Sensitivity of WRF-Chem model resolution in simulating particulate matter in South-East Asia

Abstract. Frequent haze episodes commonly caused by biomass burning has been a concern in South East Asia. One of such events was the June 2013 severe haze in the region. This study assessed the ability of WRF-Chem in capturing the spatial variability and concentrations of particulate emissions during this period. It analyzed the regional biomass burning emissions and its transport leading to higher particulate matter levels in the region. In order to analyze the effect of grid-scale, the horizontal resolution of the simulation was varied between low-resolution (100 km) and high-resolution (20 km). Evaluations of the simulations were made against meteorological observations pertinent to emission and transport of particulate matter, including surface and vertical air profile variables such as temperature, relative humidity, and wind speed and direction. Particulate matter (PM₁₀ and PM_2.5) levels were evaluated using ground measurements in Brunei and Singapore respectively. The meteorological parameters were adequately represented across the model simulations. Increasing the horizontal resolution of the simulations generally improved the representation of meteorology and air quality but some prognostic variables maintained similar or better performance with coarse resolution simulation. With the high-resolution simulation, PM₁₀ concentration in Brunei had a correlation coefficient around 0.4, and the simulated PM_2.5 level in Singapore had correlation coefficient around 0.9. Whereas, the low-resolution simulation had correlation coefficients around 0.2 and 0.8 for PM₁₀ and PM_2.5 levels at Brunei and Singapore, respectively. Both simulations could not repeat aerosol optical depth (AOD) from reanalysis unless the biomass burning emissions were enhanced. An enhancement factor of 6 with high-resolution simulation gave PM₁₀ and PM_2.5 correlations around 0.6 and 0.9 in Brunei and Singapore respectively.