Multimodel ensembles improve predictions of crop–environment–management interactions

Global Change Biology - Tập 24 Số 11 - Trang 5072-5083 - 2018
Daniel Wallach1, Pierre Martre2, Bing Liu3,4, Senthold Asseng3, Frank Ewert5,6, Peter J. Thorburn7, M.K. van Ittersum8, Pramod Aggarwal9, Mukhtar Ahmed10,11, Bruno Basso12,13, Davide Cammarano14, Andrew J. Challinor15,16, Giacomo De Sanctis17, Benjamin Dumont18, Ehsan Eyshi Rezaei19,5, Glenn J. Fitzgerald20,21, Sebastian Gayler3, Margarita García‐Vila22, Christine Girousse23, Gerrit Hoogenboom3,24, Heidi Horan7, R. C. Izaurralde25,26, Curtis D. Jones26, Belay T. Kassie3, Kurt Christian Kersebaum27, Ann‐Kristin Koehler16, Andrea Maiorano2, Sara Minoli28, Christoph Müller28, Claas Nendel29, Garry J. O’Leary30, Taru Palosuo31, Eckart Priesack32, Reimund P. Rötter33,34, Mikhail A. Semenov35, Pierre Stratonovitch35, Thilo Streck36, Fulu Tao37,31, Heidi Webber38, Zhao Zhang39
1UMR AGIR, INRA, 31326 Castanet-Tolosan, France
2UMR LEPSE, INRA, Montpellier SupAgro, Montpellier, France
3Agricultural and Biological Engineering Department, University of Florida, Gainesville, Florida
4National Engineering and Technology Center for Information Agriculture, Key Laboratory for Crop System Analysis and Decision Making, Ministry of Agriculture, Jiangsu Key Laboratory for Information Agriculture, Jiangsu Collaborative Innovation Center for Modern Crop Production, Nanjing Agricultural University, Nanjing, Jiangsu, China
5Institute of Crop Science and Resource Conservation (INRES), University of Bonn, Germany
6Leibniz Centre for Agricultural Landscape Research, Müncheberg, Germany
7CSIRO Agriculture and Food Brisbane, St Lucia, Queensland, Australia
8Plant Production Systems group, Wageningen University, Wageningen, The Netherlands
9CGIAR Research Program on Climate Change, Agriculture and Food Security, BISA-CIMMYT, New Delhi, India
10Biological Systems Engineering, Washington State University, Pullman, Washington
11Department of Agronomy, Pir Mehr Ali Shah Arid Agriculture University, Rawalpindi, Pakistan
12Department of Earth and Environmental Sciences, Michigan State University, East Lansing, Michigan
13W.K. Kellogg Biological Station, Michigan State University, East Lansing, Michigan
14James Hutton Institute Invergowrie, Dundee, Scotland, UK
15CGIAR-ESSP Program on Climate Change, Agriculture and Food Security, International Centre for Tropical Agriculture (CIAT), Cali, Colombia
16Institute for Climate and Atmospheric Science, School of Earth and Environment, University of Leeds, Leeds, UK
17European Food Safety Authority, GMO Unit, Parma, Italy
18Department Terra & AgroBioChem, Gembloux Agro-Bio Tech, University of Liege, Liege, Belgium
19Center for Development Research (ZEF), Bonn, Germany
20Agriculture Victoria Research, Department of Economic Development, Jobs, Transport and Resources, Ballarat, Victoria, Australia
21Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Creswick, Victoria, Australia
22IAS-CSIC and University of Cordoba, Cordoba, Spain
23UMR GDEC INRA Université Clermont Auvergne Clermont‐Ferrand France
24Institute for Sustainable Food Systems, University of Florida, Gainesville, Florida
25Department of Geographical Sciences, University of Maryland, College Park, Maryland
26Texas A&M AgriLife Research and Extension Center Texas A&M University Temple Texas
27Institute of Landscape Systems Analysis, Leibniz Centre for Agricultural Landscape Research, Müncheberg, Germany
28Potsdam Institute for Climate Impact Research, Potsdam, Germany
29Centre for Environment Science and Climate Resilient Agriculture, Indian Agricultural Research Institute, IARI PUSA, New Delhi, India
30Grains Innovation Park, Department of Economic Development, Jobs, Transport and Resources, Agriculture Victoria Research, Horsham, Victoria, Australia
31Natural Resources Institute Finland (Luke), Helsinki, Finland
32Institute of Biochemical Plant Pathology, Helmholtz Zentrum München, German Research Center for Environmental Health, Neuherberg, Germany
33Centre of Biodiversity and Sustainable Land Use (CBL), University of Göttingen, Göttingen, Germany
34Tropical Plant Production and Agricultural Systems Modelling (TROPAGS) University of Göttingen Göttingen Germany
35Computational and Systems Biology Department, Rothamsted Research, Harpenden, Herts, UK
36Institute of Soil Science and Land Evaluation, University of Hohenheim, Stuttgart, Germany
37Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Science, Beijing, China
38Plant Production Systems, Wageningen University, Wageningen, The Netherlands
39State Key Laboratory of Earth Surface Processes and Resource Ecology, Faculty of Geographical Science, Beijing Normal University, Beijing, China

Tóm tắt

AbstractA 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.

Từ khóa


Tài liệu tham khảo

10.18174/odjar.v1i1.14746

10.1038/nclimate2470

10.1038/nclimate1916

10.1111/gcb.12520

10.18637/jss.v067.i01

10.1016/j.tplants.2017.02.003

10.1175/JCLI-D-13-00030.1

10.1002/2014GL060133

10.1016/j.advwatres.2006.11.014

10.1016/j.envsoft.2014.12.003

10.1111/gcb.13411

10.5194/bg-2016-527

10.1111/j.1600-0870.2005.00103.x

10.1038/s41598-017-13582-y

IPCC, 2014, Climate change 2014: Impacts, adaptation, and vulnerability. Part A: Global and sectoral aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 1

10.1111/gcb.12758

10.1038/nclimate3115

Maiorano A., 2016, Crop model improvement reduces the uncertainty of the response to temperature of multi‐model ensembles, Field Crops Research

10.1016/j.jcs.2013.02.003

10.18174/odjar.v3i1.15766

10.1111/gcb.12768

10.1111/gcb.12830

10.1016/j.eja.2011.05.001

Porter J. R., 2014, Climate Change 2014: Impacts, adaptation, and vulnerability. Part A: Global and sectoral aspects. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, 485

R Core Team, 2012, R: A language and environment for statistical computing

Raftery A. E. Balabdaoui F. Gneiting T. &Polakowski M.(2003).Using Bayesian Model Averaging to Calibrate Forecast Ensembles. Retrieved fromhttp://www.stat.washington.edu/www/research/reports/2003/tr440.pdf

10.1073/pnas.1222463110

10.1016/j.agrformet.2012.09.011

10.1038/nclimate1152

10.1016/j.fcr.2012.03.016

Scheffé H., 1959, The analysis of variance

10.5194/acp-15-2535-2015

Surowiecki J., 2005, The wisdom of crowds

10.1098/rsta.2007.2076

10.1007/s10584-016-1803-1

10.1007/s00382-008-0460-0

10.1002/qj.210

10.1016/j.eja.2016.12.009

10.1029/2005GL023513

Thông tin
Thông tin xuất bản

Global Change Biology

Tập 24 Số 11

5072-5083

Thông tin tác giả