What matters when? Temporal development of drivers and sources of nitrous oxide emissions in winter wheat
- 1ETH Zurich (Zurich, Switzerland)
Abstract
Agriculture is the largest anthropogenic source of nitrous oxide (N ₂ O), primarily due to nitrogen (N) fertilization. Understanding how the influence of key drivers and the relative contribution of source processes change throughout cropping seasons is crucial for developing effective strategies to mitigate N ₂ O emissions. In this study, we combined high-resolution eddy covariance (EC) flux measurements and explainable machine learning (ML) across the 2022–2023 winter wheat season and the subsequent summer cover crop at a temperate site in Switzerland, complemented by stable isotope analyses conducted in March 2023 following a fertilization event. We used Shapley Additive exPlanations (SHAP) to interpret an XGBoost model trained to predict N ₂ O fluxes, enabling a time-resolved assessment of driver contributions. Crop establishment and early spring emerged as particularly vulnerable periods, characterized by mismatches between N supply and plant demand that led to surplus soil mineral N and elevated N ₂ O fluxes under favorable moisture and temperature conditions. Elevated gross primary productivity (GPP, used as a proxy for crop activity and N demand) showed a consistent negative effect on N ₂ O fluxes, supporting the role of active vegetation as a competing N sink. Stable isotope source partitioning identified denitrification as the dominant N ₂ O production pathway during the early-spring post-fertilization period, consistent with poor soil drainage and high soil moisture. Cumulative N ₂ O emissions over the nine-month winter wheat season reached 5.5 kg N ₂ O-N ha ⁻¹ , with yield-scaled emissions of 1.1 kg N ₂ O-N Mg ⁻¹ grain, both values exceeding those typically reported for wheat. Our findings highlight the value of combining EC, explainable ML, and stable isotope approaches to resolve the temporal dynamics of N ₂ O emissions and identify vulnerable windows for targeted mitigation in cropping systems.
Versions
Related
- Precession and meltwater control of greenhouse gas emissions from terrestrial sources during the last glacial period
- Investigation of Drivers and Sources of Potential Harmful Algal Blooms in Lake Anna, Virginia, 2023–24
- Importance, drivers and trends of inland water greenhouse gas emissions
- Annual mean methane emissions from individual oil-gas point sources worldwide inferred from multi-satellite plume-imaging observations
- The role of financial development on carbon emissions in Sub-Saharan African countries in a nonlinear framework
BibTeX
@article{turco2026what,
title = {{What matters when? Temporal development of drivers and sources of nitrous oxide emissions in winter wheat}},
author = {Fabio Turco and Iris Feigenwinter and Lorenz Allemann and Lukas Hörtnagl and Frank Liebisch and Nina Buchmann},
journal = {Biogeosciences},
year = {2026},
month = oct,
doi = {10.5194/bg-23-7151-2026},
url = {https://bg.copernicus.org/articles/23/7151/2026/},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - What matters when? Temporal development of drivers and sources of nitrous oxide emissions in winter wheat
AU - Fabio Turco
AU - Iris Feigenwinter
AU - Lorenz Allemann
AU - Lukas Hörtnagl
AU - Frank Liebisch
AU - Nina Buchmann
PY - 2026
DA - 2026/10/12
JO - Biogeosciences
DO - 10.5194/bg-23-7151-2026
UR - https://bg.copernicus.org/articles/23/7151/2026/
AB - Agriculture is the largest anthropogenic source of nitrous oxide (N ₂ O), primarily due to nitrogen (N) fertilization. Understanding how the influence of key drivers and the relative contribution of source processes change throughout cropping seasons is crucial for developing effective strategies to mitigate N ₂ O emissions. In this study, we combined high-resolution eddy covariance (EC) flux measurements and explainable machine learning (ML) across the 2022–2023 winter wheat season and the subsequent summer cover crop at a temperate site in Switzerland, complemented by stable isotope analyses conducted in March 2023 following a fertilization event. We used Shapley Additive exPlanations (SHAP) to interpret an XGBoost model trained to predict N ₂ O fluxes, enabling a time-resolved assessment of driver contributions. Crop establishment and early spring emerged as particularly vulnerable periods, characterized by mismatches between N supply and plant demand that led to surplus soil mineral N and elevated N ₂ O fluxes under favorable moisture and temperature conditions. Elevated gross primary productivity (GPP, used as a proxy for crop activity and N demand) showed a consistent negative effect on N ₂ O fluxes, supporting the role of active vegetation as a competing N sink. Stable isotope source partitioning identified denitrification as the dominant N ₂ O production pathway during the early-spring post-fertilization period, consistent with poor soil drainage and high soil moisture. Cumulative N ₂ O emissions over the nine-month winter wheat season reached 5.5 kg N ₂ O-N ha ⁻¹ , with yield-scaled emissions of 1.1 kg N ₂ O-N Mg ⁻¹ grain, both values exceeding those typically reported for wheat. Our findings highlight the value of combining EC, explainable ML, and stable isotope approaches to resolve the temporal dynamics of N ₂ O emissions and identify vulnerable windows for targeted mitigation in cropping systems.
SN - 1726-4189
ER -