Galaxy Distributions

1
Galaxy Distributions

• Analysis of Large-scale Structure Using
  Visualization and Percolation Technique on the
  SDSS Early Data Release Database

Yuk-Yan Lam August 9, 2001
2
General Background

• Sloan Digital Sky Survey
• Astrophysics principles
• Large-scale structure

3
Mapping out the SkySloan Digital Sky Survey

• ¼ of the sky
• gt100 million celestial objects
• EDR database 42,000 galaxies
• Projected galaxy spectra 1 million

Apache Point Observatory, New Mexico
4
Pi in the sky
Total Solid Angle 10,000 square degrees
5
Astrophysical Principles

• Redshift
• Spectral classification
• Galaxy types

6
Cosmological Redshift

• Stretching of light wavelength due to Universe
  expansion
• (received wavelength of light) (emitted
  wavelength) x (1 z)
• At time of light emission, Universe was a factor
  of (1 z) smaller in its linear dimension

7
Galaxies with varying redshifts
Low ? High
8
Fingerprints Spectra
QUASAR SPECTRA

• Spectrogram photograph of a spectrum
• SDSS spectrograph splits light from each source
  into hundreds of spectra, which are photographed
  by digital cameras.

GALAXY SPECTRA
9
Galaxy Types

• Two Important Characteristics
• How much light?
• How is the light divided between the bulge and
  the disk?
• Bulge spherical center
• Disk flat portion

10
The Elliptical Galaxy

• Massive
• Spectra similar to a cool star
• No discernible disk

11
The Spiral Galaxy

• Massive
• Almost flat spectra
• Disks with spiral patterns
• Three classes Sa Sb Sc with increasing
  prominence of disk and arms

12
Other Galaxy Types
Blue Compact Galaxy
Irregular Galaxy
Lenticular Galaxy No arms but a disk
Dwarf Galaxies
13
Large-Scale Structure

• Different Types of Topology Possible
• Percolation Statistics

14
Filamentary Network Topology
Filamentary Galaxies concentrated to
one-dimensional threads forming a kind of
three-dimensional network.
More connected as implied by a filamentary
geometry
15
Cellular Bubble Topology
Cellular Galaxies concentrated in narrow walls
separating large, isolated voids.
16
Swiss Cheese Topology
AKA Sponge
Both kinds of regions are connected, so it is
possible to swim to any point through the holes
or burrow to any point inside.
17
Meatball Topology
Clumpier
18
Phase 1 Visualization

• Two dimensions
• Three dimensions

19
Sample
In redshift space

• Galaxy query on EDR
• 41,946 galaxies with their spectra found

(projected into two dimensions)
20
Position In Relation to Spectrum

• Goal See how a galaxys location in the sky may
  be related to its recorded spectrum
• Plot using the following color code
• RED spectrum lt -0.4
• YELLOW -0.4 lt spectrum lt -0.2
• GREEN -0.2 lt spectrum lt 0.0
• BLUE 0.0 lt spectrum lt 0.3
• LIGHT BLUE 0.3 lt spectrum lt 0.6
• WHITE spectrum gt 0.6

21
(No Transcript)
22
CAVE Components
Stereo Glasses
Stereographic Projectors
Interactive Wand
23
Phase 2 Percolation

• Preview Percolation principles
• Step 1 Initial Conditions
• Step 2 Percolation Occurs
• Step 3 Evolution Statistics

24
Percolation Principles

• The Simple Cubic Lattice
• The On/Off State -gt Filling Factor
• Method of Expansion
• Method of Grouping

25
Two-Dimensional Analogy Square lattice on/off
state
Initial Lattice
On/Off State
26
How to expand?
27
Grouping
Neighbors

• Neighbors
• Groups
• Largest Cluster Statistic (LCS)
• Mergers

Merging Groups
28
Percolation Statistics Related to Large-Scale
Topology Standards

• Faster growth of LCS -gt More connected
  distribution
• Slower growth of voids than Gaussian -gt Bubble
  topology
• Faster growth of clusters than Gaussian -gt
  Network topology
• Slower growth of clusters than Gaussian -gt
  Meatball topology

29
Sample Simulations
Network
Poisson
LCS
Meatball
Radius Size
30
Step 1 Initial Conditions
Size used 350 FF .000468665
In red-shift space
31
Step 2 E-x - p - a - n - d

• Percolation Visualized
• Filling Factor versus Radius Size

32
(No Transcript)
33
(No Transcript)
34
Step 3 Evolution Statistics

• Number of Groups vs. Radius
• LCS vs. Radius

35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
Where now?

• Compare with simulations
• Improve grid accuracy
• Boundary effects
• Incorporate additional surveys
• Visualization of technique
• Comparison with other techniques
• Application to other areas

39
Acknowledgements

• Mark Subbarao, Astronomy and Astrophysics
  Department, University of Chicago
• Joshua Frieman, Astronomy and Astrophysics
  Department, University of Chicago
• Mark Hereld, MCS Division, Argonne National
  Laboratory
• Sloan Digital Sky Survey http//www.sdss.org