Aethalometer Measurements of Equivalent Black Carbon in the Arctic

Fate of Aethalometer measurements at the Arctic Sites:
What are the steps forward?

A candidate topic for the Arctic Report Card
Authors of Briefing Paper: 23 April 2013
Sangeeta Sharma, Climate Research Division
Science and Technology Branch
Environment Canada



Introduction
Carbonaceous aerosol (CA), includes black carbon (BC) and organic carbon (OC), constitute a major proportion to the fine ambient particulate matter. BC aerosols play an important role in the climate forcing as they are the most important contributor to the light absorption by aerosols (Jacobson, 2001; Bond et al., 2013; Andreae and Ramanathan, 2013). The origin of CA is in the incomplete combustion of fossil fuels and various types of wood combustion including the wild fires. The definition of black carbon is complicated (Bond et al., 2013) and not as direct due to the fact that BC aerosols change during their transport from source regions to the remote locations. They go through the aerosol growth and ageing process where other chemical components attach to them that may have either absorption or scattering capabilities. This in turn changes absorption capabilities of black carbon. This effect needs to be determined for accurate determination of absorption by black carbon alone. This quantity is important in the calculation of the aerosol radiative forcing which depends on the accurate aerosol extinction measurements (absorption+scattering).


The aethalometer is the earliest and most common method (Hansen et al., 1982) used for the absorption derived real time Equivalent Black Carbon measurements (EBC) and with the longest measurement records at various Arctic locations. This instrument measures absorption by all components of the aerosol besides black carbon over the broad region of visible spectrum and that’s why given a terminology EBC. This measurement is a filter-based method where air is drawn at a sample rate of Q through a sampling area A and increase in attenuation is detected with increasing aerosol loading on the filter. EBC is determined by the following equation:

where σabs is the particle absorption coefficient, αabs is the mass absorption cross section of BC. I0 is the light intensity through clean reference spot on the filter and I is the light transmitted through the sample exposed area on the filter. Values of αabs have been specified by Magee Scientific as 19 m2g-1 for model # AE1 to AE6, 16 m2g-1 for AE16 and 14625/λ m2g-1 for 7λ, AE30 and AE31 where λ is the wavelength. It is realized that EBC and ΔATN are not related linearly due to the following:



1) Enhancement in absorption due to scattering by fibrous filter matrix (instrument artifact).
2) Enhancement in absorption due to embedded scattering aerosols (scattering effect).
3) Enhancement in absorption due to accumulation of particles on the filter (shadowing effect).


As a result, the uncorrected aethalometer measurements are biased higher (Arnott et al., 2005; Weingartner et al., 2003). Several methods have been developed to make such a correction to EBC data (Arnott et al., 2005; Weingartner et al., 2003; Schmid et al., 2006; Virkkula et al., 2007). Collaud Coen developed a new correction method (2010) based on previous four methods of correction due to filter loading and scattering effects. All of these empirical corrections are obtained by comparing aethalometer uncorrected data with the reference method of determining the absorption. In most cases, absorption measured by a photoacoustic spectrometer and/or the difference of extinction (long-path cell) minus scattering (integrating nephelometer) was used as the reference method. Various experiments were performed to include the impact of white scattering aerosols on a clean filter. This warrants a need to apply all these corrections to the aethalometer data to improve on some confidence levels in the EBC measurements. However, we also need to keep in mind how accurate do these absorption measurements have to be in order to satisfy the modeling community considering there are huge uncertainties in the model outputs.

Measurements in the Arctic
There are several observatories in the Arctic (Figure 1) that have been measuring EBC with Aethalometers: Alert (Canada), Barrow (USA), Ny-Ålesund (Norway), Summit (Greenland), Pallas (Finland), Tiksi (Russia) and Station Nord (Greenland). The longest records of EBC exist at Barrow and Alert, while some sites only have a few years of EBC measurements (Table 1). Most of these sites are members of Global Atmospheric Watch program and have obligations towards submitting these data to the World Data Center. Not all sites have a documented protocol for quality control of these measurements, and some provide raw data-sets without any data removal flags. There are several papers published on the trend analysis of EBC data from individual sites (Barrow - Bodhaine et al., 1995; Alert - Sharma et al., 2004; Ny-Ålesund -Eleftheriadis et al., 2009) and combined sites (Barrow, Alert and Ny-Ålesund – Sharma et al., 2006; 2013; Hirdman et al., 2010; Collaud Coen et al, 2013). Sharma et al. and Eleftheriadis et al. used elemental carbon measurements from the thermal technique to determing αabs on a seasonal basis, and adjust the aethalometer measurements using that value. Up to this point, there are no scattering and loading corrections applied to any of these data except for EBC data from Pallas site.
In order to combine the EBC analyses from all 7 sites, there needs to be agreement on common grounds to apply flags to these data-sets so that more meaningful trends and model validations can be obtained.


Approach for use in the Arctic Report Card
The Arctic Report Card is a widely read publication that provides an annual status on key indicators of Arctic Climate. EBC would be a new and important submission to this publication. A typical submission includes an essay placing a climatological index in context for a general science audience and accompanying time series demonstrating how the current year compares to past years. With the long time series from the 7 observatories, we could provide a climatological background of EBC and highlight regional differences. In order to do so, we would need to generate a homogenous, quality checked data product across all participating observatories. It is also important to make the homogenous dataset accessible. The EBAS site does provide access to original, usually uncorrected raw aethalometer data for many of these sites. The IASOA data access portal (using the NOAA Arctic archive) can host an EBC data product developed for each observatory. We suggest the following approach:
1) Raw data from aethalometer should be submitted in Level-0 format to the World Data Center for Aerosols (WDCA, see http://www.gaw-wdca.org/SubmitData/AdvancedDataReporting/Level0/FilterAb...). The WDCA Level-0 format includes provisions for reporting crucial metadata and the results of quality control checks.
Data should be flagged for local contamination (visual inspection and/or wind sector) and for instrument malfunction, using the WDCA flag specifications.
2) Document the level of quality control that has been performed on both the archived measurements and the real-time measurements for a particular year.
– exposed area spot size correction applied when measured as compared to what is in the firmware
– flowmeter calibration, if any, over the years
– identification of local contamination sector from the measurements. The Level-0 data should be flagged for local contamination (visual inspection and/or wind sector) and for instrument malfunction, using the WDCA flag specifications.
– setting for the transmittance when next spot size is changed? Is it factory configured or manually chosen?
3) Application of a suitable correction factor (Collaud Coen et al., 2010) to all Level-0 data to derive a best estimate of the aerosol light absorption coefficient. The same correction scheme should be used for all sites in order to derive a homogeneous data set, which means that the correction scheme should not require simultaneous light scattering data. We recommend using the Collaud Coen et al. (2010, eq. 14b) correction with prescribed values of the parameters Cref, αnew, and ω0 (eq. 14b, 13). The value of Cref should be based on a comparison of the MAAP and Aethalometer at Pallas, which is the only site with a MAAP. The value of αnew=0 (i.e., no scattering data) and ω0 should be chosen based on a long-term average single-scattering albedo for the Arctic aerosol (e.g., Delene and Ogren [2002] reported an average value for Barrow of 0.96).
4) Following the same reasoning for EBC as for light absorption coefficient, the EBC mass concentration should be calculated using the same value of the mass absorption cross-section αabs. This value can be calculated from the EC data at Alert and Ny Ålesund, but the values previously reported by Sharma et al. (2004) and Eleftheriadis et al., (2009) need to be recalculated using the light absorption coefficient from (3).
5) The best estimates from (3) and (4) should be compared with simultaneous measurements by related methods where available (e.g., PSAP, MAAP, and photoacoustic spectrometer for light absorption coefficient, and Single Particle Soot Photometer (SP2) and COSMOS for EBC).
6) The 7λ aethalometer has tendency to separate influence of aerosols from difference source types with their wavelength dependence i.e., brown carbon (Brc) influence from wood combustion sources is more prominent at shorter wavelengths than the fossil fuel influence. This should be explored at various locations.
7) Are there any changes in trends post correction to these data as compared to pre-correction?
8) Need for resources to do these analyses.
9) Use the current year’s measurements to place that year in context of the sites climatology. The time scale over which these comparisons are made will likely be important (annual vs. seasonal vs. monthly).
This is simply a draft approach. Other ideas and approaches are encouraged. It is also recognized that any approach that is adopted should evolve in subsequent years as additional measurements and data products become available.

Potential Drawbacks
The most obvious drawbacks in providing a “chapter” for the Arctic Report Card on the Equivalent Black Carbon (EBC) are the potential errors associated with the measurements. How long does it currently take each site to perform adequate quality control on their data? Does it make sense to proceed using data that hasn’t been quality-controlled?


Sample Highlights and Graphics
Some highlights of the published results in the five papers mentioned earlier indicate that the enhancement on an aethalometer filter could be by a factor of 2 for uncoated soot and a factor of 3.6 for coated soot. The response of an aethalometer decreases as the aerosol loading on the filter increases and this response is non-linear. The correction schemes also depend on the mixing state of the aerosols. Some results from EUSAAR 2009 experiment showed that externally mixed (fresh) aerosols have higher loading correction than internally mixed (aged) aerosol (Thomas Müller, TROPOS Personal Communication).
We used EBC data from Sharma et al., 2013 analysis at Alert, Barrow and Ny-Ålesund to generate some sample graphics to consider for the Arctic Report Card. It is anticipated that similar graphics can be generated for each of the other four Arctic sites, but using different time periods for their “climatologies”. Figure 2 shows time series of EBC data from the three locations. Raw aethalometer data were taken for this analysis after removal of the local contamination sectors. No other corrections were applied to these data. The results from the three sites indicate that there has been a decline in EBC from 1990 values at Alert and Barrow. The levels in 2000s are similar at the three locations. One result that came from this comparison is that there is no change in the EBC concentrations at the three sites even after post 2000 increase of BC emissions in East Asia.
Figure 1: Seven Arctic observatories conducting EBC measurements.
Table 1: Lists and time-line for type of the measurements conducting at 7 Arctic observatories.
Figure 2: Surface BC (EBC) measurements at the three Arctic sites, Alert (82oN, 62.3oW), Barrow (71oN, 156.6oW) and Ny-Ålesund (79oN, 12oE) . There is an overall decline of 40% in BC measurements in the Arctic from 1990 to 2009. In spite of increase in BC FF emissions in the source regions since 2000, there is no increase in the observed surface BC measurements at all three locations

References
Andreae and Ramanathan, Climate’s Dark Forcings, Science, 340, No.6130, 280-281, 2013.
Arnott et al., Towards Aerosol Light-Absorption Measurements with a 7-Wavelength Aethalometer: eValuation with a Photoacoustic Instrument and 3-Wavelength Nephelometer, Aerosol Science and Technology, 39, 17-29, 2005.
Bond et al., Bounding the role of black carbon in the climate system: A Scientific assessment, J. Geophys. Res., 2013.
Bodhaine, B, Aerosol absorption measurements at Barrow, Mauna Loa and the South Pole, J. Geophys. Res, 100(D5), 8967-8975,1995.
Collaud Coen et al., Minimizing light absorption measurements artifacts of the Aethalometer: evaluation of five correction algorithms, Atmos. Meas. Tech., 3, 457-474, 2010.
Collaud Coen et al., Aerosol decadal trends – Part 1: In-situ optical measurements at GAW and IMPROVE stations, Atmos. Chem. Phys., 13, 869-894, 2013.
Delene and Ogren, Variability of Aerosol Optical Properties at Four North American Surface Monitoring Sites, J. Atmos. Sciences, 59, 1135-1150, 2002.
Eleftheriadis,K et al., Aerosol black cron in the European Arctic: Measurements at Zeppelin Station, Ny-Ålesund, Svalbard from 1998-2007, Geophys. Res. Lett., 36, L02809, 5, doi:10.1029/2008GL035741, 2009.
Hansen et al., Real-time measurement of the absorption coefficient of aerosol particles, Appl. Optics, 27, 3060, 1982.
Hirdman et al., Source identification of short-lived air pollutants in the Arctic using statistical analysis of measurement data and particle dispersion model output, Atmos. Chem. Phys., 10, 669-693, 2010.
Jacobson, M.Z., Strong radiative heating due to the mixing state of the black carbon in atmospheric aerosols, Nature, 409 (6821), 695-697, 2001.
Schmid et al., Spectral light absorption by ambient aerosols influenced by biomass burning in the Amazon Basin. I: Comaprison and field calibration of absorption measurement techniques, Atmos. Chem. Phys., 6, 3443-3462, 2006.
Sharma et al., 16-year simulation of Arctic black carbon: transport, source contribution,a nd sensitivity analysis on deposition, J. Geophys. Res/, 118, 1-22, 2013.
Sharma et al., Long-term trends of the black carbon concentrations in the Canadian Arctic, J. Geophys. Res., 109, D15203, doi:10.1029/2003JD004331, 2004.
Sharma et al., Variations and sources of the equivalent black carbon in the high Arcitc revealed by long-term observations at Alert and Barrow: 1989-2003, J. Geophys. Res., 111, D14208, doi:10.1029/2005JD006581, 2006.
Virkkula et al., A simple procedure for correcting loading effects of Aethalometer data, J. of Air and Waste Management, 57, 10, 1214-1222, 2007.
Weingartner et al., Absorption of light by soot particles: determination of the absorptioabsorption coefficient by means of aethalometers, Aerosol Science, 34, 1445-1463, 2003.