Regional characteristics of CO2 emissions from China's power generation: affinity propagation and refined Laspeyres decomposition

The Chinese power sector plays a vital role in national emissions reduction goal due to its large contribution to overall emissions. This paper explores the regional similarities and disparities of power generation emissions among 30 provinces. First, the affinity propagation (AP) algorithm is applied to find the similar characteristics in emissions among 30 provinces. The clustering results of CO2 emissions indicate that the 30 provinces are divided into five clusters in 1997 and seven clusters in 2012 based on four indicators (generation structure, energy intensity, GDP per capita and electricity intensity). Then, the regional differences in driving forces on CO2 emissions from power industry are examined using refined Laspeyres decomposition model. Results showed that there are significant contribution differences of five indicators (power generation emission coefficient, generation structure, electricity intensity, economy and population) on power generation emissions among different provinces. The provincial emissions reduction target and supporting policies for power industry should be customised and consistent with the actual situations considering the similarity and differences in emission characteristics.