A Study of The Diurnal Cycle With DNS and LES, Towards Hybrid Modelling?

In this work we assess the feasibility of combining both large eddy simulations (LES) and direct numerical simulations (DNS) to study the diurnal cycle. LES is an efficient and well tested computational tool for studying convective and weakly stable turbulent atmospheric boundary layer flows. In the strongly stable regime however, LES relies heavily on the adopted turbulence closure (i.e. subgrid-scale model). Alternatively, DNS does not rely on such closure and may be suitable for studying the very stable boundary layer. However, the latter method is computationally much more expensive particularly at higher Reynolds numbers. As such, in the atmospheric context, it has been primarily applied to strongly stratified flows. Because of the fact that, during day-night transition the turbulence regime may shift from highly convective to strongly stable, using solely LES or DNS for this part of the day appears not to be a feasible option. As a result, the day-night transition largely remains a terra incognita. Here, we aim to describe and test a method to combine the “best of both worlds” as a new approach to study day-night transitions in future research.