Better Measurements of Carbon Aerosol Help Study Climate Effects
[Part 1 of this series, in the Spring 2004 issue of EETD News, examined the contributions of Tihomir Novakov and his research group at Lawrence Berkeley National Laboratory (Berkeley Lab) measuring carbon aerosol particles in the atmosphere and elucidating the significance of these particles. Part 2 describes continuing research at Berkeley Lab to improve our understanding of carbon aerosol particles and our ability to measure their mass and light-absorbing effects.]
Now that researchers in the atmospheric sciences community have come to believe, thanks to the work of Berkeley Lab's Tihomir Novakov and his colleagues, that carbon particles play a significant role in the atmosphere, it is important to understand the history of these particles and to measure their concentrations with more accuracy than has been possible in the past. Among the discoveries in which Novakov and fellow researchers have recently participated is that black carbon (BC) may be contributing to climate change by absorbing the sun's heat and thus helping to warm the atmosphere.
In May 2003, Novakov, along with Makiko Sato, James Hansen, and seven other researchers at the National Aeronautics and Space Administration (NASA) Goddard Institute for Space Studies, published a paper inferring the global concentration of BC in the atmosphere based on data from the Aerosol Robotic Network (AERONET) of 250 sunphotometers, which measures the optical depth of the atmosphere around the world.
The researchers found that, to be consistent with AERONET data, the concentration of BC in the atmosphere had to be two to four times greater than climatologists believed (Figure 1). This higher concentration suggested that BC must be having a greater effect on climate change than previously thought; BC would contribute to climate change by increasing the amount of the sun's heat absorbed by the atmosphere. The news of this research attracted considerable attention from the media.
128 Years of Black Carbon
Working with colleagues Tom Kirchstetter, Jonathan Sinton, and Jayant Sathaye of Berkeley Lab, as well as Hansen, Sato, and V. Ramanathan of the Scripps Institution of Oceanography, Novakov has also documented historical changes in atmospheric BC concentrations from 1875 to the present in the six regions of the world that account for most of today's BC aerosol emissions: China, India, the former Soviet Union, Germany, the United Kingdom, and the United States. Using the historical record of coal and transportation fuel burning, the group estimated how much BC might have been emitted to the atmosphere. Calculations were based on emissions rates from today's power plants and combustion engines and assumptions of past emissions rates.
The group's preliminary estimate of historical atmospheric BC concentrations shows a large increase during the second half of the 20th century. BC concentrations increased rapidly during the late 1800s, leveled off during the first half of the 1900s, and then began to accelerate during the past 50 years (Figure 2). Industrialization in China and India contributed a substantial portion of this recent increase.
Climate scientists now are using the AERONET and historical data to improve computer models' representation of BC's effects on climate change.
Improvements in Aerosol Sampling
Researchers would like global atmospheric carbon measurements to be as accurate as possible so that computer models can accurately estimate BC's climate effects. Since 1999, Kirchstetter, a scientist in the Environmental Energy Technologies Division (EETD) who works with Novakov, has been honing the accuracy of current sampling methods for both BC and organic carbon (OC).
The usual way to measure the mass concentration of carbon particles in the air is to force air through a quartz filter for a period of time and then heat the filter to drive off volatile organic compounds and combust the BC. Finally, the evolved carbon is oxidized over a catalyst to form carbon dioxide (CO2), a gas that is easy to measure. "The evolved gas analysis [EGA] system we use for this purpose was developed and refined at Berkeley Lab," says Kirchstetter. The instrument (Figure 3) measures the amount of CO2 released as the temperature rises, creating a thermogram, which shows the mass of the carbon in the sample and can be used to distinguish between black and organic carbon.
Kirchstetter doesn't even have to leave the building to take a sample of Berkeley air because a stack from his lab connects to the roof where an air sampler is located (Figure 4). The equipment was built by Richard Schmidt, who has been a key member of Novakov's group since its beginnings in the 1970s. Kirchstetter notes that "Schmidt fabricates almost all of the instrumentation we use to do research."
"Measuring carbon aerosols in the air is a tricky business," Kirchstetter explains. "One problem is that the carbon content on the filter doesn't always reflect the carbon particle concentration in the atmosphere because of sampling artifacts. Filters adsorb gases, including organic carbon." Kirchstetter developed a simple method of correcting for adsorbed OC gas by putting a second filter behind the first in the EGA measurement setup. The second filter measures only OC gas in the air sample because carbon soot has been trapped in the first filter; the thermogram of the second filter can be subtracted from that of the first to increase the accuracy of the final measurement.
Kirchstetter's method for correcting for OC gas adsorption has increased the accuracy of atmospheric carbon measurements, but there are still problems to solve. For example, different laboratories can measure the same air sample and get different results because of sampling and analytical artifacts other than OC gas adsorption. To address this problem, Kirchstetter and colleague Lara Gundel are now working with other labs to develop "more robust sampling and analytic procedures for black and organic carbon."
From Rooftop to High-altitude Sampling
Kirchstetter has participated in a variety of air-sampling experiments around the world in recent years.
In 2000, he joined a multinational group of scientists in the South African Regional Science Initiative (SAFARI) program and spent five weeks flying on a research aircraft, along with Peter Hobbs of the University of Washington, measuring carbon aerosols in the atmosphere. "We often saw smoke plumes from burning savannah," he says. "Some were prescribed burns; some were natural fires." SAFARI provided a wealth of data from southern Africa about the prevalence of atmospheric carbon particles from burning biomass.
Kirchstetter has also participated in the Indian Ocean Experiment, INDOEX, working with colleagues from the University of Puerto Rico to analyze samples from the Indian Ocean where pollutants drift in from the Indian subcontinent. Pollutants can have a larger effect on regional climate than on overall global climate; on the Asian continent, many regional-scale brown clouds are produced by burning of forests, diesel fuel, and coal. One such cloud drifts south from Asia over the Indian Ocean. The brown haze is implicated in regional climate effects such as intensified periods of drought and rainfall, as demonstrated in climate simulations by Surabi Menon, Kirchstetter's colleague in EETD's Atmospheric Sciences Department. Kirchstetter is one of a number of scientists working with V. Ramanathan of the Scripps Institution of Oceanography, analyzing samples for the Atmospheric Brown Cloud (ABC) project, which was established to examine the climate effects of these clouds.
As part of the Mega Cities project, developed by Mario Molina of Massachusetts Institute of Technology, Kirchstetter will analyze air samples collected over large metropolitan areas, including Mexico City, to understand how aerosols and other pollutants in brown clouds affect large cities.
Organic Carbon-Contributing to Climate Change?
Researchers are now taking a close look at the OC component of carbon aerosols to determine whether, for example, differences between smoke from biomass burning and smoke from diesel fuel have implications for climate change. "In addition to the mass concentrations of these particles," says Kirchstetter, "we're studying their optical properties, so they can be represented realistically in climate models."
Recently, Kirchstetter has found evidence that OC from biomass burning ("wood smoke") behaves differently from that produced by diesel-fuel burning. Using a specially designed device, the multiple wavelength light transmission instrument (MULTI) built by Dick Schmidt, Kirchstetter and Novakov are studying a range of wavelengths of light and their absorption by OC from both wood smoke and diesel combustion. The MULTI consists of a group of light-emitting diodes (LEDs), each emitting a narrow region of the electromagnetic spectrum, and a detector (Figure 5). The MULTI incorporates elements of the aethalometer (another instrument for measuring BC, also developed at Berkeley Lab and described in Part 1 of this series) and commercial optical spectrometers. When a filter sample is placed between the LEDs and the detector, the MULTI produces a plot of the amount of light absorbed by the carbon aerosol in the filter at different wavelengths. Heating the filter removes the carbon aerosol. The sample can then be measured again and compared to the first results to show how the light absorption is affected by the presence of the carbon aerosol.
Kirchstetter has made numerous measurements with the MULTI, some of samples from his lab's roof sampler (representing ambient air in Berkeley), some of samples from the SAFARI project described above (mostly biomass smoke), and some of air drawn from the Caldecott Tunnel, a portion of Highway 24 heavily traveled by drivers in the San Francisco Bay Area (carbon aerosols in these samples are mainly from diesel smoke).
"Most climate models include black carbon as the only light-absorbing aerosol species," says Kirchstetter. "Organic carbon is assumed to be purely scattering, not absorbing." Although OC is not currently considered a contributor to global warming, "there is some indication that you can produce non-black carbon particles that are light-absorbing."
Kirchstetter and Novakov are studying the spectral dependence of aerosol light absorption, that is, how the absorption varies as a function of the light's wavelength. They found that the SAFARI samples (air affected by biomass burning), which contain a lot of OC, exhibit a spectral dependence different from that of Caldecott Tunnel samples (air with diesel particles), which contain a lot of BC. This result suggested that biomass-burning samples contain material other than BC particles that absorbs some of the sun's heat; the researchers suspected OC, and, to investigate further, used a strong solvent and extracted OC from their samples. Once they had extracted OC from the SAFARI samples, Kirchstetter and Novakov expected that those samples would behave more like the ambient air and tunnel samples, and that is exactly what they found. Their conclusion: "Biomass smoke samples actually have an organic component that absorbs some light...More generally, under certain combustion conditions, emitted organic carbon particles may contribute to light absorption," says Kirchstetter. This means that OC may be having an effect on climate change that is not accounted for in current computer models - a new and fruitful research area.
With these results, Kirchstetter is now working with EETD's Doug Black to combine the MULTI and the EGA method. The researchers plan to develop a field-deployable apparatus that would include two instruments: one to measure mass concentration of carbon aerosols and the other to measure the particles' light-absorbing effect. The two scientists will soon begin a study of "coated" particles, BC particles that have mixed with other chemicals in the atmosphere and in the process acquired an additional layer of material. These mixed particles may have an effect on climate change that is different from the sum of their components. The effect of mixing or coating currently is poorly understood.
Kirchstetter will also be returning to the Caldecott Tunnel this summer to make new measurements of pollutant emissions. He made his original measurements when gasoline in California still contained the additive methyl tertiary-butyl ether (MTBE), which has since been removed from California's gasoline because of undesirable environmental effects. With new measurements, Kirchstetter and colleagues Rob Harley (from the University of California Berkeley) and Tony Strawa (from NASA Ames Research Center) hope to understand whether the change in gasoline formulation has altered automotive pollutant emissions for better or for worse.
For more information, contact:
- Tom Kirchstetter
- (510) 486-5319; Fax (510) 486-7303
This research was supported by the Department of Energy's Office of Science.
M. Sato, J. Hansen, D. Koch, A. Lacis, R. Ruedy, O. Dubrovik, B. Holben, M. Chin, T. Novakov. 2003. Global Atmospheric Black Carbon Inferred from Aeronet. Proceedings of the National Academy of Sciences.
T. Novakov, V. Ramanathan, J. Hansen, T. Kirchstetter, M. Sato, J. Sinton, and J. Sathaye. 2003. "Large Historical Changes of fossil fuel black carbon aerosols." Geophysical Research Letters, 30: 1324-1328.
T. Kirchstetter, C. Corrigan and T. Novakov. 2001. "Laboratory and field investigation of the adsorption of gaseous organic compounds onto quartz filters." Atmospheric Environment, 35: 1663-1671.
T. Kirchstetter, T. Novakov and P. Hobbs. Evidence that the spectral dependence of light absorption by aerosols is affected by organic carbon. In press.