Open Access
Add to BibSonomy
Abstract
The vertical profiles of aerosol or mixed-phase cloud optical properties (e.g. extinction coefficient) at 1064 nm are difficult to obtain from lidar observations. Based on the techniques of rotational Raman signal at 1058 nm described by Haarig et al. [Atmos. Meas. Tech. 9, 4269 (2016) [CrossRef] ], we have developed a novel rotational Raman polarization lidar at 1064 nm at Wuhan University. In this design, we optimized the central wavelength of the rotational Raman channel to 1056 nm with a bandwidth of 6 nm to increase the signal-to-noise ratio and minimize the temperature dependence of the extracted rotational Raman spectrum. And then separated elastic polarization channels (1064 nm Parallel, P and 1064 nm Cross, S) into near range (low 1064 nm P and 1064 nm S) and far range detection channels (high 1064 nm P and 1064 nm S) to extend the dynamic range of lidar observation. Silicon single photon avalanche diodes (SPAD) working at photon counting mode were applied to improve the quantum efficiency and reduce the electronic noise, which resulted in quantum efficiency of 2.5%. With a power of 3 W diode pumped pulsed Nd:YAG laser and aperture of 250 mm Cassegrain telescope, the detectable range can cover the atmosphere from 0.3 km to the top troposphere (about 12-15 km). To the best of our knowledge, the design of this novel lidar system is described and the mixed-phase cloud and aerosol optical properties observations of backscatter coefficients, extinction coefficients, lidar ratio and depolarization ratio at 1064 nm were performed as demonstrations of the system capabilities.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Aerosol lidars are important tools to investigate the detailed transport mechanisms of aerosols, as they can be used for deriving the aerosol extinction and backscatter coefficients as well as other particle optical and microphysical properties in a high temporal and vertical resolution [1–5]. In particular, the vertical profiles of aerosol and cloud extinction coefficient in the atmosphere can only be obtained by lidar observations [6]. Usually the Raman lidar technique is widely used to extract extinction coefficient profile and lidar ratio. Because compared to the high spectral resolution technique [7], the Raman lidar technique is easier to be achieved due to its relatively low requirements of laser frequency stability. Although the intensity of return signals from the Raman lidar technique is 3-4 orders of magnitude weaker, it is simpler and more robust. Thus, the Raman lidar technique is more advantageous for long-term observations [8,9]. The polarization lidar technique is upgraded from the Mie-Rayleigh scattering lidar technique, where the Mie-Rayleigh scattered light is separated into perpendicular- or parallel-polarized channels. The particle depolarization ratio (PDR) provides information about aerosol shape (e.g. dust aerosol) and cloud phases (water cloud and ice cloud) [10–13]. The extinction-to-backscatter ratio (lidar ratio, LR) provides information about the scattering/absorption predominance of aerosol particles [1,14–16] and PDR [3,17,18]. Both are important parameters for aerosol typing [19].
The advanced multi-wavelength aerosol lidars commonly can emit laser pulses at 355 nm, 532 nm to obtain the backscatter and extinction coefficients (resulting in LR) and the PDR and emit at 1064 nm to obtain backscatter coefficient at this wavelength. It is also possible to retrieve the size distribution of atmospheric aerosols using such lidar system [20–23]. However, the most of multiwavelength aerosol lidars can not provide PDR and LR at 1064 nm [24,25] so far. Only a few recent PDR measurements at 1064 nm were reported [17,26–28]. Recently it has become possible to measure the particle extinction coefficient and lidar ratio at 1064 nm [29,30]. The profiles of aerosol properties at 1064 nm are commonly used for the investigation of smoke and volcanic ash as well as mixed-phase cloud [28,31–33] due to its higher atmospheric transmission than the shorter wavelength. The measurement of aerosol and cloud optical properties at 1064 nm in particular the depolarization ratio is critical to improving the accuracy of particle micro-physical properties retrievals, especially for coarse mode particles [34]. These properties are important for the understanding of various physical properties of the atmosphere, specifically how clouds and aerosols radiatively impact the Earth in the infrared band [32]. For more robust characterization of aerosols with lidar, in addition to 355 nm and 532 nm, the data sets of LR and PDR for 1064 nm are required.
Some methods were proposed for high-quality aerosol measurements using an interference-filter-based 532 nm or 355 nm rotational Raman lidar technique [35,36]. This method was successfully realized for particle extinction profiling at 1064 nm using a customized 1058 nm interference filter [29]. To detect the weak rotational Raman signal at 1058 nm, this lidar employed a 1064 nm laser with a very high power of 30 W and a large Cassegrain telescope with a diameter of 530 mm, but a photomultiplier (PMT) with very low quantum efficiency of 0.08% was used for the signal detection. With such a design, the rotational Raman signal at 1058 nm is still weak. Usually, one kilometer of the vertical smooth window and a long data sampling period has to be applied to obtain particle optical properties with low error bound, which makes it difficult to measure the aerosol extinction coefficients inside the atmospheric boundary layer (1.5 km) [37]. Moreover, the polarization Mie-Rayleigh signals for retrieval backscatter coefficient profiles are easy to be saturated in the near range where the dense aerosol loading is present. Due to the weaker signal-to-noise ratio (SNR) of molecular signal at 1064 nm compared to other wavelengths for these lidars, the calibration of signals at 1064 nm with pure molecular signal is also challenging [31]. According to the discussion in this study [29], for accurate aerosol backscattering profiling, the detection unit of the lidar receiver must be able to resolve 6 orders of magnitude of signal strength from 1064 nm particle backscattering to almost pure 1064 nm Rayleigh backscattering (from near surface to top of troposphere) to allow proper calibration of the 1064 nm elastic signal.
To meet the large dynamic requirement of backscattering detection and better quality of rotational Raman signals in terms of lower complete overlap and higher intensity detection, in this study, we developed a dedicated rotational Raman polarization lidar at 1064 nm for profiling aerosol and cloud properties. This lidar system was deployed at Wuhan University (30.32$^\circ$ N, 114.21$^\circ$ E), which could provide a solution for the above requirement using an interference filter with optimized parameters at 1056 nm and optical spectral technique to split the backscatter signal into two detection channels with SPAD in photon counting (PC) detection mode. In section 2., the detailed description of the polarization Raman lidar at 1064 nm development with optimized 1056 nm interference filter is presented. In section 3., the performance of the lidar signals is tested using the aerosol and mixed-phase cloud cases, including the retrieval of aerosol extinction coefficient, lidar ratio, and particle depolarization ratio at 1064 nm. Finally, section 4 gives the conclusions and an outlook for future work with this advanced lidar system.
2. Lidar instrument design
2.1 System configuration
The instrument was designed as a self-standing instrument to provide the vertical profiles of backscatter coefficient, extinction coefficient, and PDR at 1064 nm. The system consists of a 1064.14 nm transmitter (diode-pumped Nd:YAG laser), a receiver (Cassegrain telescope), a spectroscopic filter with five SPAD detectors working at PC modes for separate channels (one rotational Raman channel (R), two parallel polarization Mie-Rayleigh (P) channels and two perpendicular polarization Mie-Rayleigh (S) channels), and the data acquisition system. The reason for polarization channels separation is according to the lidar signal simulations in the following spectroscopic filter part. The parameters of the main system components are listed in Table 1.
Table 1. Main components of the 1064 nm Raman polarization lidar. The polarization beam splitter (PBS), half-wave plate (HWP), focal length (FL), and T$_{\rm P/S}$: P/S polarization transmittance, photon counting (PC), photon detection efficiency (PDE), ALA (Advanced Lidar Application Inc.).
In order to facilitate operation in extreme weather conditions, such as in strong winds, the lidar system is housed into an aluminum container and accesses the atmosphere through a near-infrared (IR) transparent window with negligible depolarizing effect. It contains an industrial-personal-computer (IPC) to control the motors, DAQ system, and high voltage of SPAD with a self-developed graphical interface. The IPC also can be fully automated and remotely operated via an Ethernet connection so that the lidar becomes possible to be a "button-pressed" system. The engineering prototype of lidar system is shown in Fig. 1, and a schematic view of 1064 nm lidar system is shown in Fig. 2
Fig. 1. (a) Photo of assembled lidar system in the outside working environment in the campus of Wuhan University (dimension: 95 cm $\times$ 65 cm $\times$ 50 cm); (b) Schematic diagram of engineering prototype for assembled lidar ; (c) Schematic diagram of engineering prototype for spectroscopic box fabricated by 3D printing; (d) Schematic diagram of engineering prototype for optical core.
Fig. 2. Schematic diagram of the designed 1064 nm rotational Raman polarization lidar system, including a telescope, laser, spectroscopic filters, detectors and data acquisition chains. IF: interference filter, PBS: polarization beam splitter, HWP: half-wave plate, ND: neutral density, DP: depolarizer plate, BS: beam splitter, SPAD: single photon avalanche diode, and IPC: industrial-personal-computer. L1064 nm P: near-range parallel polarization channel, H1064 nm P: far-range parallel polarization channel, L1064 nm S: near-range perpendicular polarization channel, L1064 nm S: far-range perpendicular polarization channel, and 1056 nm Raman: rotational Raman channel at 1056 nm.
Transmitter: The transmitter part of the system consisted of a Bright Solution diode pulsed laser (C-WEDGE pn 22701, Bright Solutions s.r.l, Cura Carpignano, Italy) at 1064 nm, a beam expander, a half-wave plate (HWP), a polarization beam splitter (PBS) and a turning mirror (Fig. 1(d) and Fig. 2). The laser operates with a 1k Hz repetition rate and pulse energy of 3 mJ at 1064 nm, and a $5\times$ beam expander was added to the laser. As the linear polarization of applied laser pulses is about 99.5%, a PBS (T$_{\rm P}$:T$_{\rm S}>$ 2000:1) is adopted to improve the purity of the linear horizontal polarization of the output laser pulses to better than 99.99%. After the PBS, an HWP is installed to calibrate the misalignment of polarization plane between the emitter and receiver, and we also use it to check the possible misalignment of the horizontal polarization plane between the laser and receiver. The main parameters of the transmitter are listed in Table 1.
Receiver: Backscattered light was gathered by a specially made f/3 Cassegrain telescope with a 250 mm aperture. With a coaxial design, the incomplete overlap range of this lidar system can be smaller than 300 m. It is covered by a window with IR transparent fused silica glass. The transmission of the window is as high as 93% for the IR. A collimated lens is placed in front of the telescope focus with a 1 mm pinhole so that the light beam enters the spectroscopic system in parallel. The main parameters of the receiver are listed in Table 1, and the engineering prototype and schematic view of the design are shown in Fig. 1(d) and Fig. 2, respectively.
Spectroscopic filter: as shown in Fig. 2, after receiving the backscatter light from the telescope, it passes two motor-controlled switchable neutral density filters (ND 1 and 2) and a depolarizer plate (DP). Two ND filters are switchable to adjust the signal attenuation according to the weather conditions, and also the same operation for other ND filters in the following detection channels correspondingly. The DP converts a polarized beam of light into a pseudo-random polarized beam of light, so it can be used to calibrate the gain ratio of polarization channels according to the method described in [38–40]. The total lidar return signal was separated into five channels with specific wavelengths and polarizations: the rotational Raman signal at 1056 nm, two Mie–Rayleigh P polarization channels at 1064 nm, and two Mie-Rayleigh S polarization channels at 1064 nm. All of the spectroscopic optical components were mounted in a 3-D printed light-tight box (Fig. 1(c)).
70% of total lidar backscatter signals are reflected by beam-splitter (BS) 1 and transmitted into the rotational Raman channel to extract the 1056 nm rotational Raman lines with a specially designed interference filter as the method described in [35]. This method is successfully applied in 532 nm rotational Raman lidar, only the applied wavelength is changed to 1064 nm. Difference with the previous design with an interference filter (IF) at 1058 nm [29], we choose the center wavelength of the interference filter (IF 1) at 1056 nm, and the transmission bandwidth is 6 nm with a transmission >70%, which avoids the loss the intensity of rotational Raman signals. The IF 1 can provide more than 4 orders of magnitude out-of-band block, which the wavelength of incident light is not sensitive to the incident angle according to the narrowband interference filter parameters according to the datasets from the Alluxa manufacturer. To avoid the cross-talk from Mie-Rayleigh in extreme cases such as the strong backscatter in the mixed-phase cloud, two pieces of this interference filter were used in the rotational Raman detection channel. The detailed analysis and optimization discussions (including the temperature dependence of rotational Raman spectrum) for the selection reasons of this interference filter are described in our other study [30]. The rest 30% backscatter light is throughout to the IF 2 at 1064 nm. The PBS 1 (CCM1-PBS25-1064-HP, Thorlabs Inc., USA) separates the infrared light into two different polarization planes (P and S polarization as Fig. 2 shown).
As the dynamic detection range of SPAD is about 4-5 orders of magnitude (estimated from dark count and maximum light count listed in Table 1), the signal from single SPAD detection is not able to cover one vertical Mie-Rayleigh profile at 1064 nm. The feature of this design is to split two polarization Mie-Rayleigh signals into two individual channels for near-range and far-range detection independently, thus the dynamic detection range of lidar can be extended. We have simulated the Mie-Rayleigh lidar signals at 1064 nm using the whole lidar designed parameters to evaluate the detection range of the channels with different optical attenuation (according to the simulation method described in [41]). The SPAD channel with stronger optical attenuation is defined as the near-range detection chain, and the weaker attenuated SPAD channel is defined as far-range detection chain. The attenuation ratio between near-range and far-range detection chains was set to be 9:1, and the dynamic range of SPAD was assumed to be 4 orders of magnitude. As the results shown in Fig. 3, the near-range chain covers detection range from 315 m to 5295 m, while the far-range covers detection range from 1215 m to 15165 m within the SPAD detection dynamic range. According to the lidar signal simulation results, the final detection range of lidar signal profiles (315 m-15165 m) can meet the requirements of 1064 nm elastic signal strength. Therefore, at the next beam splitter (BS2), the intensity of each signal was separated by a ratio of 9:1 with BS 2. The 90% intensity of P or S polarization signals is transmitted into the far-range detection channels, while the remaining 10% is reflected into the near-range detection channels. Each of these channels was connected to a fiber collimator with fibers, which is focused into the SPAD detectors (SPCM-AQRH-13, Excelitas Technologies Corp., Canada) to generate electrical signals. The outputs of the SPAD single photon avalanche are amplified and digitized by ALA CLASS with five independent PC acquisition modules (CLASS Standard, ALA-Advanced Lidar Applications s.r.l., Italy) (Fig. 2). The signals are only detected in the PC detection chain by a 120 MHz discriminator and then written to an internal RAM, allowing averaging up to 16,000 acquisition cycles. It provides a range resolution of 15 m and a maximum detection range of 52.5 km. The specifications of all optical components used in the system are summarized in Table 2.
Fig. 3. The simulated lidar signals for near-range detection and far-range detection chains. The blue curve indicates the simulated lidar signal from the near-range detection channel; while the brown one is from the far-range detection channel. The grey shadow indicates the detectable dynamic range of SPAD.
Table 2. Specifications of the components used in the spectroscopic system. The beam splitter (BS), interference filter (IF), depolarizer plate (DP), polarization beam splitter (PBS), absorptive neutral density (ND) filter, optical density (OD), transmission (T), reflection (R), T$_{\rm P/S}$: P/S polarization transmittance, R$_{\rm P/S}$: P/S polarization reflectance, and WL: Wavelength.
2.2 Data pre-process and retrieval
Commissioning of the developed 1064 nm lidar system started in August 2023 on the campus of Wuhan University. All the raw datasets of lidar measurements were sampled in one minute with a vertical resolution of 15 m. These data was pre-processed before the retrieval of aerosol optical properties, which includes pre-trigger correction, dead-time correction, subtraction of background noise, and in particular data gluing. The PC signals of H channels for far-range detection are easily saturated below about 500-1000 m (depending on the polarization channels) but the signal-to-noise ratio is better than the L channels for near-range detection. Therefore the gluing of the signals in both channels is needed to extend the lidar detection dynamic range. Signal gluing algorithm is an iterative searching algorithm and comprises of two main steps: one is searching the gluing region and the other is searching the gluing point as described in [42]. The gluing region is found based on the difference of the linearly amplified near-range signal and far-range signal within a sliding window searching from near-range to far-range data points. To gluing the signals from both channels, first, the data from L1064 nm P and S channels is scaled to fit into the H1064 nm P and S channels so that glued data is obtained by combining both PC chains. As the results are shown in Fig. 4, the signals from far-range detection channels (H1064 P and H1064 S) are saturated at the beginning of a hundred meters. After gluing with the signals from near-range detection channels (L1064 P and L1064 S), the dynamic range of entire lidar signals is extended by one order of magnitude after the data gluing. In the meanwhile, the overlap of the near-range signals was found to be about 300 m, which agrees with the simulated results.
Fig. 4. The glued raw signals from both near-range detection chain (L1064 nmP and L1064 nmS) and far-range detection chain (H1064 nmP and H1064 nmP) for parallel (a) and cross (b) polarization signals, respectively. The blue curve indicates the raw lidar signal from the far-range detection chain; The brown curve indicates the raw lidar signal from the near-range detection chain; The red curve indicates the glued raw lidar signal from both detection chains. The vertical dash lines indicate the gluing region.
It was also found that the observed results are sightly different with the simulations. The near-range detection channels (L1064 P and L1064 S) were not saturated below 300 m as expected in the simulations. It probably was caused by the low coupling efficiency between the optical fiber and SPADs. Because both of their apertures are very tiny, at the engineering stage, it is very difficult to achieve optical efficiency as high as theoretical values. It was also found that the intensity ratios between H and L channels are not about 9:1 as designed. And each dynamic range was close to 5 orders of magnitude, which performed better than the assumed 4 orders of magnitude. It means the dynamic range of real observations still has room to be extended under heavy aerosol loading conditions. Usually, each new SPAD has individual and better performances than performance quotas provided by the manufactures, but their performances will be sightly decreasing with aging. Thus, for the 1064 nm lidar system designing, the SPADs with better individual performance should be selected for H channels to ensure a larger detectable range. In the optical design of this lidar system, the complete overlap range is about 300 m according to the light-tracing calculation, so the split of H and L channels are reasonable. As the simulation results are shown in Fig. 3, it requires at least 2 orders of magnitude dynamic detection range from near surface up to 300 m. In order to obtain a lower lidar blind zone, at least 8 orders of magnitude dynamic detection range are needed. For such a case, the combination of analog mode working for near-range detection and PC mode working for far-range detection might be a good solution.
Based on the pre-processed data, profiles of backscatter coefficient, extinction coefficient, lidar ratio (LR), volume/particle depolarization ratio (PDR) at 1064 nm can be retrieved. All the data retrieval processes of these particle optical properties are performed following the data analysis tool of PollyNET Processing Chain [43]. The atmospheric temperature and pressure profiles for the calculation of molecular backscatter and extinction coefficients profiles were obtained from GDAS (Global Data Assimilation System), which is provided by the National Weather Service’s National Centers for Environmental Prediction (NCEP, NOAA’s Air Resources Laboratory ARL [44]).
3. Performance of the lidar system
The lidar measurements were carried out on 24 October 2023 during the night time. During the measurements, the weather was calm with wind speed less than 3 m/s near-surface and a patchy mixed-phased cloud was last for a long time in the upper troposphere. All lidar raw data sampled with a temporal resolution of 1 minute and range resolution of 15 m was used to retrieve the particle optical properties after pre-processing.
3.1 Quality assurance tests
Several quality tests have been processed to tackle the quality assessment of lidar measurements, such as the Rayleigh fit and depolarization calibration test, which follow the methods described in [45–47]. The Rayleigh fit tests of two polarization signals have been done and described in our companion paper [30]. In the selected periods, the cloud layer is penetrable. Also, a small perturbation with a 10% molecular backscatter coefficient at 1064 nm was added in the error analysis (Monte-Carlo simulation) and only contributed to <10% of the relative error for the backscatter coefficient. The normalization range was chosen to be from 12 km to 13 km, which is in an aerosol-free region and still with a good signal-to-noise ratio. The reference value of aerosol contributions was set to be 0. The mean relative deviation of both Rayleigh fits within the normalization range was found to be less than 5%, which indicated the good agreements between all the lidar signals and the atmospheric molecular attenuated backscattering coefficient in the free atmosphere. The raw (the range resolution of 15 m) and smoothed (the range resolution of 600 m) range corrected signal after signal gluing can be found in Fig. 5.
Fig. 5. Range corrected signal of 1064 nm polarization rotational Raman lidar for the cloud case with (a) raw resolution and (b) 600 m vertical smooth.
The polarization calibration was made by the method in [39] (Fig. 6(a)). The molecular depolarization ratio at 1064 nm is about 0.004$\pm$0.002 based on our simulation. The volume molecular depolarization ratio in the pure molecular region is 0.019$\pm$0.022, see Fig. 6(b). The mismatch between measured and simulated values should be contributed by several depolarization factors, such as the impurity of the polarization state of the laser, the misalignment of the polarization plane of the transmitting and receiving modules, and retardance, diattenuation and depolarizing effects of the optics [48]. Therefore, after the systematic uncertainties of these depolarization factors can be assessed using the method described in [49].
3.2 Assessment of derived optical properties
Two commissioning cases focusing on aerosol and mixed-phase cloud vertical structures (backscatter and extinction coefficients profiles) as well as on their type-dependent properties (LR and PDR profiles) are presented to demonstrate 1064 nm lidar performance, in particular, the rotational Raman channel at 1056 nm.
First, the mixed-phase cloud, which consists of liquid water and ice crystals, has distinctive and well-known optical properties. The obtained optical properties of the mixed-phase cloud was aimed at verifying the system capability. Usually, due to large size of the precipitate ice crystals from the mixed-phase cloud, no wavelength-dependence of optical properties between 355 nm, 532 nm and 1064 nm can be expected. In other words, the LR and PDR for these three wavelengths can be expected similar for ice virga, thus our observation results should be comparable with the results at 355 nm and 532 nm in previous studies. In addition, the suppression of elastic backscatter signals in the rotational Raman channel can be checked by the relatively strong backscatter from the liquid water layer at the cloud top, which can be distinguished by the low volume depolarization ratio. The results were interpreted using the first results of particle depolarization ratio (PDR) and LR at 1064 nm in previous study. [29]. Our lidar data were taken between 20:00 (China Standard Time UTC+8:00, CST) and 23:30 CST on 24 October 2023. In the period, the mixed-phase clouds were present in the height range of about 7 km and 10 km, which was indicated by the temporal variation of the vertical distributions of the attenuated backscatter and the volume depolarization ratio (VDR) (Fig. 7).
Fig. 7. Temporal variation of the aerosol and cloud distribution on 24 October 2023 in terms of the attenuated backscatter signal at 1064 nm (a) and the volume depolarization ratio between the two polarized Mie–Rayleigh return channels (b).
Due to the weak Raman scattering, the Raman signal needs to be accumulated from several minutes to several hours. For rotational Raman signal in UV bandwidth, the Raman backscatter cross section is 2 orders of magnitude larger than that in near-infrared bandwidth. Therefore, the accumulation time of lidar measurements is usually larger in near-infrared. The two-hour averaging profiles of backscatter and extinction coefficients are shown in Fig. 8(a) and (b). Both profiles showed a cloud layer with a thickness of about 4 km was distributed between 7 km (bottom) and 10 km, which provides an ideal condition to obtain reasonable values for mixed-phase clouds. In the retrievals, the smooth window of 600 m was adopted in order to control the relative uncertainties within 10% for backscatter coefficient and 30% for extinction coefficient. The LR at 1064 nm is present in Fig. 8(c). In this study, the multiple-scattering effect was not considered; therefore, we selected the LR values from previous studies before the multiple-scattering corrections for comparison. The value of LR for ice virga was about 30$\pm$5 sr, and that for liquid cloud top was about 20$\pm$5 sr, which are comparable with mixed-phase cloud LR in most studies (22–25 sr at cirrus base to about 35–40 sr at cirrus top, and 17-19 sr for liquid cloud.) [29,50]. The values of PDR at 1064 nm present in Fig. 8(d) also agreed with the LR observations. The peak value of PDR was found about 0.4$\pm$0.1 for the ice virga (7 km), which is the typical value of PDR (0.40-0.45) for ice crystals [13,51]. The PDR for the liquid water cloud top (8 km) was found to be 0.10$\pm$0.01, which may suggest a mixture of primary ice crystals and liquid water. Therefore, our observation fits well into the previous findings of LR and PDR.
Fig. 8. Profiles of cloud optical properties between 5 km and 11 km height, taken on 24 October 2023, at 20:00 and 22:00 CST. (a) the extinction coefficient profiles, (b) the cloud backscatter coefficient profiles retrieved by the Raman method, (c) the cloud lidar ratio profile, and (d) the cloud particle depolarization ratio profile. The blue shadows indicate the uncertainties.
An aerosol case was selected to demonstrate the capability regarding the aerosol extinction coefficient profile at 1064 nm in the lower troposphere. Figure 7 shows the aerosol layer was below 3 km at the same period of mixed-phase cloud presence. The aerosol optical properties below 4 km are shown in Fig. 9. In order to get an acceptable uncertainty, The vertical resolution of 1-hour averaging profile was 600 m. Therefore, the minimum height with valid extinction coefficient value was as low as 1 km, and the backscatter coefficient profile was as low as 200 m (Fig. 9(a) and (b)). The values of LR (80$\pm$20 sr) and PDR (below 0.10$\pm$0.01) suggested the aerosol source was aged smoke (Fig. 9(c) and (d)) around 1 km. We deduce the smoke was probably from the dense point sources and influenced by city buildings after low-level transport because there were a lot of night firms during this season, which was also observed by [52]. Similar values for aged smoke were also found in [37], which presented the averaging value. The LR has decreased to 50$\pm$10 sr over 1.5 km, and the PDR was also slightly decreased to 0.040$\pm$0.002. We assume that the vertical decrease of LR and PDR may be due to the dilution of the smoke with background aerosols. However, because the direct observations of aerosol LR at 1064 nm were very scarce, thus no relative comparisons can be made to identify the aerosol types.
Fig. 9. Profiles of aerosol optical properties at 1064 nm below 4 km, taken on 24 October 2023, at 20:00 and 21:00 CST. (a) the aerosol extinction coefficient profiles, (b) aerosol backscatter coefficient profiles retrieved by the Raman method respectively, (c) aerosol lidar ratio profile, and (d) particle depolarization ratio profile. The blue shadows indicate the uncertainties.
4. Conclusions
Design features of the rotational Raman polarization lidar at 1064 nm at Wuhan University allowed us to use it as a stand-alone device for profiling aerosol and cloud characteristics. In the commissioning, we demonstrated the functionality of the system, as well as assessed its performance using the case of mixed-phase cloud and aerosol loading within the PBL. The dynamic range of elastic signals was extended to observe particles above 300 m to over the troposphere top using the combination of two PC detection channels. The typical value of PDR at 1064 nm for ice virga and liquid cloud top was found to be 0.40$\pm$0.10 and 0.10$\pm$0.01, and its LRs were about 30$\pm$5 sr and 20$\pm$5 sr, which fitted well into typical observed values in previous studies. It demonstrates this system can be used to provide a quantitative dataset with spatio-temporal association with other observations. The aerosol layer below 3 km was observed with the new developed 1064 nm rotational Raman polarization lidar. The extinction coefficient at 1064 nm was presented in profiling results as low as 1 km over the ground with 2-hour temporal and 600 m spatial resolutions. The LR and PDR at 1064 nm for anthropogenic aerosol emitted from the surface were observed. In future, combined with 355 nm and 532 nm Raman polarization lidar measurements, it can be used for better classification of smoke, dust, industrial emission, and other aerosol types. It also will be used to correct the assumption of LR at 1064 nm made in the retrieval for the first China’s spaceborne lidar-DQ-1 and to calibrate its PDR at 1064 nm for cirrus and aerosol analysis.
Funding
National Key Research and Development Program of China (2023YFC3007801, 2023YFC3007802); National Natural Science Foundation of China (42205130, 62105248, 62275202).
Acknowledgment
We acknowledge Mr. Lei Zhao, Mr. Hengzhi Wang and Mr. Liangzhang Yin for their technical help in lidar development. It was not possible to perform this research without their assistance.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. D. Müller, A. Ansmann, I. Mattis, et al., “Aerosol-type-dependent lidar ratios observed with Raman lidar,” J. Geophys. Res.: Atmos. 112(D16), 1 (2007). [CrossRef]
2. L. Mona, A. Amodeo, and G. D’Amico, “Multi-wavelength Raman lidar observations of the Eyjafjallajökull volcanic cloud over Potenza, southern Italy,” Atmos. Chem. Phys. 12(4), 2229–2244 (2012). [CrossRef]
3. L. Wang, S. Stanič, and W. Eichinger, “Development of an automatic polarization Raman LiDAR for aerosol monitoring over complex terrain,” Sensors 19(14), 3186 (2019). [CrossRef]
4. L. Peng, F. Yi, and F. Liu, “Optical properties of aerosol and cloud particles measured by a single-line-extracted pure rotational Raman lidar,” Opt. Express 29(14), 21947–21964 (2021). [CrossRef]
5. Y. He, F. Yi, and Z. Yin, “Mega Asian dust event over China on 27–31 March 2021 observed with space-borne instruments and ground-based polarization lidar,” Atmos. Environ. 285, 119238 (2022). [CrossRef]
6. A. Ansmann and D. Müller, “Lidar and atmospheric aerosol particles,” in Lidar: range-resolved optical remote sensing of the atmosphere, (Springer, 2005), pp. 105–141.
7. D. Xiao, N. Wang, and X. Shen, “Development of ZJU high-spectral-resolution LiDAR for aerosol and cloud: Extinction retrieval,” Remote Sens. 12(18), 3047 (2020). [CrossRef]
8. G. Pappalardo, A. Amodeo, and A. Apituley, “EARLINET: towards an advanced sustainable European aerosol lidar network,” Atmos. Meas. Tech. 7(8), 2389–2409 (2014). [CrossRef]
9. R. Engelmann, T. Kanitz, and H. Baars, “The automated multiwavelength Raman polarization and water-vapor lidar pollyXT: The neXT generation,” Atmos. Meas. Tech. 9(4), 1767–1784 (2016). [CrossRef]
10. N. Sugimoto and C. Lee, “Characteristics of dust aerosols inferred from lidar depolarization measurements at two wavelengths,” Appl. Opt. 45(28), 7468–7474 (2006). [CrossRef]
11. S. Groß, M. Esselborn, and B. Weinzierl, “Aerosol classification by airborne high spectral resolution lidar observations,” Atmos. Chem. Phys. 13(5), 2487–2505 (2013). [CrossRef]
12. Z. Yin, F. Yi, and Y. He, “Asian dust impacts on heterogeneous ice formation at wuhan based on polarization lidar measurements,” Atmos. Environ. 246, 118166 (2021). [CrossRef]
13. Y. He, F. Yi, and F. Liu, “Natural seeder-feeder process originating from mixed-phase clouds observed with polarization lidar and radiosonde at a mid-latitude plain site,” J. Geophys. Res.: Atmos. 127(5), e2021JD036094 (2022). [CrossRef]
14. A. Ansmann, M. Riebesell, and C. Weitkamp, “Measurement of atmospheric aerosol extinction profiles with a Raman lidar,” Opt. Lett. 15(13), 746–748 (1990). [CrossRef]
15. A. Ansmann, U. Wandinger, and M. Riebesell, “Independent measurement of extinction and backscatter profiles in cirrus clouds by using a combined Raman elastic-backscatter lidar,” Appl. Opt. 31(33), 7113–7131 (1992). [CrossRef]
16. L. Wang, M. B. Mačak, and S. Stanič, “Investigation of aerosol types and vertical distributions using polarization Raman lidar over Vipava valley,” Remote Sens. 14(14), 3482 (2022). [CrossRef]
17. Q. Hu, H. Wang, and P. Goloub, “The characterization of Taklamakan dust properties using a multiwavelength Raman polarization lidar in Kashi, China,” Atmos. Chem. Phys. 20(22), 13817–13834 (2020). [CrossRef]
18. Y. He, Z. Yin, and A. Ansmann, “POLIPHON conversion factors for retrieving dust-related cloud condensation nuclei and ice-nucleating particle concentration profiles at oceanic sites,” Atmos. Meas. Tech. 16(7), 1951–1970 (2023). [CrossRef]
19. A. A. Floutsi, H. Baars, and R. Engelmann, “DeLiAn–a growing collection of depolarization ratio, lidar ratio and Ångström exponent for different aerosol types and mixtures from ground-based lidar observations,” Atmos. Meas. Tech. 16(9), 2353–2379 (2023). [CrossRef]
20. I. Veselovskii, A. Kolgotin, and V. Griaznov, “Inversion of multiwavelength Raman lidar data for retrieval of bimodal aerosol size distribution,” Appl. Opt. 43(5), 1180–1195 (2004). [CrossRef]
21. M. Tesche, D. Müller, and S. Gross, “Optical and microphysical properties of smoke over Cape Verde inferred from multiwavelength lidar measurements,” Tellus B 63(4), 677–694 (2011). [CrossRef]
22. S. P. Burton, E. Chemyakin, and X. Liu, “Information content and sensitivity of the 3β+ 2α lidar measurement system for aerosol microphysical retrievals,” Atmos. Meas. Tech. 9(11), 5555–5574 (2016). [CrossRef]
23. N. Wang, D. Xiao, and I. Veselovskii, “This is FAST: multivariate Full-permutAtion based Stochastic foresT method—improving the retrieval of fine-mode aerosol microphysical properties with multi-wavelength lidar,” Remote Sens. Environ. 280, 113226 (2022). [CrossRef]
24. S. Qi, Z. Huang, and X. Ma, “Classification of atmospheric aerosols and clouds by use of dual-polarization lidar measurements,” Opt. Express 29(15), 23461–23476 (2021). [CrossRef]
25. L. Wang, Z. Yin, and Z. Bu, “Quality assessment of aerosol lidars at 1064 nm in the framework of the MEMO campaign,” Atmos. Meas. Tech. 16(18), 4307–4318 (2023). [CrossRef]
26. S. Burton, J. Hair, and M. Kahnert, “Observations of the spectral dependence of linear particle depolarization ratio of aerosols using NASA Langley airborne High Spectral Resolution Lidar,” Atmos. Chem. Phys. 15(23), 13453–13473 (2015). [CrossRef]
27. A. Tsekeri, V. Amiridis, and A. Louridas, “Polarization lidar for detecting dust orientation: system design and calibration,” Atmos. Meas. Tech. 14(12), 7453–7474 (2021). [CrossRef]
28. M. Haarig, A. Ansmann, and R. Engelmann, “First triple-wavelength lidar observations of depolarization and extinction-to-backscatter ratios of Saharan dust,” Atmos. Chem. Phys. 22(1), 355–369 (2022). [CrossRef]
29. M. Haarig, R. Engelmann, and A. Ansmann, “1064 nm rotational Raman lidar for particle extinction and lidar-ratio profiling: cirrus case study,” Atmos. Meas. Tech. 9(9), 4269–4278 (2016). [CrossRef]
30. A. Wang, Z. Yin, and S. Mao, “Measurements of particle extinction coefficients at 1064 nm with lidar: temperature dependence of rotational Raman channels,” Opt. Express 32(3), 4650–4667 (2024). [CrossRef]
31. M. Vaughan, A. Garnier, and D. Josset, “CALIPSO lidar calibration at 1064 nm: version 4 algorithm,” Atmos. Meas. Tech. 12(1), 51–82 (2019). [CrossRef]
32. R. M. Pauly, J. E. Yorks, and D. L. Hlavka, “Cloud-aerosol transport system (CATS) 1064 nm calibration and validation,” Atmos. Meas. Tech. 12(11), 6241–6258 (2019). [CrossRef]
33. S. Li, H. Di, and Q. Wang, “Retrieval of the aerosol extinction coefficient of 1064nm based on high-spectral-resolution lidar,” J. Quant. Spectrosc. Radiat. Transf. 256, 107298 (2020). [CrossRef]
34. J. Gasteiger and V. Freudenthaler, “Benefit of depolarization ratio at λ= 1064 nm for the retrieval of the aerosol microphysics from lidar measurements,” Atmos. Meas. Tech. 7(11), 3773–3781 (2014). [CrossRef]
35. I. Veselovskii, D. Whiteman, and M. Korenskiy, “Use of rotational Raman measurements in multiwavelength aerosol lidar for evaluation of particle backscattering and extinction,” Atmos. Meas. Tech. 8(10), 4111–4122 (2015). [CrossRef]
36. P. Ortiz-Amezcua, A. E. Bedoya-Velásquez, and J. A. Benavent-Oltra, “Implementation of UV rotational Raman channel to improve aerosol retrievals from multiwavelength lidar,” Opt. Express 28(6), 8156–8168 (2020). [CrossRef]
37. M. Haarig, A. Ansmann, and H. Baars, “Depolarization and lidar ratios at 355, 532, and 1064 nm and microphysical properties of aged tropospheric and stratospheric Canadian wildfire smoke,” Atmos. Chem. Phys. 18(16), 11847–11861 (2018). [CrossRef]
38. W. H. Hunt, D. M. Winker, and M. A. Vaughan, “CALIPSO lidar description and performance assessment,” J. Atmospheric Ocean. Technol. 26(7), 1214–1228 (2009). [CrossRef]
39. V. Freudenthaler, “About the effects of polarising optics on lidar signals and the Δ90 calibration,” Atmos. Meas. Tech. 9(9), 4181–4255 (2016). [CrossRef]
40. Y. Chen, Z. Bu, and X. Wang, “Development and calibration of 532 nm standard aerosol lidar with low blind area,” Remote Sens. 16(3), 570 (2024). [CrossRef]
41. S. Mao, A. Wang, and Y. Yi, “Polarization Raman lidar for atmospheric correction during remote sensing satellite calibration: instrument and test measurements,” Opt. Express 30(7), 11986–12007 (2022). [CrossRef]
42. S. Mao, Z. Yin, and L. Wang, “Aerosol optical properties retrieved by polarization Raman lidar: Methodology and strategy of a quality-assurance tool,” Remote Sens. 16(1), 207 (2024). [CrossRef]
43. Z. Yin and H. Baars, “PollyNET/Pollynet_Processing_Chain: Version 3.0,” Zenodo [code] doi.org/10.5281/zenodo.5571289, 10 (2021).
44. Z. Yin, Q. Chen, and Y. Yi, “Global data assimilation system archive information,” National Weather Service’s National Centers for Environmental Predictionhttps://www.ready.noaa.gov/gdas1.php (2023).
45. U. Wandinger, V. Freudenthaler, and H. Baars, “EARLINET instrument intercomparison campaigns: overview on strategy and results,” Atmos. Meas. Tech. 9(3), 1001–1023 (2016). [CrossRef]
46. CARS, “High Power Lidar: Standard Quality Assurance Procedures for NF operation,” ACTRIS [Document], https://www.actris.eu/topical-centre/cars/announcements-resources/documents p. Jan. (2024).
47. L. Belegante, J. A. Bravo-Aranda, and V. Freudenthaler, “Experimental techniques for the calibration of lidar depolarization channels in EARLINET,” Atmos. Meas. Tech. 11(2), 1119–1141 (2018). [CrossRef]
48. J. A. Bravo-Aranda, F. Navas-Guzmán, and J. L. Guerrero-Rascado, “Analysis of lidar depolarization calibration procedure and application to the atmospheric aerosol characterization,” Int. J. Remote Sensing 34(9-10), 3543–3560 (2013). [CrossRef]
49. J. A. Bravo-Aranda, L. Belegante, and V. Freudenthaler, “Assessment of lidar depolarization uncertainty by means of a polarimetric lidar simulator,” Atmos. Meas. Tech. 9(10), 4935–4953 (2016). [CrossRef]
50. Y. Hu, “Depolarization ratio–effective lidar ratio relation: Theoretical basis for space lidar cloud phase discrimination,” Geophys. Res. Lett. 34(11), 1 (2007). [CrossRef]
51. I. Veselovskii, P. Goloub, and T. Podvin, “Spectral dependence of backscattering coefficient of mixed phase clouds over west Africa measured with two-wavelength Raman polarization lidar: Features attributed to ice-crystals corner reflection,” J. Quant. Spectrosc. Radiat. Transf. 202, 74–80 (2017). [CrossRef]
52. Z. Yin, Q. Chen, and Y. Yi, “Low blind zone atmospheric lidar based on fiber bundle receiving,” Remote Sens. 15(19), 4643 (2023). [CrossRef]
