SLIDE 1
What is a photometric redshift?
Photometric redshift estimation is a mapping from the photometric space {Fi} to redshift.
redshift space
Dim = 1
Photometric space
Dim ~ 10
Clustering-based redshift estimation with LSST & DESI Mubdi - - PowerPoint PPT Presentation
Clustering-based redshift estimation with LSST & DESI Mubdi - - PowerPoint PPT Presentation
Clustering-based redshift estimation with LSST & DESI Mubdi Rahman Alex Mendez Brice Mnard Ryan Scranton Johns Hopkins University, Samuel Schmidt Kavli IPMU Tokyo University Vivien Scottez What is a photometric redshift? Photometric
SLIDE 2
SLIDE 3
What is a photometric redshift?
Photometric redshift estimation is a mapping from the photometric space {Fi} to redshift. p(z | {Fi} ) does not apply to a given object but to a class of objects statistically indistinguishable. Color-based photometric redshifts are the same for all galaxies with similar colors (set by photometric errors). Any measured property in {Fi} has an associated noise estimate. A photometric source is not just a point in {Fi} but a region defined by the noise level.
SLIDE 4
redshift space
Dim = 1
Photometric space
Dim ~ 10
f
brightness ra,dec size ellipticity … N~4 colors
What is a photometric redshift?
SLIDE 5
redshift space
Dim = 1
Mapping the photometric space to redshift space
Photometric space
Dim ~ 10
brightness ra,dec size ellipticity … N~4 colors
Photometric Redshifts
SEDs or Training Sets
f
ˆ
Hildebrandt et al.
SLIDE 6
redshift space
Dim = 1
Mapping the photometric space to redshift space
Photometric space
Dim ~ 10
brightness ra,dec size ellipticity … N~4 colors
Photometric Redshifts
SEDs or Training Sets
f
ˆ
SLIDE 7
redshift space
Dim = 1
Mapping the photometric space to redshift space
Clustering Redshifts
Spatial Correlation with Reference Set
f
ˆ
Photometric space
Dim ~ 10
brightness ra,dec size ellipticity … N~4 colors
Photometric Redshifts
SEDs or Training Sets
f
ˆ
< ∂ . ∂ref(ra,dec) > < color . Fref(λ) >
Schneider et al. (2006) Ho et al. (2008) Newman (2008, 2010) Ménard et al. (2013) Schmidt et al. (2013) McQuinn & White (2013)
SLIDE 8
< ∂i . ∂unknown >
?
SLIDE 9
< ∂i . ∂unknown >
?
SLIDE 10
redshift space
Dim = 1
Photometric space
Dim ~ 10
Photometric Redshifts
SEDs or Training Sets
f
ˆ
brightness ra,dec size ellipticity … N~4 colors
< color . Fref(λ) >
Clustering Redshifts
Spatial Correlation with Reference Set
f
ˆ
< ∂ . ∂ref(ra,dec) > Mapping the photometric space to redshift space ~ b(z) x dN/dz
SLIDE 11
SDSS
- ptical
2MASS near infrared WISE infrared CFHT-LS
- ptical
Applications of clustering redshifts
SLIDE 12
SDSS
- ptical
2MASS near infrared WISE infrared CFHT-LS
- ptical
Applications of clustering redshifts
~100 million galaxies at
spectroscopic galaxy sample r < 18 mag 1 million objects
x ∆z
SLIDE 13
sample 1: 0.5 < g-r < 0.6 sample 2: 1.3 < g-i < 1.4 sample 3: 1.2 < g-r < 1.3 ~ 6,300 galaxies ~ 10,000 galaxies ~ 2,500 galaxies
= 1
Photometrically-selected galaxies
Rahman et al. (2015)
SLIDE 14
Rahman, BM et al. (2015)
= 1
Generalization to one million galaxies
SLIDE 15
15
Rahman, BM et al. (2015)
Generalization to one million galaxies
SLIDE 16
Comparison to photometric redshifts
SDSS KD-tree photometric redshifts
sample 2 sample 3
Rahman, BM et al. (2015)
SLIDE 17
SDSS
- ptical
2MASS near infrared WISE infrared CFHT-LS
- ptical
Applications of clustering redshifts
~100 million galaxies at
x ∆z
Entire photometric sample r < 22 mag 100 million objects
SLIDE 18
information used: (u,g,r,i,z) + (ra,dec) u-g g-r r-i i-z color color color color What can I do with 100 million photometric galaxies? color color color color
SLIDE 19
Rahman, Mendez, BM et al. (2015)
arXiv:1512.03057
SLIDE 20
Rahman, Mendez, BM et al. (2015)
arXiv:1512.03057
SLIDE 21
SLIDE 22
Comparison clustering-z vs photo-z
Rahman et al. (2015)
SLIDE 23
2MASS near infrared SDSS
- ptical
WISE infrared CFHT-LS
- ptical
Applications of clustering redshifts 2MASS extended sources K < 14 mag 1.5 million objects J H K
λ [µm]
1.0 1.5 2.0 2.5
SLIDE 24
Observations: 1997-2001, J, H & K bands
Skrutskie et al. (2006)
mean cluster-z
Rahman et al. (2015)
SLIDE 25
2MASS near infrared SDSS
- ptical
WISE infrared CFHT-LS
- ptical
Applications of clustering redshifts 2MASS extended sources K < 14 mag 1.5 million objects 2MASS extended sources point sources K < 14 mag 1.5 million objects J H K
λ [µm]
1.0 1.5 2.0 2.5
SLIDE 26
clustering redshifts for extended & point sources
Rahman et al. (2015)
SLIDE 27
Difficult sources for photometric redshifts
interlopers from ≠ z galaxy/AGN mergers strongly lensed galaxies emission-line driven sources dust-reddened
- bjects
log N
~300,000
SLIDE 28
Clustering redshifts
We have a new tool in hand to characterize the mapping between the photometric space and redshift
redshift Photometric space Photo-z
Cluster-z
summary
color color color color cluster-z