Under the combined influence of climate change and human activities, the complexity of watershed water systems has intensified, and traditional hydrological research methods face new challenges. Artificial intelligence (AI) technologies, represented by big data and deep learning, have developed rapidly and profoundly transformed the research and application paradigms of watershed hydrology. This study reviews the evolution of watershed hydrological research from empirical and statistical methods to mechanistic models and data-driven paradigms. It focuses on the research progress of AI in fields including complex hydrological process simulation, multi-source observation and fusion, and digital twin watershed construction. Overall, the advantages of AI in high-dimensional feature extraction, nonlinear relationship identification, and cross-watershed knowledge transfer significantly enhance the accuracy, generalization, and rapid response capability of hydrological simulation, providing new perspectives and tools for understanding complex hydrological systems and coping with environmental change. However, the in-depth application of AI in watershed hydrology faces several challenges, including the inherent sparsity and heterogeneity of hydrometeorological data, insufficient physical consistency and extrapolation robustness of models, and operational integration. In the future, promoting the deep integration of data-driven methods and hydrological mechanisms, developing a watershed system research framework supported by multi-source information, and building an interdisciplinary research and application ecosystem are key directions for AI to empower innovation in watershed hydrology and support the construction of national water networks and the development of smart water conservancy.
As a key driver of the new technological revolution, artificial intelligence (AI) is profoundly reshaping the research paradigms and analytical tools in resource and environmental economics. Against the backdrop of increasingly tight global resource and environmental constraints and intensifying climate change challenges, a systematic review of the research trajectories, evolutionary pathways, and key issues of AI in resource and environmental economics is of great significance for promoting theoretical innovation and methodological advancement. Based on the Web of Science Core Collection database, this study retrieves relevant literature on AI and resource and environmental economics from 2001 to 2025. Using bibliometric analysis and CiteSpace visualization methods, this study systematically analyzes the knowledge structure and development trends of this field from multiple dimensions, including countries, institutions, keywords, and temporal evolution. The results show that: (1) the application of AI in resource and environmental economics generally undergoes three stages: “technology introduction and method exploration”, “model deepening and policy evaluation expansion”, and “system integration and causal inference strengthening”. The research paradigm gradually shifts from prediction-oriented approach to mechanism identification and policy evaluation-oriented method. (2) Global research exhibits a dual-core pattern centered on China and the United States, and institutional collaboration networks continue to deepen. However, cross-regional and interdisciplinary collaboration still has room for improvement. (3) Current research hotspots primarily focus on monitoring, simulation, and prediction of environmental and economic systems, identification of causal effects of environmental policies, AI and sustainable transition, and distributional effects and just transition of climate policies. Based on the above findings, future research should focus on five directions: multimodal data fusion, model interpretability, climate well-being assessment, circular economy applications, and human-machine collaborative policy governance. Furthermore, relying on the National Comprehensive Observation System for Natural Resources, AI should be embedded into key scenarios such as risk early warning, policy evaluation, and regional collaborative emission reduction.
With the continuous advancement of the dual carbon goals and new energy systems, the large-scale development and grid integration of renewable energy make energy systems exhibit more complex spatiotemporal variation patterns. Artificial intelligence, due to its advantages in multimodal data fusion and nonlinear modeling, is becoming an important method for supporting the planning, operation, and risk management of new energy systems. This study systematically reviews the research progress of artificial intelligence technologies in the construction of new energy systems and risk response. Existing studies have shown that artificial intelligence methods significantly improve the accuracy of energy spatial information identification and system flexibility in areas such as wind and solar resource inversion, power grid modeling, demand response, and energy storage scheduling, providing essential technical support for the intelligent operation of new energy systems. However, current studies still face several key challenges, including the shortage of high-quality samples and reliable labels, high energy consumption during model training and inference processes, limited edge resources, and insufficient interpretability and credibility of artificial intelligence models. Future research needs to construct a geospatial foundation oriented system toward targeted energy demands, develop knowledge-enhanced large models and spatially interpretable reasoning methods, and build digital twin and intelligent decision making frameworks for uncertain scenarios, thereby providing innovative spatial decision-making support for the precise development of new energy systems and for the improvement of climate resilience of energy systems.
Artificial intelligence (AI) is increasingly becoming a pivotal driver of paradigmatic shifts in geographical research. With the widespread application of artificial intelligence technologies, geographical inquiry is undergoing a profound transformation from traditional empirical analysis to intelligent, data-driven research modalities. Using content analysis and the Latent Dirichlet Allocation (LDA) topic model, this paper systematically analyzes the scientific paradigms and frontier issues of AI-related geographical literature in recent years, and proposes priority directions for AI to empower future geographical research. The main findings are as follows: (1) At the paradigmatic level, the geographical empirical science research paradigm dominates existing literature, reflecting the field’s current focus on empirical synthesis and exploratory trajectory identification. Literature on the geographical system science research paradigm remains sparse, indicating AI application in complex geosystem simulation needs further deepening. Meanwhile, literature on the geographical positivist science research paradigm and geographical big data research paradigm is evenly distributed, showing parallel development of techno-empirical and data-driven approaches. (2) At the thematic level, 10 research themes are identified via the LDA model, including geographical intelligent application, urban spatial governance, and geographic data mining. Both Chinese and English corpora exhibit diversified themes, underscoring AI’s extensive penetration across geography’s sub-disciplines. (3) Regarding theme intensity, Chinese and English corpora show significant differences across research stages, each with distinct core themes. Chinese literature centers on geographic data analysis, urban spatial analysis, and spatiotemporal data analysis, emphasizing technological application and spatiotemporal expansion; English literature focuses on geographic data analysis, social space analysis, and spatial perception data analysis, reflecting a priority on socio-spatial dimensions and data sensing capabilities. (4) Concerning thematic evolution, Chinese and English literature present distinct divergent trajectories. Chinese literature follows a balanced evolutionary path of multidirectional diffusion and progressive deepening, while English literature shows stronger directional convergence, with topics gradually clustering around fields like human-environment interaction. (5) In terms of hot topics, three core research clusters are identified: geographic data mining and analysis, urban and socio-spatial applications, and human-environment relationship research, covering both technological methodological innovations and the discipline’s core focus on human-environment dynamics. For future research, AI’s empowerment of geography should prioritize deepening human-environment relationship research, optimizing Earth system research paradigms, breaking spatial analysis technology bottlenecks, and advancing regional research transformation.
[Objective] To advance urban energy transition and support the implementation of national development and security strategies, this study focuses on artificial intelligence and analyzes its impact on energy transition, with the aim of providing theoretical support and empirical experience for China’s urban energy transition. [Methods] This study constructed a theoretical analytical framework for artificial intelligence and urban energy transition. Based on panel data from 284 prefecture-level and above cities in China from 2011 to 2023, this study employed a two-way fixed effects model to empirically examine the impact of artificial intelligence on urban energy transition and its mechanisms. [Results] (1) Artificial intelligence could promote urban energy transition, and this effect was mainly achieved through three channels: labor optimization effect, green technological innovation effect, and virtual agglomeration effect. (2) Climate risk played a positive moderating role in the impact of artificial intelligence on urban energy transition, thereby strengthening the enabling effect of artificial intelligence. (3) The promoting effect of artificial intelligence on urban energy transition was more pronounced in central regions, resource-based cities, and shrinking cities. Moreover, the impact of artificial intelligence on urban energy transition exhibited a single-threshold effect, characterized by diminishing marginal returns. [Conclusion] Artificial intelligence is a core technological driver of urban energy transition. Therefore, it is necessary to accelerate its deployment and application in the energy sector, leverage its technological advantages to improve energy efficiency and reduce pollution emissions, and empower high-quality urban energy transition and sustainable development through digital and intelligent approaches.
With the intensification of climate change, emerging and re-emerging vector-borne diseases have become increasingly active worldwide, posing a continuous threat to human health. These diseases, mainly transmitted by arthropod vectors such as mosquitoes, ticks, and mites, have epidemiological processes that are generally influenced by natural environmental conditions, human factors, and biological factors, exhibiting pronounced regional distribution characteristics. In recent years, artificial intelligence (AI) technologies, driven by big data and algorithms, have developed rapidly, providing new opportunities for studying epidemiological patterns of vector-borne diseases. Based on a review of the spatiotemporal epidemiological characteristics of various common vector-borne diseases, this study reviews the development of AI-enabled research from three dimensions: data acquisition, analysis of influencing factors, and epidemic risk early warning, while also summarizing the challenges faced in this field. Finally, this study offers prospects for the application of AI in the study of epidemiological patterns of vector-borne diseases.
[Objective] To respond to the call of top-level design to cultivate leading enterprises in the “AI+” low-altitude economy, this study reveals the synergistic mechanisms of factors in the incubation of various types of technology-based enterprises, thereby providing theoretical and practical guidance for the application and cultivation of technology-based enterprise designations among “AI+” low-altitude economy enterprises. [Methods] A sample of 141 domestic listed companies simultaneously involving both artificial intelligence and the low-altitude economy was selected. Based on sample data from 2021 to 2024, fuzzy-set qualitative comparative analysis (fsQCA), necessary condition analysis (NCA), and machine learning methods were employed to investigate the necessary conditions and configurational pathways for the application of six types of technology-based enterprise designations: high-tech enterprises (HTE), enterprise technology centres (ETC), specialized-refined-distinctive-innovative enterprises (SRDI), manufacturing single-champion enterprises (MSCE), engineering technology research centres (ETRC), and technological innovation demonstration enterprises (TIDE). [Results] (1) At the element level, fsQCA and NCA indicated that no single element constituted a necessary condition for the cultivation of any designation type. Through comparison of path co-occurrence, it was found that intellectual property and enterprise qualifications played significant core driving roles in the pathways. Machine learning validated the non-necessity of all individual elements, as well as the relative importance and synergistic driving effects of intellectual property and qualifications. (2) At the pathway and enterprise level, fsQCA identified that ETC formed two incubation pathways, while HTE evolved four incubation pathways that could be grouped into one category according to core driving elements, and exhibited a configurational substitution effect with ETC pathways. SRDI, MSCE, TIDE, and ETRC each evolved one configurational pathway. The cultivation pathways of the first three tended to converge, with identical element composition, typical cases, and case numbers. [Conclusion] “AI+” low-altitude economy enterprises need to achieve technology-based transformation and upgrading by constructing technological advantages with property rights and qualifications, coordinating internal and external resources, and carrying out dynamic adaptation in conjunction with contingency factors, relying on equivalent incubation pathways to realize the cultivation and application of diversified technology-based designations.
[Objective] International low-carbon technology sharing is an important pathway to achieve global carbon neutrality. Comprehensively evaluating the carbon reduction potential of international low-carbon technology sharing, as well as its effects on reducing carbon neutrality costs, mitigating climate change, and promoting technological progress, holds significant theoretical and policy implications for future global climate cooperation. [Methods] This study constructed a climate-economy-technology integrated assessment model, CIECIA-GT, to comprehensively evaluate the carbon emission reduction potential of international low-carbon technology sharing policies and their climate governance effectiveness under the constraint of carbon neutrality from 2025 to 2100. [Results] (1) The global surface temperature rise in 2100 will decrease significantly as the degree of low-carbon technology sharing increases. Under the scenario of complete technology sharing, it can be controlled within 2.62 °C. (2) Under the scenario of complete technology sharing, developed economies will fully acquire low-carbon technology transfer due to their superior learning capabilities. In contrast, developing economies will be constrained by their own technological innovation and learning capabilities, leading to limited improvements in their technology levels. (3) The implementation of international low-carbon technology sharing policies can help countries reduce their reliance on carbon cap measures under the constraint of carbon neutrality, and increase cumulative economic utility significantly, and control their carbon neutrality costs. (4) Additional technological support from developed economies can effectively accelerate technological imitation in developing economies and further reduce carbon neutrality costs, but the marginal effect will be limited. [Conclusion] International low-carbon technology sharing holds significant potential for technological advancement and carbon emission reduction, but it cannot directly achieve climate change mitigation goals. Reducing patent protection duration can promote the participation of developed economies in international low-carbon technology transfer, and is more effective in mitigating climate change than reducing patent breadth. International low-carbon technology sharing can significantly reduce carbon neutrality costs, and additional technological support from developed economies to developing economies can further enhance the feasibility and equity of global carbon neutrality pathways, and therefore is of clear significance to climate governance. Combining low-carbon technology sharing with other high-cost emission reduction measures is a reasonable design approach for international carbon emission reduction cooperation.
[Objective] Controlling China’s agricultural non-point source pollution (ANPSP) faces multiple pressures, including technological innovation, government regulation, and agricultural structure adjustment. Investigating the impact of agricultural green transition on ANPSP and its spatial spillover effects provides scientific support for advancing regional collaborative governance of the agricultural ecological environment. [Methods] Using agricultural economic data and primary agricultural pollutant load data from 31 provincial-level units in China from 2004 to 2022, this study employed an extended spatial panel STIRPAT model to empirically examine the impact of agricultural green transition on ANPSP from the perspectives of technology, regulation, and structural factors. [Results] (1) Both agricultural technology application and innovation exhibited a rebound effect in their direct impact on ANPSP, but only agricultural technology innovation showed positive spatial spillover effects. (2) Command-and-control regulation showed no significant impact on ANPSP in the local province and neighboring provinces, whereas economic incentive-based regulation could significantly reduce ANPSP emissions in the local province and neighboring provinces. (3) Industrial structure adjustment had an inhibiting effect on ANPSP in the local province but a promoting effect in neighboring provinces. Crop structure adjustment increased the emission intensity of ANPSP in the local province and neighboring provinces due to the expansion of the non-grain cropping. By contrast, multiple cropping structure adjustment significantly increased the emission intensity of ANPSP in the local province. (4) From the perspective of heterogeneity, technology factors had a significant impact on ANPSP in the western region. The direct effect of economic incentive-based regulation on ANPSP was smaller in the eastern region than in the western region, while the indirect effect was largest in the central region, followed by the eastern and then the western region. Agricultural structure adjustment exerted varying magnitudes and directions of influence on ANPSP in the local province and neighboring provinces. [Conclusion] The impact of agricultural green transition on ANPSP is characterized by significant regional disparities and spatial correlations. Therefore, it is essential to appropriately reduce income disparities across regions, strengthen adaptive research and development of agricultural technology, optimize regional agricultural environmental policy instruments, and gradually establish a regional integrated collaborative governance mechanism for ANPSP control.
[Objective] Against the backdrop of ecological civilization construction imposing explicit requirements on agricultural green development, this study aims to evaluate the enhancement effect of the Mountain-Waters Project pilot policy on agricultural green development welfare. [Methods] Based on the theoretical framework of factor reallocation-value enhancement-environmental regulation, and using panel data from 1438 households in the Yangtze River Economic Belt from 2012 to 2022, a multi-period DID model was employed for empirical testing. [Results] (1) The Mountain-Waters Project pilot policy significantly improved agricultural green development welfare. (2) Mechanism analysis showed that the policy synergistically increased agricultural output and farmers’ income and reduced negative environmental externalities through the triple paths of production optimization, market value enhancement, and environmental protection, thereby significantly improving agricultural green development welfare. (3) Heterogeneity analysis showed that the policy effects were more pronounced in the upper reaches of the Yangtze River, non-provincial boundary areas, and non-major grain-producing areas, highlighting the key roles of policy prioritization, administrative coordination, and differences in agricultural structure in policy transmission. [Conclusion] The Mountain-Waters Project pilot policy plays a positive role in promoting agricultural green development welfare, but the policy effect exhibits significant heterogeneity. Therefore, policy design should be optimized according to regional differences, and green technical support and cross-regional collaborative governance should be strengthened, so as to provide a scientific basis and policy implications for ecological civilization construction and sustainable agricultural development.
[Objective] The agro-pastoral ecotone is ecologically sensitive with a diverse and complex food supply. Clarifying its food supply patterns and resource-environmental effects is crucial for ensuring regional food security. [Methods] Taking Hailar District, Manzhouli City, and Ergun City in Hulunbuir City, Inner Mongolia as case sites, this study quantitatively calculated the regional food supply quantities, along with the quantities of inflows and outflows, based on food source data from the previous year obtained from field surveys in 2023, combined with statistical data and literature sources. The spatial patterns of food supply were analyzed, and the ecological footprint indicator was applied to evaluate the resource and environmental costs caused by the bidirectional food flows. [Results] (1) The total food supply of the case sites in 2022 was 494800 tonnes, dominated by plant-based foods (82.45%). The overall supply structure was characterized by “self-sufficiency with surplus in grains and dairy but deficit in vegetables and fruits.” (2) Food supply exhibited a bidirectional flow pattern, both internal and external, with local foods supply accounting for only 15.21%, indicating high dependence on Heilongjiang, Shandong, and Liaoning provinces. (3) The total ecological footprint of food supply in the case sites reached 316000 hm2, equivalent to 1.55 times the cultivated land area of the three areas. Animal-based food accounted for 67.95%, and its contribution to the cropland footprint reached 59.39%, revealing a “grain-for-meat” resource conversion pattern. (4) The ecological footprint was predominantly local (56.50%), with significant outward transfer characteristics. The outflow footprint was 4.79 times the inflow footprint, making the case sites a “net output area” of ecological footprint. The land use types were consistent with the semi-agro and semi-pastoral production characteristics. [Conclusion] Food supply in the agro-pastoral ecotone shows clear structural mismatches and spatial shifts in resource-environmental costs. It is recommended to improve local vegetable and fruit supply capacity, diversify supply channels, and promote green transformation of the supply chain to ensure the sustainable and stable supply of the regional food system.
[Objective] This study aims to investigate how to ensure the sustainable supply security of China’s strategic mineral resources under the dual challenges of intensified major power intervention and rising policy uncertainty in mineral resource-rich countries [Methods] This study constructed a tripartite evolutionary game model involving mineral resource counties, the United States, and China to analyze the interactive evolution paths of strategies among the three parties and the conditions for their stability. [Results] (1) Under the coupling of multiple interests and dynamic equilibrium, the system exhibited a stable equilibrium strategy in which the mineral resource-rich countries maintained cooperation, the United States refrained from intervention, and China responded conservatively. (2) The intervention behavior of the United States had a significant benefit-driven characteristic, and the increase in its marginal benefits strengthened the intervention motivation, which might prompt mineral resource-rich countries to alter their cooperation pattern with China. (3) Mineral resource-rich countries faced internal structural constraints in policy transformation, and the presence of resource nationalism costs significantly reduced the likelihood of changing their cooperation patterns with China. (4) China’s combination of “incentive-punishment” policy tools could effectively curb the strategic shift tendency of mineral resource-rich countries and constrain the intervention space of the United States. [Conclusion] This study reveals that in the geopolitical game of strategic mineral resources, the economic constraints of the U.S. intervention and the internal conflict costs of mineral resource-rich countries are the two key mechanisms for maintaining cooperation stability. Therefore, China should establish a strategy system that responds dynamically to the game state, decisively employing response strategies—such as countermeasures and interest binding—when risks are high to ensure security, and returning to cost-optimal conservative response strategies when stability is approaching, so as to maximize long-term interests.
[Objective] This study aims to reveal the well-being effects of the integrated development of the cultural and tourism industry, thereby providing a theoretical basis and decision-making reference for optimizing the policy system of cultural and tourism integration and achieving the sustainable improvement of residents' well-being. [Methods] Based on the panel data from 31 provinces in China from 2006 to 2022, this study constructed a theoretical analytical framework for enhancing residents' well-being through the integrated development of the cultural and tourism industry. It comprehensively applied panel regression and threshold models to systematically explore the impact of cultural and tourism integration on residents' well-being and its spatiotemporal heterogeneity. [Results] (1) During the study period, the level of cultural and tourism industry integration showed an overall fluctuating upward trend, and spatially exhibited a stepwise declining distribution pattern from east to west. (2) The level of residents' well-being generally showed a stable year-by-year growth trend, and spatially exhibited a differentiation pattern characterized by high-value agglomeration in the southeastern coastal areas and low-value agglomeration in the northwestern inland areas. (3) The integrated development of the cultural and tourism industries had a significant promoting effect on residents' well-being, and its impact differed across different dimensions, with the intensity showing a differentiated pattern of social well-being effect > environmental well-being effect > economic well-being effect. (4) The integrated development of the cultural and tourism industry had a significant double-threshold effect on residents' well-being, and exhibited a nonlinear relationship characterized by “steady advancement—significant enhancement—slowing growth”. The regional heterogeneity analysis showed that the promoting effect of the integrated development of the cultural and tourism industry on residents' well-being in the eastern and central regions was significantly higher than that in the western region, with the eastern region having the largest well-being effect. [Conclusion] The integrated development of the cultural and tourism industry has a promoting effect on the enhancement of residents' well-being. It is necessary to scientifically determine the appropriate range of its development, formulate differentiated development pathways for cultural and tourism integration according to local conditions, and adhere to the enhancement of social well-being as the core objective, so as to effectively promote the improvement of residents' livelihood and realize common prosperity.
[Objective] This study, from an endogenous development perspective, reveals the impact of pluriactivity on fishermen’s willingness to exit fishing and provides micro-level evidence for optimizing policy instruments and implementation pathways for fishing exit and livelihood transition. [Methods] Using 2024 survey data from 822 fishing households in three coastal cities and eight counties of Zhejiang Province, this study employed OLS, ordered Probit, and an instrumental variable approach to examine the impact and mechanisms of pluriactivity on fishermen’s willingness to exit fishing. [Results] (1) Pluriactivity significantly increased fishermen’s willingness to exit fishing, indicating that pluriactivity was not only an adaptive strategy for fishermen to respond to changes in the external fishery environment, but also an important factor influencing their willingness to exit fishing. (2) Mechanism analysis showed that pluriactivity effectively increased fishermen’s willingness to exit fishing by reducing risk perception of unemployment and improving human capital. (3) The impact of pluriactivity exhibited heterogeneity across different groups of fishermen. First, the effect of pluriactivity on fishermen’s willingness to exit fishing was more pronounced among fishers with higher education levels and local fishers. Second, the effect of pluriactivity was more pronounced among fishermen engaged in non-fishing sectors. [Conclusion] This study reveals the key role of pluriactivity in shaping fishermen’s willingness to exit fishing, enriches the understanding of livelihood transition mechanisms, and provides empirical evidence for constructing endogenous development-oriented policy pathways. Accordingly, it is recommended that policies should focus on exploring and cultivating the endogenous development capacity of fishermen engaged in fishing, rather than focusing only on the resettlement of fishermen who have exited fishing.
[Objective] This study aims to construct an accounting framework for carbon emission responsibilities from agricultural energy consumption that integrates production, consumption, and income responsibility perspectives. It proposes an allocation scheme for carbon emission responsibilities that fully reflects the upstream and downstream relationships in the agricultural production chain, thereby providing a basis for formulating fair and effective agricultural carbon emission reduction policies. [Methods] Using the environmentally extended multi-regional input-output (EE-MRIO) model, this study measured carbon emissions from agricultural energy consumption in 30 provinces of China in 2017 from production-, consumption-, and income-based responsibility perspectives. A shared tensor was introduced to modify the Leontief model and the Ghosh model to assess the carbon emission responsibilities of agricultural energy consumption under the shared responsibility principle. [Results] (1) The total carbon emissions under the producer responsibility perspective were higher than those under the consumer responsibility perspective, whereas emissions under the income-based responsibility perspective were the lowest. (2) Compared with the integrated principle, both the income-weighted responsibility principle and the consumption-weighted responsibility principle failed to achieve a reasonable allocation of carbon emission responsibilities. (3) Under the integrated principle, the carbon emission responsibility of major grain-producing areas was 69.2237 million tons, representing a reduction of 22.04% and 23.14%, respectively, compared with the income-weighted responsibility principle and the consumption-weighted responsibility principle. The carbon emission responsibility of major grain-consuming areas was 19.64 million tons, representing increases of 4.34% and 17.53%, respectively, compared with the income-weighted responsibility principle and the consumption-weighted responsibility principle. [Conclusion] The allocation of carbon emission responsibilities from agricultural energy consumption should comprehensively consider the upstream and downstream impacts of the industrial chain and establish an allocation mechanism based on the integrated responsibility principle. It is recommended to formulate and improve policies related to agricultural carbon emission reduction, clarify carbon emission reduction targets and responsibilities of each responsible entity, and promote cross-regional agricultural cooperation and technological exchange to jointly advance the process of agricultural carbon emission reduction.
[Objective] Under the background of accelerating domestic economic circulation, this study investigates how carbon emission reduction technologies influence spatial correlations of carbon emissions in urban agglomerations through regional production networks, thereby facilitating the low-carbon transformation of regional economic circulation. [Methods] This study constructed a multi-sector general equilibrium model incorporating a pollution production function to elucidate the spillover mechanisms of carbon emission reduction technologies in the regional production network. By integrating spatial econometric models and input-output analysis, it investigated the network correlation effects of carbon emission reduction technology spillovers, the spatial structure characteristics of carbon emissions under technology spillovers, and their transmission pathways in the Yangtze River Delta urban agglomeration from 2006 to 2021. [Results] (1) Carbon emissions in the Yangtze River Delta urban agglomeration exhibited significant spatial agglomeration effects and temporal pathway dependence under production network transmission. The technological progress of carbon emission reduction generated cross-regional synergistic emission reduction effects through the production network, with demand-side-driven spillover effects being more prominent than supply-side-driven ones. (2) Spatial heterogeneity was observed in carbon emission reduction technology spillover effects, which were significant in central cities and economically developed regions, whereas peripheral areas showed weaker spillover effects due to constraints on innovation resources. Economic development and population agglomeration showed positive network spillover effects on carbon emissions. (3) The carbon emission network in the Yangtze River Delta urban agglomeration showed pronounced intra-administrative clustering characteristics. At the interprovincial level, Jiangsu and Anhui functioned as carbon emission exporters, while Shanghai and Zhejiang acted as carbon emission importers. In the carbon emission transmission path starting from Jiangsu and ending at Shanghai and international demand, Zhejiang and Shanghai assume important intermediate production functions and become the key hub nodes connecting production and consumption. [Conclusion] Carbon emission reduction technologies generate significant spatial spillover effects through the production network, serving as a crucial mechanism for promoting coordinated regional emission reduction. It is recommended to strengthen the construction of interregional green technology diffusion channels, optimize cross-regional carbon emission responsibility-sharing mechanisms and joint prevention and control systems, and refine full-chain carbon reduction solutions based on the production network, thereby advancing the green transformation of the overall socioeconomic development of the Yangtze River Delta urban agglomeration.
[Objective] Driven by both the “dual carbon” goals and the development of new quality productive forces, industrial transformation has become a core approach for enhancing carbon emission efficiency. From the perspective of industrial evolution, this study aims to investigate how the process of urban industrial transformation and upgrading affects carbon emission efficiency. [Methods] Taking the Yangtze River Delta urban agglomeration as an example, this study constructed a panel model based on Qichacha enterprise data and carbon emission data from 2013 to 2022 to reveal the impact of urban industrial pathway evolution (relatedness density of newly entered industries) and industrial capacity evolution (complexity of newly entered industries) on urban carbon emission efficiency and their heterogeneous effects. [Results] (1) Carbon emission efficiency exhibited significant spatial differentiation, and high-efficiency cities were mainly distributed along the Shanghai-Nanjing-Hangzhou corridor, coastal and bay areas, and certain peripheral cities. (2) Both industrial paths and industrial capacity showed differentiated evolutionary characteristics across cities, with evident divergence in their pace of evolution. (3) Urban industrial pathway breakthrough was positively associated with carbon emission efficiency, whereas the leap of industrial capacity toward high-complexity fields was negatively associated with carbon emission efficiency. (4) Grouped regression further showed that the relationship between urban industrial evolution and carbon emission efficiency had stage heterogeneity. The improvements in industrial capacity were generally negatively correlated across different stages of urban development, and the strength of the negative correlation gradually weakened as economic development increased. In cities at the medium economic development stage, the negative impact of industrial pathway dependence on carbon emission efficiency was more pronounced. [Conclusion] The findings reveal the complex relationship between industrial transformation and upgrading and carbon emission efficiency, provide supplements and corrections to the understanding that “industrial upgrading inevitably promotes green and low-carbon development”, and offer theoretical and practical references for evaluating the carbon reduction effects of industrial policies.
