Edwin Haas

Agricultural soils are the dominant source of anthropogenic N₂O emissions, yet their high spatial and temporal heterogeneity provides a major challenge for accurately quantifying emissions and evaluating mitigation options. Most national greenhouse gas inventories rely on empirical Tier-1 or Tier-2 emission-factor approaches and therefore do not fully capture the effects of climate variability, soil properties, or management practices. Here, we present a transferable, process-based modelling framework based on the biogeochemical model LandscapeDNDC for determining direct and indirect N₂O emissions from major crops cultivated on mineral soils at the national scale. We apply the method to Germany making use of high-resolution input data provided by the national reporting agencies, estimating N₂O emissions of 35 (29–44) kt N yr^-1(2017–2022 average). This is 28 % higher than the national inventory report (submission 2025), but well within the uncertainty range. In contrast to conventional inventory methods, the framework explicitly accounts for interannual climate variability and can be spatially disaggregated at high resolution, taking into account local variations in soil type, weather and agricultural management practices. Because the model simulates coupled carbon and nitrogen cycling, it also quantifies multiple nitrogen loss pathways and potential changes in carbon stocks simultaneously, providing a consistent basis for evaluating mitigation strategies and their potential trade-offs. Our results demonstrate that process-based modelling can substantially improve the spatial and temporal resolution of agricultural N₂O emissions and provide a platform for developing next-generation national greenhouse gas inventories. While further work is required before the framework fully satisfies all IPCC Tier-3 requirements, it offers a pathway towards a more mechanistic and policy-relevant assessment of agricultural greenhouse gas emissions.

Agricultural soils are the dominant source of anthropogenic N₂O emissions, yet their high spatial and temporal heterogeneity provides a major challenge for accurately quantifying emissions and evaluating mitigation options. Most national greenhouse gas inventories rely on empirical Tier-1 or Tier-2 emission-factor approaches and therefore do not fully capture the effects of climate variability, soil properties, or management practices. Here, we present a transferable, process-based modelling framework based on the biogeochemical model LandscapeDNDC for determining direct and indirect N₂O emissions from major crops cultivated on mineral soils at the national scale. We apply the method to Germany making use of high-resolution input data provided by the national reporting agencies, estimating N₂O emissions of 35 (29–44) kt N yr^-1(2017–2022 average). This is 28 % higher than the national inventory report (submission 2025), but well within the uncertainty range. In contrast to conventional inventory methods, the framework explicitly accounts for interannual climate variability and can be spatially disaggregated at high resolution, taking into account local variations in soil type, weather and agricultural management practices. Because the model simulates coupled carbon and nitrogen cycling, it also quantifies multiple nitrogen loss pathways and potential changes in carbon stocks simultaneously, providing a consistent basis for evaluating mitigation strategies and their potential trade-offs. Our results demonstrate that process-based modelling can substantially improve the spatial and temporal resolution of agricultural N₂O emissions and provide a platform for developing next-generation national greenhouse gas inventories. While further work is required before the framework fully satisfies all IPCC Tier-3 requirements, it offers a pathway towards a more mechanistic and policy-relevant assessment of agricultural greenhouse gas emissions.

Agricultural soils are the dominant source of anthropogenic N₂O emissions, yet their high spatial and temporal heterogeneity provides a major challenge for accurately quantifying emissions and evaluating mitigation options. Most national greenhouse gas inventories rely on empirical Tier-1 or Tier-2 emission-factor approaches and therefore do not fully capture the effects of climate variability, soil properties, or management practices. Here, we present a transferable, process-based modelling framework based on the biogeochemical model LandscapeDNDC for determining direct and indirect N₂O emissions from major crops cultivated on mineral soils at the national scale. We apply the method to Germany making use of high-resolution input data provided by the national reporting agencies, estimating N₂O emissions of 35 (29–44) kt N yr^-1(2017–2022 average). This is 28 % higher than the national inventory report (submission 2025), but well within the uncertainty range. In contrast to conventional inventory methods, the framework explicitly accounts for interannual climate variability and can be spatially disaggregated at high resolution, taking into account local variations in soil type, weather and agricultural management practices. Because the model simulates coupled carbon and nitrogen cycling, it also quantifies multiple nitrogen loss pathways and potential changes in carbon stocks simultaneously, providing a consistent basis for evaluating mitigation strategies and their potential trade-offs. Our results demonstrate that process-based modelling can substantially improve the spatial and temporal resolution of agricultural N₂O emissions and provide a platform for developing next-generation national greenhouse gas inventories. While further work is required before the framework fully satisfies all IPCC Tier-3 requirements, it offers a pathway towards a more mechanistic and policy-relevant assessment of agricultural greenhouse gas emissions.

Agricultural soils are the dominant source of anthropogenic N₂O emissions, yet their high spatial and temporal heterogeneity provides a major challenge for accurately quantifying emissions and evaluating mitigation options. Most national greenhouse gas inventories rely on empirical Tier-1 or Tier-2 emission-factor approaches and therefore do not fully capture the effects of climate variability, soil properties, or management practices. Here, we present a transferable, process-based modelling framework based on the biogeochemical model LandscapeDNDC for determining direct and indirect N₂O emissions from major crops cultivated on mineral soils at the national scale. We apply the method to Germany making use of high-resolution input data provided by the national reporting agencies, estimating N₂O emissions of 35 (29–44) kt N yr^-1(2017–2022 average). This is 28 % higher than the national inventory report (submission 2025), but well within the uncertainty range. In contrast to conventional inventory methods, the framework explicitly accounts for interannual climate variability and can be spatially disaggregated at high resolution, taking into account local variations in soil type, weather and agricultural management practices. Because the model simulates coupled carbon and nitrogen cycling, it also quantifies multiple nitrogen loss pathways and potential changes in carbon stocks simultaneously, providing a consistent basis for evaluating mitigation strategies and their potential trade-offs. Our results demonstrate that process-based modelling can substantially improve the spatial and temporal resolution of agricultural N₂O emissions and provide a platform for developing next-generation national greenhouse gas inventories. While further work is required before the framework fully satisfies all IPCC Tier-3 requirements, it offers a pathway towards a more mechanistic and policy-relevant assessment of agricultural greenhouse gas emissions.

Accurately quantifying crop yield reductions from heat and drought stress is essential for predicting how a changing climate will impact future crop production. Here, we assess the relative contributions of heat and drought stress on crop yield losses across Germany using the process-based ecosystem model LandscapeDNDC. This is achieved via a two-step calibration process, in which key model parameters governing the growth of wheat and maize (the two major crops in Germany) are optimized to simulate crop yields in both stressed and unstressed conditions. We show that recent yield losses in Germany were predominantly caused by drought stress, with heat stress playing only a negligible role. Consequently, yield losses could have been largely eliminated through irrigation, with an average of 156 ± 60 mm of water in 2018 across Germany necessary. We show that irrigation demand increases exponentially with decreasing plant-available water, indicating that the importance of this mitigation measure will grow under future climate conditions. The models’ ability to accurately capture the consequences of recent extreme climatic events on crop yields makes it a suitable candidate for evaluating the implications of climate change on both crop production and, more generally, on the carbon, nitrogen and water cycles of cropping systems in Central Europe.

Agricultural soils are the dominant source of anthropogenic N2O emissions, yet their high spatial and temporal heterogeneity provides a major challenge for accurately quantifying emissions and evaluating mitigation options. Most national greenhouse gas inventories rely on empirical Tier-1 or Tier-2 emission-factor approaches and therefore do not fully capture the effects of climate variability, soil properties, or management practices. Here, we present a transferable, process-based modelling framework based on the biogeochemical model LandscapeDNDC for determining direct and indirect N2O emissions from major crops cultivated on mineral soils at the national scale. We apply the method to Germany making use of high-resolution input data provided by the national reporting agencies, estimating N2O emissions of 35 (29–44) kt N yr-1(2017–2022 average). This is 28 % higher than the national inventory report (submission 2025), but well within the uncertainty range. In contrast to conventional inventory methods, the framework explicitly accounts for interannual climate variability and can be spatially disaggregated at high resolution, taking into account local variations in soil type, weather and agricultural management practices. Because the model simulates coupled carbon and nitrogen cycling, it also quantifies multiple nitrogen loss pathways and potential changes in carbon stocks simultaneously, providing a consistent basis for evaluating mitigation strategies and their potential trade-offs. Our results demonstrate that process-based modelling can substantially improve the spatial and temporal resolution of agricultural N₂O emissions and provide a platform for developing next-generation national greenhouse gas inventories. While further work is required before the framework fully satisfies all IPCC Tier-3 requirements, it offers a pathway towards a more mechanistic and policy-relevant assessment of agricultural greenhouse gas emissions.

Optimising Danish arable management is critical for reducing greenhouse-gas (GHG) emissions and nitrogen (N) losses while maintaining or even improving crop productivity and soil health. Process-based models such as LandscapeDNDC can simulate the effects of management on agroecosystem functioning. However, their computational demand limits large-scale optimisation. Here we present mLDNDCv1.0, a tree-based machine-learning surrogate of LandscapeDNDC that allows for the rapid exploration of large decision spaces while maintaining high fidelity to the parent process-based model's input-output behaviour. We generated a synthetic training set of >45 million LandscapeDNDC simulations from a full factorial of soils, climate (2011–2020), and management options for winter wheat. We benchmarked gradient-boosted tree algorithms (LightGBM, XGBoost, CatBoost) on predictive performance. XGBoost and LightGBM outperformed CatBoost and achieved similar predictive performance for the core indicators in this study. XGBoost, selected as the final model for its much faster inference in our implementation, achieved: soil N ₂ O emissions ( R ² =0.81), NO3- leaching ( R ² =0.84), yield ( R ² =0.93), and for soil-organic-carbon stock changes ( R ² =0.86). When evaluated on real field activity data from Denmark, the surrogate model closely reproduced the process-based model outputs, with particularly strong agreement for crop yield, which was further corroborated by independent observational data. Coupling mLDNDC with the multi-objective evolutionary algorithm NSGA-II allowed us to optimise millions of management combinations within a predefined decision boundary across all winter wheat fields in Denmark. Pareto-optimal solutions reduced N ₂ O emissions by 27.5±4.5 % , NO3- and leaching by 27±3.0 % . These solutions also increased grain yield by 8.5±1.5 % and soil-organic-carbon stocks by 1.2±0.1 % , and improving nitrogen-use efficiency (NUE) by 10±2 % , while turning the system into a net GHG sink (2200±400MgCO2-eqyr-1). These gains were achieved without increasing total fertiliser input. They arose from re-allocating mineral and organic fertliser N input, adjusting incorporation depth, and optimising residue, catch-crop, and irrigation practices. Thus, mLDNDC provides a scalable, transparent framework for country-wide scenario comparison and strategic planning at an annual time scale in climate-smart agriculture.

In this study, we have simulated inventories of arable production and soil carbon and nitrogen cycling on a national scale (0.25 × 0.25-degree) for Greece with the bio-geochemical ecosystem model LandscapeDNDC. Based on observation data, we have aggregated for each grid cell 4 most likely crop rotations, including nitrogen and manure fertilization, tilling and irrigation. The arable management was continuously projected into the future until 2100, while plant phenology was adapted to local conditions, general properties of the arable management were kept constant into the future, such as the selection of crops or the share of irrigated arable land. To understand the impacts of climate change, we used the EURO-CORDEX-11 regional climate ensemble to drive the LandscapeDNDC impact model under scenarios RCP4.5 (16 datasets) and RCP8.5 (32 datasets). The simulation timespan was from 1990 until 2100, using the first 10 years as spin-up to obtain equilibrium in the model's internal carbon and nitrogen pools from the model initialization. Arable production declines from 2045 onwards by 9.5% or 144 kg C ha^-1 yr^-1 under RCP4.5 and by 29% or 484 kg C ha^-1 yr^-1 towards 2100. At present, the ensemble results show an average soil carbon loss of 122.1 kg C ha^-1 yr^-1 versus 139.7 kg C ha^-1 yr^-1 in the future. The gaseous outfluxes of the ensemble simulations show N₂O emissions of 0.494 to 0.453 kg N₂O–N ha⁻¹ yr⁻¹, NO emissions of 0.031 kg NO–N ha⁻¹ yr⁻¹, N₂ emissions of 4.806 to 3.377 kg N₂–N ha⁻¹ yr⁻¹, NH₃ emissions of 24.662 to 35.040 or 34.205 kg NH₃–N ha⁻¹ yr⁻¹ and nitrate leaching losses from 54.304 to 58.213 kg NO₃-N ha⁻¹ yr⁻¹ comparing present versus future conditions. The overall nitrogen balance of the ensemble simulations reveals a mean nitrogen loss of 4.7 versus 5.7 kg-N ha^-1 yr^-1 comparing present to future conditions.

In this study, we have simulated inventories of arable production and soil carbon and nitrogen cycling on a national scale (0.25 × 0.25-degree) for Greece with the bio-geochemical ecosystem model LandscapeDNDC. Based on observation data, we have aggregated for each grid cell 4 most likely crop rotations, including nitrogen and manure fertilization, tilling and irrigation. The arable management was continuously projected into the future until 2100, while plant phenology was adapted to local conditions, general properties of the arable management were kept constant into the future, such as the selection of crops or the share of irrigated arable land. To understand the impacts of climate change, we used the EURO-CORDEX-11 regional climate ensemble to drive the LandscapeDNDC impact model under scenarios RCP4.5 (16 datasets) and RCP8.5 (32 datasets). The simulation timespan was from 1990 until 2100, using the first 10 years as spin-up to obtain equilibrium in the model's internal carbon and nitrogen pools from the model initialization. Arable production declines from 2045 onwards by 9.5% or 144 kg C ha^-1 yr^-1 under RCP4.5 and by 29% or 484 kg C ha^-1 yr^-1 towards 2100. At present, the ensemble results show an average soil carbon loss of 122.1 kg C ha^-1 yr^-1 versus 139.7 kg C ha^-1 yr^-1 in the future. The gaseous outfluxes of the ensemble simulations show N₂O emissions of 0.494 to 0.453 kg N₂O–N ha⁻¹ yr⁻¹, NO emissions of 0.031 kg NO–N ha⁻¹ yr⁻¹, N₂ emissions of 4.806 to 3.377 kg N₂–N ha⁻¹ yr⁻¹, NH₃ emissions of 24.662 to 35.040 or 34.205 kg NH₃–N ha⁻¹ yr⁻¹ and nitrate leaching losses from 54.304 to 58.213 kg NO₃-N ha⁻¹ yr⁻¹ comparing present versus future conditions. The overall nitrogen balance of the ensemble simulations reveals a mean nitrogen loss of 4.7 versus 5.7 kg-N ha^-1 yr^-1 comparing present to future conditions.

Optimising Danish arable management is critical for reducing greenhouse‐gas (GHG) emissions and nitrogen (N) losses while maintaining or even improving crop productivity and soil health. Process-based models such as LandscapeDNDC can simulate the effects of management on agroecosystem functioning. However, their computational demand limits large-scale optimisation. Here we present mLDNDCv1.0, a tree-based machine-learning surrogate of LandscapeDNDC that allows for the rapid exploration of large decision spaces without sacrificing mechanistic fidelity. We generated a synthetic training set of >45 million LandscapeDNDC simulations from a full factorial of soils, climate (2011–2020), and management options for winter wheat. We benchmarked gradient-boosted tree algorithms (LightGBM, XGBoost, CatBoost) on predictive performance. XGBoost delivered the most accurate and stable predictions for the core indicators in this study: soil N₂O emissions (R² = 0.81), NO₃⁻ leaching (R² = 0.84), yield (R² = 0.93), and for soil-organic-carbon stock changes (R² = 0.86). The model maintained high accuracy when confronted with real management and environmental settings that reflected true operating conditions. Coupling mLDNDC with the multi-objective evolutionary algorithm NSGA-II allowed us to optimise millions of management combinations across all winter wheat fields in Denmark. Pareto-optimal solutions reduced N₂O emissions by 27.5 ± 4.5 %, NO₃⁻ and leaching by 27 ± 3.0 %. These solutions also increased grain yield by 8.5 ± 1.5 % and soil-organic-carbon stocks by 1.2 ± 0.1 %, and improving nitrogen-use efficiency (NUE) by 10 ± 2 %, while turning the system into a net GHG sink (2200 ± 400 Mg CO₂-eq ha⁻¹ yr⁻¹). These gains were achieved without increasing total fertiliser input. They arose from re-allocating mineral and organic fertliser N input, adjusting incorporation depth, and optimising residue, catch-crop, and irrigation practices. Thus, mLDNDC therefore provides a scalable, transparent framework for country-wide optimisation and real-time decision support in climate-smart agriculture.