PALM-Simulated Urban Wind Fields over Camden for Scientific Machine Learning

This dataset contains 41 PALM-generated NetCDF files used for building-resolved urban wind field reconstruction over the London Borough of Camden. The dataset was used to train and evaluate conditional classical and hybrid quantum-classical surrogate models in the associated manuscript.

Each file corresponds to one inflow scenario defined by a wind speed and a wind direction. File names follow the format extracted_wXX_dYY_3d.000.nc, where wXX denotes the inflow wind speed in m/s. The direction label dYY ranges from d03 to d10 and represents eight inflow wind directions at 45° intervals, starting from north and increasing clockwise. For example, extracted_w08_d05_3d.000.nc represents the PALM output for an 8 m/s easterly inflow condition. Besides, an extrapolation scenario with a wind speed of 7m/s and a southeastern wind direction (120 degree) is included in extracted_w07_deg120_3d.000.nc.