Numerical simulation of urban thermal plume merging

Densely packed urban buildings trap outgoing long-wave radiation, leading to reduced surface cooling and increased building surface temperature. In calm conditions, poor natural ventilation causes both thermal comfort and air quality issue. The buoyancy flow generated by heated urban surfaces is the main driving of the urban flow and pollutant dispersion. A 3D numerical modelling is conducted to investigate the thermal plumes merging and buoyancy-driven airflow in urban areas. The performances of four different turbulence models, i.e., two URANS (Unsteady Reynolds-averaged Navier–Stokes equations) models and two LES (Large-Eddy Simulation) models are evaluated by comparing the velocity field with previous water tank measurements. Validation results show that all four turbulence models can capture the bending of thermal plumes toward the centre, and LES models provide a better prediction on the vertical velocity profiles, while both URANS models show underestimation. The plume merging mechanism is analysed with the high accuracy LES results. Both pressure difference and swaying motion caused by mean flow and turbulence are important for plume merging. The turbulence coherent structure of plume merging is analysed by a quadrant analysis, which shows ejection and sweep events could significantly change with the building density. A case study with complex urban geometry is conducted to show the impact of thermal plumes merging in the real high-density urban areas. The convergence airflow at the pedestrian level is estimated to 2 m/s under a surface-air temperature difference of 5 °C, which is comparable to wind-driven ventilation and beneficial to thermal comfort and air quality.