Satellite-borne laser radar calibration method and system based on background noise iteration

Abstract

The invention provides a satellite-borne laser radar calibration method and system based on background noise iteration. The method comprises the following steps: firstly, constructing an ideal background atmosphere total attenuation backscattering coefficient, obtaining night background noise and a calibration coefficient by adopting a robust linear fitting iteration method, and realizing calibration of a satellite-borne laser radar night detection signal; in combination with a strategy of a CALIPSO official calibration method, a day and night calibration conversion area is set, simple level detection is carried out, and a clean atmosphere calibration conversion area is screened out; calculating an attenuation scattering ratio of a day and night clean atmosphere calibration conversion area, obtaining a calibration conversion coefficient, and combining a night calibration coefficient to complete iterative calculation of a day-to-day calibration coefficient and background noise, thereby realizing calibration of a satellite-borne laser radar day-to-day detection signal; according to the invention, satellite-borne laser radar day and night calibration can be realized without collecting background noise, and the application range is expanded compared with a traditional method. The satellite-borne laser radar on-orbit calibration method is suitable for satellite-borne laser radar on-orbit calibration under different height detection conditions, and has practicability and high accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of radar technology, and in particular relates to a satellite-borne laser radar calibration method and system based on background noise iteration. Background Art

[0002] Atmospheric aerosols refer to solid or liquid particles between 0.001 and 100 μm suspended in the air, which play a very important role in the physical and chemical processes of the atmosphere. Therefore, it is necessary to conduct high-resolution horizontal and vertical continuous observations of global clouds and aerosols. Due to its short working wavelength, lidar can directly scatter with clouds and aerosol particles in the atmosphere, and has the characteristics of high directivity, high monochromaticity, and high temporal and spatial resolution. It meets the needs of regionalization and vertical profile detection of the atmosphere and has become an effective means of studying the vertical distribution of clouds and aerosols. At present, the most representative satellite-borne lidar is the CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations) satellite launched by NASA in 2006 and its CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) lidar. CALIOP has been in orbit for about 17 years, proving the long life and high reliability of satellite-borne lidar in orbit. However, due to performance degradation, the radar stopped working in August 2023. On April 16, 2022, my country successfully launched the atmospheric environment monitoring satellite DQ-1. Its onboard lidar payload, the Aerosol and Carbon Dioxide Detection Lidar (ACDL), used high spectral resolution cloud and aerosol detection technology in satellite detection for the first time, and has great potential for high-precision detection of clouds and aerosols.

[0003] The interaction between clouds and aerosols has an important impact on global radiative forcing. To accurately quantify and assess the interaction between clouds and aerosols, it is first necessary to correctly calibrate the cloud and aerosol detection sensors to obtain high-precision cloud and aerosol physical quantity detection data. At present, there are many calibration methods based on spaceborne lidar, which can convert the detection electrical signal of the spaceborne lidar into an attenuated backscattering coefficient that is only related to the optical properties of the atmospheric components. Based on the attenuated backscattering coefficient product obtained by calibration, cloud and aerosol layer detection, layer classification, and optical parameter inversion can be carried out to obtain the spatial distribution, type, and content information of clouds and aerosols, and ultimately serve the quantitative evaluation of the interaction between clouds and aerosols.

[0004] The calibration of satellite-borne laser radar is usually divided into three parts: background noise subtraction, range gain correction and on-orbit radiation calibration. First, background noise subtraction is based on the assumption that there is no atmospheric molecular load in the high-altitude detection signal. The mean value of the high-altitude detection signal is collected as the background noise, and the background noise component is subtracted from the detection signal of the entire altitude range; the range gain correction is based on the laser radar detection equation to perform distance square correction and system gain coefficient correction, and normalize the variables that are not related to the optical characteristics of the atmospheric components; on-orbit radiation calibration is to simulate the ideal atmospheric attenuation backscatter signal based on the atmospheric model data, calculate the calibration coefficient of the detection signal using the ratio of the actual detection signal to the ideal signal, and use the calibration coefficient for correction to obtain an accurate attenuation backscatter coefficient.

[0005] At present, there are many calibration methods for satellite-borne lidar. The most representative one is the CALIPSO official day and night calibration method. This method has undergone multiple versions of iterations and is divided into two sub-methods: night calibration and day calibration.

[0006] The official night calibration method of CALIPSO is to calculate the calibration coefficient based on the atmospheric signal estimated by the atmospheric molecular scattering model to achieve the best match between the measured signal and the estimated signal in the calibration area. In the latest V4 version of the official night calibration method of CALIPSO, firstly, considering that there is almost no backscattering component from atmospheric components in the detection signal at an altitude of 97-112km, the average signal of this area is collected as background noise for deduction; secondly, due to the distance attenuation effect of the laser radar echo signal and the amplifier gain effect, the detection signal after deducting the background noise signal is corrected for distance gain to obtain the uncalibrated signal; the latest V4 version method constructed by the official believes that the influence of clouds and aerosols in the altitude range of 34-38km is almost negligible, so this area is used as the calibration area, and the ratio of the uncalibrated signal to the atmospheric molecular attenuation backscattering signal estimated by the atmospheric molecular scattering model is calculated as the calibration coefficient, and applied to all signals to achieve the calibration of CALIPSO night detection signals.

[0007] During the day, the signal-to-noise ratio of the satellite-borne lidar detection signal drops sharply compared to nighttime detection, which makes it difficult to apply the nighttime calibration method to the daytime. Therefore, the official CALIPSO daytime calibration method is based on the basic assumption that the aerosol content in the high-altitude clean atmosphere area is short-term constant during the day and night, and a set of latitude-related proportional factors are applied to the nighttime calibration coefficients to estimate the daytime calibration coefficients. In the latest V4 version of the official CALIPSO daytime calibration method, the same strategy as at night is still used to subtract background noise; secondly, the stratospheric area with low frequency of cloud and aerosol is selected as the calibration conversion area, and the day-to-night attenuation scattering ratio in the calibration conversion area is used to construct the calibration conversion coefficient; finally, the latitude-related calibration conversion coefficient is applied to the nighttime calibration coefficient under the adjacent detection scene, and the daytime calibration coefficient under the daytime scene is converted to achieve the calibration of the CALIPSO daytime detection signal.

[0008] Although the current mainstream satellite-borne lidar calibration method, such as the official calibration method of CALIPSO, has been widely used in academic research and actual production, this method has limitations in principle. Specifically:

[0009] In the official CALIPSO day and night calibration method, the collection and deduction of background noise are based on the assumption that the atmospheric components in the laser echo detection signal in the extremely high altitude area are thin and there is almost no backscattering component of the atmospheric components. Therefore, the mean value of the extremely high altitude detection signal is collected as the background noise. This requires that the detection signal height recorded by the satellite-borne laser radar should be high enough (such as the background noise deduction height specified in the official CALIPSO calibration method is 97-112km). For satellite-borne laser radars with limited recording detection signal height, when using this type of calibration method, the load of atmospheric molecules in the detection area will increase due to the low signal recording height, resulting in the assumption that there is almost no backscattering component of the atmospheric components in the detection signal. At this time, if the mean value of the detection signal is still collected and deducted as background noise, a part of the atmospheric backscattering signal will be deducted, that is, the background noise is over-deducted. And because the calculation of the calibration coefficient requires the detection signal after the background noise is deducted, when the background noise is over-deducted, the calibration coefficient of the satellite-borne laser radar will also be calculated incorrectly, which will eventually cause incorrect calibration. In addition, mainstream calibration methods have relatively strict requirements on the maximum signal recording altitude of satellite-borne lidar, which increases the burden on signal storage and on-orbit data transmission, and has relatively strict requirements on the hardware level.

[0010] In order to solve the above problems and contradictions, it is necessary to establish a more accurate day and night calibration method and theoretical system for satellite-borne laser radar. Therefore, this patent, based on the advantages of the classic CALIPSO official day and night calibration method, proposes a satellite-borne laser radar calibration method and system based on background noise iteration. This method divides the satellite-borne laser radar detection signal into night detection signal and day detection signal according to the detection time. For the night detection signal, according to the relationship between the original detection signal and the attenuated backscattering coefficient, the ideal attenuated backscattering coefficient of the calibration area constructed based on the atmospheric scattering model is used, and the robust linear fitting iterative method is used to directly estimate the background noise and calibration coefficient, thereby realizing the night calibration of the satellite-borne laser radar detection signal; the daytime detection signal uses an iterative method to complete the estimation of the background noise to ensure the accuracy of the daytime calibration result. Summary of the invention

[0011] In order to solve the problem that the signal detection height relied on by the background noise collection method in the mainstream calibration method of satellite-borne laser radar is too high, the present invention proposes a satellite-borne laser radar calibration method and system based on background noise iteration.

[0012] The present invention absorbs the advantages of mainstream calibration methods of spaceborne lidar and calibrates the signals separately for nighttime detection signals and daytime detection signals according to the signal detection time. However, the traditional background noise collection method is no longer used. Instead, the ideal background atmospheric molecule scattering signal in the calibration area is evaluated by fitting atmospheric model data. The background noise and calibration coefficient are estimated simultaneously based on a robust linear fitting method and an iterative method. The day and night calibration method is optimized, which can avoid the excessive deduction of background noise caused by the traditional method and improve the calibration accuracy.

[0013] The technical solution of the method of the present invention is a satellite-borne laser radar calibration method based on background noise iteration, comprising the following steps:

[0014] Step 1: Construct the ideal background atmospheric total attenuation backscatter coefficient, the linear relationship between the ideal background atmospheric total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the two unknown quantities in the calibration, namely the background noise and the calibration coefficient, so as to realize the calibration of the night detection signal of the space-borne lidar;

[0015] Step 2: During daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, each calibration conversion area under certain environmental factors is set, and a simple layer detection is performed on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion areas where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion areas are retained;

[0016] Step 3: For each clean atmosphere calibration conversion area, calculate the night attenuation scattering ratio and the day attenuation scattering ratio respectively, and then calculate the calibration conversion coefficient. Apply the calibration conversion coefficient to the average calibration coefficient of the adjacent previous night scene to obtain the unknown calibration coefficient in the daytime scene. Then use the linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism to solve the unknown background noise. Iteratively update the calibration coefficient and the unknown background noise, and finally realize the calibration of the daytime detection signal of the space-borne lidar.

[0017] Preferably, in step 1, an attenuated backscattering coefficient profile under a clean atmosphere state is constructed according to an atmospheric scattering model, and the attenuated backscattering coefficient of clean atmospheric molecules is specifically calculated as follows:

[0018]

[0019] Among them, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The clean atmospheric molecule attenuation backscattering coefficient; β m (z,k p ) represents the backscattering coefficient of atmospheric molecules; and Represents the two-way transmittance of atmospheric molecules and ozone molecules respectively; z is the altitude of the detection point, k p is the profile index number.

[0020] The attenuation-scattering ratio is defined as follows:

[0021]

[0022] Among them, R′(z, k p ) represents the altitude of the detection point is z and the profile index is k p The attenuation scattering ratio, β′ total (z,k p ) represents the altitude of the detection point is z and the profile index is k p The attenuated backscattering coefficient contributed by the clean atmospheric molecules and particles; β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; The altitude of the detection point is z and the profile index is k p The bidirectional transmittance of particles; z is the altitude of the detection point, in km; k p is the profile index number;

[0023] The ideal background atmosphere total attenuation backscattering coefficient profile of the calibration area is as follows:

[0024] β′ ideal (z,k p ) = R′(z,k p )·β′ m (z,k p ) = β(z, k p )·T 2 (z,k p )

[0025] Among them, β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The total attenuation backscattering coefficient of the ideal background atmosphere; R′(z, k p ) represents the altitude of the detection point is z and the profile index number is k p The attenuation scattering ratio, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The atmospheric molecule attenuation backscattering coefficient profile constructed by the atmospheric scattering model; β(z, k p ) and T 2 (z,k p ) represent the total backscattering coefficient and the two-way transmittance, respectively, which are contributed by the clean atmospheric molecules and clouds and aerosol particles. The formula is as follows:

[0026] β(z,k p )=β m (z,k p )+β p (z,k p )

[0027]

[0028] Among them, β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; and They represent the altitude of the detection point, z, and the profile index number, k. p The bidirectional transmittance of atmospheric molecules, ozone molecules, and cloud and aerosol particles.

[0029] The original detection signal received by the satellite-borne laser radar is defined as:

[0030]

[0031] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at is contributed by the solar background noise and the dark current noise; r is the distance between the satellite-borne lidar and the detection point, in km.

[0032] The relationship between the original detection signal and the ideal total attenuation backscatter coefficient is constructed as shown in the definition of the original detection signal received by the space-borne lidar, which is a linear relationship of the form y = k x + b, where y = P (z, k p ), k and b are the calibration coefficients C(k p ) and background noise N Back (k p );

[0033] Therefore, a calibration area can be selected, and the corresponding detection point height set is Z Cali , for a specific profile index number k p , will belong to Z Cali P(z, k) at different heights z p )and As input, the profile is subjected to robust linear fitting iteration according to the linear relationship defined by the original detection signal received by the laser radar, and finally the calibration coefficient C(k p ) and background noise N Back (k p ) and output it.

[0034] Based on this principle, we can avoid collecting background noise based on extremely high altitude detection signals, as in the official CALIPSO calibration method, and complete the iterative estimation of the fitting of background noise and calibration coefficients at lower altitudes.

[0035]

[0036] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of detection points corresponding to the selected calibration area;

[0037] The robust linear fitting iteratively estimates the calibration coefficient and background noise as follows:

[0038] Step 1.1: First, perform horizontal averaging on the detection signal and the ideal total attenuation backscatter coefficient to improve the signal-to-noise ratio and obtain the average profile under a certain range of horizontal resolution;

[0039] Step 1.2: For each average profile, the regression coefficient is iteratively calculated based on the robust linear fitting method of M estimation. When the regression coefficient converges, the estimated calibration coefficient C(k p ) and background noise N Back (k p ), where C(k p ) and N Back (k p ) represent the profile index number k p Calibration coefficients and background noise at ;

[0040] Step 1.3: Use a two-dimensional space-time sliding window to calibrate the coefficients C(k p ) to perform sliding average;

[0041] Step 1.4: For the calibration coefficient C(k obtained after smoothing in step 1.3 p ), update the background noise N Back (k p ), the calculation formula is as follows:

[0042]

[0043] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z.

[0044] The calibration coefficient C(k p ) and background noise N Back (k p ) is applied to the detection signal of the space-borne laser radar, and the background noise subtraction, range gain correction and radiation calibration process are completed at the same time, and the attenuated backscatter coefficient profile β′(z, k p ), as follows:

[0045]

[0046] Among them, β′(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The attenuated backscatter coefficient obtained after calibration at ; P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k pThe updated value of the background noise at ; r is the distance between the satellite-borne laser radar and the detection point, in km;

[0047] Preferably, the step 2 sets each calibration conversion area under certain environmental factors, as follows

[0048] Select a certain temperature and a certain height at the bottom of the stratosphere to construct multiple calibration conversion areas in the horizontal direction. Each calibration conversion area extends a certain distance in the horizontal direction along the direction of the satellite-borne lidar scene.

[0049] As a preference, the calculation of the night attenuation scattering ratio in step 3 The details are as follows:

[0050]

[0051] Calculate the daytime attenuation scattering ratio as described in step 3 The details are as follows:

[0052]

[0053] Among them, β′ night (z,k p ) represents the altitude z and the profile index number k p The night attenuated backscatter coefficient obtained by calibrating the space-borne lidar at β′ m，day (z,k p ) and β′ m，night (z,k p ) represent the altitude z and the profile index k respectively p The attenuation backscattering coefficients of atmospheric molecules during the day and night calculated according to the atmospheric scattering model, represents the mean value of the calibration coefficients recorded in the previous night scene adjacent to the daytime scene of the spaceborne lidar; the angle brackets indicate all the profiles k located in the calibration conversion region p The average value of all heights z; X day (z,k p ) represents the altitude z and the profile index number k p The uncalibrated daytime signal at which background noise subtraction and gain distance correction have been completed is defined as follows:

[0054]

[0055] Among them, X day (z,k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The background noise at P is deducted and the gain distance corrected daytime uncalibrated signal; day (z,kp ) represents the daytime detection of the satellite-borne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and they are all known quantities; N Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km;

[0056] The calibration conversion coefficient W is obtained according to the obtained day and night attenuation scattering ratios of different calibration conversion areas. The specific calculation formula is as follows:

[0057]

[0058] Where W represents the calibration conversion factor; and Represents the average attenuation-scattering ratio of the clean atmosphere calibration conversion area at night and day corresponding to similar latitudes.

[0059] In order to obtain a stable and accurate calibration conversion coefficient W, the calibration conversion coefficients accumulated in a period of time before the time point corresponding to the daytime scene currently detected by the space-borne laser radar are averaged to obtain an estimated value of the calibration conversion coefficient of the scene;

[0060] Taking into account the spatial differences in the detection environment, the estimated value of W is interpolated into a latitude-related sequence. Since the detection scene of the spaceborne laser radar in the polar region will pass through the same latitude many times and is difficult to distinguish, the estimated value of W after latitude interpolation will be reported as the number of seconds W(t) when the scene starts to be detected, so as to distinguish the different calibration conversion coefficients reported at the same latitude in the polar region; where t represents the time the spaceborne laser radar has spent on the detection of the scene;

[0061] Multiply the time-varying and space-time calibration conversion coefficient W(t) by the average of the calibration coefficients of the previous night scene adjacent to the day scene. The number of profile indexes corresponding to different times t in this scenario is k p If there is no calibration conversion area that meets the requirements or the area is not covered by the night scene, the calibration coefficients of the adjacent night scenes are used for time series interpolation, as follows:

[0062]

[0063] Among them C day (k p) represents the profile index number k obtained by the calibration conversion factor p The daytime calibration coefficient; W(t) represents the calibration conversion coefficient obtained after latitude interpolation that changes with the detection time t; Represents the average value of the calibration coefficients of the previous nighttime scenes adjacent to this daytime scene.

[0064] The background noise N is updated according to the relationship between the calibration coefficient and the background noise. Back (k p ), the specific calculation formula is as follows;

[0065]

[0066] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z.

[0067] Since the daytime calibration factor C day (k p ) involves the calculation of the uncalibrated signal during the day, and it is necessary to assume in advance that the background noise N Back (k p ), and then update the estimated value of the background noise according to the calculated calibration coefficient;

[0068] The calculation of calibration coefficients and background noise in daytime detection signal calibration is essentially an iterative update process;

[0069] First, the initial estimate of the background noise is calculated using the detection signal of the satellite-borne lidar and the atmospheric backscattering model. In order to obtain the optimal background noise, the iterative termination discriminant function δ(i) is defined. When the termination discriminant function is less than the threshold, the iteration is terminated and C is obtained. i (k p ) and Used for calculation of daytime attenuation backscatter coefficient, daytime calibration is completed; the initial estimate of the background noise The calculation formula of the iteration termination discriminant function δ(i) is as follows:

[0070]

[0071]

[0072] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; Represents the number of profile indexes as k p The initial estimate of the background noise at ; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z; and represents the updated value of the background noise in the entire detection scene during the i-th and i-1-th iterations; δ(i) represents the iteration termination function value calculated during the i-th iteration.

[0073] The satellite-borne laser radar calibration method based on background noise iteration is as follows:

[0074] The night calibration module is specifically as follows:

[0075] For the nighttime detection signal of spaceborne lidar, the ideal attenuation backscattering coefficient is constructed based on the atmospheric scattering model;

[0076] A robust linear fit is performed iteratively on the detection signal and the ideal attenuated backscatter coefficient, and the calibration coefficient and background noise are estimated at the same time.

[0077] Combining the obtained background noise with the calibration coefficient, the detection signal is used to calculate the night attenuation backscatter coefficient to complete the night calibration;

[0078] The daytime calibration module is specifically as follows:

[0079] For the daytime detection signals of spaceborne lidar, a clean atmosphere calibration conversion area is constructed;

[0080] Based on the day-night clean atmosphere calibration conversion area, the day-night attenuation scattering ratio is calculated by the uncalibrated signal during the day and the attenuated backscattering coefficient obtained by night calibration, and then the latitude-related calibration conversion coefficient is obtained. The daytime calibration coefficient and background noise are iteratively calculated according to the calibration coefficient of the adjacent night scene.

[0081] The detection signal is calculated based on the background noise and the calibration coefficient to calculate the daytime attenuation backscatter coefficient and complete the daytime calibration.

[0082] The present invention also proposes a satellite-borne laser radar calibration system based on background noise iteration, comprising:

[0083] The night detection signal calibration module is used to construct the ideal background atmosphere total attenuation backscatter coefficient, the linear relationship between the ideal background atmosphere total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the unknown quantities in the calibration of background noise and calibration coefficient, thereby realizing the calibration of the satellite-borne laser radar night detection signal;

[0084] The clean atmosphere calibration conversion region construction module is used to set each calibration conversion region under certain environmental factors during daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, and perform simple layer detection on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion regions where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion regions are retained;

[0085] The daytime detection signal calibration module is used to calculate the nighttime attenuation scattering ratio and the daytime attenuation scattering ratio for each clean atmosphere calibration conversion area, and then calculate the calibration conversion coefficient. The calibration conversion coefficient is applied to the average calibration coefficient of the adjacent previous nighttime scene to obtain the unknown calibration coefficient in the daytime scene. The linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism is used to solve the unknown background noise. The calibration coefficient and the unknown background noise are iteratively updated to finally realize the calibration of the daytime detection signal of the spaceborne lidar.

[0086] The advantage of the present invention is that the present invention can realize the calibration of satellite-borne laser radar detection signals during the day and at night without collecting background noise in extremely high altitude areas. Compared with mainstream methods such as the official calibration of CALIPSO, the satellite-borne laser radar calibration method based on background noise iteration proposed in this patent no longer relies on the background noise of detection signal collection in extremely high altitude areas. Therefore, it avoids the stringent requirements for data storage and data transmission, and also avoids the problem of excessive background noise subtraction when the detection range of satellite-borne laser radar is low using mainstream methods, thereby improving the accuracy of calibration. These improvements enable the new method to enable satellite-borne laser radar to operate effectively under detection conditions at different altitudes, expanding its scope of application. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 : A flow chart of a method according to an embodiment of the present invention;

[0088] Figure 2 : A schematic diagram of the clean air calibration conversion area construction according to an embodiment of the present invention;

[0089] Figure 3 : Flow chart of iterative calculation of background noise of daytime detection signal and calibration coefficient according to an embodiment of the present invention;

[0090] Figure 4 : Graph showing calibration results of satellite-borne laser radar nighttime detection signals according to an embodiment of the present invention;

[0091] Figure 5 : Graph showing calibration results of satellite-borne laser radar daytime detection signals according to an embodiment of the present invention. DETAILED DESCRIPTION

[0092] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0093] In specific implementation, the method proposed in the technical solution of the present invention can be implemented by technical personnel in this field using computer software technology to realize automatic operation process. System devices for implementing the method, such as computer-readable storage media storing the corresponding computer program of the technical solution of the present invention and computer equipment running the corresponding computer program, should also be within the scope of protection of the present invention.

[0094] The present invention takes China's spaceborne aerosol and cloud high spectral resolution lidar ACDL as an example, which has fixed system hardware parameters.

[0095] Combine the following Figures 1 to 5The specific implementation mode of the present invention is a satellite-borne laser radar calibration method and system based on background noise iteration.

[0096] like Figure 1 As shown, a specific implementation of the method of the present invention is a satellite-borne laser radar calibration method based on background noise iteration, comprising the following steps:

[0097] Step 1: Construct the ideal background atmospheric total attenuation backscatter coefficient, the linear relationship between the ideal background atmospheric total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the two unknown quantities in the calibration, namely the background noise and the calibration coefficient, so as to realize the calibration of the night detection signal of the space-borne lidar;

[0098] According to the atmospheric scattering model described in step 1, the attenuated backscattering coefficient profile under the clean atmospheric state is constructed. The specific calculation method of the attenuated backscattering coefficient of clean atmospheric molecules is:

[0099]

[0100] Among them, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The clean atmospheric molecule attenuation backscattering coefficient; β m (z,k p ) represents the backscattering coefficient of atmospheric molecules; and Represents the two-way transmittance of atmospheric molecules and ozone molecules respectively; z is the altitude of the detection point, k p is the profile index number;

[0101] The attenuation-scattering ratio is defined as follows:

[0102]

[0103] Among them, R′(z,k p ) represents the altitude of the detection point is z and the profile index is k p The attenuation scattering ratio, β′ total (z,k p ) represents the altitude of the detection point is z and the profile index is k p The attenuated backscattering coefficient contributed by the clean atmospheric molecules and particles; β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; The altitude of the detection point is z and the profile index is k p The bidirectional transmittance of particles; z is the altitude of the detection point, in km; k p is the profile index number;

[0104] The ideal background atmosphere total attenuation backscattering coefficient profile of the calibration area is as follows:

[0105] β′ ideal (z,k p ) = R′(z,k p )·β′ m (z,k p ) = β(z, k p )·T 2 (z,k p )

[0106] where β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The total attenuation backscattering coefficient of the ideal background atmosphere; R′(z, k p ) represents the altitude of the detection point is z and the profile index number is k p The attenuation scattering ratio, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The atmospheric molecule attenuation backscattering coefficient profile constructed by the atmospheric scattering model; β(z, k p ) and T 2 (z,k p ) represent the total backscattering coefficient and the two-way transmittance, respectively, which are contributed by the clean atmospheric molecules and clouds and aerosol particles. The formula is as follows:

[0107] β(z,k p )=β m (z,k p )+β p (z,k p )

[0108]

[0109] Among them, β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; and They represent the altitude of the detection point, z, and the profile index number, k. p The bidirectional transmittance of atmospheric molecules, ozone molecules, and cloud and aerosol particles.

[0110] The original detection signal received by the laser radar is defined as:

[0111]

[0112] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at is contributed by the solar background noise and the dark current noise; r is the distance between the satellite-borne lidar and the detection point, in km.

[0113] The relationship between the original detection signal and the ideal total attenuation backscatter coefficient is constructed as shown in the definition of the original detection signal received by the space-borne lidar, which is a linear relationship of the form y = k x + b, where y = P (z, k p ), k and b are the calibration coefficients C(k p ) and background noise N Back (k p );

[0114] Therefore, a calibration area can be selected, and the corresponding detection point height set is z Cali = 34-38 km, for a specific profile index number k p , will belong to Z Cali P(z, k) at different heights z p )and As input, the profile is subjected to robust linear fitting iteration according to the linear relationship defined by the original detection signal received by the laser radar, and finally the calibration coefficient C(k p ) and background noise N Back (k p ) and output;

[0115] Based on this principle, we can avoid collecting background noise based on extremely high altitude detection signals, as in the official CALIPSO calibration method, and complete the iterative estimation of the fitting of background noise and calibration coefficients at lower altitudes.

[0116]

[0117] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali =34-38km represents the height set of detection points corresponding to the selected calibration area;

[0118] The robust linear fitting iteratively estimates the calibration coefficient and background noise as follows:

[0119] Step 1.1: First, perform horizontal averaging on the detection signal and the ideal total attenuation backscatter coefficient to improve the signal-to-noise ratio and obtain the average profile under a certain range of horizontal resolution;

[0120] The horizontal resolution of the certain range is 5 km;

[0121] Step 1.2: For each average profile, the regression coefficient is iteratively calculated based on the robust linear fitting method of M estimation. When the regression coefficient converges, the estimated calibration coefficient C(k p ) and background noise N Back (k p ), where C(k p ) and N Back (k p ) represent the profile index number k p Calibration coefficients and background noise at ;

[0122] Step 1.3: Use a certain two-dimensional space-time sliding window to calibrate the coefficients C(k p) to perform sliding average;

[0123] The certain two-dimensional space-time sliding window spans 5 adjacent scenes and has 125 smooth contour lines in the scene;

[0124] Step 1.4: For the calibration coefficient C(k obtained after smoothing in step 1.3 p ), iteratively update the background noise N Back (k p ), the calculation formula is as follows:

[0125]

[0126] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the set of detection point heights corresponding to the selected calibration area; angle brackets indicate the heights belonging to z Cali =The average value of the calculated values ​​obtained at different altitudes z of 34-38km.

[0127] The calibration coefficient C(k p ) and background noise N Back (k p ) is applied to the detection signal of the space-borne laser radar, and the background noise subtraction, range gain correction and radiation calibration process are completed at the same time, and the attenuated backscatter coefficient profile β′(z, k p ), as follows:

[0128]

[0129] Among them, β′(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k pThe attenuated backscatter coefficient obtained after calibration at ; P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at ; r is the distance between the satellite-borne laser radar and the detection point, in km;

[0130] Step 2: During daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, each calibration conversion area under certain environmental factors is set, and a simple layer detection is performed on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion areas where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion areas are retained;

[0131] like Figure 2 As shown, step 2 sets each calibration conversion area under certain environmental factors, as follows

[0132] Select a certain temperature and a certain height at the bottom of the stratosphere to construct multiple calibration conversion areas in the horizontal direction. Each calibration conversion area extends a certain distance in the horizontal direction along the direction of the satellite-borne lidar scene.

[0133] The certain temperature and certain height are 400K potential temperature line and 2-6km respectively;

[0134] The certain distance is 200km;

[0135] Step 3: For each clean atmosphere calibration conversion area, calculate the night attenuation scattering ratio and the day attenuation scattering ratio respectively, and then calculate the calibration conversion coefficient. Apply the calibration conversion coefficient to the average calibration coefficient of the adjacent previous night scene to obtain the unknown calibration coefficient in the daytime scene. Then use the linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism to solve the unknown background noise. Iteratively update the calibration coefficient and the unknown background noise, and finally realize the calibration of the daytime detection signal of the space-borne lidar.

[0136] Step 3: Calculate the nighttime attenuation scattering ratio The details are as follows:

[0137]

[0138] Calculate the daytime attenuation scattering ratio as described in step 3 The details are as follows:

[0139]

[0140] Among them, β night (z,k p ) represents the altitude z and the profile index number k p The night attenuated backscatter coefficient obtained by calibrating the space-borne lidar at β′ m，day (z,k p ) and β′ m，night (z,k p ) represent the altitude z and the profile index k respectively p The attenuation backscattering coefficients of atmospheric molecules during the day and night calculated according to the atmospheric scattering model, represents the mean value of the calibration coefficients recorded in the previous night scene adjacent to the daytime scene of the spaceborne lidar; the angle brackets indicate all the profiles k located in the calibration conversion region p The average value of all heights z; X day (z,k p ) represents the altitude z and the profile index number k p The uncalibrated daytime signal at which background noise subtraction and gain distance correction have been completed is defined as follows:

[0141]

[0142] Among them, X day (z,k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The background noise at P is deducted and the gain distance corrected daytime uncalibrated signal; day (z,k p ) represents the daytime detection of the satellite-borne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and they are all known quantities; N Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km;

[0143] The calibration conversion coefficient W is obtained according to the obtained day and night attenuation scattering ratios of different calibration conversion areas. The specific calculation formula is as follows:

[0144]

[0145] Where W represents the calibration conversion factor; and Represents the average attenuation-scattering ratio of the clean atmosphere calibration conversion area at night and day corresponding to similar latitudes.

[0146] In order to obtain a stable and accurate calibration conversion coefficient W, the calibration conversion coefficients accumulated in a period of time before the time point corresponding to the daytime scene currently detected by the space-borne laser radar are averaged to obtain an estimated value of the calibration conversion coefficient of the scene;

[0147] Taking into account the spatial differences in the detection environment, the estimated value of W is interpolated into a latitude-related sequence. Since the detection scene of the spaceborne laser radar in the polar region will pass through the same latitude many times and is difficult to distinguish, the estimated value of W after latitude interpolation will be reported as the number of seconds W(t) when the scene starts to be detected, so as to distinguish the different calibration conversion coefficients reported at the same latitude in the polar region; where t represents the time the spaceborne laser radar has spent on the detection of the scene;

[0148] Multiply the time-varying and space-time calibration conversion coefficient W(t) by the average of the calibration coefficients of the previous night scene adjacent to the day scene. The number of profile indexes corresponding to different times t in this scenario is k p If there is no calibration conversion area that meets the requirements or the area is not covered by the night scene, the calibration coefficients of the adjacent night scenes are used for time series interpolation, as follows:

[0149]

[0150] Among them C day (k p ) represents the profile index number k obtained by the calibration conversion factor p The daytime calibration coefficient; W(t) represents the calibration conversion coefficient obtained after latitude interpolation that changes with the detection time t; Represents the average value of the calibration coefficients of the previous nighttime scenes adjacent to this daytime scene.

[0151] The background noise N is updated according to the relationship between the calibration coefficient and the background noise. Back (k p ), the specific calculation formula is as follows;

[0152]

[0153] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z.

[0154] like Figure 3 As shown, due to the daytime calibration factor C day (k p ) involves the calculation of the uncalibrated signal during the day, and it is necessary to assume in advance that the background noise N Back (k p ), and then update the estimated value of the background noise according to the calculated calibration coefficient;

[0155] like Figure 3 As shown, the calculation of calibration coefficients and background noise in the daytime detection signal calibration is essentially an iterative update process;

[0156] First, the initial estimate of the background noise is estimated using the detection signal of the space-borne lidar and the atmospheric backscattering model. In order to obtain the optimal background noise, the iterative termination discriminant function δ(i) is defined. When the termination discriminant function is less than a certain threshold, the iteration is terminated and C is obtained. i (k p ) and Used for calculation of daytime attenuation backscatter coefficient, daytime calibration is completed;

[0157] The certain threshold is 0.001;

[0158] The initial estimate of the background noise The calculation formula of the iteration termination discriminant function δ(i) is as follows:

[0159]

[0160]

[0161] Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; Represents the number of profile indexes as k p The initial estimate of the background noise at ; r is the distance between the satellite-borne laser radar and the detection point, in km; z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z; and represents the updated value of the background noise in the entire detection scene during the i-th and i-1-th iterations; δ(i) represents the iteration termination function value calculated during the i-th iteration.

[0162] The satellite-borne laser radar calibration method based on background noise iteration is as follows:

[0163] The night calibration module is specifically as follows:

[0164] For the nighttime detection signal of spaceborne lidar, the ideal attenuation backscattering coefficient is constructed based on the atmospheric scattering model;

[0165] A robust linear fitting iteration is performed on the detection signal and the ideal attenuated backscatter coefficient, and the calibration coefficient and background noise are calculated at the same time.

[0166] Combining the obtained background noise with the calibration coefficient, the detection signal is used to calculate the night attenuation backscatter coefficient to complete the night calibration;

[0167] The daytime calibration module is specifically as follows:

[0168] For the daytime detection signals of spaceborne lidar, a clean atmosphere calibration conversion area is constructed;

[0169] Based on the day-night clean atmosphere calibration conversion area, the day-night attenuation scattering ratio is calculated by the uncalibrated signal during the day and the attenuated backscattering coefficient obtained by night calibration, and then the latitude-related calibration conversion coefficient is obtained. The iterative calculation of the daytime calibration coefficient and background noise is completed according to the night scene calibration coefficient.

[0170] Combining the obtained background noise with the calibration coefficient, the detection signal is used to calculate the daytime attenuated backscatter coefficient to complete the daytime calibration.

[0171] The embodiment of the system of the present invention is a satellite-borne laser radar calibration system based on background noise iteration, comprising:

[0172] The night detection signal calibration module is used to construct the ideal background atmosphere total attenuation backscatter coefficient, the linear relationship between the ideal background atmosphere total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the unknown quantities in the calibration of background noise and calibration coefficient, thereby realizing the calibration of the satellite-borne laser radar night detection signal;

[0173] The clean atmosphere calibration conversion region construction module is used to set each calibration conversion region under certain environmental factors during daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, and perform simple layer detection on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion regions where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion regions are retained;

[0174] The daytime detection signal calibration module is used to calculate the nighttime attenuation scattering ratio and the daytime attenuation scattering ratio for each clean atmosphere calibration conversion area, and then calculate the calibration conversion coefficient. The calibration conversion coefficient is applied to the average calibration coefficient of the adjacent previous nighttime scene to obtain the unknown calibration coefficient in the daytime scene. The linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism is used to solve the unknown background noise. The calibration coefficient and the unknown background noise are iteratively updated to finally realize the calibration of the daytime detection signal of the spaceborne lidar.

[0175] According to the above steps, the ACDL day and night detection data on July 15, 2022 are calibrated to obtain the calibrated day and night attenuated backscattering coefficients. The attenuated scattering ratio (ASR) test is often used to test the accuracy of the attenuated backscattering coefficients obtained by the calibration of the spaceborne lidar. The attenuated scattering ratio test compares the attenuated backscattering coefficient obtained by calibration with the ideal atmospheric molecule attenuated backscattering coefficient derived from the atmospheric scattering model. For clean atmospheric areas, due to the absence of clouds and aerosol layers, the actual calibrated attenuated backscattering coefficient should be consistent with the result derived from the model, so the ideal value of the attenuated scattering ratio should be 1. Figure 4 As shown, sub-figure (a) is a nighttime detection signal scene diagram received by the satellite sensor, with a horizontal resolution of 5km and a vertical resolution of 24m; (b) is the scene horizontal average attenuation scattering ratio profile calculated by using the mainstream method, namely the official calibration method of CALIPSO; (c) is the scene horizontal average attenuation scattering ratio profile calculated by using the method of the present invention. Since the highest detection area of ​​the detection signal in the ACDL nighttime detection scene shown is about 40km, which is much lower than CALIPSO, the backscattering signal after the interaction between the laser photons and the atmospheric molecules is difficult to ignore. If the mean value of the detection signal of about 40km is directly used as the background noise according to the mainstream calibration method, the background noise will be overestimated, which will cause the background noise to be over-subtracted. The nighttime attenuation backscattering coefficient obtained by calibration is obviously smaller or even negative at high altitudes, and the corresponding attenuation scattering ratio is also obviously smaller and abnormal. The theoretical results calculated by the present invention correspond to the scene average attenuation scattering ratio in the clean atmosphere area, which is very close to 1. This shows that the attenuation backscattering coefficient obtained by calibration in the clean atmosphere area is in good agreement with the result calculated by the atmospheric scattering model, which can effectively avoid the problem of excessive subtraction caused by the background and effectively make up for the limitations of the official calibration algorithm of CALIPSO at night. Figure 5As shown, (a) is a scene diagram of the daytime detection signal received by the satellite sensor, with a horizontal resolution of 5km and a vertical resolution of 24m; (b) is the scene horizontal average attenuation scattering ratio profile calculated by using the mainstream method, namely the CALIPSO official calibration method; (c) is the scene horizontal average attenuation scattering ratio profile calculated by using the new method of the present invention. The maximum detection altitude of ACDL for daytime detection is limited to about 40km. The calibration result obtained by using the mainstream daytime calibration algorithm is obviously too small above 25km, and the phenomenon of excessive background noise deduction also occurs. In contrast, the theoretical result calculated by the present invention jitters around 1 in the average attenuation scattering ratio of the clean atmosphere region, indicating that the attenuation backscattering coefficient obtained by calibration in the clean atmosphere region is in good agreement with the result calculated by the atmospheric scattering model, which effectively makes up for the limitations of the CALIPSO daytime official calibration algorithm. Therefore, the present invention can accurately estimate the background noise and calibration coefficient of the day and night detection signal of a given satellite-borne laser radar system, and effectively complete the calibration task of the satellite-borne laser radar in actual engineering measurement.

[0176] It should be understood that parts not elaborated in detail in this specification belong to the prior art.

[0177] It should be understood that the above description of the embodiments is relatively detailed and cannot be regarded as limiting the scope of patent protection of the present invention. Under the enlightenment of the present invention, ordinary technicians in this field can also make substitutions or modifications without departing from the scope of protection of the claims of the present invention, which all fall within the scope of protection of the present invention. The scope of protection requested by the present invention shall be based on the attached claims.

Claims

1. A space-borne laser radar calibration method based on background noise iteration, characterized in that: The following steps are involved: Step 1: Construct the ideal background atmospheric total attenuation backscatter coefficient, the linear relationship between the ideal background atmospheric total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the two unknown quantities in the calibration, namely the background noise and the calibration coefficient, so as to realize the calibration of the night detection signal of the space-borne lidar; Step 2: During daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, each calibration conversion area under certain environmental factors is set, and a simple layer detection is performed on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion areas where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion areas are retained; Step 3: For each clean atmosphere calibration conversion area, calculate the night attenuation scattering ratio and the daytime attenuation scattering ratio respectively, and then calculate the calibration conversion coefficient. Apply the calibration conversion coefficient to the average calibration coefficient of the adjacent previous night scene to obtain the unknown calibration coefficient in the daytime scene. Then use the linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism to solve the unknown background noise. Iteratively update the calibration coefficient and the unknown background noise to finally realize the calibration of the daytime detection signal of the spaceborne lidar.

2. The satellite-borne laser radar calibration method based on background noise iteration according to claim 1 is characterized in that: According to the atmospheric scattering model described in step 1, the attenuated backscattering coefficient profile under the clean atmospheric state is constructed. The specific calculation method of the attenuated backscattering coefficient of clean atmospheric molecules is: Among them, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The clean atmospheric molecule attenuation backscattering coefficient; β m (z,k p ) represents the backscattering coefficient of atmospheric molecules; and Represents the two-way transmittance of atmospheric molecules and ozone molecules respectively; z is the altitude of the detection point, k p is the profile index number; The attenuation-scattering ratio is defined as follows: Among them, R′(z, k p ) represents the altitude of the detection point is z and the profile index is k p The attenuation scattering ratio, β′ total (z,k p ) represents the altitude of the detection point is z and the profile index is k p The attenuated backscattering coefficient contributed by the clean atmospheric molecules and particles; β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; The altitude of the detection point is z and the profile index is k p The bidirectional transmittance of particles; z is the altitude of the detection point, in km; k p is the profile index number.

3. The satellite-borne laser radar calibration method based on background noise iteration according to claim 2 is characterized in that: The ideal background atmosphere total attenuation backscattering coefficient profile of the calibration area is as follows: β′ ideal (z,k p )=R′(z,k p )·β′ m (z,k p )=β(z,k p ) T 2 (z,k p ) Among them, β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The total attenuation backscattering coefficient of the ideal background atmosphere; R′(z, k p ) represents the altitude of the detection point is z and the profile index number is k p The attenuation scattering ratio, β′ m (z,k p ) represents the altitude of the detection point is z and the profile index is k p The atmospheric molecule attenuation backscattering coefficient profile constructed by the atmospheric scattering model; β(z, k p ) and T 2 (z,k p ) represent the total backscattering coefficient and the two-way transmittance, respectively, which are contributed by the clean atmospheric molecules and clouds and aerosol particles. The formula is as follows: β(z,k p )=β m (z,k p )+β p (z,k p ) Among them, β p (z,k p ) and β m (z,k p ) represent the altitude of the detection point, z, and the profile index number, k. p The backscattering coefficients of particles and atmospheric molecules; and They represent the altitude of the detection point, z, and the profile index number, k. p The bidirectional transmittance of atmospheric molecules, ozone molecules, and cloud and aerosol particles.

4. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 3 is characterized in that: The original detection signal received by the satellite-borne laser radar is defined as: Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at is contributed by the solar background noise and the dark current noise; r is the distance between the satellite-borne lidar and the detection point, in km.

5. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 4, characterized in that: The relationship between the original detection signal and the ideal total attenuation backscatter coefficient is constructed as shown in the definition of the original detection signal received by the space-borne laser radar, which is a linear relationship of the form y = k x + b, where k and b are the calibration coefficients C(k p ) and background noise N Back (k p ); Therefore, a calibration area can be selected, and the corresponding detection point height set is Z Cali , for a specific profile index number k p , will belong to Z Cali P(z, k) at different heights z p )and As input, the profile is subjected to robust linear fitting iteration according to the linear relationship defined by the original detection signal received by the laser radar, and finally the calibration coefficient C(k p ) and background noise N Back (k p ) and output; Based on this principle, we can avoid collecting background noise based on extremely high altitude detection signals, such as the official calibration method of CALIPSO, and complete the fitting estimation of background noise and calibration coefficients at a lower altitude. Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; β ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p Calibration coefficient at Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the set of detection point heights corresponding to the selected calibration area.

6. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 5, characterized in that: The robust linear fitting iteratively estimates the calibration coefficient and background noise as follows: Step 1.1: First, perform horizontal averaging on the detection signal and the ideal total attenuation backscatter coefficient to improve the signal-to-noise ratio and obtain the average profile under a certain range of horizontal resolution; Step 1.2: For each average profile, the regression coefficient is iteratively calculated based on the robust linear fitting method of M estimation. When the regression coefficient converges, the estimated calibration coefficient C(k p ) and background noise N Back (k p ), where C(k p ) and N Back (k p ) represent the profile index number k p Calibration coefficients and background noise at ; Step 1.3: Use a two-dimensional space-time sliding window to perform sliding average on the calibration coefficients C(kp) of different scenes and different profiles; Step 1.4: For the calibration coefficient C(kp) obtained after smoothing in step 1.3, iteratively update the background noise N Back (k p ), the calculation formula is as follows: Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z; The calibration coefficient C(k p ) and background noise N Back (k p ) is applied to the detection signal of the space-borne laser radar, and the background noise subtraction, range gain correction and radiation calibration process are completed at the same time, and the attenuated backscatter coefficient profile β′(z, kp) is calculated as follows: Among them, β′(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The attenuated backscatter coefficient obtained after calibration at ; P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, respectively, and are all known quantities; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km.

7. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 6, characterized in that: Step 2 sets each calibration conversion area under certain environmental factors, as follows A certain temperature and a certain height at the bottom of the stratosphere are selected to construct multiple calibration conversion areas along the horizontal direction. The horizontal direction of each calibration conversion area extends a certain distance along the direction of the satellite-borne lidar scene.

8. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 7, characterized in that: Step 3: Calculate the nighttime attenuation scattering ratio The details are as follows: Calculate the daytime attenuation scattering ratio as described in step 3 The details are as follows: Among them, β night (z,k p ) represents the altitude z and the profile index number k p The night attenuated backscatter coefficient obtained by calibrating the space-borne lidar at β′ m，day (z,k p ) and β′ m，night (z,k p ) represent the altitude z and the profile index k respectively p The attenuation backscattering coefficients of atmospheric molecules during the day and night calculated according to the atmospheric scattering model are represents the mean value of the calibration coefficients recorded in the previous night scene adjacent to the daytime scene of the spaceborne lidar; the angle brackets indicate the average value at all heights z of all profiles kp in the calibration conversion area; X day (z,k p ) represents the altitude z and the profile index number k p The uncalibrated daytime signal at which background noise subtraction and gain distance correction have been completed is defined as follows: Among them, X day (z,k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The background noise at P is deducted and the gain distance corrected daytime uncalibrated signal; day (z,k p ) represents the daytime detection of the satellite-borne laser radar at an altitude of z and a profile index number of k p The original detection signal at K, E(k p ) and G A represent the system constant, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and they are all known quantities; N Back (k p ) represents the profile index number k p The background noise at the location; r is the distance between the satellite-borne laser radar and the detection point, in km; The calibration conversion coefficient W is obtained according to the obtained day and night attenuation scattering ratios of different calibration conversion areas. The specific calculation formula is as follows: Where W represents the calibration conversion factor; and Represents the average attenuation-scattering ratio of the clean atmosphere calibration conversion area at night and day corresponding to similar latitudes; In order to obtain a stable and accurate calibration conversion coefficient W, the calibration conversion coefficients accumulated in a period of time before the time point corresponding to the daytime scene currently detected by the space-borne laser radar are averaged to obtain an estimated value of the calibration conversion coefficient of the scene; Taking into account the spatial differences in the detection environment, the estimated value of W is interpolated into a latitude-related sequence. Since the detection scene of the spaceborne laser radar in the polar region will pass through the same latitude many times and is difficult to distinguish, the estimated value of W after latitude interpolation will be reported as the number of seconds W(t) when the scene starts to be detected, so as to distinguish the different calibration conversion coefficients reported at the same latitude in the polar region; where t represents the time the spaceborne laser radar has spent on the detection of the scene; Multiply the time-varying and space-time calibration conversion coefficient W(t) by the average of the calibration coefficients of the previous night scene adjacent to the daytime scene The number of profile indexes corresponding to different times t in this scenario is k p If there is no calibration conversion area that meets the requirements or the area is not covered by the night scene, the calibration coefficients of the adjacent night scenes are used for time series interpolation, as follows: Among them C day (k p ) represents the profile index number k obtained by the calibration conversion factor p The daytime calibration coefficient; W(t) represents the calibration conversion coefficient obtained after latitude interpolation that changes with the detection time t; Represents the average value of the calibration coefficients of the previous nighttime scenes adjacent to this daytime scene.

9. The method for calibrating a satellite-borne laser radar based on background noise iteration according to claim 8, characterized in that: The background noise N is updated according to the relationship between the calibration coefficient and the background noise. Back (k p ), the specific calculation formula is as follows; Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; C(k p ) represents the profile index number k p The calibration coefficient after smoothing; N Back (k p ) represents the profile index number k p The updated value of the background noise at; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z; Since the daytime calibration factor C day (k p ) involves the calculation of the uncalibrated signal during the day, and it is necessary to assume in advance that the background noise N Back (k p ), and then update the estimated value of the background noise according to the calculated calibration coefficient; The calculation of calibration coefficients and background noise in daytime detection signal calibration is essentially an iterative update process; First, the initial estimate of the background noise is calculated using the detection signal of the satellite-borne lidar and the atmospheric backscattering model. In order to obtain the optimal background noise, the iterative termination discriminant function δ(i) is defined. When the termination discriminant function is less than the threshold, the iteration is terminated and C is obtained. i (k p ) and Used for calculation of daytime attenuation backscatter coefficient, daytime calibration is completed; the initial estimate of the background noise The calculation formula of the iteration termination discriminant function δ(i) is as follows: Among them, P(z, k p ) represents the spaceborne laser radar at an altitude of z and a profile index of k p The original detection signal at K, E(k p ) and G A They represent the system constants, laser pulse energy and amplifier gain coefficient of the spaceborne laser radar, and are all known quantities; β′ ideal (z,k p ) represents the altitude of the detection point is z and the profile index number is k p The ideal background atmosphere total attenuation backscattering coefficient; Represents the number of profile indexes as k p The initial estimate of the background noise at ; r is the distance between the satellite-borne laser radar and the detection point, in km; Z Cali Represents the height set of the detection points corresponding to the selected calibration area; the angle brackets indicate the heights belonging to Z Cali The average value of the calculated values ​​obtained at different heights z; and represents the updated value of the background noise in the entire detection scene during the i-th and i-1-th iterations; δ(i) represents the iteration termination function value calculated during the i-th iteration; The night calibration module is specifically as follows: For the nighttime detection signal of spaceborne lidar, the ideal attenuation backscattering coefficient is constructed based on the atmospheric scattering model; A robust linear fitting iteration is performed on the detection signal and the ideal attenuated backscatter coefficient to estimate the calibration coefficient and background noise. Combining the obtained background noise with the calibration coefficient, the detection signal is used to calculate the night attenuation backscatter coefficient to complete the night calibration; The daytime calibration module is specifically as follows: For the daytime detection signals of spaceborne lidar, a clean atmosphere calibration conversion area is constructed; Based on the day-night clean atmosphere calibration conversion area, the day-night attenuation scattering ratio is calculated by the uncalibrated signal during the day and the attenuated backscattering coefficient obtained by night calibration, and then the latitude-related calibration conversion coefficient is obtained. The daytime calibration coefficient and background noise are iteratively calculated according to the calibration coefficient of the adjacent night scene. The detection signal is calculated based on the background noise and the calibration coefficient to calculate the daytime attenuation backscatter coefficient and complete the daytime calibration.

10. A satellite-borne laser radar calibration system based on background noise iteration, characterized in that: include: The night detection signal calibration module is used to construct the ideal background atmosphere total attenuation backscatter coefficient, the linear relationship between the ideal background atmosphere total attenuation backscatter coefficient and the original detection signal in the detection mechanism, and perform robust linear fitting iteration to obtain the unknown quantities in the calibration of background noise and calibration coefficient, thereby realizing the calibration of the satellite-borne laser radar night detection signal; The clean atmosphere calibration conversion region construction module is used to set each calibration conversion region under certain environmental factors during daytime detection of the satellite-borne lidar, combined with the strategy of the official daytime calibration method of CALIPSO, and perform simple layer detection on the attenuated backscatter coefficient obtained by nighttime calibration and the uncalibrated signal during the daytime; the calibration conversion regions where cloud and aerosol layers are detected are eliminated, and the remaining clean atmosphere calibration conversion regions are retained; The daytime detection signal calibration module is used to calculate the nighttime attenuation scattering ratio and the daytime attenuation scattering ratio for each clean atmosphere calibration conversion area, and then calculate the calibration conversion coefficient. The calibration conversion coefficient is applied to the average calibration coefficient of the adjacent previous nighttime scene to obtain the unknown calibration coefficient in the daytime scene. The linear relationship between the total attenuation backscattering coefficient of the ideal background atmosphere and the original detection signal in the detection mechanism is used to solve the unknown background noise. The calibration coefficient and the unknown background noise are iteratively updated to finally realize the calibration of the daytime detection signal of the satellite-borne laser radar.

Images