Characterization of atmospheric nitrous oxide emissions from global agricultural soils

Nitrous oxide (N2O) is a potent greenhouse gas with an atmospheric lifetime of similar to 114 years. Agriculture activities are the main sources for N2O emission into the atmosphere by human activities. Global N2O emissions into the atmosphere are projected to increase in the coming years as demand for food, fibre and energy increases owing to increasing global population. Here, a statistical model (N2O_STAT) is developed for characterizing atmospheric N2O emissions from agricultural sources. We obtained N2O emissions and physicochemical variables (i.e. air temperature, soil temperature, soil moisture, soil pH, and N input to the soil) from published journal articles since 2000. A statistical model was developed by expressing a multiple linear regression equation between N2O emission and the physicochemical variables. The model was evaluated for 2012 N2O emissions. Results of the model are compared with other global and regional N2O models (e.g. EDGAR, EPA/USGS, and FAOSTAT). In comparison with other data sets, the model generates a lower global N2O estimate by 920% (N2O_STAT: 3.75 Tg N yr(-1); EDGAR: 4.49 Tg N yr(-1); FAO: 4.07 Tg N yr(-1)), but is similar to 25% higher when compared to Bouwman et al. (Glob Biogeochem Cycles 16:19. https://doi.org/10.1029/2001gb001812,2002) (2.80 Tg N yr(-1)). We also performed a region-based analysis (USA, India, and China) using the N2O_STAT model. For the USA, our model produces an estimate that ranges from - 13 to + 32% in comparison with other published data sets. Meanwhile, the N2O_STAT model estimate for India shows N2O emissions between - 56 and + 14% when compared to other data sets. A much lower estimate is seen for China, where the model estimates N2O emissions 38-177% lower than other data sets. The N2O_STAT model provides an opportunity to predict future N2O emissions in a changing world.