Interaction of aerodynamic roughness length and windflow conditions and its parameterization over vegetation surface

Surface aerodynamic roughness length is usually taken as a constant. In fact, it displays a remarkable dynamic change over underlying vegetation surfaces, because of the coupling of land surface roughness elements and windflow conditions. Current international research on this dynamic change and associated mechanisms is very limited. Using observations from different underlying surfaces (including forest, farmland and grassland) provided by a northern China coordinated observation test, the variation of aerodynamic roughness length, along with wind speed and friction velocity, is analyzed. We introduce two relationship fits, between aerodynamic roughness length and wind speed u, and dynamic variable u ²/u*. Results show that aerodynamic roughness length has a clear dynamic change, and has complicated interactions with near-surface windflow. Further, the relationship fits between aerodynamic roughness length, u and u ²/u*, are not only related to the roughness properties of the underlying vegetation surface (e.g. plant height), but also to plant dynamic response characteristics (e.g. flexibility). Aerodynamic roughness length decreases with increasing wind speed, because near-surface windflow conditions can change both plant roughness properties and airflow. However, the change of aerodynamic roughness length with friction velocity is complicated, and its sensitivities and transition points significantly depend on vegetation type. For underlying surfaces of forest and corn, with relatively substantial vegetative cover, roughness length correlates well with wind speed. For a surface with short vegetative cover, like natural lawn, the correlation is low. However, for all of the three vegetative surfaces, there is a close relation between roughness length and u ²/u*, and their coefficients of fit from testing essentially represent the plant height and flexibility of different vegetation types. The test results also indicate that the parameterized relationships of roughness length over the underlying vegetation surface hold prospects for application.