3.5. Time series and correlation analysis
It can be seen from Figs. 1 and 2 that the time series of OC, the anhydrosugars and the nitroaromatic compounds closely follow each other. The correlation coefficients between all these variables were mostly between 0.7 and 1.0 in both campaigns; only a few of the correlations among the nitroaromatic compounds were somewhat lower, but still around 0.6. Figs. 3 and 4 show the time series for the monosaccharides glucose and fructose, the polyols mannitol and arabitol, and the isoprene SOA markers. Glucose and fructose, which are both related with microorganism and plant emissions, closely follow each other and were very highly correlated (r > 0.9) with each other in both years. Similar very high correlations (all r > 0.9) were observed between the polyols mannitol and arabitol, as could be expected because of their common origin (fungal spores), between 2MT and 2ME, and among the three alkenetriols (alk2, alk3, and alk1). The correlations between 2MG and the other compounds were less consistent; in 2012, 2MG was highly correlated (all r > 0.7) with the two monosaccharides and the three alkenetriols, but in 2013, those correlations were only
in the range 0.4e0.7.