Flow simulations around randomly generated urban geometries (Random Buildings)
Simulation3.6 GB
Dataset of microscale atmospheric flow simulations around randomly generated urban geometries for several stability classes. Useful for training and testing ML surrogates of urban flow.
Large-eddy simulations of turbulent airflow over 314 realistic urban neighbourhoods from 62 cities in 20 countries, with plan-area densities from 0.05 to 0.75 and several wind directions. Provides canopy-averaged statistics and flow fields, which are useful for canopy parameterisation, model validation and machine learning of urban wind.
720 time-mean lattice-Boltzmann LES cases at 10 m resolution, 144 per city, spanning 16 wind directions, 3 reference speeds and 3 seasonal profiles, with 3D mean velocity and TKE plus building, terrain and footprint geometry. Very large (about 25 TB).
LES velocity fields over urban canopies with inputs and U-net predicted fields, including perturbed cases. Useful for training and testing ML surrogates of urban flow.
From Zenodo, added 2026-11-07.
Listed there as “Random Buildings dataset”. Tags proposed by AI and not yet checked by hand. Spotted a mistake? Tell the portal contacts.