We developed an algorithm to estimate the vertical profiles of extinction coefficients at 355 nm for three aerosol components (black carbon, dust, and air pollution particles other than black carbon), using the extinction coefficient, backscatter coefficient, and depolarization ratio at 355 nm for total aerosols derived from data measured with a high spectral resolution lidar (HSRL) with a depolarization measurement function installed on EarthCARE. We also developed a similar algorithm to analyze 532 nm polarization HSRL data from ground-based and airborne measurements. The mode radii, standard deviations, and refractive indexes for each aerosol component were prescribed; the optical properties of each aerosol component were computed and modeled on the assumption that dust was spheroidal and the other components were spherical. We performed sensitivity studies on retrieval errors due to measurement uncertainty or due to assumptions of the algorithms, and characterized the algorithm performance. We demonstrated the ability of the algorithm by applying it to the 532 nm HSRL data measured at Tsukuba, Japan from retrieved plumes consisting of black carbon, dust, air pollution aerosols or some mixture thereof, with results that were consistent with previous studies. We further compared the distributions of the aerosol components retrieved in this study with those retrieved by the aerosol classification retrieval algorithm using the 532 nm HSRL data and the lidar signal at 1064 nm developed in the previous study. We found that the estimates probably agreed in lower layers below 3 km but the dust and air pollution aerosol extinction coefficients in the upper layers differed. The simulation of 1064 nm lidar signals suggests that the mode radius of dust differs in the lower and upper layers.
