Estimation of hourly black carbon aerosol concentrations from glass fiber filter tapes using image reflectance-based method

Abstract

Black carbon (BC) is a carbonaceous component of fine particulate matter (PM2.5) that is quantified via light absorption. We demonstrate a low-cost method for quantifying BC using cell phone camera images. BC concentration is correlated with red light absorbance that is measured from photographs of particle filter samples following image processing steps to account for distortion (geometric calibration) and lighting conditions (color calibration). We trained multiple Red channel to BC models using ambient air filter samples and found that the exponential model best explains the correlation (R2 ∼ 0.94). Our approach has an effective minimum detection limit of 0.15 μg m−3 of ambient BC for an hourly sample collected at 1 m3 h−1. This detection limit should be sufficient to quantify BC concentrations at high time resolution at locations worldwide. We demonstrate the performance of our optical BC method using a combination of filter samples and filter tapes from beta attenuation monitors (BAMs) operating at two sites in an urban setting in Pittsburgh, Pennsylvania, USA. Our approach compares favorably to reference filter-based methods, suggesting that post-analysis of BAM tapes collected worldwide can be a valuable source for PM composition data, especially in countries in the Global South where there is insufficient air quality data to make informed policy decisions.