Download presentation
Presentation is loading. Please wait.
Published byMelinda Ward Modified over 9 years ago
1
Statistical biases in extragalactic distance determinations. G. Paturel, Observatoire de Lyon In collaboration with P. Teerikorpi IHP 28-29 Avril 2005
2
The difficult problem of Ho
3
When does a bias appear ? 1) When the absolute magnitude is expressed as a function of an observable parameter with a scatter Examples : Tully-Fisher Relation Period-Luminosity Relation 2) When the sample has a limiting apparent magnitude m lim
4
Let us explain in a simple case The « sosie » method (e.g.for TF relation) We select galaxies with the same logV M Through they should constitute « Standard Candles » with the same absolute magnitude. Are standard candles free of bias ? No !
5
Graphical explanation of two kinds of bias (1) Classical Malmquist bias
6
Graphical explanation of two kinds of bias (2) Incompleteness bias (at a constant distance modulus)
7
How Cepheid Period Luminosity relation looks like ? The Cepheid PL relation fill the conditions to have a bias at a constant distance
8
Does the bias affect the Cepheid Period-Luminosity relation ? The bias should exist but it can be small due to the small scatter of the PL relation. How to test the existence of a possible bias in the PL relation ?
9
A simple simulation shows that a bias can exist Is it possible to use redshift as a relative distance indicator ? YES
10
The Hubble law: V= H.r The original discovery: V < 1200 km/s
11
Small dispersion around the regular Hubble flow In 1957 de Vaucouleurs noted that deviations from Hubble law are small ( <100 km/s) In 1972 Sandage and coworkers (ApJ 172, 253) found still maller value ( <60 km/s) In 1999 Ekholm et al. confirm that the Hubble law works at small scale In 2001 Ekholm et al. and Karachentsev et al., independently found a still smaller dispersion ( <40 km/s) of the very local expansion
12
Comparison with models
13
We have a tool to have accurate, relative distances
14
Hubble diagram with Cepheid distances
15
Use of the 2-parameter bias model by Teerikorpi (1975) to check a bias diagram
16
A bias diagram When the absolute limiting Magnitude allows us to see faint Cepheids the bias is Negligible. The fit of the bias model Leads to : logH=1.76 ; H=56 (km/s)/Mpc
17
Influence of the correlation of errors The fit of the bias model Leads to : logH=1.80 ; H=63 (km/s)/Mpc The reality could be H=60 (km/s)/Mpc
18
Comparison of corrected and uncorrected distances, using HST anf ground-based Cepheid distances Ground-based HST
19
How the Hubble diagram changes
20
What about long range distances
21
The final question by A.Blanchard : Are the SN- standard candle affected in a similar way The possibility exists because the reference sample is not similar to the distant one (mixing several luminosities) Another effect could be due to evolution effect
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.