Atmospheric Dispersion from Elevated Sources in an Urban Area: Comparison between Tracer Experiments and Model Calculations

TLDR

The study site was a mainly residential area with a roughness length of 0.6 m. The authors conducted tracer experiments in Copenhagen, releasing SF₆ from a 115‑m tower and sampling ground‑level concentrations 2–6 km downwind, then compared dispersion parameters derived from the data with those calculated by wind‑variance‑based and stability‑based Gaussian models. The results show that wind‑variance‑based Gaussian methods outperform stability‑based ones in predicting lateral spread and centerline concentrations, while the assumed vertical Gaussian profile often fails to match measured cross‑wind integrated tracer concentrations.

Abstract

Atmospheric dispersion experiments were carried out in the Copenhagen area under neutral and unstable conditions. The tracer sulphurhexafluoride was released without buoyancy from a tower at a height of 115 m and then collected at ground-level positions in up to three crosswind series of tracer sampling units, positioned 2–6 km from the point of release. The site was mainly residential having a roughness length of 0.6 m. The meteorological measurements performed during the experiments included the three-dimensional wind velocity fluctuations at the height of release. Dispersion parameters estimated from the measured tracer concentrations were compared with dispersion parameters calculated by various standard methods. These included methods based on the measured wind variances at the experiments and methods based on a stability classification of the atmospheric conditions. The wind variance-based methods are seen to be better than the stability-based methods in predicting the variation of σy. In addition to being a function of downwind distance, the normalized lateral spread estimated from the measurements was found also to vary as a function of the wind speed as already demonstrated by Doran and others for low-level sources. The centerline ground-level concentrations were calculated by use of σy and σz derived according to the aforementioned methods and assuming the ordinary Gaussian concentration profile. These were compared with the corresponding measured concentrations; in this case also, the wind variance-based methods were seen to be better. In some of the experiments, the assumed Gaussian tracer concentration profile in the vertical direction was found to be inconsistent with the measured crosswind-integrated tracer concentrations.