AIR QUALITY ASSESSMENT IN THE URBAN AREAS WITH MULTIVARIATE STATITICAL ANALYSIS AT THE EAST OF THAILAND

Two statistical multivariate analysis techniques had been applied to assess and predict whether the AQI in the urban areas at the east of Thailand exceeded the standard level.Cluster analysis was utilized to group the five main pollutants also classify the monitoring stations.To evaluate the air quality group, discriminant analysis was then built to determine the prediction model.The results present that there are three clusters of the main pollutants.Cluster #1 mentions SO 2 .Cluster #2 contains NO 2 , O 3 and PM 10 while only CO is in the cluster #3.For classifying of monitoring stations, it exhibits that there are three distinct clusters: Cluster #1 consists five stations: General Education Office, Laem Chabang Municipal Stadium, Field Crop Research Center, Health Promotion Hospital Maptaput and Agricultural Office.Ta Sit Subdistrict Administrative and Sriracha Municipal Youth Center are the only one station in the cluster #2 and the cluster #3, respectively.A correct classification rate (CCR) was then measured to evaluate performance of the predictive model.The discriminant model performs very well with the average CCR 80.10% for the training data set and 78.97% for the validation data set.