Last week, China announced that it will reduce carbon "intensity" by 40 to 45 per cent below 2005 levels by the year 2020. The emissions intensity (emissions/$GDP) approach taken by the Canada and the US in the past has been much maligned here as a dishonest dodge.
It is, however, a reasonably fair way to bring a reluctant developing nation like China into an international emissions control framework. The problem of course is that actual emissions target depends entirely on how much the Chinese economy grows by 2020. So 40-45% sounds impressive, but won't amount to an actual reduction in emissions.
The graph above shows a spectrum of possibilities. Unless the growth rate is less than ~4%/year - highly unlikely - Chinese carbon emissions will be higher in 2020 than 2005. If China keeps up the planned 8%/year growth, emissions in 2020 will be 74-90% higher than 2005 levels.
And, just like in the US and Canada cases, Chinese emissions intensity will naturally decrease over time without any policy intervention. It decreased 10% from 2000 to 2005, and well over 40% from 1990 to 2005.