AN INTRODUCTION TO DYNAMIC METEOROLOGY: THIRD EDITION 11 Solution: J. R. Holton PROBLEM SOLUTIONS ‘Neglecting the latimdinal variation in the radius of the earth, calculate the angle between the gravitational force and gravity vectors at the surface of the earth a5 a function of Jatinide, Let 9 be latimde and a the angle between g* and g. Then from the law of sines: gl, 248) lng 7 maa BY IR| = acos . 2 asi . sina = @ = Q% asing SSP = 2° asin2g lel 28 A a’R Fig. Li 1.2 Calculate the altitude at which an artificial satellite orbiting in the equatorial plane can be a synchronous satellite (i.e. can remain above the same spot on the surface of the earth). Solution: From (1.7) g*+ QR =0, but g* = -($) (rk) and R=r ‘GM for 9=0. = 42,2900 kan, z = r-a = 36,000 km.1.3 1.4 = = ef Fe = e =~ = e = = oe = ee An artificial satellite is placed into a natural synchronous orbit above the equator and is attached to the carth below by a wire, A second sateilite is attached to the first by a wire of the same length and is placed in orbit directly above the first at the same smpalar veloc, Assuming thatthe wires hve zero mins, call de nonin maass of satellite. Could this tension be used to lift objects into energy? obit wid So additonal expends of Solution: Tension/unit mass = T = 0%, - © & where 7, is the distance of the outer satelite tom the center ofthe earth (71 = 78,000 km). .349 N kg-l. ‘This tension could be used once to lift a mass T= tess than F = 3.56% of the mass ofthe outer satellite into orbit — but only by displacing the satellites and wire further into space, (8 = Tm, where m, and m, are the mass to be lifted and the ‘mass of the ower satellive, respectively.) A tain is running smoothly along a curved wack at the rate of 50m s-!, A passenger standing on a set of scales observes that his weight is 10% greater than when the train is at rest. The track is banked so that the force acting on the passenger is normal tothe floor of the train, What isthe radius of curvature of the mack? Solution: From the force diagram 2» aa may? = (na)? + (™) re Lr Fig. 14 VURATVATATIVTAUTATANATITTITITTTTTTTLs 1.6 LT If a baseball player throws a ball a horizontal distance of 100 m at 30" Jatitade in 4s, by how much is it deflected horizontally as a result of the rotation of the earth? Solution: Assume ball thrown eastward and w =x/t is constant. Then from 12) 2Qut(sin 9). Integrating in time gives 2 = -2n(6ing) or angst a ~5) (100)(4) ~2066Ing)> = (7.292 x 10 Se y=-146em. ‘Two balls 4 cm in diameter are placed 100 m apart on a frictionless horizonta) place at 43° N. If the balls are impulsively propelled directly at each other with equal speeds, at what speed must they travel so that they just miss each other? Solution: From problem 1.5 the lateral deflection for each ball is, bl 2aeae)e = 2cm for the balls to just miss each other. Solving for t: bl (2)(2)(107) (2Qsingx) (10) (50) = = = 6.25ms" Thus u = wus zs x _ (50m) t A locomotive of 2 x 10° ~kg mass travels 50 m s™ along a straight horizontal track at 43°N. What lateral force is exerted on the rails? Compare upward reaction force ‘exerted by the rails for cases where the locomotive is traveling eastward and west- ward, respectively. Solution: The lateral force on the rails is just the reaction to the deflection force due to the Cériolis effect. Thus for a locomotive of mass M : F =~-2Q (sing) Mu =—(10~) (2 x 10°) x (60 ms“) =-10° N (Girected southward for train moving eastward in Northern -Hemispheré). The upward reaction force is just the weight minus thevertical component of the Coriolis force (eq. 1-10). Thus Fypward = M (g- 22 cose). AF = Foestwart~Feastward = 4QcosplulM = 2x 10° N.& 1.8 Find the horizontal Uisplacement of a body dropped from a fixed platform at a height 4 at the equator neglecting the effects of air resistance. What is the numerical value of the displacement for A = 5 km? Solution: di From (2.11 we have abe equator (!} = 20», But 2 a so that w=—gr. su = wo r= ory =(4) isthe total time of the fall from height A. From these we get du a # & 2098; thus, u = Og? ==. a 2Ogr, thus, u = Qgr* z= ey Integrating again: 4.9 A bullets fired vertically upward with initial speed wo at latitude g. Neglecting air resistance, by what distance will it be displaced horizontally when it returns to the ground? (Neglect 2Qu cosd compared to g inthe vertical equaticn.) Solution: Now & =g sothat w= wy ~ gr =0 attop of trajectory. 2 Thus, total time of flight (upward + downward) is i = —2. 8 From (1.11) a = 22 cos w. ft 2 . w= -22 cosd woe 22) and since 25 awe But 1 = 2%. So x= 2.000( . ig” 1.10 A block of mass M = 1 kg is suspended from the end of a weightless string. The other end of the string is passed through a small hole in a horizontal piatform and a ball of mass m= 10 kg is atached. At what angular velocity must the ball rotate on the horizontal platform to balance the weight of the block if the horizontal distance of the ball from che hole is 1m? While the ball is rotating, the block is pulled down 10cm. What is the new angular velocity of the ball? How much work is done in pulling down the block? TEVTAGTITAAATITTTT THAR ERE RRRADAGY 4 VETATTITAIGGGATSolution: Force balance for equilibrium: Mg = mwZa,, where @, is angular velocity of the ball, a, = 1m. % 4 @, = (2) = C3) = Us} Ifthe block f is pulled down ma 10 10 cm then the distance of the ball from the hole is 0.9 m. By conservation of angular momentum @a*,= «a7, where a, ay are final radii and angular velocity, respectively. Thus @, of) = 1.222. The rota work done is given by if the sum of the changes in kinetic and potential energy: AW = AK + AP. Now AK = (Z)llo) ~ (oa,)'] = 115 J; AP =Mgdh where Ah =-10.cm height change. -. AP = 1.98, W=115-0.98 =0.17 J. +S 6 MOe M wie Fig. 1.10 1.11 A particle is free to slide on a horizontal frictionless plane located at latinude ¢ on the earth. Find the equation governing the path of the particle if it is given an impulsive eastward velocity u=uy st ¢= 0. Give the solution for the position of the particle as a function of tim:1.12 1.13 B/S 0 ? Solution: Lenting f = 20 sin 9, from (19) and aay sccinatine y vives 2! = 7 Eliminating v gives Ty = ~/'u. General solution is u =A cos ft + B sin f. du But =0, =U, VEO. 2 ut ar £=0, w= dg, v=. 2. a 4 Wsuigcos ff and v=—up in fi, But 7 = uycos ft, . Thus A=uy, B= 0. gives x= x) + (2) fr, and 2 = —Lpsin fi which gives y =y + (‘2 ooua Let x =y ‘Then trajectory is circle about origin with radius ? and period of rotation: zs ‘Trajectory is clockwise in Northem Hemisphere. Calculate the 100-50-kPa (1000-500 mb) thicknesses for isothermal conditions with temperatures of 273 K and 250 K, respectively. Solution: From (2.19) az= (2 )inf2 Ve) \ _ 28NQ73 (9.8) oar In2 = 5536m, dz = PEDESO) sn = sovom. (9.8) Isolines of 100-50-kPa (1000-500 mb) thickness are drawn on a weather map using a contour interval of 60 m. What is the corresponding layer mean temperature imerval? ghz _, (9.8)(60) 7 (m)) (287129 Cy) = 2.96 = 3°C. Solution: From (1.19) F = WINNT TTT ENTE T ETT RTTITaiTatetaseteteetiaeswsuevwrvevwvuevrvwvvwvvwrveweeeeeTeeerne™ 1.14 1.15 1.16 Show that a homogeneous atmosphere (density independent of height) has a finite height witich depends oniy on the temperature at the lower boundary. Compute the height of a homogeneous atmosphere with surface temperature Ty = 273 K and surface pressure 100 kPa. (Use the ideal gas law and hydrostatic balance.) Solution: Now & —pg so for p foro fom Dy _ Rly . (28710273) But pp = sozy = 2 Po = PoRTy ee (9.81) u constant. u Poste = 7.987 km. For the conditions of Problem 14 compute the variation of the temperature with respect to height. This is referred to as an autoconvective lapse rate. Solution: From problem (1.14) p = pp— pgz when py = surface pressure. But p__(Po~ ps2) (Po) (2) = (2) P= Po PR) pal (8), = 7 -(£ GR) BR er (a) (a)? temperature decreases with height at rate & = °C strongly super- adiabatic! a Show that for an atmosphere with uniform lapse rate 7| 7 = -<). the _ % geopotential height at pressure level p, is given by Z = (2) 1- (2) A where To and po are the sea level temperature and pressure, respectively. Solution: From hydrostatic equation and ideal gas law: ding = -(E-)az but, T=T)— 72 Tous, fen = PoVA4FI 8 = = A) Cene a P S So solving for z: oe “-@)"] = 1.17 Calculate the 100-50-kPa (1000-500 mb) thickness for a constant lapse rate atmosphere with y= 6.5 K km and y= 273 K. Compare your results with the results in Problem 12. Solution: From the result of problem 1.16: 237(6.5 x10) Tp A 273 Az=2|1-|20 = fi-2 9 | = 5187m. y| ba 65 x10 1.18 Derive an expression for the variation of density with respect to height in a constant lapse rate atmosphere. Solution: Now p= . Letting p, be density at height z; Pp. po. Tp. be surface values (z = 0) we have p= Taiz z) (from 1.16) [er ) ty Pon py = poll -% & ) bat ey Poe PI oo % 1.1 Denive an expression for wre aitiwde variation ut ine pressure change op waich ‘occurs when an atmosphere with constant lapse rate is subjected to a height independent temperamre change OT while surface pressure remains constant. At ‘what height is the magnitude of the pressure change a maximum? VAASAVEATTTTATTSHGINGI99ANGINGNIGGSolution: From problem 16 we get PV Getal T . ss 6, Zmax < 2 for € > 0 iLe., for realistic stratification. Y STS U STOUT VUES TSESSESCECSCSCSESCSSESESSSESCEECHEEECHEEEE:WAH 10 2 = \ Chapter 2 = = = a . 1 « e A ship steaming northward at arte of 10 km bl The surface pressure increases ge towards the northwest at the rate of 5 Pa kmé!. What is the pressure tendency recorded at a nearby island station if the pressure aboard the ship decreases at a rate Pa = of 100° h? ion: Dp Dp __ (100), 1 Solution: =—-V-Vp but —=-|—|Pah’, f PO De ( 3 2 a V- Vp =|V} |¥p}cosa, 2 a = ~33.33 ~ a0r0s¥. Fig. 2.1 (2:3) The temperature at a point 50 km north of a station is 3°C cooler than at the station. If the wind is blowing from the northeast at 20 m s~ and the air is being heated by radiation at the rate of 1°Ch~!, whats the local temperarure change at the station? ar . oT oF - OT _y.vr, . DT Solution: —=Q=1°CA olution: @ 3 Dr 5x <| °C V- VT = (20 [aaa]. .47 x 10, (20) WE 8.47 x s TT _ WC _ 305°C OT _ VC _ 305°C °C on h h b -_ - — —_ - = = —_ = =_ = = ee = F = = Yp = er = — e - e = = - = - e- Ee - e- eo = e =n Fig. 2.2 2.3. Derive the relationship @x (Qxr) = —2?R which was used in Eq, (2.7). Solution: |R}=1rl sine and R is perpendicular to Q, so by definition of cross product Qxr = Rxr =/QIIRIn, where n is unit vecior pointing into paper perpendicular to plane of Q and R. Then Qx(Qxr) = Axn |QIIRI, but Qxn = i Qx(Qxr)=-Q2R. t 2 R | Fig. 23 2.4 Derive the expression given in Eq. (2.13) for the rate of change of k following the motion. a Dk Now [| (3) Solution: 2% = tim|{ 128) Do ey yy Now Vel [5k| = and Sk is directed parallel to j, CECE F12 By fimilar arguments 2 y gu x k+gk Fig. 2.4 wand Compare the magnitudes of the curvature term. and the Coriolis force for a a ballistic missile fired eastward with a velocity of 1000 m st at 45* latitude. If the missile travels 1000 km by how much is it deflected from its eastward path due t0 both these terms? Can the curvature term be neglected in this case? & ( zr + #4) where f = 2Qsing. Solution: From (2.20) — 2 2 urtand v -( -a\e Integrating twice: a 6 m 10ms! where ¢ = = 10's + io 6.37 x 10° 2 y= (et (1.03 x 0) ($) m = 130km. Suppose a 1-kg parcel of dry air is rising at a constant vertical velocity. If the parcel is being heated by radiation at the rate of 10-' W kg™, what must the speed of rise be to maintain the parcel ar a constant temperarure? DT Dp Solution: From (2:42) cy—~ - @— olution: From (2:42) cy ~~ a>" Applying hydrostatic equation: 2 =-ps. DT (2.42) can be rewiten as cy E> + HU VUVVVUN WVU BCREDCTTETGED VVV CVV GVV VV VU VV VVSSddsS Sd: 3 DT For constant temperature parcel [= = 0. (10) W kg Samet = 0002ms or | +1.02ems7 |. .8ms @) Derive an expression for the density p that results when an air parcel initially at — pressure p, and density p, expands adiabatically wo pressure p. Solution: For adiabatic expansion @ = (2h is conserved. Thus P e=T= 7(2). or substirutingT = 2, = Pp PR Ps or p= (2) " where we have used c, = An air parcel that has a temperature of 20°C at the 1000 mb level is lifted dry adiabatically. What is its density when it reaches the 500 mb level? Solution: From problem 7, p = (2/2) 10°Pa ay" “3 or p(500 mb) = atts) (3) = 0.725 kgm Suppose an air parcel starts from rest at the 80 kPa (800-mb) level and rises vertically to 50 kPa ( 500 mb) while maintaining a constant 1°C temperature excess over the environment. Assuming that the mean temperature of the 80-50-kPa layer is 260 K, compute the energy released due to the work of the buoyancy force. Assuming that all the released energy is realized as kinetic energy of the parcel, ‘what will the vertical velocity of the parcel be at 50-kPa? Solution: Using (2.44) to express r.h.s. of (2.51) in terms of T we get (Tr -T) buoyancy force / unit mass = but energy released/unit mass = force x distance.4 Distance = Az = (#4) In @) where py = 80-kPa, Pa If all 135 J kg-} goes into kinetic energy, then 2 = 135 J kgo}. *. w= 16.43 ms“! at 50 kPa! 0) Show that for an atmosphere with an adiabatic lapse rate (i.t., constant potential ve temperature) the geopotential height is given by Z = Hf - (2 I + Where po \ Po. is the pressure at Z = 0 and He 8 is the total geopotential height of the atmosphere. Solution: In problem (1.16) let y =-2- apse sate of adiabatic amosphere ? and note that for py = 1000 mb Ty=@. ‘Then immediately o-nfi-(2)4 \ Po. Gud In the isentropic coordinate system potential temperature is used as the vertical ~~" coordinate, Since in adiabatic flow potential temperature is conserved following the motion, isentropic coordinates are useful for tracing the actual paths of travei of individual air parcels. Show that the transformation of the horizontal pressure gradient force from z 108 coordinates is given by f 2) Vp = Volt, where M= ‘pT + ® isthe Montgomery streamfunction. Solution: In equation (1.22) let s= 6 (2) = (2), + (2): op 1\(a ting 2 = pp vee 2)(2) (8 VUVAVTVTVATAATAAT ATTA T ATTA ATA TTT TATE AT TETrs eee eee eee eee eee CSU UTES SEES SUIT U SUITES where we have used gdz = d®. Now (3) (2), = m(2z2) Bot from equation (2.44) In = In ~ 4 Inp + constant, « (3), - EX), % OKs) on), (2), win ee (2)(2),=(26) ote we +016 Chapter 3 3.1 Am aircraft flying a heading of 60° (Le., 60° east of north) at air speed 200 mn moves relative to the ground due cast (90°) at 225 m s“'. If the plane is flying at constant pressure, what is its rate of change in altitude in meters per kilometer of horizontal distance assuming a steady pressure field, geostrophic winds, and falots!, = 2%, Thus, assuming thatthe wind 8 is geostophic we need the y component to compute the changein height in x-direction. Now | Vai |= 200 ms | Veroung ! = 200 m st, From diagram Vg =j- Vwind = -100 ms Solution: Late (%) 0ots)\(100ms) sm ~ (3), ~ O8ms a Vv * V wind V ground Fig. 31 (3.2° The actual wind is directed 30° to the right of the geostrophic wind. If the geostrophic wind is 20 m s~!, what is the rate of change of wind speed? Let felts, Solution: From (39) 2¥ = me. -(2)2. ® +s component of Vp | to V. ples’ 2 se But |Vp| = p/lVe\, swvtassmninsenansantnnattiaat nanan a =-f {Vel sinZ = -(10~s"\(20 ms(2) =-10? ms. VETAATTTAANS,CCSSSISSSEESESSSSCSE SETS SESESECTSESESESESESLSEIETT ee Ip a tt 7 P %, Po +P . v vp Fig. 3.2 3.3) A tornado rotates with constant angular velocity @. Show that the surface pressure org at the center of the tomado is given by P = Po oxo - 928, where po is the surface pressure at a distance rg from the center and T' is the temperature (assume constant), If the temperature is 288 K, pressure at 100 m from the center is 10 kPa, and wind speed at 100 m from center is 100 m s~!, what is the central pressure? Solution: Letting r be the distance form the axis of rotation of the tomado, equation (3.14) can be rewritten as ve la wT Or But p B * Jone Po or P = Pp exp| 2 (2RT), 2 2,2 2 since 2 cet, p= py [1 - 2B) — 94 kPa. 2RT 2RT (3.4 Calculate the geostrophic wind speed in meters per second for a pressure gradient of Taal and compare with all possible gradient wind speeds for the same pressure gradient anda radius of curvature of $500 km. Let p= 1 km mand f= 10+ s™ |4 AN Sotution: From (3.11) V, From (3.15) RY (pK "3 Vigat = -(8) + (& + s,) = f:25 + (625 + s00y4]. ‘Noting that V,,,, > 0 for permitted solutions: Ved = 8.5 ms“ regular low, Vyraa = 13.8 m s“ regular high, Vena = 58.5 m s~) anomalous fow, V,qua = 36.2 ms anomalous high. FDAFERTATAL ENE 3.5 Determine the maximum possible ratio of the normal anticyclonic gradient wind speed to the geostrophic wind speed for the same pressure gradient. pa Solution: Fornormal high R <0 and Vig a) - (‘= + m,) = 2 (Vout) = -£) Les IRV, + eet = 0 for maximum E tye n Uewlae Meal on = ' 2 3.6 — Show that the geostrophic balance in isothermal coordinates may be written S - FV, =k x Vp(RT inp+®), fh C=) We=) .(=))(2) Solution: From the diagram: z (2-8) - Be, we NE), GIB) + where we have used hydrostatic balance Maa \ If weeliminate p using p = 2, ra BEY 2) + (2) = (2) [kT inp + ©}, POs, then fV, = VATS ETT TATA , \19 << 75 Fig. 3.6 Determine the radii of curvature for the trajectories of air parcels located 500 km to the east, north, south, and west of the center of a circular low-pressure system, respectively. The system is moving eastward at 15 m s“'. Assume geostrophic flow with a uniform tangential wind speed of 15 ms, Solution: From (3.24) R, = a - essen V = 15ms"! is wind speed. R, = 500 kam. ¢ = 15 ms" ~ speed of storm. North of center West of center South of center R, > #0 (7 = 0); East of center { shows y due east of storm center } c Fig. 3.7 CII SIIT Ese Sse SSS eee S TSK Seer ESS SUES CTS KUUI EES:20 3.8 Determine the gradient wind speeds for the four air parcels in Problem 7 and compare these speeds with the geostrophic speed. (Let f= 10s“) Solution: Gradient wind speed for normal low: 2 Yona = (B)+ (& * mu) = 10.5 ms” North of center, Verag = 12.1 ms“! West and East of center; Vyag = 15 ms“ South of center. [Note: for case R, > e (South of center) we get from equation G7) V=V,) 3.9 Show that as the pressure gradient approaches zero the gradient wind reduces to the geostrophic wind for a normal anticyclione and to the inertia circle for an anomalous anticyclone. Solution: For pressure grad > 0 V,—>0. av, 4 20: a= (4) + (Bf + Me 2 2 RR av, ye 2V, as V, ~ 0 (1) ~1+(%) wR RR Wyte = (A) ors (0- 2 L Positive root gives Vyraa= Vy. Negative root gives Vpay=—JR. (See equation 3.12.) (3.10) The mean temperature in the layer berween 75 and 50 kPa decreases eastward by 3°C per 100 kam. If the 75-kPa geostrophic wind is from the southeast at 20 m s~!, what is the geostrophic wind speed and direction at 50 kPa? Let f= 10s“! Solution: From (3.33) usp = ups + Up = —14.1 ms + up, Y5q = vas + op = +14. ms! + vy, v--(IEB)- -34.5 ms! CUVTUNTUDTEULGTVUUTD DTT TT TTT T TTT AT TATA TTT T ATTASEIS IIIT IIIT TST y esses sees sees e sess ddd 21 *. V, (50 kPa) = (-14.1, -20.4) ms“! or 1V,1=25ms-1, From 34” east of north, ‘31D What is the mean temperature advection in the 75-50-kPa layer in Problem 10? Solution: Temperaure advection = -V - VT at where 7 (4a * 4) -14.1 mst oT ec 5 “1 3x10 cm 00 km 4.23 x 107°C s = Es 3.1P Suppose that a vertical column of the atmosphere at 43°N is initially isothermal from 90 10 50 kPa. The geostrophic wind is 10m s~ from the west at 70 kPa, and 20 ms“! from the.west at 50 kPa. Calculate the mean horizontal temperature gradients in the two layers 90-70 kPa and 70-50 kPa. Compute the rate of advective temperature change in each layer. How long would this advection panremn have 10 persist in order 10 establish a dry adiabatic lapse rate between 60 and 80 kPa? (Assume that the lapse rate is linear from 90-50 kPa and that the 80-60 kPa layer thickness is 2.25 km.) Solution: Wind components given are Pp My Y% ur Yr 900 0 10 > 10 -10 700 10 0 > 0 0 500 20 0For layer 900 — 700 mb: For layer 700 — 500 mb: ae c Rate of temperature change is S =- Average wind in 900 — 700 mb layer is (5, 5) ms”. Average wind in 700 — 500 mb layer is (15, 0) ms~’, “Thus for 900-700 mb layer a = 139 x 104°C 54 For 700 ~ 500 mb layer a =0. ‘Thus che temperature difference between 800 and 600 mb increases by 0.5°C hr! Assuming a 2.25 km 800-600 thickness (standard atmosphere value) we find that an adiabatic lapse rate (+982) would be established in about 44 h. 3.13 An airplane pilot crossing the ocean at 45°N latitude has both a pressure altimeter and a radar altimeter, the latter measuring his absolute heigi above the sea. Flying at an air speed of 100 m s“! he maintains altitude by referring to his pressure altimeter set at an altimeter setting of 101.3 kPa. He holds an indicated 6000-m altimude. At the beginning of a one-hour period he notes that his radar altimeter reads 5700 m and at the end of the hour he notes that it reads 5950 m. In what direction and approximately how far has he drifted from his heading? Solution: Let V4 be velocity of plane wart. air, Vg be velocity of plane wr. ground, Vp be velocity of wind. Assume Vp is geostrophic so Il to lines of constant height.. The height change of plane in distance d is proportional to component of wind .L to V4. For convenience assume that the heading (hence V,) is eastward. Then northward component of geostrophic wind v = (g)o52. But d=|V4|t. where t is elapsed time. Drift of plane from its heading is D = vr. 2 po( 2) 8) (282) -4) anomalous high 2 = root: not permitted (v < -£) regular high Bo on + root antibaric flow Two roots: not permitted (anomalous low) ~ root: not permitted In the so-called geostrophic momennun approximation (Hoskins, 1975) the gradient wind formula for steady circular flow (3.17) is replaced by the approximation VV, + £V = £V,. Compare the speeds V computed using this approximation ‘wich those obtained in Problem & using the gradient wind fermui. V, Solution: Solving for V: V = 7—*y-y . North of low R = 250. 1+—8 AR24 \. V = 9.38 ms", Bast and west of low R = 500. V=1154ms", South oflow Re, +. V=15ms*. For finite positive R, V underestimates gradient wind. J, . 3.16 How large can the ratio je be before the geostrophic momentum approximation differs form the gradient wind approximation by 10% for cyclonic flow? ¥, av, y4 ‘ . feed 1 s Solution: From equation (3.17) 22 = +Ol- “a m2 GN eR (positive root for regular low.) Fi biem 15; Yemen (+4) here V, ‘rom problem 15: where V, is rp JR R geostrophic velocity. Thus: V, Vv, Zz = Difference SR 0.2 0.167 oa 0.286 os: 0.333 3.17 The planet Venus rotates about its axis so slowly that to a reasonable approximation the Coriolis parameter may be set equal to zero. For steady, frictionless motion parallel (0 latitude circles the momentum equation (2.20) then reduces to a type of a pla isobaric coordinates show that the “thermal wind” equation in this case can be ; * | ~Rinf 2 a expressed in the form @2(p,) - @7(P9) = 2 (@sing cose) 2, - cyclostrophic balance: “22 — (3) ® By wransforming this expression to . where R is the gas constant, @ the radius of the planet, and @, = is the AVVVAAVAALIARIDV DUTT TENNENT acos¢ angular velocity. How must (the vertically averaged temperature) vary withCOOP PETE CPE rPe Pree | | | respect to latitude in order that q@, be a'function only of pressure? If the zonal velocity at = 60-km height above the equator (p; = 2.9 x 105 Pa) ig 100 ms and the zonal velocity vanishes at the surface of the planet (po = 9.5 x 10 Pa) what is the vertically averaged temperamre difference between the equator and pole assuming that @, depends only on pressure? ‘The planetary radius is a = 6100 kin, and the gas constaat is R = 187 J kg! K"!. Solution: (z*tang) a 2acosd sing = {\. Since Gor 1 2( = Zz |. Integrating w. r. t. p: op acosgsing a ) ve Bree P edpr) a R or do? = ———_—_ = dh Jo cose sing fg "? ,(79) Po Po R uf an 2 2 arly) ~ @-(Po) = psing oy” For w? 10 depend on p alone (Z)< T > = cosg sing Rial falar or =T,-T'sin?g. Then @3(p) - @3(p9) = A @ [o7(r) - @7(po)]a? _ 1007 2166 K. ain) easn(in(23)) : Da 29 3.18% Suppose that during the passage of a cyclonic storm the radius of curvature of he isobars is observed to be +800 km at a station where the wind is veering (ruming clockwise) at a rate of 10° per hour. What is the radius of curvature of the trajectory for an air parcel which is pagsing over the station? (The wind speed is 20 m s“.) Solution: From eauation (3.23) 2, = aT but V = 20ms”, (Fa) a yh ) “1 8x 10° m. (2) sé5 rads".18/3600) * (@x 105) 3.19 Show that the divergence of the geostrophic wind in isobaric coordinates on the . cot 1 \fa@ \(‘cos¢ 28) herical earthis given by V - Vv, =(--|(%% ‘coro | * draty vv, =(-7\( 3) SEs) a (Use the spherical coordinate expression for the divergence operator given in ‘Appendix C.) sein 9 BE ( Sra = sh) + seal open BU) (2) where dr = acoso da 3.20) The following wind data were received from 50 km to the east, north, west and south of a station, respectively: 90°, 10 ms"; 120°, 4ms-!; 90°, 8ms-; 60°,4ms-!, Calculate the approximate horizontal divergence at the station. Solution: Ax = Ay = 105m OH MAT le , Ya= YD xy ay = HOF) 242 a erode, 10° * . VWUVAVATATATATAS ARRAN 4 aitTITTIES STU UT a Ue eee deeded edd ee eee deddddees 27 NN 4ms! B -1 4 8 ms c +—\—- 10ms"! NT ae Fig, 3.20 (3.2) Suppose thar the wind speeds given in Problem 20 are each in error by +10%. What would be the percent error in the calculated horizontal divergence in the worst case? Solution: For the worst case we let ua = 9mS, uc = 88ms, ve = 22ms1, vw=-22mst; %) (2) 5 then | — | +]—] = 4.2 x 10° “(5 (5 110 %. 1. Sorthe estimate is in error by 22 The divergence of the horizontal wind at various pressure levels above a given station is shown in the following table Pressure (kPa), VV G10 st) 100 +09 85 +0.6 70 +03 50 . 0.0 30 0.6 10 -1.0 Compute the vertical velocity at each level assuming an isothermal aumosphere with temperature 260 K and letting w= 0 at 100 kPa (1000mb).28 Solution: From equation 3.38 0(p,) = (00) + (vo-ol(2)-(2)] au) (Hy _ 1 f(a) (a ou) (oe where (@) * (5) “2 {2G 1-3) If; (100 kPa) = 0 then integrating upwards 11.25 x 107 Pas? 18.00 x 107? Pas 21.00 x 10 Pas“ @(30} = 15.00 x 107? Pa s™ (10) = —1.00 x 107? Pas” 2B e)-po-m) But @ = -pew. 3.23. Suppose that the lapse rate at the 85-kPa (850-mb) level is 4 K km). If the temperature at a given location is decreasing at a rate of 2 K hv', the wind is vgsterly at 10.m s~l, and the temperature decreases towards the west ata rate of 5 [G0 K™, compute the vertical velocity at the 85-kPa level using the adiabatic method. . - ol Dy. (e Solution: From equation 3.41 @ = 55’ (® + (2) (2) ty -T) o bur S,, a, @ = —pgw, (2), (2) +0 2) ws [aed * Kae) * Vand] \e yy (Te-T) (Cy - 1) = 9.8 x 107% —4 x 10% = 5.8 x 10K m! FEFFFIFTAVTDV ATTA TATTLE TATAIN 4ig % : n & x 8 x & oe S = a 4 : 8 Ss alg a _ = a Te (8 s le i ee rd ' fe ft 3 ale SS bs4a 4.3 30 i Chapter 4 What is the circulation about a square of 1000 km on a side for an easterly (that is, westward flowing) current whic decreases in magnitude toward the north at a rate toms? 2 What is the a in the square? S00 ea is the mean relative vorticiy in the sq Solution: ¢ = (2) - (¥) =0- te ar) -- e ~ fom = CmA = (-2% 10%) 10" mj a A cylindrical column of air at 30°N with radius 100 km expands to twice its original radius. If the air is initially at rest, what is the mean tangential velocity at the perimeter after expansion? Solution: C + 2Qsing A = Constam. + Cina = 22 Sing (Apiiat ~ Aina) + Crit» OU Anioa = 7 5 Agna = 177 = 4m, since ry = 2y. Cena = 2.2sino(-3272 = Ltn ial = 22sino (317) bur v = Lint 2m, Ave 2sine( 2 5.5 ms“ (anticyclonic). An air parcel at 30°N moves northward conserving absolute vorticity. If its inisial relative vorticity is 5 x 10° s-, whar is is relative vorticity upon reaching 90°N? Solution: ¢ + f= Constant. Conn = —Sra + (6 + Pion 2 Sinus = -22 + (5 x 10% so! + 2) = -2.3 « 10" An air column at 60°N with ¢ = 0 initially stretches from the surface to a fixed tropopause at 10-km height. If the air column moves until it is over a mountain barrier 2.5 km high ar 45°N, what is its absolute vorticity and relative voracity as it asses the mountain top? NAAMASAVAAALAAD ANS HN GENT A TT NANA NA TANTRATITTTTTSITIII TTT SSeS ESSEC G ESSEC SC ETEK SUES ETES 4.5 31 Sotution: *L) = constant. Now’ £ = 0 initially so. (C+ flags = { Pc -(2) 1.263 x 10“) = 9.47 ys st (f=) Fn To (E263 x 1074) = 9.473 x 10° st, Sfinat= 9-473 x 107 10.312 x 108s"! = ~8.4 x 105"). Find the average vorticity within a cylindrical annulus of inner radius 200 km and outer radius 400 km if the tangential velocity distribution is given by vA, where A= 10%m?s"! and r is in meters. What is the average vorticity r within the inner circle of radius 200 kan? How is the vorticity distributed within the inner circle? Solution: The line integral for computing the circulation in the annuius is shown in the figure. j 108 6 in? gt For the outer circle Co ~ (22%9)| ——] = 2 x 108 ms ") 10° 6 m2 so! 2m)| “| = 2a x 106m? s. (Note Ci taken opposite to 0 direction of V.) -- Cy +C; ). So Vorticity = 0 in annular region. Cc _(2% «10° ms) Now z, a 6 for inner cylinder. Fig. 4.54.6 4 7 32 Show that the anomalous gradient wind cases discussed in Section 3.2.5 have negative absolute circulation in the Northern Hemisphere and hence have negative average absolute vorticity. Solution: C, = 2aVR + f xR? = ana(v +f &) Here R <0 for the anomalous cases (see Table 3.1). (3.15) and Table 3.1 But from v+B>0. Thus, C, <0, hence § + f <0. Compute the rae Of change of circulation about a square in the x.y plane with sides of 1000-1on length if temperature increases eastward aa rate of 425 C cm and pressure increases northward at a rate of ar km. The pressure at the origin (the lower left cornet) is 1000 mb. Solution: oe = -fer dinp but inp changes only for N - § legs so 2 en) Dr Be _ av eo) Po . a = ~(287)(.005)(5°C) = -7.2 ms? P, = 1005 mb VAN seerunnrntntmmnmnt cantanteSOESTCTUECTTESs — ——_-—— TT TT TTT SSIS IIIT GUGM sss sdesddus 33 4.8 Verify the identity (4.18) by expanding the vectors in cartesian components, Solution: (V+ VW =u(Z)v = (Sv 2) (ZH) se) -f232) (BE BVAGEG) - (2) +0( 2) ) + af #) +02 : Ve = kx tie] ES] nor k= 5, kx joni weve = ifa(Z)- f% SE )- (3) v \dding V| Adding ( 3 Vy ek WE =(V-9)V QED. 4.9 Derive a formula for the dependence of depth on radius for an incompressible fluid in solid-body rotation in a cylindrical tank. Let H be the depth ar the center of the. tank, @ the angular velocity of the tank, and a the radius of the tank. Solution: _Cenwrifngal force must balance the pressure gradient forces, so oh Qrar Dy = (2) dha; or 8 Myr a fa- [2 ha) =H+ 27. g 2g34 4.10 By how much does the relative vorticity change for a column of fluid in a rotating cylinder if the column is moved from the center of the tank to a distance 50 cm from the center? The tank is rotating at the rate of 20 revolutions per minute, the depth of the fluid at the center is 10 cm, and the fluid is inisally in solid-body rotation. Solution: Potential vorticity is conserved so (2324) = é fo 3,2 Hg = 10cm, Hy = Ho +5 (Grom problem 8) initaly Co = Thus ¢, -(42 = Ho) oq = SSDGAB) _ 93 51, Ho (10) 4.11 A cyclonic vortex is in cyclostrophic balance with a tangential velocity profile given by the expression V = v{é "0. the distance rp from the vortex center, Compute the circulation about a streamline at radius r, the vorticity at radius r, and the pressure at radius r. (Let po be the pressure at rq and assume that density is a constant) ] where Vo is the tangential velocity component at net Solution: C = 2nvr =22¥0" % henge VHRTAASAVVVA VV TAVITATETTTTI ssa Ue reise hese reed ede dedddds 4.12 413 35 A. westerly zonal flow at 45° is forced to rise adiabatically over a north-south oriented mountain bartier. Before striking the mountain che westerly ‘wind “I increases linearly toward the south at a rate of = ms . The crest of the mountain range is at the 80kPa (800-mb) level and the tropopause, located at 30 kPs (300 mb), remains undisnirbed by the forced ascent of the air. The surface pressure to the west of the mountain barrier is 100 kPa (1000 mb). What is the initial relative vorticity of the air? What is its relative vorticity when it reaches the crest if it is deflected 5° latiude toward the south during the forced ascent? If the current assumes a uniform speed of 20 m s“! during its ascent to the crest, what is the radius of curvanure of the streamlines at the crest? Solution: eee = Constant. (+o 2) pL & ® ° fy = 2Qsin(45°) = 1,03 x 10~ 5“) Gq = 70 kPa. f = 242 sin40° = 9.37 x 10s, dp, = 50 KPa. : Vv but for uniform current £, = = 2 G = 129 x 107 so R= 2-154 x 10% of R = ~1546 km 1 (anticylonic curvature). A cylindrical vessel of radius a and constant depth H rocating at an angular velocity 2 about its vertical axis of symmetry is filled with a homogeneous, incompressible fluid with is inisially at rest with respect to the vessel. A volun of fluid Vis then ‘withdrawn through a point sink at the center of the cylinder, thus creating a vortex. Neglecting friction, desive an expression for the resulting relative azimuthal velocity as a function of radius 1.c., the velocity in a coordinate system rotating with the tank. (Assume that the motion is independent of depth and that V << ma*H). Also compute the relative vorticity and the relative circulation. Solution: For V << ma?H we can neglect change in depth when considering change, 6r, in radial position of fluid parcels, ‘Thus at radius r, . ¥ 7 -- ~~) mnentu ér Tei bt by Conservation of angular momentum Qo? = OF + ur where ro = original radius and 6 = ro * & is final radius, and u is the tangential relative velocity.where ¢ — 0 and C= 2am = 2 4.14 (a) How far must a zonal ring of air initially at rest with respect to ine earth's surface at 60° latitude and 100-km height be displaced latitudinally in order to acquire an easterly (east to west) component of 10 ms! with respect to the earth's surface? (b) To what height must it be displaced vertically in order to acquire the same velocity? Assume a frictionless atmosphere. Solution: An approximate solution can be obtained from conservation of angular omentum se page 16 intext) an? = [2 ae + ey expanding this expression gives 64 = -2Q8R, thus, for lateral displacement 6: wea: (6ysingg ~ —&); for vertical displacement dr = -—=54 _; (6 cosop = +R). 2.2 sin Go) &y = -79km or = 0.71° equatorward & = +138 km. 4.15 The horizontal motion within a cylindrical annulus with permeable walis of inner radius 10 cm, outer radius 20 cm, and 10-em depth is independent of height and azimuth and is represented the expressions u=7—0.2r, v= 40 +2r, winere u and v are the radial and tangential velocity components in em s“!, positive oupward and counterclockwise, respectively, and 7 is distance from the center of the annulus in cm. Assuming an incompressible Aluid, find (a) the circulation about the annular ting, (b) the average vorticity within the annular ring, (c) the average divergence within the annular ring, and (d) the average’ vertical velocity at the top of the annulus if it is zero at the base. VAVNAT NALA TAA TATA AANASCIGES SOPPEPP PPP RECEP PERE PPP rr rer T eee 37 Solution: C = 2mV(n) — 204 V(n) = 2(20)(80) ~ 2(10)(60) = 2m x 10° em?s“! 2 (a) .C = 6280 = C_ 2xx10° “1 Ae = 6.6757. A (wg - my) © vival ou Jor mae duena-quina. ~ (b) $= (vv... ~ [emamln) —2amule)) 1 1 <6 ogg 5 = [ae - a] aw C @ > Fig. 4.14 4.16 Prove that, as stated below Equation ( 4.38), the average isentropic vorticity on an isentropic surface that does not intersect the ground must be zero. Show that the same result holds for the isobaric vorticity on an isobaric surface.38 Solution: From the vector identity above (4,38) Lg =k + (V_ x V) = V5 -(¥ xk) 2 Jéoe= Jro-crxaar=o since the divergence vanishes when integrated over the sphere. The isonaric cass is proved in the same way. VUELTA TAAIaTaaaaaaaaaaaiitioteie:39 Chapter 5 5.1 Verify by direct substirution that the Ekman spiral expression (5.31) is indeed a solution of the boundary layer equations (5.26) and (5.27) for the case ¥,= 0. Solution: u=u,(1-<%cosp), v= uze*sing, 2, > (2) = 2u,y?sin yw e™, (S) =~ 2u,y?cose ™, gives -27°Ku, ~ f(-u,)=0. QED. i 5.2 Derive the Ekman spiral solution for the more general case where the geostophic 4 wind has both x and y components (u, and v,, respectively) which are independent of height. Solution: In this case the general solution of 5.20 is wvnaaceel ZY] + 2e0{-(E)%] +t =) for the N.H. boundary conditions 5.19 gives A = 0 and B = ~+(u, +iv,). Taking real and imaginary parts we get: Letang the Coriolis parameter and density be contants, show that Equation (5.38) is | correct for the more general Ekman spiral solution obtained in Problem. 2.S.4 ot 40 Solution: For p = constant D, dD, ‘Ou, oy, oy, ou, Re: or -1{ 38 +8) fe ~ eos) ae [St -St) Je ia hee 3 3 o, ou, But eae for constant f, and tate ¢, on tS) (Northern Hemisphere). For laminar flow in a rotating cylindrical vessel filled with water (molecular kinematic viscosity v= 0.01 cms"), compute the depth of the Ekman layer and the spindown time if the depth of che fluid is 30 cm and the rotation rate of the tank is ten revolutions pet minute, How small would the radius of the tank have to be in order that the time scale for viscous diffusion from the side wails be comparable to the spin-down time? i Solution: De == where 7 = (2) out a= 2% 7 v 6 v= —0lom?st, % De = 0.307 em. Spin down time: t = uz) = 2935 aye 2 ‘The diffusive time scale is & where L is the radius of the tank. v Thus £ = (293v) = 1.7 em. ‘Suppose that in a certain region the geoswophic wind is westerly ac 15m! Compute the net cross isobaric transport in the planetary boundary layer using besh the mixed layer solution (5.22) and the Ekman layer solution (5.31). Let f= 104s! h =De = 1km. p = Lkgm™, and x = 0.015m's. You may let }¥)= a, in (5.22), Solution for mixed {ayer case: . M = poh, But from (5.22) Vwi WVVAVVAIIAIIVAIAATIAAAAII 44 i41 | Solution for Ekman layer: >, 2 M= fone = Journ sin yoda 2 3 cM = pu, (27) (1 +e") = 2.49 x 10? kgm where recall 7 = "fp, | $.6 Derive an expression for the wind driven surface Ekman layer in the ocean. i i Assume that the wind stress +, is constant and directed along the x axis. 1 \ ‘Continuity of stress at the air-sea interface (z= (0) requires that the wind sess equal the water stress so that the boundary condition at the surface becomes / eK i | | «(%) = 0 az =0, where K is the eddy viscosity in the ocean (assumed constant). As a lower boundary condition assume that u,v + 0 as z > —e. If K = 10% ms! what is the depth of the surface Ekman layer at 45°N latitude? Solution: For u, =v, = 0. General solution of (5.20) is wire aal(Z)'s] + aen[-(£)| but w+ iv 0 as 2 -c which gives B = 0. nel} EB) .-o ml 2 Asai d= Se where r=(£y STHFTITCETIT Ess SSS SSS sssSTsSSSK STITH: = A,e7(cos yz + sin 72), v = A,e™(—cosyz + sin 72) ye De== (RE) = 14.05 mow 42 Show that the vertically integrated mass transport in the wind driven oceanic surface Ekman layer is rected 90° to the right of the surface vind stress in the Northen Hemisphere. Explain this result physically. Solution: The only forces acting on the water column are the wind smess and the Coriolis force. Thus the Coriolis force on the water column must oppose the wind stress vector, hence the mass transport must 90° to the right of the surface wind stress, From the solution to 5.6: ° 0 Jone = Jovertcon + siny)] dz = 0. 0 ° Jone = Jonertosorr + sin 2] dz = — * kgm ‘A homogeneous barotropic ocean of depth H = 3km has a zonally symmeic 2 L) where U=1 ms! and L = 200 km are constants. Compute the vertical velocity produced by convergence in the Ekman layer at the ocean bottom and show that the meridional profile of the secondary cross stream motion forced in che interior is the same as the meridional profile of Z,. What are the maximum value of and if K = 103 ms and f= 10 s-!? (Assume that i and the viscous stress vanish at the surface.) geostrophic jet whose profile is given by the expression, — vea|-( Solution: #p. = s/f oo = 9.6 «10% mms? VPOTTETT AVL43 Using the approximate zonally averaged momentum equation 2 = f¥ compute % the "spin-down” time for the zonal jet in Problem 11. | i Solution: % (2) (HY 2p (Azzy. 1) u,(t = 0) exp(-+), T= 1.34 x 107s = 155 days. 5.10 Derive a formula for the vertical velocity at the top of the planetary boundary layer using the mixed layer expression (5.22). Assume that |V| = 10 ms~', is indepen- dent of x and y and that Z, = @,(y). [f A =1kmand f = 10s", what value must x, have if the result is to agree with the vertical velocity derived from the Ekman layer solution with K,, = 5 m? s-!? Solution: From (5.22) a a{i+ VP) is F< [Vly (t+ VEY But from continuity . ww =-[(Z +3) de = AW a (1 + e1¥f) y For equality with Ekman layer result (5.38) x{V|a Solving for xs we find xs = 0.016 me! s (the other root of the quadratic is rejected since it leads to an unrealistic distribution for a, ¥ ). SET ISI STITT SUT k kee Sees eee eee e sees e ceded ddds ——— |5.11 Showthat K,, = ku in the surface layer. Solution: From (5.32) a2 4 #22 nu, #=%. Kn, =u But S Thus, Ky = kzity, QED. VAAAALEVERVAVVADED PETE VIED VAVVVVTTA TAT TATAT ATAFIIFTITIES ITTV SSS SSeS GUS SSS SESE SIEEES 45 Chapter 6. ‘Show that the static stability parameter. o = nt may be written in terms of Das . Solution: 1n@ = In of - 2) + In@ + constant, ©. ‘Show that for an isothermal atmosphere, as defined in Problem 6.1, varies inversely as the square of the pressure. Solution: R =D. sotor T = Cons. 2% =~ RT=Const. p oop = (BE) GE)- (ENF) i a » —— |6.3 6.4 46 Suppose that on the 50-kPa (500 mb) suface the relative vorticity at a certain location at 45°N latitude is increasing at a rate of 3 x 10-® s~ per 3h. ‘The wind is from the southwest at 20 m s“! and the relative vorticity decreases towards the northeast ar a rate of 4x 10-6 s~! per 100 km. Use the quasi-geostrophic vorticity ‘equation to estimate the horizontal divergence at this location on a B-plane. Solution: y- v= 43) +: ¥) (6, + a} Ky _ Bx) _ 2.78 x 1071s, & ~~ (3 x 3600) a 4 x 10% V, Vo, = vee = an (3 2"* Br, = (1.6 x 10™)(14.1) = 2.28 x 10-5? ~ VV = —(10%s)[2.78 x 107 -8 x 10719 = 2.28 x 10°")} = 2.84 x 10571 Given the following expression for the geopotential field © = Op) + ca{-[c0e{2) + 1 + eTsink(x — onl where @q is a function of p alone, c is a constant speed, k a zonal wave number, and pg = 100 k Pa (1000 mb): (a) Use the quasi-geostrophic vorticity equation to obtain the horizontal divergence field consistent with this ® field. (ssn that ¢ =0. ) (b) Assuming that © (po) = 0 obtain an expression for (x, y, p.r) by integrating the continuity equation with respect to pressure. (c) Sketch the geopotential fields at 75 k Pa (750 mb) and 25 k Pa (250-mb). Indicate regions of maximum divergence and convergence and positive and negative vorticity advection. (2 (a Solution: V-V= (+)TREES T SEITE SS ESP ESSE EH HSCS EUCEEEE: a7 ~kesink(x ~ ct), 3)-b-(g] 22 @) 2 VeVss Eee (2 coon ~ ct), Po 22 | ) Bow Ve = ~( 22) cor sin (B) costs ~e, i | 1 | ——> Con +— 250 mb H — | | y 750 mb . w a — dv —> Fig. 6.4 6.3 For the geopotential distibution of Problem 6.4 obnain an alternative expression for coby using the adiabatic thermodynamic energy equation (6.13), Assume that o is constant, For what value of k does this expression for @ agree with that obtained in Problem 4? wen: (E-A2)B) oof) om {t2)=(3 Po Po _(1\f a0) oO Go) \ oe6.6 (ol5)3) fale oe-9 x fo2\( =) Agrees with solution in 6.4(c) if k? = (4) ) : o Kp) As an additional check on the results of Problem 6.4 and 6.5 use the omega equation (6.29) to obtain an expression for «2 Note that the three expressions for agree only for one value of k. Thus, the geopotential field © (x, y, p, 1) of Problem 4 is consistent with quasi-geostrophic dynamics only for one value of the zonal wave number. som (+S o- (NRE) Assuming @ = weol 2) cosk(x — ct) get 2) BS) + (=) (2) xo 2) cosk(x ~ ct). \oFo)\ po Po (45) (2) sin( 2 coset - ct). 4 HVGVREVVVVV VATE ET TATA ‘4FAIIATSIIIGTTISATIIFSSSacosscocvovvsccsnsrres” ” 6.7 49 ‘Suppose that the geopotential distribution at a certain time has the form = ™@ =| 5 (x, y, P) = Po(p) ~ for cos| | + Fock sin kx Po. Where Uo is a constant zonal speed and all other constants are as in Problem 6.4 Assuming that f andor are constants, show by evaluating the terms in the ight hand side of the tendency equation (6.23) that z= 0 provided that 2 Lox) Po J Make qualitative sketches of the geopotential fields at 750 mb and 250 mb for this casé. Indicate regions of maximum positive and negative vatticity advection at each level. (Note: the wavelength corresponding to this value of & is called the radius of deformation.) Solution: For the condition given the terms on the right in (6.23) are ~ fo, v(z}re] ~«{23)- geos{ 2 2) (Foc coske), (ae E13) [Bm (5) (3 - 2\ (2 - Fos x = 0 these must sum to zero. Thus &? = (2)(2} oJ pe For ke = have (+, -) vorticity advection at (750, 250) mb.6.8 750 mb H L 250 mb pt 0 © 2x oO = au kx —» Fig. 6.7 For the geopotential field of Problem 6.7 use the omega equation (6.29) to find an expression for @ for the conditions in which x=0. Hint: let @ = Wo cos kx \ sin (2 | where Wo is a constant to be determined. Sketch a cross section in the Po 2 p plane indicating trough and ridge lines, vorticity maxima and minima, vertical motion and divergence patterns, and locations of maximum cold and warm temperature advection. lee e-em Evaluating the r.h.s. in (6.29): (2)(g)m we (VBE) «(Loa rere! (ef Fle ollol)a(2 [Seals )outels) Cold Advection at & = x due to Northerly wind acting on meridional temperature gradient. Subsidence warming balances cold advection. Divergence pattern balances vorticity advection everywhere so that x=0. Solutio: ela I e& = = = = = = eS C E if= Cer eT PPE ese) 6.9 — Given the following expression for the geopotential field, (x,y, p) = o(p) + fou? + ©ve0{ 2] sink(x - | Po where U,V, and care constant speeds, use the quasi-geostrophic vorticity equation (6.19) to obtain an estimate of «a Assume that ¢ = Bisa constant (not zero) and that @ vanishes for p = pp. Solution: ¢,= 4\ve = —W cos| # | sink(x ~ cr), = Po z + ee =RPV(e- Uso 2 ease - ct), ie = BY cos (2) cosk(x — ct), ae =V[P(c - u) + By (2) cosk(x — ct), Po o= (ay) [ete -U) +B] an(2 cose - a).52 6.10 For the conditions given in Problem 6.9, use the adiabatic thermodynamic energy equation to obain an alternative estimate for @ Determine the value of ¢ for which this estimate of @ agrees with that found in Problem 9. tne oS IES) = (sor) (-ke+ Uk) so( 2) cosk{x - ct), Po Po o= (2) wu ~osia( 2 cont - ct). 0 Po Comparing with @ in 9 we require (2) [Pl - 0) + B] = (# ne 8) y -<) 6.11 For the conditions given in Problem 6.9, use the omega equation (6.29) 10 obiain an expression for @. Verify that this result agrees with the results of Problems 9 and 10. Sketch the phase relationship between @ and @ at 250 mb and 750 mb. What is the amplitude of @ if B =2x 10! m's*,u = Sme'V=8ms", win’ fo = 10S, o =2 x 10%Pa*m?s%, and pp = 107k Pa? Solution: SUVA TTA T AAT ETS ETAT itTOT TCC C TTT I II U GIGI U SE See e eee Ree e seeder eee CECEs Fig. 6.11 (0% (* + fen :) Ps. 6.12 Compute the Q-vector distributions corresponding to the geopotential fields given in problems 6.4 and 6.7. [Oban = = 0.0102Pa s™ Solution: ‘The components of Q can be determined from (6.36). In the case of problems 6.4 and 6.7 T is independent of x as can be verified by crating = 82 Also, v, is independent of y. Thus Q. P 2, and g, = — Me, Po Fo Mm py yp x Thus, for problem (6.4) Q, = -Ffs2on{ 2) cosk(x ~ ct). Po \ Po. For problem (6.7) Q; = ~ fate snl 2 costx = ct) Py \ Po!34 6.13 Show that the isallobaric wind may be expressed in the form Veatt = ~ fo? VX where x Solution: From (6.40) VisaTA 7,2 55 Chapter 7 ‘Show that the Fourier component F (x) = Re{C exp(ims)] can be written as F(x) = |C|cosm(x + x9) where xp = miso) and C; stands for the immaginary part of C. Solution: F(x) = Re{Ce*} = C;cosmx — C;sinme, where C = C, + iC;. (a) F(x) = \C| (cos mxcos mxg — sin mxsin mxo) = |C|cos[m(x + x9)]- Define C, =|Clcosmx, \Clsin max. Thus mxy = sit In the study of atmospheric wave motions it is often necessary t0 consider the possibility of amplifying or decaying waves. In such a case we might assume that a solution has the form ¥ = Acos(la — vt ~ keg) exp(ar) where A is the initial amplirude, ce the amplification factor, and xp the inital phase. Show that this expression can be written more concisely as y=Re [pew - “) where both B and c are complex constants. Determine the real and imaginary pars of B and c interms of 4, ak, V, and x0. Solution: y = Re {Bexplik(x — ct)]} = Re {Bexp[ik(x - c1)]}e*", but from 7.1 we can write this as Y =|B)oos (kx — vr — key) e**” = Acos(ix - ur keg)e™, where A= (Bl, @ = ke;, So B, = Acnskx, B;73 56 Several of the wave types discussed in this chapter are governed by equations that are generalizations of the wave equation ay _aey aE “This equation can be shown to have solutions corresponding to waves of arbitrary profile moving at the speed ¢ in both the positive and neganive x directions. We consider an arbitrary initial profile of the field y: Vef@ a 520 If the profile is translated in the positive x direction at speed ¢ without change of shape then y = f(x), where x’ isa coordinate moving at speed cso that x=2'+ cr, Thus, in terms of the fixed coordinate x we can write y=f(x—ct), corresponding to a profile which moves in the positive x direction at speed without change of shape. Verify that y=/ (x ct) is a solution for any arbitrary continuous profile f(x~ct). Hint: Let x—ct=x' and differentiate f using the chain rule. Solution: Assuming that the pressure perturbation for a one-dimensional acoustic wave is given by (7.15) find the corresponding solutions of the zonal wind and density perturbations. Express the amplitude and phase for u' and p° in terms of the amplitude and phase of p’. ) a i Solution: ( sa2)e= -3)2 - (8) geet jt(z ~ et) Assume asolution of form u' = B e' then ik(i —c)B = -(2)4 or Pp7.5 Show that for isothermal motion (= THSTTIIII I Ss Ise Ia sss ssdsddsseseda 357 Maa), Thus: C = 0) the acoustic wave speed is given by igh)? where H = 52 isthe scale height. g Solution: p = pRT gives in p = Ing +InR + nT. , Dine _ Dine Dt ‘Comparing with (7.9) we see that for she isothermal case the equations are identical to the isentropic case except for the factor 7. . Thus c= @ + (RTM? =a + (gh)? In subsection 7.3.1 the linearized equations for acoustic waves were devcloped for the special situation of one-dimensional propagation in a horizontal mbe. Although this simmation does not appear to be directly applicable to the annosphere, there is a special atmospheric mode, the Lamb wave, which is a horizontally propagating acoustic mode with no vertical velocity perturbation (w’ = 0). Such oscillations have been observed following violent explosions such as volcanic eruptions and atmospheric nuclear tests. Using (7.12), (7.13) plus the linearized forms of the hydrostatic equation and the continuity equation (7.7) derive the height dependence of the perturbation fields for the Lamb mode in an isothermal basic state atmosphere assuming that the pressure perurbation at the lower boundary (z = 0) has the form 7.15), Determine the vertically integrated kinetic energy density per unit horizontal area for this mode,Solution: Equations (7.36) and (7.38) together give 58 WAI - Cae Ee Linearizing (7.7) and combining with (7.13) to eliminate u’ yields = 2652) - 22 o22)p = ($+#2)o Bla” a — e which implies that p’ = c2p’, ie j € Combining (i) and (i) to eliminate p’ gives vertical structure equationg=* for p’: Ee a € a cP a Thus p’=A ene exp [ik(x — c1)]. S From substitution into (7.12) we obtain acel? (ii where we have used c? = RT and H = RE. g ae? 2 2 put 2(4)-i4 Thos, fi 2 22\3)" 6 8 a VAAVAVAVAVVVAVVVVIVVIAIGG_ POPPE PPP r Ree err b 59 IF the surface height perturbation in a shallow water gravity wave is given by Wr = Re[ Ae 9} find the corresponding velocity perturbation u'(x, 1). Sketch the phase relationship between A’ and u’ foran eastward propagating wave. Solution: Let w = RefBeH= 9} Substituting into (7.21) gives ik(@ — c)A =~ ikHB. (c- aK S-: ee Fig. 7.7 Assuming tha the vertical velocity perturbation for a two-dimensional internal gravity wave is given by (7.43), obtain the corresponding solution for the u’, 7’. and & fies, Use these results to verify the approximation | v| & & | 9 | >> @ which was used in (7.36). Solution: From (7.39) # 30" | 8" dB) 8g _ (id (ae), From (7.40) 38 (2) = 0 =-(5 (2) a From (7.37) (218) =o, a Po. ‘These may be checked by substitution into (7.38) and use of (7.44)w, Be : “f| Clearly the ratio is large for L, << Hand |m| >> |k| where L, = 2% m Itis also large for m => 0. - 7.9 For the situation in Problem 8, express the vertical flux of horizontal momentum, Pou'w’, in terms of the amplitude A of the vertical velocity perturbation. Hence, show that the momentum flux is positive for waves in which phase propagates eastward and downward. Solution: po(u’w’) = -( since (cos*tx) Note Po(u’w’} > 0 for m <0, which corresponds to downward phase propagation if v > 0 and ¥ > 0. 7.10 Show that if (7.38) is replaced by the hydrostatic equation (i.¢., x i neglected) the resulting frequency equation for internal gravity waves is just the is asymptotic limit of (7.44) for waves in which 11 <0 and c,<0 v nm but cf = U2 so magnitudes of zonal phase and group velocities are equal. Fig. 7.11 be . FITIT TISAI GS eS SSS SCI CCU ESE S CSET STC CCCCEETETS62 7.12. Determine the perturbation horizontal and vertical velocity fields for stationary gravity waves forced by flow over sinusoidally varying topography, given following conditions: the height of the ground is k= hy os kx, where hy = SU su is aconstant; N=2x 102s; @=5ms7; and k=3 x 103m Hint: For small amplitude topography (ik << 1) we can approximate the lower boundary condition by Solution: Lower boundary condition gives w’ From (7.43) general solution is w’ = Acos(kx+mz) +B sin(ke + mz) 2M where m? = > - k? from (7.46). Lower boundary condition @ requires A = 0, B = —iibyy, Thus w’ = ~iikhy sin ice + mz); ul! = iimhysin(kx+mz). So w’ ~(0.750ms") sin (lec-+m: 7.13. Verify the group velocity relationship for inertio-gravity waves in (7.67). Solution: For {= 0 (7.66) gives v? = f? + = ‘ah 7.14 Show that if Z=0 the wave-number vector * for an intemal gravity wave is perpendicular to the group velocity vector. % Vn 4 Solution: « = (k, Bur from (7.45): (c5,, ¢ =k) ‘Thus, K+ c, = 0, which shows that x and ¢, are orthogonal. AMAA VAANSECIS SCETESSSSESSSETISUSSSESETITSCS SU IUTCETTUUN 63 7.15 Using the linearized form of the potential vorticity equation (4.26) and the B-piane approximation, derive the Rossby wave speed for a homogeneous incompressible ocean of depth A. Assume a motionless basic state and small perturbations that depend only on x and 1, usu(xt), vev(at =H +h) where Hs the mean depth ofthe ovean, With the aid ofthe continuity equation for a homogeneous layer (7.21) and the geostrophic wind relationship v’ = gf5! = ‘Show that the perturbation vorticity equation can be written in the form ~ a(a fi), , gan SSH) ed and that A’ = hye**~ ) isa solution provided that oe 8 If the ocean is 4 km deep, what is the Rossby wave speed at latitude 45° for a wave of 1000 km zonal wavelength? Solution: From the quasi-geostrophic vorticity eq. (6.10) we have BB) Bem e-3)-° For conditions of present problem the linearized version is ee) ne ou »_ {a)(any Si) + oss Wao bw yy = (2S au 1) an From (7.21) ze \(#). we ofa? 8 I _ fy MY _ oy which easily Thus we have a(4 Se) ap ao wl ;“4 reduces to form given in problem. Letting h’ = Aexp[ik(x — cr)] we get wine]? - (é } + ikB = =157 x 107" Z ne 10 | |e x10) + Sayaao thus, ¢ = -24.3ms7. 7.16 In Section 4.3 we showed that for a homogeneous incompressible fluid a decrease in depth with latirade has the same dynamic effect as a latitudinal dependence of the Coriolis parameter. Thus. Rossby-type waves can be produced in a rot cylindrical vessel if the depth of the fluid is dependent on the radial coordinate. To determine the Rossby wave speed formula for this equivalent Preffect, we assume that the Flow is confined berween rigid lids in an annular region whose distance from the axis or rotation is large enough so that the curvature terms in the equations can be neglected. We can then refer the motion to cartesian coordinates with x directed azimuthally and y directed toward the axis of rotation. If the system is rotating at angular velocity and the depth is linearly dependent on y, HO) =Ho- show that the perrurbation continuiry equation can be written and that the perturbation quasi-geostrophic vorticity equation is thus, ag Lvty + pM = Sw +B 0 where ’ is the perturbation geostrophic streamfunction and = 1s the Rossby wave speed in this situation forwaves of wavelength 100 m in both thex and y directions if @ = 1s", Hy=20cm,and y = 0.05? Hint: Assume that the velocity field is geostrophic except in the divergence term. |errr sere ee > — : Solution: 65 Now, H =Hp— yy socontinuity equation lines.ized about a state of rest is by analogy to (7.71) 268) BE) or mw Hl x \-0 when || << Hy. x ral oy Now (=) + (3) + ({&) + 20 -< ) avy’ ( ou’ Vey’ soli + 22) Y" solinearizing: > & but aw’ +B =0 avey’ a and substituting from continuity: when = oy ox Leaing y’ = A explik(x- ct)] exp(ily) gives c = ery jal 2 $)(05)(100 =) 2m)(2 x zal Im? = - 0.633 ems"! 7.17 Show by scaling arguments that if the horizontal wavelength is much greater than the depth of the fluid, two-dimensional surface gravity waves will be hydrostatic so that the "shallow water” approximation applies. Solution: Let U, W, H, L be scales for zonal velocity, vertical velocity, depth and length. Ww UH u From (7.18) 2 ~% so w ~ UA rom (7.18) a? : From zonal momentum equation: (2) af ( * ae) * py ae we get 2 ~~ where P is pressure scale and + is time scale. Pot6 Verical momentum equion () + 2) tee s o aw) Wo Un then gives (%) ~— ~ 22 mee (f)-¥-8 1 2) Pub —|S)->y-qG © (2 PoH tH Comparing (a) and (b) we see that = can be neglected when 2 <<. << 7.18 The linearized form of the quasi-geostrophic vorticity equation (6.18) can de written as follows: 9-902, oy’ — |v —=-f)5V-V (Srag)ey +e --% ‘Suppose that the horizontal divergence field is given by V-V = Acoslk(x - ct)] where A is a constant, Find a solution for the corresponding relati field. What is the phase relationship berween vorticity and divergene: value of ¢ does the vorticity become infinite? Solution: Assume solution of form w= 8 sin (k (x cr)}. Thus (3) +e(S)v -6(%) = (Re — Uk? + Be}Bcosk(x-cr) = foA Gosk(x - ct) . 8B SoA (¢ - BR +Orr eee errr errr errr rrr 67 . Vy = aa” k(x - ct). Vey - 0 for (c= ak? +B =0 oF exa-§ which is the free Rossby speed. For (¢ = a) 8 vorticity leads divergence by © radians, (ce., maximum vorticity is to the east of maximum divergence. iB z For (e ~ i) <—Dy vorticity lags divergence by radians fe an expression for the group velocity of a barotropic Rossby wave. Show that for stationary waves the group velocity always has an eastward zonal component relative to the earth. Hence, Rossby wave energy propagation must oe downstream of topographic sources. Solution: From (7.91) the frequency is v = ak - —* Now for stationary waves v = 0, or @ = _bl@ +2) plit-A) and U, = ~~ + ——>- fo (eeryp (e+e 28K Thus, U, = wea? 0.8.1 68 \ Chapter 8 ‘Show using Equation (8.26) that the maximum growth rate for baroclinic instability when f=0 occurs for = 20JT=1). How long does it take the most rapidly growing wave to amplify by a factor of e! if A = JEx% 10% m7 and Uz = 20ms!? Solution: Maximum a occurs for the & at which 2 y _ aa? (ae +2 } pata ef 2a? or k* + 4474? - 444 = 0. Thus Cmax = 222(V2 ~ 2 2 then Cima = baw tel eRe = = YEAUy (V2 - 1), A=2x10% mw for Up=20ms 2.34 x 10° + Gmax e! amplification time is ami, = 4.3 x 10° s or = 12 hours. Solve for Wy’ and w' in terms of y/; for a baroclinic Rossby wave whose phase speed satisfies (8.24). Explain the phase relationship between yy. Wx, and @z in terms of the quasi-geostrophic theory. (Note that Up=0 in this case). Solution: Here c - U,, bana UO, = Us= Une (KF + 227) From (8.19) and (8.20) it is then clear that A=0 sothar yy=-yi * Then from (8.12) (f+ Um (vs - v4) = FoPPP PPPOE PPP PPP ry, 6 2ikfo Bui (ys) or of = (vi ~ ¥§) or of = TO sa) Thus y' and Wy’ are 180° out of phase, a, lags wy’ by 90° phase. Maximum upward motion occurs 90° east of the ridge at 250 mb and 90° cast of the trough at 750 mb. The convergence- divergence pattern thus generates vorticity changes which partly cancel the planetary vorticity advection (fv), 8.3 Forthe case Uy=—U; and =A? solve for y’3 and @ interms of w, for marginally stable waves (i.e., 5 = 0 in (8.22)]. Solution: Here Uj, = 0, Uy= -Uy= Ur butsince 6 =0 and 22 =% have from equation (8.22) we A B= v3 ’ t+ Ws . + =1) Hence, yf — yj = Wit vA) w-wh. so Vj, WS are in phase. From (8.10) (Up — ¢) (ky) + kB w’ = 2 ay op Substituting for Uy, k, and c this yields: isp 4, (1 3) ’ w= Py ¥ = B | i Maximum upward motion occurs 90° west of ridge. Divergence- convergence field tends to compensate vorticity advection at both 250 and 750 mb levels. 4 8.4 Forthe case 6=0, = 22, and U,=+Uz solve for y's and a’, in terms of v;. Explain the phase relationships berween @'x,, yy, and 3 in terms of the energetics of quasi-geostrophic waves for the amplifying wave. |8.5 70 Solution: Here U,, = Ur so Uy = 2Uq, U3 = 0. From (8.25): ¢ = a + 3) w From (8.16) we have for this case +2(C — Un) Vm — PU Vr 20, OF Vr = we v3 Taos, (vi v3) = Jy (WE YS) and “wr From (8.10) 03 = (2) a, ~ el (-?) ¥% 0 (Z)wu.(4+ 3) so, 05 =| Jitu, | -1+ = 2 (). B Thus, 7/3 leads y/, by 60° and a2 leads v/; by 60° Suppose that a baroclinic fluid is confined between two rigid horizontal lids in a rotating tank in which f= 0 but friction is present in the form of linear drag proportional to the velocity (i-e., Fr =-— V). Show thar the two-level model perturbation vorticity equations in Cartesian coordinates can be written as (24,2 \ eu Lore (Frayer ese wan? where perturbations are assumed in the form given in (8.9). Show that the phase speed satisfies a relationship similar to (8.22) with B replaced everywhere by ijuk, and that as a result the condition for baroclinic instability becomes es (2a? = )# Ur > Solution: Ifa linear drag— Vy is included equations (8.5) and (8.6) become, TVUVVATVT TUT TUT VTTTV VDEVVLUVVLVUDT VD DT TUTE EEECEP errere ree er eer rrer eee rrr rer ree 71 2 \g2 2 fe 2 hv Vs. V(V2yy) =| 22 \o, - pv’ (3) 3+ Vs- ¥(¥2ws) Bp lor BV Ws. Linearizing by letting Wa Uyt Wi Gt), Ys=—Usy + Vt), we trivially obtain the forms given in the statement of the problen.. Defining Yj, and yp fields as in (8.15), these can be rewritten in 2 the form (8.16), (8.17) with pd replaced by 0, Assuming that solutions exist of the form (8.18) then gives: ik(c - U,) — w]k2A — dP Uy B= 0 fr [ik(c ~ Up) (2 + 24°) - we | B~ ikUr ‘This yields a quadratic equation in ¢ that is equivalent to (8.21) if we replace B by ink. 227) A sm ecu, em tot ot ws, C= Uy “Ee ay] wat sbat-2) b=- iP + 227) Oe 24’) [ For instability ¢ must have a positive imaginary part. Thus 6 must be negative and 2 1p (2 + 4?) 2 (eb +2y |e" |> "Te? +a) o l8l> a fee + 22} 2 which implies Up? = for marginal stability. (oH? i . eB we ‘Thus the marginal stability curve satisfies —y = 1- —“—y. 22 (qua?) Noe for 3G Uo Be ‘on marginal curve and for k= 32 Uy? — © on marginal curve.8.6 Unstable wu Fig. 85 For the case B= 0 determine the phase difference between the 250-mb and 750-mo. geopotential fields for the most unstable baroclinic wave (see Problem 1). Show that the 500-mb geopotential and thickness fields are 90° out of phase. Solution: For k? = 24?(VZ — 1) equation (8.25) gives c-U, = Up (VE — 1)? From (8.19) we then have [le - Uni? 4 = [Up(v2 - IP] a = PUpE Thus, B = i(VZ - 1) 4 or vp = iV 1)", so that Yip lags W_ by 90° (coldest air is one quarter cycle west of trough). Then yj = v5 = i (VE - 1)? (yi + v4) or v5 = va[(v2 ~ 1) - 2 (v2 -1)"7). v2(vz - 1)7) Hence @ = sin (WE =1) 265.5" So y3 field leads wf by 65.5°CICTESTESSSSSSESEEESESSSSSTEESEESEESESECESTEECEE i‘ 3 For the conditions of Problem 8.6, given that the amplitude of y,, is A = 107 m2 sw} solve the system (8.19)-(8.20) to obtain B. Let A? =2x 10-2 m?, and Uyp=i1S mst. Solution: From problem 8.6: B = i(V2 - 1)7 A. ‘Thus, the amplimde of B depends only on A, noton Uz or 42, |B) = (v3 - 1 4 = 6.436 x 10m? s Note also that | v5] =| y{| in this case. For the situation of Problem 8.7 compute wz using the expression (8.29) Solution: Now, ¢, = V7 y,,. Thus, if we let exp [ik(x —cx)], and a5 = C exp [ik(x - ct), then substimuting into (8.29) gives Sf i + 20) C = (2 )en200 A which after leting &? = 247 (VZ - 1) i4fo 4 3/2 229 4 (v=) UPA aap yr Up Thus, if A is real so that Yq = A cosk(x ~ ct) yields C = - then of = oan = 1)? UA sinklx ~ ct). So maximum upward motion is 90° in phase to east of $00-mb trough.8.9 14 @, Phase diagram for problems 8.6, 8.7, 8.8 Compute the total potential energy per unit cross-sectional area for an atmosphere with an adiabatic lapse rate given that the temperature and pressure at the ground are p=10°Pa and T= 300K, respectively. ¢ ¢ Solution: Ep + E;=—£, = Joe = [ot & & ° ° ABe is constant and equal to temperature Ty at p,. 2 zy" - fae Ps \Ta. 1 g TEUEEERERUNUEOUUULILE DERE O ROALD DADRA A AAA AADTTT ITI TTS TITS III III IE IGS SEIT I SSeS SUVS S UCC CCE EUI OE erate ste ie 75 (27 cp) (mempert} (ry (Fon T e/R* . 0 0°) = 2.38 x 10° Sm, 8.10 Consider two air masses at the uniform potential temperarures 9, = 320 K and 0, = 340 K which are separated by a vertical partition as shown in Fig. 8.5, Each air mass occupies a horizontal area of 10*m* and extends from the surface (Bq = 10° Pa) to the top of the atmosphere. What is the available potential energy for this system? What fraction of the total potential energy is available in this case? Solution: Using results of problem 8.9 we have for the total energy the initial Sate a) hes [4 + 6] (eo?)a00syo" m?) [320 + 340] = 5.259 x 1013 8) (1+ Re,) oat + 3) To compure energy in final state note that with the aid of the hydrostatic equation: - @ fro = = fro me Hey but T= 6 (2] » Thus for unit area But in the final state we must multiply by area 2A = 2x 10¢ m?. Thus total potential energy is76 foe) [a(n - ) + 040.5) (2,{10*)(1004)(10°) Q [320(.580) + 340(.4102)] = 5.230 x 107, Available Potential Energy = Initial Potential Energy — Final Potential Energy = 5.259 x 10° ~ 5.230 x 10% = 29x 107 0.029 5.259 VATED ATTA VY « 107 = 0.055% of total energy is available. 8.11 For the unstable baroclinic wave which satisfies the conditions given in Problems 8.7 and 8.8 compute the energy conversion terms in (8.38) and (8.39) and hence obtain the instantaneous rates of change of the perturbation kinetic and available potential energies. i aK’ (2fo\ Solution: From (8.38) — {2) Opyr Let y, = Acosk(x — ct) then from problem 8,7: vr =—(v2 ~ 1)" A sink(x— ct) and from problem 8.8: “ihn ~ 1)? Uy sink(x~ ct) But sin?k(—cr) = so OEY; (WE - vtar(222) a0, and & = 4(VE = 1242 PUp7 a where we have used 1? = 20 (08 Then if 2? = 2x 107, Up =15, A= 107 & = 2.91 x 1077 kgs, Similarly, from (8.39) the eddy potential energy generation term is 2A? UpA*k (V2 — 1)! = 247UrAt VE (V2 - 1) = 4.97 x 1037 kgs Thus & = 4.57 x10 = 2.91 x 107 At = 2.06 x 107 kgs 8.12 Starting with (8.63) and (8.64) derive the phase speed c for the Eady wave given in (8.69). Solution: Substimting from (8.65) into (8.63) and (8.64) gives: 2 ik(Az ~ ae +P =F] 0 R(Az - 0) Se RWA = 0. dz From which (8.66) and (8.67) follow immediately. Substiuring (8.68) into (8.67) at z= 0 gives Substituting (8.68) into (8.67) at z=H gives osha H — £ sinhatt)78 ~A(sinnart - coshart | =0 which can be rewritten as 2 aye waite» (SHE coshalt A) =o @ oH sinhaH From which it follows that AH “hi - 4coshaH _ AM, AH), _ 4 cosha 2 2 @H sinha H 8.13 Unstable baroclinic waves play an important role in the global heat budget by transferring heat poleward. Show that for che Eady wave solution the poleward heat flux averaged over a wavelength, is independent of height and is positive for a growing wave. How does the ‘magnitude of the heat flux at a given instant change if the mean wind shear is doubled? Solution: Wecan let yw’ = ¥/(z) costly exp[ik(x — cr)]. [since sin*(x) = cos*(x) = } §in(xyeos(a) = 0 TTT TUTTTTTTTTTTTTTTTTTT TTTTST T TTT STITT T IIS II ISS esse CSS SSS ESSEC SSK KK CK CESC 79 But from (8.68) ¥(z) = A sinhaz + Bcoshaz and B = Thus, if we iet A be real Y= A [sina se Acoshoz, av, ad, cA? + yw, Sie (cosh? cz — sin#? cz) az a A 242 = AEA’ (positive for ¢; > 0) A independent of height. From (8.69 ¢; = A, thus, VT « 4 does not depend on magnitude of A.80 Chapter 9 Show by transforming from 6-coordinates to height corrdinates that the zonal mean Entel potential vorticity P is proportional to F?N? ~ S*. (See equation (9.28). Solution: P= as - (%2 in @ - coordinates Oi, ou, i, (30 -|—t om Pe —/S oH) (FH) -S3), di, _ 9%, (96 _ -200 (a6)' Sts 7a(2) ~ =e a] (2) #3 (2) where we have used (9.10). Starting with the linearized Boussinesq equations for steady waves in a basic state zonal flow that is a function of height, derive (9.31) and verify the form given for the Scorer parameter. Solution: For steady waves in mean flow become: equations (7.37) - (7.40) a LRRVVVVVVVVAATATVAT ATV VV AT AVAT VAT ATTNTTT TTT ST TT EET IIS U ISIE IIS S si S SSSI ESSEC SKE EKEEECES 9.3 81 ae + 4 x @ ES . = (3) ae” ‘ is ) Subtracting —— 2D rom 2 (a _ av (F & -%)- “0 © Using (3) and (4) we can eliminate u’ and 8’ in (5) to get An air parcel at 920 mb with temperarare 20° C is saturated (mixing ratio 16 g kg~). Compute 6, for the parcel. Solution: 6, = oer e= 7(2)* ? in 8, vor(2)n (2) + (oe) (er) . 2) jy (100) , (25% 10%(t6 x 10) = In(293.15) + (2) m( 92 ) * ( [aooa (asa) |" In 0, = 5.84, 8, = 344K. Suppose that the mass of air in an entraining cumulus updraft increases ‘exponentially with height so that m = mge*/” where H = 8 km and mg is the mass at arreference level. If the updraft speed is 3 m s“! at 2 km height, what is its value ata height of 8 km assuming that the updraft has zero net buoyancy?82 Solution: From equation (9.52) a neutrally buoyant updraft has d 2) _ _(2dinm (SZ) abe = -( Ee) for m = exo 7 eZ ~»=-2 [ow] 227 G8 -2=-F but w=3ms" atz=2 Thus w(z = 8)=142 ms". 9.5 Verify the approximate relationship between moist static energy and @e given by 9.41). Solution: Differentiatiang the 2nd line of the solution to 9.3 and multiplying through by ¢,T yields ~RT pT dinBy = cyaT dp + Ledgy~ Ley di0T Substituring from the ideal gas and hydrostatic equations gives aE ap = Lap = gaz P ” Pp Bor de = BG ang [OL ce 1 £.4,dinT = dint dinT (as can be verified from a thermodynamic chart), Thus ¢,T din@, = cydT + gdz + Lda, = dh 9.6 ‘The azimuthal velocity component in some hurricanes is observed to have a radial 2 dependence given by v4, = ¥(2) for distances from the center given by r > 1. r Letting Vp = 50 m s"! and 7p = 50 km find the total geopotential difference between the for field (F — o) andr = rq assuming gradient wind balance and fy = 5 x 10° s, At what distance from the center does the Coriolis force equal the centrifugal force? a8 2 Solution: From equation (9.61) “OZ. + yo = 2 r HEEVEVIVA veil PUVVVS VTLS TAA \ rd \UNVVhhIIIT TESST SSeS eee SESS ESS SKIS ESSE SE SEC CETE! 83 Integrating in r gives Jee = iva Jee + Md fre o -% 2 + Moro Thus O( 2 P(r) + or D(%) = P(r) = x + (5 x 10°)(50)(5 x 10) = 750 m*s, ‘The pwo terms on the left in (9.61) are equal in this case when Starting with (9.61) derive the angular momentum form of the gradient wind balance for an axisymmetric vortex given by (9.62). fe By deSinition My = yar +,84 1 Chapter 10 10.1 Staring with the thermodynamic energy equation (2.42), derive the log-pressure version (10.5). Solution: Using the ideal gas law (2.42) can be written. br RT, _f Dt cp ep - - 2) H Dz Dinp where here w* = —— = -H. Dt Dt Substimating from (2) into (1) and noting that x = gives op or xt os yet Dr #H op 10.2. Show that in the o-coordinate system # mass element ppdxdydz takes the form -gpsdxdydo. Solution: By definition o = 2. But p, P(X Ys #) Pr P08 (from hydrostatic equation). Poads -. pydedydz = =F dead. 10.3 Compute the mean zonal wind Z at the 200 mb level at 30°N under the assumptions that Z = 0 at the equator, and that the absolute angular momentum is independent of latitude, What is the implication of this result for the role of eddy motions? VATVATAT TVATTTTTETEESESIESSSSCSSSSAESESCKHKSCESCECTTSE 85 Solution: By conservation of angular t.omentum: bur if Ro jis distance from axis of rotation at 0° and R, at 30°, ten = 24 8(8-#)_ am _ yyy m u a ra 4> ‘This is far greater than the observed velocity, and indicates that there must be an eddy momentum flux divergence (see Figure 10.6) between the equator and 30°N to balance the tendency of the mean Hadley cell to produce a constant angular momentum profile in the upper troposphere. 10.4 Show by scale analysis that advection by the mean meridional circulation can be neglected in the zonally averaged equations (10.11) and (10.12) for quasi- geostrophic motions. Solution: To estimate terms in (10.11) let scales be Z ~ 10 ms“, ‘time scate) ~ 10° s, and from Figure 10.6 note that aw) < 10% ms? Also let y scale by L, ~ 106m, vertical scale be H ~ 10* m. Then terns in (10.11) scale as follows: ou S - 104 ms? ‘And for quasi - geostrophic flow oH] ops and | S| < te Thus since fy = 10s“, Jo] sms“. then @ ~ 15 ~ 10% ms, c86 195m ods Thus ,— ~ 107° ms Loy Hence leading terms are a oe. For thermodynamic equation (10.12) we first observe that from 5 S08 . . ; -_@ hypsomemric equation magnitude of T ~ > but by geostophy @ ~ foUL hence term ee (imsy(%) ~3x 10%, but 3x 10%. Nin _ {107)(10~)10* Re 287 ~ sin | z\.| Z| vertical advection of 7 can also be ignored. 10.5 Show that for quasi-geostrophic eddies the next to last term in square brackets on the right hand side in (10.15) is proportional to the vertical derivative of the eddy meridional relative vorticity flux. Solution: By geostrophy u’ = >; ae 0 Visa » iTITTIES SII Ieee Sc Sse esEES CESK E KEK EECES ous Me 87 10.6 Starting from equations (10.16)-(10.19), derive the governing equation for the residva) streamfuaction (10.21), ko Solution: Taking fo 0017) + ae (10.18) and applying the thermal wind relationship we = RH But by definition py¥* = ~22", pow* = so after multiplying (A) by pp we get 10.7 Using the observed data given in Fig. 10.1 in the text, compute the time required for each possible energy transformation or loss to restore or deplete the observed ‘energy stores. (A watt equals LJ s~!.) Solution: Dividing the energy stores by each possible transformation and source/sink we get replacement/depletion times as shown in the diagram (all times in seconds): 1.67E7 2.2386 ~ | 22366 — 1.04E6 2.256 — Po = 1.67E7 | T.Q9ES 1.856 | 4.0E5 x 3.656 — | 1256 2.9E6 4.6E5 « k | ca > 10.788 10.8 Compute the surface torque per unit horizontal area exerted on the atmosphere by topography for the following distribution of surface pressure and surface height: Ps = Po + Psinke, h = Asin(kx - y) where py = 10?kPa, p = 1 kPa, & = 2.5 x 10°m, =z = Y= Grad and = Coe)” where $ 24 rad is the latitude, and a is the radius of the earth. Express the answer in kg 52. hkcos(x — y) where Solution: Torque = ~(ec0s0) (p,2) but 2 & = (acoso). * (acoso) = heos(kx - 7) +. torque = ~ ph (sinkxcos(kx - 7) =-ph ((sinkecoske + sin?kx) sin’ (2) siny. Since (sin?kx) and (sin kxcosk) = 0 3 Torque = -[o (222) sin ()- 6.25 x 10°kg 5? AT= 89 10.9 Starting from (10.66) and (10.67) show that the group velocity relative to the ground for stationary Rossby waves is perpendicular to the wave crests and has a magnitude given by (10.68). Solution: Fig. 10.9 | @ is the angle between group velocity and x-axis, @ is the angle between wave crests and x-axis From (7.91) we see that for stationary waves (v= 0) =o. (P+?) | Substituting into (10.66) and (10.67) we get - Wee 2 \ 2k sat a! i = core ‘Thus. 9+ @ = 90°, as was to be shown, TIETITIII TEs I see i sds ddddcedddddddddddddddedsK ee90 10.10 Derive the expression (10.76) for the thermal wind in the dishpan experiments. Solution: Differentiating the geostrophic relationship with respect to z and substituting from the hydrostatic equation gives 202M 21 vB a2 x Vp oz Po H Py but from equation (10.75) k x Vp = poek x VT ov, . ts Se 1 . “> 29 ** VF QED. 10.11 Consider a thermally stratified liquid contained in a rotating annulus of inner radius 0.8 m, outer radius 1.0 m, and depth 0.1 m. The temperature at the bottom boundary is held constant at Tp. The fluid is assumed to satisfy the equation of state (10.75) with pp = 108 kg mr? and ¢=2x 10 K-!. If the temperature increases linearly with height along the outer radial boundary at a rate of I°C cnr! and is constant with height along the inner radial boundary, determine the geostrophic velocity at the upper boundary for a rotation rate of = 1 rad s-! (Assume that the temperanure depends linearly on radius at each level.) Solution: From equation (10.76) 44GB) 91 GD] mrs az) | the distance from the inner wal} and 27 ig the thermal gradient Sls along the outer wall. PEMA AANA ATTA EEE EEL TTA » \“Gale. . E x ws ovo(25)[ 10.12 Show by considering 4 for small perturbations about the equilibrium points in Fig. 10.17 that point B is an unstable equilibrium point while points A and C are stable. ot Solution: DW - Kz - U) For equilibrium D(z) -x(i-U) From Fig. 10.17 itis clear that for points A and C if 52 >0 then D(a) > K(Z-U) 0 H 0 then D@) < x(@-U) 0 S>0 if Si <0 then D(i) > (a ~U) so <0 In both cases. 2 is same sign as 57 so perturbation amplifies. and equilibrium is stable.Lt 11.2 92 Chapter 11 Suppose that the relative vorticity at the top of the Ekman layer at 15°N is G=2x 105s, Let the eddy viscosity coefficient be Km = 10 m*s~!, and the water vapor mixing ratio at the top of the Ekman layer be 12 gkg~!, Use the method of Section 11.3 to estimate the precipitation rate owing to moisture convergencs in the Ekthan layer. Ka)" Solution: P = o(& q os 2 -5.-1)|__10.m?s™ |” 3 = (Ike m™)(2 x 10s fox 10} (12 x 10°) = 7.59 x 10%kg ms Dividing by the density of water p,, = 10°kg m> gives 2 = 7.59 x 10-*ms™ or 6.56 mm/day. ‘As mentioned in Section 11.1.3, barotropic instability is a possible energy source for some equatorial disturbances. Consider the following profile for an easterly jet stear the equator: where up, yo, and J are constants and y is the distance from the equator. Determine the necessary conditions for this profile to be barotopically unstable. Solution: |B - a < 0 somewhere for instability () = ~upsin"[2(y - »)]. H, = ~2upisin{i(y ~ )] cos{i(y ~ »)], i, = 2upl?{sin*[I(y ~ y)]- cos? [!(y - »)]} a 2up!. Hence 2up!? > B for instability. VEDA VAT V100099999999099959999999999999009999COC CECCOE EERE Peer eee errccec eRe ih 93 11.3 Show thar the nonlinear terms in the balance equation (11.15) 1 Gy) = -v(ivy . vv) +¥-(Vyv2y) may be writen in canes coordinates as own) 5-3] Solution: Let A= -vi(2e rmnm a9 =2[ 2% (Fey 11.4 With the aid of the results of Problem 11.3, show that if f is assumed to be constant the: balance equation (11.15) is equivalent to the gradient wind equation (3.15) for a circularly syametric geopotential perturbation given by94 where @p is a constant geopotential and J. a constant length scale, Hint: Assume that ¥ (x,y) Solution: has the same functional dependence on (x, y) as does ®. For isobaric coordinates (3.15) becomes wf ,(fP_ 9 v-%. 23/2, x) where r= |ix + jy| is the distance from the axis From (IL15) V7 = f¥?y +2 4@y _ 4, By , Thus Soo = Atos ve. Solving for yo yields 272 v2 vo=-Lt 3(42 ae + 22) 4 72 Multiplying B by 5 we see that expression B is proportonal to A, as was to be shown. 11.5 Starting from the perturbation equations (11.29) — (11.31) show that the sum of kinetic plus available potential energy is conserved for equatorial waves. Hence, show that for the Kelvin wave there is an equipartition of energy between kinetic and available potential energy. Solution: Take u’ times (1129) + v’ times (1130) + w times (11.32); 8! the result is 1a or a wane palsy Sou os ej=0 VORA AVY 090000999999999999090099999990000099COCCCCCSE TETTTT TTT TESTE sTTIELEsTescdecddccdscdseds 95 ‘Then averaging over a wavelength in x and for —= < y <=, and using the boundary conditions that perrurbations vanish for ly| > o we get 2 Thus (# tute =) is conserved where < > denotes Be the area average. 11.6 Solve for the meridional dependence of the zonal wind and geopotential perturbations for a Rossby-gravity mode in terms of the meridional velocity distribution (11.40). Solution: From (11.37) kB) iver, (% ah? \ dy ivgh, (_B_ wie ie But from (11.41) . B(y) = ivy 50) ‘Then from (11.33): Byt - ik@ = (B + kv) ys 1. ity) = iv (B + kv)yd.96 11.7. Use the linearized mode! (21.49) — (11.50) to compute the meridional distribution of divergence in the mixed layer for a situation in which the geostrophic wind is given by _ py? uy = wen 2"), ¥,=0 where uo and ¢ are constants. Solution: From (11.49)-(11.50) aa ~ Byy = 0 av + Byu = Byty Toss eB, andy = 2. 2(, BY) _ oe) 2 al (. ~) py 2 for y=0 &-he TUVFVUVGATAIALTTAT TANNOY 4 AAA7 Chapter 12 12.1 Suppose that temperature increases linearly with height in the layer between 20 and 50 kam at a rate of 2 K kn}. If the temperature is 200 K at 20 kan, find the value of the scale height H for Which the log-pressure height z* coincides with actual height at 50 km. (Assume that z* coincides with actual height at 20 km and let g bea constant) Solution: Recalling that d® = gdz we get integrating equation (10,3): feel when z= 7*=20kn = T = Tp + y(z- %) y=2x 107 Km. & ~ 8) = fie Beret —a)) Y 1% soto have z} — 29 = 30 km must have R(zr - 2) (287)(30)(2) H = — = 6697. m. T+ 7@r-20)] 9.8) in| 22 = To 200, 12.2. Find the Rossby critical velocities for zonal wave numbers 1, 2, and 3 (i.c., for 1, 2, and 3 wavelengths around a latitude circle). Let the motion be referred t0 a B-plane centered at 45°N, scale height H = 7 km, buoyancy frequency N=2x 107s", and infinite meridional scale (I= 0). ‘ , fi Solution: Fi 12.13) U, = B)k? +742, | where 1 = 0 yn: From equation (12.13) Byer a) oF 5 2Qcose ke but k= cose)" a98 + Ue = 1.619 x 107M [52(4.93 x 1074) + 1.356 x wy sel U, = 87.54 mss =2 U, = 48.65 ms"; s=3 U, = 27.95 ms". 12.3. Find the geopotential and vertical velocity fluctuations for a Kelvin wave of zonal wave number 1, phase speed 40 im s~, and zonal velocity perturbation amplitude Sms}. Let N= 4x 104s? Solution: From equation (12.26) @” Thus | ©’| = (40 ms“)(5 ms“) = 200 m? s* From equation (12.25) w" = 4 o 1 2_ Ne (410%) _ -$ since |m| >> Gay also m? = Hy = Te = 25 x 10 » [ml =5 x 104 1 4 40), ———,, (5 x 10*)(200) ( Lasse) x10") 5 AT NW esr) (4 « 10%) |= 157 x 10% mst 12.4 For the situation of Problem 12.3 compute the vertical momentum flux M = pou’w’. Show that M is constant with height. HH etl (uw) but u’ = Upcoskxe Solution: M and from problem 12.3 2 w (e x Upcoskxe@?/# AAT ATAeer SCTCTSTEETTITIIESESESSSSCSSCSCSESS SSS SEC CCCEU | | (9 40)*(6.37 x 10°) (5 x 107)¢-5) (4 « 10>) = 3.92 « 107 m? st 12.5 Determine the form for the vertical velocity perturbation for the mixed Rossby- gravity wave corresponding to the u’, v’, and @ perturbations given in (12.32). Solution: From equations (1225) and (12.32): vm wor rs 2, i. , vim Bln) aw oe ye} = 12.6 For a Rossby-gravity wave of zonal wave number 4 and phase speed ~20 ms! determine the latitude at which the vertical momentum flux M = pou’w" maximum. uw isa Solution: From problem 12.5 and equation (12.32) 2 (u’w’) = (constam)y? o| Blas") ‘And using (12.30): 24, 2 iy By ra 0> °F YMax = ar s(i+ 2) CB} veck=S ‘ 20)(4) es # 20)(4 fs max = # yy ~ (1.458 x 10~) (664)? = : (665) 2a(1+ 2) Juax= 677.5 km = MAK = 46.1 i100 1 12.7 Suppose that the mean zonal wind shear in the descending westerlies of the equatorial QBO can be represented analytically on the equatorial B-plane in the form & - seo) where L = 1200 km. Determine the approximate meridional dependence of the corresponding temperature anomaly for |y| << 1. Solution: From (12.33) a He = & exp (=) ‘An approximate solution can be obtained by lering y xy (z)=(e}~ where Ty and Tz are constants to be determined. 7 =I Tew Substituting into (12.33) and equating all terms with the same y dependence we get 12.8 Estimate the TEM residual vertical velocity in the westerly shear zone of the ‘equatorial QBO assuming that radiative cooling can be approximated by Newronian cooling with a 20 day relaxation time, that the vertical shear is 20 m s~* per 5 km, and the meridional half-width is 12° latitude. Solution: In this case the temperature tendency term can be neglected and (12.2) is an approximate balance between adiabatic warming and Newtonian cooling: iyCCCCCCTTVEC Po Ge TCC S SCS CCST REESE S ECCS S ESTOS SEES 101 _ 5 (L3x 108s) "(2.29 x 10Mm"s)(4 x 1073s )(L33 x 108)" 6 (2 x 10s") = 3.9 x 10“ms* or ®* = 3.37 mday7). cr102 Chapter 13 13.1 Show that for the barotropic vorticity equation on the Cartesian f-plane (13.26) ‘enstrophy and Kinetic energy are conserved when averaged over the whole domain, i.e. that the following integral constraints are satisfied a © dy = 4 | | YY VY aay = 2 ffEaw <0 All zy ea=0 Hint: to prove energy conservation multiply (13.26) through by -v and use the chain rule of differentiation. Solution: We assume periodic boundaries in x and let y = constant for y=0, D. Enguophy conservation: Take ¢ times (13.26): Ba (av seo 2 The first part vanishes after integrating in x, the second part vanishes after integrating in y. \ wybe PI FTTTTTT TT TEC SII CSCS SSCS ESCH S ES SP CCE CTSKTCEREE 103 Energy: Left hand side: gives avy _1a a(v-v ~O* ek Vy) = 2/—— Wy Sh avy YH) = S( ) Right hand side: “WW W=-V (Vy vs) + cYV Vy + Vyo Vy (), Q) @) (1) _ vanishes after integration over domain. 2) vanishes since V, is nondivergent (B) vanishes since Vy, is orthogonal to Vw 2 Finally, py = . which vanishes after integrating in x. m Jf5c 13.2 Verify the expression (13.31) in the text. Solution: Deb DD reese teed a) BY Yee Senet mae Sane (2)104 +2 SY Wares - @) Then when term (2) is summed in y alternating terms cancel including boundary terms. 13.3. The Euler backward method of finite differencing the advection equation is a two step method consisting of a forward prediction step, followed by a backward corrector step. A complete cycle is thus defined by: where 7, 3 Solution: is the first guess for the time step s+ 1. Use the method of Section .3 to determine the necessary condition for stability of this method. Let Gy = Ave", G,,, = Brom Then substituting gives . AS — BY = —BS(ot sinp) BY! — BY = —AS(aisinp) where we have used 2isinp =e? +e? APR susvevunsoansiensoeree rons seeetsssianseTHTETETTECSCSSCKECKCEECSECCECKES EC ECEECEES 105 Thus, A‘ = B5(1 of sinp) and B*** = B [1 ~(oisinp)(l - ai sinp)]} letting 2 = oO sinp, we get B=1~ju(\- in) =1-ip-p? But |! ~ iz — | 5 1 for stability of {a =u wis which requires u? ¢ 1, of o <1. 13.4 Carry out truncation error analyses analogous to that of Table 13.1 for the cases o = 095 anda = 0.25. Solution: For o = 0.95 /. 1D Yi P cl 2 X = oo 4 % 1.253 0.840 0.5238 8 y, 0.737 0.988 0.149 16 % 0.3721 0.997 0.035 32 % 0.186 0.997 0.009For o = 0.25 Fx P h 2 x « 7 4 % 0.253 0.644 0.016 8 %, 0.178 0.907 0.008 16 ™% 0.096 0.976 0.002 32 ye 0.049 0.994 0.0006 13.5 Suppose that the streamfunctiony is given by a single sinuscidal wave vz) =A sinkx. Find an expression for the error of the finite difference approximation PY Ym i= 2m t Yn at ad for kor = &, and x. Here x = mér with m= 0,1, 2,... 2 jon: 2M = apres aps Solution: SY = —At? sinkx = ~Al? sinknd Vmei~ 2m + Wo _ A i tt = A [sink (m + Id ~ 2sindond + sink(m ~ 1)d) = 24 sinmid (coskd -1) _ 2A sin mkd(coskd — 1) a? (—AKsin bd) Fraction error = WAV’ \ 1 VW PAVAVVVVVVVVVVVVTT TAT VATA ATTATIITTETTESCESSESSSSCSSSCECEESCECEC EC SE ECE CUE 107 -) = x Ye g & a onla ala | & 13.6 Using the method given in section 13.3.3 evaluate the computational stability of the following two finite difference approximations to the one-dimensional linear advec- tion equation: @) Emset—bms =~ 9 (Ems ~ Snes) (o) 1~ bn eo (Sess Sms) where o =céi/éx > 0. [The schemes labelled (a) and (b) are referred to as upstream and downstream differencing, respectively.] Show that scheme (a) damps the advected field, and compute the fractional damping rate per time step for o = 0.25 and kéx = 1/8 fora field with the initial form ¢ = exp(iix). Solution: Let ¢,, = V,e™. Then upstream differencing gives y -V, = ov,{1- 4, Clearly |1 - o(1 - eM4)|'< for stability. ~ oft e Thus [1 - o(1- coséd)]? + 0? sin?kd <1 or 1 = 20(1 ~ coskd)? + 67(1 + cos*kd + sin?kd ~ 2coskd) = 1 = 2a(t - coskd)(J — 0) $1 =. OS 1 forstability.108 For'downstream differencing Y, . 1 (et yy ‘taen [1 - ofe —1)P = (1 - ocoska + oF + (0 sina? = 1+ 20(1- coskd) + 07(1 - coskd)? + o*sin2kd 21 for all ¢ so absolutely unstable. For damping rare of upstrearn method note - oft - «4 -1)] ft - as(. ~ cos yf + ( 25 sin 2) \" = 0.985. Thus the amplitude decreases by | ~ 985 =.015 per time step. 13.7 Using a staggered horizontal grid analogous to that shown in Fig. 13.5 (but for an equatorial B-plane geomery) express the linearized shallow water equations (11.29)-(11.31) in finite difference form using the finite differencing and averaging operator notation introduced in Section 13.6.1. Solution: Grid stencil is as follows: Vint nh Vn nth Oman Um. Pn n Unmet n Vinee n Vom Um m= Pm nod U mst, nl (ms ben) «am Define 2 *(m, n) = ine Z*(m, 1 3 ie [alms a) (m4) 6,u(m, n) = = 2 24 AAVVAVTTTLAV TVA AAAS T TANT T TATA ATAPF TTTTTTT TTT TTT TERI LESS SS SSeS SSCS SSeS Kssecsedddd: 109 ‘Then (11.29) - (11.31) can be expressed as: Syv(m a) + B F (metin $)=-8,0(m -4) mune Pha BZ OH BD 5, O(a, 2) + the [du(m+ 3, a) + 5y(m, n+3)| =0 43.8. Verify the equaliny exp(-2ie,) given in (13.22) Solution: L=isn@ _ cos@ ~ ising _ &? T+iwnd cos@ rising oh 13.9 Compute the ratio of the numerical phase speed to the tue phase speed, £ implicit differencing scheme of (13.19) for p = Let ¢=0.75 and o= 1.25, Compare your results to those of Table 13.1. . <2 6p Solution: Here — = —— =—— " c ckbt op = ar*( snp)110 ao =0.75 o=125 Yeo | = & | & 2 = = 0 z 0 4 % 0.359 0.609 0.559 0.569 8 %, 0.259 0.880 0.416 0.848 16 % 0.143 0.968 0.235 0.957 32 | 0.073 0.992 0.121 0.989 ‘Thus phase speed error increases as @ is increased. 13.10 Starting with the assumed solution (13.61), obtain the normal mode solutions to the linearized versions of (13.58)-(13.60) and hence verify (13.62) and (13.63). Solution: Substituting from (13.61) into (13.58)-(13.60) with €= 0, we get -ivi — fo+ik® =0 a) -iW + fa+ild =0 2) iv + c [iki + ils] = 0 @) Multiplying (2) by &, (1) by J and subracting gives wiv (Sk - iil) + f (ka + 8) =0 4) Multiplying (1) by #, (2) by / and adding gives ~iv(ki + 15) - f(ko = la) + i(k? + P)S =0 (5) Eliminating (ki — li) between (4) ahd (5): (7? - v7) (ea + 8) + v(2 + P)S =0 ©AL Combining (6) and (3) to eliminate (ki + 19) gives v[e- vee +A] =0 So y =0 and v fre Ae + ape Here v = 0 isthe Rossby mode = vp and v = #[f? + 0?(k? + ey are the gravity modes v,.., ¥, For Rossby mode set v = Q in (1) ~ (2), get iif"b,, tig = If Dp. ‘These also satisfy (3) with v = 0. Yee For Gravity modes: Multiply (1) by f, (2) by iv and add to get 5 1 5 = (vo.- 2) [vex - if] Then multiply (1) by iv, (2) by f and subtract: (vB. — PY" Levon + wf] 13.11 Show that the projection of the observations onto the Rossby normal mode (13.66) is equivalent to requiring that the quasi-geostrophic potential voricity of the Rossby mode be proportional to the sum of the observed relative vorticity minus the observed geopotential multiplied by the factor (4}. Hint: linearize (4.26) and assume that fis constant. Solution: ‘The barotropic potential vorticity is where A= H+ — g and H isthe mean height and © js the perturbation geopoten pastl ® A If [| << f and isl << | we can linearize as ‘i { TJTTTTETTTTTTTTITIIIIS So sosoooecerrnn nnn12 But Co = ao - 3 = i(kiy — lig), 2 thus, Hs +f- [ree =i) - g 2a which differs from (13.66) by a constant. ‘ iy 1 sain N 1 ATUSAAGUANANGUANEANTATDAA TATE RAN UEE EATEN