Aerosol Typing from Linear estimations for the Analytical Separation (ATLAS) of complex aerosol mixtures and improved...

We developed a mathematical model that is used for describing the relationships between the intensive parameters (IPs) of arbitrary external aerosol mixtures (AMs) and the intensive parameters of the aerosol components in these AMs. We denote this procedure as aerosol typing and aerosol characterization, respectively. The IPs of AMs can be found from multiwavelength lidar measurements of backscatter coefficients (β) at 355, 532, and 1064 nm; extinction coefficients (α) at 355 and 532 nm; particle linear depolarization ratios (δ) at 355, 532, and 1064 nm; and fluorescence backscattering coefficients (β _F). The IPs of the AM components are known a priori. The mathematical model allows us to obtain by analytical linear estimations the fractions of all components that are contained in complex AMs in terms of extinction and backscatter coefficients at the available lidar-measurement wavelengths. We carry out numerical simulations to test the correctness of this approach of Aerosol Typing from Linear estimations for the Analytical Separation (ATLAS) of aerosol mixtures. We test the sensitivity of ATLAS to measurement errors and incorrectly given IPs of AM components. Our numerical simulations show that the uncertainty of ATLAS is proportional to measurement errors and the uncertainty of the given IPs.