Modulation of European air quality by Euro-Atlantic weather regimes

Abstract

This paper analyses the dependency of regional air pollution in Europe on the atmospheric circulation as represented by a set of 8 weather regimes (WRs). For this purpose, daily average PM2.5 concentrations and daily maximum concentrations of O3, NO2 and CO obtained from an atmospheric composition reanalysis have been used. The results indicate that the pollutant concentrations change substantially with the occurrence of each WR according to the associated circulation patterns. In the extended winter, the responses of PM2.5, NO2 and CO are quite similar, with zonal circulation regimes leading to better air quality than anticyclonic regimes linked to steady flows over Europe. On the other hand, the spatial patterns of O3 anomalies under most WRs resemble those of temperature in the extended summer as temperature enhancements favour the photochemical production of O3. The WR characterized by anticyclonic anomalies over Scandinavia leads to the highest concentrations at continental scale and is concurrent with an increased probability of exceeding the poor air quality limits of winter PM2.5 and summer O3 in most European countries. It is also found that a multiple linear regression model on the monthly frequency of WRs can explain a considerable fraction of the intermonthly variability of concentration anomalies over large parts of Europe regardless of the pollutant, with R2 values around 0.6–0.8 over large regions. Overall, the application of these models to sub-seasonal weather forecasts does not seem to bring major improvements to the prediction of month-ahead concentration anomalies as compared to the climatology. The poor skill of sub-seasonal forecasts in predicting the WRs for lead times beyond 15 days limits the forecast capability. However, these results open the possibility of extending current air quality forecast to scales of 5–10 days at a low cost in a near future.