TL;DR
This study quantifies the mixing state of black carbon (BC) particles emitted from traffic using advanced mass spectrometry. They found that BC particles are not uniform but exist in distinct "HOA-rich" and "BC-rich" types, with about 90% of BC mass residing in the latter. Accurately accounting for this initial mixing state in models significantly impacts calculated aerosol optical properties, specifically increasing single scattering albedo by about 0.1.
What Problem Does This Paper Address?
Black carbon (BC) is a potent absorber of solar radiation and a significant contributor to climate warming. However, its precise climate impact is poorly understood, largely because its optical properties are highly sensitive to its physical and chemical characteristics at the particle level. A key aspect of this is the "mixing state" – how BC is distributed within individual aerosol particles, especially when mixed with other aerosols like organic compounds. Previous studies have struggled to simultaneously quantify the mass-based mixing state of BC and its co-emitted species. This paper aims to fill this gap by providing quantitative measurements of BC mixing state from traffic emissions and using these to assess their impact on aerosol optical properties.
Methods
The researchers employed a soot-particle aerosol mass spectrometer (SP-AMS), equipped with a light-scattering module, to analyze individual particles. This instrument allows for the detection of refractory black carbon (rBC) by measuring its characteristic C+ fragments and simultaneously characterizing associated non-refractory species like hydrocarbon-like organic aerosol (HOA).
Key methodological aspects include:
- Single Particle Analysis: The SP-AMS was used in single-particle mode to identify distinct particle types based on their mass spectra. This allowed for the classification of particles into "HOA-rich" and "BC-rich" categories.
- Ensemble Measurements: For a broader population-level view, ensemble SP-AMS data were analyzed using Positive Matrix Factorization (PMF) to identify major sources and types of rBC-containing aerosols.
- Particle-Resolved Box Modeling: The quantitative mixing state data from the SP-AMS was used to drive a PartMC-MOSAIC particle-resolved aerosol box model. This model simulates the evolution of aerosol populations over 24 hours, allowing for a direct comparison of optical properties under different initial mixing state assumptions.
- Two Simulation Scenarios:
- Uniform Mixing State: All particles are assumed to have the same composition at emission, based on the average mass fraction of BC.
- Measurement-Constrained Mixing State: Particles are assigned compositions reflecting the observed HOA-rich and BC-rich classes.
Datasets
The study utilized data collected from two urban field campaigns in downtown Toronto, Canada:
- Non-roadside study: Air was sampled away from major roadways.
- Roadside study: Air was sampled adjacent to a busy road to capture fresh vehicle emissions.
The SP-AMS measurements from these studies served as the primary dataset for characterizing the mixing state of traffic-emitted aerosols. This data was then used as input for the PartMC-MOSAIC model. No new benchmark datasets were introduced.
Key Insights and Ideas
- Traffic emissions produce distinct BC particle types: The most significant finding is the clear identification of two primary types of fresh BC-containing particles from traffic: "HOA-rich" (low BC mass fraction, <0.1) and "BC-rich" (high BC mass fraction, ~0.8). This challenges a simplified view of uniform BC emissions.
- BC mass is concentrated in "BC-rich" particles: Approximately 90% of the total BC mass is found within the "BC-rich" particle class, highlighting that these particles are the dominant carriers of BC.
- Initial mixing state matters for optical properties: The study demonstrates that accurately representing the initial mixing state of BC particles at emission significantly influences their optical properties, even after a day of atmospheric processing.
- Model sensitivity to mixing state: The particle-resolved model simulations showed that assuming a uniform mixing state leads to an underestimation of single scattering albedo (SSA) compared to using the measured, distinct particle types.
Key Results
- Particle Classification: SP-AMS measurements identified two distinct particle classes: HOA-rich (average BC mass fraction < 0.1) and BC-rich (average BC mass fraction ~0.8).
- BC Mass Distribution: Approximately 90% of the total BC mass was found in the BC-rich particle class.
- Optical Property Differences: The particle-resolved model simulations showed that accurately accounting for the initial mixing state (measurement-constrained case) resulted in an increase in single scattering albedo (SSA) of approximately 0.1 compared to assuming a uniform initial composition. This difference persisted throughout the 24-hour simulation period.
- No significant CCN differences: Despite differences in optical properties, no significant differences in calculated cloud condensation nuclei (CCN) activity were observed, attributed to the similar hygroscopicity of rBC and HOA species.
Limitations and Future Work
- SP-AMS Limitations: The authors acknowledge potential underestimation of BC mass fraction (mfrBC) due to uncertainties in recommended relative ionization efficiencies (RIEs) for organic species and the possibility of incomplete vaporization of coatings.
- Roadside Study Limitation: In the roadside study, only rBC-containing particles were detected by the SP-AMS, while in the non-roadside study, it was possible that HOA-rich particles did not contain detectable rBC ("externally mixed" HOA).
- Mie Theory Assumptions: The study used Mie calculations assuming a core-shell structure for optical property calculations. They acknowledge that real BC-containing aerosols may not always be spherical or exhibit this structure, and that enhancements in BC absorption due to non-absorbing coatings might be overestimated in some cases.
- Future Work: The authors suggest that the approach developed here for quantitative assessment of BC mixing state from traffic can be extended to other sources and used to explore the evolution of mixing state during atmospheric processing. Such measurements are crucial for driving more accurate model assessments of BC's climate impacts on a broader scale.
Summary updated