Multimodel ensembles improve predictions of crop–environment–management interactions
A recent innovation in assessment of climate change impact on agricultural production has been to use crop multimodel ensembles (MMEs). These studies usually find large variability between individual models but that the ensemble mean (e‐mean) and median (e‐median) often seem to predict quite well. H...
| Autores principales: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Formato: | Journal Article |
| Lenguaje: | Inglés |
| Publicado: |
Wiley
2018
|
| Materias: | |
| Acceso en línea: | https://hdl.handle.net/10568/97157 |
| _version_ | 1855514493455433728 |
|---|---|
| author | Wallach, Daniel Martre, Pierre Liu, Bing Asseng, Senthold Ewert, Frank Thorburn, Peter J. Ittersum, Martin K. van Aggarwal, Pramod K. Ahmed, Mukhtar Basso, Bruno Biernath, Christian Cammarano, Davide Challinor, Andrew J. Sanctis, Giacomo de Dumont, Benjamin Eyshi Rezaei, Ehsan Fereres, Elias Fitzgerald, Glenn J. Gao, Y García Vila, Margarita Gayler, Sebastian Girousse, Christine Hoogenboom, Gerrit Horan, Heidi Izaurralde, Roberto César Jones, Curtis D. Kassie, Belay T. Kersebaum, Kurt-Christian Klein, Christian Köhler, Ann-Kristin Maiorano, Andrea Minoli, Sara Müller, Christoph Naresh Kumar, Soora Nendel, Claas O’Leary, Garry J. Palosuo, Taru Priesack, Eckart Ripoche, Dominique Rötter, Reimund P. Semenov, Mikhail A. Stöckle, Claudio O. Stratonovitch, Pierre Streck, Thilo Supit, Iwan Tao, Fulu Wolf, Joost Zhang, Zhao |
| author_browse | Aggarwal, Pramod K. Ahmed, Mukhtar Asseng, Senthold Basso, Bruno Biernath, Christian Cammarano, Davide Challinor, Andrew J. Dumont, Benjamin Ewert, Frank Eyshi Rezaei, Ehsan Fereres, Elias Fitzgerald, Glenn J. Gao, Y García Vila, Margarita Gayler, Sebastian Girousse, Christine Hoogenboom, Gerrit Horan, Heidi Ittersum, Martin K. van Izaurralde, Roberto César Jones, Curtis D. Kassie, Belay T. Kersebaum, Kurt-Christian Klein, Christian Köhler, Ann-Kristin Liu, Bing Maiorano, Andrea Martre, Pierre Minoli, Sara Müller, Christoph Naresh Kumar, Soora Nendel, Claas O’Leary, Garry J. Palosuo, Taru Priesack, Eckart Ripoche, Dominique Rötter, Reimund P. Sanctis, Giacomo de Semenov, Mikhail A. Stratonovitch, Pierre Streck, Thilo Stöckle, Claudio O. Supit, Iwan Tao, Fulu Thorburn, Peter J. Wallach, Daniel Wolf, Joost Zhang, Zhao |
| author_facet | Wallach, Daniel Martre, Pierre Liu, Bing Asseng, Senthold Ewert, Frank Thorburn, Peter J. Ittersum, Martin K. van Aggarwal, Pramod K. Ahmed, Mukhtar Basso, Bruno Biernath, Christian Cammarano, Davide Challinor, Andrew J. Sanctis, Giacomo de Dumont, Benjamin Eyshi Rezaei, Ehsan Fereres, Elias Fitzgerald, Glenn J. Gao, Y García Vila, Margarita Gayler, Sebastian Girousse, Christine Hoogenboom, Gerrit Horan, Heidi Izaurralde, Roberto César Jones, Curtis D. Kassie, Belay T. Kersebaum, Kurt-Christian Klein, Christian Köhler, Ann-Kristin Maiorano, Andrea Minoli, Sara Müller, Christoph Naresh Kumar, Soora Nendel, Claas O’Leary, Garry J. Palosuo, Taru Priesack, Eckart Ripoche, Dominique Rötter, Reimund P. Semenov, Mikhail A. Stöckle, Claudio O. Stratonovitch, Pierre Streck, Thilo Supit, Iwan Tao, Fulu Wolf, Joost Zhang, Zhao |
| author_sort | Wallach, Daniel |
| collection | Repository of Agricultural Research Outputs (CGSpace) |
| description | A recent innovation in assessment of climate change impact on agricultural production has been to use crop multimodel ensembles (MMEs). These studies usually find large variability between individual models but that the ensemble mean (e‐mean) and median (e‐median) often seem to predict quite well. However, few studies have specifically been concerned with the predictive quality of those ensemble predictors. We ask what is the predictive quality of e‐mean and e‐median, and how does that depend on the ensemble characteristics. Our empirical results are based on five MME studies applied to wheat, using different data sets but the same 25 crop models. We show that the ensemble predictors have quite high skill and are better than most and sometimes all individual models for most groups of environments and most response variables. Mean squared error of e‐mean decreases monotonically with the size of the ensemble if models are added at random, but has a minimum at usually 2–6 models if best‐fit models are added first. Our theoretical results describe the ensemble using four parameters: average bias, model effect variance, environment effect variance, and interaction variance. We show analytically that mean squared error of prediction (MSEP) of e‐mean will always be smaller than MSEP averaged over models and will be less than MSEP of the best model if squared bias is less than the interaction variance. If models are added to the ensemble at random, MSEP of e‐mean will decrease as the inverse of ensemble size, with a minimum equal to squared bias plus interaction variance. This minimum value is not necessarily small, and so it is important to evaluate the predictive quality of e‐mean for each target population of environments. These results provide new information on the advantages of ensemble predictors, but also show their limitations. |
| format | Journal Article |
| id | CGSpace97157 |
| institution | CGIAR Consortium |
| language | Inglés |
| publishDate | 2018 |
| publishDateRange | 2018 |
| publishDateSort | 2018 |
| publisher | Wiley |
| publisherStr | Wiley |
| record_format | dspace |
| spelling | CGSpace971572025-03-13T09:43:58Z Multimodel ensembles improve predictions of crop–environment–management interactions Wallach, Daniel Martre, Pierre Liu, Bing Asseng, Senthold Ewert, Frank Thorburn, Peter J. Ittersum, Martin K. van Aggarwal, Pramod K. Ahmed, Mukhtar Basso, Bruno Biernath, Christian Cammarano, Davide Challinor, Andrew J. Sanctis, Giacomo de Dumont, Benjamin Eyshi Rezaei, Ehsan Fereres, Elias Fitzgerald, Glenn J. Gao, Y García Vila, Margarita Gayler, Sebastian Girousse, Christine Hoogenboom, Gerrit Horan, Heidi Izaurralde, Roberto César Jones, Curtis D. Kassie, Belay T. Kersebaum, Kurt-Christian Klein, Christian Köhler, Ann-Kristin Maiorano, Andrea Minoli, Sara Müller, Christoph Naresh Kumar, Soora Nendel, Claas O’Leary, Garry J. Palosuo, Taru Priesack, Eckart Ripoche, Dominique Rötter, Reimund P. Semenov, Mikhail A. Stöckle, Claudio O. Stratonovitch, Pierre Streck, Thilo Supit, Iwan Tao, Fulu Wolf, Joost Zhang, Zhao climate change cambio climático crop models ensamble mean simulation models environment ecology A recent innovation in assessment of climate change impact on agricultural production has been to use crop multimodel ensembles (MMEs). These studies usually find large variability between individual models but that the ensemble mean (e‐mean) and median (e‐median) often seem to predict quite well. However, few studies have specifically been concerned with the predictive quality of those ensemble predictors. We ask what is the predictive quality of e‐mean and e‐median, and how does that depend on the ensemble characteristics. Our empirical results are based on five MME studies applied to wheat, using different data sets but the same 25 crop models. We show that the ensemble predictors have quite high skill and are better than most and sometimes all individual models for most groups of environments and most response variables. Mean squared error of e‐mean decreases monotonically with the size of the ensemble if models are added at random, but has a minimum at usually 2–6 models if best‐fit models are added first. Our theoretical results describe the ensemble using four parameters: average bias, model effect variance, environment effect variance, and interaction variance. We show analytically that mean squared error of prediction (MSEP) of e‐mean will always be smaller than MSEP averaged over models and will be less than MSEP of the best model if squared bias is less than the interaction variance. If models are added to the ensemble at random, MSEP of e‐mean will decrease as the inverse of ensemble size, with a minimum equal to squared bias plus interaction variance. This minimum value is not necessarily small, and so it is important to evaluate the predictive quality of e‐mean for each target population of environments. These results provide new information on the advantages of ensemble predictors, but also show their limitations. 2018-11 2018-09-11T20:45:13Z 2018-09-11T20:45:13Z Journal Article https://hdl.handle.net/10568/97157 en Limited Access Wiley Wallach, D.; Martre, P.; Liu, B.; Asseng, S.; Ewert, F.; Thorburn, P J. ; van Ittersum, M.; Aggarwal, P K. ;Ahmed, M. ; Basso, B.; Biernath, C.; Cammarano, D.; Challinor, Andrew J. ;De Sanctis, G.; Dumont, B. ; Eyshi Rezaei, E. ;Fereres, E. ; Fitzgerald, G J. ; Gao, Y. ; Garcia-Vila, M.; Gayler, S. ; Girousse, C.; Hoogenboom, G.; Horan, H.; Izaurralde, R C.; Jones, C D. ; Kassie, B T. ; Kersebaum, K C. ; Klein, C.; Koehler, A K.; Maiorano, A.; Minoli, S.; Müller, C.; Naresh Kumar, S.; Nendel, C.; O'Leary, G J.; Palosuo, T.; Priesack, E.; Ripoche, D.; Rötter, R P.; Semenov, M A.; Stöckle, C.; Stratonovitch, P.; Streck, T.; Supit, I.; Tao, F.; Wolf, J.; Zhang, Z. (2018). Multi-model ensembles improve predictions of crop-environment-management interactions. Global Change Biology, (January). 24(11): 5072-5083. |
| spellingShingle | climate change cambio climático crop models ensamble mean simulation models environment ecology Wallach, Daniel Martre, Pierre Liu, Bing Asseng, Senthold Ewert, Frank Thorburn, Peter J. Ittersum, Martin K. van Aggarwal, Pramod K. Ahmed, Mukhtar Basso, Bruno Biernath, Christian Cammarano, Davide Challinor, Andrew J. Sanctis, Giacomo de Dumont, Benjamin Eyshi Rezaei, Ehsan Fereres, Elias Fitzgerald, Glenn J. Gao, Y García Vila, Margarita Gayler, Sebastian Girousse, Christine Hoogenboom, Gerrit Horan, Heidi Izaurralde, Roberto César Jones, Curtis D. Kassie, Belay T. Kersebaum, Kurt-Christian Klein, Christian Köhler, Ann-Kristin Maiorano, Andrea Minoli, Sara Müller, Christoph Naresh Kumar, Soora Nendel, Claas O’Leary, Garry J. Palosuo, Taru Priesack, Eckart Ripoche, Dominique Rötter, Reimund P. Semenov, Mikhail A. Stöckle, Claudio O. Stratonovitch, Pierre Streck, Thilo Supit, Iwan Tao, Fulu Wolf, Joost Zhang, Zhao Multimodel ensembles improve predictions of crop–environment–management interactions |
| title | Multimodel ensembles improve predictions of crop–environment–management interactions |
| title_full | Multimodel ensembles improve predictions of crop–environment–management interactions |
| title_fullStr | Multimodel ensembles improve predictions of crop–environment–management interactions |
| title_full_unstemmed | Multimodel ensembles improve predictions of crop–environment–management interactions |
| title_short | Multimodel ensembles improve predictions of crop–environment–management interactions |
| title_sort | multimodel ensembles improve predictions of crop environment management interactions |
| topic | climate change cambio climático crop models ensamble mean simulation models environment ecology |
| url | https://hdl.handle.net/10568/97157 |
| work_keys_str_mv | AT wallachdaniel multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT martrepierre multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT liubing multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT assengsenthold multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT ewertfrank multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT thorburnpeterj multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT ittersummartinkvan multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT aggarwalpramodk multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT ahmedmukhtar multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT bassobruno multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT biernathchristian multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT cammaranodavide multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT challinorandrewj multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT sanctisgiacomode multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT dumontbenjamin multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT eyshirezaeiehsan multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT ferereselias multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT fitzgeraldglennj multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT gaoy multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT garciavilamargarita multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT gaylersebastian multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT giroussechristine multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT hoogenboomgerrit multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT horanheidi multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT izaurralderobertocesar multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT jonescurtisd multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT kassiebelayt multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT kersebaumkurtchristian multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT kleinchristian multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT kohlerannkristin multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT maioranoandrea multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT minolisara multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT mullerchristoph multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT nareshkumarsoora multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT nendelclaas multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT olearygarryj multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT palosuotaru multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT priesackeckart multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT ripochedominique multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT rotterreimundp multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT semenovmikhaila multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT stockleclaudioo multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT stratonovitchpierre multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT streckthilo multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT supitiwan multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT taofulu multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT wolfjoost multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions AT zhangzhao multimodelensemblesimprovepredictionsofcropenvironmentmanagementinteractions |