Skip to Main content Skip to Navigation
Journal articles

Multi-model evaluation of phenology prediction for wheat in Australia

Daniel Wallach 1 Taru Palosuo 2 Peter Thorburn 3 Zvi Hochman 3 Fety Andrianasolo 4 Senthold Asseng 5, 6 Bruno Basso 7, 8 Samuel Buis 9 Neil Crout 10 Benjamin Dumont 11, 12 Roberto Ferrise 13, 14 Thomas Gaiser 15, 16 Sebastian Gayler 17, 18 Santosh Hiremath 19 Steven Hoek 20 Heidi Horan 21 Gerrit Hoogenboom 5, 22, 6 Mingxia Huang 23, 24 Mohamed Jabloun 10 Per-Erik Jansson 25 Qi Jing 26, 27 Eric Justes 28, 29, 30 Kurt Christian Kersebaum 31, 32 Marie Launay 33 Elisabet Lewan 34 Qunying Luo 35 Bernardo Maestrini 8, 20 Marco Moriondo 36 Jørgen Eivind Olesen 37, 38 Gloria Padovan 13, 14 Arne Poyda 39, 40 Eckart Priesack 41, 42, 43 Johannes Wilhelmus Maria Pullens 37, 38 Budong Qian 26, 27 Niels Schütze 44 Vakhtang Shelia 5, 22 Amir Souissi 45, 46 Xenia Specka 31 Amit Kumar Srivastava 15, 16 Tommaso Stella 31 Thilo Streck 17, 18 Giacomo Trombi 13, 14 Evelyn Wallor 31 Jing Wang 23, 47 Tobias K.D. Weber 17, 18 Lutz Weihermüller 48, 49 Allard de Wit 20 Thomas Wöhling 44, 50 Liujun Xiao 5, 51, 52 Chuang Zhao 5, 22 Yan Zhu 51, 52 Sabine Seidel 15, 16
Abstract : Highlights: • A large multi-model study predicting wheat phenology in Australia was performed. • Calibration and evaluation datasets were independently drawn from the same population. • Mean absolute prediction error ranged from 6 to 20 days (median 9 days). • Two thirds of modeling groups predicted better than a simple temperature sum. • Variability between groups using the same model structure was substantial. Predicting wheat phenology is important for cultivar selection, for effective crop management and provides a baseline for evaluating the effects of global change. Evaluating how well crop phenology can be predicted is therefore of major interest. Twenty-eight wheat modeling groups participated in this evaluation. Our target population was wheat fields in the major wheat growing regions of Australia under current climatic conditions and with current local management practices. The environments used for calibration and for evaluation were both sampled from this same target population. The calibration and evaluation environments had neither sites nor years in common, so this is a rigorous evaluation of the ability of modeling groups to predict phenology for new sites and weather conditions. Mean absolute error (MAE) for the evaluation environments, averaged over predictions of three phenological stages and over modeling groups, was 9 days, with a range from 6 to 20 days. Predictions using the multi-modeling group mean and median had prediction errors nearly as small as the best modeling group. About two thirds of the modeling groups performed better than a simple but relevant benchmark, which predicts phenology by assuming a constant temperature sum for each development stage. The added complexity of crop models beyond just the effect of temperature was thus justified in most cases. There was substantial variability between modeling groups using the same model structure, which implies that model improvement could be achieved not only by improving model structure, but also by improving parameter values, and in particular by improving calibration techniques.
Document type :
Journal articles
Complete list of metadata
Contributor : Samuel Buis Connect in order to contact the contributor
Submitted on : Friday, November 26, 2021 - 5:50:22 PM
Last modification on : Wednesday, January 26, 2022 - 5:53:27 PM


Files produced by the author(s)


Distributed under a Creative Commons Attribution 4.0 International License



Daniel Wallach, Taru Palosuo, Peter Thorburn, Zvi Hochman, Fety Andrianasolo, et al.. Multi-model evaluation of phenology prediction for wheat in Australia. Agricultural and Forest Meteorology, 2021, 298-299, ⟨10.1016/j.agrformet.2020.108289⟩. ⟨hal-03119039⟩



Les métriques sont temporairement indisponibles