Galaxy feedback in 3D
Roberto Maiolino
Feedback is the key:- reducing SF efficiency
(gas heating+turbulence)- removing baryons
Discrepancy between halo mass functions and stellar/baryons mass function
Star formation feedbackSNeRadiation pressurePhotoionization
TurbulenceVelocity dispersion
AGN feedback Quasar modeRadio mode
Blast waveRadiation pressure
(Q>1)
Star formation feedback
SNe & radiation pressuredriven outflows
Herschel PACSM82
velocity field velocity dispersiongalaxyrotation
turbulentoutflow
X-rays(Chandra)
NIRNGC 253
Ha
X-rays
CO~170 km/s(ALMA)
Mout ~ 1-2 SFR.
Contursi+13, Kreckel+14
Bolatto+13
PAH(Spitzer)
[CII]158um (Herschel)
The importance of tracingoutflows in multiple phases
Investigating outflows as a function of M* and SFR(breaking the M*-SFR degernarcy)
160,000 SDSS galaxies sorted into 50 bins in the M*- SFR parameter space
Lowest M* ≈ 2 x 107 M Lowest SFR ≈ 2 x 10-3 Myr-1
Cicone+14
MainSequence
7
THE LINE-OF-SIGHT VELOCITY DISTRIBUTIONS
[OIII]
Hα+[NII]
STARS
v [km/s]
Trends with M*
at a fixed SFR
Cicone+14
8
THE LINE-OF-SIGHT VELOCITY DISTRIBUTIONS
[OIII]
Hα+[NII]
STARS
v [km/s]
Trends with SFR
at a fixed M*
9
OUTFLOW PROPERTIES
Stacking grid colour-coded with the excess of gas velocity with respect to stars which traces non-virial motion of gas (due to outflows)
• vGAS – vSTAR increases with SFR and sSFR Expected for star formation-
driven outflows
• vGAS – vSTAR decreases with M* for fixed SFR Qualitatively agrees with
models
Agreement with indirect observations of feedback (e.g. mass-metallicity relation)
v[OIII] - vSTARS
vHα- vSTARS
[OIII]
Hα
10
OUTFLOW PROPERTIES
Outflows are detected only above the main sequence of star formation
• Sharp increase of outflows at SFR > SFRMS
• Outflows eventually quench star formation and lower the SFR, bringing the galaxy back on the MS
Self-adjusting mechanism that may explain the tightness of the MS
v[OIII] - vSTARS
v [OIII
] – v
STAR
S [k
m/s
]
ΔMS = offset from the Main Sequence
Cicone+14
11
Next steps:- 3D on large samples of galaxies required to obtain outflow rate and PK SDSS4-MANGA- Investigate the Atomic (NaI) and Molecular components
in 3D
v [OIII
] – v
STAR
S [k
m/s
]
ΔMS = offset from the Main Sequence
• Sharp increase of outflows at SFR > SFRMS
• Outflows eventually quench star formation and lower the SFR, bringing the galaxy back on the MS
Self-adjusting mechanism that may explain the tightness of the MS
v[OIII] - vSTARS
Star formation feedback
at high redshift
Steidel+10Bradshaw+13Diamond-Stanic+12
z~3
Integrated
Outflow load factor M
SFR~ 1-2
.
3D resolved
z~2
Genzel+12
The need of feedback becomes dramatic at high-z
Photoionization the main driver?
Need to trace both ionized and neutral ISM at high-z
e.g. Lya, Ha e.g. [CII], [CI], CO
Feedback in the early Universe
[CII]158mm (ALMA)
Lya emitter at z=4.7SFR ~ 20 M/yr
Williams+14
QSO
SMG
Neutral-cold gas ([CII]) probably account just for a small fraction of the whole galaxy traced by Lya
Lya
Lya (FORS2)
But peculiar environment (QSO)…
Y-band(Lya+UV)
Vanzella+12
Star forminggalaxy at z=7.1(reionization
epoch)
SFR ~ 9 M/yr
[CII]158mm(~neutral cold gas)detected at z=7.1with ALMA
5 kpc
beam 0.5”x0.7”
Maiolino+14
Y-band(Lya+UV)
Star forminggalaxy at z=7.1(reionization
epoch)
SFR ~ 9 M/yr
FWHM8 km/s!
5 kpc
beam 0.5”x0.7”
[CII](ALMA)
Star forminggalaxy at z=7.1(reionization
epoch)
SFR ~ 9 M/yr
FWHM8 km/s!
5 kpc
beam 0.5”x0.7”
velocity vel. dispersion
km/s
km/s
Mdyn ~ 3 107 M
[CII](ALMA)
FWHM8 km/s!
[CII]
Lya+UV
5 kpc
3 kpc
beam
[CII](ALMA)
offset~3 kpc
…as expectedby modelsof primeval
galaxiesVallini+13
galaxy mostlyphotoionized+ accreting
clumpsof cold gas
AGN feedback
“Quasar mode”
Quasar-driven outflows detected in theatomic neutral and ionized component
Rupke+11,13, Spoon+11, Muller-Sanchez+11
Wavelength (Å)
Narrow Ha
[NII][NII]
Broad Ha (outflow)
F l (n
orm
alize
d)
Mrk231Mrk231
NaI
2kpcve
loci
ty [k
m/s
]
lambda6100 6160
1200 km/s
velo
city
[km
/s]
Evidence for massive quasar-drivenmolecular outflows
CO(1-0) high velocity wings
Massive molecular outflows (~1000 M/yr)Extended on kpc scalesPK ~ 0.05 LAGN
Momentum rate ~ 20 LAGN/c
OH P-Cygni profiles
Fischer+10Sturm+11
1200 km/s
5kpc
Feruglio+10,13Cicone+14
blue wing
…as expected(and required)by models
l (mm)
Mrk231
The importance of high resolution3D spectroscopy
Recurrent explosive processEnergy driven quasar outflows1kpc
1kpc600 – 800 km/s 800 – 1000 km/s
Mrk231
600 – 800 km/s
CO(1-0)
CO(1-0)
CO(2-1)Feruglio+, in prep.
IRAM - PdBI
Simulations: only diffuse gas is ejected, no dense gas… AGN not really capable of quenching star formation?
Gabor+14
Actually a lot of dense (~105 cm-3) gas is observed in the outflow…
Aalto+12
…actually there is more dense gas in the outflow(relative to diffuse) than in the disk!
Alternative scenario: molecular outflowing cloudscondensed out of the shocked ouflowing gas…… would be consistent with AGN only ejecting diffusegas and the detection of dense outflowing gas…
Zubovas & King ‘13
Expected Observed
log (critical density) [cm-3]
outfl
ow si
ze [
kpc]
Cicone+12
but 3D info:
n n
AGN feedbackat high-z
(that’s what is reallyneeded to account for
local massive ellipticals)
V[OIII] >300 km/s(outflow)
Ha (SFR)
Cano Diaz+12
Alexander+10Forster-Schreiber+13Cimatti et al. 13Weiss+13
Bourget + 2012, 2013
z=3.0 z=2.4
z=3
z=6.4
Maiolino+12, Cicone+14
[CII] Moutflow up to >3000 M/yr .
Sizes >10 kpc
Depletion timescales < 107 yr
Keeping the galaxy clean
“Radio mode”
“Quasar mode”
Cleaning the galaxy
radio
optical + radio
Evidence for radio-mode AGN feedback in LOCAL galaxies(maintenance mode)
Heating ~ CoolingIt works!
… Really?
Having, ON AVERAGE, balance betweenheating and cooling does not
mean that some regions cannot undergooverheating and other OVERCOOLING
Salome’+11Russel+14Hamer+14
Streams ofcooling molecular gas
towards the centerof Perseus
“Quasar-feedback” phase driven by jets?
Wagner & Bicknell 2011, 2012
4C12.50
SCIENCE (today)
blueshifted 21cmabsorption
see alsoMorganti+03,05, Emonts+03
ALMA CO(1-0)
Observational evidence and characterizationof feedback in galaxies:
green field
lots of open questions
rapid growth expected in the coming few yearsthanks to the new 3D facilities