Large-eddy simulations with a dynamic explicit vegetation model

Abstract:

We coupled the Regional Atmospheric Modeling System (RAMS)-based Forest Large-Eddy Simulation (RAFLES) and a modified version of the Ecosystem Demography model version 2 (ED2) to form a dynamic, high resolution, physiologically driven large eddy simulation. RAFLES represents both drag and volume restriction by the canopy over an explicit 3-D domain. We conducted a sensitivity analysis of uplift and circulation patterns at the front and back of a rectangular barrier to the representation of the canopy volume. We then used this model to perform a virtual experiment using combinations of realistic heterogeneous canopies and virtual homogenous canopies combined with heterogeneous and homogenous patterns of soil moisture to test the effects of the spatial scaling of soil moisture on the fluxes of momentum, heat, and water in heterogeneous environments at the tree-crown scale. Further simulations were performed to test the combined effects of canopy structure, soil moisture heterogeneity, and soil water availability. We found flux dynamics of momentum, heat, and water to be significantly influenced by canopy structure, soil moisture heterogeneity, and soil water availability. During non-plant-limiting soil-water conditions, we found canopy structure to be the primary driver of tree-crown scale fluxes of momentum, heat, and water, specifically through modification of the ejection sweep dynamics. However, as soil water conditions became limiting for latent heat flux from plants, tree-crown scale fluxes of momentum and heat became influenced by the spatial pattern of soil moisture, whereas soil moisture became a significant driver of tree-crown scale fluxes of water along with canopy structure.