Author(s)

O. Connan, D. Maro, P. Germain, D. Hébert, M. Rozet, C. Lac, C. Camail & F. Bonnardot

Abstract

In order to reduce uncertainties and enhance the knowledge of atmospheric dispersion in urban environments at a macroscopic scale, various experimental releases of SF6 tracer were conducted in Toulouse (France) urban center in March and July 2004, during the CAPITOUL project. Close field SF6 measurements were made on the ground and at altitude (only ground data will be addressed in this paper). During experiments, the atmospheric conditions were unstable (class B according to Pasquill classification). The Atmospheric Transfer Coefficients (ATC) vary from 2.3 10 -5 s.m -3 to 7.4 10 -8 s.m -3 on the ground for distances from the release source between 350 and 3140 m and for wind velocities ranging from 2 to 4 m.s -1 . Experimental ATC were compared to the results of Doury, Pasquill and Pasquill-Urban Gaussian models. For these urban conditions, the comparison shows that the Pasquill-Urban model is the most appropriate in most cases. Average ATC are particularly overestimated by Doury and Pasquill models. On the other hand, all models, including Pasquill-Urban, generally underestimate the horizontal dispersion by a factor of approximately 2 to 4. These results show that it is necessary to acquire more experimental data on urban environments, for various stability conditions and various urbanization levels, in order to better characterize the horizontal and vertical dissemination in these specific environments, and thus be able to meet the operational requirements. Keywords: atmospheric dispersion, urban environment, tracer release, SF6, Gaussian models, atmospheric transfer coefficient.

Keywords

atmospheric dispersion, urban environment, tracer release, SF6, Gaussian models, atmospheric transfer coefficient.