The main sequence from stars to galaxies pc kpc Mpc. Questions

Size: px

Start display at page:

Download "The main sequence from stars to galaxies pc kpc Mpc. Questions"

Lee Ellis
6 years ago
Views:

The main sequence from stars to galaxies pc kpc Mpc Questions Is the SFR M scaling-law or main sequence of star forming galaxies probing the bulk of the formation of

self-regulation vs infall Does the main sequence reflect a true universality of star-formation?

1 The main sequence from stars to galaxies pc kpc Mpc Questions Is the SFR M scaling-law or main sequence of star forming galaxies probing the bulk of the formation of present-day stars? Can we learn information on micro (pc-scale) and/or macro (Mpc-scale) physics? self-regulation vs infall Does the main sequence reflect a true universality of star-formation? role of starbursts and mergers David Elbaz (CEA Saclay) Corentin Schreiber, Maurilio Pannella, Tao Wang, Xinwen Shu, Roger Leiton, Mark Dickinson, Hanae Inami, Benjamin Magnelli, Matthieu Béthermin, Mark Sargent, Dieter Lutz & GOODS-Herschel/CANDELS-Herschel teams David Elbaz The main sequence from stars to galaxies

The main sequence of star forming galaxies SFR Elbaz +07

2 The main sequence of star forming galaxies SFR Elbaz +07 specific SFR (ssfr) : ssfr= SFR/M* = SFR/<SFR>/age = b /age SFR/M* ~ 0.06 Gyr-1 (MW) à τ 20 Gyr x 10 = 0.5 Gyr-1 (M82) à τ 200 Myr x 200 = 10 Gyr-1 (Arp220) à τ 100 Myr SFR ULIRGs LIRGs U-G Red rms~0.3 dex Blue Brinchman +07, Elbaz +07 David Elbaz M universality of star-formation in galaxies The M Herschel view on the B M

3 SFR(IR) / SFR(UV,corrected) UV-corrected 1.5<z<2.5 main sequence Compilation from Madau & Dickinson ssfr / ssfr MS SFR IR = SFR UV x starburstiness starburstiness= (ssfr/ssfr MS ) David Elbaz The Herschel view on the universality of star-formation in galaxies 3

4 IRAS cm 41" fwhm 100 µm ISO cm 100µm Spitzer cm Herschel cm FWHM 100µm=6.7 arcsec David Elbaz The main sequence from stars to galaxies GOODS + CANDELS-Herschel - robust SFR - M dust à M gas à SFE - SED: Tdust, FIR/MIR - à signature of mergers/seds 4

IRAS 1985 57 cm 100 µm 41" fwhm cf Schreiber & Inami talks ISO 1995 GOODS-H CANDELS-H PI Elbaz PI Dickinson 100µm 60 cm Spitzer 2003 85 cm GOODS + CANDELS-Herschel - robust SFR -

5 IRAS cm 100 µm 41" fwhm cf Schreiber & Inami talks ISO 1995 GOODS-H CANDELS-H PI Elbaz PI Dickinson 100µm 60 cm Spitzer cm GOODS + CANDELS-Herschel - robust SFR - Mdust à Mgasà SFE - SED: Tdust, FIR/MIR ~6000 Herschel sources -à signature of mergers/seds Herschel cm FWHM 100µm=6.7 arcsec David Elbaz The main sequence from stars to galaxies 5

6 SFR M : the Herschel view on the main sequence to avoid issues with IR selection : mass selected sample WFC3 + nearly complete in z phot & z spec Schreiber SFR M M M SFR SFR David Elbaz The main sequence 6 from stars to galaxies

SFR M : the Herschel view on the main sequence SFR M main sequence in place over at least 90% of the Universe age rms~0.3 dex of MS à universality of SF à self-regulation? (cf P.

7 SFR M : the Herschel view on the main sequence SFR M main sequence in place over at least 90% of the Universe age rms~0.3 dex of MS à universality of SF à self-regulation? (cf P.Padoan) linear slope of log(sfr) log(m ) à ssfr universal at any given epoch break of the slope à evidence for bulge formation à slow quenching of SF SFR Schreiber SFR M M M SFR log 10 (ssfr) [Gyr-1] τ < 500 Myr bulges David Elbaz The main sequence from stars to galaxies redshift 7

The enigma of star formation in the Milky Way If no external supply is available, the gaz in the solar vicinity will be exhauted about 700 million years from now SFR(MW) ~ 7,5 M yr -1 (Diehl et al.

8 The enigma of star formation in the Milky Way If no external supply is available, the gaz in the solar vicinity will be exhauted about 700 million years from now SFR(MW) ~ 7,5 M yr -1 (Diehl et al Nature) Molecular gas (H 2 ): Neutral gas (HI): 10 9 M 5x10 9 M à Gas refilled (e.g. infall, return gas frac7on) = Van den Bergh (1957) à SFR reduced with reduced gas content: SFR ~ ρ gas = Schmidt (1959) David Elbaz The main sequence from stars to galaxies 8

The MS is probably the best evidence for the need of infall A M =5x10 10 M galaxy at z~2 has M gas ~5x10 10 M On the MS its SFR~90M yr -1 à gas exhausted in <500 Myr 500Myr= time z=2.

9 The MS is probably the best evidence for the need of infall A M =5x10 10 M galaxy at z~2 has M gas ~5x10 10 M On the MS its SFR~90M yr -1 à gas exhausted in <500 Myr 500Myr= time z=2.2 and 1.9! τ<500myr But >70% galaxies in MS in this z range => τ dep >500Myr Longer duty cycle evidence for infall! Saintonge +13 David Elbaz The main sequence from stars to galaxies 9

10 Number of galaxies Elbaz +11 IR8 R~S1/3 L8 ~ PAH, PDR ~ 2D LIR~ continuum ~ 3D S x 10 à single region: 2D/3D reduced by ~2 cool dust HII region warm dust IR8 PDR (PAHs) David Elbaz The main sequence from stars to galaxies 10

Elbaz +11 IR8 cool dust HII region warm dust IR8 PDR (PAHs)

11 Number of galaxies R~S1/3 L8 ~ PAH, PDR ~ 2D LIR~ continuum ~ 3D S x 10 à single region: 2D/3D reduced by ~2 log10(ssfr/ssfrms) Elbaz +11 IR8 cool dust HII region warm dust IR8 PDR (PAHs) David Elbaz log10(ir8/ir8ms) The main sequence from stars to galaxies 11

Analogy between the stellar and galactic main sequences mass Hertzsprung Russell O B A F G K M 30000 10000 6000 3000 K Stars spend most of

$Galaxies have galactic winds Stars spend a short fraction of their life Galaxies spend a short fraction of their life in a more luminous phase$ 007 Δm c 2 à light SFR = 4x10 8 M dense The most massive stars produce black holes The most massive galaxies produce black holes that kill

007 Δm c 2 à light SFR = 4x10 8 M dense The most massive stars produce black holes The most massive galaxies produce black holes that kill

12 Analogy between the stellar and galactic main sequences mass Hertzsprung Russell O B A F G K M K Stars spend most of their life on the MS Galaxies spend most of their life on the MS Massive stars die first Stars have stellar winds Massive galaxies die first Galaxies have galactic winds Stars spend a short fraction of their life Galaxies spend a short fraction of their life in a more luminous phase (giants, AGB) in a more luminous starburst phase Universal efficiency of light production : Universal efficiency of light production : Δm c 2 à light SFR = 4x10 8 M dense The most massive stars produce black holes The most massive galaxies produce black holes that kill them that kill them (?) Dying stars = 10 9 times more luminous Dying galaxies more luminous? (starbursts?) David Elbaz The main sequence from stars to galaxies

13 Redshift evolution of the % of starbursts merger rate (Lotz +11) (Xu +12) ssfr > 4x <ssfr> MS starbursts x4 No evidence for a change of starburst % with z à self-regulation vs mergers? Schreiber +14 David Elbaz The main sequence from stars to galaxies 13

$Merger-driven starbursts in local vs distant galaxies In low gas frac?$

14 Merger-driven starbursts in local vs distant galaxies In low gas frac?on local galaxies : turbulence ~5 km/s merger à compressive turbulence à ~50km/s à strong SFR SFR 1st pass 2nd pass Renaud+2014 Bournaud+2015 M merger rate (Lotz +11) In high gas frac?on distant galaxies (z~2, 40-50%): turbulence already at ~50 km/s à rela?ve impact of compressive turbulence induced by merger = minor (see Perret, Renaud, Bournaud +2014) (Xu +12) ssfr > 4x <ssfr>ms David Elbaz Universalité de la forma7on d étoiles dans les galaxies dans la toile cosmique Académie des sciences 1er avril

$width of MS = lognormal starbursts Gabor & Bournaud 2013 infall à turbulence turbulence around the galaxy à lognormal law determining the fraction of «dense gas» in the center of the halo, i.e. the fraction of gas penetrating the galaxy David Elbaz Cosmology and Galaxy evolu?$

15 Let s extend the small/large scale analogy further : is turbulence responsible for the shape of the width of the MS? width of MS = lognormal starbursts Gabor & Bournaud 2013 infall à turbulence turbulence around the galaxy à lognormal law determining the fraction of «dense gas» in the center of the halo, i.e. the fraction of gas penetrating the galaxy David Elbaz Cosmology and Galaxy evolu?on : a major niche for SPICA SPICA Science Workshop 23 May

16 Radio IR correlation : robustness of L IR as a SFR indicator Timescale of FIR: 100 Myr Timescale or radio: 10 Myr Both SFR tracers agree independently of timescales Pannella +14 See also Ivison +10, Magnelli +12; Casey +12, Strazzullo +10, Del Moro +13 David Elbaz Cosmology and Galaxy evolu?on : a major niche for SPICA SPICA Science Workshop 23 May

Elbaz +10, Genzel +10 Elbaz +07, Daddi +07, Noeske +07.

17 The main sequence paradigm SFR-M : "starburstiness" SFR-M gas : efficiency SFR dust: compactness Daddi, Elbaz +10, Genzel +10 Elbaz +07, Daddi +07, Noeske Elbaz +11 David Elbaz The main sequence from stars to galaxies 17

2009) based on the color deconfusion : 500µm flux was wrongly associated to a nearby source

18 Taking advantage of the IR MS to «deconfuse» Herschel Shu et al. (2014, TBS, see poster) Identification of a z=4.05 galaxy (GN10, Daddi et al. 2009) based on the color deconfusion : 500µm flux was wrongly associated to a nearby source 31 candidate z>2 in GOODS-N David Elbaz Cosmology and Galaxy evolu?on : a major niche for SPICA SPICA Science Workshop 23 May

19 Conclusions A SFR M main sequence of star forming galaxies has been in place over at least 90% of the Universe age Gas content of galaxies not enough to maintain such SF more than ~500 Myr à probably the best evidence for the need for infall at all epochs. The small rms~0.3 dex of the MS => universality of SF : 68% of galaxies have the same SFR at given M* à self-regulation? (cf P.Padoan) The linear slope of log(sfr) log(m ) à ssfr universal at any given epoch The break of the slope (initially interpreted as evidence for a sublinear slope) is probably evidence for bulge formation in the MS à slow quenching of SF The fraction of outliers with high ssfr, i.e. starbursts, is small and does not vary with redshift à can be explained if relative impact of mergers is compressive turbulence whose relative role on highly turbulent gas rich distant galaxies is small. David Elbaz The main sequence from stars to galaxies 19

Remember : it all started with island universes...

Remember : it all started with island universes... If no external supply is available, the gaz in the solar vicinity will be exhauted about 700 million years from now SFR(MW) ~ 7,5 M yr -1 (Diehl et al.