Large-Eddy Simulation of Energy Balance Closure in fully Heterogeneous Terrain

The lack of energy balance closure at the interface between the earth's surface and the atmospheric boundary layer is a long-standing problem in eddy-covariance measurements and it has important implications for boundary-layer models that use the surface heat fluxes as boundary conditions. There is evidence that this energy balance closure problem is exacerbated in complex terrain, which points towards secondary circulations originating from the surface heterogeneity as a possible cause of the imbalance. By resolving these boundary-layer motions by large-eddy simulation techniques we attempt to reconstruct the experimentally found imbalance. However, earlier LES studies of the energy balance closure have primarily considered fully homogeneous domains or domains that are heterogeneous in only one dimension. One advantage of analyzing these idealized terrains is that they allow area-averaging or line-averaging to deduce the "true" turbulent flux (derived from spatial correlations) to which the observations of the turbulent flux (from temporal correlations at a virtual tower in the simulation domain) can then be compared. Directly following this "spatial averaging" approach is not feasible in fully complex terrain. Instead we will apply a different approach to study the energy balance, by focusing on a control volume around the virtual tower and calculating all terms: including those theoretically present in addition to those experimentally measured. We validate our approach by comparing with the spatial averaging approach for fully homogenous and 1D heterogeneous domains. Then we simulate a fully horizontally heterogeneous field site and compare the results with in-situ measurements. The site under study is located at an agricultural research facility. The terrain is flat, which allows us to concentrate on the circulations arising from the spatially variable surface heat fluxes and surface roughness of different crops instead of differences in topography.