two main different types of NETCDF files are contained inside any time section. One type consists of the raw LES data (time series) for eight downwind positions and for each downwind position a grid volume covering Nx = 5, Ny = 512 and Nz = 514 grid nodes is included. The second type consists of the time average for several turbulent flows and scalar statistics covering the full 3D grid. Some ancillary data are also available in separate directories containing a similar data structure for lower resolution simulations as used in the publications by Ardeshiri et al. (2020) and Kylling et al. (2020)