Mixed Land Use and Its Relationship with CO₂ Emissions: A Comparative Analysis Based on Several Typical Development Zones in Shanghai

Abstract

While development zones are the main locations of the urban industrial economy and sources of energy consumption, studies on the interactive relationship between mixed land use and CO₂ emissions in these zones have not attracted much attention. In this paper, the development zone is selected as the research object, and a honeycomb grid with a side length of 50 meters is used as the unit to measure multiple dimensions of mixed land use. The efficiency and intensity of CO₂ emissions are estimated for these units, and the coupling model is used to analyse the interactive relationship between these two factors. The results show that (1) the mixed land use degree of different types of development zones differs; the bonded zone has the highest degree, while the economic and technological development zone and the high-tech development zone have higher degrees than the industrial development zones. (2) The CO₂ emissions capacities of economic and technological development zones and high-tech development zones are better than those of industrial development zones and bonded areas. (3) There is a strong interaction between the mixing degree of land use and the intensity of CO₂ emissions; the relationship between the two may be diverse and complex in different development zones. (4) The coordinated development level between mixed land use and CO₂ emissions in economic and technological development zones and bonded areas is better than that in high-tech development zones, which in turn is better than that in industrial development zones. Based on existing measurements of mixed land use, the index of land use intensity is introduced in this paper and the land use classification in development zones is refined to supplement the land use compatibility matrix. The results of this study have implications for the development zone to enhance mixed land use and low-carbon development.

1. Introduction

In the middle and late 20th century, the phenomenon of urban sprawl intensified in the west. Led by the idea of functional zoning, single land use was common, and residents mostly relied on cars to travel, which led to energy waste and urban centre decay. As a way to improve the convenience of residents’ daily travel, mixed land use has gradually received attention. It is believed that urban mixed-use development can improve the efficiency of land use and the quality of life for residents. Therefore, a combination of different functions (such as residence, business, office, entertainment, etc.) is advocated in urban planning to reduce commuting demands and improve community accessibility and convenience [1,2]. The rise of urban renewal and sustainable development movements in the United States and European countries has led mixed land use to be recognized as an important tool for enhancing urban vitality and achieving urban sustainability. The definition, measurement indicators, methods, and impact of mixed land use have become the focus of academic research.

Most scholars believe that the core idea of mixed land use is to emphasize the integrated development of multiple functions in cities to meet diverse human development needs [3,4,5]. Some scholars also note that the core concept of compact cities is to promote moderate mixed land use [4,6,7]. Furthermore, mixed land use has been proposed as an important means of implementing low-carbon city strategies [8]. The development of high-density land use along transportation corridors is one of the basic characteristics of intensive cities [9]. Therefore, compact cities, low-carbon cities, and intensive cities are closely associated with the idea of mixed land use.

The degree of land use mixing is one of several important indices for measuring regional land use development patterns and land use structures. Previous measurements mostly focused on the diversity of land use types and used single indicators such as Shannon’s entropy to calculate the degree of land use mixing [10,11,12]. Although the quantity or proportion of land types forms the basis of the degree of land use mixing, a reliance solely on land use diversity as a measurement indicator has limitations in understanding the connotation of the degree of land use mixing and fails to achieve spatial optimization objectives. Talen [13] noted that the study of land use mixing degree at the neighbourhood scale should consider the adjacent locations and mutual influence of different land uses. The interactions between different land use types should also be considered [14,15]. Based on these considerations, the concept of compatibility has been proposed; compatibility is the basis of the degree of land use mixing, and mixing incompatible land types can cause spatial and behavioural chaos. Therefore, the compatibility between land use types is included in the calculation of the land mixed use index [16,17].

Previous studies have mainly focused on the relationship between mixed land use and travel behaviour, suggesting that mixed land use patterns can help reduce travel distances, thus decreasing average travel time and reliance on private cars. Mixed land use is also beneficial for developing a diversified urban transportation system that accommodates multiple modes of transportation [18,19]. Therefore, some scholars used “accessibility” to reflect the proximity degree of different land use types and then measured the mixing degree of land use [20]. Generally, in areas with a high mixing degree of land use, the multiple land types required by residents’ daily lives are generally close together, so the degree of land use mixing can be represented by assessing the walking environment [18,21]. Conversely, pedestrian flows can also be estimated based on the mixing degree of land use [22]. In recent years, the relationships between mixed land use and public health, urban crime [23], urban housing prices and rents [24] have gradually been favoured by scholars.

Traditional mixed land-use research focuses on either a single spatial scale or a single time scale. However, in terms of space, mixed land use is multi-scale, including plot or block scale, subdistrict scale and city scale, and any single spatial scale measurement is not comprehensive. In terms of time, mixed land use is dynamically changing, and any single time point measurement may ignore the stage characteristics of its evolution. Another research direction in recent years involves combining time and space to simulate the mixed land use structure and its dynamic change, accounting for the land use interaction and geographical scale performance. By introducing land use interaction and geographic scale as well as a temporal element of land use mixing, Gehrke and Clifton [25] established the research agenda for a spatial-temporal land use mix measure to evaluate the impact of land use mix on travel behaviour and assess more temporal policies. Liang et al. [26] developed a mixed-cell cellular automata (CA) (MCCA) to simulate the spatio-temporal dynamics of mixed land use structures, and enable mixed land use research to leap from static analysis to dynamic simulation. By combining time and space, this new approach provides a more comprehensive understanding of the degree of land use mixing and its complexities over time, allowing for a better analysis of the interactions and patterns between different land use types.

Traditional measurement methods often use the Shannon diversity index or an entropy-based weighted land use mix index to reflect the degree of land use mixing [10]. In recent years, new approaches have emerged for measuring the degree of land use mixing or simulating the dynamic evolution of mixed land uses. These include the weighted land use mix index [22], mixed-cell CA (MCCA) [26], information entropy of land use structure (IELUS) [27] and multilabel (ML) convolutional neural network CA (ML-CNN-CA) model [28]. These approaches provide more advanced and nuanced ways to measure the mixing degree of land use or simulate the complex dynamics of mixed land uses.

Overall, although scholars have realized that mixed land use is a multidimensional concept that goes beyond any single index and have attempted to quantitatively measure mixed land use from different angles, the focus has remained on the aspects of mixed land use and building complexity or the dimensions of distance, quantity and attributes. To date, no standardized measurement index system or method has been established to evaluate the spatial distribution of mixed land use at urban or regional scales [2,10]. On the other hand, compared to those on mixed urban land use, few studies on mixed rural land use and its influencing factors have been conducted [29]. In addition, due to a lack of methodological guidance and policy support, the practical application of the mixed land use model is hindered.

Research on the relationship between land use and carbon dioxide emissions has yielded abundant results. First, regarding land use types, Dalal and Allen [30] found that a decrease in vegetation area leads to reduced carbon storage and absorption through photosynthesis. Research on Urumqi City showed that transportation land, grassland, and garden land had the highest correlations with total and per capita carbon dioxide emissions, and intensity of carbon dioxide emissions between 2001 and 2015, while construction land had a significant association with carbon dioxide emissions [31]. Second, considering land use structure, a study on the Sichuan Basin found that carbon dioxide emissions and land use structure were spatially autocorrelated [32]. The information entropy of land use structures positively influenced the intensity of carbon dioxide emissions. Carbon dioxide emissions exhibited positive spillover effects, while changes in land use structure did not have a significant regional impact on surrounding areas. It was suggested that potential threshold areas for the impact of land use structure changes on carbon dioxide emissions may exist. Third, in terms of spatial relationships, a study indicated a significant positive spatial correlation between the recessive land use morphology in an advanced situation and the intensity of carbon dioxide emissions. The output and intensity of land use showed stronger positive correlations with carbon dioxide emissions than the input of land use and changes in land property rights [33]. Some researchers have also focused on the relationship between urban spatial structure and carbon dioxide emissions [34]. Fourth, with respect to the factors influencing carbon dioxide emissions, existing studies have identified population size, industrial structure, energy structure, technological development, and concentration of construction land as major factors. However, previous research has paid less attention to the potential relationship between the mixed land use and carbon dioxide emissions. Li et al. [27] found that the relationship between information entropy of land use structure and CO₂ emissions presents a positive U-shaped curve. However, they only use a single index, information entropy of land use structure, to measure mixed land use.

Development zones serve as the main carriers of both the industrial economy and energy consumption and offer important experimental fields for mixed land use. The adoption of development zones as the research object to explore the relationship between mixed land use and carbon dioxide emissions has good representativeness and practical value. The research objectives of this paper are as follows: (1) clarify the meaning of mixed land use, construct an evaluation index system for mixed land use, and measure and analyse the selected research areas accordingly; and (2) estimate the overall carbon dioxide emissions of the development zones, construct a coupling model to analyse the relationship between mixed land use and carbon dioxide emissions, and reveal the differences among different types of development zones. This paper provides a reference for improving the mixing degree of land use and reducing carbon dioxide emissions in development zones.

2. Materials and Methods

2.1. Study Area

Shanghai is located in the eastern region of China, between 120°52′–122°12′ E and 30°40′–31°53′ N. It is the core city of the Shanghai metropolitan area. It is also an international centre of economy, finance, trade, shipping, and technological innovation. As the core spatial carriers of economic activities and windows for opening to the outside world, development zones have made significant contributions to China’s economic growth. The construction of national-level development zones in Shanghai is mature, making them the most important growth anchor in the city. The municipal-level development zones are focused on attracting investment in six key industries (electronic information product manufacturing, biomedical manufacturing, complete equipment manufacturing, fine steel manufacturing, petrochemical and fine chemical manufacturing and automobile manufacturing), and the dominant industries in the zones have begun to take shape. Currently, the development zones in Shanghai are gradually stabilizing their development space and moving towards specialization and distinctiveness. This study selects seven development zones as the research objects (

Table 1. Basic information on Shanghai sample development zones.

Development Zone	Location	Level	Area (ha)	Establishment Year	Main Industries
Minhang Economic and Technological Development Zone	Minhang	National	352.69	1986	Machinery, light industry, pharmaceuticals, heavy equipment industry
	Lingang		1391.60
Caohejing High-Tech Park	Xuhui	National	537.75	1984	Integrated circuits, software, new energy, aerospace, digital content, new materials, life sciences
	Minhang		973.41
Shanghai Chemical Industry Park	Jinshan, Fengxian	National	2903.03	2012	Petrochemical industry
Shanghai Zizhu High-Tech Park	Minhang	National	846.57	2011	Integrated circuits, software, new energy, aerospace, digital content, new materials, life sciences
Shanghai Xinghuo Industrial Park	Jinshan	Provincial	740.24	1984	Optoelectronics, automotive and parts, medical devices
Shanghai Zhujing Industrial Park	Fengxian	Provincial	247.80	2006	Special equipment manufacturing (light industry), textile, clothing and leather products industry, chemical fibre manufacturing, metal products industry
Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	National	1123.00	1990	Processing, manufacturing, shipping logistics, international trade

); their spatial distribution is shown in Figure 1.

2.2. Data Source and Processing

2.2.1. Land Use Data

The basic data of this article is the result of vectorization according to the actual land use of the development zones, including the development zone name, land type, user, service life, building area and other attributes, and the accuracy is plot level.

The data on land use in the seven development zones used in this study in 2020 were obtained from the Shanghai Institute of Geological Survey. Based on the built-up status within the development zones, the land in the development zones was classified into three major categories: built-up areas, unbuilt-up areas, and unbuildable areas, which were further divided into 24 subcategories (

Table 2. Classification of land use, diversity and compatibility, and utilization degree.

Built-Up Status	Land Type	Diversity Classification	Compatibility Classification	Valuation of Utilization Intensity
Built-up	Residential land	1	1	4
	Class 1 industrial land	2	2	4
	Class 2 industrial land	2	3	4
	Class 3 industrial land	2	4	4
	Storage land	3	5	4
	Street land	4	6	4
	Other transport and communication land	5	7	4
	Commercial and service land	6	8	4
	Park and green land	7	9	4
	Other public management and service land	8	10	4
Unbuilt-up	Rural residential land	1	1	2
	Rural industrial land	2	3	2
	Rural storage land	3	5	2
	Rural road	4	6	2
	Rural commercial and service land	6	8	2
	Rural other public management and service land	8	10	2
	Cultivated land (including paddy field and irrigated land)	9	11	1
	Forest land (including timber forest and other forest land)	10	9	1
	Grassland	11	9	1
	Vacant land	12	12	3
	Agricultural facility land (including facility agricultural land and water conservancy facility land)	13	13	1
	Ponds and water surfaces	14	11	1
	Inland shoals and tidal flats	14	14	1
Unbuildable	River and lake land and flood retention area land	14	14	0

Note: part of the unbuilt land is under construction in the remote sensing image. This paper classifies this kind of land according to its construction uses in terms of diversity and compatibility and assigns the value of utilization intensity to 3.

). For the built-up areas, the land use types are subdivided into residential land, industrial land, storage land, street land, other transportation land, public and service land, park and green land, other public service land and so on. For the unbuilt-up areas, land use types are subdivided into rural residential land, rural industrial land, rural storage land, rural road, cultivated land, forest land, agricultural facility land and so on. In development zones, they are mainly based on whether the conditions for land supply are met.

Due to the long processing period of the Third National Land Survey data, there may be inconsistencies with the land use status at the end of 2020, so the historical remote sensing image data was used to make some corrections. In addition, due to the rapid development of the development zones, some areas are actually under construction, and this paper argues that their utilization degree is different from that of built-up areas and other rural lands, and the identification of these plots is also based on remote sensing images. The historical remote sensing imagery data were obtained from Esri’s World Imagery Wayback (https://livingatlas.arcgis.com/wayback, accessed on 1 December 2022).

In the compatibility analysis, different types of industrial land have varying degrees of impact on the surrounding environment. According to China’s land use classification standard (GB/T 21010-2017), industrial land was classified into three categories: Class 1, Class 2, and Class 3, based on the characteristics of land users. Class 1 industrial land causes minimal disturbance and pollution for residential and public facility land, and it can accommodate industries such as banking, electronics, and handicraft manufacturing. Class 2 industrial land refers to land that causes certain disturbances, pollution, and safety hazards for residential and public environments during industrial production, such as agricultural and food processing, construction and installation, furniture manufacturing, pharmaceutical manufacturing, and other industries. Class 3 industrial land causes significant disturbances and pollution to residential and public environments. Typical industries in this category include petroleum processing, coking and nuclear fuel processing, manufacturing of chemical raw materials and chemical products, mining and selection of black or nonferrous metals, and large-scale machinery manufacturing. The specific classification of land use is shown in Table 2. The serial number only represents the secondary classification results of various types of land use under different indicators. For example, residential land in built-up areas and rural residential land in unbuilt-up areas are both residential types, where the diversity categories belong to the same category, but the built-up status is different, so the utilization intensity categories are different. The utilization intensity is divided into five levels. See Section 2.3.4 for specific classification and assignments.

2.2.2. Carbon Dioxide Emissions Data

This study uses industrial energy consumption data from the development zones in Shanghai to estimate the carbon dioxide emissions from built-up areas. Additionally, considering the impact of the COVID-19 pandemic on social aspects, this study chooses to analyse the development zone energy consumption data in 2019 (

Table 3. Industrial energy consumption and total industrial output value of seven sample development zones in Shanghai in 2019.

Development Zone	Location	Industrial Energy Consumption (Tonnes of Standard Coal)	Total Industrial Output Value (10,000 CNY)
Minhang Economic and Technological Development Zone	Minhang, Lingang	202,281	5,579,552
Caohejing High-Tech Park	Xuhui	147,876	2,424,200
	Minhang	98,376	4,700,256
Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	9,774,535	10,662,090
Shanghai Zizhu High-Tech Park	Minhang	86,970	1,683,704
Shanghai Xinghuo Industrial Park	Jinshan	1,018,396	2,078,431
Shanghai Zhujing Industrial Park	Fengxian	41,967	558,972
Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	209,391	4,494,885

). The data sources include the “Shanghai Statistical Yearbook” [35], “Shanghai Industrial Energy Efficiency Guide” [36], and statistical yearbooks of various districts in Shanghai. Due to the difficulty of data acquisition and statistical analysis, this paper considers the entire development zone as the minimum scale for calculating CO₂ emissions. Therefore, the resolution of mixed land use results is reduced to calculate the overall mixed utilization of the development zone, and some conclusions are drawn which seem to have no particularity. However, different development zones and different areas within the same development zone have different characteristics of mixed use. This paper aims to verify the importance of the mixed use of land by studying the relationship between mixed use of land and CO₂ emissions in development zones.

It should be noted that in specific areas such as the development zone, especially the built-up areas within the development zone, industrial economic activity is the main function, as well as the main energy consumer and CO₂ emitter; industrial energy consumption accounts for the majority of all types of energy consumption. In addition, in the unbuilt areas of the development zone, the CO₂ emissions from land use also play a certain role, but the proportion is very small. Therefore, it is appropriate to use industrial energy consumption to estimate CO₂ emissions.

2.2.3. Other Data

This study also utilized urban transportation station data and road network data. Using web scraping techniques, we collected public transportation station data (including bus stops, subway stations, and other urban transportation facilities) in Shanghai from AMAP (a Chinese mapping service). The collected data include attributes such as the name, address, coordinates, and category of each transportation facility. We performed coordinate correction and removed invalid and duplicate stations, as well as the entrances/exits of subway stations, to obtain the refined dataset for this study.

2.3. Measurement Index and Method of Mixed Land Use

In this study, the degree of mixed land use within the development zones is calculated using a grid-based approach. The size of the grid is determined based on the minimum area required to satisfy the functional spatial unit requirements. In this study, a hexagonal grid with a side length of 50 m is chosen for the calculations. A regular hexagon is a polygon with the maximum number of sides that can form a uniform grid pattern. It is also the most circular-like polygon. Compared to rectangular grids, hexagonal grids can more naturally display curved patterns in the data and provide more neighbouring connections in compatibility calculations. The mixing degree of land use is calculated by combining the measurement results of four indices: diversity, accessibility, compatibility and utilization intensity.

2.3.1. Diversity Evaluation

The Shannon entropy index, as the most widely used measure of land use mixing degree [37,38], effectively combines the characteristics of diversity and evenness, meeting the requirements of diversity indicators in mixed land use measurement [39]. This index assumes that land use is most balanced when all land use types have equal proportions [40]. The calculation formula is as follows:

$$SEI=-\sum _{i=1}^n{p}_i{log}_{10}(p_i)$$

(1)

In the equation, n represents the number of land use types, and p_i represents the probability of type i occurring (in this study, it is represented by the proportion of land area occupied by that type). When the land area proportions of different types are equal, the index reaches its maximum value, indicating maximum diversity.

2.3.2. Accessibility Evaluation

Based on the characteristics of the development zone, this study selects bus stops and subway stations as target points and utilizes the shortest distance and opportunity accumulation models to assess transportation accessibility in the study area. The calculation steps for accessibility are as follows:

• Step 1: calculate the distance accessibility value for each grid cell within the development zone.
• Step 2: estimate the information point accumulation opportunity for each grid cell within the development zone.
• Step 3: the two accessibility indicators are combined to obtain the overall accessibility value for the development zone.

First, the minimum proximity distance between a point and a destination is calculated using the Euclidean distance. With each cell as the starting point and the bus station and subway station as the end points, the distance between each grid pixel and the nearest target point is calculated through proximity analysis. Finally, the traffic accessibility of the development zone is evaluated by the statistics of the minimum proximity distance. The accessibility index is calculated as follows:

$${ADI}_1=1-\frac{d_i(x,y)}{D_{max}}$$

(2)

where D_max is a constant, which is determined to be 3000 meters in this paper, and di(x,y) is the shortest distance between the raster element i and the target point. If di(x,y) exceeds 3000 meters, accessibility is 0.

Second, based on the opportunity accumulation model, the number of traffic stations that can be reached within a certain time or distance is calculated. The larger the number is, the higher the accessibility. According to the “15-minute walking life circle” and the starting point of each cell centre, the threshold value in this paper is set at 1000 meters. The total number of traffic stations within the radius is calculated, and the obtained number value is used as the basis for measuring the accessibility of the cell (Figure 2); thus, the accessibility value is calculated. To compare accessibility differences within development zones, the calculation formula is as follows:

$$a_i={\sum }_j{O}_{jt}$$

(3)

$${ADI}_2=\frac{a_i}{a_{max}}$$

(4)

where a_i is the accessibility of point i; j is the node whose distance from point i is less than threshold t; the threshold t is 1000 meters; and O_jt is the number of opportunities provided by node j. The accessibility indicator ADI₂ is expressed by the relative opportunity cumulative value.

Finally, by combining the above two results, an accessibility value is obtained and assigned to the honeycomb. The accessibility index is based on distance accessibility, supplemented by quantity accessibility, and the traffic accessibility of the development zone is evaluated as a whole. The larger the value is, the higher the accessibility. The calculation formula is as follows:

$${ADI}_i={0.7\times ADI}_{1i}+0.3\times {ADI}_{2i}$$

(5)

2.3.3. Compatibility Evaluation

In this study, a hexagonal cell with a 50 m side length is taken as the measurement unit. The specific method of measurement is as follows: according to the compatibility matrix, the compatibility relation value of each land type in the adjacent cell is calculated, and then the compatibility value of two adjacent cells is obtained by multiplying the area weight. Finally, the average compatibility value of each cell and the neighbourhood cell is calculated as the compatibility index of the cell (Figure 3). The calculation formula is as follows:

$$COM=1-\frac{1}{2}\times {\sum }_{i=1}^n\frac{c_i}{n}$$

(6)

where COM is the compatibility value of a single cell, n is the number of adjacent cells, and c_i is the compatibility value of adjacent cells calculated based on area weighting.

2.3.4. Utilization Intensity Evaluation

In this paper, graded values of different land types are used to represent utilization intensity. According to the territorial planning of the development zone, the land should be developed and built up accordingly; hence, this paper divides the degree of development and utilization into five levels according to the degree of completion and assigns values. The specific classification is as follows: all land types in the built land are assigned a value of 4; the open space in the unbuilt land and construction land in the unbuilt land that can be built up in the short term is assigned a value of 3; the rural industrial and mining storage land and other rural construction land in the uncompleted land that have basically met the conditions of being converted into construction land are assigned a value of 2; the cultivated land, forestland and grassland in the unbuilt land, which bear the ecological function of agricultural production, that may not be converted into construction land in the short term are assigned a value of 1; the land of rivers and lakes and their flood storage areas that belong to unbuildable land are assigned a value of 0 (see Table 1 for specific land classification and assigned value).

The formula for calculating land use intensity is:

$$LUI={\sum }_{i=1}^n{x}_i\times \frac{S_i}{S}$$

(7)

where LUI is the land use intensity, x_i is the assignment of the i type of land use intensity, n is the number of land use types within the honeycomb, S_i is the area of the i type of land use, and S is the total land use area of the development zone.

2.3.5. Measurement of Mixed Land Use

The mixing degree of land use is calculated by combining the measurement results of four indices: diversity, accessibility, compatibility and utilization intensity. The specific calculation steps are as follows: (1) at the cellular grid scale, the calculation results of diversity, accessibility, compatibility and availability are standardized. (2) This paper considers that the four indicators are equally important for the improvement of the land use mixing degree, so the average value of the four indicators is taken as the calculation result of the land use mixing degree. (3) At the scale of the development zone, the land use mixing degree of the whole development zone is calculated, and the natural break classification is used to divide the development zone into four grades: high, relatively high, relatively low and low.

$$LMDI=\frac{SEI+ADI+COM+LUI}{4}$$

(8)

In the formula, LMDI is the land use mixing degree, and the calculated result is between 0 and 1. The closer the mixing degree is to 1, the higher the mixing degree of land use is.

2.4. Carbon Dioxide Emissions Calculation Method

2.4.1. Calculation of Direct Carbon Dioxide Emissions

Existing research results [33,41] were combined to establish a land use carbon budget measurement system. The carbon dioxide emissions coefficients of cultivated land, forestland, grassland and water area were determined to be 0.422 ${\mathrm{t}/\mathrm{h}\mathrm{m}}^2·\mathrm{a}$, −0.644 ${\mathrm{t}/\mathrm{h}\mathrm{m}}^2·\mathrm{a}$, −0.022 ${\mathrm{t}/\mathrm{h}\mathrm{m}}^2·\mathrm{a}$ and −0.253 ${\mathrm{t}/\mathrm{h}\mathrm{m}}^2·\mathrm{a}$, respectively (a negative number indicates that the class is mainly responsible for carbon absorption). According to the National Standard of the People’s Republic of China, Code for Park Design GB51192-2016 (https://www.doc88.com/p-5803806810585.html, accessed on 1 March 2023), the planting area of garden plants in all types of parks and green spaces should be at least 65%, that is, the rate of green land; hence, its carbon sequestration capacity should be slightly lower than that of forestland. Historical remote sensing images were combined, and the carbon absorption coefficient was finally determined to be −0.457 ${\mathrm{t}/\mathrm{h}\mathrm{m}}^2·\mathrm{a}$ a (

Table 4. Carbon dioxide emissions coefficients for land use.

Land Use Type	Carbon Emissions Coefficient (${\mathbf{t}/\mathbf{h}\mathbf{m}}^2·\mathbf{a}$)
Cropland	0.422
Forest land	−0.644
Grassland	−0.022
Green space	−0.457
Water	−0.253

). Direct carbon dioxide emissions were calculated as follows:

$$LC=\sum C_i=\sum S_i\times {\delta }_i$$

(9)

where LC is the total carbon dioxide emissions directly calculated and C_i is the total carbon dioxide emissions of species i. S_i is the area of land type i, and ${\delta }_i$ is the carbon dioxide emissions/absorption coefficient of land type i.

2.4.2. Calculation of Indirect Carbon Dioxide Emissions

For built-up land, indirect carbon dioxide emissions are calculated as its combined energy consumption (standard amount) multiplied by energy use efficiency. The calculation formula is as follows:

$$EC=\sum E_t\times \theta$$

(10)

where $E_t$ is the energy consumption of the development zone (Table 3), and $\theta$ is the carbon dioxide emissions coefficient of standard coal. The carbon dioxide emissions coefficient of standard coal refers to the carbon produced by the complete combustion of 1 ton of standard coal (ton carbon/ton of standard coal, tc/tce). The recommended value of the Energy Research Institute of the National Development and Reform Commission is 0.67, the reference value of the Institute of Energy Economics of Japan is 0.68, and the reference value of the Energy Information Administration of the United States Department of Energy is 0.69. In this paper, 0.67 tc/tce is used for the calculation.

2.4.3. Calculation of Total Carbon Dioxide Emissions

The total carbon dioxide emissions (TCE) estimate of the development zone is obtained by adding the direct carbon dioxide emissions (LC) and indirect carbon dioxide emissions (EC). The calculation formula is as follows:

$$TCE=LC+EC$$

(11)

2.5. Coupling Degree and Coupling Coordination Degree Analysis Model

The coupling degree model was used to evaluate the interaction between mixed land use and carbon dioxide emissions, and the coupling coordination degree model was used to evaluate the coordinated development state between the two. Due to the differences in the area size, industrial layout and energy utilization efficiency of each development zone, the contribution rates of carbon dioxide emissions output value and carbon dioxide emissions per unit area were selected as indicators to measure the carbon dioxide emissions level of the development zone. The coupling degree and coupling coordination degree of mixed land use with these two indicators were calculated, and the relationship between them was cross-verified.

According to Israeli research [42], the coupling degree can be calculated as follows:

$$C=2\times {\left[\frac{U·V}{{(U+V)}^2}\right]}^{\frac{1}{2}}$$

(12)

where C is the coupling degree between the two systems, the value range is [0, 1], and U and V are standardized values for the two systems. The closer the coupling degree is to 1, the better the coupling. With reference to the research of relevant scholars [43], this paper divides the evolution process of coupling land mixed use and carbon dioxide emissions systems into four stages: (1) 0 < C ≤ 0.3, low-level coupling stage; (2) 0.3 < C ≤ 0.5, antagonistic stage; (3) 0.5 < C ≤ 0.8, running-in stage; and (4) 0.8 < C ≤ 1, high-level coupling stage.

However, the evaluation result of the coupling degree model can reflect the interaction degree of the mixed land use system and carbon dioxide emissions systems, but it cannot reflect the coordination degree of development between them. Therefore, the coupling coordination degree model is selected to reflect the coordination development degree between the two systems. Its calculation formula is as follows:

$$\mathrm{D}=\sqrt{C\times T}$$

(13)

$$T=\alpha \times U+\beta \times V$$

(14)

where D is the coupling coordination degree, T is the comprehensive contribution degree of the two systems, and $\alpha$ and $\beta$ are the weight coefficients, representing the relative importance of the two systems. In this paper, 0.5 is selected for both. The value range of D is also [0, 1]. The closer it is to 1, the higher the coordination. Referring to previous studies [41,44], in this paper, the coupling coordinated development states of mixed land use and carbon dioxide emissions systems in development zones are divided into four types: (1) 0.0 < D ≤ 0.3, the system is in an extremely dysfunctional stage; (2) 0.3 < D ≤ 0.5, the system is in the out-of-balance stage; (3) 0.5 < D ≤ 0.8, the system is in the coordination stage; and (4) 0.8 < D ≤ 1.0, the system is in a highly coordinated stage.

3. Results

3.1. Measurement Results of Mixed Land Use in Development Zones

3.1.1. Diversity Calculation Result

From the above seven development zones (

Table 5. Results of standardization of land type diversity in Shanghai development zones.

No.	Development Zone	Location	Diversity
			Min	Max	Mean
(a)	Minhang Economic and Technological Development Zone	Minhang	0	0.70	0.21
		Lingang	0	0.89	0.24
(b)	Caohejing High-Tech Development Zone	Xuhui	0	0.86	0.24
		Minhang	0	1.00	0.29
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	0	0.78	0.16
(d)	Shanghai Zizhu High-Tech Park	Minhang	0	0.92	0.25
(e)	Shanghai Xinghuo Industrial Park	Fengxian	0	0.85	0.17
(f)	Shanghai Zhujiang Industrial Park	Jinshan	0	0.90	0.30
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	0	0.93	0.21

), the minimum value of diversity is 0, and the maximum value is between 0.64 and 0.82. The natural break classification is used to divide the index into four levels: high, relatively high, relatively low and low. In terms of the types of development zones, the diversity of economic and technological development zones and high-tech development zones is higher than that of industrial development zones and bonded areas. The diversity index can roughly reflect the perfection of other living service facilities in the development zone: the greater the diversity is, the greater the degree or potential of facilities; the lesser the diversity is, the more likely the development zone has a single function land and potentially lacks service facilities. For example, the diversity evaluation result of Shanghai Zizhu High-Tech Park is relatively high, indicating that the living service function in the development zone is relatively perfect, and it belongs to the development zone of integration of industry and city.

3.1.2. Accessibility Calculation Results

After the accessibility results of the seven development zones were unified and standardized, the minimum, maximum and average accessibility values of each development zone were obtained (

Table 6. Statistical results of land type accessibility in Shanghai development zones.

No.	Development Zone	Location	Accessibility
			Min	Max	Mean
(a)	Minhang Economic and Technological Development Zone	Minhang	0.51	0.84	0.74
		Lingang	0.27	0.81	0.66
(b)	Caohejing High-Tech Development Zone	Xuhui	0.61	0.98	0.85
		Minhang	0.59	0.86	0.74
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	0	0.78	0.36
(d)	Shanghai Zizhu High-Tech Park	Minhang	0.51	0.84	0.71
(e)	Shanghai Xinghuo Industrial Park	Fengxian	0.55	0.79	0.67
(f)	Shanghai Zhujiang Industrial Park	Jinshan	0.48	0.76	0.65
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	0.27	1.00	0.79

). From the perspective of spatial distribution, the Xuhui area of the Caohejing High-Tech Development Zone and Shanghai Waigaoqiao Free Trade Zone are basically highly accessible; the Minhang area of the Minhang Economic and Technological Development Zone, Minhang area of the Caohejing High-Tech Development Zone and Shanghai Zizhu High-Tech Park have relatively high accessibility; and the accessibility of the Shanghai Chemical Industry Development Zone, Shanghai Xinghuo Industrial Park and Shanghai Zhujing Industrial Park is low.

3.1.3. Compatibility Calculation Result

Based on the compatibility judgement matrix, the cellular compatibility value in the development zone is calculated. The value ranges from 0 to 1. The higher the value is, the better the compatibility. The compatibility of the whole development zone is relatively high, and the difference is not significant (

Table 7. Statistical results of land type compatibility in Shanghai development zones.

No.	Development Zone Name	Location	Compatibility
			Min	Max	Mean
(a)	Minhang Economic and Technological Development Zone	Minhang	0.51	1	0.90
		Lingang	0.19	1	0.91
(b)	Caohejing High-Tech Development Zone	Xuhui	0.22	1	0.81
		Minhang	0	1	0.80
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	0	1	0.92
(d)	Shanghai Zizhu High-Tech Park	Minhang	0.31	1	0.90
(e)	Shanghai Xinghuo Industrial Park	Fengxian	0.11	1	0.87
(f)	Shanghai Zhujiang Industrial Park	Jinshan	0.20	1	0.81
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	0.31	1	0.88

). From the perspective of development zone type, the compatibility of economic and technological development zones is the best; the overall compatibility of bonded areas is relatively high; and the compatibility of industrial development zones is relatively low.

3.1.4. Utilization Intensity Calculation Results

The maximum value of internal utilization intensity in the seven development zones is 1, and the minimum value ranges from 0 to 0.18 (

Table 8. Statistical results of land use intensity in Shanghai development zones.

No.	Development Zone	Location	Utilization intensity
			Min	Max	Mean
(a)	Minhang Economic and Technological Development Zone	Minhang	0.38	1	0.96
		Lingang	0.02	1	0.69
(b)	Caohejing High-Tech Development Zone	Xuhui	0.21	1	0.90
		Minhang	0	1	0.70
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	0.07	1	0.68
(d)	Shanghai Zizhu High-Tech Park	Minhang	0	1	0.74
(e)	Shanghai Xinghuo Industrial Park	Fengxian	0.11	1	0.87
(f)	Shanghai Zhujiang Industrial Park	Jinshan	0.18	1	0.77
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	0.07	1	0.91

). From the perspective of space, most of the early development zones are located in the central city and the inner suburbs, with a longer development time and a relatively high overall development and utilization degree. In newly expanded development areas, the land use degree is relatively low.

3.1.5. Calculation Result of Mixed Land Use

According to the diversity, accessibility, compatibility and utilization intensity of the development zones, the mixing degree of land use was calculated after standardization. It can be seen from Figure 4 that the land use mixing degree of the development zones has the following main characteristics: (1) the Minhang Economic and Technological Development Zone (Figure 4a) and Caohejing High-Tech Development Zone (Figure 4b) span different zones in space, so the mixing degree of the two areas is quite different, and the mixing degree of the original main area is better than that of the newly developed area. (2) In the Shanghai Chemical Industry Development Zone, honeycomb with low mixing degree accounted for most (Figure 4c), but in the six other development zones, the honeycombs with a high mixing degree accounted for more than half of the development zones. The leading industry in the Shanghai Chemical Industry Development Zone is petrochemical. Although the compatibility is high, its diversity, accessibility and utilization intensity are the lowest among the seven development zones, so its mixing degree of land use is the lowest. (3) The leading industry in the Shanghai Waigaoqiao Free Trade Zone is the service industry, with high utilization intensity, diversity, accessibility and compatibility, so its degree of mixing is the best (Figure 4g). (4) There is a strong correlation between accessibility and land use mixing degree (

Table 9. The mixing degree of land use and its grades in seven development zones in Shanghai.

No.	Development Zone	Location	Diversity	Accessibility	Compatibility	Utilization Intensity	Mixing Degree
(a)	Minhang Economic and Technological Development Zone	Minhang	Relatively low	Relatively high	Relatively high	High	High
		Lingang	Relatively high	Relatively low	High	Low	Relatively low
(b)	Caohejing High-Tech Development Zone	Xuhui	Relatively high	High	Low	Relatively high	Relatively high
		Minhang	High	Relatively high	Low	Low	Relatively low
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	Low	Low	High	Low	Low
(d)	Shanghai Zizhu High-Tech Park	Minhang	Relatively high	Relatively high	Relatively high	Relatively low	Relatively high
(e)	Shanghai Xinghuo Industrial Park	Fengxian	Low	Relatively low	Relatively low	Relatively high	Relatively low
(f)	Shanghai Zhujing Industrial Park	Jinshan	High	Relatively low	Low	Relatively low	Relatively low
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	Relatively low	Relatively high	Relatively high	High	High

). The area with higher accessibility is usually closer to the central urban area, and its land use mixing degree is also higher.

3.2. Analysis of Carbon Dioxide Emissions in Development Zones

Based on the economic contribution coefficient of CO₂ emissions [45], this paper uses the industrial output contribution rate (IOCR) of CO₂ emissions to characterize the relative carbon production efficiency of development zones. When IOCR > 1, it indicates that the carbon production efficiency of the development zone is higher than the average level of Shanghai. When IOCR < 1, it means that the carbon production efficiency of the development zone is lower than the average level of Shanghai. The second index chosen in this paper to measure CO₂ emissions of the development zone is the carbon dioxide emissions per unit area (CUA) of the development zone.

As can be seen from

Table 10. Estimation results of CO₂ emissions in the seven development zones of Shanghai.

No.	Development Zone	Location	Industrial CO2 Emissions (t)	Land Use CO2 Emissions (t)	Total CO2 Emissions(t)	IOCR	CUA (t)
(a)	Minhang Economic and Technological Development Zone	Minhang, Lingang	135,528.27	15.49	135,543.76	4.06	77.71
(b)	Caohejing High-Tech Development Zone	Xuhui	99,077.05	−28.44	99,048.61	2.41	184.19
		Minhang	65,911.92	−54.42	65,857.50	7.02	67.66
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	6,548,938.50	−219.26	6,548,719.20	0.16	2255.82
(d)	Shanghai Zizhu High-Tech Park	Minhang	058269.90	−105.17	58,164.73	2.85	68.71
(e)	Shanghai Xinghuo Industrial Park	Fengxian	682,325.32	−19.96	682,305.36	0.30	921.74
(f)	Shanghai Zhujiang Industrial Park	Jinshan	28,117.89	−13.18	28,104.71	1.96	113.42
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	140,291.97	−23.75	140,268.22	3.16	124.90

Notes: ① $IOCR=\frac{{GVIO}_i}{{EC}_i}/\frac{GVIO}{EC}$, ${GVIO}_i \mathrm{and} {EC}_i$ are the total industrial output value and industrial carbon emission of the development zone i, respectively, while GVIO and EC are the total industrial output value and industrial carbon emission of Shanghai, respectively. ② CUA = carbon emission of per unit area.

, the IOCR of the seven development zones is mostly higher than the average level of Shanghai, and only the Shanghai Chemical Industry Development Zone and Shanghai Xinghuo Industrial Park are lower than the average level of Shanghai. Among them, the leading industry in the Shanghai Chemical Industry Development Zone is the petrochemical industry with high energy consumption, so its carbon production efficiency is very low. The production technology of industrial enterprises in Shanghai Xinghuo Industrial Park is relatively backward, and the level of low-carbon development is low.

From the perspective of CUA (Table 10), its value is generally opposite to the IOCR, that is, the development zone with high IOCR has a correspondingly low CUA, which is also consistent with the common cognition. Among them, the IOCR and CUA in the Shanghai Zizhu High-Tech Park and Shanghai Zhujing Industrial Park are lower than those of Shanghai Waigaoqiao Free Trade Zone. This may be due to the low land use intensity and fewer industrial land and enterprises in these two development zones, resulting in lower total and average values of CO₂ emissions. In addition, the Shanghai Chemical Industry Development Zone has the largest amount of carbon absorption, but because its indirect CO₂ emissions are very high, its carbon absorption plays a small role, and the CUA is about 2–33 times that of other development zones.

3.3. Coupling Coordination Analysis of Mixed Land Use and Carbon Dioxide Emissions

As shown in

Table 11. Coupling degree and coupling coordination degree of mixed land use and carbon dioxide emissions.

No.	Development Zone	Location	LMDI	${\mathit{C}}_1$	${\mathit{D}}_1$	${\mathit{C}}_2$	${\mathit{D}}_2$
(a)	Minhang Economic and Technological Development Zone	Minhang, Lingang	0.59	0.99	0.67	0.99	0.88
(b)	Caohejing High-Tech Development Zone	Xuhui	0.64	0.89	0.54	1.00	0.91
		Minhang	0.50	0.96	0.76	0.96	0.76
(c)	Shanghai Chemical Industry Development Zone	Jinshan, Fengxian	0.25	0.75	0.02	0.76	0.02
(d)	Shanghai Zizhu High-Tech Park	Minhang	0.62	0.93	0.58	0.99	0.91
(e)	Shanghai Xinghuo Industrial Park	Fengxian	0.45	0.47	0.12	0.99	0.54
(f)	Shanghai Zhujiang Industrial Park	Jinshan	0.50	0.93	0.39	0.97	0.75
(g)	Shanghai Waigaoqiao Free Trade Zone	Pudong New Area	0.70	0.92	0.66	1.00	0.98

Note: since only the total carbon dioxide emissions data of the Minhang Economic and Technological Development Zone are available, the two areas are not distinguished, so the two areas are calculated together.

, (1) except for Shanghai Xinghuo Industrial Park, the coupling degree (C₁) of the mixing degree of land use and the output value contribution rate of carbon dioxide emissions of other development zones are above 0.7, which also verifies the strong interaction between the two. In Shanghai Xinghuo Industrial Park, the coupling degree between mixed land use and the output value contribution rate of the carbon dioxide emissions is only 0.47, which is in the antagonistic stage, indicating that the development of mixed land use and the output value contribution rate of the carbon dioxide emissions is unbalanced, and the mutual influence between the two systems cannot be highlighted. The Shanghai Chemical Industry Development Zone is in the running-in stage, and other development zones are in the high-level coupling stage. (2) The coupling coordination degree (D₁) of mixed land use and the output value contribution rate of the carbon dioxide emissions in each development zone mainly includes four stages from extreme imbalance to intermediate coordination. The development zones in the extremely dysfunctional stage are the Shanghai Chemical Industry Development Zone and Shanghai Xinghuo Industrial Park. In the stage of imbalance is the Shanghai Zhujing Industrial Park. At the primary stage of coordination are the Shanghai Zizhu High-Tech Park and Xuhui area of the Caohejing High-Tech Development Zone. The coupling coordination degrees of the Minhang area of the Caohejing High-Tech Development Zone, Minhang Economic and Technological Development Zone and Waigaoqiao Free Trade Zone belong to the intermediate coordination stage.

Except for the Shanghai Chemical Industry Development Zone, the coupling coordination degree (C₂) between mixed land use and carbon dioxide emissions per unit area in other development zones is above 0.9 (Table 10). Only the Shanghai Chemical Industry Development Zone is in the running-in stage, and the other development zones are in the high-level coupling stage. Similarly, the coupling coordination degree (D₂) of mixed land use and carbon dioxide emissions per unit area in the Shanghai Chemical Industry Development Zone is in an extremely dysfunctional stage. Shanghai Xinghuo Industrial Park is in the primary coordination stage; Shanghai Zhujing Industrial Park and the Minhang Area of the Caohejing High-Tech Development Zone are in the intermediate coordination stage. The remaining development zones are in a highly coordinated stage, and the overall development is more reasonable and synchronized.

In summary, the sample development zones can be divided into the following three categories according to their coupling coordination development: (1) high-level coupling coordination development, including the Minhang Economic and Technological Development Zone and Shanghai Waigaoqiao Free Trade Zone; (2) medium-level coupling coordination development, including the two areas of Caohejing High-Tech Development Zone and Shanghai Zizhu High-Tech Park, which are high-tech development zones; and (3) low-level coupling coordination development, including the Shanghai Chemical Industry Development Zone, Shanghai Xinghuo Industrial Park and Shanghai Zhujing Industrial Park, which are industrial development zones.

4. Discussion

4.1. The Relationship between Location and Mixed Land Use Is Diverse and Heterogeneous

Our research found that, among the four dimensions of mixed land use, there is a strong correlation between accessibility and land use mixing degree, that is, location is an important factor affecting the mixing degree of land use in development zones. However, the relationships between the other three factors and location showed significant heterogeneity.

Overall, although the degree of diversity and compatibility is not closely related to location, most development zones show the opposite result of compatibility and diversity; that is, diversity is high, but compatibility is relatively low. For example, the Caohejing High-Tech Development Zone and Shanghai Zhujing Industrial Park have high land use diversity but low compatibility. The land use diversity of the Shanghai Chemical Industry Development Zone is low, but the compatibility is high. To some extent, this also reflects that in the initial stage of site selection and construction of development zones in China, it is difficult to ensure the compatibility of land use types while improving diversity, perhaps due to insufficient consideration of the impact on the surrounding land. In the early stage of China’s reform and opening up to the world, both environmental impact assessment and social impact assessment in economic construction were missing, let alone the compatibility assessment of land use.

There is a certain correlation between land use intensity and location in development zones. Within the outer ring road, the land use intensities of the Shanghai Waigaoqiao Free Trade Zone and Xuhui area of the Caohejing High-Tech Development Zone are higher. However, outside the outer ring, the utilization intensity of the development zone varies from high to low. The main reason is that the intensity of land use is related to the establishment time of the development zone. The development zones established earlier have a relatively high utilization intensity; they include the Minhang area in the Minhang Economic and Technological Development Zone (1986), the Xuhui area in the Caohejing High-Tech Development Zone (1984), Shanghai Xinghuo Industrial Park (1984) and the Shanghai Waigaoqiao Free Trade Zone (1990). The Lingang area of the Shanghai Minhang Economic and Technological Development Zone and the Minhang area of the Caohejing High-Tech Development Zone are new expansion areas, so the current land use intensity is relatively low.

4.2. Can Diversity Be Used as a Key Index to Measure the Mixing Degree of Land Use?

Land use diversity is the main measure of traditional mixed land use [46,47]. First, as early as more than half a century ago, diversity has been used as one of the indicators of urban vitality [48], and it has a similar effect on urban development with the mixing degree of land use. Second, its calculation method is simple and the data are easy to obtain. However, our measurement results show that there is no strong interaction between diversity and mixing degree of land use. In other words, diversity is not the key factor affecting the mixing degree of land use. For example, the land use diversities of the Minhang area of the Minhang Economic and Technological Development Zone and the Shanghai Waigaoqiao Free Trade Zone are low, but their mixing degrees of land use are high; on the contrary, the land use diversities of the Lingang area of the Minhang Economic and Technological Development Zone, the Minhang area of the Caohejing High-Tech Development Zone, and Shanghai Zhujing Industrial Park are high or relatively high, but their mixing degrees of land use are low or relatively low. This indicates that there is no consistent corresponding relationship between the two. Although our research sample is limited, we can at least conclude that using a single index of land use diversity to measure the multi-dimensional land use mixing degree has obvious one-sidedness.

4.3. “Low Degree of Mixing” Does Not Mean That the Overall Structure of the Development Zone Is Unreasonable

Among the seven sample development zones, the land use mixing degree of the Shanghai Chemical Industry Development Zone is the lowest, mainly because it is restricted by the characteristics of leading industries. However, its spatial layout is mainly urban production space, and the ecological spatial distribution is relatively uniform. The leading industry of the Shanghai Chemical Industry Development Zone is the petrochemical industry, which is prone to accidental explosions, so enterprises should not be overly concentrated and need to retain a certain buffer area. The diversity, accessibility and utilization intensity of land use are all affected by this factor to a certain extent; thus, the development level of industry-city integration cannot be compared with other industrial parks and needs to be treated differently. Therefore, the quality of the land use mixing degree should not only be judged according to the size of the value; it should also account for the particularity of industrial categories. In other words, a low degree of mixing does not mean that the overall structure of the development zone is unreasonable.

4.4. The Relationship between Mixed Land Use and Carbon Dioxide Emissions

The results of this study show that there is a significant interaction between land use mixing degree and carbon dioxide emissions in most development zones of Shanghai, that is, increasing land use mixing degree can help reduce the total amount and intensity of carbon dioxide emissions. Of course, the relationship between land use mixing degree and carbon dioxide emissions is not fixed, it is not only related to the leading industrial characteristics of the development zones, but also closely related to the development stage of the development zones. In other words, the relationship between the two may be diverse and complex, and does not simply lead to a unified conclusion. The research results of this paper are basically consistent with those of some scholars. Zhong et al. [49] studied the relationship between urban traffic carbon emissions and land use in Changchun city, and the results showed that improving land use intensity and increasing land mixed use can reduce traffic carbon emissions, and the implementation effect of increasing land diversity is better than that of improving land use intensity. Tan et al. [50] also confirmed that CO₂ emissions from passenger transport in traffic analysis zones (TAZs) at the community level in Shenzhen international low-carbon city were associated with mixed land use. The research results of Xu et al. [51] demonstrated that the land use mixing degree and the decoupling state of CO₂ emissions in the Hohhot-Baotou-Ordos-Yulin urban agglomeration have spatial and temporal heterogeneity.

5. Conclusions and Prospect

5.1. Conclusions

In this study, several typical development zones in Shanghai were selected as the research object, a honeycomb grid with a side length of 50 metres was used as the unit to measure mixed land use, the development zone was used as the unit to estimate carbon dioxide emissions, and the coupling coordination model was used to analyse the coordinated development relationship between the two. The following conclusions are drawn:

(1) The level and constraints of mixed land use in different types of development zones are different. The mixing degree of land use in the development zone is affected by many factors. The mixing degree of the bonded zone is the highest, and the mixing degree of the economic and technological development zones and the high-tech development zones are higher than those of the industrial development zones. Affected by location and leading industries, the mixing degree of land use in the Shanghai Chemical Industry Development Zone is the lowest.

(2) The carbon dioxide emissions capacity of economic and technological development zones and high-tech development zones is better than that of industrial development zones and bonded areas. The coupling model results show that there is a strong interaction between mixed land use and carbon dioxide emissions. The results of the coupling coordination model show that in terms of the coordinated development level between mixed land use and carbon dioxide emissions, the economic and technological development zones and bonded areas > high-tech development zones > industrial development zones.

(3) The development zones with high-level coordinated development are economic and technological development zones and bonded areas; the medium-level coordinated development zones are high-tech development zones; and the low-level coordinated development zones are mainly industrial development zones. However, the mixed use of land is closely related to industrial development, and the possible externalities of the second and third types of industrial land are the main factors restricting the mixing degree of land use, resulting in a weak integration relationship between urban production space and other land uses.

5.2. Prospect

The main innovations of this study are as follows: first, on the basis of the existing mixing degree measurement of land use, the land use intensity index is introduced, the land classification is refined based on the development zone, and the land use compatibility matrix is supplemented. Second, the interaction between mixed land use and carbon dioxide emissions efficiency and intensity is discussed by using the coupling coordination model. Third, while the focus of previous research has been the administrative spatial scale (community scale, district or town scale and city scale) [5], the focus of this paper is the functional spatial scale (development zone scale). The following limitations impact this paper. First, only the mixing degree of land use on the plane is studied; research on the mixing of different building uses and building monomers on the opposite body layer is lacking. Second, due to the limitations of data collection and data accuracy, only the industrial carbon dioxide emissions and land use carbon dioxide emissions of the development zones were estimated, while the carbon dioxide emissions of residents and the carbon dioxide emissions of the tertiary industry were not included in the statistical analysis, which may have a certain impact on the final evaluation results. Future research will aim to address these questions.