It has become common to compare crop model results in multi-model simulation experiments. In general, one observes a large variability in such studies, which reduces the confidence one can have in such models. It is important to understand the causes of this variability as a first step toward reducing it. For a given data set, the variability in a multi-model study can arise from uncertainty in model structure or in parameter values for a given structure. Previous studies have made assumptions about the origin of parameter uncertainty, and then quantified its contribution, generally finding that parameter uncertainty is less important than structure uncertainty. However, those studies do not take account of the full parameter variability in multi-model studies. Here we propose estimating parameter uncertainty based on open-call multi-model ensembles where the same structure is used by more than one modeling group. The variability in such a case is due to the full variability of parameters among modeling groups. Then structure and parameter contributions can be estimated using random effects analysis of variance. Based on three multi-model studies for simulating wheat phenology, it is found that the contribution of parameter uncertainty to total uncertainty is, on average, more than twice as large as the uncertainty from structure. A second estimate, based on a comparison of two different calibration approaches for multiple models leads to a very similar result. We conclude that improvement of crop models requires as much attention to parameters as to model structure.

The joint project 'Measures to reduce direct and indirect climate-impacting emissions caused by denitrification in agricultural soils - MinDen' addresses the topics of reducing nitrous oxide emissions and improving nitrogen efficiency through modeling, the evaluation of possible mitigation measures and the evaluation of denitrification on spatial scale. Gaseous emissions from denitrification cause N losses relevant to crop cultivation and cause direct N2O emissions from crop cultivation. Climate protection measures in crop production in the areas of fertilization, soil cultivation and crop rotation have hardly been researched with regard to the role of denitrification. Crop management that optimizes N efficiency and minimizes N emissions at the same time has therefore not yet been reliably defined. The overall objective of the present project is to identify practicable crop management measures to minimize N2 and N2O emissions from denitrification for arable cropping systems in Germany by improving the knowledge on denitrification-related N losses through field and laboratory studies and using it for parameterization, validation and application of simulation models. Our objectives are as follows: Regionalization of N losses due to denitrification in Germany based on existing models Determination of the effect of crop protection measures on N2 and N2O losses on field scale Testing of mitigation options on the model, laboratory and field scale, taking into account the topsoil and subsoil for different soils Further development of denitrification models to improve the mapping of mitigation measures using existing and new field data Testing of mitigation options for Germany using the improved models, taking into account yield, economic efficiency, technology requirements, N2O emissions, N efficiency, fertilizer requirements, NH3 emissions and nitrate leaching. We provide an overview of the approach and the current status of the joint project, which started at the beginning of 2023.

Process-based soil-crop models are widely used in agronomic research. They are major tools for evaluating climate change impact on crop production. Multi-model simulation studies show a wide diversity of results among models, implying that simulation results are very uncertain. A major path to improving simulation results is to propose improved calibration practices that are widely applicable. This study proposes an innovative generic calibration protocol. The two major innovations concern the treatment of multiple output variables and the choice of parameters to estimate, both of which are based on standard statistical procedure adapted to the particularities of soil-crop models. The protocol performed well in a challenging artificial-data test. The protocol is formulated so as to be applicable to a wide range of models and data sets. If widely adopted, it could substantially reduce model error and inter-model variability, and thus increase confidence in soil-crop model simulations.

Diffuse nitrogen losses from agriculture in Germany continue to cause regionally increased nitrate concentrations in groundwater. Groundwater quality monitoring cannot be a timely indicator of the effects of mitigation measures being applied in agriculture, due to frequently long transport routes and high residence times of the leachate. Instead, nitrate leaching potential is often determined at field and farm scale by monitoring soil mineral nitrogen contents at 0-90 cm depth in autumn (SMNa), i.e. before the start of the annual leachate period. In this study, we developed an understanding of the controls on the soil mineral nitrogen content at the start of winter. In an on-farm approach, extensive data was collected from 48 farms in five nitrate-sensitive regions in Germany from 2017 to 2020. From this data set, 25 management and site factors were evaluated with regard to their significance for SMNa by means of a random forest model. With the random forest regression, we identified the role of the factors on SMNa with an acceptable model accuracy with R2 = 0.56. The results show that the cultivated crop is the most important factor influencing SMNa. Potatoes, oilseed rape and maize produced the highest SMNas, whereas SMNas were lowest after spring barley, sugar beet and winter barley. Among site factors, soil type and texture as well as precipitation in October were most decisive. The effects of N fertilisation parameters such as rate and timing were masked by these site factors. The results show that the reduction of nitrogen-intensive crops in crop sequences can be a promising measure for the reduction of nitrate loads. On the other hand, our analysis makes clear that soil-related factors controlling nitrogen release and risk of leaching, as well as weather, can significantly mask the effect of cultivation.

A major effect of environment on crops is through crop phenology, and therefore, the capacity to predict phenology for new environments is important. Mechanistic crop models are a major tool for such predictions, but calibration of crop phenology models is difficult and there is no consensus on the best approach. We propose an original, detailed approach for calibration of such models, which we refer to as a calibration protocol. The protocol covers all the steps in the calibration workflow, namely choice of default parameter values, choice of objective function, choice of parameters to estimate from the data, calculation of optimal parameter values, and diagnostics. The major innovation is in the choice of which parameters to estimate from the data, which combines expert knowledge and data-based model selection. First, almost additive parameters are identified and estimated. This should make bias (average difference between observed and simulated values) nearly zero. These are “obligatory” parameters, that will definitely be estimated. Then candidate parameters are identified, which are parameters likely to explain the remaining discrepancies between simulated and observed values. A candidate is only added to the list of parameters to estimate if it leads to a reduction in BIC (Bayesian Information Criterion), which is a model selection criterion. A second original aspect of the protocol is the specification of documentation for each stage of the protocol. The protocol was applied by 19 modeling teams to three data sets for wheat phenology. All teams first calibrated their model using their “usual” calibration approach, so it was possible to compare usual and protocol calibration. Evaluation of prediction error was based on data from sites and years not represented in the training data. Compared to usual calibration, calibration following the new protocol reduced the variability between modeling teams by 22% and reduced prediction error by 11%.

Predicting phenology is essential for adapting varieties to different environmental conditions and for crop management. Therefore, it is important to evaluate how well different crop modeling groups can predict phenology. Multiple evaluation studies have been previously published, but it is still difficult to generalize the findings from such studies since they often test some specific aspect of extrapolation to new conditions, or do not test on data that is truly independent of the data used for calibration. In this study, we analyzed the prediction of wheat phenology in Northern France under observed weather and current management, which is a problem of practical importance for wheat management. The results of 27 modeling groups are evaluated, where modeling group encompasses model structure, i.e. the model equations, the calibration method and the values of those parameters not affected by calibration. The data for calibration and evaluation are sampled from the same target population, thus extrapolation is limited. The calibration and evaluation data have neither year nor site in common, to guarantee rigorous evaluation of prediction for new weather and sites. The best modeling groups, and also the mean and median of the simulations, have a mean absolute error (MAE) of about 3 days, which is comparable to the measurement error. Almost all models do better than using average number of days or average sum of degree days to predict phenology. On the other hand, there are important differences between modeling groups, due to model structural differences and to differences between groups using the same model structure, which emphasizes that model structure alone does not completely determine prediction accuracy. In addition to providing information for our specific environments and varieties, these results are a useful contribution to a knowledge base of how well modeling groups can predict phenology, when provided with calibration data from the target population.

Calibration, the estimation of model parameters based on fitting the model to experimental data, is among the first steps in many applications of process-based models and has an important impact on simulated values. We propose a novel method of developing guidelines for calibration of process-based models, based on development of recommendations for calibration of the phenology component of crop models. The approach was based on a multi-model study, where all teams were provided with the same data and asked to return simulations for the same conditions. All teams were asked to document in detail their calibration approach, including choices with respect to criteria for best parameters, choice of parameters to estimate and software. Based on an analysis of the advantages and disadvantages of the various choices, we propose calibration recommendations that cover a comprehensive list of decisions and that are based on actual practices.

Calibration, the estimation of model parameters based on fitting the model to experimental data, is among the first steps in many applications of system models and has an important impact on simulated values. Here we propose and illustrate a novel method of developing guidelines for calibration of system models. Our example is calibration of the phenology component of crop models. The approach is based on a multi-model study, where all teams are provided with the same data and asked to return simulations for the same conditions. All teams are asked to document in detail their calibration approach, including choices with respect to criteria for best parameters, choice of parameters to estimate and software. Based on an analysis of the advantages and disadvantages of the various choices, we propose calibration recommendations that cover a comprehensive list of decisions and that are based on actual practices. Highlights We propose a new approach to deriving calibration recommendations for system models Approach is based on analyzing calibration in multi-model simulation exercises Resulting recommendations are holistic and anchored in actual practice We apply the approach to calibration of crop models used to simulate phenology Recommendations concern: objective function, parameters to estimate, software used

Few parametric expressions for the soil water retention curve are suitable for dry conditions. Furthermore, expressions for the soil hydraulic conductivity curves associated with parametric retention functions can behave unrealistically near saturation. We developed a general criterion for water retention parameterizations that ensures physically plausible conductivity curves. Only 3 of the 18 tested parameterizations met this criterion without restrictions on the parameters of a popular conductivity curve parameterization. A fourth required one parameter to be fixed. We estimated parameters by shuffled complex evolution (SCE) with the objective function tailored to various observation methods used to obtain retention curve data. We fitted the four parameterizations with physically plausible conductivities as well as the most widely used parameterization. The performance of the resulting 12 combinations of retention and conductivity curves was assessed in a numerical study with 751 days of semiarid atmospheric forcing applied to unvegetated, uniform, 1 m freely draining columns for four textures. Choosing different parameterizations had a minor effect on evaporation, but cumulative bottom fluxes varied by up to an order of magnitude between them. This highlights the need for a careful selection of the soil hydraulic parameterization that ideally does not only rely on goodness of fit to static soil water retention data but also on hydraulic conductivity measurements. Parameter fits for 21 soils showed that extrapolations into the dry range of the retention curve often became physically more realistic when the parameterization had a logarithmic dry branch, particularly in fine-textured soils where high residual water contents would otherwise be fitted.

Converting pasture to cropping is common in many of the world's agricultural systems. This conversion results in substantial net mineralisation of soil organic matter that builds up during a phase of pasture. A few studies have shown that this mineralisation leads to increased nitrous oxide (N2O) emissions compared to long-term pasture or long-term cropping. Understanding of interactions leading to these significant emissions is still scarce but is needed to identify mitigation options for this situation. In this study, the Agricultural Production Systems sIMulator (APSIM) was used to investigate the optimal timing of pasture termination (relative to crop planting) and management of nitrogen (N) in crops after pasture termination to maximise crop yield and limit N2O emissions. Beforehand, APSIM's performance in simulating yields and N2O emissions was tested against data from field experiments conducted in the temperate high-rainfall zone of southern Australia where N2O emissions were monitored with automatic gas collection chambers during the first year of cropping wheat after terminating long-term pasture on two adjacent sites in two consecutive years. Field experiments and simulation scenarios showed very high N2O emissions (up to 48 kg N2O-N ha(-1) yr(-1)) in the first year of wheat after pasture termination, even without N fertiliser application. Measured cumulative N2O emissions, crop yields and soil mineral N and water content dynamics were simulated well with APSIM. Including a routine into APSIM to account for N2O transport through the soil profile improved the simulation of daily N2O emissions considerably, leading to up to 67% of the measured variability in daily N2O emissions being explained by the model. We predicted that a short fallow between termination of pasture and sowing wheat, instead of a long fallow which is the common practice, reduces N(2)Oemissions by more than half in the first year of cropping without a noteworthy impact on crop yield. Reducing N fertiliser applications in the first few years after pasture termination by taking available soil mineral N into account, and applying the, fertiliser six to twelve weeks after sowing instead of at sowing was predicted to further reduce N2O emissions. Since the model was calibrated against experimental data during the first year after pasture termination only, experiments determining N2O emissions in the first two or three years after terminating pasture are needed to confirm our predictions. (C) 2016 Elsevier B.V. All rights reserved.

Nitrous oxide (N2O) emissions from Australian grain cropping systems are highly variable due to the large variations in soil and climate conditions and management practices under which crops are grown. Agricultural soils contribute 55% of national N2O emissions, and therefore mitigation of these emissions is important. In the present study, we explored N2O emissions, yield and emissions intensity in a range of management practices in grain crops across eastern Australia with the Agricultural Production Systems sIMulator (APSIM). The model was initially evaluated against experiments conducted at six field sites across major grain-growing regions in eastern Australia. Measured yields for all crops used in the experiments (wheat, barley, sorghum, maize, cotton, canola and chickpea) and seasonal N2O emissions were satisfactorily predicted with R-2 = 0.93 and R-2 = 0.91 respectively. As expected, N2O emissions and emissions intensity increased with increasing nitrogen (N) fertiliser input, whereas crop yields increased until a yield plateau was reached at a site-and crop-specific N rate. The mitigation potential of splitting N fertiliser application depended on the climate conditions and was found to be relevant only in the southern grain-growing region, where most rainfall occurs during the cropping season. Growing grain legumes in rotation with cereal crops has great potential to reduce mineral N fertiliser requirements and so N2O emissions. In general, N management strategies that maximise yields and increase N use efficiency showed the greatest promise for N2O mitigation.

23% of global agricultural land are situated in the subtropics. Nitrous oxide (N2O) emissions were estimated to be higher under subtropical than under temperate climates. So mitigation of N2O emissions from subtropical farming systems can make an important contribution to reducing global warming. Accordingly, in this study we explored long-term N2O emissions and possible mitigation options for representative subtropical cropping systems (e.g., summer versus winter crops, inclusion of a legume in the rotation) and management practices (nitrogen fertilizer, irrigation) by calculating scenarios with the agricultural systems model APSIM. The model was tested against high temporal frequency data from experiments conducted on an oxisol and a vertisol in subtropical Australia, which comprised a number of fertilization and irrigation treatments. The threshold of water filled pore space above which denitrification starts was calibrated on a subset of the data while the rest of the large number of parameters controlling the carbon and nitrogen cycles were kept to default values. The validity of the model was confirmed with 11 validation data sets for yields of four different crops (R-2=0.92) and 16 validation data sets for seasonal N2O emissions during crop and fallow periods (R-2 = 0.77). While these results show that the model performs well in sub-tropical environments, this modeling skill might not translate to other environments and the model would benefit from wider testing. In the scenario analyses, long-term average N2O emissions from wheat, cotton, maize and sorghum were predicted to vary between 0.2 and 6.1 kg N ha(-1) yr(-1) and showed large interannual variability of N2O emissions. This highlights the risk that results from short-term experiments may not be representative for the long-term behavior of these agro-ecosystems, and thus the value simulation studies add to experiments. The scenario analysis revealed that long-term average yields and N2O emissions increased in response to the same management practices (e.g., increase in nitrogen rate), leading to a trade-off between maximizing yield and minimizing N2O emissions. When crop yields were limited due to water stress either by low seasonal rainfall or by lack of irrigation, average N2O emissions increased. Given the annual variability in climate and soil nitrogen stocks, mitigating N2O emissions without compromizing in yield is not a simple task but requires an optimal nitrogen management considering other limiting factors such as water supply. (C) 2015 Elsevier B.V. All rights reserved.

Agricultural crop and livestock production contribute 78 % to global anthropogenic nitrous oxide (N 2 O) emissions. Therefore, mitigation of N 2 O emissions from agricultural systems is important. In this study, the trade-off between crop yield and N 2 O emissions in subtropical cereal and fibre cropping systems under various management practices was investigated to identify potential mitigation options. For this, the APSIM model was first tested against data from a field experiment on a vertosol comprising several irrigation treatments in subtropical Queensland. One soil hydraulic parameter and one parameter in the denitrification sub-model were calibrated against a subset of the data. The validity of the model was confirmed with the remaining data. For the six datasets we found a good correlation between measured and predicted yields (R 2 = 0.90) and seasonal N 2 O emissions (R 2 = 0.49). Long-term scenarios were then calculated with the validated model. Long-term average yield and N 2 O emissions both increased under increased nitrogen (N) supply from legumes or extra N fertiliser, and so, there was a trade-off between maximising yield and minimising N 2 O emissions. N 2 O emissions also increased when crop yields were limited due to water stress, because of increased mineral N availability. Given the annual variability in climate and soil nitrogen availability, mitigating N 2 O emissions is not a simple task. A tool for proper yield forecasting could therefore be of great benefit for estimating the amount of nitrogen required and thus assist in managing N efficiently and in mitigating N 2 O emissions.

Sensitivity analysis is an important step in understanding how uncertainty is propagated through complex computer models. Unfortunately, the most reliable sensitivity analysis techniques take a significant amount of time to execute due to the large number of computer model evaluations required. Emulators can be used to speed up the process by replacing the computer model with a statistical model that mimics the computer model and is computationally efficient. In this manuscript we propose two emulator-based sensitivity index estimators that require minimal set-up and are computationally inexpensive to compute. We demonstrate their accuracy with computer models that have known sensitivity index values and illustrate their application in practice with the agriculture systems simulator APSIM.

Nitrous oxide (N2O) is a potent greenhouse gas with a contribution of approximately 6% to the global anthropogenic greenhouse gas effect. The majority of anthropogenic N2O emissions stem from agricultural production systems, which highlights the necessity for mitigation strategies. Modelling analyses of experimental information increase our understanding of processes and factors driving N2O production and emission, which is indispensable for developing mitigation strategies. Reliable and robust models are needed for long-term simulations to investigate mitigation strategies. In this study, the Agricultural Production Systems Simulator (APSIM) was calibrated and validated for the prediction of N2O emissions with experimental data from different field sites, crops and treatments. We used numerical optimisation to estimate soil moisture parameters for each site to obtain good predictions of measured soil water dynamics and yields, and changed the critical water content at which denitrification starts to accurately predict large measured emission peaks of N2O. All other values used in the models were default values supplied in the general release of APSIM. Following the model set-up, the water dynamics, yields and cumulative N2O emissions could be well predicted for various sites, treatments, crops and seasons. Predicted denitrification was very sensitive to the water content where denitrification starts. This had to be adjusted for each site, which suggests its possible dependency on soil type. The robust prediction of the above mentioned target variables without having to adjust multiple nitrogen parameters demonstrates the capability of APSIM to reliably simulate N2O emissions under a range of environmental conditions.