Accounting for the effect sub-grid scale emission variability on upper-air concentration

Heterogeneously distributed emissions characterize urban and con-urban environments. The emission sources can be linear (roads), surface (fields or urban areas) or punctual (factory). Emission inventories are created by averaging the emissions over the grid cell where the source is located. After the averaging procedure, the sources are considered as uniform over the entire grid cell. This means that not only emission spatial variability is lost, but also that it will not be accounted for in the upper atmosphere. This can represent a serious issue in case of passive as well as chemically reactive species, or for the estimation of short or long-term exposures.

We propose a novel approach to the problem including a formulation for the sub-grid variability of pollutant concentrations. This formulation takes into account the spatial heterogeneity of the emissions. The founding concept of the approach is the assumption that average emission acts as a source term of average concentration, while emission fluctuations are a source for the concentration variance. The model has been implemented in an existing mesoscale model. We used large-eddy simulation data to test the parameterization. The results show and excellent agreement between the models.

For the first time an error bar is associated with the pollutant time evolving concentration. This error-bar accounts for the way in which sub-grid scale heterogeneity of the emissions affects the concentration. The concentration variance is presented as an extra attribute to the mean concentrations in Reynolds-average models. This approach has applications from the mesoscale to the global scale. An example of the application to the Lombardia region (Italy) is also presented.