Abstract
The traditional calibration approach for process-based models, such as DayCent, consists of the iterative adjustment of model parameters and comparison of the simulated total N 2 O flux to measured observations. However, the contributions of individual production pathways, namely nitrification and denitrification, are uncertain. Here, N 2 O emissions from the soil of sugar beet plots with control (Null) and mineral (NPK) fertilizer treatments were measured by a static chamber technique. The isotopic composition of emitted N 2 O was analyzed to identify the N 2 O production pathways. The latter showed that denitrification was the predominant source of N 2 O emissions at this site. The model’s default settings strongly overestimated the contributions of nitrification. This incorrect allocation of N 2 O emissions to nitrification could partly be explained by the model’s tendency to underestimate the soil water content during the growing season. DayCent model parameters were also manually adjusted to better represent the observation derived contributions of nitrification and denitrification. Although, this “expert-informed approach”, showed a slightly lower performance concerning the cumulative N 2 O flux (Null: RMSE = 0.37 kg N ha −1 yr −1 , NPK: RMSE = 0.50 kg N ha −1 yr −1 ) than the traditional calibration (Null: RMSE = 0.15 kg N ha −1 yr −1 , NPK: RMSE = 0.10 kg N ha −1 yr −1 ), it may be considered as more representative because it better reflected the higher contribution from denitrification shown by the isotope data. This study demonstrates that the inclusion of observational methods, such as isotope measurements, can provide important insight into model function and improve pathway-specific estimation of N 2 O emissions in DayCent.