betaFeedback

WorksMathematics

Liouville-type results for time-dependent stratified water flows over variable bottom in the ββ-plane approximation

Abstract

We consider here time-dependent three-dimensional stratified geophysical water flows of finite depth over a variable bottom with a free surface and an interface (separating two layers of constant and different densities). Under the assumption that the vorticity vectors in the two layers are constant, we prove that bounded solutions to the three-dimensional water waves equations in the ββ-plane approximation exist if and only if one of the horizontal components of the velocity, as well as its vertical component, are zero; the other horizontal component being constant. Moreover, the interface is flat, the free surface has a traveling character in the horizontal direction of the nonvanishing velocity component, being of general type in the other horizontal direction, and the pressure is hydrostatic in both layers. Unlike previous studies of three-dimensional flows with constant vorticity in each layer, we consider a non-flat bottom boundary and different constant vorticity vectors for the upper and lower layer.

Introduction

Geophysical fluid dynamics (GFD) is the study of fluid motion characterized by the incorporation of Coriolis effects in the governing equations. The Coriolis force is a result of the Earth’s rotation and plays a substantial role in the resulting dynamics. While the nonlinear GFD equations are able to capture a plethora of oceanic and atmospheric flows [12, 13, 14, 21, 22, 23, 24] their high level of difficulty greatly challenges the available mathematical techniques.

The possible course of action is the recourse to simpler approximate models that are justified by oceanographical considerations. One of these approximations refers to the linearization of Coriolis forces in the tangent plane approximation, a procedure that-despite the spherical shape of the Earth-is valid due to the moderate spatial scale of the motion: the region occupied by the fluid can be approximated by a tangent plane and the linear term of the Taylor expansion captures the β\beta-plane effect, cf. the discussions in Constantin [9], Cushman-Roisin & Beckers [26], Gill [38], Pedlosky [63], Salmon [66]. The paper [9], by Constantin, presented for the first time three-dimensional explicit and exact solutions (in Lagrangian coordinates) to the equatorial β\beta-plane model. These solutions described equatorially trapped waves propagating eastward in a stratified inviscid flow. The latter β\beta-plane model was modified to incorporate centripetal terms by Constantin & Johnson [13] whereby the authors also established the existence of equatorial purely azimuthal solutions to the GFD equations in spherical, cylindrical and β\beta-plane coordinates. A further significant extension was realized by Henry [40] who presented an exact and explicit solution (to the GFD equations in the β\beta-plane approximation with Coriolis and centripetal forces) representing equatorially trapped waves propagating in the presence of a constant background current. Yet, another improvement in the realm of the GFD equations in the β\beta-plane was obtained by Henry [41] and concerns the addition of a gravitational-correction term in the tangent plane approximation. Historically, this type of approximation was proposed by Rossby et al. [65] as a conceptual model for motion on a sphere. Recent results concerning flows in the β\beta-plane approximation were obtained in [28, 31]. For subsequent solutions to the GFD equations concerning a variety of geophysical scenarios, we refer the reader to [2, 10, 11, 16, 19, 42, 45, 46, 55, 58, 60, 61, 62].

In the quest to derive explicit and exact solutions to the GFD equations in the β\beta-plane approximation with Coriolis and centripetal terms, we start from the assumption that the vorticity vector is constant in each of the two layers of the fluid domain which is assumed to be stratified: the water flow, bounded below by a bottom (varying) boundary and above by the free surface, is split by an interface into a layer adjacent to the bottom of some constant density (say ρ\rho) which sits below the layer adjacent to the free surface of constant density ρˉ<ρ\bar{\rho}<\rho. Stratification is an important aspect in the ocean science: stratified layers act as a barrier to the mixing of water, which impacts the exchange of heat, carbon, oxygen and other nutrients, cf. [48]. The discontinuous stratification (of the type we consider here) gives rise to internal waves, an aspect that has attracted much attention lately from the perspective of exact solutions describing large-scale geophysical (and non-geophysical) flows [1, 12, 13, 14, 17, 18, 32, 33, 34, 35, 39, 42, 58] or of qualitative studies of intricate features underlying the dynamics of coupled surface and internal waves, cf. [43].

The structural consequences of constant vorticity in water flows satisfying the three-dimensional equations were discussed in a handful of papers by Constantin & Kartashova [5], Constantin [7], Craig [25], Wahlén [70], Stuhlmeier [67], Martin [53, 54, 56, 57]: the main outcome is that occurrence of constant vorticity in a flow that satisfies the three-dimensional nonlinear governing equations is possible if and only if the flow is two-dimensional and if the vorticity vector has only one non-vanishing component that points in the horizontal direction orthogonal to the direction of wave propagation. In particular, constant vorticity gives a good description of tidal currents; cf. [64]. These are the most regular and predictable currents, and on areas of the continental shelf and in many coastal inlets they are the most significant currents; cf. [50].

After introducing the governing equations we state and prove a Liouville-type result in the context of a two-layer fluid domain bounded below by a bottom boundary z=b(y)z = b(y) (for some given function (x,y)↦b(y)(x,y) \mapsto b(y)) and above by the free surface z=η~(x,y,t)z = \widetilde{\eta}(x,y,t) (for some unknown function η~\widetilde{\eta}), and split by an interface z=η(x,y,t)z = \eta(x,y,t) (for some unknown function η\eta) into two layers. We prove that if the vorticity vectors in the two layers are constant then bounded solutions to the three-dimensional water waves equations in the β\beta-plane approximation (with the associated boundary conditions) exist if and only if one of the horizontal components of the velocity, as well as its vertical component, are zero; the other horizontal component being constant. Moreover, the interface is flat, the free surface has a traveling character in the horizontal direction of the nonvanishing velocity component, being of general type in the other horizontal direction, and the pressure is hydrostatic in both layers. We would like to underline that allowing for an extra xx-dependence in the bottom function bb, leads only to unbounded solutions, cf. Remark 2.3.

We would also like to note that, unlike previous Liouville-type results, [4, 5, 7, 53, 54, 56, 57, 59, 67, 70], our analysis here does not require the bottom boundary to be flat. It is also worth to note that results similar in outcome, but for the Euler- or Navier-Stokes equations without a free boundary, were obtained recently and relatively recently, cf. [36, 37]. The ethos of the previously mentioned studies is that under integrability conditions or conditions concerning the mean oscillation of the velocity field, the latter vanishes identically, or that it displays less complexity.

The three dimensional water wave problem in the β\beta-plane approximation

We choose to work in a rotating framework with the origin at a point on the Earth’s surface which is approximated by a sphere of radius R=6378R = 6378 km and denote with (x,y,z)(x,y,z) the Cartesian coordinates where the spatial variable xx refers to the longitude, the variable yy to latitude, and the variable zz stands for the local vertical, cf. Fig. 1.

The fluid domain, bounded below by a bottom boundary z=b(x,y)z = b(x,y) (for some differentiable function (x,y)↦b(x,y)(x,y) \mapsto b(x,y)) and above by the free surface z=η~(x,y,t)z = \tilde{\eta}(x,y,t) is split by an interface, denoted z=η(x,y,t)z = \eta(x,y,t), into two layers: a layer adjacent to the bottom, written as

Dη(t):={(x,y,z):b(x,y)≤z≤η(x,y,t)},D_{\eta}(t) := \{(x,y,z) : b(x,y) \le z \le\eta(x,y,t)\},

which sits below a layer adjacent to the surface, written as

Dη,η~(t):={(x,y,z):η(x,y,t)≤z≤η~(x,y,t)}.D_{\eta,\tilde{\eta}}(t) := \{(x,y,z) : \eta(x,y,t) \le z \le\tilde{\eta}(x,y,t)\}.

We denote with u(x,y,z,t)=(u(x,y,z,t),v(x,y,z,t),w(x,y,z,t))\mathbf{u}(x,y,z,t) = (u(x,y,z,t),v(x,y,z,t),w(x,y,z,t)) (respectively u~(x,y,z,t)=(u~(x,y,z,t),v~(x,y,z,t),w~(x,y,z,t))\tilde{\mathbf{u}}(x,y,z,t) = (\tilde{u}(x,y,z,t),\tilde{v}(x,y,z,t),\tilde{w}(x,y,z,t))) the velocity field in Dη(t)D_{\eta}(t) (respectively in Dη,η~(t)D_{\eta,\tilde{\eta}}(t)), with P(x,y,z,t)P(x,y,z,t) (respectively P~(x,y,z,t)\tilde{P}(x,y,z,t)) the pressure in Dη(t)D_{\eta}(t) (respectively in Dη,η~D_{\eta,\tilde{\eta}}), and with gg the gravitational acceleration. We denote with ρ\rho the density in the lower layer DηD_{\eta} and with ρ~\tilde{\rho} the density in the upper layer Dη,η~D_{\eta,\tilde{\eta}}. Then the motion of incompressible and inviscid three-dimensional water flows in the β\beta-plane approximation near the Equator, (e.g. Constantin [10], Constantin & Johnson [13] and Dellar [29]), is governed in DηD_{\eta} by the Euler equations

ut+uux+vuy+wuz+2ωw−βyv=−Pxρ,vt+uvx+vvy+wvz+βyu+ω2y=−Pyρ,wt+uwx+vwy+wwz−2ωu−ω2R=−Pzρ−g.(1)\begin{aligned} u_t + uu_x + vu_y + wu_z + 2\omega w - \beta yv &= -\frac{P_x}{\rho},\\ v_t + uv_x + vv_y + wv_z + \beta yu + \omega^2 y &= -\frac{P_y}{\rho},\\ w_t + uw_x + vw_y + ww_z - 2\omega u - \omega^2 R &= -\frac{P_z}{\rho} - g. \tag*{(1)} \end{aligned}

and by the incompressibility condition

ux+vy+wz=0.(2)u_x + v_y + w_z = 0. \tag*{(2)}

Above, tt denotes the time variable, ω=7.29⋅10−5 rad s−1\omega= 7.29 \cdot10^{-5}\ \mathrm{rad}\ \mathrm{s}^{-1} is the rotational speed of Earth round the polar axis toward east and β:=2ωR\beta:= \frac{2\omega}{R}.

Image not extracted

Figure 1. The rotating frame of reference, with the x axis chosen horizontally due east, the y axis chosen horizontally due north, and the z axis chosen upward: x corresponds to longitude, y to latitude, and z to the local vertical.

Remark 2.1. The β\beta-plane effect, captured by the quantity βy\beta y, appears in the first and second equation in (1), and is a result of linearizing the Coriolis force in the tangent plane approximation. In spite of the spherical shape of the Earth, the before-mentioned linearization procedure is legitimate due to the moderate spatial scale of the motion, cf. the discussions in Cushman-Roisin & Beckers [26] and Constantin [9]. We would like to note that the β\beta-plane equations (1) represent a consistent approximation to the governing equations only near the Equator, cf. Dellar [29].

Likewise, in Dη,η~D_{\eta,\tilde{\eta}} the flow motion obeys the equations

u~t+u~u~x+v~u~y+w~u~z+2ωw~−βyv~=−P~xρ~v~t+u~v~x+v~v~y+w~v~z+βyu~+ω2y=−P~yρ~w~t+u~w~x+v~w~y+w~w~z−2ωu~−ω2R=−P~zρ~−g(3)\begin{aligned} \tilde{u}_t+\tilde{u}\tilde{u}_x+\tilde{v}\tilde{u}_y+\tilde{w}\tilde{u}_z+2\omega\tilde{w}-\beta y\tilde{v}&=-\frac{\tilde{P}_x}{\tilde{\rho}}\\ \tilde{v}_t+\tilde{u}\tilde{v}_x+\tilde{v}\tilde{v}_y+\tilde{w}\tilde{v}_z+\beta y\tilde{u}+\omega^2 y&=-\frac{\tilde{P}_y}{\tilde{\rho}}\\ \tilde{w}_t+\tilde{u}\tilde{w}_x+\tilde{v}\tilde{w}_y+\tilde{w}\tilde{w}_z-2\omega\tilde{u}-\omega^2 R&=-\frac{\tilde{P}_z}{\tilde{\rho}}-g \tag*{(3)} \end{aligned}

and

u~x+v~y+w~z=0.(4)\tilde{u}_x+\tilde{v}_y+\tilde{w}_z=0. \tag*{(4)}

We specify now the boundary conditions which conclude the formulation of the water wave problem. We start with the kinematic boundary conditions, stating the impermeability of the boundaries: on the free surface z=η~(x,y,t)z=\tilde{\eta}(x,y,t) we require

w~=η~t+u~η~x+v~η~yonz=η~(x,y,t)(5)\tilde{w}=\tilde{\eta}_t+\tilde{u}\tilde{\eta}_x+\tilde{v}\tilde{\eta}_y \quad\text{on}\quad z=\tilde{\eta}(x,y,t) \tag*{(5)}

On the interface z=η(x,y,t)z=\eta(x,y,t) the conditions

w~=ηt+u~ηx+v~ηyonz=η(x,y,t)w=ηt+uηx+vηyonz=η(x,y,t)(6)\begin{aligned} \tilde{w}&=\eta_t+\tilde{u}\eta_x+\tilde{v}\eta_y \quad\text{on}\quad z=\eta(x,y,t)\\ w&=\eta_t+u\eta_x+v\eta_y \quad\text{on}\quad z=\eta(x,y,t) \tag*{(6)} \end{aligned}

while on the bed it holds that

w=ubx+vbyonz=b(x,y).(7)w=ub_x+vb_y \quad\text{on}\quad z=b(x,y). \tag*{(7)}

The balance of forces at the interface is encoded in the continuity of the pressure across z=η(x,y,t)z=\eta(x,y,t), that is

P(x,y,η(x,y,t),t)=P~(x,y,η(x,y,t),t)for all x,y,t.(8)P(x,y,\eta(x,y,t),t)=\tilde{P}(x,y,\eta(x,y,t),t)\quad\text{for all }x,y,t. \tag*{(8)}

Lastly, the dynamic boundary condition, states the continuity of the pressure across the free surface, that is, we require that

P~=p(x+ωR2t,y)onz=η~(x,y,t),(9)\tilde{P}=p\left(x+\frac{\omega R}{2}t,y\right)\quad\text{on}\quad z=\tilde{\eta}(x,y,t), \tag*{(9)}

for some given differentiable function (X,Y)↦p(X,Y)(X,Y)\mapsto p(X,Y). The local rotation in the flow is captured by the vorticity vector, defined as the curl of the velocity field, that is,

Ω=(wy−vz,uz−wx,vx−uy)=:(Ω1,Ω2,Ω3) in Dη,\Omega=(w_y-v_z,u_z-w_x,v_x-u_y)=:(\Omega_1,\Omega_2,\Omega_3)\text{ in }D_\eta,
Ω~=(w~y−v~z,u~z−w~x,v~x−u~y)=:(Ω~1,Ω~2,Ω~3) in Dη,η~.(10)\tilde{\Omega}=(\tilde{w}_y-\tilde{v}_z,\tilde{u}_z-\tilde{w}_x,\tilde{v}_x-\tilde{u}_y)=:(\tilde{\Omega}_1,\tilde{\Omega}_2,\tilde{\Omega}_3)\text{ in }D_{\eta,\tilde{\eta}}. \tag*{(10)}

Remark 2.2. The (piecewise) constant vorticity is instrumental in describing wave-current interactions in sheared flows [49, 50, 64, 69]. For a comprehensive treatment of two-dimensional water flows with discontinuous vorticity (from the point of view of exact solutions describing waves of small and large amplitudes) we refer the reader to the work by Constantin & Strauss [8]. However, the landscape of rotational three-dimensional water flows is much less understood. Nevertheless, it is known that in three-dimensional flows the constant vorticity significantly steers the dimensionality of the velocity field and of the pressure [5, 7, 53, 54, 56, 57, 67, 70]. Compared with previous studies on rotational three-dimensional water flows, we move here one step further and assume that the vorticity vectors, Ω\Omega and Ω~\widetilde{\Omega}, respectively, are constant. As we shall see, this assumption will have massive consequences on the velocity field.

We are now ready to state the main result for whose proof we will rely on rather direct partial differential equations methods, unlike more sophisticated Hamiltonian tools (and other structure-preserving methods) used recently in the context of layered domains [3, 15, 18, 71, 72, 73], which are not known to be available in our three-dimensional setting of the β\beta-plane.

Theorem 2.1. If the velocity field satisfies (u,v,w)∈L∞(Dη)3(u,v,w) \in L^\infty(D_\eta)^3 and (u~,v~,w~)∈L∞(Dη,η~)3(\widetilde{u},\widetilde{v},\widetilde{w}) \in L^\infty(D_{\eta,\widetilde{\eta}})^3 then the flow has constant vorticity Ω∈Dη\Omega\in D_\eta and Ω~∈Dη,η~\widetilde{\Omega} \in D_{\eta,\widetilde{\eta}} if and only if

u(x,y,z,t)=u(t),v(x,y,z,t)=0,w(x,y,z,t)=w(t),for all x,y,z,t,u(x,y,z,t)=u(t),\quad v(x,y,z,t)=0,\quad w(x,y,z,t)=w(t),\quad\text{for all }x,y,z,t,
u~(x,y,z,t)=u~(t),v~(x,y,z,t)=0,w~(x,y,z,t)=w~(t),for all x,y,z,t,\widetilde{u}(x,y,z,t)=\widetilde{u}(t),\quad\widetilde{v}(x,y,z,t)=0,\quad\widetilde{w}(x,y,z,t)=\widetilde{w}(t),\quad\text{for all }x,y,z,t,

that is, u,u~,w,w~u,\widetilde{u},w,\widetilde{w} depend only on tt, while vv and v~\widetilde{v} vanish.

If, in addition, the bottom defining function bb depends only on yy, and PP and P~\widetilde{P} are bounded in Ω\Omega, and Ω~\widetilde{\Omega}, respectively, then

u=u~=−ωR2andv=w=v~=w~=0.u=\widetilde{u}=-\frac{\omega R}{2}\quad\text{and}\quad v=w=\widetilde{v}=\widetilde{w}=0.

If, moreover, ρ≠ρ~\rho\ne\widetilde{\rho} then

η(x,y,t)=0for all x,y,t,\eta(x,y,t)=0\quad\text{for all }x,y,t,

and

η~(x,y,t)=−p(x+ωR2t,y)ρ~g,\widetilde{\eta}(x,y,t)=-\frac{p\left(x+\frac{\omega R}{2}t,y\right)}{\widetilde{\rho}g},

where pp is the given function from (9).

Proof. Eliminating the pressure between the three equations in (1) yields the vorticity equation

Ω1ux+(Ω2+2ω)uy+(Ω3+βy)uz=0,Ω1vx+(Ω2+2ω)vy+(Ω3+βy)vz=0,Ω1wx+(Ω2+2ω)wy+(Ω3+βy)wz=βv,(11)\begin{aligned} \Omega_1u_x+(\Omega_2+2\omega)u_y+(\Omega_3+\beta y)u_z&=0,\\ \Omega_1v_x+(\Omega_2+2\omega)v_y+(\Omega_3+\beta y)v_z&=0,\\ \Omega_1w_x+(\Omega_2+2\omega)w_y+(\Omega_3+\beta y)w_z&=\beta v, \tag*{(11)} \end{aligned}

cf. eg. [6, 51, 52]. Applying the operator Δ\Delta to the first equation above and using that uxu_x, uyu_y and uzu_z are harmonic functions we find that Δ(yuz)=0\Delta(yu_z)=0. The latter equation can be expanded as

yΔuz+2uyz=0y\Delta u_z+2u_{yz}=0

which leads to

uyz≡0.(12)u_{yz}\equiv0. \tag*{(12)}

Following the same line of proof we conclude that

vyz(x,y,z,t)=wyz(x,y,z,t)=0(13)v_{yz}(x,y,z,t)=w_{yz}(x,y,z,t)=0 \tag*{(13)}

at all points (x,y,z)(x,y,z) of DηD_\eta. Utilizing the definitions of Ω1\Omega_1 and Ω2\Omega_2 and recalling (13) we obtain

wyx=wyy=wyz=0,(14)w_{yx}=w_{yy}=w_{yz}=0, \tag*{(14)}

which proves that wyw_y is constant throughout DηD_\eta. That is, there is a differentiable function t→f(t)t\to f(t) such

wy=f(t),w_y=f(t),

from which we infer the existence of another differentiable function t→g(t)t\to g(t) such that

vz=g(t).(15)v_z=g(t). \tag*{(15)}

We apply now the operator of differentiation with respect to yy in the vorticity equation (11) and obtain, via (14) that

wz=vy within Dη.(16)w_z=v_y\ \text{within }D_\eta. \tag*{(16)}

We now see from above that the equalities

wzz=vyz=0andvyy=wyz=0(17)w_{zz}=v_{yz}=0\quad\text{and}\quad v_{yy}=w_{yz}=0 \tag*{(17)}

hold at all points of DηD_\eta. Differentiating the equation of mass conservation (2) with respect to zz we get that uxz+vyz+wzz=0u_{xz}+v_{yz}+w_{zz}=0, and so we have via (17) that

uxz≡0,(18)u_{xz}\equiv0, \tag*{(18)}

while from the differentiation of (2) with respect to yy we infer that

uxy≡0.(19)u_{xy} \equiv0. \tag*{(19)}

Differentiating with respect to zz in the first equation of (11) and recalling that uxz=uyz=0u_{xz}=u_{yz}=0 we see that (Ω3+βy)uzz=0(\Omega_3+\beta y)u_{zz}=0 for all (x,y,z)(x,y,z) in DηD_\eta. Hence,

uzz=wxz=0(20)u_{zz}=w_{xz}=0 \tag*{(20)}

at all points of DηD_\eta. We claim now that

uxx≡0within Dη.(21)u_{xx}\equiv0 \quad\text{within } D_\eta. \tag*{(21)}

To prove (21) we distinguish two cases as follows:

  1. Ω1=0\Omega_1=0. Then the second equation in (11) becomes

(Ω2+2ω)vy+(Ω3+βy)vz=0.(\Omega_2+2\omega)v_y+(\Omega_3+\beta y)v_z=0.

After a differentiation with respect to xx in the previous equation and accounting also for (15), we obtain that

vxy=uyy=0within Dη,(22)v_{xy}=u_{yy}=0 \quad\text{within } D_\eta, \tag*{(22)}

so that the harmonicity of uu and relation (20) provides the claim in (21).

  1. Ω1≠0\Omega_1\ne0. In this case we differentiate the first equation of (11) with respect to xx and avail of uxy=uxz=0u_{xy}=u_{xz}=0 to conclude the correctness of the claim in (21).

We can infer now from (21), (20) and the harmonicity of uu that

uyy=0within Dη.(23)u_{yy}=0 \quad\text{within } D_\eta. \tag*{(23)}

An inspection of the previous considerations shows that ∇u\nabla u, ∇v\nabla v and ∇w\nabla w are (vectorial) functions that depend only on tt. Corroborating the latter with the boundedness of uu, vv, ww yields that uu, vv, ww are functions that depend only on tt. The latter implies now that the spatial gradients of uu, vv and ww vanish identically within DηD_\eta. This last information yields via (11) that

v=0throughout Dη.(24)v=0 \quad\text{throughout } D_\eta. \tag*{(24)}

Since ww is constant within DηD_\eta we conclude via the bottom condition (7) that w=0w=0 throughout the lower layer DηD_\eta. Consequently, the Euler equations are now written as

Px=−ρu′(t),Py=−ρ(βu(t)+ω2)y,Pz=ρ(2ωu(t)−g+ω2R),(25)\begin{aligned} P_x&=-\rho u'(t),\\ P_y&=-\rho\left(\beta u(t)+\omega^2\right)y,\\ P_z&=\rho\left(2\omega u(t)-g+\omega^2R\right), \tag*{(25)} \end{aligned}

so that

P(x,y,z,t)=−ρ[u′(t)x+(βu(t)+ω2)y22+(g−2ωu(t)−ω2R)z].(26)P(x,y,z,t)=-\rho\left[u'(t)x+(\beta u(t)+\omega^2)\frac{y^2}{2}+(g-2\omega u(t)-\omega^2R)z\right]. \tag*{(26)}

The boundedness of the pressure implies now that

u(t)=−ω2β=−ωR2u(t)=-\frac{\omega^2}{\beta}=-\frac{\omega R}{2}

for all tt, and therefore for all x,y,tx,y,t and all −d≤z≤η(x,y,t)-d\le z\le\eta(x,y,t) it holds

P(x,y,z,t)=−ρgz.(27)P(x,y,z,t)=-\rho gz. \tag*{(27)}

Utilizing the vorticity equation for the domain Dη,η~D_{\eta,\tilde{\eta}} and employing an argument analogous to the one that led to the constancy of u,v,wu,v,w we conclude that u~\tilde{u} and w~\tilde{w} are constants of the time tt and v~\tilde{v} vanishes throughout Dη,η~D_{\eta,\tilde{\eta}}. Hence, the Euler equations in the upper layer are written as

P~x=−ρ~(2ωw~(t)+u~′(t)),P~y=−ρ~(βu~(t)+ω2)y,P~z=ρ~(2ωu~(t)−g+ω2R).(28)\begin{aligned} \tilde{P}_x&=-\tilde{\rho}(2\omega\tilde{w}(t)+\tilde{u}'(t)),\\ \tilde{P}_y&=-\tilde{\rho}(\beta\tilde{u}(t)+\omega^2)y,\\ \tilde{P}_z&=\tilde{\rho}(2\omega\tilde{u}(t)-g+\omega^2R). \tag*{(28)} \end{aligned}

which yields that the pressure in the upper layer Dη,η~D_{\eta,\tilde{\eta}} is given as

P~(x,y,z,t)=−ρ~[(2ωw~(t)+u~′(t))x+(βu~(t)+ω2)y22−(2ωu~(t)−g+ω2R)z].(29)\tilde{P}(x,y,z,t)=-\tilde{\rho}\left[(2\omega\tilde{w}(t)+\tilde{u}'(t))x+(\beta\tilde{u}(t)+\omega^2)\frac{y^2}{2}-(2\omega\tilde{u}(t)-g+\omega^2R)z\right]. \tag*{(29)}

Owing to the boundedness of P~\tilde{P} we infer from the previous formula that

u~(t)=−ω2β=−ωR2\tilde{u}(t)=-\frac{\omega^2}{\beta}=-\frac{\omega R}{2}

and w~(t)=0\tilde{w}(t)=0 for all tt,

P~(x,y,z,t)=−ρ~gz.(30)\tilde{P}(x,y,z,t)=-\tilde{\rho}gz. \tag*{(30)}

and for all x,y,z,tx,y,z,t such that η(x,y,t)≤z≤η~(x,y,t)\eta(x,y,t)\le z\le\tilde{\eta}(x,y,t) it holds

P~(x,y,z,t)=−ρ~gz.(31)\tilde{P}(x,y,z,t)=-\tilde{\rho}gz. \tag*{(31)}

Applying now the balance of forces at the interface we obtain (availing also of ρ≠ρ~\rho\ne\tilde{\rho}) that η(x,y,t)=0\eta(x,y,t)=0 for all x,y,tx,y,t. With this finding we immediately see that the kinematic conditions on the interface (6) are also verified.

The kinematic condition on the free surface (5) becomes

η~t−ωR2η~x=0,\tilde{\eta}_t-\frac{\omega R}{2}\tilde{\eta}_x=0,

whose general solution is given as η~(x,y,t)=F(x+ωR2t,y)\tilde{\eta}(x,y,t)=F\left(x+\frac{\omega R}{2}t,y\right) for some function (X,Y)↦F(X,Y)(X,Y)\mapsto F(X,Y). But, from the dynamic boundary condition (9) we have that F=−pρ~gF=-\frac{p}{\tilde{\rho}g}.

We conclude by revealing the reason for choosing the bottom defining function bb to depend only on yy.

Remark 2.3. Maintaining the hypotheses from Theorem 2.1 and assuming that the bottom defining function presents a most general dependence (x,y)→b(x,y)(x,y) \to b(x,y) with bx≠0b_x \ne0 we obtain that no solution with bounded pressure in the lower layer DηD_\eta exists. Indeed, proceeding like in the proof of Theorem 2.1 we notice that the arguments and the conclusions therein hold verbatim also in this scenario (with a more general bottom) until (and including) formula (2.24). Then, utilizing the bottom condition w=ubx+vbyw = u b_x + v b_y on z=b(x,y)z = b(x,y) we get w(t)=u(t)bx(x,y)w(t) = u(t)b_x(x,y) for all x,yx,y and for all tt. Assuming, ad absurdum that there is t0t_0 such that u(t0)≠0u(t_0) \ne0, we obtain from the previous equality that bxx=byy=0b_{xx} = b_{yy} = 0. Since bb is bounded, it follows that bx(x,y)=0b_x(x,y) = 0 for all x,yx,y, which is a contradiction with the hypothesis that bx≠0b_x \ne0. Thus, u(t)=0u(t) = 0 for all tt. From w(t)=u(t)bxw(t) = u(t)b_x it follows that also w(t)=0w(t) = 0 for all tt. The governing equations now yield that Py=−ρω2yP_y = -\rho\omega^2 y which clearly impedes the boundedness of PP.

Conclusion

We explored here the impact of (piecewise) constant vorticity on the dimension reduction of the velocity field in water flows satisfying the three-dimensional governing equations. This is part of a broader research agenda [1, 7, 10, 11, 16, 54, 56, 57, 67, 70] that aims to deepen the analytical understanding of the three-dimensional water waves equations with vorticity.

In addition to the previous aspects concerning the vorticity, our analysis takes into account the presence of geophysical effects in the form of the (equatorial) β\beta-plane approximation: this procedure consists in linearizing the Coriolis force in the tangent plane approximation, this course of action being justified by the moderate scale of motion. Unlike the ff-plane approximation, the β\beta-plane approximation [63] displays the essential characteristic that the Coriolis parameter is not constant in space. This is a feature that not only makes the equations of motion more tractable but also allows the study of many phenomena in the atmosphere and ocean; in particular, Rossby waves, the most important type of waves for large-scale atmospheric and oceanic dynamics, depend on the variation of the Coriolis parameter as a restoring force cf. [44].

Under the assumptions (on the vorticity) stated before, we showed that the bounded solutions have zero vertical and one horizontal velocity components, while the other horizontal component is constant. We have also proved that the interface is flat, the free surface has a traveling character in the horizontal direction of the nonvanishing velocity component, being of general type in the other horizontal direction, and the pressure is hydrostatic in both layers.

Different from other scenarios investigated before [5, 7, 53, 54, 56, 57, 59, 67, 70], we allow a varying bottom boundary. A further improvement, when compared with previous studies dealing with three-dimensional flows with piecewise constant vorticity [59], the vorticity vectors corresponding to the upper and lower layer, respectively, need not be parallel.

Our conclusion (regarding the dimension reduction of the flow) is reinforced in the study by Xia and Francois [74], which shows that large-scale structures in thick fluid layers can suppress vertical eddies and reinforce the planarity of the flow. In connection with planar flows we would also like to mention the result on the geometric structure of flows by Sun [68].

Appendix

To justify the consideration of the system (1) and for the sake of self-containedness, we recall in this section a derivation of the governing equations for the β\beta-plane, as presented by Constantin & Johnson [13].

We start from the setting of cylindrical coordinates in which the equator is replaced by a line parallel to the xx axis: the equator is “straightened” and the body of the sphere is represented by a circular disc, which is mapped out by the corresponding polar coordinates (x,θ,z)(x,\theta,z). Here, xx stands for the azimuthal direction and points from west to east, the direction of increasing θ\theta is from north to south, and zz represents the local vertical. Then, the governing equations for water wave propagation written in cylindrical coordinates (x,θ,z)(x,\theta,z) in a rotating framework are the momentum conservation equations

ut+uux+vvθR+z+wuz+2ω(wcos⁡θ−vsin⁡θ)=−pxρ,vt+uvx+vvθR+z+wwzR+z+2ωusin⁡θ+(R+z)ω2sin⁡θcos⁡θ=−pθρ(R+z),wt+uwx+vwθR+z+wwz−v2R+z−2ωucos⁡θ−(R+z)ω2cos⁡2θ=−pzρ−g,(32)\begin{aligned} u_t+uu_x+\frac{vv_\theta}{R+z}+wu_z+2\omega(w\cos\theta-v\sin\theta)&=-\frac{p_x}{\rho},\\ v_t+uv_x+\frac{vv_\theta}{R+z}+\frac{ww_z}{R+z}+2\omega u\sin\theta+(R+z)\omega^2\sin\theta\cos\theta&=-\frac{p_\theta}{\rho(R+z)},\\ w_t+uw_x+\frac{vw_\theta}{R+z}+ww_z-\frac{v^2}{R+z}-2\omega u\cos\theta-(R+z)\omega^2\cos^2\theta&=-\frac{p_z}{\rho}-g, \tag*{(32)} \end{aligned}

and the equation of mass conservation

ux+vθR+z+((R+z)w)zR+z=0,(33)u_x+\frac{v_\theta}{R+z}+\frac{\bigl((R+z)w\bigr)_z}{R+z}=0, \tag*{(33)}

where u,v,wu,v,w are the components of the velocity field corresponding to the xx, θ\theta and zz variable, respectively.

Since we are interested in the behavior of the flow close to the equator we will confine the discussion to small θ\theta. Our considerations also take into account the fact that the radius of Earth, with respect to the depth of the oceans, is extremely large. Thus, it is justified to perform the approximations

sin⁡θ≈θ,zR≈0.(34)\sin\theta\approx\theta,\qquad\frac{z}{R} \approx0. \tag*{(34)}

Setting y:=Rθy := R\theta and disregarding the terms smaller than O(θ)O(\theta) in the expansion of the trigonometric functions, the equations of momentum conservation (32) become

ut+uux+vuy+wuz+2ω(w−yRv)=−pxρ,vt+uvx+vvy+wvz+2ωyRu+ω2y=−pyρ,wt+uwx+vwy+wwz−2ωu−Rω2=−pzρ−g,(35)\begin{aligned} u_t + u u_x + v u_y + w u_z + 2\omega\left(w-\frac{y}{R}v\right) &= -\frac{p_x}{\rho},\\ v_t + u v_x + v v_y + w v_z + 2\omega\frac{y}{R}u+\omega^2 y &= -\frac{p_y}{\rho},\\ w_t + u w_x + v w_y + w w_z - 2\omega u-R\omega^2 &= -\frac{p_z}{\rho}-g, \tag*{(35)} \end{aligned}

while the equation of mass conservation acquires the form

ux+vy+wz=0,(36)u_x+v_y+w_z=0, \tag*{(36)}

which are precisely the equations (1) and (2).

Acknowledgements: The support of the Austrian Science Fund (FWF) under research grant P 33107 N is gratefully acknowledged. The author is indebted to the referees whose comments and suggestions have improved the quality of the manuscript.

Data Availability Statement: No data is associated to this manuscript.

References

  1. [1]B. Basu. On an exact solution of a nonlinear three-dimensional model in ocean flows with equatorial undercurrent and linear variation in density. Discrete Cont. Dyn. Syst. Ser. A 39 (2019) no. 8, 4783-4796.DOI
  2. [2]B. Basu. Some numerical investigations into a nonlinear three-dimensional model of Pacific equatorial dynamics. Deep-Sea Res. II 160 (2019), 7-15.DOI
  3. [3]T. J. Bridges. Multi-symplectic structures and wave propagation. Mathematical Proceedings of the Cambridge Philosophical Society vol. 121 (1997) no. 1, 147-190.
  4. [4]R. M. Chen, L. Fan, S. Walsh and M. Wheeler. Rigidity of three-dimensional internal waves with constant vorticity. J. Math. Fluid Mech. 25 (2023), Art. 71.DOI
  5. [5]A. Constantin and E. Kartashova. Effect of non-zero constant vorticity on the nonlinear resonances of capillary water waves, Europhysics Letters 86 (2009) 29001.DOI
  6. [6]A. Constantin. Nonlinear water waves with applications to wave-current interactions and tsunamis, volume 81 of CBMS-NSF Regional Conference Series in Applied Mathematics. Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2011.DOI
  7. [7]A. Constantin. Two-dimensionality of gravity water flows of constant non-zero vorticity beneath a surface wave train, Eur. J. Mech. B/Fluids 30 (2011), 12–16.
  8. [8]A. Constantin and W.A. Strauss, Periodic traveling gravity water waves with discontinuous vorticity, Arch. Ration. Mech. Anal. 202 (2011), no. 1, 133-175.
  9. [9]A. Constantin. An exact solution for equatorially trapped waves. J. Geophys. Res. 117 (2012), C05029.
  10. [10]A. Constantin. Some three-dimensional nonlinear equatorial flows, J. Phys. Oceanogr. 43 (2013) no. 1, 165-175.
  11. [11]A. Constantin. Some nonlinear, equatorially trapped, nonhydrostatic internal geophysical waves, J. Phys. Oceanogr. 44 (2014) no. 2, 781–789.DOI
  12. [12]A. Constantin and R. S. Johnson. The dynamics of waves interacting with the Equatorial Undercurrent. Geophysical and Astrophysical Fluid Dynamics. 109 (2015), no. 4, 311–358.
  13. [13]A. Constantin and R. S. Johnson. An exact, steady, purely azimuthal equatorial flow with a free surface. J. Phys. Oceanogr. 46 (2016), no. 6, 1935-1945.DOI
  14. [14]A. Constantin and R. S. Johnson. An exact, steady, purely azimuthal flow as a model for the Antarctic Circumpolar Current. J. Phys. Oceanogr. 46 (2016), 3585–3594.
  15. [15]A. Constantin, R. I. Ivanov and C.-I. Martin. Hamiltonian Formulation for Wave-Current Interactions in Stratified Rotational Flows. Arch. Ration. Mech. Anal. 221 (2016) no. 3, 1417-1447.DOI
  16. [16]A. Constantin and R. S. Johnson, A nonlinear, three-dimensional model for ocean flows, motivated by some observations of the Pacific Equatorial Undercurrent and thermocline, Physics of Fluids 29 (2017), 056604.DOI
  17. [17]A. Constantin and R. S. Johnson. Steady large-scale ocean flows in spherical coordinates. Oceanography 31 (2018), no. 3, 42-50.DOI
  18. [18]A. Constantin and R. I. Ivanov. Equatorial wave-current interactions. Comm. Math. Phys. 370 (2019) no. 1, 1-48.
  19. [19]A. Constantin and R. S. Johnson. On the nonlinear, three-dimensional structure of equatorial oceanic flows, J. Phys. Oceanogr. 49 (2019) no. 8, 2029-2042.DOI
  20. [21]A. Constantin and R. S. Johnson. On the modelling of large-scale atmospheric flows. J. Differential Equations 285 (2021), 751–798.
  21. [22]A. Constantin and R. S. Johnson. On the propagation of waves in the atmosphere. Proc. Roy. Soc. A 477 (2021), 20200424.
  22. [23]A. Constantin and R. S. Johnson. On the propagation of nonlinear waves in the atmosphere. Proc. Roy. Soc. A 478 (2022), 20210895.
  23. [24]A. Constantin and R. S. Johnson. On the dynamics of the near-surface currents in the Arctic Ocean. Nonlinear Analysis: Real World Applications 73 (2023), Art. No. 103894.DOI
  24. [25]W. Craig. Non-existence of solitary water waves in three dimensions. Recent developments in the mathematical theory of water waves (Oberwolfach, 2001). R. Soc. Lond. Philos. Trans. Ser. A Math. Phys. Eng. Sci. 360 (2002), no. 1799, 2127-2135.
  25. [26]B. Cushman-Roisin & J. M. Beckers. Introduction to Geophysical Fluid Dynamics: Physical and Numerical Aspects. Academic Press 2011.
  26. [28]J. Davies, G. G. Sutyrin and P. Berloff. On the spontaneous symmetry breaking of eastward propagating dipoles. Physics of Fluids 35 (2023), 041707.
  27. [29]P. J. Dellar. Variations on a β-plane: derivation of non-traditional beta-plane equations from Hamilton’s principle on a sphere. J. Fluid Mech. 674 (2011), 174-195.DOI
  28. [31]A. H. Duran Colmenares and L. Zavala Sanson. Anisotropic Lagrangian dispersion in zonostrophic turbulence in a closed basin. Physics of Fluids 34 (2022), 106605.DOI
  29. [32]J. Escher, A.-V. Matioc and B.-V. Matioc. On stratified steady periodic water waves with linear density distribution and stagnation points. J. Differential Equations 251 (2011), no. 10, 2932-2949.DOI
  30. [33]J. Escher, P. Knopf, C. Lienstromberg and B.-V. Matioc. Stratified periodic water waves with singular density gradients. Ann. Mat. Pura Appl. 199 (2020), no. 5, 1923-1959.
  31. [34]A. Geyer and R. Quirchmayr. Shallow water models for stratified equatorial flows. Discrete Cont. Dyn. Syst. Ser. A. 39 (2019) no. 8, 4533-4545.
  32. [35]A. Geyer and R. Quirchmayr. Weakly nonlinear waves in stratified shear flows. Comm. Pure Appl. Anal. 21 (2022) no. 7, 2309-2325.
  33. [36]Y. Giga and H. Miura. On vorticity directions near singularities for the Navier-Stokes flows with infinite energy. Comm. Math. Phys. 303 (2011) 289-300.
  34. [37]Y. Giga. A Liouville theorem for the planar Navier-Stokes equations with the no-slip boundary condition and its application to a geometric regularity criterion. Comm. Part. Diff. Equations. 39 (2014), 1906-1935.DOI
  35. [38]A. E. Gill. Atmosphere Ocean Dynamics. Academic Press, 1982.
  36. [39]D. Henry and B.-V. Matioc. On the existence of steady periodic capillary-gravity stratified water waves. Ann. Sc. Norm. Sup. Pisa-Classe di Scienze. 12 (2013), no. 4, 955-974.arxiv.org/abs/1305.5802
  37. [40]D. Henry. Equatorially trapped nonlinear water waves in a β-plane approximation with centripetal forces. J. Fluid Mechnics. 804 (2016), pp. R11-R111.DOI
  38. [41]D. Henry. A modified equatorial β-plane approximation modelling nonlinear wave-current interactions. J. Differential Equations. 263 (2017) no. 5, 2554 - 2566.DOI
  39. [42]D. Henry and C. I. Martin. Exact, free surface equatorial flows with general stratification in spherical coordinates. Arch. Ration. Mech. Anal. 233 (2019) 497-512.
  40. [43]D. Henry and G. Villari. Flow underlying coupled surface and internal waves. J. Differential Equations 310 (2022), 404-442.
  41. [44]J. R. Holton ang G. J. Hakim. An Introduction to Dynamical Meteorology. Academic Press, 2004.
  42. [45]D. Ionescu-Kruse. An exact solution for geophysical edge waves in the f-plane approximation. Nonlinear Analysis: Real World Applications 24 (2015), 190-195.
  43. [46]D. Ionescu-Kruse. An Exact Solution for Geophysical Edge Waves in the β-Plane Approximation. J. Math. Fluid Mech. 17 (2015), no. 4, 699-706.
  44. [48]G. Li, L. Cheng, J. Zhu, K. E. Trenberth, M. E. Mann, J. P. Abraham. Increasing ocean stratification over the past half-century. Nature Climate Change 10 (2020) no. 12, 1116-1123.
  45. [49]J. Lighthill. Waves in Fluids, Cambridge University Press, 1978.
  46. [50]I. G. Jonsson. Wave-current interactions, in: B. Le Méhauté (Eds.), The Sea, in Ocean Eng. Sc., vol. 9(A), Wiley, 1990, pp. 65-120.
  47. [51]R. S. Johnson. A modern introduction to the mathematical theory of water waves. Cambridge University Press, 1997.
  48. [52]A. J. Majda and A. L. Bertozzi. Vorticity and Incompressible Flow (Cambridge University Press, Cambridge), 2002.
  49. [53]C. I. Martin. Resonant interactions of capillary-gravity water waves. J. Math. Fluid Mech. 19 (2017), no. 4, 807-817.
  50. [54]C. I. Martin. Non-existence of time-dependent three-dimensional gravity water flows with constant non-zero vorticity. Physics of Fluids 30 (2018) no. 10, 107102.
  51. [55]C. I. Martin. On the vorticity in mesoscale ocean currents, Oceanography, 31 (2018), no. 3, 28-35.
  52. [56]C. I. Martin. On constant vorticity water flows in the β-plane approximation. J. Fluid Mech. 865 (2019), 762–774.
  53. [57]C. I. Martin. Constant vorticity water flows with full Coriolis term. Nonlinearity 32 (2019) no. 7, 2327-2336.
  54. [58]C. I. Martin. Azimuthal equatorial flows in spherical coordinates with discontinuous stratification. Phys. Fluids. 33 (2021), 026602.
  55. [59]C. I. Martin. Liouville-type results for the time-dependent three-dimensional (inviscid and viscous) water wave problem with an interface. J. Differential Equations 362 (2023), 88-105.DOI
  56. [60]A.-V. Matioc. An exact solution for geophysical equatorial edge waves over a sloping beach, J. Phys. A 45 (2012), no. 36, 365501, 10 pp.DOI
  57. [61]A.-V. Matioc. Exact geophysical waves in stratified fluids, Appl. Anal. 92 (2013), no. 11, 2254-2261.DOI
  58. [62]A.-V. Matioc. On the particle motion in geophysical deep water waves traveling over uniform currents, Quart. Appl. Math. 72 (2014), no. 3, 455-469.
  59. [63]J. Pedlosky. Geophysical Fluid Dynamics. Springer-Verlag, 1992.
  60. [64]D. H. Peregrine. Interactions of water waves and currents, Adv. Appl. Mech. 16 (1976), 9-117.
  61. [65]C. G. Rossby and Collaborators. Relation between variations in intensity of the zonal circulation of the atmosphere and the displacements of the semi-permanent centers of action. J. Mar. Res. 2 (1939), 38-55.
  62. [66]R. Salmon. Lectures on Geophysical Fluid Dynamics. Oxford University Press. 1998.
  63. [67]R. Stuhlmeier. On constant vorticity flows beneath two-dimensional surface solitary waves, J. Nonlinear Math. Phys. 19 (2012), suppl. 1, 1240004, 9 pp.DOI
  64. [68]C. Sun. Geometric structure of pseudo-plane quadratic flows, Physics of Fluids 29, (2017), 036602.DOI
  65. [69]G. P. Thomas and G. Klopman. Wave-current interactions in the nearshore region, in: J. N. Hunt (Ed.), Gravity Waves in Water of Finite Depth, Computational Mechanics Publications, WIT, Southampton, United Kingdom (1997), pp. 255-319.
  66. [70]E. Wahlen. Non-existence of three-dimensional travelling water waves with constant non-zero vorticity, J. Fluid Mech. 746 (2014), doi:10.1017/jfm.2014.131DOI
  67. [71]W. Hu, Z. Deng, S. Han, W. Zhang. Generalized multi-symplectic integrators for a class of Hamiltonian nonlinear wave PDEs. Journal of Computational Physics, vol. 235 (2013), 394-406.
  68. [72]W. Hu, J. Ye and Z. Deng. Internal resonance of a flexible beam in a spatial tethered system. Journal of Sound and Vibration. 475 (2020), 115286.DOI
  69. [73]W. Hu, Z. Han and T. J. Bridges. Multi-symplectic simulations of W/M-shape-peaks solitons and cuspons for FORQ equation, Applied Mathematics Letters, 145, 108772.
  70. [74]H. Xia and N. Francois, Two-dimensional turbulence in three-dimensional flows, Physics of Fluids 29, (2017), 111107.

Paper details

vlnc:W_rxzv255h
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.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents