To understand the effects of interactions and mergers on cluster galaxies, we start with lessons learned from the study of collisions in the field environment. In Section 3, we will then ask how the cluster environment modifies these results. In this discussion, we focus largely on two aspects of encounters: the triggering of starbursts and nuclear activity, and the late-time evolution of tidal debris and possible reformation of gaseous disks.
The role of galaxy interactions in driving activity and evolution of field spirals has been well documented through a myriad of observational and theoretical studies. Models of interactions have demonstrated the basic dynamical response of galaxies to close encounter (e.g., Toomre & Toomre 1972; Negroponte & White 1983; Barnes 1988, 1992; Noguchi 1988; Barnes & Hernquist 1991, 1996; Mihos & Hernquist 1994, 1996). Close interactions can lead to a strong internal dynamical response in the galaxies, driving the formation of spiral arms and, depending on the structural properties of the disks, strong bar modes. These non-axisymmetric structures lead to compression and inflow of gas in the disks, elevating star formation rates and fueling nuclear starburst/AGN activity. If the encounter is sufficiently close, dynamical friction leads to an eventual merging of the galaxies, at which time violent relaxation destroys the dynamically cold disks and produces a kinematically hot merger remnant with many of the properties found in the field elliptical galaxy population (see, e.g., Barnes & Hernquist 1992).
Observational studies support much of this picture. Interacting systems show preferentially elevated star formation rates, enhanced on average by factors of a few over those of isolated spirals (Larson & Tinsley 1978; Condon et al. 1982; Keel et al. 1985; Kennicutt et al. 1987). Nuclear starbursts are common, with typical starburst mass fractions that involve a few percent of the luminous mass (Kennicutt et al. 1987). More dramatically, infrared-selected samples of galaxies reveal a population of interacting "ultraluminous infrared galaxies," where star formation rates are elevated by 1-2 orders of magnitude and dust-enshrouded nuclear activity is common (Soifer et al. 1984; Lawrence et al. 1989). These systems are preferentially found in late-stage mergers (e.g., Veilleux, Kim, & Sanders 2002) and have been suggested as the precursors of luminous quasars (Sanders et al. 1988). This diversity of properties for interacting systems argues that the response of a galaxy to a close interaction is likely a complicated function of encounter parameters, galaxy type, local environment, and gas fraction.
Figure 2. The morphological response of galaxies to a close encounter. The galaxy models are viewed one rotation period after the initial collision. Top panels show the response to a slow, parabolic encounter, while the bottom panels show the response to fast encounters. The left columns show the response of a pure disk system, the middle panels show a disk/bulge system, and the right panels show a low-density, dark matter dominated disk.
Can we isolate the different determining factors to understand what drives the strong response in interacting systems? Numerical modeling of interactions has shown that gaseous inflow and central activity in an interacting disk is driven largely by gravitational torques acting on the gas - not from the companion galaxy, but by the developing non-axisymmetric structures (spiral arms and/or central bar) in the host disk (Noguchi 1988; Barnes & Hernquist 1991; Mihos & Hernquist 1996). This result argues that the structural properties of galaxies play a central role in determining the response to interactions. In particular, disks that are stable against the strong growth of disk instabilities will experience a weaker response, exhibiting modestly enhanced, disk-wide star formation (unless and until they ultimately merge). This stability can be provided by the presence of a centrally concentrated bulge (Mihos & Hernquist 1996) or a lowered disk surface density (at fixed rotational speed; Mihos, McGaugh, & de Blok 1997). In contrast, disk-dominated systems are more susceptible to global bar modes and experience the strongest levels of inflow and nuclear activity (see Section 3).
Interacting and merging galaxies also show a wide variety of tidal features, from long thin tidal tails to plumes, bridges, and other amorphous tidal debris. The evolution of this material was first elegantly described by the computer models of Toomre & Toomre (1972) and Wright (1972). Gravitational tides during a close encounter lead to the stripping of loosely bound material from the galaxies (see Fig. 3); rather than being completely liberated, the lion's share of this material (> 95%) remains bound, albeit weakly, to its host galaxy (Hernquist & Spergel 1992; Hibbard & Mihos 1995). Material is sorted in the tidal tails by a combination of energy and angular momentum - the outer portions of the tails contain the least bound material with the highest angular momentum. At any given time, material at the base of the tail has achieved turn-around and is falling back toward the remnant. Further out in the tail, material still expands away, resulting in a rapid drop in the luminosity density of the tidal tails due to this differential stretching. As a result, the detectability of these tidal features is a strong function of age and limiting surface brightness; after a few billion years of dynamical evolution, they will be extremely difficult to detect (Mihos 1995).
Figure 3. Evolution of the tidal debris in an equal-mass merger of two disk galaxies occurring in isolation. Each frame is approximately 0.9 Mpc on a side. Note the sharpness of the tidal debris, as well as the loops that form as material falls back into the remnant over long time scales.
In mergers, the gas and stars ejected in the tidal tails fall back onto the remnant in a long-lived "rain" that spans many billions of years (Hernquist & Spergel 1992; Hibbard & Mihos 1995). In the merger simulation shown in Figure 3, this fallback manifests itself as loops of tidal debris that form as stars fall back through the gravitational potential of the remnant. Tidal gas will follow a different evolution, as it shocks and dissipates energy as it falls back. The most tightly bound gaseous material returns to the remnant over short time scales and can resettle into a warped disk (Mihos & Hernquist 1996; Naab & Burkert 2001; Barnes 2002). Such warped H I disks have been observed in NGC 4753 (Steiman-Cameron, Kormendy, & Durisen 1992) and in the nearby merger remnant Centaurus A (Nicholson, Bland-Hawthorn, & Taylor 1992). Over longer time scales, the loosely bound, high-angular momentum gas falls back to ever-increasing radii, forming a more extended but less-organized distribution of gas outside several effective radii in the remnant. Many elliptical galaxies show extended neutral hydrogen gas, sometimes in the form of broken rings at large radius, perhaps arising from long-ago merger events (e.g., van Gorkom & Schiminovich 1997).
The ultimate fate of this infalling material is uncertain, but may have important ramifications for interaction-driven galaxy evolution models. If efficient star formation occurs in this gas, such as that observed in the inner disk of the merger remnant NGC 7252 (Hibbard et al. 1994), this may present a mechanism for building disks in elliptical galaxies. If the amount of gas resettling into the disk is significant, in principle the remnant could evolve to become a spheroidal system with a high bulge-to-disk ratio, perhaps forming an S0 or Sa galaxy (e.g., Schweizer 1998).