Multi-source open data-based urban carbon emission hectometer spatial distribution gridding method

Abstract

The invention provides a multi-source open data-based urban carbon emission hectometer spatial distribution gridding method. The method comprises the steps of obtaining multi-source open big data of a target city, constructing a hectometer grid multi-dimensional information index and calculating the total amount of urban carbon emission; initial hectometer grid carbon emission is obtained according to the hectometer grid multi-dimensional information index and the total urban carbon emission, and initial district-level total carbon emission and initial town scale total carbon emission are obtained according to the initial hectometer grid carbon emission; obtaining a district-level reference total carbon emission amount and a town scale reference total carbon emission amount, constructing a fitness function according to the district-level reference total carbon emission amount, the town scale reference total carbon emission amount, the initial district-level total carbon emission amount and the initial town scale total carbon emission amount, and carrying out iterative solution on the fitness function by using a multi-agent genetic algorithm, and obtaining the final urban hectometer grid carbon emission corresponding to the minimum fitness function. According to the method, urban information can be comprehensively integrated to calculate the carbon emission spatial distribution pattern, and the grid precision of the carbon emission spatial distribution pattern is effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of urban planning and carbon emission technology, and more specifically, to a method for gridding the hundred-meter-level spatial distribution of urban carbon emissions based on multi-source open data. Background Art

[0002] The development of urbanization is accompanied by a large amount of greenhouse gas emissions. Ecological problems such as climate warming and extreme weather have become major obstacles to sustainable development. Spatialization of carbon emissions is conducive to quickly understanding the carbon emission pattern of urban space, carrying out carbon reduction-oriented urban space optimization layout work, building high-quality urban development, utilization and protection paths, and promoting high-quality urbanization and the realization of "dual carbon" goals.

[0003] At present, the research on refined spatial simulation of carbon emissions mainly uses data such as electricity, fossil fuels, and night lights to construct a spatial distribution model of carbon emissions from top to bottom. For example, some studies identify urban land use functional zoning based on the density of various POI points, and spatialize urban carbon emissions based on the land use types of different zonings combined with the total urban carbon emissions. There are also studies that combine night light data and municipal electricity consumption data to obtain the carbon emission ratio of urban zoning for carbon emission spatialization. Different research teams have obtained carbon emission spatial grid data based on the top-down spatial inversion path, but the existing carbon emission grid data is not accurate enough to reflect the carbon emission pattern within the city. In addition, constructing a carbon emission allocation model in proportion based on a single data source has limitations in reflecting the spatial pattern of carbon emissions. Summary of the invention

[0004] In order to overcome the defect of low accuracy of carbon emission grid data in the above-mentioned prior art, the present invention provides a method for gridding the hundred-meter-level spatial distribution of urban carbon emissions based on multi-source open data.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] The present invention proposes a method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data, comprising the following steps:

[0007] Obtain multi-source open big data of the target city, and construct a 100-meter grid multi-dimensional information index based on the multi-source open big data of the target city;

[0008] Calculate the total carbon emissions of the city based on multi-source open big data of the target city;

[0009] The initial 100-meter grid carbon emissions were obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions were obtained based on the initial 100-meter grid carbon emissions.

[0010] Obtain the district-level reference total carbon emissions and the town-scale reference total carbon emissions, construct a fitness function based on the district-level reference total carbon emissions, the town-scale reference total carbon emissions, the initial district-level total carbon emissions and the initial town-scale total carbon emissions, and use a multi-agent genetic algorithm to iteratively solve the fitness function to obtain the final urban 100-meter grid carbon emissions corresponding to the minimum fitness function.

[0011] Furthermore, the target city multi-source open big data includes original data of population distribution layer, original data of natural environment layer, original data of architectural environment layer, original data of economic business layer and historical yearbook data;

[0012] The original data of the population distribution layer include WorldPop population grid data and population census data;

[0013] The raw data of the natural environment layer include DEM data and Sentinel-2 image data;

[0014] The raw data of the building environment layer include: building footprint data, Sentinel-2 four-season image data;

[0015] The original data of the economic business layer includes night light data, land parcel vector boundary data, energy consumption data, and POI data.

[0016] Furthermore, the target city multi-dimensional information index includes a 100-meter grid population distribution index, and the method for extracting the 100-meter grid population distribution index includes:

[0017] The WorldPop population grid data was corrected according to the census data to obtain the corrected WorldPop population grid data, and ArcGIS was used to map the corrected WorldPop population grid data to the grid data to obtain the 100-meter grid population distribution index.

[0018] Furthermore, the target city multi-dimensional information index also includes a 100-meter grid natural environment index. The method for extracting the 100-meter grid natural environment index includes:

[0019] According to the DEM data, the average elevation of the city in each grid data is counted and connected to the grid to obtain the 100-meter grid elevation data;

[0020] Use ArcGIS, a spatial analysis software, to calculate DEM data to obtain urban slope data, and use spatial analysis methods to connect grid data and urban slope data to obtain 100-meter grid slope data.

[0021] Use the ENVI tool to calculate the Sentinel-2 image data to obtain the urban NDVI vegetation index and NDWI vegetation index, and establish connections between the grid data and the urban NDVI vegetation index and NDWI vegetation index to obtain the 100-meter grid NDVI vegetation index and NDWI vegetation index;

[0022] The Shannon entropy weight method was used to construct the 100-meter grid natural environment index based on the 100-meter grid elevation data, 100-meter grid slope data, 100-meter grid NDVI vegetation index and NDWI vegetation index.

[0023] Furthermore, the target city multi-dimensional information also includes a 100-meter grid building environment index. The method for extracting the 100-meter grid building environment index includes:

[0024] Constructing a deep learning-based target detection neural network SEASONet, inputting the Sentinel-2 imagery of four seasons composite images and target city building footprint data into the deep learning-based target detection neural network SEASONet to extract the building shadow index;

[0025] Performing spatial analysis on building outline data with high-rise buildings in the building footprint data to obtain building data, wherein the building data includes building area percentage, building average height and building volume ratio;

[0026] A 100-meter grid building environment index is constructed based on the building shadow index, building data and grid data.

[0027] Furthermore, the target city multi-dimensional information also includes a 100-meter grid economic business index, and the method for extracting the 100-meter grid economic business index includes:

[0028] Preprocess the night light data to obtain preprocessed night light data, calculate the average brightness value of the night light data based on the preprocessed night light data, and then connect the grid data to obtain the 100-meter grid light data;

[0029] Referring to the universality and consistency of urban land classification standards and POI classification, the vector boundary data of plots are divided into multiple types of plots, including public service plots, management plots, transportation, residential plots, commercial function plots, recreational plots and industrial plots. Kernel density cluster analysis is performed on each type of plot, and the proportion of each type of POI kernel density in each plot is counted to obtain functional plot data. The functional plot data and grid data are connected to obtain 100-meter grid functional plot data.

[0030] Based on the 100-meter grid lighting data and the 100-meter grid functional plot data, the 100-meter grid economic business index is obtained.

[0031] Furthermore, the total amount of urban carbon emissions is calculated based on the multi-source open big data of the target city, including:

[0032] Based on historical yearbook data, we can obtain data on actual fossil energy consumption of different industries and types in the target city, urban electricity consumption, and total urban solid waste;

[0033] The carbon emissions from fossil energy consumption are calculated based on the actual fossil energy consumption of different industries and types in the target cities, specifically:

[0034]

[0035] Among them, FC i represents the consumption of the i-th type of fossil energy, CEF i represents the carbon emission factor of the i-th fossil energy;

[0036] The carbon emissions from electricity consumption are calculated based on the city's electricity consumption, specifically:

[0037]

[0038] Among them, EC j represents the electricity consumption of the jth category; EEF represents the carbon emission factor of electricity, and m represents different categories;

[0039]

[0040] Among them, SW k Expressed as the total amount of solid waste of type k, SEF k represents the carbon emission factor of the kth type of solid waste, and p represents the total number of solid waste categories;

[0041] The total carbon emissions of the city are:

[0042] C=C1+C2+C3

[0043] Among them, C represents the total carbon emissions of the city.

[0044] Furthermore, the initial 100-meter grid carbon emissions are obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions are obtained based on the initial 100-meter grid carbon emissions, including:

[0045] The multi-dimensional information index of the 100-meter grid is normalized to obtain the normalized multi-dimensional information index of the 100-meter grid. According to the normalized multi-dimensional information index of the 100-meter grid, the entropy weight method is used to determine the initial weight of the multi-dimensional information index of the 100-meter grid. The normalized multi-dimensional information index of the 100-meter grid is weighted and summed to obtain the initial carbon emissions of the 100-meter grid. The calculation formula is as follows:

[0046]

[0047] Among them, s j represents the carbon emission of the jth initial 100-meter grid, D i,norm(j) represents the i-th index in the normalized multi-dimensional information index of the 100-meter grid corresponding to the j-th initial 100-meter grid, D 1,norm(j) , D 2,norm(j) , D 3,norm(j) and D 4,norm(j) They represent the normalized population distribution index, natural environment index, building environment index, and economic business index of the jth initial 100-meter grid, respectively. i represents the initial weight of the i-th index in the 100-meter grid multidimensional information index, that is, α1, α2, α3 and α4 represent the initial weights of the population distribution index, natural environment index, building environment index and economic business index in the 100-meter grid multidimensional information index, respectively. total represents the total carbon emissions of the city;

[0048] The initial 100-meter grid carbon emissions within the town and street area are aggregated and calculated to obtain the total carbon emissions at the initial town and street scale;

[0049] The initial 100-meter grid carbon emissions within the district level were aggregated and calculated to obtain the initial district-level total carbon emissions.

[0050] Furthermore, obtaining the district-level reference total carbon emissions and the town-level reference total carbon emissions includes:

[0051] Obtaining district-level reference carbon emissions includes:

[0052] According to the historical yearbook data, the socio-economic structure data and the total carbon emissions at the district level in recent years were obtained. The socio-economic structure data included the historical population, total GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district. A multivariate linear regression model was established based on the socio-economic structure data and the district-level carbon emissions data, with the socio-economic structure data as the independent variable and the district-level carbon emissions as the dependent variable, specifically:

[0053] Y=β0+β1X1+β2X2+…+β n X n +ε

[0054] Among them, Y is the total carbon emissions at the district level; X1, X2, …, X n They represent the historical population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district; β0, β1, …, β n is the parameter to be estimated, ε is the error term;

[0055] Substitute the socio-economic structure data and district-level carbon emissions data of recent years into the multivariate linear regression model to calculate the parameters to be estimated;

[0056] The population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district in the reference year were obtained and input into the multivariate linear regression model to obtain the district-level reference carbon emissions;

[0057] Obtaining the total reference carbon emissions at the town and street level includes:

[0058] The EDGAR carbon emission grid dataset of the reference year was obtained and calibrated for the total urban carbon emissions to obtain the corrected EDGAR carbon emission dataset. Based on the corrected EDGAR carbon emission dataset, the total carbon emission data at the town and street scale were statistically analyzed to obtain the reference carbon emissions at the town and street scale.

[0059] Furthermore, the method for determining the fitness function includes:

[0060]

[0061] Among them, Y0 and Y logical They represent the initial district-level carbon emissions and district-level reference carbon emissions respectively; Y1 and Y EDGAR They represent the initial total carbon emissions at the town and street scale and the reference total carbon emissions at the town and street scale respectively. D represents the administrative district set, and E represents the town and street set.

[0062] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0063] The present invention proposes a method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data, including obtaining multi-source open big data of a target city, constructing a hundred-meter grid multi-dimensional information index according to the target city multi-source open big data, and calculating the total urban carbon emissions; obtaining an initial hundred-meter grid carbon emissions according to the hundred-meter grid multi-dimensional information index and the total urban carbon emissions, respectively obtaining an initial district-level carbon emissions and an initial town-scale carbon emissions according to the initial hundred-meter grid carbon emissions; obtaining a district-level reference carbon emissions and a town-scale reference carbon emissions, constructing a fitness function according to the district-level reference carbon emissions, the town-scale reference carbon emissions, the initial district-level carbon emissions, and the initial town-scale carbon emissions, and using a multi-agent genetic algorithm to iteratively solve the fitness function to obtain the final urban hundred-meter grid carbon emissions corresponding to the minimum fitness function. The present invention can more comprehensively integrate urban information to infer the spatial distribution pattern of carbon emissions, and effectively improve the grid accuracy of the spatial distribution pattern of carbon emissions. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 This is a flow chart of the method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data described in Example 1;

[0065] Figure 2 This is the flow chart of the multi-agent genetic algorithm described in Example 3. DETAILED DESCRIPTION

[0066] The drawings are for illustrative purposes only and are not to be construed as limiting the present patent;

[0067] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0068] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0069] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0070] Example 1

[0071] This embodiment provides a method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data, such as Figure 1 As shown, the following steps are included:

[0072] Obtain multi-source open big data of the target city, and construct a 100-meter grid multi-dimensional information index based on the multi-source open big data of the target city;

[0073] Calculate the total carbon emissions of the city based on multi-source open big data of the target city;

[0074] The initial 100-meter grid carbon emissions were obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions were obtained based on the initial 100-meter grid carbon emissions.

[0075] Obtain the district-level reference total carbon emissions and the town-scale reference total carbon emissions, construct a fitness function based on the district-level reference total carbon emissions, the town-scale reference total carbon emissions, the initial district-level total carbon emissions and the initial town-scale total carbon emissions, and use a multi-agent genetic algorithm to iteratively solve the fitness function to obtain the final urban 100-meter grid carbon emissions corresponding to the minimum fitness function.

[0076] In the specific implementation process, we first construct a multi-dimensional information index of a 100-meter grid and calculate the total urban carbon emissions based on the multi-source open big data of the target city. Then, we obtain the initial district-level carbon emissions and the initial town-scale carbon emissions. The district-level reference carbon emissions and the town-scale reference carbon emissions are iteratively solved by a multi-agent genetic algorithm to obtain the final urban 100-meter grid carbon emissions corresponding to the minimum fitness function.

[0077] Example 2

[0078] Based on Example 1, this example further limits the acquisition of district-level reference total carbon emissions and town-level reference total carbon emissions.

[0079] The acquisition of district-level reference carbon emissions and town-level reference carbon emissions includes:

[0080] Obtaining district-level reference carbon emissions includes:

[0081] According to the historical yearbook data, the socio-economic structure data and the total carbon emissions at the district level in recent years were obtained. The socio-economic structure data included the historical population, total GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district. A multivariate linear regression model was established based on the socio-economic structure data and the district-level carbon emissions data, with the socio-economic structure data as the independent variable and the district-level carbon emissions as the dependent variable, specifically:

[0082] Y=β0+β1X1+β2X2+…+β n X n +ε

[0083] Among them, Y is the total carbon emissions at the district level; X1, X2, …, X n They represent the historical population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district; β0, β1, …, β n is the parameter to be estimated, ε is the error term;

[0084] Substitute the socio-economic structure data and district-level carbon emissions data of recent years into the multivariate linear regression model to calculate the parameters to be estimated;

[0085] The population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district in the reference year were obtained and input into the multivariate linear regression model to obtain the district-level reference carbon emissions;

[0086] Obtaining the total reference carbon emissions at the town and street level includes:

[0087] The EDGAR carbon emission grid dataset of the reference year was obtained and calibrated for the total urban carbon emissions to obtain the corrected EDGAR carbon emission dataset. Based on the corrected EDGAR carbon emission dataset, the total carbon emission data at the town and street scale were statistically analyzed to obtain the reference carbon emissions at the town and street scale.

[0088] Example 3

[0089] This embodiment provides a method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data, including the following steps:

[0090] Obtain multi-source open big data of the target city, and construct a 100-meter grid multi-dimensional information index based on the multi-source open big data of the target city;

[0091] The target city multi-source open big data includes original data of population distribution layer, original data of natural environment layer, original data of architectural environment layer, original data of economic business layer and historical yearbook data;

[0092] The original data of the population distribution layer include WorldPop population grid data and population census data;

[0093] The raw data of the natural environment layer include DEM data and Sentinel-2 image data;

[0094] The raw data of the building environment layer include: building footprint data, Sentinel-2 four-season image data;

[0095] The original data of the economic business layer includes night light data, land parcel vector boundary data, energy consumption data, and POI data.

[0096] The target city multi-dimensional information index includes a 100-meter grid population distribution index. The method for extracting the 100-meter grid population distribution index is as follows:

[0097] The WorldPop population grid data is used as the population distribution layer for the study after correction with the census data, and the open source population grid data is proportionally corrected according to the official census data. Use ArcGIS and other spatial analysis software to spatially connect the population of the study area and map it to the grid to complete the mapping of population data and grid. Obtain the census data in the study area in the study year, count the total number of WorldPop population grids in different districts and counties, scale the population data proportionally, and obtain the 100-meter grid population distribution index. The formula is as follows:

[0098]

[0099] Among them, m i Represents the adjusted grid population, n i represents the population of the ith grid, M represents the total population of different districts and counties, and p represents the number of grids in a specified district or county.

[0100] The target city multi-dimensional information index also includes a 100-meter grid natural environment index. The method for extracting the 100-meter grid natural environment index is specifically as follows:

[0101] According to the DEM data, the grid, elevation image and slope raster data are superimposed respectively, and the average elevation of the city in each grid data is counted and connected to the grid to obtain the 100-meter grid elevation data;

[0102] Use ArcGIS, a spatial analysis software, to calculate DEM data to obtain urban slope data, and use spatial analysis methods to connect grid data and urban slope data to obtain 100-meter grid slope data.

[0103] The ENVI tool was used to calculate the Sentinel-2 image data to obtain the urban NDVI vegetation index and NDWI vegetation index. The grid data and the urban NDVI vegetation index and NDWI vegetation index were connected to obtain the 100-meter grid NDVI vegetation index and NDWI vegetation index. The calculation formulas of the NDVI vegetation index and NDWI vegetation index are:

[0104]

[0105] Among them, B3, B4, and B5 represent the remote sensing data of the 3rd, 4th, and 5th bands respectively.

[0106] According to the 100-meter grid elevation data, 100-meter grid slope data, 100-meter grid NDVI vegetation index and NDWI vegetation index, the Shannon entropy weight method is used to construct the 100-meter grid natural environment index. The formula is as follows:

[0107] Data Standardization:

[0108]

[0109] Among them, x′ ij is the original value of the i-th sample in the j-th category indicator.

[0110] Calculate the ratio value:

[0111]

[0112] Here, m is the total number of samples.

[0113] The entropy value e of the j-th indicator j for:

[0114]

[0115] in, Used to ensure that the entropy value ranges between 0 and 1. If p ij =0, then define p ij ln(p ij )=0.

[0116] The entropy value d of the j-th type difference coefficient j for:

[0117] d j =1-e j

[0118] Among them, the coefficient of difference reflects the size of the data. The larger the difference, the greater the entropy and the greater the amount of information.

[0119] Calculate the weight w of the j-th indicator according to the difference coefficient j for:

[0120]

[0121] Where n is the total number of indicators.

[0122] Urban natural environment index layer:

[0123]

[0124] Among them, M i It is the grid value of the natural environment index layer of the kth grid.

[0125] The target city multi-dimensional information also includes a 100-meter grid building environment index. The method for extracting the 100-meter grid building environment index is specifically as follows:

[0126] Constructing a deep learning-based target detection neural network SEASONet, inputting the Sentinel-2 imagery of four seasons composite images and target city building footprint data into the deep learning-based target detection neural network SEASONet to extract the building shadow index;

[0127] Perform spatial analysis on the building outline data with high-rise buildings in the building footprint data to obtain building data, wherein the building data includes indicators such as building area percentage, average building height and building volume ratio to comprehensively describe the urban building environment;

[0128] A 100-meter grid building environment index is constructed based on the building shadow index, building data and grid data.

[0129] The target city multi-dimensional information also includes a 100-meter grid economic business index. The method for extracting the 100-meter grid economic business index is as follows:

[0130] Use the Raster Calculator tool in ArcGIS to filter out abnormal light spots in the night light data study area, and perform radiation calibration and adjust pixel values ​​as needed. Use the Zonal Statistics as Table tool to obtain the average brightness value of the night light data under the grid, and then connect the grid data to obtain the 100-meter grid light data. The 100-meter grid light data represents the economic vitality of the city;

[0131] Referring to the universality and consistency of urban land classification standards and POI classification, the vector boundary data of plots are divided into multiple types of plots, including public service plots, management plots, transportation, residential plots, commercial function plots, recreational plots and industrial plots. Kernel density cluster analysis is performed on each type of plot, and the proportion of each type of POI kernel density in each plot is counted to obtain functional plot data. The functional plot data and grid data are connected to obtain 100-meter grid functional plot data. The 100-meter grid functional plot data represents economic intensity.

[0132] Based on the 100-meter grid lighting data and the 100-meter grid functional plot data, the 100-meter grid economic business index is obtained.

[0133] From the historical yearbook data, we can mine the data on coal, fuel oil, gasoline, diesel, heat and electricity consumption in various industrial sectors, agricultural fertilizer use, pesticide use, diesel use, total rural electricity consumption, agricultural film use, and the number of poultry, cattle, sheep and pigs on hand and slaughtered, and total urban solid waste. Based on the carbon accounting formula provided by the IPCC, we calculate the total urban carbon emissions based on the multi-source open big data of the target city, multiply the energy consumption of different types by the corresponding carbon emission factors, and multiply the use of different types of factors in agriculture by the corresponding carbon emission factors to obtain the carbon dioxide emissions generated in agricultural production.

[0134] Based on historical yearbook data, we can obtain data on actual fossil energy consumption of different industries and types in the target city, urban electricity consumption, and total urban solid waste;

[0135] The carbon emissions from fossil energy consumption are calculated based on the actual fossil energy consumption of different industries and types in the target cities, specifically:

[0136]

[0137] Among them, FC i represents the consumption of the i-th type of fossil energy, CEF i represents the carbon emission factor of the i-th fossil energy;

[0138] The carbon emissions from electricity consumption are calculated based on the city's electricity consumption, specifically:

[0139]

[0140] Among them, EC j represents the electricity consumption of the jth category; EEF represents the carbon emission factor of electricity, and m represents different categories;

[0141]

[0142] Among them, SW kExpressed as the total amount of solid waste of type k, SEF k represents the carbon emission factor of the kth type of solid waste, and p represents the total number of solid waste categories;

[0143] The total carbon emissions of the city are:

[0144] C=C1+C2+C3

[0145] Among them, C represents the total carbon emissions of the city.

[0146] The initial 100-meter grid carbon emissions were obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions were obtained based on the initial 100-meter grid carbon emissions:

[0147] The multi-dimensional information index of the 100-meter grid is normalized to obtain the normalized multi-dimensional information index of the 100-meter grid. According to the normalized multi-dimensional information index of the 100-meter grid, the entropy weight method is used to determine the initial weight of the multi-dimensional information index of the 100-meter grid. The normalized multi-dimensional information index of the 100-meter grid is weighted and summed to obtain the initial carbon emissions of the 100-meter grid. The calculation formula is as follows:

[0148]

[0149] Among them, s j represents the carbon emission of the jth initial 100-meter grid, D i,norm(j) represents the i-th index in the normalized multi-dimensional information index of the 100-meter grid corresponding to the j-th initial 100-meter grid, D 1,norm(j) , D 2,norm(j) , D 3,norm(j) and D 4,norm(j) They represent the normalized population distribution index, natural environment index, building environment index, and economic business index of the jth initial 100-meter grid, respectively. i represents the initial weight of the i-th index in the 100-meter grid multidimensional information index, that is, α1, α2, α3 and α4 represent the initial weights of the population distribution index, natural environment index, building environment index and economic business index in the 100-meter grid multidimensional information index, respectively. total represents the total carbon emissions of the city;

[0150] The initial 100-meter grid carbon emissions within the town and street area are aggregated and calculated to obtain the total carbon emissions at the initial town and street scale;

[0151] The initial 100-meter grid carbon emissions within the district level were aggregated and calculated to obtain the initial district-level total carbon emissions.

[0152] Obtain the district-level reference total carbon emissions and the town-scale reference total carbon emissions, construct a fitness function based on the district-level reference total carbon emissions, the town-scale reference total carbon emissions, the initial district-level total carbon emissions and the initial town-scale total carbon emissions, and use a multi-agent genetic algorithm to iteratively solve the fitness function to obtain the final urban 100-meter grid carbon emissions corresponding to the minimum fitness function.

[0153] Get the district-level reference carbon emissions:

[0154] According to the historical yearbook data, the socio-economic structure data and the total carbon emissions at the district level in the past five years were obtained. The socio-economic structure data included the historical population, total GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district. A multivariate linear regression model was established based on the socio-economic structure data and the district-level carbon emissions data, with the socio-economic structure data as the independent variable and the district-level carbon emissions as the dependent variable, specifically:

[0155] Y=β0+β1X1+β2X2+…+β n X n +ε

[0156] Among them, Y is the total carbon emissions at the district level; X1, X2, …, X n They represent the historical population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district; β0, β1, …, β n is the parameter to be estimated, ε is the error term;

[0157] Substitute the socio-economic structure data and district-level carbon emissions data of recent years into the multivariate linear regression model to calculate the parameters to be estimated;

[0158] The population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district in the reference year were obtained and input into the multivariate linear regression model to obtain the district-level reference carbon emissions;

[0159] Get the total reference carbon emissions at the town and street scale:

[0160] The EDGAR carbon emission grid dataset of the reference year is obtained and the total urban carbon emissions are calibrated to obtain the corrected EDGAR carbon emission dataset. Based on the corrected EDGAR carbon emission dataset, the total carbon emission data at the town and street scale are statistically analyzed to obtain the reference carbon emissions at the town and street scale. The calculation formula is as follows:

[0161]

[0162] Among them, y i represents the adjusted carbon emissions per kilometer grid, o irepresents the carbon emissions of the current i-th grid, Y′ represents the total carbon emissions of the city, and M1 represents the total carbon emissions of the kilometer grid.

[0163] The method for determining the fitness function is:

[0164]

[0165] Among them, Y0 and Y logical They represent the initial district-level carbon emissions and district-level reference carbon emissions respectively; Y1 and Y EDGAR They represent the initial total carbon emissions at the town and street scale and the reference total carbon emissions at the town and street scale respectively. D represents the administrative district set, and E represents the town and street set.

[0166] like Figure 2 As shown in the figure, it is a flow chart of a multi-agent genetic algorithm. In this embodiment, the multi-agent genetic algorithm abstracts the spatial distribution problem of carbon emissions into individual solutions of different spatial scales. Each agent works independently and optimizes the local carbon emission distribution through evolutionary strategies. At the same time, through the crossover and mutation operations of the genetic algorithm, the collaboration between different individuals is realized, and finally the optimization of the entire spatial distribution is realized. The termination condition of the multi-agent genetic algorithm iteration is that the set number of iterations is reached or the error between the total urban carbon emissions at the town and street scale obtained by summarizing the carbon emissions of the final 100-meter grid output by the algorithm and the initial town and street scale carbon emissions is less than 5%.

[0167] The same or similar reference numerals correspond to the same or similar components;

[0168] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0169] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data, characterized in that: The following steps are involved: Obtain multi-source open big data of the target city, and construct a 100-meter grid multi-dimensional information index based on the multi-source open big data of the target city; Calculate the total carbon emissions of the city based on multi-source open big data of the target city; The initial 100-meter grid carbon emissions were obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions were obtained based on the initial 100-meter grid carbon emissions. Obtain the district-level reference total carbon emissions and the town-scale reference total carbon emissions, construct a fitness function based on the district-level reference total carbon emissions, the town-scale reference total carbon emissions, the initial district-level total carbon emissions and the initial town-scale total carbon emissions, and use a multi-agent genetic algorithm to iteratively solve the fitness function to obtain the final urban 100-meter grid carbon emissions corresponding to the minimum fitness function.

2. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data according to claim 1 is characterized in that: The target city multi-source open big data includes original data of population distribution layer, original data of natural environment layer, original data of architectural environment layer, original data of economic business layer and historical yearbook data; The original data of the population distribution layer include WorldPop population grid data and population census data; The raw data of the natural environment layer include DEM data and Sentinel-2 image data; The raw data of the building environment layer include: building footprint data, Sentinel-2 four-season image data; The original data of the economic business layer includes night light data, land parcel vector boundary data, energy consumption data, and POI data.

3. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data according to claim 2 is characterized in that: The target city multi-dimensional information index includes a 100-meter grid population distribution index, and a method for extracting the 100-meter grid population distribution index includes: The WorldPop population grid data was corrected according to the census data to obtain the corrected WorldPop population grid data, and ArcGIS was used to map the corrected WorldPop population grid data to the grid data to obtain the 100-meter grid population distribution index.

4. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data according to claim 2 is characterized in that: The target city multi-dimensional information index also includes a 100-meter grid natural environment index. The method for extracting the 100-meter grid natural environment index includes: According to the DEM data, the average elevation of the city in each grid data is counted and connected to the grid to obtain the 100-meter grid elevation data; Use ArcGIS, a spatial analysis software, to calculate DEM data to obtain urban slope data, and use spatial analysis methods to connect grid data and urban slope data to obtain 100-meter grid slope data. Use the ENVI tool to calculate the Sentinel-2 image data to obtain the urban NDVI vegetation index and NDWI vegetation index, and establish connections between the grid data and the urban NDVI vegetation index and NDWI vegetation index to obtain the 100-meter grid NDVI vegetation index and NDWI vegetation index; The Shannon entropy weight method was used to construct the 100-meter grid natural environment index based on the 100-meter grid elevation data, 100-meter grid slope data, 100-meter grid NDVI vegetation index and NDWI vegetation index.

5. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data according to claim 2 is characterized in that: The target city multi-dimensional information also includes a 100-meter grid building environment index. The method for extracting the 100-meter grid building environment index includes: Constructing a deep learning-based target detection neural network SEASONet, inputting the Sentinel-2 imagery of four seasons composite images and target city building footprint data into the deep learning-based target detection neural network SEASONet to extract the building shadow index; Performing spatial analysis on building outline data with high-rise buildings in the building footprint data to obtain building data, wherein the building data includes building area percentage, building average height and building volume ratio; A 100-meter grid building environment index is constructed based on the building shadow index, building data and grid data.

6. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data according to claim 2 is characterized in that: The target city multi-dimensional information also includes a 100-meter grid economic business index. The method for extracting the 100-meter grid economic business index includes: Preprocess the night light data to obtain preprocessed night light data, calculate the average brightness value of the night light data based on the preprocessed night light data, and then connect the grid data to obtain the 100-meter grid light data; Referring to the universality and consistency of urban land classification standards and POI classification, the vector boundary data of plots are divided into multiple types of plots, including public service plots, management plots, transportation, residential plots, commercial function plots, recreational plots and industrial plots. Kernel density cluster analysis is performed on each type of plot, and the proportion of each type of POI kernel density in each plot is counted to obtain functional plot data. The functional plot data and grid data are connected to obtain 100-meter grid functional plot data. Based on the 100-meter grid lighting data and the 100-meter grid functional plot data, the 100-meter grid economic business index is obtained.

7. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data according to claim 2 is characterized in that: The total amount of urban carbon emissions calculated based on multi-source open big data of the target city includes: Based on historical yearbook data, we can obtain data on actual fossil energy consumption of different industries and types in the target city, urban electricity consumption, and total urban solid waste; The carbon emissions from fossil energy consumption are calculated based on the actual fossil energy consumption of different industries and types in the target cities, specifically: Among them, FC i represents the consumption of the i-th type of fossil energy, CEF i represents the carbon emission factor of the i-th fossil energy; The carbon emissions from electricity consumption are calculated based on the city's electricity consumption, specifically: Among them, EC j represents the electricity consumption of the jth category; EEF represents the carbon emission factor of electricity, and m represents different categories; Among them, SW k Expressed as the total amount of solid waste of type k, SEF k represents the carbon emission factor of the kth type of solid waste, and p represents the total number of solid waste categories; The total carbon emissions of the city are: C=C1+C2+C3 Among them, C represents the total carbon emissions of the city.

8. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data according to any one of claims 2 to 7, characterized in that: The initial 100-meter grid carbon emissions are obtained based on the 100-meter grid multi-dimensional information index and the total urban carbon emissions. The initial district-level carbon emissions and the initial town-level carbon emissions are obtained based on the initial 100-meter grid carbon emissions, including: The multi-dimensional information index of the 100-meter grid is normalized to obtain the normalized multi-dimensional information index of the 100-meter grid. According to the normalized multi-dimensional information index of the 100-meter grid, the entropy weight method is used to determine the initial weight of the multi-dimensional information index of the 100-meter grid. The normalized multi-dimensional information index of the 100-meter grid is weighted and summed to obtain the initial carbon emissions of the 100-meter grid. The calculation formula is as follows: Among them, s j represents the carbon emission of the jth initial 100-meter grid, D i,norm(j) represents the i-th index in the normalized multi-dimensional information index of the 100-meter grid corresponding to the j-th initial 100-meter grid, D 1,norm(j) , D 2,norm(j) , D 3,norm(j) and D 4,norm(j) They represent the normalized population distribution index, natural environment index, building environment index, and economic business index of the jth initial 100-meter grid, respectively. i represents the initial weight of the i-th index in the 100-meter grid multidimensional information index, that is, α1, α2, α3 and α4 represent the initial weights of the population distribution index, natural environment index, building environment index and economic business index in the 100-meter grid multidimensional information index, respectively. total represents the total carbon emissions of the city; The initial 100-meter grid carbon emissions within the town and street area are aggregated and calculated to obtain the total carbon emissions at the initial town and street scale; The initial 100-meter grid carbon emissions within the district level were aggregated and calculated to obtain the initial district-level total carbon emissions.

9. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level using multi-source open data according to claim 8 is characterized in that: The acquisition of district-level reference carbon emissions and town-level reference carbon emissions includes: Obtaining district-level reference carbon emissions includes: According to the historical yearbook data, the socio-economic structure data and the total carbon emissions at the district level in recent years were obtained. The socio-economic structure data included the historical population, total GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district. A multivariate linear regression model was established based on the socio-economic structure data and the district-level carbon emissions data, with the socio-economic structure data as the independent variable and the district-level carbon emissions as the dependent variable, specifically: Y=β0+β1X1+β2X2+…+β n X n +e Among them, Y is the total carbon emissions at the district level; X1, X2, …, X n They represent the historical population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district; β0, β1, …, β n is the parameter to be estimated, ε is the error term; Substitute the socio-economic structure data and district-level carbon emissions data of recent years into the multivariate linear regression model to calculate the parameters to be estimated; The population, GDP, total energy consumption of enterprises above designated size, and the number of listed companies in different industries in each district in the reference year were obtained and input into the multivariate linear regression model to obtain the district-level reference carbon emissions; Obtaining the total reference carbon emissions at the town and street level includes: The EDGAR carbon emission grid dataset of the reference year was obtained and calibrated for the total urban carbon emissions to obtain the corrected EDGAR carbon emission dataset. Based on the corrected EDGAR carbon emission dataset, the total carbon emission data at the town and street scale were statistically analyzed to obtain the reference carbon emissions at the town and street scale.

10. The method for gridding the spatial distribution of urban carbon emissions at the hundred-meter level based on multi-source open data according to any one of claims 8 to 9, characterized in that: The method for determining the fitness function includes: Among them, Y0 and Y logical They represent the initial district-level carbon emissions and district-level reference carbon emissions respectively; Y1 and Y EDGAR They represent the initial total carbon emissions at the town and street scale and the reference total carbon emissions at the town and street scale respectively. D represents the administrative district set, and E represents the town and street set.

Images