Modelling fertiliser significance in three major crops

Abstract : We present work using two long term climate datasets to show that nitrogen fertiliser is an important aspect of yield projection for three major crops. The ability of linear models using climate variables as predictors to accurately project the yield of maize, rice and wheat over multi-decadal scales is improved with the addition of fertiliser as an input. Highly productive nations including Argentina, India, Poland and South Africa show significant improvement in yield simulations and show that fertiliser use should not be discounted when estimating yield variability. The use of nitrogen fertiliser in the generalised linear models improves yield forecast by 18% using the Princeton climate dataset and 23% using the WFDEI climate dataset. This work therefore supports the use of additional predictors than climate for improving the ability of statistical models to reconstruct yield variability.
Type de document :
Article dans une revue
European Journal of Agronomy, Elsevier, 2017, 90, pp.1-11. 〈10.1016/j.eja.2017.06.012〉
Liste complète des métadonnées

http://hal.upmc.fr/hal-01581108
Contributeur : Gestionnaire Hal-Upmc <>
Soumis le : lundi 4 septembre 2017 - 12:58:08
Dernière modification le : jeudi 11 janvier 2018 - 06:19:44

Fichier

Parkes_2017_Modelling.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Citation

Ben Parkes, Benjamin Sultan, Philippe Ciais, Xuhui Wang. Modelling fertiliser significance in three major crops. European Journal of Agronomy, Elsevier, 2017, 90, pp.1-11. 〈10.1016/j.eja.2017.06.012〉. 〈hal-01581108〉

Partager

Métriques

Consultations de la notice

109

Téléchargements de fichiers

8