Clustering of atmospheric and emission conditions that lead to modelled peak ozone concentrations

MAZZEO, N.A.; PINEDA ROJAS, A.L.

Bolonia

Conferencia; 18th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes (Harmo18); 2017

Harmo

An application of a simple urban air quality model (DAUMOD-GRS) shows that summer maximum O3 hourly concentrations (Cmax) above 40 ppb occur outside the Metropolitan Area of Buenos Aires (MABA) where the absence of observations impedes model testing. In addition, those relatively high values present the greatest model uncertainty caused by possible errors in model input variables. In order to tackle this issue, a probability assessment of Cmax values exceeding 40 ppb is performed applying a Monte Carlo analysis. On the other hand, a non-hierarchical (k-means) clustering analyses is applied to analyse the Monte Carlo outcomes. Results show three main clusters with a marked spatial distribution resembling that of the ozone precursor species emissions, which highlights an important role of the emissions on the regimes under which modelled Cmax values in the MABA can occur.