A Report of IAU Symposium 270 Computatioal Star Formation M. Yamada - - PowerPoint PPT Presentation

A Report of IAU Symposium 270 “Computatioal Star Formation”

1

• M. Yamada

Introduction/Contents

✦ Program of the symposium: a wide variety in scales, observations

1) individual star formation (low-/high-mass), clusters 2) feedback, triggered star formation 3) star formation over Galactic/extragalactic/cosmological scales 4) ... and so on (incl. numerical techniques and hardwares)

✦ Highlights (personally biased)

1) CMF/IMF and turbulent fragmentation 2) radiative feedbacks (high-/low-mass) 3) synthetic observations 4) turbulent ISM on (Extra-)Galactic Scales 5) misc.

✦ Keywords

✦

radiation - radiative (magneto)-hydrodynamics, synthetic observation

✦

fragmentation - too many fragments? mechanisms?

✦

statistics - can it derive correct information? 2

I.CFM/IFM & Turbulent Fragmentation

3

CMF & IMF I: Observation

4 CMF/IMF in Pipe Nebula (Alves et al.2007)

IMF dense core mass function

✦Observation of Pipe Nebula

✦mass function of dense core

seems to be identical to IMF with a shift in mass ⇒ “uniform” star formation efficiency (M*/Mcore) indep. of core mass?

✦theorists‘ job become now:

✦reproduce functional form for

the mass function of synthetic “cores” [mostly in large scale simulation, e.g., whole molecular cloud or greater] - why are CMF and IMF look alike?

✦search for mechanisms to

determine the uniform star formation efficiency (=M*/Mcore) [mostly in individual cores studies]

✦Observation of mass function of dense core:

✦studies of different kinds of cores (starless, prestellar, etc.) ✦updates in obs./data-analysis techniques ✦submm. dust thermal emission (850μm) gives a better identification of true starless

prestellar cores compared with AV measurement in optical (Andre’s talk)

✦ Av cores seem to be gravitationally unbound, while dust emission cores seem to

be bound [true prestellar?] (→Pavlyuchenkov’s talk?)

✦deep imaging of Pipe Nebula by Herschel (Alves’ talk) ✦AV (or NH) PDF <-> log-normal distribution down to low AV regime, consistent

with Wada’s simlation

✦more than 90% of the mass is in low AV (Σ=46Msun/pc2)

✦Observational studies of IMF

✦many samples - cluster IMF seems largely the Salpeter-Scalo type (Ascenso’ talk) ✦indep. of metallicity, stellar density, environments...

⇒IMF may not provide a good constraints to models?

✦numbers of dense clusters can have very massive star (M*>150Msun) if universal

IMF is assumed, but no such massive star has been discovered ⇒ stars determine their mass locally?

CMF & IMF I: Observation (2)

5

✦“Turbulent fragmentation”: ✦ sheet-like gas which is compressed by shock waves induced by supersonic turbulent

flows (λfrag~λsheet~L*(ρ1/ρ0)~L/MA)

✦results are dependent on Mach numbers, but it succeeded in reproduce observed IMF

(fragment mass function+ M*=Mfrag*fJ; [fJ is defined as a fraction of gravitationally unstable core mass fraction having ρ∝r-2])

✦Different from classical view: magnetically subcritical/supercritical

CMF & IMF II: Theory (1)

6 CMF/IMF in Pipe Nebula (Alves et

IMF dense core mass function

Synthesized “IMF” in 3D ISM simulation (Padoan & Nordlund 2002)

CMF & IMF II: Theory (2)

7 AV map of Pipe Nebula (Alves et al. 2007) synthesized density structure (MHD, Padoan et al. 2006)

Simulation looks really like to real clouds, but..?

CMF & IMF II: Theory (2)

7 AV map of Pipe Nebula (Alves et al. 2007) synthesized density structure (MHD, Padoan et al. 2006)

Simulation looks really like to real clouds, but..?

✦“Turbulent fragmentation”: criticism and problems ✦ mostly isothermal simulation <-> multi-phase ISM in real universe

✦ origin of supersonic turbulence? -> avoids by adding turbulent fields by hand ✦ boundary conditions? <-> no apparent periodicity in real ISM ✦larger scale simulation so as to set a more realistic boundary? (how?) ✦what kind of quantities are to be compared with observations?

✦ISM turbulence simulation anlaysis: criticism and problems

✦ power spectrum? -> P(k)∝k-a, a=5/3 (Kolmogorov), a=2(Burgers) ✦random phase assumption? <-> anisotropic flow (e.g., by magnetic filed) ✦ structure function? Sp(l) = <|u(r+l)-u(r)|p>

✦Statistical analysis sometimes erases important information..

✦different models can give the (almost) same power spectrum

CMF & IMF II: Theory (3) - analysis -

8 Power spectrum (Padoan et al. 2006) 1-st order velocity structure function (Kritsuk et al. 2007)

✦Thermal instability makes multi-phase ISM w/ turbulent velocity

✦ tiny cold clumps embedded in warm diffuse phase ✦ (basically) solves the dissipation problem ✦efficient enough? relation to the star formation?

✦Time-dependent pressure can work as regulation of SFE? (Ostriker’s talk)

✦SFE up→ Pex increases to thermally stable ISM → rapid SF → termination of stellar

feedback lowers Pex to thermally unstable phase → cold phase form stars

CMF & IMF III: Theory (4) multi-phase ISM

9 2D MHD sim. of SNR (Inoue et al. 2009) Kritsuk’s talk & Ostriker’s talk

CMF & IMF III: Theory (5) theory of IMF

3

✦Analytic modeling of IMF w/ Press-Schechter formalism

✦ PDF of over-dense regions: log-normal in non-gravitational turbulent ISM

<-> Gaussian in cosmology

✦ assume “star formation” when average density exceeds a threshold (e.g., CO gas)

→ analytic formula for mass function

✦it has the same shortcomings as PS,

but it succeeded in describing synthetic IMF taken from simulation

✦convenient for examining what process

is essential for determining IMF Hennebelle’s talk δc: free parameter to specify physical process of SF Hennebelle & Chabrier (2008) 10

• II. Radiative Feedbacks &

Individual SF

11

✦Spitzer c2d project:

✦age estimates of Class I/0 is longer than previous ones (0.54Myr/0.35Myr) ✦but it does not solve the “luminosity problem” (Lobs<<Lacc(theory)~10-6Msun/yr) ✦intermittent accretion?/competitive accretion? (→McKee’s talk)

✦ Review of disks of YSO (Duchene’s talk)

✦ existence of disk and basic properties are almost independent of the mass of protostar

(0.5Msun<M*<5Msun) or of feedbacks from nearby massive stars

✦disk mass of class 0 seems to be larger than that of class 1

→opposite sense to the classical picture?

✦Spitzer 8μm obs. of envelopes of Class 0 objects (Tobin’s talk)

✦found significantly irregular morphology @1,000 AU ✦velocity sub-structure inside a core: dynamics of the very early phase ✦origin? →

Low-mass SF I: Observation

12 Evans’ talk

✦Spitzer c2d project:

✦age estimates of Class I/0 is longer than previous ones (0.54Myr/0.35Myr) ✦but it does not solve the “luminosity problem” (Lobs<<Lacc(theory)~10-6Msun/yr) ✦intermittent accretion?/competitive accretion? (→McKee’s talk)

✦ Review of disks of YSO (Duchene’s talk)

✦ existence of disk and basic properties are almost independent of the mass of protostar

(0.5Msun<M*<5Msun) or of feedbacks from nearby massive stars

✦disk mass of class 0 seems to be larger than that of class 1

→opposite sense to the classical picture?

✦Spitzer 8μm obs. of envelopes of Class 0 objects (Tobin’s talk)

✦found significantly irregular morphology @1,000 AU ✦velocity sub-structure inside a core: dynamics of the very early phase ✦origin? →

Low-mass SF I: Observation

12 Evans’ talk 1) dynamically unstable initial condition? 2) mis-alignment of (global) B-field and rotation axis?

✦Spitzer c2d project:

✦age estimates of Class I/0 is longer than previous ones (0.54Myr/0.35Myr) ✦but it does not solve the “luminosity problem” (Lobs<<Lacc(theory)~10-6Msun/yr) ✦intermittent accretion?/competitive accretion? (→McKee’s talk)

✦ Review of disks of YSO (Duchene’s talk)

✦ existence of disk and basic properties are almost independent of the mass of protostar

(0.5Msun<M*<5Msun) or of feedbacks from nearby massive stars

✦disk mass of class 0 seems to be larger than that of class 1

→opposite sense to the classical picture?

✦Spitzer 8μm obs. of envelopes of Class 0 objects (Tobin’s talk)

✦found significantly irregular morphology @1,000 AU ✦velocity sub-structure inside a core: dynamics of the very early phase ✦origin? →

Low-mass SF I: Observation

12 Evans’ talk 1) dynamically unstable initial condition? 2) mis-alignment of (global) B-field and rotation axis? Tobin et al. (2010) ↓contour: SCUBA 850µm ↓contour: τ

✦“c2d” MHD simulation: longer time calc. w/ a sink particle (Machida’s talk)

✦circumstellar disk-> originates in the

flattened first core

✦second core(=protostar) is formed in the

center of the first core

✦in the early phase, “proto”circumstellar

disk is massive and grav. unstable → formation of first planet?

✦ “c2d” MHD simulation II (Duffin’s talk) ✦ discovery of warped disk and precession of jet in a single star formation

✦very new discovery, mechanism unidentified yet ✦ precession may not be an indicator of binary

Low-mass SF II: Theory

13 Inutsuka et al. (2010)

✦“cloud2disk“ - a very large RMHD simulation: (Offner’s talk)

✦turbulent ISM simulation, AMR, down to proto-circumstellar disk scale ✦radiation suppresses the disk fragmentation

→radiative feed back is also important in low-mass formation (cf. Tomida’s poster)

✦turbulent fragmentation is the dominant mechanism for binary (multiple) systems ✦intermittent (time-dependent) acc. (cf. McKee’s talk)

Low-mass SF II: Theory (2)

14

Offner et al. 2008

✦ expanding HII region RCW 120 w/ Spitzer & APEX (Deharveng’s talk)

✦multi-wavelength obs.: neutral matter surrounding HII region ✦APEX 870μm obs: dust cores on the shell, Mshell~a few 10 Msun!

◎massive stars can be formed by the shell around HII region

✦fragment mechanism? different from grav. instability? ✦other sources: a trend of later type stars from the triggering source (O->B->..)

Feedback I: Observation

15 Gray: APEX 870micron red: APEX 870micron, blue:Hα

✦ expanding HII region RCW 120 w/ Spitzer & APEX (Deharveng’s talk)

✦multi-wavelength obs.: neutral matter surrounding HII region ✦APEX 870μm obs: dust cores on the shell, Mshell~a few 10 Msun!

◎massive stars can be formed by the shell around HII region

✦fragment mechanism? different from grav. instability? ✦other sources: a trend of later type stars from the triggering source (O->B->..)

Feedback I: Observation

15 Gray: APEX 870micron red: APEX 870micron, blue:Hα

✦ Perseus deep imaging survey (Arce’s talk) COMPLETE w/FCRAO

✦deep imaging of 13CO: numbers newly discovered outflows ✦Eoutflow ~ 10% of Eturb at most in the crowded region ✦contradicts some MHD simulation results (e.g., Nakamura & Li, 2008) ✦mom. transfer efficiency should be examined in more detail!

Feedback I: Observation (2)

16

The gray-scale image shows the 13CO integrated intensity. Star symbols indicate the position of candidate YSOs from the c2d survey (red) and known

• utflow and IRAS sources (orange),

while (green) diamonds represent HH

• bjects and H2 outflow shock emission.

Arce et al. 2010

✦ R(M)HD studies are IN! (talks of Bonnell, Krumholz, Susa, Nordlund...)

✦two scenarios: competitive acc. v.s. fragmentation & merger, both suffer over-

fragmentation

✦how to stabilize the core? (radiation heating is the key, but..) ✦gray approx., FLD, ... still computationally expensive to go beyond ✦Simulation results depend on sink particles criteria ✦complementary studies of massive protostar (e.g., Hosokawa & Omukai 2009)

High-mass SF I: Theory

17 Krumholz et al.(2010): AMR RMHD, color: column density

• III. Synthetic Observation

18

✦ Many (M)HD studies now adopt RT experiments!

✦most of them focus on continuum [implementation on AMR, SPH..] ✦usually easier than lines, but scattering or polarized transfer are still yet to be solved

(review: talks of Steinnacker & Juvela)

✦Application to real observation (Pavlyuhenkov’s talk)

✦IRDC(Infrared Dark Clouds) cont. calc. fitting of Spitzer images of IRDC-320.27+29,

IRDC-321.73+005 →succeeded in re-construction of density structures & found protostars in both of IRDC

✦Development of Pipe-line Tool (Padovani’s talk)

✦ARTIST: extension of 1D on-line LVG analyzer RADEX to 3D calc.

(http://www.strw.leidenuniv.nl/~moldata/radex.html)

✦non-uniform grids available ✦one of the extensive tool for ARC-let (Germany?) ✦for ALMA, Herschel, SOFIA.... etc.

S y n t h e t i c O b s e r v a t i

• n

s

19

✦ Many (M)HD studies now adopt RT experiments!

✦most of them focus on continuum [implementation on AMR, SPH..] ✦usually easier than lines, but scattering or polarized transfer are still yet to be solved

(review: talks of Steinnacker & Juvela)

✦Application to real observation (Pavlyuhenkov’s talk)

✦IRDC(Infrared Dark Clouds) cont. calc. fitting of Spitzer images of IRDC-320.27+29,

IRDC-321.73+005 →succeeded in re-construction of density structures & found protostars in both of IRDC

✦Development of Pipe-line Tool (Padovani’s talk)

✦ARTIST: extension of 1D on-line LVG analyzer RADEX to 3D calc.

(http://www.strw.leidenuniv.nl/~moldata/radex.html)

✦non-uniform grids available ✦one of the extensive tool for ARC-let (Germany?) ✦for ALMA, Herschel, SOFIA.... etc.

S y n t h e t i c O b s e r v a t i

• n

s

19

✦ Many (M)HD studies now adopt RT experiments!

✦most of them focus on continuum [implementation on AMR, SPH..] ✦usually easier than lines, but scattering or polarized transfer are still yet to be solved

(review: talks of Steinnacker & Juvela)

✦Application to real observation (Pavlyuhenkov’s talk)

✦IRDC(Infrared Dark Clouds) cont. calc. fitting of Spitzer images of IRDC-320.27+29,

IRDC-321.73+005 →succeeded in re-construction of density structures & found protostars in both of IRDC

✦Development of Pipe-line Tool (Padovani’s talk)

✦ARTIST: extension of 1D on-line LVG analyzer RADEX to 3D calc.

(http://www.strw.leidenuniv.nl/~moldata/radex.html)

✦non-uniform grids available ✦one of the extensive tool for ARC-let (Germany?) ✦for ALMA, Herschel, SOFIA.... etc.

S y n t h e t i c O b s e r v a t i

• n

s

19

✦ Multi-phase ISM over galactic scales (talks of Wada,Ostriker, Dobbs)

✦ simulation succeeded in reproduce turbulent multi-phase ISM ✦feedback from stellar activities and dynamics of galaxies can provide enough energy to

supersonic turbulent flows But how?

✦ high-res. simulations can treat dynamics of SNR/HII region expansion! ✦... but a big uncertainty still remains due to IMF/SFE (“micro” processes)

✦Multi-phase ISM & SFR (talk of Ostriker)

✦assumption - external pressure controls the onset of collapse ✦2-phase & single-phase ISM play a regulation of SFR(→previous slide)

✦PDF & structure of ISM - Gaussian v.s. Log-Normal? (?’s talk)

✦additive variables - Gaussian ⇔ multipliable variables - Log-Normal

→PDF may be able to give information on additive/multipliable nature

✦but Gaussian/log-normal distribution is established after numerous interactions so that

initial conditions dissapear → Very dangerous to rely too much upon PDF to derive physical processes!

✦Spiral Arm Dynamics (Wada’s talk)

✦hydro. simulation shows spiral pattern is quite time-variable and does not have steady

angular phase velocity (different from classical picture of density waves)

✦physical mechanisms not very clear ✦revision is necessary to SF over galactic scale (e.g. interaction with GMC and spiral arms)

Star Formation on Galactic Scales

20

Summary

✦ numerical simulations grow bigger and bigger

✦

some models seem to provide useful information to observers

✦

new algorithms, new hardwares... <-> new obs. facilities..

✦ Radiation is the key!

✦

radiation feedbacks do matter dynamically (low- and high-mass SF, triggered SF...)

✦

synthetic obs. becomes a fashion - tighter collab. with observers for a big progress toward the near future

✦ Remaining questions

✦

IMF, CMF & role of supersonic turbulent flows

✦

SFE, SFR? (in a single molecular core to galactic scale; validity of KS-law?)

✦ It is time to face fundamental questions to numerical studies

(Monaghan’s summary talk)

✦

do we solve the correct set of equations under reasonable assumptions? (boundary conditions, initial conditions....) 21

“Numbers do not matter, but new insights” (R. Larson)