Un article vient de paraître dans la revue « Agricultural and Forest Meteorology » sur la croissance du chêne sessile et du pin sylvestre. L’objectif de ce travail était d’intégrer des variables climatiques dans les modèles de manière à mieux prédire la croissance dans le contexte des changements climatiques. Ce travail a été réalisé en utilisant des données de l’IGN et des données rétrospectives de largeur de cerne provenant du dispositif OPTMix.
Vallet, P. and T. Perot (2018). « Coupling transversal and longitudinal models to better predict Quercus petraea and Pinus sylvestris stand growth under climate change. » Agricultural and Forest Meteorology 263: 258-266. doi: 10.1016/j.agrformet.2018.08.021
- Large-scale NFI data provide growth models including silvicultural effects.
- Tree rings data provide models for annual modulation of growth by climate.
- Coupling both models allowed to develop climate-dependent stand growth models.
Climate change has swept away the former general principles of long-term stability in forest productivity. New types of models are needed to predict growth and to plan forest management under future climate conditions. These models must remain robust for silvicultural practices and variations in climate. In this study, we present a new type of model development to achieve these goals. Our study focused on pure and mixed stands of Quercus petraea and Pinus sylvestris in central France. We used National Forest Inventory (NFI) data: respectively, 525 and 548 pure plots of Quercus petraea and Pinus sylvestris, and 68 plots of mixed species. We also used 108 tree cores from an experimental site of the same species. The cores cover the period from 1971 to 2013, making a total of 4572 individual annual increments. We coupled two types of models. One was developed with NFI data (transversal data). This model takes into account mean diameter and stand density effects on stand growth. It includes a set of biophysical factors accounting for stand fertility. The other one was developed with the data from tree cores (longitudinal data), and provides a climate modulation thanks to the correlation between ring width and yearly climate. The model with tree core data reveals the influence of December to July rainfalls on yearly variability in stand growth for Quercus petraea and of May to August rainfalls for Pinus sylvestris. We obtained a coupled model that allowed us to project growth up to 2100 for all the different IPCC scenarios but one; the model was outside its area of validity beyond 2060 for the RCP 8.5 scenario.