The confined nature of the debris from the Sagittarius dwarf to a narrow trail on the sky has recently prompted the suggestion that the dark matter halo of our Galaxy should be nearly spherical (Ibata et al. 2001; Majewski et al. 2003). This would seem to be in strong contrast with predictions from cold dark matter (CDM) simulations, where dark halos are found to have typical density axis ratios of 0.6 to 0.8. Here I present numerical simulations of the evolution of a system like the Sagittarius dSph in a set of Galactic potentials with varying degrees of flattening. These simulations show that the Sagittarius streams discovered so far are too young dynamically to be sensitive to the shape of the dark halo of the Milky Way. The data presently available are entirely consistent with a Galactic dark matter halo that could either be oblate or prolate, with density axis ratios c/a that range from 0.6 to 1.6 within the region of the halo probed by the orbit of the Sagittarius dwarf.

We present a statistical analysis of the interactions between satellite galaxies in cosmological dark matter halos taken from fully self-consistent high-resolution simulations of galaxy clusters. We show that the number distribution of satellite encounters has a tail that extends to as many as three to four encounters per orbit. On average 30% of the substructure population had at least one encounter (per orbit) with another satellite galaxy. However, this result depends on the age of the dark matter host halo with a clear trend for more interactions in younger systems. We also report a correlation between the number of encounters and the distance of the satellites to the centre of the cluster — satellite galaxies closer to the centre experience more interactions. However, this can be simply explained by the radial distribution of the substructure population and merely reflects the fact that the density of satellites is higher in those regions.

In order to find substructure galaxies we applied (and present) a new technique based upon the N-body code MLAPM. This new halo finder MHF (MLAPM’s halo finder) acts with exactly the same accuracy as the N-body code itself and is therefore free of any bias and spurious mismatch between simulation data and halo finding precision related to numerical effects.

Hierarchical clustering represents the favoured paradigm for galaxy formation throughout the Universe; due to its proximity, the Magellanic system offers one of the few opportunities for astrophysicists to decompose the full six-dimensional phase-space history of a satellite in the midst of being cannibalised by its host galaxy. The availability of improved observational data for the Magellanic Stream and parallel advances in computational power has led us to revisit the canonical tidal model describing the disruption of the Small Magellanic Cloud and the consequent formation of the Stream. We suggest improvements to the tidal model in light of these recent advances.

We present a modified TREE-SPH code to model galaxies in three dimensions. The model includes a multi-phase description of the interstellar medium which combines two numerical techniques. A diffuse warm/hot gas phase is modelled by SPH, whereas a cloudy medium is represented by a sticky particle scheme. Interaction processes (such as star formation and feedback), cooling, and mixing by condensation and evaporation, are taken into account. Here we apply our model to the evolution of a Milky Way type galaxy. After an initial stage, a quasi-equilibrium state is reached. It is characterised by a star formation rate of ∼1 M⊙ yr–1. Condensation and evaporation rates are in balance at 0.1–1 M⊙ yr–1.

Ultra-high resolution hydrodynamic simulations using 10243 grid points are performed of a very large supernova burst in a forming galaxy, with properties similar to those inferred for Lyman Break Galaxies (LBGs). Explosions produce kpc-sized expanding hot bubbles enclosed by cool, dense shells, and the engulfed gas is polluted with freshly-synthesised heavy elements. We show that the resultant inhomogeneous mixing produces a large spread ([Fe/H] ≈ –1 to –5) of metallicities, which affects the subsequent galactic chemical evolution and leaves its imprint on metal-poor stars. By combining a spectral synthesis model with the numerical results, we predict Lyα emission from such galaxy at z = 3. We find that the simulated galaxy, whose peak star formation rate is ≈200 M⊙ yr–1, produces a Lyα luminosity Lα = 9.7 × 1042 erg s–1. This value favorably matches the observed one, but some discrepancies are left for the Lyα line width, the metallicity, and X-ray properties. Since the results of the simulation is applicable only at the very early epoch of the galaxy formation, the metallicity is still lower than that of LBGs. However, the analysis presented here demonstrates a way to enable a systematic comparison with observational data.

It is concluded that LBGs are optimal objects to scrutinise the early self-enrichment in forming galaxies. In the future, the predicted bubbly structure carved by SNe may be directly detected by high resolution observations with JWST.

The kinematic properties of tidal debris from an orbiting Galactic satellite is presented, on the assumption that its central part once contained the most massive Galactic globular cluster, ω Cen. We simulate dynamical evolution of a satellite galaxy that follows the present-day and likely past orbits of ω Cen, and analyse the kinematic nature of their tidal debris and randomly generated Galactic stars comprising spheroidal halo and flat disk components. It is found that the debris stars show a retrograde rotation at ∼–100 km s–1, which may accord with a recently discovered stellar stream at a radial velocity of ∼300 km s–1 towards the Galactic longitude of ∼270°. These stars also contribute, only in part, to a reported retrograde motion of the outer halo in the north Galactic pole direction, without significantly modifying local halo kinematics near the Sun. The prospects for future debris searches and the implications for the early evolution of the Galaxy are briefly presented.

We describe the star formation histories of the Milky Way dwarf spheroidal galaxy and the globular cluster ω Centauri in terms of an inhomogeneous chemical evolution model developed for the Galactic halo. The observed abundance trends seen in neutron-capture elements together with α-elements constrain our models to shed light on the histories of these nearby galaxies and ω Cen. The origin of low-α stars and a new picture of the globular cluster formation scenario induced by cloud–cloud collisions are also presented.