本文运用“LMDI模型”, 对我国1995～2011年的能源消费碳排放进行了时空序列的因素分解研究。在国家层面上, 对8个子时间区间以及29个省份的碳排放影响因素进行了分解并对各因素的空间差异性影响进行了对比研究, 共分解成4类效应的11种影响因子；在产业层面上, 以占我国碳排放量最大比重的工业为典型研究部门, 进一步对我国工业部门的碳排放影响因素进行时空序列的解析, 并进一步分析比较了各因素对6大行业的累积效应, 共分解成5种影响因子。在上述因素解析的基础上, 本文选取对我国能源消费碳排放具有显著影响的因子作为情景分析指标并设定了4类碳排放情景, 对中国2020年前近、远期的人口、经济发展、能源消费及碳排放等变化趋势进行了预测与分析, 并提出了我国低碳转型的路径选择。究结果表明：（1）国家层面上, 各因素对我国能源消费碳排放的作用方向在时空序列上显示出良好的一致性, 最大驱动因素均为人均GDP（PCG）, 最大抑制因素为生产部门能源强度（EIP）, 其他因子也表现出一致的正、负驱动作用。但是, 各因素在不同省份之间的贡献值差异较大, 总体上正向驱动作用大于负向抑制作用, 各地区的碳排放均出现不同幅度的增长。（2）在产业层面上, 工业在6大行业碳排放总量中所占比重始终处于绝对的主导地位。PCG是中国工业部门在时空序列上碳排放增长的最主要正向驱动因素，人口（P）、经济机构（YS）、能源结构（ES）也表现出一定的正向效应, 但贡献值不大。能源强度（EI）是中国工业碳排放增长的唯一负向抑制因素, 极大地减少了工业部门在全国和地区水平上的能耗碳排放。（3）各影响因素对6大行业碳排放具有较大的差异性影响, PCG对所有行业均表现为正向驱动效应且累积贡献值最大, EI均表现为负向抑制效应, 其他因素则因不同行业表现出不同的效应和贡献值。（4）低碳（LC）情景是未来我国实现低碳转型最合适也最具可行性的方案, 能最大程度的促进经济、社会、能源与环境系统的协调发展。在LC情景下, 我国未来可以通过技术创新路径、产业结构优化路径、能源结构优化路径和政策引导路径进一步推进低碳转型进程。

This paper presents a decomposition analysis of energy-related carbon emissions in China from 1995-2011 using the Logarithmic Mean Disivia Index (LMDI) approach. At a national level, a decomposition analysis on factors affecting CO2 emissions was carried out in eight time sub-periods and in 29 provinces, and comparative research was conducted on the spatial differences of each factor. Eleven factors from four types of effects were decomposed. At the industrial level, taking the sector which caused the most of China’s CO2 emissions as the representative research department, time and spatial series decomposition was carried out on factors affecting CO2 emissions from industrial departments. Further comparative analysis was carried out on the cumulative effect of each factor on six sectors, which decomposed in to five influencing factors. Based on the decomposed factors mentioned above, factors which highly influence the CO2 emissions of China were selected as scenario analysis indexes and four types of emission scenarios were established. Predictions and analyses are presented for changes and trends in short-term and long-term population numbers, economic development, energy consumption and CO2 emissions in China up to 2020. Finally, path selections for a low carbon transition in China are proposed based on the scenario analysis. The results show that: (1) At the national level, the influences of the factors in time and spatial series were highly consistent, with PCG being the dominant positive driving factor and EIP being the dominant negative restricting factor. Other factors also showed consistent positive and negative effects. However, the contributing value of the factors in different provinces varied significantly, where in general the positive driving effects outweighed the negative inhibiting effects, and CO2 emissions increased by different degrees in different provinces. (2) At the industry level, PCG was also the largest positive driving factor for industrial CO2 emissions growth in both time and spatial series. P, YS and ES also played a positive driving role, albeit a relatively weak one. As the only negative inhibiting factor, EI significantly reduced the energy-related CO2 emissions from industrial sectors at national and provincial level. (3) Different influencing factors had differential impacts on CO2 emissions in the six sectors. PCG was always a positive driving factor in all sectors, and had the largest cumulative contribution value, while EI was always a negative inhibiting factor. Other factors had different effects and contribution values depending on the sector. (4) The LC scenario was found to be the most appropriate and feasible scheme for China’s low-carbon transition by enhancing to the greatest extent the harmonious development of the economy, society, energy and environmental system. The low-carbon transition process in this scenario can be further promoted via technical innovation, optimization of industrial and energy structures, and policy guidance.