Underground Cosmology

1
Underground Cosmology
The stuff that Chung-Pei doesnt want you to know!
(ie. Topics or discoveries that wouldnt be
covered in a normal cosmology class)
Brad Hagan
2
Outline

• Variation of the fine structure constant, ?
• Theoretical basis and implications
• Original observations
• Follow-up observations
• MOND Modified Newtonian Dynamics
• Theoretical Basis and Implications
• Does it mesh with all observations?
• Is the universe a soccer ball?
• The original paper
• The WMAP teams response

3
Letting ??vary with time

• (it already varies with the energy being probed
  it increases as energy increases)
• Its difficult to consider variations of just one
  parameter like c or e because those variations
  are indistinguishable from each other. One must
  only consider variation of dimensionless coupling
  constants
• First considered in detail by Dirac when he
  noticed that
• (EM force between p and e)/(Gravitational force)
  
• (H0/c)/(classical e radius) 1040 (probably
  just a coincidence)
• Quintessence models, in which a scalar field
  couples to matter and gauge fields, can cause
  coupling constants and masses to vary as the
  field slow-rolls (this is a model for dark
  energy)
• Time varying coupling constants may be related to
  how large extra dimensions evolve (Marciano, PRL,
  52, 489)
• The bottom line it wont have a major effect on
  most of cosmology but could verify some
  fundamental high-energy physics (string theory,
  slow-roll quintessence, etc.) ?it could be the
  only evidence they ever see!

4
Initial Limits

• Cosmology results are strongly coupled to ??
• BBN works so well that ????????????over the age
  of the universe (??linked to BBN via
  proton-neutron mass difference)
• Temperature of CMB gives a similar limit at
  recombination time (through H-binding energy)
• ???????????
• Local limits
• Oklo natural fission reactor inAfrica finds no
  variation in 2 Gyr (very small error bars ? tight
  constraint on the recent past)
• Isotopes in meteorites no variation in 4.6 Gyr
  (solar system formation), but error bars are
  large enough to admit the interesting QSO results
  (see below)
• Atomic clocks (hyperfine and electronic
  transitions) are another independent test, but
  provide only loose constraints on our present-day
  value Oklo constraint is the most powerful
  non-cosmological constraint

5
Using QSO Spectra

• Cant use absolute values of lines those simply
  determine the z of the absorber or the z of the
  emitting QSO
• Instead, use splitting and measure relative
  locations. 2 main methods
• Alkali doublets in emission lines (e.g. Bahcall,
  et al. 2004)--no significant variation found
• Many multiplet method (e.g. Murphy, Webb,
  Flambaum, 2003)--the really interesting one

From Murphy, et al. 2003
6
MM Method and Results

• Used Keck/HIRES for all spectra
• Analyzed a large number of absorption systems
• Unlike AD, looked at many different lines and
  isotopes
• In principle, you get more data, but the models
  you match to arent very clean
• You see 100s of lines from many distinct clouds
  along the line of sight
• Vulnerable to systematic errors

Raw data from Murphy,et al. 2003
7
Discussion

• The matching procedure lays two spectra on top of
  one another one has the absorption features, the
  other does not
• The comparison looks at relative positions of
  lines (which can be very complex)
• The alkali-doublet method (e.g. Bahcall et al.)
  is very clean (requires no relativistic
  corrections and examines one emission system
  OIII), but it suffers from less data (lower S/N)
  and thus places looser constraints

8
After some analysis . . .
9
Do people believe the results?

• Using the ESO KUEYEN/UVES and going out to
  smaller redshift (2.3 rather than 3.7), Chand, et
  al. (2004) find no significant variation
• Others (Ashenfelter, Matthews, and Olive, 2004)
  think the method measures Mg isotopic abundances
  more than ?
• Assumed the same abundance as the Sun
• Primitive clouds may have different numbers of
  heavy isotopes of Mg (heavier than 24Mg)
• If theres no 25Mg or 26Mg in the clouds, both
  sets of observers find significant variation in ?
• Future work find correlations with stellar
  evolution using other heavy elements
• Plenty of room for more work!
• I think the basic reaction is that were going to
  need more data for such a strong, revolutionary
  claim

10
MOND

• MOdified
• Newtonian
• Dynamics

(cant modify gravity on certain mass or
length scales because we see the effects of DM on
many different and disparate scales)
11
MOND no CDM

• Proposed in 1983 by Moti Milgrom to fix flat
  rotation curves in galaxies
• Alters Newtons 2nd Law so we dont need CDM

• Thus, aMOND aNewton for accelerations larger
  than a0, but aMOND (aNewtona0)1/2 for aMOND ltlt
  a0
• Empirically, a0 1.2 x 10-8 cm s-2, a value too
  small to measure directly (also, its curiously
  close to cH0)
• Found from fitting these equations to galactic
  rotation curves
• The result accelerations are higher than
  expected given the same tiny force ? stars have
  higher circular velocities ? bump up the rotation
  curve where it would otherwise be decreasing

12
Galactic Civil War What side are you on?

• Pros
• Rotation curves match many galaxies given their
  M/L, even some cases where CDM seems to fail
• Produces correct predictions that it was not
  invented to make (LSB galaxy rotation curves in
  1998 predicted in 1983 by Milgrom) ?looks like
  real science
• No need for exotic forms of matter (WIMPs, etc.)
• Cons
• You need to modify physical laws you may need to
  abandon equivalence principle and Lorentz
  invariance (at least on certain scales)
• How can we extend it to microscopic scales where
  a gtgt a0? Stars are made of many particles that
  certainly experience high accelerations. It
  seems like we will analyze the dynamics
  differently if we start with different scales.
• Still need dark energy (may actually need more to
  fit WMAP)
• All bodies become bound to every other body
  (potential goes as log)
• Clear trouble with clusters and Lyman-? forest
  (MOND should work here, but it predicts clumps of
  gas in the IGM to be too dense and predicts an
  incorrect temperature profile of hot gas in
  clusters)
• No definitive theoretical framework in which to
  answer some straight-forward questions One can
  always just add another parameter or dependence
  on mass, v, etc. Where does it end?

13
MONDers Response

• CDM also has some serious problems explaining
  observations (its also just as ad hoc!)
• A few setbacks cant undermine the whole theory
• People are trying (without success so far) to
  formulate a relativistic theory and connect it to
  a larger framework. The two can be reconciled in
  some way, but there has been no complete success.
• Stacy McGaugh (jokingly) responds to the Scott,
  et al. paper
• There have been plenty of times in human
  history when people thought they had a pretty
  good handle on cosmology. Are we really so much
  better now? Or are we just priests of a cold,
  dark religion?

My response You do not know the power of the
Dark Side.
14
The Size and Shape of the Universe
15
Quick Lesson in Topology

• Its possible (in a mathematical sense) to have a
  flat (2-D) space that is topologically equivalent
  to a torus, for example. (you can travel in two
  directions and end up exactly where you started)
• The topology defines a size the distance you
  need to travel to end up in the same place -
  anything larger makes no sense

3-torus

• CMB implications youll find a deficit on large
  scales because universe is too small to support
  large wavelengths

16
The result dodecahedron (a.k.a. soccer ball)

• Due to Luminet, Weeks, Riazuelo, Lehoucq, Uzan
• Matches better than WMAP at l 2, 3, and 4
  (longest wavelengths)
• Reproduces spectrum at smaller scales because the
  smaller wavelengths are unaffected
• Dodecahedral topology offers no free parameters
  if the edges are to fit together properly
  (contrast this to the cube which has 6 free
  parameters defining the lengths and angles of the
  sides)
• Globally homogeneous no preferred observers

White soccer-ball model Black WMAP data Gray
Usual infinite universe model
17
WMAP lays the smack down

• Paper by WMAP crew says that the anomalously low
  power is not troubling
• (de Oliveira-Costa, Tegmark, Zaldarriaga,
  Hamilton, 2004, PRD, 69, 063516)
• Small chance that cosmic variance and galactic
  orientation conspire to give us exactly what we
  see (about 1 in 24,000)
• This is small a priori but not if one considers
  the huge number of ways we could have looked for
  anomalous things
• Chances are, were going to see something
  anomalous
• If our universe had to pick one anomalous
  microstate, the probability of getting that
  particular state is small, but the significance
  is diminished because we had to pick something
• Another WMAP data-crunching team looked for
  telltale patterns and ruled the dodecahedral
  universe out
• (Cornish, Spergel, Starkman, Komatsu, 2004, PRL,
  92, 201302)
• One should find rings (of correlated temperature
  anisotropies) where the last scattering sphere
  bulges out from the soccerball and intersects
  itself. They saw no such rings.
• Also find that the smallest size of the universe
  consistent with WMAP is 24 Gpc (given the
  topologies they consider)
• More extensive tests of a wide range of other
  topologies are in preparation

18
The End!