Data_Sheet_1_Improving Source Apportionment of Urban Aerosol Using Multi-Isotopic Fingerprints (MIF) and Positive Matrix Factorization (PMF): Cross-Validation and New Insights.PDF

Urban air pollution is a matter of concern due to its health hazards and the continuous population growth exposed to it at different urban areas worldwide. Nowadays, more than 55% of the world population live in urban areas. One of the main challenges to guide pollution control policies is related to pollutant source assessment. In this line, U.S. Environmental Protection Agency's Positive Matrix Factorization (EPA-PMF) has been extensively employed worldwide as a reference model for quantification of source contributions. However, EPA-PMF presents issues associated to source identification and discrimination due to the collinearities among the source tracers. Multi-Isotopic Fingerprints (MIF) have demonstrated good resolution for source discrimination, since urban sources are characterized by specific isotopic signatures. Source quantification based on total aerosol mass is the main limitation of MIF. This study reports strategies for PMF and MIF combination to improve source identification/discrimination and its quantification in urban areas. We have three main findings: (1) cross-validation of PMF source identification based on Pb and Zn isotopic fingerprints, (2) source apportionment in the MIF model for total PM mass, and (3) new insights into potential Zn isotopic signatures of biomass burning and secondary aerosol. We support future studies on the improvement of isotopic fingerprints database of sources based on diverse elements or compounds to boost advances of MIF model applications in atmospheric sciences.

城市空气污染因其健康危害以及全球各城区持续增长的暴露人口而备受关注。目前，全球超55%的人口居住于城市区域。指导污染管控政策的核心挑战之一，与污染物源解析密切相关。在此研究方向下，美国环境保护署（U.S. Environmental Protection Agency）的正矩阵分解法（Positive Matrix Factorization, EPA-PMF）已在全球范围内被广泛用作源贡献量化的参考模型。然而，由于源示踪剂间存在共线性问题，EPA-PMF在源识别与区分方面存在局限。多同位素指纹图谱（Multi-Isotopic Fingerprints, MIF）因城市污染源具有特定同位素特征，已被证实具备优异的源区分分辨率。基于总气溶胶质量的源量化，是MIF的主要局限性。本研究提出了PMF与MIF的联用策略，以提升城市区域的污染源识别/区分能力及其量化精度。本研究取得三项核心发现：（1）基于铅（Pb）与锌（Zn）同位素指纹图谱的PMF源识别交叉验证；（2）针对总颗粒物质量的MIF模型源解析；（3）关于生物质燃烧与二次气溶胶潜在锌同位素特征的全新认知。本研究可为未来相关研究提供支撑：通过基于多种元素或化合物的污染源同位素指纹图谱数据库优化，推动MIF模型在大气科学领域的应用发展。