| 基于颗粒物化学组分粒径分布特征的源解析方法构建与评估 |
| 摘要点击 2125 全文点击 721 投稿时间：2019-07-19 修订日期：2019-08-28 |
| 查看HTML全文
查看全文 查看/发表评论 下载PDF阅读器 |
| 中文关键词 大气颗粒物 多粒径因子解析 化学组分 粒径分布 敏感性 |
| 英文关键词 particulate matter size-resolved factor analysis chemical constituent size distribution sensitivity |
|
| 中文摘要 |
| 大气颗粒物污染来源解析可以为大气污染防控提供科学支撑.大部分颗粒物源解析研究是以化学组分为基础，此外，粒径特征也是颗粒物的重要属性之一，通过分析化学组分的粒径特征，有助于提高解析结果准确性.针对目前组分粒径分布信息利用不充分的问题，在三维多粒径因子分析模型（ABB）基础上，将已知源类标识组分粒径分布信息作为约束限制，构建基于组分粒径分布特征的多粒径源解析模型（SDABB）.通过模拟数据集对模型进行了评估，探究了SDABB模型对源谱共线性和源贡献粒径分布相似性的敏感程度.结果表明，ABB模型对源谱共线性和源贡献粒径分布相似性较敏感，将粒径分布规律纳入后的SDABB模型，两种情景效果均明显变优，即SDABB模型可以较好解析共线性源类，并且对贡献粒径分布相似性不敏感. |
| 英文摘要 |
| The analysis of the sources of atmospheric particulate pollution can provide scientific support for the prevention and control of air pollution. Most particulate matter (PM) source analysis studies are based on the chemical composition of PM. In addition, particle size characteristics are also one of the important properties of PM. The accuracy of analytical results can be improved by analyzing the particle size characteristics of chemical components. In this study we aim to to solve the problem of insufficient utilization of component particle size information by using a the three-dimensional multi-particle size factor analysis model (ABB), where the particle size distribution of marked components is regarded as the constraint limit, and a multi-particle size source analytical model (SDABB) based on the characteristics of the components particle size distribution is constructed. The sensitivity of the SDABB model to the collinearity of the source spectrum and the similarity of the particle size distribution of the source contributions are investigated by evaluating the model through the simulation of the data set. The results showed that the ABB model was sensitive to the collinearity of the source spectrum and to the similarity of the particle size distribution of the source contributions. When particle size distribution rules were incorporated into the SDABB model, the effects of the two scenarios were significantly improved, that is, the SDABB model was able to better analyze collinear source spectrum and was insensitive to the similarity of the contribution particle size distribution. |
