Open Access
Add to BibSonomy
Abstract
In this work, we present an in situ online aerosol recognition scheme by synchronized parallel polarization scattering analysis. By theoretical simulations, we select the feasible scattering angles and evaluate the potential of Stokes parameters to identify aerosols. Correspondingly, we develop a measurement system based on multi-angle optical scattering and multidimensional polarization analyzing technique. We construct two index groups based on non-normalized and normalized polarization parameters respectively, and employ their frequency distribution histograms instead of the simple average values to identify and classify different types of aerosols. The experimental verification confirms a future way of a multi-dimensional polarization parameter group applied in a fast and effective air pollutants monitoring.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Ambient air contains a variety of aerosols such as dust, bacteria, and other particles of biological and nonbiological origin. Aerosols are involved in various atmospheric processes such as ice nuclei formation, precipitation and global climate effects [1]. In the past few years, atmospheric pollutants attract increasing attention in China, due to their serious impact on human health and climate change [2,3]. In Chinese major cities, various water-soluble salts, soot, crustal elements and biomass-burning emissions generated from anthropogenic activities and industrial production contribute to most of PM2.5 pollutants [4–6], and how to obtain their optical and microphysical properties is a significant challenge for environment pollution control.
Traditional methods of aerosol assessment usually include microscopic inspection or molecular analysis of filters, suffering from low time resolution due to periodical and relatively long analytical procedures [7]. Among a variety of techniques, light scattering measurements with a fast response seem suitable for an online in situ observation of aerosols. Light scattering methods do not involve filter collection and data extraction. This reduced handling delays, and enables real time evaluation of air quality [8,9].
The scattering intensity distribution of an aerosol sample can be related with various optical properties, such as size, refractive indices, shape and surface texture [10–12]. By proper selection of optical scattering index and detection mechanism, we can understand some inherent characteristics of aerosols and then identify and classify them. For example, there are some researches indicating forward scattering measurement to aerosol size and lateral scattering analysis to discriminate different refractive indexes [13–15].
However, the chemical composition and mixing state of aerosols are complex. The factors influencing the light scattering include aerosol size, shapes, optical absorption and fluorescence et.al [8,16,17]. Considering the composite microphysical properties and large uncertainties of real aerosols, multimodal optical scattering techniques combined with multi-wavelength measurement, or fluorescence analysis, or Raman spectrum, or polarization scheme maybe can provide a better solution for the effective identification of atmospheric particulates [18–20].
A non-polarization instrument can be upgraded to the polarization version by introducing the polarization state generator and analyzer in the optical path [21]. In many fields, polarization measurements can be useful for inferring the physical features of micro aerosols like cells, floating algae, and aerosols [22–25]. The light scattering and polarization change has increasingly become an important research subject across several disciplines, including atmospheric and oceanic science, astrophysics, bioscience, and medical science [26–28].
Polarization measurements contain valuable information about aerosol microphysical properties [29,30]. In the environmental field, with the help of polarization parameters, remote sensing systems greatly enhanced observational capability of surface and atmosphere [31–33]. Especially on polarization LIDAR techniques, theoretical studies, laboratory measurements, and field campaigns have been carried out to understand the depolarization capabilities of aerosols [34], and permit identifying different aerosol types, such as ice crystals, dust and bioaerosols [35–37]. The backscattering linear depolarization ratio is useful in aerosol characterization [38]. If extracting more polarization characters, like Stokes vectors, more potential individual information of suspended aerosols will be possibly distinguished and identified [39–41].
In recent years, the development of aerosol lidar with polarization techniques makes it possible to derive optical and microphysical properties of aerosols and to resolve aerosol types and mixtures as a function of height. CALIPSO measures the depolarization ratio to indicate the irregularly shaped particles such as ice crystals or dust particles [42]. High Spectral Resolution Lidar, like ATLID, allows us to discriminate different aerosol types (marine, dust, haze). Multi-wavelength HSRL technique can obtain many microphysical parameters, such as effective radius, refractive index, and mass concentration [43]. In addition, Aerosol Robotic Network (AERONET) can retrieve aerosol size distribution and refractive index from the multispectral and multi-angular polarimetric measurements. Xu and Wang demonstrate how polarization measurements increase the retrieval accuracy [44]. A. Arola1 et al. use AERONET measurements to estimate the amount of light–absorbing organic carbon and black carbon from [45].
In this paper, based on our multi-angle polarization scattering instrument, we will present the potential of aerosol characterization by various polarization characters. The paper is organized in five sections: section 2 introduces the theoretical basis and experimental device; section 3 presents the theoretical simulations, indicating that different polarization parameters at different scattering angle have their applicable situations respectively; section 4 shows our experimental results and verifies the effectiveness of theoretical calculations; and section 5 summarizes the conclusions.
2. Theory
2.1. Stokes vectors and Mueller matrix
Polarization states of light can be denoted by the four components “Stokes vector”, S [46]. The four Stokes components are labeled as I, Q, U and V, where I denotes the total intensity; Q is the intensity difference between linear polarization states at 0° and 90°; U is the intensity difference between linear polarization states at 45° and 135°; V is the intensity difference between the right-handed (RCP) and the left-handed (LCP) circular polarization states, respectively.
The Mueller matrix is to describe the polarization transformation from the incident to output optical signals during light-matter interaction, i.e. SS = M × Si (where Si and Ss are incident and scattered Stokes vectors, respectively). For a polarized incident light, the scattering matrix of randomly oriented particles (with mirror symmetry) that determines the change of the incident Stokes vector [Ii, Qi, Ui, Vi] to the scattered vector [Is, Qs, Us, Vs] is given by [47,48]:
where the elements in the scattering matrix are functions with the scattering angle ranging from 0 to 180 deg. This paper focuses on the optical properties of spheroids with aspect ratios close to unity and the suspended aerosols in the air have a random spatial orientation, so we extract the polarization parameters based on (2), and the follow-up research involving aerosols with an obvious anisotropy is under way.2.2. Simulations of aerosol polarization scattering
Modeling of light scattering from an individual aerosol plays an important role in extracting appropriate characterization parameters. The Lorenz-Mie theory was a classical method to calculate the scattering properties of a single homogeneous aerosol [12]. It was not only used to calculate scattering and absorption cross-sections but also the angular scattering distribution for a spherical particle with a given size, refractive index, and wavelength of incident light.
In this paper, we consider only spherical aerosols and employ our program based on BHMIE calculation routine to study the polarization scattering process [49]. By this program, we can predict possible scattered Stokes vectors and the scattering matrix of aerosols before experiments, and compare the theoretical characters with those measured.
The lognormal function is used to describe the size distribution of the PM sample [50,51],
where d is the diameter of an aerosol, two parameters characterize an individual lognormal mode: geometric variance σ2 and geometric mean diameter d0. During our simulations, we set the wavelength of incident light as 532nm consistent with the following experiments, the geometric mean diameter d0 from 0.3μm to 5μm and the geometric variance σ as 1.196 [52]. The refractive index settings for four simulated aerosols are listed below [53–55]:3. Experimental setup and materials
3.1. Aerosol samples
In order to be as consistent as possible with the simulation, we choose three kinds of aerosols as our experimental samples: sodium sulfate (Na2SO4) (Tianxiang Co.), Soot (black carbon is the main ingredient) (JCNANO Tech Co., Ltd) and Arizona dust (SiO2 is the main ingredient) (Shanghai Rainbow Trading Co., Ltd). Their corresponding refractive indexes setting for simulations are listed in Tab.1. Before the experiments, the graphite and dust samples are dehydrated in a vacuum drying oven (DZF-6021, Keelrein) for 4 h at 100°C, then dispersed with a ultrasonic cleaner (JP020S, Skymen).
Table 1. Refractive indexes about Na2SO4, C, SiO2 and PbO
3.2. Experimental Set-up
Figure 1(a) shows our experimental set-up. The light source is a 50mW solid state laser with the wavelength of 532nm (MSL-III-532, Changchun New Industries Optoelectronics Technology Co., Ltd,). An lens group is used to control the spot size less than 1mm and a linear polarizer is used to keep the incident beam with a 45 degree linear polarization state. The detection channel can be installed on any one of four scattering channels corresponding to the 10, 60, 115 and 160 scattering angles in the sealed chamber. Scattered pulses by individual suspended aerosols are received by a silicon-based avalanche photo detector (YN-SP01, Richysensing), then recorded by a digital oscilloscope (TBS1000, Tektronix, 50MHz).
Fig. 1 (a) Schematic of our experimental setup. PSG module controls the incident light. There are four detection channels, and a nozzle keeps the aerosol perpendicular to the optical path. Optical trap absorbs the stray beam. D1- D4 are the signal receivers, each consists of the PSA and photodetector. The scattering angles aligned by various detectors are: D1: 30 deg; D2: 60 deg; D3: 115 deg; D4:160 deg. Aerosol flow is focused by the nozzle after the dilution system. (b) Photo graph of the sub-four optical fiber bundle for transferring the analyzed scattering light. (c) The key component for the polarization analyzer, the top assembles four pieces of polarized film with the specific analyzing directions. The bottom assembles the fiber bundles. (d) The schematic of (c), FB is the optical fiber bundles, FP is film polarizer, the direction of polarization has been labeled by dot lines, FBI is the fiber bundle integration.
As shown in Figs. 1(b)-1(d), our PSA (polarization state analyzer) channel is an optical beam splitter with four fiber bundle channels. The scattered light is detected synchronously by four bundle ends with integrated linear polarization films. Each fiber bundle is 1.5mm in diameter and is connected to different detectors to obtain the corresponding polarization analyzing signals. As marked with the dotted lines in Fig. 1 (d), the scattered light is analyzed by four diaphragm type linear polarizers at 0°, 90°, 45° and 135°.
During experiments, the airflow needs to be controlled stably and the samples need to be dispersed completely. Here we employ a nozzle with the diameter of 0.5mm and the light detection area has the height of 1mm and the radius of 0.5mm. The velocity ν of vacuum pump is 1L/min. In order to disperse different types of aerosols, two sets of aerosol generator systems have been designed applying for the water soluble inorganic salts and chemical stable particulates, respectively (Fig. 2). The spray generator can control the size distribution by adjusting the suspension concentration, and the dust generator just uniformly disperses aerosols but cannot change the original size distribution of samples. During our experiments, the two generating systems share the same air compressor. The graphite sample (Dongguan XGL Graphite) and silicon dioxide sample (Shanghai Naiou Nano Technology) are dispersed by the dust generator, and sodium sulfate aerosols (Tianxiang) are generated by the spray generator. The mean diameter of all the test aerosols is 2 microns.
4. Simulation and experimental results
4.1. Simulations
Here we employ the 45 degree linear polarization (the incident Stokes vector can be denoted as (1, 0,1,0) as the incident polarized state, and according to (2) the Stokes vector of scattered light can be deduced as (S11, S12, S33, -S34). In the first version of prototype used in this paper, we can just measure I, Q, and U items of Stokes vector, so in the following experiments, we simulate and discuss the Q and U items of scattered light. We calculate the polarization scattering angular curves of various aerosol samples and pay particular attention to these typical scattering angel: 60 deg, 115 deg and 160 deg at the scattering plane, because we notice that these angles have been selected in the commercial aerosol analysis equipment based on optical scattering (Continuous Particulate Analyzer, Environnement S.A).
Q and U of four kinds of aerosols were simulated with scattering angles between 30 deg to 180 deg (Figs. 3(a) and 3(b)). Values of Q vary obviously between 30 to 60 degree, and 160 to 180 degree, and change smoothly between 60 and 160 degree (Fig. 3 (a)). Also values of U demonstrate the similar regulation (Fig. 3 (b)). So the Stokes parameters at these three angles (60, 115 and 160 degree) are relatively stable with the fluctuation of scattering angle, proper for the realization of experimental measurements. So we select the values of Q and U at these angles to observe the possibility of Stokes items to differentiate aerosols shown as (Figs. 3(c) and 3(d)).
Fig. 3 (a) Q with the scattering angle between 30 degree and 180 degree for Na2SO4, SiO2, PbO and C, (b) U with the scattering angle, (c) Column diagram of Q at the scattering angles of 60 degree, 115 degree and 160 degree upon Na2SO4, SiO2, PbO and C, (d) Column diagram of U at the scattering angles of 60 degree, 115 degree and 160 degree upon Na2SO4, SiO2, PbO and C
The simulated four aerosol samples have different complex refractive index, which is the basic optical feature related to their own composition. The real parts of refractive indexes of Na2SO4, SiO2 and PbO increase gradually, causing respective scattering and then polarization angular spectrum. Compared with other aerosols, black carbon has a larger imaginary part of refractive index, meaning additional strong absorption and then causing some particularity on polarization characters. From Figs. 3(c) and 3(d), Q at 60 and 115 deg can provide a good contrast among different aerosols, U only at 160 degree can also identify aerosols as a possible indicator.
The above simulations sets the geometric mean diameter as 2 μm, then we need check whether Q and U can act as a good character for aerosol recognition when the mean size change (0.4, 0.7, 1, 2, 3, 4 and 5 μm), as shown in Figs. 4(a) and 4(b). Figure 4 shows an apparent increase of polarization parameters with the increasing size, which can be explained by the stronger scattering intensity by larger aerosols. It seems that Q still work as an identification indicator for the aerosols with a diameter larger than 1 μm. Especially for PbO samples with a higher refractive index and black carbon samples with a stronger absorption, the Q parameter shows a good contrast to other aerosols. However, the real aerosol samples generally have a wide aerosol size range, so the important influence of aerosol size on the value of Q and U items imply the difficulty of using them in a field test. About this, next we will observe the normalized polarization parameters divided by the scattering intensity I.
Fig. 4 (a) Q at 115 deg from four simulated samples with the aerosol size of 0.4, 0.7, 1, 2, 3, 4 and 5 um, (b) U at 115 deg from four simulated samples with the aerosol size of 0.4, 0.7, 1, 2, 3, 4 and 5 um
Figures 5(a) and 5(b) show the scattering angular spectrum of Q/I and U/I for four simulated samples. Apparently by normalization, the curve of polarization parameters of black carbon is markedly different from the others, and U/I seems has a better distinguishing ability. By observing the fluctuation of polarization parameters with the scattering angle, the value of Q/I and U/I at 60° and 115° is relatively stable suitable for the real measurements. As shown in Figs. 5(c) and 5(d), the polarization contrast at three specific scattering angles confirm the Q/I and U/I at 60° and 115° have a better discrimination to black carbon sample, and the potential to identify other kinds of aerosols still need to be checked by experiments.
Fig. 5 (a) Q/I with the scattering angle between 30 degree and 180 degree for Na2SO4, SiO2, PbO and C, (b) U/I with the scattering angle between 30 degree and 180 degree for Na2SO4, SiO2, PbO and C, (c) Column diagram of Q/I at the scattering angles of 60 degree, 115 degree and 160 degree upon Na2SO4, SiO2, PbO and C, (d) Column diagram of U/I at the scattering angles of 60 degree, 115 degree and 160 degree upon Na2SO4, SiO2, PbO and C
In Fig. 6, we observe the simulated Q/I and U/I at 60° and 115° for various mean diameter settings. It is clear that all the polarization parameters are relatively stable for aerosols larger than 2 μm. Polarization characters at different detected angles have respective aerosol recognition performance. Among them U/I at 115° shows a negative value due to the strong absorption independent with the size fluctuation, and can be thought as a specific index to identify the black carbon samples.
Fig. 6 (a) Q/I at 60 degree for different aerosol sizes; (b) Q/I at 115 degree for different aerosol sizes; (c) U/I at 60 degree for different aerosol sizes; (d) U/I at 115 degree for different aerosol sizes.
After removing carbon samples using U/I at 115°, U/I at 60° can also be used to differentiate heavy metal type PbO aerosols from other samples, due to the enough contrast of the simulated results between PbO with others. However, for PbO aerosols, considering the value change smaller than the wavelength, the polarization characterization still depend on the capability of identifying small aerosols with the size of subwavelength range. Correspondingly, we divide aerosols into different size groups according to the forward scattering intensity at about 10° in our optical scattering measurements, the similar operation has been confirmed in [13].
These simulation results imply that the scattered Stokes parameters can distinguish aerosols with different refractive indexes. For some kinds of aerosols with a strong optical absorption or a high refractive index, the corresponding polarization contrast can be enough for a feasible identification. For the other two kinds of aerosols, the simulation results show the complexity related to particle size. Correspondingly, the valid of polarization parameters to differentiate different samples should be investigated case by case through experimental tests.
4.2. Experimental results
Na2SO4, Arizona dust and soot powders are chosen as the experimental samples, and they are dispersed by spray and dust generators completely. By detecting the polarization analyzing signals of these poly-dispersed aerosol samples, we can get statistical results of two polarization indexes, Q and U at 60 and 115 degree, upon three samples (Fig. 7 and Fig. 8). Figure 7 and Fig. 8 show the average value and the frequency distribution histogram of the measured polarization indexes, Q and U, respectively. It should be noted that the average value in Fig. 7 and Fig. 8 cannot be quantitatively compared with the simulated value in Fig. 3 and Fig. 4. Because without the normalization by intensity, the simulated values depend on the simulated photon number and the experimental measurements depend on the incident light.
Fig. 7 Experimental results of Q at 60 and 115 degree from Soot, Arizona dust and Na2SO4 samples, (a) the average values of Q; (b) the frequency distribution of Q at 60 deg; (c) the frequency distribution of Q at 115 deg.
Fig. 8 Experimental results of U at 60 and 115 degree from Soot, Arizona dust and Na2SO4 samples, (a) the average values of U; (b) the frequency distribution of U at 60 deg; (c) the frequency distribution of U at 115 deg.
By observing Fig. 7 and Fig. 8, the experimental results are basically as predicted by the simulations. For these two polarization characters, the distinguishing ability of Q seems better than U, and U can be a specific indicator of soot aerosols because of its bigger contrast of value between soot with other samples. For these two scattering angles, the combination of two polarization Stokes elements at 60 deg can recognize three kinds of aerosols very well. Among these three samples, the dust sample shows a relatively different measured polarization features with the simulation, probably due to the complex composition of Arizona dust except for the SiO2.
After normalization by the scattered light intensity, we can get Fig. 9 from Fig. 7 and Fig. 10 from Fig. 8. Judging from the recognition ability of particulate matter, the normalized polarization indexes show a better identifying feasibility. By combination of Fig. 9 and Fig. 10, whether we use the two polarization characters at the same detection angle, or we use the same polarization character at both angles, the differentiation of three kinds of samples can be realized. More concretely, the polarization features of soot samples always have a better distinction with other type samples, implying that polarization characterization can be used as a specific recognition method for carbonaceous atmospheric matters.
Fig. 9 Experimental results of Q/I at 60 and 115 degree from Soot, Arizona dust and Na2SO4 samples, (a) the average values of Q/I; (b) the frequency distribution of Q/I at 60 deg; (c) the frequency distribution of Q/I at 115 deg.
Fig. 10 Experimental results of U/I at 60 and 115 degree from Soot, Arizona dust and Na2SO4 samples, (a) the average values of U/I; (b) the frequency distribution of U/I at 60 deg; (c) the frequency distribution of U/I at 115 deg.
For the other two samples, the dust samples can be observed with submicron-level mesoporous structure under electron microscope, and the Na2SO4 as a water soluble salt can show an unstable refractive index due to different water content. Our experiment confirms that measurement of single scattering Stokes parameters at fixed angle can differentiate dust and Na2SO4 samples. This cannot be shown by our theoretical derivation and more measurements are required to confirm our results and explain the discrepancy.
The above Fig. 7-10 show two groups of polarization parameters of three kinds of aerosol samples. One group includes Q and U at 60 and 115 deg, and the other group includes Q/I and U/I at two scattering angles. The different average value and FDH show the potential of non-normalized and normalized index group to realize an online aerosol classification and recognition. We can generate a mixed sample composed of three types of aerosols by a smog chamber simultaneously connected with the spray generator and the dust generator shown in Fig. 2. We designed three mixed samples as shown in Fig. 11. For three stack columns of each mixed sample, column A represents the mixing ratio of three kinds of aerosols during sample preparation, column B and C show the classification results based on two groups of polarization parameters, respectively. In order to estimate the aerosol ratio in a mixed sample, we employ the FDH curves of various polarization parameters shown in Fig. 7-10. The classification algorithm has been present in our previous paper [41], is equivalent to solving the following function with the condition of minimizing the error parameter,ε, where α represents the percentage:
By comparing experimental results in Fig. 11, we can see that the classification errors for all mixed samples are no more than 5%. Specifically, for sample I with a high proportion of Arizona dust, both polarization index groups overestimate the carbon content a little. For sample III with a high proportion of carbon, non-normalized index group underestimate the carbon content a little. For sample II with a similar proportion of different types of aerosols, non-normalized index group obviously can get a better classification result. The experiments of mixed samples confirm two points: one is the potential of non-normalized polarization parameters in aerosol identification; the other is the enhancement of aerosol classification by a parallel synchronous detection of polarization parameters.
Finally focusing on the carbon aerosols in the real atmosphere, we have realized a preliminary comparison with other on-line devices during a field test [40]. Here we employed our prototype and used U/I as an indicator to characterize absorbed aerosols regardless of their size. Our equipment was placed in an online monitoring station of Chinese Research Academy of Environmental Science in Beijing. The measured aerosols passed through the optical area along the gas path, and their multi-angle scattering pulses and polarization analysis were recorded one by one in real time. According to the U/I value of individual aerosol, we can identify them one by one and then complete the soot counting. During experiments, according to the airflow per unit time and the soot counting, we can get the number concentration with time in Fig. 12. Figure 12 is a qualitative comparison of the dynamic change of the carbon aerosol content between our prototype and the Thermal-Optical Analysis for Organic and Elemental Carbon (OC/EC, Elemental Analysis, Inc.). Figure 13 is photos of this on-site measurement. Here the collected suspended particulates passed through a PM10 cutting head and then were measured by our polarization optical device and the Thermal-Optical Analysis device synchronously. It should be noted that the measurement unit is number concentration per unit time in Fig. 12(a), different from the mass concentration per unit time in Fig. 12(b). Apparently, our polarization optical technique showed a similarly changing tendency with the traditional OC/EC. As OC/EC uses the filter to collect carbonaceous aerosols, the measurement curve will be stopped for a while like the gap in Fig. 12(b) when the filter is replaced regularly. Moreover, our prototype adopts a non-invasive detection mode, can keep working without interruption more than 20 days, and can refresh our analysis with a time interval less than 1 minute, suitable for a local on-line long term observation of aerosols.
Fig. 12 (a) Number Concentration of BC measured by the prototype, (b) Mass Concentration of BC measured by OC/EC
5. Conclusion
In this paper, we investigate the feasibility of aerosol recognition and classification based on polarization parameter groups. We develop an online polarization analysis instrument to record Stokes parameters at multiple scattering angles synchronously. When suspended aerosols pass through the optical detection area, we can extract non-normalized Q and U and normalized Q/I and U/I at 60 and 115 degree one by one in real time. The suitable detection scheme and the potential of various indexes have been evaluated by theoretical simulations of polarization scattering characteristics of several types of aerosol samples.
Experimental and simulation results show that Stokes parameters at different scattering angles have respective capabilities of distinguishing aerosols, thus an appropriate combination of polarization indexes can enhance the effective aerosol differentiation. Except for the experimental average value, we propose the frequency distribution of a polarization character and confirm the validity to identify and classify aerosol type by experiments of single type samples and mixed samples. The polarization index groups in this paper are composed of non-normalized and normalized Stokes parameters, respectively. Whichever parameter group we use, we can get a good agreement between the classification results and the preset ratio, which confirms a good application prospect of synchronously measured polarization parameters applied in online air quality monitoring. The difference between experiments and simulations also indicate that we need develop a complicated modeling system focusing on the complex physical properties of suspended atmospheric matters.
Funding
National Key R&D Program of China (2016YFF0103000), and National Natural Science Foundation of China (NSFC) (41475125).
References
1. M. Leśkiewicz, M. Kaliszewski, M. Wlodarski, J. Mlynczak, Z. Mierczyk, and K. Kopczynski, “Improved real-time bio-aerosol classification using artificial neural networks,” Atmos. Meas. Tech. 11(11), 6259–6270 (2018). [CrossRef]
2. Q. Zhang, K. He, and H. Huo, “Policy: cleaning China’s air,” Nature 484(7393), 161–162 (2012). [PubMed]
3. L. Han, W. Zhou, W. Li, and L. Li, “Impact of urbanization level on urban air quality: A case of fine particles (PM(2.5)) in Chinese cities,” Environ. Pollut. 194, 163–170 (2014). [CrossRef] [PubMed]
4. R.-J. Huang, Y. Zhang, C. Bozzetti, K.-F. Ho, J.-J. Cao, Y. Han, K. R. Daellenbach, J. G. Slowik, S. M. Platt, F. Canonaco, P. Zotter, R. Wolf, S. M. Pieber, E. A. Bruns, M. Crippa, G. Ciarelli, A. Piazzalunga, M. Schwikowski, G. Abbaszade, J. Schnelle-Kreis, R. Zimmermann, Z. An, S. Szidat, U. Baltensperger, I. E. Haddad, and A. S. H. Prévôt, “High secondary aerosol contribution to particulate pollution during haze events in China,” Nature 514(7521), 218–222 (2014). [CrossRef] [PubMed]
5. X. Y. Zhang, J. Z. Wang, Y. Q. Wang, H. L. Liu, J. Y. Sun, and Y. M. Zhang, “Changes in chemical components of aerosol particles in different haze regions in China from 2006 to 2013 and contribution of meteorological factors,” Atmos. Chem. Phys. 15(22), 12935–12952 (2015). [CrossRef]
6. L. Huang and G. Wang, “Chemical characteristics and source apportionment of atmospheric particles during heating period in Harbin, China,” J. Environ. Sci. (China) 26(12), 2475–2483 (2014). [CrossRef] [PubMed]
7. E. Järvinen, O. Kemppinen, T. Nousiainen, T. Kociok, O. Möhler, T. Leisner, and M. Schnaiter, “Laboratory investigations of mineral dust near-backscattering depolarization ratios,” J. Quant. Spectrosc. Radiat. Transf. 178, 192–208 (2016). [CrossRef]
8. X. Xu, W. Zhao, B. Fang, J. Zhou, S. Wang, W. Zhang, D. S. Venables, and W. Chen, “Three-wavelength cavity-enhanced albedometer for measuring wavelength-dependent optical properties and single-scattering albedo of aerosols,” Opt. Express 26(25), 33484–33500 (2018). [CrossRef] [PubMed]
9. D. A. Fischer and G. D. Smith, “A portable, four-wavelength, single-cell photoacoustic spectrometer for ambient aerosol absorption,” Aerosol Sci. Technol. 52(4), 393–406 (2018). [CrossRef]
10. P. H. McMurry, “A review of atmospheric aerosol measurements,” Atmos. Environ. 34(12-14), 1959–1999 (2000). [CrossRef]
11. Y.-L. Pan, C. Wang, L. A. Beresnev, A. J. Yuffa, G. Videen, D. Ligon, and J. L. Santarpia, “Measurement of back-scattering patterns from single laser trapped aerosol particles in air,” Appl. Opt. 56(3), B1–B4 (2017). [CrossRef] [PubMed]
12. Y.-Y. Kao, S.-H. Huang, C.-C. Chen, C.-S. Tsai, C.-H. Wu, and W.-Y. Lin, “Measurement and evaluation of elastic light scattering from a single levitated irregular particle,” Aerosol Air Qual. Res. 17(5), 1256–1266 (2017). [CrossRef]
13. J.-B. Renard, F. Dulac, G. Berthet, T. Lurton, D. Vignelles, F. Jegou, T. Tonnelier, M. Jeannot, B. Coute, R. Akiki, N. Verdier, M. Mallet, F. Gensdarmes, P. Charpentier, S. Mesmin, V. Duverger, J.-C. Dupont, T. Elias, V. Crenn, J. Sciare, P. Zieger, M. Salter, T. Roberts, J. Giacomoni, M. Gobbi, E. Hamonou, H. Olafsson, P. Dagsson-Waldhauserova, C. Camy-Peyret, C. Mazel, T. Decamps, M. Piringer, J. Surcin, and D. Daugeron, “LOAC: a small aerosol optical counter/sizer for ground-based and balloon measurements of the size distribution and nature of atmospheric particles–Part 1: Principle of measurements and instrument evaluation,” Atmos. Meas. Tech. 9(4), 1721–1742 (2016). [CrossRef]
14. P. H. Kaye, E. Hirst, R. S. Greenaway, Z. Ulanowski, E. Hesse, P. J. Demott, C. Saunders, and P. Connolly, “Classifying atmospheric ice crystals by spatial light scattering,” Opt. Lett. 33(13), 1545–1547 (2008). [CrossRef] [PubMed]
15. J.-B. Renard, C. Thaury, J.-L. Mineau, and B. Gaubicher, “Small-angle light scattering by airborne particulates: Environnement SA continuous particulate monitor,” Meas. Sci. Technol. 21(8), 085901 (2010). [CrossRef]
16. M. Kahnert, “Optical properties of black carbon aerosols encapsulated in a shell of sulfate: comparison of the closed cell model with a coated aggregate model,” Opt. Express 25(20), 24579–24593 (2017). [CrossRef] [PubMed]
17. G. Xu, P. G. Stegmann, S. D. Brooks, and P. Yang, “Modeling the single and multiple scattering properties of soot-laden mineral dust aerosols,” Opt. Express 25(24), A990–A1008 (2017). [CrossRef] [PubMed]
18. J. Redemann, R. P. Turco, K. N. Liou, P. B. Russell, R. W. Bergstrom, B. Schmid, J. M. Livingston, P. V. Hobbs, W. S. Hartley, S. Ismail, R. A. Ferrare, and E. V. Browell, “Retrieving the vertical structure of the effective aerosol complex index of refraction from a combination of aerosol in situ and remote sensing measurements during TARFOX,” J. Geophys. Res. D Atmospheres 105(D8), 9949–9970 (2000). [CrossRef]
19. L. Li, X. Zheng, Z. Li, Z. Li, O. Dubovik, X. Chen, and M. Wendisch, “Studying aerosol light scattering based on aspect ratio distribution observed by fluorescence microscope,” Opt. Express 25(16), A813–A823 (2017). [CrossRef] [PubMed]
20. A. Kalume, L. A. Beresnev, J. Santarpia, and Y.-L. Pan, “Detection and characterization of chemical aerosol using laser-trapping single-particle Raman spectroscopy,” Appl. Opt. 56(23), 6577–6582 (2017). [CrossRef] [PubMed]
21. Y. Wang, H. He, J. Chang, N. Zeng, S. Liu, M. Li, and H. Ma, “Differentiating characteristic microstructural features of cancerous tissues using Mueller matrix microscope,” Micron 79, 8–15 (2015). [CrossRef] [PubMed]
22. L. Bi, W. Lin, D. Liu, and K. Zhang, “Assessing the depolarization capabilities of nonspherical particles in a super-ellipsoidal shape space,” Opt. Express 26(2), 1726–1742 (2018). [CrossRef] [PubMed]
23. W. Espinosa, L. A. Remer, O. Dubovik, L. Ziemba, A. Beyersdorf, D. Orozco, G. Schuster, T. Lapyonok, D. Fuertes, and J. V. Martins, “Retrievals of aerosol optical and microphysical properties from Imaging Polar Nephelometer scattering measurements,” Atmos. Meas. Tech. 10(3), 811–824 (2017). [CrossRef]
24. H. He, L. Ran, N. Zeng, P. Li, Z. Chen, X. Liu, and H. Ma, “Mueller matrix polarimetry - an emerging new tool for characterizing the microstructural feature of complex biological specimen, IEEE/OSA,” J. Lightwave Technol., doi:. [CrossRef]
25. Y. Wang, R. Liao, J. Dai, Z. Liu, Z. Xiong, T. Zhang, H. Chen, and H. Ma, “Differentiation of suspended particles by polarized light scattering at 120°,” Opt. Express 26(17), 22419–22431 (2018). [CrossRef] [PubMed]
26. W. Dunkhorst, P. Lipowicz, W. Li, C. Hux, Q. Wang, and W. Koch, “In-situ characterization of e-cigarette aerosols by 90°-light scattering of polarized light,” Aerosol Sci. Technol. 52(7), 717–724 (2018). [CrossRef]
27. H. He, C. He, J. Chang, D. Lv, J. Wu, C. Duan, Q. Zhou, N. Zeng, Y. He, and H. Ma, “Monitoring microstructural variations of fresh skeletal muscle tissues by Mueller matrix imaging,” J. Biophotonics 10(5), 664–673 (2017). [CrossRef] [PubMed]
28. X. Li, R. Liao, J. Zhou, P. T. Y. Leung, M. Yan, and H. Ma, “Classification of morphologically similar algae and cyanobacteria using Mueller matrix imaging and convolutional neural networks,” Appl. Opt. 56(23), 6520–6530 (2017). [CrossRef] [PubMed]
29. M. I. Mishchenko and L. D. Travis, “Satellite retrieval of aerosol properties over the ocean using polarization as well as intensity of reflected sunlight,” J. Geophys. Res. 102(D14), 16989–17013 (1997). [CrossRef]
30. B. Cairns, B. E. Carlson, A. A. Lacis, and E. E. Russell, “An analysis of ground-based polarimetric sky radiance measurements, in polarization: measurement, analysis, and remote sensing,” Proc. SPIE 3121, 382–393 (1997). [CrossRef]
31. D. Daugeron, J.-B. Renard, B. Gaubicher, B. Couté, E. Hadamcik, F. Gensdarmes, G. Basso, and C. Fournier, “Scattering properties of sands. 1. Comparison between different techniques of measurements,” Appl. Opt. 45(32), 8331–8337 (2006). [CrossRef] [PubMed]
32. M. Gao, P.-W. Zhai, B. Franz, Y. Hu, K. Knobelspiesse, P. J. Werdell, A. Ibrahim, F. Xu, and B. Cairns, “Retrieval of aerosol properties and water-leaving reflectance from multi-angular polarimetric measurements over coastal waters,” Opt. Express 26(7), 8968–8989 (2018). [CrossRef] [PubMed]
33. A. N. Pavlov, E. Zubko, O. G. Konstantinov, K. Shmirko, A. Y. Mayor, and G. Videen, “Videen, Gorden. Vertical profile of polarization over Vladivostok using horizon shadowing: Clues to understanding the altitude variation of reflectance of aerosol particles,” J. Quant. Spectrosc. Radiat. Transf. 204, 94–102 (2018). [CrossRef]
34. L. Bi, W. Lin, Z. Wang, X. Tang, X. Zhang, and B. Yi, “Optical modeling of sea salt aerosols: The effects of nonsphericity and inhomogeneity,” J. Geophys. Res. D Atmospheres 123(1), 543–558 (2018). [CrossRef]
35. Claus Weitkamp. Lidar: Range-Resolved Optical Remote Sensing of the Atmosphere. Springer series in optical sciences, v. 10, New York: Springer Science+Business Media, 2005; pp. 19–42.
36. V. Freudenthaler, M. Esselborn, M. Wiegner, B. Heese, M. Tesche, A. Ansmann, D. Mueller, D. Althausen, M. Wirth, A. Fix, G. Ehret, P. Knippertz, C. Toledano, J. Gasteiger, M. Garhammer, and A. Seefeldner, “Depolarization ratio profiling at several wavelengths in pure Saharan dust during SAMUM 2006. Tellus. Ser. B Chem,” Phys. Meteorol. 61, 165–179 (2009).
37. G. Roy, “Lidar polarization discrimination of bioaerosols,” Opt. Eng. 49(11), 116201 (2010). [CrossRef]
38. K. Sassen, “The polarization lidar technique for cloud research: A review and current assessment,” Bull. Am. Meteorol. Soc. 72(12), 1848–1866 (1991). [CrossRef]
39. K. Knobelspiesse, B. Cairns, M. Mishchenko, J. Chowdhary, K. Tsigaridis, B. van Diedenhoven, W. Martin, M. Ottaviani, and M. Alexandrov, “Analysis of fine-mode aerosol retrieval capabilities by different passive remote sensing instrument designs,” Opt. Express 20(19), 21457–21484 (2012). [CrossRef] [PubMed]
40. D. Li, N. Zeng, D. Zhan, Y. Chen, M. Zeng, and H. Ma, “Differentiation of soot particulates in air using polarized light scattering method,” Appl. Opt. 56(14), 4123–4129 (2017). [CrossRef] [PubMed]
41. Y. Chen, N. Zeng, S. Chen, D. Zhan, Y. He, and H. Ma, “Study on morphological analysis of suspended particles using single angle polarization scattering measurements,” J. Quant. Spectrosc. Radiat. Transf. 224, 556–565 (2019). [CrossRef]
42. S. P. Burton, R. A. Ferrare, M. A. Vaughan, A. H. Omar, R. R. Rogers, C. A. Hostetler, and J. W. Hair, “Aerosol classification from airborne HSRL and comparisons with the CALIPSO vertical feature mask,” Atmos. Meas. Tech. 6(5), 1397–1412 (2013). [CrossRef]
43. A. Ansmann, D. Müller, U. Wandinger, and R. E. Mamouri, Lidar profiling of aerosol optical and microphysical properties from space: overview, review, and outlook, Proc. SPIE 8795, First International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2013), 879502, 2013. [CrossRef]
44. X. Xu and J. Wang, “Retrieval of aerosol microphysical properties from AERONET photopolarimetric measurements: 1. Information content analysis,” J. Geophys. Res. D Atmospheres 120(14), 7059–7078 (2015). [CrossRef]
45. A. Arola, G. Schuster, G. Myhre, S. Kazadzis, S. Dey, and S. N. Tripathi, “Inferring absorbing organic carbon content from AERONET data,” Atmos. Chem. Phys. 11(1), 215–225 (2011). [CrossRef]
46. J. G. Webster, Polarization measurement, Instrumentation and Sensors Handbook, CRC, 1998, Chap. 60.
47. G. David, B. Thomas, E. Coillet, A. Miffre, and P. Rairoux, “Polarization-resolved exact light backscattering by an ensemble of particles in air,” Opt. Express 21(16), 18624–18639 (2013). [CrossRef] [PubMed]
48. J. B. Renard, J. C. Worms, T. Lemaire, E. Hadamcik, and N. Huret, “Light scattering by dust particles in microgravity: polarization and brightness imaging with the new version of the PROGRA2 instrument,” Appl. Opt. 41(4), 609–618 (2002). [CrossRef] [PubMed]
49. C. F. Bohren and D. R. Huffman, Absorption and scattering of light by small particles. New York: John Wiley and Sons; 1983.
50. P. Kulkarni, P. A. Baron, and K. Willeke. Size Distribution Characteristics of Aerosols, in Aerosol measurement: principles, techniques, and applications, John Wiley & Sons, 2011, Chap. 4.
51. T. Hussein, M. Dal Maso, T. Petaja, I. K. Koponen, P. Paatero, P. P. Aalto, K. Hameri, and M. Kulmala, “Evaluation of an automatic algorithm for fitting the particle number size distributions,” Boreal Environ. Res. 10(5), 337 (2005).
52. C. N. Davies, “Size distribution of atmospheric particles,” J. Aerosol Sci. 5(3), 293–300 (1974). [CrossRef]
53. J.-B. Renard, C. Thaury, J.-L. Mineau, and B. Gaubicher, “Small-angle light scattering by airborne particulates: environment S.A. continuous particulate monitor,” Meas. Sci. Technol. 21(8), 85901 (2010). [CrossRef]
54. G. W. Grams, I. H. Blifford Jr., D. A. Gillette, and P. B. Russell, “Complex index of refraction of airborne soil particles,” J. Appl. Meteorol. 13(4), 459–471 (1974). [CrossRef]
55. Malvern Instruments Ltd, Sample Dispersion & Refractive Index Guide, Worcestershire, UK, 1997.
