Download Feedback Regulates Star Formation

Feedback Regulates Star Formation
Phil Hopkins
F RE
Feedback In Realistic Environments
Norm Murray, Eliot Quataert, Chris Hayward, Dusan Keres, Claude Faucher-Giguere, Xiangcheng Ma
What Does Turbulence Do?
Giant Molecular Clouds:
Stars & Pre-Stellar Gas Cores:
MW
LMC
SMC
M33
M31
IC10
Number
Number (dN/dM)
Bastian
Chabrier
DM Halos?!
Blitz,
Rosolowski et al.
Mass [M ]
Turbulent Fragmentation: From GMCs to Stars
PFH 2011
JUST COUNTING “CLOUDS IN CLOUDS”
Averaging Scale R [pc]
Log[ Density / Mean ]
Thermal+
Magnetic
Turbulence
Angular Momentum
Turbulent Fragmentation: From GMCs to Stars
PFH 2011
JUST COUNTING “CLOUDS IN CLOUDS”
STRUCTURE FORMATION
STAR FORMATION
The “First Crossing” Mass Function
VS GIANT MOLECULAR CLOUDS
dn
/M
dM
↵
e
(M/MJ )
Number (>Mcloud)
GMCs
PFH 2011
log[ Mcloud / Msun ]
The “Last Crossing” Mass Function
VS PROTOSTELLAR CORES
vturb (R) < cs :
⇢
vturb
⇠
⌧1
⇢
cs
(Hennebelle & Chabrier,
Padoan & Nordlund,
PFH 2012)
Clustering of Stars/Cores
CLUSTERING IS INEVITABLE
PFH 2012
From Cores to Stars
D. Gusjenov
YOU CAN TAKE TURBULENT FRAGMENTATION TO THE IMF
“Fragmentation Tree”:
Time
arXiv tomorrow!
What About Starbursts (Extreme Environments)?
BOTTOM-HEAVY: TURBULENCE WINS!
MJeans is bigger
but MSonic is smaller
ULIRG
(bigger clouds with
more fragmentation)
Van Dokkum & Conroy
(nearby elliptical centers)
TextKroupa
Chabrier
MW
PFH 2012
(see Giles’s talk,
recent HC 2014 work)
Is Star Formation Self-Regulating?
Not Without Feedback!
TURBULENT FRAGMENTATION LEADS TO RUNAWAY COLLAPSE
(ALSO, WHAT DRIVES THE TURBULENCE?)
Fraction Collapsed Per Crossing Time
Simulation
Simulations (Cooling+Gravity+MHD)
Padoan & Nordlund
Vazquez-Semadeni PFH 2011
Federrath et al.
(see Paulo’s talk)
But Star Formation is Slow!
Q: WHY IS STAR FORMATION SO INEFFICIENT?
Log
SFR
Ṁ⇤ ⇡ 0.017 Mgas /⌧dyn
Kennicutt 1998
Log
gas
/
dyn
Galactic Scales: How Can We Do Better?
Ø
High-resolution (~1-10 pc), molecular/metal cooling (~10 K), SF at nH > 100 cm-3
Ø
Energy/Mass/Metal Injection:
Ø
Ø
SNe (II & Ia)
Ø
Stellar Winds (O & AGB)
Ø
Photoionization (HII) & Photoelectric
Momentum Flux:
Ø
Radiation Pressure
Ṗrad
Ø
Ø
L
⇠ (1 +
c
IR )
SNe
1
ṖSNe ⇠ ĖSNe vejecta
Stellar Winds
ṖW ⇠ Ṁ vwind
Ø
(also MHD, anisotropic conduction, diffusion)
Stars (Hubble image):
Blue: Young star clusters
Red: Dust extinction
Gas: Magenta: cold
Green: warm (ionized)
Red: hot
Is Star Formation Self-Regulating?
The Kennicutt Law Emerges
INDEPENDENT OF SMALL-SCALE SF LAW
No
Feedback
PFH et al. (arXiv:1311.2073)
(also Agertz+ 1404.2613)
Kennicutt-Schmidt relation emerges naturally
ISOLATED GALAXIES
⌃˙ ⇤ ⇠ ⌃gas /⌧dyn
no feedback
PFH, Quataert, & Murray, 2011a
⌃˙ ⇤ ⇠ 0.02 ⌃gas /⌧dyn
with feedback
Shetty & Ostriker ’12
CAFG et al. ’13
Kennicutt-Schmidt relation emerges naturally
Ø
Efficient cooling
Ṗdiss
Ø
the gas disk dissipates its support:
Mgas vturb
⇠ Mgas
⇠
tcrossing
disk
⌦
Collapse stops when momentum input from feedback:
Ṗ⇤ ⇠ Ṗdiss
L
⇠ ✏⇤ Ṁ⇤ c
Ṗ⇤ ⇠ few ⇥
c
✓
◆
˙⇤ ⇠
⌃
⌃gas ⌦ ⇠ 0.02 ⌃gas ⌦
✏⇤ c
(Galactic) Star Formation Rates are INDEPENDENT of how stars form!
Efficiency (SF per tdyn)
Ø
Index (SFR ~
)
SF Density Threshold
Set by feedback (SFR) needed to maintain marginal stability
Hopkins, Quataert, & Murray 2011
also Saitoh et al. 2008
How Does Star Formation Self-Regulate?
SELF-ADJUST THE MASS IN DENSE GAS
Efficiency (SF per tdyn)
Ø
SF Density Threshold
Pile up more dense gas until the SFR “needed” is obtained!
Hopkins, Quataert, & Murray 2011
Are We Done?
No! Star Formation is Inefficient In the Integral
Q: WHAT KEEPS GAS OUT OF GALAXIES?
Moster 2009
Number Density
Log( Stellar/Halo Mass )
of baryons
No10%
Feedback
Stellar
Feedback
SMBH
Feedback
Observed
Log( Halo Mass )
Mstars [M ]
How Efficient Are Galactic Super-Winds?
WHAT MECHANISMS DRIVE THEM?
No Feedback
S. Muratov et al., in prep
100
10
1
0.1
With Feedback
Does Stellar Feedback Explain the Mass Function?
HOW EFFICIENT ARE GALACTIC WINDS?
M⇤
observed
(Behroozi)
=
fb
n
o
y
ar
alo
h
M
observed
(Moster)
log (Mhalo / M )
PFH et al. (arXiv:1311.2073)
Sub-Grid Is Not Enough
PHASE STRUCTURE AND RECYCLING OF OUTFLOWS MUST BE CAPTURED!
S. Muratov
(stay tuned)
Proto-MW: Gas Temperature:
Insert Winds “By Hand” (Sub-Grid)
Following Full Feedback
Faucher-Giguere,
arXiv:1409.1919
Feedback Determines the Halo Gas Properties
ABSORBERS FALL OUT NATURALLY (EXCEPT QUASARS)
16
Faucher-Giguère et al.
16
Faucher-Giguère et al.
z=2
No feedback
m09
Feedback
With
feedback
(Quasars)
No feedback
NHIHI (cm-2-2)
kpc
kpc
Figure B1. Top: HI maps for our m12i simulation with full stellar feedback at
Dwarf Metallicities: Revealing Feedback
DEPENDS ON DETAILS OF INFLOW-OUTFLOW INTERACTIONS
observed
sims
“sub-grid” sims
SAMs
“Bathtub Models”
Ø
Outflows suppress “new” infall of pristine material
Ø
Metal-rich gas preferentially re-accretes in fountains
Xiancheng
Ma
Observed Starlight
Molecular
X-Rays
Star Formation
Ø
“Turbulent Fragmentation” at all scales:
Ø GMCs: universal mass function (power-law 1.x + cutoff at Toomre)
Ø Larson’s Laws: trace the turbulence
Ø Cores: universal slope 2.x, turnover at sonic scale
Ø IMF: fragmentation from cores: Salpeter slope & weak dependence of Mpeak
Ø Stellar Clustering: ~0.1pc - kpc, follows directly from turbulent fluctuations
Ø
Star formation is Feedback-Regulated:
Ø KS law is independent of small-scale SF physics!
Ø SF is not “slow because of turbulence”: is slow because feedback unbinds gas!
Ø
Cosmologically:
Ø Winds determine IGM enrichment, temperature, & subsequent inflow
Ø Resolved feedback 6= sub-grid feedback!
Ø Mass-metallicity, SFHs, morphology not the same