Scale Dependence of the Deviation of Single-Point Measurements from Spatially Averaged Flow Statistics within a Walnut Orchard Canopy

The exchange of momentum, heat, gases and particles between plant canopies and the atmosphere depends on the dynamic interactions between canopy elements and the surrounding flow. In order to bypass needing to calculate boundary conditions on the surfaces of individual roughness elements, the net effect of canopy elements on the flow within a control volume can be represented as a drag force. Dimensional analysis suggests that the spatially averaged drag force depends on the canopy density, the square of velocity, and a drag coefficient characterizing the efficiency of the canopy elements at extracting momentum from the flow. The drag coefficient is expected to vary with flow conditions, such as velocity and acceleration. As reported by existing studies, accounting for the dependence of drag coefficient on velocity is critical for large-eddy simulation to reproduce field measurements of velocity skewness, quadrant events, and the dispersion of particles. However, developing a drag coefficient model from first principles is non-trivial, calling for empirical estimates from experimental data.

While spatially averaged flow statistics are required for empirical estimates of the drag coefficient, field experimental data are typically sampled at fixed points in space. Investigating the spatial representativeness of single-point measurements is prerequisite for meaningful analysis of the drag-wind relationship. In this work, data from the Canopy Horizontal Array Turbulence Study (CHATS) are used to study the scale dependence of the deviation of single-point measurements from spatially averaged flow statistics. During CHATS, two levels of sonic anemometers, with nine sonic anemometers at each level, were deployed in the crosswind direction between two or three rows of a walnut orchard. We select periods of statistically stationary flows to study the mean, variance and skewness of turbulence statistics at various heights, as well as their sensitivity to the scale of the spatial average analysis. Statistics of filtered flow fields are used to identify the characteristic scales of turbulent kinetic energy and quadrant events that contribute to turbulent transport in the vertical direction. This work explores the link between the spatial variation of flow statistics and the scale of turbulent motions within plant canopies.