Brown carbon (BrC) has a substantial direct radiative effect, but current estimates of its impact on radiative balance are highly uncertain due to a lack of measurements of its light-absorbing properties, such as mass absorption efficiency (MAE). Here, we present a new analytical paradigm based on a Bayesian inference (BI) model that takes multiwavelength aethalometer measurements and total carbon data to resolve the concentrations of black carbon and BrC, and MAEs of BrC on a sample-by-sample basis. Hourly MAEs, unattainable in previous studies, can now be calculated, enabling the first-time observation of the darkening-bleaching dynamics of BrC in response to photochemical transformation. We demonstrate the application of this BI model to analyze measurements collected over one year (2021-2022) in Hong Kong. Diel variations in MAE370 nm of BrC reveal a darkening-to-bleaching transition occurring between 8 and 10 O'clock when the solar irradiance ranges from 30 to 400 W m-2. Furthermore, we consistently observe an increase in MAE370 nm of BrC with nitrogen oxide concentrations, suggesting the enhanced formation of nitrogenous organics. This BI model-based data analysis would bring forth a breakthrough in amassing observation data of BrC and its MAEs in diverse ambient environments and with high time resolution.
Keywords: Bayesian inference; aethalometer; ambient aerosol; brown carbon; light absorption.