Internal structure of superclusters of galaxies from pattern recognition
The Large-Scale Structure (LSS) of the Universe is a homogeneous network of
galaxies separated in dense complexes, the superclusters of galaxies, and
almost empty voids. The superclusters are young structures that did not have
time to evolve into dynamically relaxed systems through the age of the
Universe. Internally, they are very irregular, with dense cores, filaments and
peripheral systems of galaxies. We propose a methodology to map the internal
structure of superclusters of galaxies using pattern recognition techniques.
Our approach allows to: i) identify groups and clusters in the LSS distribution
of galaxies; ii) correct for the "fingers of God" projection effect, caused by
the partial knowledge of the third space coordinate; iii) detect filaments of
galaxies and trace their skeletons. In this paper, we present the algorithms,
discuss the optimization of the free parameters and evaluate the results of its
application. With this methodology, we have mapped the internal structure of 42
superclusters in the nearby universe (up to $z=0.15$).