Estimating PM2.5 concentrations with statistical distribution techniques for health risk assessment in Bangkok

Abstract

Ambient PM_2.5 data in the Central Business District (CBD) of Bangkok monitored by Pollution Control Department and Bangkok Metropolitan Administration were collected over three years in Bangkok from 2015 to 2017. The other air pollutions data were used as the dependent variables to develop mathematic models with statistical distribution technique. Multiple linear regression technique was selected as the main statistical distribution methodology for estimating PM_2.5 concentrations in non-monitored areas. The predicted PM_2.5 concentrations were validated against the measured PM_2.5 concentrations by various statistical techniques. The validation found that the model had strong significant correlations for ambient and roadside area with R ²?=?0.88 and 0.96, respectively. The non-carcinogenic health risk assessment of PM_2.5 was quantified as the hazard quotient (HQ) from both the measured and predicted data. The risk areas and HQ were compared using the inverse distance weighting interpolation technique and illustrated as GIS-based maps. During December to February, the HQ values of PM_2.5 were exceed 1 (HQs?>?1) at all area of CBD; however, the highest HQ was found in the southern part of CBD. The finding could be used for residential health awareness in that area.