Atmosphere (Jul 2023)
Particle Microphysical Parameters and the Complex Refractive Index from 3β + 2α HSRL/Raman Lidar Measurements: Conditions of Accurate Retrieval, Retrieval Uncertainties and Constraints to Suppress the Uncertainties
Abstract
We study retrieval methods in regard to their potential to accurately retrieve particle microphysical parameters (PMP) from 3β + 2α HSRL/Raman lidar measurements. PMPs estimated with these methods are number, surface-area and volume concentrations, the effective radius, and complex refractive index of the investigated particle size distribution (PSD). The 3β + 2α optical data are particle backscatter coefficients at 355, 532 and 1064 nm and extinction coefficients at 355 and 532 nm. We present results that are fundamental for our understanding of how uncertainties of the optical data convert into uncertainties of PMPs. PMPs can only be retrieved with preset accuracy if the input optical data are accurate to at least eight significant digits, i.e., 10−6%. Such measurement accuracy cannot be achieved by currently existing lidar measurement techniques and the fact that atmospheric conditions are not static during lidar observations. Our analysis of the results derived with the novel approach shows that (a) the uncertainty of the retrieved surface-area concentration increases proportionally to the measurement uncertainty of the extinction coefficient at 355 nm, (b) the uncertainty of the effective radius is inversely proportional to the measurement uncertainty of the extinction-related Ångström exponent, (c) the uncertainty of volume concentration is close to the one of the effective radius, and (d) the uncertainty of number concentration is proportional to the inverse of the square value of the uncertainty of the effective radius. The complex refractive index (CRI) cannot be estimated without introducing extra constraints, even if measurement uncertainties of the optical data are as low as 1−3%. We tested constraints and their impact on the solution space, and in how far these constraints could allow us to restrict the retrieval uncertainties. For example, we used information about relative humidity that can be measured with Raman lidar. Relative humidity is an important piece of information that allows for more accurate aerosol typing and thus plays a vital role in any kind of aerosol characterization. The measurement example we used in this study shows that such a constraint can reduce the retrieval uncertainty of single scattering albedo (SSA) to as low as ±0.01–±0.025 (at 532 nm), on the condition that the uncertainty of the input optical data stays below 15%. The results will be used for uncertainty analysis of data products provided by future versions of the Tikhonov Advanced Regularization Algorithm (TiARA). This algorithm has evolved into a standard tool for the derivation of microphysical particle properties from multiwavelength High-Spectral-Resolution Lidar (HSRL)/Raman lidar operated in Europe, East Asia, and the US.
Keywords
