Rfallow and related environmental, management and sampling factors from published studies.

Croplands account for ~1/3 of global anthropogenic nitrous oxide (N2O) emissions. A number of recent field experiments found substantial fallow-period N2O emissions, which have been neglected for decades. However, the global contribution of the fallow-period emissions and the associated drivers remain unclear. Based on 360 observations across global agroecosystems, we simulated the ratio of the fallow to the whole-year N2O emissions (Rfallow) by developing a mixed-effect model and compiling cropping-system-specific input data. Our results revealed that the mean global gridded Rfallow was 44% (15-75%, 95% confidence interval), with hotspots mainly in the northern high latitudes. For most cropping systems, soil pH was the dominant driver of global variation in Rfallow. Global cropland emission factors (i.e., the percentage of fertilizer N emitted as N2O, EFs) in EF-based models doubled to 1.9% when the fallow-period N2O emissions were included in our simulation, similar to EFs estimated by process-based and atmospheric inversion models (1.8-2.3%). Overall, our study highlights the importance of fallow-period N2O emissions in annual totals, especially for single cropping systems and croplands in acidic areas. To accurately estimate N2O emissions for national greenhouse gas inventories, it is crucial to update current EFs with full consideration of the fallow-period N2O emissions in the Intergovernmental Panel on Climate Change (IPCC) Tier 1 method.