Study On The Spatial-Temporal Pattern Of Energy-Related Carbon Emission And Its Influencing Factors In Prefecture-Level Of Guangdong Province

At present,global warming is a hot issue of concern to the international community,and one of the important reasons is the increasing concentration of CO₂（carbon dioxide）in the atmosphere.The carbon emissions of China have remained high for a long time,under the constraints of the national"carbon peak"and"carbon neutral"targets,China is facing the double pressure of economic development urgency and carbon emission reduction tasks.However,in terms of data sources for CO₂emissions,the incomplete energy data in the statistical yearbooks collected in China in the past decades,it is only possible to calculate the national or provincial scale using the IPCC methods.Therefore,carbon emission data at the urban or smaller scale are more difficult to obtain.Accurate estimation of urban CO₂ emissions and exploration of main driving factors are of great importance for achieving low-carbon development.Though NTL has been widely used as a proxy for estimation and spatialization of carbon emissions,This paper applied DMSP-OLS-like,which have immense potential to effectively evaluate carbon emissions.However,the data set disregarded the saturation effect of DMSP-OLS data during the construction of the long time-series data set which prevents its application and estimation accuracy.In view of the weakness of DMSP-OLS-like nighttime light data,this paper proposed an index named VANUI,based on NDVI product data.Then exploring the spatiotemporal dynamics at prefecture level in Guangdong during the last 20 years.Finally,the influencing factors and their effects change were identified by the random forest model.The results are as follows:（1）We have developed a vegetation-adjusted nighttime light index VANUI,which combines DMSP-OLS-Like,NDVI and water body data.It not only effectively reflects the spatial changes in human activities but also calibrate the saturation and overflow effects of DMSP/OLS nighttime light data.（2）Based on VANUI,a method for rapid and accurate monitoring of carbon emissions of energy consumption in city-level was constructed.This article first uses the IPCC accounting method to systematically calculate the carbon emissions from energy consumption in Guangdong Province from 2000 to 2019.Then combing the index named VANUI to conduct a panel model.As a result,realizing rapid and accurate estimation of 1km grid-scale carbon emissions in Guangdong Province from 2000-2019.Meanwhile,the city-level carbon emission was estimated in Guangdong,providing effective data support for long-term carbon emission research of prefecture-level cities.（3）We explore the spatial and temporal patterns of carbon emissions of energy consumption in Guangdong prefecture-level cities.The results show that,there was an overall upward trend in carbon emissions of energy consumption in Guangdong Province,but the intensity of energy consumption carbon emissions gradually decreased during the period2000-2019.And based on the trend index SLOPE,it can be seen that there are large differences in the growth rate of carbon emissions among the prefecture-level cities in Guangdong Province,among which Guangzhou,Foshan and Huizhou are of the rapid growth type.The global Moran’s I index showed that the spatial clustering of total energy carbon emissions in Guangdong province initially increased and then decreased during the past 20 years.LISA index was further used to conduct a local analysis of CO₂ emissions in Guangdong Province.It can be found that the spatial pattern of local autocorrelation of carbon emissions at the city-level of Guangdong Province is basically unchanged,with most prefecture-level cities in the PRD region of Guangdong Province being of the high-high agglomeration type and Meizhou City being of the low-low agglomeration type;in terms of quantity,the number of high-high agglomeration types has decreased,indicating that the transformation of the socio-economic development model in high-carbon emission areas of the Pearl River Delta has resulted in a slowdown of carbon emissions growth,leading to a weakening of the local autocorrelation binary pattern of carbon emissions in Guangdong province.（4）Research on the factors influencing the carbon emission intensity of energy consumption in Guangdong Province was carried out.based on the random forest model in machine learning.As results indicated,among the key influencing factors,number of population,foreign investment,and the proportion of secondary production are more important than others.In the importance rankings over time,the importance of road network density and per capita GDP has increased significantly,while the importance of foreign investment has decreased significantly.In addition,the results of partial dependence analysis showed that there was a nonlinear response between carbon emissions and the driving factors.Ultimately,it provides a scientific basis for the differentiated and targeted policy of carbon emission reduction in Guangdong Province.