OPENAIRE OpenAire Datacite Zurich Open Repository and Archive Sygma Bielefeld Academic Search Engine (BASE) HAL UPEC HAL-INSU HAL-IRD DOAJ Crossref HAL INRAE HAL Sorbonne Université HAL - Université de Lille NIBIO Brage Agritrop HAL-CEA European Union Open Data Portal https://www.nature.com/articles/s43247-023-00830-5.pdf https://doi.org/11250/3082084 Communications Earth &amp; Environment Open Access 2023-05-08 2023-01-01 2023-05-10 Abstract<p>Numerical models are crucial to understand and/or predict past and future soil organic carbon dynamics. For those models aiming at prediction, validation is a critical step to gain confidence in projections. With a comprehensive review of ~250 models, we assess how models are validated depending on their objectives and features, discuss how validation of predictive models can be improved. We find a critical lack of independent validation using observed time series. Conducting such validations should be a priority to improve the model reliability. Approximately 60% of the models we analysed are not designed for predictions, but rather for conceptual understanding of soil processes. These models provide important insights by identifying key processes and alternative formalisms that can be relevant for predictive models. We argue that combining independent validation based on observed time series and improved information flow between predictive and conceptual models will increase reliability in predictions.</p P33 - Chimie et physique du sol [SDE] Environmental Sciences P40 - Météorologie et climatologie analyse de séries chronologiques 01 natural sciences 2300 General Environmental Science http://aims.fao.org/aos/agrovoc/c_3041 carbone organique du sol http://aims.fao.org/aos/agrovoc/c_389fe908 GE1-350 910 Geography & travel modélisation 0105 earth and related environmental sciences QE1-996.5 http://aims.fao.org/aos/agrovoc/c_230ab86c U10 - Informatique, mathématiques et statistiques 1900 General Earth and Planetary Sciences Geology 04 agricultural and veterinary sciences 15. Life on land technique de prévision http://aims.fao.org/aos/agrovoc/c_28778 Environmental sciences 10122 Institute of Geography [SDE]Environmental Sciences 0401 agriculture, forestry, and fisheries U30 - Méthodes de recherche Le Noë, Julia, Manzoni, Stefano, Abramoff, Rose, Bölscher, Tobias, Bruni, Elisa, Cardinael, Rémi, Ciais, Philippe, Chenu, Claire, Clivot, Hugues, Derrien, Delphine, Ferchaud, Fabien, Garnier, Patricia, Goll, Daniel, Lashermes, Gwenaëlle, Martin, Manuel, Rasse, Daniel, Rees, Frédéric, Sainte-Marie, Julien, Salmon, Elodie, Schiedung, Marcus, Schimel, Josh, Wieder, William, Abiven, Samuel, Barré, Pierre, Cécillon, Lauric, Guenet, Bertrand, 2023-05-08 journalpaper Abstract<p>Numerical models are crucial to understand and/or predict past and future soil organic carbon dynamics. For those models aiming at prediction, validation is a critical step to gain confidence in projections. With a comprehensive review of ~250 models, we assess how models are validated depending on their objectives and features, discuss how validation of predictive models can be improved. We find a critical lack of independent validation using observed time series. Conducting such validations should be a priority to improve the model reliability. Approximately 60% of the models we analysed are not designed for predictions, but rather for conceptual understanding of soil processes. These models provide important insights by identifying key processes and alternative formalisms that can be relevant for predictive models. We argue that combining independent validation based on observed time series and improved information flow between predictive and conceptual models will increase reliability in predictions.</p Soil organic carbon models need independent time-series validation for reliable prediction 11250/3082084 101000289