Grand Challenges in Global Circulation Dynamics

Transcription

1 Grand Challenges in Global Circulation Dynamics Tapio Schneider ETH Zurich, Caltech (Source: CLAUS,

2 Grand Challenges in Global Circulation Dynamics Tapio Schneider ETH Zurich, Caltech Easterlies (Trades) (Source: CLAUS,

3 Grand Challenges in Global Circulation Dynamics Tapio Schneider ETH Zurich, Caltech Westerlies Easterlies (Trades) Westerlies (Source: CLAUS,

4 (Hadley 75)

5 Post Newton: Hadley s (75) view of the winds 0 N Equator 0 S

6 Post Newton: Hadley s (75) view of the winds 0 N Equator 0 S

7 Post Newton: Hadley s (75) view of the winds 0 N Equator 0 S

8 Post Newton: Hadley s (75) view of the winds 0 N Equator 0 S

9 Post Newton: Hadley s (75) view of the winds Surface easterlies in the tropics must be compensated by westerlies elsewhere; otherwise Earth s rotation rate would change. 0 N 0 S Equator

10 Eastward wind (January) a Pressure (mbar) Easterlies 0 Westerlies Westerlies 0 Latitude 50 (Schneider, Ann. Rev. Earth Planet. Sci., 2006)

11 Eastward wind (January) a Pressure (mbar) Easterlies 0 Westerlies Westerlies 0 Latitude 50 (Schneider, Ann. Rev. Earth Planet. Sci., 2006)

12 Eastward wind (January) Pressure (mbar) a mph 8 Easterlies 0 Westerlies Westerlies 0 Latitude 50 (Schneider, Ann. Rev. Earth Planet. Sci., 2006)

13 40 Eastward wind (January) Pressure (mbar) Ferrel cells Hadley cells 8 Easterlies 0 Westerlies Westerlies 0 Latitude 50 (Schneider, Ann. Rev. Earth Planet. Sci., 2006)

14

15

16 20th century: Hadley circulation only in tropics 0 N Equator 0 S

17 20th century: Hadley circulation only in tropics 0 N Equator 0 S Extratropical macroturbulence transports angular momentum out of tropics into extratropics

18 Macroturbulence in control Any theory of atmospheric circulations and of climate must be based on a theory of atmospheric macroturbulence. Because we have no complete theory of macroturbulence, the causes of the General Winds still have not been fully explained by any of those who have written on that Subject (Hadley). The Hadley circulation was generally thought not to depend strongly on atmospheric macroturbulence. But that is not the case.

19 The ideal Hadley circulation... Conserves angular momentum m in upper branch v y m 0 Since y m µ f + z, this implies with local Rossby number Is energetically closed (no heat export) Responds directly to! variations in thermal driving Result: h gh 0 /2 t 0 2 a 2 h 0 0 max (H0 t a 0h )5/2 (Schneider 977; Schneider & Lindzen 976, 977; Held & Hou 980)

20 Ideal Hadley circulation theory... Is intuitively appealing (direct reponse to thermal driving) Appears to account for extent of circulation in Earth s atmosphere But does it account for variations in Hadley circulation as climate varies?

21 January streamfunction and angular momentum Ro. 0.2 Ro. 0.5 (Schneider 2006)

22 Earth-like Hadley circulations... In the annual mean or during equinox are close to limit Ro! 0 Do not respond directly to variations in thermal driving but respond via changes in eddy fluxes We need to rethink Hadley circulation response, for example, to ENSO and global warming

23 Simulate circulations with idealized GCM... A 0.2 Sigma W e B 0.2 Sigma W e! Latitude Zonal wind (magenta) and potential temperature (blue) (Schneider and Walker 2006)

24 Wider circulations with slower rotation rates (Walker and Schneider 2006)

25 Hadley circulation width as a function of rotation rate (Walker and Schneider 2006)

26 Hadley circulation width as a function of other parameters Wider for more stable stratification (Walker & Schneider 2006; Schneider 2006; Korty & Schneider 2008)

27 Hadley circulation strength in idealized GCM Convective lapse rate gg d = g(g/c p ) (Walker & Schneider 2006; Schneider 2006)

28 Hadley circulation strength in moist GCM Hadley cell strength (0 9 kg s ) Glacial climates dry Earth-like climate moist Equable climates Surface temperature (K) Variations in optical thickness of longwave absorber (O Gorman & Schneider 2008; Schneider et al. 200)

29 Terrestrial Hadley circulations During equinox, summer, and in annual mean controlled by eddy fluxes Eddy scaling imprinted on scalings Weaker in warmer and (much) colder climates Changes in width likewise eddy-controlled (but slowly rotating wider, and less influenced by eddies) Need theory that takes eddy effects into account (intermediate Rossby number)

30 Jupiter from Cassini (Cassini Imaging Team 2000)

31 Jupiter from Cassini (Cassini Imaging Team 2000)

32 Winds on giant planets 90º Jupiter Saturn Uranus Neptune º Latitude (planetocentric) 0º 0º -0º -º -90º Zonal Velocity (m s -) (Based on data from Voyager, Cassini, HST; Liu & Schneider 200)

33 Energy budget of giant planets Emit more energy than they receive from the sun Internal heat flux can generate convection Differential solar radiative heating from above pole Equator pole

34 Giant planet properties Have similar radii and rotation rates Differ in energy budgets Very different flows: Jupiter, Saturn superrotating Uranus, Neptune subrotating Differences in flows likely caused by differences in energy budgets and dissipation. How?

35 D simulation of all giant planets (Liu & Schneider 200)

36 Jupiter control simula0ons 0. u, Eddy momentum divergence T, N Uniform solar radia0on No internal hea@lux 0. (Schneider & Liu 2009)

37 Why is Uranus subrota/ng? Almost no internal heat flux (0.042 W m 2 ), the atmosphere is stably stra/fied. u T, N (Liu & Schneider 200)

38 How about Neptune? Has significant internal heat flux (0.4 W m 2 ), the atmosphere is neutrally stradfied below tropopause. u T,N (Liu & Schneider 200)

39 Zonal wind Neptune control simulaon (a) (b) (a) Neptune s insolaon and Saturn s internal heat flux 2.0 W m 2 (b) Uniform insolaon and Neptune s internal heat flux 0.4 W m 2 (Liu & Schneider 200)

40 Equatorial superrotation favored when... Planetary rotation rate low Convective (intrinsic) heating strong Baroclinicity low Width and strength of SR jets can be understood from vorticity homogenization arguments (Schneider & Liu 2009, Liu & Schneider 200, 20)

41 no-drag region,representation the drag time scale is set a constant τ0 with respect to latitude. We va ) is used as a simple of the MHD dragtothe flow on giant planets experiences at The equatorial equatorial no-drag region extends to φe5=d 26 latitude in each hemisphere. the day) to drag time scale τ0 from to 000 d (where d = 86400Outside s Earth region, the drag scale is set to aonconstant τ0 with respect to latitude. We vary the offthe effect oftime the bottom drag the off-equatorial jets d to 00020d (where d = s Earth day) to investigate ial drag time scale τ0 from ct of the bottom drag on the off-equatorial jets. Drag dependence of off-equatorial jets Drag Pressure (bar) stronger weaker u: Latitude -80 u: N: Latitude Pressure (bar) Latitude Latitude ms x ms - s -2 -

42 no-drag region,representation the drag time scale is set a constant τ0 with respect to latitude. We va ) is used as a simple of the MHD dragtothe flow on giant planets experiences at The equatorial equatorial no-drag region extends to φe5=d 26 latitude in each hemisphere. the day) to drag time scale τ0 from to 000 d (where d = 86400Outside s Earth region, the drag scale is set to aonconstant τ0 with respect to latitude. We vary the offthe effect oftime the bottom drag the off-equatorial jets d to 00020d (where d = s Earth day) to investigate ial drag time scale τ0 from ct of the bottom drag on the off-equatorial jets. Drag dependence of off-equatorial jets Drag Pressure (bar) stronger 20 Junjun Liu s poster! weaker u: Latitude -80 u: N: Latitude Pressure (bar) Latitude Latitude ms x ms - s -2 -

43 Conclusions Terrestrial tropical circulations are influenced by eddies, but mean meridional AM fluxes also play a role, so they are in intermediate Rossby number regime (theoretical terra incognita) Still need general theory for Hadley circulation Equatorial superrotation arises when baroclinicity is weak enough, heating strong enough, and rotation rate low enough Off-equatorial jets can be baroclinically generated (difficult to generate them otherwise!) Scaling of off-equatorial jets not entirely clear. Rossby radius and Rhines scale play a role; inverse cascades not necessarily