Adaptation and evaluation of a mechanistic grass growth simulation model for grass-based systems

Cristina Hurtado-Uria (1, 2) , Deirdre Hennessy (1) , Luc Delaby (3) , Declan O'Connor (2) , Laurence Shalloo (1)

1 Moorepark Food Research Centre
2 CIT - Cork Institute of Technology
3 PEGASE - Physiologie, Environnement et Génétique pour l'Animal et les Systèmes d'Elevage [Rennes]

An accurate grass growth model would be a valuable tool in anticipating grass growth and grass utilization at farm level. Ideally, a grass growth simulation model must be accurate, dynamic, use realistic input parameters and incorporate meteorological data. The objective of this study was to parameterize the grass growth model developed by Jouven et al. (2006) to increase its accuracy of grass growth simulation in the south of Ireland. The model was parameterized using an optimization technique where a number of the parameters in the model were optimized with the objective function of minimizing the root mean square error (RMSE). Both meteorological and grass growth data for the period 2005 to 2009 were included in the optimization process. During validation the Jouven Model was compared to the Adapted Model. RMSE was reduced from 20.45 kg DM ha-1 day-1 with the Jouven Model to 14.62 kg DM ha-1 day-1 with the Adapted Model. MSPE was reduced from 476 to 183. The proposed changes to the Jouven Model improved grass growth simulation in the south of Ireland. The adapted version of the Jouven Model can be used for grass growth simulation ; albeit without perfect simulation.

Domaines: Sciences du Vivant [q-bio]

Cristina Hurtado-Uria, Deirdre Hennessy, Luc Delaby, Declan O'Connor, Laurence Shalloo. Adaptation and evaluation of a mechanistic grass growth simulation model for grass-based systems. 17th Symposium of the European Grassland Federation (EGF), Jun 2013, Akureyri, Iceland. European Grassland Federation, Grassland Science in Europe, 18, 2013, Grassland Science in Europe. ⟨hal-01210590⟩