Abstract
Spectral aerosol light absorption is an important parameter for the assessment of the radiation budget of the atmosphere. Although on-line measurement techniques for aerosol light absorption, such as the Aethalometer and the Particle Soot Absorption Photometer (PSAP), have been available for two decades, they are limited in accuracy and spectral resolution because of the need to deposit the aerosol on a filter substrate before measurement. Recently, a 7-wavelength (?) Aethalometer became commercially available, which covers the visible (VIS) to near-infrared (NIR) spectral range (?=450?950 nm), and laboratory calibration studies improved the degree of confidence in these measurement techniques. However, the applicability of the laboratory calibration factors to ambient conditions has not been investigated thoroughly yet. As part of the LBA-SMOCC (Large scale Biosphere atmosphere experiment in Amazonia ? SMOke aerosols, Clouds, rainfall and Climate) campaign from September to November 2002 in the Amazon basin we performed an extensive field calibration of a 1-? PSAP and a 7-? Aethalometer utilizing a photoacoustic spectrometer (PAS, 532 nm) as reference device. Especially during the dry period of the campaign, the aerosol population was dominated by pyrogenic emissions. The most pronounced artifact of integrating-plate type attenuation techniques is due to multiple scattering effects within the filter matrix. For the PSAP, we essentially confirmed the laboratory calibration factor by Bond (1999). On the other hand, for the Aethalometer we found a multiple scattering enhancement of 5.23 (or 4.55, if corrected for aerosol scattering), which is significantly larger than the factors previously reported (~2). While the exact reason for this discrepancy is unknown, the available data from the present and previous studies suggest aerosol mixing (internal versus external) as a likely cause. While it is well-known that RH may (moderately) affect aerosol absorption, we found no dependence of either PSAP or Aethalometer on relative humidity (RH) for 30%<RH<55% and 40%<RH<80%, respectively. However, a substantial decrease in PSAP sensitivity was observed for low RH (20%<RH<30%). In addition, while the PSAP demonstrated no sensitivity to gaseous adsorption, the Aethalometer response was clearly positively correlated with the gradient in pollution level. Hence, although very similar in measurement principle, the PSAP and Aethalometer require markedly different correction factors, which is probably due to the different filter media used. Although on-site calibration of the PSAP and Aethalometer is suggested for best data quality, we recommend a set of PSAP and Aethalometer correction factors for ambient sampling based on the data from the present and previous studies. For this study, the estimated accuracies of the absorption coefficients determined by the PAS, PSAP and Aethalometer were 10, 15 and 20%, respectively.