Estimation of Cloud Condensation Nuclei Concentration from Aerosol Optical Quantities: Influential Factors and Uncertainties

Cloud condensation nuclei (CCN) is a key variable for understanding cloud formation, but it is hard to obtain on large scales on a routine basis, whereas aerosol optical quantities are more readily available. This study presents an in-depth investigation on the relationship between CCN and aerosol optical quantities in regions of distinct aerosol types using extensive measurements collected at multiple Atmospheric Radiation Measurement (ARM) Climate Research Facility (CRF) sites around the world. The influences of relative humidity (RH), aerosol hygroscopicity (fRH) and single scattering albedo (SSA) on the relationship are analyzed. Better relationships are found between aerosol optical depth (AOD) and CCN at the Southern Great Plains (US), Ganges Valley (India) and Black Forest sites (Germany) than those at the Graciosa Island and Niamey (Niger) sites, where sea salt and dust aerosols dominate, respestively. In general, the correlation between AOD and CCN decreases as the wavelength of AOD measurement increases, suggesting that AOD measured at a shorter wavelength is a better proxy of CCN. The correlation is significantly improved if aerosol index (AI) is used together with AOD. The highest correlation exists between CCN and aerosol scattering coefficients (σsp) and scattering AI measured in-situ. The CCN-AOD (AI) relationship deteriorates with increasing RH. If RH exceeds 75%, the relationship becomes almost invalid for using AOD as a CCN proxy, whereas a tight σsp-CCN relationship exists for dry particles. Aerosol hygroscopicity has a weak impact on the σsp-CCN relationship. Particles with low SSA are generally associated with higher CCN concentrations, suggesting that SSA affects the relationship between CCN concentration and aerosol optical quantities. It may thus be used as a constraint to reduce uncertainties in the relationship. A significant increase in σsp and decrease in CCN with increasing SSA is observed, leading to a significant decrease in their ratio (CCN/ σsp) with increasing SSA. The relationships and major influential factors are parameterization for improving CCN estimation with varying amount of information on RH, particle size and SSA.