A method for assessing pollution reduction and carbon reduction in urban agglomerations

Abstract

The present invention discloses a method for assessing pollution reduction and carbon reduction in urban agglomerations for data analysis, comprising the following steps: obtaining pollution reduction and carbon reduction intensity variables and pollution reduction and carbon reduction intensity driving variable data, standardizing the variable data to obtain standard variable data; using software to regress measured city coordinates and variable data to obtain a GTWR model, inputting the pollution reduction and carbon reduction intensity variable to obtain the spatiotemporal driving characteristics of the pollution reduction and carbon reduction intensity indicator; constructing a RF model based on the variable data, inputting the pollution reduction and carbon reduction intensity variable, and obtaining RF results; wherein the mean of N RF results is obtained as the driving factor importance change characteristic; and using the spatiotemporal driving characteristics and the driving factor importance change characteristics to assess the pollution reduction and carbon reduction intensity of urban agglomerations. The present invention adopts RF to overcome the shortcomings of traditional methods and uses GTWR to account for the spatiotemporal non-stationarity of influencing factors. This method allows for variations in the regression coefficients of different influencing factors and performs a significance test for each research unit.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a method for evaluating pollution reduction and carbon reduction in urban agglomerations. Background Art

[0002] With the continuous acceleration of urbanization and industrialization, urban development faces multiple pressures: addressing climate change, protecting the ecological environment, and achieving economic growth. CO₂ emissions from urban areas primarily stem from human activities in urban economies, urban construction, and urban transportation. Cities are also relatively concentrated sources of various pollutant emissions. Studies have shown that greenhouse gas and environmental pollutant emissions share common origins and processes. For example, the combustion of fossil fuels such as coal emits air pollutants such as particulate matter and SO₂, as well as greenhouse gases such as CO₂ and black carbon. Therefore, leveraging this characteristic to simultaneously reduce pollutant emissions (referred to as "pollution reduction") and greenhouse gas emissions (referred to as "carbon reduction") is an effective environmental management strategy. However, effective evaluation methods for applying this strategy at the city level are lacking. Exploring the spatiotemporal driving forces and evolutionary characteristics of pollution reduction and carbon reduction at the scale of urban agglomerations will help better understand the interactive relationship between urbanization and the environment and contribute to sustainable urban development.

[0003] Research shows that the spatial, demographic, and economic agglomeration characteristics of cities have a definite impact on carbon emissions, and there is an inverted U-shaped curve between urbanization and carbon emissions, and urban spatial agglomeration contributes to carbon emission reduction to a certain extent. Overall, population, GDP, and NDVI are all positive driving forces for urban CO2 emissions, but temperature and precipitation have negative effects; in terms of pollutant emission drivers, it also shows stable spatial agglomeration characteristics, and various characteristics of urbanized areas are significantly correlated with pollutant emission levels. For example, PM 2.5 There are significant links with both land use and economic and industrial structures, but the characteristics vary across different stages of economic development. While existing methods for quantifying the synergistic effects of pollution reduction and carbon reduction can provide more effective technical pathways for pollution reduction and carbon reduction, case studies analyzing the drivers of pollution reduction and carbon reduction remain limited, and analysis of the underlying mechanisms of these drivers is insufficient. Furthermore, at this stage, urban agglomeration-level pollution reduction and carbon reduction lack measurement indicators, making it difficult to support integrated regional planning, deployment, and assessment. Furthermore, while regression analysis is widely used in existing driver analysis studies, traditional multivariate linear regression cannot compare the importance of influencing factors.

[0004] Therefore, those skilled in the art urgently need to provide a new method for studying the driving factors of pollution reduction and carbon reduction in urban agglomerations to solve the problems existing in the above-mentioned existing technologies. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for evaluating pollution reduction and carbon reduction in urban agglomerations to solve the problems existing in the above-mentioned prior art.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A method for assessing pollution reduction and carbon reduction in urban agglomerations comprises the following steps:

[0008] Obtaining pollution reduction and carbon reduction intensity variables and pollution reduction and carbon reduction intensity driving variable data, and standardizing the pollution reduction and carbon reduction intensity driving variable data to obtain standard pollution reduction and carbon reduction intensity driving variable data;

[0009] Using software to regress the measured city coordinates with the pollution reduction and carbon reduction intensity driving variable data to obtain a GTWR model, input the pollution reduction and carbon reduction intensity variable, and obtain the spatiotemporal driving characteristics of the pollution reduction and carbon reduction intensity index;

[0010] Building an RF model based on the pollution reduction and carbon reduction intensity driving variable data, inputting the pollution reduction and carbon reduction intensity variable, and obtaining an RF result; wherein the average of the N RF results is the driving factor importance change characteristic;

[0011] The spatiotemporal driving characteristics and the characteristics of the change in importance of the driving factors are used to evaluate the pollution reduction and carbon reduction intensity of urban agglomerations.

[0012] Preferably, it also includes performing a multicollinearity test on the standard pollution reduction and carbon reduction intensity driving variable data.

[0013] Preferably, the pollution reduction and carbon reduction intensity variables include: representative indicators and driving indicators of various sub-items of urban pollution reduction and carbon reduction;

[0014] Among them, representative indicators of each sub-item of urban pollution reduction and carbon reduction include industrial wastewater discharge efficiency, industrial sulfur dioxide emission efficiency, industrial smoke emission efficiency, and carbon dioxide emission intensity;

[0015] The driving indicators are specific indicators in terms of economic development level, industrial structure, population, land use structure, energy consumption level and climate change.

[0016] Preferably, the GTWR model expression is:

[0017] Y i =β0(μ i ,v i ,t i )+∑ k β k (μ i ,v i ,t i )X it +ε i ;

[0018] Among them, (μ i ,vi ,t i ) is the spatial-temporal coordinate of the i-th city in the Yangtze River Delta urban agglomeration, μ i ,v i ,t i are the longitude, latitude and time of the i-th city respectively; β0(μ i ,v i ,t i ) represents the regression constant of the i-th city, that is, the constant term in the model; ε i is the residual; β k (μ i ,v i ,t i ) is the kth regression parameter of the i-th city; X it is the matrix composed of independent variables of driving factors.

[0019] Preferably, the k-th regression parameter of the i-th city is estimated using the following expression:

[0020]

[0021] in, β k (μ i ,v i ,t i ) is the estimated value; X is the matrix of independent variables; X t is the transpose of matrix X; Y is the matrix value of the pollution reduction and carbon reduction intensity measurement index of the Yangtze River Delta urban agglomeration under the time scale; W(μ i ,v i ,t i ) is the spatiotemporal weight matrix.

[0022] Preferably, the Gaussian distance function is selected and the bi-square spatial weight function is used to obtain the spatiotemporal weight matrix, which is expressed as:

[0023]

[0024] Among them, d ij The spatial and temporal distance between sample i and sample j; δ is the bandwidth.

[0025] Preferably, the step of constructing the RF model based on the pollution reduction and carbon reduction intensity driving variable data includes:

[0026] Randomly acquiring a training set and a test set based on the pollution reduction and carbon reduction intensity driving variable data;

[0027] The RF model is constructed, and is trained using the training set and tested using the test set.

[0028] Preferably, the training set for training the RF model includes:

[0029] Train multiple CARTs;

[0030] Perform traversal of the pollution reduction and carbon emission reduction intensity variables for a single CART, and determine the optimal cutting variable and cutting point according to the impurity of the nodes after cutting to obtain the result of a single tree;

[0031] Obtain the RF model by integrating all the tree results.

[0032] Preferably, the calculation expression of the node impurity is:

[0033]

[0034] Among them, x represents the cutting variable; y is the cutting value of x; N s is the number of all training samples; X left is the data set composed of y i (y i <y); X rigjt is the data set composed of y i (y i (y is the average value of X left ; is the average value of X rigjt ;

[0035] Preferably, the importance of the pollution reduction and carbon emission reduction intensity variables can be quantitatively measured through the RF model:

[0036] The expression for the RF model to estimate the pollution reduction and carbon emission reduction intensity variables through out-of-bag error samples is:

[0037]

[0038] Among them, IMp(var i ) is the importance of variable i; errOOB1 ij is the error calculated according to the out-of-bag data of variable i in CART j ; errOOB2 ij is the error calculated according to the out-of-bag data of variable i in CART j plus noise interference; n is the number of CARTs.

[0039] From the above content, it can be seen that compared with the prior art, the beneficial effects of the present invention are as follows:

[0040] GTWR introduces a time factor based on geographically weighted regression, addressing the limited sample size of cross-sectional data while also accounting for temporal and spatial nonstationarity, effectively estimating factor parameters. While this does not reduce the performance of individual CART models on their corresponding training datasets, it does reduce the correlation between constructed trees, thereby reducing the variance of the pollution and carbon reduction model after averaging multiple decision trees. The RF model's measure of variable importance, IMp, eliminates the indirect effects of other variables. Specifically, the GTWR model demonstrates good spatiotemporal fitting of pollution and carbon reduction intensity, with significant spatiotemporal effects. Furthermore, the model output reveals the driving mechanisms and spatial patterns of pollution and carbon reduction in urban agglomerations. Furthermore, based on these findings, rapid reductions in pollution and carbon reduction in urban agglomerations can be promoted by improving energy efficiency, strengthening the development of the tertiary industry, and optimizing land use structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0042] Figure 1 This is a block diagram of the method of the present invention. DETAILED DESCRIPTION

[0043] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] Example 1:

[0045] This embodiment discloses a method for assessing pollution reduction and carbon reduction in urban agglomerations. In this embodiment, the Yangtze River Delta urban agglomeration is taken as the research object, and a pollution reduction and carbon reduction intensity index is constructed based on carbon emission levels and pollutant emission data. The spatiotemporal geographic weighted regression (GTWR) method and the random forest (RF) method are used to analyze the spatiotemporal driving characteristics of economic development, industrial structure, land use structure, population, and climate change on the pollution reduction and carbon reduction intensity index, as well as the changing characteristics of the importance of driving factors.

[0046] like Figure 1 As shown, specifically:

[0047] Obtain pollution reduction and carbon reduction intensity variables and pollution reduction and carbon reduction intensity driving variable data, standardize the pollution reduction and carbon reduction intensity driving variable data, and obtain standard pollution reduction and carbon reduction intensity driving variable data; use software to regress the measured city coordinates and pollution reduction and carbon reduction intensity driving variable data to obtain the GTWR model, input the pollution reduction and carbon reduction intensity variable, and obtain the spatiotemporal driving characteristics of the pollution reduction and carbon reduction intensity index; construct an RF model based on the pollution reduction and carbon reduction intensity driving variable data, input the pollution reduction and carbon reduction intensity variable, and obtain the RF results; among them, the average of the N RF results is the driving factor importance change characteristic; use the spatiotemporal driving characteristics and driving factor importance change characteristics to evaluate the pollution reduction and carbon reduction intensity of urban agglomerations.

[0048] Among them, the pollution reduction and carbon reduction intensity variables include: representative indicators and driving indicators in various sub-items of urban pollution reduction and carbon reduction; representative indicators in various sub-items of urban pollution reduction and carbon reduction include industrial wastewater discharge efficiency, industrial sulfur dioxide emission efficiency, industrial smoke emission efficiency, and carbon dioxide emission intensity; driving indicators are specific indicators in terms of economic development level, industrial structure, population, land use structure, energy consumption level, and climate change.

[0049] In this embodiment, in order to characterize the overall effectiveness of urban pollution reduction and carbon reduction, a pollution reduction and carbon reduction intensity index (IPCR) is constructed, which incorporates the emission efficiency of multiple pollutants and greenhouse gas emissions into one index. The calculation formula is as follows:

[0050]

[0051] Among them, IPCR represents the urban pollution reduction and carbon reduction intensity index, a dimensionless unit with a value range of 0-1. The closer it is to 0, the more significant the pollution reduction and carbon reduction effect is, and it also indicates a cleaner level of economic development. i Represents the city's pollution reduction and carbon reduction sub-item indicators. The sub-item indicators are normalized to 0-1 to eliminate the influence of extreme values. At the same time, the sub-item indicators can be adjusted according to the characteristic conditions of the study area; q i Represents the weight corresponding to the sub-item indicator, the sum of the weights is 1, and the weight is determined by the expert scoring method.

[0052] Taking into account the availability of data and representative indicators in each sub-sector of pollution reduction and carbon reduction, the sub-indicators selected were industrial wastewater discharge efficiency, industrial sulfur dioxide emission efficiency, industrial smoke / powder emission efficiency, and carbon dioxide emission intensity (see Table 1). Considering the current importance of pollution reduction and carbon reduction, the weight of both is set at 0.5;

[0053] Table 1 Pollution reduction and carbon reduction indicators and data sources

[0054]

[0055] Furthermore, in this embodiment, based on ArcGIS 10.5 software, the GTWR plug-in in the prior art is adopted, and the bandwidth is set using AICc optimization to realize the study of the GTWR model for pollution reduction and carbon reduction intensity; before making the GTWR model, all pollution reduction and carbon reduction intensity variables need to be standardized; at the same time, in order to avoid the occurrence of pseudo-regression during regression, all standardized pollution reduction and carbon reduction intensity variables are tested for multicollinearity.

[0056] The GTWR model expression is:

[0057] Y i =β0(μ i ,v i ,t i )+∑ k β k (μ i ,v i ,t i )X it +ε i ;

[0058] Among them, (μ i ,v i ,t i ) is the spatial-temporal coordinate of the ith city in the Yangtze River Delta urban agglomeration, μ i ,v i ,t i are the longitude, latitude and time of the i-th city respectively; β0(μ i ,v i ,t i ) represents the regression constant of the i-th city, that is, the constant term in the model; ε i is the residual; β k (μ i ,v i ,t i ) is the kth regression parameter of the i-th city; X it is the matrix composed of independent variables of driving factors.

[0059] The k-th regression parameter of the i-th city is estimated using the following expression:

[0060]

[0061] in, β k (μ i ,v i ,t i ) is the estimated value; X is the matrix of independent variables; X t is the transpose of matrix X; Y is the matrix value of the pollution reduction and carbon reduction intensity measurement index of the Yangtze River Delta urban agglomeration under the time scale; W(μ i ,vi ,t i ) is the spatio-temporal weight matrix.

[0062] In this embodiment, the Gaussian distance function is selected, and the spatio-temporal weight matrix is obtained by using the bi-square spatial weight function. The expression is:

[0063]

[0064] where d ij is the spatio-temporal distance between sample i and sample j; δ is the bandwidth.

[0065] In addition, the steps of constructing the RF model based on the data of the pollution reduction and carbon emission reduction intensity driving variables in this embodiment include:

[0066] Randomly obtain the training set and the test set based on the data of the pollution reduction and carbon emission reduction intensity driving variables;

[0067] Construct the RF model, and train it with the training set and test it with the test set.

[0068] Preferably, training the RF model with the training set includes:

[0069] Train multiple CARTs;

[0070] Traverse the pollution reduction and carbon emission reduction intensity variables for a single CART, and determine the best cutting variable and cutting point according to the impurity of the nodes after cutting to obtain the result of a single tree;

[0071] Obtain the RF model by integrating all the tree results. It should be noted that in this embodiment, the RF regression model is implemented using the R software package "randomforest" and the R programming language; 80% of the data set is randomly selected as the training set, and the remaining 20% of the data set is selected as the test set. By setting a fixed random value in this embodiment, the RF results can be guaranteed to be reproducible, and at the same time, the mean value of 10 results is used as the characteristic of the importance change of the driving factors.

[0072] The calculation expression of the node impurity is:

[0073]

[0074] where x represents the splitting variable; y is the splitting value of x; N s is the number of all training samples; X left is the data set composed of y i (y i <y); X rigjt is the data set composed of y i (y i (y is the average value of X left ; is the average value of Xrigjt The importance of variables can be quantitatively measured through the RF model:

[0075] The expression of the RF model estimating the variable through the out-of-bag error sample is:

[0076]

[0077] Where IMp(var i ) is the importance of variable i; errOOB1 ij According to CART j Error in calculating the out-of-bag data of variable i; errOOB2 ij According to CART j The error of the calculation of the variable i is the out-of-bag data plus the noise interference; n is the number of CARTs.

[0078] The results of the scheme disclosed in this embodiment show that the variance inflation factors of the 11 indicators are all less than 10, but the inflation coefficients of GDP and garden green area are relatively high, which are 9.16 and 9.38 respectively. Table 2 summarizes the accuracy evaluation results of the regression model. It can be seen that after considering the time effect, the GTWR model R 2 With Adjusted R 2 are all higher than 0.95, which greatly improves the fitting accuracy and goodness compared with the GWR model. 2 The GTWR regression model improved by 0.1931, the residual decreased by 0.6493, and the AICc decreased by 398.24. This shows that the GTWR regression model can well fit the linear relationship between the 11 specific driving indicators and the dependent variable of pollution reduction and carbon reduction intensity in both temporal and spatial scales.

[0079] Table 2 Accuracy evaluation of pollution reduction and carbon reduction driving model

[0080]

[0081] Spatial characteristics of regression coefficients of driving factors

[0082] The spatial distribution of the GTWR local regression coefficients shows that the local regression coefficients for all 27 cities were significant at the 1% level (p < 0.01). GDP showed a significant negative correlation with pollution reduction and carbon reduction intensity in Yancheng and Jinhua; in Anqing and Chuzhou, it was most significantly affected by per capita GDP, with higher per capita GDP associated with lower pollution reduction and carbon reduction intensity. From an industrial structure perspective, Shanghai, Jiaxing, Ningbo, Shaoxing, Zhoushan, Taizhou, and Wenzhou were significantly positively driven by the proportion of secondary industry value added in GDP, while Chuzhou, Yangzhou, Nanjing, Ma'anshan, and Changzhou were more negatively driven by the proportion of tertiary industry value added in GDP. From a land use perspective, Shanghai, Suzhou, and nine cities in Zhejiang Province were negatively driven by the area of ​​gardens and green spaces, meaning that increases in their gardens and green spaces drove decreases in pollution reduction and carbon reduction intensity. Chuzhou, Hefei, Ma'anshan, Wuhu, Tongling, and Chizhou were negatively driven by the area of ​​construction land. From a population perspective, population density significantly negatively impacts pollution reduction and carbon reduction intensity in Chuzhou, Hefei, and Anqing. Integrating permanent population also negatively impacts pollution reduction and carbon reduction intensity in Anqing, Chizhou, and Wenzhou. However, permanent population positively impacts pollution reduction and carbon reduction intensity in Yancheng, Nantong, Shanghai, Suzhou, Wuxi, and Jiaxing. Total energy consumption significantly negatively impacts pollution reduction and carbon reduction intensity in Anqing, Chizhou, Tongling, and Hefei, while total energy consumption positively impacts pollution reduction and carbon reduction intensity in Shaoxing, Taizhou, Jinhua, and Wenzhou in Zhejiang Province. In terms of climate change response, annual average temperature and annual precipitation significantly negatively impact pollution reduction and carbon reduction intensity in southern cities.

[0083] Table 3 shows that the Var explanation of the RF model for the three periods of 2003-2007, 2008-2012, and 2013-2017 is 75.86%, 79.95%, and 78.61%, respectively, all exceeding 70%. This means that the model has a good fitting effect and has good explanatory power for each period. The model is also very accurate, with the RMSE for the three periods being less than 8×10 -4 .

[0084] Table 3 Accuracy evaluation of RF model for pollution reduction and carbon reduction intensity in the Yangtze River Delta urban agglomeration at different periods

[0085]

[0086] As shown in Table 4, total energy consumption ranked first in importance in all three periods (2003-2007, 2008-2012, and 2013-2017). The proportion of tertiary industry value added to GDP remained stable at second in importance during the periods 2003-2007 and 2008-2012, but declined to third in importance during the period 2013-2017. Furthermore, the area of ​​garden green space gradually rose from fifth in importance during the period 2003-2017 to second in importance during the period 2013-2017, indicating that the increasing area of ​​garden green space has a growing impact on carbon reduction and pollutant purification. The importance of construction land area has also been increasing. Per capita GDP ranked third and fourth in importance during the periods 2003-2007 and 2008-2012, respectively, but dropped to eighth in importance during the period 2013-2017. The importance of GDP has also been declining. While population density and total population rank low in importance, their overall importance is on the rise. In terms of climate change response, average annual temperature and annual precipitation consistently rank last, suggesting that climate change has little impact on pollution and carbon reduction.

[0087] Table 4 Changes in the importance of driving factors for pollution reduction and carbon intensity reduction

[0088]

[0089]

[0090] The above description of the disclosed embodiments will enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for assessing pollution reduction and carbon reduction in urban agglomerations, characterized by: The following steps are involved: Obtaining pollution reduction and carbon reduction intensity variables and pollution reduction and carbon reduction intensity driving variable data, and standardizing the pollution reduction and carbon reduction intensity driving variable data to obtain standard pollution reduction and carbon reduction intensity driving variable data; Using software to regress the measured city coordinates with the pollution reduction and carbon reduction intensity driving variable data to obtain a GTWR model, input the pollution reduction and carbon reduction intensity variable, and obtain the spatiotemporal driving characteristics of the pollution reduction and carbon reduction intensity index; Building an RF model based on the pollution reduction and carbon reduction intensity driving variable data, inputting the pollution reduction and carbon reduction intensity variable, and obtaining an RF result; wherein the average of the N RF results is the driving factor importance change characteristic; The pollution reduction and carbon reduction intensity of urban agglomerations is assessed using the spatiotemporal driving characteristics and the characteristics of changes in the importance of the driving factors; Construct the calculation formula of pollution reduction and carbon reduction intensity index IPCR: Among them, IPCR represents the urban pollution reduction and carbon reduction intensity index; E i Represents the city’s pollution reduction and carbon reduction sub-item indicators; q i Represents the weight corresponding to the sub-item indicator.

2. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 1, characterized in that: It also includes a multicollinearity test on the standard pollution reduction and carbon reduction intensity driving variable data.

3. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 1, characterized in that: The pollution reduction and carbon reduction intensity variables include: representative indicators and driving indicators of each sub-item of urban pollution reduction and carbon reduction; Among them, representative indicators of each sub-item of urban pollution reduction and carbon reduction include industrial wastewater discharge efficiency, industrial sulfur dioxide emission efficiency, industrial smoke emission efficiency, and carbon dioxide emission intensity; The driving indicators are specific indicators in terms of economic development level, industrial structure, population, land use structure, energy consumption level and climate change.

4. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 1, characterized in that: The GTWR model expression is: Y i =β0(μ i ,v i ,t i )+∑ k b k (m i ,v i ,t i )X it +e i ; Among them, (μ i ,v i ,t i ) is the spatial-temporal coordinate of the ith city in the Yangtze River Delta urban agglomeration, μ i ,v i ,t i are the longitude, latitude and time of the i-th city respectively; β0(μ i ,v i ,t i ) represents the regression constant of the i-th city, that is, the constant term in the model; ε i is the residual; β k (μ i ,v i ,t i ) is the kth regression parameter of the i-th city; X it is the matrix composed of independent variables of driving factors.

5. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 4, characterized in that: The k-th regression parameter of the i-th city is estimated using the following expression: in, β k (μ i ,v i ,t i ) is the estimated value; X is the matrix of independent variables; X t is the transpose of matrix X; Y is the matrix value of the pollution reduction and carbon reduction intensity measurement index of the Yangtze River Delta urban agglomeration under the time scale; W(μ i ,v i ,t i ) is the spatiotemporal weight matrix.

6. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 5, characterized in that: Select the Gaussian distance function and use the bi-square spatial weight function to obtain the spatiotemporal weight matrix, which is expressed as: Among them, d ij The spatial and temporal distance between sample i and sample j; δ is the bandwidth.

7. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 1, characterized in that: The steps of constructing the RF model based on the pollution reduction and carbon reduction intensity driving variable data include: Randomly acquiring a training set and a test set based on the pollution reduction and carbon reduction intensity driving variable data; The RF model is constructed, and is trained using the training set and tested using the test set.

8. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 7, characterized in that: The training set training the RF model includes: Training multiple CARTs; A single CART performs pollution reduction and carbon reduction intensity variable traversal, and determines the optimal cutting variables and cutting points according to the impurity of the nodes after cutting to obtain a single tree result; The RF model is obtained by integrating all the tree results.

9. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 8, characterized in that: The calculation expression of the node impurity is: Among them, x represents the slitting variable; y is the slitting value of x; N s is the number of all training samples; X left is the data set composed of y i (y i < y); X rigjt is the data set composed of y i (y i > y); is the average value of X left ; is the average value of X rigjt ; 10. The method for evaluating pollution reduction and carbon reduction in urban agglomerations according to claim 1, characterized in that: The importance of the pollution reduction and carbon reduction intensity variables can be quantitatively measured by the RF model: The expression of the variable estimated by the RF model through out-of-bag error samples is: Where IMp(var i ) is the importance of variable i; errOOB1 ij According to CART j Error in calculating the out-of-bag data of variable i; errOOB2 ij According to CART j The error of the calculation of the variable i is the out-of-bag data plus the noise interference; n is the number of CARTs.

Images