Temporal And Spatial Variation Characteristics And Driving Forces Of Carbon Emission From Energy Consumption In China Using DMSP/OLS Nighttime

The problems caused by global warming have raised extensive concern of the international community.The emission of greenhouse gases,especially CO2,is the most critical factor of global warming.Therefore,reducing greenhouse gas emissions,decreasing global warming and developing low-carbon economy will be the goal of all countries in the future.China,the world's largest developing country with the most CO2 emissions,faces not only the pressure of reducing emissions from the industries,but also the problems of sustainable of economic and environmental development.Therefore,the 17^th,18^th,19^th National Congress of the Communist Party of China has proposed to promote the reform of the ecological civilization system successively,develop low-carbon economy,and build socialist modernization powerful country which is"rich,strong,democratic,civilized,harmonious and beautiful".At present,research on carbon emissions mainly based on energy statistics of the National and Local Statistical Bureaus in China.But the statistical yearbooks only publish data on energy consumption at the national and provincial levels,while cities and even smaller-scale county and township energy consumption data are not available,which makes it impossible to have studied on carbon emission in smaller scales.At the same time,there are still some problems in energy statistics data,such as differences among statistical departments in different areas or time lag.Therefore,it is urgent to use new resources to achieve carbon emission assessment on a smaller scale,which can analyze the spatial pattern of carbon emissions at different scales,quantify the driving force of carbon emission,and provide a scientific reference for China government to formulate a rational emission reduction policy and develop a green and low-carbon economy.Based on the statistical data from 1997 to 2012,this paper uses the nighttime light data to construct a provincial-fitting model to achieve carbon emission calculation in a smaller scale,which constitute for the limitations of statistical data in China,and to be important for reducing emissions and developing a low-carbon economy.This paper also analyse the basic characteristics of carbon emissions,including carbon emission,per capita carbon emissions and carbon emission intensity in provincial,city and county scales in 2000,2005 and 2010,respectively.Then,the spatial properties of carbon emissions studied from the change of barycenter,trend change and spatial agglomeration.Finally,based on the traditional STIRPAT model,spatial geolocation and time effects are introduced to construct an improved STIRPAT model to explore the driving forces of carbon emissions in different regions of China.The main research results are as follows:?1?The carbon emission accounting model fits well in this paper.Two factors,the total nighttime intensity?T?and normalized value?G?,are selected to fit the carbon consumption of energy consumption.Although the Pearson Correlation Coefficient of both is above 0.9,the correlation coefficient of T is slightly higher than that of G.Variance of T is less than the G variance,so the factor T chosen as the fitting index.Considering the spatically socio-economic differences,a province-fitting model is constructed to improve the accuracy of the national carbon emission spatialization estimation,and the provincial carbon emission data is applied for linear adjustment to correct the preliminary estimated carbon emission value of each grid,generating a national spatial distribution map of carbon emissions.?2?The total carbon emissions and per capita carbon emissions show an increasing trend,and the carbon emission intensity shows a downward trend.From the national scale,the total carbon emissions and per capita carbon emissions from 1997to 2012 increase from 3.02 Gt to 9.3 Gt,2.47 t/person to 6.93 t/person,and the energy consumption intensity decreases from 3.93 t/10⁴ yuan to 1.62t/10⁴ yuan;From the provincial scale,the number of provinces with carbon emissions less than 1.5×1.04million tons is 23,15 and 8 in 2000,2005 and 2012,respectively,the number of provinces with per capita carbon emissions less than 3t/person is 19,17 and 7,the number of provinces with the carbon emission intensity less than 3t/10⁴ yuan is 14,17and 22 respectively;From the municipal scale,the carbon emissions are 3.08t,4.28t and 7.84t,the per capita carbon emissions are 3.08t/person,4.28t/person,7.84t/person,the emission intensity is 5.26t/10⁴ yuan,3.8t/10⁴ yuan,3.38t/10⁴ yuan in 2000,2005 and 2010,respectively;From the county-level scale,carbon emissions are 119.83×10⁴t,174.98×10⁴t,309.81×10⁴t,per capita carbon emissions are3.4t/person,4.82t/person,8.3t/person,and carbon emission intensity is 5.3t/10⁴ yuan,4.29t/10⁴ yuan,3.58t/10⁴ yuan in 2000,2005,2010,respectively.?3?The barycenter of carbon emissions is moving westward.The barycenter of carbon emissions falls in Henan Province from 1997 to 2012,and its curve is rotated180°clockwise,that is,the carbon emission barycenter moved westward,indicating that the overall growth rate of carbon emissions in western China is greater than that of other parts of the country.?4?The trend of carbon emissions is slow.The trend of different scales of provinces,cities and counties in the past 16 years is mainly slow-changing,accounting for 76.67%,79.13%,and 79.48%,respectively,and the fast-changing types account for the least,respectively 3.3%,9%,and 7.19%and the fast-changing types are concentrated in most of regions of North China,Central China and some large cities in the Midwest China.?5?Carbon emissions appear significant spatial agglomeration.The global Moran's I of provincial,municipal and county carbon emissions is greater than 0,showing a spatial positive correlation,and the smaller the scale,the stronger the spatial agglomeration;the local autocorrelation of provincial carbon emissions is relatively stable,The HH province is dominant,mainly concentrated in the circum-BoHai-sea of China,and LL distributed in the western region.Carbon emission aggregation types of municipal and county-level in China are mainly LL types.?6?There are obvious regional differences in the driving forces of carbon emissions.This paper selects eight variables include the total population?P?,labor force population?FP?,the rate of urbanization?UR?,GDP per capita?PP?,total amount of foreign investment?FIA?,resident consumption level?RC?,industrial structure?IP?and energy consumption intensity?CP?to construct a spatio-temporal dynamic STIRPAT model.The results show that the regression coefficients of P,PP,IP and CP are positive,and their effects on carbon emissions are CP>PP>IP>P.The spatial distribution of P is consistent with Hu Huanyong's line.The greater the population density,the more carbon emissions in the region;The impact of PP on carbon emissions in the western and northeastern China is greater than that in the eastern region;The impact of CP on carbon emissions in the midwest regions is greater than that in the eastern regions;the regression coefficient of CP to carbon emissions is decreasing from the east to west.