Canopy Structure and Environment

Turbulence

Air flow changes abruptly when it encounters a canopy. Air currents are smooth and horizontal above a stand, but are decelerated and re-directed in three-dimensional space upon entering a stand, creating canopy turbulence (Finnigan 2000). Air is the vector for the exchange of heat, gases, and airborne particles between vegetation and the surrounding atmosphere. Accurate models of turbulence are therefore of great importance for researchers interested in net ecosystem processes, including global change parameters such as energy exchange and carbon flux.

Canopy height, density, species composition and continuity influence air movement; since these are all variables that tend to be mapped at a broad geographic scale, most turbulence models are designed for relatively large areas. Early interpretations of canopy wind dynamics used a boundary layer model, which represents the canopy surface as a topographic sheet (Finnigan 2000). However, this model does not account for complex vertical structure within the canopy, which has considerable effects on local air movement. Vertical profiles of canopy turbulence are better predicted by the plane mixing layer model (illustrated below), a concept taken from engineering sciences.

Figure 1. A plane mixing layer model of wind-canopy interactions. Turbulence is created by a series of interactions between wind and canopies: (a) a large-scale wind gust blows above the canopy; (b) horizontal air flow is caught by the canopy, forming "rollers" at regular intervals; (c) secondary instability in the rollers creates complex three-dimensional eddies and airstream interactions. From Finnigan (2000).

The plane mixing layer is "the free shear layer that forms when two airstreams of different velocity merge" (Raupach et al. 1996). Figure 1 depicts the formation of turbulance in three stages, according to a plane mixing layer model. Figure 2 is an idealized graph of the mean (2a) and standard deviation (2b) of air momentum above and within the canopy, and Figure 3 shows real data collected from natural and artificial canopies (the latter simulated using wind tunnels). The trends in Figure 3 generally conform to the idealized graph, with differences due primarily to variation in foliage distribution (Raupach et al. 1996). The Eucalypt stand (3f) is a notable exception, causing no observable reduction in air momentum.

 

Figure 2. Expected mean (1a) and variance (1b) of air momentum above and within a canopy. From Raupach et al. (1996).

 

Figure 3. A "family portrait" of canopy turbulence for canopies "a" through "j" in Table 1. Canopy heights (y-axis) are normalized; values greater than 1 are above the canopy, values less than 1 are within it. Air momentum is shown along the x-axis. From Raupach et al. (1996).