The ISM in the Galactic Centre, molecular gas, star formation, the - - PowerPoint PPT Presentation

The ISM in the Galactic Centre, molecular gas, star formation, the Fermi Bubbles and high energy events

Ise-Shima Winter School Lecture 2

Roland Crocker Australian National University

The Galactic Centre ISM

2

In General: Extreme ISM in GC...

• SFR density over central ~200 pc ≳ 3 orders of

magnitude larger than mean in disk (∂tΣ∗ ~ 2 M⊙ yr-1 kpc-2)

• this activity (stellar winds, supernovae) sustains an

energy density in the different GC Interstellar Medium (ISM) phases about 2 orders of magnitude larger than typically found in the local ISM GC: UB ~ Uturb ~ Uplasma ~ UISRF ~ 100 eV cm-3 local: UB ~ Uturb ~ Uplasma ~ UISRF ~ 1 eV cm-3

3

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI

EAST Galactic Plane

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI

EAST Galactic Plane

~100 pc

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI SNR

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI SNR NTF

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI SNR NTF radio arc

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI SNR NTF radio arc HII (free free)

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI SNR NTF radio arc HII (free free) Sgr A complex

The Galactic Centre ISM

• Why is there so much SF activity?

...because there is lots of molecular gas

• Why is there lots of gas?

...because the Galactic Centre is the end-of-the- line: angular momentum loss suffered by gas in the disk will inevitably bring it into the GC eventually

6

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI

Tsuboi (1999)

Gas Density Through GC

purple: 20 cm radio

• range: 1.1 mm (cold dust)

cyan: IR (PAHs) Image courtesy of NRAO/AUI

Tsuboi (1999)

Gas Density Through GC

Tsuboi (1999)

Gas Density Through GC

Molecular line observations trace gas density and show up Central Molecular Zone of giant molecular cloud complexes bound in tight orbits (~100 pc) around GC - these contain ~ 5% of the Galaxy’s molecular gas (~3 × 107 solar mass) and at volumetric-average number densities of nH2 ~ 102 cm-3

Tsuboi (1999)

Gas Density Through GC

GC molecular clouds are unusually

• dense (~104 cm-3),
• turbulent (velocity dispersion > 15 km/s)
• warm (10’s K)

...when compared with Galactic Disk clouds

8

GC molecular clouds are unusually

• dense (~104 cm-3),
• turbulent (velocity dispersion > 15 km/s)
• warm (10’s K)

...when compared with Galactic Disk clouds

8

High density required for the clouds to resist tidal shearing in the GC environment

How does the gas reach the GC?

9

10Andrea Stolte

We live in a barred spiral galaxy

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space bar

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space bar

X1

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space bar

X1 X2

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space bar

X1 X2 X1

How does the gas reach the GC?

• Contopoulos & Mertzanides (1977): there are two classes
• f closed orbits in barred spirals

–X1 elongated along the bar, outside the ILR –X2 rounder, elongated perpendicular to the bar, interior to ILR

11

based on Morris 2008

phase space bar

X1 X2 X1 X2

How does the gas reach the GC?

• ILR = Inner Lindblad Resonance, at r ~ few 100 pc in

Galaxy, roughly corresponds to smallest X1 orbit that is non-self-intersecting

• 12

based on Morris 2008

At the ILR the torque is zero, inside (X2 orbits) torque is positive (gas driven outwards) Outside the ILR, the bar’s non-axisymmetric gravitational potential exerts a negative torque on the gas sending it inwards

How does the gas reach the GC?

• ILR = Inner Lindblad Resonance, at r ~ few 100 pc in

Galaxy, roughly corresponds to smallest X1 orbit that is non-self-intersecting

• 12

based on Morris 2008

At the ILR the torque is zero, inside (X2 orbits) torque is positive (gas driven outwards) Outside the ILR, the bar’s non-axisymmetric gravitational potential exerts a negative torque on the gas sending it inwards

How does the gas reach the GC?

• Gas therefore accumulates on a ring near the outer X2
• rbits (Wada & Habe 1992)
• Binney et al. 1991: the gas becomes molecular at the ILR

where it undergoes a compressive shock

• We expect that there should be a 100-200 pc ring of H2

inside the ILR; this seems to be consistent with

• bservations

13

Molinari et al. 2011, Herschel observations of a twisted, 100 pc ring around the GC

How does the gas reach the GC?

• Gas therefore accumulates on a ring near the outer X2
• rbits (Wada & Habe 1992)
• Binney et al. 1991: the gas becomes molecular at the ILR

where it undergoes a compressive shock

• We expect that there should be a 100-200 pc ring of H2

inside the ILR; this seems to be consistent with

• bservations

13

Molinari et al. 2011, Herschel observations of a twisted, 100 pc ring around the GC

How does the gas reach the GC?

• Gas therefore accumulates on a ring near the outer X2
• rbits (Wada & Habe 1992)
• Binney et al. 1991: the gas becomes molecular at the ILR

where it undergoes a compressive shock

• We expect that there should be a 100-200 pc ring of H2

inside the ILR; this seems to be consistent with

• bservations

13

Molinari et al. 2011, Herschel observations of a twisted, 100 pc ring around the GC

How does the gas reach the GC?

• the velocity difference between X1 and X2 orbits is up

to ~100 km/s

• this provides for high-velocity collisions between giant

molecular clouds on the different orbits

• ...these collisions may trigger star formation
• helical nebulae phenomena found in the GC may also

reveal this process in action (Matsumura et al. 2012)

• perhaps mechanism behind ‘pearls-on-a-string’ nature

(van der Laan et al. 2013) of star formation in nuclear stellar rings found in barred spiral galaxies

14

Matsumura et al. 2012

NGC 1512 true-color + red representing the F658N Maoz et al. 2001 250 pc

The GC’s molecular gas is a target for cosmic rays

17

pp ➞ Pion Decay ➞ secondaries

18

pions

π0

γ γ

π+ π- π+

neutron

e.g.

e+ ννν

Credit: HESS Collab

Credit: HESS Collab

Credit: HESS Collab

Credit: HESS Collab CS contours from Tsuboi et al. (1999)

• Correlated with gas column
• Hard spectrum: Fγ∝ Eγ-2.3
• Fγ ~ 1035 erg/s

GC Diffuse TeV Emission

There is also a diffuse, hard- spectrum (non-thermal) radio continuum flux from the GC

24

Diffuse, non-thermal RC

Diffuse, non-thermal RC

Diffuse, non-thermal RC

Diffuse, non-thermal RC

Diffuse, non-thermal RC

Diffuse, non-thermal RC

Diffuse, non-thermal RC

LaRosa et al. 2005

10 GHz Nobeyama data - spectrum is non- thermal up to 10 GHz

Diffuse Non-thermal Source (DNS)

10 GHz Nobeyama data - spectrum is non- thermal up to 10 GHz

Diffuse Non-thermal Source (DNS)

Q: What are the diffuse, non-thermal signals (γ-rays, radio continuum) we detect from the GC telling us? 27

Q: What are the diffuse, non-thermal signals (γ-rays, radio continuum) we detect from the GC telling us? 27 A: There must be diffuse, hard-spectrum populations of cosmic ray protons (and heavier ions) and cosmic ray electrons inhabiting the GC

Far Infrared-Radio Continuum Correlation (FIR-RC) Far Infrared-γ-ray Correlation (FIR-γC)

28

FIR-RC

L60μm = 1.3 × 108 L☀ L1.4 GHz = 1 . 2 × 1 019 W a t t / H z

Yun et al. 2001 ApJ 554, 803 fig 5

RC in deficit wrt expectation from FIR HESS system is 1 dex (> 4σ) off correlation i.e. GHz RC emission of HESS region only ~10% expected

Sidebar: origin of FIR- RC?

• correlation between FRC and RC

ultimately tied back to massive star formation (Voelk 1989)

• massive stars → UV → (dust) → IR
• massive stars → supernovae → SNRs →

acceleration of CR e’s → (B field) → synchrotron

FIR-γ-ray Scaling?

• SNR also accelerate CR p’s (and heavier

ions)

• there should exist a global scaling b/w FIR

and gamma-ray emission from region (Thompson et al. 2007): LGeV ~ 10-5 LTIR (assuming 1050 erg per SN in CRs)

• Given scaling, GeV emission only

~10% expected, TeV emission of HESS region only about 1% expected

FIR-γ-ray Scaling?

• SNR also accelerate CR p’s (and heavier

ions)

• there should exist a global scaling b/w FIR

and gamma-ray emission from region (Thompson et al. 2007): LGeV ~ 10-5 LTIR (assuming 1050 erg per SN in CRs)

• Given scaling, GeV emission only

~10% expected, TeV emission of HESS region only about 1% expected

FIR-γ-ray Scaling?

• SNR also accelerate CR p’s (and heavier

ions)

• there should exist a global scaling b/w FIR

and gamma-ray emission from region (Thompson et al. 2007): LGeV ~ 10-5 LTIR (assuming 1050 erg per SN in CRs)

• Given scaling, GeV emission only

~10% expected, TeV emission of HESS region only about 1% expected does

Martin 2011, Fermi collab

Martin 2011, Fermi collab

Martin 2011, Fermi collab

GC

Why is GC’s non-thermal emission much less than expected given its FIR?

• Explanation 1: a star-burst occurred more

recently than the lifetime (~107 years) of the massive stars which produce most UV and whose lives end in supernovae

• Explanation 2: GC SNRs are intrinsically low-

efficiency CR-accelerators

• Explanation 3: some transport process

removing non-thermal particles from system

Explanation 1: Starburst?

NO:

• Star-formation history of GC is a subject of

debate and we expect stochastic variation in SFR at some level

• BUT stellar population studies show GC star-

formation has been sustained over long timescales (2 Gyr) at more-or-less current rate (Figer et al 2004)...there has been no recent ‘starburst’ in the GC

• In any case, both Quintuplet and Central

cluster are old enough to have experienced core-collapse supernovae

Explanation 2: Low efficiency of SN as CR accelerators in GC?

• NO: detailed modelling shows that GC

supernovae act with at least typical efficiency as cosmic ray accelerators

Explanation 3: CR Transport

Explanation 3: CR Transport ✓

Explanation 3: CR Transport

BUT: Flat spectrum of in-situ electron and proton population → transport is advective not diffusive, i.e. via a wind [contrast situation in Galactic plane]

✓

Properties of the Outflow

Properties of the Outflow

Properties of the Outflow

CRs do not penetrate into densest gas BUT they can heat/ionize the low-density (warm) H2

...there is a wind of plasma, cosmic rays and magnetic field escaping the GC

• Modelling of broadband emission from GC

suggests that star-formation-related processes launch ≳1039 erg/s in CRs ions (and ≳1038 erg/s in CRs electrons) into an an outflow of a few 100 km/s wind

Galactic Centre Outflows

39

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

• RC studies show extended emission (1.2◦)

north of the plane whose spectrum steepens with distance (Law 2010)

• extended NIR emission mirroring RC (Bland-

Hawthorn and Cohen 2003)

• X-rays → apparent, diffuse, very hot plasma in

inner ~100 pc ... cf. external star-burst systems

• very extended X-ray emission (10’s degrees)

(Sofue 2000, Bland-Hawthorn and Cohen 2003)

Observational GC Wind Evidence

Herschel SPIRE 250 μm (Molinari et al. 2011)

Herschel SPIRE 250 μm (Molinari et al. 2011)

Herschel SPIRE 250 μm (Molinari et al. 2011)

Herschel SPIRE 250 μm (Molinari et al. 2011)

Herschel SPIRE 250 μm (Molinari et al. 2011)

Herschel SPIRE 250 μm (Molinari et al. 2011)

HESS TeV (Aharonian et al 2006)

Herschel SPIRE 250 μm (Molinari et al. 2011)

HESS TeV (Aharonian et al 2006)

Herschel SPIRE 250 μm (Molinari et al. 2011)

HESS TeV (Aharonian et al 2006)

Herschel SPIRE 250 μm (Molinari et al. 2011) Ring collimates outflow -

• utflow ablates cold gas

HESS TeV (Aharonian et al 2006)

Herschel SPIRE 250 μm (Molinari et al. 2011) Ring collimates outflow -

• utflow ablates cold gas

HESS TeV (Aharonian et al 2006) 2.7 GHz unsharp-masked Pohl et al 1992

Observational GC Wind Evidence

GC Spur (Sofue et al. 1989)

• Non-thermal

Collimated (~300 pc width if at the Galactic Centre) ...what is this thing? an AGN jet?

Fermi Bubbles

43

Fermi Bubbles

Su, Slatyer and Finkbeiner 2010 (ApJ)

Fermi Bubbles

Su, Slatyer and Finkbeiner 2010 (ApJ)

Fermi Bubbles

Su, Slatyer and Finkbeiner 2010 (ApJ)

Fermi Bubbles

Su, Slatyer and Finkbeiner 2010 (ApJ)

Fermi Bubbles

• 2 x 1037 erg/s [1-100 GeV]
• hard spectrum, but spectral down-break

below ~ GeV in SED

• uniform intensity
• sharp edges
• vast extension: ~7 kpc from plane
• coincident emission at other wavelengths

Microwave ‘Haze’

(Finkbeiner 2004)

Dobler 2012

Microwave ‘Haze’

(Finkbeiner 2004)

Leptonic Scenarios

• ~GeV γ-ray emission from IC by

hypothesised population of hard-spectrum ~TeV electrons

• same population synchrotron-radiates into

microwave frequencies

• BUT short cooling time (<Myr) ⇒

relativistic transport OR in situ acceleration (Cheng et al. 2011; Mertsch & Sarkar 2011)

• related to AGN-type activity(?): Su et al.

2010; Guo & Matthews 2011;

Gamma-ray Jet(?)

Su and Finkbeiner 2012 ‘jet’ detected with 5σ confidence

Extra Slides

51

23 GHz Polarized Intensity from WMAP

Fermi Bubbles

Fermi Bubbles

Fermi Bubbles

Fermi Bubbles

~100 pc

Fermi Bubbles

~100 pc

Spitzer, VLA 20 cm, CSO 1.1 mm BOLOCAM

NRAO/AUI

Fermi Bubbles

~100 pc

Spitzer, VLA 20 cm, CSO 1.1 mm BOLOCAM

NRAO/AUI

Fermi Bubbles

Fermi Bubbles