The error in agricultural systems model prediction depends on the variable being predicted

Wallach, Daniel; Thorburn, Peter J.

doi:10.1016/j.envsoft.2014.08.001

Cited by 13 publications

(5 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ensemble modelling provides insights into uncertainty Rosenzweig et al, 2013) as demonstrated by one paper in the Thematic Issue (Marin et al, 2014). However, there was only one paper (Wallach and Thorburn, 2014) dealing with statistical approaches to quantifying performance in agricultural production systems models, and none on statistical approaches to quantifying uncertainty. Uncertainty and sensitivity analyses are computationally intensive so the slow run-time and large number of parameters in many agricultural systems models makes uncertainty and sensitivity analyses challenging.…”

Section: Reflections From This Thematic Issuementioning

confidence: 97%

“…In this Thematic Issue, this increase in complexity is very evident and has arisen from a range of different sources. Huth et al (2014) and Raes et al (2014) discuss the need to develop or parameterise crops with little or minimal information while (Wallach and Thorburn, 2014) show an alternative approach to evaluate model performance. Archontoulis et al (2014) and consider the need to rapidly estimate values for phenological parameters in crop models.…”

Section: Reflections From This Thematic Issuementioning

confidence: 98%

See 1 more Smart Citation

Agricultural production systems modelling and software: Current status and future prospects

Holzworth

Snow

Janssen

et al. 2015

Environmental Modelling & Software

197

View full text Add to dashboard Cite

Section: Reflections From This Thematic Issuementioning

confidence: 97%

Section: Reflections From This Thematic Issuementioning

confidence: 98%

Agricultural production systems modelling and software: Current status and future prospects

Holzworth

Snow

Janssen

et al. 2015

Environmental Modelling & Software

197

View full text Add to dashboard Cite

“…This study provides the first assessment of process‐based simulation models used for simultaneous estimates of crop and pasture productivity and of N 2 O emissions in response to climate, soil and management conditions. The statistical approach of model error adopted in this study is based on predictions averaged over space (means of replicate measurements) and time (seasonal and annual means) (Wallach & Thorburn, ). Compared to Willmott, Robeson, and Matsuura (), where model performance metric (index of agreement, d r ) ranges from 0 to 1, our dimensionless indicator (Z m,i ) scales the model performance by considering the uncertainties in the measurements and allows for assessing model estimates on an observed SD basis.…”

Section: Discussionmentioning

confidence: 99%

Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N₂O emissions

Ehrhardt¹,

Soussana²,

Bellocchi³

et al. 2017

Global Change Biology

122

112

View full text Add to dashboard Cite

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 (N 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 N 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 N O emissions. At stage 1, ensembles formed by the three lowest prediction model errors predicted both yields and N 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 N O emissions. Yield-scaled N O emissions (N 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 N O emissions at field scale is discussed.

show abstract

“…We recognize there are a number of major limitations that need to be addressed before N recommendation tools can be developed and employed with an acceptable level of confidence ( Asseng et al, 2013 ; Wallach and Thorburn, 2014 ). Current models may lack accurate mechanistic processes and model parameters for predicting environmental N losses and crop yields ( Brilli et al, 2017 ).…”

Section: Challengesmentioning

confidence: 99%

A Vision for Incorporating Environmental Effects into Nitrogen Management Decision Support Tools for U.S. Maize Production

et al. 2017

View full text Add to dashboard Cite

Meeting crop nitrogen (N) demand while minimizing N losses to the environment has proven difficult despite significant field research and modeling efforts. To improve N management, several real-time N management tools have been developed with a primary focus on enhancing crop production. However, no coordinated effort exists to simultaneously address sustainability concerns related to N losses at field- and regional-scales. In this perspective, we highlight the opportunity for incorporating environmental effects into N management decision support tools for United States maize production systems by integrating publicly available crop models with grower-entered management information and gridded soil and climate data in a geospatial framework specifically designed to quantify environmental and crop production tradeoffs. To facilitate advances in this area, we assess the capability of existing crop models to provide in-season N recommendations while estimating N leaching and nitrous oxide emissions, discuss several considerations for initial framework development, and highlight important challenges related to improving the accuracy of crop model predictions. Such a framework would benefit the development of regional sustainable intensification strategies by enabling the identification of N loss hotspots which could be used to implement spatially explicit mitigation efforts in relation to current environmental quality goals and real-time weather conditions. Nevertheless, we argue that this long-term vision can only be realized by leveraging a variety of existing research efforts to overcome challenges related to improving model structure, accessing field data to enhance model performance, and addressing the numerous social difficulties in delivery and adoption of such tool by stakeholders.

show abstract

The error in agricultural systems model prediction depends on the variable being predicted

Cited by 13 publications

References 36 publications

Agricultural production systems modelling and software: Current status and future prospects

Agricultural production systems modelling and software: Current status and future prospects

Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N₂O emissions

A Vision for Incorporating Environmental Effects into Nitrogen Management Decision Support Tools for U.S. Maize Production

Contact Info

Product

Resources

About

The error in agricultural systems model prediction depends on the variable being predicted

Cited by 13 publications

References 36 publications

Agricultural production systems modelling and software: Current status and future prospects

Agricultural production systems modelling and software: Current status and future prospects

Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N2O emissions

A Vision for Incorporating Environmental Effects into Nitrogen Management Decision Support Tools for U.S. Maize Production

Contact Info

Product

Resources

About

Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N₂O emissions