alternative gravity vs. cdm jerry sellwood. settling the argument requires clear predictions that...

Alternative gravity vs. CDM

Jerry Sellwood

Settling the argument

• Requires clear predictions that distinguish one from the other– consistency with one or the other is not enough if

both make similar predictions

• Alternative gravity is more easily falsifiable– e.g. Milgrom predicted TFR for LSBs

• not yet regarded as decisive by the CDM folks

– but predictions must be well-worked out!

WMAP 3-year data

• Rules out all no DM models?

• No!

Falsifiable predictions of AG

• Baryonic mass should be correlated with dynamical mass. Vulnerable to:– one rogue galaxy rotation curve– similar light distributions with very diff. M/L– etc.

• The shape of luminous matter should be reflected in the shape of the mass– no misalignments or offsets, etc.

Other concerns

• Galaxy clusters

• Dwarfs & globular clusters

• Dynamical friction and galaxy mergers

• ….

Challenging CDM

• Gauntlet already thrown down:– TFR for LSBs– Why does MOND work?

• Issues involving gastrophysics are too murky

• Somewhat firm predictions of DM halos– cusp/core issue – still no surrender!– absolute density scale

• But target just moved!– baryon/dark mass fraction– tilted or running spectral index

The greatest challenge to CDM• Spherically averaged density of dark matter halos

seems to approximate the form:

(r) = s rs3 / [r(r+rs)3-]

• i.e. a broken power law, with 1 < < 1.5 = 1 is “NFW”

Concentration

s is directly related to the concentration parameter

c = r200/rs

• c correlates with mass – halos are predicted to be a 1-parameter family (e.g. Bullock et al.)

Halo density

• Dark matter halos are not as dense as predicted

• Plot from Alam et al.v/2 is the mean density

inside the radius at which the DM rotation curve reaches vmax/2

• Points are estimates from real galaxies

• Heavy curve is for NFW and standard CDM

Tilted or running power spectrum

• Zentner & Bullock (2002):

• Lower values of v/2 predicted

– by about a factor 10 in their most extreme model (n.b. 8 0.65)

1 practical difficulty• How much mass

should be assigned to the stars?

• Disk-halo degeneracy

• Low surface-brightness galaxies and dwarfs are more dominated by DM

Measure disk mass dynamically

Measure disk mass dynamically

Magnitude of discrepancy

• Weiner’s work gets around uncertainty in M/L

• Milky Way similar (Binney & Evans 2001)

• Better data are in worse agreement

• Halos are under-dense by factor > 30 for n=1 models> 5 for extreme tilted

power spectra

• assumes =1 and ignores compression!

Effect of halo compression

• Conservative values:– NFW halo

– baryon fraction fb=0.05

– disk scale: rs/Rd=5

• Value of v/2 increased by factor 4

• In Weiner’s cases, it would be a factor > 30(decompression is hard)

Bar-halo friction• Consistent with

Debattista’s work on dynamical friction

• Rlast is Rc/aB when the simulation was stopped

• Rc/aB > 1.4 quickly in high-concentration models

• Bars stay fast for 30 disk rots only if c < 6

Reduce DM density?

• Feedback – Gnedin & Zhao– points vs. dashed– maximum possible

effect – factor 2– for a disk of

reasonable size

Reduce DM density?• Feedback – Gnedin & Zhao

• Binary BHs – Milosavljevic & Merritt– DM particles ejected as the binary hardens– removes about as much mass as the BHs– but only to a radius of a few hundred pc

Reduce DM density?• Feedback – Gnedin & Zhao

• Binary BHs – Milosavljevic & Merritt

• Bars – Weinberg & Katz

Bar-halo interaction• Holley-

Bockelmann, Weinberg & Katz (2005)

• Smaller changes reported by Weinberg & Katz (2006)– argue problem is

very challenging numerically

Density reductions

• 5 skinny, massive bars of different lengths

• flatten the cusp to about 1/3 bar length

• interesting, but unreasonable bar required

Rapid convergence

with N• Use the shortest bar

– 104 N 107

– dotted curve for unequal mass particles

• Number of terms in expansion, fine grid, etc. all make no diff.

• No evidence to support WK05 worries

Weaker bars• Flattening of the cusp occurs only for bars that are both

– strong: axis ratio 4:1 or greater, and

– massive: Mb > 40% of enclosed halo mass

• Sudden change in density – a collective effect• Smaller and more gradual density change for slightly

weaker bars – but over a greater radial range

Maximum effect

• Rigid bar highly artificial– increase MoI by factor 5– more significant density reduction

• Reduction in v/2 is only by 39% in most extreme case– Angular momentum transferred: 0.01– i.e. most of that in the baryons

• And this was for a huge bar (a = rs)

Conclusions

• Best data on halos in galaxies indicate densities lower than LCDM prediction by factor >10– assumes =1 and neglects compression

• No internal dynamical mechanism can reduce the density by much– maximum 40% for most extreme bars– results from careful simulations can be trusted

• Simply cannot unbind the halo– not enough energy can be extracted from the baryons– trying to make the tail wag the dog!