Introduction

Background

The twenty-first century has been marked by the rapid advancement and extensive application of internet technology, which has significantly accelerated the development of the digital economy (DE). The trajectory of new industrialization is increasingly intertwined with the evolution of digital media, with digital products playing a pivotal role in shaping the future1. However, alongside the growth of DE, global warming driven by carbon emissions (CE) remains a critical challenge2. Despite collective international efforts to achieve the goals of the Paris Agreement3, CE continues to rise, with no evident sign of having reached its peak4.

China’s urbanization, characterized by high energy consumption and significant pollution, poses urgent challenges for energy conservation and emission reduction5. The Yangtze River Economic Belt (YREB), a key region in China’s “T”-shaped territorial spatial planning framework, encompasses over 100 prefecture-level cities and municipalities along the Yangtze River. This economically vibrant region features diverse city types and industrial structures, making it a prime example of the urgent need for coordinated environmental and economic development. By 2020, CE in the YREB had exceeded 900 million tons, driven primarily by ore smelting and thermal power industries. Economic growth in this region has significantly contributed to rising regional carbon emissions6. In recent years,the YREB has been actively adjusting its industrial structure, which has brought about some subtle changes in the region’s economy and environment.

Progress in DE studies

DE represents an innovative application of digital technology to traditional financial products7. The concept of DE can be traced back to the 1990s, when Tapscott introduced the term “digital economy era”8. Moulton later refined this definition, emphasizing DE’s inclusion of information technology developments within e-commerce9. Lane expanded this perspective by describing DE as an economic model driven by information technology, which has transformed both e-commerce and traditional business models10. Cheon & Kim subsequently clarified DE’s boundaries11. At the 2016 G20 Hangzhou Summit, the consensus on DE highlighted digital knowledge and information as crucial factors of production. DE enhances investment efficiency and facilitates the upgrading of traditional industries12, while also making positive contributions to economic growth13, industrial transformation14, green economy development15, and employment structure adjustment16.

Research on DE predominantly follows two approaches.First, industry-specific studies explore DE’s impact but often overlook its internal structural changes. Second, index-based studies develop multidimensional DE evaluation systems. For instance, Maji & Laha analyzed DE in the Asia Pacific region using panel data17, while Hanna constructed a DE evaluation index system based on digital infrastructure, platforms, governance, and data processing18. Other studies have examined DE’s relationship with regional economic growth and total factor productivity19,20.

Progress and gaps in CE studies

Despite the rapid expansion of DE research, the environmental dimensions of digitalization, and CE in particular, remain comparatively underexplored and lack integrated examination. Concurrently, global climate change has drawn significant academic attention, emphasizing the need for effective CE research21. Scholars have explored CE’s spatiotemporal dimensions22,23. For example, Zheng et al. analyzed CE trends across 30 Chinese regions, revealing an upward trajectory and regional disparities24. Similarly, Zhang & Li examined rural China’s CE evolution25, while Gao & Ge focused on regional spatiotemporal CE changes26. These studies confirm CE’s spatiotemporal heterogeneity. Driving factors of CE have also been extensively studied27,28. Liu & Song demonstrated that fiscal development can curb CE,recommending targeted financial support for CE projects29. The environmental Kuznets curve (EKC) hypothesis, which posits an inverted U-shaped relationship between CE and GDP, has been validated by multiple studies30,31. For instance, Wang confirmed EKC’s applicability in China from 2005 to 201732, while Shen et al. identified economic growth as a key CE driver33. However,some scholars question EKC’s universality34,35.

Most CE studies focus on macro scales (e.g., global or provincial)36,37, with limited attention to meso scale or micro scale analyses, and most major countries use regulations to control the output of carbon emissions. Common research methods include SDM models38, difference models39, and LMDI models40,41, which employ quantitative analysis through regression and decomposition techniques. Findings consistently highlight CE’s concentration in central and eastern coastal provinces42.

As discussed above, we observe that DE, as a crucial engine of global economic growth, is rapidly transforming production and consumption patterns while simultaneously exerting complex impacts on CE. This interplay has become a key issue in achieving sustainable development and green transition. Globally, the widespread application of digital technologies contributes to carbon emission reduction by enhancing energy efficiency and optimizing resource allocation. On the other hand, the energy consumption and embedded carbon footprint of digital infrastructure may also intensify emission pressures. This dual effect has become a focal point of international academic and policy-making circles. Existing research has preliminarily explored the relationship between DE and CE, primarily focusing on the role of DE in promoting economic growth, industrial transformation, and energy efficiency, as well as examining the spatiotemporal patterns, influencing factors, and emission reduction policies of CE. However, most studies still exhibit significant limitations. Firstly, research on DE and CE often remains segregated, lacking a systematic analysis of their coupled coordination relationship. Secondly, existing predominantly operate at national or provincial macro scales, with insufficient attention paid to urban units, making it difficult to reveal localized mechanisms and spatial heterogeneity. Furthermore, in terms of theoretical frameworks, most studies rely on traditional economic or environmental models and fail to integrate multidimensional factors to explain the interaction mechanisms between DE and CE.

To fill these research gaps, this study examines the YREB, a vital economic-ecological corridor in China, as a case study. Based on panel data from 110 cities during 2012–2021, we establish a comprehensive evaluation framework to assess the development trajectories of DE and CE. By employing a coupling coordination degree (CCD) model and an improved the diamond model (DM), we aim to uncover the coordination relationship between DE and CE and its driving mechanisms. This study addresses three key questions: (1) What is the degree of coupled coordination between DE and CE in the YREB? (2) What are the characteristics of its spatiotemporal evolution? (3) Which factors drive the spatial differentiation of CCD? The findings will provide a theoretical basis and policy references for regionally synergizing digitalization and low-carbon transition, while also offering insights for similar regions globally.

Theoretical framework

We find that DE influences CE through two mechanisms: On one hand, digital technologies such as cloud computing, the internet of things, and Artificial Intelligence (AI) influence CE in sectors like manufacturing and logistics by improving energy efficiency, optimizing industrial processes, and achieving smart energy management. On the other hand, increased energy demand due to data center power consumption, electronic waste, and the expansion of digital infrastructure, if reliant on fossil fuels, could exacerbate carbon emissions.

Despite substantial research on DE and CE, existing studies primarily focus on horizontal measurements and macro scale influencing factors43,44. The YREB, though a crucial region, remains underexplored and lacks integrated analyses of DE and CE. Firstly, grounded in the Coupling Coordination Theory45, this study conceptualizes the DE and CE systems as two complex, interactive subsystems. The theory posits that the degree of mutual influence and synergistic development between systems can be measured through a CCD model. This provides a theoretical basis for quantifying whether the growth of the digital economy and the pressure of carbon emissions are in a state of dysregulation or mutual coordinated development. This study addresses this gap by constructing a novel index system and applying the CCD model, exploratory spatial data analysis to examine the CCD between DE and CE and its driving factors. The findings contribute to a deeper understanding of DE-CE interactions and provide theoretical support for achieving China’s “dual carbon” goals and advancing YREB’s ecological protection and digital development strategies. Secondly, insights from Environmental Economics, particularly the theory of externalities and the decoupling theory, offer critical lenses to interpret the mechanisms. On one hand, the DE generates positive environmental externalities by improving resource allocation efficiency and promoting green technological innovation . On the other hand, it can also have negative externalities, such as the high energy consumption of digital infrastructure itself. The decoupling theory further provides a framework to assess whether the growth of DE can be separated from or even lead to a decline in CE pressure, which is the core of achieving sustainable development goals.

By integrating the CCD model with the Diamond Model (DM), an innovative multi-dimensional theoretical framework has been constructed to systematically reveal the dynamic interaction mechanisms between the DE and carbon emissions CE. Compared to previous studies that isolated macro relationships, this research focuses on spatiotemporal heterogeneity at the city scale, elucidating the bidirectional interaction mechanisms between DE and CE, thereby deepening theoretical understanding of how digitalization can drive or mitigate environmental effects. It also holds significant practical implications for promoting the coordinated development of the digital economy and environmental sustainability. Firstly, in this study, the entropy weight method was used for data preprocessing of DE and CE, and the spatiotemporal evolution of DE and CE in the YERB region was analyzed. Secondly, we analyzed the CCD of DE and CE in the YREB. Then, we applied the improved DM to analyze the driving factors affecting CCD. Finally, draw a conclusion (Fig. 1).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Research theoretical framework.

Data sources and research methods

Data sources

This study focuses on the YREB in China, covering the period from 2012 to 2021. The data used in this research are sourced from authoritative publications, including the China Statistical Yearbook, China Energy Statistical Yearbook, China Science and Technology Statistical Yearbook, China Environmental Statistical Yearbook, Tianyan Survey, and provincial and municipal statistical yearbooks from various years. To ensure the scientific rigor, completeness, and reliability of the dataset, missing data for certain regions were addressed using interpolation methods. This approach guarantees a robust foundation for analyzing the CCD between DE and CE in the YREB.

Study area

The YREB, linked by the Yangtze River, encompasses 11 provinces and municipalities, including Shanghai, Jiangsu, Zhejiang, Anhui, Jiangxi, Hubei, Hunan, Chongqing, Sichuan, Yunnan, and Guizhou. Spanning the eastern, central, and western regions of China, the YREB covers an area of 2.052 million square kilometers, representing 21.4% of the nation’s total land area. It accounts for 43.1% of China’s permanent population and contributed 46.6% of the national GDP in 2021. For analytical purposes, the study area is divided into three sub-regions: upstream, midstream, and downstream. The geographic distribution of these regions is illustrated in Fig. 2. This division facilitates a detailed examination of the spatial heterogeneity in the coupling and coordination between DE and CE within the YREB.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Study area. Note: The map is based on ArcGIS Desktop 10.7 (http://www.esri.com/software/arcgis), and the base map adopts the standard map service website of China National Administration of Surveying, Mapping and Geoinformation under the supervision of the Ministry of Natural Resources of China. The review number of the base map is GS (2020) 4619, and the base map has not been modified.

Methods

  1. (1)

    Entropy weight TOPSIS method

Entropy weight TOPSIS method is an objective weighting method, which is mostly used in the evaluation of the comprehensive level of multiple schemes. Its advantage is that it can effectively eliminate the bias caused by subjective factors and overcome the attribute repeatability caused by too many indicators in complex systems46.

  1. (2)

    Coupling coordination degree (CCD) model

This study aims to explore the synergistic effect of CE and DE by constructing a CCD model47. After calculation, CCD is classified, as shown in Table 1. The specific formula is shown as follows.

$$C = \frac{{\sqrt {U_{1} \times U_{2} } }}{{(U_{1} + U_{2} )/2}}$$
(1)
$$T = {\text{a}}U{}_{1} + bU_{2}$$
(2)
$$D = \sqrt {C \times T}$$
(3)
Table 1 CCD classification.
Full size table

where, C is the coupling degree of CE and DE in YREB, and U1 is the CE level; U2 Measure DE development level. a and b are coefficients of uncertainty. The value of C ranges from 0 to 1. A higher value of C indicates that the trends of the two subsystems are more organized. T is a comprehensive evaluation index to measure DE and CE. CE is considered to be as important as DE, so the value a = b = 0.5 is assigned. D indicates the CCD between CE and CE. The value ranges from 0 to 1.

Index system construction

Carbon emission index

Due to the lack of CE data published in official statistics, the formula provided by the United Nations Intergovernmental Panel on Climate Change (IPCC) is generally used to calculate greenhouse gas emissions data. Considering the greenhouse gas emissions generated by energy consumption at the production end, eight energy sources, including raw coal, coke, crude oil, gasoline, kerosene, diesel, fuel oil and natural gas, are selected as terminal consumption and CE are calculated based on consumption. Meanwhile, regional carbon emission intensity (CI) is represented by the ratio of regional carbon emission to regional GDP, that is, CO2 emission per unit GDP. Regional carbon emission density (CD) is represented by the ratio of regional CE to regional square area, that is, CO2 emission per unit area of a region. The ratio of CE in each region to regional population, that is, CO2 emissions per unit population in the region, represents regional per capita CE (CP). In this way, CE data for 110 prefecture-level cities in the YREB were obtained. Then, the entropy weight method was applied to calculate the indices, and the CE index system was subsequently constructed (Table 2).

$$CE = \sum\limits_{i = 1}^{8} {CO_{2,ijt} } = \sum\limits_{i = 1}^{8} {M_{ijt} } \times K_{j} \times Z_{j} \times \frac{44}{{12}}$$
(4)
$$\begin{gathered} CI_{{{\text{it}}}} = CE_{{{\text{it}}}} /{\text{GDP}}_{{{\text{it}}}} \hfill \\ CD_{{{\text{it}}}} = CE_{{{\text{it}}}} /UA_{it} \hfill \\ CP_{{{\text{it}}}} = CE_{{{\text{it}}}} /RP_{{{\text{it}}}} \hfill \\ \end{gathered}$$
(5)
Table 2 CE coefficient and standard coal conversion coefficient.
Full size table

where t is the year; i is the region; j is for energy; CE is CO2 emission; CI is carbon emission intensity; \(M_{{{\text{ijt}}}}\) is energy consumption; Kj is the CO2 emission coefficient. Zj is CO2 standard discount coal coefficient; CD is carbon emission density; UA is the area; RP is per capita carbon emission; RP is the total population of the region (Table 2).

Digital economy index

This study establishes a comprehensive evaluation system for the development level of the DE, focusing on four key dimensions, including digital infrastructure, digital industry development, digital innovation capability, and digital inclusive finance (Table 3). Digital infrastructure is assessed through broadband Internet, mobile Internet, and international Internet foundations, measured by the number of Internet users per 100 people, mobile phone users (in tens of thousands of households), and total Internet users (households), respectively. Digital industry development, the core of DE, is characterized by the output of the telecommunication industry, postal industry growth, and the information industry foundation, represented by the total value of telecommunication business, postal business (both in tens of thousands), and the number of computer practitioners. Digital innovation capability is evaluated by DE-related patents and science and technology expenditure, reflecting innovation levels and support. Digital inclusive finance, a vital driver of resource allocation and digital life, is measured by its coverage breadth, depth of use, and degree of digitalization. Together, these indicators provide a robust framework for assessing DE development.

Table 3 Evaluation index system of regional DE development level and CE level.
Full size table

Diamond model (DM)

DM was originally proposed by American scholar Porter in 1990. It was mainly used in the field of management and built around four key factors: resource conditions, demand conditions, related and supporting industries, and industry competition. Building on the work of Li et al. (2024) and considering the unique aspects of this study, the improved DM is employed to explore the factors influencing the coupling relationship between DE and CE (Table 4). Specifically, this study has improved it by adding environmental stress factors and government policy dimensions, thereby more comprehensively capturing the complex interaction mechanism between the digital economy and carbon emissions—covering both traditional economic drivers and key dimensions such as environmental constraints and policy interventions. This multi-dimensional framework is more suitable for analyzing the coupling relationship between the “economy-environment” dual system, rather than relying solely on the production function centered on capital and labor or the endogenous growth theory of technology.

Table 4 Index system of CCD influencing factors of DE and CE in the YREB.
Full size table

This innovative approach not only extends the theoretical application of the DM but also provides a valuable new methodology for future research in this field. Multiple linear regression is a classic statistical modeling method used to analyze the linear relationship between a continuous dependent variable and multiple independent variables. Accordingly, we apply multiple linear regression to examine the relationship between the factors in the DM (independent variables) and the coupling coordination index (dependent variable), in order to assess the influence of each factor and further explore the underlying impact mechanisms. In addition, as shown in Table 4, each factor consists of multiple indicators. We applied the entropy weight method to assign weights to the indicators, conducted weighted aggregation, and finally combined them into a single factor.

Result analysis

Eastern dominance in carbon emissions

Figures 3 and 4 show a distinct spatiotemporal distribution pattern, characterized by “high in the east and low in the west.” Specifically, in Fig. 3, the average CE index in the YREB exhibited a fluctuating upward trend over the study period. The downstream region consistently had a significantly higher average CE index compared to the entire YREB, although the middle reaches experienced a faster average growth rate than both the upstream and downstream regions. Figure 4 illustrates that the downstream CE index was markedly higher than other regions, with values increasing progressively from west to east. The 41 downstream cities accounted for 34.6% of the total CE index, showing a steady upward trend.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Multi-year change of CE index of YREB from 2012 to 2021.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Temporal and spatial evolution of CE in 2012, 2015, 2018 and 2021.

This trend can be attributed to regional industrial structures and consumption patterns. Developed regions focus on high-end manufacturing, which demands advanced technology and equipment, resulting in significant energy consumption. For example, electronic chip manufacturing requires energy intensive operations and environmental controls. Additionally, higher incomes and stronger consumption power in developed areas drive greater demand for goods and services, amplifying production, energy use, and CE across the supply chain. Activities like frequent online shopping boost the logistics industry, increasing transportation mileage and energy consumption. Diverse travel demands, high private car ownership, and intensive air traffic further exacerbate CE.

Imbalanced growth and spatial agglomeration of the digital economy

The average DE development level in the YREB shows an overall upward trend, with significant spatial and zonal differences. The average DE index in the middle and downstream regions exceeds the overall YREB average. Across different regions, the downstream region exhibits a rapid growth rate, significantly outpacing the YREB average, while the middle region shows a year-by-year decline. In contrast, the upstream region maintains a consistently low level, despite a gradual upward trend, remaining well below the YREB average (Fig. 5). DE development forms “clusters” in the Yangtze River Delta and Chengdu-Chongqing areas, with a “core–edge” spatial distribution (Fig. 6). A few key cities act as core hubs, while others depend on these high-level cities, highlighting the need for strong foundational support for DE growth.

Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Time evolution of DE index of YREB from 2012 to 2021.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Temporal and spatial evolution of DE in 2012, 2015, 2018, and 2021 for YREB.

The disparities in DE development stem from factors such as digital infrastructure, regional economic foundations, talent resources, and policy environments. Developed regions benefit from robust network infrastructure, such as extensive 5G coverage, facilitating data transmission and digital transformation. Conversely, underdeveloped areas face weak network construction, hindering digital advancements. Economically advanced regions attract investments in technology, foster digital industrial clusters (e.g., Shanghai, Suzhou), and draw digital talent due to better resources, job opportunities, and living conditions. In contrast, less-developed areas struggle with limited financial resources, weak talent retention, and brain drain. Policy support plays a crucial role in DE growth. Progressive regions, like Zhejiang, have implemented targeted policies, established industrial parks, and provided tax incentives and financial subsidies to promote digital industrialization and transformation. However, some regions lack sufficient policy backing, impeding DE development and guidance.

Upgrading coupling coordination with persistent regional imbalance

Figure 7 shows that CCD between DEand CE in the YREB is on an annual upward trend. The average CCD value across regions maintains positive year-on-year growth, with downstream areas significantly higher than the basin-wide average and showing upward fluctuations. Although upstream areas have also grown concurrently, their coordination remains low, which may be closely related to the limited scale of digital economic output and the lack of expansion momentum in this region. From the perspective of spatial patterns, CCD values exhibit significant regional imbalances (Fig. 8). Cities with low coordination are mostly distributed in the middle and upper reaches, while downstream areas form high-cohesion coordination clusters.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

CCD time evolution of DE and CE in YREB from 2012 to 2021.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Space–time evolution of CCD levels of DE and CE in YREB in 2012, 2015, 2018 and 2021.

In 2012, most cities along the YREB were in a state of imbalance, with only Shanghai entering the high-cohesion zone; by 2021, more cities had achieved a leap in coupling levels, and almost all cities had transitioned from moderate to mild imbalance. However, regional CCD differences remain prominently evident in the middle and upper reaches. Notably, the differentiated impact of carbon neutrality policies on digital cities may have exacerbated this spatial divergence. In core cities downstream, the carbon neutrality goal is enforced through mandatory energy efficiency standards and carbon trading mechanisms, driving the green transformation of digital infrastructure (such as data centers using renewable energy), thereby mitigating the marginal effects of DE expansion on CE. Whereas in the middle and upper reaches, due to insufficient policy enforcement, digital city construction still relies on traditional energy systems, weakening emission reduction outcomes. The root cause lies in the multiple obstacles faced by mid-to-upstream cities during the transition from traditional industries to an eco-protective green development model. On one hand, the integration of Industrial Internet of Things (IoT) technology lags behind, making it difficult for high-energy-consuming industries to achieve real-time monitoring of production processes and precise emission control. On the other hand, the regional innovation system is weak, limiting environmental efficiency improvements. The collaborative innovation mechanism of digital economy and clean technology has yet to be established in mid-to-upstream regions. In addition, the accumulation of new DE momentum has been slow, failing to effectively substitute traditional high-carbon economic activities with platform economies and sharing models.

In summary, the coordinated development of DE and CE in the YREB has continued to improve during the study period, reflecting the intrinsic connection between the expansion of the digital economy and the growth in carbon emissions. This trend can be traced to the phased characteristics of China’s transition from high-speed growth to high-quality development. Under the impetus of the “dual carbon” goals, the digital economy simultaneously drives the expansion of energy-intensive infrastructure (e.g., data centers and electronic equipment) and strengthens environmental governance through innovative applications such as smart energy systems and digital twin technologies. The outcome is a complex coupling mechanism characterized by the coexistence of “technological lock-in” and “emission reduction breakthroughs.” In the future, three key areas need to be prioritized. First, strengthening institutional synergy between carbon neutrality policies and digital city construction, establishing a blockchain-based carbon footprint traceability system. Second, accelerating the penetration of industrial IoT in key emission industries, achieving online optimization of process carbon emissions through increased device connectivity rates. Third, building green digital technology innovation alliances, promoting the research and application of AI algorithm optimization and edge computing energy-saving technologies, to resolve the “efficiency-growth” paradox in the DE-CE system.

The driving factors have obvious spatial heterogeneity and priority

Based on the DM, the driving factors of the coupling coordination degree of DE and CE in the YREB were analyzed (Fig. 9), and it was found that the influence of each driving factor on CCD showed obvious strip spatial differentiation. At the same time, according to the spatial differentiation characteristics, and based on this, we divided the five drivers into four categories of priority according to the number of positive values of the city, and drew CCD relationship diagram between the drivers and DE and CE in the YREB (Fig. 10). In the sequence, IF > GP > DC > PF = EF.

  1. (1)

    Production factors (PF) and Environmental stress factors (EF). EF and PF affected the same number of cities, the number of cities with positive coefficient is slightly exceeds that of cities with negative coefficients (Fig. 9a, d). EF and PF in the middle and upper reaches were significantly higher than those in the middle and lower reaches, and the negative regions were more obvious than the positive regions, showing a trend of higher distribution in the west and lower distribution in the east. This may be attributed to the continuous improvement of PF is conducive to the continuous absorption of manufacturing labor and the improvement of supporting facilities, resulting in the increase of EF, but promote the orderly flow of capital, labor and data elements among departments, improve the efficiency of factor allocation, and promote the development of regional digital economy. At the same time of the industrial structure transformation in the upstream region, the transformation of traditional industries is slow, and the infrastructure restricts the energy transformation. For example, the limited compatibility of traditional thermal power stations and transmission–distribution networks with renewable energy. Meanwhile, industrial production requires a lot of fossil fuel energy, and the combustion process will release a lot of pollutants. In CE, the high energy consumption of digital infrastructure and the pollution from electronic waste have intensified environmental pressures. The implied carbon footprint has expanded with the surge in computing power demand. While digital technologies enhance production efficiency in traditional industries, they simultaneously offset energy conservation efforts through scaled-up manufacturing operations. Moreover, excessive consumption fueled by platform economies accelerates the circulation of high-carbon goods, while regional energy structure lock-in driven by computing power competition further amplifies the decarbonization challenges at the production end.

  2. (2)

    Innovation factor (IF). IF correlated high values are most widely distributed, and IF in the study area has a positive development trend, which plays a crucial role in CCD of DE and CE in the whole region (Fig. 9b). It can be seen that IF has a significant positive effect on CCD of DE and CE, especially in the middle reaches. Two main explanations account for this result. First, scientific and technological innovation supports the establishment of intelligent decision-making, intelligent management, and related technical service systems, thereby advancing the digital economy. Second, technological innovation simultaneously stimulates the manufacturing and use of electronic equipment, leading to higher regional carbon emissions.

  3. (3)

    Demand conditions(DC). Cities with positive coefficient of demand conditions in the middle and downstream regions display a continuous distribution trend, and positive cities are more than negative cities on the whole (Fig. 9c). The continuous improvement of regional demand can lead to the agglomeration effect and spatial spillover effect of capital, technology, talent and other factor resources, thereby increasing population concentration and infrastructure investment. This process not only raises regional carbon emissions but also enhances the allocation efficiency of digital economy and carbon-related factors, ultimately strengthening the coupled coordination between the digital economy and carbon emissions.

  4. (4)

    Government Policy (GP). The government policy coefficient is predominantly positive, with urban clusters concentrated in the middle reaches, whereas negative concentrations are more evident in the downstream and upstream regions (Fig. 9e). This may be attributed to the decentralized system, under which the inherent advantages of government regulation are pronounced, constituting an important exogenous variable for altering path dependence. The government has promoted the CCD of DE and CE through a series of policy measures. However, excessive government intervention may lead to a phenomenon that is not conducive to the coordinated development of the two.

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

Spatial distribution pattern of DE and CE drivers in YREB.

Fig. 10
Fig. 10The alternative text for this image may have been generated using AI.
Full size image

CCD driver relationship between DE and CE in YREB.

Discussion

In this study, we examine the relationship between CE and DE in this region to provide reference for the subsequent coordinated development of economic and environmental factors in developing countries. The spatial and temporal distribution of CE and DE in the YREB was evaluated by entropy weight TOPSIS method. The results reveal a distribution pattern of CE and DE marked by higher levels in the east and lower levels in the west. These regions are characterized by strong economic performance and robust growth momentum, which is broadly consistent with the findings of other scholars48,49,50,51,52.

Furthermore, we examined the coupling relationship between CE and DE to determine whether the two elements mutually reinforce each other in their development. Drawing on extensive existing literature, the coupling model constructed in this paper reveals that DE and CE are coupled in a relationship characterized by mutual dependence and reinforcement. Although it is recognized that DE is an economic model that helps to promote green development53, it does not directly reduce CE. Conversely, there is a positive correlation. No effective decoupling phenomenon between DE and CE was found, which contradicts the findings of other scholars54.. For instance, the construction of digital infrastructure has become the basis for the growth of DE55,56. At the same time, the extensive use and maintenance of data centers and data base stations have led to a sharp increase in energy consumption and an increase in carbon emissions57.

In addition, at present, there is a severe shortage of digital talents, which leads to significant regional differences in China’s digital transformation. The strategic directions of each region are not clear, and the technical standards are inconsistent. China is still in the early stage of DE construction at this stage. It is precisely because of the immaturity of digital technology that enterprises’ digital transformation is relatively slow during their development, and the penetration rate of DE is low, which makes the impact of digitalization on the growth of enterprise benefits negligible58,59. Therefore, at the current stage of digital economic development in China, the growth of the digital economy has exacerbated carbon emissions. It also indirectly shows that China’s growth model is still significantly affected by high levels of carbon emissions.

We also observed that in certain downstream cities, CE expanded rapidly at first but decelerated in the later stage, while DE maintained steady growth. This divergence can be attributed to the misalignment between regional economic development policies and environmental policies, along with the broad adjustment orientation of environmental regulations. At the regional scale, CE in the YREB has continued to rise, whereas DE has shown a fluctuating trajectory with relatively slow growth. Importantly, this phenomenon does not compromise the validity of the CE–DE coupling results.

More importantly, the concept of management DM has been incorporated into geographical research, allowing for a systematic analysis of the factors influencing coupling. The results have good applicability. Based on the basic coupling characteristics, the DM is improved by adding environmental pressure factor. Under the influence of many factors in the new DM, production factors exhibit a relatively limited influence on the coupling coordination between DE and CE, constituting a secondary effect in essence, this indicates that the adjustment of production factors has an insufficient impact on regional economic and environmental development, while there is a clear regional imbalance in the impact of other drivers on CCD of DE and CE. It is recognized that only adjustment can promote a benign development pattern of economic growth and environmental sustainability60. During the whole research process, we observed some subtle changes. For example, in 2021, there is no moderate imbalance zone between DE and CE in the YREB, and it has fully entered the mild imbalance zone, indicating that regional economic development and environmental development gradually show an upward trend of mutual promotion.

Owing to inconsistencies in the measurement standards of carbon efficiency across studies, some researchers have employed night light data as a supplementary indicator for estimating regional carbon emissions61. In global research, agriculture62,63, construction64,65, and transportation66 are also identified as key drivers of carbon emissions. China, however, exhibits comparatively higher efficiency in carbon emission control. Future research should therefore consider integrating these sectoral variables into regional carbon emission estimates. At the same time, efforts should also be made to reduce carbon emissions by starting from the industrial structure and energy structure16. Meanwhile, the income level of the region, the educational level and the tax on green industries should be continuously reduced to decouple the DE and CE in the region17.

Based on the above discussion, we propose three policy recommendations. First, in the upstream region, capitalizing on clean energy advantages to advance green digital infrastructure. Prioritize developing low-carbon data centers powered by renewable energy sources like hydropower and wind power, while enhancing digital technologies for ecological conservation and establishing intelligent carbon emission monitoring systems. Policy measures could include green electricity subsidies, restrictions on high-energy-consuming digital industries, and integration of digital economy with eco-economy development. Second, in the midstream region, focus on efficiency enhancement and structural optimization during industrial digital transformation. Support smart manufacturing upgrades through industrial internet to reduce energy consumption per unit output, while strictly controlling overcapacity induced by digital platforms. Establish regional carbon markets incorporating digital enterprises into quota management to incentivize low-carbon technology adoption. Third, in the downstream region, leverage innovation strengths to integrate digital technologies with deep decarbonization. Encourage applications of AI and blockchain in optimizing energy systems to create zero-carbon smart city clusters; strengthen carbon footprint tracking in digital consumption sectors by implementing “carbon labeling” systems on e-commerce platforms to guide low-carbon purchasing. Promote cross-regional technology transfer to drive coordinated carbon reduction across entire river basins.

Conclusion

This study uses panel data from 110 cities in the YREB (2012–2021) to evaluate DE and CE systems through entropy weighted method. The CCD model, spatial data analysis, and DM were applied to explore the CCD relationship and driving factors. The key findings can be summarized as follows. First, between 2012 and 2021, CE increased steadily, exhibiting a distribution pattern of “high in the east and low in the west,” with the downstream region contributing 34.6% of the total. While DE showed regional imbalance, with coastal downstream areas exhibiting higher, more volatile values,the growth rate of DE in the downstream region is more than 50% of the average growth rate of the whole region. Second, DE and CE exhibited coordinated development, with annual improvement in coupling, but significant regional imbalance. The CCD index of the middle and lower regions is more than 50% of the overall regional index. Transitions from slight imbalance to mild coupling occurred in 2014 and 2017, with more medium coordination cities, mainly in the downstream. Third, spatiotemporal differentiation is shaped by multiple drivers. Production factors primarily influence the upstream and middle reaches; innovation and demand are more significant in the midstream and downstream; environmental stress displays a clear east–west variation; and government policies exert their strongest effects in the middle reaches.

While this study systematically analyzes the coupling and coordination relationship between DE and CE in the YREB, it still has certain limitations. In terms of data, carbon emission data for some cities rely on estimates, which may contain measurement errors. Methodologically, although the coupling coordination model and Diamond Model are explanatory, they fail to cover all potential influencing factors (such as international trade and consumer behavior). Additionally, the research does not consider full lifecycle carbon emissions from digital infrastructure or inter-regional technology spillover effects. Future studies could incorporate higher-resolution data, compare multi-scale models, and utilize machine learning methods to identify nonlinear relationships, thereby enhancing the robustness and generalizability of the conclusions.