Ionised Gas in the Interstellar Medium

1
Ionised Gas in theInterstellar Medium
NASA, ESA, N. Smith (University of California,
Berkeley) and the Hubble Heritage Team
(STScI/AURA)
Greg Madsen Research Scientist, University of
WisconsinVisiting Scientist, University of Sydney
2
Outline

• Introduction and motivation
• HII regions
• structure, recombination radiation
• Discovery of the Warm Ionised Medium
• How to detect faint ionised gas Fabry-Perot
  interferometry
• The Wisconsin H-alpha Mapper (WHAM)
• WHAM Northern Sky Survey
• Physics of the Warm Ionised Medium
• collisional excitation
• example Perseus Superbubble
• Other WHAM projects
• WHAM-South

3
Why is the ISM important?

• galaxies are cosmic factories where stars are
  continuously formed and destroyed
• the ISM mediates this cycle of stellar birth and
  death
• least understood part of the cycle
• extreme physical conditions not reproducible in
  a lab (e.g. million-degree vacuum)
• fascinatingly complex, combines many fields of
  physics and chemistry

Courtesy Peter Woitke, Leiden Univ.
4
HII Regions

• the ISM is largely ionised by UV photons from
  hot stars (hn gt 13.6 eV, l gt 912Å)
• they carve out a cavity of ionised gas (H II),
  idealised as a Stromgren sphere (Stromgren
  1939)

• size is given by balancing ionisation and
  recombination

Q0 ionising photons per secondRs radius of
Stromgren spherear volume recombination raten
density
using typical values for Q, n, T

• the mean free path of an electron given by L
  1/(nesn)

for n 1 cm-3, L 0.05 pc -gt very sharp edge!
Rosette Nebula
(cf. Lecture 6,9)
5
Recombination Radiation
I don't like electrons they've always had a
negative influence on society -- Chris Lipe

• ionised gas radiates when electrons recombine
  (free-bound transition)
• volume emissivity given by

• one of brightest and most easily observed
  recombination lines is when n 3 -gt 2, called
  Balmer-a or Ha wavelength is 6562.56Å
• it is convenient to define the emission measure

that is directly proportional to the Ha surface
brightness (T4 T/104 K )

• for typical HII region (n10 cm-3, dl30 pc),
  EM 3000 cm-6 pc
• interstellar ionised gas ought to be confined
  to HII regions!

6
Subverting the Dominant Paradigm
Australian Journal of Physics, vol. 16, p.1, 1963
George Ellis

• in 1963, Hoyle and Ellis saw absorption of radio
  synchrotron radiation that they attributed to
  presence of low-density free electrons

Astrophysical Journal vol. 192, L53, 1974
Ron Reynolds

• in 1974, Reynolds was studying Ha emission from
  Earths atmosphere and noticed contamination
  from the Galaxy, EM 5
• took more than 15 years of observations and
  analysis to convincingly prove the existence of
  huge layer of interstellar ionised gas -gt
  called Reynolds layer or Warm Ionised Medium
  (WIM)

7
Why is the WIM important?
This wide-spread H II is one of the principal
phases of the interstellar medium and impacts our
understanding of the

• Morphology and structure of the ISM
• relationship to other phases (cold molecules,
  warm and cold neutrals)
• transport of ionising radiation
• feedback from massive star formation
• Heating and ionisation processes within galactic
  disks and halos
• Interpretation of extra-galactic observations
• contamination of cosmic backgrounds
• IR thermal emission from interstellar dust
• FUV 2-photon decay from n 2 of H
• CMB free-free emission
• quasar absorption lines
• intergalactic radiation field escape of
  ionizing photons

8
Fabry-Pérot Interferometry

• Challenge is to isolate the faint H-alpha
  emission line from the night sky
• Standard spectroscopic tools (prisms/gratings)
  throw away a lot of photons, not good for diffuse
  sources
• Fabry-Pérot etalons give you high throughput
  and high spectral resolution
• Operate on principal of interferometry
• Maximum transmission occurs when the path length
  difference is an integer multiple of the photon
  wavelength

A. Pérot (1863-1925)
C. Fabry (1867-1945)

• Tune an etalon to wavelength of interest by
  changing n or l
• Good etalons are very expensive, require special
  coatings

9

• Wisconsin H-Alpha Mapper
• 15 cm, dual etalon Fabry-Perot on a dedicated 60
  cm siderostat located at Kitt Peak and fully
  remotely operated from Madison and Sydney
• 1º diameter beam on sky
• High spectral/velocity resolution (R
  25,000, or 12 km/s)

• Can be tuned to any optical wavelength
• High sensitivity (EM 0.1 cm-6 pc in 30 s)
• Observes distribution and kinematics of diffuse
  ionised hydrogen

10
The WHAM Northern Sky Survey
Ha Map
37, 565 Ha Spectra
Ha Spectrum
Total Intensity
Haffner et al. 2003
11
The WHAM Northern Sky Survey

• faint, diffuse, warm ionized gas pervades the
  Galaxy
• ne 101 cm-3
• T 104 K
• nearly fully ionized
• volume filling fraction f 20
• scale height 1000 pc
• the only sufficient power source are UV photons
  from hot stars
• how do they penetrate neutral gas?

Orion-Eridanus Bubble
Northern Filament
Perseus Superbubble
12
Collisional Excitation

• bound electrons can be excited to higher levels
  via collisions with free electrons
• if the density is below a critical density,
  the excited electron will be radiatively
  de-excited, emitting a forbidden line, e.g.
  NIIl6583, SII l6716, OIII l5007
• the density of the ISM is generally lower than
  the critical density for several lines
• ionised gas cools primarily through these
  collisional lines (transfer thermal energy of
  free electrons into photons that escape)
• use the strength of observed emission lines to
  infer physical conditions of gas

In

• ratios of lines often minimise dependence of
  unknown parameters

NII / Ha
SII / NII
cf. Lecture 6
13
The Perseus Superbubble

• discovered a bipolar loop in the Perseus spiral
  arm D 2 kpc -gt loop extends 1 kpc above
  plane!
• group of massive stars at the base HI maps show
  a cavity
• whole area observed in line of NIIl6583
• want to compare WIM to HII regions
• found detailed anti-correlation between I(Ha)
  and NII/Ha I(Ha) high -gt NII/Ha low
  I(Ha) low -gt NII/Ha high
• whats going on?

I(Ha)
NII / Ha
14
The Perseus Superbubble

• Ionisation potentials H0 13.6eV He0
  24.6eV N0 14.5eV N 29.6eV S0 10.4eV
  S 23.4eV
• energetics imply thatN/N H/H and S is
  either S or S
• NII/Ha measures temperature
• SII/NII measures S/S
• diagram separates the two
• compared to HII regions, low-density WIM is
  hotter and in lower ionisation state

S/S 1.00
S/S 0.75
S/S 0.50
S/S 0.25
Temperature
15
Collisional Excitation

• bound electrons can be excited to higher levels
  via collisions with free electrons
• if the density is below a critical density,
  the excited electron will be radiatively
  de-excited, emitting a forbidden line, e.g.
  NIIl6583, SII l6716, OIII l5007
• the density of the ISM is generally lower than
  the critical density for several lines
• ionised gas cools primarily through these
  collisional lines (transfer thermal energy of
  free electrons into photons that escape)
• use the strength of observed emission lines to
  infer physical conditions of gas

In

• ratios of lines often minimise dependence of
  unknown parameters

NII / Ha
SII / NII
cf. Lecture 6
16
Physical Conditions of the WIM

• Use optical nebular emission line diagnostics
• Ha, Hb, N II l6583, l5755, S II l6716, O I
  l6300, OIII l5007, He I l5876
• Based on observations or upper limits toward a
  few sightlines or maps of limited coverage
  (Reynolds et al 1985, 1995, 1998, Tufte 1997,
  Haffner et al. 1999, Hausen et al. 2002)
• Compared to classical HII regions, the WIM is
  characterized by
• higher N II / Ha and SII / Ha -gt higher
  temperature (T 8000 K)
• higher NII l5755 / NII l6583 -gt confirms
  higher temperature
• lower He I / Ha -gt softer ionizing spectrum
• lower O III / Ha -gt lower ionization state
• higher O I / Ha -gt nearly fully ionized (H/H
  gt 0.9)

• Pure photoionization models have difficulty
  reproducing all of the line ratios and their
  variations (e.g. Sembach et al. 2000, Mathis
  2000)
• need to include 3-D structure of WIM, extra
  heating (Wood Mathis 2004)

Madsen, Reynolds Haffner 2006
17
An Optical Window into the Inner Galaxy

• Ha emission seen out to 7 kpc (RG 4 kpc)
• Quantify attenuation with Ha / Hb
• Infer scale height, density, ionization state

35 lt VLSR lt 65
Madsen Reynolds 2005
18
Survey of Large Planetary Nebulae

• Large, evolved PN trace the latest stages of PN
  evolution
• low surface brightness, misidentification as HII
  regions, SNRs, ISM
• Traditional long-slit spectroscopy difficult

Sh 2-200

• 45 targets observed to date (r gt 5)
• Accurate fluxes in Ha, NII, OIII
• High resolution spectroscopy
• Identify impostor PN
• Nebular morphology
• Emission line ratios
• Velocity, line widths
• WHAM imaging produces velocity channel maps

V -50 km/s (PN)
Bonafide PN
Ionized ISM
V 0 km/s (ISM)
Madsen Frew 2006
19
Ionized gas in High-Velocity Clouds

• HI clouds moving anomalously with respect to
  Galactic rotation
• Distances and metallicities not generally known
• Origin of clouds remain widely debated (Galaxy,
  Local Group)

• Physical conditions of HVCs can probe the
  Galactic halo
• Complex, multi-phase nature (e.g., Tripp et al
  2003, Fox et al 2005)
• Pressure confined in a hot halo?
• Ha measures incident Lyman continuum flux
• Escape of ionising radiation

HI Map of Complex A
Wakker et al (2001, 2003)
20
Kinematics of Zodiacal Dust Cloud

• Observed reflection of solar MgI absorption line
  
• Velocity amplitude and line widths are powerful
  discriminators
• non-circular orbits
• radiation pressure
• inclination
• Elliptical orbits required to match the data -gt
  cometary origin
• High ecliptic latitude observations suggest
  inclinations up to 30-40 º

Madsen et al 2007
21
Extragalactic astronomy with WHAM

• Ionisation of the extended HI disk of M 31
  (Madsen et al 2001)

• Search for ionised gas in dwarf spheroidal
  galaxies (Gallagher et al 2003)

• No detectable H I
• Recently formed stars?
• Upper limits of Ha emission constrain total gas
  mass

22
WHAM - South

• Form a more complete picture of the ionised gas
  in the Galaxy
• Velocity-resolved maps of Gum Nebula, Magellanic
  Clouds, Magellanic Stream
• Complementary to existing Ha, HI surveys (UKST,
  SHASSA, GASS)
• Include nod-and-shuffle, improve arcminute-scale
  imaging capabilities
• Moving to Cerro Tololo in Chile in September 2008