Search

Background: Globally, rice systems are a major source of atmospheric CH4 and for major rice-producing countries, such as Vietnam, CH4 as well as N2O emissions from agricultural land used for rice production may represent about one-fourth of total national anthropogenic greenhouse gas (GHG) emissions. However, national-scale estimates of GHG emissions from rice systems are uncertain with regard to its magnitude, spatial distribution, and seasonality. Aims: Here, we used the biogeochemical model LandscapeDNDC to calculate emissions of CH4 and N2O from rice systems in Vietnam (Tier 3 IPCC approach). Our objectives were to identify hotspot regions of emissions and to assess the contribution of N2O to the total non-CO2 (CH4+N2O) GHG balance of rice systems as well as the seasonal and interannual variability of fluxes in dependence of uncertain input data on field management. Methods: The biogeochemical model LandscapeDNDC model was linked to publicly available information on climate, soils, and land management (fertilization, irrigation, crop rotation) for calculating a national inventory in daily time steps of CH4 and N2O emissions from rice systems at a spatial resolution of 0.083° × 0.083°. Uncertainty in management practices related to fertilization, use of harvest residues or irrigation water, and its effects on simulated CH4 and N2O fluxes was accounted for by Latin Hypercube Sampling of probability distribution functions. Results: Our study shows that CH4 and N2O fluxes from rice systems in Vietnam are highly seasonal, with national CH4 and N2O emissions totaling to about 2600 Gg CH4 year–1 and 42 Gg N2O year–1, respectively. Highest emissions were simulated for double and triple rice cropping systems in the Mekong Delta region. Yield-scaled emissions varied largely in a range of 300–3000 kg CO2-eq Mg–1 year–1, with CH4 emissions during the rice season(s) dominating (>82%) the total annual non-CO2 GHG balance of rice systems. In our study, uncertainty in field management information (nitrogen fertilization, ratio synthetic to organic fertilization, residue management, availability of irrigation water) were major drivers of uncertainty of the national CH4 and N2O emission inventory. Conclusions: Our study shows that Tier 3 approaches, that is, process-oriented model approaches combined with GIS databases, for estimating national-scale GHG emissions from rice systems are ready to be applied at national scale. Generally, this approach is powerful as it allows to identify regions with elevated emissions, thereby accounting not only for CH4, but as well for N2O emissions. However, our study also shows that specifically better information on land management is required to narrowing uncertainties.