Statistical Classification and an Optimized Red-Sequence Technique for the Determination of Galaxy Clusters

This study presents a novel method for characterizing galaxy clusters by integrating statistical classification techniques with an optimized adaptation of the red sequence approach. The proposed algorithm employs Gaussian mixture models to analyze the distribution of three key variables: r magnitude, g–r color index, and redshift z. To enhance cluster discrimination, we incorporate Mahalanobis distance metrics and modify the conventional red sequence technique by adopting the principal eigenvector as the slope of the cluster. A sample of 114 galaxy groups and clusters within the redshift range 0.002<z<0.45 was used to validate the method. Comparative analyses demonstrate that the proposed approach achieves comparable or, in certain cases, superior performance in cluster characterization relative to the standard red sequence technique. These results highlight the algorithm’s potential as a robust tool for the exploratory identification and initial parameter determination of galaxy clusters, particularly in large-scale surveys. The methodology bridges statistical rigor with established astrophysical techniques, offering a promising avenue for advancing cluster detection in observational cosmology.