Agroecosystem Model Research Articles

An open platform to build, evaluate and simulate integrated models of farming and agro-ecosystems

Due to significant changes in agro-ecological contexts, farmers need new solutions to produce goods. Modelling complements field experiments in the design of new farming systems. French researchers involved in such design issues developed a specific modelling platform to help model, simulate and evaluate cropping systems. After testing several existing environments, the RECORD platform was developed under the VLE environment, allowing the design of atomic and coupled models. It integrates different time steps and spatial scales and proposes some standard formalisms used to model agro-ecosystems (e.g. difference equations, differential equations, state charts…). A graphic user interface was designed to simplify coding tasks. A variety of research projects already use this platform. Examples are given showing the ability to recode simple models, encapsulate more complex models, link with GIS and databases, and use the R statistical package to run models and analyse simulation outputs. The option to use web interfaces enables application by non-scientist end-users. As the models follow a given standard, they can be placed in a repository and used by other researchers. Linking RECORD to other international platforms is now a compelling issue.

Impacts of Urbanization on Ecosystem Goods and Services in the U.S. Corn Belt

In this study, three cities located in the U.S. Corn Belt are evaluated for impacts of past (1992–2001) and projected (2001–2030) urban expansion on ecosystem goods and services, with a specific focus on changes in energy balance, hydrology, and productivity. Scenarios for high-, medium- and low-density urban areas are simulated using a dynamic agro-ecosystem model (Agro-IBIS), by incorporating new parameterizations for impervious surfaces and turf grass. Moderate Resolution Imaging Spectroradiometer (MODIS) 500-m albedo data and remote sensing-derived 30-m resolution maps from the U.S. National Land Cover Database are used as model input to simulate biogeochemical, thermodynamic, and hydrological cycles for a range of land-cover types in each region. The results show that the expanding urban areas have a significant impact on each city’s capacity to regulate climate and flooding. High-density urban areas, for instance, have soil surface temperatures up to 6°C higher than soils within natural and managed ecosystems. Expansion of turf grass in residential areas could require an additional 8–105 million m3 of water use annually, which increases runoff by 15–48% and reduces the capacity to respond/adapt to flooding. Finally, the analysis shows that net primary productivity (NPP) decreases as expected due to the removal of cropland, forests, and grasslands in favor of development, but increased urban turf grass provides an annual offset of 40–210 g C m−2. Urban expansion through 2030 is estimated to lower total annual crop production by 8.1, 8.6, and 16.7% for the Madison, Peoria, and Indianapolis regions, respectively. Given current projections for city growth to exceed 2–3% per year in the north-central U.S., urban expansion across a nine-state region in the Corn Belt could potentially take an additional 210,000–310,000 ha of farmland out of production annually at a time when demand for food, fuel, and fiber is increasing. Because conversion of cropland to urban uses is nearly always unidirectional, any changes to ecosystem goods and services due to urbanization are likely to be permanent and irreversible.

Global sensitivity and uncertainty analysis of a dynamic agroecosystem model under different irrigation treatments

Savings in consumptive use through limited or deficit irrigation in agriculture has become an increasingly viable source of additional water for places with high population growth such as the Colorado Front Range, USA. Crop models provide a mechanism to evaluate various management methods without performing costly and time-consuming experiments, e.g., field studies investigating irrigation scheduling and timing effects on crop growth. Few studies have focused on CERES-Maize crop model parameterization with respect to water-stressed conditions, and the model has previously been shown to overestimate evapotranspiration (ET) for limited irrigation treatments (stress during vegetative stage). It is therefore desirable to quantify the effects of CERES-Maize input parameters on model output responses typically used for calibration and/or important in limited irrigation management, including vegetative growth, crop yield, and ET. A sensitivity analysis (SA) utilizing the Morris one-at-a-time screening and Sobol’ variance-based methods was performed on CERES-Maize v4.5 input parameters affecting water balance and crop growth including soil hydraulic properties, phenological growth properties, and radiation use efficiency. CERES-Maize output responses of interest for the SA included anthesis date, maturity date, leaf number per stem, maximum leaf area index, yield, and cumulative ET. The SA study utilized five years of multi-replicate field management data (both full and limited irrigation treatments) for each combination of random input parameters. Results comparing the Morris mean and the Sobol’ total sensitivity index showed very high correlation between the two, indicating that in this study the computationally cheaper Morris method could have been used as an effective indicator of input parameter sensitivity. For the full irrigation treatment, CERES-Maize output responses were mostly sensitive to crop cultivar parameters. For the limited irrigation treatment, CERES-Maize leaf area index, yield, and ET output responses were highly influenced by soil lower limit and drained upper limit input parameters, which define water holding capacity. There was also a greater amount of interaction between input parameters for the limited irrigation treatment than for full irrigation. An uncertainty analysis was also conducted using model outputs from the Sobol’ SA method. In some cases, cumulative ET had higher values for limited irrigation than for full irrigation, further indicating the need to evaluate model processes governing ET under water stress. A new methodology for systematic calibration of CERES-Maize, based on the Morris and Sobol’ sensitivity indices for the two irrigation treatments, is proposed for future model evaluation as sensitivity differences between treatments indicates that existing CERES-Maize calibration procedures (typically based on non-stressed crops) may need to be reconsidered in cases of water stress.

Understanding Crop Response to Climate Variability with Complex Agroecosystem Models

Agroecosystem modeling studies often rely on relatively short time-series historical records for training/tuning empirical parameters and to predict long-term variation in crop production associated with trends in climate and hydrological forcing. While ecosystem models may exhibit similar prediction skill in validation studies, their sensitivity to climate variability can differ significantly. Such discrepancy often arises due to the need to tradeoff model complexity with data availability. We examine the sensitivity in predicting spring wheat crop productivity across agricultural sites with differing soil and climate conditions where long-term agronomic and climate records are available. We report significant changes in the model sensitivity accompanying changing climatic regime. If not corrected for, this can lead to substantial predictive error when simulating across time and space. Our findings lend further support for a hierarchical (componentwise) approach for reducing model complexity and improving prediction skill.

Development of a regional crop growth model for Northeast China

State-of-the-art crop growth models are powerful tools to assess the response of crops to altered environmental conditions and cultivation practices. In order to develop effective strategies for a sustainable management of resources and for adapting to environmental conditions, these models are applied not only on the field scale but also on the regional scale. So far the regional DANUBIA crop growth model has been successfully validated and applied in Germany. A new challenge is the extension and transfer of this process-based and object-oriented model to an intensively used agricultural region in Northeast China. This study area is part of the Sanjiang Plain in the Heilongjiang (Amur) river region, where a few decades ago, wetland ecosystems were converted to mainly large-scale irrigated rice fields. In this contribution, the generic DANUBIA crop growth model will be presented. Included are (i) the necessary steps for extending the model for simulating rice growth as well as (ii) a concept for integrating optical remote sensing and laser scanning data into the model. This method aims at improving the model results by analyzing and updating observed vegetation characteristics during a model run. Assimilating these data into a state-of-the-art crop growth model provides a large innovative potential for regional agro-ecosystem modelling.

Influence of Agro-Ecosystem Modeling Approach on the Greenhouse Gas Profiles of Wheat-Derived Biopolymer Products

An approach is presented to include a wider range of factors involved in the nitrogen and carbon cycles in agro-ecosystems than is typical of many Life Cycle Assessments (LCAs) of agriculture-based products. This use results from the process-oriented Denitrification-Decomposition (DNDC, modified version) model. Here we evaluate the effects of using site-specific N(2)O emissions derived from the DNDC model rather than the values derived from the commonly used Intergovernmental Panel on Climate Change (IPCC) Tier 1 empirical model on the results of whole life cycle greenhouse gas (GHG) profiles for wheat-based biopolymer products. Statistical methods were also used to analyze the quality of the DNDC and IPCC outputs and to characterize the uncertainty in the GHG results. The results confirm that the GHG profiles of the wheat-derived biopolymer products are sensitive to how the agricultural system is modeled and uncertainty analyses indicate that DNDC is preferred over the IPCC Tier 1 approach for site-specific LCAs. The former allows inclusion of a wider range of important site-specific agricultural parameters in the LCA, provides for improved quality in the LCA data, and permits better calibration of uncertainty in the LCA inventory.

A coffee agroecosystem model: I. Growth and development of the coffee plant

This paper is the first of three on the coffee production system consisting of (1) the coffee plant, (2) coffee berry borer (CBB) and (3) the role of CBB parasitoids. A previous simulation model of the coffee plant was developed using data from Brazil where coffee phenology is characterized by distinct seasonal flowering ( Gutierrez et al., 1998). In contrast, flowering in Colombia is continuous with low seasonality. To capture the differences in coffee phenology and growth in the two climatic regions, the Gutierrez et al. (1998) model was modified using new data from Colombia. The modifications to the model include: (1) The effect of solar radiation on floral buds initiation; (2) An age structure population model to track the daily input and development of the floral buds; (3) The effect of leaf water potential on breaking dormancy in flower buds, and hence on the timing and intensity of flowering; (4) The incorporation of both the vegetative and the reproductive demands to predict the photosynthetic rate. (5) The effect of low temperature on photosynthesis and defoliation. Other aspects of the model were re-interpreted and refinements made to generalize its structure for use across coffee varieties and geographic areas. The model, without modification, realistically simulates field data from Brazil and two Colombian locations having different varieties, patterns of rainfall and hence flowering phenology. The model will be used as the base trophic level for incorporating CBB and high tropic levels effects, and for the analysis of management options in the coffee production system.

Predicting and mitigating the net greenhouse gas emissions of crop rotations in Western Europe

Nitrous oxide, carbon dioxide and methane are the main biogenic greenhouse gases (GHGs) contributing to net greenhouse gas balance of agro-ecosystems. Evaluating the impact of agriculture on climate thus requires capacity to predict the net exchanges of these gases in a systemic approach, as related to environmental conditions and crop management. Here, we used experimental data sets from intensively monitored cropping systems in France and Germany to calibrate and evaluate the ability of the biophysical crop model CERES-EGC to simulate GHG exchanges at the plot-scale. The experiments involved major crop types (maize–wheat–barley–rapeseed) on loam and rendzina soils. The model was subsequently extrapolated to predict CO2 and N2O fluxes over entire crop rotations. Indirect emissions (IE) arising from the production of agricultural inputs and from use of farm machinery were also added to the final greenhouse gas balance. One experimental site (involving a maize–wheat–barley–mustard rotation on a loamy soil) was a net source of GHG with a net GHG balance of 670kgCO2-Ceqha−1yr−1, of which half were due to IE and half to direct N2O emissions. The other site (involving a rapeseed–wheat–barley rotation on a rendzina) was a net sink of GHG for −650kgCO2-Ceqha−1yr−1, mainly due to high C returns to soil from crop residues. A selection of mitigation options were tested at one experimental site, of which straw return to soils emerged as the most efficient to reduce the net GHG balance of the crop rotation, with a 35% abatement. Halving the rate of N inputs only allowed a 27% reduction in net GHG balance. Removing the organic fertilizer application led to a substantial loss of C for the entire crop rotation that was not compensated by a significant decrease of N2O emissions due to a lower N supply in the system. Agro-ecosystem modeling and scenario analysis may therefore contribute to design productive cropping systems with low GHG emissions.

Evaluating the ability of four crop models to predict different environmental impacts on spring wheat grown in open-top chambers

We used the modeling package Expert-N to investigate the ability of four generic-mechanistic crop models that were originally developed under field conditions to simulate the plant growth of spring wheat grown in open-top chambers (OTC) under different environmental conditions. We focus on the impacts of water limitation and elevated atmospheric CO 2 concentration on biomass production. Expert-N facilitates the comparison of the components of agro-ecosystem models because it allows the exchange of single modules while leaving the rest of the model unchanged. The crop growth part of the models SPASS, CERES-Wheat, SUCROS and GECROS were combined with the Penman–Monteith equation for potential evapotranspiration, the HYDRUS-1D model for water transport and the LEACHN model for nitrogen transport and turnover simulation. The models were applied to a data set provided by OTC experiments with spring wheat ( Triticum aestivum L. cv. ‘Minaret’) that was grown under two atmospheric CO 2 concentrations (ambient/elevated), two irrigation schemes (non-limited water supply/water limitation) and two soil types (Cambisol/Chernosem) in two subsequent vegetation periods (1998/1999). Based on the model calibration using experimental and literature data, the best simulation results describing the impact of the considered environmental conditions were obtained using the SUCROS model followed by the SPASS, GECROS and CERES models. The study depicts the shortcomings of the underlying processes in all of the models. These shortcomings need to be addressed when models are applied on regional scales or for prediction under climate change conditions.

Modeling water and soil quality environmental impacts associated with bioenergy crop production and biomass removal in the Midwest USA

The removal of corn stover or production of herbaceous crops such as switchgrass ( Panicum virgatum) or big bluestem ( Andropogon gerardii) as feedstocks for bioenergy purposes has been shown to have significant benefits from an energy and climate change perspective. There is potential, however, to adversely impact water and soil quality, especially in the United States Corn Belt where stover removal predominantly occurs and possibly in other areas with herbaceous energy crops depending upon a number of geo-climatic and economic factors. The overall goal of this research was to provide a thorough and mechanistic understanding of the relationship between stover and herbaceous crop production management practices and resulting range of impacts on soil and water quality, with a focus on eastern Iowa, USA. Comparisons of the production of herbaceous bioenergy crops to continuous corn ( Zea mays L.) and corn-soybean ( Glycine max L.) rotations on five different soils representative of the region were performed. Indices for total nutrient (nitrogen and phosphorus) loss to surface water and groundwater, total soil loss due to water and wind erosion, and cumulative soil carbon loss were derived to assess long-term sustainability. The Agricultural Policy/Environmental eXtender (APEX) agroecosystem model was used to quantify the sustainability indices and to generate sufficient data to provide a greater understanding of variables that affect water and soil quality than previously possible. The results clearly show the superiority of herbaceous crop production from a soil and water quality perspective. They also show, however, that compared to traditional cropping systems (e.g., corn-soybean rotations with conventional tillage), soil and water quality degradation can be reduced under certain conditions at the same time stover is removed.

Crop management and phenology trends in the U.S. Corn Belt: Impacts on yields, evapotranspiration and energy balance

Crop yields are affected by many factors, related to breeding, management and climate. Understanding these factors, and their relative contributions to historical yield increases, is important to help ensure that these yield increases can continue in the future. Two important factors that can affect yields are planting dates and the crop's growing degree day (GDD) requirements. We analyzed 25 years of data collected by the USDA in order to document trends in planting dates, lengths of the vegetative and reproductive growth periods, and the length of time between maturity and harvest for corn and soybeans across the United States. We then drove the Agro-IBIS agroecosystem model with these observations to investigate the effects of changing planting dates and crop GDD requirements on crop yields and fluxes of water and energy. Averaged across the U.S., corn planting dates advanced about 10 days from 1981 to 2005, and soybean planting dates about 12 days. For both crops, but especially for corn, this was accompanied by a lengthening of the growth period. The period from corn planting to maturity was about 12 days longer around 2005 than it was around 1981. A large driver of this change was a 14% increase in the number of GDD needed for corn to progress through the reproductive period, probably reflecting an adoption of longer season cultivars. If these changes in cultivars had not occurred, yields around 2005 would have been 12.6 bu ac −1 lower across the U.S. Corn Belt, erasing 26% of the yield increase from 1981 to 2005. These changes in crop phenology, together with a shortening of the time from maturity to harvest, have also modified the surface water and energy balance. Earlier planting has led to an increase in the latent heat flux and a decrease in the sensible heat flux in June, while a shorter time from maturity to harvest has meant an increase in net radiation in October.

The MONICA model: Testing predictability for crop growth, soil moisture and nitrogen dynamics

A fundamentally revised version of the HERMES agro-ecosystem model, released under the name of MONICA, was calibrated and tested to predict crop growth, soil moisture and nitrogen dynamics for various experimental crop rotations across Germany, including major cereals, sugar beet and maize. The calibration procedure also included crops grown experimentally under elevated atmospheric CO 2 concentration. The calibrated MONICA simulations yielded a median normalised mean absolute error ( nMAE) of 0.20 across all observed target variables ( n = 42) and a median Willmott's Index of Agreement ( d) of 0.91 (median modelling efficiency ( ME): 0.75). Although the crop biomass, habitus and soil moisture variables were all within an acceptable range, the model often underperformed for variables related to nitrogen. Uncalibrated MONICA simulations yielded a median nMAE of 0.27 across all observed target variables ( n = 85) and a median d of 0.76 (median ME: 0.30), also showing predominantly acceptable results for the crop biomass, habitus and soil moisture variables. Based on the convincing performance of the model under uncalibrated conditions, MONICA can be regarded as a suitable simulation model for use in regional applications. Furthermore, its ability to reproduce the observed crop growth results in free-air carbon enrichment experiments makes it suited to predict agro-ecosystem behaviour under expected future climate conditions.

The next generation models for crops and agro-ecosystems

Growth in population, decrease in arable land area, and change in climate are endangering our food security. Precision agriculture has the potential to increase crop productivity thorough tailored agricultural practices for different growing areas. Many models of crops and agro-ecosystems capable of predicting interaction between plants and environments have been developed for precision agriculture. Currently, there are several representative categories of crop and agro-ecosystem models, including the de Wit school models, the DSSAT series models and the APSIM series models, which have contributed substantially to improvement of agricultural practices. However, these models are weak in predicting performances of crops under environmental and genetic perturbations are generally weak, which severely limits the application of these models in guiding precision agriculture. We need to develop the next generation crop and agro-ecosystems models with a high level of mechanistic basis, which can be integrated with high throughput data and can predict the heterogeneity of environmental factors inside canopy and dynamic canopy photosynthesis. In developing such a model close collaboration is inevitably required among scientists from different disciplines. The successful development and application of such models will undoubtedly advance precision agriculture through providing better agronomical practices tailored for different growing environments. These models will also form a basis to identify breeding targets for increased productivity at given location with given soil and climatic conditions.

The effect of situational variability in climate and soil, choice of animal type and N fertilisation level on nitrogen leaching from pastoral farming systems around Lake Taupo, New Zealand

Agricultural systems with grazing animals are increasingly under scrutiny for their contribution to quality degradation of waterways and water bodies. Soil type, climate, animal type and nitrogen (N) fertilisation are contributors to the variation in N that is leached through the soil profile into ground and surface water. It is difficult to explore the effect of these factors using experimentation only and modelling is proposed as an alternative. An agro-ecosystem model, EcoMod, was used to quantify the pastoral ecosystem responses to situational variability in climate and soil, choice of animal type and N fertilisation level within the Lake Taupo region of New Zealand. Factorial combinations of soil type (Oruanui and Waipahihi), climate (low, moderate and high rainfall), animal type (sheep, beef and dairy) and N fertilisation level (0 or 60 kg N/ha/yr) were simulated. High rainfall climates also had colder temperatures, grew less pasture and carried fewer animals overall which lead to less dung and urinary N returned. Therefore, even though a higher proportion of N returned ultimately leached at the higher rainfall sites, the total N leached did not differ greatly between sites. Weather variation between years had a marked influence on N leaching within a site, due to the timing and magnitude of rainfall events. In this region, for these two highly permeable soil types, N applied as fertiliser had a high propensity to leach, either after being taken up by plants, grazed and returned to the soil as dung and urine, or due to direct flow through the soil profile. Soil type had a considerable effect on N leaching risk, the timing of N leaching and mean pasture production. Nitrogen leaching was greatest from beef cattle, followed by dairy and sheep with the level of leaching related to urine deposition patterns for each animal type and due to the amount of N returned to the soil as excreta. Simulation results indicate that sheep farming systems with limited fertiliser N inputs will reduce N leaching from farms in the Lake Taupo catchment.

Nitrogen cycle model of agroecosystem in the karst region of Guizhou Province

According to the transport theory of soil solutes and the conditions of soil, geology and climate in the karst region of Guizhou Province, a numerical simulation model of edaphic nitrogen element transport under field conditions is initially established. In this model, NO−3 and NH+4 are regarded as soil solutes. Transformation mechanisms such as biological release, bio-immobilization, ammonium adsorption-desorption, nitration-denitrification and factors of crop root uptaking are considered in this model. It is hoped that the data from this model could directly be used to guide agricultural production in this region and offer feasible ways to improve the use of nitrogen element, sustainable development of agriculture in karst mountainous areas and natural environment.

Estimating soil strength in the rooting zone of wheat

When roots abstract water thus drying the soil, crop growth may be reduced by increasing strength of soil as well as the lack of water. Strong soil impedes root growth, restricting access to deeper water. As a result, there is a need to estimate soil strength in order to model crop response to dry soil correctly. The strength of soil can be routinely assessed with a penetrometer but measurements are time consuming and hard work to acquire at the frequency required to understand soil-water-plant relations. To make progress, a published relationship that derives penetrometer pressure from both water relations in soil and density was improved to take account of the effects of depth including the friction that results from the increasing hydrostatic pressure. These relationships were then incorporated into an agroecosystem model so that the dynamics of strong soil and its effect on wheat could be simulated. The combined model requires the moisture release curve (but this can be derived from other commonly-measured soil properties), daily rainfall, temperature, and potential evaporation and the agronomy of the crop. Modelled values of penetrometer pressure were simulated well compared with measured values in artificially strengthened (compacted) and weakened (irrigated) soils. Simulations of the strength of soil and the matric potential before anthesis are compared with measured total dry-matter yields of winter wheat in experimental fields. The results lend weight to the hypothesis that wheat yield is limited by the strength of soil in the field and that soil strength, rather than soil matric potential, better explains differences between soils.

Predicting the net carbon exchanges of crop rotations in Europe with an agro-ecosystem model

Carbon fluxes between croplands and atmosphere are highly conditioned by farmer practices that involved intense atmospheric CO 2 uptake during crop growing season compared to other terrestrial ecosystems. Modelling and measuring land-atmosphere carbon exchanges from arable lands are important tasks to predict the influence of vegetation dynamics on climate change and its retroactive effects on crop productivity. We tested the agro-ecosystem model CERES-EGC against gap-filled daily net CO 2 exchanges over crop rotations monitored in three arable sites in Europe. The model parameters were estimated using Bayesian calibration and the model prediction accuracy was assessed with two supplementary independent data sets. As a result, the calibrated model allowed us to compute the net ecosystem production (NEP) and net biome production (NBP) for entire crop rotations. The Bayesian calibration method results in an improvement of goodness of fit compared to initial parameter-based simulations. The calibrated model was accurate to estimate the NEP from daily time scale to aggregated NEP for entire crop rotation. The carbon returns from application of organic manure and the carbon uptake from catch crops and crop volunteers generated an important C sink effect on the NBP. Adding the nitrous oxide and methane fluxes from soils to the CO 2 balance will allow us to compute the total greenhouse gas budget of agro-ecosystems.

Application of the DNDC model to the Rodale Institute Farming Systems Trial: challenges for the validation of drainage and nitrate leaching in agroecosystem models

Ecosystem models are increasingly used to guide natural resource management policy decisions. In this study, we build on available agroecosystem policy modeling tools by testing two methodologies for applying the Denitrification-Decomposition (DNDC) model to naturally-drained, temperate grain cropping systems. We used long-term observations from the Rodale Institute Farming Systems Trial (FST) to validate the DNDC model for application to grain cropping systems on silty clay loam soils typical of mid-Atlantic farmlands. Based on modeling efficiency (EF), Theil’s Inequality (U2), and correlation coefficient (r) metrics, the DNDC model showed moderate fit between observations and simulations at annual time scales for drainage (EF = 0.34, U2 = 0.12, r = 0.74) and nitrate leaching (EF = −0.05, U2 = 0.4, r = 0.86). Replication of observed seasonal water flux and nitrate leaching trends were difficult to capture in model simulations, resulting in a weak fit between observations and simulations for drainage (EF = −1.2, U2 = 0.89, r = 0.28) and nitrate leaching (EF = −2.5, U2 = 2.1, r = 0.3). Our comparison of observations and model outcomes highlights the challenge of scaling up belowground fluxes to farm or watershed scales. Ecosystem model representation of water transport generally assumes highly homogeneous soil conditions. In contrast, data from lysimeter sampling represents a small percentage of the total study area and is unlikely to capture average soil field properties. Additionally, our Rodale work highlights the limitation of biogeochemistry models which use vertical mass movement to describe water drainage and nitrate leaching. The application of the DNDC model to the Rodale FST demonstrates that model studies are not a simple substitute for field observation. The predictive utility of model outcomes can only be broadened through rigorous testing against long-term field observations.

High-resolution inventory of NO emissions from agricultural soils over the Ile-de-France region

Arable soils are a significant source of nitric oxide (NO), a precursor of tropospheric ozone, and thereby contribute to ozone pollution. However, their actual impact on ozone formation is strongly related to their spatial and temporal emission patterns, which warrant high-resolution estimates. Here, we combined an agro-ecosystem model and geo-referenced databases to map these sources over the 12 000 km 2 administrative region surrounding Paris, France, with a kilometric level resolution. The six most frequent arable crop species were simulated, with emission rates ranging from 1.4 kg N–NO ha −1 yr −1 to 11.1 kg N–NO ha −1 yr −1. The overall emission factor for fertilizer-derived NO emissions was 1.7%, while background emissions contributed half of the total NO efflux. Emissions were strongly seasonal, being highest in spring due to fertilizer inputs. They were mostly sensitive to soil type, crops' growing season and fertilizer N rates.

Bayesian calibration of the nitrous oxide emission module of an agro-ecosystem model

Nitrous oxide ( N 2 O ) is the main biogenic greenhouse gas contributing to the global warming potential (GWP) of agro-ecosystems. Evaluating the impact of agriculture on climate therefore requires a capacity to predict N 2 O emissions in relation to environmental conditions and crop management. Biophysical models simulating the dynamics of carbon and nitrogen in agro-ecosystems have a unique potential to explore these relationships, but are fraught with high uncertainties in their parameters due to their variations over time and space. Here, we used a Bayesian approach to calibrate the parameters of the N 2 O submodel of the agro-ecosystem model CERES-EGC. The submodel simulates N 2 O emissions from the nitrification and denitrification processes, which are modelled as the product of a potential rate with three dimensionless factors related to soil water content, nitrogen content and temperature. These equations involve a total set of 15 parameters, four of which are site-specific and should be measured on site, while the other 11 are considered global, i.e. invariant over time and space. We first gathered prior information on the model parameters based on the literature review, and assigned them uniform probability distributions. A Bayesian method based on the Metropolis–Hastings algorithm was subsequently developed to update the parameter distributions against a database of seven different field-sites in France. Three parallel Markov chains were run to ensure a convergence of the algorithm. This site-specific calibration significantly reduced the spread in parameter distribution, and the uncertainty in the N 2 O simulations. The model’s root mean square error (RMSE) was also abated by 73% across the field sites compared to the prior parameterization. The Bayesian calibration was subsequently applied simultaneously to all data sets, to obtain better global estimates for the parameters initially deemed universal. This made it possible to reduce the RMSE by 33% on average, compared to the uncalibrated model. These global parameter values may be used to obtain more realistic estimates of N 2 O emissions from arable soils at regional or continental scales.