Eddy structure near the plant canopy interface

We have proposed a model for the eddy structure of the RSL-canopy layer that explains qualitatively the unique features of this region and why it differs from the inertial sublayer above. The components of the model are firstly, the inviscid instability of the mean velocity profile at the canopy top, which in turn is a consequence of the momentum absorption over a depth h_C through the canopy rather than at a plane surface. This instability selects the scale of the most-amplified two- and three-dimensional disturbances and relates them to the vorticity thickness at canopy top. Secondly we invoke symmetry breaking by two competing processes. Coherent 'Stuart' vortex tubes generated by the inviscid instability can be deflected up or down by ambient turbulent motions. Downward deflected vortex loops or hairpins experience greater rotation and straining by the mean shear and consequent amplification of the vorticity in the legs of the hairpins than upward deflections. In contrast, large upward deflections are more probable except for deflections by eddies of scale ≤ h_C at the canopy top. In this region the greater strain experienced by downward deflections produces head-down sweep-generating hairpins that are more energetic than ejection-generating head-ups. Far enough above the canopy, this advantage is reversed. Laminar instability theory also predicts that hairpins with a lateral spacing similar to their streamwise spacing will be the most amplified. The result of this scale selection is eddies that are significantly more coherent and effective in transport than in the inertial sublayer above. Finally we note that this simple picture is strongly modulated by the larger scale turbulence in the boundary layer above, which can increase (or decrease) the canopy-top shear and the growth rates of instability modes in spatial patches with horizontal extents large compared to h_C (Raupach et al., 1996). This model is substantially supported by our conditional sampling and EOF analysis of the LES canopy model wind field presented above.