Importance of in-situ measurements of both N2O and N2 emissions to calibration of biogeochemical models to simulate N budgets

Denitrification is a key process in the global nitrogen (N) cycle, causing nitrous oxide (N₂O) and dinitrogen (N₂) emissions. Even though denitrification is assumed to be a major N loss pathway from agroecosystems, field-scale estimates of both N₂O and N₂ are scarce, reflecting methodological difficulties in measuring and upscaling N₂ emissions. Mechanistic biogeochemical models allow estimates of seasonal denitrification losses at the field scale, extrapolating important yet often limited experimental results. However, such predictions rely mostly on N₂O data, meaning that the lack of N₂ data hinders the validation of overall denitrification rates, which remain a major uncertainty for N budgets.

This study investigated denitrification losses and N budgets in two subtropical sugarcane systems using Agricultural Production Systems sIMulator (APSIM) with unique datasets of both N₂O and N₂ emissions measured in the field with the ¹⁵N gas flux method and upscaled over the growing season. Five key soil N parameters in APSIM were identified as influential on N₂O and N₂ emissions via global sensitivity analysis, followed by generalised likelihood uncertainty estimation to determine their posterior distributions using (i) both N₂O and N₂ data and (ii) N₂O data only.

For both approaches, the calibration of APSIM led to larger denitrification (N₂O+N₂) losses and a shift towards N₂ compared to the use of default parameters. Simulated N₂O emissions did not differ between the different calibration approaches. However, simulated N₂ emissions were larger and agreed better with the observed values when calibrated with both N₂O and N₂ consistently across sites. This approach also improved the simulation of fertiliser N losses via denitrification, leaching and runoff, compared to the observed fertiliser ¹⁵N loss at harvest.

These findings indicate that biogeochemical models commonly used with default soil N parameters or calibration limited to N₂O data are likely to underestimate denitrification losses, producing a bias in simulations of N budgets. Our findings also highlight the importance to integrate in-situ measurements of N₂O and N₂ in simulation exercises, and demonstrate how innovative isotope methods can be used to inform biogeochemical models, ensuring more accurate N budget estimates across scales.