A high-resolution assimilation inversion system and method for methane emission sources

Abstract

The present invention discloses a high-resolution assimilation inversion system and method for methane emission sources, including a data preprocessing module, a regional atmospheric transmission mode module, a methane emission source diagnosis module, an ensemble Kalman smoothing module, a sub-sector accounting module of assimilated sources, and an output module. The ensemble forecast methane concentration field under the regional atmospheric transmission mode is obtained through the regional atmospheric transmission mode module; the methane emission source is constructed through the methane emission source diagnosis module; the methane assimilation source is obtained through the ensemble Kalman smoothing module; the bottom-up methane emission inventory is used as a priori estimation through the sub-sector accounting module of the assimilated source, and the assimilated emissions of each department are obtained on the basis of the methane assimilated source. The present invention has the advantages of high estimation accuracy, improved spatiotemporal resolution, etc., and can solve the uncertainty problem of bottom-up priori emission source accounting.

Description

Technical Field

[0001] The invention relates to a high-resolution assimilation inversion system and method for methane emission sources, belonging to the technical field of methane monitoring. Background Art

[0002] Methane is the second most powerful greenhouse gas after carbon dioxide. Its importance lies in its greater warming potential (28 times greater than carbon dioxide over 100 years), shorter atmospheric lifetime (9.1 ± 0.9 years), and higher chemical activity (compared to NO x , CO, O3 and other chemical reactions). Therefore, reducing methane emissions is an effective way to reduce greenhouse gas concentrations and their associated radiative forcing. Due to the limited transparency and accuracy of bottom-up inventories, the IPCC (Intergovernmental Panel on Climate Change) recommends top-down verification, that is, using independent atmospheric concentration monitoring to invert emission sources, especially using satellite monitoring, to support the international community in responding to the climate change crisis.

[0003] The resolution and accuracy of methane inversion depend on the quality of observations, the accuracy of the transport model, and the accuracy of the inversion scheme. At present, most estimates of my country's methane emissions based on atmospheric inversion use global atmospheric transport models to correlate emissions with observations at relatively coarse temporal and spatial resolutions, or use Lagrangian models to quantify the sensitivity of receptors to sources. On the one hand, methane emissions come from the superposition of emissions from different sectors. The main anthropogenic sources include fossil fuel extraction, agriculture, waste, etc., and the main natural sources come from wetlands. The range of anthropogenic emissions given in various bottom-up inventories in my country is 44.4 to 57.5 , while the official report of our country in 2014 was 53.57 , there is still great uncertainty in the sources of anthropogenic methane emissions. On the other hand, methane is a short-lived greenhouse gas, and OH radicals are its main sink (sink means absorption or conversion from the atmosphere). The Eulerian atmospheric chemical transport model can realize the chemical conversion of OH radicals to methane, but the global atmospheric chemical transport model is often simplified. Processing of atmospheric data in 2020 found that 34% of the changes in atmospheric methane concentrations were caused by changes in OH radicals. Small-scale inversions based on Lagrangian models often ignore chemical reactions during atmospheric transport. Therefore, if the uncertainty of the forward simulation of the atmospheric chemical transport model can be reduced, and the assimilation scheme suitable for high-resolution methane emissions can be carefully designed, the inversion of methane emission sources will be more accurate. Summary of the invention

[0004] Purpose of the invention: In order to overcome the deficiencies in the prior art, the present invention provides a high-resolution assimilation inversion system and method for methane emission sources, which has the advantages of high estimation accuracy and improved spatiotemporal resolution, and can solve the uncertainty problem of bottom-up a priori emission source accounting.

[0005] Technical solution: To achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A high-resolution assimilation inversion method for methane emission sources comprises the following steps:

[0007] Step 1: Obtain a priori emission sources and satellite methane column concentrations.

[0008] Step 2: Preprocess the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation. Set the regional meteorological model to provide the meteorological field.

[0009] Step 3: The emission, transport, diffusion and chemical transformation processes of methane in the regional atmospheric transmission model are constructed to obtain the regional atmospheric transmission model of methane, so that the chemical transformation process includes not only NO X The impact of changes in CO on methane and the increase in the reaction of O3 related to methane are calculated. The methane regional atmospheric transport model obtains an ensemble forecast methane concentration field based on the prior emission sources, the methane assimilation sources at the previous moment, and the meteorological field.

[0010] Step 4: construct a methane emission source diagnosis model based on the prior emission sources, the methane assimilation sources at the previous moment, and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission sources.

[0011] Step 5: construct a methane ensemble Kalman smoothing module based on the ensemble forecast methane concentration field and the ensemble forecast methane emission source, take the satellite methane column concentration as satellite observation, realize methane emission assimilation inversion, and obtain the methane assimilation source at this moment.

[0012] Step 6: By constructing a sub-sector accounting module for fine-scale assimilation sources, the bottom-up methane emission inventory is used as a priori estimate of sub-sector emissions. Based on the methane assimilation sources at this moment, the emission model of each sector after assimilation is established according to the sector proportion in the prior emissions, and then the assimilated emissions of each sector are obtained.

[0013] Preferably: the emission models of each department after assimilation are as follows:

[0014] ;

[0015] in, Indicates time t The grid point position of the time mode ( i, j ) of the updated department ip Emissions estimates, ip represents the emission sector, Indicates time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip emissions; Indicates time t The grid point position of the time mode ( i, j ) of the total prior emissions, Indicates time t The grid point position of the time mode ( i, j ) of total assimilated emissions.

[0016] Preferably, the methane simulation time step output by the methane regional atmospheric transport model is 1 hour, and the horizontal resolution range is 1-64 km.

[0017] Preferably, the input data of the methane emission source diagnosis model include a priori emission sources, ensemble forecast methane concentration field and ensemble average analysis field of methane emissions at the same time every day in the previous several days.

[0018] Preferably: the emission sectors include coal methane emission sectors, petroleum methane emission sectors, natural gas methane emission sectors, animal husbandry methane emission sectors, rice field methane emission sectors, and waste methane emission sectors.

[0019] A high-resolution assimilation inversion system for methane emission sources includes an input module, a data preprocessing module, a regional atmospheric transmission model module, a methane emission source diagnosis module, a methane ensemble Kalman smoothing module, a sub-sector accounting module for assimilated sources, and an output module, wherein:

[0020] The input module is used to input a priori emission sources and satellite methane column concentrations.

[0021] The data preprocessing module is used to preprocess the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation, and to set the regional meteorological model to provide the meteorological field.

[0022] The regional atmospheric transmission model module is used to construct the emission, transportation, diffusion and chemical conversion processes of methane in the regional atmospheric transmission model to obtain a methane regional atmospheric transmission model. The methane regional atmospheric transmission model obtains an ensemble forecast methane concentration field based on a priori emission sources, methane assimilation sources at the previous moment and meteorological fields.

[0023] The methane emission source diagnosis module is used to construct a methane emission source diagnosis model according to the prior emission source, the methane assimilation source at the previous moment and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission source.

[0024] The methane ensemble Kalman smoothing module implements methane emission assimilation inversion based on the ensemble forecast methane concentration field, the ensemble forecast methane emission source and the satellite methane column concentration to obtain the methane assimilation source at this moment.

[0025] The sub-sector accounting module of the assimilated source uses the bottom-up methane emission inventory as a priori estimate of sub-sector emissions, and on the basis of the methane assimilated source at this moment, establishes an emission model for each sector after assimilation according to the sector proportion in the prior emissions, and then obtains the emissions of each sector after assimilation.

[0026] The output module is used to output the assimilated emissions of each sector.

[0027] Preferably: the emission models of each department after assimilation are as follows:

[0028] ;

[0029] in, Indicates time t The grid point position of the time mode ( i, j ) of the updated department ip Emissions estimates, ip represents the emission sector, Indicates time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip emissions; Indicates time t The grid point position of the time mode ( i, j ) of the total prior emissions, Indicates time t The grid point position of the time mode ( i, j ) of total assimilated emissions.

[0030] Preferably, the methane simulation time step output by the methane regional atmospheric transmission model module is 1 hour, and the horizontal resolution range is 1-64 km.

[0031] Preferably, the input data of the methane emission source diagnosis module includes a priori emission sources, ensemble forecast methane concentration field and ensemble average analysis field of methane emissions at the same time every day in the previous several days.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. The present invention realizes the inversion of methane emissions based on the regional atmospheric chemical transport model, which takes into account the fine-scale methane emission, transport, diffusion, and chemical conversion processes, and can better obtain high temporal and spatial resolution methane concentration forward simulation results.

[0034] 2. The present invention develops methane emission source assimilation based on the ensemble Kalman smoothing module, which can achieve high-resolution assimilation at hourly and kilometer scales. Its emission source diagnostic model has hourly resolution, its regional atmospheric chemistry model has hourly resolution, and its horizontal resolution range is 1-64km. Its assimilation scheme adopts the coordinated assimilation of concentration and flux, which can reduce the mutual influence of the uncertainties of the two.

[0035] 3. The bottom-up calculation of the methane emission inventory in the present invention can reach the kilometer scale and hourly scale, which is significantly higher than the current emission source inversion accuracy based on the global atmospheric chemical transport model and Lagrangian model.

[0036] In summary, the present invention can be used as a technical channel for satellite monitoring of methane emissions verification, with the advantages of high estimation accuracy and improved temporal and spatial resolution, and can solve the uncertainty problem of bottom-up a priori emission source accounting. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a flow chart of Example 1 of the present invention.

[0038] Figure 2 This is a structural block diagram of embodiment 2 of the present invention. DETAILED DESCRIPTION

[0039] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent forms of modifications to the present invention by those skilled in the art all fall within the scope defined by the claims attached to this application. Example 1

[0040] This example proposes a high-resolution assimilation inversion method for methane emission sources, such as Figure 1 As shown, the following steps are included:

[0041] Step 1, data acquisition: obtain a priori emission sources and satellite methane column concentrations.

[0042] Step 2, data preprocessing: preprocess the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation.

[0043] Preprocess the satellite methane column concentration into the format required for assimilation. Set the simulation area and horizontal and vertical resolution of the regional meteorological model RAMS (full name Regional Atmospheric Modeling System), and conduct regional meteorological model RAMS simulation to provide a meteorological field, which corresponds to the horizontal and vertical resolution of the regional atmospheric transport model CMAQ (full name Community Multi-scale Air Quality). Complete the emission source and initial boundary condition settings required for methane simulation in the regional atmospheric transport model CMAQ. The methane simulation time step output by the methane regional atmospheric transport model is 1 hour, and the horizontal resolution range is 1-64km.

[0044] Step 3: Obtain the ensemble forecast methane concentration field:

[0045] The regional atmospheric transport model CMAQ is an Eulerian regional atmospheric chemical transport model. The traditional CB05 (full name Carbon Bond Mechanism V) scheme includes 52 chemical substances and 156 core reactions, among which methane is set as a fixed constant. In view of this, in order to improve the forward simulation process in assimilation, we constructed the emission, transportation, diffusion and chemical transformation process of methane in the regional atmospheric transport model CMAQ to obtain the methane regional atmospheric transport model, so that in the chemical transformation process, not only NO X The impact of changes in CO on methane also increases the reaction of O3 related to methane. By newly adding the chemical reaction mechanism of methane, the spatiotemporal distribution of atmospheric methane concentration with high spatiotemporal resolution can be more accurately described. The methane regional atmospheric transmission model is input into the methane regional atmospheric transmission model based on the prior emission sources, the methane assimilation sources at the previous moment, and the meteorological field to obtain the ensemble forecast methane concentration field under the regional atmospheric transmission model. The forward model of the concentration in the methane regional atmospheric transmission model has a spatiotemporal resolution of kilometers and hours.

[0046] Step 4: Obtain ensemble forecast of methane emission sources:

[0047] In order to construct the methane emission source on the basis of the methane prior emission source, a methane emission source diagnosis model is constructed. The methane emission source diagnosis module constructs the methane emission source on the basis of the methane prior emission source through the methane emission source diagnosis model. The methane emission source diagnosis model is constructed according to the prior emission source, the methane assimilation source at the previous moment, and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission source. The ensemble average analysis field of methane emissions at the same time every day in the previous few days is obtained through the prior emission source and the methane assimilation source at the previous moment. Therefore, the input data of the methane emission source diagnosis model includes the prior emission source, the ensemble forecast methane concentration field, and the ensemble average analysis field of methane emissions at the same time every day in the previous few days.

[0048] Methane emission source diagnostic model:

[0049] ;

[0050] in, represents the ensemble forecast of methane emission sources, represents the covariance inflation factor, represents the ensemble forecast methane concentration field at time t, represents the ensemble average analysis field of methane emissions at the same time of day in the previous several days, , , N is the total number of sample sets, represents the ensemble average of the CH4 concentration scale parameter at time t, represents the a priori emission source, Indicates the previous The ensemble mean analysis field of CH4 flux per day, , represents the analysis field, Indicates the forecast field.

[0051] The methane emission source diagnostic model has hourly resolution and highlights the hourly variations of methane emission sources.

[0052] Step 5: Assimilation and inversion to obtain the methane assimilation source:

[0053] Ensemble Kalman Smooth module EnKS (short for Ensemble Kalman Smooth).

[0054] The assimilation algorithm of this embodiment is developed based on the Ensemble Square RootFilter square (EnSRF), which is derived from the Ensemble Kalman Filter (EnKF) assimilation system operated by NOAA (National Oceanic and Atmospheric Administration). Considering that the Ensemble Kalman Smooth (EnKS) module allows sequential processing of observations within a time window, which is more in line with the impact mechanism of emission sources on concentration, this embodiment further expands the Ensemble Kalman Root Filter EnSRF with the characteristics of the Ensemble Kalman Smooth module EnKS, and further develops the assimilation inversion of methane emission sources within the assimilation framework. The methane Ensemble Kalman Smooth module is constructed based on the ensemble forecast methane concentration field and the ensemble forecast methane emission source. The assimilation system designs the coordinated assimilation of methane concentration and flux, thereby reducing the mutual influence of the uncertainties of the two, and its state variables , represents the methane concentration, Represents methane flux. The methane ensemble Kalman smoothing module enables the coordinated assimilation of methane concentration and flux, and realizes the assimilation inversion of methane emissions based on emission sources, ensemble forecast methane concentration fields, and satellite observations (satellite methane column concentrations), and obtains methane assimilation sources. Therefore, under the framework of this assimilation algorithm, and based on the above-mentioned ensemble background fields (including emission sources and concentration fields), satellite observations (satellite methane column concentrations), etc., the assimilation inversion of methane emissions is realized, and the impact of upstream emission sources on the current concentration is comprehensively considered.

[0055] Step 6: Sector-by-sector accounting of assimilated sources:

[0056] In this embodiment, the assimilated source sector accounting module uses the bottom-up methane emission inventory as a priori estimate, and on the basis of the methane assimilated source at this moment, establishes the assimilated emission model of each department according to the sector proportion in the priori emission, and then obtains the assimilated emissions of each department. The bottom-up methane emission inventory refers to a list obtained by classifying the country's major carbon emission sources, and then constructing sub-directories under the department classification until all emission sources are included for accounting.

[0057] Emission models of various sectors after assimilation:

[0058] ;

[0059] in, Indicates time (month) t The grid point position of the time mode ( i, j ) of the updated department ip Emission estimation, the assimilated emissions of each sector are Time (month) t The grid point position of the time mode ( i, j ) of the updated department ip Emissions estimates , ip It indicates the emission sector, which includes coal methane emission sector, oil methane emission sector, natural gas methane emission sector, animal husbandry methane emission sector, rice field methane emission sector, waste methane emission sector, etc. Indicates time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip emissions; Indicates time t The grid point position of the time mode ( i, j ) of the total prior emissions, time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip Emissions and timet The grid point position of the time mode ( i, j ) of the total prior emission Derived from bottom-up methane emissions inventories, Indicates time t The grid point position of the time mode ( i, j ) of total assimilated emissions, time t The grid point position of the time mode ( i, j ) of total assimilated emissions Refers to time t Time methane assimilation source.

[0060] In this way, combining the bottom-up methane emission inventory and the top-down (methane assimilation source) assimilation inversion, a further methane sectoral emission accounting is given. The difference between the a priori and assimilated sectoral emission sources can be attributed to the assimilation constraint effect of satellite observations. Example 2

[0061] This example proposes a high-resolution assimilation inversion system for methane emission sources, such as Figure 2 As shown, it includes an input module, a data preprocessing module, a regional atmospheric transmission model module, a methane emission source diagnosis module, a methane ensemble Kalman smoothing module, a sub-sector accounting module of assimilated sources, and an output module, among which:

[0062] The input module is used to input a priori emission sources and satellite methane column concentrations.

[0063] The data preprocessing module is used to preprocess the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation, and to set the regional meteorological model to provide the meteorological field.

[0064] The regional atmospheric transmission model module is used to construct the emission, transportation, diffusion and chemical conversion process of methane in the regional atmospheric transmission model to obtain the methane regional atmospheric transmission model, so that the chemical conversion process includes not only NO X The impact of changes in CO on methane and the increase in the reaction of O3 related to methane. The methane regional atmospheric transport model obtains an ensemble forecast methane concentration field based on the prior emission sources, the methane assimilation sources at the previous moment, and the meteorological field.

[0065] The methane emission source diagnosis module constructs a methane emission source diagnosis model according to the prior emission source, the methane assimilation source at the previous moment and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission source.

[0066] The methane ensemble Kalman smoothing module is used to realize the coordinated assimilation of methane concentration and flux, and realizes the assimilation inversion of methane emissions according to the ensemble forecast methane concentration field, the ensemble forecast methane emission source and the satellite methane column concentration to obtain the methane assimilation source at this moment.

[0067] The sub-sector accounting module of the assimilated source uses the bottom-up methane emission inventory as a priori estimate of sub-sector emissions, and on the basis of the methane assimilation source at this moment, establishes an emission model for each sector after assimilation according to the sector proportion in the prior emissions, and then obtains the assimilated emissions of each sector. At this point, the sub-sector emissions of methane constrained by top-down atmospheric observations are given.

[0068] The emission models of each sector after assimilation are as follows:

[0069] ;

[0070] in, Indicates time t The grid point position of the time mode ( i, j ) of the updated department ip Emissions estimates, ip represents the emission sector, Indicates time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip emissions; Indicates time t The grid point position of the time mode ( i, j ) of the total prior emissions, Indicates time t The grid point position of the time mode ( i, j ) of total assimilated emissions.

[0071] The output module is used to output the assimilated emissions of each sector. Example 3

[0072] This embodiment uses the method in Example 1 to perform experimental simulation, and preprocesses the 2020 GOSAT (Greenhouse gases observing satellite) satellite methane column concentration into the format required for assimilation. The regional meteorological model RAMS simulation area and initial boundary data settings are completed, and the 2020 RAMS simulation is carried out to provide a meteorological field that corresponds to the horizontal and vertical resolutions of the regional atmospheric transport model CMAQ.

[0073] The obtained RAMS meteorological field is used for CMAQ methane simulation. The emission, transport, diffusion, and chemical conversion process of methane in CMAQ are carefully designed. The chemical conversion includes not only NO XThe impact of changes in CO on it also increases the reaction of O3 related to methane. The emission sources and initial boundary conditions required for the methane simulation in my country in 2020 in the methane CMAQ model were set. The methane simulation time step output by the CMAQ model is 1 hour, and the horizontal resolution is set to 64km.

[0074] The obtained CMAQ simulated methane concentration and other information are used to construct a methane emission source diagnosis model. The input data of the diagnosis model include: prior emission sources, ensemble forecast methane concentration field, and ensemble average analysis field of methane emissions at the same time every day in the previous few days. The diagnosis model has hourly resolution.

[0075] The EnSRF with EnKS features was extended, and the assimilation inversion of methane emission sources was further carried out within the assimilation framework. The assimilation system designed the coordinated assimilation of methane concentration and flux to reduce the mutual influence of the uncertainties of the two. Its state variables are Therefore, under the framework of this assimilation algorithm, the satellite observations and ensemble background fields (including emission sources and concentration fields) are used for methane emission assimilation inversion. The assimilation results are shown in Table 1.

[0076] Table 1 Statistics of assimilated sources (in bold) and a priori sources in my country by sector in 2020 (in units) and )

[0077] coal oil natural gas Animal Husbandry Rice field emissions Waste resident transportation Total January 1.16 / 1.73 0.02 / 0.03 0.10 / 0.14 0.60 / 0.83 0.82 / 1.17 0.72 / 1.02 0.49 / 0.69 0.02 / 0.02 4.01 / 5.74 February 1.44 / 1.62 0.03 / 0.03 0.12 / 0.13 0.46 / 0.53 0.78 / 0.93 0.84 / 0.96 0.43 / 0.49 0.02 / 0.02 4.20 / 4.80 March 1.74 / 1.73 0.04 / 0.03 0.14 / 0.14 0.23 / 0.29 0.41 / 0.60 0.90 / 1.02 0.29 / 0.31 0.02 / 0.02 3.87 / 4.26 April 1.47 / 1.67 0.03 / 0.03 0.12 / 0.13 0.35 / 0.40 0.49 / 0.57 0.90 / 0.99 0.18 / 0.20 0.02 / 0.02 3.66 / 4.12 May 1.69 / 1.73 0.03 / 0.03 0.13 / 0.14 0.51 / 0.63 0.60 / 0.82 0.85 / 1.02 0.11 / 0.14 0.02 / 0.02 4.04 / 4.65 June 1.32 / 1.67 0.03 / 0.03 0.12 / 0.13 0.77 / 1.04 0.73 / 1.10 0.72 / 0.99 0.09 / 0.13 0.01 / 0.02 3.87 / 5.23 July 1.26 / 1.73 0.03 / 0.03 0.11 / 0.14 0.85 / 1.04 1.01 / 1.43 0.73 / 1.02 0.09 / 0.13 0.02 / 0.02 4.19 / 5.66 August 1.14 / 1.73 0.03 / 0.03 0.12 / 0.14 0.88 / 1.19 1.25 / 1.85 0.72 / 1.02 0.09 / 0.13 0.02 / 0.02 4.32 / 6.24 September 1.18 / 1.67 0.03 / 0.03 0.12 / 0.13 1.11 / 1.52 1.69 / 1.97 0.89 / 0.99 0.12 / 0.13 0.02 / 0.02 5.25 / 6.58 October 1.67 / 1.73 0.04 / 0.03 0.14 / 0.14 0.68 / 0.74 1.32 / 1.53 0.93 / 1.02 0.17 / 0.18 0.02 / 0.02 5.08 / 5.52 November 1.60 / 1.67 0.04 / 0.03 0.13 / 0.13 0.72 / 0.77 1.31 / 1.55 0.89 / 0.99 0.31 / 0.34 0.02 / 0.02 5.12 / 5.62 December 2.09 / 1.73 0.04 / 0.03 0.16 / 0.14 0.70 / 0.69 0.78 / 0.90 1.02 / 1.02 0.74 / 0.73 0.02 / 0.02 5.69 / 5.38 annual 17.76 / 20.42 0.40 / 0.41 1.51 / 1.61 7.86 / 9.69 11.20 / 14.41 10.10 / 12.07 3.12 / 3.60 0.22 / 0.25 53.30 / 63.80

[0078] The bottom-up methane emission inventory was used as a priori estimate, and based on the above methane assimilation sources, the assimilated emissions of each sector were inferred according to the sector proportion in the prior emissions.

[0079] After assimilation, the emission models of each department are as follows:

[0080] ;

[0081] in, Indicates time (month) t The grid point position of the time mode ( i, j ) of the updated department ip Emissions estimates, ip It indicates the emission sector, which includes coal methane emission sector, oil methane emission sector, natural gas methane emission sector, animal husbandry methane emission sector, rice field methane emission sector, waste methane emission sector, etc. Indicates time t The grid point position of the time mode ( i, j ) of the sectors in the prior emission sources ip emissions; Indicates time tThe grid point position of the time mode ( i, j ) of the total prior emissions, Indicates time t The grid point position of the time mode ( i, j ). In this way, combined with the bottom-up methane emission inventory and the top-down assimilation inversion, the methane emission accounting by sector is further given. The difference between the a priori and assimilated sub-sector emission sources can be attributed to the assimilation constraints of satellite observations. The results are shown in Table 1.

[0082] Assimilation source verification: Since there is no direct observation of emission sources, the CMAQ model is driven by assimilation sources and a priori sources respectively, and the assimilation effect is evaluated by the concentration simulation effect (see Table 2). Independent observations come from 8 station detection values. The stations include: AMY (Anmyeon-do, South Korea), DSI (Dongsha Island, China), JGS (JejuGosan, Jeju Island, South Korea), RYO (Ryori, Ayari, Japan), ULD (Ulleungdo, Ulleungdo, South Korea), UUM (UlaanUul, Ulan Ula, Mongolia), WLG (Mt. Waliguan, China), YON (Yonagunijima, Yonaguni Island, Japan). After verification, it is shown that the standard error RMSE and deviation of the assimilated concentration analysis field are better than the a priori source simulation value CTRL and the assimilated source simulation value FC, indicating that the performance of the assimilation analysis field is the best. Compared with other stations, the standard errors RMSE of the assimilated concentration analysis field AN, assimilated source simulation value FC and prior source simulation value CTRL at the Waliguan station are 68.11, 81.14 and 81.81 ppb, respectively, and the correlation coefficients are 0.74, 0.71 and 0.70.

[0083] Table 2 Assimilation effect evaluation (unit: ppb)

[0084] Latitude (°N) / Longitude (°E) Number of observations Observation frequency <![CDATA[Average( Obs. / AN / FC / CTRL)]]> RMSE(AN / / CTRL) Imp.RateAN / (%) Deviation (AN / / CTRL) Site Overview AMY 36.54 / 126.33 8330 Hourly <![CDATA[ 2033.12 / 2063.45 / 2098.56 / 2098.59]]> 112.01 / / 149.64 25.15 / 30.33 / / 65.47 Coastal (South Korea) DSI 20.70 / 116.73 52 Weekly <![CDATA[ 1935.80 / 2061.20 / 2084.56 / 2084.68]]> 174.40 / / 199.47 12.57 / 125.40 / / 148.88 Ocean (China) JGS 33.30 / 126.21 8349 Hourly <![CDATA[ 1987.72 / 2050.29 / 2076.35 / 2076.85]]> 129.40 / / 160.78 19.52 / 62.57 / / 89.13 Coastal (South Korea) RYO 39.03 / 141.82 8402 Hourly <![CDATA[ 1967.53 / 2027.40 / 2046.36 / 2046.37]]> 100.77 / / 131.34 23.27 / 59.87 / / 78.84 Takayama (Japan) ULD 37.48 / 130.90 8026 Hourly <![CDATA[ 1977.99 / 2042.29 / 2066.83 / 2066.89]]> 121.61 / / 156.32 22.20 / 64.30 / / 88.90 Ocean (Korea) UUM 44.45 / 111.08 37 Weekly <![CDATA[ 1967.09 / 1950.82 / 1959.86 / 1959.78]]> 54.27 / / 62.32 12.91 / −16.27 / / 7.31 Steppe (Mongolia) WLG 36.28 / 100.90 47 Weekly <![CDATA[ 1951.68 / 1965.97 / 1969.91 / 1970.42]]> 68.11 / / 81.81 16.74 / 14.29 / / 18.75 Qinghai-Tibet Plateau (China) YON 24.47 / 123.02 8502 Hourly <![CDATA[ 1936.90 / 2055.53 / 2080.44 / 2080.44]]> 164.03 / / 193.48 15.22 / 118.63 / / 143.54 Ocean (Japan)

[0085] Note: The table includes the a priori source simulation value (CTRL), assimilated source simulation value (FC, Italic ), assimilated concentration analysis field (AN, bold). "Obs." indicates the observed value, marked with a dotted line, and "Imp. Rate" indicates the RMSE improvement rate, where the RMSE improvement rate of the assimilated concentration analysis field standard error is ( ) / RMSE CTRL × 100%, the RMSE improvement rate of the standard error of the assimilated source simulation value is ( ) / RMSE CTRL × 100%. RMSE CTRL Represents the standard error of the prior source simulation value, RMSE ANRepresents the standard error of the assimilated concentration analysis field, RMSE FC represents the standard error of the simulated value of the assimilated source, The annual average is obtained from the hourly values.

[0086] In addition, we further analyze the role of satellite observations in the bottom-up calculation of methane emission inventories. Considering that the information of a priori sources is mainly reflected in the a priori source simulation value (CTRL), and the observation information is mainly reflected in the assimilated concentration analysis field (AN), a comparative evaluation of the standard error RMSE of the eight ground stations was conducted. The improvement rate of the standard error RMSE of the assimilated concentration analysis field is defined as ( ) / RMSE CTRL × 100%, where RMSE CTRL Represents the standard error of the prior source simulation value, RMSE AN Represents the standard error of the assimilated concentration analysis field. As can be seen from Table 2, the standard error RMSE between the two groups of experimental methane concentrations, the assimilated concentration analysis field AN and the priori source simulation value CTRL, has improved by 12.57% to 25.15%. This shows that the present invention has played a positive role in the top-down methane emission source accounting in the GOSAT satellite-based experiment, effectively improved the simulation effect of the ground concentration, and improved the inversion accuracy of the methane emission source.

[0087] The present invention first improves the methane forward simulation effect by improving the methane emission, transmission, and diffusion in the methane regional atmospheric chemical transport model CMAQ, and adding the chemical transformation process of methane to the traditional CB05 gas phase chemical mechanism. Secondly, under the framework of coordinated assimilation of concentration and flux, methane emission source assimilation based on carbon satellite monitoring is carried out to reduce the impact of initial field and emission uncertainty. In particular, high-resolution methane emission assimilation technology is developed based on high-resolution methane concentration forward simulation, which fills the gap in top-down methane inversion to a certain extent.

[0088] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A high-resolution assimilation inversion method for methane emission sources, characterized in that: The following steps are involved: Step 1, obtain the prior emission sources and satellite methane column concentrations; Step 2: pre-process the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation; and set the regional meteorological model to provide the meteorological field; Step 3, constructing the emission, transport, diffusion and chemical conversion processes of methane in the regional atmospheric transmission model to obtain a methane regional atmospheric transmission model, wherein the methane regional atmospheric transmission model obtains an ensemble forecast methane concentration field based on a priori emission sources, methane assimilation sources at the previous moment and meteorological fields; Step 4: construct a methane emission source diagnosis model based on the prior emission sources, the methane assimilation sources at the previous moment, and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission sources; The methane emission source diagnosis model is as follows: ， in, represents the ensemble forecast of methane emission sources, represents the covariance inflation factor, express The ensemble forecast of methane concentration field at any moment, represents the ensemble average analysis field of methane emissions at the same time of day in the previous several days, , , is the total number of sample sets, represents the a priori emission source, Indicates the previous The ensemble mean analysis field of CH4 flux per day, , represents the analysis field, It indicates the forecast field; Step 5: construct a methane ensemble Kalman smoothing module based on the ensemble forecast methane concentration field and the ensemble forecast methane emission source, take the satellite methane column concentration as observation, realize methane emission assimilation inversion, and obtain the methane assimilation source at this moment; Step 6: By constructing a sub-sector accounting module for fine-scale assimilation sources, using the bottom-up methane emission inventory as a priori estimate of sub-sector emissions, and on the basis of the methane assimilation sources at this moment, the assimilated emission model of each sector is established according to the sector proportion in the prior emissions, and then the assimilated emissions of each sector are obtained; The emission models of each sector after assimilation are as follows: ， in, represents the updated sector ip emission estimate for the model grid point location (i, j) at time t, where ip represents the emitting sector, represents the emissions of sector ip in the a priori emission source at the model grid point location (i, j) at time t; represents the total prior emission at the model grid point location (i, j) at time t, represents the total assimilated emissions at the model grid point location (i, j) at time t.

2. The high-resolution assimilation inversion method for methane emission sources according to claim 1, characterized in that: The methane simulation time step output by the methane regional atmospheric transport model is 1 hour, and the horizontal resolution ranges from 1 to 64 km.

3. The high-resolution assimilation inversion method for methane emission sources according to claim 2, characterized in that: The emission sectors include the coal methane emission sector, the oil methane emission sector, the natural gas methane emission sector, the animal husbandry methane emission sector, the rice field methane emission sector, and the waste methane emission sector.

4. A high-resolution assimilation inversion system for methane emission sources, characterized by: It includes input module, data preprocessing module, regional atmospheric transmission model module, methane emission source diagnosis module, methane ensemble Kalman smoothing module, assimilation source sub-sector accounting module and output module, among which: The input module is used to input a priori emission sources and satellite methane column concentrations; The data preprocessing module is used to preprocess the satellite methane column concentration to obtain the methane column concentration in the format required for assimilation; and set the regional meteorological model to provide the meteorological field; The regional atmospheric transmission model module is used to construct the emission, transportation, diffusion and chemical conversion process of methane in the regional atmospheric transmission model to obtain a methane regional atmospheric transmission model, and the methane regional atmospheric transmission model obtains an ensemble forecast methane concentration field based on a priori emission sources, methane assimilation sources at the previous moment and meteorological fields; The methane emission source diagnosis module is used to construct a methane emission source diagnosis model according to the prior emission source, the methane assimilation source at the previous moment and the ensemble forecast methane concentration field to obtain the ensemble forecast methane emission source; The methane emission source diagnosis model is as follows: ， in, represents the ensemble forecast of methane emission sources, represents the covariance inflation factor, express The ensemble forecast of methane concentration field at any moment, represents the ensemble average analysis field of methane emissions at the same time of day in the previous several days, , , is the total number of sample sets, express The ensemble average of the CH4 concentration scale parameters at the time, represents the a priori emission source, Indicates the previous The ensemble mean analysis field of CH4 flux per day, , represents the analysis field, It indicates the forecast field; The methane ensemble Kalman smoothing module is used to realize methane emission assimilation inversion according to the ensemble forecast methane concentration field, the ensemble forecast methane emission source and the satellite methane column concentration, and obtain the methane assimilation source at this moment; The sub-sector accounting module of the assimilated source uses the bottom-up methane emission inventory as a priori estimate of sub-sector emissions, and on the basis of the methane assimilated source at this moment, establishes an emission model for each sector after assimilation according to the sector proportion in the priori emissions, and then obtains the emissions of each sector after assimilation; After assimilation, the emission models of each department are as follows: ; in, represents the updated sector ip emission estimate for the model grid point location (i, j) at time t, where ip represents the emitting sector, represents the emissions of sector ip in the a priori emission source at the model grid point location (i, j) at time t; represents the total prior emission at the model grid point location (i, j) at time t, represents the total assimilated emissions at the model grid point location (i, j) at time t; The output module is used to output the assimilated emissions of each sector.

5. The high-resolution assimilation inversion system for methane emission sources according to claim 4, characterized in that: The methane simulation time step output by the methane regional atmospheric transport model module is 1 hour, and the horizontal resolution range is 1-64km.

Images