Correction method for temperature measurements inside clouds using rotational Raman lidar

Qimeng Li; Qimeng Li; Huige Di; Huige Di; Ning Chen; Xiao Cheng; Jiaying Yang; Yan Guo; Dengxin Hua; Dengxin Hua

doi:10.1364/OE.507673

1. Introduction

Clouds play a pivotal role in weather observation and forecasting and are key to studying climate change. Moreover, they play a central role in the water cycle and regulate the distribution of global water resources. In particularly, temperature variations within clouds affect their growth and dissipation processes and serve as the foundation for conducting research on phase changes within clouds and precipitation formation [1–6]. However, high-precision detection of temperature in clouds is a challenging task. Currently, the primary methods for atmospheric temperature measurements include infrared radiation techniques, global positioning system (GPS) measurements, radiosondes, microwave radiometers, and lidar.

Instruments employing infrared radiation measurements can obtain a continuous temperature distribution [7,8]. However, the temperature detection process within the cloud layers is influenced by boundary cloud droplets, often referred to as the cloud surface brightness temperature. Consequently, these instruments are generally unable to capture the temperature distribution within cloud layers [9–11], and their vertical resolution is relatively limited owing to the detection mechanism. GPS radio occultation utilizes a wireless radio signal transmission process, influenced by atmospheric refractivity, to probe the Earth’s atmosphere. A major advantage of this technology is that it yields vertical atmospheric profile data. However, the vertical resolution of GPS is constrained by the propagation path, typically reaching only several hundred meters within the troposphere. This limitation poses a challenge in obtaining high spatiotemporal resolution temperature data within clouds, thereby hindering fine-scale research and analysis of cloud thermodynamics and dynamics [12–14]. Currently, meteorological agencies primarily employ radiosondes to obtain vertical atmospheric temperature profiles [15,16]. Although in situ measurements can offer higher precision and stability, achieving continuous and real-time directional detection of specific targets remains challenging. Furthermore, microwave radiometers can retrieve temperature profiles from the Earth’s surface up to 10 km; however, their spatial resolution is insufficient for recognizing the details of temperature variations, such as inversion phenomena.

Lidar, an active remote sensing technology, offers high spatial and temporal resolutions for detecting atmospheric temperature profiles. Temperature measurement with lidar can be categorized into relative and absolute measurement mechanisms, with the latter requiring more advanced spectral separation techniques and posing greater challenges in terms of system implementation [17,18]. Wuhan University has designed a spectral system for extracting a single Raman line in the 532 nm band, eliminating the influence of oxygen molecule rotational Raman (RR) spectral lines on the second-order polynomial calibration function during system calibration [19] and achieving absolute measurement of atmospheric temperature. Relative measurements mainly encompass pure RR temperature measurement techniques and vibration-rotation Raman temperature measurement techniques. However, achieving a higher signal-to-noise ratio (SNR) and detection sensitivity is challenging because of the weaker vibration Raman signals and lower correlation with temperature [20–22]. The dual-channel pure RR technique is the primary approach for atmospheric temperature measurements using lidar. Cooney first proposed the principle of RR temperature measurement in 1972, utilizing the dependence of the RR scattering spectra of nitrogen and oxygen molecules on atmospheric temperature to measure the temperature [23]. The detection principle and system structure of RR temperature measurement lidar are relatively simple, leading to its rapid development. Typically, the intensity of RR scattering is at least 3–4 orders of magnitude lower than that of elastic scattering; therefore, achieving RR spectroscopy separation technology with a high elastic scattering suppression ratio (ESSR) is crucial. Currently, the mainstream choice for system construction is to use narrowband interference filters and diffraction gratings as the core splitters, which have been experimentally verified under clear-sky conditions [24–35].

Clouds comprise tiny water droplets, ice crystals, or a mixture of both, and exhibit strong elastic scattering that directly impacts the acquisition of Raman scattering signals. The primary research approach for Raman temperature measurements within clouds involves utilizing a spectrometric system with a high ESSR to effectively suppress elastic scattering signals. According to existing literature and preliminary simulations, achieving a suppression of at least 10⁸ is deemed necessary [21]. Behrendt and Reichardt developed a dual RR temperature measurement system based on a laser source with 532 nm, determining that an ESSR greater than 70 dB is suitable for temperature measurement under conditions of a few thin clouds [36]. Moreover, Reichardt et al. constructed an RR spectroscopy system based on edge and narrowband interference filters, achieving cloud temperature measurements with a backscatter ratio of less than 60 [37]. However, the system structure is complex and may be susceptible to fluorescence. When the ESSR is sufficiently high, the RR lidar can accurately measure the tropospheric temperature, including the cases when aerosols and optically-thin clouds are present. However, the ESSR of several RR lidar spectroscopic systems fails to satisfy this requirement. Particularly, when the laser wavelength is 354.7 nm, the center wavelength (CWL) of the high and low quantum-number RR channels (CH-H and CH-L) corresponding to the anti-stokes branch has a small spectral spacing compared with elastic scattering, with the spectral spacing of the CH-L only ranging from 0.6 to 0.9 nm, which hinders effective suppression of strong elastic scattering signals. Therefore, researchers have proposed temperature correction methods based on elastic scattering echo signals that can achieve temperature correction in clouds to a certain extent. However, these methods require certain assumptions, involve complex correction process, and the correction coefficients are highly sensitive to assumed conditions, which can lead to overcorrection or undercorrection [38]. Nevertheless, this study presents a research approach for temperature detection in clouds.

To achieve temperature measurements in clouds using RR lidar, we propose a temperature correction technique based on the backscatter ratio. The main premise of this method relies on the correlation between the ESC and the backscatter ratio. This involves determining the magnitude of the ESC in the RR channels from the measured backscatter ratio and subsequently correcting the temperature data within the clouds. The effectiveness of this correction method was analyzed through numerical simulations. Moreover, experimental observations were conducted at the Jinghe National Basic Meteorological Observing Station based on the temperature and aerosol detection lidar of the Xi'an University of Technology (XUT), and the feasibility of the temperature correction technique in clouds was verified using the measured data.

2. Temperature and aerosol detection lidar system

Figure 1 presents the schematic layout of the temperature and aerosol detection lidar system. The lidar transmitter employs an injection-seeded Nd: YAG laser as the light source. Based on the third-harmonic generation (THG) technique, it achieves a working wavelength of 354.7 nm, a repetition frequency of 20 Hz, an energy per pulse of approximately 240 mJ, and a pulse duration of 9 ns. A Cassegrain telescope is employed as the optical receiver, and narrowband interference filters are utilized as core filter devices to finely separate the Raman spectra.

Fig. 1. Schematic diagram of the Raman lidar system.

Download Full Size | PDF

The system consists of four detection channels: nitrogen vibrational Raman channel (CH-V), CH-L, CH-H and Mie-Rayleigh scattering (M-Ray) channel. The first edge filter (EF) separates the incident signal light into two parts: one part is transmitted to the narrowband interference filter (IF1), and finally received by PMT-1, and the other part is reflected into the beam splitter (BS). BS reflects most of the signal light onto the CH-H, and the signal light reflected by IF2a is incident onto the CH-L. A small amount of signal light transmitted through the BS is incident onto the M-Ray channel. Two RR channels are employed for dual-RR temperature measurements, whereas the other two signals are used to detect aerosols. Each channel includes corresponding filters, a focusing lens, and a photomultiplier tube (PMT). Data are acquired using a data acquisition card and subsequently transferred to a computer for processing. The ESSRs of CH-H, CH-L, and CH-V are approximately 65, 50 and 75 dB, respectively. Table 1 summarizes the main system parameters of the overall temperature and aerosol detection lidar system.

Table 1. Main Simulation Parameters^a

View Table

3. Measurement method

The high-quantum-number (high-J) and low-quantum-number (low-J) RR lines exhibit a simple temperature correlation and follow a Boltzmann distribution, which is the theoretical basis for temperature measurement [26]. The signal intensities of the high-J RR lines increase with increasing temperature, whereas those of the low-J RR lines exhibit the opposite behavior. Nitrogen vibrational Raman scattering signals are generally considered to exhibit a weak temperature correlation and can be used to retrieve parameters such as the extinction coefficient and backscatter ratio, combined with elastic scattering. The signal intensities corresponding to the RR, CH-V and M-Ray channels can be expressed as follows:

(1)$${X_r}(T,z,{\lambda _r}) = \frac{{C(z)}}{{{z^2}}}{\eta _r}({\lambda _r}){\beta _r}(T,z,{\lambda _r})\exp \left\{ { - \int_0^z {[{\alpha (z^{\prime},{\lambda_0}) + \alpha (z^{\prime},{\lambda_r})} ]} \textrm{d}z^{\prime}} \right\}, $$

(2)$${X_v}(T,z,{\lambda _v}) = \frac{{C(z)}}{{{z^2}}}{\eta _v}({\lambda _v}){\beta _v}(T,z,{\lambda _v})\exp \left\{ { - \int_0^z {[{\alpha (z^{\prime},{\lambda_0}) + \alpha (z^{\prime},{\lambda_v})} ]} \textrm{d}z^{\prime}} \right\}, $$

(3)$${X_e}(z,{\lambda _0}) = \frac{{C(z)}}{{{z^2}}}{\eta _e}({\lambda _0})({\beta _m}(z,{\lambda _0}) + {\beta _a}(z,{\lambda _0}))\exp \left\{ { - 2\int_0^z {[{\alpha (z^{\prime},{\lambda_0})} ]} \textrm{d}z^{\prime}} \right\}, $$

(4)$$C(z) = {X_0}\frac{{c\tau }}{{{z^2}}}{A_T}O(z), $$

(5)$${\beta _r}(T,z,{\lambda _r}) = N(z)\sigma (T,z,{\lambda _r}), $$

where X₀ is the mean laser power per pulse, c and τ are the velocity of light and pulse duration, respectively. A_T and O(z) represent the surface area of the receiving telescope and the overlap function of the laser beam and the receiver field-of-view. η_r, η_v and η_e are system efficiency for each channel and include factors such as the reflectivity of the telescope, the transmission of conditioning optics, the transmission of any filters, and the quantum efficiency of the detector. β_r and β_v represent the backscatter coefficients of the RR channels and CH-V, respectively, β_m and β_a represent those of molecules and aerosols corresponding to elastic scattering, respectively. T and z correspond to the atmospheric temperature and altitude, respectively, and λ₀ is the excitation wavelength, λ_r and λ_v are the CWL of the RR channels and CH-V, respectively. The exponential factor gives the two-way atmospheric transmission, and α denotes the extinction coefficient at different wavelengths, which is used to calculate atmospheric transmittance. N(z) is the nitrogen molecule number density, and σ is the differential backscatter cross-section. A ratio operation was performed on the CH-H and CH-L signals as follows:

(6)$$R(T,z) = \frac{{X{}_{rh}(T,z,{\lambda _{rh}})}}{{{X_{rl}}(T,z,{\lambda _{rl}})}} = \frac{{\sum\limits_{{J_h}} {[{\eta ({\lambda_{rh}}) \cdot {\sigma_h}(T,z,{\lambda_{rh}})} ]} }}{{\sum\limits_{{J_l}} {[{\eta ({\lambda_{rl}}) \cdot {\sigma_l}(T,z,{\lambda_{rh}})} ]} }}\exp \left\{ { - \int_0^z {[{\alpha (z^{\prime},{\lambda_{rh}}) - \alpha (z^{\prime},{\lambda_{rl}})} ]} \textrm{d}z^{\prime}} \right\}, $$

where J denotes the rotational angular-momentum quantum number, and the subscripts h and l correspond to CH-H and CH-L.

System calibration for temperature retrieval can be achieved based on the atmospheric temperature profile of the radiosonde under the same spatiotemporal conditions, and the system calibration function is usually formulated as follows [36]:

(7)$$R(T,z) = \exp (\frac{\textrm{A}}{{{T^2}(z)}} + \frac{\textrm{B}}{{T(z)}} + \textrm{C}), $$

where A, B and C are the system constants for temperature retrieval to be calibrated.

When the laser beam encounters clouds, Behrendt et al. believe that CH-L closer in wavelength to the elastic scattering signal has a higher probability of crosstalk. Therefore, they made corrections for ESC in CH-L [39]. However, under strong elastic scattering conditions, both RR channels may experience ESC. Therefore, the ratio within the cloud layer can be expressed as [38]:

(8)$${R_e}(T,z) = \frac{{X{}_{rh}(T,z,{\lambda _{rh}}) + {k_h} \cdot X{}_e(z,{\lambda _0})}}{{{X_{rl}}(T,z,{\lambda _{rl}}) + {k_l} \cdot {X_e}(z,{\lambda _0})}}, $$

where k_h and k_l denote the amount of leakage in CH-H and CH-L, respectively, which are referred to as crosstalk coefficients. Based on Equations (1)–(3), its expression can be derived as η_rh(λ₀)/η_e(λ₀) and η_rl(λ₀)/η_e(λ₀), respectively. The backscatter ratio can be calculated using the following relationship, where the extinction coefficient is used for transmittance compensation during the backscatter ratio derivation [40]. The wavelength difference between nitrogen vibration Raman scattering and elastic scattering is significant. Therefore, when using the method proposed by Whiteman to approximate the backscatter ratio, it is necessary to calculate the difference of atmospheric transmittance in CH-V and M-Ray channels based on the extinction coefficient. The extinction coefficient can be solved using the Raman method [41].

(9)$$Br(z,{\lambda _0}) = \frac{{{\beta _a}(z,{\lambda _0}) + {\beta _m}(z,{\lambda _0})}}{{{\beta _m}(z,{\lambda _0})}}, $$

(10)$${\beta _a}(z,{\lambda _0}) = \frac{{{\eta _v}({\lambda _v})}}{{{\eta _e}({\lambda _0})}}\frac{{{X_e}(z,{\lambda _0})}}{{{X_v}(T,z,{\lambda _v})}}{\beta _v}(T,z,{\lambda _v})\exp \left\{ {\int_0^z {[{\alpha (z^{\prime},{\lambda_0}) - \alpha (z^{\prime},{\lambda_v})} ]} \textrm{d}z^{\prime}} \right\} - {\beta _m}(z,{\lambda _0}), $$

(11)$${\alpha _a}(z,{\lambda _0}) = \frac{{\frac{\textrm{d}}{{\textrm{d}z}}[\ln (\frac{{N(z)}}{{{X_v}(z,{\lambda _v}) \cdot {z^2}}})] - {\alpha _m}(z,{\lambda _0}) - {\alpha _m}(z,{\lambda _v})}}{{1 + {{(\frac{{{\lambda _0}}}{{{\lambda _v}}})}^K}}}$$

where K is the Angstrom exponent, typically taken as 1 [41].

The inherent deviation of the second-order polynomial calibration function can be ignored. ESC is the primary factor affecting temperature inversion in clouds, and the extent of crosstalk depends on the ESSR of the RR channel.

When the backscatter ratio increases, ESC is present in both CH-H and CH-L, causing both the molecular and denominator terms in the measured rotational Raman ratio of clouds (R_e) to be affected by leakage. R₀ is defined as the theoretical value of rotational Raman ratio without ESC, which can be calculated using the system constants and radiosonde data. The elastic scattering crosstalk ratio (ESCR) G(z) is calculated by the ratio operation of R₀ and R_e, and it denotes the amount of ESC in the rotational Raman ratio. Although the ESC within CH-H or CH-L and backscatter ratio are linearly correlated, the relationship between R_e and ESC is nonlinear. Therefore, when both CH-H and CH-L exhibit varying degrees of ESC, quadratic or other nonlinear functions should be performed for nonlinear fitting. Based on previous data statistics and analysis, the relationship between G(z) and the backscatter ratio of clouds seems to follow a quadratic function distribution. Therefore, we use quadratic functions as fitting functions to achieve the calibration process.

(12)$$G(z) = \frac{{{R_0}(T,z)}}{{{R_e}(T,z)}} = \textrm{a} \cdot Br{(z)^2} + \textrm{b} \cdot Br(z) + \textrm{c}, $$

where Br(z) is the backscatter ratio of clouds, and a, b and c are the calibration coefficients for rotational Raman ratio correction to be calibrated. The calibration coefficients are obtained by fitting the quadratic function using the least squares method. Notably, during the function fitting process, the low backscatter ratio region near the cloud boundary will be unable to accurately distinguish between the effects of system noise and ESC. Therefore, it is necessary to set a backscatter ratio threshold and discard data from low SNR regions to filter out the data with high correlation for the function fitting. The setting of the backscatter ratio threshold depends on the detection performance of the system. Theoretically, when the system parameters are stable, the correction constants do not change. Thus, the correction method enables the correction of the measured rotational Raman ratio within clouds, thereby allowing the retrieval of vertical atmospheric temperature profiles in cloud layers based on the principle of RR relative temperature measurements.

Figure 2 illustrates the specific steps of temperature retrieval in clouds. The M-Ray signal, nitrogen vibrational Raman signal, and two RR signals are obtained through the temperature and aerosol detection lidar. Firstly, based on the above equation, the backscatter ratio of clouds and rotational Raman ratios (R and R_e) are calculated. The system calibration process of dual rotational Raman temperature measurement is achieved based on radiosonde data and R, and the system constants for temperature retrieval are obtained. Then, combined with the system constants and the temperature profile of radiosonde, the theoretical value (R₀) of the rotational Raman ratio corresponding to the temperature of clouds is calculated, and G(z) is obtained. Furthermore, based on the backscatter ratio of clouds and G(z), a quadratic function fitting relationship is constructed to calibrate the correction function for rotational Raman ratio correction, and the calibration coefficients are obtained. Combined the calibration coefficients, R_e and the backscatter ratio of clouds, the rotational Raman ratio is corrected. Finally, based on the corrected ratio and the system constants of temperature retrieval, the temperature profile inside the clouds can be calculated.

Fig. 2. Schematic diagram of the temperature correction in clouds.

Download Full Size | PDF

4. Simulation and analysis

We conducted a simulation analysis based on the aforementioned theoretical relationships and system parameters. The simulation process primarily focused on the cloud layers, and the influence of the bottom layer aerosols was not considered. We simulated cloud layers with a cloud base height of 3 km by setting the backscatter ratio condition (≤ 120). Gaussian white noise was injected to simulate the trend of signal variation as the SNR decreased. The atmospheric echo signal obtained from the system simulation is described as follows:

Figure 3(a) plots the backscatter ratio conditions. In Fig. 3(b), the red solid curve indicates the range-square-corrected signal (RSCS) corresponding to the M-Ray signal. The purple dashed curve represents the CH-V signal, whereas the blue and black solid curves correspond to the CH-H and CH-L signals with ESC, respectively. The blue and black dashed curves represent the signals without ESC. In particular, the different ESSR conditions result in varying degrees of ESC in the CH-H and CH-L, with signal intensities in the clouds region being higher than the values under the no-crosstalk conditions.

Fig. 3. Simulated signals. (a) Backscatter ratio; (b) RSCS of atmospheric echo signals.

Download Full Size | PDF

Calculations were performed to obtain the ratios of CH-H to CH-L (R-HL), CH-H to CH-V (R-HV) and CH-L to CH-V (R-LV), respectively, as shown in Fig. 4. The blue dashed curve represents the ratios under the ESC conditions, whereas the black solid curve represents the ratios without crosstalk. Notably, transmittance correction is necessary for R-HV and R-LV.

Fig. 4. Ratios. (a) R-HL; (b) R-HV; (c) R-LV.

Download Full Size | PDF

Typically, real atmospheric environments do not exhibit absolute clean atmospheric conditions. During the simulation process, a backscatter ratio greater than 2 within the cloud layers was taken for correction, which namely the threshold condition for initiating the correction.

In Fig. 5(a), the black solid curve represents G(z) and the blue dashed curve represents the backscatter ratio. When the backscatter ratio increases, resulting in the presence of ESC in both CH-L and CH-H, the backscatter ratio of clouds and G(z) exhibits a nonlinear relationship. Due to the decrease in SNR, the G(z) above 4.6 km experienced significant fluctuations. Figure 5(b) presents the quadratic function fitting relationship between G(z) and the backscatter ratio. The red box discrete points represent the deviations caused by a low SNR, which were excluded during the fitting process. It is evident that these two variables exhibit a strong correlation, especially within the range wherein the backscatter ratio is less than 45, which can be approximated as a linear relationship. Because the strength of the elastic scattering signal directly determines the magnitude of the backscatter ratio, weaker elastic scattering is insufficient to cause crosstalk in CH-H. When the backscatter ratio decreases, the CH-H can effectively suppress the elastic scattering, and the temperature deviation can be approximated as the effect of the ESC in the CH-L, which is consistent with the previous research conclusions of Behrendt et al. (2002).

Fig. 5. Fitting results. (a) ESCR and backscatter ratio; (b) fitting relationship of Eq. (12).

Download Full Size | PDF

Based on this quadratic function fitting relationship, combined with R_e, the temperature correction constants and the backscatter ratio, the true ratio corresponding to the temperature within clouds can be obtained, thereby achieving temperature measurements in clouds, as shown in Fig. 6. The black solid and blue dotted curves in Fig. 6 (a) correspond to the specific temperature model and the temperature retrieval results with the ESC, respectively. The red dashed curve represents the corrected temperature results in the clouds, and Fig. 6(b) plots the temperature deviations. The simulation results indicate that the maximum uncorrected temperature deviation within the cloud layers is approximately 23 K. However, the corrected results in the interval of 3.2 to 4.5 km are less than 0.2 K, with a significant deviation in the low SNR interval, which is consistent with theoretical findings. This method can effectively correct the deviations caused by ESC and achieve the temperature correction within the clouds.

Fig. 6. Simulated results. (a) Temperature; (b) deviations.

Download Full Size | PDF

5. Experimental observations and results

Experimental observations were performed based on the temperature and aerosol detection lidar of XUT at the Jinghe National Basic Meteorological Observing Station (34.43°N, 108.97°E). This site provides radiosonde data twice a day, in the morning and evening, and is used for system calibration and data validation.

5.1 System calibration

The system calibration consists of two steps. The first step involves calibration of the system constants for atmospheric temperature retrieval. In this study, the system calibration is performed using the cloud-free lidar data and the temperature profile of the radiosonde during the same period [24–37]. The second step involves the calibration of the temperature correction functions within the clouds. The measured lidar data and the radiosonde data on the night of September 27 and 28, 2022, were selected as the source data for the system calibration of temperature retrieval and the calibration of correction functions in clouds. And the calibration process assumes a backscatter ratio greater than 2 as the threshold condition.

Figure 7(a) and (d) illustrate the RSCSs of the measured atmospheric echo signals, the red thin solid curve is the M-Ray channel signal, the purple dashed curve is the nitrogen vibrational Raman signal, the black thick solid and the blue dotted curves are the CH-L and CH-H signals, respectively. In Fig. 7(b) and (e), the black solid curve represents R_e, and the blue dashed curve represents R₀ based on the calculation of the radiosonde temperature. Figure 7(c) and (f) present the black solid curves as G(z), and the blue dashed curve represents the backscatter ratio.

Fig. 7. Measured results on the 27^th and 28^th. (a) and (d) RSCS; (b) and (e) rotational Raman ratio; (c) and (f) ESCR and backscatter ratio.

Download Full Size | PDF

According to the elastic scattering signal, a cloud layer with a thickness of approximately 1.5 km begins to appear at 5.5 km, with a maximum backscatter ratio of 115, and another cloud layer with a thickness of approximately 0.6 km begins to appear at 3.6 km, with a maximum backscatter ratio of 40. In the strong elastic scattering region, all the inelastic scattering channels exhibit significant signal attenuation. In addition, the CH-L signal experiences significant distortion, and the distortion trend is consistent with the M-Ray signal. That is, the ESC occurs in this region.

Figure 8(a) illustrates the functional fitting relationship, in which the blue circle data represents the original data to be fitted on the 27^th, the red box corresponds to the original data to be fitted on the 28^th, and the black solid curve represents the fitting result. G(z) and the backscatter ratio of clouds exhibited a significant correlation, and the calibration coefficients were obtained with a correlation coefficient of 0.9956. In Fig. 8(b) and(d), the black dotted and dashed curves represent the temperature retrieval results without correction and radiosonde temperature, respectively. The red solid curve represents the corrected temperature profile within the cloud layers, and Fig. 8(c) and (e) depict the deviations compared with the radiosonde temperature. Notably, the maximum uncorrected temperature deviation inside the clouds is approximately 75 K. However, in the correction results, over 85% of the deviations are less than 1.5 K. A local deviation of nearly 3 K occurs near 6.5 km in Fig. 8(c), which may be due to the inconsistency between the ESC of the RR channels and the measured backscatter ratio. This inconsistency is mainly constrained by SNR and system stability. Therefore, under the condition that the lidar system has high stability, based on this correction relationship, effective correction of the atmospheric temperature profile inside the strong elastic scattering region of the clouds can be achieved.

Fig. 8. Fitting relationship and temperature results on the 27^th and 28^th. (a) Fitting relationship of Eq. (12); (b) and (e) temperature results; (c) and (f) deviations. UCT, uncorrected temperature; RT, radiosonde temperature; CT, corrected temperature.

Download Full Size | PDF

5.2 Corrected results

Based on the aforementioned correction functions, a temperature correction was applied to the measured data under cloud conditions for September 29, 2022. Figures 9(a) and (b) represent the atmospheric scattering echo signals and backscatter ratio, respectively. The subject of the illustration in Fig. 9 is identical to that in Fig. 7 and Fig. 8. A cloud layer with a thickness of approximately 3.5 km appears at 5.6 km, and with a maximum backscatter ratio of approximately 60. Without correction, the maximum temperature deviation inside the clouds is approximately 70 K, and the deviations below 7.7 km do not exceed 1 K. Owing to the signal attenuation under strong elastic scattering conditions, the SNR of the system decreases, resulting in significant temperature deviations above 7.7 km. Compared with radiosonde data, the feasibility of the temperature correction method within clouds was verified.

Fig. 9. Data and retrieval results. (a) RSCS; (b) backscatter ratio; (c) temperature; (d)deviation.

Download Full Size | PDF

Figure 10 represents the continuous observation results for September 29, 2022. The RSCS of the M-Ray channel is shown in Fig. 10(a). Approximately 16 min of the data were missing for objective reasons. Starting from 19:00, continuous high-altitude clouds appeared above 5.5 km, and the clouds at 7 km exhibited a relatively strong signal attenuation between 0:00 and 2:00. Figure 10(b) displays the corrected continuous temperature distribution. The black dashed curve with a star pattern corresponds to the cloud base height. The black solid curve represents the corrected vertical atmospheric temperature profile, and the red dashed curve represents the uncorrected temperature profile within the clouds.

Fig. 10. Continuous observation results. (a) RSCS of M-Ray signal and the cloud base height; (b) temperature results and the cloud base height.

Download Full Size | PDF

Significant temperature deviations occur in the altitude range corresponding to the cloud layers and are significantly correlated with the ESC. Based on the correction results, it can be concluded that this temperature correction method can effectively eliminate the impact of ESC on Raman relative temperature measurements, allowing for the retrieval of vertical atmospheric temperature profiles inside the clouds. Notably, owing to the strong signal attenuation effect of clouds, the lidar signal cannot achieve complete penetration through thicker clouds.

6. Discussion and conclusions

To achieve temperature measurements in clouds based on RR lidar, a cloud temperature correction technique for strong elastic scattering conditions was proposed by utilizing the correlation between the backscatter ratio of clouds and ESC. Through data comparison and analysis, we discovered that the ratio relationships of the theoretical and measured Raman ratios corresponding to the temperature within the clouds have a high correlation with the backscatter ratio, which conforms to a quadratic function relationship. Based on this finding, we conducted a theoretical simulation analysis based on the existing system parameters, revealing that the aforementioned content was fully validated within the clouds. Temperature correction within the clouds was achieved based on the calibrated correction function. Moreover, it was observed that this correction process was not suitable for the low SNR conditions, which aligned with theoretical expectations. This challenge may stem from the tiny RR signals in the low SNR region, which makes it difficult to eliminate ESC without distortion. For experimental validation, observations were performed based on the temperature and aerosol Raman lidar of XUT. First, the calibration of temperature correction function within the clouds and the system constant calibration for atmospheric temperature retrieval were conducted on the existing system. The preset conditions for initiating the correction were established, and the calibration relationship of the measured data inside the clouds met the expected results. Additionally, the corrected cloud temperature exhibited a high degree of agreement with the simultaneously collected radiosonde data. Although on September 29^th, temperature deviations after correction began to exhibit significant discrepancies (>3 K) in the upper part of the cloud layer, primarily due to low SNR resulting from optical attenuation. The temperature deviations within the correction region were basically maintained within 1.5 K. Furthermore, long-term continuous observation results underscored the effectiveness of this correction technique in achieving vertical atmospheric temperature measurements within cloud layers, contributing valuable insights into the study of cloud microphysics and the hydrological cycle within the Earth's atmosphere.

The effective measurement height of the Raman temperature measurement lidar is constrained by atmospheric transmission conditions, causing to a decrease in SNR, which poses the greatest challenge to cloud temperature correction. Notably, the cloud temperature correction requires knowledge of the backscatter ratio and the measured Raman ratio within the clouds, which will involve the reception of optical signals from four channels. And the inter-channel consistency is a prerequisite to ensure the implementation of temperature correction within the clouds. Moreover, the cloud temperature correction process requires the signals from each channel to be subjected to the same preprocessing conditions. The theoretical basis for temperature correction relies on the correlation between the ESC and the backscatter ratio. The ESSR of the RR channel is a key parameter determining the degree of crosstalk, resulting in slight variations in the temperature correction constant in the cloud owing to differences in parameters, such as the lidar spectroscopic system. In addition, the detection performance of the system is also affected by parameters such as laser energy, axiality of the optical transceiver system, and photoelectric conversion efficiency. Changes in these parameters directly impact the stability of the system, thereby influencing the temperature correction constant within the clouds. This highlights the necessity of a lidar system with high stability for the successful implementation of this correction technique. Therefore, it may be necessary to recalibrate the correction function for different system parameters.

Funding

National Natural Science Foundation of China (2021ZDLSF06-07, 41627807, 42130612).

Acknowledgment

The authors gratefully acknowledge the support of the Jinghe National Basic Meteorological Observing Station. Writing assistance was provided by Huige Di and Dengxin Hua.

Disclosures

The authors declare no conflicts of interest.

Data availability

The 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. V. Wulfmeyer, R. M. Hardesty, D. D. Turner, et al., “A review of the remote sensing of lower tropospheric thermodynamic profiles and its indispensable role for the understanding and the simulation of water and energy cycles,” Rev. Geophys. 53(3), 819–895 (2015). [CrossRef]

2. A. Baron, P. Chazette, and J. Totems, “Extreme temperature events monitored by Raman lidar: Consistency and complementarity with spaceborne observations and modelling,” Meteorol. Appl. 29(5), e2062 (2022). [CrossRef]

3. G. Martucci, F. Navas-Guzmán, L. Renaud, et al., “Validation of pure rotational Raman temperature data from the Raman Lidar for Meteorological Observations (RALMO) at Payerne,” Atmos. Meas. Tech. 14(2), 1333–1353 (2021). [CrossRef]

4. J. Sedlar, M. D. Shupe, and M. Tjernström, “On the Relationship between Thermodynamic Structure and Clouds Top, and Its Climate Significance in the Arctic,” J. Clim. 25(7), 2374–2393 (2012). [CrossRef]

5. L. Sun and Y. Fu, “A new merged dataset for analyzing clouds, precipitation and atmospheric parameters based on ERA5 reanalysis data and the measurements of the Tropical Rainfall Measuring Mission (TRMM) precipitation radar and visible and infrared scanner,” Earth Syst. Sci. Data 13(5), 2293–2306 (2021). [CrossRef]

6. F. Pithan, B. Medeiros, and T. Mauritsen, “Mixed-phase clouds cause climate model biases in Arctic wintertime temperature inversions,” Clim. Dyn. 43(1-2), 289–303 (2014). [CrossRef]

7. W. Cooper, “Effects of coincidence on measurements with a forward scattering spectrometer probe,” J. Atmos. Oceanic Technol. 5(6), 823–832 (1988). [CrossRef]

8. H. H. Aumann and M. T. Chahine, “Infrared multidetector spectrometer for remote sensing of temperature profiles in the presence of clouds,” Appl. Opt. 15(9), 2091–2094 (1976). [CrossRef]

9. R. P. Lawson and R. H. Cormack, “Theoretical design and preliminary tests of two new particle spectrometers for clouds microphysics research,” Atmos. Res. 35(2-4), 315–348 (1995). [CrossRef]

10. T. Y. Nakajima, T. Tsuchiya, H. Ishida, et al., “Clouds detection performance of spaceborne visible-to-infrared multispectral imagers,” Appl. Opt. 50(17), 2601–2616 (2011). [CrossRef]

11. E. Wong, K. D. Hutchison, S. Steve Ou, et al., “Cirrus clouds top temperatures retrieved from radiances in the National Polar-Orbiting Operational Environmental Satellite System—Visible Infrared Imager Radiometer Suite 8.55 and 12.0 µm bandpasses,” Appl. Opt. 46(8), 1316 (2007). [CrossRef]

12. S. Yang and X. Zou, “Temperature Profiles and Lapse Rate Climatology in Altostratus and Nimbostratus Clouds Derived from GPS RO Data,” J. Clim. 26(16), 6000–6014 (2013). [CrossRef]

13. L. Lin, X. Zou, R. Anthes, et al., “COSMIC GPS Radio Occultation Temperature Profiles in Clouds,” Mon. Wea. Rev. 138(4), 1104–1118 (2010). [CrossRef]

14. S. Yang and X. Zou, “Lapse rate characteristics in ice clouds inferred from GPS RO and Cloudsat observations,” Atmos. Res. 197, 105–112 (2017). [CrossRef]

15. D. Wang, J. Guo, A. Chen, et al., “Temperature Inversion and Clouds Over the Arctic Ocean Observed by the 5th Chinese National Arctic Research Expedition,” J. Geophys. Res. Atmos. 125(13), e2019JD032136 (2020). [CrossRef]

16. J. Zhang, “Clouds-top temperature inversion derived from long-term radiosonde measurements at the ARM TWP and NSA sites,” Atmos. Res. 246, 105113 (2020). [CrossRef]

17. P. Piironen and E. W. Eloranta, “Demonstration of a high-spectral-resolution lidar based on an iodine absorption filter,” Opt. Lett. 19(3), 234–236 (1994). [CrossRef]

18. D. Hua, M. Uchida, and T. Kobayashi, “Ultraviolet high-spectral-resolution Rayleigh-Mie lidar with a dual-pass Fabry–Perot etalon for measuring atmospheric temperature profiles of the troposphere,” Opt. Lett. 29(10), 1063–1065 (2004). [CrossRef]

19. M. Weng, F. Yi, F. Liu, et al., “Single-line-extracted pure rotational Raman lidar to measure atmospheric temperature and aerosol profiles,” Opt. Express 26(21), 27555–27571 (2018). [CrossRef]

20. W. S. Heaps, J. Burris, and J. A. French, “Lidar technique for remote measurement of temperature by use of vibrational-rotational Raman spectroscopy,” Appl. Opt. 36(36), 9402–9405 (1997). [CrossRef]

21. Q. Li, H. Di, D. Hua, et al., “Temperature measurement of clouds or haze layers based on Raman rotational and vibrational spectra,” Opt. Express 30(13), 23124–23137 (2022). [CrossRef]

22. 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]

23. J. Cooney, “Measurement of Atmospheric Temperature Profiles by Raman Backscatter,” J. Appl. Meteorol. Climatol. 11(1), 108–112 (1972). [CrossRef]

24. P. Di Girolamo, R. Marchese, D. N. Whiteman, et al., “Rotational Raman Lidar measurements of atmospheric temperature in the UV,” Geophys. Res. Lett. 31(1), L01106 (2004). [CrossRef]

25. D. Wu, Z. Wang, P. Wechsler, et al., “Airborne compact rotational Raman lidar for temperature measurement,” Opt. Express 24(18), A1210–1223 (2016). [CrossRef]

26. F. Liu, R. Wang, F. Yi, et al., “Pure rotational Raman lidar for full-day troposphere temperature measurement at Zhongshan Station (69.37°S, 76.37°E), Antarctica,” Opt. Express 29(7), 10059–10076 (2021). [CrossRef]

27. F. Liu, F. Yi, Y. Zhang, et al., “Double-Receiver-Based Pure Rotational Raman LiDAR for Measuring Atmospheric Temperature at Altitudes Between Near Ground and Up To 35 km,” IEEE Trans. Geosci. Remote Sens. 57(12), 10301–10309 (2019). [CrossRef]

28. M. Radlach, A. Behrendt, and V. Wulfmeyer, “Scanning rotational Raman lidar at 355 nm for the measurement of tropospheric temperature fields,” Atmos. Chem. Phys. 8(2), 159–169 (2008). [CrossRef]

29. P. Achtert, M. Khaplanov, F. Khosrawi, et al., “Pure rotational-Raman channels of the Esrange lidar for temperature and particle extinction measurements in the troposphere and lower stratosphere,” Atmos. Meas. Tech. 6(1), 91–98 (2013). [CrossRef]

30. E. Hammann, A. Behrendt, F. Le Mounier, et al., “Temperature profiling of the atmospheric boundary layers with rotational Raman lidar during the HD(CP)2 Observational Prototype Experiment,” Atmos. Chem. Phys. 15(5), 2867–2881 (2015). [CrossRef]

31. J. Totems, P. Chazette, and A. Baron, “Mitigation of bias sources for atmospheric temperature and humidity in the mobile Raman Weather and Aerosol Lidar (WALI),” Atmos. Meas. Tech. 14(12), 7525–7544 (2021). [CrossRef]

32. S. Chen, Z. Qiu, Y. Zhang, et al., “A pure rotational Raman lidar using double-grating monochromator for temperature profile detection,” J. Quant. Spectrosc. Radiat. Transfer 112(2), 304–309 (2011). [CrossRef]

33. J. Jia and F. Yi, “Atmospheric temperature measurements at altitudes of 5–30 km with a double-grating-based pure rotational Raman lidar,” Appl. Opt. 53(24), 5330–5343 (2014). [CrossRef]

34. V. Zuev, V. Gerasimov, V. Pravdin, et al., “Tropospheric temperature measurements with the pure rotational Raman lidar technique using nonlinear calibration functions,” Atmos. Meas. Tech. 10(1), 315–332 (2017). [CrossRef]

35. J. He, S. Chen, Y. Zhang, et al., “A novel calibration method for pure rotational Raman lidar temperature profiling,” J. Geophys. Res. Atmos. 123(19), 10925–10934 (2018). [CrossRef]

36. A. Behrendt and J. Reichardt, “Atmospheric temperature profiling in the presence of clouds with a pure rotational Raman lidar by use of an interference-filter-based polychromator,” Appl. Opt. 39(9), 1372–1378 (2000). [CrossRef]

37. J. Reichardt, U. Wandinger, V. Klein, et al., “RAMSES: German Meteorological Service autonomous Raman lidar for water vapor, temperature, aerosol, and clouds measurements,” Appl. Opt. 51(34), 8111–8131 (2012). [CrossRef]

38. J. Su, M. P. McCormick, Y. Wu, et al., “Cloud temperature measurement using rotational Raman lidar,” J. Quant. Spectrosc. Radiat. Transfer 125, 45–50 (2013). [CrossRef]

39. A. Behrendt, T. Nakamura, M. Onishi, et al., “Combined Raman lidar for the measurement of atmospheric temperature, water vapor, particle extinction coefficient, and particle backscatter coefficient,” Appl. Opt. 41(36), 7657 (2002). [CrossRef]

40. D. Whiteman, “Examination of the traditional raman lidar technique. II. Evaluating the ratios for water vapor and aerosols,” Appl. Opt. 42(15), 2593 (2003). [CrossRef]

41. A. Ansmann, M. Riebesell, U. Wandinger, et al., “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]

Specification	Value
Excitation wavelength	EW: 354.7 nm; EPP: ∼240 mJ; PD: < 9 ns
Half field of view	HFOV: 0.5 mrad; TA: 400 mm
CH-H	IF2a, IF2b	FWHM: 1.0 nm @352.5 nm, ESSR ∼65 dB @354.7 nm
CH-L	IF3a, IF3b	FWHM: 0.5 nm @353.9 nm, ESSR ∼50 dB @354.7 nm
CH-V	IF1	FWHM: 0.5 nm @386.7 nm, ESSR ∼75 dB @354.7 nm
M-Ray	IF4	FWHM: 0.5 nm @354.7 nm
Edge filter (EF)	EF	Tr > 93% @381∼950 nm, Re > 98% @327∼371 nm
PMT	Quantum efficiency: 0.26, Hamamatsu
Integration time	2 min

Correction method for temperature measurements inside clouds using rotational Raman lidar

Abstract

1. Introduction

2. Temperature and aerosol detection lidar system

3. Measurement method

4. Simulation and analysis

5. Experimental observations and results

5.1 System calibration

5.2 Corrected results

6. Discussion and conclusions

Funding

Acknowledgment

Disclosures

Data availability

References

Data availability

Cited By

Figures (10)

Tables (1)

Equations (12)

Optics Express