Zichong Chen, Daniel J. Jacob, Haipeng Lin, Nicholas Balasus, Andy Hardy, James East, Yuzhong Zhang, Benjamin Runkle, Sarah Hancock, Charles Taylor, Xinming Du, Bjoern Ole Sander

Rice agriculture is a major source of atmospheric methane, but current emission inventories are highly uncertain, mostly due to poor rice-specific inundation data. Inversions of atmospheric methane observations can help to better quantify rice emissions but require high-resolution prior information on the location and timing of emissions. Here we use Landsat satellite data at 30 m resolution to identify flooded vegetation and combine this information with a 30 m global cropland database, rice-specific data, and a recent global dataset of emission factors (EFs) per unit of rice paddy area. The resulting Global Rice Paddy Inventory (GRPI) provides methane emission estimates at 0.1o× 0.1o (~10 km ×10 km) spatial resolution and monthly resolution. Evaluation of GRPI with independent rice area data and FLUXNET-CH4 eddy flux measurements shows good agreement. Our global emission of 39.3 ± 4.7 Tg a-1 for 2022 (best estimate and error standard deviation) is higher than previous inventories that use outdated rice maps and IPCC-recommended EFs now considered too low. GRPI shows the largest discrepancy from previous inventories in South Asia, where rice agriculture has rapidly developed but outdated rice maps fail to represent it. China is the largest rice emitter in GRPI (8.2 ± 1.0 Tg a-1), followed by India (6.5 ± 1.0 Tg a-1), Bangladesh (5.7 ± 1.2 Tg a-1), Vietnam (5.7 ± 1.0 Tg a-1), and Thailand (4.4 ± 0.9 Tg a-1). These five countries together account for 78% of global total rice emissions. The seasonality of emissions varies between countries depending on local climate and cultivation practices. We define a rice methane intensity (methane emission per unit of rice produced) to assess the potential of mitigating methane without compromising food security. We find national methane intensities ranging from 10 to 120 kg methane per ton of rice produced (global mean 51) for major rice-growing countries. Countries can achieve low intensities with high-yield cultivars, upland rice agriculture, water management, and organic matter management.