Spatiotemporal Dynamics of the Eco-Innovation Level of China’s Marine Economy

Abstract

This paper aims to provide a reliable basis for formulating, revising, and selecting sustainable marine economic development plans through a scientific and comprehensive evaluation of the eco-innovation level of China’s marine economy. Based on the analysis of the three-stage theoretical model of marine economic eco-innovation, an index system for evaluating the eco-innovation level of China’s marine economy is first constructed. Also, an integration EWM-HDEMATEL method that balances objective and subjective weighting is introduced to determine the index weights. The proposed methods are applied to analyze the temporal and spatial variations in the eco-innovation level of China’s marine economy in 2006–2021, and the following conclusions are drawn. First, between 2006 and 2021, the average level of marine economic ecological innovation across all regions exhibited a steady upward trajectory. Second, regional imbalances in ecological innovation intensified over the study period, with the maximum disparity widening from a 1.6-fold difference in 2006 to a 2.5-fold difference in 2021. Third, although differences among the three principal marine economic zones were not pronounced, significant heterogeneity persisted within each zone, underscoring the need for targeted policies and interventions to achieve coordinated development. Fourth, regions performed better on the support environment and performance dimensions of marine economic ecological innovation than on the capacity and activity dimensions. These findings identify critical leverage points for policy action and carry important implications for promoting the balanced and sustainable development of marine economic ecological innovation efficiency.

1. Introduction

China has experienced rapid economic growth over the past several decades as the world’s second-largest economy and a major maritime nation [1]. The marine economy has significantly contributed to China’s economic expansion, with gross ocean product (GOP) having reached CNY 9.9 trillion in 2023, accounting for 7.9% of the national GDP (the State Council of the People’s Republic of China, 2024) [2]. However, the rapid development of the marine economy has exerted significant pressure on marine resources and the environment; issues such as overfishing, overexploitation, and marine pollution have led to ecological degradation and the loss of biodiversity [3,4]. In response, Chinese authorities have implemented various measures to promote the transition of the marine economy towards greener and more ecological practices [5]. In the context of this study, based on multiple concepts of eco-innovation, a comprehensive evaluation of the eco-innovation efficiency of the marine economy is conducted, which is crucial for optimizing marine resource allocation, promoting the transformation of marine economic growth models, and finally contributing to achieving the sustainable development of the marine economy.

Eco-innovation refers to the form of innovation aimed at achieving sustainable development by reducing environmental burdens, enhancing adaptability to environmental pressures, or improving the efficiency of natural resource utilization through technological, managerial, and organizational innovations [6,7,8]. Under modern society’s complex and interrelated environmental challenges, eco-innovation has played a crucial role in driving the transition towards a green economy and society, becoming an essential pathway to achieving sustainable development goals [9,10]. Meanwhile, eco-innovation is increasingly recognized as an effective tool for measuring the synergistic development of the economy and environment [11,12]. Its driving factors and economic or environmental impacts have likewise been thoroughly analyzed [13,14,15,16,17].

Research on the measurement and evaluation of eco-innovation has reached a relatively mature stage. Existing studies have conducted in-depth analyses from multiple perspectives. Regarding research scope, scholars have explored the measurement and evaluation of eco-innovation from micro to macro levels, encompassing companies [18,19], industries [20,21], regions [22,23], and nations [24,25]. On the other hand, numerous methodological approaches have been extensively explored and applied in measuring eco-innovation. One widely adopted approach is the development of specific scales tailored to research needs. For example, García-Granero et al. [26] developed a multidimensional measurement method for corporate eco-innovation, encompassing product, process, organization, and marketing dimensions. Also, efficiency analysis methods are commonly used to evaluate eco-innovation, such as Data Envelopment Analysis (DEA) [27] and Slacks-Based Measure (SBM) [23]. Furthermore, subjective evaluation methods like the Analytical Hierarchy Process (AHP) [28] and DEMATEL [29] have also been employed to evaluate eco-innovation levels. In summary, although various methods have been applied to study eco-innovation across different scopes, research explicitly focusing on eco-innovation in the marine economy still needs to be expanded.

The primary objective of this study is to develop a theoretically grounded evaluation framework and to employ rigorous scientific methodologies for a comprehensive assessment of the ecological innovation efficiency of China’s marine economy. Firstly, a targeted evaluation index system is developed for eco-innovation in the marine economy, considering the various aspects of eco-innovation and the unique characteristics of the marine economy. After that, a combined objective and subjective evaluation method, the integration of the Entropy Weight Method and Hierarchical Decision Making Trial and Evaluation Laboratory (EWM-HDEMATEL) method, is established to assess the eco-innovation levels of the marine economy in China’s coastal regions. Finally, this study analyzes the spatial and temporal evolution of China’s marine economic eco-innovation levels. This will provide a theoretical basis for optimizing resource allocation and upgrading the marine economy, ultimately promoting sustainable development in China and worldwide. The major novelties and contributions of this paper are summarized as follows:

(1)

Based on the three-stage theoretical model of “basic”, “advance”, and “adaptation”, a systematic index system is constructed to assess the level of eco-innovation in the marine economy and explore eco-innovation from four dimensions: capacity, support environment, activities, and performance.

(2)

This paper proposes a new evaluation method combined with objective and subjective properties—the integration EWM-HDEMATEL method—which rebalances subjective and objective judgments while enhancing assessment efficiency.

(3)

A detailed spatiotemporal dynamic analysis of the eco-innovation levels in China’s marine economy is conducted from 2006 to 2021. It reveals the evolution trends and spatial distribution characteristics of marine economic eco-innovation levels from both overall and dimensional perspectives as well as from the perspective of the economic circle.

The rest of this paper is organized as follows. Section 2 describes the construction of the evaluation index system. Section 3 introduces the proposed method that includes the integration EWM-HDEMATEL method to determine the index weights and the calculation method of the evaluation value. Section 4 analyzes the index weights and the temporal and spatial variations in the eco-innovation level of China’s marine economy in 2006–2021. A discussion of the results is presented in Section 5. Finally, Section 6 concludes the paper and illustrates the main contributions of this study.

2. Construction of the Evaluation Index System

2.1. Theoretical Model

Horizon 2020 introduced the word “systemic” as a prefix of eco-innovation, which indicates that constructing the evaluation index system for the marine economic eco-innovation level based on the systemic principle is the right thing to do [30]. Drawing upon previous studies, a three-stage theoretical model for evaluating the marine economic eco-innovation level, including “basic”, “advance”, and “adaptation”, is introduced in this research, as shown in Figure 1.

The three stages are the main groupings of indices covering the key concepts and issues of marine economic eco-innovation. First, the “basic” stage includes the dimensions of eco-innovation capacity and eco-innovation supporting environment. Eco-innovation capacity measures the capacity of a region to realize and continue the sustainable development of the marine economy, while the eco-innovation supporting environment measures the quality of the supporting system for the sustainable development of the regional marine economy. Second, the “advance” stage includes the eco-innovation activity dimension, which measures the intensity of the initiatives taken by the region to achieve sustainable development of the marine economy. Third, the “adaptation” stage involves the eco-innovation performance dimension, which measures the level of the region’s current status in terms of sustainable development of the marine economy.

The “basic” stage is the foundation for the operation of subsequent stages, which can provide dynamism and sustained influence to the “advance” stage. The “advance” stage is the source of the content of the “adaptation” stage, which determines the level of status of the “adaptation” stage. The “adaptation” stage is the result of the combined effect of the previous two stages, and it also feeds back into the “basic” stage, providing continuous enhancement and optimization for the “basic” stage. Ultimately, a virtuous cycle of marine economic eco-innovation is formed when all three stages of the framework are successfully operated. Therefore, constructing the evaluation index system based on this three-stage theoretical model can provide a comprehensive and scientific perspective for understanding the eco-innovation level of China’s marine economy.

2.2. The Evaluation Index System

Based on the above analysis of the three-stage theoretical model, guided by the principles of comprehensiveness, scientificity, representativeness, and data availability, the index system for evaluating the eco-innovation level of China’s marine economy is constructed. This evaluation index system includes 22 secondary indices covering the four dimensions of the capacity, supporting environment, activity and performance of marine economic eco-innovation, as shown in

Table 1. Evaluation index system of eco-innovation level of China’s marine economy.

Dimension	Index	Description	Unit	Attribute
Eco-innovation Capacity	e1. Marine science and technology patents granted	Number of patents officially granted in the field of marine science and technology	Number	+
	e2. Share of the added value of marine scientific research, education, management and service	Share of marine scientific research, education, management and service in the value added of marine and related industries	%	+
	e3. Seawater desalination project scale	Total daily production of freshwater from desalination of seawater	Million tons	+
	e4. Charge for sea area use	Amount of royalties received by the government through the legal granting of the right to use the sea area	Million CNY	+
Eco-innovation Supporting Environment	e5. Scientific and technical staff in marine research institutions	Number of persons engaged in scientific and technological activities in marine scientific research institutions	Number	+
	e6. Coastal observation stations by coastal regions	Number of laboratories and systems for continuous observation of marine environment at the coastline or on offshore islands	Number	+
	e7. Green finance index	Performance of financial institutions or investment products in terms of environmental sustainability	Score	+
	e8. Strength of financial support	The proportion of government funds in intramural expenditure on R&D of marine R&D institutions	%	+
	e9. Awareness of sustainability management	Regulations, standards and policies formulated by the government in environmental protection and the strength of their implementation	Score	+
	e10. Losses from major marine disasters	Property damage caused by major maritime natural disasters to human production, life, and social development	Million yuan	-
Eco-innovation Activity	e11. R&D projects of marine R&D institutions	Number of marine science R&D projects conducted by marine R&D institutions	Number	+
	e12. Digital economy index	Performance of the digital economy in the region	Score	+
	e13. Seawater direct utilization	Direct adoption of seawater as a substitute for freshwater use in human production and life	Million tons	+
	e14. Offshore wind project scale	Total amount of electricity resources that can be produced by wind projects built offshore	Megawatts	+
	e15. Completed investment in treatment of industrial pollution	Amount of investment completed during the year in projects to treat industrial pollution	Million yuan	+
	e16. Marine protected areas	Number of marine protected areas set aside for the exclusive protection of marine resources, environment and ecology	Number	+
Eco-innovation Performance	e17. Gross output value of marine industries	Total value of marine products produced by resident units in the marine industry sector	Billion yuan	+
	e18. Industrial solid waste generation	Total solid waste generated in industrial production activities	Million tons	-
	e19. Total CO2 emissions	Total carbon dioxide emissions from major energy consumption in production and domestic use	Million tons	-
	e20. Wastewater discharged directly to the sea	Total amount of production and domestic wastewater discharged directly into the sea without adequate treatment	Billion tons	-
	e21. Health of marine ecosystems	State of health of marine ecosystems in the regional priority monitoring areas	Score	+
	e22. Primary and secondary water quality in nearshore waters	Proportion of the checkpoints of near-shore marine waters whose seawater quality meets national Class I and II standards	%	+

.

In the eco-innovation capacity dimension, the marine science and technology patents granted (e1) and the share of the added value of marine scientific research, education, management and service (e2) represent the scientific and technical capacity of marine economic eco-innovation, while the seawater desalination project scale (e3) and the charge for sea area use (e4) represent the management capacity. These indicators were selected to balance measurable output (e1–e2) with policy-driven management tools (e3–e4), ensuring both innovation and governance capacities are represented.

Social, policy, and natural supporting environments are the three main aspects under the eco-innovation supporting environment dimension. The social supporting environment could be characterized by the scientific and technical staff in marine research institutions (e5), the coastal observation stations by coastal regions (e6), and the green finance index (e7). The policy supporting environment could be indicated by the strength of financial support (e8) and the awareness of sustainability management (e9). The losses from major marine disasters (e10) is the main content of the natural supporting environment. Emphasis on institutional and policy factors (e8–e9) underscores the role of governance frameworks, while social and natural metrics (e5–e7, e10) contextualize innovation within broader environmental and societal support.

The eco-innovation activity dimension can be described through three aspects: science and technology innovation, resource development and pollution control. Science and technology innovation consists of the R&D projects of marine R&D institutions (e11) and the digital economy index (e12). Resource development is represented by the seawater direct utilization (e13) and the offshore wind project scale (e14). Completed investment in the treatment of industrial pollution (e15) and the marine protected areas (e16) represent pollution control. The combination of project-based (e11) and index-based (e12) measures alongside development and regulatory compliance indicators (e13–e16) provides comprehensive coverage of activity stages and policy impacts.

Economic, social, and ecological performances are the main elements of the eco-innovation performance dimension. The economic performance could be described by the gross output value of marine industries (e17). The social performance consists of the industrial solid waste generation (e18), the total CO₂ emissions (e19), and the wastewater discharged directly to the sea (e20). The health of marine ecosystems (e21) and the primary and secondary water quality in nearshore waters (e22) represent the ecological performance of marine economic eco-innovation. Linking output (e17) with environmental externalities (e18–e20) and ecosystem health (e21–e22) ensures that performance metrics reflect both economic gains and environmental sustainability.

3. Materials and Methods

3.1. Research Area and Data Sources

According to the outline of the 14th Five-Year Plan (2021–2025), there are three central marine economic circles in China: the Northern, Eastern, and Southern Marine Economic Circles (Figure 2). The three marine economic circles cover a total of 11 provincial administrative regions, and their combined gross regional product has accounted for more than 50 percent of China’s GDP for many consecutive years. The related annual data from 2006 to 2021 for the above 11 provinces are used in this paper to study the spatiotemporal dynamics of the eco-innovation level of China’s marine economy. The research data used in this paper are mainly derived from the China Marine Economy Statistical Yearbook, China Environment Statistical Yearbook, China Marine Statistical Yearbook, China Statistical Yearbook, and other relevant documents published by the Ministry of Ecology and Environment, the National Bureau of Statistics, the Ministry of Natural Resources, and local government departments.

3.2. Determination of Index Weights

Objective weighting and subjective weighting are the two main methods for determining index weights [31]. Objective weights, being derived from the indicator data, capture the inherent characteristics of the data but do not account for the relationships among indicators. Subjective weights are determined by expert experience, which are helpful in reflecting the impact relationships between indices, but they are susceptible to subjective arbitrariness. Therefore, as shown in Figure 3, an integration EWM-HDEMATEL method that balances objective and subjective weighting is introduced in this subsection to determine the index weights for evaluating the eco-innovation level of China’s marine economy.

3.2.1. Index Data-Driven Objective Weighting

Drawing on the concept of thermodynamics, the average amount of information excluding redundancy can be called information entropy [32]. According to this definition, the entropy value can be used to determine the degree of dispersion of a certain index. The smaller the value of information entropy, the greater the degree of dispersion of the index and the greater the impact (i.e., weight) of the index on the comprehensive evaluation [33]. EWM avoids the interference of human factors in the weighting process; thus, it can guarantee the objectivity of the index weights [34]. Therefore, in this subsection, the data of all indices will be processed using the EWM to obtain the objective weights.

Considering the differences in attributes in dimension and order of magnitude, it is necessary to standardize the initial data of the indices as follows:

$$o_{ikt}^{\prime }=\left\{\begin{array}{l}{\displaystyle \frac{o_{ikt}-\min \left\{o_i\right\}}{\max \left\{o_i\right\}-\min \left\{o_i\right\}}},{ o}_i \mathrm{is} \mathrm{a} \mathrm{positive} \mathrm{index}\\ {\displaystyle \frac{\max \left\{o_i\right\}-o_{ikt}}{\max \left\{o_i\right\}-\min \left\{o_i\right\}}},{ o}_i \mathrm{is} \mathrm{a} \mathrm{negative} \mathrm{index}\end{array}\right.$$

(1)

where $o_{ikt}$ ($i=1,2, \cdots , N$; $k=1,2, \cdots , K$; $t=1,2, \cdots , T$) represents the value of the $i_{th}$ index of the $k_{th}$ province in the $t_{th}$ year, and $\min \left\{o_i\right\}$ and $\max \left\{o_i\right\}$ are the minimum and maximum values of the $i_{th}$ index of all provinces in all years, respectively.

The proportion $p_{ikt}$ of the $i_{th}$ index of the $k_{th}$ province in the $t_{th}$ year is calculated as follows:

$$p_{ikt}={\displaystyle \frac{o_{ikt}^{\prime }}{{\displaystyle {\sum }_{k=1, t=1}{o}_{ikt}^{\prime }}}}$$

(2)

The value of information entropy $e_i$ of the $i_{th}$ index is calculated as follows:

$$e_i=-{\displaystyle \frac{1}{\ln (K\times T)}}\times {\displaystyle {\sum }_{k=1, t=1}(p_{ikt}\times \ln p_{ikt})}$$

(3)

It is obvious that $0\le e_i\le 1$, and, in particular, if $p_{ikt}$ equals 0, then $\ln p_{ikt}$ is defined to be equal to 0.

The value of the information entropy redundancy of the $i_{th}$ index is $1-e_i$; thus, the objective weights of each index $W^o$ can be obtained as follows:

$$W^o={\left[w_i^o\right]}_{1\times N}={\displaystyle \frac{1-e_i}{{\displaystyle \sum _{i=1}^N(1-e_i)}}}$$

(4)

3.2.2. Expert Experience-Driven Subjective Weighting

DEMATEL is a methodology that uses diagrams to analyze the interrelationships between different system components and identify critical influences [35]. It is known for its ability to solve sophisticated problems, particularly those that involve interactive elements. However, the original DEMATEL method was applied to simple systems with a limited number of factors but was not used not to identify the importance of a large number of factors in complex systems [36]. There are more than 20 indices in the index system for evaluating the eco-innovation level of the marine economy constructed in Section 2.2, and the complexity of the internal impact relationships of the indices increases the difficulty of implementing the DEMATEL method and, consequently, undermines its effectiveness. HDEMATEL is a good choice to solve this problem; its hierarchical decomposition rules and super IDR matrices can effectively deal with the influence relationships in complex index systems [37]. The steps to realize expert experience-driven subjective weighting using the HDEMATEL method are as follows.

Step 1. The hierarchical decomposition of the evaluation index system.

The purpose of hierarchical decomposition is to divide the complex evaluation index system into multi-level subsystems, so that the direct impacts between various indices in each subsystem are clear enough. According to Table 1, capacity, supporting environment, activity, and performance are the four main dimensions of the evaluation index system, so they can be considered as the four subsystems of the index system. There are 4–7 specific indices under each subsystem, and the impact relationships between them is easy to directly determine. As a result, as shown in Figure 4, the evaluation index system is decomposed into a structure of one level of subsystems to allow for a more efficient and accurate determination of the impact relationships between indices within the system.

Step 2. Direct influence analysis.

Based on the hierarchical structure of the evaluation index system constructed above, direct influence relationships may explicitly exist between indices within each subsystem, which can be directly judged by experts in the HDEMATEL method. Meanwhile, direct influence relationships between indices within different subsystems may be difficult to judge, so they will be dealt with in the next step rather than being judged directly by experts here.

As shown in Figure 4, the evaluation index system $F$ is classified into four subsystems, $F_q=\left\{f_i^q\left|i=1, 2, \dots , N_q\right.\right\}$, and $f_i^q$ is the $i_{th}$ index in $F_q$, $q=1, \cdots , 4$. It is obvious to find that $F=F_1\cup F_2\cup F_3\cup F_4=\left\{f_i\left|i=1,2, \cdots , N\right.\right\}$, $N={\displaystyle {\sum }_q{N}_q}$. In this structure, subsystems $F_1, \cdots , F_4$ are the elements within the system $F$, and indices $f_1^q, \cdots , f_{N_q}^q$ are the elements within subsystem $F_q$. Accordingly, the direct influence degrees should be judged by experts not only for the indices $f_n^q$ within subsystem $F_q$ but also for the subsystems $F_1, \cdots , F_4$ within system $F$.

Experts are invited to judge the direct influence degrees between each pair of elements with 0–4 scales, i.e., 0, 1, 2, 3, and 4 represent “no influence”, “low influence”, “medium influence”, “high influence” and “very high influence”, respectively. The direct influence degree on element $f_i$ impacting $f_j$ (denoted by $f_i\to f_j$) is given as $x_{ij}\in 0,1,2,3,4$, $i,j=1, \cdots , N$. The initial direct relation (IDR) matrix that describes direct influence degrees between all pairs of elements in the system is defined as $X={\left[x_{ij}\right]}_{N\times N}$, where $x_{ij}=0$ for $i=j$.

Step 3. The construction of the super IDR matrix.

The super IDR matrix, as the inputs of the hierarchical DEMATEL, is defined as a matrix that includes all the direct influence degrees between the elements of complex systems. In other words, in the super IDR matrix, the direct influence degrees of indices are required to be obtained for all pairs of indices in the system, not only the indices included in each subsystem but also the indices involved in different subsystems. The direct influence degree between indices on the same subsystem and different subsystems ${\overline{x}}_{ij}^{q{q}^{\prime }}$ can be summarized as follows:

$${\overline{x}}_{ij}^{q{q}^{\prime }}=\left\{\begin{array}{l}{\displaystyle \frac{x_{q{q}^{\prime }}}{{\displaystyle \sum _i{\displaystyle \sum _j{x}_{ij}^{q{q}^{\prime }}}}}}{x}_{ij}^{q{q}^{\prime }},q=q^{\prime }\\ {\displaystyle \frac{z_i^q{z}_j^{q^{\prime }}}{{\displaystyle \sum _i{\displaystyle \sum _j{z}_i^q{z}_j^{q^{\prime }}}}}}{x}_{q{q}^{\prime }},q\ne q^{\prime }\end{array}\right.i=1, \cdots , N_q,j=1, \cdots , N_{q^{\prime }}$$

(5)

where $x_{q{q}^{\prime }}$ is the direct influence degree on subsystems $F_q\to F_{q^{\prime }}$, $x_{ij}^{q{q}^{\prime }}$ is the direct influence degree on $f_i^q\to f_j^{q^{\prime }}$, and $f_i^q$ and $f_j^{q^{\prime }}$ are the indices included in two subsystems $F_q=\left\{f_i^q\left|i=1, \cdots , N_q\right.\right\}$ and $F_{q^{\prime }}=\left\{f_j^{q^{\prime }}\left|i=j, \cdots , N_{q^{\prime }}\right.\right\}$, respectively. $z_i^q$ and $z_j^{q^{\prime }}$ constitute the prominence that describes the importance of index $f_i^q$ in subsystem $F_q$ and index $f_j^{q^{\prime }}$ in subsystem $F_{q^{\prime }}$, respectively.

The IDR matrix on a pair of subsystems, $F_q\to F_{q^{\prime }}$, can be constructed as ${\overline{X}}_{q{q}^{\prime }}={\left[{\overline{x}}_{ij}^{q{q}^{\prime }}\right]}_{N_q\times N_{q^{\prime }}}$, $q,q^{\prime }=1, \cdots , 4$. Super IDR matrix $\overline{X}$ is the integration of IDR matrices on all pairs of subsystems and can be constructed as follows:

$$\begin{array}{ll}\overline{X}& ={\left[{\overline{x}}_{ij}\right]}_{N\times N}=\left[\begin{array}{ccc}{\overline{X}}_{11}& \cdots & {\overline{X}}_{14}\\ ⋮& \ddots & ⋮\\ {\overline{X}}_{41}& \cdots & {\overline{X}}_{44}\end{array}\right]\\ & =\left[\begin{array}{ccc}\left[\begin{array}{ccc}{\overline{x}}_{11}^{11}& \cdots & {\overline{x}}_{1N_1}^{11}\\ ⋮& \ddots & ⋮\\ {\overline{x}}_{N_{1}1}^{11}& \cdots & {\overline{x}}_{N_1{N}_1}^{11}\end{array}\right]& \cdots & \left[\begin{array}{ccc}{\overline{x}}_{11}^{14}& \cdots & {\overline{x}}_{1N_q}^{14}\\ ⋮& \ddots & ⋮\\ {\overline{x}}_{N_{1}1}^{14}& \cdots & {\overline{x}}_{N_1{N}_q}^{14}\end{array}\right]\\ ⋮& \ddots & ⋮\\ \left[\begin{array}{ccc}{\overline{x}}_{11}^{41}& \cdots & {\overline{x}}_{1N_1}^{41}\\ ⋮& \ddots & ⋮\\ {\overline{x}}_{N_{q}1}^{41}& \cdots & {\overline{x}}_{N_q{N}_1}^{41}\end{array}\right]& \cdots & \left[\begin{array}{ccc}{\overline{x}}_{11}^{44}& \cdots & {\overline{x}}_{1N_q}^{44}\\ ⋮& \ddots & ⋮\\ {\overline{x}}_{N_{q}1}^{44}& \cdots & {\overline{x}}_{N_q{N}_q}^{44}\end{array}\right]\end{array}\right]\end{array}$$

(6)

Step 4. The calculation of the weights for each index.

Based on the super IDR matrix obtained above, the normalized IDR matrix $\overline{M}$ of $\overline{X}$ is computed as follows:

$$\overline{M}={\left[{\overline{m}}_{ij}\right]}_{N\times N}={\displaystyle \frac{\overline{X}}{\max ({\max }_{1\le i\le N}{\displaystyle \sum _j{\overline{x}}_{ij},\epsilon +{\max }_{1\le j\le N}{\displaystyle \sum _i{\overline{x}}_{ij}}})}}$$

(7)

where $\epsilon$ is a non-Archimedean infinitesimal.

The total relation matrix $\overline{T}$ of $\overline{M}$ that reflects the degrees of both direct and indirect influences between indices is computed as follows:

$$\overline{T}={\left[{\overline{t}}_{ij}\right]}_{N\times N}=\overline{M}{(\overline{I}-\overline{M})}^{-1}$$

(8)

where $\overline{I}$ is the $N\times N$ identity matrix.

Suppose $\overline{r}$ and $\overline{d}$ denote the summation of rows and columns of the total relation matrix $\overline{T}$, respectively. According to $\overline{T}={\left[{\overline{t}}_{ij}\right]}_{4\times 4}$, ${\overline{r}}_i$ is the total influence given by index $f_i$ to other indices, $\overline{r}={\left[{\overline{r}}_i\right]}_{N\times 1}={\displaystyle {\sum }_j{\overline{t}}_{ij}}$, $i,j=1, \cdots , N$. ${\overline{d}}_i$ is the total influence received by index $f_i$ from other indices, and $\overline{d}={\left[{\overline{d}}_j\right]}_{1\times N}={\displaystyle {\sum }_i{\overline{t}}_{ij}}$, $i,j=1, \cdots , N$. ${\overline{r}}_i+{\overline{d}}_i$ is the prominence, indicating the degree of the importance of index $f_i$ in the evaluation index system. Finally, the subjective weights of each index $W^s$ can be obtained by normalizing the prominence of all indices as follows:

$$W^s={\left[w_i^s\right]}_{1\times N}={\displaystyle \frac{({\overline{r}}_i+{\overline{d}}_i)}{{\displaystyle \sum _{i=1}^N({\overline{r}}_i+{\overline{d}}_i)}}}$$

(9)

3.2.3. Integration of Subjective and Objective Weights

It has been proved that the weighted result of the EWM does not consistently represent the amount of information and significance of the indices. The HDEMATEL method, on the other hand, has significant advantages in characterizing the overall effect of indices in complex systems. The integration EWM-HDEMATEL method makes it possible to take into account both data characteristics and index impact relationships in the final weights. Therefore, the final weights of evaluation indices are calculated by integrating the objective weights with the subjective weights as follows:

$$W={\left[w_i\right]}_{1\times N}=\alpha w_i^o+(1-\alpha )w_i^s, i=1, 2, \cdots , N$$

(10)

The objective weights and subjective weights are regarded as equally important to balance the role of data characteristics and impact relationships of indices, so $\alpha =0.5$. It is obvious to find that ${\displaystyle {\sum }_i{w}_i}=1$ and $w_i\ge 0$ for $i=1, 2, \cdots , N$.

3.3. Calculation of the Evaluation Value

Based on the normalized index data shown in Equation (1) and the integration index weights shown in Equation (10), the eco-innovation level of the marine economy in the $q_{th}$ dimension $v_{kt}^q$ of the $k_{th}$ province in the $t_{th}$ year can be obtained as follows:

$$v_{kt}^q={\displaystyle \sum _{i=1}^{N_q}{o}_{ikt}^{\prime }\times w_i}, k=1, 2, \cdots , K, t=1, 2, \cdots , T$$

(11)

where $N_q$ is the number of component indices for the $q_{th}$ dimension, $q=1, \cdots , 4$. The overall eco-innovation level of the marine economy $v_{kt}$ of the $k_{th}$ province in the $t_{th}$ year is

$$v_{kt}={\displaystyle \sum _{q=1}{v}_{kt}^q}={\displaystyle \sum _{i=1}^N{o}_{ikt}^{\prime }\times w_i}$$

(12)

For the temporal analysis, to explore the overall development and internal differences in the eco-innovation level of the marine economy, the mean $\overline{v_t}$ and coefficient of variation $CV$ for $v_{kt}$ are calculated as follows:

$$\overline{v_t}={\displaystyle \frac{1}{K}}{\displaystyle \sum _{k=1}^K{v}_{kt}}, t=1, 2, \cdots , T$$

(13)

$$CV={\displaystyle \frac{\sigma }{\overline{v_t}}}\times 100, t=1, 2, \cdots , T$$

(14)

where $\sigma$ is the standard deviation of the dataset. $\overline{v_t}$ represents the central tendency of the eco-innovation level of the marine economy in the $t_{th}$ year for each province, while $CV$ quantifies the extent of variability relative to the mean, thereby shedding light on the degree of heterogeneity among the provinces.

For the spatial analysis, the Natural Breaks Classification (NBC) method is introduced to further process the eco-innovation levels of the marine economy of various regions. Specifically, the eco-innovation level of the marine economy is classified into five categories using the NBC method: low level, relatively low level, medium level, relatively high level, and high level. The Jenks optimization algorithm is employed to determine the optimal breaking points for the classification.

4. Results

4.1. Index Weight Analysis

As presented in Section 3.2, the index weights for evaluating the eco-innovation level of China’s marine economy were determined by the integration EWM-HDEMATEL method. First, based on the negative correlation between the value of information entropy and index weights, index data-driven objective weighting was realized with the EWM. Second, based on the positive correlation between the total influence degree in the system and index weights, expert experience-driven subjective weighting was realized with HDEMATEL. Third, the final weights of evaluation indices were obtained according to the assumption that the objective weights and subjective weights are equally important.

Table 2. Results of objective weighting driven by index data.

Index	${\mathit{e}}_{\mathit{i}}$	${\mathit{w}}_{\mathit{i}}^{\mathit{o}}$	Index	${\mathit{e}}_{\mathit{i}}$	${\mathit{w}}_{\mathit{i}}^{\mathit{o}}$
e1	0.8545	0.0762	e12	0.9715	0.0149
e2	0.9579	0.0221	e13	0.8113	0.0989
e3	0.7619	0.1248	e14	0.5578	0.2317
e4	0.9040	0.0503	e15	0.9309	0.0362
e5	0.9412	0.0308	e16	0.8860	0.0597
e6	0.9317	0.0358	e17	0.9251	0.0392
e7	0.9824	0.0092	e18	0.9884	0.0061
e8	0.8932	0.0560	e19	0.9873	0.0067
e9	0.9505	0.0259	e20	0.9939	0.0032
e10	0.9973	0.0014	e21	0.9838	0.0085
e11	0.9076	0.0484	e22	0.9734	0.0139

shows the results of index data-driven objective weighting using the EWM. The information entropy value of the offshore wind project scale (e14) is the smallest among all indices, and thus its objective weight value, 0.2317, is the largest. Meanwhile, the seawater desalination project scale (e3) is another index with an objective weight value greater than 0.1. On the contrary, the information entropy value of losses from major marine disasters (e10) is the largest among all indices; and therefore, its objective weight value is the smallest, only 0.0014. The green finance index (e7), industrial solid waste generation (e18), total CO2 emissions (e19), wastewater discharged directly to the sea (e20), and health of marine ecosystems (e21) are also indices with an objective weight value less than 0.01. The dispersion degrees of data for other indices are moderate, and their objective weight values are between 0.01 and 0.1.

Based on the hierarchical decomposition of the evaluation index system in Section 3.2.2, five qualified experts were invited to judge the direct influence relationships between the internal elements of the two levels, including two scholars with a long career in eco-innovation research, two managers of enterprises related to the maritime economy, and one government employee familiar with the development of the maritime economy. Although the expert panel comprised only five individuals, such sample sizes are relatively common in the recent literature and have been demonstrated to yield valid results [38,39]. The entire judgment process was guided by the Delphi Method, and the final IDR matrices that describe direct influence degrees between all pairs of elements in the system are shown in Figure 5.

According to Equations (5)–(9), based on the IDR matrices obtained above, the results of expert experience-driven subjective weighting using HDEMATEL are shown in

Table 3. Results of subjective weighting driven by expert experience.

Index	$\mathbf{(}{\overline{\mathit{r}}}_{\mathit{i}}\mathbf{+}{\overline{\mathit{d}}}_{\mathit{i}}\mathbf{)}$	${\mathit{w}}_{\mathit{i}}^{\mathit{s}}$	Index	$\mathbf{(}{\overline{\mathit{r}}}_{\mathit{i}}+{\overline{\mathit{d}}}_{\mathit{i}}\mathbf{)}$	${\mathit{w}}_{\mathit{i}}^{\mathit{s}}$
e1	4.8226	0.0616	e12	2.9813	0.0381
e2	5.2899	0.0675	e13	2.8172	0.0360
e3	2.2587	0.0288	e14	3.0116	0.0384
e4	4.2669	0.0545	e15	3.9871	0.0509
e5	3.0583	0.0390	e16	4.1999	0.0536
e6	3.3317	0.0425	e17	4.1536	0.0530
e7	3.5116	0.0448	e18	2.8337	0.0362
e8	3.7754	0.0482	e19	2.9032	0.0371
e9	3.8288	0.0489	e20	2.7611	0.0352
e10	3.0648	0.0391	e21	3.4864	0.0445
e11	4.4558	0.0569	e22	3.5423	0.0452

. The prominence of the share of the added value of marine scientific research, education, management and service (e2) is the largest among all indices, and accordingly, its subjective weight value is the largest, 0.0675. The following indices with relatively high prominence are marine science and technology patents granted (e1), R&D projects of marine R&D institutions (e11), charge for sea area use (e4), marine protected areas (e16), the gross output value of marine industries (e17), and completed investment in the treatment of industrial pollution (e15), and accordingly, their subjective weight values are all greater than 0.05. The extremely small prominence of the seawater desalination project scale (e3) indicates the low degree of importance it has in the evaluation index system, which suggests that its subjective weight value 0.0288 is the smallest among all indices.

The index weights describe the importance of the indices in the evaluation of the eco-innovation level of the marine economy. The larger the weight of an index, the more important role of the index in the evaluation process; when the weight of an index is low, the index is accordingly considered insignificant. Figure 6 presents the final index weights after the fusion of the objective and subjective weights described above. At the dimension level, the weights are ranked in the following order: activity (0.38) > capacity (0.24) > supporting environment (0.21) > performance (0.16). This suggests that eco-innovation activity is the primary dimension in the evaluation of the eco-innovation level of the marine economy, while eco-innovation performance is of relatively minimal importance. At the index level, the seawater desalination project scale (e3), strength of financial support (e8), offshore wind project scale (e14), and gross output value of marine industries (e17) are the most heavily weighted indices of the four dimensions, respectively. Among them, the offshore wind project scale (e14) is the most highly weighted of all the indices and the only one with a weight value greater than 0.1. There are four indices with a weight value of 0.02, losses from major marine disasters (e10), industrial solid waste generation (e18), total CO₂ emissions (e19), and wastewater discharged directly to the sea (e20), and they are also the indices with the smallest weight values of all the indices.

4.2. Temporal Analysis

Based on the data and weights of the indices mentioned above, the evaluation value of the eco-innovation level of China’s marine economy was calculated using the method introduced in Section 3.3. The detailed evaluation results for each region from 2006 to 2021 can be found in

Table A1. The regional eco-innovation level of China’s marine economy.

	The Northern Marine Economic Circle					The Eastern Marine Economic Circle				The Southern Marine Economic Circle
	Liaoning	Hebei	Tianjin	Shandong	Average	Jiangsu	Shanghai	Zhejiang	Average	Fujian	Guangdong	Guangxi	Hainan	Average
2006	0.2042	0.1624	0.1578	0.2159	0.1851	0.1892	0.1804	0.1741	0.1812	0.1964	0.2520	0.1901	0.2398	0.2196
2007	0.1733	0.1636	0.1563	0.2233	0.1791	0.2408	0.1958	0.1691	0.2019	0.1980	0.2770	0.2038	0.2187	0.2244
2008	0.1720	0.1621	0.1616	0.2711	0.1917	0.1926	0.1955	0.1767	0.1882	0.1969	0.2651	0.1915	0.2092	0.2157
2009	0.1676	0.1354	0.1564	0.2221	0.1704	0.1880	0.1993	0.1513	0.1796	0.1956	0.2657	0.1786	0.1929	0.2082
2010	0.1566	0.1282	0.1466	0.2159	0.1618	0.1841	0.1939	0.1401	0.1727	0.1757	0.2593	0.1640	0.1856	0.1961
2011	0.1567	0.1368	0.1519	0.2325	0.1695	0.1956	0.2134	0.1495	0.1862	0.1752	0.2545	0.1609	0.2016	0.1981
2012	0.1900	0.1609	0.2087	0.2590	0.2047	0.2041	0.2156	0.1971	0.2056	0.1883	0.3237	0.1657	0.2206	0.2246
2013	0.1945	0.1528	0.1632	0.2761	0.1967	0.2152	0.2293	0.2223	0.2222	0.2093	0.3137	0.1759	0.2065	0.2263
2014	0.2507	0.2323	0.2201	0.3918	0.2737	0.2249	0.2564	0.3034	0.2616	0.2226	0.3407	0.1729	0.2274	0.2409
2015	0.2563	0.1909	0.2256	0.3233	0.2490	0.2446	0.2746	0.2693	0.2628	0.2591	0.3834	0.1857	0.2085	0.2592
2016	0.2681	0.1775	0.2083	0.3673	0.2553	0.2434	0.2574	0.2786	0.2598	0.2515	0.3982	0.1806	0.2147	0.2613
2017	0.2256	0.1842	0.2126	0.3704	0.2482	0.2517	0.2576	0.2772	0.2622	0.2608	0.4262	0.1755	0.2176	0.2700
2018	0.2456	0.2699	0.2733	0.3985	0.2968	0.3086	0.2616	0.3191	0.2964	0.2820	0.4885	0.2127	0.2413	0.3061
2019	0.2622	0.2557	0.2524	0.4160	0.2966	0.3297	0.2644	0.3465	0.3135	0.2634	0.4984	0.2080	0.2368	0.3016
2020	0.2582	0.2617	0.2450	0.4399	0.3012	0.3462	0.3013	0.3753	0.3410	0.2778	0.5681	0.2323	0.2525	0.3327
2021	0.2984	0.2948	0.2480	0.4848	0.3315	0.4093	0.3217	0.4015	0.3775	0.3183	0.6113	0.2421	0.2533	0.3562

in Appendix A. In this subsection, an overall temporal variation analysis is performed first, followed by a temporal variation analysis for the evaluation dimensions.

4.2.1. Overall Temporal Variation Analysis

As shown in Figure 7, the average eco-innovation level of the marine economy in each region of China shows an overall steady upward trend over the period 2006–2021. The year 2010 marks an important turning point in the evolution of China’s overall marine economic eco-innovation level. Before 2010, the average eco-innovation level of the marine economy is still in a state of continuous slight decline; after 2010, the average eco-innovation level of the marine economy enters into a state of significant and continuous increase. Specifically, 2012, 2014, and 2018 are the three years with the most significant growth trends, and they all grow at an annual rate of more than 15%. The average eco-innovation level of China’s marine economy reaches a maximum value of 0.353 in 2021.

However, the coefficient of variation of the eco-innovation level of the marine economy among regions also shows an upward trend over the period 2006–2021, which implies that the imbalance of development between regions is in a growing trend. During the research period, 2011, 2015, and 2018 are the only three years that achieve negative growth in the coefficient of variation. However, 2007, 2014, 2016 and 2020 are the years in which the growth of the coefficient of variation exceeded 10%, being the years in which the growth of imbalance in the development of the marine economic eco-innovation level between the regions is the most severe. In 2021, the coefficient of variation reaches a peak value of 0.3247.

Figure 8 shows the temporal variations in the regional eco-innovation level of China’s marine economy. As shown by the trend of the average values above, the eco-innovation levels of the marine economy in all regions generally first experience a downward trend from 2006 to 2010 and then enter an upward trend from 2010 to 2021. The marine economic eco-innovation level of Guangdong has been the highest over the research period, with the exception of 2008 and 2014. Shandong has always been in the top three, even surpassing Guangdong in 2008 and 2014 to rank first. In addition, it is obvious that the marine economic eco-innovation level varies considerably between regions, and there is an increasingly significant evolutionary trend in this variation. In 2006, the marine economic eco-innovation level in the most advanced region was 1.6 times that of the least advanced region, while it was 2.5 times greater in 2021.

4.2.2. Dimensional Temporal Variation Analysis

Figure 9 illustrates the dimensional temporal variations in the eco-innovation level of China’s marine economy. Overall, the difference in the contribution of the four dimensions in the evaluation of the eco-innovation level of China’s marine economy varies from large to small over time. In 2006, the difference in the share of the contribution between the most important and least important dimensions amounts to 37% compared to only 8% in 2021.

Specifically, the contribution of eco-innovation performance to the evaluation results was the highest in the period 2006–2020 but has shown an overall decreasing trend. Eco-innovation supporting environment has the next highest contribution, which has remained relatively stable since 2009, and overtook eco-innovation performance as the dimension with the highest contribution in 2021. The contribution rate of eco-innovation activity shows a continuous growth trend as a whole, exceeding 20% in 2012 and even reaching 26% in 2017. Eco-innovation capacity is the dimension with the relatively lowest contribution rate in the evaluation of the eco-innovation level of China’s marine economy. Its contribution rate was only at about 10% during 2006–2013 and has increased significantly since 2014 and remained at about 20%.

4.3. Spatial Analysis

As with Section 4.2, the results presented in this subsection are also calculated according to the method introduced in Section 3.3. In the following, an overall spatial variation analysis is performed first, followed by a spatial variation analysis of the evaluation dimensions.

4.3.1. Overall Spatial Variation Analysis

In order to compare the spatial differences in eco-innovation levels of the marine economy between different regions more intuitively, the ArcGIS (version10.8) software was used to create maps of eco-innovation levels for each region in 2006, 2011, 2016 and 2021, with a 5-year interval between each map. The resulting maps are shown in Figure 10.

On the whole, there are obvious differences in the marine economic eco-innovation level among the regions within each of the three main marine economic circles, which indicates that there is a serious imbalance of development within the region. Specifically, Guangdong is the most advanced region in terms of marine economic eco-innovation, and it is at a high level in all four of the above years. Shandong is the most shining area in the Northern Marine Economic Circle, and its marine economic eco-innovation level is stable at a relatively high level and high level. The situation in Jiangsu and Zhejiang within the Eastern Marine Economic Circle is good, with Jiangsu consistently at a medium level and relatively high level, while Zhejiang has made rapid progress, evolving from a relatively low level in 2006 and 2011 to a relatively high level in 2016 and 2021. Hainan is the region with the most serious regression in the marine economic eco-innovation level, from a high level in 2006 to a relatively high level in 2011, to a relatively low level in 2016, and finally to a low level in 2021. Also regressing severely is Guangxi, from a medium level in 2006 to a low level in 2021. Liaoning’s status is less stable, with a relatively high level in 2006 and 2016 but a relatively low level in 2011 and 2021. In addition, in all four of the above years, the marine economic eco-innovation level of Shanghai and Fujian are basically at a medium level and above, while that of Hebei and Tianjin are basically at a medium level and below.

4.3.2. Dimensional Spatial Variation Analysis

Figure 11 presents the spatial differences among various dimensions of the eco-innovation level of China’s marine economy in 2005, 2010, 2015, and 2020. This figure provides an intuitive representation of the comparative advantages and disparities among different dimensions across regions at different time points. AN, AS, and AE represent the average values for the Northern, Southern, and Eastern Marine Economic Circles, respectively.

Overall, the marine economic eco-innovation levels of each region in the supporting environment dimension and performance dimension are better than that in the capability dimension and activity dimension. Moreover, the level differences between regions are not significant in the supporting environment dimension and performance dimension, while it is quite remarkable in the capability dimension and activity dimension, which are the main sources of development imbalances between regions.

Specifically, in the dimension of eco-innovation capacity, the Northern Maritime Economic Circle shows the most pronounced increase in level among the three marine economic circles, from the lowest level in 2006 to the highest level in 2016 and 2021. Shandong, Guangdong and Zhejiang are consistently the top three in the eco-innovation capacity dimension. In the dimension of eco-innovation supporting environment, the eco-innovation level of each region first declined in the period of 2006–2011 and then maintained a continuous increase in the period of 2011–2021. As for the eco-innovation activity dimension, Shandong and Guangdong have always been two of the regions with the highest level, while Tianjin and Hebei have always been the two with the lowest level. It is worth noting that Jiangsu’s level in the eco-innovation activity dimension has been on an upward trend, even once overtaking Shandong to rank second in 2021. In the dimension of eco-innovation performance, the Southern Marine Economic Circle has always been the highest level of the three marine economic circles. In particular, on the eco-innovation performance dimension, not only are there small differences in relative levels between regions, but the regions themselves are very stable in absolute levels.

5. Discussion

In this study, the EWM-HDEMATEL method was utilized to determine the indicator weights for the evaluation system of China’s marine economic eco-innovation levels. The spatiotemporal changes in major coastal provinces of China from 2006 to 2021 were assessed. Through this analysis, we provide new insights into the constituent factors, development levels, and future policy recommendations for China’s marine economic eco-innovation.

Regarding indicator system, this paper integrates the characteristics of the marine economy and, based on existing studies of eco-innovation evaluation [12,40], constructs an indicator system tailored to the eco-innovation assessment of China’s marine economy. From the dimension weight results, eco-innovation activities emerge as the primary aspect of eco-innovation in the marine economy, which is consistent with the findings of Cai and Li [41]. This is likely because these activities are not only the direct means of achieving eco-innovation outcomes but also reflect the intensity of policy support and investment efforts. The indicator weight results show that the wind power project scale significantly leads to comprehensive weight. Also, the seawater desalination scale is more critical in objective weight, while the added value of marine science research, education, management, and services is the most important in subjective weight. This indicates the importance of clean energy use and water resource protection in eco-innovation and reflects the long-term and indirect contribution of research and education services [24,42].

From a temporal perspective, this study confirms the overall upward trend in eco-innovation levels across Chinese regions from the marine economic industry perspective [43]. The empirical results indicate that 2012, 2014, and 2018 saw the most significant growth in eco-innovation levels. This growth may be attributed to policy shifts, such as the emphasis on “ecology first, green development” for marine economic development in 2012, the inclusion of “new development concepts, ecological civilization, and building a beautiful China” in the 2018 Constitution, as well as various government policy documents promoting environmental protection [44], collectively substantiating the decisive role of policy in promoting ecological innovation within the marine economy. However, the development of eco-innovation levels is uneven across different regions, with a clear differentiation trend. This indicates the need to focus on slower-developing regions and implement targeted policies and interventions to narrow the gap and achieve synergistic development of eco-innovation efficiency in the marine economy across regions. From the perspective of dimension contribution rates, the four dimensions of capacity, support environment, activities, and performance show a converging trend over time. This suggests that balanced development across all four dimensions of eco-innovation efficiency in the marine economy should be emphasized.

Regarding spatial differences, some regions, such as Guangdong and Shandong, consistently demonstrate high levels of eco-innovation in the marine economy, while others, such as Guangxi and Hainan, perform poorly. Additionally, certain regions exhibit significant fluctuations in their rankings. This further illustrates the evident imbalance and instability in the development of eco-innovation levels within China’s marine economy [45]. Notably, some regions considered to have high levels of eco-innovation in other studies, such as Jiangsu and Zhejiang [43], were rated only medium or relatively low in our early marine economic assessments. This discrepancy may be due to differences in evaluation index systems, which overlook these regions’ natural resources and economic development advantages. At the dimensional level, certain regions exhibit significant strengths in specific dimensions. For example, Guangdong excels in the eco-innovation activities dimension, Shandong stands out in the eco-innovation capacity dimension, and the Southern Marine Economic Circle leads in the eco-innovation performance dimension. These regional strengths emphasize the importance of leveraging regional advantages and addressing weaknesses to promote balanced and sustainable development of eco-innovation efficiency in the marine economy.

6. Conclusions

Considering the important role of eco-innovation in addressing the paradox of achieving endless economic growth with limited resources, it is urgent to provide a reliable basis for formulating, revising, or selecting sustainable marine economic development plans by scientifically and comprehensively evaluating the eco-innovation level of the marine economy. This paper analyzes the eco-innovation level of China’s marine economy in 2006–2021 by providing a systematic evaluation method, which will not only provide valuable insights into the sustainable development of China’s marine economy but will also contribute to the exploration of sustainable economic growth in the world.

The main contributions of this paper can be summarized as follows. First, this study develops an evaluation indicator system for China’s marine economic ecological innovation by integrating the characteristics of China’s marine economy with a three-stage theoretical model of ecological innovation. Second, it introduces an EWM–HDEMATEL integration method that balances subjective and objective weights, enhancing objectivity and accuracy while improving efficiency, to determine the indicator weights scientifically. Third, it analyzes the spatiotemporal evolution of China’s marine economic ecological innovation levels from 2006 to 2021, thereby identifying strategic directions for achieving sustainable marine economic development and contributing to global assessments of ecological innovation and marine economic growth.

However, owing to data limitations, this study was unable to evaluate certain regions or include all pertinent indicators. Moreover, the subjective–objective hybrid evaluation approach employed retains a degree of subjectivity, and the underlying determinants were not examined in depth. Therefore, future research should refine and extend the methodological framework, investigate the external drivers, internal incentives, and policy measures affecting marine economic ecological innovation efficiency, and situate these analyses within a broader global context.