Evolution of turbulent flow characteristics in a hedgerow vineyard during the growing season

The characteristics of turbulent flow at the land-atmosphere interface and within plant canopies are strongly modified by the interactions with vegetation elements. However, only few experimental studies were conducted so far on the effect of changing leaf area on within-canopy turbulence statistics. This aspect is important for deciduous forests and perennial woody crops (orchards and vineyards), where the variation of leaf area is recurrent and substantial during the growing season. Increasing the understanding on canopy turbulence is fundamental to improve the parameterization of multi-layer models to predict vegetation-atmosphere exchanges. In this context, we conducted an experimental campaign in a hedgerow vineyard in North-East Italy carrying out measurements with a vertical array of 3D sonic anemometers, together with canopy structure characterization, in order to analyze the evolution of turbulent flow characteristics from a leafless canopy to full development. Additionally, the effects of wind direction with respect to rows and of atmospheric stability were analyzed. We found that the aerodynamic properties of the vineyard were not only influenced by canopy height and total leaf area, but also by the vertical distribution of leaf density, with the thickness of the mid-upper layer playing a major role in determining the canopy roughness and the mean level of momentum absorption. The characteristics of within-canopy turbulent flow became more similar to a well-defined mixing-layer type flow as foliage developed, but only for diagonal and across-row wind. In contrast, with wind parallel to rows the effect of increasing leaf density was lower and the vineyard was more similar to an open canopy. The influence of atmospheric stability was less important compared to wind direction or leaf density, except for the free convection class. Nevertheless, a variation of canopy aerodynamic parameters with stability was observed and this should be taken into account in atmospheric models.