You conveniently used a month with an unusually sharp drop as your end of decade, which make me think I'm feeding a troll, but nevertheless:

The correct decade is May 2002 to April or May 2012, which gives you 272 gigatons per year as the average. There is pretty obvious seasonality in the data, so you can't just jump between seasons.

You could also use September 2002 to September 2012, if you insist on including the outlier month, which would work out out to 309 gigatons/year.