Evaluation of multi-component ensemble simulation of PM2.5 in the Pearl River Delta region

Sulfate, nitrate, ammonium, organic carbon and black carbon are the key components of PM_2.5, but their simulations are still facing high uncertainty. Exploring the sources of such uncertainty is important for the modeling of PM_2.5 and the understanding of atmospheric chemical processes. This study aims to evaluate and investigate the modeling uncertainty of these aerosols over Pearl River Delta (PRD) region based on Monte Carlo simulations of a Nested Air Quality Prediction Modeling System (NAQPMS) during 2015. Emission inventory as one of the most important uncertainty sources are perturbed according to their uncertainties to derive 50 ensemble simulations of NAQPMS with 15km horizontal resolution. The surface observations of sulfate, nitrate, ammonium, OC and BC from 10 sites in PRD region for one year are used to evaluation the performance of the ensemble mean estimation of the simulations. The results suggested that the ensemble mean could well reproduce the spatial and temporal variations of nitrate, ammonium, OC and BC with the correlation coefficients above 0.74 and their mean bias less than 2μg·m^-3 . However, the model has poor skills in the sulfate modeling with the correlation coefficients 0.26 and remarkable underestimation in winter. Further analysis for such modeling uncertainties suggested that the uncertainties in emissions can explain most of modeling uncertainties for BC and OC. However, the biases in sulfate and ammonium modeling especially during the wintertime are probably caused by the uncertainty in heterogeneous reaction modeling. The above results provide an overall assessment of the uncertainty in inorganic aerosol modeling over PRD region and can serve a basis for its simulation improvement.