Application of Time Series Analysis and Geographic Information Systems for Particulate Matter Less Than or Equal to 10 Microns in Diameter (PM10) Assessment in Bangkok

Abstract

The study aims to forecast the particulate matter less than or equal to 10 microns in diameter (PM10) in 31 stations in Bangkok and surrounding provinces. The data of monthly average PM10 from 1997 to 2014, were analyzed to evaluate the dispersion of PM10 and demarcate the risk areas in Bangkok. The estimated data imputation technique was applied to represent the non- measurement data. The monthly average PM10 in 2015 was predicted using time series analysis technique. The calculated yearly average PM10 was analyzed in Geographic Information Systems (GIS) using data interpolation technique to classify five levels of risk areas, and then the size of areas. The number of affected people were analyzed by the overlay technique. The result showed that the 2015 yearly average PM10 was between 15.5 to 80.3 µg/m³. The highest value was found at the station of Vibhavadi Forest Park Din Daeng, Chulalongkorn Hospital, and Phra Khanong District Office, respectively. For the size of risk areas in Bangkok, most of the areas (80.14%) were classified within moderate risk level and the highest risk area (1.95%) was 30.673 km². For the analysis of the number of affected people, 156,175 people would be affected in high risk areas (the annual average PM10 exceeding standard) and 17,262 persons were in the highest risk areas. This study showed that the Time Series Analysis is a good technique to forecast the monthly average PM10. The Mean Absolute Percentage Errors (MAPE) was in the range of 10.5 to 21.82. In addition, GIS could be used to estimate the concentration of PM10 in the non-measurement data area. The results of the study could support the planning, decision making on effective prevention and air pollution management.