On the equivalency of Black Smoke / British Smoke and historic Elemental Carbon

Abstract

Elemental Carbon (EC) is a good indicator for aerosol air quality and a main component in global warming; therefore data on its historic concentration would be highly valuable for an assessment of historic levels. There were already major sources of EC in Europe in the early 20th century, but the component has only been measured since the late 1980s when analysis procedures were developed and even until recently data are scarce. However, historic EC-concentrations may be deducible from the parameter British Smoke (BrS), respectively the equivalent parameter Black Smoke (BS; in continental Europe), monitored on a daily basis since the late 1940s. Values of the parameter were based on the light-absorptivity of aerosol filter samples and these should be a measure for EC, because it is/was the dominant light-absorbing species in the countries in which the method was used. We show here that BrS was a linear proxy for EC in the standardisation studies performed in the early 1960s, based on the approach of “proportional” sampling. In this procedure filter samples were collected in parallel sampling lines with increasing filter loadings of the same aerosol. The light-absorption as a function of mass loading, of a variety of aerosol types, followed a generic curve. This graph was normalised in the “BrS-curve” by setting BrS equal to the average mass loading of the urban aerosol in the tests. Because of the proportional sampling the mass-loading of the component EC followed, by necessity, the BrS-curve as a function of light-absorption. The proportionality factor between BrS and EC is not known because the latter parameter was not analysed then. When EC became measured, the major source of urban EC was diesel exhaust. We show that filter samples with urban aerosols containing EC deriving from diesel combustion and from domestic coal use give similar light-absorption signals when probed with the BrS/BS-method. This implies that the equivalency factor between EC and BS, as assessed in recent years, must also have applied in the past. In the intercomparison studies that we found in literature the correlation factor (R2) between EC and BS was on average 0.85, but the absolute EC values were quite different because of the different analysis methods used for EC. We converted the EC-data to values comparable to the new EU-standard “EUSAAR2-TOT”. We then arrive at an equivalency factor of 0.15 (±25%) respectively 0.18 (±25%) to translate historic BS or BrS data to concentrations of EC.