区域间资源禀赋、主体功能、产业格局及经济发展水平的梯度差异往往推动了城市间产业转移。纵观国内外区域间产业转移，我们发现无论是产业发展需求还是政策引导，产业转移倾向于将高耗能、高排放企业转移至环境标准较低地区，这对区域间经济发展、资源利用造成深远影响，使得区域碳排放格局发生转变。珠三角城市群是粤港澳大湾区的重要组成部分，是国家建设世界级城市群和参与全球竞争的重要空间载体，相应的绿色低碳发展规划需要进一步细化落实。 基于此，本研究以珠三角地区为研究案例，从空间地理学角度构建区域产业转移碳排放空间关联理论模型，对2005年至2015年珠三角城市群区域间产业碳排放空间关联进行实证研究。本研究通过区域间空间差异、时空演变等因素，揭示产业碳排放空间动态变化特征，可以通过探究空间关联的内在机理以及发展规律，把握区域间经济发展规律，促进城市群低碳协调发展。主要研究内容及结论如下： 产业碳排放空间关联模型构建。本研究提出的产业碳排放空间关联模型包括碳排放核算、驱动机制、时空分布和空间关联四部分组成。通过产业转移理论、碳排放测度模型，解析产业碳排放结构特征。考虑空间差异情况，利用多区域因素分解法对城市群产业碳排放演变驱动机制研究剖析，利用标准差椭圆与Theil系数法研究产业碳排放空间动态分布。结合以上研究，构建修正的引力模型，形成产业碳排放空间关联研究框架。 城市群产业能源碳排放测度及其结构特征研究。首先从产业结构方面入手，利用LMDI(对数均值迪氏分解方法)方法从产业层面区域碳排放整体格局进行分析，得出引起行业能源碳排放变化的主要影响因素。同时利用产业关联分析中的假设抽取法对产业间隐含碳排放进行分析，从供给侧与需求侧探寻产业部门主要碳排放消费端，从而解析出产业转移部门类型中区域碳排放变化的驱动因素。研究发现，人口增加因素、经济增长因素总体上驱动碳排放量增加，由于技术进步带来的能源强度的减少，根本上抵消了经济增长因素对碳排放增加量的影响。 城市群产业碳排放时空演变驱动机制研究。根据产业结构分析，从时空演变方面，本部分从经济增长、产业结构和空间分布三个方面研究了城市群产业转移下的能源消费碳排放，讨论了不同要素密集型行业碳排放的模式和趋势，利用多区域LMDI方法分解碳排放的驱动力，以计算产业转移两大特征中经济增长和产业升级所产生的碳排放变化。研究发现在广州、佛山等发达城市，产业结构因素在减少碳排放方面发挥了正向作用，而在接收资本密集型产业转移的欠发达城市，如江门、惠州等，产业结构因素引起碳排放增加。产业转移不仅导致工业总规模的变化，使经济落后地区的增长速度超过发达地区，还降低了整体碳排放强度。发达地区通过降低要素成本来改进技术创新以降低能源消耗，经济落后地区通过接收资本密集型工业造成碳排放继续增加。 城市群产业碳排放空间差异研究。结合时空演变分析，本部分从空间差异角度进行分析对不同区域间工业碳排放不平等研究，测度城市年际间工业碳排放空间随时间变化情况，探讨碳排放量时空分布特征，建立人均工业碳排放不平等研究模型，检验城市产业发展中各个因素对城市间碳排放不平等现状的主要影响因素。在对各区域分配减排责任时，应考虑各地区的能源强度差异以及资源禀赋的差异，不能仅仅考虑地区间经济发展水平的不同。 城市群产业转移碳排放空间关联研究。综合碳排放核算-驱动机制-空间差异分析，本部分为对碳排放的空间格局和集群现象有更好的描述，通过选取珠三角城市群及周边城市各市面板数据，以碳排放作为衡量指标，从空间关联效应对产业转移下各城市碳排放的空间格局和集群现象进行了更加深入的分析，以揭示城市间碳排放之间的空间关系。研究发现，产业碳排放空间关联网络密度整体上呈现广州-深圳两极，佛山-东莞支撑的特点，显著的“核心—过渡—边缘”三级圈层结构特征。核心与边缘城市在产业发展的空间差异性和不平衡性，使得产业结构调整存在梯级转移诉求，对周边城市提供了承接产业转移动力，影响产业碳排放空间格局。借助空间经济理论与方法构建的产业碳排放空间关联模型能有效指导低碳空间管控策略的选择，可以促进产业转移下生产要素空间配置效率的提升，从而优化区域间产业碳排放的空间格局，具有一定的理论意义和实践价值。

Due to the large differences in the resource endowments, the main functions, the industrial patterns and the economic development levels between regions, the regional economic gradient emerged, it provides an intrinsic motivation for industrial transfers between cities in China. The process of the industrial transfer tends to shift high-energy, high-emissions enterprises to the areas with lower environmental standards, it also has far-reaching effects on the inter-regional economic development and the resources utilization, which has transformed the regional carbon emission patterns. Therefore, while arranging regional carbon emission responsibilities, the regional policy-makers should consider the transfer of carbon emissions concealed in the industrial transfer. It meas that when consider the pattern of industrial carbon emission in the regions, the various factors in the attention to regional sparital linkages should also be estimated. Thus, the study of the spatial linkages of carbon emissions will help to clarify the emission reduction responsibility more fairly and objectively, as well as coordinating economic development and sustainable use of resources from the regional macro level. As one of the three major metropolitan areas in China, the Pearl River Delta urban agglomeration aim to achieve low-carbon development. The target set for urban agglomeration low-carbon development should also consider bout the relationship with the cities around. Based on the target, this study takes 9 cities in the urban agglomeration of Pearl River Delta from 2005 to 2015 as the research sample, supplemented by the other 12 peripherial cities in Guangdong and constructs a theoretical model of carbon spaitial linkage analysis under the regional industrial transfer from the perspective of geoeconomics. Inter-regional spatial differences and factors circulation were analyzed to conduct the empirical study on the spatial linkage analysis of carbon emissions between different regions. The main research contents and conclusions are followed: The carbon spatial linkage model under industrial transfer of the urban agglomeration. The spatial linkage model of industrial carbon emission firstly presented in this study includes four parts: carbon emission accounting, driven mechanism, space-time distribution and spatial linkage analysis. The characteristics of industrial carbon emission structure were firstly analyzed through industrial transfer theory and carbon emission measurement model. Considering the spatial difference, the multi-region factor decomposition method is used to analyze the driving mechanism of urban carbon emission evolution in urban agglomerations. The standard deviation ellipse and Theil coefficient method are used to study the spatial dynamic distribution of industrial carbon emissions. Combined with the above research, a modified gravity model is constructed to form a spatial correlation research framework for industrial carbon emissions. Research I. Research on the regional industrial energy carbon emission measurement and its structural characteristics. From the scale of the industrial structure, the LMDI method (Logarithmic Mean Divisia Index) was used to analyze the overall pattern of regional carbon emissions, and to analyze the main influencing factors which caused the changes in the carbon emissions. Meanwhile, industry linkage analysis was conducted and the hypothetical extraction method was used to identify the implicit carbon emissions between sectors, so as to analyze the driving factors of regional carbon emission change from the types of industrial transfer sectors. The study found that the carbon emissions from the fossil consumption continued to increase during the “11th Five-Year Plan” period, but the growth rate slowed down and remained unchanged during the “12th Five-Year Plan period”. Among which the industrial sectors were the most important contributor to the total amount of the carbon emissions, accounting for more than 80%, and the industrial factors are also the key point for industrial restructuring and low-carbon development. The population increase factors and economic growth factors generally drove the increase of carbon emissions. Due to the reduction of energy intensity brought by technological progress, the increase in carbon emissions by the economic growth factors are basically offset. Research II: Research on the driving mechanism of industrial carbon emission evolution in urban agglomerations. This part discussed the carbon emissions changes under the industrial transfer in the urban agglomerations from three aspects: economic growth, industrial structure and spatial distribution. The patterns and trends of carbon emissions in different industrial sectors were discussed, and the multi-regional LMDI method was used to decompose carbon emissions. The driving forces were to be calculated for the changes in carbon emissions generated by economic growth and industrial structure, the two characteristics of industrial transfer. The study found that both economic growth effects and industrial structure effects played important roles in the changes in carbon emissions. Among them, the economic growth effect has led to rapid growth of carbon emissions, and industrial structure effects, especially for labor and technology-intensive industries, helped to alleviate the increasing carbon emissions. From a spatial perspective, the regional differences in urban structure carbon emissions were rather obvious. In developed cities such as Guangzhou and Foshan, this effect has played a positive role in reducing carbon emissions, while in less developed cities which accepted capital-intensive industries, such as Jiangmen and Huizhou, the industrial structure effect led to an increase in carbon emissions. In general, industrial transfer not only led to the changes in the total size of the industry, but also made the economically backward regions grow faster than the developed regions, and reduced the overall carbon intensity. Developed regions improved technological innovation by reducing factor costs while developing regions continued to increase carbon emissions through the acceptance of the capital-intensive industries. Research III: Difference analysis of the spatial patter of the industrial carbon emissions. This study focused on the spatial difference of industrial carbon emissions between the urban agglomeration and peripherial cities of the Pearl River Delta, by measuring the spatial variation of industrial carbon emissions among cities, to explore the temporal and spatial distribution characteristics of the industrial carbon emissions, and to establish a research model of per capita industrial carbon emission inequality. To test the main influencing factors of various factors in urban industrial development on the inequality of carbon emissions among cities. The study found that there is a significant difference between the movement trajectory of the gravity centers of the industrial value added and the industrial carbon emissions. During the “11th Five-Year Plan period”, the displacement continued changing, and the movement during the “12th Five-Year Plan” period was relatively stable. Among them, the gravity center of the industrial added value is expanding to the north, and the gravity center of the industrial carbon emissions is expanding toward the east. The long-axis standard deviation of the standard deviation ellipse has been increasing, indicating that the carbon emissions in the region during the study period are scattered in the northeast-southwest direction, it mainly due to the rapid increase of industrial carbon emissions in the Eastern Guangdong and Western Guangdong regions. For the analysis of per capita carbon emission inequality, the inequality of industrial carbon emissions in the study area generally declined, and remained stable during the “12th Five-Year Plan period”. The inequality between the urban agglomeration and the peripheral cities was close to zero. When analyzed the driving forces affecting the inequality of the per capita industrial carbon emissions, it is found that the differences of the economic development level effect and the industrial structure effect between regions are the primary causes of the decline in regional inequality, contributing more than 200%. On the contrary, carbon emissions inequality caused by differences in energy intensity effects between regions continues to increase. The analysis of the spatial pattern of industrial carbon emissions provides a basis for formulating regionally differentiated energy consumption and carbon emission reduction policies. Research IV: The carbon spatial linkage research under the industrial transfer. In order to have a better description of the spatial pattern and clustering phenomenon of carbon emissions in various cities, the data in this paper were mainly panel data of various cities in Guangdong. The pattern and clustering phenomenon were further analyzed to reveal the spatial relationship between carbon emissions in cities. The study found that the spatial connection between urban agglomeration and peripheral cities in the Pearl River Delta has significant hierarchical characteristics. With the further development of the inter-regional economy, the carbon-emission spatial linkage network has been transformed from a “linear radiation”, such as Guangzhou-Shenzhen, to a “facial radiation”, a multi-center network. By 2015, a prominent “core-halfedge-edge” structure and a three-level circle structure with Guangzhou as the core were formed. Among them, Guangzhou and Shenzhen, the first pole of the urban agglomeration of the Pearl River Delta, have been at the core of the network, and the dominant radiation in Foshan, Dongguan and other regions is more significant gradually. The second level serves as a bridge, including the surrounding city within the Pearl River Delta, and the third level is an industrial transfer inflow area. It is composed of Zhaoqing and the peripheral cities of Western, Northen and Eastern Gauangdong, and its industrial spatial linkage is weak. The spatial differences and imbalances in the development of the urban agglomeration and the perpirpheral cities have led to the transfer of industrial structure adjustments, and provided the driving force for industrial transfer to the devleloping cities. The study of the carbon spatial linkage can be used as a reference for adjusting the spatial organization model of regional economic development and energy carbon emission. It can help to optimize the spatial allocation mode of regional production factors, to promote the spatial allocation efficiency of production factors under industrial transfer, and to provide new opportunities for regional economic growth.