Transcript Lensing by Cosmic Large Scale structure

Introduction to Observational Cosmology
on ftp.mpia-hd.mpg.de /pub/rix/Vatican2003/Lecture22-Vatican2003-v5.ppt
The basic pillars of our cosmological picture
(i.e. we are starting with the answer first)
1. Averaged over sufficiently large scales, the universe is
nearly homogeneous and isotropic (=cosmological principle)
2. The universe, i.e. space itself, is expanding so that the
distance D between any pairs of widely separated points
increases as dD/dt~D (=Hubble law)
3. (?) the universe expanded from a very dense, hot initial
state (=big bang)
4. The expansion of the universe is determined by its
mass/energy content and the laws of General Relativity
5. On “small scales” (<10-100 Mpc) a great deal of structure
has formed through gravitational self-organization.
Vatican 2003 Lecture 20 HWR
Historical Milestones of
Cosmology
• 1920s : Galaxies (=“nebulae”) are distinct entities of stars
like our Milky Way  Universe // Milky Way
Note: Nearly co-identical with Einstein's General Relativity
(1916)
 DAnd  0.5 Mpc
Distances in the Universe
First measure of characteristic distance to other galaxies:
Öpik (1922) measured the rotation speed of the Andromeda
galaxy and determined the distance at which its mass to
light ratio equals the Milky Way's: (M/L)And  (M/L)* in MW
 R  v  / G and R    D
with M ~
obs
And
2
circ
 DAnd  0.5 Mpc
L
lD
Hubble (1929) measured distances to galaxies using
Cepheids (M31, M33) and brightest stars, obtaining similar
distances.
2
And
Vatican 2003 Lecture 20 HWR
Expansion of the Universe
• 1929 and 1931 Hubble
(and Humason)
• Fainter (and hence more distant) galaxies recede from the
Milky Way at a rate proportionate to their (estimated)
distance.
10,000 km/s
1000 km/s
Mean apparent magnitude
• Big problem: what is the proportionality constant, the
distance scale? (Hubble got it wrong by a factor of 8)
Vatican 2003 Lecture 20 HWR
Redshift
•
Observational result:
The measured redshift (often falsely interpreted as
"recession velocity") of an object
z
em  0
(0 = rest wavelength)
0
is proportional to the (independently) measured distance D
of an object (but with some scatter!!)
•
Longstanding problem:
We need the absolute distance measure to get the slope of
the relation v = H0 · D ; the proportionality constant is called
"Hubble constant".
More appropriate qualitative interpretation:
The universe/space has expanded by a factor of (1+z)
between emission and observation of the light.
Vatican 2003 Lecture 20 HWR
Quasars
• In 1963 Maarten Schmidt (+Jesse Greenstein) found that a
radio source was actually a galaxy nucleus at z @ 0.16
that was by far the most distant object at the time
• 1965 Quasars with redshifts z>2 found!
 cosmological look-back!
Vatican 2003 Lecture 20 HWR
Cosmic Microwave Background (CMB)
• “Afterglow” of the
big-bang
• Discovered by Penzias
and Wilson (1965)
 glimpse of the universe
in its infancy
<T> = 2.73°K
Small temperature
fluctuations arising from
early weak fluctuations
in the matter
distribution
Vatican 2003 Lecture 20 HWR
Distance Measurements
• Measuring distances is one of the most
fundamental tasks of astronomy, needed to:
–
–
–
–
Convert angular sizes to physical sizes
Convert apparent flux to absolute luminosity
Determine relative geometry of different ojects
Determine look-back time
• Measuring astronomical distances is difficult,
there are three types of methods:
– Geometric methods
– Standard Candles
– Direct physical estimates (gravitational lensing, CMB)
Vatican 2003 Lecture 20 HWR
Geometric Distance Measurements
• Parallax measurements
i.e. Earth’s motion reflex
• State of the art: Hipparchos (ESA Mission in the 1990s)
Parallax accuracy of 0.9 milli-arcseconds  distances to 1 kpc
Application
Calibrate nearby (~ 100pc)
stars of the appropriate
metallicity
 match to globular
cluster main-sequence
 Distance to globular
clusters (to 5%), which
contain cepheids
Reid 1997
Vatican 2003 Lecture 20 HWR
Standard Candles
Approach:
• Select objects whose intrinsic luminosity can be estimated,
either from physical first principles, from empirical
calibrations of nearby examples or can be inferred from
another distance-independent observable.
• Instrinsic luminosity + apparent flux  distance (modulus)
Examples:
• Cepheids: luminosity estimate from their pulsation period
• Spiral Galaxies: luminosity estimate from their disk rotation
curve
• Type-Ia Supernovae (SNIa): luminosity estimate from their
light-curve shape
Vatican 2003 Lecture 20 HWR
Cepheid Distances
• E.g. HST Key-Project to measure Hubble constant, H0
(Freedman, Kennicutt,Mould, et al.)
lightcurves
• Compare Cepheid brightness in M81 to LMC
and local Milky Way Cepheids  DM81=3.63+-0.34
• This way we can measure distances to
Galaxies with where Supernovae exploded.
Note: for nearby (<50Mpc) galaxies distance and
redshift are correlated with considerable scatter
 Measuring H0 is not easy
Vatican 2003 Lecture 20 HWR
Supernova Type Ia Distances
• SN Ia: white dwarf stars near the Chandrasekhar mass limit
(1.4 Msun), where Carbon and Oxygen burn explosively.
• Most luminous variety of Supernova. Can be seen to z>1!
Perlmutter etal
2002

Vatican 2003 Lecture 20 HWR
SN Ia as Pseudo-Standard Candles
Phillips, Hamuy, Ries, Kirshner and others ~1996
– Intrinsically bright SN Ia decline slower
 SN Ia: H0=67+-5 km/s/Mpc
(Current estimate (all methods): H0=70+-3)
Note:
with correction
- still needs Cepheid calibration
- Galaxy velocities differ from the local
mean by ~300 km/s  systematic uncertainty in H0
Vatican 2003 Lecture 20 HWR
Distant Supernovae
• The distance modulus M-m
to a certain redshift z
depends on the expansion
history, not just the current
expansion rate.
• Type-Ia Supernovae can be
seen to great distances: z>1
 probes of the expansion
history.
• 1998: expansion of the
Universe is accelerating (!?)
•
•
Riess etal 1998, AJ, 116, 1009
Perlmutter et al. 1999, ApJ, 517, 565
Vatican 2003 Lecture 20 HWR
Surveying Galaxies
Goal:
• We want to draw up a comprehensive picture of the galaxy
population in either the nearby, present-day universe, or in
the distant, earlier universe.
• Such a “picture” includes:
– the frequency of galaxies as a function of their
•
•
•
•
•
Luminosity
Structure (Size, bulge-to-disc ratio, etc..)
Dynamical mass
Star formation rate
Color (or spectral energy distribution) of the stellar population
– the correlation of these parameters with each other.
• Tully-Fisher relation, fundamental plane, star formation – color
– The connection of individual galaxy properties with their
“environment” (e.g. cluster or field, etc..)
Ideally: we would like to know “everything” about all galaxies
in a given, large volume ………
Vatican 2003 Lecture 20 HWR
Observables in Galaxy (or Star) Surveys
Observables in large surveys:
– Flux in a given aperture for several bands (e.g. IR, optical X-ray)
– Spectra (for D/<> ~0.5) and redshift (Distance +- 4Mpc)
– Some morphology/structure information.
Observational constraints in galaxy/cosmology surveys
– For compact or unresolved objects surveys are flux-limited
(separate flux limit in each observational band).
– For extended objects surveys are surface brightness limited (i.e.
limited by the contrast to the background)
– For imaging surveys, the observed bands correspond to different
rest-frame wavelengths.
– As galaxies have different intrinsic spectral energy distributions,
some objects will be above the survey limit in one band, but not the
other.
How to deal with these difficulties?
Vatican 2003 Lecture 20 HWR
Some Consequences
• Impracticability of nearby,
volume limited samples:
– Given a flux limit flim, any
object of luminosity L, can be
seen only within an volume
V=dW*(L/(4pflim,))3/2
– In practice, it means one only
needs to know the maximum
volume within which the
object would have in the
survey
From the 2DFRS
Vatican 2003 Lecture 20 HWR
K-Corrections
L(1+z) 

K(z)=-2.5log (1+z)
L 

Vatican 2003 Lecture 20 HWR
Galaxy Surveys:
the present-day luminosity function
•
Luminosity function:
– basic, long-standing statistic
to describe the galaxy
population: described the
space density of galaxies at a
given luminosity (or absolute

 L   L 
 L 
d 
   *
 e
 L*  L*
 L*

 L 


 L* 
 L 
d

 L*
magnitude):
– Often parameterized as a
“Schechter function” (1976),
to be a power-law at the faint
end and an exponential at the
brightest end, with a
characteristic luminosity L*
Vatican 2003 Lecture 20 HWR
The LF from SDSS
(+2MASS and 2DFRS)
Sloan Digital Sky Survey:
Five-color digital map of the
Northern sky (8000 sq.
deg) with
• photometry on
>10,000,000 galaxies
spectra (redshifts) for
nearly 1,000,000 galaxies.
Vatican 2003 Lecture 20 HWR
How the Sloan Digital Sky Survey (SDSS) works
Vatican 2003 Lecture 20 HWR
Vatican 2003 Lecture 20 HWR
www.sdss.org
Vatican 2003 Lecture 20 HWR
SDSS Luminosity Function
(5% of Data)
Blanton et al 2001
 bright
about 1 L* galaxy / 100 Mpc3 ; M* (now)=-20.8 in r-band (Note: by convention H0=100)
Schechter function is a good representation
LF (= abundance distribution) depends on the intrinsic color of the galaxy
Vatican 2003 Lecture 20 HWR
The Galaxy Mass (in Stars) Function
• Bell et al. 2003:
use galaxy colors to
estimate (M/L)stars
and hence Mstars
Early type == concentrated light
Vatican 2003 Lecture 20 HWR
Compare this to the local group
Vatican 2003 Lecture 20 HWR
The local size distribution of galaxies
•
Shen et al 2003 from SDSS
Note: Size Specific Angular Momentum
Vatican 2003 Lecture 20 HWR
Present-day color distribution of galaxies
(Strateva etal 2001, Hogg etal 2002)
Vatican 2003 Lecture 20 HWR
The Population of Galaxies at Earlier Epochs
• Basic look-back approach to galaxy
evolution:
– Select earlier epoch (=redshift
slice) and identify galaxies.
– Measure their properties and
compare to “now”
• Milestones of such surveys:
– Canada-French-Redshift Survey
(CFRS): Lilly, LeFevre etal. 1990’s
– LDSS and Auto-Fib Survey: Ellis,
Broadhurst et al 1990’s
– Hubble Deep Field: Williams et al.,
Madau etal.
• Basic Data:
– Redshifts, luminosities,
SFRs(?),SEDs
Vatican 2003 Lecture 20 HWR
COMBO-17 in practice
30´x30´
Vatican 2003 Lecture 20 HWR
COMBO-17
Technique
Eff.
[%]
Wavelength [nm]
Wolf, etal. 2001, A&A, 681;
Wolf, Meisenheimer, Roeser, 2001, A&A, 660
Vatican 2003 Lecture 20 HWR
Fit
SED and z
simultaneously
Wolf et al 2001
dz~0.015
works well to z~1
Vatican 2003 Lecture 20 HWR
Redshift Histograms in COMBO-17 Fields
>25,000 redshifts
(Wolf et al. 2002)
Fields: 30´x30´each
Existing comparable
surveys:
CFRS,CNOC2,Keck
Incl. GOODS !
<1000 redshifts
(Wm/WL)=(0.3/0.7)
Vatican 2003 Lecture 20 HWR
Recent Evolution of the Galaxy Population
using “COMBO-17” as an example
• Sample: 30,000 galaxies to
mr~24 and to z~1.2
• How did the luminosity
function of galaxies evolve
over the last 8x109 years?
– This depends very much on
their SED-type (~color),
assumed to be non-evolving!!
– Red galaxies used to be
much more rare!
- Luminous, blue galaxies used
to be much more common
Vatican 2003 Lecture 20 HWR
Global averages
What Light from what type of galaxies?
Starbursts
Sbc-S.B.
Sa-Sbc
E-Sa
Vatican 2003 Lecture 20 HWR
COMBO-17: How did galaxy colors
change since z~1 (Bell etal 2003)
.
Vatican 2003 Lecture 20 HWR
The (Red) Color-Magnitude
Relation
• The CMR zero-point
evolution is consistent
with passive aging of
ancient stellar populations.
• But: the total mass in star
in the “red-and-dead” is
now twice as high as it was
at z~1.
Vatican 2003 Lecture 20 HWR
GEMS: Key to “internal structure”
(Galaxy Evolution from Morphology and SEDs)
•
Large HST program: 125+50 orbits PI: H.-W. Rix
to image
“extended-Chandra-Deep-Field-South”
–
–
–
–
10,000 redshifts from COMBO-17
9x9 ACS tiles  150 x HDF
V and z
Limit: mz~27.5
Vatican 2003 Lecture 20 HWR
GEMS 58
1.5% of
total
Vatican 2003 Lecture 20 HWR
COMBO-17 (~0.7”) vs.
HST/ACS
Vatican 2003 Lecture 20 HWR
GEMS: Color vs Morphology
•
What kinds of galaxies do
we select?
Bulge-Dominated
Disk-Dominated
Interacting/Peculiar
Irregulars/Clumpy
.
Vatican 2003 Lecture 20 HWR
Summary
• Hubble constant (=distance scale) now known to 5-10%
• Expansion history can now be mapped with SN Ia to z>1.
• New generation of surveys (e.g. SDSS) now can present the
properties of the present day galaxy population as a function
of e.g. luminosity, size, color, (environment), etc..
• Properties of the galaxy population can now be mapped
(samples of >10,000) to z~1 and beyond:
– disappearance of luminous blue galaxies
– Red galaxies are becoming redder, etc..
Vatican 2003 Lecture 20 HWR