Implementation of an aggregate index to elucidate the influence of atmospheric synoptic conditions on air quality in Athens, Greece

To evaluate air quality in Athens (Greece), hourly concentrations of PM10, NO2, O-3, and CO from an urban background monitoring station and a suburban sampling site, covering the years 2014-2018, were used for the calculation of individual daily air quality indices (AQIs), reflecting the impact of each pollutant on air quality. Individual AQIs were then integrated in a previously reported formula for the computation of a daily aggregate AQI (AQI(agg)) expressing the overall air stress. According to the results of AQI(agg), a generally moderate air quality was estimated in Athens; however, an increased likelihood of extreme AQI(agg) episodes was calculated during the warm period (16 April-15 October), due to the combination of enhanced PM10 and O-3 concentrations. Approximately 83% of AQI(agg) episodes occurred during the warm period and only 17% in cold period (16 October-15 April). Peaks of AQI(agg) in cold period were primarily related to desert dust intrusions from North Africa. AQI(PM10) and AQI(O3) were the primary contributors to AQI(agg), whereas AQI(NO2) and AQI(CO) remained at low levels. AQI(PM10) and AQI(O3) had the highest contribution to AQI(agg) on 72% and 28% of total days respectively. Atmospheric circulation patterns produced by a K-means cluster analysis procedure revealed that extended low pressure areas centered over the UK were associated with dust importation in Athens due to the triggering of southwestern airflows by the cyclonic activity. Strong correlations between the individual and aggregate AQIs calculated at the urban background and the suburban sampling site were found, indicating relatively uniform air quality conditions inside the Athens basin. Mean annual AQI(agg) values at the suburban station for the years 2001-2018 showed a gradual amelioration and stabilization of air quality in Athens after 2008, related to the status of Greek economy and depollution measures.