Whether urban elements are spatially clustered and how they are clustered are the basic prerequisites for judging the spatial layout within a city. They are also important factors affecting the spatial structure within a city. Therefore, this section depicts the spatial distribution pattern of producer services and carbon emissions then explores the spatial association between producer services and carbon emissions.
Spatial pattern of producer services in the YREB
Analysis of the spatial distribution pattern of producer services in the YREB
The closest proximity index of each industry in the producer services is calculated using ArcGIS10.8, developed by Environmental Systems Research Institute. The results are shown in Table 2. The range of the closest distance index is 0.0979–0.1862, which is less than 1. The Z-value is less than − 4055.63, which passes the test at 1% significance level. Significant agglomeration exists in all sectors of producer services.
From the clustering degree of each industry in the producer services, the wholesale and retail service industry has the highest clustering degree. The average nearest neighbor distance of the financial and insurance industry samples is 211.35 m, indicating a high average density. NNI is 0.1258, suggesting that it is highly concentrated in local areas. The YREB is one of the most active regions in China in terms of economic development. As one of the critical features of sustainable urban development, financial agglomeration can generate external scale effects and support development63. The average nearest neighbor distance of the real estate industry is 221.92 m, and NNI is 0.1857, indicating that the real estate industry is not widely but more uniformly distributed. The average nearest neighbor distance of the leasing and business service industry is 640.96 m, and NNI is 0.1759, indicating that it is widely but less uniformly distributed. The average nearest neighbor distance of the scientific research and technical service industry is 492.03 m, and NNI is 0.1848, indicating that it is widely distributed but not quite uniform. The average closest distance of the wholesale and retail service industry is 34.95 m, and NNI is 0.0979, indicating that it is not widely distributed and not uniform, which is consistent with the highest degree of aggregation of the wholesale and retail service industry. The average closest distance of transportation, storage, and postal industry is 301.21 m, and NNI is 0.1862, which indicates that it is not widely distributed but more uniform.
Analysis of spatial aggregation characteristics of producer services in the YREB
This paper uses ArcGIS 10.8 to visualize the point of interest. In this paper, the kernel density is divided into six grades by using the natural subsection point method, and the kernel density of each grade is shown in the graph according to the gradient color. The nuclear density distribution of producer services is shown in Fig. 3, where the red area indicates the high concentration area of each industry in the producer services.
Kernel density of producer service.
Financial and insurance industry. It is estimated that the peak core density for the financial and insurance industry is between 2.15 and 3.70 per square kilometer. The financial and insurance industry in the YREB presents a decentralized and combined spatial distribution pattern, which is in line with the multi-pole nuclear agglomeration pattern. The main concentration centers of the financial and insurance industry are located in ten cities, such as Shanghai and Suzhou, showing a “ten-center” spatial concentration structure. Seven small nuclei are formed in Hefei and Taizhou, etc. As the main pillar industry of the YREB, the financial and insurance industry in Shanghai, Jiangsu, and Zhejiang are more prominent. Improving the specialization level of the financial and insurance industry in the central city, can produce the agglomeration radiation effect on the surrounding cities.
Real estate industry. Its peak density is between 11.32 and 18.75 per square kilometer. The high-density agglomeration area shows a radial expansion-type aggregation pattern, spreading along the east-central-west direction. There are only four independent small polar cores formed in Shanghai, Hangzhou, Wuhan, and Chengdu. The real estate industry is one of the most concerned industries in producer services. Areas, with their development, resource advantages, and strategic planning, have a more diversified real estate industry pattern. It deserves our further attention.
Leasing and business service industry. Its peak core density of it is between 0.432 and 0.683 per square kilometer. The high-density agglomeration area shows the spatial pattern of concentrated clusters, in line with the distribution characteristics of “large clustering-small dispersal”. The clustering centers in Shanghai and Wuxi form a continuous cluster, while the clustering in Chengdu is faceted. This may be related to the small number of POI in leasing and business services. In addition, Nanjing, Zhenjiang, and Wuhan have fragmented clusters. The spatial heterogeneity of the industry is not significant.
Scientific research and technical service industry. The peak nuclear density of the scientific research and technology service industry is between 0.86 and 1.31 per square kilometer. The high-density agglomeration area shows dispersed and combined spatial distribution patterns, in line with the agglomeration pattern of multi-polar nuclear clusters. Compared with other cities, central cities’ nucleus density of scientific research and technology service industry is higher. Talent is an important guarantee to promote economic and social development and a key factor in developing advanced productive forces. To provide richer talent and technology reserves, cities develop industries such as scientific and technological consultation and information technology to serve the manufacturing industry.
Wholesale and retail service industry. The peak nuclear density of the wholesale and retail service industry is between 74.83 and 156.40 per square kilometer. It shows a belt-like aggregation pattern extending outward from the high nuclear density area in the scale of the YREB. Eighteen high-density agglomerations are formed in cities such as Shanghai and Suzhou, where wholesale and retail industries are highly concentrated and spread to peripheral cities. In addition, sub-clustering centers are formed in 11 cities, such as Nantong City and Yangzhou City, which promote the polycentric development of the wholesale and retail service industry.
Transportation, storage, and postal industry. The peak nuclear density of transportation, storage, and postal industry is between 2.44 and 4.22 per square kilometer. It shows the agglomeration pattern of multi-pole nuclear clusters. The main clusters of it are distributed in the Yangtze River Delta region. It forms a chain of clusters spreading in Shanghai, Jiangsu, and Zhejiang. In Zhejiang Province, Hangzhou, Ningbo, and four other cities are agglomeration centers. This corresponds to the reality that the development of cross-border e-commerce in Zhejiang is in the leading position. This shows that the development of the transportation, storage, and postal industry has gradually lowered the barriers to cooperation between enterprises of the Yangtze River Delta Economic Zone, and the trend of cross-regional enterprises cooperation is obvious.
In short, the polycentric agglomeration phenomenon is most obvious and widely distributed in the wholesale and retail service industry, which may be related to the number of POI of the wholesale and retail service industry. Most agglomeration and sub-glomeration centers of each industry in the producer services are located in Chinese municipalities such as Shanghai and Chongqing and provincial capitals such as Nanjing and Hangzhou. This indicates that the producer services have the location preference of large cities with obvious economic, social and human capital pointers. The reason is closely related to the new development pattern of “an axis—two wings—three levels and multiple points” in the YREB. Among them, “an axis” plays a central role in Shanghai, Wuhan, and Chongqing; “two wings” refers primarily to the extension of the radiation-driving effect from the YREB’s main axis, northern and southern, respectively. “Three poles” refer to the Yangtze River Delta, Triangle of Central China, ChengYu; “multi-point” refers to the support role of prefecture-level cities. With the development of social economy, the “point-axis” will undoubtedly develop into the “point-axis-colony”. This is a typical “point-axis system” theory. The key to achieving the rise of the YREB is to play the role of agglomeration and radiation of producer services in central cities.
Analysis of the development trend of producer services clustering in the YREB
ArcGIS 10.8 is used to analyze the development trend of producer services agglomeration. The standard deviation ellipse parameter (Table 3) and standard deviation ellipse diagram (Fig. 4) is obtained. From the shape of the standard deviation ellipse, the producer services are distributed southwest-northeast, which points to the development axis of “Chongqing-Wuhan-Shanghai”. There is little difference in the shape of the six major industries. The oval area of the standard deviation of each industry does not show the phenomenon of “secondary industry dominance”, which indicates that the simultaneous centrifugation has promoted the formation of a polycentric city structure under the radiation drive of the main axis of the Yangtze River. The X-axis of the transportation, storage, and postal industry is the longest at 871,280.42, which means that they have the most apparent distribution directions. The X-axis of the wholesale and retail service industry is the shortest at 823,074.13, with the least obvious distribution direction. The industry’s distribution is in the middle of these two. The Y-axis of the wholesale and retail service industry is the longest at 319,315.59. Despite its extensive distribution, the wholesale and retail service industry has the least obvious centripetal. Their distribution is the widest, but their centripetal is the smallest. The Y-axis of the real estate industry is the shortest at 284,046.08, indicating that its centrality of it is the most obvious. The deflection angle for the real estate industry is 76.03°, the highest of any industry. It shows that the shape of the YREB significantly impacts the agglomeration of the real estate industry. The deflection angle for the wholesale and retail service industry is 71.27°, the smallest of any industry. The deflection angles for the rest of the industry are in the middle of these two. From the perspective of development trend, the total number of spatial associations within the Yangtze River Delta urban agglomeration had increased from 108 at the beginning of the observation period to 133 at the end, with a maximum total of 650. Additionally, the overall network density has also increased from 0.166 to 0.205.
Standard deviation ellipse of producer services.
Spatial patterns of carbon emissions in the YREB
Analysis of the spatial distribution characteristics of carbon emissions in the YREB
The carbon emissions data were imported into ArcGIS10.8. The research uses the natural discontinuous point classification method to classify the carbon emissions into six levels and exports. The results are shown in Fig. 5. Concerning carbon emission intensity, the YREB has relatively high disequilibrium. There are fewer cities in the high-value areas of carbon emissions, but the value of carbon emissions is relatively high. There is only one city in the world with an emission intensity of 87.44–174.13 kg/m3: Chongqing. Five cities have carbon emission intensities between 87.44 and 174.13 kg/m3, namely Ganzhou City, Huaihua City, Zunyi City, Qujing City, and Pu’er City. Except Chongqing, all the municipalities and provincial capitals in the YREB are located in the ow-value zone. The spatial distribution of carbon emissions mainly shows a “High west–Low east” pattern. The middle and high-value areas are primarily located in the upper and middle reaches of the Yangtze River. China’s energy and mineral resources are distributed in regions with high carbon emissions, such as southern Hunan and southern Jiangxi. There are many places where coal is the primary energy source, including Guizhou, western Yunnan, and other regions. These regions should vigorously advocate using clean energy and promote energy-saving, environmentally friendly technologies in the subsequent development process to create a good low-carbon and environmental protection atmosphere in the whole society and improve the status quo of high carbon emissions.
Current status of carbon emissions.
Analysis of the spatial aggregation characteristics of carbon emissions in the YREB
The Global Morans I calculated by Geoda (https://geodacenter.github.io/download_windows.html) is 0.188. The p-value of 0.001. The Z-index test value is greater than 2.58. It shows that the YREB has significant spatial agglomeration characteristics for carbon emissions. Meanwhile, the Moran scatter plot and spatial distribution map demonstrate the correlation characteristics (Fig. 6). The first quadrant is the High-High agglomeration area, including Lincang, Dazhou, and Ganzhou, etc., indicating that these cities’ carbon emissions is high and the surrounding cities are also high. The second quadrant, Gangan, is Low–High agglomerations, indicating that there are high carbon emissions in cities surrounding these areas. The third quadrant is the Low-Low agglomeration area, including Shanghai, Nanjing, Suzhou, etc., indicating that these cities are low-value areas for carbon emissions. The fourth quadrant is the High-Low agglomeration area, including Hanggang, Fuzhou, Shangrao, etc., meaning that these cities have higher carbon emissions and are surrounded by cities with lower. The results show that the High-High concentration is in the western part of the YREB, and the Low-Low concentration is in the eastern part of the YREB, which is highly similar to the pattern of “High west–Low east”.
Spatial autocorrelation of carbon emissions.
The spatial distribution characteristics of the producer services show significant clustering. The wholesale and retail service industry has the highest degree of agglomeration. Financial and insurance industry agglomerations are concentrated in 10 central cities, such as Shanghai, Suzhou, and Wuxi. The main agglomeration centers of the real estate industry are in Shanghai, Hangzhou, Wuhan, and Chengdu. The main agglomeration centers of the leasing and business service industry are in Shanghai, Wuxi, Suzhou, and Hangzhou. The main agglomeration centers of scientific research and technology service industry are in 8 central cities, including Shanghai, Suzhou, and Nanjing. The main agglomeration centers of the wholesale and retail service industry are in Shanghai, Chongqing, and seven provincial capitals such as Nanjing and Hangzhou. The spatial distribution of carbon emissions mainly shows a “High west–Low east” pattern. The Emission intensity of carbon dioxide is not evenly distributed. Fewer cities are in high-value areas, but the value of carbon emissions is relatively high. All the municipalities and provincial capitals in the YREB, except Chongqing, are located in the low-value zone. It is noteworthy that the central city is the agglomeration center of the producer services and the low-value area of carbon emissions. This reflects that producer services and carbon emissions are spatially correlated, which sets the stage for the subsequent analysis.
Spatial correlation analysis of producer services agglomeration and carbon emissions
This paper uses a Factor detector, Interaction detector, and Risk detector to verify the impact of producer services agglomerations on carbon emissions. Y is a dependent variable that measures carbon emissions. X1 is the number of POI in the financial and insurance industry (FI). X2 is the number of POI in the real estate industry (RE). In the leasing and business service industry (LB), X3 represents the number of POI. The number of POI in the scientific research and technical service industry (ST) is X4. X5 is the number of POI in the wholesale and retail service industry (WR). X6 is the number of POI in the transportation, storage, and postal industry (TS). Since the independent variables are numerical quantities, this paper uses ArcGIS 10.8 to discretize the independent variable X into six tiers based on the natural intermittent point hierarchy. GeoDetector (http://www.geodetector.org/) was used to calculate and analyze the grading results of the six indicators. The impact of agglomerating of different industries on carbon emissions is compared based on the results of the geographic detector.
Factor detection
Utilizing a Factor Detector to determine the extent to which each producer services explains carbon emissions spatial heterogeneity. The results of the factor detector are shown in Table 4. The q-values of the factor detection are as follows: wholesale and retail service industry (0.322) > finance and insurance industry (0.243) > transportation, storage and postal industry (0.238) > scientific research and technology service industry (0.233) > leasing and business service industry (0.204) > real estate industry (0.160). The nature of the producer services is different, and so is the impact on carbon emissions. The p-values are all less than 0.01. They indicate a statistically significant association between producer services agglomeration and carbon emissions. The largest q-value of the wholesale and retail service industry indicates that its agglomeration is the most important control factor of spatial heterogeneity. The agglomeration of the financial and insurance industry ranks second, indicating that its agglomeration also has a more significant impact on the spatial heterogeneity of carbon emissions. This is because the finance and insurance industry is more specialized and risky. It is an environment-friendly industry. Improving the development and scale of the financial and insurance industry contributes to reducing carbon emissions. The agglomeration of the wholesale and retail service industry has twice as much decisive power on the spatial heterogeneity of carbon emissions as the real estate industry. The real estate industry has the smallest q value, indicating it has the least influence, explaining 16% of the spatial differentiation. The q-values of other industries are in between, indicating that they also have an essential influence on the spatial heterogeneity of carbon emissions. Based on the above findings, it can be inferred that a practical and feasible solution to achieve carbon reduction targets without excessive impact on the economic is to regulate the agglomeration of the wholesale and retail industry as well as financial and insurance industry.
Interaction detection
Exploring the influence of the two-pair synergistic agglomeration is using the interaction detector. The results are shown in Table 5. The interaction values of all combinations, such as “real estate industry-wholesale and retail service industry, leasing and business service industry-wholesale and retail service industry” are between the maximum value of two-factor q and the sum of two-factor q. They have a two-factor enhancement effect. This indicates that all interaction factors have a greater effect on the spatial differentiation of carbon emissions than any single factor. This is because the synergistic clustering of industries will increase the degree of industrial association, reducing wasted resources and decreasing transaction costs, thus improving economic and environmental efficiency. The interaction value of the “leasing and business service industry -wholesale and retail service industry” combination reaches 0.421, indicating that the “leasing and business service industry -wholesale and retail service industry” combination is the key interaction factor with the highest determination power on spatial heterogeneity. It is shown that the synergistic agglomeration of the leasing and business service industry and the wholesale and retail industry significantly affect the spatial pattern of carbon emissions.
Many factors contribute to the status quo of carbon emissions along the YREB. Factor and interaction detection show that some factors that initially have weak explanatory power for the spatial heterogeneity will produce a two-factor enhancement when spatially superimposed with other elements. The combination of factors greatly enhances its ability to explain spatial differentiation in carbon emissions.
Risk detection
The results of the risk detector are shown in Fig. 7, which presents the influence trend of each producer services on carbon emissions. The horizontal coordinate represents the discrete concentration level of each sector in the producer services. The higher the class represents the higher concentration of the producer services. The vertical coordinate represents the average of the carbon emissions levels. Carbon emissions is linked to producer services. As the agglomeration of the financial and insurance industry, leasing and business service industry, transportation, storage, and postal industry have expanded, carbon emissions has increased, fallen, and then increased. Carbon emissions reaches the minimum level when the agglomeration reaches level 5. Carbon emissions peaks at the maximum power when the agglomeration comes level 6. With the agglomeration of the real estate industry, scientific research and technical service industry, carbon emissions shows a fluctuating trend. A peak in carbon emission intensity is reached at level 6 after the intensity bottoms out at level 5. As the wholesale and retail service industry agglomerate, carbon emissions rises, falls, and then rises again. Despite this, it is always higher than the carbon emissions of the first rank. Notably, the carbon emissions is the smallest when the concentration of industries other than the wholesale and retail service industry reaches the 5th level. It indicates that the concentration of producer services significantly inhibits carbon emissions to a certain extent. As a low-carbon industry, producer services agglomeration development will inevitably lead to improve of technological innovation and labor productivity in the agglomeration area. Within a certain agglomeration range, carbon emissions will reduce as producer services agglomeration increases.
