1. Introduction

The atmosphere condition is expressed by its composition such as water vapor, aerosol and clouds. Water vapor has a great impact on weather and climate due to its role in the radiative energy transfer, hydrological cycle, and atmospheric chemistry processes [1]. The complexity of atmospheric aerosols results from their highly variable particle number concentrations, multimodal size distributions, variable shape characteristics, complex chemical composition and mixing behavior. And the correspondingly large temporal and spatial (horizontal and vertical) variability in the aerosol characteristics is the main reason for the high uncertainties in our quantitative understanding of the role of atmospheric aerosol in environmental, weather, and climate-related processes [2]. The International Panel on Climate Change (IPCC) Fourth Assessment Report [3] has identified aerosol radiative forcing and the impact of aerosols on cloud and precipitation processes as one of the major unknowns in our understanding of climate change. Practically all long-range transport of aerosols occurs at elevated height levels decoupled from the ground. A global climatology of the mesoscale and large-scale aerosol transport based on long-term data sets of vertically resolved aerosol distributions does, however, not exist [2].

For the detection of the water vapor profile, two lidar techniques have been applied: the Differential Absorption Lidar (DIAL) and the Raman lidar technique. In terms of the DIAL, two laser pulses at different wavelengths, called “on-line” and “off-line”, respectively, are emitted to the atmosphere [4–7] for the measurement. In this paper, the lidar system applies Raman technique. This technique was firstly used by Melfi [8, 9] and Cooney [10] and the profiles of water vapor mixing ratio were retrieved and provided. A water vapor Raman lidar was built at the Table Mountain Facility (TMF) of the Jet Propulsion Laboratory (JPL) in California, in 2005 [11]. And more than a year of routine 2-h-long nighttime measurements 4–5 times per week have been completed. In July, 2007, the front-end of the lidar receiver was redesigned to permanently suppress the fluorescence identified in the fiber optic [12]. In 2005, the German Meteorological Service (DWD) complemented the suite of remote-sensing instruments of its Richard Aßmann Observatory at Lindenberg, Brandenburg, with the automated water-vapor Raman lidar system RAMSES (Raman lidar for atmospheric moisture sensing). In 2010, the RAMSES was therefore redesigned and expanded to become a powerful day- and night-time multi-parameter Raman lidar for long-term monitoring of the atmosphere [13]. In 2008, Caeli was set up in Netherland as a high-performance, multiwavelength Raman lidar. The instrument provides backscatter and extinction profiles, depolarization and water vapor profiles [14]. Since the deployment at CESAR (Cabauw Experimental Site for Atmospheric Research) in May 2008, Caeli contributes to observation programmes and studies. The instrument is part of European Aerosol Research Lidar Network (EARLINET), has provided data during the Intensive Observation Period at Cabauw Tower (IMPACT) and Nitrogen Dioxide measuring Instrument (CINDI) campaigns and provides correlative measurements for Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO). Recently, the instrument was compared to its peers in the EARLINET Reference Lidar Intercomparison (EARLI09) in Leipzig. In 2009, RAman Lidar for meteorological Observations (RALMO) was developed as fully automated, eye-safe instrument for operational use by the Swiss meteorological service – MeteoSwiss [1, 15]. The lidar is able to provide vertical profiles of water vapor mixing ratio with time resolution from 5 to 30 min and vertical resolution from 15 m in boundary layer and 75 – 500 m in free troposphere. The daytime vertical operational range of the lidar extends from about 50 m to mid-troposphere and the detection limit is 0.5 g·kg⁻¹. In night-time conditions the vertical operational range extends up to the tropopause with 0.01 g·kg⁻¹ detection limit [1]. In 2010, Bo Liu and Zhien Wang measured the water vapor and aerosol both during nighttime and daytime by using an airborne Raman lidar [16].

For the detection of the aerosol and cloud, the polarization lidar technique [17–19] is a well-established method. Ansmann et al. [20, 21] set up a polarization lidar to distinguish ice clouds from water clouds and to identify layers with ice crystals in mixed–phase clouds. In order to study the evolution of contrails, Freudenthaler et al. also applied a scanning polarization lidar in 1996 [22]. The technique has been used to identify the type of polar stratospheric clouds [19,23,24] and volcanic ash in the troposphere and stratosphere [25–27]. The polarization lidar can also be used to do some research on aerosol profiling [27–31] and help to classify desert dust from other aerosols such as biomass burning aerosols, fresh smoke, etc. The polarization lidar allows us to unambiguously discriminate desert dust from other aerosols [32–37]. Based on model calculations, it has been demonstrated that the spectral dependence of the dust linear depolarization ratio is sensitive to the size distribution of the nonspherical scatterers [38]. Thus, observations of the linear depolarization ratio at several wavelengths may be used in retrieval schemes [39] to improve the estimation of the microphysical properties of dust from optical measurements [40]. The first dual-wavelength aerosol polarization lidar measurements were presented by Sugimoto et al. [41]. Freudenthaler et al. [42] measured the vertical profiles of the linear particle depolarization ratio of pure dust clouds during the Saharan Mineral Dust Experiment (SAMUM-1) at Ouarzazate, Morocco (30.9$°$N, –6.9$°$E), close to source regions in May–June 2006, with four lidar systems at four wavelengths (355 nm, 532 nm, 710 nm and 1064 nm). Tesche et al. 2010 [43] used multiwavelength aerosol Raman lidar in combination with polarization lidar at Praia (14.9$°$N, 23.5$°$W), Cape Verde, to separate the optical properties of desert dust and biomass burning particles as a function of height in the mixed dust and smoke plumes over the tropical North Atlantic west of the African continent. The observations were performed during the Saharan Mineral Dust Experiment (SAMUM-2) in January and February 2008. In order to classify different multi-component mixtures of atmospheric particles during long-range transport, David et al. 2013 [44] proposed a new methodology based on combining a sensitive and accurate UV-VIS polarization lidar experiment with T-matrix numerical simulations and air mass back trajectories. Tian Zhou and Jianping Huang et al. 2013 [45] proposed a new method based on the depolarization–attenuated backscatter relationship. And it can significantly improve the classification of cloud and dust plumes and can supplement the current space-borne LIDAR discrimination approach, especially over dust source regions. Yun He and Fan Yi, 2014 [46] measured the vertical distribution, horizontal range, and optical properties of Asian dust by using a ground-based depolarization lidar and CALIPSO at Wuhan (30.5$°$N, 114.4$°$E), China. Two years of Asian dust transported over long distances were simultaneously observed and analyzed.

For purpose of combined meteorological research and aiming at the detection of water vapor mixing ratio, depolarization ratio, extinction and backscatter coefficient and cloud information, the multifunctional WACAL is developed. The lidar system was developed from June 2012 to July 2013. The WACAL is set up on an optical platform and housed in a weather-proof cabinet of 6000 mm$\times$3000 mm$\times$2000 mm. Through the design for transportation platform of standard container, it can be transported and deployed easily in the field. The WACAL can measure the aerosol and cloud optical properties as well as the water vapor mixing ratio. It is useful for studying the direct and indirect effects of the aerosol on the climate change. Based on its multifunctional design, the WACAL can also be capable of classifying the aerosol type and cloud phase.

In this paper, the lidar technique is discussed in section 2 and the design and explanation of the WACAL are provided in details. In section 3, the methodology and retrieval method are described and the measurement examples, data analysis and conclusions are presented as well.

2. Water vapor, cloud and aerosol lidar

This lidar hardware consists of four subsystems, including transmitter, receiver, data acquisition and auxiliary system. The whole system is shown in Fig. 1 and the schematic diagram is provided in Fig. 2.

 

Fig. 1 The photo of the whole system.

Download Full Size | PDF

 

Fig. 2 Schematic diagram of WACAL system

Download Full Size | PDF

2.1. Transmitter

The laser transmitter of WACAL, Continuum Powerlite 9030, is a high peak power flash lamp pumped Nd:YAG laser with three wavelengths of 354.7 nm, 532 nm and 1064 nm. And the pulse energy is 410 mJ, 120 mJ and 700 mJ, respectively. The flash lamp-pumped Nd:YAG laser transmitter generates light pulses at the wavelength of 1064 nm. And after the second harmonic generator (SHG) and third harmonic generator (THG), the wavelengths of 532 nm (frequency doubled) and 354.7 nm (frequency tripled) are generated. With the residual light at wavelength 1064 nm, all these three beams are transmitted to the atmosphere. The basic parameters are listed in Table 1. The light with the wavelength of 354.7 nm is used for exciting Raman backscatter of nitrogen and water vapor molecule. Meanwhile the backscattered light excited by the light at the wavelengths of 532 nm and 1064nm are utilized for the detection of aerosol and cloud. For purpose of decreasing divergence angle, two beam expanders are designed. As shown in Fig. 2, the transmitter includes laser, one half-wave plate, one reflecting prism, one mirror, two beam expanders and two windows with anti-reflective coating. The expanded beams with 90 mm diameter transmit into the atmosphere on an axis closed to the receiver axis.

Table 1. Component parameters of the WACAL system

View Table | View all tables in this article

In order to decrease the beam divergence angle and ensure the high optical transmittance of the unit, the beam expander is used. Unfortunately, it is difficult to use a commercial reliable coating lens that can stand for such a peak power. Consequently, for the practical consideration, two separated beam expanders are used for the light at wavelengths of 354.7 nm and 532 nm/1064 nm, respectively. The laser at wavelength of 354.7 nm is directed to the vertically installed beam expander by a Pellin-Broca (PB) prism. To minimize the Fresnel reflection losses the input and output surfaces of the prism are at Brewster angle, and the horizontal laser polarization is rotated to vertical with a multiple order half-wave plate, which is installed before the prism. Considering the light at wavelength of 354.7nm is used to excite the weak Raman backscatter, in order to decrease the beam divergence angle, it is transmitted to a 10-times beam expander-A. The beam expander-A consists of two lens. The first lens is a plano-concave lens with the focal length of −60 mm. After this lens, the beam is diverged. The laser arrives at the second plano-convex lens with the focal length of 600 mm and the diameter of 120 mm and then the beam is collimated. After the beam expander, the full beam divergence is decreased to 0.05 μrad. In consideration of the size and diameter of the expander and the purpose of exciting the Mie and Rayleigh backscatter, the laser at wavelengths of 532 nm and 1064 nm are transmitted to a 6-times beam expander-B. A similar work with beam expander-A, after the crescent lens with a focal length of −100 mm (at wavelength of 532 nm) and the plano-convex lens with the focal length of 600 mm (at wavelength of 532 nm) and diameter of 100 mm, the lasers at wavelength of 532 nm and 1064 nm are collimated as well. In these two beam expanders, all of the optical components are coated with anti-reflecting (AR) film to prevent the reflections. After the two expanders, two fused silica windows with anti-reflective coating are installed for protecting the optical components in the cabinet.

2.2. Receiver

After a laser pulse is transmitted to the atmosphere, molecules and particles scatter the light in all directions. A portion of the light is scattered backwards to the lidar. The light is collected by telescopes and then transmitted to the detection system. In order to increase the amount of collected light, this system deployed four Newtonian telescopes with the diameter of 300 mm and the focal length of 1524 mm. The primary mirror of Newtonian telescope is a parabolic mirror while the secondary mirror is a plane mirror. The 4 telescopes assembly served as a telescope array with an equivalent receiver aperture of about 610 mm. The design of the array has better practicability for detecting the signal from near field and far field. Moreover, it takes the collection efficiency of the strong elastic backscatter light and the weak Raman backscatter light into consideration. This design also makes the system easy to transport and suitable for field experiments. However, it makes the system more complicated to align the telescope and to determine the overlap function.

After collected by the telescope array, the scattered light is transmitted into 5 fibers, including 4 far-field fibers and 1 near-field fiber. Considering the overlap function and the collection efficiency of near-field signal, the near-range fiber is designed specially as shown in Fig. 3.

 

Fig. 3 Schematic of the telescope array and fibers.

Download Full Size | PDF

As a significant estimation factor for the WACAL, the overlap function of WACAL is provided in Fig. 4. In the WACAL, the distance between the axes of telescope and laser beam is about 250 mm. The overlap of the WACAL is obtained by ZEMAX simulation and experiment respectively. Figure 4(a) shows the overlap function of the Raman channel by simulation. For detection of the water vapor mixing ratio below 5 km, the telescope direction and entrance of fiber position are adjusted for the different range detection. Moreover, by applying the near-range fiber, the near field signal is available and enhanced and the detection range can be extended from 1.3 km to lower than 0.5 km. The thick black line indicates the equivalent overlap of the whole Raman channel. Figure 4(b) presents the overlap of the polarization and infrared channel by simulation. Because of the blocking of the secondary mirror and the other components, the full overlap is about 0.88. The distance of full overlap function is about 5 km. In Fig. 4(c) and Fig. 4(d), the overlaps of Raman channel and polarization and infrared channel are determined by experiment.

 

Fig. 4 Overlap of the WACAL. (a) Overlap function of the Raman channel simulated by ZEMAX; (b) Overlap function of the Raman channel measured by experiment ;(c) Overlap function of the polarization and infrared channel simulated by ZEMAX and (d) Overlap function of the polarization and infrared channel measured by experiment.

Download Full Size | PDF

2.2.1. Polychromator of Raman channel

The optical setup of the polychromator of the Raman channel is provided in Fig. 2 (c). In this channel, since the Raman backscatter cross section (~10⁻³⁰ cm²· sr⁻¹) is several orders of magnitude smaller than Rayleigh backscatter cross section (~10⁻²⁸ $c{m}^2·s{r}^{-1}$) and Mie backscatter cross section (~10⁻¹⁰ cm²· sr⁻¹), the collection of the Raman signal at wavelength of 386.7 nm and 407.5 nm is difficult. In order to increase the amount of collected light, we deploy the four telescopes as shown before. In order to avoid the interference from the elastic backscatter signal, band-pass filters of FF01-390/40-25 from Semrock are used ahead of the fibers. The center wavelength of the filters is 390 nm and the FWHM is 44.6 nm. The transmission between 370 nm and 410 nm is > 93% and the Optical Density (OD) is >5 for light at the wavelength of 354.7 nm and 532 nm. After the filters, four fibers are mounted for the coupling of the signal. The core of the fibers is 200 microns and the numerical aperture is 0.22. After the coupling of fiber, the Raman signal is delivered to a grating-based polychromator system and separated as Raman signal of nitrogen and water vapor.

When the signal is transmitted to the polychromator system, the light is dispersed and then collimated by the convex lens with the focal length of 50.0 mm. Through the reflection of the reflecting prism, the parallel light arrived at the diffraction grating. The groove density of the grating is 1302 l/mm and the blaze wavelength is 400 nm. According to the grating equation [47,48] $\sin i+\sin {\gamma }_m=mG\lambda$(where $i$ is the incidence angle, $G$ is the grating groove frequency per unit length, being the reverse of the groove spacing, $m$ is the grating order and $\lambda$ is incident light wavelength. In order to avoid symbol confusion, the symbols used in this paper are different from the original literature. The Raman signal of nitrogen at 386.7 nm and water vapor at 407.5 nm are separated with different diffraction angle ${\gamma }_m$. Then the filters are used to ensure the purity of each signal at 386.7 nm and 407.5 nm. The center wavelength (CWL) of the filter-1 from Barr is 407.5$\pm$0.1 nm. Meanwhile, the CWLs are 386.7 nm$\pm$0.1 nm and 354.7 nm$\pm$0.08 nm for filter-2 and filter-3, respectively. The ‘$\pm$’ stands for the uncertainty of the central wavelength. The FWHM of all filters is 0.5 nm$\pm$0.10 nm and peak transmittance is > 50% and OD is 5 when out of band blocking from 200 nm to 1200 nm. Note that together with the filters before the fibers, the total OD in the Raman channel is >10 to eliminate the interference from the elastic backscatter signal. Whiteman evaluated the temperature-dependent lidar equations of the traditional Raman lidar in 2003 [49]. According to the study of Whiteman, the ratio of transmitted intensities between 200 K and 300 K for water vapor passbands at central wavelength of 407.50 nm is close to 1. As a consequence, the temperature has slight impact on WACAL and the temperature dependence can be ignored. The parallel signal after the filter is then focused by a plano-convex lens with the focal length of 100 mm. Finally, the signals are acquired by the photomultiplier tubes (PMTs, H10721P-110 from Hamamatsu), which are mounted at the focal point of the plano-convex lens. The specifications of the optical elements of this channel are shown in Table 2.

Table 2. Specifications of the optical elements in Receiver

View Table | View all tables in this article

2.2.2. Polarization channel and infrared channel

Considering the high power laser and the relatively strong Rayleigh and Mie scattering, only one telescope of the telescope assembly is utilized for the collection of elastic backscatter light in the polarization channel and the infrared channel. The optical setup is also described in Fig. 2(b). The details of the transmitting system have been explained in section 2.1. One adjustable aperture is installed at the focal point of the telescope ahead of the polarization channel. Since the fiber is not used in this channel, the field of view (FOV) is determined by adjusting the pinhole aperture. In this measurement, the aperture is fixed as 2 mm and the FOV is determined as 1.3 mrad. After the collimating lens with a focal length of 50 mm, the divergent light is changed to parallel signal. By utilizing a dichroic long-pass filter-1 with reflection wavelength of 350 nm to 430 nm and transmission wavelength of 470 nm to 1600 nm, the elastic signal at wavelength of 354.7 nm and Raman signal are reflected and the signal at wavelengths longer than 470 nm transmitted through the filter. Similarly, the elastic signal of 532 nm and 1064 nm is separated by the dichroic long-pass filter-2 with reflection wavelength of 460 nm to 570 nm and transmission wavelength of 625 nm to 1600 nm. The signal at wavelength of 1064 nm is detected by the Avalanche Photo Diode (APD) of G8931-20 from Hamamatsu. The parallel and perpendicular backscatter signals at 532 nm are collected to retrieve the depolarization ratio. The parallel-polarized and perpendicular-polarized signals are separated by the polarizing beamsplitter (PBS) and detected by PMTs respectively. The transmissivity of the PBS for parallel-polarized signal is > 90% while the reflectivity for perpendicular-polarized signal is > 99.5%. The specifications of the optical elements of this channel are also listed in Table 2.

2.3. Data acquisition

Data acquisition consists of PMT, APD, analog acquisition and photon-counting transient recorder and Computer. After the detection of the PMT and APD, the optical signal is transferred to electrical signal and stored by the analog acquisition and photon-counting transient recorder from Licel which has 6 analog/photon-counting channels and 2 trigger interfaces. The channel C1, C2 and C3 are used for the Raman signal detection. C4 and C5 are utilized for the parallel/perpendicular polarization measurement and C6 is for the infrared scattering signal acquisition. The trigger signal to synchronize the acquisition is generated by an optoelectronic detector near the exist window of the laser transmitter.

In the Raman channel, polarization channel and infrared channel, five PMTs and one APD are mounted for the detection of the backscatter signal. The spectral response range of PMT is between 230 nm and 700 nm. The peak wavelength is 400 nm and the sensitivity can reach up to 110 mA/W. The gain is proportional to the control voltage. In this system, the control voltage is set as 1.1 V and the gain is >$4\times {10}^6$. On the basis of characteristics of PMT, it is suitable for measuring the signal at the wavelengths of 354.7 nm, 386.7 nm, 407.5 nm and 532 nm.

However, since PMT is unavailable for the measurement of infrared signal, APD is taken into consideration. The advantages of APD are low voltage operation (40 V to 60 V), low dark current (max. 200 nA), low capacitance (1.5 pF) and high sensitivity (0.9 A$·$W⁻¹ @ peak sensitivity wavelength of 1550 nm). The spectral response range is between 950 nm and 1700 nm. For the signal at 1064 nm, the photo sensitivity is 0.5 A$·$W⁻¹. As a consequence, it is available for the detection of infrared signal at 1064 nm. In this channel, only analog acquisition is used.

2.4. Auxiliary system

Figure 1 has shown the exterior and the opened instrument of the WACAL. The weather proof cabinet is with double layer walls filling with insulated foam. The cabinet includes three cabins: lidar cabin, electricity and water chiller cabin and working cabin. In Fig. 5, the pictures of three cabins are provided.

 

Fig. 5 The photos of the three cabins: (a). Lidar cabin; (b). Electricity and water chiller cabin; (c). Operating cabin.

Download Full Size | PDF

In the lidar cabin, all of the optical components including laser head, beam expanders and the receiver optics with the telescopes and the three channels are supported by the optical platform. For the consideration of making full use of the space in the cabin, one aluminum frame to hold the polychromator is mounted on the optical bench above the laser. One air conditioner is mounted on the roof of the cabin to control the temperature of cabin and protect the optical components from moisture. In order to avoid the dust air exchange between the cabin inside and the environment, two fused silica windows with diameter of 160 mm are mounted over the two beam expanders and four fused silica windows with diameter of 330 mm are mounted over the four telescopes. In order to deploy the system in harsh environment, a Fan Filter Unit (FFU) is installed in the lidar cabin to filter dust and protect optics coating under high peak power laser, which is commonly used for optical clean room. In case of power outage, one high power uninterruptible power supply (UPS) of 10 kVA is to manage power cuts and enable the lidars to send alert message and close down software safely for up to 15 minutes without power.

In the electricity and water chiller cabin, the acquisition system, laser power supply and the water chiller are mounted. Since the laser chiller inside the cabin generate a lot of heat, which is harmful for the operation of the laser, it is essential to cool the air in this cabin. The ventilation facility with high ventilation rate fan is taken into consideration, which played a very practical role in the high elevation and low air pressure field experiment, e.g. the later Tibetan Plateau field campaign (described in a separated paper). Moreover, one filtering window is installed to prevent the infiltration of the dust and keep the cabin clean.

In the operation cabin, several computers are placed for the controlling of the whole system.

3. Measurement examples, data analysis and conclusions

In this section, the methodology and inversion process of the water vapor, backscatter coefficient, depolarization ratio and cloud height are described. Moreover, in order to verify the capability of the WACAL system, some case studies are presented. In this section, because of the lower SNR of Raman signal during daytime, the water vapor mixing ratio is only measured during nighttime. The measurements of aerosol and cloud are operated all the time during daytime and nighttime.

3.1. Methodology

3.1.1. Water vapor mixing ratio

The Raman lidar equation can be described as Eq. (1) [1]:

According to Eq. (1), the backscatter signal of $N_2$ and $H_{2}O$ are obtained as $P(z,{\lambda }_{N_2})$ and$P(z,{\lambda }_{H_{2}O})$. The water vapor mixing ratio can be calculated by Eq. (3):

The $\alpha (z\text{'},{\lambda }_{N_2})$ and $\alpha (z\text{'},{\lambda }_{H_{2}O})$ can be calculated by Raman method [50]. The calibration constant is retrieved using regression to a vertical water vapor mixing ratio profile obtained by a reference radiosonde of GTS1 type. The radiosonde provides temperature accuracy of $\pm 0.2°C$, relative humidity accuracy of $\pm 5\%$ and pressure accuracy of $\pm 1hPa$. . Additionally, the pressure and relative humidity profiles are also obtained. The Eq. (5) is used to obtain a mixing ratio profile from radiosonde data. In this equation, the temperature, pressure and relative humidity profiles are used and the mixing ratio WR (g·kg⁻¹) is then estimated.

The calibration constant for this comparison was retrieved using regression to a vertical water vapor mixing ratio profile obtained by a reference radiosonde as shown in Fig. 4. The text routines will be listed in Table 3. The lidar water vapor mixing ratio ($W_{Lidar}$) profile was calculated according to Eq. (3) with a calibration constant set to one. We assume that the relationship between lidar data $W_{Lidar}=\Delta T({\lambda }_{N_2},{\lambda }_{H_{2}O},z)·P(z,{\lambda }_{H_{2}O})/P(z,{\lambda }_{N_2})$ and radiosonde data $W_{Sonde}^{}$ as Eq. (7):

Table 3. Test routines of the validation observations

View Table | View all tables in this article

Where $C$ is the calibration constant and $D$ is the offset. Using the linear model, the lidar and radiosonde profiles are fitted. The slope from the fit is a direct estimation of the lidar calibration constant $C$. According to Fig. 6, $C$ is found to be equal to 219. $D$ is the offset and determined as −0.34. Result from the different observation stations of the WACAL and radiosonde and the WACAL system error, the offset $D$ exists. The correlation coefficient of these two system is 0.83. The standard deviation is 1.4 and the number of samples is 169. After the calibration the water vapor mixing ratio can be rewritten as Eq. (8):

 

Fig. 6 (a). Distance between sites of WACAL and radiosonde; (b). Regression of WACAL mixing ratio profile to radiosonde measurement.

Download Full Size | PDF

Aiming at validating the calibration of water vapor mixing ratio, the scatter diagram based on the calibrated Lidar data and radiosonde data measured is drawn in Fig. 7. And the test routines of validation are different from the calibration test routines. The validation test was operated from 10 July 2014 to 16 August 2014. According to the figure, the correlation coefficient can reach up to 0.94 and mean deviation is 0.77 $g·k{g}^{-1}$. Considering the range difference (16.7 km) between the observation of the WACAL and radiosonde, the correlation coefficient and deviation are acceptable. As a conclusion, the calibration of the water vapor mixing ratio is satisfied for the routine observation.

 

Fig. 7 Validation of the calibrated water vapor mixing ratio (red dashed line is 1:1 curve and black line is fitting curve).

Download Full Size | PDF

3.1.2. Determination of depolarization ratio and backscatter coefficient

The SNR of the signal in this channel is calculated as Eq. (9):

According to the lidar equation, the $P_{\parallel }$ and $P_{\perp }$ can be described as Eq. (11) and Eq. (12), respectively:

$C_{\parallel }$ and $C_{\perp }$represent the different receiver efficiency on the two channels including the PBS cross-talk effect, and the lidar system depolarization effects including non-linear polarized laser source and optical device depolarization effect. Moreover, the depolarization ratio can be affected by the misaligned optical arrangement and depolarization effect of telescope, long-pass filter and IF. As a result, ${\delta }^v$cannot be used directly to classify the aerosol types and cloud phases without calibration. In this paper, the half wave plate method [42, 52] is used for calibration and the calibrated depolarization ratio ${\delta }_{cal}^v$ can be described as Eq. (13). During the calibration process, the cross talk from PBS and the gain ratio between parallel and perpendicular channel are solved.

Based on the signal of the polarization lidar at 532nm, the aerosol extinction coefficient can be calculated by the Fernald method [53], and can be written as Eq. (14):

The algorithm for determination of the $z_0$ is known as “Minimum value method”. In this method, the $z_0$ is corresponding to the height $z$ where the value of $P_{{\lambda }_0}(z)z^2/{\beta }_{{\lambda }_0}^{mol}(z)$ is minimum. For the determination of the ${\alpha }_{{\lambda }_0}^{aer}(z_0)$, as the reference height $z_0$ is fixed, the iteration of Fernald algorithm is taken into consideration. The ${\alpha }_{{\lambda }_0}^{aer}(z_0)$ starts with the value of 4 × 10⁻⁹. From the derived extinction coefficient result, the mean value ${\alpha }_{{\lambda }_0,mea}^{aer}(z_0)$ is computed. For each iteration step, the input value of ${\alpha }_{{\lambda }_0}^{aer}(z_0)$ is increased by 10% until the relative difference between ${\alpha }_{{\lambda }_0}^{aer}(z_0)$ and ${\alpha }_{{\lambda }_0,mea}^{aer}(z_0)$ is less than 5% [2].

3.1.3. Cloud height measurement and cloud statistics

WACAL is capable of detecting of cloud base height (CBH) and even cloud top height (CTH) when the clouds can be penetrated by laser. For purpose of determining the cloud height, the algorithm that combines “differential zero-crossing algorithm” [54] and “threshold algorithm” [54] is taken into account. Generally cloud height can be determined directly from these zero crossings of the first derivative of backscatter intensity $dP(z)/dz$. However, in order to exclude interference of spurious zero crossings, The $dP(z)/dz$ is determined from a least-squares fit by using a multipoint window that slides through successive points from $z_0$ to the end of the lidar profile. Besides, a range-dependent threshold is defined, based on the noise in the recorded lidar signal. Signal excursion above the threshold value is identified as a cloud. By this algorithm, cloud layers apparent in the return signal are identified. However, if the lower clouds were too thick to be penetrated, the higher clouds will not be detected and cannot be identified.

3.2. Results

3.2.1. Water vapor mixing ratio

Here we provide case studies of water vapor mixing ratio in Fig. 8. Several intercomparisons between lidar derived vertical profiles and radiosonde measurements are presented. In Fig. 8, the blue dashed line indicates the water vapor mixing ratio measured by lidar and the horizontal lines show the error bar of the data. The red line shows the data of water vapor mixing ratio measured by radiosonde. All of the water vapor mixing ratio profiles are averaged every 60 minutes and the range resolution is 75 m. According to Fig. 8, the water vapor mixing ratio on 26 and 31 May 2014 are less than those on 6 and 7 June 2014. Due to the high spatial resolution of WACAL system, details of the vertical profiles of water vapor mixing ratio can be noticed. On 31 May 2014, one cloud layer is measured and the water vapor mixing ratio in the cloud is bigger than that of surrounding atmosphere. From Fig. 8, the results retrieved by WACAL and radiosonde show a good consistency.

 

Fig. 8 Water vapor mixing ratio study cases in Qingdao on 26 and 31 May, 6 and 7 June 2014.

Download Full Size | PDF

The water vapor mixing ratio can be obtained by the Raman channel, which is insensitive to the overlap correction because of the slight different optical path setup for the Raman channels. This allows that the analysis of water vapor mixing ratio down to 400 m AGL. Defined as the ratio of perpendicular-polarized and parallel-polarized signal, depolarization ratio also allows to be calculated down to 400 m. In contrast, affected by the overlap function, the backscatter coefficient and extinction coefficient below 1000 m AGL should be determined seriously.

3.2.2. Depolarization ratio, Backscatter coefficient and color ratio of cloud and aerosol

In this section, the depolarization ratio and backscatter coefficient of cloud and aerosol will be discussed.

One case study of 31 May 2014 including information about depolarization ratio and backscatter coefficient is provided. For the consideration of the reliability of the data, only the signal of SNR >10 is used. The temporal development of data products is presented in Fig. 9, with temporal and spatial resolution of 16 s and 3.75 m respectively.

 

Fig. 9 Depolarization ratio and backscatter coefficient measured by WACAL on 31 May 2014: (a) Temporal development of ${\delta }_{cal}^v$, (b) Vertical profile of ${\delta }_{cal}^v$ at 20:25 LST, (c) Temporal development of ${\beta }_a$, (d) Vertical profile of ${\beta }_a$ at 20:25 LST.

Download Full Size | PDF

The vertical profile of depolarization ratio which is drawn in Fig. 9(b) is corresponding to the data represented by the black line at about 20:25 LST (LST = UTC + 8h) in Fig. 9(a). Meanwhile, the backscatter coefficient at wavelength of 532 nm is calculated and the vertical profile of backscatter coefficient which is drawn in Fig. 9(d) is corresponding to the data represented by the black line at about 20:25 LST in Fig. 9(c). The figures provide the temporal development of ${\beta }_a$ and ${\delta }_{cal}^v$. It can be seen that the depolarization ratio of clouds located at AGL height of 6 km to 8 km is 22.6% ± 5.6% while the backscatter coefficient is 0.095 ± 0.03 km⁻¹sr⁻¹. Similarly, the ${\delta }_{cal}^v$ and ${\beta }_a$ of the aerosol layer suspended in the range of 2 km to 4 km are 7.2% ± 0.7% and 0.0039 ± 0.0015 km⁻¹sr⁻¹, respectively.

Owning to the multi-wavelength design of WACAL, the color ratio (${\beta }_a^{532}/{\beta }_a^{355}$) can be obtained. In Fig. 10, the color ratio on 28 November 2015 is provided. From Fig. 10, the color ratio of aerosol is around 1 and the color ratio of cloud which reaches up to 2.5 is much higher than that of aerosol.

 

Fig. 10 The color ratio of cloud and aerosol on 28 November 2015.

Download Full Size | PDF

3.2.3. Cloud base height

The cloud heights measured by WACAL and ceilometer (VAISALA CL31) are compared. Figure 11 shows the comparison of CBH measured by WACAL and ceilometer (The CBH-F,S and T stand for the first, second and third cloud layer base height respectively). The data were measured in 26 December 2013. From this figure, the standard deviation and correlation coefficient are 0.076 km and 86% respectively. Considering the difference of emission wavelengths of WACAL (532 nm) and VAISALA CL31 ceilometer (910 nm), the deviation between these two systems is acceptable.

 

Fig. 11 Comparison of cloud base height measured by WACAL and ceilometer. (a) is the temporal development of the cloud base height and (b) is the scatter diagram based on the measurement of WACAL and ceilometer.

Download Full Size | PDF

However, because of the higher laser power and receiver capability, the detection capability of the WACAL is superior to the ceilometer. Consequently, WACAL is able to detect high altitude cirrus or optical thin but ceilometer cannot. A case study will be provided in Fig. 12.

 

Fig. 12 Contrast of cloud base height by WACAL and ceilometers.

Download Full Size | PDF

According to Fig. 12, the clouds before 19:41 LST are detected by WACAL but however, not detected by the ceilometer. This phenomenon may result from the fact that the clouds located at high altitude and the optical depth were small.

3.2.4. Case study of continuous observation experiment

For purpose of verifying the capability and stability of WACAL, one continuous observation experiment result on 13 June 2014 in Qingdao is provided. The measurement time was between 19:00 and 21:00 LST (LST = UTC + 8 h).

As shown in the Fig. 13 above, in the troposphere between 8 km and 13 km height AGL, two cloud layers are observed. The linear volume depolarization ratio of clouds located at height AGL of 8 km to 10 km is 31.2% ± 2.6% while the backscatter coefficient of the clouds is 0.083 ± 0.036 km⁻¹sr⁻¹. It is found that the value of backscatter coefficient is greater than expected. The clouds at such high altitude might be classified as cirrus or cirro-cumulus and the backscatter coefficient is acceptable according to the study of Liu [56]. Similarly, the ${\delta }_{cal}^v$ and ${\beta }_a$ of the clouds located at height AGL of 11km to 13 km are 40.4% ± 3.5% and 0.029 ± 0.018 km⁻¹sr⁻¹ respectively.

 

Fig. 13 Temporal development of the (a) depolarization ratio, (b) backscatter coefficient and (d) cloud base height between 19:00 and 21:00 LST, 31 May, 2014. And (c) the vertical profile of water vapor mixing ratio.

Download Full Size | PDF

In the lower troposphere, a lofted aerosol layer suspended in the range of 1.3 km and 1.7 km can be detected. The ${\delta }_{cal}^v$ and ${\beta }_a$ are 7.5% ± 0.7% and 0.0071 ± 0.0012 km⁻¹sr⁻¹, respectively. Moreover, the Raman backscatter light of water vapor is collected for 1 h from 20:00 LST to 21:00 UTC and the water vapor mixing ratio is calculated. As Fig. 13(c) shows, water vapor mixing ratio in the aerosol layer reaches up to 9.36 g$·$kg⁻¹. Combining with the optical properties of depolarization ratio and backscatter coefficient, the layer might be classified as fog and the residual layer of the atmospheric boundary layer. It is found that the layer is stable with the development of time. According to the backscatter coefficient, the height of stable boundary layer decreasing with the development of time. In Fig. 13(d), the cloud base heights are measured during the measurement time period, where CBH-F, CBH-S and CBH-T corresponding to the first, second and third cloud base height, respectively. From Fig. 13(d), the clouds at this time period develop fast and the cloud base heights are unstable.