我国碳排放省域差异及影响因素研究

发布时间：2018-07-01 10:58

本文选题：碳排放 + 省域差异　；参考：《江西财经大学》2017年硕士论文

【摘要】：随着经济的快速发展,全球的气候问题也越来越突出,吸引了全国人民的广泛关注,其中突出的是二氧化碳等温室气体,这也是造成地球温度普遍升高的原因之一。因此实现碳的减排是应对气候变化的重中之重。首先,对1995-2015年我国30个省区碳排放量进行核算,构建面板数据库,并将30个省区按碳排放、人均碳排放、碳排放强度分别聚类,研究了各类别所属省区的碳排放差异。研究发现碳排放量、人均碳排放总体上升,碳排放强度总体下降,但不同省域在不同时期有较小的波动。通过对碳排放、能源消耗和GDP三个变量建立多变量聚类模型,得到了3类不同省域。从该3类不同省域来看,第3类省域中山东、辽宁、广东等省区碳排放量及人均碳排放量远远高于其他地区,而碳排放强度却最低,第1类省域中内蒙古、新疆等省区碳排放强度最高。其次,研究了3类省域碳排放、人均碳排放、碳排放强度的异同。在此基础上,利用扩展的STIRPAT模型对我国总体影响因素进行静态面板模型和动态GMM模型估计。在静态面板模型当中,除技术水平外其余因素对碳排放均有显著的正向影响,而技术水平对碳减排有显著的正向影响。经济弹性系数最大,即人均GDP变动对碳排放影响最明显。在动态面板模型当中,上一期碳排放量对当期具有显著的路径依赖特征。人口规模、经济水平对碳排放也显著为正,但其弹性系数有所下降。另外,经济水平依然是几个弹性中最大的,且只有人口、经济水平上一期的值对当前碳排放量有明显的负向抑制作用。再次,对3类不同省域的碳排放影响因素作分解。在静态面板模型中,3类省域回归结果与总体一致,其中第1类省域模型中的人口弹性最大,第2类和第3类省域模型中的经济弹性最大。在动态面板模型中,滞后一期的碳排放回归系数按从大到小排列的省域依次为第1类(1.129)、第3类(1.101)和第2类(1.087)省域。最后,根据实证结果对全国和3类省域提出相应的碳减排建议。其中,对于全国而言,应控制人口过度增长,倡导低碳生活,提高能源利用率,广泛开展节能工作,并且出台能源消费的有关法律政策,保证政策得以有力实施。对各类省域而言,第1类省域应控制人口增长、加快城镇化进程,第2类省域应促进经济增长、加快城镇化进程和提高技术水平,第3类省域应持续控制人口增长、提高技术水平。
[Abstract]:With the rapid development of economy, the global climate problem is becoming more and more prominent, attracting the widespread attention of the whole country, especially carbon dioxide and other greenhouse gases, which is one of the causes of the global warming. Therefore, carbon emission reduction is the top priority in tackling climate change. Firstly, the carbon emissions of 30 provinces and regions in China from 1995 to 2015 are calculated, and the panel database is constructed, and the carbon emissions of 30 provinces and autonomous regions are clustered according to carbon emissions, per capita carbon emissions and carbon emission intensities, respectively, and the differences of carbon emissions between the provinces and autonomous regions of each category are studied. It is found that the carbon emissions per capita increase and the intensity of carbon emissions decrease, but there are small fluctuations in different provinces in different periods. Through the establishment of multi-variable clustering model for three variables, carbon emission, energy consumption and GDP, three kinds of different provincial regions are obtained. From the perspective of these three different provinces, the carbon emissions and per capita carbon emissions of Shandong, Liaoning, Guangdong and other provinces in the third category are far higher than those in other regions, but the intensity of carbon emissions is the lowest. Xinjiang and other provinces and regions the highest intensity of carbon emissions. Secondly, the similarities and differences of three kinds of provincial carbon emission, per capita carbon emission and carbon emission intensity are studied. On this basis, the static panel model and dynamic GMM model are used to estimate the total influencing factors in China by using the extended STIRPAT model. In the static panel model, except for the technical level, the other factors have significant positive effects on carbon emissions, while the technical level has a significant positive impact on carbon emission reduction. The coefficient of economic elasticity is the largest, that is, the change of GDP per capita has the most obvious effect on carbon emissions. In the dynamic panel model, the last period of carbon emissions has a significant path-dependent characteristics of the current period. Population size and economic level were also significantly positive for carbon emissions, but its elastic coefficient decreased. In addition, the economic level is still the largest of several elasticity, and only the population, the economic level of the last period of the value of the current carbon emissions have a significant negative inhibition. Thirdly, the factors affecting carbon emission in three different provinces are decomposed. In the static panel model, the regression results of three types of provincial domain are consistent with the whole. The population elasticity of the first type of provincial model is the largest, and the economic elasticity of the second and the third type of provincial model is the largest. In the dynamic panel model, the regression coefficients of carbon emissions in the lag period are the first (1.129), the third (1.101) and the second (1.087) provincial domains according to the provincial range from large to small. Finally, according to the empirical results, the corresponding carbon emission reduction recommendations for the whole country and three kinds of provinces are put forward. Among them, for the whole country, we should control the excessive growth of population, advocate low carbon living, improve energy efficiency, widely carry out energy conservation work, and introduce the relevant laws and policies on energy consumption to ensure that the policies can be carried out effectively. For all kinds of provinces, category 1 should control population growth and speed up the process of urbanization, category 2 should promote economic growth, speed up the process of urbanization and raise the level of technology, and category 3 should continue to control population growth. Raise the technical level.
【学位授予单位】：江西财经大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：X321

【参考文献】