Photochemical smog modeling of PM2.5 for assessment of associated health impacts in crowded urban area of Southeast Asia

Abstract

Integrated approach of emission inventory (EI) and dispersion modeling was used to generate annual gridded PM2.5 over Bangkok Metropolitan Region (BMR) for the health effect assessment. Road traffic, rice straw open burning, and industry were the major anthropogenic sources of PM2.5 and its precursors in BMR in 2016. The EI results were used to prepare input data for 3D photochemical smog WRF-CAMx modeling with nested domains. WRF model satisfactorily reproduced meteorological data observed at 2 international airports. WRF-CAMx simulated PM2.5, gridded at 2x2 km, were evaluated against the available observed data at 3 monitoring sites in BMR which showed better agreement for longer averaging time (daily, monthly, and annual) than the hourly values. Hourly simulated PM2.5 could capture well the range of the observed values but not the temporal variations. The daily and monthly PM2.5 were satisfactorily simulated with all the statistical parameters of the performance evaluation meeting the suggested goals. The simulated weekly PM2.5 mass and compositions of sulfate and nitrate were also reasonably compared with the available observed data at a suburban site. Especially, a good agreement between the simulated annual PM2.5 and the observed, with relative errors of 8%–18%, justified the use of the simulated data for the health impact assessment. The health impact analysis using BenMAP-CE showed that reducing annual PM2.5 to meet the WHO guideline would help avoid 1,415 deaths (1,068 – 1,758) in BMR with 911 in Bangkok alone. Further reducing PM2.5 to the counterfactual concentration of 2.4 μg/m3 would help avoiding 2,990 PM2.5-attributable mortality in BMR with 1,770 in Bangkok alone. Global Exposure Mortality Model (GEMM) incorporating the data from global cohort studies was useful for calculation of relative risk values for the health assessment in BMR.