Surveys of nearby molecular clouds with Spitzer Lori
Surveys of nearby molecular clouds with Spitzer Lori Allen (NOAO) with much help from Mike Dunham (Yale), Neal Evans (Texas), Tom Megeath (Toledo), Rob Gutermuth (Smith), Dawn Peterson (SAO)
Spitzer Surveys of Nearby Clouds (large area imaging surveys within 1 kpc) c 2 d: Legacy Program (Evans et al. 2009) Gould Belt: Legacy Program (Allen et al. in prep. ) Taurus: Legacy Program (Rebull et al. 2010) Orion: GTO program (Megeath et al. in prep. ) Embedded Clusters: GTO program (Gutermuth et al. 2009) Cygnus-X: Cycle X Legacy Program (Hora et al. in prep. ) Ceph. OB 3, NAN, Mon, S 140 (Gutermuth in prep; Rebull et al. 2010) c 2 d/GB Survey Characteristics ll: 3. 6, 4. 5, 5. 8, 8. 0, 24 and 70 micron (GB includes 160 micron) Completeness: 90% to 0. 05 Lsun; 50% to 0. 01 Lsun (1 -30 micron) Data Products Available through SSC (Popular Data Products) and IRSA
GB / Cepheus 19. 09. 11 mw 2011 [Kirk et al. 2009] 3
Spitzer Legacy data products http: //data. spitzer. caltech. edu/popular/ • Point source catalogues Spitzer and 2 MASS fluxes, quality, characterization • Image mosaics • Ancillary data Extinction maps Re-sampled SWIRE catalogs § Documentation 19. 09. 11 mw 2011 4
c 2 d+GB : census of nearby YSO ~17 clouds, > 1600 YSO (Ser/Aql TBD) Large, uniform sample I: Flat: III: 19. 09. 11 mw 2011 5
Serpens South embedded cluster [Gutermuth et al. 2008] 0. 5 pc Protostar dominated cluster: 60% protostars (Class I + Flat) 70% protostars in core Core median density ~600 YSO/pc 2 19. 09. 11 mw 2011 6
Serpens South Spitzer and Herschel [Courtesy Vera Konyves] Herschel rgb 350/160/100 mm Spitzer 8 mm 19. 09. 11 mw 2011 7
Serpens South 1. 1 mm continuum map (Az. TEC/ASTE) Contours: 30 m. Jy / bm (x 2 per level) Central peak ~4 Jy / bm M(tot) > 1000 Msun [Gutermuth in prep. ] [Also Maury et al. 2011]
Serpens South 1. 1 mm continuum map (Az. TEC/ASTE) Contours: 30 m. Jy / bm (x 2 per level) Central peak ~4 Jy / bm Millimeter emission follows absorption feature
GB Number of YSO by Cloud and Class: c 2 d+GB Cloud I Flat II III* Total I Flat II III* Perseus 87 42 225 31 385 76 35 233 41 Ophiuch. 35 47 176 34 292 28 43 171 50 Lup I, III, IV 5 10 52 27 94 4 8 49 33 Cham II 6 10 63 7 86 4 5 62 15 Serpens 36 23 140 28 227 33 19 130 45 Cor Aus 13 3 31 7 54 9 6 22 17 Cepheus 23 11 65 4 103 18 11 61 13 Cham I, III 7 10 63 7 86 5 5 62 18 Lup V, VI 0 1 5 73 88 0 1 9 78 Musca 1 0 1 11 13 1 0 1 11 Auriga 36 20 81 10 147 36 10 81 21 Scorpius 1 1 4 3 9 1 1 3 4 IC 5146 29 12 87 4 132 28 10 79 15 Total 275 181 958 observed 246 1660 242 19. 09. 11 mw 2011 148 918 352 extinction corrected * Class III grossly incomplete c 2 d Census 10
Timescales: previous work • Previous estimates for timescales – t(I) ~ t(II) ~ 0. 4 Myr • In Ophiuchus (Wilking et al. 1989) – t(I) ~ t(flat) ~ 0. 1 -- 0. 2 Myr; Class II 1 -2 Myr • In Taurus (Kenyon and Hartmann 1995) • Note t(II) ~ 10 t(I) – t(0) ~ 0. 01 Myr (Andre, Oph) Rapid early accretion – Or t(0) ~ 0. 1 Myr (Visser, larger sample) – Issues • Small number statistics! • Differences between clouds! 19. 09. 11 mw 2011 11
Clouds are all different Auriga 165 IC 5146 132 Perseus 385 Large sample N > 1000 Serpens 227 Ophiuchus 292 Lupus I-III 94 Average over several clouds Evans et al. 2008 Allen et al. in prep.
17 clouds, 1660 YSO c 2 d + GB surveys I 21% F III Assumptions: Time is the only variable (all masses co-eval) Star formation continuous Class II lasts 2 Myr (half-life) THEN t(Class I) = 0. 47 Myr t(Flat) = 0. 38 Myr 55% II (bit longer than most previous estimates) I: Flat: III: 14% 9% 55% 21% Caveats: Depends on how is calculated Class 0 mixed with Class I
Timescales for YSO evolution • SED class may not map directly to evolutionary stage For example, are all Class I’s embedded? 19. 09. 11 mw 2011 14
MISidentified YSOs from SED FITS (MISFITS; Heiderman et al. ) • Checking Class I sources in c 2 d and GB surveys HCO+ J=3 -2 line, CSO and applied to APEX • So far, 7% Flat and 42% Class I detected from sample of ~100 (northern clouds). • Clouds: Aur, Cep, IC 5146, Lup, Oph, Per, Sco, Ser-Aqu, Musca 19. 09. 11 mw 2011 15
Protostellar Evolution: the “Luminosity Problem” Many Class I sources are under-luminous for predicted infall rates (10 -6 Mo/yr) Suggests episodic accretion in early stages. (See Kenyon et al. 1990 &1994; Kenyon & Hartmann 1995; Dunham et al. 2009; Evans et al. 2009; Enoch et al. 2009)
Many are under-luminous Predicted L = GM(d. M/dt)/R= 1. 6 Lsun for standard (Shu) accretion onto M = 0. 08 Msun, R = 3 Rsun. Most (59%) are below this. [Dunham et al. 2010 a] 19. 09. 11 mw 2011 17
With 2 D, Outflows, Episodic Accretion 19. 09. 11 [Dunham et al. 2010 a] mw 2011 18
All(? ) surveys out to 1 kpc Cloud(s) I Flat II III* Total Reference c 2 d 169 132 656 127 1084 Evans et al. 2009 GB 109 58 336 108 >619 Allen et al. 2011 Taurus 47 (23) 33 (16) 139 (38) Orion 428 2882 141 (70) 360 (147) Rebull et al. 2010 3310 Megeath et al. 2011 N. Am. 224 194 1087 237 1742 Rebull et al. 2009 Cep OB 3 26 170 1835 141 2172 Gutermuth &Pipher S 140 14 42 434 36 526 Gutermuth&Pipher 1143 1627 12445 6121 21336 Cygnus-X Total Hora et al. 2011 >31, 000 No corrections for extinction or completeness 19. 09. 11 mw 2011 19
19. 09. 11 [G He Ev mw 2011 See talks by Heiderman (l Megeath (Thursday), Lada SFR – gas relations
YSO Surface densities for nearby clouds (ordered by total number of YSO) log (YSO per sq. pc)
Census within 1 kpc: YSO 10 k 14 cloud complexes, 20 clouds > 10, 000 YSO Fraction in clusters (N>10) = 67% (N>100) = 10% But varies greatly from cloud to cloud: Fewer than half the nearby clouds contain clusters of N>100 Statistics less complete than for c 2 d+GB
For cloud complexes within 1 kpc, the fraction of young stars in clusters as a function of cluster size (3363) (10) 19. 09. 11 mw 2011 23
Distribution of surface densities [Bressert et al. 2010] Smooth distribution of YSO surface densities See talk by Bressert (Thursday) 19. 09. 11 mw 2011 24
Summary • There’s a lot of Spitzer data out there. ll 3. 6 -160 mm complements Herschel • Lifetimes of Class I, Flat stages longer Based on N>1600, highly complete sample • More evidence for episodic accretion early Observed protostar luminosities low • SFR ~ 10 x exgal relations Variations from cloud to cloud factor of 100 19. 09. 11 mw 2011 25
How “Efficient” is Star Formation? • Not very for the cloud as a whole – 2% to 5% of mass with AV > 2 is in dense cores • (Enoch et al. 2007) – 3% to 6% is in stars • Large variations in SF rate – Serpens is 360 times more productive than Cha III – Normalized to area, Serpens is 100 times Cha III • Quite efficient in dense gas – Current TOTAL M* similar to Mdense – Depletion time is 0. 6 to 2. 9 Myr 19. 09. 11 mw 2011 26
Similar Relation in Regions forming High Mass Stars Green line is (YSOs) ~ 2 (gas) Red points are Class I (more likely to be in formation region) Gutermuth 19. 09. 11 mw 2011 27
Lessons from Nearby Clouds • SFR ~ 10 times prediction of relations for galaxies on average • Variations from cloud to cloud factor of 100 • These regions are forming only low mass stars – Would not even be seen in most exgal SFR tracers • On scales where SF actually happens… – Dependence on Smol may be very strong • SFR determined on sub-pc scales << exgal resolution 19. 09. 11 mw 2011 28
Summary • Stars form mostly in clustered environments – Regions of high extinction, dense gas – About half in large (N>100) clusters • Mass function may trace IMF of stars • Timescales for Class 0 and I longer • Shu inside-out collapse consistent, except – Luminosities imply accretion is episodic • Efficiencies in clouds are low (3 -6%) – Efficiencies in dense cores are high (>25%) • Exgal prescriptions do not work locally 19. 09. 11 mw 2011 29
Number of clouds Median density of young stars for 15 molecular clouds YSO per sq. pc Median density not a strong function of total number of YSO
Number of clouds Median density of young stars for 15 molecular clouds YSO per sq. pc Median density not a strong function of cloud mass
Number of clouds Peak density of young stars for 15 molecular clouds YSO per sq. pc Peak density a weak function of number of YSO
Fraction of young stars in clusters (> 10 per sq pc) Fraction in clusters not strongly correlated with total number of YSO.
Fraction of young stars in clusters (> 10 per sq pc) Fraction in clusters …nor with cloud mass
Conclusions We used Spitzer to identify >7000 YSO in 15 molecular clouds within 500 pc § Molecular clouds form stars with a wide range of surface densities. § The median surface density does not correlate with the total number of YSO nor with the cloud mass. § The fraction of YSO in clusters (N >10 pc-2) does not depend strongly on the total number of YSO, nor on the total cloud mass. § The clouds divide into low and high fraction of clustering (e. g. Taurus-like and Orion-like). § Peak surface densities are highest in the two most massive clouds.
Fraction of young stars in groups of > 10 per sq pc > 100 per sq pc Fraction in “clusters” weakly correlated with global average gas density
Rates and efficiencies Star formation rates range from cloud to cloud, from 5 Msun/Myr to 1000 Msun/Myr SF efficiencies from 0. 6% to 8% (cloud-wide) Taken together, for all 10 k YSO within 1 kpc: SFR = 2700 Msun/Myr SFE = 2%
Timescales for earlier stages • For 3 clouds with millimeter maps (Enoch 08) – – Use Tbol to separate Class 0 from I Absence of IR source to say starless Largest sample to date: 200 cores N(0) = 0. 44 N(I), so t(0) = 0. 16 Myr • Not consistent with fast, early infall (Andre et al. ) • Except Oph: 0. 04 Myr, Oph was basis of low t(0) – Oph has faster evolution or not continuous – N(SL) = 0. 8 N(0+I), so t(SL) ~ 0. 46 Myr • After <n> > 2 x 104 cm– 3 • t(SL) ~ 3 tff; between predictions of fast and slow 19. 09. 11 mw 2011 38
Starless Core Lifetimes 19. 09. 11 mw 2011 Enoch et al. 2008 39
Comparison to Shu model • Assume inside-out collapse at 0. 19 km/s – Sound speed at 10 K • In 0. 57/2 Myr, rinf = 0. 057 pc – Consistent with some sizes – Mean separation in clusters 0. 072 pc (Gutermuth) • At d. M/dt = 1. 6 x 10– 6 Msun/yr, M* ~ f 0. 90 Msun – If f ~ 0. 3, get 0. 27 Msun ~ modal mass • Consistent with assumptions, most data • Picture holds together, except… 19. 09. 11 mw 2011 40
Alternative “classification” Class and Tbol in nearby clouds 19. 09. 11 mw 2011 41
Star formation efficiency § In dense cores SFE is high, 15 -30% (Note filamentary structure) §Cloud-wide SFE is low, 3 -6% [Enoch et al. 2007] Perseus: Bolocam 1. 3 mm
- Slides: 42