An Evaluation of three methods for measuring black carbon at Alert, Canada

The measurement of BC is complicated by the lack of a simple definition of BC and the absence of techniques that are uniquely sensitive to BC (e.g. Petzold et al., 2013). At the Global Atmosphere Watch baseline observatory at Alert, Nunavut (82.5^oN), BC mass is estimated in three ways, none of which fully represent BC: conversion of light absorption measured with an Aethalometer to give equivalent black carbon (EBC), thermal desorption of elemental carbon from weekly-integrated filter samples to give elemental carbon (EC), and measurement of incandescence from the refractory black carbon (rBC) component of individual particles using a Single Particle Soot Photometer (SP2). Based on measurements between March 2011 and December 2013, EBC and EC are 2.7 and 3.1 times higher than rBC respectively. The EBC and EC measurements are influenced by factors other than just BC, and higher estimates of BC are expected from these techniques. Some bias in the rBC measurement may result from calibration uncertainties that is difficult to estimate here. Considering a number of factors, our best estimate of averaged BC mass at Alert are 49±28 ng m^-3, 30±26 ng m^-3, 22±13 ng m^-3, and 29±9 ng m^-3, for winter, spring, summer, and fall respectively. Average coating thicknesses estimated from the SP2 are 25% to 40% of the 160-180 nm diameter rBC core sizes. For particles of approximately 200-400 nm optical diameter, the fraction containing rBC cores is estimated to be between 10% and 16%, but the possibility of smaller undetectable rBC cores in some of the particles cannot be excluded. Mass absorption coefficients (MAC)±uncertainty at 550 nm wavelength are 8±4, 8±4, 5±2.5 and 9±4.5 m² g^-1, for winter, spring, summer and fall respectively. Adjusted to better estimate absorption by BC only, the winter and spring values of MAC are 7.6±3.8 and 7.7±3.8 m²g^-1.