Spectrally Decomposed Diffusion Models for Generative Turbulence Recovery
Authors:
Mohammed Sardar,
Alex Skillen,
Małgorzata J. Zimoń,
Samuel Draycott,
Alistair Revell
Abstract:
We investigate the statistical recovery of missing physics and turbulent phenomena in fluid flows using generative machine learning. Here we develop a two-stage super-resolution method using spectral filtering to restore the high-wavenumber components of a Kolmogorov flow. We include a rigorous examination of generated samples through the lens of statistical turbulence. By extending the prior meth…
▽ More
We investigate the statistical recovery of missing physics and turbulent phenomena in fluid flows using generative machine learning. Here we develop a two-stage super-resolution method using spectral filtering to restore the high-wavenumber components of a Kolmogorov flow. We include a rigorous examination of generated samples through the lens of statistical turbulence. By extending the prior methods to a combined super-resolution and conditional high-wavenumber generation, we demonstrate turbulence recovery on a $8\times$ upsampling task, effectively doubling the range of recovered wavenumbers.
△ Less
Submitted 22 December, 2023;
originally announced December 2023.
The influence of spectral bandwidth and shape on deep-water wave breaking onset
Authors:
M. L. McAllister,
N. Pizzo,
S. Draycott,
T. S. van den Bremer
Abstract:
Deep-water surface wave breaking affects the transfer of mass, momentum, energy and heat between the air and sea. Understanding when and how the onset of wave breaking will occur remains a challenge. The mechanisms that form steep waves, i.e. nonlinearity or dispersion, are thought to have a strong influence on the onset of wave breaking. In two-dimensions and on deep-water, spectral bandwidth is…
▽ More
Deep-water surface wave breaking affects the transfer of mass, momentum, energy and heat between the air and sea. Understanding when and how the onset of wave breaking will occur remains a challenge. The mechanisms that form steep waves, i.e. nonlinearity or dispersion, are thought to have a strong influence on the onset of wave breaking. In two-dimensions and on deep-water, spectral bandwidth is the main factor that affects the roles these mechanism play. Existing studies, in which the relationship between spectral bandwidth and wave breaking onset is investigated, present varied and sometimes conflicting results. We perform numerical simulations of two-dimensional focused wave groups on deep-water to better understand this relationship, with the aim reconciling existing studies. We show that the way in which steepness is defined, may be the main source confusion in the literature. At breaking onset, locally defined steepness reduces as a function of bandwidth, and globally defined steepness increases. The relationship between global breaking onset steepness and spectral shape (using the parameters bandwidth and spectral skewness) is too complex to parameterise in a general sense. However, we find that the local surface slope of maximally steep non breaking waves, of all spectral bandwidths and shapes (constant-steepness, constant-amplitude, and JONSWAP), approaches a limit of $1/\tan(π/3)\approx0.5774$. This slope based threshold, is simple to measure and may be used as an alternative to existing kinematic breaking onset thresholds. There is a potential link between slope based and kinematic breaking onset thresholds which future work should seek to better understand.
△ Less
Submitted 15 May, 2023;
originally announced May 2023.
