Multi-source data fusion high-resolution carbon emission list construction method based on GWR model

Abstract

The invention discloses a multi-source data fusion high-resolution carbon emission list construction method based on a GWR model, and the method comprises the steps: collecting the carbon emission statistical data of electric power, industrial, traffic and civil departments, combining the night light, road density, population density, point source emission position and other space agent data, and improving a conventional downscaling method. The spatial heterogeneity of carbon emission is dynamically captured by using a GWR model, and an enhanced spatial proxy index is generated. Experimental results show that the method is remarkably superior to a traditional method in spatial resolution and prediction precision, and technical support is provided for precise carbon emission management and policy making.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission accounting, and in particular to a method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of a GWR model. Background Art

[0002] As the global climate change problem becomes increasingly serious, carbon emission control has become an important issue in global environmental protection and economic development. Accurate carbon emission inventories are of great significance for formulating reasonable emission reduction policies and evaluating the effectiveness of policy implementation. However, traditional carbon emission inventories are usually based on administrative regions, with low resolution, and it is difficult to reflect the spatial heterogeneity of actual emission distribution, especially in areas with a high degree of urbanization, where the spatial pattern of carbon emissions shows significant unevenness. Therefore, making a high-resolution carbon emission inventory is one of the important technical challenges in current carbon emission research.

[0003] In recent years, data methods based on remote sensing and geographic information systems have provided new possibilities for the spatialization of carbon emissions. Among them, night light data, population density data, industrial emission location data, etc. have become important spatial proxy variables for studying the spatial distribution of carbon emissions. Night light data can be used as a proxy indicator of human activity intensity, which has a strong correlation with carbon emissions; population density data can reflect the spatial distribution of residents' activities and energy consumption; road density data is closely related to traffic emission intensity; point source emission location data provides direct spatial distribution information of high-pollution emission industries and facilities.

[0004] Although the above data provide rich basic information for the preparation of high-resolution carbon emission inventories, how to integrate these multi-source data and effectively improve the prediction accuracy remains a key issue. The geographically weighted regression model is a tool widely used in spatial statistical analysis. By introducing the weight of spatial location, it can effectively capture the spatial heterogeneity between variables and their local variation characteristics. Therefore, the use of the GWR model combined with night lights, population density, road density and point source emission data can provide a higher accuracy and spatial resolution solution for the preparation of carbon emission inventories.

[0005] Existing research mainly focuses on the spatial processing of carbon emissions from a single data source or atmospheric inversion model, which results in insufficient resolution and accuracy in the generated inventory and the inability to capture the spatial characteristics of different emission sources. The multi-source data integration method based on the GWR model can significantly improve the accuracy of the carbon emission inventory by utilizing the complementarity and synergy of different data sources, and provide a scientific basis for the spatial management and policy support of carbon emissions. In addition, this method can also adapt to the application needs of different scales and regions, and has strong universality and promotion value.

[0006] At present, the production of global and regional carbon emission inventories has formed several mature representative methods and databases, including ODIAC, EDGAR and MEIC. These inventories are widely used in carbon emission research and policy making, but there are still limitations in practical applications. For example, ODIAC provides a global carbon emission inventory with a spatial resolution of 1KM. Although it has a high resolution on a global scale, it does not use spatial proxy data to characterize the line source emission characteristics of the transportation sector. And due to the openness of the data, the location of industrial emissions in some regions is difficult to obtain, so the current emission inventory lacks point source emission data and cannot meet the needs of precise emission reduction at the local scale. Similarly, the EDGAR and MEIC inventories invert carbon emissions through atmospheric inversion models, resulting in a coarser spatial resolution and unable to reflect detailed emission characteristics such as transportation line source emissions and industrial point source emissions.

[0007] There are also many studies on methods for constructing high-resolution carbon emission inventories that have not yet formed an inventory, but most of them use simple allocation weights or linear regression methods, and do not fully integrate the collaborative information of multiple source data such as night lights, road density, and point source emission locations. This method is difficult to accurately reflect the spatial heterogeneity between different emission sources, resulting in insufficient prediction accuracy. Especially in areas with significant regional differences, model results are prone to systematic errors. Summary of the invention

[0008] The production of accurate and high-resolution carbon emission inventories has become a key link in carbon emission reduction research. More accurate characterization of carbon emission spatial patterns can facilitate more accurate atmospheric model simulation and emission tracing. Therefore, the purpose of this invention is to:

[0009] 1. Improve the spatial resolution of the carbon emission inventory and refine the emission characteristics of different regions;

[0010] 2. In the process of modeling point, line and surface multi-source emissions, make full use of the collaborative information of various types of data and optimize the multivariate modeling method;

[0011] 3. Overcome the problem that existing models are not able to handle local heterogeneity and enhance the model's ability to depict regional carbon emission differences;

[0012] This paper provides a method for constructing a high-resolution carbon emission inventory by integrating multi-source data with the GWR model. By integrating statistical data and spatial proxy data, a 1km spatial resolution grid distribution of carbon emissions is achieved, providing support for carbon emission research and management at different scales.

[0013] To achieve the above object, the present invention provides the following technical solutions:

[0014] The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model includes the following specific steps:

[0015] S-1. Data collection and processing: Collect energy consumption and spatial distribution data of the four sectors of electricity, industry, transportation, and civil use; multiply the energy consumption by the corresponding carbon emission factor to obtain the carbon emissions of each sector, and add up the carbon emissions of the four sectors to obtain the total carbon emissions of the region.

[0016] S-2. Establish a carbon emission downscaling model: Use the geographic weight regression (GWR) model to build the relationship between the total carbon emissions and the spatial proxy index, and distribute the total carbon emissions to a grid with a resolution of 1 km. The formula is as follows:

[0017]

[0018] in, represents the regression value of carbon emissions of sector i in region k in year t; subscripts i = 1, 2, 3, 4 correspond to the four sectors of electricity, industry, transportation, and civil use, respectively, and year t is the natural year;

[0019] β n (u k ,v k ) is the slope, which represents the sensitivity of carbon emissions to spatial distribution data and reflects the increase in carbon emissions caused by changes in spatial distribution data per unit; where (u k ,v k ) represents the geographical coordinates of region k; It represents the spatial distribution data of the power and industrial sectors obtained by extracting the night light index of the power and industrial point locations in the kth region of the i-th sector in the t-th year; It represents the spatial distribution data of the transportation department or civil department of the i-th department in the k-th region in the t-th year obtained by coupling the formula previously. 0.34 is the default critical value of the night lighting value.

[0020] S-3. Correction of carbon emission regression value: Adjust the regression value obtained in the regression model. The formula is:

[0021]

[0022] in, represents the carbon emission grid value of the kth region after correction, represents the high-resolution grid value of carbon emission regression of the kth region obtained by using the GWR regression model; represents the total predicted carbon emission value of the kth region obtained by using the GWR regression model, It means that the energy consumption is multiplied by the corresponding carbon emission factor to get the calculated carbon emission value of each department in the kth region, thus obtaining the carbon emission inventory.

[0023] Preferably, the power department uses public global power plant location data to obtain spatial distribution data, extracts the night light values ​​corresponding to the power plant locations, and uses the extracted night light values ​​as the spatial distribution data of the power department; the night light values ​​use the global 500-meter resolution "NPP-VIIRS-like" night light data.

[0024] Preferably, the location data of the industrial sector is obtained by using NASA's VIIRS 375m active fire product, and the scattered fire burning points are removed by applying the density-based noise spatial clustering algorithm DBSCAN, and the industrial point locations with similar distances are preliminarily screened out. Then, combined with the land use data, the fire points on the artificial surface are screened again to effectively distinguish industrial heat sources from other thermal anomalies. In order to avoid repeated statistics, the industrial heat source points within the power plant are excluded, thereby obtaining the accurate location of the industrial sector; the night light values ​​corresponding to the processed industrial department locations are extracted, and the extracted night light values ​​are used as the spatial distribution data of the industrial sector.

[0025] Preferably, the transportation department calculates road density data by using the public global road data OpenStreetMap data, and couples it with nighttime light data to obtain spatial distribution data;

[0026] The resident department obtains spatial distribution data by coupling the global population density data and nighttime light data published by the LandScan database;

[0027] Preferably, the spatial distribution data coupling formula of the transportation and resident sectors is as follows:

[0028]

[0029] Among them, D i NTL represents spatial distribution data, i represents the traffic or residential sector, D represents road or population density data, NTL represents the night lighting index, and 0.34 is the default critical value of the night lighting value.

[0030] Preferably, the k region is a country, province or city; the spatial distribution data is: night light index.

[0031] Preferably, the density-based noise application spatial clustering algorithm DBSCAN is used, wherein the algorithm clustering distance is set to 800 meters and the minimum number of clusters of point elements is set to 5.

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

[0033] 1. The present invention takes into account the integrity of statistical data and the refined characteristics of spatial proxy data. By optimizing the traditional downscaling method, it significantly improves the spatial resolution and prediction accuracy of the carbon emission inventory.

[0034] 2. The present invention makes full use of multi-source spatial proxy data such as nighttime lights, population density, road density and point source emission locations to establish an efficient mapping relationship between carbon emission statistics and spatial proxy data. Compared with the traditional downscaling method based on single proxy data, the method of the present invention achieves accurate allocation of carbon emission spatial characteristics based on power, industry, transportation and civil sectors at a spatial resolution of 1km;

[0035] 3. This paper introduces the GWR model in the construction of the carbon emission inventory. By considering the influence of spatial position on the model parameters through the weighted matrix, it can dynamically capture the spatial heterogeneity of carbon emissions in different regions. Compared with the traditional global regression model, this method can better characterize the changes in carbon emission characteristics within the region and provide more refined carbon emission distribution data for areas with significant regional differences.

[0036] 4. The present invention significantly improves the model operation efficiency by calculating spatial proxy data compared with the atmospheric transmission model and the bottom-up statistical method to calculate carbon emissions. In addition, this method supports multi-scale applications and can adapt to the construction requirements of national, provincial and municipal carbon emission inventories, providing refined data support for carbon emission reduction policies at different management levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall method flow of the present invention. DETAILED DESCRIPTION

[0038] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0039] Example:

[0040] See also Figure 1 , the present invention provides one of the technical solutions:

[0041] The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model includes the following specific steps:

[0042] S-1. Data collection and processing: Collect energy consumption and spatial distribution data of the four sectors of electricity, industry, transportation, and civil use; multiply the energy consumption by the corresponding carbon emission factor to obtain the carbon emissions of each sector, and add up the carbon emissions of the four sectors to obtain the total carbon emissions of the region.

[0043] The power sector uses public global power plant location data to obtain spatial distribution data, extracts the night light values ​​corresponding to the power plant locations, and uses the extracted night light values ​​as the spatial distribution data of the power sector; the night light values ​​use the global 500-meter resolution "NPP-VIIRS-like" night light data.

[0044] The location data of the industrial sector is obtained by using NASA's VIIRS 375m active fire product. The scattered fire burning points are removed by applying the density-based noise spatial clustering algorithm DBSCAN. The algorithm clustering distance is set to 800 meters and the minimum number of clusters of point elements is set to 5.

[0045] The locations of industrial points with similar distances are preliminarily screened out, and then combined with land use data, the fire points on the artificial surface are screened again to effectively distinguish industrial heat sources from other thermal anomalies. In order to avoid repeated statistics, the industrial heat source points within the power plant are excluded to obtain the accurate location of the industrial sector; the night light values ​​corresponding to the processed industrial department locations are extracted, and the extracted night light values ​​are used as the spatial distribution data of the industrial sector.

[0046] The transportation department uses the public global road data OpenStreetMap to calculate road density data, and couples it with nighttime light data to obtain spatial distribution data;

[0047] The residential sector obtains spatial distribution data by coupling the global population density data and nighttime light data published by the LandScan database;

[0048] The coupling formula for the spatial distribution data of transportation and resident sectors is as follows:

[0049]

[0050] Among them, D i NTL represents spatial distribution data, i represents the traffic or residential sector, D represents road or population density data, NTL represents the night lighting index, and 0.34 is the default critical value of the night lighting value.

[0051] S-2. Establish a carbon emission downscaling model: Use the geographic weight regression (GWR) model to build the relationship between the total carbon emissions and the spatial proxy index, and distribute the total carbon emissions to a grid with a resolution of 1 km. The formula is as follows:

[0052]

[0053] in, represents the regression value of carbon emissions of sector i in region k in year t; subscripts i = 1, 2, 3, 4 correspond to the four sectors of electricity, industry, transportation, and civil use, respectively, and year t is the natural year;

[0054] β n (u k ,v k ) is the slope, which represents the sensitivity of carbon emissions to spatial distribution data and reflects the increase in carbon emissions caused by changes in spatial distribution data per unit; where (u k ,v k ) represents the geographical coordinates of region k; It represents the spatial distribution data of the power and industrial sectors obtained by extracting the night light index of the power and industrial point locations in the kth region of the i-th sector in the t-th year; It represents the spatial distribution data of the transportation department or civil department of the i-th department in the k-th region in the t-th year obtained by coupling the formula previously. 0.34 is the default critical value of the night lighting value.

[0055] S-3. Correction of carbon emission regression value: Adjust the regression value obtained in the regression model. The formula is:

[0056]

[0057] in, represents the carbon emission grid value of the kth region after correction, represents the high-resolution grid value of carbon emission regression of the kth region obtained by using the GWR regression model; represents the total predicted carbon emission value of the kth region obtained by using the GWR regression model, It means that the energy consumption is multiplied by the corresponding carbon emission factor to obtain the calculated carbon emission value of each department in the kth region.

[0058] This paper adopts a multi-source data fusion method to organically combine traditional statistical data with high-resolution remote sensing data, providing a new technical solution for building a high-precision carbon emission inventory. By combining the coordinated use of multiple spatial proxy data, the refined spatial allocation of carbon emissions in the power, industrial, transportation and civil sectors is achieved, filling the technical gap of insufficient spatial resolution in the production of traditional carbon emission inventories.

[0059] This paper proposes a downscaling method based on the geographically weighted regression model, combines the geographically weighted regression model with the task of downscaling carbon emission statistical data, and proposes a modeling method that considers the global carbon emission statistical characteristics and captures local spatial heterogeneity. By introducing the weighting matrix of geographically weighted regression, the present invention can dynamically adjust the model parameters, capture the local variation characteristics of carbon emissions within the region, and greatly improve the adaptability and prediction accuracy of the model.

[0060] In addition, the present invention combines NASA VIIRS fire point data with the DBSCAN clustering algorithm for the first time in point source emission identification, and combines the method of identifying industrial heat source points to extract the spatial location of industrial point sources, thereby improving the efficiency and accuracy of point source emission modeling, and to a certain extent solving the problem that the location of industrial emission point sources cannot be obtained.

[0061] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for constructing a high-resolution carbon emission inventory based on multi-source data fusion based on the GWR model, characterized in that: The specific steps include: S-1. Data collection and processing: Collect energy consumption and spatial distribution data of the four sectors of electricity, industry, transportation, and civil use; multiply energy consumption by the corresponding carbon emission factor to obtain carbon emissions of each sector, and add up the carbon emissions of the four sectors to obtain the total carbon emissions of the region; S-2. Establish a carbon emission downscaling model: Use the geographic weight regression (GWR) model to construct the relationship between the total carbon emissions and the spatial proxy index, and distribute the total carbon emissions to a grid with a resolution of 1 km. The formula is as follows: in, represents the regression value of carbon emissions of sector i in region k in year t; subscripts i = 1, 2, 3, 4 correspond to the four sectors of electricity, industry, transportation, and civil use, respectively, and year t is the natural year; β n (u k ,vk) is the slope, which represents the sensitivity of carbon emissions to spatial distribution data and reflects the carbon emission increment caused by the change of spatial distribution data per unit; k ,vk) represents the geographical coordinates of region k; It represents the spatial distribution data of the power and industrial sectors obtained by extracting the night light index of the power and industrial point locations in the kth region of the i-th sector in the t-th year; It represents the spatial distribution data of the transportation department or civil department of the i-th department in the k-th region in the t-th year obtained by coupling the formula previously. 0.34 is the default critical value of the night lighting value; S-3. Correction of carbon emission regression value: Adjust the regression value obtained in the regression model. The formula is: in, represents the carbon emission grid value of the kth region after correction, represents the high-resolution grid value of carbon emission regression of the kth region obtained by using the GWR regression model; represents the total predicted carbon emission value of the kth region obtained by using the GWR regression model, It means that the energy consumption is multiplied by the corresponding carbon emission factor to obtain the calculated carbon emission value of each department in the kth region; thus, the carbon emission inventory is obtained.

2. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 1 is characterized by: The power department uses public global power plant location data to obtain spatial distribution data, extracts the night light values ​​corresponding to the power plant locations, and uses the extracted night light values ​​as the spatial distribution data of the power department; the night light values ​​use the global 500-meter resolution "NPP-VIIRS-like" night light data.

3. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 1 is characterized by: The location data of the industrial sector is obtained by using NASA's VIIRS 375m active fire product. The density-based noise spatial clustering algorithm DBSCAN is used to remove scattered fire burning points, and the locations of industrial points with similar distances are preliminarily screened out. Then, combined with land use data, the fire points on the artificial surface are screened again to effectively distinguish industrial heat sources from other thermal anomalies. In order to avoid repeated statistics, the industrial heat source points within the power plant are excluded to obtain the accurate location of the industrial sector; the night light values ​​corresponding to the processed industrial sector locations are extracted, and the extracted night light values ​​are used as the spatial distribution data of the industrial sector.

4. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 1 is characterized by: The transportation department calculates road density data by using the public global road data OpenStreetMap data, and couples it with the night light data to obtain spatial distribution data; the resident department obtains spatial distribution data by coupling the public global population density data and night light data from the LandScan database.

5. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 4 is characterized by: The coupling formula of the spatial distribution data of the transportation and residential sectors is as follows: Among them, D i NTL represents spatial distribution data, i represents the traffic or residential sector, D represents road or population density data, NTL represents the night lighting index, and 0.34 is the default critical value of the night lighting value.

6. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 1 is characterized by: The k region is the country, province or city; the spatial distribution data is: night light index.

7. The method for constructing a high-resolution carbon emission inventory based on multi-source data fusion of the GWR model according to claim 2 is characterized by: The density-based noise spatial clustering algorithm DBSCAN was applied, where the algorithm clustering distance was set to 800 meters and the minimum number of clusters of point features was set to 5.

Images