betaFeedback

WorksPhysics and Astronomy

Effective transport by 2D turbulence: Vortex-gas theory vs. scale-invariant inverse cascade

Abstract

The scale-invariant inverse energy cascade is a hallmark of 2D turbulence, with its theoretical energy spectrum observed in both direct numerical simulations (DNS) and laboratory experiments. Under this scale-invariance assumption, the effective diffusivity of a 2D turbulent flow is dimensionally controlled by the energy flux and the friction coefficient only. Surprisingly, however, we show that such scaling predictions are invalidated by numerical solutions of the 2D Navier-Stokes equation forced at intermediate wave number and damped by weak linear or quadratic drag. We derive alternate scaling-laws for the effective diffusivity based on the emergence of intense, isolated vortices causing spatially inhomogeneous frictional dissipation localized within the small vortex cores. The predictions quantitatively match DNS data. This study points to a universal large-scale organization of 2D turbulent flows in physical space, bridging standard 2D Navier-Stokes turbulence with large-scale geophysical turbulence.

Introduction

Many turbulent flows are effectively described by two-dimensional (2D) equations as a result of small aspect ratio, global rotation or external magnetic field [1, 2, 3, 4, 5, 6, 7, 8, 9, 10], in situations ranging from large-scale flows in oceans and planetary atmospheres [11, 12, 13, 14] to magneto-hydrodynamics [1, 15, 16, 17, 18, 19] and active matter systems [20, 21]. The phenomenology of 2D turbulence strongly departs from that of 3D turbulence. The conservation of both energy and enstrophy by 2D flows induces inverse energy transfers, from smaller to larger scales, together with forward enstrophy transfers, from larger to smaller scales. When energy is supplied at wavenumber kfk_f and removed at large scales by some damping mechanism, the Kraichnan-Leith-Batchelor (KLB) scaling theory predicts an inverse energy cascade characterized by an energy spectrum proportional to k−5/3k^{-5/3}, where kk denotes the wavenumber [22].

The k−5/3k^{-5/3} energy spectrum has been observed in both laboratory experiments and Direct Numerical Simulations (DNS) [8, 17, 23, 24, 25, 26], making the KLB theory standard textbook material [27, 28]. Beyond the sole energy spectrum, however, one is often interested in the large-scale organization of the flow and in its effective transport properties as it acts on a much larger-scale tracer distribution: given a large-scale background gradient of some passive tracer, what is the mean scalar flux induced by the 2D turbulent flow, or equivalently, what is the effective diffusivity of the turbulent flow? On the one hand, there is a vast literature on inertial-range statistics of concentration fluctuations in passive scalar turbulence [29, 30, 31, 32]. On the other hand, the effective diffusivity has been characterized only within the analytically tractable regime of laminar flows at moderate Péclet number [33, 34, 35, 36]. By contrast, the effective diffusivity of a fully turbulent 2D flow has received far less attention, although it plays a central role in the context of oceanic and atmospheric flows (see discussion section).

In the following we thus report a numerical and theoretical study on the effective diffusivity of 2D turbulent flows. The DNS flow is driven at a scale much smaller than the domain size. It is damped by a drag force, either linear or quadratic in velocity. Linear drag is motivated by the Hartmann friction of quasi-2D magnetohydrodynamic flows [17] and the Ekman friction of quasi-2D rapidly rotating flows [11, 12, 13]. To better model the turbulent drag force on the roughness elements of the Earth surface and Ocean floor, atmospheric and Ocean models often resort to a quadratic drag force instead, proportional to the local velocity squared [37, 38, 39].

The core assumptions of the KLB theory – a scale-invariant inverse-energy cascade – readily provide standard dimensional estimates à la Kolmogorov for the effective diffusivity of the flow: the latter is dimensionally constrained by the energy flux and the drag coefficient arresting the cascade only, while being independent of the forcing scale kf−1k_f^{-1} [38, 40]. We refer to the resulting scaling-laws for the effective diffusivity as the scale-invariant inverse cascade (SIIC) scaling predictions. Surprisingly, however, the present DNS data indicate that the effective diffusivity of the flow strongly departs from such SIIC dimensional estimates. We attribute the failure of the scale-invariance argument to the emergence of coherent vortices [41, 42, 43, 44]: the vortices come with two characteristic scales, the inter-vortex distance and the vortex-core radius, thus breaking the scale-invariance of the flow.

Also suggestive of the importance of coherent vortices is the fact that a successful vortex-based scaling theory for freely evolving 2D turbulence has been around for decades [45, 46]. Finally, Chang & Held noticed that the KLB dimensional estimates are not satisfied by the quasi-2D ‘baroclinic’ turbulence arising in models of large-scale ocean and atmospheric flows with quadratic drag [39]. In this context, the crucial role of large-scale vortices [47] is the starting point of a scaling theory put forward by Gallet & Ferrari [48], [GF in the following), but a connection to standard 2D Navier-Stokes turbulence has so far remained elusive. The goal of this Letter is to establish this connection through the development of a quantitative vortex-based theory for the eddy diffusivity of 2D Navier-Stokes turbulence with drag.

Forced 2D turbulence with drag.— We consider a two-dimensional incompressible flow u=(u,v)=(−ψy,ψx)\mathbf{u}=(u,v)=(-\psi_y,\psi_x) inside a doubly periodic domain [0,2πL]2[0,2\pi L]^2. The flow is driven at small scale by the curl of a body-force, f(x,y,t)f(x,y,t), peaked around a wavenumber kfk_f. The governing equation for the vorticity ζ=Δψ\zeta=\Delta\psi reads

∂tζ+J(ψ,ζ)=−D(ψ)+f(x,y,t)−νΔ4ζ,(1)\partial_t\zeta+J(\psi,\zeta)=-\mathcal{D}(\psi)+f(x,y,t)-\nu\Delta^4\zeta, \tag*{(1)}

where J(g,h)=gxhy−gyhxJ(g,h)=g_xh_y-g_yh_x denotes the Jacobian operator and ν\nu denotes the hyperviscosity coefficient. The drag term D(ψ)\mathcal{D}(\psi) corresponds to (the curl of) either a linear or a quadratic drag force. Denoting as κ\kappa (resp. μ\mu) the linear (resp. quadratic) drag coefficient, the drag term reads:

D(ψ)={κΔψlinear dragμ[(∣∇ψ∣ψx)x+(∣∇ψ∣ψy)y]quadratic drag.\mathcal{D}(\psi)= \begin{cases} \kappa\Delta\psi& \text{linear drag}\\ \mu\left[(\lvert\nabla\psi\rvert\psi_x)_x+(\lvert\nabla\psi\rvert\psi_y)_y\right] & \text{quadratic drag.} \end{cases}

We consider two small-scale forcing protocols. Following the most widely adopted setup in the literature [24, 25, 26, 38, 40], in a first suite of numerical runs f(x,y,t)f(x,y,t) is white-noise-in-time isotropic forcing inside a narrow band of wavenumbers centered around kfk_f. In a second suite of numerical runs f(y)f(y) is steady ‘Kolmogorov’ forcing proportional to cos⁡(kfy)\cos(k_fy) [49, 50, 51]. For the former forcing the mean energy injection rate per unit mass of fluid ϵ\epsilon is a control parameter of the system, while for the latter forcing ϵ\epsilon is an emergent parameter. Regardless, we use ϵ\epsilon to characterize the strength of the forcing in both cases.

To diagnose the effective diffusivity of the flow, we consider the advection of a passive tracer subject to a uniform background gradient [38, 40]. The tracer evolution equation being linear, without loss of generality we set the large-scale gradient to minus one and denote as τ(x,y,t)\tau(x,y,t) the doubly periodic departure from the background gradient. The evolution equation for τ\tau then reads

∂tτ+J(ψ,τ)=ψx−ντΔ4τ,(2)\partial_t\tau+J(\psi,\tau)=\psi_x-\nu_\tau\Delta^4\tau, \tag*{(2)}

where we employ the same hyperdiffusion coefficient for momentum and scalar concentration, ντ=ν\nu_\tau=\nu (see Supplemental Material for a Table of all parameter values). We perform DNS of equations (1) and (2) using a pseudospectral solver running on GPU with resolution up to 8192×81928192\times8192. Once the system has reached a statistically steady state, we extract ϵ\epsilon together with the effective diffusivity of the flow. Because the background tracer gradient is set to −1-1, the effective diffusivity equals the tracer flux, D=⟨ψxτ⟩D=\langle\psi_x\tau\rangle, where the angular brackets denote time and space average. The energy spectra, provided as End Matter, are very similar to the ones reported in previous studies [23, 24, 25, 26], with best-fit exponents close to −5/3-5/3 for k<kfk<k_f. We focus on the large-domain small-hyperviscosity limit where DD is independent of both LL and ν\nu (as assessed by running DNS with various LL and ν\nu). We illustrate this regime in Figure 1, where we show snapshots of the tracer departure field τ\tau, of the vorticity ζ\zeta and of the Okubo-Weiss parameter Q=ψxy2−ψxxψyyQ=\psi_{xy}^2-\psi_{xx}\psi_{yy} [52, 53].

Snapshots of the passive tracer departure field, vorticity field, and Okubo-Weiss parameter

Figure 1. Snapshots of the passive tracer departure field τ\tau (upper panels), of the vorticity field Δψ\Delta\psi (bottom left) and of the Okubo-Weiss parameter QQ (bottom right) for a typical 4096×40964096\times4096 run with quadratic drag μ/kf=5×10−4\mu/k_f=5\times10^{-4} and Kolmogorov forcing at kfL=400k_fL=400 (passive tracer values in red, negative values in blue, zero is white).

The inverse energy transfers are arrested by the drag term at a scale smaller than LL. This prevents the formation of a large-scale condensate and rules out any ‘anomalous diffusion’, as discussed in Ref. [36]. The effective diffusivity DD depends only on the injection wavenumber kfk_f, the energy input rate ϵ\epsilon and the drag coefficient κ\kappa or μ\mu. In dimensionless form, we thus seek the dependence of the dimensionless diffusivity D^=Dϵ−1/3kf4/3\widehat{D}=D\epsilon^{-1/3}k_f^{4/3} on the dimensionless drag coefficient κ^=κϵ−1/3kf−2/3\widehat{\kappa}=\kappa\epsilon^{-1/3}k_f^{-2/3} or μ^=μ/kf\widehat{\mu}=\mu/k_f.

Diffusivity according to scale invariance.— As discussed in Refs. [38, 40], a simple dimensional estimate for the effective diffusivity can be deduced from the KLB assumption of SIIC: as for any large-scale quantity, the effective diffusivity should depend on the energy flux ϵ\epsilon and on the friction coefficient arresting the inverse energy cascade, while being independent of the small forcing scale kf−1k_f^{-1}. Dimensional analysis then yields D∼ϵ/κ2D\sim\epsilon/\kappa^2 for linear drag and D∼ϵ1/3/μ4/3D\sim\epsilon^{1/3}/\mu^{4/3} for quadratic drag, which we recast in terms of the dimensionless quantities as [38]:

D^∼κ^−2,D^∼μ^−4/3,(3)\widehat{D}\sim\widehat{\kappa}^{-2},\qquad\widehat{D}\sim\widehat{\mu}^{-4/3}, \tag*{(3)}

for linear and quadratic drag, respectively.

We compare these predictions to the DNS data in Figure 2. Surprisingly, the numerical data depart from the SIIC predictions (3) for both linear and quadratic drag, as clearly illustrated by the compensated plots in the insets. In stark contrast with the assumption of complete scale-invariance, the diffusivity of the turbulent flow clearly retains some dependence on the small injection scale kf−1k_f^{-1}.

Dimensionless effective diffusivity as a function of dimensionless drag coefficient

Figure 2. Dimensionless effective diffusivity D^\hat{D} as a function of the dimensionless drag coefficient. Symbols are DNS data (+: white-noise forcing; ∘\circ: Kolmogorov forcing). In the left-hand panel we plot both D^\hat{D} and D^log⁡2D^\hat{D}\log^{2}\hat{D} as functions of κ^\hat{\kappa} to test the KLB prediction and the vortex-gas prediction, respectively. In the right-hand panel we plot the effective diffusivity as a function of the quadratic drag coefficient, together with the KLB prediction (dashed line) and the vortex-gas prediction (solid line). Inset: compensated plots using both the KLB prediction (red) and the present theory (gray).

Diffusivity according to vortex-gas dynamics.—A striking flow feature in physical space is the emergence of coherent vortices, as highlighted by the snapshot of QQ in Figure 1. With the goal of including them in a scaling theory, we model the coherent vortices using an idealized vortex-gas framework, in a similar fashion to GF [48, 54]. We consider a population of identical vortices of circulation ±Γ\pm\Gamma and core radius rr, the typical vorticity within the vortex cores being ζcore∼Γ/r2\zeta_{\mathrm{core}} \sim\Gamma/r^2. The vortex-gas is dilute, with a typical inter-vortex distance ℓ≫r\ell\gg r. The vortices wander around as a result of mutual induction with a typical velocity V∼Γ/ℓV \sim\Gamma/\ell, the latter being also the typical large-scale velocity in the inter-vortex region between vortices. Assuming that the transport properties of a single dipole of oppositely signed vortices correctly reflect transport within the vortex gas, GF show that:

D∼ℓV∼Γ.(4)D \sim\ell V \sim\Gamma. \tag*{(4)}

In the inter-vortex region, energy injection and inverse transfers proceed following the standard KLB phenomenology [44, 55]. We thus estimate the large-scale velocity VV at the large inter-vortex distance ℓ\ell based on the KLB expression for the velocity increment at scale ℓ\ell:

V∼(ϵℓ)1/3.(5)V \sim(\epsilon\ell)^{1/3}. \tag*{(5)}

Combining this relation with (4) leads to expressions for VV and ℓ\ell in terms of DD and ϵ\epsilon:

V∼D1/4ϵ1/4,ℓ∼D3/4ϵ−1/4.(6)V \sim D^{1/4}\epsilon^{1/4}, \qquad\ell\sim D^{3/4}\epsilon^{-1/4}. \tag*{(6)}

The last two scaling arguments are based, respectively, on the energy and enstrophy power integrals. The spatially intermittent nature of frictional damping comes into play through the energy power integral. Hyperviscous energy dissipation being negligible, the energy power integral reads ϵ=κ⟨u2⟩\epsilon= \kappa\langle\mathbf{u}^2\rangle for linear drag and ϵ=μ⟨∣u∣3⟩\epsilon= \mu\langle|\mathbf{u}|^3\rangle for quadratic drag. Following GF, we estimate the second and third moments of the vortex-gas velocity field by considering a single isolated Rankine vortex of core radius rr and circulation Γ\Gamma located at the center of a disk-shaped domain of radius ℓ\ell. Computing the space average of u2\mathbf{u}^2 and ∣u∣3|\mathbf{u}|^3 over the disk-shaped domain for this idealized velocity field yields ⟨u2⟩∼V2log⁡(ℓ/r)\langle\mathbf{u}^2\rangle\sim V^2\log(\ell/r) and ⟨∣u∣3⟩∼V3ℓ/r\langle|\mathbf{u}|^3\rangle\sim V^3\ell/r, which we substitute into the energy power integral to obtain:

ϵ∼κV2log⁡(ℓr),ϵ∼μV3ℓr.(7)\epsilon\sim\kappa V^2\log\left(\frac{\ell}{r}\right), \qquad\epsilon\sim\mu V^3\frac{\ell}{r}. \tag*{(7)}

for linear and quadratic drag, respectively. Finally, an estimate for the core radius rr can be deduced from the enstrophy power integral. Following the vortex-gas description of freely evolving turbulence in Refs. [46, 56], we adopt the picture of an isolated vortex being reinforced through continuous mergers with smaller intense vorticity structures [41]. This process leads to filamentation and enstrophy dissipation in the vicinity of the vortex core, a region of both strong vorticity and strong strain. The flux of enstrophy to small dissipative scales through this process is the product of the typical enstrophy in this region, ζcore2\zeta_{\mathrm{core}}^2, with the local strain rate Γ/r2∼ζcore\Gamma/r^2 \sim\zeta_{\mathrm{core}}. We thus estimate the enstrophy flux as ζcore3\zeta_{\mathrm{core}}^3 within the near-core regions, that is, over a fraction r2/ℓ2r^2/\ell^2 of the domain. The contribution of the near-core regions to the space-averaged enstrophy dissipation rate is thus estimated as ζcore3r2/ℓ2\zeta_{\mathrm{core}}^3 r^2/\ell^2. Demanding that the latter be less than or equal to the enstrophy injection rate kf2ϵk_f^2\epsilon leads to the following inequality for the vortex core radius:

r≲kfϵ1/2ℓζcore3/2.(8)r \lesssim\frac{k_f\epsilon^{1/2}\ell}{\zeta_{\mathrm{core}}^{3/2}}. \tag*{(8)}

As discussed in Ref. [46], isolated vortices within the vortex gas reinforce and expand through mergers, conserving energy and core vorticity, while dissipating enstrophy. They can do so as long as enstrophy dissipation by the mergers is smaller than enstrophy injection by the forcing. In the equilibrated state, we thus expect inequality (8) to be saturated. We thus replace the inequality (8) with an equality, before combining the resulting relation with (4) and (6), together with ζcore∼Γ/r2∼D/r2\zeta_{\mathrm{core}} \sim\Gamma/r^2 \sim D/r^2. This leads to the following estimate for the dimensionless core radius:

kfr∼D^3/8.(9)k_f r \sim\hat{D}^{3/8}. \tag*{(9)}

Starting from the energy power integral (7), one finally substitutes the estimates (6) and (9) for VV, ℓ\ell and rr. This leads to the following scaling predictions for the eddy diffusivity of the flow in terms of the dimensionless drag coefficient:

D^log⁡2D^∼κ^−2,D^∼μ^−8/9,(10)\hat{D}\log^2\hat{D} \sim\hat{\kappa}^{-2}, \qquad\hat{D} \sim\hat{\mu}^{-8/9}, \tag*{(10)}

for linear and quadratic drag, respectively. In contrast with the assumption of scale invariance, reverting to dimensional quantities indicates that the predicted eddy diffusivity (10) explicitly involves the forcing scale: Dlog⁡2[Dϵ−1/3kf4/3]∼ϵ/κ2D\log^2[D\epsilon^{-1/3}k_f^{4/3}] \sim\epsilon/\kappa^2 or D∼ϵ1/3kf−4/9μ−8/9D \sim\epsilon^{1/3}k_f^{-4/9}\mu^{-8/9}. As shown in Figure 2, the predictions (10) capture the numerical data with excellent accuracy in the low-drag regime, for both types of drag and both forcing protocols.

Discussion: connection to baroclinic turbulence.—Interestingly, the present scaling theory sheds new light on the phenomena of baroclinic turbulence. on traditional approaches to parameterize ‘baroclinic’ turbulence in oceans and atmospheres [39, 40, 57, 58]. Baroclinic turbulence is a form of quasi-2D turbulence driven by an instability mechanism that injects kinetic energy around a small spatial scale λ\lambda (the Rossby deformation radius [11, 12, 13, 59]). The 2D velocity field is coupled to a temperature field τ\tau, which has dimensions of a streamfunction. The latter is subject to a background meridional gradient denoted as UU in this context. The source of kinetic energy in baroclinic turbulence is the meridional transport of heat, hence a direct proportionality relation between the mean meridional heat flux and the mean kinetic energy dissipation rate of the flow, ϵ=DU2/λ2\epsilon= DU^{2}/\lambda^{2}. The sink of kinetic energy is linear or quadratic drag. The dimensional control parameters are UU, λ\lambda and the drag coefficient in this context, and scaling theories aim at relating the dimensionless diffusivity D∗=D/(Uλ)D_{*} = D/(U\lambda) to the dimensionless drag coefficient κ∗=κλ/U\kappa_{*} = \kappa\lambda/U or μ∗=μλ\mu_{*} = \mu\lambda. Although the equations governing baroclinic turbulence depart from the standard 2D Navier-Stokes equation, the traditional approach to deriving a scaling theory is based on a parallel between the two systems: one assumes that the forcing wavenumber is kf∼λ−1k_{f} \sim\lambda^{-1} and combines the relation ϵ=DU2/λ2\epsilon= DU^{2}/\lambda^{2} with estimates for DD based on the phenomenology of 2D turbulence. However, combining this relation with the KLB prediction (3) for quadratic drag leads to D∗∼1/μ∗2D_{*} \sim1/\mu_{*}^{2}, a prediction clearly invalidated by DNS of baroclinic turbulence [39, 48, 54, 60] (a similar failure arises for linear drag). In parallel, GF and later Hadjerci & Gallet ([54], HG in the following)

designed a purely vortex-based theory that captures the behavior of D∗D_{*} with drag κ∗\kappa_{*} or μ∗\mu_{*} with excellent accuracy. The puzzle remained why baroclinic turbulence would behave so differently from standard 2D Navier-Stokes turbulence. The present study solves this puzzle by insisting that emergent coherent vortices are, in fact, a crucial ingredient of standard 2D Navier-Stokes turbulence too. When the relations ϵ=DU2/λ2\epsilon= DU^{2}/\lambda^{2} and kf∼λ−1k_{f} \sim\lambda^{-1} are combined with the new scaling predictions (10) for the 2D Navier-Stokes diffusivity, one recovers the successful predictions of HG for the dependence of D∗D_{*} on κ∗\kappa_{*} or μ∗\mu_{*}.

Conclusion.— We have reported a suite of DNS of 2D Navier-Stokes turbulence with drag, focussing of the effective diffusivity of the flow. By reaching lower drag than previous studies [38, 40], we identified clear departures from the SIIC dimensional estimates. We introduced an alternate scaling theory that takes into account the coherent vortices that populate 2D turbulent flows. The vortices derail the SIIC scaling predictions because they induce localized frictional dissipation within the small vortex cores. The combination of broken scale-invariance and spatially intermittent energy dissipation is somewhat reminiscent of intermittency effects in the forward energy cascade of 3D turbulence. This parallel shall be drawn with care, however, as velocity structure functions are known to display no intermittency corrections in the inertial range of the 2D inverse energy cascade [23].

End matter: Energy spectra

In Figure 3 we plot the energy spectra extracted from DNS performed with white-noise forcing and both types of drag, together with a k−5/3k^{-5/3} eyeguide. As expected, the spectra are very similar to the ones reported in previous numerical studies with white-noise forcing and (linear) drag, see e.g. Ref. [25]. To extract a spectral slope in the inverse cascade range, we first introduce an energy-containing wavenumber kL2=⟨u2⟩/⟨ψ2⟩k_L^2 = \langle u^2\rangle/\langle\psi^2\rangle, before performing a power-law fit to each spectrum over the range k∈[kL,kf]k \in[k_L,k_f]. The resulting power-law exponent α\alpha is compared to the KLB value −5/3-5/3 in the inset of each panel, where we plot the ratio δ=α/(−5/3)\delta= \alpha/(-5/3). The best-fit exponent α\alpha is within 15% percents of the theoretical value −5/3-5/3 at low drag.

Energy spectra from sample simulations with linear and quadratic drag

Figure 3. Energy spectra from sample simulations with linear (top) and quadratic (bottom) drag, driven by white-noise-in-time forcing. The spectra are non-dimensionalized using ϵ\epsilon and kfk_f. The inset displays the ratio δ\delta of the best-fit exponent α\alpha to the KLB prediction −5/3-5/3.

References

  1. [1]J. Sommeria and R. Moreau. Why, how and when, mhd turbulence becomes two-dimensional. Journal of Fluid Mechanics, 118:507–518, 1982.DOI
  2. [2]P. A. Davidson. Turbulence in rotating, stratified and electrically conducting fluids. Cambridge University Press, 2013.DOI
  3. [3]E. Deusebio, G. Boffetta, E. Lindborg, and S. Musacchio. Dimensional transition in rotating turbulence. Physical Review E, 90(2):023005, 2014.DOI
  4. [4]B. Gallet and C. R. Doering. Exact two-dimensionalisation of low-magnetic-reynolds-number flows subject to a strong magnetic field. Journal of Fluid Mechanics, 773:154–177, 2015.DOI
  5. [5]B. Gallet. Exact two-dimensionalization of rapidly rotating large-reynolds-number flows. Journal of Fluid Mechanics, 783:412–447, 2015.DOI
  6. [6]S. J. Benavides and A. Alexakis. Critical transitions in thin layer turbulence. Journal of Fluid Mechanics, 822:364–385, 2017.
  7. [7]K. Seshasayanan and A. Alexakis. Condensates in rotating turbulent flows. Journal of Fluid Mechanics, 841:434–462, 2018.DOI
  8. [8]A. Alexakis and L. Biferale. Cascades and transitions in turbulent flows. Physics Reports, 767-769:1–101, 2018. Cascades and transitions in turbulent flows.DOI
  9. [9]A. van Kan and A. Alexakis. Condensates in thin-layer turbulence. Journal of Fluid Mechanics, 864:490–518, 2019.DOI
  10. [10]K. Seshasayanan and B. Gallet. Onset of three-dimensionality in rapidly rotating turbulent flows. Journal of Fluid Mechanics, 901:R5, 2020.DOI
  11. [11]R. Salmon. Lectures on geophysical fluid dynamics. Oxford University Press, USA, 1998.DOI
  12. [12]J. Pedlosky. Geophysical fluid dynamics. Springer Science & Business Media, 2013.
  13. [13]G. K. Vallis. Atmospheric and oceanic fluid dynamics. Cambridge University Press, 2017.
  14. [14]L. Miller, B. Deremble, and A. Venaille. Gyre turbulence: Anomalous dissipation in a two-dimensional ocean model. Physical Review Fluids, 9(5):L051801, 2024.
  15. [15]C. F. Driscoll and K. S. Fine. Experiments on vortex dynamics in pure electron plasmas. Physics of Fluids B: Plasma Physics, 2(6):1359–1366, 1990.DOI
  16. [16]C. F. Driscoll, D. Z. Jin, D. A. Schecter, and D. H. E. Dubin. Vortex dynamics of 2d electron plasmas. Physica C: Superconductivity, 369(1-4):21–27, 2002.DOI
  17. [17]J. Sommeria. Experimental study of the two-dimensional inverse energy cascade in a square box. Journal of fluid mechanics, 170:139–168, 1986.DOI
  18. [18]B. Gallet, J. Herault, C. Laroche, F. Pétrélis, and S. Fauve. Reversals of a large-scale field generated over a turbulent background. Geophysical & Astrophysical Fluid Dynamics, 106(4-5):468–492, 2012.DOI
  19. [19]G. Michel, J. Herault, F. Pétrélis, and S. Fauve. Bifurcations of a large-scale circulation in a quasi-bidimensional turbulent flow. Europhysics Letters, 115(6):64004, 2016.
  20. [20]B. Martínez-Prat, R. Alert, J. Ignés-Mullol F. Meng, J-F Joanny, J. Casademunt, R. Golestanian, and F. Sagués. Scaling regimes of active turbulence with external dissipation. Physical Review X, 11(3):031065, 2021.
  21. [21]D. Bárdfalvy, and C. Nardini V. Skultéty, A. Morozov, and J. Stenhammar. Collective motion in a sheet of microswimmers. Communications Physics, 7(1):93, 2024.DOI
  22. [22]R. H. Kraichnan. Inertial ranges in two-dimensional turbulence. Physics of fluids, 10(7):1417, 1967.DOI
  23. [23]G. Boffetta, A. Celani, and M. Vergassola. Inverse energy cascade in two-dimensional turbulence: Deviations from gaussian behavior. Physical Review E, 61(1):R29, 2000.pubmed.ncbi.nlm.nih.gov/11046365
  24. [24]S. Chen, R. E. Ecke, G. L. Eyink, M. Rivera, M. Wan, and Z. Xiao. Physical mechanism of the two-dimensional inverse energy cascade. Physical review letters, 96(8):084502, 2006.pubmed.ncbi.nlm.nih.gov/16606186
  25. [25]G. Boffetta and S. Musacchio. Evidence for the double cascade scenario in two-dimensional turbulence. Physical Review E—Statistical, Nonlinear and Soft Matter Physics, 82(1):016307, 2010.pubmed.ncbi.nlm.nih.gov/20866725
  26. [26]G. Boffetta and R. E. Ecke. Two-dimensional turbulence. Annual review of fluid mechanics, 44:427–451, 2012.DOI
  27. [27]M. Lesieur. Introduction to turbulence in fluid mechanics. Turbulence in Fluids: Fourth Revised and Enlarged Edition, pages 1–23, 2008.
  28. [28]U. Frisch. Turbulence: the legacy of AN Kolmogorov. Cambridge university press, 1995.
  29. [29]R. H. Kraichnan. Anomalous scaling of a randomly advected passive scalar. Physical review letters, 72(7):1016, 1994.pubmed.ncbi.nlm.nih.gov/10056596
  30. [30]A. Celani and M. Vergassola. Statistical geometry in scalar turbulence. Physical review letters, 86(3):424, 2001.pubmed.ncbi.nlm.nih.gov/11177846
  31. [31]J. Cardy, G. Falkovich, and K. Gawedzki. Non-equilibrium Statistical Mechanics and Turbulence. London Mathematical Society Lecture Note Series. Cambridge University Press, 2008.DOI
  32. [32]B. I. Shraiman and E. D. Siggia. Scalar turbulence. Nature, 405(6787):639–646, 2000.pubmed.ncbi.nlm.nih.gov/10864314
  33. [33]G. I. Sivashinsky. Weak turbulence in periodic flows. Physica D: Nonlinear Phenomena, 17(2):243–255, 1985.DOI
  34. [34]M. N. Rosenbluth, H. L. Berk, I. Doxas, and W. Horton. Effective diffusion in laminar convective flows. The Physics of fluids, 30(9):2636–2647, 1987.
  35. [35]B. Dubrulle and U. Frisch. Eddy viscosity of parity-invariant flow. Physical Review A, 43(10):5355, 1991.
  36. [36]L. Biferale, Andrea Crisanti, M Vergassola, and Angelo Vulpiani. Eddy diffusivities in scalar transport. Physics of Fluids, 7(11):2725–2734, 1995.DOI
  37. [37]J. R. Holton and G. J. Hakim. An introduction to dynamic meteorology, volume 88. Academic press, 2013.
  38. [38]N. Grianiк, I. M. Held, K. S. Smith, and G. K. Vallis. The effects of quadratic drag on the inverse cascade of two-dimensional turbulence. Physics of Fluids, 16(1):73–78, 01 2004.DOI
  39. [39]C. Y. Chang and I. M. Held. The control of surface friction on the scales of baroclinic eddies in a homogeneous quasigeostrophic two-layer model. Journal of the Atmospheric Sciences, 76(6):1627 – 1643, 2019.DOI
  40. [40]K. S. Smith, G. Boccaletti, C. C. Henning, I. Marinov, C. Y. Tam, I. M. Held, and G. K. Vallis. Turbulent diffusion in the geostrophic inverse cascade. Journal of Fluid Mechanics, 469:13–48, 2002.DOI
  41. [41]V. Borue. Inverse energy cascade in stationary two-dimensional homogeneous turbulence. Physical review letters, 72(10):1475, 1994.pubmed.ncbi.nlm.nih.gov/10055618
  42. [42]S. Danilov and D. Gurarie. Nonuniversal features of forced two-dimensional turbulence in the energy range. Physical Review E, 63(2):020203, 2001.pubmed.ncbi.nlm.nih.gov/11308456
  43. [43]R. K. Scott. Nonrobustness of the two-dimensional turbulent inverse cascade. Physical Review E—Statistical, Nonlinear and Soft Matter Physics, 75(4):046301, 2007.pubmed.ncbi.nlm.nih.gov/17500987
  44. [44]B. H. Burgess, D. G. Dritschel, and R. K. Scott. Vortex scaling ranges in two-dimensional turbulence. Physics of Fluids, 29(11), 2017.
  45. [45]J. C. McWilliams. The vortices of two-dimensional turbulence. Journal of Fluid mechanics, 219:361–385, 1990.DOI
  46. [46]G. F. Carnevale, J. C. McWilliams, Y. Pomeau, J. B. Weiss, and W. R. Young. Evolution of vortex statistics in two-dimensional turbulence. Phys. Rev. Lett., 66:2735–2737, May 1991.pubmed.ncbi.nlm.nih.gov/10043603
  47. [47]A. F. Thompson and W. R. Young. Scaling baroclinic eddy fluxes: Vortices and energy balance. Journal of physical oceanography, 36(4):720–738, 2006.
  48. [48]B. Gallet and R. Ferrari. The vortex gas scaling regime of baroclinic turbulence. Proceedings of the National Academy of Sciences, 117(9):4491–4497, 2020.DOI
  49. [49]J. M. Burgess, C. Bizon, W. D. McCormick, J. B. Swift, and H. L. Swinney. Instability of the Kolmogorov flow in a soap film. Physical Review E, 60(1):715, 1999.pubmed.ncbi.nlm.nih.gov/11969812
  50. [50]M. Rivera and X-L Wu. External dissipation in driven two-dimensional turbulence. Physical review letters, 85(5):976, 2000.pubmed.ncbi.nlm.nih.gov/10991453
  51. [51]Y. K. Tsang and W. R. Young. Forced-dissipative two-dimensional turbulence: A scaling regime controlled by drag. Phys. Rev. E, 79:045308, Apr 2009.pubmed.ncbi.nlm.nih.gov/19518293
  52. [52]A. Okubo. Horizontal dispersion of floatable particles in the vicinity of velocity singularities such as convergences. In Deep sea research and oceanographic abstracts, volume 17(3), pages 445–454. Elsevier, 1970.DOI
  53. [53]J. Weiss. The dynamics of enstrophy transfer in two-dimensional hydrodynamics. Physica D: Nonlinear Phenomena, 48(2-3):273–294, 1991.DOI
  54. [54]G. Hadjerci and B. Gallet. Vortex core radius in baroclinic turbulence: Implications for scaling predictions. Phys. Rev. Fluids, 8:094501, Sep 2023.arxiv.org/abs/2308.15398
  55. [55]J. Wang, J. Sesterhenn, and W. C. Müller. Coherent structure detection and the inverse cascade mechanism in two-dimensional navier–stokes turbulence. Journal of Fluid Mechanics, 963:A28, 2023.DOI
  56. [56]J. B. Weiss. Punctuated hamiltonian models of structured turbulence. Semi-Analytic Methods for the Navier–Stokes Equations (Montreal, Canada, 1995)(ed. K. Coughlin). CRM Proc. Lecture Notes, 20:109–119, 1999.DOI
  57. [57]I. M. Held. The macroturbulence of the troposphere. Tellus A, 51(1):59–70, 1999.
  58. [58]V. D. Larichev and I. M. Held. Eddy amplitudes and fluxes in a homogeneous model of fully developed baroclinic instability. Journal of Physical Oceanography, 25(10):2285 – 2297, 1995.doi.org/10.1175/1520-0485(1995)025<2285:eaafia>2.0.co;2
  59. [59]B. Gallet, B. Miquel, G. Hadjerci, K. J. Burns, G. R. Flierl, and R. Ferrari. Transport and emergent stratification in the equilibrated eddy model: the vortex-gas scaling regime. Journal of Fluid Mechanics, 948:A31, 2022.DOI
  60. [60]G. Hadjerci and B. Gallet. Two-layer baroclinic turbulence with arbitrary layer depths. Physical Review Fluids, 9(5):L051802, 2024.arxiv.org/abs/2406.02564

Paper details

vlnc:W_vwyf5yfr
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents