5.8 Appendix
5.8.4 Comparison with the XMM-LSS survey
The first part of the XMM-LSS survey (the initial 5 deg2, Pierre et al. 2007; Pacaud et al. 2006, 2007) offers an excellent match to our survey not only with respect to the area, but also to the typical depth (having only slightly higher average exposure times). Since the XMM-LSS
characterization pipeline, we make here an effort to compare results derived from our XMM-BCS pipeline with their published results.
Cluster detection comparison
A full comparison of the source detection pipelines would be only of limited use and is currently impossible since only a small part of the XMM-LSS extended sources have also been spectro- scopically confirmed up to now (the so-called C1 sample of Pacaud et al. (2007)13). Therefore, we restrict ourselves to the reanalysis of the C1 sample.
We downloaded all the XMM-LSS fields with C1 detections14 and fully reanalyzed them with the XMM-BCS pipeline. We confidently detected all the C1 clusters and they are among our highest ranked extended source detections.
In Fig. 5.16 we compare their detection and extent likelihoods with their respective XMM- LSS variants (SB Detect Likelihood and SB Extent Likelihood). Both sets of paramet- ers exhibit a strong correlation, showing good consistency between both detection approaches (XMM-LSS uses a single band wavelet detection scheme). The scatter between the parameters is caused by small differences in the data reduction process, background estimation and source detection algorithms.
The C1 sample is defined bySB Detect Likelihood> 32,SB Extent Likelihood> 33. We fit a linear relation in the two log-log planes and use these cuts to convert the XMM-LSS thresholds to our parameters obtaining: det ml> 16.4 (equivalent to ∼ 5.4σ detection in our scheme) andext ml> 8.3 (i.e. ∼3.7σextent significance).
X-ray photometry comparison
In Fig. 5.17 we compare the fluxes in the 0.5−2 keV band and 0.5 Mpc aperture measured by the XMM-LSS and by us using the growth curve method (Sect. 5.3.2). Being interested only in the flux estimation we have fixed the redshift and temperature to their spectroscopic values provided by XMM-LSS. Both methods give fluxes that are in good agreement and no significant bias is found.
We have checked the dependence of flux residuals defined here as (fX MM−LS S
X − f
X MM−BCS
X ) /
fX MMX −BCS on several parameters: the flux itself, cluster redshift, off-axis angle, fraction of missing pixels (due to chip gaps etc.), background correction factors and amount of extrapolation. We did not find any systematic effects in either PN or MOS fluxes.
This agreement is encouraging, if we take into account that the two pipelines utilize prin- cipally different approaches to the flux measurement. XMM-LSS utilizes a beta model fit to the cluster’s surface brightness integrated out to a fiducial radius, while our method is completely non-parametric (except for a typically small extrapolation factor if the required aperture is larger
13Catalog available at :heasarc.gsfc.nasa.gov/W3Browse/all/xmmlssoid.html
14XMM OBSIDs: 0037980301, 0037980701, 0037981001, 0037981101, 0037981201, 0037981501,
0037981601, 0037981801, 0037982501, 0037982601, 0109520201, 0109520301, 0109520601, 0111110301, 0111110401, 0112680101, 0112680201, 0112680301, 0112680401, 0112680501, 0147110101, 0147110201.
5.8. Appendix 109
than the range where the cluster emission is detected directly). Background estimation in both approaches is also markedly different.
The fluxes do not agree within the error bars for the brightest cluster in this sample (XLSS- J022145.2-034617), with our flux being by ∼ 15% higher. The brightest outlier in the other direction (i.e. our flux lower than the one from XMM-LSS) is XLSS-J022609.9-045805. In this case we found excessive contamination from point sources in the X-ray photometry aperture. We carefully checked and manually adjusted the automatic point source removal, which led to a net decrease of measured flux.
Interestingly, if we decide to rely only on a temperature derived from the L−T the flux estim-
ation precision is practically unchanged (the average difference is only 3%). The temperatures are also in good agreement, although the error bars are large. The mean temperature residuals are
< 1% with a standard deviation of∼ 23%, comparable to measurement errors. This shows that the L−T scaling relation and its evolution adopted in this work from Pratt et al. (2009) is suitable
for cluster samples drawn from surveys of this type. We do not find any systematic dependence of the temperature residuals on redshift, flux or flux residuals.
The cluster mass is not a direct observable in either of the two surveys. XMM-LSS gives rough estimates based on their spectroscopic measurement and beta model fit using the relation from Ettori (2000). Our estimates, using the L−M relation of Pratt et al. (2009), give on average
almost 40% higher masses. The mass residuals strongly amplify the temperature residuals where a unit increment of temperature residual increases the temperature more than a unit decrement of temperature residual would decrease it. This leads to a net increase of mass with respect to the XMM-LSS value.
Finally, we also check the consistency of the beta model fits between the two pipelines. Since the core radius rcoreand theβexponent of the beta model are strongly degenerate, especially for
the case of low counts profiles, our fitting procedure keepsβfixed to the canonical value of 2/3. The XMM-LSS pipeline carries out fits with both the rcore andβas free parameters. Despite this
Figure 5.16: Comparison of detection (left panel) and extent likelihoods (right panel) between our pipeline (x-axis) and the XMM-LSS pipeline Pacaud et al. (y-axis 2007). The derived likeli- hoods are well correlated and the red line shows the best fit relations.
Figure 5.17: Comparison of measured X-ray fluxes of the C1 subsample of the XMM-LSS survey in the 0.5 − 2 keV band and a 0.5 Mpc aperture (Pacaud et al. 2007, y-axis) and the fluxes measured by our pipeline (x-axis). The red line marks equality. See Sect. 5.8.4 for details of this comparison.
5.8. Appendix 111
Figure 5.18: Beta model core radii for the XMM-LSS C1 sample as estimated by our pipeline (x-axis) and by the XMM-LSS estimates. Red line marks equality. The core radii are typically highly uncertain given the relatively low photon statistics. Despite this the agreement between the two estimates is good. Note that the XMM-LSS values are fitted with the beta value as a free parameter, while we fix its value to 2/3.
Chapter 6
XMM-Newton detection of two clusters of
galaxies with strong SPT
Sunyaev-Zel’dovich effect signatures
R. ˇSuhada, J. Song, H. B¨ohringer, B. A. Benson, J. Mohr, R. Fassbender, A. Finoguenov, D. Pierini, G. W. Pratt, K. Andersson, R. Armstrong and S. Desai
A&A, 514 (2010), 3
Abstract
We report on the discovery of two galaxy clusters, SPT-CL J2332-5358 and SPT-CL J2342-5411, in X-rays. These clusters were also independently detected through their Sunyaev-Zel’dovich effect by the South Pole Telescope, and in the optical band by the Southern Cosmology Sur- vey. They are thus the first clusters detected under survey conditions by all major cluster search approaches. The X-ray detection is made within the frame of the XMM-BCS cluster survey utilizing a novel XMM-Newton mosaic mode of observations. The present study makes the first scientific use of this operation mode. We estimate the X-ray spectroscopic temperature of SPT-CL J2332-5358 (at redshift z = 0.32) to be T = 9.3+3.3
−1.9 keV, implying a high mass,
M500 =8.8±3.8×1014M⊙. For SPT-CL J2342-5411, at z=1.08, the available X-ray data do not
allow us to directly estimate the temperature with good confidence. However, using our measured luminosity and scaling relations we estimate that T=4.5±1.3 keV and M500 =1.9±0.8×1014M⊙.
We find a good agreement between the X-ray masses and those estimated from the Sunyaev- Zel’dovich effect.