4.3. Galaxy Mergers
A major, but still largely under-studied, aspect of galaxy formation is the role of galaxy mergers. The merging of galaxies to form larger systems is the cornerstone of the idea behind the modern galaxy formation model, dark matter, and cosmology (e.g., Cole et al. 2000). Constraining this process observationally is just now being done. The first problem is identifying, confidently, which galaxies at high redshift are undergoing a major merger (defined as a merger mass ratio 3:1 or lower). There are a few methods for finding galaxies which are merging, or which soon will. The traditional method for identifying mergers is to identify galaxies in kinematic pairs that are close (< 20 kpc), and at nearly the same radial velocity (with < 500 km s-1). Identifying pairs at high redshift in this manner is difficult due to the inability to determine radial velocities for complete samples, and it is thus largely applicable only at redshifts z < 1.
A promising method explored in Conselice (2003) is identifying galaxy mergers which are in progress, that is systems that have already merged and are undergoing dynamical relaxation. One way to identify these systems is through their chaotic kinematic structures revealed through integral field spectroscopy or velocity curves (e.g., Erb et al. 2004). Alternatively, and more commonly, is to use the stellar structures of galaxies. In Section 2 the methods and reasoning behind using the asymmetry index to find mergers are explained, and the results of these techniques applied to high redshift galaxies are described in Section 3.4.
Since the modern paradigm for forming galaxies implicitly assumes massive galaxies form by merging, it is important to test this idea. The agreement between Cold Dark Matter based models and the data, shown in Figure 6, is good at high redshift, but fails by a significant amount to reproduce the merger fractions for bright galaxies at lower redshifts. This is likely because massive galaxies are forming earlier than low mass galaxies, which may or may not be an effect of environment - massive galaxies are also more likely to be found in dense areas (Dressler 1980). We can examine this in more detail to determine how the stellar masses of high redshift galaxies are built up through mergers and the star formation induced by this merging. The amount of stellar mass added to a galaxy during an observed major merger, assuming a mass ratio of 1:1, can be calculated from the star forming properties of galaxies and the mass accretion rate from merging (e.g., Papovich et al. 2001; Figure 6). Based on the merger rates calculated in Conselice et al. (2003a), a typical Lyman-break galaxy at z ~ 3 will undergo ~ 5 major mergers, with accompanying star formation, before z ~ 1 and is unlikely to have a major merger at z < 1. The mass added by each of these mergers, plus the likely amount of new stars produced in star formation, is enough to create a > M* (Cole et al. 2001) galaxy by z = 1. This also suggest that the galaxy structure-redshift relationship can be described as a cooling of the galaxy population from an era of rapid mergers that has been steadily declining since z ~ 3.
Minor mergers are harder to constrain, yet are likely a major method for adding material to normal galaxies at z < 1 (e.g., Patton et al. 2002; Bundy et al. 2004). This method, or secular evolution, are the most likely possibilities for driving evolution in the galaxy population at z < 1, when up to 50% of all stellar mass formed.