Pre-filter analysis for retrieval of microphysical particle parameters: a quality-assurance method applied to 3 backscatter (ß) +2 extinction (α) optical data taken with HSRL/Raman lidar.

We analyze the solution space of 3ß+2α optical data inferred from lidar measurements, i.e., backscatter coefficients at three wavelengths and extinction coefficients at two wavelengths. These optical data are governed by microphysical parameters that can be expressed in terms of particle size distribution, effective radius, and complex refractive index (CRI). In our analysis, we consider two scenarios of the solution space. First, it can be expressed in terms of monomodal particle size distributions represented either by fine modes or by coarse modes. Secondly, the particle size distributions contain a fine mode as well as a coarse mode. Consideration of both scenarios and different values of the effective radius and CRI allows us to find synthetic 3ß+2α optical data and corresponding intensive parameters (IPs) such as lidar ratios, backscatter- and extinction-related Ångström exponents at the available measurement wavelengths. Based on interdependencies between synthetic IPs and various microphysical properties, the qualitative and quantitative criteria for the optical data quality-assurance tool are developed. We derive the conditions of smoothness, closeness, convergence, and stability of the solution space for the quantitative criteria to test the quality of the 3ß+2α optical data. Our novel methodology, to the best of our knowledge, can be used not only for particles of spherical shape, but also for cases in which particles are irregularly shaped. Another strength of our methodology is that it also works for the case of a size-dependent and wavelength-dependent CRI. We show the potential of this methodology for a measurement case from the ORACLES campaign. Data were taken with NASA Langley's airborne HSRL-2 instrument on September 24, 2016.

Simultaneous profiling of dust aerosol mass concentration and optical properties with polarized high-spectral-resolution lidar.

Dust particles originating from arid desert regions can be transported over long distances, presenting severe risks to climate, environment, social economics, and human health at the source and downwind regions. However, there has been a dearth of continuous diurnal observations of vertically resolved mass concentration and optical properties of dust aerosols, which hinders our understanding of aerosol mixing, stratification, aerosol-cloud interactions, and their impacts on the environment. To fill the gap of the insufficient observations, to the best of our knowledge, this work presents the first high-spectral-resolution lidar (HSRL) observation providing days of continuous profiles of the mass concentration, along with particle linear depolarization ratio (PLDR), backscattering coefficient, extinction coefficient and lidar ratio (LR), simultaneously. We present the results of two strong dust events observed by HSRL over Beijing in 2021. The maximum particle mass concentrations reached (1.52 ± 3.5) x103 µg/m3 and (19.48 ± 0.36) x103 µg/m3 for the two dust events, respectively. The retrieved particle mass concentrations and aerosol optical depth (AOD) agree well with the observation from the surface PM10 concentrations and sun photometer with correlation coefficients of 0.90 and 0.95, respectively. The intensive properties of PLDR and LR of the dust aerosols are 0.31 ± 0.02 and 39 ± 7 sr at 532 nm, respectively, which are generally close to those obtained from observations in the downwind areas. Moreover, inspired by the observations from HSRL, a universal analytical relationship is discovered to evaluate the proportion of dust aerosol backscattering, extinction, AOD, and mass concentration using PLDR. The universal analytical relationship reveals that PLDR can directly quantify dust aerosol contribution, which is expected to further expand the application of polarization technology in dust detection. These valuable observations and findings further our understanding of the contribution of dust aerosol to the environment and help supplement dust aerosol databases.

Implementation of UV rotational Raman channel to improve aerosol retrievals from multiwavelength lidar.

Vibrational Raman effect is widely used in atmospheric lidar systems, but rotational Raman present several advantages. We have implemented a new setup in the ultraviolet branch of an existing multiwavelength lidar system to collect signal from rotational Raman lines of Oxygen and Nitrogen. We showed that, with an appropriate filter wavelength selection, the systematic error introduced in the particle optical properties due to temperature dependence was less than 4%. With this new setup, we have been able to retrieve aerosol extinction and backscatter coefficients profiles at 355 nm with 1-h time resolution during daytime and up to 1-min time resolution during nighttime.

Characterization of smoke and dust episode over West Africa: comparison of MERRA-2 modeling with multiwavelength Mie-Raman lidar observations.

Observations of multiwavelength Mie-Raman lidar taken during the SHADOW field campaign are used to analyze a smoke-dust episode over West Africa on 24-27 December 2015. For the case considered, the dust layer extended from the ground up to approximately 2000 m while the elevated smoke layer occurred in the 2500-4000 m range. The profiles of lidar measured backscattering, extinction coefficients, and depolarization ratios are compared with the vertical distribution of aerosol parameters provided by the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). The MERRA-2 model simulated the correct location of the near-surface dust and elevated smoke layers. The values of modeled and observed aerosol extinction coefficients at both 355 and 532 nm are also rather close. In particular, for the episode reported, the mean value of difference between the measured and modeled extinction coefficients at 355 nm is 0.01 km-1 with SD of 0.042 km-1. The model predicts significant concentration of dust particles inside the elevated smoke layer, which is supported by an increased depolarization ratio of 15 % observed in the center of this layer. The modeled at 355 nm the lidar ratio of 65 sr in the near-surface dust layer is close to the observed value (70 ± 10) sr. At 532 nm, however, the simulated lidar ratio (about 40 sr) is lower than measurements (55 ± 8 sr). The results presented demonstrate that the lidar and model data are complimentary and the synergy of observations and models is a key to improve the aerosols characterization.

Assessing the temperature dependence of narrow-band Raman water vapor lidar measurements: a practical approach.

Narrow-band detection of the Raman water vapor spectrum using the lidar technique introduces a concern over the temperature dependence of the Raman spectrum. Various groups have addressed this issue either by trying to minimize the temperature dependence to the point where it can be ignored or by correcting for whatever degree of temperature dependence exists. The traditional technique for performing either of these entails accurately measuring both the laser output wavelength and the water vapor spectral passband with combined uncertainty of approximately 0.01 nm. However, uncertainty in interference filter center wavelengths and laser output wavelengths can be this large or larger. These combined uncertainties translate into uncertainties in the magnitude of the temperature dependence of the Raman lidar water vapor measurement of 3% or more. We present here an alternate approach for accurately determining the temperature dependence of the Raman lidar water vapor measurement. This alternate approach entails acquiring sequential atmospheric profiles using the lidar while scanning the channel passband across portions of the Raman water vapor Q-branch. This scanning is accomplished either by tilt-tuning an interference filter or by scanning the output of a spectrometer. Through this process a peak in the transmitted intensity can be discerned in a manner that defines the spectral location of the channel passband with respect to the laser output wavelength to much higher accuracy than that achieved with standard laboratory techniques. Given the peak of the water vapor signal intensity curve, determined using the techniques described here, and an approximate knowledge of atmospheric temperature, the temperature dependence of a given Raman lidar profile can be determined with accuracy of 0.5% or better. A Mathematica notebook that demonstrates the calculations used here is available from the lead author.

Vertical profiles of pure dust and mixed smoke-dust plumes inferred from inversion of multiwavelength Raman/polarization lidar data and comparison to AERONET retrievals and in situ observations.

We present for the first time vertical profiles of microphysical properties of pure mineral dust (largely unaffected by any other aerosol types) on the basis of the inversion of optical data collected with multiwavelength polarization Raman lidar. The data were taken during the Saharan Mineral Dust Experiment (SAMUM) in Morocco in 2006. We also investigated two cases of mixed dust-smoke plumes on the basis of data collected during the second SAMUM field campaign that took place in the Republic of Cape Verde in 2008. Following the experience of the Aerosol Robotic Network (AERONET), the dust is modeled as a mixture of spherical particles and randomly oriented spheroids. The retrieval is performed from the full set of lidar input data (three backscatter coefficients, two extinction coefficients, and one depolarization ratio) and from a reduced set of data in which we exclude the depolarization ratio. We find differences of the microphysical properties depending on what kind of optical data combination we use. For the case of pure mineral dust, the results from these two sets of optical data are consistent and confirm the validity of the spheroid particle model for data inversion. Our results indicate that in the case of pure mineral dust we do not need depolarization information in the inversion. For the mixture of dust and biomass burning, there seem to be more limitations in the retrieval accuracy of the various data products. The evaluation of the quality of our data products is done by comparing our lidar-derived data products (vertically resolved) to results from AERONET Sun photometer observations (column-averaged) carried out at the lidar field site. Our results for dust effective radius show agreement with the AERONET observations within the retrieval uncertainties. Regarding the complex refractive index a comparison is difficult, as AERONET provides this parameter as wavelength-dependent quantity. In contrast, our inversion algorithm provides this parameter as a wavelength-independent quantity. We also show some comparison to results from airborne in situobservation. A detailed comparison to in situ results will be left for a future contribution.

Spatial distribution of doubly scattered polarized laser radiation in the focal plane of a lidar receiver.

Depolarization lidars are widely used to study clouds and aerosols because of their ability to discriminate between spherical particles and particles of irregular shape. Depolarization of cloud backscattered radiation can be caused also by multiple scattering events. One of the ways to gain information about particle parameters in the presence of strong multiple scattering is the measurement of radial and azimuthal dependence of the polarization patterns in the focal plane of receiver. We present an algorithm for the calculation of corresponding polarized patterns in the frame of double scattering approximation. Computations are performed for various receiver field of views, for different parameters of the scattering geometry, e.g., cloud base and sounding depth, as well as for different values of cloud particle size and refractive index. As the spatial distribution of cross-polarized radiation is of cross shape and rotated at 45 degrees with respect to laser polarization, the use of a properly oriented cross-shaped mask in the receiver focal plane allows the removal of a significant portion of the depolarized component of the backscattered radiation produced by double scattering. This has been verified experimentally based on cloud depolarization measurements performed at different orientations of the cross-shaped mask. Results obtained from measurements are in agreement with model predictions.

Information content of data measured with a multiple-field-of-view lidar.

Lidars with multiple fields of view (MFOVs) are promising tools for gaining information on cloud particle size. We perform a study of the information content of MFOV lidar data with the use of eigenvalue analysis. The approach we have developed permits an understanding of the main features of MFOV lidars and provides a way to relate the accuracy of particle size estimation with the measurement uncertainty and the scattering geometry such as the cloud-base height and the lidar sounding depth. Second-order scattering computations are performed for an extended range of particle sizes and for a wide range of lidar fields of view (FOVs). The results obtained allow us to specify the areas of possible applications of these lidars in cloud studies. Comparison of results obtained with polarized and cross-polarized scattered components demonstrate that the cross-polarized signal should provide a more stable retrieval and is preferable when double scattering is highly dominant. Our analysis allows for the estimation of the optimal number of FOVs in the system and their angular distribution, so this work can be a useful tool for practical MFOV lidar design.

Information content of multiwavelength lidar data with respect to microphysical particle properties derived from eigenvalue analysis.

The multiwavelength Raman lidar technique in combination with sophisticated inversion algorithms has been recognized as a new tool for deriving information about the microphysical properties of atmospheric aerosols. The input optical parameter sets, provided by respective aerosol Raman lidars, are at the theoretical lower limit at which these inversion algorithms work properly. For that reason there is ongoing intense discussion of the accuracy of these inversion methods and the possibility of simultaneous retrieval of the particle size distribution and the complex refractive index. We present results of the eigenvalue analysis, used to study the information content of multiwavelength lidar data with respect to microphysical particle properties. Such an analysis provides, on a rather mathematical basis, more insight into the limitations of these inversion algorithms regarding the accuracy of the retrieved parameters. We show that the effective radius may be retrieved to 50% accuracy and the real and imaginary part of the complex refractive index to +/- 0.05 and +/- 0.005i, if the imaginary part is < 0.02i. These results are in accordance with the classic approach of simulation studies with synthetic particle size distributions. Major difficulties are found with a particle effective radius of < 0.15 microm. In that case the complex refractive index may not be derived with sufficient accuracy. The eigenvalue analysis also shows that the accuracy of the derived parameters degrades if the imaginary part is > 0.02i. Furthermore it shows the importance of the simultaneous use of backscatter and extinction coefficients for the retrieval of microphysical parameters.

Numerical simulation of light backscattering by spheres with off-center inclusion. Application to the lidar case.

A Mie backscattering model for spherical particles with off-center inclusion has been developed and tested. The program is capable of dealing with size parameter values up to approximately 1000, thus allowing one to simulate the optical behavior of a large variety of atmospheric aerosols, as well as cloud and precipitation particles. On the basis of this model, we simulated the optical properties of polydisperse composite atmospheric particles as observed by ground-based and airborne lidar systems. We have characterized optical properties in terms of host and inclusion radii, considering water particles with different composition inclusions. The performed modeling provides some insight into the so-called lidar bright- and dark-band phenomenon.

Inversion of multiwavelength Raman lidar data for retrieval of bimodal aerosol size distribution.

We report on the feasibility of deriving microphysical parameters of bimodal particle size distributions from Mie-Raman lidar based on a triple Nd:YAG laser. Such an instrument provides backscatter coefficients at 355, 532, and 1064 nm and extinction coefficients at 355 and 532 nm. The inversion method employed is Tikhonov's inversion with regularization. Special attention has been paid to extend the particle size range for which this inversion scheme works to approximately 10 microm, which makes this algorithm applicable to large particles, e.g., investigations concerning the hygroscopic growth of aerosols. Simulations showed that surface area, volume concentration, and effective radius are derived to an accuracy of approximately 50% for a variety of bimodal particle size distributions. For particle size distributions with an effective radius of < 1 microm the real part of the complex refractive index was retrieved to an accuracy of +/- 0.05, the imaginary part was retrieved to 50% uncertainty. Simulations dealing with a mode-dependent complex refractive index showed that an average complex refractive index is derived that lies between the values for the two individual modes. Thus it becomes possible to investigate external mixtures of particle size distributions, which, for example, might be present along continental rims along which anthropogenic pollution mixes with marine aerosols. Measurement cases obtained from the Institute for Tropospheric Research six-wavelength aerosol lidar observations during the Indian Ocean Experiment were used to test the capabilities of the algorithm for experimental data sets. A benchmark test was attempted for the case representing anthropogenic aerosols between a broken cloud deck. A strong contribution of particle volume in the coarse mode of the particle size distribution was found.

Angle- and size-dependent characteristics of incoherent Raman and fluorescent scattering by microspheres. 1. General expressions.

A dipole model is used to simulate incoherent Raman and fluorescent scattering by microspheres. The use of the addition theorem for spherical harmonics circumvents the need to evaluate double sums in the final formulas, thereby drastically reducing computational effort. Special attention is paid to consideration of backscattering geometry, which is important for lidar applications. The formulas derived for backscattering geometry decrease the computation time for size parameter x approdximately 100 by a factor of 200 compared with the time for calculations performed at other angles.

Angle- and size-dependent characteristics of incoherent Raman and fluorescent scattering by microspheres. 2. Numerical simulation.

The results of numerical simulation of inelastic scattering by microspheres with the use of a dipole model are presented. The formulas that are derived speed up the computation, thereby permitting larger-sized microspheres to be studied. The angular scattering cross section and depolarization are calculated for a wide range of size parameters as well as for different orientations of incident wave polarization. Calculations performed with small incremental changes in size permit the influence of morphology-dependent resonance (MDR) on the power and angular distribution of scattered radiation to be studied. TM and TE types of MDR produce enhanced scattering of the incident wave with vertical and horizontal polarization; the corresponding shape of the phase function becomes oscillatory. Special attention is paid to the simulation of backward scattering by water droplets, which is important for Raman lidar applications.

Inversion with regularization for the retrieval of tropospheric aerosol parameters from multiwavelength lidar sounding.

We present an inversion algorithm for the retrieval of particle size distribution parameters, i.e., mean (effective) radius, number, surface area, and volume concentration, and complex refractive index from multiwavelength lidar data. In contrast to the classical Tikhonov method, which accepts only that solution for which the discrepancy reaches its global minimum, in our algorithm we perform the averaging of solutions in the vicinity of this minimum. This averaging stabilizes the underlying ill-posed inverse problem, particularly with respect to the retrieval of number concentration. Results show that, for typical tropospheric particles and 10% error in the optical data, the mean radius could be retrieved to better than 20% from a lidar on the basis of a Nd:YAG laser, which provides a combination of backscatter coefficients at 355, 532, and 1064 nm and extinction coefficients at 355 and 532 nm. The accuracy is improved if the lidar is also equipped with a hydrogen Raman shifter. In this case two additional backscatter coefficients at 416 and 683 nm are available. The combination of two extinction coefficients and five backscatter coefficients then allows one to retrieve not only averaged aerosol parameters but also the size distribution function. There was acceptable agreement between physical particle properties obtained from the evaluation of multiwavelength lidar data taken during the Lindenberg Aerosol Characterization Experiment in 1998 (LACE 98) and in situ data, which were taken aboard aircraft.