Spatial-Temporal Evolution of the Coupling Coordination Relationship between Urbanization and Atmospheric Environment in the Yangtze River Economic Belt, China

Based on the coupling coordination degree model and the exploratory spatial data analysis method, we established the coupling coordination relationship between urbanization and atmospheric environment and explored the spatial-temporal evolution characteristics of the coupling coordination degree of 11 provinces in the Yangtze River Economic Belt (YEB) from 2003 to 2017. The results indicated the following: (1) The level of urbanization increases linearly, and the atmospheric environment level shows a uctuating upward trend. (2) The types of coordination gradually change from “Seriously uncoordinated development with urbanization lag” to “Superiorly coordinated development with atmospheric environment lag”. The spatial distribution of coordination shows the characteristics of “high in the eastern region and low in the central and western region”. Regarding temporal evolution, the coupling coordination degree of the region gradually increases, and the spatial differences between provinces gradually narrow. (3) Signicant spatial autocorrelation is observed between the coordination of urbanization and atmospheric environment, which weakens over time. The local agglomeration mode shows that the “High-High” cluster areas are in the lower reaches and the “Low-Low” cluster areas are mainly in the middle or upper reaches. This study contributes to promoting the sustainable development of the Yangtze River Economic Belt and provides basic data and research perspectives for further investigation of the relationship between urbanization and atmospheric environment.


Introduction
Since the second half of the 20th century, urbanization has been rapidly increasing around the world (Gupta, 2002). After the reforms and opening-up of the economy, China's urbanization rate increased from 17.9-63.89% in 1978-2020 (Lin and Zhu, 2018). According to the theory by Northam, China is currently in a phase of accelerated development urbanization (Northam, 1975). Urbanization in China has resulted in improved socioeconomic levels. However, urbanization inevitably causes negative impacts on the environment, especially the impacts on the atmosphere resulting from the coal industry. Regional climate change and local atmospheric environmental pollution have become very prominent. Haze and acid rain occur frequently in China and have an impact on climate change and human health (Yuan et al., 2010).
Since the environmental Kuznets curve (EKC), a model simulating the relationship between urban economic growth and environmental quality, was initially proposed (Grossman and Krueger, 1995), various empirical studies have established a nonlinear relationship by employing it (Diao et al., 2009;Dogan and Turkekul, 2016;Ulucak and Bilgili, 2018). However, EKC also has some limitations. For example, it assumes the independence of the economic and environmental systems and ignores their interactions. Scholars have applied some new theories and methods to explore the relationship between urbanization and the environment. For instance, an improved entropy model was utilized to evaluate the impact of urbanization on the atmospheric environment (Wang et al., 2012), and a gravity model was used to analyze the spatial correlation of pollutants and CO 2 emissions in Henan (Chen et al., 2018).
Principal component analysis and regression analysis were applied to study the relationship between the urban population, economy, space, and environment interactions (Zhang et al., 2008). A nonparametric Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js kernel density model and spatial econometric analysis method were used to explore the dynamic evolution and spatiotemporal process of regional green competitiveness systems (Cheng et al., 2019), while Zhang et al. explored the dynamics of carbon emissions via spatial-temporal analysis (Zhang et al., 2020). Originating in physics, coupling represents the interactions of two or more systems (Li et al., 2012). In recent years, coupling has been widely used to investigate the nonlinear relationship between urbanization and the ecological environment ( Liao, 1999;Wang et al., 2014;Xing et al., 2019).
The atmospheric environment is an important part of the human living environment and can be regarded as one of the contributing factors suppressing the sustainability of human society. Studies have shown that there are spatial spillover effects between urbanization development and atmospheric environment problems (Han and Yu, 2016;Meng et al., 2017), which means that the development of a city will drive the surrounding cities and the air pollution will also affect neighborhoods. Most current research pays close attention to the change in the coupling relationship of a single city in the time series, while few studies focus on the spatial pattern of the regional area (Ding et al., 2015;Jiang et al., 2019). Hence, it is essential to explore the spatial interaction of urbanization-atmospheric environment systems among different areas from the perspective of a large watershed. This study examined the Yangtze River Economic Belt (YEB), collected o cial data from 2003 to 2017 and established an index system of urbanization and atmospheric environment. Then, the entropy method was utilized to quantitatively describe the development level of urbanization and atmospheric environment. Furthermore, a coupling coordination degree model was used to explore the relationship between urbanization and atmospheric environment.
Finally, exploratory spatial data analysis was introduced to analyze the spatiotemporal evolution process of the coupling coordination degree at regional and provincial scales. This study aims to promote the sustainable development of the YEB; moreover, it provides basic data and research perspectives for further work on urbanization and atmospheric environments in other areas facing similar situations.

Study Area
The YEB (97°20′~123°30′E, 21°30′~35°20′N) is a major national strategic development area of China. Its administrative area is approximately 2.05 million square kilometers, accounting for 21.38% of the land of China. The YEB consists of three regions: upper, middle and lower reaches, including Sichuan, Yunnan, Guizhou and Chongqing; Hubei, Hunan, Jiangxi and Anhui; Jiangsu, Zhejiang and Shanghai, respectively ( Fig. 1). As China's largest and most dense river basin economic zone, the regional GDP is approximately CNY 45.16 trillion, accounting for 46.4% of the country's total GDP in 2020. The average urbanization rate was 61.7% in 2019, higher than the national average (60.6%). However, long-term extensive economic development increases atmospheric pollutants, far beyond the environmental carrying capacity, and the deterioration of the atmospheric environment will in turn restrict the development of urbanization in the YEB ( Taking into account the availability and continuity of research data, combined with the speci c circumstances of the provinces of the YEB, we selected 2003-2017 as the research interval. The data were derived from the China Statistical Yearbook, China Urban Statistical Yearbook, China Regional Statistical Yearbook, China Environmental Yearbook and China Environmental Statistics Yearbook from 2002 to 2018, as well as the Government Work Report and Statistical Bulletin of National Economy and Social Development of each province. The interpolation method of adjacent years was adopted to supplement the missing individual data. The extreme-range method (Formulas (1)-(2)) was used to standardize the original data. Depending on the promoting or inhibiting effect, the indicators were divided into positive and negative indicators: Positive indicator: Urbanization and atmospheric environment are two independent, complex and interactive dynamic systems. In this paper, the weight value was determined by the entropy method (Zou et al., 2006), and the comprehensive level of urbanization and atmospheric environment were calculated by weighted summation according to the weight and standardized value of each index. Since the range method will produce a value of 0 when the index data are standardized and a logarithm is needed in the calculation of the entropy method, it cannot be directly used. Previous research (Han and Ma, 2013) found that for the extreme value 0, given a small change of 10 − 5 or less, the calculated weight is almost unchanged, so we decided to give the extreme value a small change of 10 − 5 .
The steps were as follows (Formulas (3)- (7)): Proportion of the indicator j in year i:  Table 2.

Exploratory spatial data analysis
Exploratory spatial data analysis (EDSA) is a collection of various spatial data processing methods and technologies, with spatial autocorrelation analysis as the main function (Anselin, 1996). Spatial autocorrelation refers to the spatial dependence of a variable with spatial attributes in different geographical locations, including global autocorrelation and local autocorrelation. Moran's I, with statistical superiority, is widely used in spatial autocorrelation analysis (Getis, 2007).
The calculation formula of global Moran's I is as follows (Cliff and Ord,1981): The calculation formula of local Moran's I is as follows (Anselin, 1995):   Combined with the variation tendency and the evolution process of type, we divided the coupling coordination process of urbanization and atmospheric environment into four stages: (1) 2003-2006: The coupling coordination degree presents a linear growth trend (y = 0.1624x-325.34, R 2 = 0.9959), and the coupling coordination type changes from "Seriously uncoordinated development with lagging urbanization" to "Barely coordinated development with lagging urbanization". During this period, the growth rate of the urbanization level index is greater than that of the atmospheric environment, but its value is smaller. Urbanization development in the YEB is in its infancy, which causes some interference with the atmospheric environment, though it is not serious. A radar map was drawn to re ect the spatiotemporal evolution of the coupling coordination degree in the YEB more intuitively from a global view (Fig. 6).   During the investigated period, most provinces are included in the H-H and L-L zones, indicating that the spatially positive autocorrelation of the coupling coordination degree in the YEB is remarkable.
Comparing the results over representative years, most provinces do not make a quadrant transition, presenting a certain spatial stability.
(2) The LISA diagram represents the spatial connection mode of the coupling coordination degree between adjacent regions in geographical space and can further test the signi cance level of spatial agglomeration. As shown in Fig. 8, the agglomeration mode of the provinces of the YEB in the representative years has the following characteristics: In 2003, the "High-High" cluster mode includes 3 provinces (i.e., Jiangsu, Shanghai, and Zhejiang), and Yunnan Province is in the "Low-Low" cluster mode. The rest of the provinces in the YEB do not have signi cant spatial agglomeration characteristics. In 2006, the provinces in the "High-High" and "Low-Low" cluster modes remain unchanged, and Anhui Province transforms into the "Low-High" outlier mode. The LISA diagram of the spatial agglomeration mode in 2009 is the same as that in 2006. In 2012, the "High-High" cluster and "Low-High" outlier modes remain constant. The provinces of the "Low-Low" cluster mode are increased to 2, with Guizhou Province being added. In 2014, Jiangsu and Shanghai provinces are included in the "High-High" cluster, Anhui Province is included in the "Low-High" outlier mode, and Sichuan Province transforms from the "Low-Low" cluster mode to the "High-Low outlier mode. In 2017, there are only two agglomeration modes on the diagram: "High-High" cluster mode with Jiangsu and Shanghai provinces and "Low-High" outlier mode with Anhui Province.
Based on the above analysis, the "High-High" cluster appears in the provinces of the lower regions, showing the characteristics of a high coupling coordination degree and a signi cant spatial aggregation effect. These provinces have a good social and economic foundation in promoting urbanization and atmospheric environment governance. "Low-Low" cluster provinces are mainly in the middle and upper regions of the YEB. Lagging urbanization development and inadequate atmospheric environmental protection affected by geographical location and economic structure results in a state of low coupling coordination. The existence of "High-High" and "Low-Low" cluster provinces re ects the spatial dependence of the coupling coordination degree between urbanization and atmospheric environment. In various years, some provinces showing "Low-High" and "High-Low" outlier modes re ect the spatial heterogeneity of the coordination between urbanization and atmospheric environment. With the evolution of time, fewer provinces show signi cant agglomeration, and the "Low-Low" cluster areas and "High-Low" outlier areas gradually disappear. This phenomenon shows that the overall coupling coordination degree of the YEB is on the rise, and the differences and imbalance between regions are reducing. diagram depicts, "High-High" cluster areas are in the upper regions of the YEB, and "Low-Low" cluster areas are mainly in the middle and lower regions. With evolution time, there are fewer provinces showing signi cant agglomeration, and the "Low-Low" cluster areas and "High-Low" outlier areas gradually disappear. This phenomenon shows that the overall coupling coordination degree of the YEB maintains an upward trend, and the difference and imbalance between regions have been reducing.
As a major strategic development region in China, the overall level of urbanization in the YEB has developed relatively fast in recent decades, while the development level of the atmospheric environment is relatively backward. This competing trend makes it necessary to increase investment in atmospheric environmental governance and enhance people's awareness of energy conservation and emission reduction. We recommend that the Chinese government designate the main towns (i.e., Shanghai, Wuhan and Chongqing) along the Yangtze River as the nodes and build a green development axis to encourage the gradient development of the economy from the coast to upstream and realize the coordinated development of the YEB. In view of the "spillover effect" among provinces, the local government could enhance the industrial interaction among regions and promote exchanges for air pollution prevention and control technologies, thereby providing complementary advantages to achieve a continuous optimization of the overall coupling coordination degree in the YEB.
This paper has important practical and application value for promoting new urbanization construction and improving the atmospheric environment in the YEB. Nevertheless, some issues remain to be further studied: 1) in the future, a variety of methods besides entropy method can be used to improve the process of index construction; 2) in addition to the provincial spatial scale, smaller scales, such as the prefecture level, should be considered for a more accurate analysis; 3) Except in addition to the rook adjacency matrix, distance or other space weight de nition standards should be used to nd an optimal one.

Declarations
Author contributions Yuxia Deng wrote the paper, analyzed the data and made gures. Ting Huang conceptualized, analyzed the data and revised the paper.
Xianglian Wang and Ya Liu calculated the results. Lian Zeng and Xiangwen Zhang collected data. Daishe Wu organized resources. All the authors read and contributed to the submitted version of the manuscript.

Funding
This work was nancially supported by the National Natural Science Foundation of China (41402312) and the Natural Science Foundation of Jiangxi Provincial (2018BAB213017).

Data Availability
All relevant data are within the manuscript and available from the corresponding author upon request.  Figure 1