An experimental solution to the mechanism and parameterization of ventilation and pollutant dispersion in the atmospheric surface layer

Abstract

Air pollution has aroused increasing public concerns in urban areas where the building morphology complicates transport processes in the atmospheric surface layer (ASL). Air quality management requires an advanced understanding of transport mechanism and practical dispersion models, formulating efficient pollution mitigations. However, the current knowledge of morphological effect on urban air quality is still inadequate for the interpretation of flows and dispersion processes, hindering the assessment of ventilation and pollutant plume dispersion over urban areas.

This study attempts to develop parameterizations and elucidate the mechanisms of ventilation and pollutant dispersion in response to the wind flows and turbulence over rough surfaces. Wind-tunnel modeling is used as an experimental platform to simulate the turbulent boundary layer (TBL). Different configurations of idealized urban and vegetation roughness elements are adopted. Water vapor is used to develop the tracer plume in the experiments. The mean and turbulent properties of velocity and concentration are measured to validate the parametrizations and demystify the mechanisms of street-level ventilation, wind profiles in roughness sublayer (RSL) and pollutant dispersion in the TBL.

At the street-level scale, it is found that the air exchange rate (ACH) is proportional to the square root of drag coefficient Cd ( ACH''/U∞b ∝Cd^1/2 ; where ACH'' is the turbulent ACH, U∞ the freestream velocity and b the street width). The newly suggested formulation demonstrates that the ventilation ability improves with rougher surfaces. It is because the aerodynamic resistance induced by rough surfaces promotes the extreme events (ejection Q2 and sweep Q4) and strengthens the large-scale motions, which partly explains the street-level ventilation mechanism. The formulation could then be used as a parameterization of ventilation.

At the RSL scale, the new analytical model of mean-wind profiles for RSL and inertial sublayer (ISL) is applicable to both urban and vegetation canopies. Its key parameter, the RSL constant, is found to be in the range of 1.8 to 3.1. The more uniform mean-wind speed and enhanced turbulent mixing in the RSL are signified by the faster increasing integral length scale in the RSL than that in the ISL. This finding enriches our understanding of the RSL eddy diffusivity.

At the boundary layer scale, the conventional Gaussian model could mostly estimate the concentrations over urban rough surfaces. However, the near-ground maximum concentrations are elevated and the vertical dispersion coefficient σz is enhanced due to the large aerodynamic resistance induced by surface roughness. It is found that σz can be estimated in terms of the downwind distance x, the TBL thickness δ and the drag coefficient Cd, i.e. σz ∝ x^1/2×δ^1/2×Cd^1/4. Comparing the horizontal advection flux and vertical turbulent flux, it is found that rougher surfaces enhance the turbulent flux so do the dispersion processes, explaining the transport mechanism in the ASL over urban areas.

The handy parameterizations and advanced understanding of transport mechanisms benefit the practice of environmental impact assessment, urban planning and sustainable city management in the highly urbanized and densely built cities, such as Beijing, Hong Kong and Shanghai.