Presentation of the ARCHES Integrated cluster finder made at the ARCHES scientific workshop (Paris, 30.11-2.12 2015)

1. The integrated cluster finder - a part of
the ARCHES project
Alexey Mints, Axel Schwope and ARCHES consortium
{Leibniz-Institut f¨ur Astrophysik Potsdam (AIP)}
November 30, 2015

2. ARCHES Integrated cluster finder
Goal
Search for galaxy clusters and
estimate their parameters (redshift,
sizes) in multi-wavelength
photometric and spectroscopic data,
using X-ray information on the
expected cluster positions.

3. Catalog table
catalogue
Frequency
range
Covered
Area
(deg.2)
Number
of objects
3XMMe
clusters
overlap
Catalogs used
AllWISE MIR 41000 747,634,026 1543
UKIDSS JHK 4000 82,655,526 298
SDSS (DR9) ugriz 14555 932,891,133 959
CFTHLS-Wide photo-z 157 35,651,677 149
CFTHLS-Deep photo-z 5.25 2,293,851 38
ALHAMBRA photo-z 4 441,303 18
Spectroscopic catalogs used
SDSS-BOSS spec - 859,322 -
VIPERS spec - 57,204 -

4. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)

5. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)
Utilize color-redshift relation to estimate redshift, a.k.a.
redMaPPer (Rykoff et al., 2014)

6. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)
Utilize color-redshift relation to estimate redshift, a.k.a.
redMaPPer (Rykoff et al., 2014)
Use spectral observations to calibrate color-redshift
relation

7. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)
Utilize color-redshift relation to estimate redshift, a.k.a.
redMaPPer (Rykoff et al., 2014)
Use spectral observations to calibrate color-redshift
relation
Estimate background and spurious detection probability

8. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)
Utilize color-redshift relation to estimate redshift, a.k.a.
redMaPPer (Rykoff et al., 2014)
Use spectral observations to calibrate color-redshift
relation
Estimate background and spurious detection probability
Inputs: position (X-ray source coordinates)

9. Cluster finder basics
Use optical AND infrared colors ⇒ we need a
cross-match tool (ARCHES Xmatch)
Utilize color-redshift relation to estimate redshift, a.k.a.
redMaPPer (Rykoff et al., 2014)
Use spectral observations to calibrate color-redshift
relation
Estimate background and spurious detection probability
Inputs: position (X-ray source coordinates)
Assumptions: luminosity function, density profile,
color-redshift relation...

10. Color-redshift relation

11. Limitations of SDSS, UKIDSS and WISE

12. Cluster membership probability
λ(z) =
r<R
P(z, r, m, χ2
(z, C))
λ – multiplicity;
z – redshift, m – magnitude, C – colors, r – distance from the
X-ray source;
χ2
– the probability of the galaxy with colors C to have
redshift z (incomplete set of colors can be used);

13. Cluster membership probability
λ(z) =
r<R
P(x = (z, r, m, χ2
(z, C))) =
λ(z)u(x)
λ(z)u(x) + b(x)
u(x) – density profile of the cluster (NFW ⊗ LF);
Background is tabulated as b(z, m, χ2
);
Solved iteratively for λ for each redshift on a pre-defined grid
(from 0.02 to 0.8 with a step of 0.01).

14. Example of λ(z).

15. Spurious detection probability
Based on approximation of the distribution of spurious
detections in λ and rNFW . Confidence = 1 - pspurious

16. Validation
Tests against other cluster catalogs.
Cluster catalogue Number of objects Subset used for testing Reference
z range Used objects Recovered
Wen and Han 1757 0.16-0.8 524 313 (60%) Wen et al. (2011)
Takey et al. 530 0.03-0.7 515 491 (95%) Takey et al. (2013)

17. 3XMMe cluster sample
1543 extended X-ray sources (from 3XMMe), 850 with
SDSS photometry;
Run ICF on these sources: 729 detections;
509 detections after duplicate removal (361 with
spectroscopic redshift);
Select X-ray spectra from XMM archive;
Fit temperature and luminosity for spectra;

18. Redshift distribution

19. Products
Integrated Cluster Finder server
Command-line client for ICFs (Python and bash);
ICF web interface (http://serendib.unistra.fr/icf), Hands-on
session tomorrow;
Integrated Cluster catalog
List of cluster candidates;
List of possible cluster members;
Associations with other cluster catalogs;
Images (SDSS colour + XMM contours);

20. Thank you for the attention!

21. Colours

22. Peak detection

23. photo-z
pν(z) =
1
Σphotoz 2πσ2
photoz
exp −
(z − zphot)2
2σ2
photoz
(1)
σ2
photoz = δz2
phot + 4∆z2
(2)
Σphotoz = erf ∆z
2
σ2
photoz
(3)

24. Weighted radius
The inverse cumulative NFW function
F−1
(t) : F−1
(F(r)) = r.
rNFW = F−1
n
i=1 F(ri )
n
(4)
rNFW ≈ 0.5 if members are distributed perfectly at random.

25. Extra numbers
729 detections in 516 fields 509 detections in 440 fields after
duplicate removal (361 (329) with spectroscopic redshift);

26. 3XMMe cluster cuts
1. Observations with high background, hotspots and
corrupted mosaic mode data were removed;
2. Low exposure (< 5ks) observations were removed;
3. 0 < EP EXTENT < 80 arcseconds. This only considers
detections with real extent that is below the upper limit
of 80 arcsecs imposed in the source detection step within
the standard XMM-Newton pipeline processing.
4. EP EXTENT ERR < 10. Excludes poorly constrained
extent values.
5. The galactic latitude must satisfy the constraint
|bII | > 20.3 degrees
6. EP 9 DET ML > 10. Demands a minimum detection
likelihood value of 10 in band 9 (XID band = 0.5-4.5 keV)
7. SUM FLAG < 2. Excludes manually flagged detections
and also detections with sum flag = 2 – generally
detections that are extended and close to other sources or