Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N2O emissions

Fiona Ehrhardt, Jean-François Soussana, Gianni Bellocchi, Peter Grace, Russel McAuliffe, Sylvie Recous, Renáta Sándor, Pete Smith, Val Snow, Massimiliano D. A. Migliorati, Bruno Basso, Arti Bhatia, Lorenzo Brilli, Jordi Doltra, Christopher D. Dorich, Luca Doro, Nuala Fitton, Sandro J. Giacomini, Brian Grant, Matthew T. HarrisonStephanie K. Jones, Miko U. F. Kirschbaum, Katja Klumpp, Patricia Laville, Joël Léonard, Mark Liebig, Mark Lieffering, Raphaël Martin, Raia Silvia Massad, Elizabeth Meier, Lutz Merbold, Andrew D. Moore, Vasileios Myrgiotis, Paul Newton, Elizabeth Pattey, Susanne Rolinski, Joanna Sharp, Ward N. Smith, Lianhai Wu, Qing Zhang

Research output: Contribution to journalArticlepeer-review

105 Citations (Scopus)
7 Downloads (Pure)

Abstract

Simulation models are extensively used to predict agricultural productivity and greenhouse gas (GHG) emissions. However, the uncertainties of (reduced) model ensemble simulations have not been assessed systematically for variables affecting food security and climate change mitigation, within multispecies agricultural contexts. We report an international model comparison and benchmarking exercise, showing the potential of multimodel ensembles to predict productivity and nitrous oxide (N2 O) emissions for wheat, maize, rice and temperate grasslands. Using a multistage modelling protocol, from blind simulations (stage 1) to partial (stages 2-4) and full calibration (stage 5), 24 process-based biogeochemical models were assessed individually or as an ensemble against long-term experimental data from four temperate grassland and five arable crop rotation sites spanning four continents. Comparisons were performed by reference to the experimental uncertainties of observed yields and N2 O emissions. Results showed that across sites and crop/grassland types, 23 to 40% of the uncalibrated individual models were within two standard deviations (s.d.) of observed yields, while 42 (rice) to 96% (grasslands) of the models were within one s.d. of observed N2 O emissions. At stage 1, ensembles formed by the three lowest prediction model errors (RRMSE) predicted both yields and N2 O emissions within experimental uncertainties for 44 and 33% of the crop and grassland growth cycles, respectively. Partial model calibration (stages 2 to 4) markedly reduced prediction errors of the full model ensemble E-median for crop grain yields (from 36% at stage 1 down to 4% on average) and grassland productivity (from 44 to 27%) and to a lesser and more variable extent for N2 O emissions. Yield-scaled N2 O emissions (N2 O emissions divided by crop yields) were ranked accurately by 3-model ensembles across crop species and field sites. The potential of using process-based model ensembles to predict jointly productivity and N2 O emissions at field scale is discussed. This article is protected by copyright. All rights reserved.

Original languageEnglish
Pages (from-to)603-616
Number of pages14
JournalGlobal Change Biology
Volume24
Issue number2
Early online date24 Nov 2017
DOIs
Publication statusPublished - Feb 2018

Bibliographical note

FE acknowledges support through a grant from ADEME (n° 12-60-C0023). This study was coordinated by the Integrative Research Group of the Global Research Alliance (GRA) on agricultural greenhouse gases and was supported by five research projects (CN-MIP, Models4Pastures, MACSUR, COMET-Global and MAGGNET) funded by a multi-partner call on agricultural GHGs with support of FACCE JPI. LM was supported by the Swiss National Science Foundation under the 40FA40_154245 / 1 grant agreement; JD participated in the framework of Red REMEDIA. The authors wish to thank Dr. Alex Ruane (NASA GISS) for provision of AgMERRA weather data; Benjamin Loubet (INRA) and Kathrin Fuchs (ETH-Zürich) for their contribution to the consolidation of experimental datasets; Laura Cardenas (Rothamsted Research), Marco Carozzi (INRA) and DairyMod team (Karen Christie, Brendan Cullen, Rachel Meyer, Richard Eckard, Richard Rawnsley) for their modelling efforts; Marco Bindi (University of Florence), Rich Conant (Colorado State University), Heinrike Mielenz (Julius Kühn Institute) and Kairsty Topp (SRUC), for their help as supervisors.

Keywords

  • Journal Article
  • greenhouse gases
  • climate change
  • agriculture
  • benchmarking
  • biogeochemical models
  • nitrous oxide
  • yield
  • soil

Fingerprint

Dive into the research topics of 'Assessing uncertainties in crop and pasture ensemble model simulations of productivity and N2O emissions'. Together they form a unique fingerprint.

Cite this