Development and Evaluation of Plant Growth Models: Methodology and Implementation in the PYGMALION platform

Abstract

Abstract. Mathematical models of plant growth are generally characterized by a large number of interacting processes, a large number of model parameters and costly experimental data acquisition. Such complexities make model parameterization a difficult process. Moreover, there is a large variety of models that coexist in the literature with generally an absence of benchmarking between the different approaches and insufficient model evaluation. In this context, this paper aims at enhancing good modelling practices in the plant growth modeling community and at increasing model design efficiency. It gives an overview of the different steps in modelling and specify them in the case of plant growth models specifically regarding their above mentioned characteristics.Different methods allowing to perform these steps are implemented in a dedicated platform PYGMALION (Plant Growth Model Analysis, Identification and Optimization). Some of these methods are original. The C++ platform proposes a framework in which stochastic or deterministic discrete dynamic models can be implemented, and several efficient methods for sensitivity analysis, uncertainty analysis, parameter estimation, model selection or data assimilation can be used for model design, evaluation or application.Finally, a new model, the LNAS model for sugar beet growth, is presented and serves to illustrate how the different methods in PYGMALION can be used for its parameterization, its evaluation and its application to yield prediction. The model is evaluated from real data and is shown to have interesting predictive capacities when coupled with data assimilation techniques.

SUNLAB: A functional-structural model for genotypic and phenotypic characterization of the sunflower crop

“…Generally, our approach fits into a current general trend of development of modular models, with generic modules that can be shared by other modelers. This trend goes hand in hand with the increasing number of modeling platforms: Pygmalion in our case (Cournède et al, 2013), OpenAlea (Pradal et al, 2004), GroIMP (Kniemeyer et al, 2006) It is expected that mechanistic description of ecophysiological processes improves the model predictive capacities and their ability to differentiate between genotypes (Allen et al, 2005;Minchin and Lacointe, 2005;Bertheloot et al, 2011 (Christopoulos and Michael, 2000). Moreover, the parameterization effort of these more and more complex models should always be taken into account when improving their mechanistic description, to prevent from a high level of uncertainty in the parameters which may hinder the original purposes of the model in terms of prediction and genotypic differentiation.…”

Analysis of Parameter Sensitivity and Identifiability of Root Zone Water Quality Model (RZWQM) for Dryland Sugarbeet Modeling

“…Because no model was available in DSSAT for sugarbeet, Anar et al (2015) modified and incorporated CERES-Beet (Leviel, 2000) into DSSAT 4.6.1, and the resultant sugarbeet model is termed CSM-CERES-Beet. Baey et al (2014) showed that CERES-Beet provided overall good predictions of plant growth and yield for sugarbeets after comparing CERESBeet with four other sugarbeet models: GreenLab (Vos et al, 2007), LNAS (Cournede et al, 2013), STICS (Brisson et al, 1998), and Pilote (Taky, 2008).…”

Leaf Segmentation and Tracking in Arabidopsis thaliana Combined to an Organ-Scale Plant Model for Genotypic Differentiation

“…Mathematically speaking, if we consider the system of interest as the plant in its environment (or a population of plants, or a specific part of the plant for models at smaller scales) plant growth models could formally be represented in the very generic following form: $Y=ffalse(\theta ,Efalse)$ where: Y represents all the phenotypic traits of interest, and is generally a real-valued function of space and time. f represents the functional equations (usually dynamical, see for example the description of plant growth models as dynamic state space models and hidden Markov models in Cournède et al, 2013). …”

“…f represents the functional equations (usually dynamical, see for example the description of plant growth models as dynamic state space models and hidden Markov models in Cournède et al, 2013). …”