Chapter 11 More on wave motions, filtering

 Consider a homogeneous layer of inviscid fluid on an f-plane confined between rigid horizontal boundaries. z y x f Inertial waves  Suppose that the entire layer is set impulsively in motion in the y-direction with the constant velocity v = at t = 0.

Equations The same equation is satisfied also by u. constant  = ± f The full solution is:

The vector velocity has magnitude V where The direction changes periodically with time with period 2  /f. called the inertial period f is sometimes referred to as the inertial frequency.

 The perturbation velocity is independent of spatial position.  All fluid parcels move with the same velocity V at any instant --- assumes that the flow domain is infinite and unconstrained by lateral boundaries.  => at each instant, the layer moves as would a solid block. Some notes  Consider a fluid parcel initially at the point (x 0, y 0 ) and suppose that it is at the point (x,y) at time t.  Integrating the velocity =>

The parcel executes a circular path, an inertia circle, with centre at and radius. ft parcel trajectory The motion is anticyclonic in sense. The period of motion 2  /  = 2  /f = half a pendulum day, => the time for a Focault pendulum to turn through 180°.

centrifugal force Coriolis force There is no pressure gradient in the flow - the only forces are the centrifugal and Coriolis forces. In circular motion these must balance. This is possible only in anticyclonic motion. Force balance on a fluid parcel undergoing pure inertial oscillations (northern hemisphere).

 Pure inertial oscillations do not seem to be important in the earth's atmosphere, but  Time spectra of ocean currents often exhibit significant amounts of energy at the inertial frequency (see Holton, p.60).  Nevertheless, inertial effects are observed in the atmosphere.  Examples are: - sea breeze circulations ( => next pictures) - the nocturnal low level jet Inertial effects

The Morning Glory

A Gulf cloud line

Coral Sea Cape York Peninsula Gulf of Carpentaria 800 km Sea breeze circulations over Cape York Peninsula

1200 h

1600 h

1800 h

2000 h

2200 h

0000 h

0200 h

Movie by Gerald Thomsen

The profile of wind component in the direction of the geostrophic wind (u g ) showing a nocturnal jet, compared with the profile the previous afternoon oCoC z m 1500 day night day 50 0 ugug u m s  z m

Shallow water model configuration  In pure inertial wave motion, horizontal pressure gradients are zero.  Consider now waves in a layer of rotating fluid with a free surface where horizontal pressure gradients are associated with free surface displacements. f H(1 +  ) H papa z = 0 p = p a +  g(H(1 +  ) – z) undisturbed depth Inertia-gravity waves

Consider hydrostatic motions - then independent of z The fluid acceleration is independent of z. If the velocities are initially independent of z, then they will remain so.

Linearized equations - no basic flow Consider wave motions which are independent of y. A solution exists of the form if

These algebraic equations for have solutions only if

The solution with  = 0 corresponds with the steady solution ( ¶ / ¶ t = 0) of the equations and represents a steady current in strict geostrophic balance in which The other two solutions correspond with so-called inertia- gravity waves, with the dispersion relation The phase speed of these is The waves are dispersive

 The existence of a positive and negative root for  (or c p ) shows that these waves can propagate in either x-direction.  In the limit as k ® 0,  ® f and the x-dependence of the motion drops out.  Then the motion corresponds with the pure inertial wave discussed earlier.  In the absence of rotation (i.e., f = 0) the theory reduces to that for small amplitude surface gravity waves on shallow water.  Such waves are non-dispersive:

Note that the hydrostatic assumption restricts the validity of the whole analysis to long waves on shallow water; strictly for waves with wavelength  >> H. Rotation is important if => the Rossby radius of deformation for a homogeneous fluid

In the deep ocean, At 45° latitude, f = 10  4 s  1 Rotation effects are important only on the planetary scale. and.

 Replace the v-momentum equation by the vorticity equation => where Assume that f can be approximated by its value f 0 at a particular latitude, except when differentiated with respect to y in the vorticity equation. This is justified provided that meridional particle displacements are small. Inclusion of the beta effect

 Again assume that ¶ / ¶ y º 0 and consider travelling-wave solutions of the form: These are consistent only if the determinant is zero

Now expanding by the second row A cubic equation for  with three real roots. When  >>  /k, the two non-zero roots are given approximately by the formula This is precisely the dispersion relation for inertia-gravity waves. When  2 << gHk 2, there is one root given approximately by

 For wavelengths small compared with 2  times the Rossby length reduces to the dispersion relation for nondivergent Rossby waves. the effects of divergence due to variations in the free surface elevation become important, or even dominant. For longer wavelengths,

For extremely long waves These are nondispersive.  The importance of divergence effects on the ultra-long waves explains why the calculated phase speeds for planetary waves of global wave-numbers-1 and -2 were unrealistically large.  The latter were obtained from the phase speed formula for nondivergent waves.  In the atmosphere, divergence effects are not associated with the displacement of a free surface!

 To understand the effect of divergence on planetary waves, we introduce the meridional displacement  of a fluid parcel.  This is related to the meridional velocity component by v = D  /Dt, or to a first approximation by  Substituting for v in and integrating with respect to time, assuming that all perturbation quantities vanish at t = 0, gives This formula is equivalent (within a linear analysis) to the conservation of potential vorticity (see exercise 11.2).

holds for a fluid parcel which has  and  initially zero so that  = - . In the absence of divergence, The term f 0  represents the increase in relative vorticity due to the stretching of planetary vorticity associated with horizontal convergence (positive  ).

1 11 0   HILO  = kx   t The relative phases of the quantities v, ,  and  over one wavelength northern hemisphere case The relative vorticity tendency due to stretching opposes that due to meridional motion and thereby reduces the restoring tendency of the induced velocity field. Phase diagram for Rossby waves

This has the solution, where Equations: If ¶ / ¶ y º 0 as before, the vorticity equation reduces to an equation for , and since  = ¶ v/ ¶ x => Filtering

 There is no other solution for  as there was before.  In other words, making the geostrophic approximation when calculating v has filtered out in the inertia-gravity wave modes from the equation set, leaving only the low frequency planetary wave mode.  This is not too surprising since the inertia-gravity waves, by their very essence, are not geostrophically-balanced motions. Dispersion relation for a divergent planetary wave

 The idea of filtering sets of equations is an important one in geophysical applications.  The quasi-geostrophic equations are often referred to as 'filtered equations' since, as in the above analysis, the consequence of computing the horizontal velocity geostrophically from the pressure or stream-function suppresses the high frequency inertia-gravity waves which would otherwise be supported by the Boussinesq equations.  Furthermore, the Boussinesq equations themselves form a filtered system in the sense that the approximations which lead to them filter out compressible, or acoustic waves. Filtered equations

 The full nonlinear form of the shallow-water equations in a layer of fluid of variable depth h(x,y,t) is relative vorticity horizontal divergence  The u- and v-equations can be replaced by the vorticity and divergence equations: => The Balance Equations

vorticity equation divergence equation  = df/dy Combine the vorticity equation and the continuity equation to form the potential vorticity equation: potential vorticity is q = (  + f)/h

Equations These represent an equivalent system to the equations

 Choose representative scales: U for the horizontal velocity components u, v, L for the horizontal length scale of the motion, H for the undisturbed fluid depth,  H for depth departures h  H, f 0 for the Coriolis parameter, and T = L/U - an advective time scale.  Define two nondimensional parameters: the Rossby number, Ro = U/(f 0 L), and the Froude number, Fr = U 2 /(gH). Scale analysis

where y is nondimensional and. Let For middle latitude systems  ' is of order unity. Treatment of f

The nondimensional forms of the u-momentum and continuity equations are:  Let us examine first the quasi-geostrophic scaling:

 In quasi-geostrophic motion (Ro << 1), the term proportional to Ro can be neglected.  Then g  H/f 0 UL  O(1).  We could choose the scale H so that this quantity is exactly unity, i.e.  H = f 0 UL/g. Then the term  H/H may be written

 to lowest order in Ro Then the momentum equations a consistent scaling for quasi-geostrophic motion is that  H/H = Ro -1 Fr = O(Ro) so that the term proportional to  H/H ® 0 as Ro ® 0. v = ¶ x h and u = ¶ y h D = 0 [or, more generally, D £ O(Ro) ].

and where J(u,v) is the Jacobian  In nondimensional form, the vorticity and divergence equations are The vorticity and divergence equations

At lowest order in Rossby number The quasi-geostrophic approximation where D = RoD 1 When  H/H = Ro  1 Fr, the continuity equation  where  = Ro - 2 Fr and D = Ro D 1.

 These are the quasi-geostrophic forms of the vorticity, divergence and continuity equations.  Note that the divergence equation has reduced to a diagnostic one relating the fluid depth to the vorticity which is consistent with  obtained from

 D = 0 at O(Ro 0 ) => there exists a streamfunction  such that u = -¶ y  and v = ¶ x  and from  = h +constant (e.g. h - H).

 The potential vorticity equation in nondimensional form can be obtained by eliminating D 1 between and a single equation for  (or h) => analogous to the form for a stably-stratified fluid. Note: in the case with a continuous vertical stratification, the  term would be replaced by a second-order vertical derivative of . The potential vorticity equation

 The quasi-geostrophic approximation leads to an elegant mathematical theory, but calculations based upon it tend to be inaccurate in many atmospheric situations such as when the isobars are strongly curved.  In the latter case, we know that centrifugal forces are important and gradient wind balance gives a more accurate approximation.  We try to improve the quasi-geostrophic theory by including terms of higher order in Ro in the equations.  is the streamfunction.  the velocity potential  We decompose the horizontal velocity into rotational and divergent components

and It follows that  This decomposition is general (see e.g. Holton, 1972, Appendix), but is not unique :   One can add equal and opposite flows with zero vorticity and divergence to the two components without affecting the total velocity u.  Consistent with the quasi-geostrophic scaling, where c is zero to O(Ro 0 ), we scale c with ULRo so that in nondimensional form

Equations and

The idea is to neglect terms of order Ro 2 and Ro 3 in these equations, together with the equivalent approximation in the continuity equation In dimensional form they may be written u = u  + u . Note: the divergence equation has been reduced to a diagnostic one relating  to h.

is by the total wind u and not just the nondivergent component of u as in the quasi-geostrophic approximation. Moreover the advection of in and of h in

 These equations are called the balance equations.  They were first discussed by Charney (1955) and Bolin (1955).  It can be shown that for a steady axisymmetric flow on an f- plane, the second equation reduces to the gradient wind equation.  => we may expect the equations to be a better approximation than the quasi-geostrophic system for strongly curved flows. The Balance Equations

 Unfortunately it is not possible to combine the balance equations into a single equation for y as in the quasi- geostrophic case and they are rather difficult to solve.  Methods of solution are discussed by Gent and McWilliams (1983).  Note: although the balanced equations were derived by truncating terms of O(Ro 2 ) and higher, the only equation where approximation is made is the divergence equation.  => the equations represent an approximate system valid essentially for sufficiently small horizontal divergence.  As long as this is the case, the Rossby number is of no importance. Methods of solution

 Elimination of between gives the potential vorticity equation: and Thus an alternative form of the balance system is

Given q, the first two equations can be regarded as a pair of simultaneous equations for diagnosing  and h, subject to appropriate boundary conditions. Equation enables the prediction of q, while may be used to diagnose c.

Under certain circumstances [e.g. Ro O(1)], the nonlinear terms in Eq.(11.47) may be neglected in which case the equation becomes With this approximation, the system (11.46), (11.48) and (11.49) or (11.30), (11.31), (11.48), (11.49) constitute the linear balance equations. These systems are considerably easier to solve than the balance equations. The Linear Balance Equations

The End