Related Experiment Video
Updated: Aug 4, 2025
Composition and Distribution Analysis of Bioaerosols Under Different Environmental Conditions
Published on: January 7, 2019
Source apportionment of fine particulate matter at a megacity in China, using an improved regularization supervised
1State Environmental Protection Key Laboratory of Urban Ambient Air Particulate Matter Pollution Prevention and Control, Tianjin Key Laboratory of Urban Transport Emission Research, College of Environmental Science and Engineering, Nankai University, Tianjin 300350, PR China; CMA-NKU Cooperative Laboratory for Atmospheric Environment-Health Research (CLAER), College of Environmental Science and Engineering, Nankai University, Tianjin 300350, PR China.
This study introduces a supervised Positive Matrix Factorization (PMF) model for accurate atmospheric particulate matter source apportionment. The improved method effectively identifies pollution sources and their contributions using prior knowledge.
Area of Science:
- Environmental Science
- Atmospheric Chemistry
- Data Science
Background:
- Atmospheric particulate pollution requires effective source apportionment for mitigation.
- Positive Matrix Factorization (PMF) is a common but challenging source apportionment model.
- Integrating prior knowledge into PMF can improve result reliability.
Purpose of the Study:
- To develop an improved source apportionment method using a regularization supervised PMF (RSPMF) model.
- To leverage actual source profiles to guide the factor profile identification process.
- To achieve rapid and automatic identification of source categories and quantification of contributions.
Main Methods:
- Proposed a regularization supervised PMF (RSPMF) model.
- Utilized actual source profiles to guide the RSPMF factor profiles.
- Applied the RSPMF model to high-resolution online datasets.
Main Results:
- The RSPMF model successfully identified seven factors closely matching actual source profiles.
- Source contribution estimates from RSPMF agreed well with EPA-PMF.
- Key contributors included secondary nitrate (26%), secondary sulfate (23%), and coal combustion (18%).
- RSPMF demonstrated good generalizability across different pollution episodes.
Conclusions:
- The supervised PMF model effectively embeds prior knowledge for more reliable source apportionment.
- RSPMF offers a superior approach for accurate and timely identification of particulate matter sources.
- This method enhances the understanding and management of atmospheric particulate pollution.
