Evolution of passive scalar mixing layers in stratified and unstratified homogeneous turbulence
Stephen M. de Bruyn Kops 1, Peter N. Blossey 2 and James J. Riley 3 · 2026
Read Evolution of passive scalar mixing layers in stratified and unstratified homogeneous turbulence by Stephen M. de Bruyn Kops 1, Peter N. Blossey 2 and James J. Riley 3 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-03843.txt.
Introduction
Stephen M. de Bruyn Kops, Peter N. Blossey, and James J. Riley simulate how passive scalars mix in stratified and unstratified turbulence. They use high-resolution large-eddy simulations of decaying flows. Two passive scalar mixing layers, one vertical and one transverse, model a plume much larger than the flow's length scale. Stable layering should suppress vertical mixing, and the simulations quantify that effect.
In the transverse direction, stratified and unstratified evolution are broadly similar, though the scalar spreads slightly faster and fluctuates more strongly when stratified, with a more broken-up interface. In the vertical direction, the stratified layer barely mixes: it grows early until its width tracks the horizontal velocity's vertical integral length, then nearly stops spreading.
Catalog record
| CREATOR | Stephen M. de Bruyn Kops 1, Peter N. Blossey 2 and James J. Riley 3 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.03843
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Evolution of passive scalar mixing layers in stratified and unstratified homogeneous turbulence Stephen M. de Bruyn Kops 1, Peter N. Blossey 2 and James J. Riley 3 Abstract High-resolution large-eddy simulations of decaying stratified and unstratified homogeneous turbulence are used to understand the mixing of passive scalars in stably stratified flows. Two passive scalar mixing layers, one in the vertical direction and the other in the transverse direction, are a model for a plume that is very large relative to the length scale of the velocity. In the transverse direction, the evolution of the passive scalar is broadly similar in the stratified and unstratified cases, although it does spread slightly faster when stratified. Also, the intensity of the scalar fluctuations is higher in the stratified case, and the turbulent/non-turbulent interface is more intermittent. In the vertical direction, though, the stratified case has almost no mixing because the stratification prevents large-scale stirring. Initially, the stratified passive layer grows until its width is proportional to the vertical integral length of the horizontal velocity, which is itself constrained to maintain the vertical Froude number order one. After this early growth, there is little additional spreading of the passive scalar. Modelling of the stratified scalar flux in the transverse direction is done effectively with a one-constant model if the mean profile is known, and a two-constant model if the profile shape must be assumed. In the latter case, the model is good only if the scalar is in quasi-equilibrium with the velocity field such that the length scale of the scalar can be scaled from the kinetic energy. In this study, the Prandtl number of the active and passive scalars is 0.7. It is anticipated that the reverse buoyancy flux resulting from higher Prandtl numbers will affect the passive scalar mixing. † † affiliation: 1 Department of Mechanical and Industrial Engineering, University of Massachusetts Amherst, Amherst, Massachusetts, USA 2 Department of Atmospheric and Climate Science, University of Washington, Seattle, Washington, USA 3 Department of Mechanical Engineering, University of Washington, Seattle, Washington, USA © 2026 by Stephen M. de Bruyn Kops, Peter N. Blossey and James J. Riley 1 Introduction The dispersion of particles in the atmosphere and oceans can be of great consequence in many situations. For example, it is often important to understand and predict the behaviour of aerosol particles (such as dust, soot, pollen and volcanic ash), cloud droplets and ice crystals in the upper tropopsphere and stratosphere. This is also true of the behaviour of pollutants, microplastics, heat, and various small organisms in the oceans. The understanding and prediction of the behaviour of these is made more difficult because the upper tropopsphere, stratosphere and the ocean below the mixed layer are stably stratified in density. The stable stratification has a strong effect on the turbulence and flow in these regions. For example, vertical motions are inhibited by the stable stratification, whereas the media can support the propagation of internal waves. Much has been learned about how turbulence with stable density stratification affects the dispersion and mixing of the density field itself ( 6 , see, e.g., the review of) , but much less on the dispersion of the various particles mentioned above. The dispersion of particles depends very much on how the particles have been introduced, for example as contrails from an airplane ( 54 , e.g.,) , or microplastics in a river outflow ( 16 , e.g.,) . A major advance in understanding and predicting the dispersion of particles in non stratified flows was made by Taylor (1921), who considered the dispersion of fluid particles in a homogeneous, isotropic, stationary turbulent flow. He related the turbulent diffusivity D to the particle velocity autocorrelation function and found that it asymptotes as D=\langle u^{2}\rangle T_{L} , where \langle u^{2}\rangle is the Lagrangian mean square velocity, and T_{L} is the Lagrangian velocity integral time scale. 9 followed Taylor’s approach and attempted to estimate the effect of stable stratification on the particle velocity autocorrelation and hence on Lagrangian dispersion. He found that the standard deviation of the vertical fluid particle displacement tends asymptotically to a constant value due to the suppression of the vertical motion by the stable stratification. This result was in qualitative agreement with some existing field data. 40 assumed a balance between the production and dissipation terms in the equation for the mean square temperature fluctuations in the ocean, and found that the temperature diffusivity D_{T} can be approximated as D_{T}=\frac{\epsilon_{T}}{\left({\partial\left<T\right>}\over{\partial z}\right)^{2}}\,, where \langle T\rangle is the mean temperature profile, and \epsilon_{T}=\kappa_{T}\langle\nabla T\cdot\nabla T\rangle is the mean square temperature dissipation rate. Based upon a result by 31 regarding vertical particle displacement in a stratified flow, 32 determined that the vertical eddy diffusivity for fluid particles D in a stable stratified fluid can be given as: D={\epsilon_{p}\over N^{2}}\,, where \epsilon_{p} is the dissipation rate of mean square buoyancy fluctuations, and N is the buoyancy frequency. Lindborg and Fedina were then able to demonstrate good quality in this model by comparing the results of the model with those from direct numerical simulations by considering the case of the vertical diffusion of a horizontal ‘slab’ of a scalar in a forced, statistically steady, stably-stratified turbulent fluid flow. 45 show that the assumptions of homogeneity and stationarity can be relaxed if density-sorted coordinates are used ( 58 ; 57 ) . When information about \epsilon_{p} is not available, the kinetic energy dissipation rate \epsilon_{k} can be used to estimate the vertical eddy diffusivity of heat: D={\Gamma\epsilon_{k}\over N^{2}}\,, where \Gamma=\epsilon_{p}/\epsilon_{k} is a mixing parameter, whose value is uncertain but taken to be 1/3 for stratospheric turbulence based on arguments about the flux Richardson number ( 30 ) . While vertical mixing is limited by stratification, constraints on horizontal mixing are much weaker. Estimated horizontal diffusivities in stratified turbulence exceed the vertical diffusivities by orders of magnitude, and their ratio may grow over time as the horizontal length and velocity scales grow ( 42 ; 51 ) . These anisotropic diffusivities are consistent with large horizontal-to-vertical aspect ratios of turbulent patches observed in the stably stratified upper tropsphere ( 43 , e.g.,) . In this study we make the continuum approximation and treat the assemblage of particles as a continuum. Furthermore, we assume that the particles act as fluid particles. We then choose a simple configuration for studying inhomogeneous turbulent mixing, a passive scalar initialised with a step change in the scalar value across the flow. Research starting in the 1970s suggests that the scalar statistics for a mixing layer in homogeneous isotropic turbulence (HIT) at high Péclet and Reynolds numbers can be expected to approach self-similarity in time with departure from self-similarity being consistent with finite Péclet number \mathit{Pe} , which introduces additional length and time scales into the flow ( 29 ; 26 ; 25 ; 34 ; 35 ; 13 ) . In short, the scalar statistics for a mixing layer in HIT can be well modelled in terms of the integral length and time scales of the turbulence. An open question is what happens when a passive scalar mixing layer (PSML) is introduced into HIT and subjected to a stabilising density gradient such that the buoyancy force becomes comparable to the inertial force