Evaluation of energy balance closure adjustment and imbalance prediction methods in the convective boundary layer-A large eddy simulation study

The non-closure of the surface energy balance is one of the greatest challenges in quantifying the atmospheresurface exchange of energy and water. One open question associated with the energy imbalance is how to partition the residual energy, i.e., the difference between the available energy and the sum of turbulent fluxes, between the sensible and latent heat fluxes. Here, based on high-resolution large-eddy simulations (LESs), five energy balance closing methods and three imbalance prediction methods are evaluated in the convective boundary layer over idealized heterogeneous surfaces. The results show that advection fluxes are highly correlated with flux imbalances and are leading factors for the flux imbalances. Overall, the Bowen ratio closure method has a better performance than other closure methods, especially when the vertical advection flux dominates the flux imbalance where its assumption is nearly fulfilled. However, when the horizontal advection flux dominates the flux imbalance, the Bowen ratio closure method cannot correctly close the sensible and latent heat fluxes, even though it exhibits better performance than other methods. This is mainly because the heat and water vapor transported by horizontal advection originate from different sources due to surface heterogeneity, which breaks the assumption of the Bowen ratio closure method, leading to failure. Moreover, using the footprint-weighted surface true fluxes (weighted surface true fluxes within the source area of EC) calculated by the Lagrangian particle model instead of the analytical footprint models can reduce the flux imbalance and improve the performance of different closure methods. None of the existing imbalance prediction methods can correctly predict the flux imbalance for H or LE, even though the prediction method proposed over heterogeneous surfaces has a better performance than those proposed over homogeneous surfaces. Our findings provide meaningful suggestions for the selection of energy balance closing methods in practice.