Validation of the DNDC-Rice model by using CH4 and N2O flux data from rice cultivated in pots under alternate wetting and drying irrigation management

The DNDC (DeNitrification-DeComposition)-Rice model, one of the most advanced process-based models for the estimation of greenhouse gas emissions from paddy fields, has been discussed mostly in terms of the reproducibility of observed methane (CH ₄) emissions from Japanese rice paddies, but the model has not yet been validated for tropical rice paddies under alternate wetting and drying (AWD) irrigation management, a water-saving technique. We validated the model by using CH ₄ and nitrous oxide (N ₂O) flux data from rice in pots cultivated under AWD irrigation management in a screen-house at the International Rice Research Institute (Los Baños, the Philippines). After minor modification and adjustment of the model to the experimental irrigation conditions, we calculated grain yield and straw production. The observed mean daily CH ₄ fluxes from the continuous flooding (CF) and AWD pots were 4.49 and 1.22 kg C ha ⁻¹ day ⁻¹, respectively, and the observed mean daily N ₂O fluxes from the pots were 0.105 and 34.1 g N ha ⁻¹ day ⁻¹, respectively. The root-mean-square errors, indicators of simulation error, of daily CH ₄ fluxes from CF and AWD pots were calculated as 1.76 and 1.86 kg C ha ⁻¹ day ⁻¹, respectively, and those of daily N ₂O fluxes were 2.23 and 124 g N ha ⁻¹ day ⁻¹, respectively. The simulated gross CH ₄ emissions for CF and AWD from the puddling stage (2 days before transplanting) to harvest (97 days after transplanting) were 417 and 126 kg C ha ⁻¹, respectively; these values were 9.8% lower and 0.76% higher, respectively, than the observed values. The simulated gross N ₂O emissions during the same period were 0.0279 and 1.45 kg N ha ⁻¹ for CF and AWD, respectively; these values were respectively 87% and 29% lower than the observed values. The observed total global warming potential (GWP) of AWD resulting from the CH ₄ and N ₂O emissions was approximately one-third of that in the CF treatment. The simulated GWPs of both CF and AWD were close to the observed values despite the discrepancy in N ₂O emissions, because N ₂O emissions contributed much less than CH ₄ emissions to the total GWP. These results suggest that the DNDC-Rice model can be used to estimate CH ₄ emission and total GWP from tropical paddy fields under both CF and AWD conditions.