Читать книгу Galaxies - Группа авторов - Страница 23

The CVRHS system is only one approach to galaxy classification, but it has several advantages: (1) a high focus on features that are likely intimately connected to dynamics and evolution, such as bars, rings and spirals; (2) correlation with star formation history; and (3) the broadest perspective on galaxy morphology without being too unwieldy. However, in the era of large imaging surveys like the Sloan Digital Sky Survey (SDSS; Gunn et al. 1998, 2006; York et al. 2000), the number of classifiable images of galaxies available is literally in the millions. CVRHS classification by a single individual (for example, Buta 2019) will likely be impractical for samples much larger than 20,000 objects. Another approach to galaxy classification is needed.

Nair and Abraham (2010) provide classifications for 14,034 SDSS galaxies in the redshift range 0.01 < z < 0.1 in a system similar to the CVRHS. In this case, a single professional astronomer classified each galaxy twice to check for consistency. In the studies of Fukugita et al. (2007), Baillard et al. (2011) and Ann et al. (2015), 3–10 professional astronomers classified either the whole or part of samples of 2253, 4458 and 5836 SDSS galaxies, respectively. In each case, the final classification is an average of the multiple astronomers, usually after intercomparing results and taking into account “personal equations” of each classifier. One of the largest applications of the multiclassifier approach was made by Kartaltepe et al. (2016), who had the input of 65 professional astronomers.

In spite of these efforts, the number of galaxies that have been classified by experts is still relatively small. This led some professional astronomers to try “crowd-sourcing” galaxy classification by engaging with the general public and enlisting the help of “citizen scientists”. The first major crowd-sourcing effort is described by Lintott et al. (2008) and is called “Galaxy Zoo 1” (GZ1). In GZ1, a web interface was created to display SDSS color images of several hundreds of thousands of galaxies accompanied by a set of “buttons” for selecting rudimentary elements of morphology, such as whether a galaxy is disk-shaped or not. This was followed by “Galaxy Zoo 2” (Willett et al. 2013), which allowed classifiers to specify more detail, such as bars, rings and spiral arm character. The classifications in GZ2 were combined using a vote scheme, and an attempt was made to correct the classifications for biases due to resolution and distance. GZ2 provided classifications for nearly 300,000 SDSS galaxies based on about 80,000 participants. Willett et al. (2013) provide short abbreviations of these classifications that are compared with the CVRHS classifications of Buta (2019) and Buta et al. (2019) in Table 1.12. In these abbreviations, E galaxies are smooth while S galaxies are disk shaped with structure. The other codes are described in Appendix A of Willett et al. (2013). Four of the galaxies in Table 1.12 are shown in Figure 1.27.