We present our results of SPH simulations of cluster mergers. We investigate the temperature evolution explicitly considering the relaxation process between ions and electrons. The electron temperature distributions are significantly different than those of ions in the later stages of mergers. We also investigate evolution of non-thermal emission from relativistic electrons accelerated around shock fronts. Hard X-ray is luminous only while signatures of merging events are clearly seen in the thermal intracluster medium. On the other hand, extreme ultraviolet emission is still luminous after the system has relaxed.

We present results of numerical simulations of mergers of spiral galaxies using GADGET (Springel, Yoshida, & White 2001). In three of these simulations one of the progenitor galaxies contained a central supermassive black hole (BH), as well as one simulation which did not contain a BH. The merger remnants were evolved to an age of ∼ 13 Gyr to examine the evolution of the shape of each merger remnant. The results of these simulations were compared to observations of elliptical galaxies, which show that older galaxies appear rounder than younger ones (Ryden, Forbes, & Terlevich 2001).

We found that the simulations in which the BH mass was fixed throughout the evolution influence the shape of their host galaxies on timescales less than 3 Gyr. These simulations show little trend of shape with age beyond this time. In the simulations in which the BH mass increased linearly over the duration of the simulation, there is a significant evolution of the shape of the remnant throughout its lifetime, comparable to the observational trend.

Multi-scale interaction between LMC, the Galactic halo, and the disk is examined by N-body simulations. The Galactic models are constructed most realistically to satisfy available observational constraints. LMC is orbiting in a ellipse with pericentric radius of 50kpc, and the Galactic halo is assumed to be well beyond the orbit. By using hybrid algorithm I have succeeded to follow the evolution with million particles. The orbiting satellite excites density enhancement as a wake, and the wake exerts a tidal force on the disk. Through this multi-scale interaction, an integral warp is excited. Its amplitude is, however, a few 100pc, and impossible to explain the observed Galactic warp.

Type Ia supernovae are very good, but not perfect, standard candles, because their observed brightness shows a little diversity. The origin of this dibersity needs to be understood for the application to cosmology.

In close binary systems, a white dwarf must be rotating faster and faster as it gains angular momentum from the accretion disk. Its rapid rotation affects its final mass and strucure just before a supernova expolosion. Brightness of supernovae can be changed if mass of their progenitors have some diversity.

The abundance in M82 is Si and S rich, and can not be explained by the combination of conventional Type I and II supernova yields. We show that these abundance patterns, as well as the observed hot plasma energy to metal mass ratios can be explained by the hypernova nucleosynthesis. This hypothesis is consistent with the expected age after the star-burst of M82.

The structure of obscuring matter in the environment of active galactic nuclei with associated nuclear starbursts is investigated using 3-D hydrodynamical simulations, in which the multi-phase inhomogeneous interstellar matter and its interaction with the supernovae are consistently followed. A globally stable, torus-like, but highly inhomogeneous and turbulent structure is found. To achieve the high column densities as suggested by observations of some Seyfert 2 galaxies with nuclear starbursts, the viewing angle should be larger than about 70° from the pole-on for a 108M⊙ massive black hole. We propose that nuclear starbursts and the mass of the black hole that determines the geometry of the obscuring torus.

We have developed a simulation code with the techniques which enhance both spatial and time resolution of the PM method for which the spatial resolution is restricted by the spacing of the structured mesh. The adaptive mesh refinement (AMR) technique subdivides the cells which satisfy the refinement criterion recursively. On the other hand, the technique of hierarchical time steps (HTS) changes the time step, from particle to particle, depending on the size of the cell in which particles reside. Our N-body code with these AMR and HTS is fully vectorized including the operations of the mass assignment.

We present a phenomenological model to predict the clustering of dark matter halos on the light-cone. The model is constructed by combining several existing theoretical models. We test our model against the Hubble Volume N-body simulation and examine its validity. A good agreement is found in two-point correlation functions of dark matter halos between our model predictions and measurements from the simulation. The model is quite general and thus can be applied to a wider range of astrophysical objects, such as galaxies and quasars.

We perform an extensive analysis of nonlinear and stochastic biasing of galaxies and dark halos in spatially flat low-density CDM universe using cosmological hydrodynamic simulations. We identify galaxies by linking cold and dense gas particles which satisfy the Jeans criterion. We compare their biasing properties with the predictions of an analytic halo biasing model. Dark halos in our simulations exhibit reasonable agreement with the predictions only on scales larger than 10Mpc/h, and on smaller scales the volume exclusion effect of halos due to their finite size becomes substantial. Interestingly the biasing properties of galaxies are better described by extrapolating the halo biasing model predictions.