Soil organic carbon models need independent time-series validation for reliable prediction
Abstract 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 depe...
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-05-01
|
| Series: | Communications Earth & Environment |
| Online Access: | https://doi.org/10.1038/s43247-023-00830-5 |
| _version_ | 1797827323253227520 |
|---|---|
| author | Julia Le Noë Stefano Manzoni Rose Abramoff Tobias Bölscher Elisa Bruni Rémi Cardinael Philippe Ciais Claire Chenu Hugues Clivot Delphine Derrien Fabien Ferchaud Patricia Garnier Daniel Goll Gwenaëlle Lashermes Manuel Martin Daniel Rasse Frédéric Rees Julien Sainte-Marie Elodie Salmon Marcus Schiedung Josh Schimel William Wieder Samuel Abiven Pierre Barré Lauric Cécillon Bertrand Guenet |
| author_facet | Julia Le Noë Stefano Manzoni Rose Abramoff Tobias Bölscher Elisa Bruni Rémi Cardinael Philippe Ciais Claire Chenu Hugues Clivot Delphine Derrien Fabien Ferchaud Patricia Garnier Daniel Goll Gwenaëlle Lashermes Manuel Martin Daniel Rasse Frédéric Rees Julien Sainte-Marie Elodie Salmon Marcus Schiedung Josh Schimel William Wieder Samuel Abiven Pierre Barré Lauric Cécillon Bertrand Guenet |
| author_sort | Julia Le Noë |
| collection | DOAJ |
| description | Abstract 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. |
| first_indexed | 2024-04-09T12:46:05Z |
| format | Article |
| id | doaj.art-4dcab8d04662477892c658cb8612071c |
| institution | Directory Open Access Journal |
| issn | 2662-4435 |
| language | English |
| last_indexed | 2024-04-09T12:46:05Z |
| publishDate | 2023-05-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Communications Earth & Environment |
| spelling | doaj.art-4dcab8d04662477892c658cb8612071c2023-05-14T11:28:56ZengNature PortfolioCommunications Earth & Environment2662-44352023-05-01411810.1038/s43247-023-00830-5Soil organic carbon models need independent time-series validation for reliable predictionJulia Le Noë0Stefano Manzoni1Rose Abramoff2Tobias Bölscher3Elisa Bruni4Rémi Cardinael5Philippe Ciais6Claire Chenu7Hugues Clivot8Delphine Derrien9Fabien Ferchaud10Patricia Garnier11Daniel Goll12Gwenaëlle Lashermes13Manuel Martin14Daniel Rasse15Frédéric Rees16Julien Sainte-Marie17Elodie Salmon18Marcus Schiedung19Josh Schimel20William Wieder21Samuel Abiven22Pierre Barré23Lauric Cécillon24Bertrand Guenet25Laboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLDepartment of Physical Geography, Stockholm UniversityLaboratoire des Sciences du Climat et de l’Environnement, CEA-CNRS-UVSQUniversity of Paris-Saclay, INRAE, AgroParisTech, UMR EcoSysLaboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLAIDA, Univ Montpellier, CIRADLaboratoire des Sciences du Climat et de l’Environnement, CEA-CNRS-UVSQUniversity of Paris-Saclay, INRAE, AgroParisTech, UMR EcoSysUniversité de Reims Champagne-Ardenne, INRAE, FARE, UMR A 614INRAE, Biogéochimie des Ecosystèmes ForestiersBioEcoAgro Joint Research Unit, INRAE, Université de Liège, Université de Lille, Université de Picardie Jules VerneUniversity of Paris-Saclay, INRAE, AgroParisTech, UMR EcoSysLaboratoire des Sciences du Climat et de l’Environnement, CEA-CNRS-UVSQUniversité de Reims Champagne-Ardenne, INRAE, FARE, UMR A 614INRAE, InfoSolNorwegian Institute of Bioeconomy Research (NIBIO), Division of Environment and Natural ResourcesUniversity of Paris-Saclay, INRAE, AgroParisTech, UMR EcoSysUniversité de Lorraine, AgroParisTech, INRAE, SILVALaboratoire des Sciences du Climat et de l’Environnement, CEA-CNRS-UVSQDepartment of Geography, University of ZurichDepartment of Ecology, Evolution and Marine Biology, University of California Santa BarbaraClimate and Global Dynamics Laboratory, National Center for Atmospheric ResearchLaboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLLaboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLLaboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLLaboratoire de Géologie, École normale supérieure, CNRS, PSL Univ., IPSLAbstract 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.https://doi.org/10.1038/s43247-023-00830-5 |
| spellingShingle | Julia Le Noë Stefano Manzoni Rose Abramoff Tobias Bölscher Elisa Bruni Rémi Cardinael Philippe Ciais Claire Chenu Hugues Clivot Delphine Derrien Fabien Ferchaud Patricia Garnier Daniel Goll Gwenaëlle Lashermes Manuel Martin Daniel Rasse Frédéric Rees Julien Sainte-Marie Elodie Salmon Marcus Schiedung Josh Schimel William Wieder Samuel Abiven Pierre Barré Lauric Cécillon Bertrand Guenet Soil organic carbon models need independent time-series validation for reliable prediction Communications Earth & Environment |
| title | Soil organic carbon models need independent time-series validation for reliable prediction |
| title_full | Soil organic carbon models need independent time-series validation for reliable prediction |
| title_fullStr | Soil organic carbon models need independent time-series validation for reliable prediction |
| title_full_unstemmed | Soil organic carbon models need independent time-series validation for reliable prediction |
| title_short | Soil organic carbon models need independent time-series validation for reliable prediction |
| title_sort | soil organic carbon models need independent time series validation for reliable prediction |
| url | https://doi.org/10.1038/s43247-023-00830-5 |
| work_keys_str_mv | AT julialenoe soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT stefanomanzoni soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT roseabramoff soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT tobiasbolscher soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT elisabruni soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT remicardinael soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT philippeciais soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT clairechenu soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT huguesclivot soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT delphinederrien soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT fabienferchaud soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT patriciagarnier soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT danielgoll soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT gwenaellelashermes soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT manuelmartin soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT danielrasse soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT fredericrees soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT juliensaintemarie soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT elodiesalmon soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT marcusschiedung soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT joshschimel soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT williamwieder soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT samuelabiven soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT pierrebarre soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT lauriccecillon soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction AT bertrandguenet soilorganiccarbonmodelsneedindependenttimeseriesvalidationforreliableprediction |