An evaluation of some issues regarding the use of aethalometers to measure woodsmoke concentrations

Abstract

Recent papers have described the use of both seven-wavelength and two-wavelength aethalometers to estimate the concentration of woodsmoke in the atmosphere. This application depends upon the enhanced absorption of woodsmoke at UV wavelengths relative to that of traffic particles which is quantified by the aethalometer. This paper draws together evidence from a number of experimental data sources which challenges the reliability of woodsmoke concentration estimates derived from aethalometer measurements. One crucial aspect is the selection of an Ångstrom exponent ( α) for woodsmoke, and our experimental data from a wood combustion source suggest that, consistent with other published data, this is highly variable. The outputs of the "aethalometer model" for estimating woodsmoke mass are sensitive to this parameter and there is currently no way to select the optimum value of α for woodsmoke, which may vary with location as it will depend upon the type of wood fuel and the combustion conditions. Examples are included demonstrating the sensitivity of the aethalometer model to the choice of α values for traffic and woodsmoke. Additionally, analysis of data for UVPM (Delta-C) from an aethalometer network shows facets in the data which cast doubt on the reliability of the method. In particular, the small seasonal variation of UVPM at a London background site in comparison to other woodsmoke markers and its greater similarity to that of black carbon suggests that there are probably other UV absorbing contributors than woodsmoke to the aethalometer signal. Considerable caution should be exercised in interpreting aethalometer data as offering quantitative estimates of woodsmoke concentrations, and a number of questions are posed which need to be addressed before aethalometers can be used with confidence to give quantitative estimates of woodsmoke concentrations in a range of environments.