Coupling chemical transport model source attributions with positive matrix factorization: application to two IMPROVE...

Area of Science:

• Environmental Chemistry
• Atmospheric Science
• Air Quality Modeling

Background:

• Accurate source apportionment of fine particle carbon is crucial for understanding air quality.
• Traditional receptor models like positive matrix factorization (PMF) have limitations in source identification.
• Chemical transport models (CTM) offer source-specific predictions but can have uncertainties.

Purpose of the Study:

• To develop and evaluate a novel receptor-oriented hybrid model integrating CTM and PMF for improved source contribution analysis.
• To assess the model's ability to differentiate carbon sources, including wildfires and biogenic emissions.
• To compare the predictive accuracy of the hybrid model against standalone PMF and CTM approaches.

Main Methods:

• Developed a hybrid model by incorporating CTM source contributions into the PMF receptor model framework.
• Utilized the Multilinear Engine (ME-2) to adjust the influence of CTM and PMF through a weighting parameter.
• Applied the hybrid model to IMPROVE (Interagency Monitoring of Protected Visual Environments) data from Montana sites (2006-2008).

Main Results:

• The hybrid model successfully separated major wildfire carbon contributions from minor biogenic sources.
• Total carbon (TC) predictions from the hybrid model exhibited a lower cross-validated Root Mean Square Error (RMSE) than PMF or CTM alone.
• Identified two additional minor features: a soil-derived source with summer influence and a sulfate/nitrate-enriched source with sporadic contributions.

Conclusions:

• The receptor-oriented hybrid model offers enhanced source apportionment capabilities, identifying features missed by traditional PMF.
• This integrated approach improves the accuracy of total carbon source predictions in atmospheric studies.
• The methodology provides a robust framework for analyzing complex air pollution source contributions.