A model for the estimation of standard deviation of air pollution concentration in different stability conditions

We propose to estimate the standard deviations of the air pollution concentration in the horizontal and vertical direction, and , based on Pasquill's well known equation, in terms of the wind variance and the Lagrangian integral time scales, on the basis of an atmospheric turbulence spectra model. The main advantage of the spectral model is its treatment of turbulent kinetic energy spectra as the sum of buoyancy and a shear produced part, modelling each one separately. The formulation represents both shear and buoyant turbulent mechanisms characterizing the various regimes of the Planetary Boundary Layer, and gives continuous values at any elevation and all stability conditions from unstable to stable. As consequence, both the wind variance and the Lagrangian integral time scales in the dispersion parameters are more general than those found in literature, because they are not derived from diffusion experiments as most parameterizations. Furthermore, they provide a formulation continuous for the whole boundary layer resulting more physically consistent. The , parameters, included in a Gaussian model have been tested and compared with a dispersion scheme reported in the literature, using experimental data in different emission conditions (low and tall stacks) and in several meteorological conditions ranging from stable to convective. Results show that the dispersion model with the sigmas- parameterisation included, produces a good fitting of the measured ground-level concentration data in all the experimental conditions considered, performing slightly better than other-state-of art models