Large eddy simulation of the neutral atmospheric boundary layer: performance evaluation of three inflow methods for terrains with different roughness

In large eddy simulations (LES) of the atmospheric boundary layer (ABL), the choice of inflow method and the specification of surface roughness are crucial for obtaining accurate results. The consistency between inflow conditions and wall boundary condition has been investigated in depth for the RANS approach, whereas it needs further analysis for LES. This paper investigates the combined effect of inflow method and aerodynamic roughness length for LES of the neutral ABL. Three basic inflow generators in combination with three terrain types are evaluated in terms of streamwise homogeneity of the vertical profiles of mean velocity and turbulence kinetic energy. The results show that the precursor method best preserves the homogeneity of the profiles for all roughness lengths considered, with mean absolute deviations up to 0.7% for mean velocity and 3.5% for turbulence kinetic energy at the outlet of the computational domain. Among the synthetic methods, the Vortex Method performs satisfactorily for mean velocity and, to a lesser extent, turbulence kinetic energy, whereas the random flow generation (RFG) method leads to the largest deviations from the target profiles. Finally, the value of the aerodynamic roughness length is found to only weakly influence the performance of the inflow methods considered.