Turbulence in the (Cold) ISM

Transcription

1 Turbulence in the (Cold) ISM P. Hily-Blant IPAG April 14th, 2011

2 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

3 failed to catch on...

4

5 IRAS: complex cirrus clouds

6 The turbulent ISM: Observations Several components or phases (HIM, WIM, WNM, CNM, MM) Different volume fractions and thermal properties thermal pressure Vertical structure of the Milky Way gravitational potential

7 Reviews Interstellar Turbulence, Elmegreen & Scalo 2004, Scalo & Elmegreen 2004, ARAA Control of star formation by supersonic turbulence, Mac Low & Klessen 2004 Astrophysical Turbulence Modelling, Brandenburg & Nordlund 2011, Rep. Prog. Phys

8 1) Pressure Total pressure (gravity) Difference: Non-Thermal pressure Total thermal pressure (except coronal phase) Adapted from D. Cox, ARA&A 2005 Non-thermal pressure: P CR P nth P mag dyn cm 2

9 1) Pressure Jenkins & Tripp 2007 A small fraction (0.1% in mass) of the diffuse gas has P/k 10 5 K cm 3 : pressure variations occur on scales 50 AU.

10 2) Power laws in various phases Ionized gas: n 3D (e ) Armstrong et al. (1995) HII regions Gibson & Nordsieck (2003) Cold Neutral Medium (CNM) Miville-Deschênes et al. (2003)

11 3) Scaling laws in the Molecular Medium Larson (1981) Falgarone (1997) Larson (1981): σ th+nth (km s 1 ) = 1.10 L(pc) 0.38 Falgarone (1997): σ nth L h, h 1/3 All the clouds belong to a single power-law: observed motions are part of a common hierarchy of turbulent motions No preferred length-scale Do these structures belong to an inertial range of developped turbulence?

12 4) Supersonic Linewidth in the CNM σ HI 3.5 km s 1, c s (100 K) 0.8 km s 1 σ H2 1.0 km s 1, c s (10 K) 0.2 km s 1

13 5) Spatial scales and fractal structure of molecular clouds Falgarone et al 1992

14 6) Star formation and Molecular Cloud lifetimes Not clear (debated)

15 The ISM is, as a whole, turbulent The cold ISM is also turbulent is = is very likely

16 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

17 Introduction to turbulence Laminar flow Turbulent flow Photograph by Graham Jeffery (

18 Turbulent flows are highly irregular in space and time Astrophysical turbulence = extreme ordinary turbulence Reynolds number: non-linear vs. viscous terms t u + ( u ) u = 1 ρ P + f v + ν u Re = ( u ) u / ν u ul/ν Re Flow type 0 10 highly viscous laminar motion laminar transition to turbulence > 10 4 turbulent

19 Turbulence couples a wide range of scales Fully developped turbulence: cascade with inertial range

20 Tools Experiments on Earth (tunnel flows: time series): incompressible, unmagnetized, turbulence Numerical simulations: more versatile but... Statistics of scalar and vector fields (pressure, density, velocity, magnetic fields) Reynolds number are never as high as in the ISM

21 Questions Cascade: integral/dissipation scales? Energy source? Structures: sheets? filaments? Dissipation: time scale? mechanism? Type: compressible? magnetized? Consequences: hot chemistry; star formation mode and efficiency; structure of the ISM We restrict ourselves to the Cold ISM: CNM and MM (traced say by HI and CO)

22 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

23 Turbulent cascade 1) Power sources 1 galactic differential rotation (kpc): shocks of spiral arms, MRI (outer gal.) 2 Supernovae ( 100 pc) 3 star winds, expansion of HII region, massive stars ( 10 pc) 4 gravity for self-gravitating clouds (Hoyle 1953, Klessen 2001, Federrath et al 2011) (pc) 5 outflows from young stars (Nakamura et al 2011, Curtis et al 2010, Carroll et al 2010) ( 0.1 pc) Coupling factors with ISM (efficiency of energy injection) are not well known Note: for structures in the cascade, external power supply from large scales

24 2) Scales Observations: kinetic energy is on large scale motions (Larson 1981) Injection scale: open issue l cloud L (Brunt & Heyer 2009) Upper limit to the injection scale from HI filaments moving in the WIM, 100 pc (GALFA-HI survey, Heiles 2007) Dissipative scale: open issue Tiny scale molecular structures (TSMS): smaller than pc or 200 AU (Falgarone et al 2009) Tiny scale atomic structures (TSAS): 30 AU (Heiles 2007)

25 Estimates for the Reynolds number and dissipative scale (integral/dissipative) Re 4/3 : direct numerical simulations need 10 7 grid points per dimension. Ohmic dissipative scale: l m /η = (Re/Rm) 3/4, Rm = u l l/η m 10 6

26 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

27 Structures 1) Large scales atomic structures (100 pc) IRAS map towards the North Pole: filaments

28 2) Large scales molecular structures (100 pc) Taurus molecular cloud, Goldsmith et al 2008 CO emission is filamentary, lacunary: not a homogeneous sphere

29 3) TSMS Falgarone et al 2009 Filaments at milli-pc scale (PdBI)

30 3) TSMS Absorption spectra towards NRAO150 IRAM30m+PdBI, Pety et al 2010

31 Introduction Introduction to turbulence Turbulent Cascade Structures Dissipation Flavors Perspectives References 4) Pure velocity structures Hily-Blant & Falgarone 2009 Clabaut et al, in prep Filamentary (in projection) velocity structure (no density enhancement)

32 5) Sheets (from simulations) Density field Padoan & Nordlund D section Filaments are sections of sheets formed in supersonic and superalfvénic turbulence; cores form in those filaments

33 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

34 1) Time scale Dimensional grounds: t diss = τu 3 /L dissipative channels: viscosity (shocks and shear structures), reconnection τ 0.6 for all favors of isotropic and homogeneous turbulence, and including all dissipative channels (Kaneda et al 2003, Haugen et al 2004, Stone et al 1998, MacLow et al 1998) Dissipation time scale wrt crossing time of a cloud: t diss 0.5t cross

35 2) Viscous dissipative structures Rate of viscous dissipation per unit mass: ɛ d = 1 ( 2 ν vj + v ) 2 k x i x i In terms of the vorticity ω = v: ɛ d = ν v 2 Idea: trace the vorticity... but only one velocity component in two directions (v z (x, y)) Chemical tracers: CH+ as traced with the Herschel satellite (Joulain et al 1998, Godard et al 2009, Falgarone et al 2010)

36 Centroid Velocity Increments method determine the two-point statistics (Lis et al 1996, Pety et al 2003) Make a pdf of these values for various lags

37 Lis et al 1996 Large CVI regions pinpoint large shear-regions

38 Introduction Introduction to turbulence Turbulent Cascade Structures Dissipation Flavors Perspectives References E-CVI structures in the Polaris Flare Clabaut et al, in prep Moisy & Jimenez 2004

39 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

40 Magnetic fields Globally in the ISM, components in equipartition: P dyn P mag P CR dyn cm K cm 3 P dyn dyn cm 2 : σ v 6 km s 1 P mag dyn cm 2 : B 5 µg v A c s Question: pc scales (GMCs, molecular clouds)?

41 The Taurus molecular cloud Polarization catalog of Heiles IRAS100mic data

42 SuperAlfvénic paradigm of star formation (MacLow & Klessen 2004) But, measurments of velocity anisotropy favour strong incompressible MHD (Heyer et al 2008, Goldreich & Shridar 1995) Absence of increase of the B-field intensity with density: favours flows along field lines rather than compression of gas (Crutcher et al 2010) Absence of strong shocks signatures (Hily-Blant & Falgarone 2009, Falgarone et al 2009) Further problems...

43 Statistics of the velocity field Hily-Blant et al 2008 Deviations from superalfvénic predictions

44 Observations of Magnetic Fields in the Polaris Flare Question: magnetic fields in a non-star forming cloud? Polarization of background starlight by the dust Observations performed at Mont-Mégantic Observatory, Québec, with Beauty and the Beast instrument (Manset & Bastien 1995) In the R-band (766 nm), sensitivity R 15mag in 2 minutes 3 nights in March 2010

45 Polar + dust Polar + CO

46 Observational Results Morphology Field lines are not random on the large scales A main structure running SE-NW clearly correlated with dust and gas Huge variations at small scales Estimates of plane-of-sky B: B pos (µg) = 6.6 n H v δθ 5µG Magnetic to dynamic pressures ratio: ɛ mag ɛ kin = ( ) δθ

47 Outline 1 Introduction 2 Introduction to turbulence 3 Turbulent Cascade 4 Structures 5 Dissipation 6 Flavors 7 Perspectives

48 Turbulence in the ISM, though strongly suggested, is still to be proved Basic properties of the turbulence are waiting further numerical experiments and new observations with better sensitivity and spatial resolution ALMA will not be able to resolve out the viscous dissipative scale, but the question of how far do small-scale in the molecular gas is crucial. Magnetic fields observed by the Planck satellite will provide us with a big picture of the B-field configuration, as well as with estimates of its intensity at moderate scales.

49 Armstrong, J. W., Rickett, B. J., & Spangler, S. R. 1995, ApJ, 443, 209 Falgarone, E. 1997, 170, 119 Gibson, S. J. & Nordsieck, K. H. 2003, ApJ, 589, 347 Larson, R. B. 1981, MNRAS, 194, 809 Miville-Deschênes, M., Joncas, G., Falgarone, E., & Boulanger, F. 2003, A&A, 411, 109