Urban carbon emission translation and accounting method based on multi-source data
Through the integration and refinement of multi-source data, combined with carbon emission coefficient calculation, carbon emission data is directly translated from the production end to the land use end, solving the problems of low spatial resolution and no differences in land use functions in the existing technology, and achieving high Accurate carbon emission identification and spatial placement improve the accuracy and accuracy of identification results.
Patent Information
- Application Number
- CN202411100616.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-19
AI Technical Summary
The existing carbon emission statistics method is based on the departmental production side, and lacks connection with the national land space planning, resulting in low spatial resolution and inability to accurately reflect the differences in urban carbon emissions within the city, and fails to accurately consider the differences in night light data in different land use functions. , resulting in inaccurate simulation of carbon emission spatial distribution.
Using multi-source data integration method, including night light data, land use data, energy department yearbooks, etc., through refined classification and model training, high-precision urban carbon emission data are generated, combined with carbon emission coefficient calculation, and the space for carbon emissions is realized Place the position and directly translate from the production end to the land end, eliminating carbon loss during the transmission process of the consumer end.
It improves the accuracy and spatial accuracy of carbon emission estimation, solves the problem that the differences in different land use functions are not reflected, simplifies the translation path, eliminates carbon loss during the transmission process, and improves the accuracy and accuracy of the identification results.
Smart Images

Figure CN118982147A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of carbon emission calculation, and in particular to a method for translating and calculating urban carbon emissions based on multi-source data. Background Art
[0002] At present, the carbon emission statistics method is mainly based on the departmental production end, and lacks a port that can be connected with national land space planning. This poses a challenge to the formulation of land control and spatial planning strategies for low-carbon cities.
[0003] In the past, the use of night light data to identify the spatial distribution of carbon emissions usually combined night light data and energy consumption data, and through certain statistical and modeling methods, the total carbon emissions within the administrative boundary were directly converted to the pixel level to show the temporal and spatial distribution of carbon emissions. For example, polynomial regression was used to fit the fitting equation between the unit night light value and the carbon emissions, and the carbon emissions within the administrative boundary were spatialized.
[0004] However, the above method also has obvious disadvantages:
[0005] First, the spatial resolution is low, usually 500 or 1000 meters, which makes it difficult to accurately reflect the differences in carbon emissions within the city;
[0006] Second, considering the differences in carbon emissions among different functional zones within the city, it is impossible to provide accurate data support for the carbon emission quota within the city;
[0007] Third, the differences in night light data among different land use functions are rarely considered, resulting in inaccurate simulation of the spatial distribution of carbon emissions. Summary of the invention
[0008] In order to overcome the above technical problems, the purpose of the present invention is to provide a method for translating and calculating urban carbon emissions based on multi-source data. The method integrates and utilizes multi-source data such as night light data, land use data, and energy department yearbooks to calculate carbon emissions, and translates urban carbon emissions from the department production end to the high-precision planned land end. It has the characteristics of high accuracy and strong feasibility.
[0009] In order to achieve the above object, the technical solution adopted by the present invention is:
[0010] A method for translating and calculating urban carbon emissions based on multi-source data, comprising the following steps;
[0011] Step 1: Using LUCC land use data, Euluc-China construction land function data, OSM road network mapping data, Baidu POI and building outline data, and using Python and ArcGIS software, the construction land part in the LUCC land use data is refined and classified to generate raster data based on different functional types of urban areas including construction land (residential, commercial, public service, industrial), blue-green space land (arable land, forest land, grassland, water area, unused land, urban green space) and road land, and the data is used as the basic unit of carbon emission identification to improve the accuracy of carbon emission data identification;
[0012] Step 2: Based on the regional energy balance table in the China Energy Statistical Yearbook and the steel and cement production data in the government bulletin, the total carbon emissions of construction land are calculated using the IPCC carbon emission calculation method;
[0013] Through data translation, the total carbon emissions are divided into carbon emissions from industrial land, residential land, commercial land, public service land and road land; at the same time, the raster data of blue-green space land obtained in step 1 is used to calculate the area, and the total carbon emissions from grassland and unused land, cultivated land, forest land and water are calculated by multiplying the land area with the corresponding carbon emission coefficient;
[0014] Through data translation, the carbon emissions generated by various industrial sectors are converted into carbon emissions of various urban land use functions; this subdivision helps to more accurately identify and quantify the carbon emission characteristics of different urban functional areas and has strong feasibility.
[0015] Step 3: Using the resampled night light data, ArcGIS software and arcpy library are used to classify and sum the night light DN values of different functional types in step 1, and the carbon emission coefficient of each functional land is calculated by combining the total carbon emissions of each functional land in step 2;
[0016] Step 4: By multiplying the nighttime light value of each plot by its corresponding carbon emission coefficient, the carbon emissions of each plot are obtained to achieve the spatial location of carbon emissions.
[0017] The step 1 is specifically as follows:
[0018] The generation of plot form data for the city-wide functional zoning includes four steps: model training, plot division, land use identification, and fishing net cutting;
[0019] 1) Model training:
[0020] First, ArcGIS software was used to clip the Euluc-China urban functional zoning data of the target city;
[0021] Secondly, the TF-IDF weighted transformation is performed using the Baidu business POI big data after spatial correction in 2018 to generate a business weighted feature attribute table;
[0022] Secondly, using Baidu's building outline data in 2018, we calculated the highest building height, average height, average base area and floor area ratio of each plot, generated a building form feature attribute table, and then constructed a training database;
[0023] Finally, with the help of Python's sklearn library, the land use function field of Euluc-China functional zoning data is used as the dependent variable, and the business format weighted feature attribute table and building form feature attribute table are used as independent variables. The GBDT model is used for training to obtain the urban land recognition model.
[0024] 2) Land division:
[0025] First, the buffer tool in ArcGIS was used to perform buffer operations on the OSM road network mapping data of the target city in the estimation year according to the road width of different roads to generate road network surface data;
[0026] Secondly, ArcGIS software was used to clip the LUCC land use data of the target city in the estimation year; finally, the road network surface data was merged with the LUCC land use data to generate land use data containing road network segmentation, with attributes including "construction land", "road land", "cultivated land", "grassland", "forest land", "water area", and "unused land";
[0027] 3) Land Identification:
[0028] First, extract the elements with the attribute of "construction land" from the land use data generated in the previous step as the basis for land use identification;
[0029] Secondly, we use Baidu’s business POI big data and building outline data after spatial correction in the estimated year, calculate the weighted characteristic attributes of the business format and the characteristic attributes of the building form in the estimated year, and build a recognition database;
[0030] Finally, the urban land recognition model trained in step 1) is used to classify and recognize the functions of construction land, subdividing "construction land" into residential land, public service land, commercial land and industrial land, and integrating them back into the land use data to generate the global functional zoning data;
[0031] 4) Fishing net cutting:
[0032] Use ArcGIS to create a fishnet tool, with the target city as the processing range, to create fishnet data with an accuracy of 500 meters; then, use the fishnet data to intersect with the global functional zoning data to generate the global functional zoning data after the fishnet cutting.
[0033] The step 2 is specifically as follows:
[0034] The total carbon emission estimation at the land use end includes three steps: production end carbon emission estimation, land use end carbon emission translation, and blue-green space carbon emission estimation;
[0035] 1) Estimation of carbon emissions at the production end:
[0036] According to the carbon emission classification method in the IPCC's latest "2019 Revised Edition of the IPCC 2006 National Greenhouse Gas Inventory Guidelines" (hereinafter referred to as the "Guidelines"), the production-end carbon emission data is divided into four types: energy activities, industrial processes and product use, agriculture, forestry and other land use, and waste treatment for calculation; the carbon emissions from energy activities are calculated based on the provincial energy consumption data of the "China Energy Statistical Yearbook";
[0037]
[0038] Where EC is the estimated total carbon emissions from various energy consumption; i represents the type of energy consumption, including coal, tar, climate, etc.; E i is the total consumption of the i-th type of energy in each province; CF i is the calorific value; CC i is the carbon content; COF i is the oxidation factor, CF i ×CC i ×COF i It is called the carbon emission factor, and CF i ×CC i ×COF i ×3.67 is called the carbon emission factor;
[0039] Carbon emissions from industrial processes and product use are limited by data acquisition. Only the departments for which data is provided in the government bulletin of the target city can be calculated, including carbon emissions from cement production, steel production, calcium carbide production, and lime production. The formula for estimating CO2 emissions from the cement production process is as follows:
[0040] E cement =AD×EF
[0041] In the formula, E cement is the carbon emission of cement production process, AD is the cement clinker output, and EF is the average carbon emission factor of cement process;
[0042] The carbon emissions estimation formula for the steel production process is as follows:
[0043]
[0044] In the formula, E steel is the carbon emissions from steel production; AD l is the amount of limestone consumed by steel companies as a solvent; EF l is the emission factor for limestone consumption as a solvent; AD d is the amount of dolomite as solvent; EF d is the emission factor for dolomite consumption as a solvent; AD r is the amount of pig iron used in steelmaking; F r AD is the average carbon content of pig iron used for steelmaking; s is the output of steel produced by steelmaking; F s It is the average carbon content of steel products produced by steelmaking;
[0045] The carbon emission estimation formula of calcium carbide production process is as follows:
[0046] E cc =AD c ×EF c
[0047] In the formula, E cc is the carbon emission in the production process of calcium carbide; AD c is the calcium carbide output, EF c is the average carbon emission factor of the calcium carbide production process;
[0048] The carbon emission estimation formula of lime production process is as follows:
[0049]
[0050] In the formula, E lime is the carbon emission of lime production process, EF s,i is the carbon emission factor for lime of type i, M l,i is the lime production of type i, CF lkd,i is the LKD correction factor for type i lime, C h,i is the correction factor for type i slaked lime.
[0051] 2)Translation of carbon emissions at the land use end:
[0052] First, the carbon emissions of industry and construction were split to obtain the carbon emissions of construction and industry separately. Then, the total building area of each type of land was calculated using building outline data, and the energy consumption ratio of residential, commercial and public service buildings in the target city was estimated based on the reference amount of energy consumption of residential, commercial and public service buildings in the "Civil Building Energy Consumption Standard GB / T51161-2016", so as to split the carbon emissions of the construction industry into carbon emissions of residential, commercial and public service land. Finally, the remaining carbon emissions from the production end were translated to the land end to obtain the total carbon emissions of five types of land: residential, commercial, public service, transportation and industrial.
[0053] 3) Estimation of carbon emissions from blue and green spaces:
[0054] The carbon emissions of water areas, grasslands, cultivated lands, forest lands and unused lands are estimated by multiplying the area by the corresponding coefficient. Green land in urban construction land is split into forest land and grassland at a ratio of 1:1.
[0055] The step 3 is specifically as follows:
[0056] The spatialization of carbon emission data includes three steps: preprocessing and resampling of night light data, spatial statistics of night light values and land use type summary, and calculation and spatial location of carbon emission coefficients.
[0057] 1) Night light data preprocessing and resampling:
[0058] The 500-meter-resolution monthly night light dataset provided by the National Oceanic and Atmospheric Administration (NOAA) of the United States was used to process outliers and extract masks before merging the data to generate an annual night light dataset. Subsequently, the bilinear interpolation (BILINEAR) resampling technique was used to increase the resolution of the annual night light data to 10 meters to obtain more detailed spatial distribution characteristics.
[0059] 2) Spatial statistics of nighttime light values and summary of land use types:
[0060] The zoning statistics tool of ArcGIS software was used to conduct statistical analysis on the resampled nighttime light data at the plot level. Then, the data were classified and summarized according to the different land use functions, such as residential, commercial, public service, industry, transportation, grassland and unused land, cultivated land, water area and forest land, to obtain the total amount of nighttime light for each type of land use.
[0061] 3) Carbon emission coefficient calculation and spatial location:
[0062] The carbon emission coefficient per unit night light value is calculated by the ratio of the total carbon emissions of the land use type to its corresponding total night light value. Secondly, the night light value of each plot is multiplied by its corresponding land use function carbon emission coefficient to obtain the carbon emissions of each plot and realize the spatial location of carbon emissions.
[0063] Beneficial effects of the present invention:
[0064] Using this method, not only does it retain the intuitiveness of night light data in characterizing the levels of carbon emissions, but it also effectively eliminates the problem of not reflecting the differences in land use functions that may have occurred when using night light data as a whole for carbon emissions estimation in the past through precise coefficient calculations, thereby improving the accuracy of carbon emission estimates.
[0065] In step 1, two sub-steps, model training and land use identification, can finely identify the land use functions of different years, avoiding the time limitations of previous studies that used land use function data of a single year to estimate carbon emission data of different years, thereby improving the temporal accuracy of urban carbon emission identification results. The two sub-steps of plot division and fishing net cutting can improve the accuracy of night light data used in subsequent steps from 250 / 500 meters to 10 meters, thereby improving the spatial accuracy of urban carbon emission identification results.
[0066] Step 2 proposes a calculation process for carbon emission translation directly from the production end to the land use end, which simplifies the secondary translation path from the production end to the consumption end to the land use end in previous studies, and directly translates carbon emissions from the production end to the land use end, eliminating the carbon loss generated by the consumption end during the forward and backward transmission process, thereby improving the accuracy of the carbon emission translation results at the land use end.
[0067] Step 3 uses night light and land use function data to perform spatial identification of urban carbon emissions, which improves the accuracy of the identification results and solves the problem of inaccurate identification of industrial and commercial land caused by using only night light data in previous studies. It combines the high precision of night light data with the high accuracy of carbon emission data based on land use functions, thereby improving the accuracy of urban carbon emission identification results. BRIEF DESCRIPTION OF THE DRAWINGS
[0068] Figure 1 It is a schematic diagram of the technical route of multi-source data and method thereof of the present invention.
[0069] Figure 2 This is a translation diagram of carbon emissions from the production end and carbon emissions from the land use end of the present invention.
[0070] Figure 3 It is a comparison diagram of the present invention and other methods.
[0071] Figure 4Schematic diagram of the experimental method using Changchun City as an example. DETAILED DESCRIPTION
[0072] The present invention will be further described in detail below in conjunction with the accompanying drawings.
[0073] like Figure 1 As shown: A method for translating and calculating urban carbon emissions based on multi-source data, comprising the following steps:
[0074] Step 1: Using LUCC land use data, Euluc-China construction land function data, OSM road network mapping data, Baidu POI and building outline data, and using Python and ArcGIS software, the construction land part in the LUCC land use data is refined and classified to generate plot data of the city's full-area functional zoning, including construction land (residential, commercial, public service, industrial), blue-green space land (arable land, forest land, grassland, water area, unused land, urban green space) and road land, and the data is used as the basic unit for carbon emission identification, thereby improving the accuracy of carbon emission data identification;
[0075] Step 2: Based on the regional energy balance table in the China Energy Statistical Yearbook and the steel and cement production data in the government bulletin, the total carbon emissions of construction land are calculated using the IPCC carbon emission calculation method;
[0076] Through data translation, the total carbon emissions are divided into carbon emissions from industrial land, residential land, commercial land, public service land and road land; at the same time, the blue-green space land obtained in step 1 is used to calculate the area, and the total carbon emissions from grassland and unused land, cultivated land, forest land and water are calculated by multiplying the land area with the corresponding carbon emission coefficient;
[0077] Through data translation, the carbon emissions generated by various industrial sectors are converted into carbon emissions of various urban land use functions; this subdivision helps to more accurately identify and quantify the carbon emission characteristics of different urban functional areas and has strong feasibility.
[0078] Step 3: Using the resampled night light data, ArcGIS software and arcpy library are used to perform zoning statistics on the night light DN values of each plot in step 1; after summarizing the total night light DN values of each functional type, the carbon emission coefficient of each functional land is calculated in combination with the total carbon emissions of each functional land in step 2;
[0079] Finally, by multiplying the night light value of each plot by its corresponding carbon emission coefficient, the carbon emissions of each plot can be obtained, thus realizing the spatial location of carbon emissions.
[0080] The step 1 can be specifically divided into four parts:
[0081] 1) Construction land function classification model training
[0082] ArcGIS software was used to clip the Euluc-China urban function zoning data of the target city. Using the 2018 Baidu business POI big data, the business weighted feature attribute table was generated through TF-IDF weighted transformation. Using the Baidu building outline data of the same year, the building characteristics of the plot were calculated, including the highest height, average height, average base area and floor area ratio, and the building form feature attribute table was constructed. Combining the business and building feature attribute tables, a training database was constructed, and the GBDT model in Python's sklearn library was used to train the model with the land use function field as the dependent variable to form an urban land recognition model.
[0083] 2) Estimated year land use division
[0084] First, the buffer tool in ArcGIS was used to perform buffer operations on the OSM road network mapping data of the target city in the estimation year according to the road width of different roads to generate road network surface data;
[0085] Secondly, ArcGIS software was used to clip the LUCC land use data of the target city in the estimation year; finally, the road network surface data was merged with the LUCC land use data to generate land use data containing road network segmentation, with attributes including "construction land", "road land", "cultivated land", "grassland", "forest land", "water area", and "unused land";
[0086] 3) Identification of construction land functions in the estimated year
[0087] Extract the elements with the attribute "construction land" from the land use data generated in the previous step as the basis for identification. Combine the Baidu business POI big data and building outline data of the estimated year, calculate the weighted characteristic attributes of the business and the characteristic attributes of the building form, and build a new identification database. Apply the urban land identification model trained in step 1 to classify and identify the construction land, subdivide it into residential, public service, commercial and industrial land, and integrate it back into the land use data to form a global functional zoning data set.
[0088] 4) Urban functional zoning and fishing net cutting
[0089] Use ArcGIS to create fishing net data with a precision of 500 meters, covering the target city. Intersect the fishing net data with the global functional zoning data to generate the land parcel data cut by the fishing net.
[0090] The step 2 can be specifically divided into three parts:
[0091] 1) Estimation of carbon emissions at the production end
[0092] According to the IPCC's "2006 IPCC National Greenhouse Gas Inventory Guidelines 2019 Revision" (hereinafter referred to as the "Guidelines"), the production-end carbon emission data is divided into four types: energy activities, industrial processes and product use, agriculture, forestry and other land use, and waste treatment for calculation. Energy activity carbon emissions are calculated based on the provincial energy consumption data of the "China Energy Statistical Yearbook", combined with the population ratio to the target city, and according to the estimation method of the "Guidelines" and the "Guidelines for the Preparation of Provincial Greenhouse Gas Inventories". The specific formula is as follows:
[0093]
[0094] Among them, ECEC represents the estimated total carbon emissions from energy consumption; Ei is the total consumption of the i-th type of energy; CFi is the calorific value; CCi is the carbon content; COFi is the oxidation factor; and the carbon emission coefficient is CFi×CCi×COFi.
[0095] Carbon emissions from industrial processes and product use are limited by data acquisition. Only the departments for which data is provided in the government bulletin of the target city can be calculated, including carbon emissions from cement production, steel production, calcium carbide production, and lime production. The formula for estimating CO2 emissions from the cement production process is as follows:
[0096] E cement =AD×EF
[0097] In the formula, E cement is the carbon emission of cement production process, AD is the cement clinker output, and EF is the average carbon emission factor of cement process;
[0098] The carbon emissions estimation formula for the steel production process is as follows:
[0099]
[0100] In the formula, E steel is the carbon emissions from steel production; AD l is the amount of limestone consumed by steel companies as a solvent; EF l is the emission factor for limestone consumption as a solvent; AD d is the amount of dolomite used as solvent; EF d is the emission factor for dolomite consumption as a solvent; AD r is the amount of pig iron used in steelmaking; F r AD is the average carbon content of pig iron used for steelmaking; s is the output of steel produced by steelmaking; F s It is the average carbon content of steel products produced by steelmaking;
[0101] The carbon emission estimation formula of calcium carbide production process is as follows:
[0102] E cc =AD c ×EF c
[0103] In the formula, E cc is the carbon emission in the production process of calcium carbide; AD c is the calcium carbide output, EF c It is the average carbon emission factor of the calcium carbide production process.
[0104] The carbon emission estimation formula of lime production process is as follows:
[0105]
[0106] In the formula, E lime is the carbon emission of lime production process, EF s,i is the carbon emission factor for lime of type i, M l,i is the lime production of type i, CF lkd,i is the LKD correction factor for type i lime, C h,i is the correction factor for type i slaked lime.
[0107] 2)Translation of carbon emissions at the land use end
[0108] The carbon emissions of industry and construction were split to obtain the carbon emissions of construction and industry. The total building area of each type of land was calculated using the building outline data, and the energy consumption ratio of residential, commercial and public service buildings was estimated according to the "Civil Building Energy Consumption Standard GB / T51161-2016", and the carbon emissions of the construction industry were split into carbon emissions of the corresponding land. Figure 2 The remaining carbon emissions from the production end are translated to the land use end to obtain the total carbon emissions of five types of land: residential, commercial, public services, transportation and industrial.
[0109] 3) Estimation of carbon emissions from blue and green spaces
[0110] Carbon emissions from water areas, grasslands, cultivated land, forests and unused lands are estimated by multiplying the area by the corresponding coefficient. For green land in urban construction land, it is divided into forests and grasslands at a ratio of 1:1 to estimate carbon emissions.
[0111] The step 3 can be divided into three parts:
[0112] 1) Night light data preprocessing and resampling
[0113] The 500-meter-resolution monthly night light data set provided by NOAA was used to process outliers and extract masks, and then merged to generate an annual night light data set. The bilinear interpolation (BILINEAR) resampling technology was applied to increase the spatial resolution of the annual night light data to 10 meters to capture more detailed spatial features.
[0114] 2) Spatial statistics and summary of night light values
[0115] The zoning statistics tool of ArcGIS software was used to conduct spatial statistical analysis on the resampled nighttime light data at the plot level. The total amount of nighttime light for each type of land use was obtained by classifying and summarizing the different land use functions, such as residential, commercial, public service, industrial, transportation, grassland and unused land, cultivated land, water area and forest land.
[0116] 3) Carbon emission coefficient calculation and spatial location
[0117] The carbon emission coefficient per unit night light value is calculated by the ratio of the total carbon emission of the land use type to the total night light value. The carbon emission of each plot is obtained by multiplying the night light value of each plot by its corresponding land use function carbon emission coefficient, thus realizing the spatial positioning of carbon emissions.
[0118] Figure 2 for Figure 1 Supplementary explanation of the translation of land carbon emissions in step 2, that is, the correspondence between carbon emissions from energy activities on the production side and carbon emissions from spatial activities on the land side: the energy industry with carbon emissions from energy combustion in energy activities can be translated into residential energy use, commercial activity energy use, public building equipment energy use, industrial and mining energy use, and crop production energy use; the construction industry can be translated into the construction and demolition of residential, commercial and public buildings, as well as industrial and mining energy use; the transportation industry can be translated into urban passenger transport, intercity passenger transport, and freight transport; the mining industry, metal industry, and chemical industry in industrial production processes and product use can be translated into industrial and mining production processes; in agriculture, forestry and other land uses, farmland and forest land are translated into crop production and forestry and generate corresponding carbon sinks, while the remaining grasslands, waters and unused land will all produce carbon sinks.
[0119] Figure 3A comparative analysis of the research method and other existing methods in estimating the spatial distribution of carbon emissions is presented. The carbon emission distribution map provided by ODIAC data has the problems of large scale and insufficient precision. Its distribution characteristics are manifested as a circle-like pattern with high in the central area and low in the peripheral area, which is inconsistent with the actual carbon emission pattern; the method of simulating the spatial distribution of carbon emissions based on night light data, its results are completely dependent on the spatial distribution of night light data, and fail to reflect the differences in carbon emissions of different urban functional types; the estimation method combined with urban function data has solved to a certain extent the problem that the differences in urban functions cannot be reflected when only night light data are used, but due to its large scale and presentation in the form of raster data, this method has limitations in the detailed display of the spatial distribution of carbon emissions.
[0120] In contrast, this research method, based on comprehensive consideration of existing methods, uses vector data in the form of plots as the basic unit of estimation, which can not only accurately reflect the differences in carbon emissions between different land use types, but also significantly improve the accuracy of the estimation results.
[0121] Example:
[0122] According to the method steps, take the carbon emission estimation of Changchun City in 2020 as an example: First, through the four steps of model training, plot division, land use identification and fishing net cutting in step 1, the form and function zoning data of Changchun City in 2020 are obtained (see attached Figure 4 a, which is a schematic diagram of the central urban area of Changchun City's 2020 land parcel form and functional zoning data); secondly, the production-end carbon emission estimation step in step 2 is performed to obtain the production-end carbon emissions. Figure 4 e, and translate it into land-use carbon emission data, and estimate the carbon emissions of the blue-green space; then, perform the night light data preprocessing and resampling, spatial statistics of night light values, and land use type summary steps in step 3, and count the night light values of each plot unit (such as Figure 4 b), and count the total nighttime light values of various types of land; finally, perform the carbon emission coefficient calculation and spatial location steps in step 3 to locate the carbon emissions at the land end in space (as shown in Figure 4 c), the calculation of the spatial data of carbon emissions in Changchun in 2020 is completed. In addition, the result data can be integrated twice and merged into fishing net unit data or raster data ( Figure 4 d).
Claims
1. A method for translating and calculating urban carbon emissions based on multi-source data, characterized in that: The steps include: Step 1: Using LUCC land use data, Euluc-China construction land function data, OSM road network mapping data, Baidu POI and building outline data, and using Python and ArcGIS software, the construction land part of the LUCC land use data is refined and classified to generate raster data based on different functional types of urban areas including construction land, blue-green space land, and road land, and the data is used as the basic unit for carbon emission identification; Step 2: Use the IPCC carbon emission calculation method to calculate the total carbon emissions of construction land; Through data translation, the total carbon emissions are divided into carbon emissions from industrial land, residential land, commercial land, public service land and road land; at the same time, the raster data of blue-green space land obtained in step 1 is used to calculate the area, and the total carbon emissions from grassland and unused land, cultivated land, forest land and water are calculated by multiplying the land area with the corresponding carbon emission coefficient; Through data translation, the carbon emissions generated by various industrial sectors are converted into carbon emissions of various land use functions in the city; Step 3: Using the resampled night light data, ArcGIS software and arcpy library are used to classify and sum the night light DN values of different functional types in step 1, and the carbon emission coefficient of each functional land is calculated by combining the total carbon emissions of each functional land in step 2; Step 4: By multiplying the nighttime light value of each plot by its corresponding carbon emission coefficient, the carbon emissions of each plot are obtained to achieve the spatial location of carbon emissions.
2. According to the method of translating and calculating urban carbon emissions based on multi-source data in claim 1, it is characterized in that: The step 1 is specifically as follows: The data generation of plot forms for the functional zoning of the entire city area includes four steps: model training, plot division, land use identification, and fishing net cutting; 1) Model training: First, ArcGIS software was used to clip the Euluc-China urban functional zoning data of the target city; Secondly, the TF-IDF weighted transformation is performed using the Baidu business POI big data after spatial correction in 2018 to generate a business weighted feature attribute table; Secondly, using Baidu's building outline data in 2018, we calculated the highest building height, average height, average base area and floor area ratio of each plot, generated a building form feature attribute table, and then constructed a training database; Finally, with the help of Python's sklearn library, the land use function field of Euluc-China functional zoning data is used as the dependent variable, and the business format weighted feature attribute table and building form feature attribute table are used as independent variables. The GBDT model is used for training to obtain the urban land recognition model. 2) Land division: First, the buffer tool in ArcGIS was used to perform buffer operations on the OSM road network mapping data of the target city in the estimation year according to the road width of different roads to generate road network surface data; Secondly, ArcGIS software was used to clip the LUCC land use data of the target city in the estimated year; finally, the road network surface data was merged with the LUCC land use data to generate land use data containing road network segmentation, with attributes including "construction land", "road land", "cultivated land", "grassland", "forest land", "water area", and "unused land"; 3) Land Identification: First, extract the elements with the attribute "construction land" from the land use data generated in the previous step as the basis for land use identification; Secondly, we use Baidu’s business POI big data and building outline data after spatial correction in the estimated year, calculate the weighted characteristic attributes of the business format and the characteristic attributes of the building form in the estimated year, and build a recognition database; Finally, the urban land recognition model trained in step 1) is used to classify and recognize the functions of construction land, subdividing "construction land" into residential land, public service land, commercial land and industrial land, and integrating them back into the land use data to generate global functional zoning data; 4) Fishing net cutting: Use ArcGIS to create a fishnet tool, with the target city as the processing range, to create fishnet data with an accuracy of 500 meters; then, use the fishnet data to intersect with the global functional zoning data to generate the global functional zoning data after the fishnet cutting.
3. The method for translating and calculating urban carbon emissions based on multi-source data according to claim 1 is characterized in that: The step 2 is specifically as follows: The total carbon emission estimation at the land use end includes three steps: production end carbon emission estimation, land use end carbon emission translation, and blue-green space carbon emission estimation; 1) Estimation of carbon emissions at the production end: The carbon emission data at the production end is divided into four types: energy activities, industrial processes and product use, agriculture, forestry and other land use, and waste treatment for calculation; the carbon emissions from energy activities are calculated; Where EC is the estimated total carbon emissions from various energy consumption; i represents the type of energy consumption, including coal, tar, and climate; E i is the total consumption of the i-th type of energy in each province; CF i is the calorific value; CC i is the carbon content; COF i is the oxidation factor, CF i ×CC i ×COF i is called the carbon emission factor, and CF i ×CC i ×COF i ×3.67 is called the carbon emission factor; Carbon emissions from industrial processes and product use are limited by data acquisition. Only the carbon emissions from cement production and steel production processes provided by the target city government bulletin are calculated. The formula for estimating CO2 emissions from the cement production process is as follows: AND cement =AD×EF In the formula, E cement is the carbon emission of cement production process, AD is the cement clinker output, and EF is the average carbon emission factor of cement process; The carbon emissions estimation formula for the steel production process is as follows: In the formula, E steel is the carbon emissions from steel production; AD l is the amount of limestone consumed by steel companies as a solvent; EF l is the emission factor for limestone consumption as a solvent; AD d is the amount of dolomite as solvent; EF d is the emission factor for dolomite consumption as a solvent; AD r is the amount of pig iron used in steelmaking; F r AD is the average carbon content of pig iron used for steelmaking; s is the output of steel produced by steelmaking; F s It is the average carbon content of steel products produced by steelmaking; The carbon emission estimation formula of calcium carbide production process is as follows: AND cc =AD c ×EF c In the formula, E cc AD is the carbon emission in the production process of calcium carbide; c is the calcium carbide output, EF c is the average carbon emission factor of the carbide production process; The carbon emission estimation formula of lime production process is as follows: In the formula, E lime is the carbon emission of lime production process, EF s,i is the carbon emission factor for lime of type i, M l,i is the lime production of type i, CF lkd,i is the LKD correction factor for type i lime, C h,i is the correction factor for type i slaked lime; 2)Translation of carbon emissions at the land use end: First, the carbon emissions of industry and construction were split to obtain the separate carbon emissions of construction and industry. Then, the total building area of each type of land was calculated using building outline data, and the energy consumption ratio of residential, commercial, and public service buildings in the target city was estimated, and the carbon emissions of the construction industry were split into carbon emissions of residential, commercial, and public service land. Finally, the remaining carbon emissions from the production end were translated to the land use end to obtain the total carbon emissions of the five types of land: residential, commercial, public service, transportation, and industrial. 3) Estimation of carbon emissions from blue and green spaces: The carbon emissions of water areas, grasslands, cultivated lands, forest lands and unused lands are estimated by multiplying the area by the corresponding coefficient. Green land in urban construction land is split into forest land and grassland at a ratio of 1:
1.
4. The method for translating and calculating urban carbon emissions based on multi-source data according to claim 1 is characterized in that: The step 3 is specifically as follows: The spatialization of carbon emission data includes three steps: preprocessing and resampling of night light data, spatial statistics of night light values and land use type summary, and calculation and spatial location of carbon emission coefficients. 1) Night light data preprocessing and resampling: The 500-meter-resolution monthly night light dataset provided by the National Oceanic and Atmospheric Administration (NOAA) of the United States was used to process outliers and extract masks before merging the data to generate an annual night light dataset. Subsequently, the bilinear interpolation (BILINEAR) resampling technique was used to increase the resolution of the annual night light data to 10 meters to obtain more detailed spatial distribution characteristics; 2) Spatial statistics of nighttime light values and summary of land use types: Using the zoning statistics tool of ArcGIS software, we conducted a statistical analysis of the resampled nighttime light data at the plot level. Then, we classified and summarized the data according to the different land use functions to obtain the total amount of nighttime light for each type of land use. 3) Carbon emission coefficient calculation and spatial location: The carbon emission coefficient per unit night light value is calculated by the ratio of the total carbon emissions of the land use type to its corresponding total night light value. Secondly, the night light value of each plot is multiplied by its corresponding land use function carbon emission coefficient to obtain the carbon emissions of each plot and realize the spatial location of carbon emissions.
Citation Information
Cited By
Multi-source open data-based urban carbon emission hectometer spatial distribution gridding method
CN119692598A
High-value industrial land identification method and system based on land economic value
CN119963034A
City-scale-oriented multi-source heterogeneous data fusion carbon emission spatial distribution simulation method
CN120217685A
A city-scale multi-source heterogeneous data fusion carbon emission spatial distribution simulation method
CN120217685B
Carbon emission prediction method and device based on urban system model, electronic equipment and storage medium
CN120373574A
