The impacts of boundary layer processes on chemical transformations are currently of great interest in many fields of atmospheric chemistry since they affect the concentration of trace gases such as ozone, nitrogen oxids and volatile organic compounds. Thus the chemical lifetime of reacting species varies on a wide range of time-scales, the transport and the mixing of pollutants are highly dependent on the ratio between the time-scale associated to the chemistry and that characteristic of the boundary layer processes. The ratio between the turbulent and the chemical time-scales, e.g. the turbulent Damköhler number Da, is used to estimate the influence of turbulence on chemical transformations.

Figure 1:Vertical cross section of the concentration of A for different turbulent reacting flows. A is emitted at the surface with a constant and uniform flux. It reacts following a second-order reaction with a chemical species B entrained from the free troposphere.

For a reacting scalar A involved in the second order reaction A+B->C with a rate coefficient k, it reads

According to this number, the reacting flow can be classified as

• Da << 1, the reactant is homogeneously spread in the boundary layer before the reaction takes place. There is therefore little influence of the turbulent structures on the chemistry. The reacting flow is in a so-called slow chemistry regime.
• Da ~ 1, the turbulent mixing is expected to have a non-negligible impact on chemical transformations. The reacting flow is representative of moderate chemistry such as the oxidation of nitrogen oxide by ozone or the degradation of isoprene.
• Da >> 1, the flow shows a fast chemistry regime behaviour. The turbulent mixing limits the mixing of the reactants before they react or the chemistry is so active that the chemical species react in-situ and are almost not transported

Figure 2:Schematic representation of two modelling approaches: (a) the mixed-layer model that assumes instantaneous and homogeneous mixing and (b) the 3-dimensional LES model that explicitly solves the turbulent structure of the CBL (from Vinuesa and Vilà-Guerau de Arellano, 2005).

In the ABL, the transport of the chemical compounds is associated with the upward and the downward motions. As a result, two reactants involved in the same chemical reaction can be transported in the opposite or in the same direction. Since the chemistry requires a complete mixing up to molecular level, the segregation of species due to turbulent characteristics can affect the reaction rate and can lead to a slowing down of the reactivity of chemical compounds. By comparing an homogeneously mixed situation to a heterogeneous distribution of pollutants due to atmospheric turbulence (see Figure 1), one can study the relevance of accounting for inefficient mixing in reactant concentration calculation. The inability to correctly parameterize the effect of turbulence on chemical transport and chemical transformation has a profound impact in the accuracy of many facets of numerical air quality prediction ranging from short-term regional forecasts to long-range climate change studies. Errors in the parameterization of the inefficient mixing of reactants by the atmospheric turbulence typically lead to the misestimating of the segregation of chemical species, which in turn can have a very important impact (reduction or increase) on the rate of the chemical transformations, with respect to the well-established values obtained under laboratory (non-turbulent) conditions. Large-scale chemical transport models and air quality models cannot solve boundary layer turbulent eddies and, therefore, they need to parameterize the turbulent fluxes of momentum and scalars (e.g. temperature, water vapor and pollutants). In addition, the turbulent covariance, that is the variable that quantifies the influence of turbulent mixing on the chemical species, is usually neglected by assuming that the chemical species are uniformly mixed.

Figure 3: Time evolution of the concentrations. The diamonds account for the LES results. The mixed-layer model simulations using the laboratory reaction rate, which assumes a homogeneous and instantaneous mixing are plotted with solid lines. The dashed lines represent the mixed-layer model results when the heterogeneous mixing due to convective turbulence is considered by using the effective rate coefficient calculated by Vinuesa and Vilà-Guerau de Arellano (2005).

Figure 3 shows that solving turbulence explicitly by using LES decreases the mixing ratio of ozone. These differences appear at sunrise and increase during the boundary layer growth. One can notice similar behaviour for highly reactive compounds such as OH, HO2 and H2O2. For others, such as NO and, to a certain extent, RH, the segregation and the non-instantaneous mixing due to the characteristics of the CBL reduces their reactivity. The effective reaction rates as given by Vinuesa and Vilà-Guerau de Arellano (2005) for the chemical mechanism used by Krol et al. (2000) shows that the reaction rates between ozone and NO and between ozone and HO2 are enhanced by the turbulent mixing. As a result, when using these effective reaction rates with a mixed layer model, we found that the levels of ozone and HO2 are reduced leading to a better agreement with the LES results. One can notice also that the reaction rate between OH and RH is reduced by 50% improving the agreement with LES results.

References:

• Krol M. C., M. J. Molemaker, and J. Vilà-Guerau de Arellano, 2000: Effects of turbulence and heterogeneous emissions on photochemically active species in the convective boundary layer. Journal of Geophysical Research, 105, 6871-6884.
• Vinuesa J.-F., and J. Vilà-Guerau de Arellano, 2003: Fluxes and (co-)variances of reacting scalars in the convective boundary layer. Tellus, 55B, 935-949.
• Vinuesa J.-F., and J. Vilà-Guerau de Arellano, 2005: Introducing effective reaction rates to account for the inefficient mixing of the convective boundary layer. Atmospheric Environment, 39, 445-461.