EVALUATION OF THE AKSARAY PROVINCE AIR QUALITY: CONDITIONAL BIVARIATE PROBABILITY FUNCTION AND K-MEANS CLUSTERING

Abstract: This study addresses the three major questions: (1) what are the emission sources of PM10 and SO2 which are affecting the study area; (2) where do these emission sources come from; and (3) is there any temporal variation in the emission sources. In the current work K-means clustering techniques were applied directly to bivariate polar plots to identify and group similar features. The technique is analogous to clustering applied to back trajectories at the regional scale. When applied to data from a monitoring site with high source complexity it is shown that the technique is able to identify important clusters in ambient monitoring data. In Aksaray PM10 values follow a seasonal trend. The average PM10 concentration was recorded higher in the summer season and lower in the winter. It is observed that 50 µgm-3, which is the 24-hour limit value of PM10, was exceeded in both summer and winter months. The average SO2 concentrations also was detected higher during the winter months due to domestic heating and there was a decrease in concentration in summer. The winter and summer SO2 average concentrations were calculated as 7 and 2 µgm-3, respectively. Looking at the SO2 distribution over the months, it was seen that the normalized values are below 0.5 and the higher values were recorded in the period between November and February. Cluster analysis has been carried out for the PM10 and SO2 surface for clusters between 2 and 10. The choice of the number of appropriate clusters is heuristic and is best determined by post-processing the data according to cluster. 5 and 4 clusters were considered for PM10 and SO2, respectively. PM10 clusters were determined as 1 and 2- suburban emission, 3-traffic emission, 4-urban emission and 5-industrial emission. SO2 clusters were identified as 1- suburban emission, 2- industrial emission, 3- urban emission and 4- mix of urban and suburban emission.