A CO2 column concentration inversion method based on active and passive combined observations

Abstract

The present application discloses a method for inverting carbon dioxide column concentration by active and passive combined observation, which relates to the field of meteorological detection technology. The method compiles an active lidar observation data set based on month information and latitude and longitude information, calculates differential optical depth based on a priori active lidar observation data and radar equation, constructs a carbon dioxide absorption cross-section lookup table based on the line-by-line integration principle, combines a priori profile data and differential optical depth, calculates a correction coefficient matrix for the prior profile and corrected prior profile data, and calculates carbon dioxide column concentration based on passive observation data using the corrected prior profile data as a constraint based on the optimization algorithm principle. As a result, the inversion method provided by the present application includes both active lidar observations and passive observation data, effectively combining active and passive observations, providing more accurate prior profile information, and improving the accuracy of inversion.

Description

Technical Field

[0001] The present application relates to the field of meteorological detection technology, and in particular to a method for inverting carbon dioxide column concentration by active and passive combined observation. Background Art

[0002] As a major greenhouse gas, carbon dioxide plays a crucial role in global warming. Therefore, accurately assessing carbon dioxide is crucial for improving our predictions of the carbon cycle and climate change. There are two common methods for inferring carbon dioxide: empirical and physical inversion.

[0003] Empirical algorithms invert carbon dioxide column concentrations by establishing a statistical regression model between satellite observations and the target parameters to be inverted. This inversion does not involve radiation transfer processes and has the advantages of simple inversion principles and high computational efficiency. Typical empirical algorithms include statistical regression and neural network algorithms. Currently, such empirical algorithms are used in near-infrared nadir sounding sensors. However, the reliability of the inversion results of such algorithms is primarily determined by the quality of the sample set constructed from the regression model. When the total number of samples in the sample set is small or unrepresentative, the inversion results will deviate from the true value. These drawbacks limit the application of existing empirical algorithms.

[0004] Physical inversion algorithms employ a forward radiation transfer model to solve for the target parameters to be inverted, which has clear physical meaning. Specifically, physical inversion methods include optimization algorithms and DOAS algorithms. However, the inversion results must be constrained by the CO2 profile as a priori. The accuracy of this priori profile is crucial for the accuracy of the final CO2 column concentration. If the priori information deviates from the actual CO2 distribution, the inversion results will be poor.

[0005] Therefore, in view of the defects of the above-mentioned inversion method, a new inversion method is needed to solve the problems existing in the above-mentioned existing inversion method. Summary of the Invention

[0006] The purpose of this application is to provide a method for inverting carbon dioxide column concentration by active and passive combined observation, which can effectively combine active and passive observations, have more accurate prior profile information, make up for the defect of inaccurate prior profile information of existing physical inversion algorithms, and solve the problem of low inversion accuracy of existing empirical algorithms.

[0007] To achieve the above objectives, this application provides the following solutions.

[0008] The present application provides a method for inverting carbon dioxide column concentration by active and passive combined observation, which includes the following steps.

[0009] Obtain an active lidar observation dataset, a profile dataset, and passive observation information; the active lidar observation dataset is a dataset compiled based on month information and latitude and longitude information; the active lidar observation dataset includes online active lidar monitoring data, offline active lidar monitoring data, online active lidar echo data, and offline active lidar echo data; the profile dataset includes pressure profile and carbon dioxide profile data; and the passive observation information includes OCO-2 observation data.

[0010] According to the passive observation information, data of the corresponding grid is extracted from the profile data set as prior profile data.

[0011] According to the passive observation information, data of the corresponding grid is extracted from the active lidar observation data set as prior active lidar observation data.

[0012] The differential optical depth is calculated based on the radar equation according to the prior active lidar observation data.

[0013] According to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed.

[0014] A correction coefficient matrix of the priori profile and corrected priori profile data are calculated based on the priori profile data, the differential optical thickness, and the carbon dioxide absorption cross-section lookup table.

[0015] According to the principle of the optimization algorithm, the corrected prior profile data is used as a constraint and based on the passive observation information, the carbon dioxide column concentration is calculated.

[0016] Optionally, an active lidar observation dataset is obtained, specifically including the following contents.

[0017] Acquire active lidar observation data.

[0018] The active lidar observation data are screened, the active lidar observation data above the clouds are eliminated, and only the active lidar observation data of the clear sky are retained to obtain an active lidar observation data set.

[0019] Optionally, the calculation formula for the prior active lidar observation data is as follows.

[0020] ;

[0021] in, is the extracted prior active lidar observation data; It is the online active lidar monitoring data; It is the echo data of online active lidar; Offline active lidar monitoring data; It is the offline active lidar echo data.

[0022] Optionally, the calculation formula for the differential optical thickness is as follows.

[0023] ;

[0024] in, is the differential optical thickness; Monitor the peak value of the pulse signal for the online active lidar; is the peak value of the online active lidar echo pulse signal; Monitor the peak value of the pulse signal for the offline active lidar; is the peak value of the offline active lidar echo pulse signal.

[0025] Optionally, based on the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed, specifically including: calculating the carbon dioxide absorption cross-section at different temperatures, humidity, pressures and frequencies using a line-by-line integration method at preset intervals, and considering the influence of aliasing of carbon dioxide and water vapor absorption lines to construct a four-dimensional carbon dioxide absorption cross-section lookup table.

[0026] Optionally, a correction coefficient matrix of the prior profile and the corrected prior profile data are calculated based on the prior profile data, the differential optical thickness and the carbon dioxide absorption cross-section lookup table, specifically including the following contents.

[0027] Gaussian fitting is performed on the differential optical thickness to calculate the mean and standard deviation.

[0028] The differential optical thickness is quality controlled according to the mean and the standard deviation using the 3σ principle to obtain a quality-controlled differential optical thickness.

[0029] A correction coefficient matrix of the priori profile and the corrected priori profile data are calculated based on the priori profile data, the quality-controlled differential optical depth, and the carbon dioxide absorption cross-section lookup table.

[0030] Optionally, before calculating the correction coefficient matrix of the prior profile and the corrected prior profile data, the active and passive combined observation method for inverting carbon dioxide column concentration further includes the following contents.

[0031] Calculate the Doppler shift of the prior active lidar observation data after quality control. The specific calculation formula is as follows.

[0032] ;

[0033] in, The Doppler shift of the active lidar observation data after quality control; is the original emission frequency of the active lidar; is the active lidar flight speed; is the angle between the active lidar flight speed and the transmitted / received pulse echo signal; The speed of light.

[0034] Optionally, a calculation formula for the correction coefficient matrix of the prior profile is as follows.

[0035] ;

[0036] ;

[0037] in, is the correction coefficient matrix of the prior profile; is the differential optical depth after quality control; For pressure The carbon dioxide volume mixing ratio at ; For pressure The weight function when ; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the online band; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the offline band; is the molecular mass of water vapor; is the mass of dry air molecules; is the water vapor volume mixing ratio, is the acceleration due to gravity.

[0038] Optionally, the calculation formula of the corrected prior profile data is as follows.

[0039] ;

[0040] ;

[0041] in, is the corrected prior profile data; is the extracted prior profile data; is the correction coefficient of the average prior profile; For the Correction coefficients for the a priori profiles of the lidar footprints; For the The peak value of the online active lidar echo pulse signal of each lidar footprint.

[0042] Optionally, the calculation formula for the carbon dioxide column concentration is as follows.

[0043] ;

[0044] in, For the The carbon dioxide profile of the iteration; is the corrected prior profile data; For the The Jacobian matrix of the iteration; is the error covariance matrix of the observation model; is the smoothing factor; is the error covariance matrix of the corrected prior profile data; It is passive observation data; is the forward model; For the The CO2 profile for the iteration.

[0045] According to the specific embodiments provided in this application, this application discloses the following technical effects.

[0046] This application provides a method for inverting carbon dioxide column concentration using active and passive combined observations. Based on monthly and longitude and latitude information, an active lidar observation dataset is compiled. Based on the prior active lidar observation data and the radar equation, differential optical depth (DOD) is calculated. A carbon dioxide absorption cross-section lookup table is constructed based on the line-by-line integration principle. A correction coefficient matrix for the prior profile and corrected prior profile data are calculated using the prior profile data combined with the DOD. Based on the optimization algorithm principle, the carbon dioxide column concentration is calculated based on the passive observation data using the corrected prior profile data as a constraint. This method incorporates both active lidar and passive observation data, effectively combining active and passive observations and providing more accurate prior profile information. This method overcomes the inaccurate prior profile information of existing physical inversion algorithms and addresses the low inversion accuracy of existing empirical algorithms. Furthermore, a carbon dioxide absorption cross-section lookup table is constructed based on the line-by-line integration principle, thereby accelerating the calculation of the correction coefficient matrix for the prior profile and the corrected prior profile data, improving the efficiency of carbon dioxide column concentration inversion. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 This is a diagram of the application environment of a method for inverting carbon dioxide column concentration using active and passive combined observations in one embodiment of the present application.

[0049] Figure 2 A schematic flow chart of a method for inverting carbon dioxide column concentration using active and passive combined observations provided in one embodiment of the present application.

[0050] Figure 3 This is a flowchart for inverting the carbon dioxide column concentration of a nadir sensor provided in one embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0052] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0053] Traditional physical inversion algorithms are widely used for nadir sounding data. However, they require the CO2 profile as a priori constraint to constrain the inversion results. The accuracy of this priori profile is crucial for the accuracy of the final CO2 column concentration. If the priori information deviates from the actual CO2 distribution, the inversion results will be poor. The emergence of active lidar observations can provide more effective information for CO2 inversion. However, the current CO2 column concentration inversion method has not effectively combined the two.

[0054] The reliability of empirical algorithm inversion results is primarily determined by the quality of the sample set used to construct the regression model. When the sample set is small or unrepresentative, the inversion results will deviate from the true value. These drawbacks limit the application of existing empirical algorithms.

[0055] The active and passive combined observation method for inversion of carbon dioxide column concentration provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the acquired active lidar observation data set, profile data set and passive observation information to the server 104. The active lidar observation data set is a data set obtained based on the month information and longitude and latitude information; the active lidar observation data set includes: online active lidar monitoring data, offline active lidar monitoring data, online active lidar echo data and offline active lidar echo data; the profile data set includes: pressure profile and carbon dioxide profile data; the passive observation information includes: OCO-2 observation data; after the server 104 receives the active lidar observation data set, profile data set and passive observation information, the server 104 will generate the following information according to the data set. The passive observation information is used to extract data from the corresponding grid from the profile dataset as prior profile data; based on the passive observation information, data from the corresponding grid is extracted from the active lidar observation dataset as prior active lidar observation data; differential optical depth is calculated based on the radar equation based on the prior active lidar observation data; a carbon dioxide absorption cross-section lookup table is constructed based on the line-by-line integration principle; a correction coefficient matrix for the prior profile and corrected prior profile data are calculated based on the prior profile data, the differential optical depth, and the carbon dioxide absorption cross-section lookup table; and based on the principles of an optimization algorithm, the corrected prior profile data is used as a constraint to calculate the carbon dioxide column concentration based on the passive observation information. Server 104 can provide feedback of the obtained carbon dioxide column concentration to terminal 102. In addition, in some embodiments, the carbon dioxide column concentration inversion method of active and passive joint observations can also be implemented independently by the server 104 or the terminal 102. For example, the terminal 102 can directly perform carbon dioxide column concentration inversion on the active lidar observation dataset, profile dataset and passive observation information, or the server 104 can obtain the active lidar observation dataset, profile dataset and passive observation information from the data storage system, and perform carbon dioxide column concentration inversion on the active lidar observation dataset, profile dataset and passive observation information.

[0056] The terminal 102 may be, but is not limited to, various desktop computers, laptop computers, smart phones, and tablet computers. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers, or a cloud server.

[0057] In an exemplary embodiment, Figure 2As shown, a method for inverting carbon dioxide column concentration by active and passive combined observation is provided. The method is executed by a computer device, specifically a computer device such as a terminal or a server, or a terminal and a server. In the embodiment of the present application, the method is applied to Figure 1 The server 104 in the example is used for explanation, and the steps include the following steps S1 to S7.

[0058] S1: Acquire an active lidar observation dataset, a profile dataset, and passive observation information. The active lidar observation dataset is a dataset compiled based on month information and latitude and longitude information. The active lidar observation dataset includes online active lidar monitoring data, offline active lidar monitoring data, online active lidar echo data, and offline active lidar echo data. The profile dataset includes pressure profile and carbon dioxide profile data. The passive observation information includes OCO-2 observation data.

[0059] S2: According to the passive observation information, extract the data of the corresponding grid from the profile data set as prior profile data.

[0060] S3: According to the passive observation information, extract the data of the corresponding grid from the active lidar observation data set as the prior active lidar observation data.

[0061] S4: Calculate the differential optical depth based on the radar equation according to the prior active lidar observation data.

[0062] S5: Based on the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed.

[0063] S6: Calculate a correction coefficient matrix of the priori profile and the corrected priori profile data according to the priori profile data, the differential optical thickness, and the carbon dioxide absorption cross-section lookup table.

[0064] S7: According to the optimization algorithm principle, with the corrected prior profile data as a constraint, based on the passive observation information, the carbon dioxide column concentration is calculated.

[0065] By implementing steps S1 through S7 above, the original profile dataset is corrected based on active lidar observation data. This ensures that the inversion of CO2 column concentration incorporates both active lidar and passive observation information, effectively combining active and passive observations to provide more accurate prior profile information. Furthermore, a CO2 absorption cross-section lookup table is constructed based on the line-by-line integration principle, accelerating the calculation of the correction coefficient matrix for the prior profile and the corrected prior profile data, thereby improving the efficiency of CO2 column concentration inversion.

[0066] In an exemplary embodiment, in step S1, an active lidar observation dataset is acquired, which specifically includes the following steps.

[0067] S11: Acquire active lidar observation data.

[0068] S12: Filter the active lidar observation data, remove the active lidar observation data above the clouds, and retain only the active lidar observation data of the clear sky to obtain an active lidar observation data set.

[0069] In practical applications, for example, active lidar observation data can be organized based on monthly information, with a latitude interval of 5° and a longitude interval of 30°, to obtain a monthly active lidar observation dataset with a 5° latitude grid and a 30° longitude grid. This is just an example.

[0070] Since the vertical distribution of carbon dioxide has temporal and spatial variability, based on the month, longitude and latitude information of the passive observation data, data with the same month and in the same longitude and latitude band as the passive observation data are extracted as prior profile data and prior active lidar observation data.

[0071] Among them, the extracted prior profile data is , for dimensional matrix, is the number of atmospheric layers; the formula for extracting the prior active lidar observation data is as follows.

[0072] (1);

[0073] in, is the extracted prior active lidar observation data; It is the online active lidar monitoring data; It is the echo data of online active lidar; Offline active lidar monitoring data; It is the offline active lidar echo data. for dimensional matrix; for dimensional matrix; for dimensional matrix; for dimensional matrix; is the number of laser footprints; The number of sampling points of the lidar online monitoring signal; is the number of sampling points of the online echo signal; The number of sampling points for the offline monitoring signal; The number of sampling points of the offline echo signal.

[0074] In practical applications, the profile dataset comes from an existing atmospheric background library, for example, the atmospheric background library of the SCITRAN radiative transfer model. This is just an example.

[0075] It should be noted that the month, longitude, and latitude information of different passive observation data are different, and the prior profile data and the prior active lidar observation data obtained by screening are also different.

[0076] As an optional implementation, before calculating the differential optical depth, the active and passive combined observation method for inverting carbon dioxide column concentration of the present application further includes: extracting the peak value of the prior active lidar pulse signal.

[0077] This example uses a correlation filtering method. Taking the online monitoring signal of a laser footprint as an example, we first locate the extreme value of the online monitoring signal. Using the 200 sampling points before and after the extreme value as the initial interval, we perform a correlation filtering traversal using the laser radar's transmitted signal as a template. The value of the online monitoring signal at the location with the highest correlation is used as the peak value of the online active laser radar monitoring pulse signal. This is merely an example.

[0078] It should be noted that the larger the range of the initial interval, the longer the calculation process of the peak value of the prior active lidar pulse signal takes.

[0079] Specifically, after correlation filtering, the peak formula of the prior active lidar pulse signal is as follows.

[0080] (2);

[0081] in, is the peak value of the prior active lidar pulse signal; Monitor the peak value of the pulse signal for the online active lidar; is the peak value of the online active lidar echo pulse signal; Monitor the peak value of the pulse signal for the offline active lidar; is the peak value of the offline active lidar echo pulse signal. , , and Both dimensional matrix; is the number of lidar footprints.

[0082] After obtaining the peak information of the a priori active lidar pulse signal, the differential optical depth is calculated based on the radar equation.

[0083] Specifically, the formula for differential optical thickness is as follows.

[0084] (3);

[0085] in, is the differential optical thickness; Monitor the peak value of the pulse signal for the online active lidar; is the peak value of the online active lidar echo pulse signal; Monitor the peak value of the pulse signal for the offline active lidar; is the peak value of the offline active lidar echo pulse signal. for dimensional matrix, is the number of lidar footprints.

[0086] Because the absorption cross section calculation based on the line-by-line integration method is time-consuming, in order to improve the overall carbon dioxide column concentration inversion efficiency, a carbon dioxide absorption cross section lookup table is constructed in advance.

[0087] Specifically, at preset intervals, the line-by-line integration method is used to calculate the carbon dioxide absorption cross-section at different temperatures, humidity, pressures, and frequencies. The influence of the aliasing of carbon dioxide and water vapor absorption lines is taken into account to construct a four-dimensional carbon dioxide absorption cross-section lookup table.

[0088] In practical applications, for example, the CO2 absorption cross-section lookup table has a temperature range of 170.15-330.15 K, in 0.5 K increments; a pressure range of 25-105525 Pa, in 100 Pa increments; a relative humidity range of 0%-100% in 10% increments; and a frequency range of -14.9-14.9 MHz, in 0.1 MHz increments. This is just an example.

[0089] It should be noted that the determination of temperature, pressure, relative humidity and frequency range is related to the atmospheric conditions and satellite attitude during the actual observation of the active lidar. The temperature, pressure, relative humidity and frequency range should include all possible detection states of the active lidar as much as possible.

[0090] In an exemplary embodiment, step S6 specifically includes the following steps.

[0091] S61: Performing Gaussian fitting on the differential optical thickness to calculate the mean and standard deviation.

[0092] S62: performing quality control on the differential optical thickness according to the mean and standard deviation using the 3σ principle to obtain a quality-controlled differential optical thickness.

[0093] S63: Calculate a correction coefficient matrix of the priori profile and the corrected priori profile data according to the priori profile data, the quality-controlled differential optical thickness, and the carbon dioxide absorption cross-section lookup table.

[0094] The quality control of differential optical thickness is carried out according to the "3σ" principle. First, Gaussian fitting is performed on the differential optical thickness to calculate the mean μ and standard deviation σ. The active lidar observation data outside the interval (μ-3σ, μ+3σ) are eliminated to obtain the differential optical thickness after quality control. , for dimensional matrix, where is the number of lidar footprints after quality control.

[0095] As an optional implementation, before calculating the correction coefficient matrix of the prior profile and the corrected prior profile data, the active and passive combined observation method for inverting the carbon dioxide column concentration also includes: calculating the Doppler frequency shift of the prior active lidar observation data after quality control. The specific calculation formula is shown below.

[0096] (4);

[0097] in, The Doppler shift of the active lidar observation data after quality control; is the original emission frequency of the active lidar; is the active lidar flight speed; is the angle between the active lidar flight speed and the transmitted / received pulse echo signal; The speed of light.

[0098] Specifically, the calculation formula of the correction coefficient matrix of the prior profile is as follows.

[0099] (5);

[0100] (6);

[0101] in, is the correction coefficient matrix of the prior profile; for dimensional matrix, where is the number of lidar footprints after quality control; is the differential optical depth after quality control; For pressure The carbon dioxide volume mixing ratio at ; For pressure The weight function when ; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the online band; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the offline band; is the molecular mass of water vapor; is the mass of dry air molecules; is the water vapor volume mixing ratio, is the acceleration due to gravity.

[0102] In practical applications, the pressure is The carbon dioxide absorption cross section at time t is obtained by interpolating the 4D carbon dioxide absorption cross section lookup table, for example, by nearest neighbor interpolation, linear interpolation, or spline interpolation. This is just an example.

[0103] Furthermore, after obtaining the correction coefficient matrix of the prior profile After that, the correction coefficient matrix of the prior profile is weighted averaged to calculate the correction coefficient of the average prior profile, and the prior profile is scaled according to the correction coefficient of the average prior profile to obtain the corrected prior profile data.

[0104] Specifically, the calculation formula of the corrected priori profile data is as follows.

[0105] (7);

[0106] (8);

[0107] in, is the corrected prior profile data; for dimensional matrix, where is the number of atmospheric layers; is the extracted prior profile data; is the correction coefficient of the average prior profile; For the Correction coefficients for the a priori profiles of the lidar footprints; For the The peak value of the online active lidar echo pulse signal of each lidar footprint.

[0108] It should be noted that the month, longitude and latitude information of different passive observation data are different, and the corrected prior profile data calculated based on the passive observation data information are also different.

[0109] It is through the above steps that the corrected prior profile data of the same grid as the passive observation data are obtained.

[0110] As an optional implementation, the calculation formula for the carbon dioxide column concentration is as follows.

[0111] (9);

[0112] in, For the The carbon dioxide profile of the iteration, i.e., the carbon dioxide column concentration; is the corrected prior profile data; For the The Jacobian matrix of the iteration; is the error covariance matrix of the observation model; is the smoothing factor; is the error covariance matrix of the corrected prior profile data; It is passive observation data; is the forward model; For the The CO2 profile for the iteration.

[0113] The above process is illustrated using the observation data 231102 (latitude 64°S, longitude 59°W) of the OCO-2 nadir sensor in one track. The flow chart is shown in Figure 3.

[0114] Step 1: Data selection and active lidar observation dataset construction. The DQ-1 ACDL Level 2B observation data product was selected, covering the period from January to December 2023 and the spatial range from 82 degrees north latitude to 82 degrees south latitude. The active lidar data was filtered based on the altitude of the active lidar data echo signal, eliminating active lidar observation data above clouds and retaining only active lidar observation data from clear skies. This dataset was compiled into a monthly active lidar observation dataset with intervals of 5° in latitude and 30° in longitude.

[0115] Step 2: Prior data extraction. Based on the observation time and latitude of track 231102, the prior profile data and prior active lidar observation data corresponding to the observation data of track 231102 in November at the 65°S latitude and 60°W longitude were selected from the atmospheric background library of the SCITRAN radiation transfer model and the active lidar observation dataset.

[0116] Step 3: Extract the peak value of the pulse signal. Based on the correlation filtering method, with the maximum correlation coefficient as the principle, the echo pulse signal position is traversed and the value of the signal at that position is extracted as the peak value of the pulse signal; the extreme value of the pulse signal is also used as a replacement.

[0117] Step 4: Calculate differential optical depth. Substitute the peak value of the echo signal into the radar equation to calculate the differential optical depth.

[0118] Step 5: Construct an absorption cross-section lookup table. Based on the HITRAN2020 molecular absorption spectral library, the CO2 absorption cross-section is calculated using a line-by-line integration method, using the Voigt linear linearization method. Given different temperature, humidity, pressure, and frequency information, the CO2 absorption cross-sections under different atmospheric conditions are calculated and summarized into a four-dimensional CO2 absorption cross-section lookup table.

[0119] Step 6: Differential optical thickness screening: Perform quality control on the differential optical thickness according to the "3σ" principle and eliminate abnormal differential optical thickness data.

[0120] Step 7: Doppler shift calculation: Calculate the Doppler shift of the active lidar observation data based on the original transmission frequency, flight speed, and the angle between the transmitted and received pulse echo signals.

[0121] Step 8: Calculate the correction coefficient matrix for the a priori profile. The atmospheric CO2 absorption cross section of the observation data for track 231102 is obtained by interpolating the CO2 absorption cross section lookup table. Substituting the a priori profile data, differential optical depth, and interpolated CO2 absorption cross section into the correction coefficient matrix calculation formula, the correction coefficient matrix for the a priori profile is obtained.

[0122] Step 9: Obtain the corrected priori profile data. Perform a weighted average of the correction coefficient matrix of the priori profile according to the peak value of the online echo pulse signal to obtain the correction coefficient of the average priori profile. Multiply the correction coefficient of the average priori profile by the priori profile to obtain the corrected priori profile data.

[0123] Step 10: CO2 column concentration inversion. Using the corrected prior profile data as the initial CO2 profile data for optimal inversion, the CO2 column concentration results are continuously updated iteratively. When the iteration termination condition is met, the final CO2 column concentration is obtained.

[0124] Therefore, the present application provides a method for inverting carbon dioxide column concentration by active and passive combined observation. According to month information and longitude and latitude information, an active lidar observation data set is compiled. According to the prior active lidar observation data, the differential optical depth is calculated based on the radar equation. According to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed. Combined with the prior profile data and the differential optical depth, a correction coefficient matrix of the prior profile and the corrected prior profile data are calculated. According to the optimization algorithm principle, the corrected prior profile data is used as a constraint to calculate the carbon dioxide column concentration based on the passive observation data. Therefore, the inversion method of carbon dioxide column concentration provided by the present invention includes both active lidar observation information and passive observation information, effectively combining active and passive observations, and having more accurate prior profile information, which makes up for the defect of inaccurate prior profile information of the existing physical inversion algorithm and solves the problem of low inversion accuracy of the existing empirical algorithm.

[0125] In addition, based on the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table was constructed, which accelerated the calculation speed of the correction coefficient matrix of the prior profile and the corrected prior profile data, and improved the efficiency of carbon dioxide column concentration inversion.

[0126] The present application also provides an application scenario, which applies the above-mentioned active and passive combined observation method for inversion of carbon dioxide column concentration. Specifically: the active and passive combined observation method for inversion of carbon dioxide column concentration provided in this embodiment can be applied in meteorological detection scenarios. The meteorological detection scenario includes: data acquisition link, prior profile data extraction link, prior active lidar observation data extraction link, differential optical thickness calculation link, carbon dioxide absorption cross-section lookup table construction link, prior profile correction coefficient matrix and corrected prior profile data calculation link and carbon dioxide column concentration calculation link; obtaining active lidar observation data set, profile data set and passive observation information; the active lidar observation data set is a data set obtained by sorting out the month information and longitude and latitude information; the active lidar observation data set includes: online active lidar monitoring data, offline active lidar monitoring data, online active lidar echo data and offline active lidar echo data; the profile data set includes: pressure profile and carbon dioxide profile data According to the passive observation information, the data of the corresponding grid is extracted from the profile data set as the prior profile data; according to the passive observation information, the data of the corresponding grid is extracted from the active lidar observation data set as the prior active lidar observation data; according to the prior active lidar observation data, the differential optical thickness is calculated based on the radar equation; according to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed; according to the prior profile data, the differential optical thickness and the carbon dioxide absorption cross-section lookup table, a correction coefficient matrix of the prior profile and the corrected prior profile data are calculated; according to the optimization algorithm principle, the corrected prior profile data is used as a constraint, and the carbon dioxide column concentration is calculated based on the passive observation information.

[0127] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0128] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for inverting carbon dioxide column concentration by combining active and passive observation, characterized in that: The active and passive combined observation method for inverting carbon dioxide column concentration includes: Obtain an active lidar observation dataset, a profile dataset, and passive observation information; the active lidar observation dataset is a dataset compiled based on month information and latitude and longitude information; the active lidar observation dataset includes: online active lidar monitoring data, offline active lidar monitoring data, online active lidar echo data, and offline active lidar echo data; the profile dataset includes: pressure profile and carbon dioxide profile data; the passive observation information includes: OCO-2 observation data; extracting corresponding grid data from the profile data set as prior profile data based on the passive observation information; According to the passive observation information, extracting data of the corresponding grid from the active lidar observation data set as prior active lidar observation data; According to the a priori active lidar observation data, based on the radar equation, the differential optical depth is calculated; According to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed; Calculating a correction coefficient matrix of the priori profile and corrected priori profile data according to the priori profile data, the differential optical depth, and the carbon dioxide absorption cross-section lookup table; According to the principle of the optimization algorithm, the corrected prior profile data is used as a constraint and based on the passive observation information, the carbon dioxide column concentration is calculated.

2. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: Obtain active lidar observation datasets, including: Acquire active lidar observation data; The active lidar observation data are screened, the active lidar observation data above the clouds are eliminated, and only the active lidar observation data of the clear sky are retained to obtain an active lidar observation data set.

3. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: The calculation formula of the prior active lidar observation data is: ； in, is the extracted prior active lidar observation data; It is the online active lidar monitoring data; It is the echo data of online active lidar; Offline active lidar monitoring data; It is the offline active lidar echo data.

4. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: The calculation formula of the differential optical thickness is: ； in, is the differential optical thickness; Monitor the peak value of the pulse signal for the online active lidar; is the peak value of the online active lidar echo pulse signal; Monitor the peak value of the pulse signal for the offline active lidar; is the peak value of the offline active lidar echo pulse signal.

5. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: According to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed, including: At preset intervals, the line-by-line integration method is used to calculate the carbon dioxide absorption cross-section under different temperatures, humidity, pressures and frequencies. The influence of the aliasing of carbon dioxide and water vapor absorption lines is considered, and a four-dimensional carbon dioxide absorption cross-section lookup table is constructed.

6. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: Calculating a correction coefficient matrix of the priori profile and the corrected priori profile data according to the priori profile data, the differential optical depth, and the carbon dioxide absorption cross-section lookup table specifically includes: Performing Gaussian fitting on the differential optical thickness to calculate the mean and standard deviation; Performing quality control on the differential optical thickness using the 3σ principle according to the mean and the standard deviation to obtain a quality-controlled differential optical thickness; A correction coefficient matrix of the priori profile and the corrected priori profile data are calculated based on the priori profile data, the quality-controlled differential optical depth, and the carbon dioxide absorption cross-section lookup table.

7. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: Before calculating the correction coefficient matrix of the prior profile and the corrected prior profile data, the active and passive combined observation method for inverting carbon dioxide column concentration further includes: Calculate the Doppler shift of the prior active lidar observation data after quality control. The specific calculation formula is: ； in, The Doppler shift of the active lidar observation data after quality control; is the original emission frequency of the active lidar; is the active lidar flight speed; is the angle between the active lidar flight speed and the transmitted / received pulse echo signal; The speed of light.

8. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: The calculation formula of the correction coefficient matrix of the prior profile is: ； ； in, is the correction coefficient matrix of the prior profile; is the differential optical depth after quality control; For pressure The carbon dioxide volume mixing ratio at ; For pressure The weight function when ; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the online band; For pressure , the Doppler shift is The carbon dioxide absorption cross section in the offline band; is the molecular mass of water vapor; is the mass of dry air molecules; is the water vapor volume mixing ratio, is the acceleration due to gravity.

9. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: The calculation formula of the corrected prior profile data is: ； ； in, is the corrected prior profile data; is the extracted prior profile data; is the correction coefficient of the average prior profile; For the Correction coefficients for the a priori profiles of the lidar footprints; For the The peak value of the online active lidar echo pulse signal of each lidar footprint; is the number of lidar footprints after quality control.

10. The method for inverting carbon dioxide column concentration by active and passive combined observation according to claim 1, characterized in that: The calculation formula of the carbon dioxide column concentration is: ； in, For the The carbon dioxide profile of the iteration; is the corrected prior profile data; For the The Jacobian matrix of the iteration; is the error covariance matrix of the observation model; is the smoothing factor; is the error covariance matrix of the corrected prior profile data; It is passive observation data; is the forward model; For the The CO2 profile for the iteration.

Images