Modeling of the forest gaps-forest interaction

There are few detailed field experiments aimed at the study of turbulence structure and exchange processes of different substances in proximity of the forest edge interface. Such processes considerable impact the interaction between surface and the atmosphere and studying them is needed for many practical applications. In contrast with the scale of experiments required to examine the in-depth spatial variation of physical processes, for instance, downwind of a sharp forest-clearcut discontinuity, modeling is relatively cheap and useful to investigate spatially dependent complex processes. However, LES and higher-order closure models are still expensive technology to take a close look at the open place -forest patch area. K-l models existed (K is the turbulent diffusivity and l-mixing length) are simple but they only allow a gradual change from an upwind mixing length to a downwind equilibrium value. Such kind of models allow the mixing length to drop below the equilibrium value as a result of vertical advection and that is not sufficient to simulate the drop of turbulent diffusivity observed close to the edge (Kruijt, 1993). As compromise, results from a two-equation model which does not require a predefined mixing length and which includes a new parameterization for drag term are presented. The Model SCADIS based on E-ω scheme (where E is turbulent kinetic energy and ω is specific dissipation of E) capable to reproduce both turbulent diffusivity drop and decreasing of turbulent length scales (experimentally confirmed by spectral analysis) downwind of the forest. Such properties of the model would have the potential to be widely applied provided it can be resolved with sufficient reliability many questions concerning the advective contribution to fluxes measured by either gradient or eddy-covariance methods in the presence of surface patchiness. Applications of the model are presented, namely the investigation of increased heat fluxes near a forest edge observed by Klaassen et al., ( 2002), study of effect of clearcuts on footprints and flux measurement at Florida site (Leclerc et al., 2003) and that of Sogachev et al. (2005). The results suggested that an adjustment in the momentum flux does not necessarily mean an adjustment in scalar flux. The flow distortion created by the forest edge produces complex flow motions that influence the scalar distribution. A simultaneous interaction of sources located on the surface and in the canopy layer can produce a net flux enhancement over different fetches, the amplitude and distribution of which is a function of the ratio of source strengths of the surface to that of the canopy layer. It is recommended to estimate surface fluxes of forest from atmospheric observations using expanded footprint models which take into account the turbulence structure the downwind forest edge.