Influence of foliar density profile on canopy flow: A large-eddy simulation study

详细信息	查看全文 | 推荐本文 |

• 作者：Sylvain Dupont ; Yves Brunet
• 关键词：Coherent structures ; Coriolis force ; Forest morphology ; Large-eddy simulation ; Mixing-layer analogy ; Turbulent flow ; Vegetation canopy ; Wavelet transform
• 刊名：Agricultural and Forest Meteorology
• 出版年：2008
• 期刊代码：2_01681923
• 类别：abs
• 出版时间：30 June 2008
• 卷：148
• 期：6-7
• 页码：976-990
• 文件大小：1933 K

摘要

The Advanced Regional Prediction System (ARPS) has been modified so as to simulate turbulent flow at very fine scale within and above vegetation canopies using a large-eddy simulation (LES) approach. It is first shown that this new version of ARPS is able to reproduce accurately all essential features of turbulent flow over homogeneous canopies. A sensitivity study of the flow to the morphology of the canopy, i.e. the density and vertical leaf-area (LAI) distribution, is then performed numerically over three types of canopies with five levels of leaf-area index, from 1 to 5. This study confirms the universal characteristics of turbulent flow over vegetation canopies, as previously observed from wind-tunnel and in situ experiments. It shows that the typical features of canopy flow become more pronounced as canopy density increases, and quantifies the extent to which differences in canopy morphology can explain the experimental variability observed between canopies. This variability in turbulent characteristics is mostly visible in the subcanopy space, and depends significantly on the density of the upper foliated layers. The mean longitudinal separation Λ_w between adjacent coherent structures has also been computed for each canopy using the wavelet transform and approximating the convection velocity of coherent structures as 1.8 times the average wind velocity at canopy top. Except for the sparsest canopies, the analogy between the atmospheric flow near the top of a vegetation canopy and a plane mixing layer is well verified: Λ_w is directly related to the shear length scale L_s=U(h)/U^′(h), where h is canopy height, U mean velocity and U^′ is the vertical gradient dU/dz, in a way close to the prediction Λ_w/h=8.1L_s/h.