Elsevier

European Journal of Agronomy

Volume 88, August 2017, Pages 22-40
European Journal of Agronomy

Multi-model simulation of soil temperature, soil water content and biomass in Euro-Mediterranean grasslands: Uncertainties and ensemble performance

https://doi.org/10.1016/j.eja.2016.06.006Get rights and content

Highlights

  • We simulate biomass, soil water content (SWC) and temperature (ST) in grasslands.

  • We compare nine models to the multi-model median (MMM) at nine sites.

  • With model calibration, we obtain satisfactory estimates of ST, less of SWC and biomass.

  • We observe discrepancies across models in the simulation of grassland processes.

  • We improve performance with multi-model approach.

Abstract

This study presents results from a major grassland model intercomparison exercise, and highlights the main challenges faced in the implementation of a multi-model ensemble prediction system in grasslands. Nine, independently developed simulation models linking climate, soil, vegetation and management to grassland biogeochemical cycles and production were compared in a simulation of soil water content (SWC) and soil temperature (ST) in the topsoil, and of biomass production. The results were assessed against SWC and ST data from five observational grassland sites representing a range of conditions – Grillenburg in Germany, Laqueuille in France with both extensive and intensive management, Monte Bondone in Italy and Oensingen in Switzerland – and against yield measurements from the same sites and other experimental grassland sites in Europe and Israel. We present a comparison of model estimates from individual models to the multi-model ensemble (represented by multi-model median: MMM). With calibration (seven out of nine models), the performances were acceptable for weekly-aggregated ST (R2 > 0.7 with individual models and >0.8–0.9 with MMM), but less satisfactory with SWC (R2 < 0.6 with individual models and <  0.5 with MMM) and biomass (R2 < ∼0.3 with both individual models and MMM). With individual models, maximum biases of about −5 °C for ST, −0.3 m3 m−3 for SWC and 360 g DM m−2 for yield, as well as negative modelling efficiencies and some high relative root mean square errors indicate low model performance, especially for biomass. We also found substantial discrepancies across different models, indicating considerable uncertainties regarding the simulation of grassland processes. The multi-model approach allowed for improved performance, but further progress is strongly needed in the way models represent processes in managed grassland systems.

Section snippets

List of the most important symbols and abbreviations used in the paper

Symbol/abbreviationLong versionExplanation
bAridity indexAgro-climatic metric
hwHeat wave days frequencyAgro-climatic metric
GRIGrillenburgGrassland site
OENOensingenGrassland site
LAQLaqueuilleGrassland site
MBOMonte BondoneGrassland site
KEMKemptenGrassland site
LELLelystadGrassland site
MATMattaGrassland site
ROTRothamstedGrassland site
SASSassariGrassland site
AnnuGrowProcess-based model of the growth of annual plants in drylandsGrassland model
ARMOSAMonitoring and modelling nitrogen cycle and crop

Study sites

The nine long-term grassland sites used for the modelling exercise (Table 1) cover a broad range of geographic and climatic conditions (Fig. 1; see also Fig. A and Table A1 in the Supplementary material, Section 1) as well as a variety of management practices (Table A2 in the Supplementary material, Section 1). The sites represent typical grassland cultivation conditions in Europe, which include both intensive and extensive management practices for livestock production (Chang et al., 2015a).

Evaluation of soil temperature (ST) estimates (flux sites)

Fig. 2 shows the range of model results (represented by the shaded area) and the multi-model median (MMM hereinafter) together with the measured values at weekly resolution (see also Figs. B and C of Supplementary material, Section 3, with daily and monthly time resolutions, respectively).

The figure suggests that the range of model results decreased drastically after calibration. However, it is worth noting that the upper bound in Fig. 2 (left) (almost constant ST around 28 °C) is caused by

Soil temperature (ST)

All the models simulated ST relatively well, and their performance for representing ST generally improved after calibration. However, modelling efficiency (ME, at times <0) indicated problems with the quality of the results. It means that the information content of the simulations is questionable in spite of the level of explained variance, which appears high. Therefore, developments are still needed in terms of ST representation of the models to improve the quality of the simulations. Error

Conclusions and future directions

In this study, we presented a framework for proper interpretation of model performances and uncertainties obtained with a set of biophysical models (individually and in an ensemble) simulating grasslands systems at a variety of sites.

There are multiple foci when designing multi-model studies of complex ecosystems (such as grasslands) depending on the questions to be answered. We have not identified the best model for grasslands and we have not assigned probability of success to prove the

Acknowledgements

The results of this research were obtained within an international research project named “FACCE MACSUR – Modelling European Agriculture with Climate Change for Food Security, a FACCE JPI knowledge hub”, with the support of the Hungarian Scientific Research Fund (OTKA K104816) and the EU-FP7 INFRASTRUCTURES-2011-2, BioVel - Biodiversity Virtual e-Laboratory Project (project number 283359), the German Ministry of Education and Research (031A103A), the Italian Ministry of Agricultural, Food and

References (51)

  • R. Sándor et al.

    Uncertainty in simulating biomass yield and carbon-water fluxes from grasslands under climate change

    Adv. Anim. Biosci.

    (2015)
  • N. Senapati et al.

    Modelling heat, water and carbon fluxes in mown grassland under multi-objective and multi-criteria constraints

    Environ. Model. Softw.

    (2016)
  • M.K. van der Molen et al.

    Drought and ecosystem carbon cycling

    Agric. For. Meteorol.

    (2011)
  • J.-A. Vital et al.

    High-performance computing for climate change impact studies with the pasture simulation model

    Comput. Electron. Agric.

    (2013)
  • J.H.M. Wösten et al.

    Development and use of a database of hydraulic properties of European soils

    Geoderma

    (1999)
  • M. Aubinet et al.

    Eddy Covariance: A Practical Guide to Measurement and Data Analysis

    (2012)
  • G.B. Bonan et al.

    Landscapes as patches of plant functional types: an integrating concept for climate and ecosystem models

    Global Biogeochem. Cycles

    (2002)
  • N. Carvalhais et al.

    Implications of the carbon cycle steady state assumption for biogeochemical modeling performance and inverse parameter retrieval

    Global Biogeochem. Cycles

    (2008)
  • A. Cavallero et al.

    Caratterizzazione della dinamica produttiva di pascoli naturali italiani

    Riv. Agron.

    (1992)
  • J. Chang et al.

    Incorporating grassland management in ORCHIDEE: model description and evaluation at 11 eddy-covariance sites in Europe

    Geosci. Model Dev.

    (2013)
  • J. Chang et al.

    The greenhouse gas balance of European grasslands

    Global Change Biol.

    (2015)
  • J. Chang et al.

    Modelled changes in potential grassland productivity and in ruminant livestock density in Europe over 1961–2010

    PLoS One

    (2015)
  • L. Chow et al.

    Field performance of nine soil water content sensors on a sandy loam soil in New Brunswick, Maritime Region, Canada

    Sensors

    (2009)
  • W.S. Cleveland

    Robust locally weighted regression and smoothing scatterplots

    J. Am. Stat. Assoc.

    (1979)
  • E. De Martonne

    Nouvelle carte mondiale de l’indice d’aridité

    Ann. Géogr.

    (1942)
  • Cited by (0)

    1

    University of New South Wales, Climate Change Research Center, Sydney, Australia.

    2

    Catholic University of the Sacred Heart, Department of Sustainable Food Production, Piacenza, Italy.

    View full text