Open Access
Add to BibSonomy
Abstract
A compact polarization Raman lidar has been designed and constructed for using it for atmospheric correction measurements during satellite optical sensor calibration in areas with high altitude and extremely low aerosol loading. The parameters of this lidar, such as laser wavelength, telescope diameter and interference filter bandwidth, were simulated and optimized for the best observation performance. The instrument has low weight, is small in size, and requires air cooling instead of commonly used water-cooling of the laser. Thus, the instrument is suitable for autonomous operation in remote sites. The lidar prototype was installed in Lijiang (26°43’ N, 100°01’ E), China, a potential observation site for calibrations of optical sensors of satellites. This observation site has been shown to be an appropriate place for remote sensing and satellite calibration activities with low aerosol loading, thin air and a comparably high proportion of cloud-free days. A field campaign carried out between November 2019 and April 2020 allowed for thoroughly testing the instruments. The results of test observations show that complete overlap between emitted laser beam and field-of-view of the receiver unit is achieved at relatively low heights above ground. The measurement accuracy is comparably high. Thus, this instrument is suitable for operating in areas with relatively clean atmospheric conditions.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Spaceborne remote sensing is an essential method for carrying out large-scale and long-term observations of Earth’s surface and atmosphere. It plays an essential role in surveying and mapping, water conservation, atmospheric monitoring, disaster prediction, and other fields of Earth observations [1–6].
Remote sensing at levels of low light intensity is an important expansion of the field of remote sensing, which can effectively solve the challenge of detection of surface features at dawn, dusk and at nighttime in the presence of sufficient moonlight [7–10]. Remote sensing under levels of low-light conditions has experienced growing attention and there has been continuous progress. For example, the Visible Infrared Imager Radiometer Suite (VIIRS) on the Suomi National Polar-orbiting Partnership (NPP) satellite which was launched in 2011 operates a low-light channel at night [8,11]. The LJ1-01 nighttime-light remote sensing satellite has been developed by Wuhan University, China [12,13]. China's Fengyun-3E satellite also carries a low-light imager [14].
The Fengyun Meteorological Satellites Programs belongs to a series of operational, global Earth-observation satellites [15]. These satellites monitor change of the atmosphere and surface states. Data from these satellites are used for, e.g., weather forecasting, disaster warning, and environmental monitoring [16,17]. Fengyun satellites have the ability to observe the moon. Fengyun-3E satellite launched in July 2021 is equipped with a medium resolution spectral imager with low-light capability (MERSI-LL) [14]. During the operation of the Fengyun satellite, the quantitative value of the signals pertaining to radiation received from ground objects degraded due to aging of satellite optical components. In addition, the signals measured by the satellite require a correction for the atmospheric aerosol attenuation effects. Therefore, satellite sensor calibration is an urgent need to determine the relationship between signal strength, e.g., radiation value received by the satellite detectors and true value of the radiation that is emitted by the ground target [18,19], and the selection of the calibration site described in this contribution is particularly vital for this task.
Calibrations at field sites are one of the crucial methods for satellite sensor operation [20]. In recent years, technology advanced to the point that satellites can also be used to observe the radiation reflected from the moon. The moon for that reason has become an ideal target for calibration activities. However, the lunar radiation model is not accurate, which leads to inacceptable errors for satellite calibration. Additional ground-based atmospheric correction observations of moon’s radiation field are needed to obtain more accurate parameters of the lunar model [21,22]. In addition, the specific light sources, i.e., diffused light over desert [23], on the ground are generally taken as the calibration sources for low light-level remote sensing satellite. And ground-based atmospheric correction observations are needed for correcting the atmospheric aerosol effects. In this context, calibrations under cloud-free conditions are preferred over situations in which clouds or rain interferes with the measurements. Consequently, it is essential to evaluate whether the selected area is suitable as a calibration site.
Aerosol optical depth (AOD) is an important physical parameter to characterize optical attenuation by columnar aerosol concentration. AOD plays a significant role in atmospheric correction and satellite-sensors calibration [24–28]. Sun-photometer (SPM) is an instrument commonly used for measurements of AOD, and accurate AOD can be obtained from daytime observations [29,30].
Although AOD-measurements by sun-photometer (lunar photometer), if used in lunar-observation mode, can be extended to observations of the moon [28,31], accuracy of the measured AOD-values is limited at night [32,33]. Firstly, the lunar irradiance is very weak, making the precise tracking of the moon a challenge. The lunar irradiance obtained from the lunar model is not accurate for calculating AOD and this can lead to comparably large errors of AOD [33]. Secondly, lunar photometer measurements are limited by the moon phase. Lunar photometer usually observe from the first to the third quarter of the moon [32].
Thirdly, the pointing direction of the lunar photometer poses a challenge. Lunar photometer measures AOD between the observation site and the moon at night, whereas AOD between the observation site and the satellite is required in satellite calibration. This issue of pointing direction is the same for daytime measurements.
Lidar has the capability of spatial-resolved profiling of the distribution of aerosol concentration and other important parameters by means of active remote sensing. In that regard, most lidar activities focus on vertically resolved observations. However, in principle lidar can be used for the 3-dimensional mapping of the aerosol distribution by means of, e.g., a scanning unit that directs the laser beam into any direction. As lidar uses its own light source for observations of AOD and thus does not depend on the availability of an external light source, such as sun or moon. Depending on the type of lidar measurement technique that is used, AOD can be inferred with comparably high precision [34–36]. In addition, accurate polarization measurements rely on the polarization purity of the laser head [37]. Polarization measurements is important for characterizing depolarized aerosol types, which can be helpful to extend the AOD information at single wavelength (355 nm) to the wide-band range up to near infrared wavelength. Therefore, lidar has evolved as the ideal tool for atmospheric correction and thus calibration of satellite optical sensors. In addition, such lidar measurements can also be used for evaluating whether the calibration site is suitable for such type of measurements.
In order to improve the accuracy of satellite calibration, the atmospheric aerosol concentration (particles and molecules) at the calibration sites should be comparably low. In that case the impact on the measured radiances is comparably low. Therefore, the calibration site is generally selected at high-altitude and sparsely populated areas. The areas that meet these conditions generally have a complex terrain, e.g., high-altitude mountain region. For convenient transportation and maintenance, the size and weight of the lidar should be considered when designing the lidar. In this context, the lidar cannot use a high pulse energy, low repetition frequency laser, because these lasers need to be water-cooled, which will greatly increase the volume and weight of the lidar. In addition, aerosol concentration on average is higher near the surface than aloft. Thus, we need to keep the height up to which the overlap between lidar beam and receiver-field-of-view region is below 1 as small as possible.
In this study, a compact polarization Raman lidar has been developed for detection of aerosol and atmospheric correction during satellite overpasses and calibration of satellite optical sensors. Section 2 presents the theoretical background and optimization of lidar parameters, including laser wavelength, telescope diameter and interference filter bandwidth. Section 3 we present system setup and field-site measurements. We demonstrate the performance of the lidar and conclude that the designed lidar is suitable to obtain AOD for atmosphere correction during remote sensing satellite calibration. Section 4 presents results. We analyze the characteristics of monthly AOD in field-site with a summary that the candidate observation site meets the requirements of calibration site. Section 5 presents conclusions.
2. Optimization of lidar parameters
2.1 Theoretical background
The backscatter signals received from the interaction between the emitted laser light and aerosols can be expressed by the lidar equation [38]:
In this study, the molecular concentration distribution is calculated using the standard atmospheric model [40]. The molecular backscatter and extinction coefficients are obtained by semi-empirical formulas [41]. The ratio of aerosol extinction to backscatter coefficient is called lidar ratio, which reflects aerosol microphysical properties, such as particle complex refractive index (from which optical parameters such as light-absorption capacity can be inferred), particle size distribution and particle shape [42–45].
The number of photons NS(R) received by a lidar system can be expressed as [46]:
where CT is the cumulative number of pulses in the lidar data acquisition unit. TA is the transmittance between the telescope input and the photodetector. Δt is the sampling time resolution. η is the system efficiency. λ is wavelength. h is Planck's constant. c is the speed of light.The sky background noise NB(λ) can be expressed as [46–48]:
The dark count ND of a detector working in photon counting mode, can be expressed as [46]:
where RD is dark count rate.For the photon counting mode, the signal uncertainty can be expressed as [47]:
Therefore, the SNR can be expressed as [46,48,49]:
2.2 Lidar system optimization
Solving the lidar equation requires the retrieval of two parameters, i.e., the aerosol extinction coefficient and aerosol backscatter coefficient [50]. It means that, for example for a Mie-scattering lidar [51], the lidar ratio or extinction-to-backscatter ratio needs to be assumed, which will introduce retrieval uncertainty that can be quite high [52,53]. For Raman lidar [36] or High Spectral Resolution Lidar (HSRL) [54], these two parameters can be independently solved without assuming the lidar ratio. However, this capability comes at a cost arising from having to implement a molecular channel (for HSRL) or Raman channels (a N2 vibration-rotation Raman channel in the case of this study). HSRL is a complex instrument which requires high financial investments and significant expert knowledge to build and operate. Data analysis also requires highly skilled data analysis specialists. Such instruments can be quite bulk which makes their transport and installation in remote sites a challenge. Long-term stability of HSRL cannot be guaranteed unless the aforementioned investments in terms of money and staff can be made. Compared with HSRL, sophisticated Raman lidar can overcome some of these drawbacks. However, state-of-the art Raman lidars are still almost exclusively operated under conditions of nighttime as that operation mode ensures a reasonable signal-to-noise ratio (SNR) of the Raman signals.
Aerosols in the atmosphere influence the radiation field. In order to optimize our lidar system, we did a series of simulations for different lidar parameters, e.g., laser wavelength, telescope diameter, and interference filter bandwidth.
The lidar signal can be simulated based on the lidar equation described in section 2.1. We carried out simulations for different values of AODs: 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5. Wang et al. [22] showed that AOD at Lijiang was less than 0.1 at 550 nm during most of their observation period. Considering the situation at the calibration site, we showed a simulation example (Fig. 1) for AOD 0.085.
Fig. 1. The simulated scene. (a) Extinction coefficient, (b) backscatter coefficient and (c) lidar ratio at 355 nm.
As is shown above (Fig. 1), we assumed two particle layers and a thin cirrus cloud atop. Below 2 km height above ground level (agl), an aerosol backscatter coefficient of 0.5 Mm-1sr-1 and a lidar ratio of 60 sr at 355 nm are assumed to represent the aerosol condition within the planetary boundary layer. From 2.5∼3 km height agl, a lofted aerosol layer is assumed with aerosol backscatter coefficient of 1 Mm-1sr-1 and lidar ratio of 50 sr at 355 nm. Around 7 km height agl, a cirrus cloud with geometrical thickness of 500 m is simulated with backscatter coefficient of 2 Mm-1sr-1 and lidar ratio of 20 sr at 355 nm [42,55].
As stated in the introduction, the size and weight of designed lidar should be kept as small as possible. Therefore, the high pulse energy, low repetition frequency lasers are not considered. Instead, a low pulse energy, high repetition frequency laser is preferred, which can achieve similar average power, and feature a highly compacted design as well. Diode-pumped solid-state (DPSS) lasers with kHz repetition rate meet our requirements and is air-cooled. And such commercial laser heads are really limited. Therefore, we assumed laser repetition frequency of 2 kHz. Combined the commonly used laser power, the laser pulse energies were selected as 200 µJ and 400 µJ at 355 nm and 532 nm, respectively (see Table 1). Referring to the characteristics of the commonly used PMTs, the dark counts were selected as 50 s-1 for 355 nm, 386nm and 532nm, 100 s-1 for 607 nm. The detector quantum efficiency was 0.28 for 355 nm and 0.3 for 532nm. In addition, low light-level remote sensing satellite calibration is mainly carried out from dusk to dawn. The background radiance is small at night and relatively large at dusk and dawn, whose influence should be considered when optimizing the bandwidth of interference filter. In this study, the sky background radiances at dusk and dawn were used when optimizing interference filter bandwidth of elastic channel. The sky background radiances are 3*10−6 Wm-2sr-1nm-1, 5*10−6 Wm-2sr-1nm-1 respectively at 355 nm and 532 nm at dusk and dawn, and 2*10−9 Wm-2sr-1nm-1, 5*10−9 Wm-2sr-1nm-1 at 355 nm and 532 nm respectively at night [56–58].
Table 1. Values of main simulation parameters used in this study
2.2.1 Selection of laser wavelength
When retrieving the aerosol optical parameters, the molecule is taken as reference value and the noise of molecular signal will be propagated to the retrieval results. In order to obtain more accurate retrieval results, the molecular SNR should be high. As the intensity of Rayleigh scattering is inversely proportional to λ4. Therefore, the following two points hold true: 1) the shorter the wavelength, the stronger the scattering and thus the signal. 2) the shorter the wavelength, the stronger the attenuation and thus the shorter the detection range. Therefore, it is necessary to optimize a suitable wavelength.
Nd:YAG lasers are commonly used for lidar. This type of laser emits radiation at 1064 nm, 532 nm, 355 nm, 266 nm, which represent the fundamental, second, third and fourth harmonics.
The strong absorption of radiation by atmospheric ozone and oxygen at 266 nm wavelength limits the detection range and thus makes this wavelength quite unsuitable for lidar observations. In addition, as molecular scattering is inversely proportional to λ4, the Raman cross-section at 1064 nm is comparably small which makes this wavelength also unsuitable for calibration measurements. Therefore, the wavelengths of 355 nm and 532 nm are considered in this study.
We simulated the range-corrected signals (RCS) and the signal-to-noise ratio (SNR) to find out which of the two wavelengths (355 and 532 nm) were more suitable for satellite sensor calibration in view of the instrument parameters and the location. An example of the simulation results is displayed in Fig. 2.
Fig. 2. (a) RCS and (b) SNR at 355 and 532 nm, respectively; (c) RCS and (d) SNR at 386 and 607 nm, respectively. The accumulative time were 1 minute for elastic signal and 30 minutes for Raman signal.
Figure 2(a) shows that the 355 nm wavelength signal is larger than that of 532 nm below 6.5 km height agl. Within the particle layer at 2.5∼3 km height agl, the signal at 532 nm wavelength is stronger. The reason is the energy of the 532 nm laser radiation which is twice the energy of the 355nm laser. In addition, Ångström exponent is associated with particle size. For large particles, i.e., mineral dust, Ångström exponent can be as low as 0; while for small particles, i.e., urban pollution, Ångström exponent can be as large as 2.5 [55] . These two factors lead to the following simulation results. The SNR at 355 nm wavelength is higher than the SNR at 532 nm below 6.5 km height agl (see Fig. 2(b)), which will make the Rayleigh fitting more accurate and thus deliver more accurate retrieval results. And above 6.5 km, the SNR at 532 nm becomes larger than that at 355 nm. The 355 nm signal shows better performance characteristics in the near range, while the 532 nm signal performs better at far range. Figure 2(c) and (d) show that both the RCS and the SNR at 386 nm are stronger than at 607 nm. Since the concentration of aerosol particles in general is higher near the ground, i.e., closer to the lidar than to the satellite, the laser wavelength of 355 nm is preferred over the wavelength at 532 nm.
2.2.2 Selection of telescope diameter
The telescope diameter (the diameter of the telescope's primary mirror) directly affects the performance quality of the lidar system. There are two fundamental issues: (1) shadowing effect of secondary mirror on the detection efficiency; (2) effect of telescope’s focal length on full overlap range (height where the overlap between laser beam and receiver-field-of-view region achive 1). The first effect is negligible because of small size of the secondary mirror and a coaxial design. The second effect is investigated in detail in our simulations.
The distance at which full overlap between emitted laser beam and field-of-view of the telescope is achieved is proportional to the telescope diameter (see Appendix A). It means the larger the diameter, the larger the distance until full overlap is achieved. Therefore, the full overlap range at near-range and the SNR at far-range need to be considered together when optimizing the telescope diameter.
We selected six telescope diameters for our simulations: 100 mm, 150 mm, 200 mm, 250 mm, 300 mm, and 500 mm. We use a coaxial design, which has a relatively small blind area. The characteristics of near-range signal greatly rely on the overlap function. The arctangent function is used in our simulations because the shape of this function is relatively similar to the measured overlap function (Fig. 3(d)).
Fig. 3. (a) Example of RCSs for different telescope diameters and AOD 0.085. (b) Relative errors of AOD for different telescope diameters and AOD 0.085, fitting trend line with a quartic equation. (c) Optimal values of telescope diameter for different AODs. For each AOD scenario, 50 simulations were performed to obtain the average value and uncertainty of the optimal value, fitting trend line with a linear equation. (d) Measured overlap function (blue dots) with telescope diameter 200 mm at different 60 m height levels, fitting trend line with an arctangent function (dashed line).
Figure 3(a) shows an increase of signal intensity and the range at which full overlap is achieved. Figure 3(b) shows the relative error (the ratio of AOD uncertainty and AOD) of aerosol optical depth (AOD) from ground to 6 km for different telescope diameters and AOD 0.085. The dotted line is the trend line of the relative error of AOD for changing telescope diameter. The trend line reaches its minimum when the telescope diameter is around 200 mm.
Figure 3(c) shows the results of simulations for 10 different AODs: 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5. For each scenario for a given AOD we have done 50 times simulation experiments. The results showed that the optimal value of telescope diameter increases as AOD increases. The optimal telescope diameter ranges from 180 mm to 230 mm. Since AOD is generally less than 0.2 at the calibration site, the optimum telescope diameter for our system is approximately 200 mm.
Figure 3(d) present measured overlap data while the telescope diameter is selected as 200 mm. The trend line is fitted with an arctangent function. It can find that the full overlap range is approximately 150 m.
2.2.3 Selection of bandwidth of interference filter
Interference filter bandwidth is an important parameter, which can both affect the received signal power and the background noise. A wider bandwidth of an interference filter allows for passing through more signal power but also background noise. In our simulation studies, we take 355 nm as emission wavelength and 200 mm for the telescope diameter. We evaluated ten interference filter bandwidths from 0.1 nm to 1 nm. The laser spectral function was expressed in terms of a Gaussian distribution function [46,59]. In the simulation experiments we used a linewidth of 0.1 nm. The transmittance function of the interference filters were fitted by a Lorentz function [60] with central wavelength of 354.7 nm for elastic channel, 386.7 nm for Raman channel and maximum transmittance of 70%.
Figure 4(a) shows the transmittance of the light (355 nm) for different interference filter bandwidths. The red dashed line marks the maximum transmittance (70%) of the interference filter. The transmittance increases as the interference filter bandwidth increases from 0.1 nm to 3 nm. For an interference filter bandwidth larger than 3 nm, the transmittance approaches the maximum transmittance of the interference filter (70%), which means that all the laser light can pass through.
Fig. 4. (a) Transmittance of the light (355 nm), fitting trend line with an eighth order equation when the value of interference filter bandwidth is from 0.1 nm to 2 nm. (b) SNR of elastic signal (355 nm) and Raman signal (386 nm) at 6 km height agl for different interference filter bandwidth and AOD 0.085, fitting trend line with an eighth order equation. The accumulative time were 1 minute for both elastic and Raman signal.
Figure 4(b) is an example of the SNR of elastic and Raman signal at 6 km height agl for different interference filter bandwidths and AOD 0.085. The maximum SNR is 24.8 with bandwidth of 0.37 nm for elastic channel, and the maximum SNR is 6.6 with bandwidth of 0.47 nm for Raman channel. For bandwidths of 0.3 nm, 0.4 nm and 0.5 nm, the difference of their SNRs is very small.
We carried out simulations for AODs from 0.05 to 0.5 with step size 0.05. We find that optimal interference filter bandwidths for elastic (Raman) signal are 0.37 (0.48), 0.36 (0.46), 0.35 (0.45), 0.34 (0.44), 0.33 (0.43), 0.32 (0.42), 0.32 (0.41), 0.31 (0.39), 0.3 (0.38) and 0.29 (0.37), respectively. Based on the availability of commercial optical components, we selected interference filters with bandwidth of 0.5 nm both for elastic and Raman channel.
3. Equipment and experiment
3.1 System setup
Table 2 show some technical parameters of the system. These parameters have been chosen according to the simulation results with synthetic lidar profiles (see section 2).
Table 2. Main parameters of lidar (The parameters in bold in the table are optimized in this study).
Figure 5 shows the schematic and two photos of the compact polarization Raman lidar. This lidar has three receiver channels: parallel polarization, cross polarization, and Raman channel. The lidar system is composed of three main units, i.e., the laser emitting unit, the signal receiving unit, and the signal acquisition and control unit.
Fig. 5. (a) Schematic of the compact polarization Raman lidar, IF1 and IF2 are interference filters with central wavelengths of 354.7 and 386.7 nm, respectively. 355P, 355S and R386 represent parallel, cross polarization (355 nm) and Raman (386 nm) channel, respectively. (b) lidar structure and (c) outside view of the working diagram of lidar at the observation site
The laser emitting unit mainly includes the laser, beam expander (BE) and motorized actuator. The laser is from Bright Solutions. The magnification factor of the beam expander is 10. The laser is a diode-pumped solid-state (DPSS) laser with air cooling, which can greatly reduce the size and weight of the laser. The motorized actuator CONEX-AG-M100D (Newport Ltd.) is controlled by a piezo-crystal. The motor is used for the positioning of the mirrors and adjusting the alignment between the laser beam and the receiver-field-of-view of the telescope.
The signal receiving unit includes telescope, field stop and collimating lens (CL). The telescope (Cassegrain)) has a special design which makes it suitable for unattended operation in environments that undergo strong temperature changes and a relatively small blind area. The main material is aluminum whose focus point is not affected too much by temperature with special design. Therefore, there is no need for temperature control to the telescope which can greatly reduce its size and weight.
The three measurement channels are operated in photon counting mode. The signal acquisition and control unit include photomultipliers (PMT), control board, photon-counting acquisition board. The PMT type is H10721P-210 (Hamamatsu). The spectral response is from 230 nm to 700 nm. The control board is designed for controlling the laser, the power supply to the PMTs, the data acquisition board, the gain of the PMTs, and the SERVO-motors that are used to steer the positions of the depolarizer and attenuators, respectively. We use an optical trigger for controlling data acquisition. The elastic signals (scattered by aerosol particles and molecules) and Raman signals (scattered by nitrogen molecules) are received by the telescope and then separated by the dichroic beam splitter (DBS) and interference filters (IF). Parallel polarization and cross polarization signals are separated by polarizing beam splitter (PBS).
After the lidar was assembled, it has been compared to the lidar at the Naples station of the European Aerosol Research Lidar Network (EARLINET) [61] during a comparison study. The results showed that the parallel polarization, cross polarization and Raman signals with temporal resolution of 30 minutes agreed well. We had also compared our retrieval algorithms with that of EARLINET, the profiles of backscatter and extinction coefficients agreed too. We used the method of depolarizer [62] to calibrate the polarization of lidar regularly.
3.2 Field-site measurements
Figure 6 shows the location and geographical situation of the calibration field site. Lijiang observation site is located at 26°43’ N, 100°01’ E. The altitude of the observation site is 3,170 m height above sea level. The field site is characterized by comparably low aerosol loading throughout the year. The lidar station lies in a sparsely populated region. Thus, pollution from local sources, caused by human activities, can be considered low. Free-tropospheric pollution from regional and distant sources may still exist. We will carry out long-term observations which will allow us to assess the impact of free- tropospheric pollution.
Fig. 6. Location and geographical characteristics of Lijiang station. Also shown is the digital elevation model.
A first set of data from long-term observations was acquired between November 2019 and April 2020. In November and December 2019, the main target of our field campaign was to acquire data in the two weeks around the full moon. The goal of the observations carried out between January and April 2020 was to collect as much data as possible.
Table 3 shows statistics of the observation conditions in each month, where “scheduled days” means the number of days scheduled for observation. The factors that caused the lidar to not collect data normally mainly included weather reasons (such as rain and snow) and other external factors (such as power supply failure, remote control failure). As this study is mainly concerned with whether the observation site is suitable for satellite calibration, days with rain and snowfall are separately counted. The “rainy/snowy ratio” is the proportion of rainy and snowy days to the number of scheduled observation days. The proportion of clouds on a day is an important parameter for evaluating the suitability of an observation site for satellite-sensor calibration. In this study, a cloud screening algorithm [63,64] is used to distinguish whether the profiles of lidar signals contain clouds. The time resolution of the lidar profiles used for this study is one minute. If the lidar profiles contained clouds, we deleted them and calculated the proportion of the cloud-containing profiles to all lidar profiles for that day. If the proportion of clouds on that day was greater than 60%, this day was considered a cloudy day. The “observation days” are the days when the lidar collected data. The “effective days” means the number of days suitable for satellite calibration. The “effective ratio” means the ratio of effective observation days to actual observation days.
Table 3. Statistics of meteorological condition
According to Table 3, the number of scheduled days is 134. There are 43 rainy and snowy days. The total rainy/snowy proportion is 32.1%, and the monthly rainy/snowy proportion stays below 30% during all observation months except for April 2020. The lidar took data on 84 days. The number of effective observation days is 62. The total effective ratio is 73.8%, and the monthly effective observation ratio is higher than 60% except for April 2020.
Figure 7 shows one example of a lidar observation. Data were taken from 1 January 2020, 19:30 local time (LT) to 2 January 2020, 07:30 LT. The temporal resolution and spatial resolution of the raw data are 1 minute and 15 m, respectively. Figure 7(a) and (b) show the space-time distribution of the RCS and volume linear depolarization ratio (VLDR) [65], respectively. Figure 7(c) shows the result of the cloud distribution obtained by a cloud screening algorithm [63,64]. Figure 7(d), (e) and (f) show the profiles of RCS, VLDR, backscatter coefficient and accumulative AOD (Acc. AOD) on 1 January 2020, 21:00-21:30 LT respectively. The error bars of volume linear depolarization ratio (Fig. 7(e)) are derived by error propagation [66]. The error bars of backscattering coefficient (Fig. 7(f)) are derived by Monte Carlo method [66]. It was a measurement day with clouds present. AOD down to 0.015 indicates a very clean atmosphere. A considerable amount of cirrus clouds was present at approximately 6∼7 km height agl. Low-level clouds were detected at about 2 km height agl. The maximum value of the backscatter coefficient on 1 January 2020, 21:00-21:30 LT is around 0.3 Mm-1sr-1, which means that some small amount of aerosol particles was present in the lower parts of the troposphere.
Fig. 7. Space-time distribution of (a) RCS, (b) volume linear depolarization ratio (VLDR) and (c) cloud discrimination (“cld” and “ukn” are the abbreviations of cloud and unknown, respectively). The temporal resolution and spatial resolution are 1 minute and 15 m respectively for (a), (b) and (c). Profiles of (d) RCS, the red line corresponds to the Rayleigh profile to which the measured signals are fitted. (e) VLDR, the dotted line represents the typical value of VLDR and (f) backscatter coefficient and accumulative AOD on 1 January 2020, 21:00-21:30 LT. The temporal resolution are 30 minutes for (d), (e) and (f). The spatial resolution is 60 m for (d), and 180 m for (e) and (f), respectively.
Figure 8 shows an example of RCS and SNR profiles derived from the elastic and Raman signals on 8 December 2019, 21:00-22:00 LT. We accumulated the lidar signals for one hour. From Fig. 8(a) and (c), it could be seen that there was almost no aerosol above 2 km height agl. Moreover, Fig. 8(b) and (d) show that both the SNRs of the elastic and Raman signals were larger than 3 within 10 km height agl, indicating that the designed lidar possesses the capability to observe at far range. The sub-panel presents enlarged view of RCS without overlap correction at 0-2 km. It can be seen that the full-overlap range is extremely low (around 150 m).
Fig. 8. Example of the RCS of the (a) elastic signal and (c) Raman signal. The red line corresponds to the Rayleigh profile to which the measured signals are fitted. The sub-panel presents enlarged view of RCS without overlap correction at 0-2 km. SNR of the (b) elastic signal and (d) Raman signal on 8 December 2019, 21:00-22:00 LT. The temporal resolution and spatial resolution are 60 minutes and 60 m respectively.
Rayleigh fit is widely used for evaluating lidar performance with well-characterized backscatter from molecule in the lidar community [67]. It can be seen from Fig. 7(d), 8(a) and 8(c) that the signals present a good agreement with molecular signals, which proofs the lidar has good performance.
Figure 9 is an example of the optical parameters (extinction coefficient, backscatter coefficient, lidar ratio, VLDR and particle linear depolarization ratio (PLDR)), which have been retrieved from the signals on 8 December 2019, 21:00-22:00 LT. The optical parameters are obtained with the Raman method [36,68]. The maximum value of the extinction coefficient was less than 100 Mm-1 and the average value was around 40 Mm-1 between 0∼2 km, which showed that the aerosol content was very small, and the air was very clean. The average value of the lidar ratio at 355 nm was about 40 sr below 2 km height agl. The average value of the volume linear depolarization ratio and particle linear depolarization ratio [69] were approximately 1% and 6%, respectively.
Fig. 9. Optical parameters. (a) extinction coefficient; (b) backscatter coefficient; (c) lidar ratio; (d) volume linear depolarization ratio (VLDR), the dotted line represents the typical value of the VLDR; (e) particle linear depolarization ratio (PLDR) for the measurement on 8 December 2019, 21:00-22:00 LT. The temporal resolution of these five optical parameters are 60 minutes. The spatial resolution of the extinction coefficient is 60m below 2.5 km height agl and 180 m above 2.5 km height agl. The spatial resolution is 60 m for the backscatter coefficient, lidar ratio, VLDR and PLDR.
4. Results
4.1 Meteorological conditions at the calibration site
Temperature and rainfall sensors at the observation site allow for analyzing the local meteorological conditions. Figure 10 shows the timeseries of temperature and rainfall for the time from February 2019 to October 2020.
Temperature gradually rose from February 2019 to August 2019 and from January 2020 to June 2020. Temperature gradually decreased from August 2019 to January 2020 and from June 2020 to October 2020. The temperature ranged from -9 °C to 24.3 °C during this period. Minimum temperature was reached in January 2020.
Rainfall, in terms of number of rainy days shows an upward trend from February 2019 to July 2019 and from December 2019 to August 2020. A downward trend was observed from July 2019 to December 2019 and from August 2020 to October 2020. The results show that the climate in Lijiang observation site presents a periodic characteristic and rainfall is less likely to occur between October and March.
4.2 Characteristics of monthly AOD
AOD is a measure of aerosol loading in the atmosphere. AOD can be calculated by integration of the profile of the lidar-derived extinction coefficient from ground to a certain reference height. In this study, we selected the reference height as the height range in which no significant lidar signal from aerosol particles could be identified. In another step the timeseries of AOD was analyzed.
As mentioned above, the lidar profile that contains clouds is removed and the effective days are counted as is shown in Table 3. We retrieved the aerosol extinction coefficient using 60-minute accumulated data and then calculated AOD from ground to the reference height. All the AOD in every month is counted and divided by its respective numerical value. When the value of AOD is between 0 and 0.15, it is divided into three groups in increments of 0.05. Measurements with an AOD value greater than 0.15 are put into one additional group. Figure 11 shows the histogram distribution of AOD for the time from November 2019 to April 2020.
Fig. 11. Statistical chart of AOD at 355 nm. Lidar signal profiles with clouds in it are removed. AOD is calculated from ground to reference height with 60 minutes temporal resolution.
Figure 11 shows that AOD at 355 nm was rather low (less than 0.1) in November, December 2019 and January, February 2020. Average AODs were 0.021 ± 0.013, 0.074 ± 0.028, 0.044 ± 0.036 and 0.086 ± 0.062 in these months, respectively. In February, March and April 2020, there is a relative high proportion of measurements with AODs larger than 0.15. The average AODs of March, and April 2020 are 0.106 ± 0.09, and 0.109 ± 0.063, respectively. The main message we obtain from our 6 months of observations is that average AODs are less than 0.11. This result indicates that the aerosol loading in Lijiang station is low and this site is suitable for the purpose of calibration of satellite sensors.
Figure 12 finally shows the distribution of AOD at 355 nm for the observational period and the impact of clouds on the observation capabilities at Lijiang. We stress that identification of clouds may be influenced by many factors such as threshold values in the lidar signals which eventually allow for flagging a measurement as contaminated by clouds. Another very crucial point is that the lidar has a very narrow field of view compared to the swath width of a satellite sensor. Thus, even if a lidar signal (respectively profile) points to a cloud-free condition in the observational area, this may by far not be true for the satellite that overpasses the calibration site during a lidar measurement.
Fig. 12. (a) Boxplot of monthly AOD at 355 nm. The bottom and top of the box is the 25% and 75% percentiles of the AOD, and the line in the box represents the median value of AOD. The red crosses are outliers, and the two lines extending outward from the box represent the maximum and minimum values of AOD except for outliers, respectively. (b) Monthly cloud proportion, expressed in terms of proportion of cloud-containing profiles to total profiles acquired during the observation period. The temporal resolution of each profile is 1 minute.
Figure 12(a) is the boxplot of monthly AOD. All median AODs of the 6 months are below 0.1. Excluding outliers, the range of upper and lower limits is relatively small in November, December 2019 and January 2020. The monthly AOD shows an overall increasing trend from November 2019 to April 2020 (Fig. 12(a)).
Figure 12(b) is the monthly cloud proportion, which is the proportion of the number of signal profiles containing clouds to the total number of profiles during the observation period. If a certain day is considered as a cloudy day (see section 3.2), all the profiles of this day are labelled as cloud-contained profiles. The proportion of clouds in the column above the lidar on average increases, though there is significant inter-monthly variability, see Fig. 12(b). April is the month with the largest number of cloud-contaminated observations, i.e., close to 35%. Still these numbers are low enough to make the Lijiang observation station suitable for satellite sensor calibration.
5. Conclusions
A compact polarization Raman lidar system has been developed for atmospheric correction in the context of calibration of optical sensors of Earth observation satellites. The characteristics of the observation site, Lijiang (26°43’ N, 100°01’ E) are suitable for satellite as well as lidar observations. On the one hand, thin air at high altitude makes Lijiang a suitable site from the point of view of satellite calibration. On the other hand, low aerosol loading of the atmosphere in this sparsely populated, remote mountain site makes Lijiang suitable from the point of view of carrying out lidar observations in the context of satellite-sensor calibration/validation studies.
The main hardware parameters of the lidar were optimized based on simulation studies with synthetic data. The instrument features a low overlap height, which meets the needs of atmospheric correction and calibration of satellite sensors. The UV polarization Raman lidar is light (∼35 kg) and small, and features a compact design, which allows for easy transportation and deployment in areas with complex terrain and harsh environment.
Lijiang is a candidate for a calibration site for the Fengyun satellite. Lidar observations were made at the site for approximately six months in 2019/2020. The results show that the instrument can be used for capturing atmospheric aerosol distributions with high accuracy. With regard to the quantification of the aerosol load in terms of particle backscatter and extinction coefficients, we assume that a comparably high measurement accuracy can be achieved.
We analyzed the time series of AOD at 355 nm and cloud cover encountered during the 6-month measurement period. Climatic environment and atmospheric characteristics show considerable variation. AODs of six months were relatively low. Both mean and median values during the observational period were less than 0.11, especially in November 2019 and January 2020. We find that the cloud proportion is quite variable. The main result is that cloud cover remained below 35% in each month of the observational period. Therefore, both the designed lidar and the Lijiang observation site are suitable for the calibration of satellite sensors.
Appendix A
From the lens formula $\frac{1}{f} = \frac{1}{u} + \frac{1}{v}$, we can get that:
where f is the focal length of the telescope, u is object distance, v is image distance. In general, the field stop is placed at the focal plane of telescope. When object distance u=∞, image distance v = f. That is, objects at an infinite distance are imaged on field stop. When object distance is finite, such as, 500 m or 1000 m, objects are imaged under the field stop. $\frac{{v - f}}{f}$ represents the relative deviation of the image distance v of laser spot to the field stop. When the relative deviation is small, that is, the laser spot is in the field stop or very close to the field stop. As the field of view (FOV) of receiver is larger than laser divergence, lidar can receive all the scattered energy, that is, the objects is located at the full-overlap area. On the contrary, when the relative deviation is large, lidar can only receive a fraction of the scattered energy, that is, the objects is located at the incomplete overlap area. From the formula (8), the full-overlap range is proportional to the focal length f of the telescope. In practice, the minimum value of F number is limited by engineering technology, and the choices of F number are limited. In simulation, we kept F number constant and changed the focal length of the telescope by changing telescope’s diameter. Therefore, the larger the telescope’s diameter, the larger the distance until full overlap is achieved.Acknowledgments
The authors would like to express our appreciation to Prof. Detlef Müller for valuable suggestions in improving our manuscript. We also thank Mr. Zhou Fang (Lijiang Gemini Observatory) for providing the temperature and rainfall data of the site.
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. R. Srivastava and R. K. Saxena, “Technique of large-scale soil mapping in basaltic terrain using satellite remote sensing data,” Int. J. Remote Sens. 25(4), 679–688 (2004). [CrossRef]
2. G. M. Foody, “Status of land cover classification accuracy assessment,” Remote Sens. Environ. 80(1), 185–201 (2002). [CrossRef]
3. Z. Zhu and C. E. Woodcock, “Object-based cloud and cloud shadow detection in Landsat imagery,” Remote Sens. Environ. 118, 83–94 (2012). [CrossRef]
4. Z.-L. Li, B.-H. Tang, H. Wu, H. Ren, G. Yan, Z. Wan, I. F. Trigo, and J. A. Sobrino, “Satellite-derived land surface temperature: Current status and perspectives,” Remote Sens. Environ. 131, 14–37 (2013). [CrossRef]
5. A. A. Kokhanovsky, “Remote sensing of atmospheric aerosol using spaceborne optical observations,” Earth-Sci. Rev. 116, 95–108 (2013). [CrossRef]
6. N. Bellouin, O. Boucher, J. Haywood, and M. S. Reddy, “Global estimate of aerosol direct radiative forcing from satellite measurements,” Nature 438(7071), 1138–1141 (2005). [CrossRef]
7. H. D. Chen, X. X. Xiong, C. B. Sun, X. X. Chen, and K. F. Chiang, “Suomi-NPP VIIRS day-night band on-orbit calibration and performance,” J. Appl. Remote Sens. 11, 13 (2017). [CrossRef]
8. C. F. Schueler, T. F. Lee, and S. D. Miller, “VIIRS constant spatial-resolution advantages,” Int. J. Remote Sens. 34(16), 5761–5777 (2013). [CrossRef]
9. V. K. Prasad, Y. Kant, P. K. Gupta, C. Elvidge, and K. V. S. Badarinath, “Biomass burning and related trace gas emissions from tropical dry deciduous forests of India: A study using DMSP-OLS data and ground-based measurements,” Int. J. Remote Sens. 23(14), 2837–2851 (2002). [CrossRef]
10. C. D. Elvidge, K. E. Baugh, J. B. Dietz, T. Bland, P. C. Sutton, and H. W. Kroehl, “Radiance calibration of DMSP-OLS low-light imaging data of human settlements,” Remote Sens. Environ. 68(1), 77–88 (1999). [CrossRef]
11. X. Xiong, K. Chiang, J. McIntire, H. Oudrari, A. Wu, M. Schwaller, J. Butler, and and Ieee, “Early assessment of VIIRS on-orbit calibration and support activities,” in 2012 Ieee International Geoscience and Remote Sensing Symposium (2012), pp. 7189–7192.
12. G. Zhang, Y. Shi, and M. Xu, “Evaluation of LJ1-01 Nighttime Light Imagery for Estimating Monthly PM2.5 Concentration: A Comparison With NPP-VIIRS Nighttime Light Data,” IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing 13, 3618–3632 (2020). [CrossRef]
13. W. Zhai, B. Han, and C. Cheng, “Evaluation of Luojia 1802–1806 Nighttime Light Imagery for Built-Up Urban Area Extraction: A Case Study of 16 Cities in China,” IEEE Geosci. Remote Sens. Lett. 17(10), 1802–1806 (2020). [CrossRef]
14. P. Zhang, X. Hu, Q. Lu, A. Zhu, M. Lin, L. Sun, L. Chen, and N. Xu, “FY-3E: The First Operational Meteorological Satellite Mission in an Early Morning Orbit,” Adv. Atmos. Sci. 39(1), 1–8 (2022). [CrossRef]
15. G. Xingfa and T. Xudong, “Overview of China Earth Observation Satellite Programs [Space Agencies],” IEEE Geosci. Remote Sens. Mag. 3(3), 113–129 (2015). [CrossRef]
16. Y. Bao, L. Zhu, Q. Guan, Y. Guan, Q. Lu, G. P. Petropoulos, H. Che, G. Ali, Y. Dong, Z. Tang, Y. Gu, W. Tang, and Y. Hou, “Assessing the impact of Chinese FY-3/MERSI AOD data assimilation on air quality forecasts: Sand dust events in northeast China,” Atmos. Environ. 205, 78–89 (2019). [CrossRef]
17. Z. Yang, N. Lu, J. Shi, P. Zhang, C. Dong, and J. Yang, “Overview of FY-3 Payload and Ground Application System,” IEEE Trans. Geosci. Remote Sensing 50(12), 4846–4853 (2012). [CrossRef]
18. X. Hu, L. Sun, J. Liu, L. Ding, X. Wang, Y. Li, Y. Zhang, N. Xu, and L. Chen, “Calibration for the Solar Reflective Bands of Medium Resolution Spectral Imager Onboard FY-3A,” IEEE Trans. Geosci. Remote Sensing 50(12), 4915–4928 (2012). [CrossRef]
19. Y. Zhang, Z. Zheng, X. Hu, Z. Rong, and L. Zhang, “Lake Qinghai: Chinese site for radiometric calibration of satellite infrared remote sensors,” Remote Sens. Lett. 4(4), 315–324 (2013). [CrossRef]
20. P. M. Teillet, P. N. Slater, Y. Ding, R. P. Santer, R. D. Jackson, and M. S. Moran, “Three methods for the absolute calibration of the NOAA AVHRR sensors in-flight,” Remote Sens. Environ. 31(2), 105–120 (1990). [CrossRef]
21. Y. Wang, Y. Huang, S. Wang, Z. Li, Z. Zhang, X. Hu, and P. Zhang, “Ground-based Observation System Development for the Moon Hyper-spectral Imaging,” Publ. Astron. Soc. Pac. 129(975), 055002 (2017). [CrossRef]
22. Y. Wang, X. Hu, L. Chen, Y. Huang, Z. Li, S. Wang, P. Zhang, R. Wu, L. Zhang, and W. Wang, “Comparison of the Lunar Models Using the Hyper-Spectral Imager Observations in Lijiang, China,” Remote Sens. 12(11), 1878 (2020). [CrossRef]
23. F. Yu, X. Wu, M. Grotenhuis, and H. Qian, “Intercalibration of GOES Imager visible channels over the Sonoran Desert,” J. Geophys. Res.-Atmos. 119(14), 8639–8658 (2014). [CrossRef]
24. A. L. Quijano, I. N. Sokolik, and O. B. Toon, “Influence of the aerosol vertical distribution on the retrievals of aerosol optical depth from satellite radiance measurements,” Geophys. Res. Lett. 27(21), 3457–3460 (2000). [CrossRef]
25. C. J. Bruegge, R. N. Halthore, B. Markham, M. Spanner, and R. Wrigley, “Aerosol optical depth retrievals over the KONZA PRAIRIE,” J. Geophys. Res. 97(D17), 18743–18758 (1992). [CrossRef]
26. Q. Song, S. Chen, C. Xue, M. Lin, K. Du, S. Li, C. Ma, J. Tang, J. Liu, T. Zhang, and X. Huang, “Vicarious calibration of COCTS-HY1C at visible and near-infrared bands for ocean color application,” Opt. Express 27(20), A1615–A1626 (2019). [CrossRef]
27. N. T. Oneill, A. Royer, and J. R. Miller, “Aerosol optical depth determination from ground based irradiance ratios,” Appl. Opt. 28(15), 3092–3098 (1989). [CrossRef]
28. T. A. Berkoff, M. Sorokin, T. Stone, T. F. Eck, R. Hoff, E. Welton, and B. Holben, “Nocturnal Aerosol Optical Depth Measurements with a Small-Aperture Automated Photometer Using the Moon as a Light Source,” J. Atmos. Ocean. Technol. 28(10), 1297–1306 (2011). [CrossRef]
29. B. N. Holben, D. Tanré, A. Smirnov, T. F. Eck, I. Slutsker, N. Abuhassan, W. W. Newcomb, J. S. Schafer, B. Chatenet, F. Lavenu, Y. J. Kaufman, J. V. Castle, A. Setzer, B. Markham, D. Clark, R. Frouin, R. Halthore, A. Karneli, N. T. O’Neill, C. Pietras, R. T. Pinker, K. Voss, and G. Zibordi, “An emerging ground-based aerosol climatology: Aerosol optical depth from AERONET,” J. Geophys. Res.-Atmos. 106(D11), 12067–12097 (2001). [CrossRef]
30. O. Dubovik, A. Smirnov, B. N. Holben, M. D. King, Y. J. Kaufman, T. F. Eck, and I. Slutsker, “Accuracy assessments of aerosol optical properties retrieved from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements,” J. Geophys. Res. Atmos. 105(D8), 9791–9806 (2000). [CrossRef]
31. A. Barreto, E. Cuevas, B. Damiri, C. Guirado, T. Berkoff, A. J. Berjón, Y. Hernández, F. Almansa, and M. Gil, “A new method for nocturnal aerosol measurements with a lunar photometer prototype,” Atmos. Meas. Tech. 6(3), 585–598 (2013). [CrossRef]
32. R. Román, R. González, C. Toledano, Á. Barreto, D. Pérez-Ramírez, J. A. Benavent-Oltra, F. J. Olmo, V. E. Cachorro, L. Alados-Arboledas, and Á. M. de Frutos, “Correction of a lunar-irradiance model for aerosol optical depth retrieval and comparison with a star photometer,” Atmos. Meas. Tech. 13(11), 6293–6310 (2020). [CrossRef]
33. Á. Barreto, R. Román, E. Cuevas, A. J. Berjón, A. F. Almansa, C. Toledano, R. González, Y. Hernández, L. Blarel, P. Goloub, C. Guirado, and M. Yela, “Assessment of nocturnal aerosol optical depth from lunar photometry at the Izaña high mountain observatory,” Atmos. Meas. Tech. 10(8), 3007–3019 (2017). [CrossRef]
34. V. Matthais, V. Freudenthaler, A. Amodeo, I. Balin, D. Balis, J. Bösenberg, A. Chaikovsky, G. Chourdakis, A. Comeron, A. Delaval, F. De Tomasi, R. Eixmann, A. Hågård, L. Komguem, S. Kreipl, R. Matthey, V. Rizi, J. A. Rodrigues, U. Wandinger, and X. Wang, “Aerosol lidar intercomparison in the framework of the EARLINET project. 1.Instruments,” Appl. Opt. 43(4), 961–976 (2004). [CrossRef]
35. C. Bockmann, U. Wandinger, A. Ansmann, J. Bosenberg, V. Amiridis, A. Boselli, A. Delaval, F. De Tomasi, M. Frioud, I. V. Grigorov, A. Hagard, M. Horvat, M. Iarlori, L. Komguem, S. Kreipl, G. Larchevque, V. Matthias, A. Papayannis, G. Pappalardo, F. Rocadenbosch, J. A. Rodrigues, J. Schneider, V. Shcherbakov, and M. Wiegner, “Aerosol lidar intercomparison in the framework of the EARLINET project. 2. Aerosol backscatter algorithms,” Appl. Opt. 43(4), 977–989 (2004). [CrossRef]
36. G. Pappalardo, A. Amodeo, M. Pandolfi, U. Wandinger, A. Ansmann, J. Bösenberg, V. Matthias, V. Amiridis, F. De Tomasi, M. Frioud, M. Iarlori, L. Komguem, A. Papayannis, F. Rocadenbosch, and X. Wang, “Aerosol lidar intercomparison in the framework of the EARLINET project. 3. Ramanlidar algorithm for aerosol extinction, backscatter, and lidar ratio,” Appl. Opt. 43(28), 5370–5385 (2004). [CrossRef]
37. V. Freudenthaler, “About the effects of polarising optics on lidar signals and the Δ90 calibration,” Atmos. Meas. Tech. 9(9), 4181–4255 (2016). [CrossRef]
38. C. Weitkamp, Lidar, Range-Resolved Optical Remote Sensing of the Atmosphere (Springer Series in Optical Sciences, 2005).
39. F. Liu and F. Yi, “Lidar-measured atmospheric N2 vibrational-rotational Raman spectra and consequent temperature retrieval,” Opt. Express 22(23), 27833–27844 (2014). [CrossRef]
40. International Organization for Standardization, “Standard Atmosphere,” ISO 2533:1975 (1975).
41. M. G. Frontoso, “Lidar remote sensing for the characterization of the atmospheric aerosol on local and large spatial scale,” (PhD dissertation in Applied Physics, Universita degli di Napoli” Fererico II, 2007).
42. A. H. Omar, D. M. Winker, C. Kittaka, M. A. Vaughan, Z. Liu, Y. Hu, C. R. Trepte, R. R. Rogers, R. A. Ferrare, K.-P. Lee, R. E. Kuehn, and C. A. Hostetler, “The CALIPSO Automated Aerosol Classification and Lidar Ratio Selection Algorithm,” J. Atmos. Ocean. Technol. 26(10), 1994–2014 (2009). [CrossRef]
43. D. Josset, R. Rogers, J. Pelon, Y. Hu, Z. Liu, A. Omar, and P.-W. Zhai, “CALIPSO lidar ratio retrieval over the ocean,” Opt. Express 19(19), 18696–18706 (2011). [CrossRef]
44. X. Lu, Y. Jiang, X. Zhang, X. Wang, and N. Spinelli, “Two-wavelength lidar inversion algorithm for determination of aerosol extinction-to-backscatter ratio and its application to CALIPSO lidar measurements,” J. Quant. Spectrosc. Radiat. Transf. 112(2), 320–328 (2011). [CrossRef]
45. D. Müller, A. Ansmann, I. Mattis, M. Tesche, U. Wandinger, D. Althausen, and G. Pisani, “Aerosol-type-dependent lidar ratios observed with Raman lidar,” J. Geophys. Res. 112(D16), D16202 (2007). [CrossRef]
46. J. Xian, D. Sun, S. Amoruso, W. Xu, and X. Wang, “Parameter optimization of a visibility LiDAR for sea-fog early warnings,” Opt. Express 28(16), 23829–23845 (2020). [CrossRef]
47. Z. Liu, P. Voelger, and N. Sugimoto, “Simulations of the Observation of Clouds and Aerosols with the Experimental Lidar in Space Equipment System,” Appl. Opt. 39(18), 3120–3137 (2000). [CrossRef]
48. T. Y. Nakajima, T. Imai, O. Uchino, and T. Nagai, “Influence of daylight and noise current on cloud and aerosol observations by spaceborne elastic scattering lidar,” Appl. Opt. 38(24), 5218–5228 (1999). [CrossRef]
49. J. D. Spinhirne, “Micro pulse lidar,” IEEE Trans. Geosci. Remote Sensing 31(1), 48–55 (1993). [CrossRef]
50. A. Ansmann and D. Müller, “Lidar and Atmospheric Aerosol Particles,” in Lidar (SpringerLink, 2005), pp. 105–141.
51. J. N. Porter, B. R. Lienert, S. K. Sharma, and H. W. Hubble, “A small portable Mie-Rayleigh lidar system to measure aerosol optical and spatial properties,” J. Atmos. Ocean. Technol. 19(11), 1873–1877 (2002). [CrossRef]
52. D. N. Kafle and R. L. Coulter, “Micropulse lidar-derived aerosol optical depth climatology at ARM sites worldwide,” J. Geophys. Res. Atmos. 118(13), 7293–7308 (2013). [CrossRef]
53. J. D. Klett, “Stable analytical inversion solution for processing lidar returns,” Appl. Opt. 20(2), 211–220 (1981). [CrossRef]
54. S. T. Shipley, D. H. Tracy, E. W. Eloranta, J. T. Trauger, J. T. Sroga, F. L. Roesler, and J. A. Weinman, “High spectral resolution lidar to measure optical scattering properties of atmospheric aerosols. 1: Theory and instrumentation,” Appl. Opt. 22(23), 3716–3724 (1983). [CrossRef]
55. H. Baars, T. Kanitz, R. Engelmann, D. Althausen, B. Heese, M. Komppula, J. Preißler, M. Tesche, A. Ansmann, U. Wandinger, J.-H. Lim, J. Y. Ahn, I. S. Stachlewska, V. Amiridis, E. Marinou, P. Seifert, J. Hofer, A. Skupin, F. Schneider, S. Bohlmann, A. Foth, S. Bley, A. Pfüller, E. Giannakaki, H. Lihavainen, Y. Viisanen, R. K. Hooda, S. N. Pereira, D. Bortoli, F. Wagner, I. Mattis, L. Janicka, K. M. Markowicz, P. Achtert, P. Artaxo, T. Pauliquevis, R. A. F. Souza, V. P. Sharma, P. G. van Zyl, J. P. Beukes, J. Sun, E. G. Rohwer, R. Deng, R.-E. Mamouri, and F. Zamorano, “An overview of the first decade of PollyNET: an emerging network of automated Raman-polarization lidars for continuous aerosol profiling,” Atmos. Chem. Phys. 16(8), 5111–5137 (2016). [CrossRef]
56. D. H. Hohn and W. Buchtemann, “Spectral radiance in the s20-range and luminance of the clear and overcast night sky,” Appl. Opt. 12(1), 52–61 (1973). [CrossRef]
57. C. Leinert, S. Bowyer, L. K. Haikala, M. S. Hanner, M. G. Hauser, A. C. Levasseur-Regourd, I. Mann, K. Mattila, W. T. Reach, W. Schlosser, H. J. Staude, G. N. Toller, J. L. Weiland, J. L. Weinberg, and A. N. Witt, “The 1997 reference of diffuse night sky brightness,” Astron. Astrophys. Suppl. Ser. 127(1), 1–99 (1998). [CrossRef]
58. H. Neckel and D. Labs, “Improved data of solar spectral irradiance from 0.33 to 1.25µ,” Sol. Phys. 74(1), 231–249 (1981). [CrossRef]
59. Q. Yang, M. Liu, X. Xie, J. Kang, A. Guo, H. Zhu, and Q. Gao, “The influence of output pulse spectral shape and bandwidth on pulse contrast in the chirped pulse amplification,” Optik 125(3), 1448–1450 (2014). [CrossRef]
60. M. Shangguan, H. Xia, C. Wang, J. Qiu, G. Shentu, Q. Zhang, X. Dou, and J.-W. Pan, “All-fiber upconversion high spectral resolution wind lidar using a Fabry-Perot interferometer,” Opt. Express 24(17), 19322–19336 (2016). [CrossRef]
61. G. Pappalardo, A. Amodeo, A. Apituley, A. Comeron, V. Freudenthaler, H. Linné, A. Ansmann, J. Bösenberg, G. D’Amico, I. Mattis, L. Mona, U. Wandinger, V. Amiridis, L. Alados-Arboledas, D. Nicolae, and M. Wiegner, “EARLINET: towards an advanced sustainable European aerosol lidar network,” Atmos. Meas. Tech. 7(8), 2389–2409 (2014). [CrossRef]
62. C. A. Hostetler, Z. Liu, J. Reagan, M. Vaughan, and C. Trepte, “Calibration and Level 1 Data Products,” CALIOP Algorithm Theoretical Basis Document (2006).
63. C. Zhao, Y. Wang, Q. Wang, Z. Li, Z. Wang, and D. Liu, “A new cloud and aerosol layer detection method based on micropulse lidar measurements,” J. Geophys. Res. Atmos. 119(11), 6788–6802 (2014). [CrossRef]
64. Z. Yin, F. Yi, F. Liu, Y. He, Y. Zhang, C. Yu, and Y. Zhang, “Long-term variations of aerosol optical properties over Wuhan with polarization lidar,” Atmos. Environ. 259, 118508 (2021). [CrossRef]
65. 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]
66. I. Mattis, G. D’Amico, H. Baars, A. Amodeo, F. Madonna, and M. Iarlori, “EARLINET Single Calculus Chain – technical – Part 2: Calculation of optical products,” Atmos. Meas. Tech. 9(7), 3009–3029 (2016). [CrossRef]
67. V. Freudenthaler, H. Linné, A. Chaikovski, D. Rabus, and S. Groß, “EARLINET lidar quality assurance tools,” Atmos. Meas. Tech. Discuss. 2018, 1–35 (2018). [CrossRef]
68. A. Ansmann, M. Riebesell, U. Wandinger, C. Weitkamp, E. Voss, W. Lahmann, and W. Michaelis, “Combined Raman elastic-backscatter lidar for vertical profiling of moisture, aerosol extinction, backscatter, and lidar ratio,” Appl. Phys. B 55(1), 18–28 (1992). [CrossRef]
69. J. Biele, G. Beyerle, and G. Baumgarten, “Polarization lidar: Corrections of instrumental effects,” Opt. Express 7(12), 427–435 (2000). [CrossRef]
