SLIDE 1
Gas content and gas kinema/cs in low-mass galaxies
Leslie Hunt INAF-Osservatorio Astrofisico di Arcetri, Firenze
SLIDE 2 why low-mass (metal-poor) galaxies?
low mass galaxies are the most numerous at all redshiGs, very low-luminosity galaxies provide the dominant contribu/on to the luminosity density at z>6
(e.g., Bouwens+ 2011, 2012; Atek+ 2015; Sun & FurlaneUo 2016)
although dark-maUer haloes are built boUom-up, massive galaxies form their stars early
- n (downsizing), unlike low-
mass galaxies that form a large frac/on of their stars later
Galaxy Stellar Mass func/ons to z~4 from Ilbert+ 2013, data from z~0 Moustakas+ 2013, Baldry+ 2012) UV luminosity func/onsz~4, z~8 Bouwens+ 2011 Specific star forma/on rate as func/on of look- back /me for galaxies of varying masses (Thomas+ 2010)
SLIDE 3
INEFFICIENCY of star forma/on/galaxy mass assembly and the role of FEEDBACK feedback and gas/SFR scaling rela/ons (KennicuU-Schmidt) scaling rela/ons for low-mass galaxies: gas frac/ons, mass, SFR, metallicity (take note of metal-poor starbursts) feedback and the baryonic Tully-Fisher rela/on: contribu/on of molecular gas in low-mass (metal-poor) starbursts pressure support and turbulence, comparison of different tracers in low-mass galaxies Low-mass galaxies important probe of galaxy forma/on scenarios because of their vulnerability to feedback
SLIDE 4
feedback and the inefficiency of galaxy mass assembly and star forma/on
SLIDE 5 stellar and baryon frac/on in dark-maUer halos
stellar mass depends on dark-maUer halo mass, but rela/vely independent of redshiG (lookback /me)
Behroozi+ 2013a, 2013b (see also Moster+ 2013, Genel+ 2014) star-forma/on efficiency (SFE) peaks at 5x1011 M¤, characteris/c mass “sweet spot” for galaxy forma/on z=0
SLIDE 6 feedback renders galaxy forma/on inefficient
Hopkins+ 2014, FIRE simula/ons Stellar mass (M¤) Halo mass (M¤) Stellar /baryonic “sweet spot” at 5x1011 M¤ Stellar /baryonic Stellar /baryonic Halo mass (M¤)
SLIDE 7 Stellar /baryonic Stellar /baryonic Halo mass (M¤)
feedback renders galaxy forma/on inefficient
models without feedback are in severe conflict with observa/ons, especially at low masses Stellar mass (M¤) Halo mass (M¤) Stellar /Baryonic “sweet spot” at 5x1011 M¤ Hopkins+ 2014, FIRE simula/ons
SLIDE 8
feedback and gas scaling rela/ons
SLIDE 9 star forma/on is inefficient because of feedback
Hopkins+ 2014 KennicuU-Schmidt (KS) law rela/ng SFR and gas surface densi/es; observa/ons (KennicuU 1998; Bigiel+ 2008; Genzel+ 2010; Daddi + 2010) indicated by yellow shaded region. Simula/ons including various recipes for feedback (see previous slide) shown by points. Instantaneous SFE (here SF per dynamical /me) in dense gas is 100%, but global SFE is ~2%.
stellar feedback injects sufficient momentum to offset dissipa/on; otherwise too- efficient cooling would cause all gas to collapse and form stars
SLIDE 10 inefficient star forma/on in metal-poor galaxies?
lower intercepts for lines (with unit slope) in this plane mean longer gas deple/on /mes starbursts and high-z galaxies tend to have shorter gas deple/on /mescales (here αCO converted to common scale, e.g., Genzel, Tacconi+ 2010)
100 101 102 103 104 10−4 10−3 10−2 10−1 100 101 102 103 Σgas [Msun pc−2] ΣSFR [Msun yr−1 kpc−2]
10Gyr 1Gyr 100Myr SMGs Genzel+ (2010) SFGs Tacconi+ (2010, 2013) BzKs Daddi+ (2010) Spirals Kennicutt (1998) Starbursts Kennicutt (1998)
All resolved measurements (except Kennicutt)
False colors: Bigiel+ (2008)
K r u m h
z + ( 2 9 )
SLIDE 11 100 101 102 103 104 10−4 10−3 10−2 10−1 100 101 102 103 Σgas [Msun pc−2] ΣSFR [Msun yr−1 kpc−2]
10Gyr 1Gyr 100Myr
SFGs Tacconi+ (2010, 2013) BzKs Daddi+ (2010) Spirals Kennicutt (1998) Starbursts Kennicutt (1998)
All resolved measurements (except Kennicutt)
- Blue compact dwarfs (HI only)
LITTLE THINGS (HI only) SHIELD dwarfs (HI only) FIGGS dwarfs (HI only) VLA−ANGST dwarfs (HI only)
inefficient star forma/on in metal-poor galaxies?
Hunt+ in prep high ΣSFR and high ΣHI common in blue compact dwarf galaxies (BCDs), and some LiUle THINGS dwarfs, despite their low metallicity dwarfs may not all be inefficient at forming stars compare with high-redshiG popula/ons to consider the best local analogs
SLIDE 12 inefficient star forma/on in metal-poor galaxies?
Hunt+ in prep 100 101 102 103 104 10−4 10−3 10−2 10−1 100 101 102 103 Σgas [Msun pc−2] ΣSFR [Msun yr−1 kpc−2]
10Gyr 1Gyr 100Myr
NGC1140
SBS0035−052 DDO87 NGC1569 IC10
SMGs Genzel+ (2010) SFGs Tacconi+ (2010, 2013) BzKs Daddi+ (2010) Spirals Kennicutt (1998) Starbursts Kennicutt (1998)
All resolved measurements (except Kennicutt)
- Blue compact dwarfs (HI only)
LITTLE THINGS (HI only) SHIELD dwarfs (HI only) FIGGS dwarfs (HI only) VLA−ANGST dwarfs (HI only)
Ruby (Canameras+ 2017) high ΣSFR and high ΣHI common in blue compact dwarf galaxies (BCDs), and some LiUle THINGS dwarfs, despite their low metallicity IIZw40 from Kepley+ (2014, 2016): rightmost es/mate includes H2 from ALMA; NGC5253, NGC1569, IC10 also include H2 in rightmost point dwarfs may not all be inefficient at forming stars compare with high-redshiG popula/ons, some approach maximal starbursts
SLIDE 13
low-mass galaxy resolved HI samples
VLA-ANGST (Very Large Array survey of Advanced Camera for Surveys Nearby Galaxy Survey Treasury): 35 nearby galaxies from the ANGST survey (OU+ 2012; Warren+ 2012; S/lp+ 2013a, 2013b, 2013c) SHIELD (Survey of H I in Extremely Low-mass Dwarfs): 12 galaxies selected from the ALFALFA survey having HI masses < 107 M¤ (Cannon+ 2011; McNichols+ 2016; Teich+ 2016) FIGGS (Faint Irregular Galaxies GMRT Survey): 65 faint dwarf irregulars (dIrrs) within D≤10 Mpc (Begum+ 2008; Rowychowdhury+ 2014; Patra+ 2016) 37 dIrrs, 4 Blue Compact Dwarfs (BCDs) within D≤10.3 Mpc (Hunter+ 2007; Hunter+ 2012; Zhang+ 2012; Oh+ 2015; Maier+ 2017; Iorio, Fraternali + 2017) LITTLE THINGS (Local Irregulars That Trace Luminosity Extremes, The H I Nearby Galaxy Survey):
SLIDE 14
low-mass galaxy resolved HI samples
Low-metallicity starbursts: 18 BCDs with measured star-forma/on histories (Lelli, Fraternali+ 2012a, 2012b, 2014a, 2014b); 18 BCDs with single-dish CO(1-0) IRAM detec/ons (Hunt+ 2015, in prep)
SLIDE 15
low-mass starbursts: scaling rela/ons of metallicity (O/H), star- forma/on rate (SFR), stellar mass (Mstar), and gas frac/ons
SLIDE 16 7.5 8.0 8.5 9.0 10−11 10−10 10−9 10−8 10−7
12+Log(O/H) SFR/Mstar
- 106 107 108 109 1010 1011
Mstar
LVL 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
local scaling rela/ons of stellar mass and SFR
Local Volume Legacy (LVL, KennicuU+ 2008, Lee+ 2009, Dale+ 2009, Cook+ 2014) within 11 Mpc COLDGASS (CO, followup of HI sample, Saintonge+ 2011, Huang+ 2014)
volume-limited samples also follow clear “main sequence” (Noeske+ 2007, Salim + 2007) of star forma/on (slope best fit to LVL sample, specific SFR SFR/Mstar: sSFR ~ Mstar
SLIDE 17
8.0 8.5 9.0 10−11 10−10 10−9 10−8 10−7
12+Log(O/H) SFR/Mstar
Iorio+ (2017) Hunt+ (in prep)
DDO87
NGC1140 SBS0335−052 IZw18
106 107 108 109 1010 1011
Mstar
LVL 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
local scaling rela/ons of stellar mass and SFR
but some galaxies (metal-poor low-mass starbursts) fall well outside the ±1σ spread of the main sequence (here shown by shaded grey area)
Local Volume Legacy (LVL, KennicuU+ 2008, Lee+ 2009, Dale+ 2009, Cook+ 2014) within 11 Mpc COLDGASS (CO, followup of HI sample, Saintonge+ 2011, Huang+ 2014)
SLIDE 18 gas frac/ons correlate with specific SFR
0.2 0.4 0.6 0.8 1.0 10−11 10−10 10−9 10−8 SFR/Mstar Mgas/Mbaryonic
Thuan+ (2016) Lelli+ (2014) Iorio+ (2017) Teich+ (2016) Hunt+ (in prep)
Thuan+ (2016) includes single- dish HI
extremely metal- poor (XMP) dwarf galaxies, all of which are extremely gas rich
SLIDE 19
8.0 8.5 9.0 0.0 0.2 0.4 0.6 0.8 1.0
12+Log(O/H) Mgas/Mbaryonic
Lelli+ (2014) Iorio+ (2017) Hunt+ (in prep)
107 108 109 1010 1011
Mstar
gas frac/ons also correlate with Mstar (less so with O/H)
SLIDE 20
feedback and the baryonic Tully-Fisher rela/on (BTFR)
SLIDE 21
102 106 107 108 109 1010 1011 Vrot [km s−1] Mbaryonic [Msun]
NGC5253 NGC4163 UGC6456 SBS1415+437 NGC4449
Lelli+ (2014) Lelli+ (2016) McGaugh (2012) Iorio+ (2017) McNichols+ (2016) Hunt+ (in prep)
Reyes+ (2011) Lelli+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
HI samples follow a fairly /ght baryonic TFR over 5 orders
Hunt+ (in prep)
SLIDE 22
102 106 107 108 109 1010 1011 Vrot [km s−1] Mbaryonic [Msun]
NGC5253 NGC4163 UGC6456 SBS1415+437 NGC4449
Lelli+ (2014) Lelli+ (2016) McGaugh (2012) Iorio+ (2017) McNichols+ (2016) Hunt+ (in prep)
Reyes+ (2011) Lelli+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
no/ce effect of adding molecular gas mass es/mated from CO (here included)
Hunt+ (in prep)
SLIDE 23 feedback affects the BTFR by suppressing baryonic frac/ons at low mass
EAGLE/APOSTLE simula/ons (feedback via stochas/c thermal energy injec/on) by Sales+ (2017) log Vrot (km/s) log Mbaryonic (M¤) cri/cal issue where to measure Vrot?
SLIDE 24
102 106 107 108 109 1010 1011 Vrot [km s−1] Mbaryonic [Msun]
NGC5253 NGC4163 UGC6456 SBS1415+437 NGC4449
Lelli+ (2014) Lelli+ (2016) McGaugh (2012) Iorio+ (2017) McNichols+ (2016) Hunt+ (in prep)
Sales+ (2017) EAGLE/APOSTLE Reyes+ (2011) Lelli+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
steepening of slope at low masses encompasses this feedback effect, data inconsistent with predic/on
SLIDE 25
102 106 107 108 109 1010 1011 Vrot [km s−1] Mbaryonic [Msun]
NGC5253 NGC4163 UGC6456 SBS1415+437 NGC4449
Lelli+ (2014) Lelli+ (2016) McGaugh (2012) Iorio+ (2017) McNichols+ (2016) Hunt+ (in prep)
Sales+ (2017) EAGLE/APOSTLE Reyes+ (2011) Lelli+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
correc/on for pressure support* brings some points closer to correla/on
NGC 5253, NGC 4449, SBS1415+437 all contain ionized-gas
(Heckman+ 2015, 2016; Chisholm+ 2015, 2016): trouble with feedback! Is Vrot truly tracing halo virializa/on? *correc/on =
√(Vrot
2 + 2 σ2) (e.g.,
Genzel 2010)
SLIDE 26 log Vrot (km/s) log M* (M¤) mass of disk transi/on
similar behavior in Hα STELLAR TFR at z~0.2 corrected by “kinema/c parameter”: S0.5 = √(0.5 Vrot
2 + σ2), see Simons, Kassin+
(2015)
109.5 M¤
SLIDE 27
pressure support and turbulence in low-mass galaxies : comparison of different gas tracers
SLIDE 28 Wisnioski+ (2015), see also Green+ (2010, 2014) Moiseev+ (2015) While Hα traces rota/on through central velocity posi/on, it does not trace true turbulence in the gas, but rather bulk mo/ons of the individual HII regions unresolved spa/ally
energy injected into the ISM by star forma/on drives (ionized) gas turbulence
Log(SFR) Log[L(Hα)] σ (km/s) velocity dispersion σ correlates with SFR in dwarf galaxies, giant spirals, U/LIRGs σ also increases with redshiG (because of increasing gas frac/ons, but remember scaling rela/ons, sSFR)
SLIDE 29 UGC8508 DDO53
10−2 10−1 100 101 102 SFR [Msun yr−1] σHI [km s−1] SFR ~ σ5.3
- Moiseev+ (2015) (ionized gas)
Lelli+ (2014) Iorio+ (2017) McNichols+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
HI velocity dispersions roughly constant with SFR
HI velocity dispersions do not show same correla/on with SFR as ionized gas
SLIDE 30 UGC8508 DDO53
10−2 10−1 100 101 102 SFR [Msun yr−1] σHI [km s−1] SFR ~ σ5.3
- Moiseev+ (2015) (ionized gas)
Lelli+ (2014) Iorio+ (2017) McNichols+ (2016)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Gas fraction
HI velocity dispersions do not show same correla/on with SFR as ionized gas
two galaxies currently with HI and Hα velocity dispersions measured, HI always lower
HI velocity dispersions roughly constant with SFR
SLIDE 31 White+ (2017): CO measurements of DYNAMO galaxies at z~0.2 selected for high SFRs as local analogues for z~2 clumpy disks
100 10−1 100 Mgas/Mbaryonic σ/Vrot
NGC5253
Iorio+ (2017) McNichols+ (2016) Q = 1 (White+ 2017)
HI dwarf samples (starbursts, LITTLE THINGS, SHIELD): HI velocity dispersion rela/vely constant with gas frac/on
CO velocity dispersions increase with gas frac/on
but sample selec/on effects?
SLIDE 32 Mogotski+ 2016 Caldu-Primo+ 2013
CO velocity dispersions ≈ HI velocity dispersions
THINGS galaxies: stacking spa/ally-
resolved spectra gives mean ra/o σ(HI)/σ(CO) = 1.0 ± 0.2 (Caldu-Primo+ 2013) Focusing on bright molecular regions gives σ(HI)/σ(CO) = 1.4 ± 0.2 (Mogotski+ 2016) Evidence for high velocity-dispersion (“thick”) molecular disk
σ(HI)/σ(CO)
σ(HI)/σ(CO)
SLIDE 33 A cau/onary tale: global versus (too much) detail
Mrk 206 (0.2 Z¤, 12+Log(O/H) = 8.07, SMC-like) with molecular gas [CO(1-0)] observa/ons with PdBI: CO peak offset from
integrated intensity CO(1-0) SDSS thumbprint, dwarf S0
Hunt, Garcia-Burillo, Combes, et al. , in prep
SLIDE 34 Mrk 206: regularly rota/ng disk, but lopsided σ(CO)
color Vrot color σ(CO)
high CO velocity dispersion (~ 40 km/s) coincides with highest dust ex/nc/on (Cairos+ 2010)
SLIDE 35
Theore/cal considera/ons suggest that low-mass (metal-poor) dwarf galaxies are forming stars inefficiently, because of the impact of feedback KS rela/on suggests that this is not generally true and that some low-mass galaxies can form stars as efficiently as high-z massive starbursts Gas frac/ons in low-mass galaxies correlate with Mstar and specific SFR, and are comparable with high-z galaxies Baryonic TFR of low-mass galaxies inconsistent with theore/cal ΛCDM predic/ons, but adding a correc/on for pressure support/turbulence could beUer align results Beware of different tracers of gas velocity dispersion (ionized, molecular, atomic) as they do not probe the same physical processes To test galaxy forma/on schemes, HI measurements of rota/on/dispersion veloci/es are needed for low-mass galaxies: APERTIF, MEERKAT, SKA.
conclusions
