Method and system for anthropogenic co2 emission source inversion based on carbon-nitrogen homology

Abstract

This application relates to a method and system for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship, belonging to the technical field of carbon pollution synergistic governance and atmospheric emission inversion. The method includes: collecting and preprocessing multi-source data; constructing a carbon dioxide emission source assimilation system based on four-dimensional variational theory; statistically analyzing the proportions of anthropogenic carbon and nitrogen emissions by province and department based on carbon-nitrogen homology constraints, and combining this with high-resolution anthropogenic nitrogen dioxide emission sources through proportional conversion to obtain prior anthropogenic carbon dioxide emission sources. These prior anthropogenic carbon dioxide emission sources, along with the preprocessed multi-source data, are input into the assimilation system for assimilation and inversion. The carbon dioxide concentration obtained from three-dimensional variational assimilation within a buffer region larger than the study area is used as the initial and boundary conditions of the system. A four-dimensional variational algorithm is used to invert and obtain carbon dioxide emission adjustment coefficients, ultimately correcting the data to obtain high-resolution anthropogenic carbon dioxide emission data. This method can achieve accurate inversion of anthropogenic carbon dioxide emissions.

Description

Technical Field

[0001] This application relates to the fields of carbon pollution synergistic governance and atmospheric emission inversion technology, and in particular to a method and system for inverting anthropogenic carbon dioxide emission sources based on the common origin relationship of carbon and nitrogen. Background Technology

[0002] Currently, to reverse the rapid growth of carbon dioxide emissions, it is generally necessary to formulate scientific emission reduction policies based on accurate and timely carbon emission inventories. Traditional carbon emission inventories use an inventory compilation method, estimating carbon dioxide emissions through statistical emission factors and activity level data, and serve as the basic basis for carbon trading, emission reduction accounting, and climate negotiations. However, carbon dioxide emission inventories obtained through the inventory compilation method usually have a coarse spatiotemporal resolution and require a large amount of human and material resources for statistics, making it difficult to achieve timely and rapid updates.

[0003] Four-dimensional variational theory can optimize traditional emission inventories using model and observational data, resulting in higher accuracy and faster updates. Its advantage lies in using multi-source observations as constraints on state variables. However, current carbon emission retrieval studies typically only use carbon dioxide concentration observations as constraints. Due to the limited number of carbon dioxide observation stations and the low spatial coverage of carbon satellites such as GOSAT, GOSAT-2, OCO-2, OCO-3, and DQ-1, high-resolution anthropogenic carbon dioxide emission retrieval is not yet supported. Summary of the Invention

[0004] Therefore, it is necessary to provide a method and system for inverting anthropogenic carbon dioxide emission sources based on the common origin relationship of carbon and nitrogen to address the above-mentioned technical problems. This method and system can obtain prior anthropogenic carbon dioxide emission sources based on the proportion of anthropogenic carbon and nitrogen emissions by province and department and high-resolution anthropogenic nitrogen dioxide emission data. By combining the constraints of multi-source carbon satellite observation data, it can ultimately achieve high-resolution, province-by-province-by-department accurate inversion of anthropogenic carbon dioxide emissions.

[0005] A method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship, the method comprising:

[0006] Collect and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources in the study area.

[0007] A carbon dioxide emission source assimilation system is constructed based on four-dimensional variational theory, categorized by province, department, and source.

[0008] Based on the carbon and nitrogen homology constraint, statistical analysis was performed on the anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in the preprocessed multi-source data. The anthropogenic carbon and nitrogen emission ratios by province and department were obtained and an anthropogenic carbon and nitrogen emission ratio database was constructed. The anthropogenic carbon and nitrogen emission ratio database was applied to the high-resolution anthropogenic nitrogen dioxide emission sources by province and department. The corresponding high-resolution anthropogenic carbon dioxide emission sources were obtained by ratio conversion and used as the prior anthropogenic carbon dioxide emission sources.

[0009] Prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data are input into a carbon dioxide emission source assimilation system for assimilation and inversion. The carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area is used as the initial and boundary conditions of the system. The carbon dioxide emission adjustment coefficients of provinces and departments in the study area are obtained by using a four-dimensional variational algorithm, and the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments are corrected. Finally, high-resolution anthropogenic carbon dioxide emission data after assimilation are obtained.

[0010] In one embodiment, multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area are collected, including:

[0011] The study collected multi-source carbon satellite observation data, terrestrial ecosystem carbon flux data by vegetation type, marine carbon flux data, high-resolution weather forecast data, wildfire carbon emission data by province, aircraft carbon emission data by province, ship carbon emission data by province, carbon emission data generated by chemical reactions by province, anthropogenic carbon dioxide emission data by province and anthropogenic nitrogen dioxide emission data by province and sector. Among them, the emitting sectors of anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data include power plants, factories, transportation and residents; the vegetation types of terrestrial ecosystem carbon flux data include evergreen broad-leaved, evergreen coniferous, bare soil and farmland soil.

[0012] In one embodiment, the multi-source data is preprocessed, including: data format standardization processing of the multi-source data, spatiotemporal matching processing to unify the multi-source data to the same time scale and spatial resolution, outlier removal processing, and data interpolation processing.

[0013] In one embodiment, a carbon dioxide emission source assimilation system is constructed based on four-dimensional variational theory, categorized by province, department, and source, including:

[0014] Based on four-dimensional variational theory, a carbon dioxide emission source assimilation system is constructed, categorized by province, sector, and source. This system centers on regional atmospheric transport models and uses preprocessed multi-source data, including province- and sector-specific anthropogenic carbon dioxide emission data, province-specific wildfire carbon emission data, province-specific aircraft carbon emission data, province-specific ship carbon emission data, province-specific carbon emission data from chemical reactions, terrestrial ecosystem carbon flux data categorized by vegetation type, and marine carbon flux data, as state variables. The objective function of the carbon dioxide emission source assimilation system is then constructed, expressed as:

[0015] ;

[0016] in, The objective function value; For state variables, These are background emission sources, also known as prior carbon dioxide emission sources. It is the optimized emissions. This represents the ratio between the prior optimized emissions and the background emissions, at which point the emission sources have not yet been optimized. ; Representing the One state variable, Indicates the total number of state variables; Representing the time; The background error covariance of carbon dioxide emissions by province and department, its dimensions and state variables Consistent; These are carbon dioxide emission adjustment coefficients broken down by province and department; It is the first The observation vector at time t, This represents the assimilation time window, within which all multi-source carbon satellite observation data are used as constraints in the assimilation inversion. For the first The observation error covariance in diagonal matrix form at time points; For the first The time-based observation operator is used to convert carbon dioxide emissions into carbon dioxide concentration through horizontal diffusion, turbulent diffusion, dry / wet deposition, and emission processes. The concentration was converted to units consistent with multi-source carbon satellite observation data, and spatial matching between observation sites and model formats was performed; superscript Represents the transpose of a vector;

[0017] The a priori carbon dioxide emission sources include: provincial and departmental a priori anthropogenic carbon dioxide emission sources, provincial wildfire carbon emission sources, provincial aircraft carbon emission sources, provincial ship carbon emission sources, provincial chemical reaction generation sources, terrestrial ecosystem carbon flux sources by vegetation type, and marine carbon flux sources. Carbon dioxide emissions from different sources need to be uniformly processed to the same spatial and temporal resolution as the carbon dioxide emission source assimilation system. For anthropogenic carbon dioxide emissions, intraday and weekly variations are considered; for aircraft, ship, wildfire, and marine carbon fluxes, intraday variations are not considered; for terrestrial ecosystem carbon fluxes, daily and intraday variations are considered; for carbon emissions generated by chemical reactions, monthly resolution is considered, but daily and intraday variations are not considered.

[0018] In one embodiment, spatial matching between observation stations and model formats includes:

[0019] First, the vertical profile of the average nucleus number obtained from multi-source carbon satellite observation data is interpolated to the model grid. The four closest grid positions are matched, and then vertical interpolation is performed. By considering the characteristics of atmospheric vertical distribution, the profile height of the average nucleus number and the height of the model layer are converted into atmospheric pressure height and interpolated by taking the natural logarithm. The interpolated average nucleus number is multiplied by the carbon dioxide concentration of each layer simulated by the model and accumulated to obtain the model carbon dioxide column concentration that is comparable to the carbon dioxide column concentration converted from satellite observation.

[0020] In one embodiment, the observation error covariance is constructed from the statistical relationship between the carbon dioxide dry air mass column concentration in carbon satellite observation data from different sources and the carbon dioxide concentration in ground-based observations. The carbon satellite observation data from different sources are all subjected to hierarchical quality control and fusion processing to remove data with cloud obstruction and abnormal observation angles, and the observation error is allocated according to the observation accuracy of different satellites to constrain its weight ratio in the objective function.

[0021] In one embodiment, when performing statistical analysis on anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in preprocessed multi-source data based on carbon and nitrogen homology constraints to obtain the proportion of anthropogenic carbon and nitrogen emissions by province and department, the method further includes:

[0022] The pre-set activity factors for anthropogenic carbon dioxide emissions and anthropogenic nitrogen dioxide emissions are consistent; and the anthropogenic carbon and nitrogen emission ratios for power plants and factories are set separately based on their respective combustion temperature characteristics.

[0023] In one embodiment, prior artificial carbon dioxide emission sources and pre-processed multi-source data are input into a carbon dioxide emission source assimilation system for assimilation and inversion, including:

[0024] The system inputs prior artificial carbon dioxide emission sources and high-resolution meteorological forecast data from preprocessed multi-source data into the carbon dioxide emission source assimilation system. Multi-source carbon satellite observation data from the preprocessed multi-source data is then used as emission source constraints for assimilation and inversion. Simultaneously, the objective function of the carbon dioxide emission source assimilation system introduces emission source spatiotemporal continuity constraints and industry emission threshold constraints. The emission source spatiotemporal continuity constraints are used to avoid abrupt changes in emission data from neighboring provinces and related departments. The industry emission threshold constraints are used to constrain and match the upper limits of emission conditions for each department. Specifically, for province-department-source combinations with emission intensity higher than the preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.7 to 0.9, prioritizing the assimilation accuracy of high-emission provinces. For province-department-source combinations with emission intensity lower than the preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.3 to 0.6, taking into account the computational efficiency of low-emission provinces. The preset threshold is dynamically set based on industry emission benchmarks within the study area.

[0025] In one embodiment, the carbon dioxide concentration obtained from three-dimensional variational assimilation within a buffer region larger than the study area is used as the initial and boundary conditions of the system, including:

[0026] First, a global chemical model was used to continuously simulate global carbon dioxide concentrations over multiple years, with a simulation period set at 3 to 5 years. During the simulation, the trend of global carbon dioxide concentration changes was monitored in real time until a concentration-mass equilibrium state was reached. The concentration-mass equilibrium state was defined as a global average carbon dioxide concentration fluctuation range of ≤5 µmol / mol over six consecutive months, expressed as follows: In the formula, For the first The global average carbon dioxide concentration over a month. The average global carbon dioxide concentration over a continuous six-month period;

[0027] Subsequently, global carbon dioxide concentration data reaching concentration-mass equilibrium is provided to the regional model, offering accurate global background concentration support for regional, high-resolution anthropogenic carbon dioxide emission retrieval. Simultaneously, to avoid interference from errors in initial and boundary conditions of carbon dioxide concentration on the emission source assimilation results, a buffer region larger than the study area for emission source inversion is selected. The area of ​​the buffer region is 1.5 to 2.0 times the area of ​​the study area and is adapted to the topography and atmospheric transport path characteristics of the study area. A three-dimensional variational assimilation algorithm is used to specifically assimilate the carbon dioxide concentration within the buffer region, prioritizing the correction of concentration gradient deviations at the boundary between the buffer region and the study area. The carbon dioxide concentration obtained from the three-dimensional variational assimilation is then provided to the carbon dioxide emission source assimilation system as initial and boundary conditions. The objective function of the three-dimensional variational assimilation is:

[0028] ;

[0029] In the formula, The objective function value for three-dimensional variational assimilation. The carbon dioxide concentration field in the buffer region. For the concentration background field of the buffer region, This is the boundary between the buffer zone and the study area. This represents the normal concentration gradient at the boundary line. Adjust the weights for gradient correction; For the observation vector, The observation error covariance is in diagonal matrix form. For observation operators.

[0030] An anthropogenic carbon dioxide emission source inversion system based on the carbon-nitrogen homology relationship, the system comprising:

[0031] The data acquisition and preprocessing module is used to acquire and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area.

[0032] The assimilation system construction module is used to construct carbon dioxide emission source assimilation systems by province, department, and source based on four-dimensional variational theory.

[0033] The carbon and nitrogen homology constraint module is used to perform statistical analysis on anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in preprocessed multi-source data based on carbon and nitrogen homology constraints. It obtains the anthropogenic carbon and nitrogen emission ratios by province and department and constructs an anthropogenic carbon and nitrogen emission ratio database. The anthropogenic carbon and nitrogen emission ratio database is applied to high-resolution anthropogenic nitrogen dioxide emission sources by province and department. After ratio conversion, the corresponding high-resolution anthropogenic carbon dioxide emission sources are obtained as prior anthropogenic carbon dioxide emission sources.

[0034] The assimilation and inversion module is used to input prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data into the carbon dioxide emission source assimilation system for assimilation and inversion. The carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area is used as the initial and boundary conditions of the system. The four-dimensional variational algorithm is used to invert and obtain the carbon dioxide emission adjustment coefficients of provinces and departments in the study area, and to correct the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments. Finally, the assimilated high-resolution anthropogenic carbon dioxide emission data is obtained.

[0035] The above-mentioned method and system for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship have the following advantages compared to existing technologies:

[0036] 1. Fully utilize the homogeneity of carbon and nitrogen to invert anthropogenic carbon emissions: By establishing anthropogenic carbon and nitrogen emission ratios by province and sector, high-resolution anthropogenic nitrogen dioxide emission sources with spatial resolution at the kilometer level are transformed into a priori high-resolution anthropogenic carbon dioxide emission sources, significantly improving the accuracy of the prior field and solving the shortcomings of traditional methods such as low spatial resolution and missing sectoral information. Furthermore, by statistically analyzing the anthropogenic carbon and nitrogen emission ratios of different emission sectors (power plants, factories, transportation, and residents), the construction and assimilation adjustment of the prior field by sector are realized, which can accurately distinguish the emission contributions of each sector and provide accurate data support for the formulation of targeted emission reduction policies.

[0037] 2. Achieve high-resolution, province-specific, and department-specific carbon dioxide emission inversion: Based on high-resolution meteorological forecast data and prior fields constrained by carbon and nitrogen homology, and combined with four-dimensional variational assimilation technology, a carbon dioxide emission source assimilation system is constructed that is province-specific, department-specific, and source-specific. The anthropogenic carbon dioxide emission data obtained by the inversion can achieve a spatial resolution of up to the kilometer level, and can accurately capture the spatial distribution characteristics of small-scale, high-intensity carbon dioxide emission sources.

[0038] 3. By inputting prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data into the assimilation system for assimilation and inversion, multi-source observation information can be fully integrated, improving the accuracy and reliability of the anthropogenic carbon dioxide emission source inversion results. At the same time, using a buffer zone larger than the study area for three-dimensional variational assimilation and using the obtained carbon dioxide concentration as the initial and boundary conditions of the system can effectively reduce the interference of boundary effects on the inversion results and improve the stability and accuracy of anthropogenic carbon dioxide emission data within the study area. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating an anthropogenic carbon dioxide emission source inversion method based on the carbon-nitrogen homology relationship in one embodiment.

[0040] Figure 2 This is a schematic diagram of carbon dioxide emissions from a factory sector in a certain region after assimilation adjustments in April 2019, as shown in one embodiment. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0042] In one embodiment, such as Figure 1 As shown, a method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship is provided, including the following steps:

[0043] Step 1: Collect and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources in the study area.

[0044] Step 2: Construct a carbon dioxide emission source assimilation system based on four-dimensional variational theory, categorized by province, department, and source.

[0045] Step 3: Based on the carbon and nitrogen homology constraint, perform statistical analysis on the anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in the preprocessed multi-source data, obtain the anthropogenic carbon and nitrogen emission ratios by province and department, and construct an anthropogenic carbon and nitrogen emission ratio database. Apply the anthropogenic carbon and nitrogen emission ratio database to the high-resolution anthropogenic nitrogen dioxide emission sources by province and department, and obtain the corresponding high-resolution anthropogenic carbon dioxide emission sources through ratio conversion, which serve as the prior anthropogenic carbon dioxide emission sources.

[0046] Step 4: Input the prior anthropogenic carbon dioxide emission sources and the preprocessed multi-source data into the carbon dioxide emission source assimilation system for assimilation and inversion. Use the carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area as the initial and boundary conditions of the system. Use the four-dimensional variational algorithm to invert and obtain the carbon dioxide emission adjustment coefficients of provinces and departments in the study area, and correct the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments. Finally, obtain the assimilated high-resolution anthropogenic carbon dioxide emission data.

[0047] It should be understood that the four-dimensional variational algorithm can achieve optimal adjustment of prior anthropogenic carbon dioxide emission sources by minimizing the objective function of observed and simulated values.

[0048] The aforementioned method for retrieving anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship first collects and preprocesses multi-source data, including carbon emissions, nitrogen emissions, and meteorological data. Then, a carbon dioxide emission source assimilation system is constructed based on four-dimensional variational theory. Subsequently, based on the carbon-nitrogen homology constraint, the proportion of anthropogenic carbon and nitrogen emissions by province and department is statistically analyzed, and combined with high-resolution anthropogenic nitrogen dioxide emission sources, high-resolution anthropogenic carbon dioxide emission sources by province and department are obtained. Finally, the prior anthropogenic emission sources and the preprocessed multi-source data are input into the assimilation system for assimilation and inversion. This method can fully integrate multi-source observation information, improving the accuracy and reliability of the carbon dioxide emission inversion results. Simultaneously, a buffer zone larger than the study area is used for three-dimensional variational assimilation, and the resulting concentration field is used as the initial and boundary conditions of the system. This effectively reduces the interference of boundary effects on the inversion results, improves the stability and accuracy of the carbon dioxide concentration and emission fields within the inversion area, and ultimately obtains high-resolution, province-specific, and department-specific accurate anthropogenic carbon dioxide emission data. Overcome the shortcomings of existing methods for retrieving anthropogenic carbon dioxide emission sources, such as low resolution, insufficient utilization of the common source characteristics of carbon and nitrogen, and inability to accurately retrieve data by province and department, and achieve rapid updating and optimization of high-resolution anthropogenic carbon dioxide emission sources in typical regions.

[0049] In one embodiment, multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area are collected, including:

[0050] The study collected multi-source carbon satellite observation data, terrestrial ecosystem carbon flux data by vegetation type, marine carbon flux data, high-resolution weather forecast data, wildfire carbon emission data by province, aircraft carbon emission data by province, ship carbon emission data by province, carbon emission data generated by chemical reactions by province, anthropogenic carbon dioxide emission data by province and anthropogenic nitrogen dioxide emission data by province and sector. Among them, the emitting sectors of anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data include power plants, factories, transportation and residential areas; the vegetation types of terrestrial ecosystem carbon flux data include evergreen broad-leaved, evergreen coniferous, bare soil and farmland soil.

[0051] In one embodiment, multi-source data is preprocessed to ensure data quality meets the requirements of subsequent assimilation and inversion. Preprocessing includes: data format standardization, spatiotemporal matching to unify all data to the same time scale and spatial resolution, outlier removal, and data interpolation. Specifically, this can be achieved through... The criteria remove outliers from the observed data and use Kriging interpolation to fill in missing data.

[0052] In one embodiment, considering the scarcity of existing ground-based carbon dioxide observation data and the limited coverage of satellite observations due to orbital band width constraints, which cannot provide high spatiotemporal resolution observation data to meet the carbon emission inversion constraints, the application of traditional grid-by-grid constraint methods would result in insufficient spatial representativeness of the observation data. Furthermore, my country's emission reduction and pollution control policies are implemented at the provincial level, and the effectiveness of emission reduction and carbon reduction varies across different industries and sectors. Therefore, this method specifically designs the aforementioned multi-dimensional carbon dioxide emission source assimilation system, specifically constructing a province-specific, sector-specific, and source-specific carbon dioxide emission source assimilation system based on four-dimensional variational theory, including:

[0053] Based on four-dimensional variational theory, a carbon dioxide emission source assimilation system is constructed, categorized by province, sector, and source. This system centers on regional atmospheric transport models and uses preprocessed multi-source data, including province- and sector-specific anthropogenic carbon dioxide emission data, province-specific wildfire carbon emission data, province-specific aircraft carbon emission data, province-specific ship carbon emission data, province-specific carbon emission data from chemical reactions, terrestrial ecosystem carbon flux data categorized by vegetation type, and marine carbon flux data, as state variables. The objective function of the carbon dioxide emission source assimilation system is then constructed, expressed as:

[0054] ;

[0055] in, The objective function value; For state variables, These are background emission sources, also known as prior carbon dioxide emission sources. It is the optimized emissions. This represents the ratio between the prior optimized emissions and the background emissions, at which point the emission sources have not yet been optimized. ; Representing the One state variable, Indicates the total number of state variables; Representing the time; The background error covariance of carbon dioxide emissions by province and department, its dimensions and state variables Consistent; This is the carbon dioxide emission adjustment coefficient by province and department. This coefficient is set because the first and second terms on the right side of the formula are the background error term and the observation error term, respectively. Since this application uses the change in the ratio of carbon dioxide emissions from different provinces and departments as the state variable, its cost function is much smaller than the traditional cost function constructed with grid point emissions as the state variable. In order to balance the values ​​of the background error term and the observation term and facilitate the convergence of the cost function, a correction coefficient is added on the basis of the background term. It is the first The observation vector at time t, This represents the assimilation time window, within which all multi-source carbon satellite observation data are used as constraints in the assimilation inversion. For the first The observation error covariance in diagonal matrix form at time t is used to balance computational speed and accuracy. For the first The time-based observation operator is used to convert carbon dioxide emissions into carbon dioxide concentrations through processes such as horizontal diffusion, turbulent diffusion, dry / wet deposition, and emission. The concentration was converted to units consistent with multi-source carbon satellite observation data, and spatial matching between observation sites and model formats was performed; superscript This represents the transpose of a vector.

[0056] The a priori carbon dioxide emission sources include: provincial and departmental anthropogenic carbon dioxide emission sources, provincial wildfire carbon emission sources, provincial aircraft carbon emission sources, provincial ship carbon emission sources, provincial chemical reaction generation sources, terrestrial ecosystem carbon flux sources by vegetation type, and marine carbon flux sources. Carbon dioxide emissions from different sources need to be uniformly processed to the same spatial and temporal resolution as the carbon dioxide emission source assimilation system. For anthropogenic carbon dioxide emissions, intraday and weekly variations are considered; for aircraft, ship, wildfire, and marine carbon fluxes, intraday variations are not considered; for terrestrial ecosystem carbon fluxes, daily and intraday variations are considered; for carbon emissions generated by chemical reactions, which account for a relatively small proportion, monthly resolution is considered, but daily and intraday variations are not considered.

[0057] It should be understood that, compared with conventional atmospheric emission inversion and assimilation systems, the carbon dioxide emission source assimilation system constructed by this method has significantly reduced the number of state variables. Furthermore, the design of these state variables is tailored to emission sources, provinces, and vegetation types, specifically addressing the technical challenge of the inapplicability of traditional grid-by-grid carbon emission inversion methods due to the scarcity of carbon concentration observation data. This system can accurately capture the dynamic changes in carbon dioxide emission sources across different sectors and effectively analyze the temporal patterns and departmental differences in the implementation of pollution reduction and carbon lowering policies in various provinces.

[0058] In one embodiment, spatial matching between observation stations and model formats includes:

[0059] First, the vertical profile of the average kernel obtained from satellite observations is interpolated to the model grid. The four closest grid positions are matched, and then vertical interpolation is performed. By considering the characteristics of the vertical distribution of the atmosphere, the profile height of the average kernel and the height of the model layer are converted into atmospheric pressure height and interpolated by taking the natural logarithm (ln). The interpolated average kernel is multiplied by the carbon dioxide concentration of each layer simulated by the model and summed to obtain the model carbon dioxide column concentration that is comparable to the carbon dioxide column concentration converted from satellite observations.

[0060] In one embodiment, the observation error covariance is constructed from the statistical relationship between the carbon dioxide dry air mass column concentration in carbon satellite observation data from different sources and the carbon dioxide concentration observed from ground-based data. Specifically, GOSAT, GOSAT2, OCO-2, and OCO-3 carbon satellite observation data can be used. The observation errors of different carbon satellite observation data versions and their observed carbon dioxide dry air mass mixed concentrations are shown in Table 1.

[0061] Table 1. Carbon satellite observation data versions from different sources and their observation errors in the mixed concentration of carbon dioxide in dry air.

[0062]

[0063] Among them, carbon satellite observation data from different sources are all subjected to hierarchical quality control and fusion processing to remove data with cloud obstruction and abnormal observation angles, and the observation error is allocated according to the observation accuracy of different satellites to constrain its weight ratio in the objective function.

[0064] In one embodiment, when performing statistical analysis on anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in preprocessed multi-source data based on carbon and nitrogen homology constraints to obtain the proportion of anthropogenic carbon and nitrogen emissions by province and department, the method further includes:

[0065] The pre-set level activity factors for anthropogenic carbon dioxide emissions and anthropogenic nitrogen dioxide emissions are consistent to avoid the problem of ambiguous definition of cross-departmental emission source attribution and to accurately quantify the emission intensity of each emission sector. Furthermore, considering that the proportion of anthropogenic carbon and nitrogen emissions is significantly affected by the emission sector, especially by the differences in carbon and nitrogen emissions caused by temperature during combustion, the anthropogenic carbon and nitrogen emission proportions for power plants and factories are set separately based on their respective combustion temperature characteristics.

[0066] It should be understood that anthropogenic carbon dioxide emissions mainly originate from the combustion of fossil fuels. During combustion, fossil fuels also release nitrogen dioxide. Nitrogen dioxide is not only abundant in data but also has a short lifespan; its concentration near anthropogenic emission sources far exceeds background concentrations, making it a suitable tracer for anthropogenic carbon dioxide emissions. Utilizing carbon and nitrogen source-constrained synergistic inversion to derive anthropogenic carbon dioxide emissions can compensate for the shortcomings of carbon dioxide observation and solve the problems of low spatial resolution and missing departmental information in traditional prior emission sources.

[0067] In one embodiment, prior artificial carbon dioxide emission sources and pre-processed multi-source data are input into a carbon dioxide emission source assimilation system for assimilation and inversion, including:

[0068] The system inputs prior artificial carbon dioxide emission sources and high-resolution meteorological forecast data from preprocessed multi-source data into the carbon dioxide emission source assimilation system. Multi-source carbon satellite observation data from the preprocessed multi-source data is then used as emission source constraints for assimilation and inversion. Simultaneously, the objective function of the carbon dioxide emission source assimilation system introduces emission source spatiotemporal continuity constraints and industry emission threshold constraints. The emission source spatiotemporal continuity constraints are used to avoid abrupt changes in emission data from neighboring provinces and related departments. The industry emission threshold constraints are used to constrain and match the upper limits of emission conditions for each department. Specifically, for province-department-source combinations with emission intensity higher than the preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.7 to 0.9, prioritizing the assimilation accuracy of high-emission provinces. For province-department-source combinations with emission intensity lower than the preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.3 to 0.6, taking into account the computational efficiency of low-emission provinces. The preset threshold is dynamically set based on industry emission benchmarks within the study area.

[0069] In one embodiment, the carbon dioxide concentration obtained from three-dimensional variational assimilation within a buffer region larger than the study area is used as the initial and boundary conditions of the system, including:

[0070] First, a global chemical model was used to continuously simulate global carbon dioxide concentrations over multiple years, with a simulation period set at 3 to 5 years. During the simulation, the trend of global carbon dioxide concentration changes was monitored in real time until a concentration-mass equilibrium state was reached. The concentration-mass equilibrium state was defined as a global average carbon dioxide concentration fluctuation range of ≤5 µmol / mol over six consecutive months, expressed as follows: In the formula, For the first The global average carbon dioxide concentration over a month. It is the global average carbon dioxide concentration over a continuous 6-month period; the formula for judging the mass balance of this concentration can dynamically calculate the average relative fluctuation of global carbon dioxide concentration, thus achieving accurate determination of mass balance.

[0071] Subsequently, global carbon dioxide concentration data reaching concentration-mass equilibrium is provided to the regional model, offering accurate global background concentration support for regional, high-resolution anthropogenic carbon dioxide emission retrieval. Simultaneously, to avoid interference from errors in initial and boundary conditions of carbon dioxide concentration on the emission source assimilation results, a buffer region larger than the study area for emission source inversion is selected. The area of ​​the buffer region is 1.5 to 2.0 times the area of ​​the study area and is adapted to the topography and atmospheric transport path characteristics of the study area, avoiding interference from concentration transport errors between regions. A three-dimensional variational assimilation algorithm is used to specifically assimilate the carbon dioxide concentration within the buffer region, prioritizing the correction of concentration gradient deviations at the boundary between the buffer region and the study area. The carbon dioxide concentration obtained from the three-dimensional variational assimilation is then provided to the carbon dioxide emission source assimilation system as initial and boundary conditions. The objective function of the three-dimensional variational assimilation is:

[0072] ;

[0073] In the formula, The objective function value for three-dimensional variational assimilation. The carbon dioxide concentration field in the buffer region. For the concentration background field of the buffer region, This is the boundary between the buffer zone and the study area. This represents the normal concentration gradient at the boundary line. Gradient correction weights (taken as 0.8~1.0, prioritizing gradient bias correction); For the observation vector, The observation error covariance is in diagonal matrix form. For observation operators.

[0074] It should be understood that the objective function of this three-dimensional variational assimilation can specifically suppress abrupt concentration changes at the boundary between the buffer region and the study region, improve the accuracy of boundary conditions, and provide the assimilated high-precision carbon dioxide concentration data to the carbon dioxide emission source assimilation system as the initial and boundary conditions for the operation of the carbon dioxide emission source assimilation system, thereby further improving the accuracy and stability of emission source inversion.

[0075] Furthermore, to verify the performance of the method proposed in this application, the following detailed description uses the application of the steps proposed in this method in the assimilation and inversion of anthropogenic carbon dioxide emissions in my country in April 2019 as an example. The multi-source data collected by this method in this embodiment includes:

[0076] We collected anthropogenic carbon dioxide emissions data by province and sector, anthropogenic nitrogen dioxide emissions data by province and sector, aircraft carbon emissions data by province (GCPv2022.2), ship carbon emissions data by province, carbon emissions from chemical reactions by province, terrestrial ecosystem carbon flux data by vegetation type (ORCHIDEE simulation results), marine carbon flux data (CLMELS), and wildfire carbon emissions data by province (GFEDv4). We then uniformly processed these emissions data to the same spatial and temporal resolution as the assimilation system. For anthropogenic carbon dioxide emissions, we considered intraday and weekly variations; for aircraft, ship, wildfire, and marine carbon fluxes, we did not consider intraday variations; for terrestrial ecosystem carbon fluxes, we considered daily and intraday variations; and for carbon emissions from chemical reactions, we considered monthly resolution but did not consider daily or intraday variations. We collected observational data from GOSAT, GOSAT2, OCO-2, and OCO-3 multi-source carbon satellites. Preprocessing of the multi-source carbon satellite data included data format standardization, spatiotemporal matching to unify all data to the same time scale and spatial resolution, outlier removal, and data interpolation. We also collected high-resolution meteorological forecast data and processed it to provide regional atmospheric chemistry models for use in driving assimilation and forecast simulations.

[0077] In this embodiment, the steps proposed in this method utilize the prior artificial carbon dioxide emission adjustment coefficient of a certain region's factory sector to correct for carbon dioxide emission sources. The adjusted carbon dioxide emissions from the factory sector are as follows: Figure 2 As shown. By Figure 2 As shown, this method can effectively address the issue of improving the accuracy of carbon emissions; the method is simple and easy to implement, requires little computation and funding, and the anthropogenic carbon dioxide emissions obtained can also be used for carbon dioxide simulation and forecasting research, providing data support for carbon trading, emission reduction accounting, climate negotiations, and other purposes.

[0078] In one embodiment, an anthropogenic carbon dioxide emission source inversion system based on the carbon-nitrogen homology relationship is provided, comprising:

[0079] The data acquisition and preprocessing module is used to acquire and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area.

[0080] The assimilation system construction module is used to construct carbon dioxide emission source assimilation systems by province, department, and source based on four-dimensional variational theory.

[0081] The carbon and nitrogen homology constraint module is used to perform statistical analysis on anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in preprocessed multi-source data based on carbon and nitrogen homology constraints. It obtains the anthropogenic carbon and nitrogen emission ratios by province and department and constructs an anthropogenic carbon and nitrogen emission ratio database. The anthropogenic carbon and nitrogen emission ratio database is applied to high-resolution anthropogenic nitrogen dioxide emission sources by province and department. After ratio conversion, the corresponding high-resolution anthropogenic carbon dioxide emission sources are obtained as prior anthropogenic carbon dioxide emission sources.

[0082] The assimilation and inversion module is used to input prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data into the carbon dioxide emission source assimilation system for assimilation and inversion. The carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area is used as the initial and boundary conditions of the system. The four-dimensional variational algorithm is used to invert and obtain the carbon dioxide emission adjustment coefficients of provinces and departments in the study area, and to correct the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments. Finally, the assimilated high-resolution anthropogenic carbon dioxide emission data is obtained.

[0083] Specific limitations regarding the anthropogenic carbon dioxide emission source inversion system based on the carbon-nitrogen homology relationship can be found in the limitations of the anthropogenic carbon dioxide emission source inversion method based on the carbon-nitrogen homology relationship above, and will not be repeated here. Each module in the above-mentioned anthropogenic carbon dioxide emission source inversion system based on the carbon-nitrogen homology relationship can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0084] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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.

[0085] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. A method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship, characterized in that, The method includes: Collect and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources in the study area. A carbon dioxide emission source assimilation system is constructed based on four-dimensional variational theory, categorized by province, department, and source. Based on the carbon and nitrogen homology constraint, statistical analysis was performed on the anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in the preprocessed multi-source data. The anthropogenic carbon and nitrogen emission ratios by province and department were obtained and an anthropogenic carbon and nitrogen emission ratio database was constructed. The anthropogenic carbon and nitrogen emission ratio database was applied to the high-resolution anthropogenic nitrogen dioxide emission sources by province and department. The corresponding high-resolution anthropogenic carbon dioxide emission sources were obtained by ratio conversion and used as the prior anthropogenic carbon dioxide emission sources. Prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data are input into a carbon dioxide emission source assimilation system for assimilation and inversion. The carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area is used as the initial and boundary conditions of the system. The carbon dioxide emission adjustment coefficients of provinces and departments in the study area are obtained by using a four-dimensional variational algorithm, and the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments are corrected. Finally, high-resolution anthropogenic carbon dioxide emission data after assimilation are obtained.

2. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 1, characterized in that, Collect and obtain multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area, including: The study collected multi-source carbon satellite observation data, terrestrial ecosystem carbon flux data by vegetation type, marine carbon flux data, high-resolution weather forecast data, wildfire carbon emission data by province, aircraft carbon emission data by province, ship carbon emission data by province, carbon emission data generated by chemical reactions by province, anthropogenic carbon dioxide emission data by province and anthropogenic nitrogen dioxide emission data by province and department; wherein, the emitting sectors of the anthropogenic carbon dioxide emission data and the anthropogenic nitrogen dioxide emission data include power plants, factories, transportation and residential areas; the vegetation types of the terrestrial ecosystem carbon flux data include evergreen broad-leaved, evergreen coniferous, bare soil and farmland soil.

3. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 2, characterized in that, Preprocessing of multi-source data includes: standardizing the data format of multi-source data, spatiotemporal matching processing to unify multi-source data to the same time scale and spatial resolution, outlier removal processing, and data interpolation processing.

4. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 3, characterized in that, Based on four-dimensional variational theory, a carbon dioxide emission source assimilation system is constructed, segmented by province, department, and source, including: Based on four-dimensional variational theory, a carbon dioxide emission source assimilation system is constructed, categorized by province, sector, and source. This system centers on regional atmospheric transport models and uses preprocessed multi-source data, including province- and sector-specific anthropogenic carbon dioxide emission data, province-specific wildfire carbon emission data, province-specific aircraft carbon emission data, province-specific ship carbon emission data, province-specific carbon emission data from chemical reactions, terrestrial ecosystem carbon flux data categorized by vegetation type, and marine carbon flux data, as state variables. The objective function of the carbon dioxide emission source assimilation system is then constructed, expressed as: ； in, The objective function value; For state variables, These are background emission sources, also known as prior carbon dioxide emission sources. It is the optimized emissions. This represents the ratio between the prior optimized emissions and the background emissions, at which point the emission sources have not yet been optimized. ; Representing the One state variable, Indicates the total number of state variables; Representing the time; The background error covariance of carbon dioxide emissions by province and department, its dimensions and state variables Consistent; These are carbon dioxide emission adjustment coefficients broken down by province and department; It is the first The observation vector at time t, This represents the assimilation time window, within which all multi-source carbon satellite observation data are used as constraints in the assimilation inversion. For the first The observation error covariance in diagonal matrix form at time points; For the first The time-based observation operator is used to convert carbon dioxide emissions into carbon dioxide concentration through horizontal diffusion, turbulent diffusion, dry / wet deposition, and emission processes. The concentration was converted to units consistent with multi-source carbon satellite observation data, and spatial matching between observation sites and model formats was performed; superscript Represents the transpose of a vector; The a priori carbon dioxide emission sources include: provincial and departmental anthropogenic carbon dioxide emission sources, provincial wildfire carbon emission sources, provincial aircraft carbon emission sources, provincial ship carbon emission sources, provincial chemical reaction generation sources, terrestrial ecosystem carbon flux sources by vegetation type, and marine carbon flux sources. Carbon dioxide emissions from different sources need to be uniformly processed to the same spatial and temporal resolution as the carbon dioxide emission source assimilation system. For anthropogenic carbon dioxide emissions, intraday and weekly variations are considered; for aircraft, ship, wildfire, and marine carbon fluxes, intraday variations are not considered; for terrestrial ecosystem carbon fluxes, daily and intraday variations are considered; for carbon emissions generated by chemical reactions, monthly resolution is considered, but daily and intraday variations are not considered.

5. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 4, characterized in that, Spatial matching between observation sites and model formats includes: First, the vertical profile of the average nucleus number obtained from multi-source carbon satellite observation data is interpolated to the model grid. The four closest grid positions are matched, and then vertical interpolation is performed. By considering the characteristics of atmospheric vertical distribution, the profile height of the average nucleus number and the height of the model layer are converted into atmospheric pressure height and interpolated by taking the natural logarithm. The interpolated average nucleus number is multiplied by the carbon dioxide concentration of each layer simulated by the model and accumulated to obtain the model carbon dioxide column concentration that is comparable to the carbon dioxide column concentration converted from satellite observation.

6. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 4, characterized in that, The observation error covariance is constructed from the statistical relationship between the carbon dioxide dry air mass column concentration in carbon satellite observation data from different sources and the carbon dioxide concentration in ground-based observations. Among them, the carbon satellite observation data from different sources are all subjected to hierarchical quality control and fusion processing to remove data with cloud obstruction and abnormal observation angles, and the observation error is allocated according to the observation accuracy of different satellites to constrain its weight ratio in the objective function.

7. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 1, characterized in that, When performing statistical analysis on anthropogenic carbon dioxide and nitrogen dioxide emission data from preprocessed multi-source data based on carbon and nitrogen homology constraints to obtain the proportion of anthropogenic carbon and nitrogen emissions by province and department, the method further includes: The pre-set activity factors for anthropogenic carbon dioxide emissions and anthropogenic nitrogen dioxide emissions are consistent; and the anthropogenic carbon and nitrogen emission ratios for power plants and factories are set separately based on their respective combustion temperature characteristics.

8. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 1, characterized in that, Prior anthropogenic carbon dioxide emission sources and pretreated multi-source data are input into the carbon dioxide emission source assimilation system for assimilation and inversion, including: The carbon dioxide emission source assimilation system inputs prior human carbon dioxide emission sources and high-resolution meteorological forecast data from preprocessed multi-source data into the carbon dioxide emission source assimilation system. Multi-source carbon satellite observation data from the preprocessed multi-source data is then used as emission source constraints for assimilation and inversion. Simultaneously, an emission source spatiotemporal continuity constraint term and an industry emission threshold constraint term are introduced into the objective function of the carbon dioxide emission source assimilation system. The emission source spatiotemporal continuity constraint term is used to avoid abrupt changes in emission data from neighboring provinces and related departments. The industry emission threshold constraint term is used to constrain and match the upper limit of emission conditions for each department. Specifically, for province-department-source combinations with emission intensity higher than a preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.7 to 0.9, prioritizing the assimilation accuracy of high-emission provinces. For province-department-source combinations with emission intensity lower than the preset threshold, the emission source spatiotemporal continuity constraint coefficient is assigned a value ranging from 0.3 to 0.6, taking into account the computational efficiency of low-emission provinces. The preset threshold is dynamically set based on industry emission benchmarks within the study area.

9. The method for inverting anthropogenic carbon dioxide emission sources based on the carbon-nitrogen homology relationship according to claim 1, characterized in that, The carbon dioxide concentration obtained from three-dimensional variational assimilation within a buffer region larger than the study area is used as the initial and boundary conditions of the system, including: First, a global chemical model was used to continuously simulate global carbon dioxide concentrations over multiple years, with a simulation period set at 3 to 5 years. During the simulation, the trend of global carbon dioxide concentration changes was monitored in real time until a concentration-mass equilibrium state was reached. This concentration-mass equilibrium state was defined as a global average carbon dioxide concentration fluctuation range of ≤5 µmol / mol over six consecutive months, expressed as follows: In the formula, For the first The global average carbon dioxide concentration over a month. The average global carbon dioxide concentration over a continuous six-month period; Subsequently, global carbon dioxide concentration data reaching a concentration-mass equilibrium state is provided to the regional model, offering accurate global background concentration support for regional, high-resolution anthropogenic carbon dioxide emission retrieval. Simultaneously, to avoid interference from errors in initial and boundary conditions of carbon dioxide concentration on the emission source assimilation results, a buffer region larger than the study area for emission source inversion is selected. This buffer region has an area 1.5 to 2.0 times the study area and is adapted to the topographic and atmospheric transport path characteristics of the study area. A three-dimensional variational assimilation algorithm is used to specifically assimilate the carbon dioxide concentration within the buffer region, prioritizing the correction of concentration gradient deviations at the boundary between the buffer region and the study area. The carbon dioxide concentration obtained from the three-dimensional variational assimilation is then provided to the carbon dioxide emission source assimilation system as initial and boundary conditions. The objective function of the three-dimensional variational assimilation is: ； In the formula, The objective function value for three-dimensional variational assimilation. The carbon dioxide concentration field in the buffer region. For the concentration background field of the buffer region, This is the boundary between the buffer zone and the study area. This represents the normal concentration gradient at the boundary line. Adjust the weights for gradient correction; For the observation vector, The observation error covariance is in diagonal matrix form. For observation operators.

10. A system for retrieving anthropogenic carbon dioxide emission sources based on the common origin relationship of carbon and nitrogen, characterized in that, The system includes: The data acquisition and preprocessing module is used to collect and preprocess multi-source data related to the inversion of anthropogenic carbon dioxide emission sources within the study area. The assimilation system construction module is used to construct carbon dioxide emission source assimilation systems by province, department, and source based on four-dimensional variational theory. The carbon and nitrogen homology constraint module is used to perform statistical analysis on anthropogenic carbon dioxide emission data and anthropogenic nitrogen dioxide emission data in preprocessed multi-source data based on carbon and nitrogen homology constraints. It obtains the anthropogenic carbon and nitrogen emission ratios by province and department and constructs an anthropogenic carbon and nitrogen emission ratio database. The anthropogenic carbon and nitrogen emission ratio database is applied to high-resolution anthropogenic nitrogen dioxide emission sources by province and department. After ratio conversion, the corresponding high-resolution anthropogenic carbon dioxide emission sources are obtained as prior anthropogenic carbon dioxide emission sources. The assimilation and inversion module is used to input prior anthropogenic carbon dioxide emission sources and preprocessed multi-source data into the carbon dioxide emission source assimilation system for assimilation and inversion. The carbon dioxide concentration obtained by three-dimensional variational assimilation in a buffer area larger than the study area is used as the initial and boundary conditions of the system. The four-dimensional variational algorithm is used to invert and obtain the carbon dioxide emission adjustment coefficients of provinces and departments in the study area, and to correct the prior anthropogenic carbon dioxide emission sources of the corresponding provinces and departments. Finally, the assimilated high-resolution anthropogenic carbon dioxide emission data is obtained.

Images