Journal cover Journal topic
Geoscientific Model Development An interactive open-access journal of the European Geosciences Union
https://doi.org/10.5194/gmd-2017-272
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
Model evaluation paper
17 Nov 2017
Review status
This discussion paper is a preprint. It is a manuscript under review for the journal Geoscientific Model Development (GMD).
Comparison of dealiasing schemes in large-eddy simulation of neutrally-stratified atmospheric boundary-layer type flows
Fabien Margairaz1, Marco G. Giometto2, Marc B. Parlange2, and Marc Calaf1 1University of Utah, Department of Mechanical Engineering, 1495 E 100 S, Salt Lake City, Utah 84112, USA
2University of British Columbia, Faculty of Applied Science, 2332 Main Mall, Vancouver BC V6T 1Z4, Canada
Abstract. Three dealiasing schemes for large-eddy simulation of turbulent flows are inter-compared for the canonical case of pressure-drive atmospheric boundary-layer type flows. Aliasing errors arise in the multiplication of partial sums, such as those encountered when integrating the non-linear terms of the Navier–Stokes equations in spectral methods (Fourier or polynomial discrete series), and are detrimental to the accuracy of the numerical solution. This is of special relevance when using high-order schemes. In this work, a performance/cost analysis is developed for three well-accepted approaches: the exact 3/2 rule, the Fourier truncation method, and a high order Fourier smoothing method. Tests are performed within a newly developed mixed pseudo-spectral collocation - finite differences large-eddy simulation code, parallelized using a two-dimensional pencil decomposition. The static Smagorinsky eddy-viscosity model with wall damping of the model coefficient is used. A series of simulations are performed at varying resolution and key flow statistics are inter-compared among the considered dealiasing schemes. The numerical results validate the numerical performance predicted by theory when using the Fourier truncation and Fourier smoothing methods. In terms of turbulence statistics, the Fourier Truncation method proves to be over-dissipative when compared against the Fourier Smoothing method and the traditional 3/2-rule, leading to an enhanced horizontal integrated mass flux and to higher dispersive momentum fluxes. Its use in large-eddy simulation of atmospheric boundary-layer type flows is therefore not recommended. Conversely, the Fourier Smoothing method yields accurate flow statistics, comparable to those resulting from the application of the 3/2 rule, with a significant reduction in computational cost, which makes it a convenient alternative for use in the studies related to the atmospheric boundary layer.

Citation: Margairaz, F., Giometto, M. G., Parlange, M. B., and Calaf, M.: Comparison of dealiasing schemes in large-eddy simulation of neutrally-stratified atmospheric boundary-layer type flows, Geosci. Model Dev. Discuss., https://doi.org/10.5194/gmd-2017-272, in review, 2017.
Fabien Margairaz et al.
Fabien Margairaz et al.

Model code and software

LES-Utah-momentum
F. Margairaz, M. Calaf, and M. Giometto
https://doi.org/10.5281/zenodo.1048337
Fabien Margairaz et al.

Viewed

Total article views: 160 (including HTML, PDF, and XML)

HTML PDF XML Total BibTeX EndNote
131 26 3 160 3 4

Views and downloads (calculated since 17 Nov 2017)

Cumulative views and downloads (calculated since 17 Nov 2017)

Viewed (geographical distribution)

Total article views: 160 (including HTML, PDF, and XML)

Thereof 160 with geography defined and 0 with unknown origin.

Country # Views %
  • 1

Saved

Discussed

Latest update: 16 Dec 2017
Publications Copernicus
Download
Short summary
In this project, we compare three different approaches to integrate the fluid-motion equations when applied to solve atmospheric flow dynamics. Differences between the three methods reside on accuracy as well as computational cost. Results illustrate that there is an intermediate solution that performs well in terms of computational cost, while at the same time producing good enough results, as long one is not interested in the smallest turbulent scales.
In this project, we compare three different approaches to integrate the fluid-motion equations...
Share