Technical Comments

Comment on "Radiative Absorption Enhancements Due to the Mixing State of Atmospheric Black Carbon"

See allHide authors and affiliations

Science  25 Jan 2013:
Vol. 339, Issue 6118, pp. 393
DOI: 10.1126/science.1229920


Cappa et al. (Reports, 31 August 2012, p. 1078) suggest that black carbon (BC) in a mixture absorbs only ~6% more sunlight than when volatile chemicals are evaporated from the mixture, and state that "many climate models may overestimate warming by BC." However, the authors misinterpret at least some model results and omit optical focusing at high relative humidity and of involatile components. Thus, their conclusion about model error is not demonstrated.

Cappa et al. (1) suggest that, after 20 hours of aging in the ambient atmosphere, the absorption coefficient of black carbon (BC) mixed with other volatile chemicals is only ~6% greater than that when the volatile chemicals are evaporated from the BC surface. These results were obtained from measurements from two field campaigns in California. In one case, the particles were dried to 55% relative humidity (RH) before any measurements were taken. From the analysis, they concluded "The small observed values of Eabs [from their figure 2] suggest that models that assume internal mixtures in a core‐shell configuration…can substantially overestimate the atmospheric warming by BC, potentially by up to a factor of 2" [citing Jacobson (2) and another paper]. They summarize this result as a core conclusion in their abstract: "many climate models may overestimate warming by BC."

However, in (2), I never assumed an "internal mixture" as I evolved the mixing state among 18 aerosol distributions. Figure 3 of (2) shows that at 20 hours (up to when measurements in (1) were taken for CalNex and 5 hours past when they were taken for the Carbonaceous Aerosols and Radiative Effects Study), only 25% of the radiative forcing (and Eabs) of a well‐mixed internal mixture was realized. This result is within the 91st percentile of figure 2 of (1) after 20 hours. Thus, Cappa et al. cannot claim that "many climate models may overestimate warming by BC" based on that comparison.

Second, Cappa et al.'s particles were up to 15 to 20 hours old, and thus were not completely "aged" particles represented in models. The soot lifetime is about 1 week, and often more, and most soot is processed through clouds. Thus, the particles measured in (1) were not fully coated with volatile material or fully coagulated with volatile and involatile material.

Third, most soot optical focusing occurs at high RH (36). Mikhailov et al. (3) found a peak amplification factor of 3.5 for a soot–water drop mixture at 100% RH. Brem et al. (4) found an amplification factor of up to 2.7 at 530 nm for an organic aerosol core resulting from burning oak wood when the RH rose from <70% to 95%. For CalNex, particles in (1) were "sampled…at 55% RH" [supplementary materials for (1)], so one would not expect to have swollen particles like those globally, where nearly 70% of the world is covered with clouds and 74.5% with sea water or sea ice. Jacobson (6) compared a global model with AERONET (Aerosol Robotic Network) and OMI (Ozone Monitoring Instrument) AAOD (absorption aerosol optical depth) and vertical BC profiles, accounting for the full variation of RH within and outside of clouds, and found no bias, which could not occur if absorption were overestimated.

Fourth, denuding particles of volatile components does not purify BC because soot coagulates with many particles that contain involatile material, such as Na, Mg, Ca, K, Si, Fe, Al, and others. Thus, Cappa et al. may have not have accounted for optical focusing of involatile material, which models account for.

Fifth, measurements were made in limited locations, whereas trillions of mixing states exist worldwide. There is no statistical significance to global climate properties based on few measurements, just as a few weather measurements cannot predict climate trends.

Finally, figure 2 of (1) indicates that many "measured" values of Eabs were below 1.0. Although compaction of particles due to coating them can modify the optical focusing effect, the authors were not compacting particles in their experiment; rather, they were removing volatile material from particles already compacted. The authors offer no physical explanation for why Eabs < 1 in many cases, but these values seem to bias the Eabs average significantly.


View Abstract


Navigate This Article