China is a major source of anthropogenic methane (CH4) emissions because it is the world's largest producer of rice grain, nearly all of which is grown in irrigated paddies. This study sought to reduce the uncertainty in estimates of CH4 emissions from rice cultivation in China by improving the quantification of the effects of management practices (intermittent drainage and fertilizer inputs) on emissions. These results were spatially extrapolated with digital maps of type of rice using new estimates of organic matter and fertilizer inputs, as well as estimates of soil drainage. The estimated total annual CH4 emissions from rice agriculture in China in 1990 were 9.9 +/- 3.0 x 10(12) g. If intermittent drainage practices were adopted on 33% of the poorly drained soils used for rice cultivation in southern China, the estimated emissions would be 8.9 +/- 2.7 x 10(12) g CH4 yr(-1). Reducing projected organic matter inputs by 50% as a sensitivity analysis to reflect the trend for reduced use of organic fertilizer, resulted in emissions of 9.6 +/- 2.9 x 10(12) g CH4 yr(-1), with 8.7 +/- 6 x 10(12) yr(-1) emitted with 33% adoption of intermittent drainage on poorly drained paddies. Although intermittent drainage has been shown to reduce emissions by 50%, the area of rice that is relatively easy to drain and re-flood is not very large. The use of intermittent drainage with better drained paddies is limited because of problems with re-flooding and it is also limited with very poorly drained paddies that are difficult to drain. The 10% emission reduction predicted with 33% adoption of intermittent drainage practices, while not large, is conservative and may be possible to realize. These CH4 emissions results are relative estimates because the uncertainty remains large due to a lack of emissions measurements from paddies in more regions and a lack of detailed information about organic fertilizer application rates.
