UMR EcoFoG, 6th Plant Biomechanics Conference

Font Size:  Small  Medium  Large

Plant motion in heterogeneous landscapes: a coupled flow-tree simulation study

Yves Brunet

Last modified: 2009-11-06

Abstract


Yves Brunet1, Pascal Roux1, Sylvain Dupont1, Damien Sellier2

1Ephyse, INRA, Villenave d'Ornon, France
2SCION, Rotorua, New Zealand

Landscape heterogeneities such as forest edges, clearings or gaps substantially alter mean velocity and turbulence fields. They induce streamwise variations in turbulence properties that may have considerable impact on the aerodynamic sollications exerted on the trees. The heterogeneity in the damages caused by windstorms in forested landscapes can be partly attributed to such spatial variations in windload. We investigate here the impact of landscape heterogeneity on tree motion using a coupled flow-tree modelling approach. The ultimate aim of the approach is to provide guidance for sustainable forest management practices.
In order to predict the flow behaviour at fine spatial and temporal resolution in complex environments, large-eddy simulation (LES) is a powerful tool. An LES model based on the atmospheric ARPS code has been developed and validated against several data sets for the purpose of providing instantaneous flow fields at fine spatial resolution in landscape configurations with marked heterogeneities.
In order to simulate tree motions induced by instantaneous wind forces, we also developed a dynamical model of tree behaviour, based on a detailed description of its architecture and mechanical properties. Under the forcing of measured turbulent wind, it predicts swaying amplitudes and temporal patterns in good agreement with those measured in the direction of the dominant wind.
This biomechanical model has been coupled with the LES code so that the tree response to a fluctuating wind field can be predicted at any location in a virtual landscape. As an example of application, a comparison between a tree at the forest edge and a tree further downstream shows large differences in the distribution of internal strains, both in space and time.
Using this coupled model we first illustrate the response of trees submitted to an incoming flow influenced by various parameters such as stand density, tree architecture and distance to the edge. We then focus on the changes in tree loading caused by the influence of a gap in a forest stand, defined by various width-to-height ratios. Finally, we characterise the streamwise variations in tree motions along a forested hill, as caused by the variations in mean wind and turbulence.