Star Formation: How my knowledge has been shaped by observations taken by the James Clerk Maxwell Telescope

Transcription

1 Star Formation: How my knowledge has been shaped by observations taken by the James Clerk Maxwell Telescope With: J. Di Francesco, H. Kirk (NRC); S. Mairs, M. Chen (UVic); A. Pon (Leeds); R. Friesen (DI); S. Sadavoy (MPIA); and S. Schnee (NRAO). Doug Johnstone: Former Associate Director JCMT, NRC Herzberg, Univ. Victoria

2 The Early 90 s: Graduate School Days Little known about the lives of deeply embedded proto-stars simple inside-out collapse assumed for formation Singular isothermal sphere Emphasis on understanding disk formation/evolution T Tauri stars and excess FIR Charlie Lada JCMT single pixel sub-mm detector: UKT14 built for UKIRT

3 The Early 90 s: Graduate School Days Key Results: Disk masses, evidence of dust evolution

4 The Early 90 s: Graduate School Days Searching for the ever earlier analogues to T Tauri Stars Hidden from optical view by surrounding gas (dust) Investigating manner by which stars (and disks) assemble dm/dt ~ c s3 /G?!? dm/dt ~ Luminosity Andrea Isella Still JCMT single pixel sub-mm detector: UKT14

5 The Early 90 s: Graduate School Days Key Results: Young age, radial density profile,

6 The Early 90 s: Graduate School Days Key Results: Jeans instability, lifetime, density profiles,

7 The Late 90 s: Post-Doctoral Era Integral Shaped Filament in Orion Length ~ 1 degree Beam ~ 15 arcsec 50 by 250 beams 4 nights at the telescope From single sources to imaging SCUBA pixels! Imaging `large regions within molecular clouds Exploring and studying Taxonomy and Morphology Extraction algorithms Semantics: cores, clumps

8 The Late 90 s: Post-Doctoral Era Key Results: Filaments, radial profiles, dust evolution Radial Profiles for the Orion Filament

9 The Late 90 s: Post-Doctoral Era Key Results: core stability, core mass function (IMF-like) More unstable Cloud-like IMF-like

10 The New Millennium: NRC Science Officer Key Results: Environmental effect and Star Formation Key Results: Graduate Students!! JCMT Archive!! A v

11 The Present: Telescope Management From Imaging to Mapping SCUBA-2 10,000 pixels HARP 16 pixels Sensitive Mapping of square degree (and larger) regions The Era of Survey Science Synergy with other Telescopes Herschel far IR/sub-mm ALMA high spatial resolution

12 The Present: Telescope Management Observing Plan: Sensitive CO and Dust Continuum Observations of Star-Forming Regions. [Key Canadian involvement within the team, including new graduate students.]

13 The Present: Telescope Management Coordinated approach to observations. Serpens HARP CO Channels (J. Richer) Serpens Velocity Map (M. Hogerheijde) Images Courtesy of JCMT Gould Belt Survey Serpens SCUBA microns (R. Friesen)

14 The Present: Telescope Management Distributed approach to investigations. Determining Dust Properties (T, β) Comparing CO and dust emission Perseus SCUBA microns (R. Friesen) Determining mass functions; Measuring stability parameters; Examining environments; Identifying special cases Images Courtesy of JCMT Gould Belt Survey

15 The Future: East Asian Observatory Synergy with other observatories. Archive Science! New Surveys Ophiuchus: Red: SCUBA micron Green: Herschel 70 micron Blue: Spitzer 8 micron Image Courtesy of JCMT Gould Belt Survey/R. Friesen

16 JCMT results I had to skip (no time!): Evolution of disks around stars Molecular outflows from young stars Molecular line tracers of in-fall Chemical evolution in clouds and cores Magnetic fields Maunakea background Images courtesy of Daniel McVey (

17 The Evolution of Star-Forming Cores in Molecular Clouds Using Theoretical Models to Inform Observations Doug Johnstone: Former Associate Director JCMT Senior Astronomer, NRC-Herzberg Associate Professor U. Victoria, Canada With: S. Mairs, S. Schnee, S Offner H. Kirk, J. Di Francesco, S. Sadavoy B. Hendricks, G. Herczeg, S. Bruderer

18 w estern P erseus K irk et al Perseus Herschel

19 What We Think We Know About Cores Found in localized regions of cloud - Highest A v zones (highest column density) - Clustered together Distribution of core mass is steep - N M -3/2 : mass resides in small objects - Similarity to stellar IMF is intriguing - Result indep. of structure analysis form - Totals to small fraction of the cloud mass Thermal size vs. mass relation - M R 3 (Pressure-confined objects?) - Largest objects should be gravitationally unstable Johnstone et al. Motte et al. Dense material has different properties than bulk cloud. No requirement for non-thermal support.

20 A Typical Core Mass Function This is quite tantalizing but what exactly supports a massive core? Aquila: Herschel by Andre et al. 2010

21 Cores and the Jeans Mass: How do we populate the high mass end of the core mass function? Legacy Catalogue: JCMT by Sadavoy et al. 2010

22 Observing Super-Jeans Starless Cores with Interferometers Some of these Super-Jeans cores turn out to harbour Class 0 sources. CO (2-1) and C 18 O (2-1) Perseus: SMA by Schnee et al. 2012

23 Observing Super-Jeans Starless Cores in optically thick/thin Molecular Lines Infall Outflow Super-Jeans cores typically show infall signatures. HCO + and H 13 CO + Deuterated Sample: JCMT by Schnee et al. 2013

24 Synthetic Observations of the Evolution of Cores in a Molecular Cloud Simulation Hydrodynamic simulation of a 600 M O star-forming region with proscribed initial turbulence over a free-fall time (Mach 3D = 6.6, n H ~ 2000 cm -3, r = 2 pc, t ff ~ 1 Myr) by Mairs et al. 2014

25 Synthetic Observations of the Evolution of Cores in a Molecular Cloud Simulation (+) Sink Particles (protostars) Hydrodynamic simulation convolved to JCMT resolution assuming distance of Perseus (Beam ~15, ~4000 AU). Note: large-scale structure removed as well (simulating chopping).

26 Clump-finding in Space and Time: Searching for stable ( ) and unstable ( ) cores Including Sink masses in determining stability. Note: cores predominantly along filaments and at intersections.

27 GLOBAL: Evolution of Cores (Number/Mass) Number of cores grows with time until saturating after half a free-fall time Mass in cores grows very fast, reaching 16% of cloud mass by a free-fall time But, the mass in the cores is mostly contained within the protostars

28 Evolution of Core Mass Function The distribution of core masses remains roughly constant over a free-fall time.

29 Evolution of Core (+Protostar) Mass Function The distribution of core+protostar masses increases significantly over a free-fall time..

30 Following the Evolution of Core Properties Pre-stellar cores have low masses. Protostars have a wide range of masses. Protostars initially grow from low-mass cores and then may accrete further.

31 Contemplating the Collapse Mass The typical mass of a core which collapses is ~ 1 M O The Jeans mass associated with the mean density in the simulation (n o ~ 2000 cm -3 ) is M J0 ~ 8 M O If one assumes that the peak density in cores is due to ram pressure provided by supersonic shocks then n c ~ M 2 1d n o ~ (2000) cm -3 M Jc ~ 2 M O This model is undoubtedly too simple and indeed densities at the centres of cores get much higher than n c Multiple shocks? Only the highest velocity shocks?

32 Synthetic Observations of the Evolution of Cores in a Molecular Cloud Simulation 1. The masses and densities of the simulated cores are similar to real cores in Perseus 2. The simulations suggest that observed Jeans unstable cores should already contain protostars How then to reproduce the (massive) stellar IMF with prestellar cores? 3. Nearly all the simulated cores eventually form protostars 4. Nearly all the simulated cores are associated with filaments 5. Most of the final protostellar mass accretes from beyond the originally identified core boundary! Do massive prestellar cores really exist? 6. The collapse occurs from inside-out High resolution (interferometry) required to uncover detail by Mairs et al. (2014)

33 Formation of a star in one slide! Protostars PMS stars stars

34 Spectral Energy Distribution (SED) Protostars PMS stars stars

35 Mass Accretion onto Protostars Expect inside-out collapse of protostellar core (Gravity Wins!): Predict a fixed rate of accretion in simplest models (~c 3 /G : Shu 77) Energy released as mass flows into potential well Accretion Luminosity However, predicted luminosity is HIGHER than observed in most cases! More sophisticated models include internal structure/support Still maintain the same order of magnitude for accretion luminosity C2D: Spitzer by Dunham et al. 2010

36 Mass Accretion - Protostars HH212 But, lifetime of protostars DOES match simple models! Suggests that the mean accretion rate is correct Suggests mass accretes onto protostar during bursts Further supported by structure in protostellar jets Possibly mediated by circumstellar disks Numerical Simulations by Vorobyov and Basu 2005

37 Implications of Variable Accretion - I Temperature Profile of the Envelope responds to accretion luminosity Location of effective photosphere Johnstone et al. 2013

38 Implications of Variable Accretion - II Luminosity of Source gets higher and SED shifts to the blue (Warmer) Approximately linear shift with accretion ratio Approximately linear shift with temperature ratio

39 Implications of Variable Accretion - III Dust must be heated (cooled) to these new temperatures For an ideal gas C V constant Guhathakurta & Draine 1989

40 Implications of Variable Accretion - IV The light propagation time must be taken into account 1000 hrs 100 hrs Crossing time of the effective photosphere (R ph ~ 50 AU) is ~ 5 hrs

41 Implications of Variable Accretion - V The observable timescale for variability can be assessed: 500 hrs 5000 hrs Don t Trust 0.5 hrs 5.0 hrs 50 hrs

42 Implications of Variable Accretion 1. A stepwise change in accretion luminosity will be smeared out Within envelope s effective photosphere hard to detect variation Photons don t freely escape high optical depth Sets minimum timescale for process R ph ~ 50 AU -> t ph ~ 5hrs Sets peak wavelength for source SED -> T ph ~ 100 K Emission at longer wavelengths dominated by larger, colder envelope Takes ever longer to heat the enormous envelope Sets maximum timescale R env ~ 10 4 AU -> t env ~ 10,000 hrs 2. Observations of variability should be able to constrain theory Identify the underlying timescales for accretion changes Periodic?, Episodic?, Stochastic?, Structured chaos? Probe variations in amplitude of mass accretion with timescales? 3. Possibility to open up a new branch of star formation studies Johnstone et al. 2013

43 Possible Observing Strategies: 1. Monitor at short wavelengths (near peak of SED) for variations Maximal change in brightness, shortest delay times Herschel (archive) ~70 microns, JCMT ~450 microns, CCAT ~ 350 microns Lack of truly appropriate instrument available for this purpose! 2. Monitor at longer wavelengths but with high resolution In a large aperture, the longer wavelengths dominated by outer envelope In a small aperture, the longer wavelengths probe toward photosphere Present-day interferometers and since bright, can use SMA/CARMA/PdB 3. Follow-up interesting sources with ALMA Both time-dependent and wavelength dependent observations In principle can use reverberation mapping to uncover structure of envelope

44 Need for Single-Dish Sub-millimetre Telescopes Even in an ALMA world there remains a need for singledish sub-millimetre telescopes for star formation studies Most of the mass within clouds is on large scales Need to understand the role of turbulence and magnetic fields on cloud stability/dynamics Cores are not isolated but connect with large-scale filamentary structure Environment must affect the properties of cores observed with interferometers (mass, density, angular momentum)