Polarization Lidar Measurements of Dust Optical Properties at the Junction of the Taklimakan Desert–Tibetan Plateau

Abstract

Previous studies have shown that dust aerosols may accelerate the melting of snow and glaciers over the Tibetan Plateau. To investigate the vertical structure of dust aerosols, we conducted a ground-based observation by using multi-wavelength polarization lidar which is designed for continuous network measurements. In this study, we used the lidar observation from September to October 2020 at the Ruoqiang site (39.0°N, 88.2°E; 894 m ASL), located at the junction of the Taklimakan Desert–Tibetan Plateau. Our results showed that dust aerosols can be lifted up to 5 km from the ground, which is comparable with the elevation of the Tibetan Plateau in autumn with a mass concentration of 400–900 μg m⁻³. Moreover, the particle depolarization ratio (PDR) of the lifted dust aerosols at 532 nm and 355 nm are 0.34 ± 0.03 and 0.25 ± 0.04, respectively, indicating the high degree of non-sphericity in shape. In addition, extinction-related Ångström exponents are very small (0.11 ± 0.24), implying the large values in size. Based on ground-based lidar observation, this study proved that coarse non-spherical Taklimakan dust with high concentration can be transported to the Tibetan Plateau, suggesting its possible impacts on the regional climate and ecosystem.

1. Introduction

Dust is an important component of tropospheric aerosol [1,2], and it is estimated that about 2000 tons of dust are injected into the atmosphere every year [3]. It participates in various cycles of the Earth’s system [4,5,6] and directly affects the Earth’s energy budget [7,8,9,10,11]. It can also become cloud condensation nuclei and modify the microphysical characteristics of clouds, which can eventually influence the global climate [12,13,14,15]. Understanding the vertical distribution of dust properties would assist in revealing the impact of dust on air quality assessment, human health, and the climate [16,17,18,19,20].

The research conducted by Goudie et al. and Tanaka et al. demonstrated that the dust emission from East Asia is second only to the Sahara Desert, which contributes the most to the global dust budget [21,22]. The Taklimakan Desert (TD), located in Northwest China and the northern edge of the Tibetan Plateau, is the main dust source in East Asia [23,24,25]. Many studies have confirmed the transportation of dust from the TD over a long distance through the westerlies, which affected East China and even the Pacific and North America [26,27,28,29,30]. Yuan et al. [31] reported that due to the special terrain of TD and the prevailing east wind at low altitude, the dust below 5 km near the ground is not easy to be transported to the East in summer. Based on Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), Huang et al. [32] found the radial transmission of the dust aerosol from the TD in summer to the northern slope of the Tibetan Plateau which resulted in dust accumulation. Liu et al. [33] also indicated that TD dust is the main source of dust aerosol in the Tibetan Plateau. Moreover, the absorptive aerosols (e.g., dust and black carbon) transmitted to the glaciers and snow on the Tibetan Plateau can absorb more solar radiation to influence the surface radiation flux and the South Asian monsoon [34,35,36,37]. Wang et al. [38] also pointed out that the glacial retreat and snow melting in the Tibetan Plateau are closely related to the transmission of dust particles.

Earlier studies on dust in the TD mainly relied on the data of meteorological stations to analyze the frequency of dust occurrence [39,40]. The vertical distribution characteristics of dust aerosols were often studied by means of satellite remote sensing, such as CALIPSO, moderate-resolution imaging spectroradiometer (MODIS) and simulation results [41,42,43]. By analyzing the seasonal and vertical distributions of aerosols using satellite remote sensing data, Pan et al. [44] found that dust is the main type in the TD, accounting for 88.38% of all aerosols. However, due to the limitation of satellite remote sensing, it is difficult to study the diurnal variation of aerosol vertical characteristics. Besides, model simulation results still exhibit large uncertainty, especially in East Asia [45]. Ground-based lidar observation with high temporal and spatial resolutions is very useful to study the vertical characteristics of aerosols [46,47,48,49,50,51]. Zhou et al. [52] conducted a lidar observation in Northwest China, indicating that the high-intensity dust layer mainly occurs in the planetary boundary layer (PBL), and its occurrence frequency is above 88%. However, few studies have focused on dust aerosol in the TD using ground-based lidar, and most of them were concentrated in the north of TD [53,54].

In this study, we firstly investigate the vertical distribution of dust optical properties at the Ruoqiang site (88.2°E, 39.0°N) by using a ground-based polarization lidar in Autumn (from September to October 2020). Section 2 briefly describes the study area and lidar system used in this study. Two dust cases and the characteristics of dust optical properties during the study period are presented in Section 3 and Section 4. Our conclusions are then briefly summarized in Section 5.

2. Study Areas and Lidar System

2.1. Study Areas

The ground-based lidar system used in this study is located at the Ruoqiang site (39.0°N, 88.2°E; Altitude: 894 m), Xinjiang Province, China (see Figure 1). Ruoqiang lidar site is the seventh of the “Belt and Road” lidar network conducted by Lanzhou University, China. It is located in the east of the TD, and is very close to the snow cover and glaciers at the northeastern part of the Tibetan Plateau. In addition, it belongs to the warm temperate and continental desert arid climate, with much dust loading and little precipitation.

2.2. Ground-Based Multi-Wavelength Mie Polarization Lidar System (MMPL)

The multi-wavelength Mie polarization lidar system developed by Lanzhou University can collect the backscattering signals of 1064 nm, 532 nm and 355 nm simultaneously. The schematic diagram of the lidar system used in this study is shown in Figure 2. The lidar system employs a Nd: YAG laser which can emit lasers with fundamental frequency (1064 nm), double frequency (532 nm) and triple frequency (355 nm). Then, lasers are collimated and amplified by beam expanders. The telescope with a diameter of 400 mm is used to receive the backscattering signals. The signals at 532 nm and 355 nm are divided into parallel components and vertical components using polarizing beamsplitters respectively, and consequently detected by photomultiplier tubes (PMT). The 1064 nm signal is detected by an avalanche photo diode (APD). The temporal and spatial resolutions of the observed data are 2 min and 3.75 m, respectively. The lidar system is installed in a container with the temperature of around 23 °C. It is equipped with UPS that can supply power continuously for more than 8 h.

We determined the dust extinction coefficient and backscattering coefficient by use of the Fernald method [55,56]. In this study, the lidar ratio of dust aerosols is assumed to be 50 sr [57]. Then, the backscattering coefficient and extinction coefficient of dust can be retrieved from Equation (1),

$${\beta }_1\left(I\right)=-{\beta }_2\left(I\right)+\frac{X\left(I\right)\exp \left[A\left(I,I+1\right)\right]}{\frac{X\left(I+1\right)}{{\beta }_1\left(I+1\right)+{\beta }_2\left(I+1\right)}+S_1\left\{X\left(I+1\right)+X\left(I\right)\exp \left[A\left(I,I+1\right)\right]\right\}\mathrm{\Delta }Z}$$

(1)

where $A\left(I,I+1\right)=\left(S_1-S_2\right)\left[{\beta }_2\left(I\right)+{\beta }_2\left(I+1\right)\right]\mathrm{\Delta }Z$, ${\beta }_1$ and ${\beta }_2$ are the backscattering coefficient of aerosols and air molecules, respectively; while $S_1$ and $S_2$ (=8π/3) are the lidar ratios of aerosols and air molecules, respectively. X is the normalized signal after backgroud subtraption, range correction as well as overlap correction, and ∆Z is the spatial resolution.

The lidar used in this study can perform polarization measurement and distinguish spherical or non-spherical aerosols, especially in dust detection [58,59]. The particle depolarization ratio (PDR) is given by Equation (2) [57],

$${\delta }^p=\frac{\left(1+{\delta }^m\right){\delta }^{v}R-\left(1+{\delta }^v\right){\delta }^m}{\left(1+{\delta }^m\right)R-\left(1+{\delta }^v\right)}$$

(2)

where ${\delta }^m$ is the molecular depolarization ratio (the typical value is 0.00376 [60]), ${\delta }^v$ is the volume depolarization ratio, and R is the ratio of total backscattering to air molecule backscattering. The volume depolarization ratio can be calculated by multiplying the calibration factor with the ratio of vertical channel to parallel channel. The calibration factor k can be determined from the experiment [61]. In this study, the atmospheric molecular method is used. PDR is an important parameter for identifying dust aerosols. For spherical targets, the depolarization ratio is equal to 0. However, for non-spherical particles (e.g., dust aerosols), the depolarization ratio is greater than 0. For fine aerosols, the PDR is often less than 0.05 [62,63].

Ångström exponent is an important parameter to characterize the optical characteristics of atmospheric aerosols [64]. It reflects the dependence of aerosol extinction on the wavelength of incident light. It is related to the size of aerosols, and small values mean large particle sizes. The extinction-related Ångström exponent (EAE) is expressed by,

$$A_{{\lambda }_1/{\lambda }_2}^{\alpha }=\frac{\ln \left({\alpha }_{{\lambda }_1}/{\alpha }_{{\lambda }_2}\right)}{\ln \left({\lambda }_2/{\lambda }_1\right)}$$

(3)

In this study, ${\lambda }_1$ and ${\lambda }_2$ represent the wavelengths of 355 nm and 532 nm, respectively.

2.3. CALIPSO Lidar Data

The Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) is a polar orbiting satellite of the NASA A-train satellite group. It is equipped with the Cloud Aerosol Lidar With Orthogonal Polarization (CALIOP), which can provide vertical profiles of clouds and aerosols globally. Polarization measurement is conducted at the wavelength of 532 nm. The vertical resolution is 30 m below 8.2 km [65]. It is an effective tool to study dust aerosols. The level-1B data from CALIPSO products, including 532 nm total attenuated backscattering coefficient, depolarization ratio (532 nm) and attenuated color ratio, are used to confirm the vertical structure of dust aerosols observed by ground-based lidar in this study.

3. Results

3.1. Dust Case 1: 13 September 2020

A dust event which occurred in Northwest China from 13 to 14 September 2020 was also detected by the CALIPSO spaceborne lidar, as shown in Figure 3. The total attenuated backscattering coefficient at 532 nm was quite large. The depolarization ratio (532 nm) and color ratio were also high in value, indicating that it was mainly composed of particles with high non-sphericity and large particle size. It is suggested that there was an obvious thick dust layer. In addition, CALIPSO lidar found the main distribution of dust in the TD from the ground to about 2 km height with a wide horizontal scale. Moreover, dust aerosols also appeared over the Tibetan Plateau at an altitude of about 4 km near the TD.

Figure 4 shows the vertical structure of clouds and dust aerosols detected by the ground-based polarization lidar at the Ruoqiang site (39.0°N, 88.2°E, 894 m) from 13 to 14 September 2020. The attenuated backscattering coefficient both at 1064 nm and 532 nm, and the PDRs at 532 nm and 355 nm, are shown. It can be seen that there was an obvious dust layer during the observed period, mainly distributed from the ground to 5 km height. The attenuated backscattering coefficients of the dust layer at 1064 nm and 532 nm were larger than 0.003 ${\mathrm{km}}^{-1}{\mathrm{sr}}^{-1}$. PDR at 532 nm (${\delta }_{532}^p$) was greater than 0.3, and that at 355 nm (${\delta }_{355}^p$) was greater than 0.2, indicating highly non-spherical particles. It is similar to the typical values of dust aerosols observed by lidar in Dushanbe [65]. Regarding the dust event on 13 September, the dust layer was mainly distributed around 2–5.6 km height. In addition, the dust distribution exhibited an obvious stratification phenomenon. For example, the dust layer located at 2–5 km height from 8:00 to 16:00 was not distributed uniformly. On 14 September, the dust layer settled continuously with time, mainly distributed below 4 km. However, the layer was distributed at 5–6 km from 0:00 to 8:00, and then descended down to 4 km with time.

To specifically analyze the vertical structure of dust aerosols at the Ruoqiang site during the dust event, vertical profiles of dust optical properties from 13 to 14 September 2020, as shown in Figure 5. The profiles of extinction coefficients and particle depolarization ratios (PDR) both at 532 nm and 355 nm, extinction-related Ångström exponent (EAE) as well as dust mass concentration are used to discuss in detail. The uncertainty of extinction coefficient caused by the lidar ratio with error within 10% is about 8–17%, which is acceptable. On 13 September, the extinction coefficient of aerosols was small at 1–2 km height from the ground, but was large at 3 km height up to 0.27 ${\mathrm{km}}^{-1}$ at 532 nm. The variation of vertical profiles on 14 September was slightly different. Dust extinction coefficients at 532 nm reach a peak of 0.24 ${\mathrm{km}}^{-1}$ around 4 km. Moreover, it increased slightly at about 5.5 km height, which was related to another dust layer at high altitude. The variation trend of depolarization ratio profile (see Figure 5b) is consistent with the extinction coefficients on 13 September. The situation on 14 September is slightly different. The maximum value of PDR is distributed around 1 km, rather than around 4 km as with the extinction coefficient. We see that that PDR at 532 (355) nm was about 0.3 (0.2) for the dust layer. In particular, the EAE was small, indicating that the size of such particles is coarse. Wang et al. proposed a method for estimating the mass concentration of dust from lidar measurements at Dunhuang which is closed to the Ruoqiang site. So, we assume the optical properties of dust at Ruoqiang are similar to those at Dunhuang. Consequently, for calculating mass concentration of dust from extinction coefficients at 532 nm, we suppose the conversion coefficient is 0.41 [66]. In Figure 5d, we can see that the dust mass concentration was very high and reached up to 650 ${\mathsf{\mu }\mathrm{g}\mathrm{m}}^{-3}$ from 13 to 14 September.

3.2. Dust Case 2: 2 October 2020

On 2 October 2020, CALIPSO lidar captured another dust event near Ruoqiang site. Figure 6 shows that the dust concentration was high, and the depolarization ratio and color ratio also showed large values, indicating that the dust particles had high non-sphericity and large particle size. Compared with the case on 14 September, the height of the dust layer for this case was higher. In particular, some dust aerosols were very close to the Tibetan Plateau at an altitude of 4–5 km. According to ground-based observation as shown in Figure 7, the dust layer was mainly distributed from 2 km to 4 km height, and the dust concentration near the ground was low. It can be seen that the dust intensity was large, but the dust was not evenly distributed in the vertical direction. For example, there was a thin layer of dust at about 4 km height from 2:00 to 6:00 (LST), and there was a thick layer of dust at 2–3.8 km height with high concentration. PDR at 532 (355) nm is larger than 0.3 (0.2), implying that the particles were highly non-spherical.

The vertical distribution of dust optical properties (Figure 8b) shows that the dust layer was mainly concentrated at 2–4 km height from the ground, and extinction coefficients at two wavelengths were greater than 0.35 ${\mathrm{km}}^{-1}$. The extinction coefficient increased slightly between 4.2 km and 4.8 km height, mainly related to another thin dust layer. The variation of ${\delta }_{532}^p$ and ${\delta }_{355}^p$ profiles were generally consistent with those of extinction coefficients. The PDR at 355 nm of the dust layer appeared at 2–3 km height was larger than 0.25, and 0.3 for PDR at 532 nm. In addition, the EAE of the dust layer was close to zero, indicating that the particle size is very large in this layer. In particular, the mass concentration of dust was also very high, reaching up to 900 ${\mathsf{\mu }\mathrm{g}\mathrm{m}}^{-3}$.

4. Discussion

4.1. Characteristics of Dust Optical Properties

To further understand the characteristics of dust properties, we investigated the relationships among key optical properties of dust using two months’ observational data. The selection of dust data is based on the aerosol temporal and spatial distribution and PDR observed by lidar. Figure 9 presents the frequency histograms of ${\delta }_{532}^p$ (green) and ${\delta }_{355}^p$ (magenta) of dust aerosols during the observed period. It can be seen that ${\delta }_{355}^p$ is mainly concentrated in 0.2–0.3, with an average value of 0.25 and a standard deviation of 0.04. The peak of PDR is located in the range of 0.225–0.25. For the range less than 0.15 and greater than 0.3, the relative frequency of PDR at 355 nm is very small. The frequency distribution characteristics of ${\delta }_{532}^p$ are similar to those of ${\delta }_{355}^p$, but mainly concentrated in the range of 0.3–0.375, with an average value of 0.34 and a standard deviation of 0.03.

Figure 10 shows relationships between the PDR and extinction coefficient (532 nm) obtained from the polarization lidar from September to October 2020. The contoured color at panels represents the number of points for each grid. We can see that they are mainly distributed in the range of 0.1–0.4, and there is a strong linear correlation between the PDR at 355 nm and 532 nm. The averaged extinction coefficient of 532 nm is 0.22 ± 0.04 ${\mathrm{km}}^{-1}$. In particular, the extinction coefficient ${\alpha }_{532}$ increased with the PDR ${\delta }_{532}^p$, indicating that ${\alpha }_{532}$ of dust aerosol is positively correlated with non-sphericity. Analyzing the relationship between ${\alpha }_{532}$ and ${\delta }_{532}^p/{\delta }_{355}^p$ is helpful for understanding the wavelength dependence of ${\alpha }_{532}$, ${\delta }_{532}^p/{\delta }_{355}^p$ of dust aerosol is greater than 1, and mainly concentrated in 1.3–1.5, which is consistent with the results reported by Huang et al. [67]. There is a negative correlation between ${\alpha }_{532}$ and ${\delta }_{532}^p/{\delta }_{355}^p$. The ${\alpha }_{532}$ of dust aerosol is mainly about 0.15–0.3 ${\mathrm{km}}^{-1}$.

To study the relation between PDR and the size of particles, the extinction-related Ångström exponent (EAE) is used to analyze the data (as shown in Figure 11). It can be seen that the average EAE is about 0.11 ± 0.24, indicating the size of particles is quite large. The reason might be due to Ruoqiang being situated in the south-eastern part of the TD. The linear relationship between EAE and ${\delta }_{532}^p$ is not obvious, and ${\delta }_{532}^p$ is mainly concentrated in 0.3–0.4. Nevertheless, EAE decreases with the increase of PDR, and the size of dust particles with high non-sphericity is also large. In addition, ${\delta }_{532}^p/{\delta }_{355}^p$ slightly increases with EAE. It is suggested dust particles with high nonsphericity, large particle size and strong extinction can be observed at Ruoqiang. They can be lifted up to 3–5 km height and transmitted to the Tibetan Plateau, which can change the albedo of glaciers and snow, accelerate melting, and influence the radiation balance of the earth-atmosphere system [68,69].

4.2. Comparison with Previous Studies

To compare our results with other previous reports, we summarize an overview of dust PDR and the extinction-related Ångström exponent (EAE) for the TD and the Sahara Desert in literature, as shown in

Table 1. Summary of dust PDR and the extinction-related Ångström exponent (355 nm/532 nm) observed by lidar in literatures.

Dust Source	Location	PDR (355 nm)	PDR (532 nm)	Ångström Exponent	References
Taklimakan Desert	Kashi	0.28–0.32	0.36 ± 0.05	−0.01 ± 0.30	Hu et al., 2020 [70]
	Dushanbe	0.24 ± 0.003	0.33 ± 0.04	0.1 ± 0.2	Hofer et al., 2020 [71]
	Dushanbe	0.18–0.29	0.31–0.35	−0.08–0.12	Hofer et al., 2017 [72]
	Ruoqiang	0.25 ± 0.003	0.34 ± 0.04	0.11 ± 0.24	This study
Saharan Desert	Barbados	0.25 ± 0.03	0.28 ± 0.02	–	Haarig et al., 2017 [63]
	Mbour	–	0.3 ± 0.045	−0.2~0.2	Veselovskii et al., 2016 [73]
	Cape Verde	0.24–0.27	0.29–0.31	–	Groß et al., 2011 [74]
	Cape Verde	–	0.31 ± 0.1	0.2 ± 0.3	Tesche et al., 2011 [76]
	Évora	–	0.28 ± 0.04	0.0 ± 0.2	Preißler et al., 2011 [75]
	Ouarzazate	–	0.31 ± 0.02	0.04–0.35	Freudenthaler et al., 2008 [58]

. Hu et al. [70] reported that the PDR was 0.28–32 (0.36 ± 0.05) at 355 (532) nm, and the particle size was also large with EAE of −0.01 ± 0.30, observed by a Multi-wavelength lidar in Kashi located at the west of the TD. For dust aerosols in Dushanbe, the averaged PDR was 0.18–0.29 at 355 nm and 0.31–0.35 at 532 nm [71,72]. In addition, for Sahara dust, the PDR was about 0.24–0.27 at 355 nm and 0.28–0.31 at 532 nm. Moreover, the EAE of dust was between −0.2 and 0.35 [58,63,73,74,75,76]. In this study, it is represented that our results are in general agreement with those of Asian dust. The PDRs of dust aerosols are 0.25 ± 0.04 at 355 nm and 0.34 ± 0.03 at 532 nm. The Ångström exponents related to extinction coefficient are 0.11 ± 0.24.

5. Conclusions

To investigate the vertical structure of dust aerosols, we conducted a ground-based observation by use of a multi-wavelength Mie polarization lidar which was designed for continuous network measurements at Ruoqiang site in TD. The lidar system was developed with the ability to detect the total backscattering signals at 1064 nm, 532 nm and 355 nm simultaneously, and achieve polarization measurements both at visible and ultraviolet wavelengths. Based on the lidar observations from September to October 2020, we analyzed the vertical distribution of dust aerosols.

We found dust aerosols can be lifted up to 3–5 km from the ground, which is comparable with the elevation of the Tibetan Plateau in autumn (with a mass concentration of 400–900 ${\mathsf{\mu }\mathrm{g}\mathrm{m}}^{-3}$). The dust layer is quite thick with an obvious and complex structure for some dust cases. The PDR of the lifted dust aerosols is 0.34 ± 0.03 at 532 nm and 0.25 ± 0.04 at 355 nm, respectively. The extinction-related Ångström exponents are very small (0.11± 0.24). The extinction coefficient of dust layer aerosol is positively correlated with non-sphericity. Moreover, the ratio of depolarization ratios at visible and ultraviolet wavelengths (${\delta }_{532}^p/{\delta }_{355}^p$) is 1.37 ± 0.12, and has a negative correlation with the extinction coefficient of dust.

Many studies have shown that glaciers and snow near dust sources are vulnerable to dust particles, especially in Asia [77,78]. Coarse dust particles with high extinction coefficients probably have larger impacts on the melting of snow and glaciers [79,80]. This study showed that such dust particles in the TD can be lifted up to the Tibetan Plateau, implying the importance of evaluating their influence on the snow melting and radiation budget in the future.