Open Access
Add to BibSonomy
Abstract
A portable polarization lidar system based on the division-of-focal-plane scheme has been proposed for all-day accurate retrieval of the atmospheric depolarization ratio. The polarization lidar system has been designed as a T-shaped architecture consisting of a closed transmitter and a detachable large focal receiver, which is capable of outdoor unmanned measurements. The lidar system features low cost, low maintenance and short blind range (∼100 m) by utilizing a 450 nm multimode laser diode as the light source and a polarization image sensor with four polarized channels as the detector. Validation measurements have been carried out on a near horizontal path in ten consecutive days. The linear volume depolarization ratio (LVDR) as well as its measurement uncertainty has been theoretically and experimentally evaluated without employing additional optical components and sophisticated online calibrations. The offset angle can also be accurately retrieved (i.e., −0.06°) from the four-directional polarized lidar profiles with a standard deviation of ±0.02° during the whole measurement period, which contributes negligible influence on the retrieval of the LVDR. It has been found out that the uncertainty of the LVDR was mainly originated from the random noise, which was below 0.004 at nighttime and may reach up to 0.008 during daytime owing to the increasing sunlight background. The performance of the polarization lidar system has been further examined through atmospheric vertical measurements. The low-cost low-maintenance portable polarization lidar system, capable of detecting four-directional polarized lidar signals simultaneously, opens up many possibilities for all-day field measurements of dust, cloud, urban aerosol, oriented particles, etc.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Light Detection and Ranging (Lidar), as an active remote sensing tool, has been widely employed in atmospheric aerosol sensing [1,2]. By transmitting a nanosecond pulsed laser beam into the atmosphere and then detecting the backscattering light originating from aerosols and molecules with sensitive photodetectors, the range-resolved backscattering echo related to the aerosol properties can be obtained based on the time-of-flight principle. If a linearly polarized laser beam is transmitted into the atmosphere, the backscattering light would be partly depolarized and the degree of depolarization is related to the shape of the scattering particles. By measuring the polarization property of the backscattering light, one can evaluate the particle shape and identify the particle type. Based on this principle, the polarization lidar technique emerges [3]. Nowadays, there are generally two different approaches to measure the polarization property of atmospheric aerosols, namely the dual-channel polarization lidar and the single-channel polarization lidar based on the time-division multiplexing (TDM) scheme [4–7].
The dual-channel polarization lidar is deployed widely around the world, famed for the space-borne Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) [8], in which a linearly polarized laser beam is transmitted into the atmosphere. The co- and cross-polarized backscattering echoes are separated by a polarization beam splitter (PBS) in the receiver and detected by two photodetectors, respectively. Nowadays, the dual-channel polarization lidar technique has been widely employed to classify cloud phase (e.g., ice cloud and water cloud) [9–11], distinguish the dust from other atmospheric particles [12–14], identify aerosol types [15–18], evaluate aerosol transportation [19–23], etc. However, this technique faces a great challenge in the calibration of the gain ratio, which characterizes the difference of the optical and electronic gains between the co- and cross-polarized receiver channels. Many calibration methods have been proposed, such as the “clean air” method, the rotating half-wave plate (HWP) method, and pseudo-depolarizer method, the un-polarized light method, etc. These calibration methods require additional polarization optics, precise adjustments and careful data evaluations for accurate measurement of the depolarization ratio, which are inconvenient particularly for long-term unmanned operations [24].
Single-channel polarization lidar is another way to detect the atmospheric polarization information. This technique often acquires the co- and cross-backscattering signal sequentially through a single detector based on the TDM scheme. In 1994, Eloranta et al. demonstrated a single-channel polarization lidar system, which rotated the polarization state of the transmitted laser pulse by 90° alternately by using a Pockels cell at a repetition rate of 4 kHz. This allowed the measurement of the parallel and cross polarized lidar signals with a single detector [6]. In 2017, Wang et al. proposed a polarization coherent Doppler lidar for simultaneous detection of the wind velocity and the atmospheric depolarization ratio by utilizing a PBS and a delay line based on the TDM method [25]. In 2019, Sharukh et al. developed a single detector polarization lidar using a rotating polarizer in the receiver path to discriminate the incoming light into co- and cross-polarized components for deriving the range resolved depolarization measurements of tropical cirrus [26]. The rotation of the polarizer is synchronized with the laser pulse using a dedicated trigger control. In 2017, a polarization Scheimpflug lidar system based on the Scheimpflug principle has been developed by employing two orthogonal linearly polarized 808-nm laser diodes and a 45° tilted CMOS image sensor using a TDM scheme for polarization studies of urban aerosol studies [27–29]. However, the single-channel polarization lidar needs to modulate the polarization state of the transmitted or received lights accurately and quickly, leading to a great challenge for stable operation of polarization lidar system.
In recent years, the division-of-focal-plane (DoFP) polarization image sensor, which integrates micro-polarizer elements with a focal plane and features of high-integration, perfect alignment, and real-time polarization measurement, has been developed and widely used in many fields, e.g., metal surface detection, biological tissue detection, etc. [30–35]. In this paper, a polarization-sensitive imaging lidar (PSI-Lidar) based on the DoFP scheme, capable of detecting four-directional polarized lidar signals simultaneously, has been developed for all-day field measurements of the atmospheric depolarization ratio. The PSI-Lidar system has been designed as a T-shaped architecture consisting of a closed transmitter and a detachable large focal receiver, which is capable of all-day outdoor unmanned measurements that has yet been feasible in previous studies [36]. The lidar system utilizes a 450 nm linearly polarized laser diode as the light source and a polarization sensitive image sensor with a micro-polarizer array as the detector to simultaneously acquire lidar signals at four different polarization angles, from which the depolarization ratio and the offset angle can be mathematically retrieved without employing additional optical components and sophisticated online calibrations. Atmospheric measurements as well as careful evaluations have been carried out to validate the performance of the PSI-Lidar for atmospheric polarization studies.
2. Polarization sensitive imaging lidar
2.1 Optical setup
The principle of the PSI-Lidar is shown in Fig. 1. A continuous-wave linear polarized laser beam is transmitted into the atmosphere, the scattering light from molecules or aerosol particles would be depolarized. The backscattering light is collected by a telescope and then detected by a polarization-sensitive image sensor with four different polarization channels. The system schematic, the photograph and system specifications of the PSI-Lidar are shown in Fig. 2(a), (b) and Table 1, respectively. The polarization lidar system is designed as a T-shaped architecture consisting of a closed transmitter and a detachable large focal receiver.
Fig. 2. (a) The architecture of the PSI-Lidar system. (b) The picture of the portable unmanned PSI-Lidar system, the system was deployed on the rooftop of the Graduate Education Building, Dalian University of Technology (DLUT), with about 33 m above the ground.
Table 1. System specifications of the PSI-Lidar system.
The transmitter mainly includes a laser diode, collimating optics and a system controlling unit, which is closed by a metal cover with a 90 mm diameter optical window. A 450 nm laser diode (LD) with approximately 3.5-W output power is packaged in a customized aluminum alloy mount with high-power thermoelectric coolers (TECs) and a cooling fan. The LD current and the TEC driving circuits are mounted on the baseplate. The slow axis and the polarization state of the laser beam are both parallel to the baseplate. The laser beam emitted from the laser diode is reshaped by a cylindrical lens pair to reduce the divergence of the laser beam along the fast axis and thus improve the geometrical transmission efficiency. A removable 10% reflecting mirror together with a black and white camera are also mounted in a cube along the optical path to monitor the field-of-view (FOV) of the transmitting optical system during system alignment. A linear polarizer is also placed in the optical path and installed in the cube, which was utilized to improve the degree of linear polarization (DoLP) of the laser beam. Considering a typical polarization purity of 100:1 for the 450 nm multimode laser diodes, the DoLP of the transmitted laser beam could be beyond 105:1 after passing through the linear polarizer with an extinction ratio of 2400:1 at 450nm. Finally, the laser beam is folded by an elliptical mirror and then collimated by a lens (f=400 mm, $\phi$=76 mm). The divergence of the collimated laser beam is evaluated to be about 0.11 mrad (slow axis) × 0.53 mrad (fast axis) by measuring the laser beam size (e.g., 13× 60 cm) at far distance of 1140 m. The whole optical path was covered with an L-shaped shielding tube, which is fixed on the L-shaped bar. The angle between the transmitted laser beam and the receiving optics can be manually adjusted by rotating the L-shaped bar to achieve the coincidence of the FOVs between the transmitter and receiving telescope. The L-shaped aluminum bar can be locked on the baseplate after the lidar system is well aligned.
The detachable large focal receiver mainly includes a Maksutov-Cassegrain telescope (f=1000 mm, $\phi$=105 mm), an interference filter and a polarization sensitive image sensor (IMX250MZR, 2448×2048, 3.45 µm). The atmospheric backscattering signals are received by the Maksutov-Cassegrain telescope with a cone half angle of 2.9°, and then detected by a polarization-sensitive image sensor, which is placed parallel to the equivalent lens plane of the Maksutov-Cassegrain telescope. A 450 nm interference filter with 10 nm full width at half maximum (FWHM) is placed before the polarization sensitive image sensor to suppress the sunlight background. The polarization image sensor with four-directional polarizer (0°, 45°, 90° and 135°) can capture a four-directional polarization image in a single shot, which is separated into four individual images corresponding to four different polarization angles. Thus, lidar signals at four different polarized angles can be simultaneously obtained by vertically binning the pixel intensities of the corresponding polarized image. The 0° polarizer of the image sensor, regarded as the parallel polarized channel, is supposed to be consistent with the polarization state of the emitted laser beam.
System controlling as well as data acquisition is achieved by an industrial computer installing a LabVIEW-based program developed by our research group. The PSI-Lidar system weights about 20 kg and the system dimension is labeled in Fig. 2(b), which is capable of outdoor unmanned measurements.
2.2 Lidar signal acquisition and processing
The emitted light intensity of the laser diode is on-off modulated, which is synchronized with the exposure of the image sensor. The image sensor can alternately capture the laser beam image and the background image (laser off) in the region of interest (ROI, 2448×800 pixels). Real-time subtraction was carried out to eliminate the background signal. As shown in Fig. 3, the background-corrected image can be decoupled to four polarized images with a size of 1224 (H) × 400 (V). The four raw lidar signals can thus be obtained by vertically binning the pixel intensities for each polarized image. Pixel-intensity signals of each polarized channel with 1224 data points were transform into 2448 data points through interpolation based on the Piecewise Cubic Hermite Interpolating Polynomial (PCHIP), as the original polarization signals are sampled by spatially separated pixels, as shown in Fig. 1. During the measurement, the exposure time of the image sensor automatically changes between 80-500 ms to optimize the signal to noise ratio (SNR). Meanwhile, the averaging number also changes correspondingly to achieve the same measurement time (e.g., 1 minute). The SNR analysis will be discussed in detail in Section 3.
Fig. 3. Signal processing procedure of the polarized lidar signals. (a-1) and (a-2) are the original laser beam image and the background image, respectively. (b-1), (b-2), (b-3) and (b-4) are the 90°, 45°, 135°, and 0° polarized laser beam images after background correction, respectively. (c) The pixel-intensity backscattering signals. (d) The range-intensity backscattering signals.
In this work, the image sensor is placed parallel to the lens plane, which does not satisfy the Scheimpflug principle [37]. The PSI-Lidar can only acquire clear image of the transmitted laser beam in a certain region, e.g., in the far range to optimize the far-range performance. As a result, the near-range laser beam image could be blurred due to the out-of-focus imaging. Nevertheless, employing a large f-number receiver can reduce the influence of relatively small depth-of-field. The corresponding distance of a given image pixel can be determined according to the geometrical optics. The pixel-range relationship is calibrated by measuring the backscattering signal from a tall building with 2000 m away from the lidar system [37].
Here z is the measurement distance, $\Phi $ is the swing angle of the receiving lens, L is the distance of the lens from the object plane, ${L_{\textrm{IL}}}$ is the distance between the lens and image planes, ${p_\textrm{I}}$ is the pixel position. It’s noted that the center of the image sensor is supposed to be placed coinciding with the optical axis of the lens. Finally, the range-intensity backscattering signals were calculated through the pixel-range relationship in Fig. 3(d). In the present system configuration, the minimum measurement distance is about 100 m.
2.3 Retrieval of the depolarization ratio
The PSI-Lidar system, from the emitted laser beam to the image sensor, can be interpreted as multiple functional modules with the corresponding Stokes vectors and Mueller matrices [38,39]. The influence of non-ideal polarized laser beam on the linear volume depolarization ratio (LVDR) is less than 1% by employing a high polarization extinction ratio (PER) linear polarizer to improve the DoLP of the transmitted laser beam [39]. The systematic errors introduced by the transmitter, e.g., the small depolarizing effect of the achromatic lens, and the polarization crosstalk of the Maksutov-Cassegrain telescope can be neglected [40,41]. The relative QEs of the four polarization channels can be obtained from the datasheet provided by the manufacture. Thus, the LVDR (${\delta _v}$) and the offset angle ($\theta $) can be retrieved through the Stokes-Mueller formalism, which is given by
Here $E{R_{0^\circ }}$, $E{R_{90^\circ }}$, $E{R_{45^\circ }}$ and $E{R_{135^\circ }}$ are referred to as 0°-PER, 90°-PER, 45°-PER, and 135°-PER, respectively. ${V_1}$ and ${V_2}$ are defined as
Here ${I_{0^\circ }}$, ${I_{90^\circ }}$, ${I_{45^\circ }}$ and ${I_{135^\circ }}$ are the measured signal intensities at four polarized channels, respectively. The LVDR includes the linear particle depolarization ratio (LPDR) and the depolarization ratio of the air molecules. The LPDR can be calculated as [4]
Here ${\delta _m}$ is the depolarization ratio of air molecules, which can be accurately determined according to the Rayleigh scattering model (including the Cabannes line and the pure rotational Raman spectrum) [42]. The depolarization ratios of air molecules at 253 K, 273 K, 303 K for a 10 nm bandwidth with a center wavelength of 450 nm are estimated to be 0.01371, 0.01366, 0.0136, respectively. The backscatter ratio R is defined as the ratio of the total backscattering coefficient to the molecular component. The molecular backscattering coefficients can be calculated through known temperature and pressure. As the PSI-lidar is only equipped with Mie-scattering channels, the Fernald method was utilized to invert the lidar equation to obtain the backscattering coefficient or the extinction coefficient. The key issue for the Fernald-inversion algorithm is the determination of the boundary value and the aerosol lidar ratio. The Douglas–Peucker algorithm was utilized to find a subinterval range in the far distance, where the atmosphere can be considered homogeneous. The boundary value of the aerosol extinction coefficient is then obtained according to the slope method. The aerosol lidar ratio is defined by the ratio of the extinction coefficient to the backscattering coefficient of aerosol, which depends on aerosol size, shape and refractive index, etc. The aerosol lidar ratio at 450 nm was set 44, 53, 70 for dust, clean continental and polluted continental, respectively [43].
3. Measurements and discussion
3.1 Validation measurements of the PSI-Lidar
3.1.1 Horizontal atmospheric measurements
Atmospheric remote sensing was performed on a near horizontal path with an elevation angle of 0.3° from 29th October to 8th November 2020 in Dalian, Northern China. The lidar system was deployed on the rooftop of the Graduate Education Building, Dalian University of Technology (DLUT), with about 33 m above the ground. The particle concentrations of PM10 and PM2.5, the ambient temperature, and the relative humidity (RH) were reported by a micro air quality monitor (Fairsense, A108) that was located at about 10 m away from the lidar system. Atmospheric pressure was obtained from the National Meteorological Information Center (CMIC).
SNR is a key parameter indicating the quality of the lidar signal, which has been carefully evaluated in this work. A relatively smooth subset signal in the region between 500 and 800 pixels, corresponding to a measurement range of 120 -143 m, was extracted for the SNR analysis. The signal level was evaluated from the mean value of the subset signal. The subset signal was first fitted by a 5-order polynomial and then the noise level was estimated from the standard deviation of the fitting residual. The relationship among the original lidar signal, the corresponding background intensity and the SNR is shown in Fig. 4. It should be noted that the signal and background intensities have been normalized to the same exposure time of 100 ms and the same gain of 12.
Fig. 4. The relationship among the original lidar signals, the corresponding background intensities, and SNR (a) at 90° polarized channel and (b) at 0° polarized channel. During nighttime measurements, the background intensity was generally less than 2000.
As shown in Fig. 4, the SNR of the 0° polarized lidar signal during daytime can reach up to 400, while the SNR of the 90° polarized channel was generally below 100 owing to the weak depolarized signal and the strong sunlight noise. During nighttime, the SNR of the 0° and 90° polarized channel can reach up to 900 and 500, respectively. It can be found out that the SNR generally increased with the increasing of the signal intensity during daytime. Meanwhile, the SNR increased with the decreasing of the sunlight background if the intensity of the lidar signal remained. During nighttime measurements, the SNR, which was mainly determined by the photon-response non-uniformity (PRNU) noise of the image sensor, gradually increased with the increasing of the intensity of the 90° polarized lidar signal (low light level), as shown in Fig. 4(a). However, the increment of the SNR in Fig. 4(b) was not significant for 0° polarized lidar signal at nighttime, as the PRNU coefficient became a constant for high light level conditions. It can be concluded that the noise of the lidar signal at daytime was dominated by the sunlight shot noise, while it was dominated by the PRNU of the image sensor at nighttime [44]. Moreover, the SNRs of all lidar signals can be increased by a factor of about 3 by utilizing an adaptive digital filter based on the Savitzky – Golay (S–G) filter and the Fourier analysis [45].
The temporal-spatial map of the recorded atmospheric backscattering signal, the aerosol extinction coefficient and the LPDR are shown in Fig. 5. The maximum retrieval range was set to 5 km, since the range resolution deteriorated in the far range. Meanwhile, the detected range (SNR>5) was only 2 ∼ 3 km during the heavy haze due to the strong absorption at 450 nm. Figures 6(a) and 6(b) show the atmospheric parameters (e.g., RH, temperature), particulate matter concentrations (e.g., PM2.5, PM10). According to the meteorological data and particulate matter concentrations, different lidar ratios were utilized to retrieve the extinction (or backscattering) coefficient for different atmospheric conditions and symbolized by different colors in Fig. 6(b) [43]. The median aerosol extinction, the median LVDR and the median LPDR were obtained by taking the median value along the laser beam path, as shown in Figs. 6(b) and 6(c).
Fig. 5. (a) Time-range map of the total backscattering intensity, (b) the extinction coefficient and (c) the LPDR.
Fig. 6. Temporal evolution of (a) the relative humidity, the temperature, (b) the extinction coefficients and the particle concentration, (c) linear volume depolarization ratio (LVDR) and the linear particle depolarization ratio (LPDR).
Two haze events occurred during the near-horizontal measurement. The first haze event occurred during 14:00 on October 31st to 00:00 on November 1st. The peak concentrations of the PM10 and PM2.5 were 150 µg/m3 and 81 µg/m3, respectively. The relative humidity (RH) was up to 80%. The aerosol extinction coefficient varied from 0.78 km−1 to 1.2 km−1 and the LPDR stabilized at 0.02. This illustrated that spherical particles were dominant near the ground in Dalian under polluted atmospheric conditions. Later, the north wind blew away the pollution, leading to a rapid decreasing of the extinction coefficient from 1.2 km−1 to 0.5 km−1. Meanwhile, the LPDR swiftly increased from 0.02 to 0.28 at around 00:00 on November 1st, indicating that the haze was dispersing and the atmosphere was going to be dominated by dust particles. The second haze event started at 05:00 on November 6th, the aerosol extinction coefficient increased rapidly, e.g., from 0.25 to 1.55 km−1, and the LPDR decreased to 0.024. The high peak concentrations of the PM10 (244 µg/m3) and the PM2.5 (140 µg/m3) indicated that a more severe haze event occurred. At the beginning of 18:00 on November 6th, the aerosol extinction coefficient decreased to 0.43 km−1 rapidly and the LPDR increased to about 0.14. Statistical analysis indicated that the Pearson correlation coefficient between the spatially averaged extinction coefficient and the PM concentrations were beyond 0.9, indicating good reliability of the lidar measurements.
The relationship among the LPDR, the extinction coefficient and the RH is shown in Fig. 7. It can be found out that a small LPDR (less than 0.05) generally occurred in a high RH atmospheric condition. Large extinction coefficient and small LPDR were normally observed with the appearance of high RHs, indicating strong hygroscopic growth of particulate matters in polluted atmospheric conditions [46–48]. High LPDRs (e.g., above 0.2) often occurred in atmospheric conditions with low extinction coefficients (low PM concentrations) and low RHs, indicating that non-spherical particles such as dust were often dominated under low-RH and low-pollution atmospheric conditions during this measurement [49,50]. The scatter points marked by the red oval characterized the dissipation of the first haze event from 01:00 to 04:00 on November 1st, 2020. During this period, dust particles from the north land transported by the north wind mixed with the local haze particles, leading to the decreasing of the extinction coefficient as well as the increasing of the LPDR.
Fig. 7. The relationship among the linear particle depolarization ratio (LPDR), the extinction coefficient and the relative humidity.
3.1.2 Measurement uncertainties of the PSI-Lidar system
The offset angle is important for the retrieval of the depolarization ratio. In this work, the temporal variation of the offset angle of the lidar system has been quantified for ten consecutive days. The measurement data recorded during nighttime with high SNRs (from 17:00 to 05:00 in the early morning of the next day) were extracted to retrieve the offset angle according to Eq. (2). A set of offset angles were estimated from each lidar curve, which were then spatially averaged along the measurement path to retrieve a single offset angle for each measurement. Figure 8 shows the variation of the offset angle during the whole measurement period. The offset angle was estimated to be −0.06° with a standard deviation of ±0.02°, implying good alignment and excellent system stability. The small uncertainty of the offset angle (±0.02°) caused a negligible error (0.012 ‰) even for a small LVDR of 0.01. In fact, even if the offset angle was assumed to be zero, the retrieval error of the LVDR was still less than 0.2‰ [39], owing to the small value and the high measurement accuracy of the offset angle. The promising result implied that the measurement error on the LVDR owing to the offset angle can be neglected in the PSI-Lidar technique.
The uncertainty of the LVDR includes systematic and statistic errors. According above discussions, the systematic error is mainly due to the PER of the polarization-sensitive image sensor for the PSI-Lidar technique, which results to the polarization crosstalk between orthogonal lidar signals. Nevertheless, the crosstalk effect on the measured LVDR can be eliminated according to Eq. (3) by utilizing the PER of the image sensor from the datasheet provided by the manufacturer. However, the polarization characteristic may not be guaranteed in mass production, the deviation of the PER with respect to the factory value may introduce additional measurement uncertainty on the retrieved LVDR. In this work, a PER uncertainty of 20% for each polarization channel was assumed to evaluate the systematic error introduced by the PER. Monte Carlo simulation analysis illustrated that the systematic error of the LVDR introduced by the uncertainty of the PER was about 0.0005. On the other hand, the statistic error, mainly originating from the random noise of the lidar signal, can be estimated by calculating the standard deviation of the LVDR in a short region with homogeneous atmospheric conditions. The total uncertainty of the LVDR can thus be estimated by
Figure 9(a) shows typical LVDR profiles retrieved both in nighttime and daytime. As can be seen, the uncertainty of the LVDR was below 0.004 at nighttime. During daytime, the uncertainty may increase up to 0.008 within 2 km, owing to the increasing of the sunlight noise. Beyond 2 km or 3km, the background noise may lead to much larger measurement errors on the depolarization ratio during daytime. On the other hand, the uncertainty of the LPDR can be calculated based on the error propagation function, by considering the uncertainty of the LVDR. As can be seen from Fig. 9(b), the uncertainty of the LPDR was below 0.006 during nighttime, while it may increase up to 0.01 during daytime.
Fig. 9. (a) The linear volume depolarization ratio (LVDR) and (b) the linear particle depolarization ratio (LPDR).
3.2 Atmospheric vertical measurements
Atmospheric vertical measurements with an elevation of 80° have been carried out from 18:00 on 30th March to 06:00 on 2nd April, 2021. Figures 10(a) and 10(b) show the temporal evolution of the lidar profile and the LVDR, respectively. Although a 10 nm interference filter has been employed to suppress the background, the SNR of the 90° polarized lidar signals were still not very high, affected by the strong background noise during daytime. The polarized lidar signals were thus truncated at the distance with a SNR less than 2. Total backscattering lidar signals (blue line) and LVDR profiles (red line) with error bars in different hours are shown in Fig. 11. The uncertainty of the LVDR at daytime was between 0.009-0.011, which was much larger than the uncertainty (less 0.001) at nighttime. In the altitude of 2-3 km, a LVDR value as low as [0.015, 0.018] (95% confidence interval) has been observed from 18:20 to 19:36 on 30th March, which was very close to the molecular depolarization ratio (0.0137). In this region, the aerosol extinction coefficient was also very small, i.e., 0.057 km−1, evaluated with a lidar ratio of 53 Sr. From 00:00 on 31st March to 14:00 on 1st April, the LVDR at an altitude beyond 1-2 km was between 0.15-0.18, which was much larger than the LVDR values at the near ground (e.g., less 0.1 below 1 km altitude). The large difference on the LVDR between the surface aerosols and the high-altitude aerosols implied that the aerosol layer was probably transported from other areas, as has been discussed in [50,51]. Since 00:00 on 2nd April, strong backscattering echoes were observed at an altitude of 2-4 km. According to the radiosonde data, the temperature and the RH at the altitude of 3 km were 1°C and 100%, respectively [52]. Meanwhile, the corresponding LVDR was less than 0.03, indicating the presence of water cloud in this region. In summary, the promising result demonstrated in atmospheric vertical measurements indicated a great potential of utilizing the PSI-Lidar technique for all-day atmospheric polarization studies.
Fig. 11. (a)-(d) Backscattering profiles and (e)-(h) the corresponding LVDR profiles in different measurement hours.
4. Conclusion
In this work, a 450-nm portable PSI-Lidar technique based on the DoFP scheme has been developed for accurately retrieving the depolarization ratio without employing additional optical components (e.g., half wave plate) and sophisticated calibrations (e.g., gain ratio calibration). The PSI-Lidar system has been designed as a T-shaped architecture consisting of a closed transmitter and a detachable large focal receiver, which is capable of all-day outdoor unmanned measurements that has yet been feasible in previous studies. The detachable large focal receiver can also be replaced by receivers with different telescopes and image sensors dedicated for other applications. By utilizing high-power multimode 450-nm laser diodes and highly integrated polarization-sensitive image sensors, the lidar system features low cost, low maintenance and short blind range (∼100 m), which is of great importance for field and lidar-network applications. Owing to the wide wavelength availability of high-power multimode laser diodes, the PSI-Lidar system operating at other wavelengths, e.g., 405 nm, 520 nm, 808 nm can also be implemented based on the present system architecture.
Atmospheric horizontal measurements have been carried out for ten consecutive days, namely, from 29th October to 8th November 2020, for the performance validation of the PSI-Lidar system. The correlation coefficient between the spatially averaged extinction coefficient retrieved from lidar measurements and the PM concentrations was beyond 0.9, indicating reliable performance of the all-day PSI-lidar system. It has been found out that the offset angle directly retrieved from the four-directional polarized lidar signals was about −0.06° with a temporal variation of only 0.02° in the whole measurement period. The measurement error on the LVDR resulting from the offset angle can thus be neglected owing to the high accuracy and excellent reliability of the measured offset angle. The LVDR as well as its measurement uncertainty have been theoretically and experimentally evaluated. The measurement uncertainty of the LVDR, mainly determined by the random noise of the lidar signal, was below 0.004 at nighttime, and may increase up to 0.008 during daytime owing to the increasing sunlight background. Nevertheless, the measurement uncertainty can be reduced after improving the SNR of the lidar signal by employing laser diodes with higher output power (e.g., 4.75 W) and narrow-band filters (e.g., 2 nm). Besides, atmospheric vertical measurements have also been performed for polarization studies of cloud, urban aerosol, etc. In summary, the low-cost low-maintenance portable polarization lidar system, capable of detecting four-directional polarized lidar signals simultaneously, opens up many possibilities for all-day field measurements of dust, cloud, urban aerosol, oriented particles, etc.
Funding
National Natural Science Foundation of China (62075025).
Acknowledgments
The authors greatly acknowledge the help of Dr. Zhenping Yin from Wuhan University.
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. E. Giannakaki, P. G. van Zyl, D. Muller, D. Balis, and M. Komppula, “Optical and microphysical characterization of aerosol layers over South Africa by means of multi-wavelength depolarization and Raman lidar measurements,” Atmos. Chem. Phys. 16(13), 8109–8123 (2016). [CrossRef]
2. T. Nishizawa, N. Sugimoto, I. Matsui, A. Shimizu, Y. Hara, U. Itsushi, K. Yasunaga, R. Kudo, and S. W. Kim, “Ground-based network observation using Mie-Raman lidars and multi-wavelength Raman lidars and algorithm to retrieve distributions of aerosol components,” J. Quant. Spectrosc. Radiat. Transfer 188, 79–93 (2017). [CrossRef]
3. R. M. Schotland, K. Sassen, and R. Stone, “Observations by Lidar of Linear Depolarization Ratios for Hydrometeors,” J. Appl. Meteorol. 10(5), 1011–1017 (1971). [CrossRef]
4. V. Freudenthaler, M. Esselborn, M. Wiegner, B. Heese, M. Tesche, A. Ansmann, D. Muller, D. Althausen, M. Wirth, A. Fix, G. Ehret, P. Knippertz, C. Toledano, J. Gasteiger, M. Garhammer, and M. Seefeldner, “Depolarization ratio profiling at several wavelengths in pure Saharan dust during SAMUM 2006,” Tellus B 61(1), 165–179 (2009). [CrossRef]
5. V. Freudenthaler, “About the effects of polarising optics on lidar signals and the Delta 90 calibration,” Atmos. Meas. Tech. 9(9), 4181–4255 (2016). [CrossRef]
6. E. W. Eloranta and P. Piironen, “Depolarization measurements with the high spectral resolution lidar,” in 17th Int. Laser Radar Conf., Laser Radar Society of Japan, ICLAS, NASDA, and NIES, 1994. pp. 127–128.
7. H. L. Xie, T. Zhou, Q. Fu, J. P. Huang, Z. W. Huang, J. R. Bi, J. S. Shi, B. D. Zhang, and J. M. Ge, “Automated detection of cloud and aerosol features with SACOL micro-pulse lidar in northwest China,” Opt. Express 25(24), 30732–30753 (2017). [CrossRef]
8. V. Noel and H. Chepfer, “A global view of horizontally oriented crystals in ice clouds from Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO),” J. Geophys. Res. 115, D00H23 (2010). [CrossRef]
9. A. Ansmann, I. Mattis, D. Muller, U. Wandinger, M. Radlach, D. Althausen, and R. Damoah, “Ice formation in Saharan dust over central Europe observed with temperature/humidity//aerosol Raman lidar,” J. Geophys. Res. 110(D18), D18S12 (2005). [CrossRef]
10. A. Ansmann, M. Tesche, D. Althausen, D. Muller, P. Seifert, V. Freudenthaler, B. Heese, M. Wiegner, G. Pisani, P. Knippertz, and O. Dubovik, “Influence of Saharan dust on cloud glaciation in southern Morocco during the Saharan Mineral Dust Experiment,” J. Geophys. Res. 113(D4), D04210 (2008). [CrossRef]
11. K. Sassen, “The Polarization Lidar Technique for Cloud Research - a Review and Current Assessment,” Bull. Am. Meteorol. Soc. 72(12), 1848–1866 (1991). [CrossRef]
12. T. Zhou, J. P. Huang, Z. W. Huang, J. J. Liu, W. C. Wang, and L. Lin, “The depolarization-attenuated backscatter relationship for dust plumes,” Opt. Express 21(13), 15195–15204 (2013). [CrossRef]
13. T. Zhou, H. L. Xie, J. R. Bi, Z. W. Huang, J. P. Huang, J. S. Shi, B. D. Zhang, and W. Zhang, “Lidar Measurements of Dust Aerosols during Three Field Campaigns in 2010, 2011 and 2012 over Northwestern China,” Atmosphere 9(5), 173 (2018). [CrossRef]
14. T. Zhou, H. L. Xie, T. Jiang, J. P. Huang, J. R. Bi, Z. W. Huang, and J. S. Shi, “Seasonal characteristics of aerosol vertical structure and autumn enhancement of non-spherical particle over the semi-arid region of northwest China,” Atmos. Environ. 244, 117912 (2021). [CrossRef]
15. F. Cairo, G. Di Donfrancesco, A. Adriani, L. Pulvirenti, and F. Fierli, “Comparison of various linear depolarization parameters measured by lidar,” Appl. Opt. 38(21), 4425–4432 (1999). [CrossRef]
16. K. Sassen, J. Zhu, P. Webley, K. Dean, and P. Cobb, “Volcanic ash plume identification using polarization lidar: Augustine eruption, Alaska,” Geophys. Res. Lett. 34(8), 1–4 (2007). [CrossRef]
17. N. Sugimoto, T. Nishizawa, A. Shimizu, I. Matsui, and Y. Jin, “Characterization of aerosols in East Asia with the Asian Dust and aerosol lidar observation network (AD-Net),” Proc. SPIE 9262, 92620K (2014). [CrossRef]
18. L. Janicka, I. S. Stachlewska, I. Veselovskii, and H. Baars, “Temporal variations in optical and microphysical properties of mineral dust and biomass burning aerosol derived from daytime Raman lidar observations over Warsaw, Poland,” Atmos. Environ. 169, 162–174 (2017). [CrossRef]
19. T. Nishizawa, H. Okamoto, T. Takemura, N. Sugimoto, I. Matsui, and A. Shimizu, “Aerosol retrieval from two-wavelength backscatter and one-wavelength polarization lidar measurement taken during the MR01K02 cruise of the R/VMiraiand evaluation of a global aerosol transport model,” J. Geophys. Res. 113(D21), D21201 (2008). [CrossRef]
20. S. Gross, V. Freudenthaler, K. Schepanski, C. Toledano, A. Schafler, A. Ansmann, and B. Weinzierl, “Optical properties of long-range transported Saharan dust over Barbados as measured by dual-wavelength depolarization Raman lidar measurements,” Atmos. Chem. Phys. 15(19), 11067–11080 (2015). [CrossRef]
21. P. Ortiz-Amezcua, J. L. Guerrero-Rascado, M. J. Granados-Munoz, J. A. Benavent-Oltra, C. Bockmann, S. Samaras, I. S. Stachlewska, L. Janicka, H. Baars, S. Bohlmann, and L. Alados-Arboledas, “Microphysical characterization of long-range transported biomass burning particles from North America at three EARLINET stations,” Atmos. Chem. Phys. 17(9), 5931–5946 (2017). [CrossRef]
22. Y. Tian, X. Pan, Z. Wang, D. Wang, B. Ge, X. Liu, Y. Zhang, H. Liu, S. Lei, T. Yang, P. Fu, Y. Sun, and Z. Wang, “Transport Patterns, Size Distributions, and Depolarization Characteristics of Dust Particles in East Asia in Spring 2018,” J. Geophys. Res. 125(16), 1–17 (2020). [CrossRef]
23. Z. P. Yin, F. Yi, Y. He, F. C. Liu, C. M. Yu, Y. P. Zhang, and W. Wang, “Asian dust impacts on heterogeneous ice formation at Wuhan based on polarization lidar measurements,” Atmos. Environ. 246, 118166 (2021). [CrossRef]
24. Y. C. Tong, X. D. Tong, K. Zhang, D. Xiao, Y. H. Rong, Y. D. Zhou, C. Liu, and D. Liu, “Polarization lidar gain ratio calibration method: a comparison,” Chinese Optics 14(3), 685–703 (2021). [CrossRef]
25. C. Wang, H. Y. Xia, M. J. Shangguan, Y. B. Wu, L. Wang, L. J. Zhao, J. W. Qiu, and R. J. Zhang, “1.5 mu m polarization coherent lidar incorporating time-division multiplexing,” Opt. Express 25(17), 20663–20674 (2017). [CrossRef]
26. M. Sharukh, B. K. Yellapragada, and R. Ayub, Single detector polarization lidar for atmospheric studies, SPIE Future Sensing Technologies (SPIE, 2019), Vol. 11197.
27. L. Mei and P. Guan, “Development of an atmospheric polarization Scheimpflug lidar system based on a time-division multiplexing scheme,” Opt. Lett. 42(18), 3562–3565 (2017). [CrossRef]
28. Z. Kong, T. Ma, K. Chen, Z. F. Gong, and L. Mei, “Three-wavelength polarization Scheimpflug lidar system developed for remote sensing of atmospheric aerosols,” Appl. Opt. 58(31), 8612–8621 (2019). [CrossRef]
29. S. M. Zhu, E. Malmqvist, W. S. Li, S. Jansson, Y. Y. Li, Z. Duan, K. Svanberg, H. Q. Feng, Z. W. Song, G. Y. Zhao, M. Brydegaard, and S. Svanberg, “Insect abundance over Chinese rice fields in relation to environmental parameters, studied with a polarization-sensitive CW near-IR lidar system,” Appl. Phys. B 123(7), 211 (2017). [CrossRef]
30. S. K. Gao and V. Gruev, “Bilinear and bicubic interpolation methods for division of focal plane polarimeters,” Opt. Express 19(27), 26161–26173 (2011). [CrossRef]
31. S. B. Powell and V. Gruev, “Calibration methods for division-of-focal-plane polarimeters,” Opt. Express 21(18), 21039–21055 (2013). [CrossRef]
32. X. Z. Tu, O. J. Spires, X. B. Tian, N. Brock, R. G. Liang, and S. Pau, “Division of amplitude RGB full-Stokes camera using micro-polarizer arrays,” Opt. Express 25(26), 33160–33175 (2017). [CrossRef]
33. T. Y. Huang, R. Y. Meng, J. Qi, Y. D. Liu, X. J. Wang, Y. Chen, R. Liao, and H. Ma, “Fast Mueller matrix microscope based on dual DoFP polarimeters,” Opt. Lett. 46(7), 1676–1679 (2021). [CrossRef]
34. V. Gruev, R. Perkins, and T. York, “Material detection with a CCD polarization imager,” in 2010 IEEE 39th Applied Imagery Pattern Recognition Workshop (AIPR), (2010). pp. 1–7.
35. J. Qi, C. He, and D. S. Elson, “Real time complete Stokes polarimetric imager based on a linear polarizer array camera for tissue polarimetric imaging,” Biomed. Opt. Express 8(11), 4933–4946 (2017). [CrossRef]
36. Z. Kong, T. Ma, Y. Cheng, R. Fei, Z. Zhang, Y. Li, and L. Mei, “A polarization-sensitive imaging lidar for atmospheric remote sensing,” J. Quant. Spectrosc. Radiat. Transfer 271, 107747 (2021). [CrossRef]
37. Z. Kong, T. Ma, Y. Cheng, Z. Zhang, Y. Li, K. Liu, and L. Mei, “Feasibility investigation of a monostatic imaging lidar with a parallel-placed image sensor for atmospheric remote sensing,” J. Quant. Spectrosc. Radiat. Transfer 254, 107212 (2020). [CrossRef]
38. R. A. Chipman, HandBook Of Optics, 3rd ed. (McGraw-Hill, New York, 2009), pp. 478–567.
39. Z. Kong, Z. Yin, Y. Cheng, Y. Li, Z. Zhang, and L. Mei, “Modeling and Evaluation of the Systematic Errors for the Polarization-Sensitive Imaging Lidar Technique,” Remote Sens. 12(20), 3309 (2020). [CrossRef]
40. J. A. Bravo-Aranda, L. Belegante, V. Freudenthaler, L. Alados-Arboledas, D. Nicolae, M. J. Granados-Munoz, J. L. Guerrero-Rascado, A. Amodeo, G. D’Amico, R. Engelmann, G. Pappalardo, P. Kokkalis, R. Mamouri, A. Papayannis, F. Navas-Guzman, F. J. Olmo, U. Wandinger, F. Amato, and M. Haeffelin, “Assessment of lidar depolarization uncertainty by means of a polarimetric lidar simulator,” Atmos. Meas. Tech. 9(10), 4935–4953 (2016). [CrossRef]
41. H. G. Di, D. X. Hua, L. J. Yan, X. L. Hou, and X. Wei, “Polarization analysis and corrections of different telescopes in polarization lidar,” Appl. Opt. 54(3), 389–397 (2015). [CrossRef]
42. A. Behrendt and T. Nakamura, “Calculation of the calibration constant of polarization lidar and its dependency on atmospheric temperature,” Opt. Express 10(16), 805–817 (2002). [CrossRef]
43. M. H. Kim, A. H. Omar, J. L. Tackett, M. A. Vaughan, D. M. Winker, C. R. Trepte, Y. X. Hu, Z. Y. Liu, L. R. Poole, M. C. Pitts, J. Kar, and B. E. Magill, “The CALIPSO version 4 automated aerosol classification and lidar ratio selection algorithm,” Atmos. Meas. Tech. 11(11), 6107–6135 (2018). [CrossRef]
44. L. Mei, L. S. Zhang, Z. Kong, and H. Li, “Noise modeling, evaluation and reduction for the atmospheric lidar technique employing an image sensor,” Opt. Commun. 426, 463–470 (2018). [CrossRef]
45. Z. F. Liu, C. G. Yang, Z. F. Gong, H. Li, and L. Mei, “Adaptive digital filter for the processing of atmospheric lidar signals measured by imaging lidar techniques,” Appl. Opt. 59(30), 9454–9463 (2020). [CrossRef]
46. M. Haarig, A. Ansmann, J. Gasteiger, K. Kandler, D. Althausen, H. Baars, M. Radenz, and D. A. Farrell, “Dry versus wet marine particle optical properties: RH dependence of depolarization ratio, backscatter, and extinction from multiwavelength lidar measurements during SALTRACE,” Atmos. Chem. Phys. 17(23), 14199–14217 (2017). [CrossRef]
47. J. Chen, Z. Q. Li, M. Lv, Y. Y. Wang, W. Wang, Y. J. Zhang, H. F. Wang, X. Yan, Y. L. Sun, and M. Cribb, “Aerosol hygroscopic growth, contributing factors, and impact on haze events in a severely polluted region in northern China,” Atmos. Chem. Phys. 19(2), 1327–1342 (2019). [CrossRef]
48. T. Wu, Z. Q. Li, J. Chen, Y. Y. Wang, H. Wu, X. A. Jin, C. Liang, S. Z. Li, W. Wang, and M. Cribb, “Hygroscopicity of Different Types of Aerosol Particles: Case Studies Using Multi-Instrument Data in Megacity Beijing, China,” Remote Sens. 12(5), 785 (2020). [CrossRef]
49. W. Yang, A. Marshak, T. Varnai, O. V. Kalashnikova, and A. B. Kostinski, “CALIPSO observations of transatlantic dust: vertical stratification and effect of clouds,” Atmos. Chem. Phys. 12(23), 11339–11354 (2012). [CrossRef]
50. Y. He and F. Yi, “Dust Aerosols Detected Using a Ground-Based Polarization Lidar and CALIPSO over Wuhan (30.5 degrees N, 114.4 degrees E), China,” Adv. Meteorol. 2015, 1–18 (2015). [CrossRef]
51. Z. W. Huang, J. P. Huang, T. Hayasaka, S. S. Wang, T. Zhou, and H. C. Jin, “Short-cut transport path for Asian dust directly to the Arctic: a case study,” Environ. Res. Lett. 10(11), 114018 (2015). [CrossRef]
52. “Observations for Station 54662 starting 2315Z 01 Apr 2021” (University of Wyoming, 2021), retrieved http://weather.uwyo.edu/cgi-bin/bufrraob.py?datetime=2021-04-02%2000:00:00&id=54662&type=TEXT:LIST.
