Extraplanar HI and its implications

1
Extraplanar HIand its implications

• James Binney
• and
• Filippo Fraternali
• (Oxford Bologna)

TexPoint fonts used in EMF. Read the TexPoint
manual before you delete this box. AAAAAAAAA
2
Overview

• The phenomenology of extraplanar gas
• The basic fountain model
• Galactic coronae
• Fountain/coronal interaction
• Prospects
• (Fraternali JB, 2006, 2007, MNRAS)

3
Extra-planar gas in NGC 891

• Sancisi Allen 1979
• NH 5 1020 cm-2
• Swaters et al. 1997
• NH 7 1019 cm-2
• Oosterloo et al. 2005
• NH 1.7 1019 cm-2

• Sancisi Allen 1979
• NH 5 1020 cm-2
• Swaters et al. 1997
• NH 7 1019 cm-2
• Oosterloo et al. 2005
• NH 1.7 1019 cm-2

• Sancisi Allen 1979
• NH 5 1020 cm-2
• Swaters et al. 1997
• NH 7 1019 cm-2
• Oosterloo et al. 2005
• NH 1.7 1019 cm-2

4
NGC891 Low rotation of extra-planar gas
Fraternali 2005
5
NGC 2403
.Distance 3 Mpc .Type Sc .Inclination
62 .Non-interacting .Very similar to M33
6
NGC2403 Extra-planar gas
Forbidden gas
130 km/s
Extra-planar gas
Fraternali, Oosterloo, Sancisi, van Moorsel 2001
7
NGC2403 Non circular motions
Thin disc
Extra-planar gas
8
Non-circular motions
9
NGC 6946 Extra-planar gas and SF
WRST
Boomsma PhD 2005
10
Extra-planar gas and star formation
11
Summary (observations)

• Extra-planar gas detected up to 15 kpc from plane

• Rotation lower than the disc

• Global inflow motion

• High velocities (100-200 km s-1)

• Link with star formation?

• Evidence for accretion?

Too much gas for just accretion
12
How common is halo gas?

• Halo gas (HI) found and studied in 7 galaxies
• NGC891, N2403, N6946, N253 (Boomsma et al. 2005),
  
• N4559 (Barbieri et al. 2005), UGC7321 (Matthews
  Wood 2003),
• NGC2613 (Irwin Chaves 2003).

13
Dynamical models
Previous works

• A barotropic pp(r) fluid in a gravitational
  field corotates (Poincaré, 1893)

• Hydrostatic models for non-barotropic fluid show
  gradient in
• rotation velocity but also high temperatures
• (Barnabé, Ciotti, Fraternali, Sancisi, AA,
  submitted)

• Galactic fountain gas circulation
  (disc-halo-disc)
• (Shapiro Field, ApJ 1976 Bregman, ApJ 1980)

• Ballistic models disagreement between predicted
  gradient in
• rotation velocity and H? data
• (Collins, Benjamin Rand, AA 2002)

14
Fountain model(Shapiro Field, ApJ 1976
Bregman, ApJ 1980 Fraternali Binney 06 )

• Clouds ejected from circular orbits with
  distributions in v, ?
• Surface density / (HI)1.3

• Clouds move ballistically as in Collins,
  Benjamin Rand, AA 02, but may not be visible
  until zmax or rmax

• Axisymmetry exploited to build pseudo-data cube

• Clouds return to disk on first or second passage
  through z0
• lt4 of SN energy needed

15
Dynamical model

• Continuous flow of particles from the disc to
  the halo

• Initial conditions distribution of kick
  velocities

• Potential exponential discs bulge DM halo

• Integration in the (R,z) plane, then
  distribution along ?

• At each dt projection along the line of sight

• Stop at the first or second passage through the
  disc

• Pseudo-cube to be compared with HI data cube

16
Model constraint vertical distribution
Vkick 75 km s-1 Mhalo 2 109 M?
residuals
17
N891 inflow/outflow
Travel times
Energy input lt4 of energy from SNe
18
NGC 891 Lack of low angular momentum
Fast rotating gas
?NEED FOR LOW ANGULAR MOMENTUM MATERIAL
19
NGC2403 lagging gas
Vkick 70 km s-1 Mhalo 5 108 M?
20
NGC2403 inflow/outflow
Radial outflow
?NEED FOR INFALLING MATERIAL
21
Second-passage models
V?
VR
Vz
V?
VR
Vz
22
Phase-change models
NGC 2403
NGC 891
Fast rotating gas
23
Phase-change models
Vertical motions
24
N2403 substructures
25
Coronae

• Feedback efficient
  (1/3 of baryons in galaxies)
• Outflows from starburst galaxies observed
• If any gas bound to galaxies, it is at
  T' Tvir' 106K

M82 (Chandra /Hubble/Spitzer)
26
NGC 891 X-ray halo

• Fit isothermal halo (T2.7106K)

Strickland 04
27
Cloud-corona interaction

• tcool À tflow D/v ' 1 Myr
  (D100 pc, v 100 km/s)
• Drag deceleration Cv2/L
  C 1, Lm/(½¾)
• So v(t)v(0)/(1t/tdrag) where
  tdrag L/Cv(0) ' 300 tflow ' 300Myr
• Hence drag significant for HI

28
Impact on corona

• Rate of absorption of angular momentum
• For gas at zgt1.3kpc,
• ) tspin (2Mcorona/106M)Myr
• But Mcorona 3107(r/10kpc)M so
  tspin 60(r/10kpc)Myr

29
Inescapable conclusion

• Star-forming disks act like impellers of
  centrifugal pumps
• Must profoundly modify the structure of
  X-ray corona

30
Accretion

• Many arguments for accretion, including
• G-dwarf problem
• HI exhaustion
• Sustained star formation
• Net infall of HVCs around Milky Way
• Incoming gas not expected to have L parallel to L
  of existing disk (warps)

31
Accretion the fountain

• Suppose fountain clouds sweep up infalling gas so
  
• Then
• Results insensitive to (R,z) take constant
• Take mean vi either along z or along r with
  Gaussian scatter around mean ¾ ' 50 km/s

32
NGC 891

• Polar 2Gyr-1
• Radial 0.6Gyr-1

33
NGC 891 (rotation)

• Good fits to data _at_ 3.9 and 5.2 kpc

34
NGC 2403

• Produces asymmetry indicative of infall
• Polar 2 Gyr-1
• Radial 1.2 Gyr-1

P-v diags for minor axis and 2
35
NGC 2403
36
Conclusions

• Star-forming galaxies cycle their HI through halo
  several times over life
• (NGC 891 hasgt25 of HI in halo)
• Simple fountain model fails because
  (i) halo predicted to rotate too fast, and
  (ii) halo predicted to show outflow
  rather than inflow
• X-ray emitting gas at Tvir exists around NGC 891
• HI halo must act on it like a pumps impeller

37
Conclusions (cont)

• Adding low Lz infall at rate expected a priori
  solves both fast-rotation and outflow problems of
  simple fountain model
• HI clouds need to gain mass at rate '1 Gyr-1
• Next steps
  (i) to use model to explain HVC
  population of the Milky Way
  (ii) to understand
  dynamics of X-ray gas