Urban Sector Carbon Emission Estimation Methods and Systems Based on Satellite Observation and GIS

Abstract

This invention discloses a method and system for estimating carbon emissions from urban sectors based on satellite observation and GIS. The invention utilizes monthly average nighttime light data, urban functional zoning information data, point-of-interest data, and annual carbon emission data for urban functional zones such as industry, transportation, residential life, and services obtained through statistical calculations from the domestically produced professional nighttime light remote sensing satellite sensor, Luojia-1, which has higher spatial resolution. By preprocessing the satellite observation data and integrating objective data from multi-source satellite observations and geographic information systems, a high-precision gridded carbon emission database for industries such as industry, transportation, residential life, and services can be established. Combined with urban functional zoning information and a geographic weighted regression model, the spatial heterogeneity of carbon emission distribution across different urban functional zones is fully considered, resulting in a more accurate spatial representation of carbon emissions and significantly improving the estimation accuracy of CO2 emissions from urban sectors.

Description

Technical Field

[0001] This invention belongs to the field of geographic information and satellite remote sensing image information extraction and application, specifically involving a method and system for estimating carbon emissions in urban sectors based on satellite observation and GIS. Background Technology

[0002] Carbon dioxide emissions from the combustion of fossil fuels have become the most important global climate change issue. Carbon dioxide emissions primarily originate from anthropogenic industrial activities, transportation, services, residential activities, and other sources. Since a city's total carbon emissions are the sum of emissions from multiple functional zones (including industry, transportation, residential activities, and services), obtaining more detailed information on carbon dioxide emissions from different functional zones within a city can help governments formulate more effective carbon dioxide reduction policies.

[0003] In urban planning, urban functional zoning refers to the division of a city into functional areas to rationally utilize land and natural conditions and avoid mutual interference between industrial production, transportation, and residential life. Functional zoning is typically carried out based on the assessment and selection of urban land use. Generally, carbon dioxide emissions from different urban functional zones are mainly derived from statistical data. However, the lack of spatial information within the city makes it difficult to estimate carbon dioxide emissions from functional zones at high spatial resolution scales.

[0004] Since satellite-observed nighttime light data reflects the intensity of human activity to some extent, many studies have used satellite-observed nighttime light intensity (NTL) data to estimate carbon dioxide emissions, employing a "top-down" approach to spatialize carbon dioxide emissions. However, because different functional zones (industry, transportation, residential life, services, etc.) are located in different urban functional zones, carbon emissions exhibit significant spatial differences within cities. Currently, there are relatively few studies that integrate satellite observation data and point-of-interest density data to establish high-resolution carbon emission data for different sectors such as industry, transportation, residential life, and services. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method and system for estimating carbon emissions in urban sectors based on satellite observation and GIS, so as to solve the problems existing in the prior art.

[0006] To solve the above technical problems, the present invention provides the following technical solution:

[0007] This invention proposes a method for estimating carbon emissions in urban sectors based on satellite observation and GIS, comprising the following steps:

[0008] S1. Based on the Luojia-1 remote sensing satellite sensor, monthly average nighttime light intensity data is acquired and preprocessed to obtain annual average nighttime light intensity data. Then, the annual average nighttime light intensity data, the point of interest data of various sectors (including industry, transportation, residential life, and service industry) of cities across the country, and the data of functional zones (industry, transportation, residential life, and service industry) of cities across the country are resampled to a specified grid spatial resolution of 300m×300m.

[0009] S2. Based on the annual average intensity data of nighttime lights and the national urban functional zoning data from step S1, calculate the grid data of the annual average intensity of nighttime lights based on urban functional zoning.

[0010] Based on the national administrative division of prefecture-level cities, the annual average nighttime light intensity data of the nighttime light in step S1 is used to calculate the annual average nighttime light intensity data of urban functional zones at the prefecture-level city scale.

[0011] Based on the interest point data of various departments in cities across the country obtained in step S1, a fishnet grid is created, the interest point density at the grid scale of each department is calculated, and then the interest point density of different departments at the prefecture-level city scale is calculated in combination with the administrative division of prefecture-level cities across the country.

[0012] S3. Based on statistical CO2 emission data of various sectors in cities across the country, and according to the density of points of interest in different sectors at the prefecture-level city scale obtained in step S2 and the average annual intensity of nighttime light based on urban functional zones, establish a CO2 emission estimation model for urban sectors based on geographical weighted regression. By adjusting the bandwidth, determine the optimal spatial weight of the model, and then calculate the regression coefficient of CO2 emission of urban functional zones at the prefecture-level city scale across the country.

[0013] S4. For different prefecture-level cities, input the grid data of the average annual intensity of nighttime lights and the grid data of point of interest density based on urban functional zoning obtained in step S2. Calculate the initial grid CO2 emission results for each department using the regression coefficients obtained from the CO2 emission estimation model. Correct the calculated initial CO2 emission amount to obtain the final grid CO2 emission results for each department.

[0014] On the other hand, this invention also proposes a carbon emission estimation system for urban sectors based on satellite observation and GIS, comprising:

[0015] The preprocessing module is configured to perform the following actions: acquire monthly average nighttime light intensity data based on remote sensing satellite sensors and preprocess it to obtain annual average nighttime light intensity data; then resample the annual average nighttime light intensity data, the data of points of interest in various departments of cities across the country, and the data of functional zones of cities across the country to the specified grid spatial resolution.

[0016] The data calculation module is configured to perform the following actions:

[0017] Based on the annual average intensity data of nighttime light and the national urban functional zoning data, calculate the grid data of the annual average intensity of nighttime light based on urban functional zoning;

[0018] Based on the national administrative division of prefecture-level cities and the annual average intensity data of nighttime light, calculate the annual average intensity data of nighttime light in urban functional zones at the prefecture-level city scale.

[0019] Based on the data of points of interest (POIs) for various departments in cities across the country, a fishnet grid is created to calculate the density of POIs at the grid scale for each department. Then, combined with the administrative division of prefecture-level cities across the country, the density of POIs for different departments at the prefecture-level city scale is calculated.

[0020] The CO2 emission estimation model calculation module is configured to perform the following actions: Based on the statistical CO2 emission data of various sectors in cities across the country, and according to the density of points of interest in different sectors at the prefecture-level city scale obtained in step S2 and the annual average intensity data of nighttime light based on urban functional zones, a CO2 emission estimation model based on geographical weighted regression of urban sectors is established. The optimal spatial weight of the model is determined by adjusting the bandwidth, and then the regression coefficients of CO2 emissions of urban functional zones at the prefecture-level city scale across the country are calculated.

[0021] The CO2 emission grid product calculation unit is configured to perform the following actions: For different prefecture-level cities, input the grid data of the annual average intensity of nighttime lights and the grid data of point of interest density based on the urban functional zoning, use the regression coefficients obtained by the CO2 emission estimation model to calculate the initial grid CO2 emission results of each department, correct the calculated initial CO2 emission amount, and then obtain the final grid CO2 emission results of each department.

[0022] Compared with the prior art, the beneficial technical effects of the present invention in addressing the above-mentioned technical problems are as follows:

[0023] This invention utilizes monthly average nighttime light data from the Luojia-1 satellite, taken with a domestically produced satellite sensor offering higher spatial resolution, along with urban functional zoning information, point-of-interest (POI) data, and annual carbon emission data for urban functional zones obtained through statistical calculations. By preprocessing the satellite observation data and integrating objective data from multi-source satellite observations and geographic information systems, a high-precision gridded carbon emission database for functional zones can be established. Combined with urban functional zoning information and a geographically weighted regression model, it fully considers the spatial heterogeneity of carbon emission distribution across different urban functional zones, resulting in a more accurate spatial representation of carbon emissions by functional zone. This significantly improves the estimation accuracy of CO2 emissions from functional zones.

[0024] Furthermore, the domestically produced nighttime light satellite data from Luojia 1-01 offers a higher spatial resolution advantage compared to current international nighttime light data (NPP-VIIRS and DMSP-OLS satellite data). This results in high-resolution gridded carbon emission data for specific sectors (industry, transportation, residential, and service industries) that outperforms current data generated using international nighttime light data in both spatial resolution and estimation accuracy. This invention improves the accuracy of carbon dioxide emission estimation methods for functional zones, contributing to the development of policies to mitigate urban sector carbon dioxide emissions. Attached Figure Description

[0025] Figure 1 This is a flowchart of the urban sector carbon emission estimation method based on satellite observation and GIS, as described in this invention.

[0026] Figure 2 This is a block diagram of the urban sector carbon emission estimation system based on satellite observation and GIS, according to the present invention. Detailed Implementation

[0027] To better understand the technical content of the present invention, specific embodiments are described below in conjunction with the accompanying drawings.

[0028] In this invention, various aspects of the invention are described with reference to the accompanying drawings, in which numerous illustrative embodiments are shown. Embodiments of the invention are not limited to those depicted in the drawings. It should be understood that the invention is implemented through any of the various concepts and embodiments described above, as well as the concepts and embodiments described in detail below, because the concepts and embodiments disclosed herein are not limited to any particular implementation. Furthermore, some aspects of the invention disclosed may be used alone or in any suitable combination with other aspects of the invention disclosed.

[0029] like Figure 1 As shown in the flowchart of the present invention, it includes the following steps:

[0030] S1. Based on the monthly average nighttime light intensity data, nationwide urban sector (industry, transportation, residential life, and service industry) point of interest data, and urban functional zoning data (including industry, transportation, residential life, and service industry) acquired by the Luojia-1 (Luojia 1-01) remote sensing satellite sensor, data preprocessing was performed. The Luojia-1 satellite is the world's first professional nighttime illumination remote sensing satellite sensor, launched on June 2, 2018. It was jointly developed by Wuhan University and Changchun Changguang Aerospace Technology Co., Ltd. Compared with the two nighttime light products used in previous studies, DMSP / OLS (approximately 1000 meters) and Suomi-NPPVIIRS (approximately 500 meters), Luojia-1 has a finer spatial resolution (approximately 130 meters), a scanning range of 250 kilometers, and a quantization level of 14 bits. Most existing sectoral carbon emission data estimated using satellite data largely employ DMSP-OLS and Suomi-NPPVIIRS nighttime light data; therefore, the resolution of these products is usually inferior to the spatial resolution (above 500 meters) of the light data. Therefore, the newly released domestically produced remote sensing satellite sensor Luojia-1, with its superior spatial and radiometric resolution, is more suitable for establishing high-precision carbon dioxide emission grid products for different urban sectors.

[0031] As a specific embodiment of the present invention, step S1 specifically includes the following sub-steps:

[0032] S101. Perform noise reduction preprocessing on the monthly average nighttime light intensity data collected by the Luojia-1 satellite. Data with an intensity less than 1 in the monthly average nighttime light intensity data is considered noise, and data with an intensity greater than or equal to 1 are retained as valid nighttime light areas, as shown in the following formula:

[0033]

[0034] Among them, z p x represents the intensity value of nighttime lights after noise reduction for grid p. p is the original intensity value of nighttime light for grid p.

[0035] S102. Outlier correction is performed using the threshold neighborhood averaging method. To further reduce the spillover effect of some lights or the occurrence of extreme outliers, this invention uses the province as the scale and sets the maximum radiation value of the lights of the provincial capital airport or port of each province as the light threshold for that province. For areas exceeding the threshold, the threshold neighborhood averaging method is used for processing. The threshold neighborhood averaging method operator is as follows:

[0036]

[0037] S103. Average the monthly nighttime light intensity data to obtain the annual average nighttime light intensity data.

[0038] S104. In order to maintain spatial consistency, the nighttime light image data, POI (Point of Interest) data and national urban functional zoning data were resampled. Since the POI data is point data, WGS84-Albers projection was uniformly adopted based on ArcGIS 10.5 software to avoid errors, and the nearest neighbor sampling method was used to resample all data to a spatial resolution of 300m×300m.

[0039] POI data includes location data for shopping malls, restaurants, tourist attractions, hotels, and leisure and entertainment venues, provided by the online map service provider Gaode Maps. Based on functional zoning classifications for carbon emissions, the various locations of interest are categorized and merged into four main categories: industrial, transportation, urban residential, and service urban functional zones. Data on the administrative divisions of prefecture-level cities in China can be downloaded from the website of the National Geographic Information Center (http: / / www.ngcc.cn / ngcc / ). Bottom-up CO2 emission data for each functional zone (industrial, transportation, urban residential, and service) in Chinese cities is provided by the China Urban Greenhouse Gas Working Group.

[0040] S2. Based on grid scale, calculate the annual average nighttime light intensity data of grid-based functional zones such as industry, transportation, residential life, and service industry within the city; based on the administrative division of prefecture-level cities, calculate the annual average nighttime light intensity data of functional zones (industry, transportation, residential life, and service industry) across the country; based on national point of interest data, calculate the grid-based point of interest density of different sectors of industry, transportation, residential life, and service industry, and then, combined with the administrative division of prefecture-level cities across the country, calculate the point of interest density of different sectors (industry, transportation, residential life, and service industry) within each city.

[0041] As a specific embodiment of the present invention, step S1 specifically includes the following sub-steps:

[0042] S201. Based on the urban functional zoning data as a mask, the annual average nighttime light intensity data is further extracted, and the annual average nighttime light intensity data of each grid functional zone is calculated as follows:

[0043]

[0044] In the formula, a i For the information values ​​of urban functional zones (industry, transportation, residential life, and service industries) in grid s, FZ w For the value to which w belongs in the city's functional zone, SNTL s,w The nighttime light intensity of grid s belonging to functional zone w;

[0045] S202. Calculate the density of interest points in different sectors of industry, transportation, residential life, and service industries at the grid scale. Based on the created fishnet grid, the density of interest point data is calculated using the kernel density estimation method, as shown in the following formula:

[0046]

[0047] Where D(x) is the POI density index, h is the threshold, and n is the number of points within the threshold range, (xx q ) 2 +(yy q ) 2 The coordinates of the point of interest (x) q ,y q The square of the distance between the grid coordinates (x, y) within the region and the grid coordinates (x, y).

[0048] S203. Combine the administrative divisions of prefecture-level cities to calculate the density of points of interest in different sectors of industry, transportation, residential life, and service industries in all prefecture-level cities across the country.

[0049] S3. Based on the CO2 emission data of various sectors (industry, transportation, residential life, and service industry) of cities across the country obtained through statistical calculations, and according to the annual average data of point of interest density and nighttime light intensity of urban functional zones (industry, transportation, residential life, and service industry) at the prefecture-level city scale obtained in step S2, a CO2 emission estimation model based on geographically weighted regression is established. The optimal spatial weight of the model is determined by adjusting the bandwidth, and then the regression coefficients of CO2 emissions of urban functional zones at the prefecture-level city scale across the country are calculated. The geographically weighted regression model performs local regression on spatial variables, and based on different spatial bandwidths and kernel functions, combined with the generalized least squares method, the spatial heterogeneity between variables is analyzed.

[0050] The geographically regression-weighted CO2 emission estimation model is as follows:

[0051]

[0052] Among them, y i,j For the estimated carbon emissions of sector j in city i, (u i v i ) represents the geographical coordinates of city i; β represents the geographic coordinates of city i. 0,j (u i ,v i ) represents the intercept of sector j in city i, β k,j (u i v i X represents the regression parameter of the k-th explanatory variable in sector j of city i; k,i,j Let β be the value of the k-th explanatory variable for sector j in city i; 0,j (u i vi ) represents the constant term; m represents the number of explanatory variables; ε i,j This is random error.

[0053] The optimal spatial weights for the CO2 emission estimation model are determined by adjusting the bandwidth, as shown in the following formula:

[0054]

[0055] Where b represents bandwidth, w represents the distance between sample point v and regression point u. uv For weights.

[0056] S4. For different prefecture-level cities, input the annual average nighttime light intensity data and grid-scale point of interest density of urban functional zones such as industry, transportation, residential life, and service industry. Using the regression coefficients obtained from the CO2 emission estimation model, calculate the initial grid CO2 emission results for each city's industry, transportation, residential life, and service industry sectors. Correct the calculated initial CO2 emission amounts to obtain the final high-precision grid CO2 emission results for industry, transportation, residential life, and service industry sectors.

[0057] Based on the coefficients of each variable for each city calculated using a geographic regression-weighted CO2 emission estimation model, this invention will spatially visualize the coefficients of each variable for all prefecture-level cities using the ArcGIS software platform.

[0058] Based on NTL data of each grid (resampled to 300m×300m), and the density of Points of Interest (POIs) in each functional zone (industry, transportation, urban residents, and services) of each grid, combined with the coefficients of each variable in each city calculated by a geographically weighted regression-based CO2 emission estimation model, the initial CO2 emissions of different sectors in each grid within each city are calculated.

[0059] GY i，g，j =β 0，j (u i v i ) / f i，j +β 1，j (u i v i )X 1，i，g，j +β 2，j (u i v i )X 2，i，g，j +ε i，j / f i，j ;

[0060] Among them, GY i，g，jThis represents the predicted carbon emissions of sector j for a specific grid cell g within city i, (u i v i ) represents the geographical coordinates of city i; β represents the geographic coordinates of city i. 0，j (u i v i ) represents the intercept of sector j in city i, β 1，j (u i v i ) and β 2，j (u i v i X represents the regression parameters for two explanatory variables (nighttime light data and point of interest density) in sector j of city i; 1，i，g，j and X 2，i，g，j For a specific grid cell g in city sector i, the values ​​of two explanatory variables are given (nighttime light data and point of interest density); β 0，j (u i v i ) represents a constant term; f i，j ε represents the number of specific grid cells within department j of city i; ij This is random error.

[0061] For a specific city, the sum of the estimated sectoral carbon emissions of each grid within the city's administrative boundaries, based on the results of a geographically weighted regression model, may not be exactly equal to the original statistically calculated urban sectoral emissions. Although the emission differences are usually small, it is necessary to further adjust the sectoral carbon emissions of each grid so that the total gridded sectoral carbon emissions within any city boundary are completely consistent with the total sectoral carbon emissions value calculated by the city.

[0062] The initial CO2 emissions from various sectors within each city's grid, including industry, transportation, residential life, and service industries, are corrected as follows:

[0063] GE i，g，j =GY i，g，j ×RY i，j / CY i，j ;

[0064] Among them, GE i,g,j GY represents the carbon emissions of the final corrected sector (industry, transportation, residential, and service industries) j within a specific grid cell g in city i. i,g,j RY represents the predicted carbon emissions of sector j for a specific grid cell g within city i. i,j CY represents the projected total carbon emissions of sector j in city i. i,j This represents the emissions calculated by department j of city i.

[0065] This invention extracts nighttime light intensity data from different functional zones within a city, improving the accuracy of CO2 emission data reflecting the true status of these zones. Furthermore, unlike many other studies that only use nighttime light data for downscaling and grid allocation of CO2 emissions across different functional zones, this invention introduces POI density values, directly correlated with high spatial resolution grid scales, as another crucial factor in estimating and downscaling CO2 emissions across sectors (industry, transportation, residential, and service industries). In terms of the estimation model, this invention uses a geographically weighted regression model to regress spatial variables. This model improves upon traditional regression methods, revealing the spatial heterogeneity of different variables across different cities. Compared to the spatial non-stationarity issues that traditional global regression cannot address, the geographically weighted regression model can calculate variable coefficients specific to different cities, significantly improving the accuracy of CO2 emission estimation for functional zones.

[0066] like Figure 2 As shown, this invention also proposes a city sector carbon emission estimation system based on satellite observation and GIS, comprising: a preprocessing module and a data calculation module, including: an annual average nighttime light intensity data calculation unit for prefecture-level cities based on urban functional zones (industry, transportation, residential life, and service industries); an annual average nighttime light intensity data calculation unit for gridded nighttime light intensity based on urban functional zones (industry, transportation, residential life, and service industries); a sector (industry, transportation, residential life, and service industries) point of interest density calculation unit; a sector CO2 emission estimation model calculation unit; and a high-precision sector CO2 emission grid product calculation unit. Wherein:

[0067] The preprocessing module is configured to perform the following actions: preprocessing data based on monthly average nighttime light intensity data acquired by satellite, point of interest data for various urban departments, urban functional zoning data, and urban departmental CO2 emission statistics.

[0068] The data calculation module includes:

[0069] The grid-based nighttime light intensity annual average data calculation unit, based on urban functional zones (industrial, transportation, residential, and service industries), is configured to perform the following actions: calculate the annual average nighttime light intensity data at the grid scale within different urban functional zones;

[0070] The unit for calculating the annual average nighttime light intensity data of prefecture-level cities based on urban functional zones (industrial, transportation, residential, and service industries) is configured to perform the following actions: calculate the annual average nighttime light intensity data of each city within different urban functional zones based on the administrative divisions of prefecture-level cities.

[0071] The interest point density calculation unit for each sector (industry, transportation, residential life, and service industry) is configured to perform the following actions: Based on national interest point data, create a fishnet grid, calculate the interest point density at different grid scales for industry, transportation, residential life, and service industry, and then, in conjunction with the national prefecture-level city administrative divisions, calculate the interest point density for different sectors (industry, transportation, residential life, and service industry) within each city.

[0072] The CO2 emission estimation model calculation unit for the sector (industry, transportation, residential life, and service industry) is configured to perform the following actions: Based on the statistical CO2 emission data of various sectors (industry, transportation, residential life, and service industry) in cities across the country, and based on the annual average data of point of interest density and nighttime light intensity of urban functional zones (industry, transportation, residential life, and service industry) at the prefecture-level city scale obtained in step S2, a CO2 emission estimation model based on geographical weighted regression is established. The optimal spatial weight of the model is determined by adjusting the bandwidth, and then the regression coefficients of CO2 emissions of urban functional zones at the prefecture-level city scale are calculated.

[0073] The high-precision CO2 emission grid product calculation unit for sectors (industry, transportation, residential life, and service industry) is configured to perform the following actions: For different prefecture-level cities, it inputs the annual average nighttime light intensity data and the grid-scale point of interest density of urban functional zones such as industry, transportation, residential life, and service industry, and uses the regression coefficients obtained from the CO2 emission estimation model to calculate the initial grid CO2 emission results for each city's industry, transportation, residential life, and service industry sectors. It then corrects the calculated initial CO2 emission amounts to obtain the final high-precision grid CO2 emission results for industry, transportation, residential life, and service industry sectors.

[0074] While the present invention has been described above with reference to preferred embodiments, it is not intended to limit the invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for estimating carbon emissions in urban sectors based on satellite observation and GIS, characterized in that, Includes the following steps: S1. Acquire monthly average nighttime light intensity data based on remote sensing satellite sensors and preprocess the data to obtain annual average nighttime light intensity data; then resample the annual average nighttime light intensity data, the data of points of interest in various departments of cities across the country, and the data of functional zones of cities across the country to the specified grid spatial resolution. S2. Based on the annual average intensity data of nighttime lights and the national urban functional zoning data from step S1, calculate the grid data of the annual average intensity of nighttime lights based on urban functional zoning constraints. Based on the national administrative division of prefecture-level cities, the annual average nighttime light intensity data of the nighttime light in step S1 is used to calculate the annual average nighttime light intensity data of urban functional zones at the prefecture-level city scale. Based on the interest point data of various departments in cities across the country obtained in step S1, a fishnet grid is created, the interest point density at the grid scale of each department is calculated, and then the interest point density of different departments at the prefecture-level city scale is calculated in combination with the administrative division of prefecture-level cities across the country. S3. Based on the statistical data of carbon dioxide emissions from various sectors of cities across the country, and according to the density of points of interest in different sectors at the prefecture-level city scale obtained in step S2 and the average annual intensity of nighttime light based on urban functional zones, establish a carbon dioxide emission estimation model for urban sectors based on geographical weighted regression. By adjusting the bandwidth, determine the optimal spatial weight of the model, and then calculate the regression coefficients of carbon dioxide emissions from urban functional zones at the prefecture-level city scale across the country. S4. For different prefecture-level cities, input the grid data of the average annual intensity of nighttime lights and the grid data of the density of points of interest based on urban functional zoning obtained in step S2, use the regression coefficients obtained by the carbon dioxide emission estimation model to calculate the initial grid carbon dioxide emission results of each department, correct the calculated initial carbon dioxide emission, and then obtain the final grid carbon dioxide emission results of each department. The aforementioned departments include industry, transportation, residential life, and service sectors; the aforementioned urban functional zones include industrial, transportation, residential life, and service functional zones. The grid spatial resolution is 300m × 300m.

2. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 1, characterized in that, In step S1, the annual average data of nighttime light intensity is obtained as follows: S101. Perform noise reduction preprocessing on the monthly average nighttime light intensity data collected by satellite. Pixel values ​​less than 1 in the monthly average nighttime light intensity data are considered noise, and pixel values ​​greater than or equal to 1 are retained as valid nighttime light areas, as shown in the following formula: in, For grid p The intensity value of nighttime lights after noise reduction. For grid p The original intensity value of nighttime lights; S102. Outlier correction is performed using the threshold neighborhood averaging method, wherein the threshold neighborhood averaging operator is as follows: S103. Average the monthly nighttime light intensity data to obtain the annual average nighttime light intensity data; S104. Resample the nighttime light image data, POI data, and national urban functional zoning data. Based on ArcGIS 10.5 software, uniformly adopt WGS84-Albers projection and use the nearest neighbor sampling method to resample all data to the specified spatial resolution.

3. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 1, characterized in that, In step S1, the remote sensing satellite sensor is the Luojia-1 satellite.

4. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 2, characterized in that, Step S2 specifically includes the following sub-steps: S201. Based on the urban functional zoning data as a mask, the annual average nighttime light intensity data is further extracted, and the annual average nighttime light intensity data of each grid functional zone is calculated as follows: In the formula, For grid s The city functional zoning information value, Urban functional zoning w The value to which it belongs For grid s Nighttime light intensity belonging to functional zone w; S202. Calculate the density of interest points in different sectors of industry, transportation, residential life, and service industries at the grid scale. Based on the created fishnet grid, the density of interest point data is calculated using the kernel density estimation method, as shown in the following formula: in, For POI density index, h For the threshold, n The number of points within the threshold range. Coordinates of the point of interest and grid coordinates within the region The square of the distance between them; S203. Combine the administrative divisions of prefecture-level cities to calculate the density of points of interest for different departments in all prefecture-level cities across the country.

5. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 4, characterized in that, In step S3, the sectoral carbon dioxide emission estimation model based on geographic regression weighting is as follows: in, For the city i department j The estimated carbon emissions, ( u i , v i (is a city) i Geographic coordinates; For the city i department j The intercept, β k,j ( u i , v i (for the city) i department j The k Regression parameters for each explanatory variable; For the city i department j The k One explanatory variable value; β 0,j ( u i , v i ) is a constant term; m The number of explanatory variables; ε i,j This is random error.

6. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 5, characterized in that, In step S3, the optimal spatial weights of the model are determined by adjusting the bandwidth, as shown in the following formula: in, b Indicates bandwidth. Represents sample points v Regression point u The distance between them w uv As weight.

7. The urban sector carbon emission estimation method based on satellite observation and GIS according to claim 6, characterized in that, Step S4 includes the following sub-steps: S4.1 Calculate the initial carbon dioxide emissions of each grid section within each city, as shown in the following formula: ； in, Represents city i Internal Grid Cells g Forecasting Department j Carbon emissions, ( u i , v i (is a city) i Geographic coordinates; For the city i department j The intercept, and Cities i department j Regression parameters for nighttime light data and point-of-interest density; and Cities i department j Specific grid cells g Nighttime light data and point-of-interest density variable values; β 0,j ( u i , v i ) is a constant term; For the city i Internal Department j The number of specific grid cells; ε i,j This is random error; S4.2 Correct the initial carbon dioxide emissions of each grid unit within each city, as shown in the following formula: ； in, Represents city i Internal Grid Cells g The final calibration department j Carbon emissions Represents city i Internal Grid Cells g Forecasting Department j Carbon emissions Indicates the city i department j The predicted total carbon emissions, Indicates the city i department j Emissions calculated statistically.

8. A carbon emission estimation system for urban sectors based on satellite observation and GIS, characterized in that, include: The preprocessing module is configured to perform the following actions: acquire monthly average nighttime light intensity data based on satellite remote sensing sensors and perform preprocessing to calculate the annual average nighttime light intensity data; Then, the annual average intensity data of nighttime lights, the data of points of interest in various departments of cities across the country, and the data of functional zones of cities across the country are resampled to the specified grid spatial resolution. The data calculation module is configured to perform the following actions: Based on the annual average intensity data of nighttime light and the national urban functional zoning data, calculate the grid data of the annual average intensity of nighttime light based on urban functional zoning; Based on the national administrative division of prefecture-level cities and the annual average intensity data of nighttime light, calculate the annual average intensity data of nighttime light in urban functional zones at the prefecture-level city scale. Based on the data of points of interest (POIs) for various departments in cities across the country, a fishnet grid is created to calculate the density of POIs at the grid scale for each department. Then, combined with the administrative division of prefecture-level cities across the country, the density of POIs for different departments at the prefecture-level city scale is calculated. The carbon dioxide emission estimation model calculation module is configured to perform the following actions: Based on the statistical carbon dioxide emission data of various sectors in cities across the country, according to the density of points of interest in different sectors at the prefecture-level city scale obtained in step S2 and the annual average intensity data of nighttime light based on urban functional zones, a carbon dioxide emission estimation model based on geographical weighted regression of urban sectors is established. The optimal spatial weight of the model is determined by adjusting the bandwidth, and then the regression coefficients of carbon dioxide emissions of urban functional zones at the prefecture-level city scale across the country are calculated. The carbon dioxide emission grid product calculation unit is configured to perform the following actions: For different prefecture-level cities, input the grid data of the annual average intensity of nighttime lights and the grid data of the density of points of interest based on the urban functional zoning, use the regression coefficients obtained by the carbon dioxide emission estimation model to calculate the initial grid carbon dioxide emission results of each department, perform overall consistency correction on the calculated initial carbon dioxide emission, and then obtain the final grid carbon dioxide emission results of each department. The aforementioned departments include industry, transportation, residential life, and service sectors; the aforementioned urban functional zones include industrial, transportation, residential life, and service functional zones. The grid spatial resolution is 300m × 300m.

Images