Environ. Sci. Technol. 2009, 43, 8233–8239
Absorption Coefficient and Site-Specific Mass Absorption Efficiency of Elemental Carbon in Aerosols over Urban, Rural, and High-Altitude Sites in India KIRPA RAM AND M. M. SARIN* Physical Research Laboratory, Ahmedabad - 380 009, India
Received April 17, 2009. Revised manuscript received September 3, 2009. Accepted September 10, 2009.
Temporal and spatial variability in the absorption coefficient (babs, Mm-1) and mass absorption efficiency (MAE, σabs, m2g-1) of elemental carbon (EC) in atmospheric aerosols studied from urban, rural, and high-altitude sites is reported here. Ambient aerosols, collected on tissuquartz filters, are analyzed for EC mass concentration using thermo-optical EC-OC analyzer, wherein simultaneously measured optical-attenuation (ATN, equivalent to initial transmittance) of 678 nm laser source has been used for the determination of MAE and absorption coefficient. At high-altitude sites, measured ATN and surface EC loading (ECs, µg cm-2) on the filters exhibit linear positive relationship (R2 ) 0.86-0.96), suggesting EC as a principal absorbing component. However, relatively large scatter in regression analyses for the data from urban sites suggests contribution from other species. The representative MAE of EC, during wintertime (Dec 2004), at a rural site (Jaduguda) is 6.1 ( 2.0 m2g-1. In contrast, MAE at the two high-altitude sites is 14.5 ( 1.1 (Manora Peak) and 10.4 ( 1.4 (Mt. Abu); and that at urban sites is 11.1 ( 2.6 (Allahabad) and 11.3 ( 2.2 m2g-1 (Hisar). The long-term average MAE at Manora Peak (February 2005 to June 2007) is 12.8 ( 2.9 m2g-1 (range: 6.1-19.1 m2g-1). These results are unlike the constant conversion factor used for MAE in optical instruments for the determination of BC mass concentration. The absorption coefficient also shows large spatiotemporal variability; the lower values are typical of the high-altitude sites and higher values for the urban and rural atmosphere. Such large variability documented for the absorption parameters suggests the need for their suitable parametrization in the assessment of direct aerosol radiative forcing on a regional scale.
1. Introduction Black carbon (BC), produced during incomplete fossil-fuel and biomass combustion processes, is one of the major absorbing particulate species in the atmosphere and is being considered as a driver of the global warming (1, 2). The absorption and scattering properties of aerosols are the key parameters to assess direct aerosol radiative forcing and their climatic impact on a regional to global scale (3-5). The absorption coefficient (babs) is either measured using photoacoustic instruments (6, 7) or more commonly used online * Corresponding author phone: +91 79 26314306; fax: +91 79 26301502; e-mail:
[email protected]. 10.1021/es9011542 CCC: $40.75
Published on Web 09/28/2009
2009 American Chemical Society
filter-based absorption methods (8-12); whereas scattering coefficient (bscat) is mainly inferred from Nephelometer based measurements (11, 12). Nevertheless, the assessment of radiative forcing is associated with large uncertainty arising due to the lack of reliable measurements of these optical parameters (13). The filter-based online absorption measurements are affected by the shadowing and multiple scattering effects (8, 9, 11-13). Also, relevant information on the mixing state of BC in aerosols is essential as the internal mixing leads to further increase in absorption signal (2, 10, 14-16). The measurement of BC mass concentration via optical methods is relatively convenient and rapid but requires knowledge of “site-specific” mass absorption efficiency (MAE or σabs). A wide range of values for σabs (2-25 m2g-1) have been reported in the literature, derived based on independent and simultaneous measurements of EC concentration (by thermal method) and absorption coefficient by optical methods (9, 10, 15, 16). The variability in MAE has been interpreted in terms of source regions, analytical measurement protocols, chemical and optical properties of aerosols at a sampling site (13, 14). Furthermore, aging and atmospheric chemical processing can lead to an internal mixing of BC in aerosols, and thus, increases the MAE through enhancement in absorption signal for the same amount of BC (13, 15). The use of site-specific σabs for the determination of BC mass concentration by optical methods has been suggested (13, 14). However, it is common practice to use a constant value of σabs at a given wavelength. For example, the Aethalometer uses a value of 16.6 m2g-1 (at 880 nm) while particle soot absorption photometer (PSAP) uses a value of 10 m2g-1 to convert measured absorption into BC mass concentration (10). A recent review by Bond and Bergstrom (9) has suggested a value of 7.5 ( 1.2 m2g-1 at 550 nm for σabs for uncoated soot particles. Carbonaceous aerosols in south-Asian region, originating from a variety of anthropogenic sources, are gaining considerable importance because of their potential impact on regional climate (3-5). In this context, systematic measurements of relevant optical parameters (BC mass fraction, babs, bscat, and site-specific σabs) from Indian region are essential. This manuscript reports the first measurement of these parameters from northern India; documenting large temporal and spatial variability in σabs and babs over urban, rural, and high-altitude sites. Our analytical approach makes use of the optical-attenuation (ATN) measured at 678 nm, on thermo-optical EC-OC analyzer, for the simultaneous determination of babs and σabs along with the measurement of EC and OC mass concentrations.
2. Experimental Methods 2.1. Study Sites and Aerosol Sample Collection. Ambient aerosol samples were collected from urban, rural and highaltitude sites during wintertime (December 2004). Hisar (29.2°N, 75.7°E) and Allahabad (25.4°N, 81.9°E) are the two typical urban sites in northern India, mainly influenced by biomass burning emissions during wintertime. Aerosol sampling from rural site, Jaduguda (22.5°N, 85.7°E), dominated by coal-based emissions provide an appropriate location to study their optical and chemical properties. The locations of sampling sites and prevailing wind pattern (during wintertime) are shown in Figure 1 and relevant details are provided in Table 1. The aerosol loading at the two highaltitude sites, Manora Peak (29.4°N, 79.5°E, ∼1950 m above sea level (asl)) and Mt. Abu (24.6°N, 72.7°E, ∼1700 m asl) are dominated by the long-range transport as well as local VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8233
measurement is no more than 7% (except in few cases where it is as high as 13%) and is expressed as relative percentage deviation from n ) 20 measurements. The relevant parameters used in this study, along with the associated measurement errors, have been summarized in Supporting Information (SI) Section A1 and Table S1. The measurement of ATN and methodology used in this study are similar to those employed in filter-based online optical instruments (8, 12). In the latter approach, aerosols are generally collected for a short time on a small filter area at low flow rates and ATN is obtained by measuring the change in transmittance as a function of time. On the contrary, collection of aerosol samples integrated over longer time on a large filter area (as used in this study) minimizes the sample heterogeneity and enhances the ATN signal. The attenuation cross-section (σATN, m2g-1) can be directly obtained from above eq 1 by correcting the measured ATN for shadowing effect (R(ATN); as explained in later section) and dividing by surface EC loading (ECs): σATN(m2g-1) )
FIGURE 1. The typical wind pattern (@10 ms-1) during wintertime (December 2004) at the sampling locations Allahabad and Hisar (urban sites), Mt. Abu and Manora Peak (high-altitude sites) and Jaduguda (rural site). emission sources (17, 18). In addition, an extended sampling was carried out at the two high-altitude sites (Manora Peak: February 2005 to June 2007, Mt. Abu: May 2005 to February 2006). All samples were collected by operating a high-volume sampler at a flow rate of ∼1.0 m3 min-1 and using precombusted (at 500 °C for ∼6 h) tissuquartz filters (Pallflex, 2500 QAT-UP; size: 20 × 25 cm2), integrated for ∼15-20 h in order to meet the analytical requirements for the determination of carbonaceous species and a suit of chemical constituents (17, 18). 2.2. Thermal EC and Optical-Attenuation (ATN) Measurements. Sample aliquots in the form of punches (1.5 cm2) were drawn from main filters and surface EC loading (ECs; in unit of µg cm-2) was ascertained on the EC-OC analyzer (Sunset Laboratory, Forest Grove, OR) using thermal-optical transmittance (TOT) protocol (19, 20). The temperature program used in this study is described in our earlier publication (17, 18) and is similar to the base case temperature program used in the ACE-Asia intercomparison study (19). The analytical instrument (EC-OC analyzer) provides absorbance (equivalent to initial transmittance) at 678 nm by measuring intensities of incident and transmitted light through the filter loaded with aerosols. The absorption signal is represented by optical-attenuation (ATN, a unit less parameter) and is governed by the Beer-Lambert’s law, according to the following equation:
()
ATN ) -100 · ln
I I0
(1)
where I0 is the intensity of incident light and I is the transmitted light through the filter substrate and aerosols. The measured ATN signal for blank filters is zero (n ) 50). Thus, the measured ATN through the sample filter is attributed to the presence of light absorbing carbon (LAC) and is equivalent to in situ EC concentration in aerosols (15). Sciare et al. (20) had reported that absorption measurements performed on PSAP showed good agreement with those measured from Sunset EC-OC analyzer (R2 ) 0.93). Hence, simultaneous measurements of ATN and ECs can be used to determine the absorption coefficient (babs) and mass absorption efficiency of EC (σabs). The uncertainty in the ATN 8234
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 21, 2009
ATN R(ATN) · ECs
(2)
2.3. Determination of Aerosol Absorption Coefficient (babs) and Mass Absorption Efficiency of EC (σabs). The attenuation coefficient (bATN) is calculated from the measured ATN with the help of following equation: bATN )
ATN · A(m2) V(m3)
(3)
whereas absorption coefficient (babs) is related to bATN according to eq 4 as described in the literature (8, 11-13) babs )
bATN ATN A · ) C · R(ATN) V C · R(ATN)
(4)
where, A is the effective filter area (417 × 10-4 m2 for tissuquartz filter used in this study), V is the volume of air sampled (m3). C and R(ATN) are the two empirical factors for correcting the measured absorption due to the multiple scattering and shadowing effects, respectively. The value of C depends on the type of absorbing material, the filter substrate and mixing state of BC in aerosols (8, 9). A value of 2.14 ( 0.21 has been suggested for correction due to the multiple scattering effect for uncoated and externally mixed soot particles collected on tissuquartz filters (same as used in present study) (8, 9). However, much higher values have been reported for internally mixed aerosol particles (e.g., 3.6 ( 0.6 for soot particles coated with organics) (8). This introduces large uncertainty in babs values when filter-based measurements are performed. We have used a value of 2.14 ( 0.21 for C and R(ATN) has been determined using eq 5. The multiple scattering effect is due to accumulation of particles and leads to an enhancement in absorption while the shadowing effect decreases absorption by reducing the optical path length. Weingartner et al. (8) have provided a wavelength dependent parameter f to estimate R(ATN) by fitting linear empirical curve to the observed data set obtained by Aethalometer: R(ATN) )
- ln 10 +1 ( 1f - 1)( lnlnATN 50 - ln 10 )
(5)
where R(ATN) depends on parameter f and decreases with increasing value of f. It is evident from above equation that for lower values of optical-attenuation (ATN e 10%), R(ATN) can be taken as unity. However, for ATN > 10%, R(ATN) values are always less than unity and decreases with increasing ATN. For ATN > 10%, we have used f ) 1.103 during wintertime (December-March) and f ) 1.114 for rest of the seasons to
TABLE 1. Absorption Coefficient (babs) and Site-Specific Mass Absorption Efficiency (σabs) of EC at Different Geographical Locations in Northern India longitude °E
elevation m asl
σabs (m2g-1)
sampling time
location
type
latitude °N
December 04 December 04 December 04 December 04 February 05 to June 07 December 05 to February 06 May 05 to February 06
Jaduguda Hisar Allahabad Manora Peak
rural urban urban high-altitude
22.5 29.2 25.4 29.4
85.7 75.7 81.9 79.5
150 219 123 2000
7 40 19 20
3.1 ( 0.6 8.5 ( 2.2 8.1 ( 1.7 6.0 ( 1.9
69.7 ( 19.6 39.9 ( 9.1 66.1 ( 17.2 12.9 ( 4.6
6.1 ( 2.0 11.3 ( 2.2 11.1 ( 2.6 14.5 ( 1.1
3.4 6.1 7.5 12.7
9.1 16.4 16.8 16.9
Manora Peak
high-altitude
29.4
79.5
2000
47
8.6 ( 2.8
12.2 ( 6.6
12.8 ( 2.9
6.1
19.1
Mt. Abu
high-altitude
24.6
72.7
1700
14
6.1 ( 2.0
8.0 ( 5.5
10.4 ( 1.4
8.1
12.2
Mt. Abu
high-altitude
24.6
72.7
1700
37
6.1 ( 2.0
5.8 ( 4.3
9.8 ( 2.1
3.6
13.7
calculate R(ATN) from eq 5 as per values reported at 660 nm (21). It is noteworthy that use of f value equal to 1.103 or 1.114 leads to maximum change of 2% in R(ATN). The calculated values of 0.85 ( 0.03 for R(ATN) at the sampling sites (Hisar, Allahabad and Jaduguda) are similar (within errors) to that used by Chou et al. (22) for an urban site in Taipei (0.80) wheras R(ATN) is relatively high (0.92 ( 0.04) for the high-altitude sites. The propagated root-sum-square error (RSSE) for the determination of aerosol absorption coefficient (babs) is ∼23%, arising from measurements of ATN, A, V, and corrections due to the multiple scattering and shadowing effects. The attenuation cross-section (σATN) and mass absorption efficiency (σabs) are the two frequently used terms for determination of BC mass concentration via optical methods. Although, both parameters are expressed in same units (m2g-1); σATN accounts for the intrinsic absorption due to BC particles and an additional increase in light absorption due to the multiple scattering effect (C) and is related with σabs by the following equation: σabs )
σATN(m2g-1) babs(Mm-1) ) C EC(µgm-3)
(6)
Assuming an uncertainty of ∼23% in babs and 22% in EC measurements (23), the propagated RSSE in determination of σabs is estimated to be of the order of ∼32% using our approach.
3. Results and Discussion 3.1. Optical-Attenuation (ATN) and Thermal EC Concentration. The analytical data on optical-attenuation (ATN), surface EC loading (ECs; µg cm-2), EC concentration (EC; µg m-3), absorption coefficient (babs, Mm-1; 1Mm-1 ) 10-6 m-1) and mass absorption efficiency (MAE, σabs, m2g-1) for the high-altitude site, Manora Peak, are given in SI Tables S2 and S3. The measured ATN and ECs concentrations at Manora Peak varied from 8 to 134 and 0.3 to 9.3 µg cm-2 over the entire sampling period (February 2005 to June 2007); whereas the two parameters varied from 15 to 95 and 0.5 to 3.5 µg cm-2, respectively for the daily samples collected during wintertime (December 2004). The measured ATN and ECs concentration exhibit a significant linear relationship at Manora Peak (R2 ) 0.96 and 0.86 respectively in Figures 2a and b), indicating the validity of Beer-Lambert’s law and EC as a principal absorbing component in aerosols. However, this linearity does not extend for ECs exceeding 4.5 µg cm-2. Recently, Junker et al. (24) have reported that ATN measured by Aethalometer (range: 22-178) varied linearly with BC surface mass loading (in unit of µg cm-2). In a related study, it has been shown that about 90% of data fall in the linear range for babs
