Measurements Of Black Carbon Mass Concentrations Research Articles

Multiple PM Low-Cost Sensors, Multiple Seasons’ Data, and Multiple Calibration Models

In this study, we combined state-of-the-art data analysis techniques (machine learning [ML] methods) and data from state-of-the-art low-cost particulate matter (PM) sensors (LCSs) to improve the accuracy of LCS-measured PM2.5 (PM with aerodynamic diameter less than 2.5 microns) mass concentrations. We collocated nine LCSs and a reference PM2.5 instrument for 9 months, covering all local seasons, in Bengaluru, India. Using the collocation data, we evaluated the performance of the LCSs and trained around 170 ML models to reduce the observed bias in the LCS-measured PM2.5. The ML models included (i) Decision Tree, (ii) Random Forest (RF), (iii) eXtreme Gradient Boosting, and (iv) Support Vector Regression (SVR). A hold-out validation was performed to assess the model performance. Model performance metrics included (i) coefficient of determination (R2), (ii) root mean square error (RMSE), (iii) normalised RMSE, and (iv) mean absolute error. We found that the bias in the LCS PM2.5 measurements varied across different LCS types (RMSE = 9–30 µg m-3) and that SVR models performed best in correcting the LCS PM2.5 measurements. Hyperparameter tuning improved the performance of the ML models (except for RF). The performance of ML models trained with significant predictors (fewer in number than the number of all predictors, chosen based on recursive feature elimination algorithm) was comparable to that of the ‘all predictors’ trained models (except for RF). The performance of most ML models was better than that of the linear models. Finally, as a research objective, we introduced the collocated black carbon mass concentration measurements into the ML models but found no significant improvement in the model performance.

Physical and chemical properties of black carbon and organic matter from different combustion and photochemical sources using aerodynamic aerosol classification

Abstract. The physical and chemical properties of black carbon (BC) and organic aerosols are important for predicting their radiative forcing in the atmosphere. During the Soot Aerodynamic Size Selection for Optical properties (SASSO) project and a EUROCHAMP-2020 transnational access project, different types of light-absorbing carbon were studied, including BC from catalytically stripped diesel exhaust, an inverted flame burner, a colloidal graphite standard (Aquadag) and controlled flaming wood combustion. Brown carbon (BrC) was also investigated in the form of organic aerosol emissions from wood burning (pyrolysis and smouldering) and from the nitration of secondary organic aerosol (SOA) proxies produced in a photochemical reaction chamber. Here we present insights into the physical and chemical properties of the aerosols, with optical properties presented in subsequent publications. The dynamic shape factor (χ) of BC particles and material density (ρm) of organic aerosols was investigated by coupling a charging-free Aerodynamic Aerosol Classifier (AAC) with a Centrifugal Particle Mass Analyzer (CPMA) and a Scanning Mobility Particle Sizer (SMPS). The morphology of BC particles was captured by transmission electron microscopy (TEM). For BC particles from the diesel engine and flame burner emissions, the primary spherule sizes were similar, around 20 nm. With increasing particle size, BC particles adopted more collapsed/compacted morphologies for the former source but tended to show more aggregated morphologies for the latter source. For particles emitted from the combustion of dry wood samples, the χ of BC particles and the ρm of organic aerosols were observed in the ranges 1.8–2.17 and 1.22–1.32 g cm−3, respectively. Similarly, for wet wood samples, the χ and ρm ranges were 1.2–1.85 and 1.44–1.60 g cm−3, respectively. Aerosol mass spectrometry measurements show no clear difference in mass spectra of the organic aerosols in individual burn phases (pyrolysis or smouldering phase) with the moisture content of the wood samples. This suggests that the effect moisture has on the organic chemical profile of wood burning emissions is through changing the durations of the different phases of the burn cycle, not through the chemical modification of the individual phases. In this study, the incandescence signal of a Single Particle Soot Photometer (SP2) was calibrated with three different types of BC particles and compared with that from an Aquadag standard that is commonly used to calibrate SP2 incandescence to a BC mass. A correction factor is defined as the ratio of the incandescence signal from an alternative BC source to that from the Aquadag standard and took values of 0.821 ± 0.002 (or 0.794 ± 0.005), 0.879 ± 0.003 and 0.843 ± 0.028 to 0.913 ± 0.009 for the BC particles emitted from the diesel engine running under hot (or cold idle) conditions, the flame burner and wood combustion, respectively. These correction factors account for differences in instrument response to BC from different sources compared to the standardised Aquadag calibration and are more appropriate than the common value of 0.75 recommended by Laborde et al. (2012b) when deriving the mass concentration of BC emitted from diesel engines. Quantifying the correction factor for many types of BC particles found commonly in the atmosphere may enable better constraints to be placed on this factor depending on the BC source being sampled and thus improve the accuracy of future SP2 measurements of BC mass concentrations.

Changes in black carbon and PM2.5 in Tokyo in 2003-2017.

Black carbon (BC) particles cause adverse health effects and contribute to the heating of the atmosphere by absorbing visible solar radiation. Efforts have been made to reduce BC emissions, especially in urban areas; however, long-term measurements of BC mass concentration (MBC) are very limited in Japan. We report MBC measurements conducted in Tokyo from 2003 to 2017, showing that MBC decreased by a factor of 3 from 2003 to 2010 and was stable from 2010 to 2017. Fine particulate concentrations (PM2.5) decreased by a much smaller factor during 2003–2010. The diurnal variations of BC size distributions suggest that the BC in Tokyo originates mainly from local sources, even after 2010. Our three-dimensional model calculations show that BC from the Asian continent contributes a small portion (about 20%) of the annual average MBC in the Kanto region of Japan, which includes Tokyo. This indicates that continued reduction of BC emissions inside Japan should be effective in further decreasing MBC.

Comparison of Experimental and Modeled Absorption Enhancement by Black Carbon (BC) Cored Polydisperse Aerosols under Hygroscopic Conditions

The quantification of the radiative impacts of light absorbing ambient black carbon (BC) particles strongly depends on accurate measurements of BC mass concentration and absorption coefficient (β(abs)). In this study, an experiment has been conducted to quantify the influence of hygroscopic growth of ambient particles on light absorption. Using the hygroscopic growth factor (i.e., Zdanovskii-Stokes-Robinson (ZSR) approach), a model has been developed to predict the chemical composition of particles based on measurements, and the absorption and scattering coefficients are derived using a core-shell assumption with light extinction estimates based on Mie theory. The estimated optical properties agree within 7% for absorption coefficient and 30% for scattering coefficient with that of measured values. The enhancement of absorption is found to vary according to the thickness of the shell and BC mass, with a maximum of 2.3 for a shell thickness of 18 nm for the particles. The findings of this study underline the importance of considering aerosol-mixing states while calculating their radiative forcing.

Volatility of elemental carbon

A volatility technique whereby aerosol particles are heated to the relatively high temperature of 860 °C is used to infer the presence of elemental carbon in polluted air masses in the vicinity of the west coast of Ireland. The volume of elemental carbon for submicrometre sized particles contained in the aerosol is estimated from the fall off in number concentration at a critical onset temperature of about 730–735 °C, as also obtained for laboratory carbon ink aerosol. The technique permits determination of the elemental carbon volume percentage of the total fine aerosol volume, and an estimation of the abundance of elemental carbon contained within the black carbon fraction of the atmospheric aerosol. Supplementary black carbon mass concentration measurements were obtained using a thermal method and an aethalometer absorption method. The work suggests that elemental carbon can be identified using the volatility technique and that it can yield size‐segregated information on the fraction of elemental carbon in atmospheric aerosol.