Active and passive combined observation carbon dioxide column concentration inversion method

Abstract

The invention discloses an active and passive combined observation carbon dioxide column concentration inversion method, and relates to the technical field of meteorological detection.The method comprises the steps that an active laser radar observation data set is obtained through arrangement according to month information and longitude and latitude information, and differential optical thickness is obtained through calculation based on a radar equation according to priori active laser radar observation data; according to the line-by-line integration principle, a carbon dioxide absorption cross section lookup table is constructed, priori profile data and differential optical thickness are combined, a correction coefficient matrix of a priori profile and corrected priori profile data are obtained through calculation, and according to the optimization algorithm principle, the corrected priori profile data serve as constraint, and the carbon dioxide absorption cross section is obtained. And the concentration of the carbon dioxide column is calculated based on the passive observation data, so that the inversion method provided by the invention not only comprises active laser radar observation, but also comprises the passive observation data, the active observation and the passive observation are effectively combined, more accurate prior profile information is provided, and the inversion precision is improved.

Description

Technical Field

[0001] The present application relates to the technical field of meteorological detection, and particularly to a method for retrieving carbon dioxide column concentration by combining active and passive observations. Background Art

[0002] As an important component of greenhouse gases, carbon dioxide plays an important role in the process of global warming. Therefore, accurately assessing carbon dioxide is crucial for improving our prediction of the carbon cycle and climate change. There are two common methods for carbon dioxide retrieval: empirical algorithms and physical inversion algorithms.

[0003] Empirical algorithms refer to retrieving the carbon dioxide column concentration by establishing a statistical regression model between satellite observation values and the target parameters to be retrieved. The retrieval does not involve the radiation transfer process and has the advantages of simple retrieval principle and high calculation efficiency. Specifically, typical methods of empirical algorithms include statistical regression methods and neural network algorithms. Currently, this type of empirical algorithm is applied in near-infrared nadir sounding sensors. However, the reliability of the retrieval results of this type of algorithm is mainly determined by the quality of the sample set used to construct the regression model. When the total number of samples in the sample set is small or not representative, the retrieval results will deviate from the true values. Due to the above defects, the application of existing empirical algorithms is limited.

[0004] Physical inversion algorithms use a forward radiation transfer model to solve the target parameters to be retrieved and have clear physical meanings. Specifically, physical inversion methods include optimization algorithms and DOAS algorithms. However, it is necessary to use the carbon dioxide profile as the prior profile to constrain the retrieval results, and the accuracy of the prior profile plays a decisive role in the final accuracy of the carbon dioxide column concentration. If the prior information deviates from the actual distribution state of carbon dioxide, the retrieval results will be poor.

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

[0006] The purpose of the present application is to provide a method for retrieving carbon dioxide column concentration by combining active and passive observations, which can effectively combine active and passive observations, has more accurate prior profile information, makes up for the defect of inaccurate prior profile information in existing physical inversion algorithms, and solves the problem of low retrieval accuracy of existing empirical algorithms.

[0007] To achieve the above purpose, the present application provides the following solutions.

[0008] The present application provides a method for retrieving carbon dioxide column concentration by combining active and passive observations, and the method for retrieving carbon dioxide column concentration by combining active and passive observations includes the following steps.

[0009] An active lidar observation data set, a profile data set and passive observation information are obtained; the active lidar observation data set is a data set compiled based on 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.

[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] According to the a priori active laser radar observation data, the differential optical thickness is calculated based on the radar equation.

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

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

[0015] According to the optimization algorithm principle, the corrected priori 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 data set is obtained, specifically including the following contents.

[0017] Acquire active lidar observations.

[0018] The active lidar observation data are screened, the active lidar observation data on the clouds are eliminated, and only the active lidar observation data on 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] ; in, is the extracted prior active lidar observation data; It is the online active lidar monitoring data; is the online active lidar echo data; is the offline active lidar monitoring data; is the offline active lidar echo data.

[0021] Optionally, the calculation formula of the differential optical thickness is as follows.

[0022] ; where is the differential optical thickness; is the peak value of the monitoring pulse signal of the online active lidar; is the peak value of the echo pulse signal of the online active lidar; is the peak value of the monitoring pulse signal of the offline active lidar; is the peak value of the echo pulse signal of the offline active lidar.

[0023] Optionally, according to the line-by-line integration principle, a lookup table of carbon dioxide absorption cross-sections is constructed, specifically including: at a preset interval, using the line-by-line integration method, calculating the carbon dioxide absorption cross-sections at different temperatures, humidities, pressures and frequencies, and considering the influence of the overlap of carbon dioxide and water vapor absorption lines, a four-dimensional carbon dioxide absorption cross-section lookup table is constructed.

[0024] Optionally, 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 corrected prior profile data are calculated, specifically including the following content.

[0025] Perform Gaussian fitting on the differential optical thickness to calculate the mean value and standard deviation.

[0026] According to the mean value and standard deviation, use the 3σ principle to perform quality control on the differential optical thickness to obtain the quality-controlled differential optical thickness.

[0027] According to the prior profile data, the quality-controlled differential optical thickness and the carbon dioxide absorption cross-section lookup table, a correction coefficient matrix of the prior profile and corrected prior profile data are calculated.

[0028] Optionally, before calculating the correction coefficient matrix of the prior profile and the corrected prior profile data, the active and passive combined observation carbon dioxide column concentration inversion method further includes the following content.

[0029] Calculate the Doppler shift of the quality-controlled prior active lidar observation data, and the specific calculation formula is as follows.

[0030] ; wherein, is the Doppler shift of the prior active lidar observation data after quality control; is the original emission frequency of the active lidar; is the flight speed of the active lidar; is the angle between the flight speed of the active lidar and the transmitted / received pulse echo signal; is the speed of light.

[0031] Optionally, the calculation formula of the correction coefficient matrix of the prior profile is as follows.

[0032] ; ; wherein, is the correction coefficient matrix of the prior profile; is the differential optical thickness after quality control; is the carbon dioxide volume mixing ratio at a pressure of ; is the weighting function at a pressure of ; is the carbon dioxide absorption cross section in the online band at a pressure of , and a Doppler shift of ; is the carbon dioxide absorption cross section in the offline band at a pressure of , and a Doppler shift of ; is the mass of water vapor molecules; is the mass of dry air molecules; is the water vapor volume mixing ratio, is the acceleration due to gravity.

[0033] Optionally, the calculation formula of the corrected prior profile data is as follows.

[0034] ; ; wherein, is the corrected prior profile data; is the extracted prior profile data; is the correction coefficient of the average prior profile; is the correction coefficient of the prior profile of the th lidar footprint; is the peak value of the online active lidar echo pulse signal of the th lidar footprint.

[0035] Optionally, the calculation formula of the carbon dioxide column concentration is as follows.

[0036] ; Wherein, is the carbon dioxide profile for the th iteration; is the prior profile data after correction; is the Jacobian matrix for the th iteration; is the error covariance matrix of the observation model; is the smoothing factor; is the error covariance matrix of the prior profile data after correction; is the passive observation data; is the forward model; is the carbon dioxide profile for the th iteration.

[0037] According to the specific embodiments provided by the present application, the following technical effects are disclosed by the present application.

[0038] The present application provides a method for retrieving the carbon dioxide column concentration by combining active and passive observations. According to the month information and the longitude and latitude information, the active lidar observation data set is sorted out. Based on the prior active lidar observation data and the radar equation, the differential optical thickness is calculated. According to the line-by-line integration principle, a lookup table of the carbon dioxide absorption cross-section is constructed. Combining the prior profile data and the differential optical thickness, the correction coefficient matrix of the prior profile and the prior profile data after correction are calculated. According to the principle of the optimization algorithm, with the prior profile data after correction as the constraint, the carbon dioxide column concentration is calculated based on the passive observation data. Therefore, the method for retrieving the carbon dioxide column concentration provided by the present application includes both active lidar observations and passive observation data, effectively combines active and passive observations, has more accurate prior profile information, makes up for the defect of inaccurate prior profile information in the existing physical inversion algorithm, and solves the problem of low inversion accuracy of the existing empirical algorithm. In addition, according to the line-by-line integration principle, a lookup table of the carbon dioxide absorption cross-section is constructed, thereby accelerating the calculation speed of the correction coefficient matrix of the prior profile and the prior profile data after correction, and improving the efficiency of retrieving the carbon dioxide column concentration. Description of the Drawings

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0040] Figure 1 This is an application environment diagram of a method for retrieving carbon dioxide column concentration by combined active and passive observations in an embodiment of the present application.

[0041] Figure 2 This is a schematic flowchart of a method for retrieving carbon dioxide column concentration by combined active and passive observations provided in an embodiment of the present application.

[0042] Figure 3 This is a flowchart for retrieving carbon dioxide column concentration by a nadir sensor provided in an embodiment of the present application. Detailed implementation manners

[0043] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0044] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0045] Currently, traditional physical inversion algorithms are widely applied to nadir sounding data. However, it is necessary to use the carbon dioxide profile as a prior profile to constrain the inversion result, and the accuracy of the prior profile plays a decisive role in the final carbon dioxide column concentration accuracy. If the prior information deviates from the actual distribution state of carbon dioxide, the inversion result will be poor. The emergence of the active lidar observation method can provide more effective information for carbon dioxide inversion. However, currently, the two have not been well combined in the retrieval of carbon dioxide column concentration.

[0046] The reliability of the inversion result of the empirical algorithm is mainly determined by the quality of the sample set used to construct the regression model. When the total number of samples in the sample set is small or not representative, the inversion result will deviate from the true value. Due to the above defects, the application of existing empirical algorithms is limited.

[0047] The method for retrieving carbon dioxide column concentration by combined active and passive observations provided in the embodiments of the present application can be applied to, for example 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 separately, integrated on the server 104, placed on the cloud or other servers. The terminal 102 can send the obtained active lidar observation dataset, profile dataset, and passive observation information to the server 104. The active lidar observation dataset is a dataset sorted based on month information and longitude and latitude 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. After receiving the active lidar observation dataset, profile dataset, and passive observation information, the server 104 extracts the data of the corresponding grid from the profile dataset as the prior profile data according to the passive observation information; extracts the data of the corresponding grid from the active lidar observation dataset as the prior active lidar observation data according to the passive observation information; calculates the differential optical thickness based on the radar equation according to the prior active lidar observation data; constructs a lookup table of carbon dioxide absorption cross-sections according to the line-by-line integration principle; calculates the correction coefficient matrix of the prior profile and the corrected prior profile data according to the prior profile data, the differential optical thickness, and the lookup table of carbon dioxide absorption cross-sections; calculates the carbon dioxide column concentration based on the corrected prior profile data as a constraint according to the principle of the optimization algorithm and based on the passive observation information. The server 104 can feedback the obtained carbon dioxide column concentration to the terminal 102. In addition, in some embodiments, the method for retrieving the carbon dioxide column concentration by joint active and passive observations can also be implemented separately by the server 104 or the terminal 102. For example, the terminal 102 can directly retrieve the carbon dioxide column concentration for 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 retrieve the carbon dioxide column concentration for the active lidar observation dataset, profile dataset, and passive observation information.

[0048] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smartphones, and tablet computers. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0049] In an exemplary embodiment, such as Figure 2As shown, a method for retrieving carbon dioxide column concentration by combining active and passive observations is provided. This method is executed by a computer device, specifically, it can be executed alone by a computer device such as a terminal or a server, or jointly executed by a terminal and a server. In the embodiments of the present application, taking the application of this method to Figure 1 server 104 in

[0050] as an example, the method includes the following steps S1 to S7.

[0051] S1: Obtain an active lidar observation dataset, a profile dataset, and passive observation information; the active lidar observation dataset is a dataset sorted based on month information and longitude and latitude 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 profiles and carbon dioxide profile data; the passive observation information includes: OCO-2 observation data.

[0052] S2: According to the passive observation information, extract the data of the corresponding grid from the profile dataset as the prior profile data.

[0053] S3: According to the passive observation information, extract the data of the corresponding grid from the active lidar observation dataset as the prior active lidar observation data.

[0054] S4: According to the prior active lidar observation data, calculate the differential optical thickness based on the radar equation.

[0055] S5: Construct a lookup table for carbon dioxide absorption cross-sections according to the line-by-line integration principle.

[0056] S6: According to the prior profile data, the differential optical thickness, and the lookup table for carbon dioxide absorption cross-sections, calculate the correction coefficient matrix of the prior profile and the corrected prior profile data.

[0057] By implementing the above steps S1 to S7, the original profile data set is corrected according to the active lidar observation data, so that the inversion process of the carbon dioxide column concentration includes both active lidar observation information and passive observation information, effectively combining active and passive observations, and having more accurate prior profile information. At the same time, according to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed, thereby speeding up the calculation speed of the correction coefficient matrix of the prior profile and the corrected prior profile data, and improving the inversion efficiency of the carbon dioxide column concentration.

[0058] In an exemplary embodiment, in step S1, an active lidar observation data set is obtained, which specifically includes the following steps.

[0059] S11: Acquire active lidar observation data.

[0060] S12: Filter the active laser radar observation data, remove the active laser radar observation data on the clouds, and only retain the active laser radar observation data on the clear sky to obtain an active laser radar observation data set.

[0061] In practical applications, for example, according to the month information, the latitude interval is 5°, and the longitude interval is 30°, the active lidar observation data is sorted to obtain the monthly, 5° latitude grid, and 30° longitude grid active lidar observation data set. This is just an example.

[0062] 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, the data with the same month and in the same longitude and latitude band as the passive observation data are extracted as the prior profile data and prior active lidar observation data.

[0063] 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.

[0064] (1); 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 laser radar; It is offline active lidar monitoring data; It is the offline active lidar echo data. for d - dimensional matrix; is d - dimensional matrix; is d - dimensional matrix; is d - dimensional matrix; is the number of laser footprints; is the number of sampling points of the lidar online monitoring signal; is the number of sampling points of the online echo signal; is the number of sampling points of the offline monitoring signal; is the number of sampling points of the offline echo signal.

[0065] In practical applications, the profile dataset is sourced from an existing atmospheric background library. For example, the atmospheric background library of the SCITRAN radiative transfer model. This is just an example.

[0066] It should be noted that the month, longitude, and latitude information of different passive observation data are different, and the prior profile data and prior active lidar observation data obtained by screening are also different.

[0067] As an alternative implementation, before calculating the differential optical thickness, the method for retrieving the carbon dioxide column concentration by joint active and passive observations of the present application further includes: extracting the peak value of the prior active lidar pulse signal.

[0068] In this embodiment, the method of correlation filtering is adopted. Taking the online monitoring signal of a certain laser footprint as an example, first find the position of the extreme value of the online monitoring signal. Using the 200 sampling points before and after the extreme value position as the initial interval, traverse using the correlation filtering method with the transmitted signal of the lidar as the template, and use the value of the online monitoring signal at the position with the maximum correlation as the peak value of the online active lidar monitoring pulse signal. This is just an example.

[0069] 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.

[0070] Specifically, after correlation filtering, the formula for the peak value of the prior active lidar pulse signal is as follows.

[0071] (2); Wherein, is the peak value of the prior active lidar pulse signal; is the peak value of the online active lidar monitoring pulse signal; is the peak value of the online active lidar echo pulse signal; is the peak value of the pulsed signal monitored by the offline active lidar; is the peak value of the echo pulsed signal of the offline active lidar. , , and are both dimensional matrices; is the number of lidar footprints.

[0072] After obtaining the prior information on the peak value of the pulsed signal of the active lidar, the differential optical depth is calculated based on the radar equation.

[0073] Specifically, the formula for the differential optical depth is as follows.

[0074] (3); Among them, is the differential optical depth; is the peak value of the pulsed signal monitored by the online active lidar; is the peak value of the echo pulsed signal of the online active lidar; is the peak value of the pulsed signal monitored by the offline active lidar; is the peak value of the echo pulsed signal of the offline active lidar. is dimensional matrix, is the number of lidar footprints.

[0075] Because the calculation of the absorption cross-section based on the line-by-line integration method is time-consuming, in order to improve the overall inversion efficiency of the carbon dioxide column concentration, a lookup table for the carbon dioxide absorption cross-section is constructed in advance.

[0076] Specifically, at a preset interval, the line-by-line integration method is used to calculate the carbon dioxide absorption cross-section at different temperatures, humidities, pressures, and frequencies, and considering the influence of the overlap of the absorption lines of carbon dioxide and water vapor, a four-dimensional lookup table for the carbon dioxide absorption cross-section is constructed.

[0077] In practical applications, for example, the temperature range of the lookup table for the carbon dioxide absorption cross-section is 170.15 - 330.15K, with an interval of 0.5K; the pressure range is 25 - 105525Pa, with an interval of 100Pa; the relative humidity is 0% - 100%, with an interval of 10%; the frequency range is -14.9 - 14.9MHz, with an interval of 0.1MHz. This is just an example.

[0078] It should be noted that the determination of the temperature, pressure, relative humidity and frequency range is related to the atmospheric state and satellite attitude during the actual observation of the active lidar. The temperature, pressure, relative humidity and frequency range should cover as many possible detection states of the active lidar as possible.

[0079] In an exemplary embodiment, in step S6, it specifically includes the following content.

[0080] S61: Perform Gaussian fitting on the differential optical thickness, and calculate the mean value and standard deviation.

[0081] S62: According to the mean value and standard deviation, perform quality control on the differential optical thickness using the 3σ principle to obtain the differential optical thickness after quality control.

[0082] S63: According to the prior profile data, the differential optical thickness after quality control, and the lookup table of carbon dioxide absorption cross-section, calculate the correction coefficient matrix of the prior profile and the prior profile data after correction.

[0083] Perform quality control on the differential optical thickness according to the "3σ" principle. First, perform Gaussian fitting on the differential optical thickness to calculate the mean value μ and standard deviation σ, and eliminate the active lidar observation data outside the interval (μ - 3σ, μ + 3σ) to obtain the differential optical thickness after quality control. , is a dimensional matrix, where

[0084] As an alternative implementation, before calculating the correction coefficient matrix of the prior profile and the prior profile data after correction, the method for retrieving the carbon dioxide column concentration by combined active and passive observations further includes: calculating the Doppler shift of the prior active lidar observation data after quality control, and the specific calculation formula is as follows.

[0085] (4); where is the Doppler shift of the prior active lidar observation data after quality control; is the original emission frequency of the active lidar; is the flight speed of the active lidar; is the angle between the flight speed of the active lidar and the transmitted / received pulse echo signal; is the speed of light.

[0086] Specifically, the calculation formula of the correction coefficient matrix of the prior profile is as follows.

[0087] (5); (6); Wherein, is the correction coefficient matrix of the a priori profile; is -dimensional matrix, wherein, is the number of lidar footprints after quality control; is the differential optical thickness after quality control; is the carbon dioxide volume mixing ratio at pressure ; is the weighting function at pressure ; is the carbon dioxide absorption cross section of the online band at pressure , Doppler shift ; is the carbon dioxide absorption cross section of the offline band at pressure , Doppler shift ; is the mass of water vapor molecules; is the mass of dry air molecules; is the water vapor volume mixing ratio, is the acceleration of gravity.

[0088] In practical applications, the carbon dioxide absorption cross section at pressure is obtained by interpolating a four-dimensional carbon dioxide absorption cross section lookup table, for example, nearest neighbor interpolation, linear interpolation or spline interpolation. This is just an example.

[0089] Further, after obtaining the correction coefficient matrix of the a priori profile, the correction coefficient matrix of the a priori profile is weighted averaged to calculate the correction coefficient of the average a priori profile, and the a priori profile is scaled according to the correction coefficient of the average a priori profile to obtain the corrected a priori profile data.

[0090] Specifically, the calculation formula of the corrected a priori profile data is as follows.

[0091] (7); (8); Wherein, is the corrected a priori profile data; is -dimensional matrix, wherein, is the number of atmospheric layers; is the extracted a priori profile data; is the correction coefficient of the average a priori profile; For the Correction coefficients for the a priori profiles of the lidar footprints; For the Peak value of the online active lidar echo pulse signal for each lidar footprint.

[0092] It should be noted that different passive observation data have different month, longitude and latitude information, and the corrected prior profile data calculated based on the passive observation data information are also different.

[0093] It is through the above steps that the corrected prior profile data of the same grid as the passive observation data are obtained.

[0094] As an optional implementation, the calculation formula for the carbon dioxide column concentration is as follows.

[0095] (9); 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 CO2 profile for the iteration.

[0096] The above process is illustrated by taking the observation data 231102 (latitude 64°S, longitude 59°W) of the OCO-2 nadir sensor as an example, and the flow chart is shown in Figure 3.

[0097] Step 1: Data selection and construction of active lidar observation data set. DQ-1 ACDL Level2B observation data products are selected, with a time range of January to December 2023 and a spatial range from 82 degrees north latitude to 82 degrees south latitude. Active lidar data are screened according to the height of the active lidar data echo signal, and active lidar observation data on clouds are eliminated, leaving only active lidar observation data in clear sky; monthly active lidar observation data sets are sorted out with 5° latitude intervals and 30° longitude intervals.

[0098] Step 2: Prior data extraction. According to the observation time longitude and latitude information of track 231102, the prior profile data and prior active lidar observation data corresponding to the observation data of track 231102 in November, 65°S latitude band and 60°W longitude band were selected from the atmospheric background library of the SCITRAN radiation transfer model and the active lidar observation data set.

[0099] Step 3: Extract the peak value of the pulse signal. Based on the correlation filtering method, the position of the echo pulse signal is traversed and the value of the signal at this position is extracted as the peak value of the pulse signal; the extreme value of the pulse signal is also used as a substitute.

[0100] Step 4: Calculate differential optical thickness. Substitute the peak value of the echo signal into the radar equation to calculate the differential optical thickness.

[0101] Step 5: Construction of absorption cross section lookup table. Based on the HITRAN2020 molecular absorption spectrum library, the carbon dioxide absorption cross section is calculated by line-by-line integration, and the selected line shape is Voigt linear. By giving different temperature, humidity, pressure and frequency information, the carbon dioxide absorption cross section under different atmospheric conditions is calculated and summarized into a four-dimensional carbon dioxide absorption cross section lookup table.

[0102] Step 6: Differential optical thickness screening: Perform quality control on differential optical thickness according to the "3σ" principle and eliminate abnormal differential optical thickness data.

[0103] Step 7: Doppler shift calculation: Based on the original transmission frequency, flight speed, and angle between the transmission / reception pulse echo signal of the active laser radar, the Doppler shift of the active laser radar observation data is calculated.

[0104] Step 8: Calculation of correction coefficient matrix of a priori profile. The carbon dioxide absorption cross section under atmospheric conditions of the observation data of track 231102 is obtained by interpolating the carbon dioxide absorption cross section lookup table, and the a priori profile data, differential optical depth and interpolated carbon dioxide absorption cross section are substituted into the correction coefficient matrix calculation formula to obtain the correction coefficient matrix of the a priori profile.

[0105] Step 9: Obtain the corrected prior profile data. The correction coefficient matrix of the prior profile is weighted averaged according to the peak value of the online echo pulse signal to obtain the correction coefficient of the average prior profile, and the correction coefficient of the average prior profile is multiplied by the prior profile to obtain the corrected prior profile data.

[0106] Step 10: Inversion of CO2 column concentration. The corrected prior profile data is used as the initial CO2 profile data for optimal inversion, and the CO2 column concentration results are continuously updated iteratively. When the iteration termination condition is met, the final CO2 column concentration is obtained.

[0107] Therefore, the present application provides a method for retrieving carbon dioxide column concentration by combining active and passive observations. According to the month information and latitude-longitude information, an active lidar observation dataset is sorted out. Based on the prior active lidar observation data and the radar equation, the differential optical thickness is calculated. According to the line-by-line integration principle, a lookup table of carbon dioxide absorption cross-sections is constructed. Combining the prior profile data and the differential optical thickness, the correction coefficient matrix of the prior profile and the corrected prior profile data are calculated. According to the principle of the optimization algorithm, with the corrected prior profile data as a constraint, the carbon dioxide column concentration is calculated based on the passive observation data. Thus, the carbon dioxide column concentration retrieval method provided by the present invention includes both active lidar observation information and passive observation information, effectively combines active and passive observations, and has more accurate prior profile information, making up for the defect of inaccurate prior profile information in the existing physical inversion algorithms and solving the problem of low inversion accuracy of the existing empirical algorithms.

[0108] In addition, according to the line-by-line integration principle, a lookup table of carbon dioxide absorption cross-sections is constructed, thereby accelerating the calculation speed of the correction coefficient matrix of the prior profile and the corrected prior profile data and improving the carbon dioxide column concentration inversion efficiency.

[0109] The present application also provides an application scenario, which applies the above-mentioned method for retrieving carbon dioxide column concentration by combining active and passive observations. Specifically: The method for retrieving carbon dioxide column concentration by combining active and passive observations provided in this embodiment can be applied in a meteorological detection scenario. The meteorological detection scenario includes: a data acquisition link, a prior profile data extraction link, a prior active lidar observation data extraction link, a differential optical thickness calculation link, a carbon dioxide absorption cross-section lookup table construction link, a correction coefficient matrix of the prior profile, and a corrected prior profile data calculation link and a carbon dioxide column concentration calculation link; obtaining an active lidar observation data set, a profile data set, and passive observation information; the active lidar observation data set is a data set sorted based on 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: a pressure profile and carbon dioxide profile data; the passive observation information includes: OCO-2 observation data; according to the passive observation information, extracting the corresponding grid data from the profile data set as the prior profile data; according to the passive observation information, extracting the corresponding grid data from the active lidar observation data set as the prior active lidar observation data; according to the prior active lidar observation data, calculating the differential optical thickness based on the radar equation; constructing a carbon dioxide absorption cross-section lookup table according to the line-by-line integration principle; according to the prior profile data, the differential optical thickness, and the carbon dioxide absorption cross-section lookup table, calculating the correction coefficient matrix of the prior profile and the corrected prior profile data; according to the principle of the optimization algorithm, using the corrected prior profile data as a constraint, and calculating the carbon dioxide column concentration based on the passive observation information.

[0110] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered as the scope recorded in this specification.

[0111] In this article, specific examples are used to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, based on the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for retrieving carbon dioxide column concentration by combining active and passive observations, characterized in that The active and passive combined observation method for inversion of carbon dioxide column concentration includes: Obtain an active laser radar observation data set, a profile data set, and passive observation information; the active laser radar observation data set is a data set compiled based on month information and longitude and latitude information; the active laser radar observation data set includes: online active laser radar monitoring data, offline active laser radar monitoring data, online active laser radar echo data, and offline active laser radar echo data; the profile data set includes: pressure profile and carbon dioxide profile data; the passive observation information includes: OCO-2 observation data; According to the passive observation information, extracting data of corresponding grids from the profile data set as prior profile data; According to the passive observation information, extracting data of corresponding grids from the active lidar observation data set as prior active lidar observation data; According to the a priori active laser radar observation data, based on the radar equation, the differential optical thickness is calculated; According to the line-by-line integration principle, a carbon dioxide absorption cross-section lookup table is constructed; According to the priori profile data, the differential optical thickness and the carbon dioxide absorption cross-section lookup table, a correction coefficient matrix of the priori profile and the corrected priori profile data are calculated; According to the optimization algorithm principle, the corrected priori 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 retrieving carbon dioxide column concentration by combined active and passive observations according to claim 1, wherein Obtain active lidar observation datasets, including: Acquire active lidar observation data; The active lidar observation data are screened, the active lidar observation data on the clouds are eliminated, and only the active lidar observation data on the clear sky are retained to obtain an active lidar observation data set.

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

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

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

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

8. The method for retrieving carbon dioxide column concentration by combined active and passive observations according to claim 1, wherein The calculation formula for the correction coefficient matrix of the prior profile is as follows: ； ； Among them, is the correction coefficient matrix of the prior profile; is the differential optical thickness after quality control; is the volume mixing ratio of carbon dioxide at pressure; is the weighting function at pressure; is the , Doppler shift is absorption cross section of carbon dioxide in the online band; is the , Doppler shift is absorption cross section of carbon dioxide in the offline band; is the mass of water vapor molecules; is the mass of dry air molecules; is the volume mixing ratio of water vapor, is the acceleration of gravity.

9. The method for retrieving carbon dioxide column concentration by combined active and passive observations according to claim 1, characterized in that The calculation formula for the prior profile data after correction is as follows: ； ； Among them, is the corrected prior profile data; is the extracted prior profile data; is the correction coefficient of the average prior profile; is the correction coefficient of the prior profile of the th lidar footprint; is the peak value of the online active lidar echo pulse signal of the th lidar footprint.

10. The method for retrieving carbon dioxide column concentration by combined active and passive observations according to claim 1, wherein The calculation formula for the carbon dioxide column concentration is as follows: ； wherein, is the carbon dioxide profile for the th iteration; is the prior profile data after correction; is the Jacobian matrix for the th iteration; is the error covariance matrix of the observation model; is the smoothing factor; is the error covariance matrix of the prior profile data after correction; is the passive observation data; is the forward model; is the carbon dioxide profile for the th iteration.

Images