Application of cluster analysis to surface ozone, NO2 and SO2 daily patterns in an industrial area in Central-Southern Spain measured with a DOAS system

The daily variations of surface ozone, NO2 and SO2 have been investigated in a heavily industrialised area in the centre of the Iberian Peninsula (Puertollano) using hourly values recorded during two years (2008-2009) with an active LP-DOAS system. The meteorological conditions and air masses have been studied using the HYSPLIT model. The maximum hourly levels of these air pollutants exceeded 100 ppb for ozone, 150 ppb for SO2 and 210 ppb for NO2. However, mean values for ozone, NO2 and SO2 were of 49, 10 and 3 ppb respectively. Daily-monthly evolutions (defined as daily evolutions for different months) have been analysed in order to know the general daily behaviour of these species. Air pollution problems have been identified using the thresholds defined in the European Directive 2008. The limits to protect human health (human health protection limitations) have been exceeded during the study period. In order to find a set of representative daily cycles for each pollutant at different air quality regimes, a K-mean cluster technique has been applied. Five and four optimal cluster numbers have been obtained for the daily patterns of ozone and SO2 respectively. In addition, we studied the daily variation of the temperature, relative and specific humidity and wind speed associated with each air pollutant daily pattern. Ozone daily patterns showed typical daily variations with one exception of a cluster which presents a peak in the early morning. For SO2. the first two clusters present a low mixing ratio, however cluster 3 and 4 are less frequent but with higher levels. The more frequent air pollutant daily patterns do not exceed the threshold defined in the Directive. Nevertheless, clusters with lower frequency (representing between 5 and 7% of days) exceed the thresholds and could be considered as air pollution events. (C) 2012 Elsevier By. All rights reserved.