到2020年和2030年,中国的碳减排目标是否会有所改变？

Will China make a difference in its carbon intensity reduction targets by 2020and2030?Lei Xu a ,Nengcheng Chen a ,b ,⇑,Zeqiang Chen aa State Key Laboratory of Information Engineering in Surveying,Mapping,and Remote Sensing,Wuhan University,Wuhan 430079,China bCollaborative Innovation Center of Geospatial Technology,Wuhan 430079,Chinah i g h l i g h t sChina could achieve its 2020and 2030carbon intensity reduction targets under current policies. The energy efficiency in the tertiary industry remained unimproved in the last decade. CO 2emissions will be 1.64times that of 2005in 2020and 1.69times in 2030. Carbon emissions fail to meet the 450ppm scenario in 2020and 2030.a r t i c l e i n f o Article history:Received 5January 2017Received in revised form 24June 2017Accepted 28June 2017Keywords:Carbon emission STIRPAT IntensityEnergy consumption Chinaa b s t r a c tThe Chinese government has made ambitious commitments in terms of its carbon intensity reduction targets for 2020and 2030.Whether China will achieve these targets remains uncertain,especially under the context of increasing consistency in carbon emissions cut globally.This study decomposed total energy consumption into five types and modeled each of them with its influential factors based on the stochastic impacts by regression on population,affluence and technology (STIRPAT)model.Carbon emis-sions were predicted by combining economic growth forecasting,industrial structure and energy struc-ture projections.The results show that the estimated CO 2emissions in 2020were 10.05gigatonnes (Gt),with a 52.8%reduced intensity compared to 2005.And the predicted CO 2emissions in 2030were 10.39Gt,with a 70.0%reduced intensity.China’s carbon intensity reduction targets in 2020(40–45%)and 2030(60–65%)can be met under current policies.However,the total CO 2emissions fail to meet the 450ppm scenario (8.4Gt in 2020and 7.1Gt in 2030)only by the improvement of industrial structure and energy structure.New policies such as carbon trading market (CTM)and carbon capture,utilization and storage (CCUS)technology need to be developed in depth to further mitigate CO 2emissions.Ó2017Elsevier Ltd.All rights reserved.1.IntroductionThe IPCC (Intergovernmental Panel on Climate Change)fifth assessment report indicated that human activities are extremely likely to be the main reason causing global warming and climate change [1].The report confirmed the necessity of limiting the tem-perature rise to 2°C in the 21st century.Of all the factors leading to global warming,greenhouse gas (GHG)emissions contribute the most.Since 2005,China has emitted more carbon than any other nation.In 2009,the Chinese government made a commitment in the United Nations Framework Convention on Climate Change (UNFCCC)conference in Copenhagen to reduce its CO 2intensity by 40–45%in 2020compared to that in 2005.And in 2015,the Chi-nese government made a new declaration in the UNFCCC confer-ence in Paris to reduce the carbon intensity by 60–65%in 2030compared to 2005.To achieve these goals,a series of measures have been taken to decrease the consumption of fossil fuels and develop cleaner energy such as hydropower,wind energy,nuclear power [2].Whether China will fulfill these carbon intensity reduc-tion targets remains a question worth studying.Economic growth and energy consumption are two main factors influencing carbon emissions.The GDP of China has been increas-ing in the past forty years (Fig.1),whereas the annual growth rate of GDP has been fluctuating and has exhibited a declining trend in recent years.The annual economic growth rates for 2013,2014and 2015were 7.7%,7.3%and 6.9%,respectively.China is ranked as the second largest economy behind the USA since 2010.However,after/10.1016/j.apenergy.2017.06.0870306-2619/Ó2017Elsevier Ltd.All rights reserved.⇑Corresponding author at:State Key Laboratory of Information Engineering in Surveying,Mapping,and Remote Sensing,Wuhan University,Wuhan 430079,China.E-mail address:cnc@ (N.Chen).Applied Energy 203(2017)874–882Contents lists available at ScienceDirectApplied Energyj o ur na l h o me pa ge :w w w.e ls e v ie r.c o m/lo c a t e/ap en e rgya long time of rapid economic growth,unbalanced social and eco-nomic development existed among areas.The central government of China has adjusted the economic policy from ‘‘fast and sound”to ‘‘sound and fast”to ensure high-quality economic growth.Based on ‘‘The 13th five-year plan ”,China’s economy is planned to keep an annual growth rate larger than 6.5%.The energy consumption of China showed a substantial increase during 1978–2015.How-ever,China slowed its energy demand growth rate in recent years,largely resulting from the low economic growth rate,improvement of energy efficiency and upgrade of industrial structure [3].In 2014,the State Council of China announced the ‘‘Energy development strategic action plan 2014–2020’’,defining the total tar-get,policy,and mission of energy strategy for the next stage.China will make efforts to improve the proportion of non-fossil fuels in the total energy consumption,expecting to reach approximately 15%by 2020.The shares of coal and natural gas consumption will be below 62%and approximately 10%,respectively.In addition,the use of cap-and-trade emission trading schemes (ETSs)is another way conducted by the Chinese government to mitigate carbon emissions [4].In 2011,China proposed establishing a domestic car-bon trading scheme in order to effectively cope with climate change not only by relying on administrative measures,but also more through market-based mechanisms [5].Seven pilot cap-and-trade ETSs across the country were successfully implemented between 2013and 2014.The national carbon trading scheme,which will be launched in 2017,will probably make a difference in China’s CO 2reduction.The prediction of energy consumption and carbon emission is helpful for decision makers to make appropriate environmental and economic strategies.There are numerous methods to model and calculate carbon emissions according to the literature.Xiao et al.[6]divided these models into three categories:top-down,bottom-up and hybrid models.The top-down models,with partic-ular attention on the connection between energy production and consumption in the departments of the national economy from the viewpoint of an economic model,are mainly used to analyze macro economy and energy policy [7].The key technology param-eters related to energy production and consumption are regarded as exogenous variables.The bottom-up models consider the eco-nomic impact of technological alternatives in a partial equilibrium framework [8].As the top-down and bottom-up models have their own advantages and weaknesses,hybrid models were developed to link and narrow the gap between the two models [9].However,many uncertainties exist in the economic and technological parameters of these models.Some other models,such as kaya iden-tity [10],the Log-Mean Divisia Index (LMDI)[11]and STIRPAT [12],focus on influencing factor decomposition and are also widely used for energy and carbon emission analysis.However,it is difficult to apply these decomposition methods to prediction as many of the influencing factors are unpredictable.In addition,there are also some general forecasting techniques such as the autoregressive integrated moving average (ARIMA)[13],grey model (GM)[14]and artificial neural networks (ANNs)[15]that can be used for pre-diction.For these general models,the data size and quality are important factors influencing the forecasting results.The grey sys-tem is used to solve problems with uncertainty,a small data size and a lack of information.High accuracy can be achieved for short time predictions [16].As our data covers a short period of time,therefore,this study employed the grey model to forecast eco-nomic growth for the 13th five-year period of China.Many of the existing literatures employed the econometric the-ory to model the relationships among carbon emission,energy consumption and economic growth [17–21].Ozturk and Acaravci [22]studied the relationships between economic growth,carbon emissions,energy consumption,foreign trade,and employment in Cyprus and Malta based on the autoregressive distributed lag (ARDL)bounds testing and causality models.Alshehry and Belloumi [23]used cointegration theory to investigate the relation-ship between energy consumption,CO 2emissions,and economic growth.Hassan [24]studied the relationship between economic growth,CO 2emissions,and energy consumption in five ASEAN (Association of South East Asian Nations)countries by the panel smooth transition regression model.In spite of these studies,industrial structure and energy consumption structure that can significantly influence carbon emissions were not well considered.Although the relationship among carbon emission,energy con-sumption,and economic development can be modeled by the econometric method,the underlying factors influencing these vari-ables such as energy consumption structure,economic growth types,population,and technology were not well studied.A number of researches have focused on China’s commitment in its carbon intensity target.Yuan et al.[25]suggested that the 45%reduction in China’s carbon intensity target coincided with its socioeconomic development planning without further effortsinFig.1.The energy consumption and real GDP of China between 1978and 2015.L.Xu et al./Applied Energy 203(2017)874–882875CO2mitigation measures.Liu et al.[26]studied the carbon inten-sity goal in China from a thermal power perspective,and the results showed that by2020the CO2intensity will be2.1tons/104 Yuan,approximately twice the2005level.Li and Lin[27]used a nonlinear threshold cointegration method to measure the connec-tion between carbon intensity and economic growth and found that China’s carbon intensity in2020would reduce by approxi-mately40.60%when the GDP growth rate is between7%and 8.4%.Cansino[28]et al.concluded that the carbon intensity in China in2020would be reduced by50%compared to that of 2005under current policy without additional climate change mea-sures.And they indicated that the total CO2emissions in2020will be seven times the volume of2005.This may not be true because China is continuously adjusting its industrial structure to improve the share of the tertiary industry.The total CO2may not reach up to seven times.In this study,it’s less than2times.Few researches have paid much attention to the projections of China’s CO2emissions for2030.Michel et al.[29]studied China’s CO2emission in2030based on scenario analysis method and con-cluded that carbon emissions can peak before2030,with a pro-jected CO2emission ranging from11.3to11.8Gt.Luukkanen et al.[30]adopted a general Long-range Integrated Development Analysis(LINDA)model to examine economic structure on carbon emissions.The conclusion indicated that China’s emission and energy intensity targets were ambitious.Some literatures analyzed carbon emissions in specific sectors by2030,such as transporta-tion[31,32],textile industry[33]and building sectors[34].These studies provide us with a general understanding of carbon emis-sions in sectoral or national level.The existing literature,focused mainly on the relationship between carbon emissions,economic growth,and energy con-sumption[35–37].The econometric methods such as cointegration theory,panel model,and causality test were mostly used.For the cointegration method,CO2emissions,GDP,and total energy con-sumption were incorporated into one equation together.The coin-tegration method is intrinsically linear.However,carbon emissions and economic growth showed a weak decoupling relationship in recent years in China[38–40].Basically,energy consumption and its structure are two direct factors contributing to energy-related carbon emissions.Energy consumption and economic growth are related to each other in the context of rapid economic growth of China.In addition,different industries contribute to the total energy consumption differently.The heavy industry consumes more energy than agriculture and service industry.In this study, we took these differences into consideration.We decomposed total energy consumption intofive types:primary industry,heavy industry,construction,tertiary industry and household energy consumptions.Each type of energy consumption was modeled with the corresponding economic growth types or population based on STIRPAT method.Once the relationships between energy consumption and economic growth were determined,they would be combined with energy consumption structure to forecast car-bon emissions.Although some of the existing literatures conclude that China would meet its carbon intensity target even without additional carbon mitigation measures[26,41],it is uncertain how much the carbon intensity will be reduced as China is actively involved in its energy consumption structure improvement in recent years. This study investigates whether or not China will meet the carbon intensity reduction targets and450ppm(ppm)scenarios[42]in 2020and2030.Specifically,the following procedures were con-ducted:(1)exploration of the relationships betweenfive types of energy consumption and economic growth,industrial structure, population and technology,(2)projections of China’s carbon emis-sions in2020and2030to understand the carbon reduction poten-tial and limits under the context of adjustments in industrial structure and energy structure in China,and(3)discussions on the intensity targets and how to achieve the450ppm scenario. The corresponding results were discussed.In summary,the contribution of this paper is twofold.First,we used a series of decomposed STIRPAT models to examine the rela-tionships between different types of energy consumption and their determinants.Although the STIRPAT method was used in the exist-ing literature to model the carbon emissions[43,44],all the influ-encing factors were mixed together.In other words,different kinds of economic growth and different kinds of energy consumption result in different carbon emissions.However,they were not well considered in the published papers.We did a redecomposition of the energy consumption intofive types and modeled each of them separately.Second,we did a comparatively comprehensive analy-sis on the energy consumption,energy-related carbon emissions and intensities in China in2020and2030.The gaps between pro-jected emissions and the450scenario,potential solutions to meet it,and the latest technology and measures adopted by the Chinese government were investigated.This study can be regarded as a basic outlook of China’s carbon emissions from a macro perspective.The remainder of this study was organized as follows.Section2 described the relevant methods and data.Section3demonstrated the forecasting results of carbon emissions and performed discus-sions.Section4concluded this study.2.Methods and data2.1.GM(1,1)modelThe grey theory wasfirst proposed by Deng in1989[45].The GM(1,1)model is a basic model in the grey theory.Thefirst‘‘1”means thefirst order grey differential equation,and the second ‘‘1”indicates that there is one variable.The main idea of this model is to apply an accumulating generation operator(AGO)to the raw data to construct a new series with the exponential rule.Afterfit-ting the new data series with exponential models,the inverse accumulating generation operator(IAGO)is performed to obtain the predicted value of the original series.Some studies have developed dynamic grey models to improve the model accuracy of the traditional grey model[46].Namely,the grey model is applied using only recent data rather than the entire data set.For example,x(0)(k+1)is predicted using the data set, (x(0)(1),x(0)(2),...,x(0)(k))where k<n.In the next prediction pro-cess,x(0)(1)is removed from the group and x(0)(k+1)is added to the group.Thus,x(0)(k+2)is predicted using(x(0)(2),...,x(0)(k), x(0)(k+1)).The prediction process is continued until x(0)(n)is pre-dicted.As the grey model is appropriate for short-term predictions [16],this study employed the dynamic GM(1,1)model to forecast economic growth in the nextfive years.2.2.Energy consumption decompositionThe IPAT equation is a famous formula to model the impact of human activities on the environment[47,48].It wasfirst proposed by Ehrlich-Holdren-Commoner in a course debate in1970.This equation consists of three independent variables:population,afflu-ence,and technology.The basic formula is as follows:I¼PATð1Þwhere I denotes environmental impact;P represents population;A is affluence and T is technology.Energy consumption can be regarded as a kind of environmental impact because it consumes natural resources.Affluence is commonly represented by GDP per capita.The multiplication of population and GDP per capita pro-duces GDP.Technology refers to how much environmental impact876L.Xu et al./Applied Energy203(2017)874–882is generated in the process of affluence creation,e.g.CO 2emissions per unit of GDP.The IPAT equation assumes that environmental impact changes proportionally or monotonically.This may not be always the case as the environmental Kuznets curve (EKC)theory [49]points out that there is a non-monotonic relationship between environmental impact and economic development.Rosa and Dietz [50]developed a model named stochastic impacts by regression on population,affluence,and technology (STIRPAT)to empirically examine the driving forces of environmental impact.The specific model can be described as:I ¼aP b A c T d eð2Þwhere I ,P ,A and T have the same meanings as that in the IPAT equa-tion;a ,b ,c and d are parameters representing stochastic impacts;e denotes error term.After natural logarithm transformation,the STIRPAT model can be expressed as in general linear form:log I ¼a þb log P þc log A þd log T þe ð3Þwhere a and e are the logarithm forms of that in Eq.(2),respectively.In this study,we decomposed total energy consumption into five types 1:primary industry,heavy industry,construction,tertiary industry and household energy consumptions.These energy types are related to economic growth,population,and technology.Fig.2showed different types of economic growth or population with their corresponding energy consumptions during 2000–2015.The value-added of primary industry (G1)and its energy consumption (E1)don’t have an obvious linear relationship.But with the increase of G1,the relationship is nearly linear.The value-added of heavy indus-try (G2)seems to have a log-linear relationship with its energy con-sumption (E2).The value-added of construction (G3),value-added of tertiary industry (G4)and population (P)have a nearly linear rela-tionship with their corresponding energy consumptions (E3–E5).Thus,we used linear regression to model the relationship between G1,G3,G4,P and their energy consumptions.The technology effect was not considered in these relationships because the estimated coefficients were not statistically significant or didn’t coincidewithFig.2.Different types of economic growth or population with their corresponding energy consumptions.G1:value-added of primary industry;G2:value-added of heavy industry;G3:value-added of construction;G4:value-added of tertiary industry;P:population;E1:energy consumption of primary industry;E2:energy consumption of heavy industry;E3:energy consumption of construction;E4:energy consumption of tertiary industry;E5:household energy consumption.1The divisions were based on the National Industry Classification (GB/T 4754–2011)of China.The primary industry refers to agriculture,forestry,animal husbandry and fishery industries.The secondary industry includes heavy industry and construction.The tertiary industry refers to service industry.L.Xu et al./Applied Energy 203(2017)874–882877their physical meanings when incorporating technology term in our experiments.This can be explained by the fact that these industries are not so energy-intensive.The technical term was represented by time trend in this study.We assumed that technology progresses with time.For the value-added of heavy industry,the technology effect needs to be considered as the heavy industry is energy-intensive.Referring to STIRPAT formula,the specific models between five types of energy consumption and their influencing factors were shown as follows:E 1¼a 1þb 1G 1log E 2¼a 2þb 2log G 2þc 2log T 2E 3¼a 3þb 3G 3E 4¼a 4þb 4G 4E 5¼a 5þb 5P8>>>>>><>>>>>>:ð4Þwhere E i (i =1,2,3,4,5)represents a specific type of energy consump-tion;G i (i =1,2,3,4)denotes a certain type of industry;T is time,indi-cating technology;a,b and c are estimation parameters.2.3.Energy consumption structureThe historical energy consumption structures from 2000to 2015were shown in Fig.3.The shares of coal in the total energy consumption (TEC)decreased in recent years.According to national energy development strategy in China,the share of coal in the TEC will be further reduced to cut GHG emissions in the upcoming years.Here,the shares of coal till 2020were computed based on the average change rate in the last decade.Natural gas will be paid special attention because it generates less GHGs and pollutants than coal.The Chinese government has decided to improve the share of natural gas in the national energy plan.In the past five years,the share of natural gas increased from 4.6%to 5.9%,with an annual increase of 0.3%.We assumed that natural gas would evolve in this way in the next five years.Non-fossil energy resources will be further rge investments will be put into the development of renewable and nuclear energy resources according to the national mid-long term scientific andtechnological plans.The share of non-fossil energy increased from 8.4%to 12%in the last five years.A 0.9%increase in non-fossil energy share would be expectable.The projected energy structure in 2020is 59.33%,16.64%,7.53%and 16.50%for coal,oil,natural gas and non-fossil energy,respectively.2.4.CO 2emission calculationThis study employed the IPCC recommended method to calcu-late CO 2emissions [51].The formula is as follows:CE ¼X 4i ¼1EC i ÂEF i Â4412ð5Þwhere EC i (i =1,2,3,4)indicates the standard energy consumption of coal,oil,natural gas and non-fossil fuel,respectively.EC 1,EC 2,EC 3and EC 4refer to the carbon emission factors.Following Zhu et al.[41],the carbon emission factors for coal,oil,natural gas and non-fossil fuel are 0.7304,0.5630,0.4190and 0,with units of t (C)/t .All the data were collected from the China statistical yearbook and the China energy statistical yearbook.All the monetary series are deflated to 2005prices.The CO 2emissions and intensity were calculated using the predicted energy consumption as well as energy structure scenarios.3.Results and discussion 3.1.Energy consumption predictionBased on different STIRPAT models,the relationships between five types of energy consumptions and their influencing factors were shown in Table 1.All these models and their coefficients are statistically significant at 1%significance level,indicating a strong correlation between energy consumption and its influential factors.For heavy industry,the coefficient of time term is negative,suggesting a suppressing effect on energy consumption.With the advances of technology progress or the improvement of energy efficiency,less energy could produce the same amount products as before.Population has a positive effect on householdenergyFig.3.Energy consumption structures of China during 2000–2015.878L.Xu et al./Applied Energy 203(2017)874–882consumption.The time term was excluded from Model4(Table1) due to insignificance,indicating that technology plays no role in the tertiary industry.The economic growth in the service industry is only linearly correlated with its energy consumption,suggesting unimproved energy efficiency.The energy consumption in2020can be predicted based on the five established STIRPAT models.Economic growth was forecasted using dynamic GM(1,1)model.The industrial structure was projected to2020according to its average change rate in the last decade.China sustained a0.5%steady increase of population annu-ally since2006.Thus,we assumed that population growth rate remains0.5%in the upcoming years.The predicted energy con-sumption is4.90billion tons of standard coal equivalent(tce)in 2020,with an annual increase of2.65%.3.2.Carbon emissions and intensity in2020The CO2emissions and intensities from2016to2020were obtained based on the predicted energy consumption and energy structure.The results were shown in Fig.4.The energy-related CO2emission in2020will be10.05gigatonnes(Gt),1.64times that of2005(6.13Gt).The CO2emissions showed a rapid increase from 2002to2005mainly as a result of the rapid economic growth and tremendous energy consumption.In recent years,the total CO2 emission slowed down and even decreased.A continuous decline in the carbon intensity is expected.China is actively improving its industrial structure and energy consumption structure,and has implemented some new ways to mitigate its carbon emissions. One of the new ways is carbon capture,utilization and storage (CCUS).CCUS is a technique that captures CO2from the production process and reuses or stores it,keeping it away from the US has three types:chemical utilization,biological uti-lization,and geological utilization.Carbon geological utilization and storage(CGUS)uses underground minerals and conditions to mineralize CO2and promote generating new products such as oil, natural gas,and coalbed methane.China has done some basic researches and technology demonstrations in CGUS in the last dec-ades[24].Besides,China has cooperated with other countries on CCUS science and technology,such as the Carbon Sequestration Leadership Forum(CSLF),the U.S.-China Clean Energy Research Center,the Near Zero Emissions Coal(NZEC),and the China-Australia Geological Storage of CO2(CAGS),etc.[52].Although the CCUS technology faces difficulties in large-scale commercial application due to complex technology,high cost and low demand, it is a promising way to reduce carbon emissions[53].Another method to mitigate carbon emissions is carbon emission trading system(ETS).ETS relies on the market to reduce CO2emissions and has been adopted worldwide[54,55].The Chinese government started seven ETS pilots during2013–2015[56].The national ETS will probably be set up in July2017,which will play an important role in emissions cuts in the upcoming years.The predicted CO2intensity in2020was1.55tons per ten thou-sand Yuan,with a52.8%reduced intensity compared to2005(3.30 tons per104Yuan).According to this,China’s carbon intensity reduction target in2020will be fulfilled,and approximately an additional8%reduction in carbon intensity above commitment is attainable.The reason for the optimistic result mainly lies in two aspects.First,rapid economic growth produces hugeeconomicFig.4.The total CO2emissions and intensity in China from2000to2020.Table1Regression results offive types of energy consumptions and their influencing factors.Standard errors are in parentheses.Variable Model1Model2Model3Model4Model5GDP0.137***(0.019) 1.435***(0.166)0.213***(0.006)0.280***(0.012)–Population–––– 3.227***(0.082)Time–À105.395***(30.004)–––Constant3012.704***(554.346)797.277***(226.284)1010.341***(120.233)10845.694***(1561.705)À394520.459***(10853.050) R20.7780.9880.9880.9750.991F-statistic49.024518.9191106.4552.7931548.758p value0.0000.0000.0000.0000.000Note:Significance level:1%(***).L.Xu et al./Applied Energy203(2017)874–882879。