Assessing the value of air stagnation indices to reproduce PM10 variability in Europe.

Different air stagnation indices (ASIs) have been proposed to measure the atmospheric diffusion conditions. This paper undertakes a comparative analysis of three ASIs and explores their ability to capture the conditions conducive to elevated PM 10 (particulate matter ≤10 μm) at 306 background sites in Europe for the winter and summer months of 2000–2012. Despite the similar spatial patterns of stagnation frequency found for the three ASIs, the use of different meteorological variables to characterize the dilution capacity of the atmosphere (wind speeds at 10 m and 500 hPa, boundary layer height and ventilation in the boundary layer) leads to important differences in the seasonal cycles. Moreover, the response of the PM 10 concentrations to stagnation varies with the ASI. Winter PM 10 anomalies under stagnant conditions are of the same order of magnitude (on average 17.2 to 18.6 μg m−3, around 60% of the mean values) for the three ASIs. The anomalies are considerably smaller (3.8–5.7 μg m−3, around 19–28%) in summer, when one of the indices outperforms the others at most locations. The dependence of the PM 10 concentrations on the ASI components is also evaluated by using correlations and generalized additive models. The results indicate that the consideration of the large-scale circulation is particularly relevant in summer, explaining the different ASI performances during this season. We have also identified some potential improvements that could be made to two of the ASIs. Nevertheless, since the three ASIs are based on fixed thresholds, they cannot deal with non-linear relationships, which limits their ability to explain PM 10 variability. • First comparison of three different air stagnation indices in Europe. • The indices yield different seasonal cycles in the frequency of air stagnation. • The response of PM 10 to stagnation varies with the index, location and season. • Large-scale circulation consistently contributes to reproducing PM 10 variability. • Fixed thresholds limit the ability of stagnation indices to explain PM 10 variability. [ABSTRACT FROM AUTHOR]