[Paper Review] Automated search for galactic star clusters in large multiband surveys: I. Discovery of 15 new open clusters in the Galactic anticenter region

S. E. Koposov, E. V. Glushkova+1 more|ArXiv.org|Sep 10, 2007

Stellar, planetary, and galactic studies24 references50 citations

TL;DR

This paper presents an automated method to detect open star clusters in large multiband surveys using convolution-based density peak detection and isochrone fitting in color-magnitude diagrams. Applied to 2MASS data in the Galactic anticenter, the method discovered 15 new open clusters and homogeneously determined physical parameters for 12 of them, significantly expanding the sample of well-characterized clusters in this region.

ABSTRACT

Aims: According to some estimations, there are as many as 100000 open clusters in the Galaxy, but less than 2000 of them have been discovered, measured, and cataloged. We plan to undertake data mining of multiwavelength surveys to find new star clusters. Methods: We have developed a new method to search automatically for star clusters in very large stellar catalogs, which is based on convolution with density functions. We have applied this method to a subset of the Two Micron All Sky Survey catalog toward the Galactic anticenter. We also developed a method to verify whether detected stellar groups are real star clusters, which tests whether the stars that form the spatial density peak also fall onto a single isochrone in the color-magnitude diagram. By fitting an isochrone to the data, we estimate at the same time the main physical parameters of a cluster: age, distance, color excess. Results: For the present paper, we carried out a detailed analysis of 88 overdensity peaks detected in a field of $16 imes16$ degrees near the Galactic anticenter. From this analysis, 15 overdensities were confirmed to be new open clusters and the physical and structural parameters were determined for 12 of them; 10 of them were previously suspected to be open clusters by Kronberger (2006) and Froebrich (2007). The properties were also determined for 13 yet-unstudied known open clusters, thus almost tripling the sample of open clusters with studied parameters in the anticenter. The parameters determined with this method showed a good agreement with published data for a set of well-known clusters.

Motivation & Objective

• To address the significant underrepresentation of open clusters in current catalogs, with only ~20% of estimated 100,000 clusters identified.
• To overcome challenges in detecting clusters in high-extinction regions like the Galactic plane, where traditional methods fail due to dust obscuration.
• To develop a uniform, automated method for identifying and characterizing open clusters using large photometric surveys.
• To improve the homogeneity and reliability of physical parameters (age, distance, color excess) for both new and known clusters in the anticenter region.

Proposed method

• A convolution-based algorithm detects spatial overdensities in stellar catalogs by convolving star counts with Gaussian kernels of varying widths to identify potential cluster centers.
• The method applies a multi-scale approach to detect clusters of different physical sizes, from compact to diffuse systems.
• Candidate clusters are validated by testing whether their members align along a single isochrone in color-magnitude diagrams (CMDs) using (J, J−H) and (Ks, J−Ks) bands.
• Isochrone fitting is performed to simultaneously derive age, distance, color excess (E(B−V)), and cluster radius with consistent photometric data.
• The method uses Hess diagrams and radial density profiles to further confirm cluster membership and structural properties.
• All known clusters in the field are re-analyzed using the same automated pipeline to ensure homogeneity in derived parameters.

Experimental results

Research questions

• RQ1 Can an automated, convolution-based method reliably detect open clusters in high-extinction regions of the Galactic anticenter using multiband photometry?
• RQ2 To what extent can isochrone fitting in color-magnitude diagrams confirm the physical coherence of detected overdensities as real star clusters?
• RQ3 How many new open clusters can be discovered and characterized in a 16×16 degree field toward the Galactic anticenter using 2MASS data?
• RQ4 How do the derived physical parameters (age, distance, E(B−V)) of newly discovered clusters compare with existing catalog values?
• RQ5 To what extent does applying a uniform, automated method improve the homogeneity and reliability of cluster parameter estimates compared to heterogeneous past studies?

Key findings

• The method detected 88 spatial overdensity peaks in a 16×16 degree field toward the Galactic anticenter, of which 15 were confirmed as new open clusters.
• Twelve of the 15 newly discovered clusters had their physical parameters—age, distance, and color excess—successfully determined via isochrone fitting.
• Thirteen previously unstudied known open clusters in the field were re-analyzed, and their physical parameters were improved or newly derived, tripling the number of well-characterized clusters in the region.
• The method achieved good agreement (within 200–500 pc for distance, 0.10 mag for E(B−V), and 0.05 dex for log(age)) with published data for 11 well-known clusters.
• The study increased the number of open clusters in the region with reliably determined parameters from 11 (in the Dias et al. 2002 catalog) to 35, demonstrating the value of uniform re-analysis.
• Ten of the 15 new clusters were previously suspected as candidates by Kronberger (2006) and Froebrich (2007), confirming the method’s sensitivity and consistency.