Dynamic evapotranspiration in tree-resolving LES - The ED2RAFLES model

How evapotranspiration scales from individual plants to a landscape is a critical problem in hydrology. To investigate this problem, we developed a version of the RAMS-Based Forest Large Eddy Simulations (RAFLES) which includes a dynamic, multi-layered evapotranspiration and CO2 uptake. RAFLES resolves vegetation as a heterogeneous 3D field. It can run at resolutions down to 1 m^3 and can incorporate explicit canopy representations derived from lidar. The canopy influences dynamics through a drag term, proportional to the leaf density at each numerical cell, and as a solid volume, which blocks some of the volume and apertures inside the canopy space. To calculate the exchange of scalar fluxes between the canopy and the air, the Ecosystem Demography version 2 (ED2) was coupled with RAFLES. ED2RAFLES dynamically resolves sensible heat fluxes, evapotranspiration, radiative transfer, and carbon fluxes. The surface energy budget is solved at vertical levels inside the canopy to find the skin temperature of the leaves and soil surface, and evapotranspiration rates. Evapotranspiration is also driven by frictional velocity near the leaves and is further restricted by stomatal conductance, parameterized using a Ball-Berry scheme, and by soil water availability. We tested the sensitivity of modeled evapotranspiration to the dynamic representation for a virtual canopy domain, based on the spatial structure of the forest around the Ameriflux eddy-flux tower, at the University of Michigan Biological Station (UMBS). A “test” simulation used the dynamic ED2-driven evaporation. A “control” simulation prescribed sensible and latent heat fluxes as a function of LAI and light attenuation at a rate equal to the mean heat flux from the test simulation. The simulation differed in the degree of spatial heterogeneity of fluxes across the domains and in the coherency between canopy structure and evaporation. Quantifying this coherency is important to our understanding of the effects of canopy structure on fluxes, and predictions of forest function with modified canopy structures. Canopy top (green surface) and stems (brown lines) affect wind flow (white streamlines) inside the canopy sub-domain, and latent (blue scale, side walls) and latent (white-red color scale, back wall) heat fluxes