Proposal and extensive test of a calibration protocol for crop phenology models

Wallach Daniel, Palosuo Taru, Thorburn Peter J., Mielenz Henrike, Buis Samuel, Hochman Zvi, Gourdain Emmanuelle, Andrianasolo Fety, Dumont Benjamin, Ferrise Roberto, Gaiser Thomas, Garcia Cécile, Gayler Sebastian, Harrison Matthew, Hiremath Santosh, Horan Heidi, Hoogenboom Gerrit, Jansson Per-Erik, Jing Qi, Justes Eric, Kersebaum Kurt Christian, Launay Marie, Lewan Elisabet, Liu Ke, Mequanint Fasil, Moriondo Marco, Nendel Claas, Padovan Gloria, Qian Budong, Schütze Niels, Seserman Diana‑Maria, Shelia Vakhtang, Souissi Amir, Specka Xenia, Srivastava Amit Kumar, Trombi Giacomo, Weber Tobias K. D., Weihermüller Lutz, Wöhling Thomas, Seidel Sabine Julia. 2023. Proposal and extensive test of a calibration protocol for crop phenology models. Agronomy for Sustainable Development, 43:4, 14 p.

Résumé : 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%.

Mots-clés Agrovoc : modèle de simulation, modélisation des cultures, phénologie, modèle mathématique, méthode statistique, essai de variété, changement climatique, modélisation

Mots-clés géographiques Agrovoc : France, Australie

Mots-clés libres : Modelling, Methodology of calibration, Crop model, Prediction error, Protocol, Model ensemble, Variability

Classification Agris : F01 - Culture des plantes
F40 - Écologie végétale
U10 - Informatique, mathématiques et statistiques

Champ stratégique Cirad : CTS 2 (2019-) - Transitions agroécologiques

Agences de financement hors UE : Academy of Finland, Natural Resources Institute Finland, Bundesministerium für Bildung und Forschung, Deutsche Forschungsgemeinschaft, Ministerstvo Školství, Mládeže a Tělovýchovy, Agriculture and Agri-Food Canada, Ministero delle Politiche Agricole Alimentari e Forestali, Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement

Projets sur financement : (USA) Agricultural Model Intercomparison and Improvement Project, (DEU) Soil as a Sustainable Resource for the Bioeconomy, (CZE) Adaption strategies for sustainable ecosystem services and food security under adverse environmental conditions

Auteurs et affiliations

• Wallach Daniel, Universität Bonn (DEU)
• Palosuo Taru, Natural Resources Institute Finland (FIN)
• Thorburn Peter J., CSIRO (AUS)
• Mielenz Henrike, JKI (DEU)
• Buis Samuel, INRAE (FRA)
• Hochman Zvi, CSIRO (AUS)
• Gourdain Emmanuelle, ARVALIS Institut du végétal (FRA)
• Andrianasolo Fety, ARVALIS Institut du végétal (FRA)
• Dumont Benjamin, Université de Liège (BEL)
• Ferrise Roberto, University of Florence (ITA)
• Gaiser Thomas, Universität Bonn (DEU)
• Garcia Cécile, ARVALIS Institut du végétal (FRA)
• Gayler Sebastian, Universität Hohenheim (DEU)
• Harrison Matthew, University of Tasmania (AUS)
• Hiremath Santosh, Aalto University (FIN)
• Horan Heidi, CSIRO (AUS)
• Hoogenboom Gerrit, University of Florida (USA)
• Jansson Per-Erik, Royal Institute of Technology (SWE)
• Jing Qi, Ottawa Research and Development Center (CAN)
• Justes Eric, CIRAD-DG-Direction générale (FRA) ORCID: 0000-0001-7390-7058
• Kersebaum Kurt Christian, Leibniz Centre for Agricultural Landscape Research (DEU)
• Launay Marie, INRAE (FRA)
• Lewan Elisabet, Swedish University of Agricultural Sciences (SWE)
• Liu Ke, University of Tasmania (AUS)
• Mequanint Fasil, Universität Hohenheim (DEU)
• Moriondo Marco, CNR-IBE (ITA)
• Nendel Claas, Leibniz Centre for Agricultural Landscape Research (DEU)
• Padovan Gloria, University of Florence (ITA)
• Qian Budong, Agriculture and Agri-Food Canada (CAN)
• Schütze Niels, Technische Universität Dresden (DEU)
• Seserman Diana‑Maria, Leibniz Centre for Agricultural Landscape Research (DEU)
• Shelia Vakhtang, University of Florida (USA)
• Souissi Amir, Agriculture and Agri-Food Canada (CAN)
• Specka Xenia, Leibniz Centre for Agricultural Landscape Research (DEU)
• Srivastava Amit Kumar, Universität Bonn (DEU)
• Trombi Giacomo, University of Florence (ITA)
• Weber Tobias K. D., Universität Hohenheim (DEU)
• Weihermüller Lutz, Centre de recherche de Jülich (DEU)
• Wöhling Thomas, Technische Universität Dresden (DEU)
• Seidel Sabine Julia, Universität Bonn (DEU) - auteur correspondant

Source : Cirad-Agritrop (https://agritrop.cirad.fr/606064/)

Métriques

[ Page générée et mise en cache le 2024-12-18 ]