Statistical analysis of the X-ray morphology of galaxy clusters

The morphological analysis of galaxy clusters in X-rays allows a
reliable determination of their dynamical state. Substructures on
(sub-)Mpc scale influence the gravitational potential of a cluster
and manifest themselves in the X-ray surface brightness
distribution as secondary peaks or overall irregular shape. They
lead to deviations from the hydrostatic equilibrium and spherical
shape, two assumptions which are widely used in galaxy cluster
studies to derive global astrophysical properties. Analyzing the
X-ray morphology of clusters thus yields valuable information,
provided that the employed substructure measures are well-tested
and well-calibrated. In this work, the X-ray morphology of galaxy
clusters is quantified using three common substructure parameters
(power ratios, center shift and the asymmetry parameter), which are
subsequently employed to study the disturbed cluster fraction as a
function of redshift. To ensure a reliable application of these
substructure parameters on a variety of X-ray images, a detailed
parameter study is conducted. It focuses on the performance and
reliability of the parameters for varying data quality using
simulated and observed X-ray images. In particular, when applying
them to X-ray images with low photon counts such as observations of
distant clusters or survey data, it is important to know the
characteristics of the parameters. Comparing the three substructure
measures, the center shift parameter is most robust against Poisson
noise and allows a reliable determination of the clusters'
dynamical state even for low-count observations. Power ratios,
especially the hexapole P3/P0, and the asymmetry parameter, on the
other hand, are severely affected by noise, which results in
spuriously high substructure signals. Furthermore, this work
presents methods to minimize the noise bias. The results of the
parameter study provide a step forward in the morphological
analysis of high-redshift clusters and are employed in the
framework of this thesis to quantify the evolution of the disturbed
cluster fraction. The sample used for this analysis comprises 78
low-z (z < 0.3) and 51 high-z (0.3 < z < 1.08) galaxy
clusters with varying photon statistics. The low-redshift objects
were observed with the XMM-Newton observatory, contain a high
number of photon counts and are part of several well-known and
representative samples. For z > 0.3, the high-redshift subsets
of the 400d2 and SPT survey catalog are used. These objects were
mainly observed with the Chandra observatory and have low photon
counts. To ensure a fair comparison, which is independent of the
data quality, the photon statistics of the low- and high-redshift
observations are aligned before performing the morphological
analysis. In agreement with the hierarchical structure formation
model, a mild positive evolution with redshift, i.e. a larger
fraction of clusters with disturbed X-ray morphologies at higher
redshift, is found. Owing to the low photon counts and small number
of high-redshift observations, the statistical significance of this
result is low. For two of the three substructure parameters (power
ratios and center shift) the findings are also consistent within
the significance limits with no evolution, but a negative evolution
of the disturbed cluster fraction can be excluded for all
parameters.