A sodium fluorescence Doppler laser radar data quality control method and system

Through the photon counting data quality control method, including high correction, signal-to-noise ratio detection and filter correction, the error and utilization problems in the quality control of sodium fluorescence Doppler lidar are solved, and the data automation and intelligence are realized Processing improves the accuracy and reliability of data.

CN115390092BActive Publication Date: 2024-11-12NANJING XIAOZHUANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210808612.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-11
Publication Date
2024-11-12
Estimated Expiration
2042-07-11

AI Technical Summary

Technical Problem

The sodium fluorescent Doppler lidar hardware system is complex, with many observation data channels, and the parameter inversion is difficult. The inversion result depends on personal experience, resulting in large errors and insufficient data utilization, and lack of effective data quality control methods to improve data processing. Level of automation and intelligence.

Method used

The photon counting data quality control method is adopted, including high correction, sodium layer signal-to-noise ratio detection, laser frequency stability detection, filter transmittance correction, temperature ratio detection and outlier removal, so as to achieve sodium fluorescence Doppler Automatic diagnosis and batch processing of lidar data.

Benefits of technology

It improves the accuracy and reliability of lidar data, reduces human resources costs, improves data utilization, reduces misjudgment and misjudgment problems, and enhances the automation and intelligent processing capabilities of data products.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115390092B_ABST
    Figure CN115390092B_ABST
Patent Text Reader

Abstract

The invention relates to a sodium fluorescence Doppler laser radar data quality control method and system. The method comprises photon counting data quality control and data product quality control. Through the processes of original echo signal height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination, photon counting linearity detection and correction, filter transmittance correction, temperature ratio detection and abnormal value elimination, wind speed ratio detection and abnormal value elimination, zenith direction wind speed profile detection and abnormal value elimination, temperature and wind speed detection error detection and unreliable value elimination, etc., automatic diagnosis and batch processing of sodium fluorescence Doppler laser radar data are realized, and the problems of much manual intervention in the sodium fluorescence Doppler laser radar parameter inversion process, dependence of the inversion result quality on personal experience, large error in detection result, or inability to fully utilize a large amount of detection data are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a laser radar data quality control technology, and in particular to a data quality control method and system for a sodium fluorescence Doppler laser radar. Background Art

[0002] Sodium fluorescence Doppler lidar is one of the important active optical remote sensing detection methods for the mesosphere and lower thermosphere. It can realize the simultaneous detection of atmospheric sodium atomic density, sodium layer temperature and wind field, and has important value for atmospheric science and space physics research. In recent years, with the development of laser technology and photoelectric detection technology, the performance of sodium fluorescence Doppler lidar system has been gradually optimized, which has led to its further promotion and application. With the implementation of major national science and technology infrastructure projects, more and more sodium fluorescence lidars are being established and will be put into operational observation and operation, which also makes people's demand for data quality control of sodium fluorescence Doppler lidar more and more urgent. The accuracy and reliability of lidar data are the fundamental premise for conducting relevant scientific research using detection and analysis data. However, due to the complexity of the hardware system of sodium fluorescence Doppler lidar, the large number of observation data channels, the difficulty of parameter inversion, the inversion process often requires more manual intervention, and the quality of the inversion results is also highly dependent on personal experience. It is easy to miss and misjudge, resulting in large errors in the detection results or a large amount of detection data cannot be fully utilized. In addition, the frequency drift or mode hopping of the transmitting laser, the jitter of the laser pulse energy, the change of the field of view at high and low altitudes, the saturation of the photomultiplier tube, the addition of filters during all-day observations, and the change of the local weather conditions at the observation station will directly affect the accuracy of the detection data. There is currently a lack of solutions on how to set the diagnostic standards for sodium fluorescence Doppler lidar data, establish a systematic and clear data quality control method to obtain reliable atmospheric parameter detection results, and improve the automation and intelligence level of data processing. Summary of the invention

[0003] Purpose of the invention: The purpose of the invention is to provide a data quality control method for a sodium fluorescence Doppler laser radar.

[0004] Another object of the present invention is to provide a data quality control system for a sodium fluorescence Doppler laser radar.

[0005] Technical solution: A sodium fluorescence Doppler laser radar data quality control method of the present invention includes: photon counting data quality control and laser radar data product quality control, wherein:

[0006] Photon counting data quality control specifically includes the following steps:

[0007] Step SA1, respectively performing height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in three detection directions obtained by the sodium fluorescence Doppler laser radar;

[0008] Among them, the three detection directions are zenith, zenith east / west and zenith north / south, and each detection direction obtains three groups of original echo signals corresponding to three working frequencies, and nine groups of original echo signals are obtained corresponding to the three detection directions; the three working frequencies are the peak working frequency f0, the right wing working frequency f+=f0+δf and the left wing working frequency f-=f0-δf, and δf is the frequency shift;

[0009] Step SA2, performing photon counting linearity detection and correction on the nine groups of echo signals in the three detection directions processed in step SA1 respectively;

[0010] Step SA3, performing filter transmittance correction on the nine groups of echo signals in the three detection directions corrected in step SA2 respectively;

[0011] Step SA4, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer temperature ratios in the three detection directions, and respectively performing temperature ratio detection and outlier elimination;

[0012] Step SA5, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer wind speed ratios in the three detection directions, and performing wind speed ratio detection and outlier elimination;

[0013] LiDAR data product quality control specifically includes the following steps:

[0014] Step SB1, using the sodium layer temperature ratio in the zenith direction after the outliers are removed obtained in step SA4 and the sodium layer wind speed ratio in the zenith direction after the outliers are removed obtained in step SA5, the zenith direction wind speed profile is detected and outliers are removed;

[0015] Step SB2, using the sodium layer echo signal, the average value of the high-altitude background signal, the average value of the Rayleigh signal at the reference height, the sodium layer temperature ratio after the abnormal value is eliminated, and the sodium layer wind speed ratio after the abnormal value is eliminated, the temperature detection error is detected and the unreliable value is eliminated;

[0016] Step SB3, using the sodium layer echo signal, the average value of the high-altitude background signal, the average value of the Rayleigh signal at the reference height, the sodium layer temperature ratio in the zenith direction after the abnormal values ​​are removed, and the sodium layer wind speed ratio in the zenith direction after the abnormal values ​​are removed, the wind speed detection error is detected and the unreliable values ​​are removed.

[0017] Furthermore, the height correction in step SA1 is specifically as follows:

[0018] The detection height in the zenith direction after correction is Hd0=Hd0'+H0, Hd0' is the distance data column in the original echo signal of the zenith detection direction stored by the laser radar, and H0 is the altitude of the observation site; the detection height in the east / west direction of the zenith after correction is Hd1=Hd1'×cos(θ1)+H0, Hd1' is the distance data column in the original echo signal of the east / west detection direction of the zenith stored by the laser radar, and θ1 is the azimuth of the east / west detection direction of the zenith; the detection height in the south / north direction of the zenith after correction is Hd2=Hd2'×cos(θ2)+H0, Hd2' is the distance data column in the original echo signal of the south / north detection direction of the zenith stored by the laser radar, and θ2 is the azimuth of the south / north detection direction of the zenith;

[0019] The specific detection of the signal-to-noise ratio of the sodium layer is as follows:

[0020] Calculate the signal-to-noise ratio SNR_Na(d,f) of the peak value of the sodium layer echo signal corresponding to each detection direction. The calculation formula is:

[0021] SNR_Na(d,f)=(Np_Na(d,f)-Nb(d,f)) / B_Na(d,f)

[0022] Among them, Np_Na(d,f) is the peak photon count value of the sodium layer echo signal, Nb(d,f) is the average photon count value of the high-altitude background, and B_Na(d,f) is the counting statistical noise.

[0023] The reference height Rayleigh signal noise ratio detection is specifically:

[0024] Calculate the signal-high-altitude background ratio SNR_RayH(d,f), signal-low-altitude background ratio SNR_RayL(d,f) and signal-to-noise ratio SNR_Ray(d,f) of the reference height Rayleigh scattering echo signal in the echo signal corresponding to each detection direction. The calculation formulas are:

[0025] SNR_RayH(d,f)=Nm_Ray(d,f) / Nb(d,f)

[0026] SNR_RayL(d,f)=Nm_Ray(d,f) / Nb_L(d,f)

[0027] SNR_Ray(d,f)=(Nm_Ray(d,f)-Nb(d,f)) / B_Ray(d,f)

[0028] Where, Nm_Ray(d,f) is the average photon count value of the Rayleigh scattering echo signal at the reference height, Nb_L(d,f) is the average photon count value of the background in the low-altitude section, and B_Ray(d,f) is the counting statistical noise of the Rayleigh signal at the reference height.

[0029] The detection of laser frequency stability is specifically as follows:

[0030] Calculate the average value N of the number of sodium layer photons in the height range near the sodium layer peak in the echo signals corresponding to the peak operating frequency f0, the right-wing operating frequency f+, and the left-wing operating frequency f- in each detection direction. m (h_na0,d,f0), N m (h_na0,d,f+), N m (h_na0,d,f-);

[0031] The rejection of bad data is specifically as follows:

[0032] Reject the files where SNR_Na(d,f) < P1 or SNR_Ray(d,f) < P2 or SNR_RayH(d,f) < P3 or SNR_RayL(d,f) < P4 or N m (h_na0,d,f+) > N m (h_na0,d,f0) or N m (h_na0,d,f-) > N m (h_na0,d,f0). P1, P2, P3, and P4 are the set SNR thresholds for the sodium layer, the Rayleigh SNR threshold, the Rayleigh signal - high-altitude background ratio threshold, and the Rayleigh - low-altitude background ratio threshold, respectively.

[0033] Furthermore, step SA2 is specifically as follows:

[0034] Calculate the photon counting rate R(h,d,f) corresponding to each photon count value N(h,d,f) in the nine groups of echo signals in the three detection directions after being processed in step SA1. The calculation formula is:

[0035] R(h,d,f) = N(h,d,f) / (t b ×M / 3)

[0036] Where, t b is the counting gate width, and M is the cumulative number of pulses;

[0037] And compare the photon counting rate R(h,d,f) with the linear counting rate threshold R0 of the photomultiplier tube. If R(h,d,f) > R0, then let N c (h,d,f) = R c (h,d,f)×tb ×M / 3, otherwise N c (h, d, f) = N(h, d, f);

[0038] Among them, N c (h, d, f) are the photon counts after nonlinear correction; R c (h, d, f) = f{R(h, d, f)} is the photon counting rate after nonlinear correction, and f is the functional relationship corresponding to the photomultiplier tube input counting rate-output counting rate curve measured in the laboratory.

[0039] Furthermore, step SA3 is specifically as follows:

[0040] Measure the transmittance of the atomic filter in the daytime channel for the reference height Rayleigh echo signal T_Ray(f) and the transmittance of the sodium layer echo signal T_Na(f) corresponding to the three working frequencies;

[0041] For the echo signal of the daytime channel, let the reference height correction coefficient C_Ray(f)=T_Ray(f), and the sodium layer correction coefficient C_Na(f)=T_Na(f); for the echo signal of the nighttime channel, let the reference height correction coefficient C_Ray(f)=1, and the sodium layer correction coefficient C_Na(f)=1;

[0042] The nonlinearly corrected photon count value Nc(h, d, f) processed in step SA2 is integrated over a certain height and time to obtain a combined signal ∑N(h, d, f);

[0043] Select the sodium layer height segment signal of the combined signal ∑N(h,d,f) to obtain the sodium layer echo signal ∑N(h_na,d,f);

[0044] The sodium layer echo signal ∑N(h_na,d,f) is normalized by the reference height Rayleigh signal to obtain the normalized sodium layer echo signal N norm (h_na,d,f), the formula is:

[0045] N norm (h_na,d,f)=(∑N(h_na,d,f)-N m (h_b,d,f)) / (N m (h_ray,d,f)-N m (h_b,d,f))

[0046] Among them, N m (h_b,d,f) is the average value of the high-altitude background signal, N m (h_ray,d,f) is the average value of Rayleigh signal at reference height;

[0047] The normalized sodium layer echo signal N is calculated using the reference height correction factor and the sodium layer correction factor. norm (h_na,d,f) performs filter transmittance correction to obtain the sodium layer echo signal N after filter correction cor (h_na,d,f), that is:

[0048] N cor (h_na,d,f)=N norm (h_na,d,f)×C_Ray(f) / C_Na(f).

[0049] Furthermore, step SA4 specifically includes:

[0050] Calculate the sodium layer temperature ratio in the three detection directions respectively. The sodium layer temperature ratio R T The calculation formula for (h_na,d) is:

[0051]

[0052] Among them, N cor (h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction;

[0053] If R T (h_na,d) <R Tmin or R T (h_na,d)>R Tmax , then the corresponding R T (h_na,d) removed; R Tmin and R Tmax are the minimum and maximum thresholds of the temperature ratio respectively.

[0054] Furthermore, step SA5 is specifically as follows:

[0055] Using the N obtained in step SA3 above cor (h_na,d,f), calculate the sodium layer wind speed ratio in three directions respectively, the sodium layer wind speed ratio R w The calculation formula for (h_na,d) is:

[0056]

[0057] Among them, N cor(h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction;

[0058] If R w (h_na,d) <R wmin or R w (h_na,d)>R wmax , then the corresponding R w (h_na,d) removed; R wmin and R wmax are the minimum and maximum thresholds of the wind speed ratio respectively.

[0059] Furthermore, step SB1 is specifically as follows:

[0060] The sodium layer temperature ratio R in the zenith direction after removing the abnormal values ​​obtained in step SA4 T (h_na, d0) and the sodium layer wind speed ratio R in the zenith direction after the abnormal value is removed obtained in step SA5 w (h_na,d0), invert the sodium layer wind speed profile W(h_na,d0) in the zenith direction; use the sodium layer wind speed profile W(h_na,d0) in the zenith direction to calculate the average value W of the height segment near the sodium layer peak m (h_na0,d0), if |W m (h_na0,d0)|>W m0 , then W(h_na,d0) is judged as an abnormal result, and the abnormal value is removed, W m0 is the maximum wind speed error.

[0061] Furthermore, step SB2 is specifically as follows:

[0062] Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w(h_na,d) calculates the temperature detection error △T; if |ΔT|>ΔT0, the corresponding temperature inversion result Temp(h_na,d0) is eliminated; ΔT0 is the temperature detection error threshold.

[0063] Furthermore, in step SB3, the wind speed detection error is detected and unreliable values ​​are eliminated, which specifically includes the following process:

[0064] Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w (h_na,d) calculates the wind speed detection error ΔW; if |ΔW|>ΔW0, the corresponding wind speed inversion result W(h_na,d0) is eliminated; ΔW0 is the wind speed detection error threshold.

[0065] A sodium fluorescence Doppler laser radar data quality control system of the present invention comprises:

[0066] The laser radar raw data storage module is used to store the raw echo signals in three detection directions detected by the laser radar;

[0067] The photon counting data quality control module is used to perform height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in the three detection directions obtained by the sodium fluorescence Doppler lidar; then perform photon counting linearity detection and correction, filter transmittance correction, temperature ratio detection and outlier elimination, wind speed ratio detection and outlier elimination;

[0068] The laser radar data product quality control module uses the sodium layer temperature ratio and the sodium layer wind speed ratio in the zenith direction obtained in the photon counting data quality control module to detect the zenith wind speed profile and remove abnormal values; the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average value N m (h_b, d, f), the average value of the Rayleigh signal at the reference height N m (h_ray,d,f) and the sodium layer temperature ratio and the sodium layer wind speed ratio are used to detect the temperature detection error and wind speed detection error and eliminate unreliable values ​​respectively;

[0069] The lidar qualified data product storage module is used to store qualified data products processed by the photon counting data quality control module and the lidar data product quality control module.

[0070] Beneficial effects: Compared with the prior art, the present invention can realize automatic diagnosis and elimination of all-day sodium lidar observation data by performing step-by-step detection and quality control on the original photon counting data and lidar data products. It has high efficiency and can realize batch automatic processing, which reduces the cost of using human resources, reduces the problem of misjudgment or missed judgment caused by human factors, and improves the accuracy and reliability of lidar data products. In addition, the present invention fully considers the influence of photomultiplier tube saturation and the introduction of filters into the daytime detection channel during all-day observation on the detection data, corrects the photomultiplier tube saturation and filter attenuation, greatly improves the utilization rate of data, and provides strong support for the use of sodium fluorescence Doppler lidar data to carry out atmospheric and space science research. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 is a flow chart of the method of the present invention;

[0072] Figure 2 is a graph showing the change of photon number with time and height before and after the photon counting data quality control; wherein (a) is a graph showing the change of photon number with time and height before the photon counting data quality control, and (b) is a graph showing the change of photon number with time and height after the photon counting data quality control;

[0073] Figure 3 are the temperature and wind speed inversion result diagrams before data product quality control; (a) is the temperature inversion result diagram before data product quality control, and (b) is the wind speed inversion result diagram before data product quality control;

[0074] Figure 4 are the temperature and wind speed inversion result diagrams after data product quality control; (a) is the temperature inversion result diagram after data product quality control, and (b) is the wind speed inversion result diagram after data product quality control;

[0075] Figure 5 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0076] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0077] The present invention aims to establish a systematic sodium fluorescence Doppler laser radar data quality control scheme to solve the problems of frequent manual intervention in the sodium fluorescence Doppler laser radar parameter inversion process and the reliance of the inversion result quality on personal experience, which leads to large errors in the detection results or a large amount of detection data cannot be fully utilized.

[0078] like Figure 1 As shown, a sodium fluorescence Doppler laser radar data quality control method of the present invention comprises the following steps:

[0079] Perform quality control on photon counting data, including:

[0080] Step SA1, respectively performing height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in three detection directions obtained by the sodium fluorescence Doppler laser radar;

[0081] Among them, the three detection directions are d: zenith, east / west of zenith and north / south of zenith. Each detection direction obtains three groups of original echo signals corresponding to three working frequencies, and nine groups of original echo signals are obtained corresponding to the three detection directions; the three working frequencies are f: peak working frequency f0, right wing working frequency f+=f0+δf and left wing working frequency f-=f0-δf, and δf is the frequency shift.

[0082] The height correction is as follows:

[0083] The detection height in the zenith direction after correction is Hd0=Hd0'+H0, Hd0' is the distance data column in the original echo signal of the zenith detection direction stored by the laser radar, and H0 is the altitude of the observation site; the detection height in the east / west direction of the zenith after correction is Hd1=Hd1'×cos(θ1)+H0, Hd1' is the distance data column in the original echo signal of the east / west detection direction of the zenith stored by the laser radar, and θ1 is the azimuth of the east / west detection direction of the zenith; the detection height in the south / north direction of the zenith after correction is Hd2=Hd2'×cos(θ2)+H0, Hd2' is the distance data column in the original echo signal of the south / north detection direction of the zenith stored by the laser radar, and θ2 is the azimuth of the south / north detection direction of the zenith;

[0084] The specific detection of the signal-to-noise ratio of the sodium layer is as follows:

[0085] Calculate the signal-to-noise ratio SNR_Na(d,f) of the peak echo signal of the sodium layer echo signal at an altitude of 80-110km corresponding to each detection direction. SNR_Na(d,f) is defined as follows:

[0086] SNR_Na(d,f)=(Np_Na(d,f)-Nb(d,f)) / B_Na(d,f) (1)

[0087] Among them, Np_Na(d,f) is the peak photon count value of the sodium layer echo signal in the altitude range of 80 - 110 km, Nb(d,f) is the average photon value of the noise taking the altitude section of 150 - 160 km as the background, and B_Na(d,f) is the counting statistical noise.

[0088] The detection of the signal-to-noise ratio of the reference altitude Rayleigh signal is specifically as follows:

[0089] Calculate the signal-to-high-altitude background ratio SNR_RayH(d,f), signal-to-low-altitude background ratio SNR_RayL(d,f), and signal-to-noise ratio SNR_Ray(d,f) of the Rayleigh scattering echo signal at the reference altitude of 30 - 35 km in the echo signal corresponding to each detection direction, which are respectively defined as follows:

[0090] SNR_RayH(d,f) = Nm_Ray(d,f) / Nb(d,f) (2)

[0091] SNR_RayL(d,f) = Nm_Ray(d,f) / Nb_L(d,f) (3)

[0092] SNR_Ray(d,f) = (Nm_Ray(d,f) - Nb(d,f)) / B_Ray(d,f) (4)

[0093] Among them, Nm_Ray(d,f) is the average photon count value of the Rayleigh scattering echo signal corresponding to taking 30 - 35 km as the reference altitude, Nb_L(d,f) is the average photon count value of taking the altitude section of 60 - 70 km as the background, and B_Ray(d,f) is the counting statistical noise of the Rayleigh signal at the altitude of 30 - 35 km.

[0094] The detection of the laser frequency stability is specifically as follows:

[0095] Calculate the average value N of the sodium layer photon numbers in the altitude range of 85 - 95 km in the echo signals corresponding to the peak operating frequency f0, the right wing operating frequency f+, and the left wing operating frequency f- in each detection direction m (h_na0,d,f0), N m (h_na0,d,f+), N m (h_na0,d,f-);

[0096] The rejection of bad data is specifically as follows:

[0097] Let SNR_Na(d,f) < P1 or SNR_Ray(d,f) < P2 or SNR_RayH(d,f) < P3 or SNR_RayL(d,f) < P4 or N m(h_na0,d,f+)>N m (h_na0,d,f0) or N m (h_na0,d,f-)>N m The files with (h_na0,d,f0) are eliminated, and P1, P2, P3, and P4 are the set sodium layer signal-to-noise ratio threshold, Rayleigh signal-to-noise ratio threshold, Rayleigh signal-to-high altitude background ratio threshold, and Rayleigh-to-low altitude background ratio threshold, respectively.

[0098] In this embodiment, the values ​​of P1, P2, P3, and P4 are 5, 5, 1.05, and 1.05, respectively.

[0099] Step SA2, performing photon counting linearity detection and correction on the nine groups of echo signals in the three detection directions processed in step SA1 respectively;

[0100] The photon counting rate R(h, d, f) corresponding to each photon counting value N(h, d, f) in the nine groups of echo signals in the three detection directions after processing in step SA1 is calculated. The calculation method of R(h, d, f) is shown in the following formula:

[0101] R(h,d,f)=N(h,d,f) / (t b ×M / 3) (5)

[0102] Where, t b is the counting gate width, M is the accumulated pulse number, t b The calculation formula of M is as follows:

[0103] t b =2H / c; (6)

[0104] M=T×Rep (7)

[0105] In the formula, c is the speed of light, Rep is the laser radar pulse repetition frequency, H and T are the distance resolution and time resolution of each data file set when the laser radar is working;

[0106] The photon counting rate R(h, d, f) is compared with the linear counting rate threshold R0 of the photomultiplier tube. If R(h, d, f)>R0, then N c (h,d,f)=R c (h,d,f)×t b ×M / 3, otherwise N c (h, d, f) = N(h, d, f);

[0107] Among them, N c (h, d, f) are the photon counts after nonlinear correction; R c(h, d, f) = f{R(h, d, f)} is the photon counting rate after nonlinear correction, and f is the functional relationship corresponding to the photomultiplier tube input counting rate-output counting rate curve measured in the laboratory; R0 can be obtained from the above curve, or by querying the factory nominal parameters of the photomultiplier tube.

[0108] Step SA3, performing filter transmittance correction on the nine groups of echo signals in the three detection directions corrected in step SA2 respectively;

[0109] Measure the transmittance T_Ray(f) of the reference height Rayleigh echo signal and the transmittance T_Na(f) of the sodium layer echo signal of the atomic filter in the daytime channel corresponding to the three working frequencies; for the echo signal of the daytime channel, set the reference height correction coefficient C_Ray(f)=T_Ray(f), and the sodium layer correction coefficient C_Na(f)=T_Na(f); for the echo signal of the nighttime channel, set the reference height correction coefficient C_Ray(f)=1, and the sodium layer correction coefficient C_Na(f)=1;

[0110] The nonlinearly corrected photon count value Nc(h,d,f) processed in step SA2 is integrated for 1 km and 1 h to obtain a combined signal ∑N(h,d,f); the 80-110 km sodium layer height segment signal of the combined signal ∑N(h,d,f) is selected to obtain a sodium layer echo signal ∑N(h_na,d,f); the sodium layer echo signal ∑N(h_na,d,f) is normalized by the reference height Rayleigh signal to obtain a normalized sodium layer echo signal N norm (h_na,d,f), that is:

[0111] N norm (h_na,d,f)=(∑N(h_na,d,f)-N m (h_b,d,f)) / (N m (h_ray,d,f)-N m (h_b,d,f))(9)

[0112] Among them, N m (h_b, d, f) is the average value of the high-altitude background signal, which is obtained by selecting ∑N(h, d, f) signals in the 150-160km altitude segment and averaging them; N m (h_ray,d,f) is the average value of the Rayleigh signal at the reference height, which is obtained by selecting ∑N(h,d,f) signals at the 30-35km altitude segment and averaging them;

[0113] The normalized sodium layer echo signal N is calculated by using the reference height correction coefficient and the sodium layer correction coefficient. norm (h_na,d,f) performs filter transmittance correction to obtain the sodium layer echo signal N after filter correction cor(h_na,d,f), that is:

[0114] N cor (h_na,d,f)=N norm (h_na,d,f)×C_Ray(f) / C_Na(f) (10).

[0115] Step SA4, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer temperature ratios in the three detection directions, and respectively performing temperature ratio detection and outlier elimination;

[0116] The sodium layer echo signal N after filter correction obtained in step SA3 above is cor (h_na,d,f), calculate the sodium layer temperature ratio in the three detection directions respectively, the sodium layer temperature ratio R T The calculation formula for (h_na,d) is:

[0117]

[0118] Among them, N cor (h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction;

[0119] If R T (h_na,d) <R Tmin or R T (h_na,d)>R Tmax , then the corresponding R T (h_na,d) removed; R Tmin and R Tmax are the minimum and maximum thresholds of the temperature ratio respectively.

[0120] In this embodiment, 0.25 and 0.75 are taken as the minimum threshold and maximum threshold of the temperature ratio respectively. T (h_na,d)<0.25 or R T (h_na,d)>0.75, then the corresponding R T (h_na,d) is eliminated.

[0121] Step SA5, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer wind speed ratios in the three detection directions, and performing wind speed ratio detection and outlier elimination;

[0122] The sodium layer echo signal N after filter correction obtained in step SA3 above is cor (h_na,d,f), calculate the sodium layer wind speed ratio in the three detection directions respectively, the sodium layer wind speed ratio R w The calculation formula for (h_na,d) is:

[0123]

[0124] Among them, N cor (h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction;

[0125] If R w (h_na,d) <R wmin or R w (h_na,d)>R wmax , then the corresponding R w (h_na,d) removed; R wmin and R wmax are the minimum and maximum thresholds of the wind speed ratio respectively.

[0126] In this embodiment, -0.5 and 0.5 are taken as the minimum threshold and maximum threshold of the wind speed ratio respectively. w (h_na,d)<-0.5 or R w (h_na,d)>0.5, then the corresponding R w (h_na,d) is eliminated.

[0127] Figure 2 (a) shows the variation of photon counts with time and height before quality control of photon counting data. Figure 2 (b) shows the change of photon number with time and altitude after photon counting data quality control. It can be seen from the figure that after quality control, the photon number data with low signal-to-noise ratio and poor data quality are eliminated, and the photon number size also changes after atomic filter transmittance correction during the daytime observation period.

[0128] After the photon counting data quality control, the lidar data product quality control is carried out, which specifically includes the following steps:

[0129] Step SB1: Detect the wind speed profile in the zenith direction and remove abnormal values:

[0130] The sodium layer temperature ratio R in the zenith direction after removing the abnormal values ​​obtained in step SA4 above is T (h_na, d0) and the sodium layer wind speed ratio R in the zenith direction after the abnormal value is removed obtained in step SA5 w (h_na,d0), invert the sodium layer wind speed profile W(h_na,d0) in the zenith direction; use the above sodium layer wind speed profile W(h_na,d0) in the zenith direction to calculate the average value W in the 85-95km height segment m (h_na0,d0), taking the maximum wind speed error W caused by the chirp effect of the laser pulse frequency m0 =15m / s, if |W m (h_na0,d0)|>15m / s, then W(h_na,d0) is judged as an abnormal result and the outlier is removed.

[0131] Step SB2, detecting temperature detection errors and eliminating unreliable values;

[0132] Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w (h_na,d) calculates the temperature detection error △T; takes the temperature detection error threshold △T0=15K, if |△T|>15K, the corresponding temperature inversion result Temp(h_na,d0) is eliminated.

[0133] Step SB3: Detect wind speed detection errors and eliminate unreliable values:

[0134] Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w(h_na,d) calculates the wind speed detection error △W; takes the wind speed detection error threshold △W0=20m / s, if |△W|>20m / s, the corresponding wind speed inversion result W(h_na,d0) is eliminated.

[0135] Figure 3 and 4 The temperature and wind speed data product results of lidar detection before and after data product quality control are shown respectively. After data quality control, the unreliable detection values ​​of the upper and lower boundaries of the sodium layer are eliminated, and the temperature and wind speed values ​​are within a reasonable range, indicating that this method can effectively control the quality of sodium fluorescence Doppler lidar data.

[0136] like Figure 5 As shown, a sodium fluorescence Doppler laser radar data quality control system of the present invention comprises:

[0137] The laser radar raw data storage module is used to store the raw echo signals in three detection directions detected by the laser radar;

[0138] The photon counting data quality control module is used to perform height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in the three detection directions obtained by the sodium fluorescence Doppler lidar; then perform photon counting linearity detection and correction, filter transmittance correction, temperature ratio detection and outlier elimination, wind speed ratio detection and outlier elimination;

[0139] The laser radar data product quality control module uses the sodium layer temperature ratio and the sodium layer wind speed ratio in the zenith direction obtained in the photon counting data quality control module to detect the zenith wind speed profile and remove abnormal values; the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average value N m (h_b, d, f), the average value of the Rayleigh signal at the reference height N m (h_ray,d,f) and the sodium layer temperature ratio and the sodium layer wind speed ratio are used to detect the temperature detection error and wind speed detection error and eliminate unreliable values ​​respectively;

[0140] The lidar qualified data product storage module is used to store qualified data products processed by the photon counting data quality control module and the lidar data product quality control module.

[0141] In summary, the method of the present invention includes photon counting data quality control and data product quality control. Through the processes of original echo signal height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination, photon counting linearity detection and correction, filter transmittance correction, temperature ratio detection and outlier elimination, wind speed ratio detection and outlier elimination, zenith direction wind speed profile detection and outlier elimination, temperature and wind speed detection error detection and unreliable value elimination, etc., automatic diagnosis and batch processing of sodium fluorescence Doppler laser radar data are realized, and the problems of much manual intervention in the sodium fluorescence Doppler laser radar parameter inversion process and the dependence of the inversion result quality on personal experience resulting in large errors in the detection results or a large amount of detection data cannot be fully utilized are solved.

Claims

1. A sodium fluorescence Doppler laser radar data quality control method, characterized in that: include: Photon counting data quality control and lidar data product quality control, including: Photon counting data quality control specifically includes the following steps: Step SA1, respectively performing height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in three detection directions obtained by the sodium fluorescence Doppler laser radar; Among them, the three detection directions are zenith, zenith east / west and zenith north / south, and each detection direction obtains three groups of original echo signals corresponding to three working frequencies, and nine groups of original echo signals are obtained corresponding to the three detection directions; the three working frequencies are the peak working frequency f0, the right wing working frequency f+=f0+δf and the left wing working frequency f-=f0-δf, and δf is the frequency shift; Step SA2, performing photon counting linearity detection and correction on the nine groups of echo signals in the three detection directions processed in step SA1 respectively; Step SA3, performing filter transmittance correction on the nine groups of echo signals in the three detection directions corrected in step SA2 respectively; Step SA4, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer temperature ratios in the three detection directions, and performing temperature ratio detection and outlier elimination respectively; Step SA5, respectively calculating the nine groups of echo signals in the three detection directions after the filter transmittance correction in step SA3 to obtain three groups of sodium layer wind speed ratios in the three detection directions, and performing wind speed ratio detection and outlier elimination; LiDAR data product quality control specifically includes the following steps: Step SB1, using the sodium layer temperature ratio in the zenith direction after the outliers are removed obtained in step SA4 and the sodium layer wind speed ratio in the zenith direction after the outliers are removed obtained in step SA5, the zenith direction wind speed profile is detected and outliers are removed; Step SB2, using the sodium layer echo signal, the average value of the high-altitude background signal, the average value of the reference height Rayleigh signal, the sodium layer temperature ratio after the abnormal value is eliminated, and the sodium layer wind speed ratio after the abnormal value is eliminated, the temperature detection error is detected and the unreliable value is eliminated; Step SB3, using the sodium layer echo signal, the average value of the high-altitude background signal, the average value of the Rayleigh signal at the reference height, the sodium layer temperature ratio in the zenith direction after the abnormal values ​​are removed, and the sodium layer wind speed ratio in the zenith direction after the abnormal values ​​are removed, the wind speed detection error is detected and the unreliable values ​​are removed.

2. A sodium fluorescence Doppler laser radar data quality control method according to claim 1, characterized in that: The height correction in step SA1 is specifically as follows: The detection height in the zenith direction after correction is Hd0=Hd0'+H0, Hd0' is the distance data column in the original echo signal of the zenith detection direction stored by the laser radar, and H0 is the altitude of the observation site; the detection height in the east / west direction of the zenith after correction is Hd1=Hd1'×cos(θ1)+H0, Hd1' is the distance data column in the original echo signal of the east / west detection direction of the zenith stored by the laser radar, and θ1 is the azimuth of the east / west detection direction of the zenith; the detection height in the south / north direction of the zenith after correction is Hd2=Hd2'×cos(θ2)+H0, Hd2' is the distance data column in the original echo signal of the south / north detection direction of the zenith stored by the laser radar, and θ2 is the azimuth of the south / north detection direction of the zenith; The specific detection of the signal-to-noise ratio of the sodium layer is as follows: Calculate the signal-to-noise ratio SNR_Na(d,f) of the peak value of the sodium layer echo signal corresponding to each detection direction. The calculation formula is: SNR_Na(d,f)=(Np_Na(d,f)-Nb(d,f)) / B_Na(d,f) Among them, Np_Na(d,f) is the peak photon count value of the sodium layer echo signal, Nb(d,f) is the average photon count value of the high-altitude background, and B_Na(d,f) is the counting statistical noise. The reference height Rayleigh signal noise ratio detection is specifically: Calculate the signal-high-altitude background ratio SNR_RayH(d,f), signal-low-altitude background ratio SNR_RayL(d,f) and signal-to-noise ratio SNR_Ray(d,f) of the reference height Rayleigh scattering echo signal in the echo signal corresponding to each detection direction. The calculation formulas are: SNR_RayH(d,f)=Nm_Ray(d,f) / Nb(d,f) SNR_RayL(d,f)=Nm_Ray(d,f) / Nb_L(d,f) SNR_Ray(d,f)=(Nm_Ray(d,f)-Nb(d,f)) / B_Ray(d,f) Among them, Nm_Ray(d,f) is the average photon count value of the Rayleigh scattering echo signal at the reference altitude, Nb_L(d,f) is the average photon count value of the low-altitude background, and B_Ray(d,f) is the statistical noise of the Rayleigh signal count at the reference altitude. Laser frequency stability detection is specifically as follows: Calculate the average number of sodium layer photons N in the echo signal within the height range near the sodium layer peak value corresponding to the peak operating frequency f0, the right wing operating frequency f+, and the left wing operating frequency f- in each detection direction m (h_na0,d,f0),N m (h_na0,d,f+),N m (h_na0,d,f-); The specific steps for removing bad data are: Set SNR_Na(d,f)<P1 or SNR_Ray(d,f)<P2 or SNR_RayH(d,f)<P3 or SNR_RayL(d,f)<P4 or N m (h_na0,d,f+)>N m (h_na0,d,f0) or N m (h_na0,d,f-)>N m The files with (h_na0,d,f0) are eliminated, and P1, P2, P3, and P4 are the set sodium layer signal-to-noise ratio threshold, Rayleigh signal-to-noise ratio threshold, Rayleigh signal-to-high altitude background ratio threshold, and Rayleigh-to-low altitude background ratio threshold, respectively.

3. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: Step SA2 is specifically as follows: The photon counting rate R(h, d, f) corresponding to each photon counting value N(h, d, f) in the nine groups of echo signals in the three detection directions after processing in step SA1 is calculated, and the calculation formula is: R(h,d,f)=N(h,d,f) / (t b ×M / 3) Among them, t b is the counting gate width, M is the accumulated pulse number; The photon counting rate R(h, d, f) is compared with the linear counting rate threshold R0 of the photomultiplier tube. If R(h, d, f)>R0, then N c (h,d,f)=R c (h,d,f)×t b ×M / 3, otherwise N c (h, d, f) = N(h, d, f); Among them, N c (h, d, f) are the photon counts after nonlinear correction; R c (h, d, f) = f{R(h, d, f)} is the photon counting rate after nonlinear correction, and f is the functional relationship corresponding to the photomultiplier tube input counting rate-output counting rate curve measured in the laboratory.

4. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: Step SA3 is specifically as follows: Measure the transmittance of the atomic filter in the daytime channel for the reference height Rayleigh echo signal T_Ray(f) and the transmittance of the sodium layer echo signal T_Na(f) corresponding to the three working frequencies; For the echo signal of the daytime channel, let the reference height correction coefficient C_Ray(f)=T_Ray(f), and the sodium layer correction coefficient C_Na(f)=T_Na(f); for the echo signal of the nighttime channel, let the reference height correction coefficient C_Ray(f)=1, and the sodium layer correction coefficient C_Na(f)=1; The nonlinearly corrected photon count value Nc(h, d, f) processed in step SA2 is integrated over a certain height and time to obtain a combined signal ∑N(h, d, f); Select the sodium layer height segment signal of the combined signal ∑N(h,d,f) to obtain the sodium layer echo signal ∑N(h_na,d,f); The sodium layer echo signal ∑N(h_na,d,f) is normalized by the reference height Rayleigh signal to obtain the normalized sodium layer echo signal N norm (h_na,d,f), the formula is: N norm (h_na,d,f)=(∑N(h_na,d,f)-N m (h_b,d,f)) / (N m (h_ray,d,f)-N m (h_b,d,f)) Among them, N m (h_b,d,f) is the average value of the high-altitude background signal, N m (h_ray,d,f) is the average value of Rayleigh signal at reference height; The normalized sodium layer echo signal N is calculated using the reference height correction factor and the sodium layer correction factor. norm (h_na,d,f) performs filter transmittance correction to obtain the sodium layer echo signal N after filter correction cor (h_na,d,f), that is: N cor (h_na,d,f)N norm (h_na,d,f)×C_Ray(f) / C_Na(f).

5. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: The specific steps in step SA4 are as follows: Calculate the sodium layer temperature ratio in the three detection directions respectively. The sodium layer temperature ratio R T The calculation formula for (h_na,d) is: Among them, N cor (h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction; If R T (h_na,d)<R Tmin or R T (h_na,d)>R Tmax , then the corresponding R T (h_na,d) removed; R Tmin and R Tmax are the minimum and maximum thresholds of the temperature ratio respectively.

6. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: Step SA5 is specifically as follows: Using the N obtained in step SA3 above cor (h_na,d,f), calculate the sodium layer wind speed ratio in three directions respectively, the sodium layer wind speed ratio R w The calculation formula for (h_na,d) is: Among them, N cor (h_na,d,f0) is the sodium layer echo signal corresponding to the peak operating frequency f0 after filter correction, N cor (h_na,d,f + ) is the sodium layer echo signal corresponding to the right wing operating frequency f+ after filter correction, N cor (h_na,d,f - ) is the sodium layer echo signal corresponding to the right wing operating frequency f- after filter correction; If R w (h_na,d)<R wmin or R w (h_na,d)>R wmax , then the corresponding R w (h_na,d) removed; R wmin and R wmax are the minimum and maximum thresholds of the wind speed ratio respectively.

7. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: Step SB1 is specifically as follows: The sodium layer temperature ratio R in the zenith direction after removing the abnormal values ​​obtained in step SA4 T (h_na, d0) and the sodium layer wind speed ratio R in the zenith direction after the abnormal value is removed obtained in step SA5 w (h_na,d0), invert the sodium layer wind speed profile W(h_na,d0) in the zenith direction; use the sodium layer wind speed profile W(h_na,d0) in the zenith direction to calculate the average value W of the height segment near the sodium layer peak m (h_na0,d0), if |W m (h_na0,d0)|>W m0 , then W(h_na,d0) is judged as an abnormal result, and the abnormal value is removed, W m0 is the maximum wind speed error.

8. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: Step SB2 is specifically as follows: Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w (h_na,d) calculates the temperature detection error △T; if |△T|>ΔT0, the corresponding temperature inversion result Temp(h_na,d0) is eliminated; △T0 is the temperature detection error threshold.

9. The method for controlling the quality of sodium fluorescence Doppler laser radar data according to claim 1, characterized in that: In step SB3, the wind speed detection error is detected and unreliable values ​​are eliminated, which specifically includes the following process: Using the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average N in step SA3 m (h_b, d, f) and the average value of the Rayleigh signal at the reference height N m (h_ray, d, f), the sodium layer temperature ratio R after removing the abnormal values ​​obtained in step SA4 T (h_na,d) and the sodium layer wind speed ratio R after the abnormal value is removed obtained in step SA5 w (h_na,d) calculates the wind speed detection error ΔW; if |ΔW|>ΔW0, the corresponding wind speed inversion result W(h_na,d0) is eliminated; ΔW0 is the wind speed detection error threshold.

10. A sodium fluorescence Doppler laser radar data quality control system, characterized in that: include: The laser radar raw data storage module is used to store the raw echo signals in three detection directions detected by the laser radar; The photon counting data quality control module is used to perform height correction, sodium layer signal-to-noise ratio detection, reference height Rayleigh signal signal-to-noise ratio detection, laser frequency stability detection and bad data elimination on the original echo signals in the three detection directions obtained by the sodium fluorescence Doppler lidar; then perform photon counting linearity detection and correction, filter transmittance correction, temperature ratio detection and outlier elimination, wind speed ratio detection and outlier elimination; The laser radar data product quality control module uses the sodium layer temperature ratio and the sodium layer wind speed ratio in the zenith direction obtained in the photon counting data quality control module to detect the zenith wind speed profile and remove abnormal values; the sodium layer echo signal ∑N(h_na,d,f) and the high-altitude background signal average value N m (h_b, d, f), the average value of the Rayleigh signal at the reference height N m (h_ray,d,f) and the sodium layer temperature ratio and the sodium layer wind speed ratio are used to detect the temperature detection error and wind speed detection error and eliminate unreliable values ​​respectively; The lidar qualified data product storage module is used to store qualified data products processed by the photon counting data quality control module and the lidar data product quality control module.