Spatial Distribution and Source Apportionment of Air Pollutionin Bahrain using Multivariate Analysis Methods

摘要

The objective of this study is to identify the most important air pollutants based on theirindividual contribution to Air Quality Index (AQI) and to determine the major air pollutionsources in Bahrain. Data sets from seventeen air quality monitoring sites were evaluated usingXLSTAT 2014 and Statistical Package for the Social Sciences (SPSS 22) over six-and-half-yearbetween July 2006 and December 2012. Hierarchical Agglomerative Cluster Analysis (HACA)categorized the monitoring sites into three distinctive clusters based on similarities of airpollutants characteristics and meteorological parameters. Principal Component Analysis (PCA)identified major sources of air pollution in each cluster. Results demonstrated that dust storms,industrial activities, vehicular emissions, airport activities, power plants and filling stationswere major air polluters. PCA analysis showed that temperature and wind speed have positiveloading while relative humidity has negative loading. Multiple Linear Regression (MLR) analysiswas applied to develop models for prediction of AQI for every cluster based on concentrationsof key air pollutants. Results showed PM 10 and PM 2.5 highly contributed to AQI values. MLRmodels exhibited good fit with adjusted R 2 value of 0.865, 0.794 and 0.842 for Clusters 1, 2 and3 respectively. Standardized coefficient values for PM 10 succeeded by PM 2.5 were the highestin each cluster.