Improvements in source apportionment of multiple time-resolved PM_2.5 inorganic and organic speciation measurements...

Area of Science:

• Environmental Science
• Atmospheric Chemistry
• Chemical Engineering

Background:

• Positive Matrix Factorization (PMF) is widely used for source apportionment of particulate matter.
• Incorporating data with varying time resolutions into PMF can lead to mixed factor profiles and high uncertainty.
• Existing PMF models struggle with low time-resolved data, impacting accuracy.

Purpose of the Study:

• To develop and validate a dual-stage PMF modeling procedure with predefined constraints.
• To improve the accuracy of source apportionment for PM2.5 in Taipei.
• To address uncertainties arising from mixed time-resolved data in PMF analysis.

Main Methods:

• A dual-stage PMF modeling procedure with predefined constraints was proposed and implemented.
• PM2.5 inorganic and organic speciation measurements were collected over a year in Taipei, Taiwan.
• The proposed method was compared against a standard PMF approach.

Main Results:

• The dual-stage PMF approach successfully resolved eight distinct and meaningful source profiles, unlike the standard method which produced mixed or unphysical factors.
• Vehicles were identified as the primary source of PM2.5 (26%) and organic carbon (47%).
• Industrial activities were the dominant source for secondary inorganic aerosols: SO4(2-) (27%), NO3(-) (25%), and NH4(+) (31%).

Conclusions:

• The proposed dual-stage PMF method significantly enhances source apportionment accuracy for PM2.5 by resolving complex profiles and reducing uncertainties.
• Vehicle emissions and industrial activities are key sources of air pollution in Taipei, necessitating targeted regulatory actions.
• Further research is needed to account for time-lag effects in gas-to-particle conversion for more accurate secondary aerosol source apportionment.