利用ME-2模型提升PM2.5源解析效果

为探讨ME-2模型控制旋转对传统PMF模型源解析效果的提升作用,于2017年9月10日~2018年8月29日在深圳北部某工业区开展PM_2.5采样,共获得153套样品.对PM_2.5中31种化学组分进行了分析,筛选出17个物种输入模型运算.2018年深圳北部工业区大气PM_2.5年均浓度为32.3 μg/m³,利用PMF模型初步识别出9个因子,分别为二次硫酸盐、二次硝酸盐、老化海盐、土壤扬尘、工业排放、燃煤、生物质燃烧、船舶排放和机动车,PMF输出结果中"混合因子"问题显著.基于PMF解析结果及获得的先验信息,在ME-2模型中建立4个限制源谱进一步解析,结果表明,与PMF模型相比,ME-2结果的示踪物在源中分配更集中,对示踪物浓度与相应源贡献的时间序列也提供了更好的拟合效果.二次硝酸盐、老化海盐、工业排放源在PMF模型中被高估了9%~51%,而二次硫酸盐、燃煤和生物质燃烧源被低估了19%~40%.本研究中ME-2解析结果比PMF更具有环境和统计学意义,为污染防治提供了更精确的控制指向.

Improved PM2.5 source apportionment results using the multilinear engine (ME-2) model

In order to test the improvement of source apportionment of PM_2.5 using the ME-2 model on unconstrained PMF analyses, a total of 153 daily PM_2.5 samples were collected continuously from September 10, 2017 to August 29, 2018 at an industrial site in northern Shenzhen. The concentrations of 31 chemical compositions were determined, 17 of which were selected for model iteration. The annual mean concentration of PM_2.5 at the industrial site was 32.3μg/m³, and nine sources of PM_2.5 were identified by a prior PMF run, which were secondary sulfate, secondary nitrate, aged sea salt, soil dust, industrial emissions, coal combustion, biomass burning, ship emissions and vehicle emissions. However, the apportionment with PMF yielded mixed factors. Based on the PMF results and priori information obtained, four constrained factor profiles were input into the ME-2 model as a rotational control technique for model simulation. Compared to the PMF solution, tracers were more concentrated in source profiles of ME-2 results, and the ME-2 iteration provided a more significant fitting for time series of tracer concentrations and corresponding source contributions. Secondary nitrate, aged sea salt and industrial emissions were overestimated by 9% to 51% in the PMF results, while secondary sulfate, coal combustion and biomass burning were underestimated by 19% to 40%. The ME-2 result was found to be more environmentally and statistically significant than that of PMF and it could provide a more accurate scientific basis for pollution prevention and control at the same time.