Connecting the failure of K- theory inside vegetation canopies and ejection-sweep cycles by a Large Eddy Simulation (LES).

Abstract:

The failure of gradient diffusion parameterization for mean variables is a wide known issue in canopy turbulence. Especially for scalars like temperature, parameterizing the scalar flux with the mean scalar gradient with a turbulent diffusivity (K) as a proportionality parameter (hence K-theory) often fails in the canopy sub-layer. The main reasons behind this failure can be identified as (i) variable source or sink distribution inside the canopy, i.e., near field effects, (ii) lack of local balance between turbulent production and dissipation and (iii) vertical transport by coherent eddy motions of characteristic dimensions larger than the scale at which change of mean gradients takes place, i.e., non-local effects. All of these reasons can be traced to the triple moment in the turbulent scalar flux budget equation. While a prognostic approach can be taken to attempt writing closure equations for this triple moment which takes care of the first two local effects, a diagnostic approach can also be attempted to link the nonlocal effects to the scalar triple moment. The relative importance between sweep and ejections can be linked to this triple moment by means of an Incomplete Cumulant Expansion Method (ICEM), which provides the basis for a non-local closure. In the present work, a Large Eddy Simulation (LES) is used to resolve scalar turbulence in a vegetation canopy. Quadrant analysis is performed to quantify the contributions from sweep and ejections events and this ICEM closure is investigated. When the flux transport term is closed by a gradient diffusion, and when expanded by third order cumulants, and those two approximations to the same term are equated, predictions of how the relative importance of ejections and sweeps emerges from local gradient diffusion can be made. Deviations between Quadrant analysis and such predictions can also highlight the role of nonlocal transport by ejections and sweeps beyond those associated with gradient diffusion of triple moment. Thus a new closure for the triple moment anchored to the ejection sweep cycle may emerge and it will be investigated by means of field data and LES.