Advancing Vertical Wind Profile Characterization under Urban Morphological Constraints for High-Density Cities

Accurately characterizing vertical wind profiles within the urban roughness sublayer (URS) is fundamental for urban ventilation assessment, wind energy evaluation, and climate-resilient urban design. In high-density cities, conventional similarity-based formulations often fail to represent non-linear flow structures induced by heterogeneous building morphologies, variable wind directions, and non-neutral atmospheric conditions. This study develops a physics-guided, multi-source framework for retrieving high-resolution vertical wind profiles (10–300 m) by integrating ground-based Doppler LiDAR observations with near-surface micrometeorological measurements and GIS-derived urban morphological constraints. Focusing on a compact high-density urban environment in Taipei, the framework reconstructs vertical wind structure by resolving zonal and meridional momentum components prior to synthesizing horizontal wind speed, thereby preserving directional variability and anisotropic drag effects suppressed in scalar extrapolation approaches. The framework is validated through an intensive field campaign combining Doppler LiDAR, 3D ultrasonic anemometers, and micro-weather stations. Results demonstrate robust performance across multiple flow regimes (R² = 0.97, RMSE = 0.61 m/s), with substantial error reduction relative to power-law and log-law extrapolation. Feature contribution analysis identifies building height density (BHD) as a stable morphological proxy for aerodynamic roughness. Importantly, a cost-effective configuration relying only on micro-weather stations and urban morphology retains high accuracy (R² = 0.96), indicating that physically interpretable morphological parameterization can compensate for sparse dynamic measurements. The proposed framework supports applications in urban climate assessment and climate-resilient urban design, ventilation analysis, urban wind energy assessment, and the specification of realistic inflow conditions for CFD simulations.