How Standard Are Standard Candles?
Download
Report
Transcript How Standard Are Standard Candles?
How Standard are
Cosmological Standard Candles?
Mathew Smith
and Collaborators
(UCT, ICG, Munich, LCOGT and SDSS-II)
SKA Bursary Conference
02/12/2010
Introducing Type Ia Supernova
One of the (optically) brightest astrophysical
phenomena, so can be seen to large distances
Categorised through their spectral features
A limited understanding of their nature / origin
Standardisable Candles
• Type Ia SNe are observed to be an extremely homogeneous population
• Both spectra and light-curves show little variation
• Using an empirical correction (Phillips 1993), the scatter is reduced
• Peak brightness correlated with decline rate (“stretch”)
• After correction ~ 0.15 mag, or cosmological distances to 7%
• Correction reduces scatter on your Hubble diagram
• Used to infer the apparent acceleration of the Universe, and thus “Dark Energy”
• However, scatter is still seen, and needs to be improved for future surveys
Standardisable Candles
• Type Ia SNe are observed to be an extremely homogeneous population
• Both spectra and light-curves show little variation
• Using an empirical correction (Phillips 1993), the scatter is reduced
• Peak brightness correlated with decline rate (“stretch”)
• After correction ~ 0.15 mag, or cosmological distances to 7%
• Correction reduces scatter on your Hubble diagram
• Used to infer the apparent acceleration of the Universe, and thus “Dark Energy”
• However, scatter is still seen, and needs to be improved for future surveys
In order to reduce SNe scatter we need to understand
their properties.
Most previous studies focus on local or high-z SNe, so
obtain biased samples
Need to obtain a homogeneous and representative
sample, independent of galaxy type
Obtained high-quality light-curves for SNe with 0 < z <
0.5
Spectroscopically confirmed over 500 SNe Ia, with ~1,000
photometric SNe Ia – key for future surveys
Able to constrain cosmological parameters
B. Dilday
Determining Host Galaxy Properties
Sample Information:
• Produce a uniform sample,
independent of observational
“issues” (filter transmissions, etc)
• Host Galaxies of each object
determined from deep stacks
• Use Spectral Energy Distributions to
determine galaxy properties from
magnitude and redshift information
• Sample split in two groups;
‘passive and star-forming’
• Large study of the systematics of
template fitting (another talk!)
z<0.45
357 galaxies, 30% passive
z<0.21
135 galaxies, 26% passive
Do SNe know where they come from? - Sort of
• The stretch – brightness
correction varies as a function
of host galaxy type
• Bright SNe are primarily seen
in star-forming galaxies –
caused by recent SF activity?
• The distribution of extinction /
colour in SNe is not dependent on
the host galaxy type.
• True for two SNe light-curve
fitters
Cosmology – Hubble Diagrams …
Cosmology – Hubble Diagrams …
Cosmology – Galaxy Type
Standard
No Prior
Cosmology – Galaxy Type
Standard
No Prior
What’s causing this?
• There is a clear correlation
between:
- Galaxy type
- Stellar mass
- Star-formation rate
- “Stretch” / Delta
• Residuals from the Hubble diagram
seen for several galaxy properties
(independent of redshift)
• Is there a “higher-order” parameter
governing all of this
- Metallicity??
• Galaxy properties can be used for
cosmological constraints
SALT2 Distances determined using:
What’s causing this?
• There is a clear correlation
between:
- Galaxy type
- Stellar mass
- Star-formation rate
- “Stretch” / Delta
• Residuals from the Hubble diagram
seen for several galaxy properties
(independent of redshift)
• Is there a “higher-order” parameter
governing all of this
- Metallicity??
• Galaxy properties can be used for
cosmological constraints
SALT2 Distances determined using:
Other implications
• The dispersion on the Hubble diagram is smaller for passive galaxies
- Is this telling us that they are better distance estimators
- Or about “intrinsic dispersion”?
- Important for the next generation of surveys
• There are indications that SNe in different environments obey a different colour law.
• An additional parameter (such as host galaxy mass) “improves” the Hubble diagram.
An aside: SNe Ia Rates
The SN Ia rate is dependent on the specific
star-formation rate – the proportion of a galaxy’s
mass that is used to form stars
There is a dependence on host galaxy mass –
that differs for different galaxy types
The star-formation rate drives the SN Ia rate
Currently - The Maraston Models
The Maraston models give us a handle on metallicity
- The possible hidden parameter / correlation?
PEGASE has issues with accurately determining galaxy properties
However, template selection is far more complex – need to select on colour?
Metallicity (tentatively)
HIGHLY PRELIMINARY
4 estimates of metallicity
Not a continuous parameter
An offset seen with Hubble residual?
Different populations of “stretch”
Color distribution ‘uncertain’
Need to consider systematics
Metallicity estimated from the colours
Able to estimate for each galaxy, but
less accurate
Degenerate with age / extinction
The Summary
•
•
•
•
•
•
•
•
•
•
SNe Ia are important cosmological probes – we are now at the stage where we are
seriously considering systematics.
The SDSS-II Supernova Survey is complete – we have over 500 SNe with z<0.5
We have produced a large, homogeneous and representative sample with which to
study the SN Ia population
SNe in different galaxies have different absolute magnitudes (modulo template
fitting issues)
The colour law may be different for different galaxy types
‘passive’ SNe show a lower dispersion from the best-fitting Hubble diagram –
target them for future surveys?
Metallicity could be the hidden parameter
Need to test this with more advanced models and spectral information
Combining SDSS and SNLS we will be able to study this effect with redshift
Host Galaxy information maybe useful for cosmological analyses