Since the industrial revolution, the global economy has shown exponential growth. Resource-intensive industries in the manufacturing industry rely more on fossil fuel and emit a great deal of carbon dioxide, which has a series of negative effects on the ecological environment and global climate. Global warming threatens the environment on which people depend. On December 11,1997, the third Conference of the Par-ties to the ' United Nations Framework Convention on Climate Change ' held in Tokyo and adopted the ' Kyoto Protocol ', which sets standards for countries ' carbon dioxide emissions : Between 2008 and 2012, the industrial carbon emissions of the world 's major industrial countries were on average 5.2% lower than their 1990 emissions. The treaty entered into force in 2005. In 2016, 178 parties around the world jointly signed the Paris Agreement. The long-term goal is to control the increase of global average temperature within 2 degrees Celsius compared with the pre-industrial period, and to limit the increase of temperature to 1.5 degrees Celsius. On September 3,2016, the Standing Committee of the National People 's Congress approved China 's accession to the Paris Climate Change Agreement. The implementation of the agreement empha-sizes that equality, common but differentiated responsibilities and respective princi-ples will be reflected according to different national conditions.
In order to be a responsible country, China actively participates in activities to reduce carbon emissions and contribute to global warming. In September 2020, the State Council issued ' China 's Energy Development in the New Era ', proposing to ' improve energy efficiency in key areas, actively optimize the industrial structure, and vigorously develop advanced manufacturing industries with low energy consumption. ' In September 2021, the State Council issued the ' Opinions on Completely and Accu-rately Implementing the New Development Concept and Doing a Good Job in Carbon Neutralization of Carbon Peaks ', which proposed that ' Formulate carbon peak implementation plans for energy, steel, non-ferrous metals, petrochemical, chemical, building materials, transportation, construction and other industries and fields '. China 's manufacturing industry is a major carbon emitter, and carbon emissions account for 67% of China 's total carbon emissions. The effective implementation of carbon emission reduction policies in the manufacturing industry will help China 's overall social ' double carbon ' goals. The implementation of carbon emission reduction targets re-quires sub-industry and sub-regional research. The manufacturing system is huge, and the carbon reduction targets and energy conservation and emission reduction policies faced by different sub-sectors must be specifically analyzed by specific industries. Therefore, by exploring the spatial and temporal evolution rules and influencing factors of carbon emissions in China 's manufacturing industry, it is an urgent problem to explore what kind of industry policies are designed to achieve peak carbon neutrality.
The research on the spatial and temporal distribution and evolution characteristics of manufacturing carbon emissions is mainly carried out by province and industry. Using cluster analysis method, 30 provinces in China are divided into high-value and low-value types. Obviously, the spatial pattern of manufacturing carbon emission distribution and its changing trend in a certain period of time ( Wang Xia et al., 2020 ; Chen et al., 2021 ), sub-sector one is to divide the manufacturing industry into 28 sub-sectors, analyze the sub-sector 's carbon emissions over time and analyze the reasons ( Wang and Nie, 2012 ; Tang et al., 2019 ). The other is to divide each sub-industry into capital, technology and labor intensive according to factor intensity, and analyze the differences and causes of carbon emissions ( Sun et al., 2012 ; Chen et al.,2013; Wang Xia et al., 2020 ; Wang et al, 2015 ). Furthermore, through the carbon emission data of manufacturing industry in 30 provinces, the Theil index is used to study the regional differences of carbon emissions, and the spatial distribution characteristics are described by spatial autocorrelation model ( Liu et al., 2022 ). The spatial convergence of carbon emissions in manufacturing industry is also analyzed by spatial econometric model ( Wang et al., 2022 ). However, the construction of spatial weight matrix is not comprehensive enough, and the influence of the double nesting of distance and geographical distance on the spatial weight matrix is not considered comprehensively, and the selected influencing factor indicators need to be further discussed.
In terms of driving factors, it is generally believed that scale factors such as gross national product, total population and energy consumption and investment structure, energy structure and industrial structure ( Fu et al., 2021 ; Shao S et al., 2017 ; Hammond and Norman, 2012 ) have a significant impact on manufacturing carbon emissions. In the research method, it is divided into factor decomposition model and econometric model. In the factor decomposition model, the decomposition method is divided into PDA ( production theory decomposition analysis ) ( Ding et al., 2021 ), SDA ( structural decomposition method ) ( Guan et al., 2008 ; Cansino, 2016 ; Pan, 2011 ) and IDA ( index decomposition method ) ( Andreoni, 2016 ). In exponential de-composition analysis, logarithmic mean Divisia decomposition ( Ang, 2004 ; Wang et al., 2005 ; Deng et al., 2014 ) and generalized Divisia index decomposition ( Vaninsky, 2014 ; Shao et al., 2017 ; Zhang et al., 2021 ) has been widely used for its unique ad-vantages. The econometric model initially used the traditional multiple linear regression as the main model to analyze the influencing factors of carbon emissions in the manufacturing industry, such as Logistic model, STIRPAT model ( Liu and Xiao, 2018; Jin and Han, 2021 ), generalized impulse response function, generalized least squares method and support vector regression(Song J K, 2012).
The existing research has laid a good foundation for this article and provided very valuable theoretical and methodological support, but it needs to be further improved. The possible innovation of this paper is to select the data of manufacturing sub-sectors and sub-provinces in the China Carbon Emission Database ( Ceads ), analyze the temporal evolution of sub-sector data, and use the exploratory spatio-temporal data anal-ysis method to explore the spatial correlation between manufacturing carbon emissions and neighborhoods in each province. On this basis, the spatial-temporal transition measurement method is used to analyze whether the local spatial structure of manufacturing carbon emissions in each province is stable and whether there are conditions for spatial transition between 2005 and 2019. The absolute β convergence and conditional β convergence test whether the manufacturing carbon emissions at the na-tional and regional levels can tend to be stable. On this basis. The spatial Durbin model combined with spatial error and spatial lag effect is used to analyze the influencing factors of carbon emissions in China 's manufacturing industry, and the policy suggestions for carbon emission reduction in manufacturing industry are put forward.
The existing research has laid a solid foundation for this article and provided very valuable theoretical and methodological support, but it needs to be further improved. The possible innovation of this paper is to select the data of manufacturing sub-sectors and sub-provinces in the China Carbon Emission Database ( CEADs ), analyze the temporal evolution of sub-sector data, and use the exploratory spatio-temporal data analysis method to explore the spatial correlation between manufacturing carbon emissions and neighborhoods in each province. On this basis, the spatial-temporal transition measurement method is used to analyze whether the local spatial structure of manufacturing carbon emissions in each province is stable and whether there are conditions for spatial transition between 2005 and 2019. The absolute β convergence and conditional β convergence test whether the manufacturing carbon emissions at the national and regional levels can tend to be stable. On this basis. The spatial Durbin model combined with spatial error and spatial lag effect is used to analyze the influencing factors of carbon emissions in China 's manufacturing industry, and the policy suggestions for carbon emission reduction in manufacturing industry are put forward.