COMPARISON OF MEASUREMENTS OF AEROSOL OPTICAL-ABSORPTION BY FILTER COLLECTION AND A TRANSMISSOMETRIC METHOD

Light-absorbing aerosol particles are present in the atmosphere all over the world, and they play an important role in the energy transfer to the atmosphere. The main substance causing light absorption in the atmosphere is graphitic (black) carbon, the major constituent of soot. It is generated at any kind of combustion, and in source regions soot particles are mostly mixed externally with other particles. After some time the soot particles may become attached to other particles, thus becoming mixed internally. Many methods for the determination of the light absorption by particles exists. The integrating plate technique offers a fast and reproducible way to determine light absorption coefficients of aerosol particles deposited on filters. Comparisons of this method with other determinations of light absorption coefficients have given partially agreement and partially disagreement. Therefore a comparison with an absolute method has been performed in order to determine possible inaccuracies. An externally mixed aerosol was made by independently generating a light-absorbing aerosol and a non-light-absorbing aerosol and mixing them such that within the measuring period coagulation was negligible. The light-absorbing aerosol was generated by a spark discharge between carbon electrodes. The sizes of the particles were small, usually more than 90% of the particle mass was below 50 nm. The non-absorbing was generated by a constant output atomizer. For small carbon particles the light extinction is dominated by the light absorption, therefore a measurement of the light extinction coefficient in a long path photometer provides an absolute method to determine the absorption coefficient. A comparison with the absorption coefficient determined simultaneously with the integrating plate method permits the estimation of the error. For pure carbon the light absorption coefficient determined by the integrating plate method is approximately 20% higher than the ''true'' value. For mixed aerosols the deviation depends on the mixing rate, but the integrating plate method always overestimates the light absorption coefficient.