Changes in China's carbon footprint and driving factors based on newly constructed time series input-output tables from 2009 to 2016.

China is the country with the most carbon emissions and the largest foreign trade volume worldwide. China's carbon footprint, especially the carbon footprint of exports, must be studied to clarify China's responsibility for carbon emissions, avoid "carbon leakage" and develop more reasonable emission reduction policies. By comparing 6 update models, this paper adopts the GRAS, which is a typical entropy optimization method for matrix updating, to construct new time series input-output (IO) tables for China from 2009 to 2016 to analyze the carbon footprint of China's domestic final demand and exports. Then, the changes in the footprints were investigated with a structural decomposition analysis (SDA) to determine the driving factors. The carbon footprint of exports showed an upward trend from 2009 to 2012 and a downward trend from 2012 to 2016. The export share of the total consumption-based carbon footprint also increased from 2009 (17.64%) to 2012 (21.47%) and then decreased to 16.40% in 2016, reflecting a reduction in the CO2 emissions transferred to China after 2013. According to the SDA results, the emission intensity effect (664.20 Mt CO2) and primary input effect (555.21 Mt CO2) played a key role in reducing the carbon footprint, and the total export effect (1083.12 Mt CO2) contributed the most to the increase in the carbon footprint of exports. Based on further SDA, the effects of the primary input, Leontief structure and export structure dramatically varied based on China's industrial structure adjustment and technological change during the subperiod. The analysis framework and data in this paper can be applied to further study China's energy economy and environmental issues.