Limits on brane world cosmology

Systematic effects in SN cosmology

(I) Extinction by dust in host galaxy or intergalactic

medium/reddening Ongoing SCP High-

z SN search

(II) Gravitational lensing

The SDSS-II SN-search

Summary

Ariel Goobar

Stockholm University

2

Gravitational leakage into X-dimensionGravitational leakage into X-dimension

• Use SNLS (Astier et al 2005) + Baryon oscillations (Eisenstein et al 2005) to examine 5D extenction of Friedmann eqn suggested by Dvali, Gabadadge,Porrati 2000; Deffayet, Dvali, Gabadadze 2001.

2 8

3c

H GH

r

Fairbairn & AG, 2005

3

• Consider more general modifications to Friedmann eqn (as in Dvali & Turner, 2003)

Fit SNLS data + baryon oscillations AND flat universe

Gravitational leakage into X-dimension Gravitational leakage into X-dimension (2)(2)

22

8

3c

H GH

r

Fairbairn & AG, 2005

equiv

4

SNLS 1-year + BAO prior + flatnessSNLS 1-year + BAO prior + flatness

w=w0 + w1·z

5

• SN brightness evolution• Shape-brightness relation• K-corrections and SN colors

• Non-Type Ia contamination• Malmquist bias

• Host galaxy dust properties• Intergalactic dust• Gravitational lensing• Exotica:axion-photon oscillations, etc

• Instrumental corrections• Absolute calibration• Lightcurve fitting technique/host galaxy subtraction

• …

(Known) systematic effects(Known) systematic effects

Astrophysics of supernovae

Selection effects,contamination

Line of sight effects

Measurement issues

6

Dust/reddening: a real problem!Dust/reddening: a real problem!

• Dust in SN host galaxy (or along line of sight)

• Correction assumes some reddening law, typically Galactic type dust (SCP,High-Z Team) or average fit to any kind of reddening/blueing (SNLS)

• Can only be estimated for individual SNe with

a) accurate multi-wavelength data

b) good knowledge of intrinsic ”color” of SNe

• Extinction probabilty in a given galaxy depends on where the SN explosion happens

B-V colorof low-z SNe

Extinction:MB=RB·E(B-V) with RB~ 2 - 5

Extinction correction dominates measurement error!Exception: Elliptical galaxies (E/S0) have little star-formation & dust.

7

Extinction/reddening correctionsExtinction/reddening corrections

• z-dependence in reported Av ?

• Problems with K-corrections/assumed intrinsic colors in UV part of the SNIa spectrum?

• Changing dust properties ?

• Selection effects?

• Degeneracy in global fit?

• Watch out for priors on AV! Riess et al assume P(Av)~exp(-Av)

• Potential inconsistency for elliptical hosts

SN97ff: assumed extinction- free, E-host

Riess et al 2004 (gold sample)

Uncertainties ?

V

8

””GOLD”: systematics dominated?GOLD”: systematics dominated?

Spergel et al ’06

How to make progress at the highest-z?

9

galaxy type dispersion

•Spiral: Sa/Sb/Scmag•Irregular Scd/Irrmag

•Elliptical: E/S0 =0.16 mag

””Dustfree and decelerated”Dustfree and decelerated”

• 219 HST/ACS Orbits awarded (PI: Perlmutter) in C14 for rolling search for SNe on galaxy clusters 0.9<z<1.4.

• Clusters are rich on elliptical galaxies which (at low-z) only host SNIa (no contamination) and extinction by dust should be minimal.

• Expect ~20 SNe in a ”sharp(er)” Hubble diagram.

Sullivan et al 2003

10

First SN discovered in a cluster First SN discovered in a cluster in this searchin this search

Cluster RCS0221-03 at z = 1.02

Host was cataloged Cluster member.

Spectrum taken for confirmation.

ACS z bandACS I bandNicmos J band

preliminary

11

Intergalactic dustIntergalactic dust

• Large dust grains (weak wavelength dependence) may exist in the IG-medium

• Evolution of dust density: two limiting cases:

1. dust (1+z)3 [Model A]

2. dust (1+z)3 for z<0.5 &

dust(z>0.5)= dust(z=0.5) [Model B]

• SDSS QSO colors (>16000 objects, z<2) <0.1 mag extinction for SN1a at z=1; faintness of SNe cannot be only due to IG-dust

AG,Bergström & Mörtsell, A&A, 2002

Mörtsell & AG, 2003,

Östman & Mörtsell, 2005

Concordance

Model A

ModelB; M=1

Milne

IG Dust cannot explain observed faintness of SNe – but is a serious concern for precision cosmology.

SNLS: |M|<0.025; |w|<0.05

12

Lensing (de)magnification in the Lensing (de)magnification in the GOODS SN survey: a study case GOODS SN survey: a study case

• The photometric redshift catalogue for GOODS used to study the line-of-sight properties of the SNIa in the Riess et al 2004 sample

(see Gunnarsson et al ApJ 2006 and Jönsson et al ApJ 2006)

• Faber-Jackson & Tully Fischer relations used for M/L

• Galaxy halos modelled as truncated SIS or NFW

• Self-consistency loop: mass density in galaxies + unresolved matter=M

13

Magnification probabilityMagnification probability

• We find evidence for magnified and demagnified supernovae (1)

• Uncertainty computed by error progation from:

Finite field size error Redshift and position errors Scatter in FJ&TF relations Survey magnitude limit (incompleteness)

PDF built up by randomizing the contributions above according to their individual uncertainties,

• Estimate of magnification in SN1997ff smaller than in Benitez et al 2002, Riess et al 2004. This is understood, both authors now agree with our result.

z=1.27

z=1.75

14

Lensing PDFs for GOODS SN-sampleLensing PDFs for GOODS SN-sample

We found NO evidence for selection effects due to lensing in the GOODS SN sample.Negligible corrections to ’s & w.Expected lensing bias on SNLS results is also small: |M| ~0.01 in M- plane. Added uncertainty on w0 is w~0.014 for BAO prior (SNOC simulation)

15

SDSS II: intermediate-z SNeSDSS II: intermediate-z SNe

• NEW PROJECT – Since Sep 2005

• Aiming at filling in the ”gap” left by eg SNLS and ESSENCE with >300 well measured, accurately calibrated, multicolor LCs

• Repeat imaging of ~270 sq. deg. Sep-Nov 05-07

16

• 126 spectroscopically confirmed SN Ia (<z>=0.21)

• 13 spectroscopically probable SN Ia

• 6 SN Ib/c (3 hypernovae)

• 10 SN II (4 type IIn)

• 5 AGN

• ~hundreds of other unconfirmed SNe with good light curves (galaxy spectroscopic redshifts measured for ~25 additional Ia candidates)

• TO BE REPEATED IN 06 & 07 WITH EVEN BETTER FOLLOW-UP:

> 300 SNeIa in the DE dominated era!

Results from 2005Results from 2005

Courtesy of Bob Nichol

PreliminaryPreliminaryNo reddening corr.No reddening corr.

17

SummarySummary

• Lots of activities to increase the statistics of low, intermediate and high-z Type Ia supernovae

• Emphasis on high-quality data – control of systematics

• Extinction/reddening corrections remain a source of concern –especially for the highest-z data, maybe not in elliptical hosts

• Gravitational lensing (de)magnification not a problem for high-z SNe

• Concordance model in excellent shape

…so far, seems un-challenged by SN-data.