Dark Matter in Dwarf Galaxies

1
Dark Matter in Dwarf Galaxies

• High-Resolution Measurements of the Density
  Profiles of Dwarf Galaxies

Josh Simon UC Berkeley
Collaborators Leo Blitz Alberto Bolatto Adam
Leroy
2
The Central Density Problem
cusp
core

• Parameterize density profile as r(r) µ r -a
• Observations show a 0 (constant-density core)
• Simulations predict 1 ? a ? 1.5 (central cusp)

3
Improvements Over Previous Work

• 2-D velocity fields
• observations in Ha, CO, and HI
• can detect noncircular motions
• Nearby targets high spatial resolution (100
  pc)
• Multicolor optical/near-IR imaging
• better stellar disk model
• Concentrate on the simplest galaxies
• low mass, no bulges, no bars
• Test for systematics!

4
Targets
5
Targets
6
NGC 2976

• Sc dwarf galaxy in the M 81 group (D 3.5 Mpc)
• Gas-rich, no bulge, no bar, no spiral arms

• High-quality data
• 2-D velocity fields in Ha and CO
• BVRIJHK photometry to better model
  stellar disk

See Simon et al. (2003) for more details
7
NGC 2976 Velocity Field

• Fit a tilted ring model
• vobs vsys vrot cos q vrad sin q

Ha
CO
8
NGC 2976 Rotation Curve

• Rotation velocity
• Derived from combined CO and Ha velocity field

9
NGC 2976 Rotation Curve

• Significant radial motions in inner 30 (blue)

• Rotation velocity
• Radial velocity
• Systemic velocity

10
NGC 2976 Rotation Curve

• Power law provides a good fit to rotation curve
  out to 100 (1.7 kpc) (red)

• Power law fit

11
Maximum Disk Fit

• Even with no disk, dark
• halo density profile is
• r(r) 1.2 r -0.27 0.09 M?/pc3

12
Maximum Disk Fit

• Even with no disk, dark
• halo density profile is
• r(r) 1.2 r -0.27 0.09 M?/pc3

HI
H2
13
Maximum Disk Fit
stars
14
Maximum Disk Fit
dark halo
15
Maximum Disk Fit
16
What About the Systematics?

• Beam-smearing
• beam lt 100 pc gt 1100 independent data points
• Errors in geometric parameters
• center position, PA, inclination, systemic
  velocity
• Extinction
• vHa vCO
• Asymmetric drift
• After accounting for systematics, total
  uncertainty on density profile slope is 0.1

17
Targets
18
NGC 5963 The NFW Galaxy

• Larger and more distant galaxy (D 13 Mpc)
• Compact inner spiral surrounded by very LSB disk

19
NGC 5963 Rotation Curve
NFW profile also a good fit! V200 90 km s-1,
R200 130 kpc, rs 7 kpc
Best fit a 1.28 power law
20
Galaxy 3 NGC 4605

• Nearby (4.3 Mpc), LMC-mass, CO-rich pure disk
  galaxy

See Bolatto et al. (2002) and Simon et al.
(2004) for more details
21
Galaxy 4 NGC 5949

• More distant (14 Mpc), otherwise looks just like
  NGC 2976

NGC 5949
NGC 2976
See Simon et al. (2004) for more details
22
Galaxy 5 NGC 6689

• 11 Mpc away, slightly more highly inclined and
  more massive

See Simon et al. (2004) for more details
23
Is There a Universal Density Profile?

• No evidence for a universal density profile
• large scatter compared to simulations
• mean slope shallower than simulations

• Also different from previous observations,
  though
• e.g., a 0.2 0.2 (de Blok, Bosma, McGaugh
  2003)

24
Puzzles

• 1) Radial motions - whats causing them?
• Bar, triaxial dark matter halo, intrinsically
  elliptical disk
• Not only present in our sample - most 2D velocity
  fields show evidence for them
• Could have been missed in other galaxies due to
  long-slit observations . . .

25
Are Galaxy Halos Triaxial?

• Triaxial DM halos cause noncircular motions in
  disks
• 4/5 galaxies show measurable orbital ellipticity
• Lower limits on the potential ellipticity range
  from 0.5 to 3

26
Puzzles

• 1) Radial motions - whats causing them?
• Bar, triaxial dark matter halo, intrinsically
  elliptical disk
• Not only present in our sample - most 2D velocity
  fields show evidence for them
• Could have been missed in other galaxies due to
  longslit observations . . .
• 2) How can a rotation curve be fit by both a
  pseudo-isothermal profile and a cuspy power law?

27
Distinguishing Cores From Cusps
NGC 6689
NGC 5949

• Only exquisite data can distinguish cores from
  cusps in these galaxies
• Even then, the galaxies have to be very well
  behaved
• If you look for cores, you will find them. Same
  for cusps. Phrasing the debate as cores vs. cusps
  may not be the most useful approach . . .

28
Conclusions

• 1) Galaxy-to-galaxy scatter in density profile
  slope (Da 0.46) is much larger than in
  simulations

2) Mean slope (a 0.77) is shallower than
predicted
3) Disagreement between observations and
simulations is real, and systematics are only
partially responsible