Spatiotemporal Evolution of Coupling Coordination Degree Between Economy and Habitat Quality in the Shandong Peninsula Urban Agglomeration: Grid Scale Based on Night-Time Lighting Data

Abstract

The process of social globalization and urbanization has developed rapidly in China, and the tension between economic development and the eco-environment is becoming increasingly tense, posing a major challenge to the sustainable development strategy of the Shandong Peninsula Urban Agglomeration (SPUA). Coordination development between economic development and habitat quality has become essential for preserving ecological stability and advancing long-term regional sustainability. This study constructed the optimal regression model to measure GDP density using night-time lighting data and economic statistical data and calculated habitat quality at the grid scale with the InVEST model. The spatiotemporal dynamics and driving factors of the coupling coordination between economy and habitat quality (EHCCD) were revealed using the coupling coordination degree model and the Geo-detector model. The results show that (1) between 2000 and 2020, the spatial pattern of GDP density has evolved from a single-core to a multi-core networked development. (2) The habitat quality of the SPUA exhibited a spatial pattern high in the east and low in the west, showing a downward trend. (3) The synergistic effect between GDP density and habitat quality was strengthened continuously, showing an overall strengthening tendency. (4) Driving factors’ influence on the EHCCD showed evident differences; socio-economic factors such as built-up area especially had greater explanatory power for the EHCCD; the interaction factors had shifted from socio-economic dominance to synergistic dominance of natural and human factors. This study not only overcomes the limitations imposed by administrative boundaries on assessing inter-regional coupling coordination but also provides fundamental data support for cross-regional cooperation, thereby advancing the sustainable development goal of the SPUA.

1. Introduction

China has achieved rapid economic development by setting clear economic growth targets. By 2024, the GDP reached CNY 134 trillion, with a year-on-year growth rate of 5%, accounting for 17% of global GDP. This rapid economic development has significantly contributed to social progress in China, but it has also inevitably led to high population concentration [1], urban expansion [2], and a dramatic increase in human activities [3,4]. The Shandong Peninsula Urban Agglomeration is a key region for economic development in eastern China. While rapid urbanization and industrialization have driven economic growth, they have simultaneously degraded ecosystem services and threatened habitat quality. Consequently, achieving coordinated development between economic development and habitat quality has become a critical issue for sustainability, both within the study area and globally.

The current relationship between economic development and the eco-environment mainly focuses on themes such as regional economic development and urbanization. Xie et al. [5] studied the coordinated relationship between the ecological environment and economic development from the perspective of industrial parks, emphasizing that the high consumption of resources poses a great threat to the ecological balance. Lei et al. [6] utilized the coupling coordination degree model to explore the coupling coordination between urbanization and the ecological environment, revealing the differences between economic growth and ecological protection. Fang et al. [7] utilized the EKC model to explore the relationship between the social economy and ecological environment of Hainan Island, emphasizing the significance of sustainable development in social economy and ecological governance. With ecological systems facing severe challenges and the growing imbalance between socio-economic progress and environmental protection, there is an urgent need for systematic research on the economic–ecological interplay to establish a theoretical foundation for sustainable development strategies. Schirpke et al. [8] conducted a comprehensive assessment of key ecosystem services in Alpine mountain lakes, revealing the complex relationship between critical ecological services, including water provisioning and biodiversity maintenance and regional socio-economic contexts. Their findings highlight the necessity of balancing ecological conservation with economic development within a sustainable framework. However, most of the existing studies have focused on the ecological impacts of macroscopic regions or single economic factors (such as urbanization and economy) and have paid less attention to the dynamic coupling mechanism between the economy and habitat quality.

The existing research on economic development mainly builds an indicator system based on economic statistical data and studies economic development at spatial scales such as countries, provinces, cities, and counties. Islamutdinov [9] conducted an integrated analysis of regional development practices across Europe, America, and Central Asia employing the entropy method, systematically revealing both common patterns and distinctive characteristics among different regional development models while establishing a novel methodological framework for reconciling regional economic growth with ecological sustainability. Kwaku Ohene-Asare et al. [10] analyzed the bidirectional relationship between energy efficiency and economic development through a three-stage framework based on statistical data from 46 African countries. They confirmed the virtuous cycle mechanism of energy efficiency improvement and economic growth, providing empirical evidence for African countries to promote sustainable development through energy technology innovation. Han et al. [11] analyzed the spatiotemporal characteristics of China’s digital economy and eco-environment by taking each province as the research scale. Lv et al. [12] took the cities of the Central Yangtze River urban agglomeration as the research area and proposed an innovative framework to construct an index system for the level of the digital economy. Yang et al. [13] established an indicator system and employed the entropy method to analyze socio-economic development trends, using Mei county as a case study. With the progress of remote sensing technology, many scholars have analyzed the economy using night-time lighting data and achieved some results [14,15,16,17]. However, there are some problems in using only night-time lighting data to study economic development. For instance, night-time lighting data may overlook remote or sparsely populated villages with less lighting, and thus cannot accurately measure economic development. To address this limitation and accurately represent the economic development process, this study fits the night-time lighting data with economic statistical data to calculate the GDP density at the grid scale. This method breaks the administrative boundary and realizes the dynamic change of GDP density.

In the process of economic development, human intervention has brought about profound changes in habitat quality and is also regarded as one of the most direct threats to regional habitat quality. In early research, habitat quality assessment primarily relied on the data obtained from field investigations to construct an indicator system [18,19]. However, compared with the comprehensive index evaluation method, model assessment has significant advantages in the long-term monitoring of habitat quality. Some of the more common models include the HSI model [20,21], the SoIVES model [22,23,24], and the InVEST model [25,26,27]. Among numerous evaluation models, the InVEST model has been increasingly recognized and adopted by scholars in the assessment of habitat quality at spatiotemporal scales due to its advantages such as high accuracy, convenient visualization, and easy data acquisition. The InVEST model mainly focuses on the following two aspects: First, assessing the habitat quality of habitats and studying the effects of human-induced pressures and climatic changes influencing habitat quality. Second, studying the dynamic changes in habitat quality and its affecting factors, for instance land use change [28,29,30], soil erosion [31], and driving-factor analysis [32]. Current research mainly analyzes the impact on habitat quality from the perspective of land planning, while the dynamic coupling relationship between habitat quality and the economy still requires further study.

The SPUA is a key area for the development of Shandong Province, which was the first comprehensive pilot zone in China to be approved for a regional–national development strategy. In recent years, with the intensification of economic development activities such as port expansion and industrial park construction, environmental issues such as coastal sea pollution, wetland shrinkage, and decline in biodiversity have become increasingly prominent. Yet it is precisely against this backdrop of development that this region is confronted with a typical economic development and ecological contradiction during the rapid urbanization process: on the one hand, as an economic uplift zone along the coast, it plays an important role in economic growth; on the other hand, its dense population and industrial layout have made the ecosystem particularly vulnerable in the SPUA. And the coupling coordination degree between economy and habitat quality (EHCCD) remains limited, and existing studies primarily focus on qualitative or quantitative analyses of coupling coordination at administrative division levels, lacking studies at the grid scale. To address these research gaps, the primary objectives of this study have been established as follows: (1) This study constructed the optimal regression model to measure GDP density using night-time lighting data and economic statistical data and calculated habitat quality with the InVEST model. (2) The coupling and coordination relationship between economy and habitat quality was evaluated using the optimized coupling and coordination model, and the driving factors were explored. This study explores the key drivers of the economic–environmental quality coupling and coordination relationship. This paper provides a theoretical basis for the SPUA to formulate space control strategies, achieving coordinated optimization of regional economic growth and the maintenance of ecosystem service functions.

2. Materials and Methods

2.1. Study Area

The SPUA, which borders the Beijing–Tianjin–Hebei region to the north (Figure 1), is an important opening gateway in northern China and an important growth pole in the Bohai Rim region. By the end of 2023, the total area of the research region was 1.567 × 10⁵ km², with a total GDP of CNY 9206.87 billion, ranking third in China, and the industrial structure is dominated by secondary and tertiary industries. Cultivated land, woodland, grassland, wetland, construction land, and unused land are the main land use types in the study area, with cultivated land covering the largest area. The SPUA is in the warm-temperate monsoon climate zone, with an average annual temperature of 14 °C and annual average precipitation of 710 mm. The coastline stretches for a total of 3504.74 km. The general terrain is characterized by high altitude in the central regions and low altitude in the surrounding regions, with hills and mountains dominating the topography, which is complex and diverse.

2.2. Data Sources

This study adopted a total of three periods of land use data of Shandong Province between 2000 and 2020, which were derived from Data Centre for Resource and Environment Science of the Chinese Academy of Sciences. The land use types were reclassified into cultivated land, woodland, grassland, wetland, construction land, and unused land, with a spatial resolution of 30 m. Economic data was derived from the Statistical Yearbook of Shandong Province in China. Night-time lighting data used DMSP/OLS night-time data products from the National Oceanic and Atmospheric Administration’s (NOAA’s) National Geophysical Data Centre for 1992–2013 and monthly average value of NPP/VIRS lighting data for 2012–2020.

In this paper, the raster data were projected into a unified coordinate system (WGS_1984_UTM_Zone_49N), with a spatial resolution of 1 km. ArcGIS was used to clip the data to match the scope of the study area, ensuring the consistency of the projection and the coverage of the data. The data types and sources are shown in the following table (

Table 1. Data types and sources.

Data Types	Sources
LULC	Resource and Environmental Science and Data Center http://www.resdc.cn/ (accessed on 10 May 2024)
Night-time lighting data	National Oceanic and Atmospheric Administration and National Geographical Data Center, NOAA/NGDC
Administrative Divisions	National Catalogue Service for Geographic Information https://www.webmap.cn/ (accessed on 6 May 2024)
DEM	Resource and Environmental Science and Data Center http://www.resdc.cn/ (accessed on 13 May 2024)
Precipitation	National Earth System Science Data Center
Temperature	National Qinghai-Tibet Plateau Scientific Data Center
Population	Shandong Statistical Yearbook
Gross domestic product
Total retail sales of consumer goods
Built-up area
Municipal road area
Green coverage rate
Land use intensity	Calculated from land use data
Cultivated land area

).

2.3. Research Methods

The balance between economy and habitat quality in the SPUA region was sought as the starting point and end point. This study constructed the optimal regression model to measure GDP density using night-time lighting data and economic statistical data and calculated habitat quality at the grid scale with the InVEST model. Then the coupling and coordination relationship between economy and habitat quality was evaluated using the optimized coupling and coordination model, and the driving factors were explored through the Geo-detector model. The specific research framework is shown in Figure 2.

2.3.1. Construction and Verification of the Optimized Regression Model

(1) Calculation of night-time lighting data

The night-time lighting index can accurately reflect the socio-economic conditions at the county or district level [33]. According to the actual situation of the SPUA, which mainly includes the four indicators total night light, average light index, light area ratio, and compounded night light index, the formula is

$$TNL={\displaystyle \sum _{i=1}^{D{N}_{max}}\left(D{N}_i{n}_i\right)}$$

(1)

$$I={\displaystyle \frac{TNL}{\left(D{N}_{max}{N}_L\right)}}$$

(2)

$$S={\displaystyle \frac{A_n}{A}}$$

(3)

$$CNLI=I\times S$$

(4)

where ${\mathit{DN}}_i$ denotes the attribute value of i image element in the administrative unit; $n_i$ denotes the number of attribute values corresponding to the i attribute value;${\mathit{DN}}_{\mathit{max}}$ denotes the maximum value of the image element in the administrative unit; $N_L$ is the total number of image elements in the region; S is the light area ratio; and $A_N$ and A are the overall lighted images in the region and the entire area of the region, respectively.

(2) Construction of the optimized regression model

The total night-time light (TNL), light area ratio (S), compounded night light index (CNLI), and relative average light intensity (I) were introduced, which were then correlated with the gross regional domestic product (GDP_all), the second (GDP₂), the third (GDP₃), and the sum of the gross product of the second and third industries (GDP₂₃) of the SPUA. Using GDP statistics (GDP_all, GDP₂, GDP₃, and GDP₂₃) and night-time lighting data (TNL, S, I, and CNLI) of each administrative region, the linear, logarithmic, power, and exponential models were constructed for regression analysis. The best night-time lighting index with the highest regression coefficient was selected for analysis, and a 1 km GDP grid was used for modeling to construct the GDP density. According to 

Table 2. Results of fitting the night-time lighting index to GDP (R²).

GDP all	TNL	I	S	CNLI	GDP2	TNL	I	S	CNLI
Linear	0.752	0.307	0.435	0.674	Linear	0.520	0.305	0.391	0.612
logarithmic	0.730	0.253	0.388	0.590	logarithmic	0.522	0.253	0.351	0.547
Exponential	0.728	0.204	0.545	0.694	Exponential	0.595	0.223	0.443	0.595
Power	0.759	0.166	0.495	0.653	Power	0.624	0.186	0.403	0.566
GDP3	TNL	I	S	CNLI	GDP23	TNL	I	S	CNLI
Linear	0.490	0.334	0.410	0.670	Linear	0.509	0.329	0.410	0.660
logarithmic	0.472	0.276	0.361	0.571	logarithmic	0.498	0.272	0.363	0.572
Exponential	0.646	0.208	0.553	0.711	Exponential	0.640	0.225	0.516	0.680
Power	0.651	0.169	0.495	0.655	Power	0.656	0.185	0.466	0.634

, it was found that the power model constructed by GDP_all and TNL has the highest coefficient of determination R², indicating that the power model has a better fitting effect on GDP_all and TNL. Therefore, the TNL that has the highest regression coefficient within the 95% confidence interval is chosen as the optimal night-time lighting index, and its power model is regarded as the optimal regression model. The formula is

Y = 945.59x^0.6445

(5)

(3) Verification of model accuracy

The simulated GDP for the 16 prefecture-level cities in the SPUA is calculated using the corresponding model, which is constructed with the night-time lighting index and GDP data from 2000, 2010, and 2020, and then compared with GDP statistical data in corresponding years. The numerical difference between the simulated GDP and the statistical GDP is verified by the relative error and the error coefficient. The relative error is calculated by the method of the difference between both the simulated GDP and the statistical GDP divided by the statistical GDP. The error coefficient is calculated by the statistical GDP divided by the simulated GDP.

The prediction equation has a small overall error in total (Figure 3). The relative error of 70% of the cities is below 10%, and that of Jinan, Qingdao, Zibo, and Weifang is more than 15% in 2000. Between 2000 and 2020, the difference in relative error of GDP shows that 70% of cities have an increasing relative error, mainly due to interference from economic factors. From 2000 to 2020, the average relative error of each administrative region decreased from 8.43% to 8.10%. The GDP output value of each administrative region can be calculated relatively accurately and has high reference value.

2.3.2. Measurement of Habitat Quality

The InVEST model evaluates habitat quality through examining land use changes and the degree of threat to biodiversity from different land types. It is widely used for habitat quality measurement and spatiotemporal dynamics [34,35,36]. Based on the actual situation of Shandong Province and relevant studies [37], cultivated land, urban land, rural land and road and other land types with intense human activities are selected as important threat factors. In InVEST model, the spatial decay type of the different threat factors, threat factor weights, and sensitivities used were obtained by combining the actual characteristics of the study area and referring to the previous studies [38,39]. The scientific validity of the weights of all of the factors has been verified in the previous studies. Habitat suitability refers to the suitability of each LULC class to provide a habitat for biodiversity, and HQ is directly related to the habitat suitability [40,41]. The range of habitat suitability is 0 to 1, and 1 represent the highest suitability. Therefore, with land use, threat factors, and sensitivity as the input data (

Table 3. Threat-factor weights.

Threat Factors	Weight	Maximum Distance of Influence/km	Decay Types
cultivated land	0.6	4	linear
urban	1	8	exponential
village	0.8	7	exponential
road	0.6	5	linear
railway	0.8	6	linear
unused	0.6	4	linear

and

Table 4. Sensitivity of threat factors.

Habitat Types	Suitability Score	Threat
		Cultivated Land	Urban	Village	Road	Railway	Unused
Cultivated land	0.4	0.3	0.5	0.7	0.5	0.5	0.2
Woodland	0.8	0.7	0.6	0.7	0.6	0.5	0.3
Grassland	0.7	0.5	0.6	0.6	0.4	0.4	0.2
Wetland	0.9	0.5	0.8	0.6	0.6	0.5	0.3
Construction land	0	0	0	0	0	0	0
Unused land	0	0.1	0	0.1	0.2	0.2	0

), the habitat quality in the SPUA was calculated by the InVEST model and divided into 5 grades—low, relatively low, medium, relatively high, and high—with one grade divided every 0.2. The specific calculation model is as follows [42,43]:

$$i_{rxy}=1-\left({\displaystyle \frac{d_{xy}}{d_{r\max }}}\right)\mathit{if}\mathit{linear}$$

(6)

$$i_{rxy}=\exp \left[-\left({\displaystyle \frac{2.99}{d_{r\max }}}\right)d_{xy}\right]\mathit{if}\mathit{exponential}$$

(7)

$$Q_{xj}=H_j\left[1-\left({\displaystyle \frac{D_{xj}^z}{D_{xj}^z+k_z}}\right)\right]$$

(8)

where $Q_{xj}$ represents the habitat quality index of grid x in land use type j, with a value range of [0, 1]. $D_{xj}$ denotes the level of stress to which grid x in land use type j is subjected, $H_j$ denotes the ecological suitability of land use habitat type j, and the K constant is the half-saturation constant, with a default K of 0.5, and z is defined as 2.5.

2.3.3. Coupling Coordination Degree Model

Because the traditional coupling degree C is not an average distribution function between 0 and 1, the optimized coupling coordination degree model disperses the C value as widely as possible between 0 and 1, increasing the variability of the C value, making the D value of the resultant coordinated development a more accurate reflection of the coupling coordination relationship and the measurement of the level of development. Based on previous studies [44], the optimized coupling coordination degree model was introduced by combining Then the coupling coordination degree between the GDP density and the habitat quality was calculated, and the spatiotemporal relationship between them was revealed.

Assuming max U_i is U₂, the equation is as follows:

$$C=\sqrt{\left[1-\sqrt{{(U_2-U_1)}^2}\right]\times {\displaystyle \frac{U_1}{U_2}}}$$

(9)

$$T={\alpha }_1{U}_1+{\alpha }_2{U}_2,{\alpha }_1+{\alpha }_2=1$$

(10)

U₁ represents the GDP density, U₂ denotes the habitat quality, U₁ U₂ ∈ [0, 1], and C is the degree of coupling between the habitat quality and the GDP density. The larger the value C, the more coordinated development the GDP density and habitat quality.

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

(11)

D is the degree of coupling coordination; the higher the value, the higher the level of development of system coupling coordination. α and β are the weighting coefficients. Since economy and habitat quality are equally important, the formula applied is α = β = 0.5, classified into 4 types and 6 subclasses (

Table 5. Level of the coupling coordination degree index.

Type	D Value Range	Subclass
Coordinated development	(0.8–1]	Coordination
Transformation development	(0.6–0.8]	Intermediate coordination
	(0.5–0.6]	Primary coordination
	(0.4–0.5]	Basic coordination
Uncoordinated development	(0.2–0.4]	Intermediate incoordination
	(0–0.2]	Extreme incoordination

).

2.3.4. Geo-Detector Model

We further investigated the driving factors that contribute to the differences in EHCCD. Based on existing research results [45,46,47,48], and taking into account the availability of data and the actual situation of the economy and habitat quality of the SPUA, we selected 9 influencing factors. Total retail sales of consumer goods reflect the vitality of the consumer market and economic strength. Expanded consumption may increase resource consumption, thereby affecting habitat quality and economic–ecological coupling coordination. Urban population density is a core indicator of urbanization, promoting the concentration of economic activities while exacerbating traffic and environmental pressures, serving as an intermediary factor linking economic agglomeration and ecological stress. Built-up areas embody urbanization expansion and economic spatial carriers, driving economic growth while encroaching on natural ecological spaces and leading to habitat fragmentation. Municipal road area reflects infrastructure development, supporting economic connectivity and growth; however, it may fragment ecological corridors and increase carbon emissions, negatively impacting habitat quality. Land use intensity measures land development efficiency. High intensity utilization enhances economic output, while overexploitation directly compresses ecological spaces and reduces habitat quality. Average temperature directly affects ecosystem stability. Economic activities may trigger temperature anomalies, and declining habitat quality can exacerbate local climate deterioration, forming a bidirectional feedback loop. Precipitation constrains vegetation, wetland ecological functions, and biodiversity. Water resource development in economic activities may alter the water cycle, and precipitation changes can also restrict economic activities, influencing coupling coordination. Green coverage rate is a core positive indicator of habitat quality. It improves ecological quality by purifying air and maintaining biodiversity while enhancing living environments to promote high quality economic development, acting as the link for the coordination of the two systems. Cultivated land area is the core of agricultural ecosystems. Reductions in its area threaten food security and biodiversity, while rational protection sustains ecological balance, serving as a key factor in coordinating agricultural economy and ecology. We explored the main driving factors and interaction between factors of EHCCD in the SPUA with Geo-detector. In the Geo-detector analysis, factors with q value of over 0.5 were identified as primary factors, and between 0.3 and 0.5 as secondary factors.

Geo-detector serves as a statistical method for examining geographic factor influences and spatial heterogeneity patterns [47,49,50]. Through the Geo-detector and specific analysis of the explanatory power of each driving factor, the contribution of each driving factor on the EHCCD in the SPUA can be identified quantitatively. The formula is as follows:

$$q=1-{\displaystyle \frac{{\displaystyle {\sum }_{h=1}^L{N}_h{\sigma }_h^2}}{N{\sigma }^2}}$$

(12)

where q is the detection value of the driving force of each influencing factor on the EHCCD; $\sum _{h=1}^L{N}_h{\sigma }_h^2$ is the cumulative sum of the variance of each sub-region of the detected factor; $N{\sigma }^2$ is the variance of the whole study area; the range of the value of q is [0, 1]; the bigger the value of q, the greater the influence of the driving factor on the EHCCD, and the smaller vice versa. When q equals 1, it indicates that the driving factor directly controls the EHCCD. When q equals 0, it indicates that the EHCCD is independent of this factor. By comparing the influence of the factors, the dominant factors of the divergence of the EHCCD can be detected.

2.3.5. Spatial Autocorrelation Model

The spatial autocorrelation model is used to study whether the observed values in an evaluation unit are geographically correlated with the observed values in its neighboring regions [37,51]. It can be better to analyze the GDP density in the SPUA of spatial clustering characteristics. Global Moran’s I assesses interregional spatial dependence, while local Moran’s I reveals localized clustering patterns, and can be classified into 5 spatial association types: high-high (H-H), high-low (H-L), low-high (L-H), low-low (L-L), and not-significant.

3. Results

3.1. Spatiotemporal Evolution Characteristics of GDP Density

The GDP density had steadily increased in the SPUA (Figure 4,

Table 6. Average GDP density of the SPUA.

City	2000	2010	2020	Average Annual Growth Rate in 20 Years
Jinan	3.241	4.068	5.478	3.45%
Zibo	2.811	3.946	5.372	4.55%
Taian	1.966	2.992	4.310	5.5%
Liaocheng	1.864	3.147	4.698	7.6%
Dezhou	2.115	2.958	3.695	3.75%
Binzhou	2.040	3.194	4.881	6.95%
Dongying	2.195	2.944	4.806	5.95%
Qingdao	3.287	4.237	6.337	4.65%
Yantai	2.449	3.273	5.120	5.50%
Weifang	2.435	3.578	5.120	5.50%
Weihai	2.946	4.152	4.888	3.30%
Rizhao	1.934	2.915	4.449	6.50%
Linyi	1.903	2.953	4.729	7.45%
Zaozhuang	2.823.59	3.970	5.421	4.60%
Jining	2.004.84	3.606	5.021	7.50%
Heze	1.387.66	2.926	4.530	11.30%
Provincial Capital Economic Circle	2.319.37	3.321	4.749	5.25%
Jiaodong Economic Circle	2.610.78	3.631	4.947	4.50%
Lunan Economic Circle	2.029.77	3.364	4.925	7.15%

Table Note: 10⁷ CNY/km².

), with the average GDP density rising from 2.33 × 10⁷ CNY/km² to 4.855 × 10⁷ CNY/km², an increase of nearly 1.08 times in twenty years. The average annual growth rate accelerated from 4.1% during 2000 to 2010 to 4.6% during 2010 to 2020, driving the GDP density to rise from 2.338 × 10⁷ CNY/km² to 4.855 × 10⁷ CNY/km² and demonstrating a pattern of planar diffusion. Heze, Linyi, and other late-developing regions had risen to prominence, and the growth rate of the Lunan Economic Circle had surpassed that of the Jiaodong region, resulting in a balanced and regional sustainable development.

GDP density shows obvious regional differences, with higher GDP density in eastern regions and lower GDP density in central and western regions. Economic development levels demonstrate marked disparities among municipalities, forming multiple economy centers and development poles. In 2000, the GDP density showed the characteristics of a single-core, with the high-value regions mostly located in Jinan and Qingdao, while the low-value areas were widely distributed. In 2010, the economic growth center was strengthened, the region began to develop in clusters, and the surrounding areas were radiated and driven. Economic development showed trends of agglomeration and diffusion. In 2020, it entered the multi-core networked development. Jinan-Qingdao and the Jiaodong economic circle formed high-value central areas. The southern part of the Lunan economic circle relied on Linyi and Jining to form a dual-node leap. However, there were still gradients in the northwest (Dezhou and Binzhou) and southwest (Heze) of Shandong.

From the perspective of urban agglomeration, Jiaodong Economic Circle exhibited the highest GDP density, while the Provincial Capital Economic Circle and Lunan Economic Circle showed comparable levels, with all three demonstrating upward trends (Table 6). Driven by the radiation of Jinan city, the economic development of the Provincial Capital Economic Circle had developed rapidly, and the GDP density had more than doubled from 2.319 × 10⁷ CNY/km² in 2000 to 4.749 × 10⁷ CNY/km² in 2020, which was mainly due to the continuous improvement of policy support, economic activities, population migration, and infrastructure. Notably, the Lunan Economic Circle achieved the fastest economy growth at 7.15%, linked to its industrial restructuring and infrastructure development over the past two decades.

The spatial agglomeration characteristics of the GDP density in the SPUA were significant, with the agglomeration intensity increasing gradually. The Moran’s I value of the GDP density from 2000 to 2020 was 0.90, 0.92, and 0.94, passing the 99% confidence test (Z > 2.58, p < 0.01). This points out that GDP density shows a highly significant positive correlation in space. The Z-value presented a slight upward tendency, indicating that the economic development of the SPUA had a strong spatial dependence characteristic. Specifically, regions with a higher level of GDP tend to be adjacent to each other in spatial distribution, while regions with a lower level of GDP tend to be adjacent to each other.

The spatial pattern of GDP density in the SPUA presents obvious “high-high” (H-H) and “low-low” (L-L) clustering patterns (Figure 5). The H-H agglomeration area was mainly distributed in the Jiaodong Peninsula and Jinan metropolitan area, as well as other economically developed areas, revealing the significance of the radiation effect of core cities, forming a continuous high GDP density cluster. The L-L agglomeration area was concentrated in the less developed counties of southwest Shandong and Northwest Shandong. The number of H-L and L-H outlier regions was tiny, indicating that the spatial gradient difference of GDP density is significant, the transition zone is not obvious, and the regional collaborative development mechanism needs to be optimized.

3.2. Spatiotemporal Evolution Characteristics of Habitat Quality

Habitat quality exhibited marked spatial heterogeneity in the SPUA and generally showed a spatial pattern of high in the mid-eastern regions and low in the western regions from 2000 to 2020 (Figure 6). The high-grade and relatively high-grade regions were largely clustered throughout hilly-mountainous regions occupying the eastern and mid-southern parts of the SPUA, primarily attributable to greater vegetative density coupled with intensified anthropogenic pressures. Relatively low-grade habitat quality regions were spread widely, mainly in the western and northern plains. The low-grade and relatively low-grade habitat quality regions were predominantly clustered urban centers of large cities such as Jinan, Qingdao, Zibo, and coastal areas of lakes, or along the lake and coastal regions, showing a spatial distribution pattern of “lines–clusters”. In addition, there were also small parts distributed in the surrounding areas of key development towns and counties. This was primarily owing to the concentration of undeveloped land in coastal and lakefront regions, while the wide distribution of construction land was in urban areas, counties, and surrounding areas.

The overall habitat quality was low-grade and presented a slight downward trend, with a habitat quality index of 0.287 in 2000 and 0.272 in 2020. During the study period, the proportion of areas with relatively low-grade habitat quality was 64.66%, 63.99%, and 62.89%, which decreased by about 2.74% (Figure 7). The relatively low-grade habitat quality mainly transformed to medium- and relatively high-grade, and the habitat quality slightly improved. Low-grade habitat quality area increased from 15.99% to 19.73%, with an annual growth rate of 1.17%. This increase was driven by construction land expansion and cultivated land encroachment, which threatened habitat quality and resulted in a continuous decline in eco-environmental quality. In addition, low-grade and relatively low-grade habitat quality regions were concentrated in coastal and urban centers, with high levels of commercial and industrial activity, large population sizes, and rapid urbanization. During the process of development, the eco-environment was constantly destroyed, further reducing habitat quality. The proportion of areas with relatively high habitat quality decreased from 8.39% to 7.08%, while the proportion of areas with high habitat quality increased from 2.49% to 2.84%. This change highlights the partial effectiveness of the ecological protection policies. These regions were mainly concentrated in the hilly-mountainous regions in the mid-eastern part of the SPUA and the Yellow River Delta. Among them, the habitat quality conditions in the Yellow River Delta have shown marked enhancement, which is primarily attributable to the protection and restoration of the Yellow River Delta wetland, which had substantially boosted environmental functionality.

The area of habitat quality grade regions showed small inter-annual fluctuations, indicating that the effectiveness of government ecological protection policies significantly offset the negative impacts of human activities such as urban expansion and cultivated land encroachment. Core cities at different development stages exhibited differences in ecological protection measures and implementation timelines. For instance, Jinan, as the provincial capital, has continuously promoted the delineation of ecological protection redlines in the southern mountainous areas and the “Green City” construction project since 2010, improving habitat quality through strict development restrictions and vegetation restoration. Conversely, since the implementation of the “Blue Bay Restoration Initiative” in 2018, Qingdao has prioritized the rehabilitation of coastal wetland ecosystems in Jiaozhou Bay, effectively mitigating the expansion of degraded coastal habitats. In contrast, the emerging core cities in the southern part of Shandong Province (such as Jining and Heze) had a later industrialization and urbanization process. Since 2015, they have gradually mitigated environmental degradation through initiatives such as the “Grain for Green” program and the construction of eco-industrial parks.

Based on urban agglomerations analysis, the three urban agglomeration habitat quality indexes revealed an upward trend (Figure 8). Specifically, the habitat quality of all three economic circles declined significantly from 2000 to 2010, followed by a slow recovery phase from 2010 to 2020. The habitat quality indexes of the three economic circles all exhibited “first decline, then rise” fluctuations, but there were differences among the various circles. The Jiaodong Economic Circle showed only a 2% decline in habitat quality index, due to its coastal natural advantages such as abundant coastal wetlands and marine ecosystems, along with strengthened ecological protection awareness since 2010, such as the implementation of the “Shandong Province Marine Ecological Environment Protection Regulations”, which has jointly formed a more comprehensive protection system for marine ecosystems and coastal wetlands. Additionally, the restoration project of Jiaozhou Bay wetlands and the construction of the National Nature Reserve of the Yellow River Delta have effectively mitigated the pressure of habitat degradation. The habitat quality in the Provincial capital Economic Circle and the Lunan Economic Circle had decreased by 5%, with more significant changes. This was mainly due to the rapid urbanization process from 2000 to 2010, which led to the extensive occupation of natural habitats such as cultivated land and forest land. Although policies such as “reversion of farmland to forest” and “ecological redlines” were implemented after 2010 to gradually restore the habitats, the restoration rate was slower than that of the Jiaodong Economic Circle.

3.3. Spatiotemporal Evolution Characteristics of Coupling Coordination Degree of Economy and Habitat Quality (EHCCD)

The EHCCD exhibited a continuous increasing tendency, with the EHCCD increasing from 0.30 to 0.42 between 2000 and 2020 (Figure 9). The EHCCD demonstrated rapid expansion during the 2000–2010 period, reaching 22%; and the growth of the EHCCD was 16.3% from 2010 to 2020. This continued growth trend showed that the SPUA was on a healthy and sustainable trend. Meanwhile, the relationship between both economy and habitat quality across cities ranged from extreme incoherence to basic coordination, and the average value of the EHCCD was between 0.22 and 0.54. Most were in the moderate uncoordinated stage in 2000, and only Qingdao was in the basic coordination stage, with an average of 0.41. Compared with 2000, the EHCCD in each city had developed rapidly in 2010, in which Jining and Liaocheng showed the fastest growth at more than 50% (Figure 10). And Jinan, Qingdao, Zaozhuang had upgraded to the basic coordination stage. By 2020, Jinan and Qingdao further upgraded to the initial coordination stage, while the basic coordinated regions of Zibo, Zaozhuang, Weifang, Jining, Weihai, Liaocheng, Binzhou, and Heze expanded.

In terms of spatial pattern, the EHCCD exhibited a dual-core pattern; Jinan and Qingdao were leading, with a circular and gradually shrinking distribution from the inside to the outside. Among them, the highest EHCCDs were mostly located in the central region of the SPUA, with Jinan, Qingdao, and Yantai in the inner main circle and Zibo, Dongying, Weifang, Jining, and Linyi in the second circle. The EHCCD of the cities in the outermost circle was the lowest, with predominant clustering in the northwestern of the SPUA, such as Liaocheng, Heze, Dezhou, and Binzhou. Following the continual rise in urbanization expansion, the EHCCD of each city was also improving, and the overall development tended to be coordinated and sustainable.

From the perspective of urban agglomerations, the EHCCD of the three economic circles had been constantly improving; Lunan Economic Circle developed the fastest, with an annual growth rate of 2.95% (Figure 11). The Lunan Economic Circle, which started from a relatively low level, had the fastest growth rate and showed strong development potential. This was mainly because of the key policy support and in-depth implementation of regional development strategies by Shandong Province for the Lunan region. By strengthening infrastructure construction, optimizing industrial structures, and improving the level of public services, the Lunan region had effectively promoted the coupling and coordination of the region, achieving a significant improvement in EHCCD. The EHCCD of the Provincial Capital Economic Circle had steadily improved from 2000 to 2020, and the growth rate was relatively stable. This was mainly because of the radiating and driving role of the provincial capital Jinan, as well as the continuous support of government policies and the constant improvement of regional infrastructure. In addition, the cooperation between cities within the provincial capital economic circle had become increasingly close, and the achievements in industrial synergy development had been remarkable, jointly promoting the overall improvement of regional coordination.

3.4. The Driving Factors Influencing the EHCCD

3.4.1. Single-Factor Detection Results

The primary factors influencing the EHCCD were total retail sales of consumer goods, built-up area, and land use intensity in the SPUA (

Table 7. Detection results of the drivers of the EHCCD.

	2000	2010	2020	Total Changes
Total retail sales of consumer goods (X1)	0.7508	0.7328	0.8779	0.1271
Urban population density (X2)	0.3312	0.3278	0.3357	0.0046
Average Temperature (X3)	0.3502	0.3544	0.3477	−0.0044
Precipitation (X4)	0.4125	0.4521	0.4831	0.0706
Municipal road area (X5)	0.1291	0.1368	0.2378	0.1086
Bulit-up area (X6)	0.5947	0.6943	0.7087	0.1140
Green coverage rate (X7)	0.1126	0.1452	0.3190	0.2064
Cultivated land use (X8)	0.6169	0.5028	0.6514	0.0318
Land use intensity (X9)	0.4032	0.5162	0.5336	0.1304

). Within the study period, the primary factors had greater explanatory power (q > 0.5) for the EHCCD. This was essentially reflected the “economic activity–night time lighting–coupling coordination” triadic interaction mechanism: The night-time lighting data quantifies the intensity of human economic activities, becoming the core medium connecting the driving factors and EHCCD. The synergy of the three elements reveals the profound impact of economic development on the coordination of regional economy and ecosystem. The secondary factors affecting the EHCCD were urban population density, average temperature, and precipitation. The explanatory power of the three factors was between 0.3 and 0.5, which reflects that temperature and precipitation changes can improve the eco-environment, promote clean production, and save resources, which are beneficial to ecological and resource sustainability.

The explanatory power q of urban population density decreases slightly from 2000 to 2020, while the rest of the indicators exhibit differential increases. The power of green cover and land use intensity grew larger, which showed that regional development prioritized sustainable ecological management, and the role of the eco-environment in the coupling and coordination progress of the region became more obvious. Ecological civilization construction and sustainable development had been paid more attention in the process of regional development.

3.4.2. Interaction-Factor Detection Results

From 2000 to 2020, the detection results of factor interactions were sorted, and the top five factor combinations in terms of interaction influence were selected (

Table 8. Detection results of interaction of influencing factors.

2000		2010		2020
Interaction Factor	q-Value	Interaction Factor	q-Value	Interaction Factor	q-Value
X1 ∩ X5	0.9813	X1 ∩ X2	0.9847	X2 ∩ X8	0.9170
X1 ∩ X7	0.9913	X1 ∩ X4	0.9834	X3 ∩ X4	0.9958
X1 ∩ X9	0.9897	X1 ∩ X7	0.9815	X3 ∩ X5	0.9918
X4 ∩ X6	0.9826	X3 ∩ X5	0.9813	X3 ∩ X7	0.9584
X5 ∩ X9	0.9913	X5 ∩ X9	0.9843	X4 ∩ X8	0.9912

). According to the detection results, it is known that the interaction intensity of the nine influencing factors is greater than the independent single factors, and there are two-factor enhancement and nonlinear enhancement among the factors, indicating that the spatial pattern differentiation of EHCCD is the result of the combined action of multiple factors.

From 2000 to 2020, the total retail sales of consumer goods (X₁) had a relatively strong independent influence and a large interaction value with urban population density (X₂), precipitation (X₄), municipal road area (X₅), green coverage rate (X₇), and land use intensity (X₉), with an explanatory power of over 98% in all cases. Moreover, the interaction detection effect is mainly characterized by nonlinear enhancement, indicating that the interaction effect with this variable is higher than its own effect, and it has a strong explanatory power for the spatial variation of coupling coordination degree. Between 2000 and 2020, the social system of “economic development–population concentration–infrastructure” is the dominant driver of EHCCD. In 2020, the interaction between temperature (X₃) and precipitation (X₄) and municipal road area (X₅), green coverage rate (X₇), and cultivated land use (X₈) showed a nonlinear enhancement, with an explanatory power of over 95%. Natural factors such as temperature and precipitation have replaced socio-economic ones as the core of the interaction factors, indicating that the coordinated drive of “natural and human regulation” has become the new dominant force.

4. Discussion

4.1. Significance of Integrating Night-Time Lighting Data and Statistical Data to Measure the GDP Density

From 2000 to 2020, the GDP density has been continuously increasing and has presented a distinct new pattern: a structure of single-core to multi-core networks. In general, the GDP density is high in the east and low in the west, with large differences in inner-city development, and urban development in the eastern region surpasses that in the western region [52,53]. The conclusion of this research is also in line with the overall pattern of “three circles and four zones, networked development” proposed in 2017 and the active promotion of the construction of the economic circle of the SPUA. In response to regional development incoherence, Shandong province emphasized the construction of new mechanisms, and the “14th Five-Year Plan” further emphasizes the spatial development coordination by building Jinan into the central node of the Provincial Economic Capital, with Qingdao as the core of the Jiaodong Economic Circle, relying on the significant advantages of the location and vigorously advanced marine economic growth. Progressing in parallel, construction of the Lunan Economic Circle is being promoted.

4.2. Consideration on the Causes of the Fluctuating Decline in Habitat Quality

Habitat quality exhibited a fluctuating declining trend in the SPUA, showing a spatial pattern of high quality in the mid-eastern regions and low in the western regions, which was similar to the findings of studies [52,54,55]. The high-value regions of habitat quality are mainly distributed in mountainous and hilly terrain. The favorable climatic conditions in these terrains provide the best temperature and precipitation for vegetation, thereby promoting the growth and development of vegetation [38]. However, habitat quality in city centers and extensive plains is relatively low. The main reason may be due to several factors: Firstly, significant infrastructure development has been undertaken in the SPUA, including the construction of high-speed railway tunnels and relentless urban expansion [56]. These human activities have intensified land use intensity and accelerated the decline in habitat quality. Secondly, the SPUA occupies the downstream section of the Yellow River Basin. In recent years, there have been environmental deterioration phenomena such as soil degradation and soil erosion [57]. As a special semi-artificial and semi-natural ecosystem, cultivated land is not conducive to the survival of other organisms through agricultural practices such as grain planting and vegetable cultivation, causing damage to the original habitat conditions.

To address the issue of declining habitat quality, it is recommended to implement a strategy that integrates ecological conservation with restoration measures. Firstly, multi-level ecological conservation redlines should be delineated based on habitat quality assessments, with priority given to biodiversity protection. Strictly control the expansion of construction land within the redline area and the intensity of farmland development, and establish an ecological compensation mechanism to balance the interests between ecological protection and regional development. Secondly, for degraded ecosystems, particularly in vulnerable areas such as coastal zones and wetlands, restoration projects—including coastal wetland rehabilitation increasing the coverage of vegetation and hydrological connectivity restoration—should be implemented to enhance ecosystem service functions.

4.3. Scale Effect of Coupling Coordination Between Economy and Habitat Quality

As the economy progresses, its impacts on habitat quality becomes more pronounced, and environmental constraints on economic development become stronger. Various studies confirm this interaction [36,58]. The overall trend of EHCCD increased from 0.30 to 0.42, indicating that the synergistic effect between economy and habitat quality is gradually increasing. Wang et al. [59] studied that the PM_2.5 pollution in Shandong Province reached its peak in 2013, and the pollution level then began to decline. Shan et al. [60] found that the agricultural carbon emissions in Shandong Province showed a downward trend from 2012 to 2022, with an average annual decrease of 2.37%. The above research conclusions all confirm the research results of this paper.

In terms of spatial distribution, a dual-core model with Jinan and Qingdao as the core has been formed, and it shows a decreasing trend from inside to outside. At the level of urban agglomeration, the Jiaodong Economic Circle has the best performance, and the Lunan Economic Circle has the fastest growth. At the prefecture-level city level, the EHCCD of various cities has improved significantly, but the development is uneven. The studies concerning the coupling coordination degree of the grid scale break the limitations of the previous research area [52]. At the same time, new discoveries have been made, such as Lunan Economic Circle developing fastest, surpassing the other two economic circles in 2020.

To better promote the coordinated development between economy and habitat quality in the SPUA, the following strategies should be implemented. First, it is essential to establish a development philosophy of “ecological priority and zonal governance”. For the hilly and mountainous areas in the central-eastern regions, the most stringent ecological protection systems should be enforced, including delineating core ecological redline zones to restrict economic development activities while establishing cross-municipal ecological compensation mechanisms. For the western agricultural plains, a “smart agriculture plus ecological restoration” model should be adopted, with the construction of green infrastructure such as ecological ditches to mitigate habitat disturbances caused by agricultural activities. Second, the coordinated development mechanism of the urban agglomeration should be strengthened by establishing an ecological corridor network along the lower reaches of the Yellow River, with particular emphasis on restoring key ecological nodes such as the Dongying Delta wetlands to enhance regional ecological connectivity.

4.4. Uncertainty and Perspectives

This study constructed an optimal regression model to measure GDP density using night-time lighting data and economic statistical data and calculated habitat quality at the grid scale with the InVEST model, and then explored the EHCCD. This method not only provides new ideas for quantitative analysis of GDP density but also reveals the environmental response characteristics of different regions in economic development. Firstly, the statistical data employed in this study were confined to the municipal scale, and the relatively coarse spatial resolution of the research units may lead to diminished model accuracy. Future research should incorporate smaller-scale data (such as county or township levels) to enhance simulation precision. Secondly, only nine influencing factors were selected. In future research, more influencing factors (such as environmental regulation) can be considered to provide a more comprehensive explanation. These refinements would not only address current scale limitations but also provide more granular scientific support for regional sustainable development policies.

5. Conclusions

In this paper, we contributed the GDP density regression model using night-time lighting data and statistical data to acquire spatial GDP of the SPUA and calculated habitat quality with the InVEST model. Based on the GDP density and habitat quality, the EHCCD was calculated with an optimized coupling coordination degree model, and the spatiotemporal pattern and change characteristics of the EHCCD were explored. Finally, the factors to drive the variation of the EHCCD were revealed by the Geo-detector model. The major findings can be summarized as follows:

(1) The economic development has been increasing, and the macroeconomic development of the SPUA is better. The spatial distribution of GDP density presents a structure of single-core to multi-core networks, and the SPUA economy develops in a coordinated and sustainable way.

(2) The overall condition of habitat quality was poor in the SPUA and showed a downward fluctuation. The spatial distribution of habitat quality in the SPUA shows that it is higher in the mid-eastern regions and lower in the western regions, and there are significant differences in habitat quality among cities.

(3) The EHCCD of the SPUA showed an upward trend. The relationship of the EHCCD in most cities developed in the direction of coordination, and in several cities declined. The spatial distribution of the EHCCD presented the characteristics of a dual core and the trend of declining gradually from the eastern coastal area and the central region to the western and southern region. The cities around the dual core, including Jinan and Qingdao, are mainly moderately coordinated, and the EHCCD is the lowest in the west–south regions of the SPUA.

(4) Among the key factors driving the development of EHCCD, total retail sales of social consumer goods, built-up area, and land use intensity have greater explanatory power, and we should focus on these influencing factors in the development process of the SPUA to continuously improve the quality of urbanization development. And the interaction-factors approach has shifted from socio-economic dominance to synergistic dominance of natural and human factors.