Parametrizations of size distribution and refractive index of biomass burning organic aerosol with black carbon content

Abstract. Biomass burning organic aerosol (BBOA) impacts significantly on climate and regional air quality directly through scattering and absorbing solar radiation and indirectly through acting as cloud condensation nuclei. However, fundamental parameters in the simulation of BBOA radiative effects and cloud activities such as size distribution and refractive index remain poorly parameterized in models. In this study, biomass burning events were frequently observed during autumn in the Pearl River Delta region, China. Aerosol physical properties including aerosol size distributions, aerosol scattering coefficients and aerosol absorptions as well as aerosol chemical compositions were comprehensively measured during these biomass burning events. An improved absorption Ångström exponent (AAE) ratio method considering both variations and spectral dependence of black carbon AAE was proposed to differentiate brown carbon (BrC) absorptions from total aerosol absorptions. BBOA size distributions, mass scattering and absorption efficiency were retrieved based on the changes in aerosol number size distributions, scattering coefficients and derived BrC absorptions that occurred with BBOA spikes. Geometric mean diameter of BBOA volume size distribution Dgv depended largely on combustion conditions, ranging from 245 to 505 nm, and a linear relationship between Dgv and ∆BC/∆BBOA was achieved. Retrieved BBOA mass scattering efficiency, ranges from 3 to 7.5 m2/g, depending nonlinearly on Dgv (R=0.86) which was confirmed by Mie theory simulations. Retrieved real part of BBOA refractive index ranges from 1.47 to 1.64, with evidences showing that its variations might depend largely on combustion efficiency, which however requires further comprehensive investigations. Retrieved BBOA mass absorption efficiencies and imaginary parts of BBOA refractive index (mi,BBOA) correlated highly with ∆BC/∆BBOA (R>0.88), but changes almost linearly with ∆BC/∆BBOA (R>0.88) which differs much with previous findings. Consistent with results of previous studies, the variations of mi,BBOA as a function of optical wavelength λ can be well parameterized using mi,BBOA(λ) = mi,BBOA (520) × (λ/520)wBBOA . The spectral dependence parameter wBBOA ranged from 2.5 to 5.5 with an average of 4.7 which is in generally higher than wBBOA values predicted by previous parameterization schemes, however, is actually consistent with previous laboratory results of similar ∆BC/∆BBOA ranges. In addition, wBBOA is also generally linearly correlated (R=-0.51) with ∆BC/∆BBOA. These findings have significant implications for simulating BBOA climate effects and suggest that linking both BBOA refractive index and BBOA volume size dsitrbutions to black carbon content might be a feasible and a good choice for climate models.