Greenhouse gas emission and its sensitivity parameters under controlled irrigation in cold-region paddy fields

Controlled irrigation (CI) is considered a key management practice for water-saving in cold-region paddy fields; however, its impact on greenhouse gas (GHG) emissions and their influencing factors remains unclear. Therefore, a 2-year field experiment was conducted to compare GHG emissions under conventional irrigation (CK) and CI treatments. In addition, we evaluated the applicability of the denitrification–decomposition (DNDC) model in cold-region paddy fields and analyzed the sensitivity of model parameters to determine the main factors affecting GHG emissions. Results demonstrate that compared with CK, CI markedly reduced CH₄ emissions by 45.5 % while increasing rice yield by 7.0 %. This synergistic improvement resulted in a net 3.6 % reduction in global warming potential (GWP) and a 10.3 % decrease in GHG intensity. Notably, CI considerably increased N₂O emissions by 96.8 % compared with CK, highlighting the need to consider N₂O emissions in CI management practices. The DNDC model performed well in simulating CO₂ and CH₄ emission fluxes from rice paddies (coefficient of determination [R²] > 0.85 and relative root mean square error [RRMSE] < 0.5). Compared with CK, the DNDC model showed greater sensitivity to changes in field water level and sun-drying duration under cold-region CI (change in sensitivity index [SIchange] > 0.1), resulting in GHG emissions under CI varying by over 10 % when these parameters simultaneously change. In addition, combining CI with straw returning and nitrogen application is a collaborative field management practice to improve rice yield and reduce emissions. Overall, the findings suggest that CI can reduce greenhouse gas emissions while maintaining or improving rice yield, offering a promising strategy for sustainable rice cultivation in cold regions.