Filter-based Measurement and Light Absorption Enhancement of Black Carbon in Urban Environment
城市環境中黑炭氣溶膠的膜法測量和光吸收增強的研究
Student thesis: Doctoral Thesis
Author(s)
Related Research Unit(s)
Detail(s)
Awarding Institution | |
---|---|
Supervisors/Advisors |
|
Award date | 28 Mar 2019 |
Link(s)
Permanent Link | https://scholars.cityu.edu.hk/en/theses/theses(f2b66f91-de25-415a-a098-400ae675e35c).html |
---|---|
Other link(s) | Links |
Abstract
Black carbon (BC) has a unique and important role in the Earth’s climate system, while there are large differences between estimates of its climate effect in many studies. One key factor in the radiative transfer model is the light absorption enhancement induced by other aerosol composition or coatings, which tend to be internally mixed with BC as the core, whereas the non-absorbing shell or coatings can refract light towards BC, known as “lensing” effect, which amplifies the BC light absorption. This research studied the impact of coatings on the BC light absorption in different urban environment. Before that, filter-based BC measurement was explored to ensure the measured accuracy, and physiochemical properties of aerosol in different environment were characterized. The dissertation consists of following three integrated components.
First, a comparison for different filter-based BC measurement instrument was conducted. We found that all these filter-based BC instruments can’t avoid filter loading effect even for the most advanced Magee model AE33 with dual-spot technique. The comparison experiment was conducted during Chinese New Year. Result showed that BC peaks at the night of the Chinese New Year’s Eve and first day of New Year, with BC mass concentration 6 times higher than normal day due to new year celebration. Multiple peaks were then analyzed corresponding to people’s celebration activities. Significant filter loading effect were observed during analysis. A generic Python program with two correction algorithms (bin and gap method) was developed to determine the filter loading parameter K value and compensate the raw data. Results showed the absolute K value generally ranged from 0.0001 to 0.01. The bin and gap method produced different slightly K values for each instrument. The K values varied between different instruments due to the usage of different filter tape and optical system design. Therefore, care must be taken when comparing the K values. Besides, higher K was observed for each instrument during Chinese New Year peak period than normal days due to aged aerosol of long-rang transport from mainland China. After correction, AE33, AE42, AE51 and OT 21 showed rather closer concentrations, while low concentration was still observed for Met One BC1054 attributed to the application of higher multiple scattering parameter. Absorption Ångström Exponent (AAE) were also calculated separately during New Year peak period and normal days. Larger AAE during New year peak were observed, further implicating the dominance of non-local emission. Mass attenuation cross section (MAC) were also calculated with two different EC determination protocol NIOSH and IMPROVE. Result showed that MAC derived from NIOSH protocol was 1.6 times higher than IMPROVE, while MAC by IMPROVE was nearly identical to that from manufacturer.
Second, chemical characteristics of PM1 aerosol were investigated in two sites, suburban Yuen Long (YL) and Mong Kok (MK) roadside environment. Both sites exhibited similar chemical composition, dominated with organics (45% - 47%), and sulfate (21% - 26%), followed by ammonium (10%), black carbon (9%) and nitrate (7% - 8%). PMF Analysis was carefully applied to organic aerosol (OA) source apportionment. In the MK roadside site, hydrocarbon-like OA (HOA) (28%), low-volatility oxygenated OA (LVOOA) (39%) and semi-volatile oxygenated OA (24%) were apportioned, respectively. While four factors were apportioned in suburban YL site, i.e., LVOOA (42%), HOA (22%), SVOOA (12%) and OOA-3 (24%). BC mass concentration in MK roadside was much higher than suburban YL site. Implication of the age and mixing state of BC-containing particles were analyzed by Absorption Angström exponent (AAE) and filter loading parameter K value. Result showed that higher AAE and lower K were observed in suburban YL site, indicative of more aged aerosol and thus more coating fractions around BC.
Third, in addition to simultaneous measurement of BC and non-refractory PM1 aerosol chemical compositions performed above, a thermodenuder (TD) was employed to remove the semi-volatile coatings at different heating temperatures to explore the determinative coating fraction that affect BC light absorption. Light absorption enhancement (Eabs) of BC by semi-volatile coatings at seven wavelengths was evaluated. In the suburban YL site, result showed overall Eabs was ~1.6 measured at 280℃, whereas Eabs increased very little from 50 ℃ (1.02 – 1.04) to 200 ℃ (1.13 – 1.20). The lensing-effect-related Eabs was probably attributed to the presence of ammonium and sulfate. While in MK roadside site, a lower Eabs of ~1.3 was observed with heating at 280℃, implying the coatings could have different lensing effect in different environment where particle mixing state can be different. Furthermore, Mie model was applied to simulate the magnitude of ensemble coating fraction assuming a core-shell mixing state. Result showed BC-containing particles in suburban have higher coating fraction (105% on average) than roadside (57% on average) which provide quantitative estimation on the BC mixture.
First, a comparison for different filter-based BC measurement instrument was conducted. We found that all these filter-based BC instruments can’t avoid filter loading effect even for the most advanced Magee model AE33 with dual-spot technique. The comparison experiment was conducted during Chinese New Year. Result showed that BC peaks at the night of the Chinese New Year’s Eve and first day of New Year, with BC mass concentration 6 times higher than normal day due to new year celebration. Multiple peaks were then analyzed corresponding to people’s celebration activities. Significant filter loading effect were observed during analysis. A generic Python program with two correction algorithms (bin and gap method) was developed to determine the filter loading parameter K value and compensate the raw data. Results showed the absolute K value generally ranged from 0.0001 to 0.01. The bin and gap method produced different slightly K values for each instrument. The K values varied between different instruments due to the usage of different filter tape and optical system design. Therefore, care must be taken when comparing the K values. Besides, higher K was observed for each instrument during Chinese New Year peak period than normal days due to aged aerosol of long-rang transport from mainland China. After correction, AE33, AE42, AE51 and OT 21 showed rather closer concentrations, while low concentration was still observed for Met One BC1054 attributed to the application of higher multiple scattering parameter. Absorption Ångström Exponent (AAE) were also calculated separately during New Year peak period and normal days. Larger AAE during New year peak were observed, further implicating the dominance of non-local emission. Mass attenuation cross section (MAC) were also calculated with two different EC determination protocol NIOSH and IMPROVE. Result showed that MAC derived from NIOSH protocol was 1.6 times higher than IMPROVE, while MAC by IMPROVE was nearly identical to that from manufacturer.
Second, chemical characteristics of PM1 aerosol were investigated in two sites, suburban Yuen Long (YL) and Mong Kok (MK) roadside environment. Both sites exhibited similar chemical composition, dominated with organics (45% - 47%), and sulfate (21% - 26%), followed by ammonium (10%), black carbon (9%) and nitrate (7% - 8%). PMF Analysis was carefully applied to organic aerosol (OA) source apportionment. In the MK roadside site, hydrocarbon-like OA (HOA) (28%), low-volatility oxygenated OA (LVOOA) (39%) and semi-volatile oxygenated OA (24%) were apportioned, respectively. While four factors were apportioned in suburban YL site, i.e., LVOOA (42%), HOA (22%), SVOOA (12%) and OOA-3 (24%). BC mass concentration in MK roadside was much higher than suburban YL site. Implication of the age and mixing state of BC-containing particles were analyzed by Absorption Angström exponent (AAE) and filter loading parameter K value. Result showed that higher AAE and lower K were observed in suburban YL site, indicative of more aged aerosol and thus more coating fractions around BC.
Third, in addition to simultaneous measurement of BC and non-refractory PM1 aerosol chemical compositions performed above, a thermodenuder (TD) was employed to remove the semi-volatile coatings at different heating temperatures to explore the determinative coating fraction that affect BC light absorption. Light absorption enhancement (Eabs) of BC by semi-volatile coatings at seven wavelengths was evaluated. In the suburban YL site, result showed overall Eabs was ~1.6 measured at 280℃, whereas Eabs increased very little from 50 ℃ (1.02 – 1.04) to 200 ℃ (1.13 – 1.20). The lensing-effect-related Eabs was probably attributed to the presence of ammonium and sulfate. While in MK roadside site, a lower Eabs of ~1.3 was observed with heating at 280℃, implying the coatings could have different lensing effect in different environment where particle mixing state can be different. Furthermore, Mie model was applied to simulate the magnitude of ensemble coating fraction assuming a core-shell mixing state. Result showed BC-containing particles in suburban have higher coating fraction (105% on average) than roadside (57% on average) which provide quantitative estimation on the BC mixture.
- Black carbon, Aerosols, Light absorption enhancement, BC coatings