Leibniz MMS Days 2020 - Abstract

Schaefer-Rolffs, Urs

How important is a spectral drop-off near the resolution scale?

Different contradictory perspectives exist on the high-wavelength part of the resolved spectra in high-resolution global circulation models. Physically consistent are such simulations where the resolved fields of wind and temperature would lead to spectra of kinetic and available potential energy with a spectral slope of -5/3 up to the resolution scale according to measurements e.g. by Nastrom and Gage (1984). This has been done in a spectral model by Schaefer-Rolffs and Becker (2013, 2018). From a mathematical point of view, the finite spectral representation in local space leads to the Gibbs phenomenon. In areas of the atmospheric simulation where sharp gradients are present, e.g. in fronts in the atmosphere, such overshoots can lead to the spurious generation of artificial gravity waves, where no such noise is observed in nature. An additional spectral regime in the resolved scales defined by an "effective resolution" and usually referred to as the viscous subrange, can prevent the existence of such gravity waves. However, the cost of suppressing these waves is that all information beyond this "effective resolution" has no practical meaning. Therefore we want to discuss the importance of these contradictory aspects are for the modelling of the atmosphere.