SESSIONS VII and VIII: NUMERICAL MODELING

Frederic Bournaud

CEA Saclay, France

Interactions and mergers are a nursery of galactic substructures, ranging from super star clusters to decoupled cores, and central black holes to tidal dwarf galaxies. The formation of these structures is interlocked with the interstellar gas dynamics and star formation in interacting galaxies. I will review important results on the ISM dynamics in both the central regions and the outer parts of galaxy mergers, and the various processes that can lead to structure formation. Depending on their mode and efficiency of star formation, new structures as massive clusters and tidal dwarfs could be long- lived, and become substructures of relaxed early-type galaxies several billion years after the merging event. Putting these results in the context of high- to low-redshift mergers, I will review the expected signatures of the merger driven formation of early-type galaxies. Several of these signatures are likely observed in the stellar structures of early-type galaxies, and I will discuss whether or not they are unambiguous footprints of past mergers.

Takayuki Saitoh

Center for Computational Astrophysics, National Astronomical Observatory of Japan.

We present high-resolution N-body/SPH simulations of interacting disk galaxies. Our simulations resolve the multiphase nature of the ISM, and thus can be used to locate star-forming sites robustly. The largest simulation used a total of 3 X 10⁷ particles to represent two gas-rich progenitor galaxies. During the first encounter, a giant filament of 10⁹ M(sun) induced by shocks is formed at the collision interface of the colliding galaxies, and then a large-scale starburst naturally takes place at the giant filament, resulting in the formation of star clusters. Eleven self-bound star clusters are located in the giant filament after the starburst. These star clusters have masses of about 10^6-7 M(sun). They have no dark matter component. Star formation in individual star clusters is quenched in about 10⁷ yrs via type II supernovae. The formation efficiency of star clusters in the filament, M(stars)/M(filament), is found to be about 0.18 at the final stage of the star cluster formation. Star clusters formed in this manner have similar characteristics to observed metal-rich globular clusters.

Li-Hsin Chien

Institute for Astronomy, University of Hawaii

Interacting and merging galaxies are believed to play an important role in many aspects of galactic evolution. Their violent interactions are thought to be the trigger of starbursts, which lead to formation of young globular clusters.Therefore the ages of these young globular clusters can be interpreted to yield the timing of interaction-triggered events, and thus provide a key to reconstruct the star formation history in merging galaxies.The link between galaxy interaction and star formation is well established, but the triggers of star formation in interacting galaxies are still not understood. To date there are two competing formulas that describe the star formation mechanism--density-dependent and shock-induced rules.Numerical models implementing the two rules predict significantly different star formation histories in merging galaxies.

My dissertation involves the combination of these two distinct areas of astrophysics--- stellar evolution and galactic dynamics--- to investigate star formation history in galaxies at various merging stages. I will briefly describe the star formation models and illustrate the concept of using the ages of clusters to constrain the modeling.The ages of the clusters are derived from spectra that were taken with multi-object spectroscopy on Keck. I will present state of the art dynamical modeling with the prediction of star formation history and other properties for three mergers at different stages- Arp 256, NGC 4676 and NGC 7252. I will also show comparisons between the ages of the clusters and the predicted star formation histories. To conclude, I will address this important link as the key to answer the fundamental question: what is the trigger of star formation in merging galaxies?

Diederik Kruijssen

Astronomical Institute Utrecht, Utrecht University, and Leiden Observatory, Leiden University

We present the first hydrodynamical simulations of (interacting) galaxies in which the evolution of star clusters is included. These SPH simulations include self-gravity, feedback, a multiphase interstellar medium, a stellar N-body component and the analytic star cluster evolution code SPACE. Using detailed subgrid models for the cluster evolution, we can model the stellar content of galaxies over more than ten orders of magnitude. The star clusters in the simulations evolve due to stellar evolution and dynamical dissolution. We account for the coupling between the external tidal field and the rapidity of cluster dissolution. In this talk, I will present the results of our first model runs.

Igor Chilingarian

LERMA - Observatoire de Paris

The GalMer project presents the statistical approach to the studies of galaxy interactions. It comprises (1) a database (http://galmer.obspm.fr/) providing access to thousands of simulations of galaxy mergers and flybys; and (2) a set of value-added tools and services to process and analyse the results of simulations on-the-fly. We followed the evolution of the baryonic (gas and stars) and dark matter components using a Tree-SPH code. The hybrid baryonic particles allow us to trace complete star formation (SFH) and chemical enrichment (CEH) histories. We have developed an efficient algorithm to model the spectrophotometric properties of merging galaxies using pre-computed simple stellar population (SSP) models produced by the PEGASE.HR and PEGASE.2 codes and taking into account: (1) kinematics and (2) SFH and MEH of hybrid and star particles, as well as (3) simple prescriptions for the dust attenuation. We are able to model broadband FUV-to-NIR colours as well as optical spectra at a resolution of R=3000 on-the-fly for the simulations represented by 240000 particles. Our approach provides quantitative improvement over all existing efforts of applying stellar population modelling to the results of TreeSPH simulations presented in the literature. Our synthetic spectra and images can be directly compared with the results of imaging and spectroscopic observations (including 3D spectroscopy). We will provide several scientific applications of our modelling, including the studies of the E+A phenomenon.

Florent Renaud

Institute of Astronomy, Vienna (Austria) & Observatoire de Strasbourg (France)

Physics lectures always refer to the tides as a disruptive effect. However, tides can also be compressive. When the potential of two galaxies overlap, especially during a merger, fully compressive tides can develop and have a strong impact on the dynamics of substructures such as star clusters or tidal dwarf galaxies. Using N-body simulations of a large set of mergers, we noticed the importance of these tidal modes at cluster scale. With a model of the Antennae galaxies, we concluded that the positions and timescales of these tidal modes match the actual distribution of young clusters. Detailed study of the statistics of the compressive tides shows a stunning correlation between this purely gravitational effect and the observed properties of the star clusters. In my talk, I will introduce the concept of compressive tide and show its relevance in the special case of the Antennae galaxies. I will then discuss its implications on several critical points such as the infant mortality, the second generation of stars or the age distribution.

Mark Fardal

University of Massachusetts

M31 shows a Giant Southern Stream (GSS) of stars extending >150 kpc to the south, as well as a variety of other tidal features. I use N-body simulations to investigate the origin of the GSS, comparing against recent data from the SPLASH and PAndAS projects. The model suggests the stream resulted from the disruption of a large satellite galaxy less than 1 Gyr ago, and connects it to other debris structures in M31's halo. The stream pumps a significant fraction of the progenitor's mass into M31's outer reaches, demonstrating the ongoing buildup of galaxy stellar halos. Rotating models give better fits to the observed stream structure than spherical models, giving a further clue to the progenitor's nature. I use Bayesian sampling of the simulations to estimate the progenitor's initial mass, finding that it was until quite recently one of the most massive Local Group galaxies. I then discuss constraints from this and other halo tracers on M31's halo mass and concentration.

John Wallin

George Mason University

The huge data volumes across the sciences requires us to consider new approaches to data analysis and simulation. In this project, we investigate how citizen scientists from GalaxyZoo.org can be tasked to address this data flood by providing the human input to classify simulations of galaxy interactions. The primary science goal in the galaxy collision project is developing numerical models for hundreds of interacting systems. With these simulations, we can learn if these collisions are unique locations in the multi-parameter phase space that defines the input parameters to our simulations. We can also correlated these dynamical parameters with the star formation and nuclear activity of in these system. Further self-consistent, high resolution models will follow from these simple matches to compare simulations directly with the observations. A secondary goal of the galaxy collision project is to create a set of human selected matches between simulations and models that can be used as a training set for machine learning algorithms. Analysis of this data set will allow the creation of a robust fitness function for evolutionary computing algorithms that can be used to model other interacting systems.