How Modelers Model: the Overlooked Social and Human Dimensions in Model Intercomparison Studies.

There is a growing realization that the complexity of model ensemble studies depends not only on the models used but also on the experience and approach used by modelers to calibrate and validate results, which remain a source of uncertainty. Here, we applied a multi-criteria decision-making method to investigate the rationale applied by modelers in a model ensemble study where 12 process-based different biogeochemical model types were compared across five successive calibration stages. The modelers shared a common level of agreement about the importance of the variables used to initialize their models for calibration. However, we found inconsistency among modelers when judging the importance of input variables across different calibration stages. The level of subjective weighting attributed by modelers to calibration data decreased sequentially as the extent and number of variables provided increased. In this context, the perceived importance attributed to variables such as the fertilization rate, irrigation regime, soil texture, pH, and initial levels of soil organic carbon and nitrogen stocks was statistically different when classified according to model types. The importance attributed to input variables such as experimental duration, gross primary production, and net ecosystem exchange varied significantly according to the length of the modeler's experience. We argue that the gradual access to input data across the five calibration stages negatively influenced the consistency of the interpretations made by the modelers, with cognitive bias in "trial-and-error" calibration routines. Our study highlights that overlooking human and social attributes is critical in the outcomes of modeling and model intercomparison studies. While complexity of the processes captured in the model algorithms and parameterization is important, we contend that (1) the modeler's assumptions on the extent to which parameters should be altered and (2) modeler perceptions of the importance of model parameters are just as critical in obtaining a quality model calibration as numerical or analytical details.

Global evaluation of inhibitor impacts on ammonia and nitrous oxide emissions from agricultural soils: A meta-analysis.

Inhibitors are widely considered an efficient tool for reducing nitrogen (N) loss and improving N use efficiency, but their effectiveness is highly variable across agroecosystems. In this study, we synthesized 182 studies (222 sites) worldwide to evaluate the impacts of inhibitors (urease inhibitors [UI], nitrification inhibitors [NI] and combined inhibitors) on crop yields and gaseous N loss (ammonia [NH3 ] and nitrous oxide [N2 O] emissions) and explored their responses to different management and environmental factors including inhibitor application timing, fertilization regime, cropping system, water management, soil properties and climatic conditions using subgroup meta-analysis, meta-regression and multivariate analyses. The UI were most effective in enhancing crop yields (by 5%) and reducing NH3 volatilization (by 51%), whereas NI were most effective at reducing N2 O emissions (by 49%). The application of UI mitigates NH3 loss and increases crop yields especially in high NH3 -N loss scenarios, whereas NI application would minimize the net N2 O emissions and the resultant environmental impacts especially in low NH3 -N loss scenarios. Alternatively, the combined application of UI and NI enables producers to balance crop production and environmental conservation goals without pollution tradeoffs. The inhibitor efficacy for decreasing gaseous N loss was dependent upon soil and climatic conditions and management practices. Notably, both meta-regression and multivariate analyses suggest that inhibitors provide a greater opportunity for reducing fertilizer N inputs in high-N-surplus systems and presumably favor crop yield enhancement under soil N deficiency situations. The pursuit of an improved understanding of the interactions between plant-soil-climate-management systems and different types of inhibitors should continue to optimize the effectiveness of inhibitors for reducing environmental losses while increasing productivity.

Exploring management strategies to improve yield and mitigate nitrate leaching in a typical radish field in northern China.

It is currently uncertain whether process-based models are capable of assessing crop yield and nitrogen (N) losses while helping to investigate best management practices from vegetable cropping systems. The objectives of this study were to (1) calibrate and evaluate the Denitrification-Decomposition (DNDC) model in simulating crop growth and nitrate leaching in a typical field radish system; (2) optimize management practices to improve radish yield and mitigate nitrate leaching under 20-year climate variability. A five-season in-situ field experiment of spring and autumn radish in northern China was established in the autumn of 2017 and measurements of radish yield, N uptake, soil temperature, soil moisture, drainage, and nitrate leaching were obtained under different N usage. DNDC overall demonstrated "good" to "excellent" performance in simulating radish yield, total biomass, N uptake, and soil temperature across all treatments (6.4% ≤ normalized root mean square error (nRMSE) ≤ 15.5%; 0.12 ≤ Nash-Sutcliffe efficiency (NSE) ≤ 0.88; 0.80 ≤ index of agreement (d) ≤ 0.97). DNDC generally exhibited "fair" performance in estimating soil moisture and drainage (10.9% ≤ nRMSE ≤ 27.2%; -0.18 ≤ NSE ≤ 0.37; 0.69 ≤ d ≤ 0.82) and "good" performance when predicting nitrate leaching (12.4% ≤ nRMSE ≤ 26.7%; -0.59 ≤ NSE ≤ 0.51; 0.68 ≤ d ≤ 0.90). Sensitivity analyses demonstrated that optimized management practices (planting dates, irrigation amount, fertilization rate and timing) could substantially reduce N usage by 40%-50%, irrigation amount by 33%-50%, and nitrate leaching by 86%-95% compared to farmers' practice in radish planting system. This study indicated that a modelling method is helpful for evaluating the biogeochemical effects of management alternatives and identifying optimal management practices in radish production systems of China.

Regional climate influences manure temperature and methane emissions - A pan-Canadian modelling assessment.

This study explores the variation of liquid manure temperature (Tm) and CH4 emissions associated with contrasting regional climates, inter-annual weather variation, and manure storage emptying. As a case-study, six regions across Canada were used, spanning 11°32' latitude and 58°30' longitude. Annual average air temperatures ranged from 3.9 °C (prairie climate) to 10.5 °C (maritime climate), with an overall average of 6.6 °C. A model predicted Tm over 30 years, using daily weather (1971-2000), and over one "normal" year (30-year average weather). Modelled Tm was then used in Manure-DNDC to model daily CH4 emissions. Two manure storage emptying scenarios were simulated: (i) early spring and autumn, or (ii) late spring and autumn. Regional differences were evident as average Tm ranged from 8.9 °C to 14.6 °C across the six locations. Early removal of stored manure led to warmer Tm in all regions, and the most warming occurred in colder regions. Regional climate had a large effect on CH4 emissions (e.g. 1.8× greater in the pacific maritime and great lakes regions than the prairie region). Inter-annual weather variability led to substantial variation in inter-annual CH4 emissions, with coefficient of variation being as high as 20%. The large inter-annual range suggests that field measurements of CH4 emissions need to compare the weather during measurements to historical normals. Early manure storage emptying reduced CH4 emissions (vs late removal) in some regions but had little effect or the opposite effect in other regions. Overall, the results from this modelling study suggest: i) Tm differs substantially from air temperature at all locations, ii) accurate estimates of manure storage CH4 emissions require region-specific calculations using Tm (e.g. in emission inventories), iii) field measurements of CH4 emissions need to consider weather conditions relative to climate normal, and iv) emission mitigation practices will require region-specific measurements to determine impacts.

Ensemble modelling, uncertainty and robust predictions of organic carbon in long-term bare-fallow soils.

Simulation models represent soil organic carbon (SOC) dynamics in global carbon (C) cycle scenarios to support climate-change studies. It is imperative to increase confidence in long-term predictions of SOC dynamics by reducing the uncertainty in model estimates. We evaluated SOC simulated from an ensemble of 26 process-based C models by comparing simulations to experimental data from seven long-term bare-fallow (vegetation-free) plots at six sites: Denmark (two sites), France, Russia, Sweden and the United Kingdom. The decay of SOC in these plots has been monitored for decades since the last inputs of plant material, providing the opportunity to test decomposition without the continuous input of new organic material. The models were run independently over multi-year simulation periods (from 28 to 80 years) in a blind test with no calibration (Bln) and with the following three calibration scenarios, each providing different levels of information and/or allowing different levels of model fitting: (a) calibrating decomposition parameters separately at each experimental site (Spe); (b) using a generic, knowledge-based, parameterization applicable in the Central European region (Gen); and (c) using a combination of both (a) and (b) strategies (Mix). We addressed uncertainties from different modelling approaches with or without spin-up initialization of SOC. Changes in the multi-model median (MMM) of SOC were used as descriptors of the ensemble performance. On average across sites, Gen proved adequate in describing changes in SOC, with MMM equal to average SOC (and standard deviation) of 39.2 (±15.5) Mg C/ha compared to the observed mean of 36.0 (±19.7) Mg C/ha (last observed year), indicating sufficiently reliable SOC estimates. Moving to Mix (37.5 ± 16.7 Mg C/ha) and Spe (36.8 ± 19.8 Mg C/ha) provided only marginal gains in accuracy, but modellers would need to apply more knowledge and a greater calibration effort than in Gen, thereby limiting the wider applicability of models.

Modelling climate change impacts on maize yields under low nitrogen input conditions in sub-Saharan Africa.

Smallholder farmers in sub-Saharan Africa (SSA) currently grow rainfed maize with limited inputs including fertilizer. Climate change may exacerbate current production constraints. Crop models can help quantify the potential impact of climate change on maize yields, but a comprehensive multimodel assessment of simulation accuracy and uncertainty in these low-input systems is currently lacking. We evaluated the impact of varying [CO2 ], temperature and rainfall conditions on maize yield, for different nitrogen (N) inputs (0, 80, 160 kg N/ha) for five environments in SSA, including cool subhumid Ethiopia, cool semi-arid Rwanda, hot subhumid Ghana and hot semi-arid Mali and Benin using an ensemble of 25 maize models. Models were calibrated with measured grain yield, plant biomass, plant N, leaf area index, harvest index and in-season soil water content from 2-year experiments in each country to assess their ability to simulate observed yield. Simulated responses to climate change factors were explored and compared between models. Calibrated models reproduced measured grain yield variations well with average relative root mean square error of 26%, although uncertainty in model prediction was substantial (CV = 28%). Model ensembles gave greater accuracy than any model taken at random. Nitrogen fertilization controlled the response to variations in [CO2 ], temperature and rainfall. Without N fertilizer input, maize (a) benefited less from an increase in atmospheric [CO2 ]; (b) was less affected by higher temperature or decreasing rainfall; and (c) was more affected by increased rainfall because N leaching was more critical. The model intercomparison revealed that simulation of daily soil N supply and N leaching plays a crucial role in simulating climate change impacts for low-input systems. Climate change and N input interactions have strong implications for the design of robust adaptation approaches across SSA, because the impact of climate change in low input systems will be modified if farmers intensify maize production with balanced nutrient management.

Modifying fertilizer rate and application method reduces environmental nitrogen losses and increases corn yield in Ontario.

Nitrogen (N) use in corn production is an important driver of nitrous oxide (N2O) emissions and 4R (Right source, Right rate, Right time and Right place) fertilizer practices have been proposed to mitigate emissions. However, combined 4R practices have not been assessed for their potential to reduce N2O emissions at the provincial-scale while also considering trade-offs with other N losses such as leaching or ammonia (NH3) volatilization. The objectives of this study were to develop, validate, and apply a Denitrification-Decomposition model framework at 270 distinct soil-climate regions in Ontario to simulate corn yield and N2O emissions across eleven fertilizer management scenarios during 1986-2015. The results show that broadcasting fertilizer at the surface without incorporation had the highest environmental N loss which was primarily caused by NH3 volatilization. When injected at planting or at sidedress, the NH3 loss was reduced considerably. However, because more N was left in the soil, injection and sidedressing induced more losses by nitrate leaching and N2O emissions. Reduction of N rate as proposed by the DNDC model did not affect crop yield but decreased leaching and N2O emissions. Addition of inhibitors promoted a further reduction in N2O emission (11-16%) although lesser than the reduction in N rate. Overall, our results emphasize that N rate adjustment following improvements in placement, use of inhibitors, and application timings can mitigate N2O emissions by 42-57% and result in 3-4% greater yields compared to baseline scenario in Ontario corn production.

Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N2 O emissions.

Simulation models are extensively used to predict agricultural productivity and greenhouse gas emissions. However, the uncertainties of (reduced) model ensemble simulations have not been assessed systematically for variables affecting food security and climate change mitigation, within multi-species agricultural contexts. We report an international model comparison and benchmarking exercise, showing the potential of multi-model ensembles to predict productivity and nitrous oxide (N2 O) emissions for wheat, maize, rice and temperate grasslands. Using a multi-stage modelling protocol, from blind simulations (stage 1) to partial (stages 2-4) and full calibration (stage 5), 24 process-based biogeochemical models were assessed individually or as an ensemble against long-term experimental data from four temperate grassland and five arable crop rotation sites spanning four continents. Comparisons were performed by reference to the experimental uncertainties of observed yields and N2 O emissions. Results showed that across sites and crop/grassland types, 23%-40% of the uncalibrated individual models were within two standard deviations (SD) of observed yields, while 42 (rice) to 96% (grasslands) of the models were within 1 SD of observed N2 O emissions. At stage 1, ensembles formed by the three lowest prediction model errors predicted both yields and N2 O emissions within experimental uncertainties for 44% and 33% of the crop and grassland growth cycles, respectively. Partial model calibration (stages 2-4) markedly reduced prediction errors of the full model ensemble E-median for crop grain yields (from 36% at stage 1 down to 4% on average) and grassland productivity (from 44% to 27%) and to a lesser and more variable extent for N2 O emissions. Yield-scaled N2 O emissions (N2 O emissions divided by crop yields) were ranked accurately by three-model ensembles across crop species and field sites. The potential of using process-based model ensembles to predict jointly productivity and N2 O emissions at field scale is discussed.

Review and analysis of strengths and weaknesses of agro-ecosystem models for simulating C and N fluxes.

Biogeochemical simulation models are important tools for describing and quantifying the contribution of agricultural systems to C sequestration and GHG source/sink status. The abundance of simulation tools developed over recent decades, however, creates a difficulty because predictions from different models show large variability. Discrepancies between the conclusions of different modelling studies are often ascribed to differences in the physical and biogeochemical processes incorporated in equations of C and N cycles and their interactions. Here we review the literature to determine the state-of-the-art in modelling agricultural (crop and grassland) systems. In order to carry out this study, we selected the range of biogeochemical models used by the CN-MIP consortium of FACCE-JPI (http://www.faccejpi.com): APSIM, CERES-EGC, DayCent, DNDC, DSSAT, EPIC, PaSim, RothC and STICS. In our analysis, these models were assessed for the quality and comprehensiveness of underlying processes related to pedo-climatic conditions and management practices, but also with respect to time and space of application, and for their accuracy in multiple contexts. Overall, it emerged that there is a possible impact of ill-defined pedo-climatic conditions in the unsatisfactory performance of the models (46.2%), followed by limitations in the algorithms simulating the effects of management practices (33.1%). The multiplicity of scales in both time and space is a fundamental feature, which explains the remaining weaknesses (i.e. 20.7%). Innovative aspects have been identified for future development of C and N models. They include the explicit representation of soil microbial biomass to drive soil organic matter turnover, the effect of N shortage on SOM decomposition, the improvements related to the production and consumption of gases and an adequate simulations of gas transport in soil. On these bases, the assessment of trends and gaps in the modelling approaches currently employed to represent biogeochemical cycles in crop and grassland systems appears an essential step for future research.

Scenario analysis of fertilizer management practices for N2O mitigation from corn systems in Canada.

Effective management of nitrogen (N) fertilizer application by farmers provides great potential for reducing emissions of the potent greenhouse gas nitrous oxide (N2O). However, such potential is rarely achieved because our understanding of what practices (or combination of practices) lead to N2O reductions without compromising crop yields remains far from complete. Using scenario analysis with the process-based model DNDC, this study explored the effects of nine fertilizer practices on N2O emissions and crop yields from two corn production systems in Canada. The scenarios differed in: timing of fertilizer application, fertilizer rate, number of applications, fertilizer type, method of application and use of nitrification/urease inhibitors. Statistical analysis showed that during the initial calibration and validation stages the simulated results had no significant total error or bias compared to measured values, yet grain yield estimations warrant further model improvement. Sidedress fertilizer applications reduced yield-scaled N2O emissions by c. 60% compared to fall fertilization. Nitrification inhibitors further reduced yield-scaled N2O emissions by c. 10%; urease inhibitors had no effect on either N2O emissions or crop productivity. The combined adoption of split fertilizer application with inhibitors at a rate 10% lower than the conventional application rate (i.e. 150kgNha-1) was successful, but the benefits were lower than those achieved with single fertilization at sidedress. Our study provides a comprehensive assessment of fertilizer management practices that enables policy development regarding N2O mitigation from agricultural soils in Canada.

Impact of management strategies on the global warming potential at the cropping system level.

Estimating the greenhouse gas (GHG) emissions from agricultural systems is important in order to assess the impact of agriculture on climate change. In this study experimental data supplemented with results from a biophysical model (DNDC) were combined with life cycle assessment (LCA) to investigate the impact of management strategies on global warming potential of long-term cropping systems at two locations (Breton and Ellerslie) in Alberta, Canada. The aim was to estimate the difference in global warming potential (GWP) of cropping systems due to N fertilizer reduction and residue removal. Reducing the nitrogen fertilizer rate from 75 to 50 kg N ha(-1) decreased on average the emissions of N2O by 39%, NO by 59% and ammonia volatilisation by 57%. No clear trend for soil CO2 emissions was determined among cropping systems. When evaluated on a per hectare basis, cropping systems with residue removal required 6% more energy and had a little change in GWP. Conversely, when evaluated on the basis of gigajoules of harvestable biomass, residue removal resulted in 28% less energy requirement and 33% lower GWP. Reducing nitrogen fertilizer rate resulted in 18% less GWP on average for both functional units at Breton and 39% less GWP at Ellerslie. Nitrous oxide emissions contributed on average 67% to the overall GWP per ha. This study demonstrated that small changes in N fertilizer have a minimal impact on the productivity of the cropping systems but can still have a substantial environmental impact.

A Greenhouse Gas and Soil Carbon Model for Estimating the Carbon Footprint of Livestock Production in Canada.

To assess tradeoffs between environmental sustainability and changes in food production on agricultural land in Canada the Unified Livestock Industry and Crop Emissions Estimation System (ULICEES) was developed. It incorporates four livestock specific GHG assessments in a single model. To demonstrate the application of ULICEES, 10% of beef cattle protein production was assumed to be displaced with an equivalent amount of pork protein. Without accounting for the loss of soil carbon, this 10% shift reduced GHG emissions by 2.5 TgCO2e y(-1). The payback period was defined as the number of years required for a GHG reduction to equal soil carbon lost from the associated land use shift. A payback period that is shorter than 40 years represents a net long term decrease in GHG emissions. Displacing beef cattle with hogs resulted in a surplus area of forage. When this residual land was left in ungrazed perennial forage, the payback periods were less than 4 years and when it was reseeded to annual crops, they were equal to or less than 40 years. They were generally greater than 40 years when this land was used to raise cattle. Agricultural GHG mitigation policies will inevitably involve a trade-off between production, land use and GHG emission reduction. ULICEES is a model that can objectively assess these trade-offs for Canadian agriculture.