Fang Peng

Neutron stars, with their strong surface gravity, have interestingly short timescales for the sedimentation of heavy elements. Recent observations of unstable thermonuclear burning (observed as X-ray bursts) on the surfaces of slowly accreting neutron stars (<0.01 of the Eddington rate) motivate us to examine how the sedimentation of CNO isotopes affects the ignition of these bursts.

In this thesis, we present the results of a study of the effect of sedimentation on the ignition of H and He in the envelope of an accreting neutron star. A diffusion code is developed for solving the diffusion and sedimentation of isotopes, which is coupled to an accreting neutron star envelope model for tracing the evolutions of isotopes before the thermonuclear instability occurs. We estimate the burst development using a simple one-zone model with a full reaction network. At accretion rates <0.003 Eddington, the ignition of H is sufficiently unstable that the rise in temperature ignites the triple-alpha reaction and produces a vigorous flash. At higher accretion rates (but still less than 0.01 Eddington), the H ignition, although unstable, does not heat the envelope enough to trigger the triple-alpha reaction, but instead manifests itself as a weak H flash. We propose that sources accreting at these rates will build up a massive He layer and produce a long burst, as seen from SLX 1737-282, SLX 1735-269, and GX 3+1. Intriguingly, even at accretion rates >0.1 Eddington sedimentation can still play a role. Although the H, He and CNO isotopes do not completely separate, the H abundance at the base of the accumulated layer is reduced. This changes the proton-to-seed ratio for the rapid-proton capture process. In the absence of convective mixing, the partial stratification would change the composition of the ashes and might enhance the abundance of 12 C---a necessary fuel for superbursts.

Zhaoming Ma

We modify the public PM code developed by Anatoly Klypin and Jon Holtzman to simulate cosmologies with arbitrary initial power spectrum and equation of state of dark energy. With this tool in hand, we perform the following studies on the matter power spectrum.

With an artificial sharp peak at k ~ 0.2 h Mpc -1 in the initial power spectrum, we find that the position of the peak is not shifted by nonlinear evolution. An upper limit of the shift at the level of 0.02% is achieved by fitting the power spectrum local to the peak using a power law plus a Gaussian. This implies that, for any practical purpose, the baryon acoustic oscillation peaks in the matter power spectrum are not shifted by nonlinear evolution which would otherwise bias the cosmological distance estimation. We also find that the existence of a peak in the linear power spectrum would boost the nonlinear power at all scales evenly. This is contrary to what HKLM scaling relation predicts, but roughly consistent with that of halo model.

We construct two dark energy models with the same linear power spectra today but different linear growth histories. We demonstrate that their nonlinear power spectra differ at the level of the maximum deviation of the corresponding linear power spectra in the past. Similarly, two constructed dark energy models with the same growth histories result in consistent nonlinear power spectra. This is hinting, not a proof, that linear power spectrum together with linear growth history uniquely determine the nonlinear power spectrum. Based on these results, we propose that linear growth history be included in the next generation fitting formulas of the nonlinear power spectrum.

For simple dark energy models parametrized by w 0 and w a , the existing nonlinear power spectrum fitting formulas, which are calibrated for ACDM model, work reasonably well. The corrections needed are at percent level for the power spectrum and 10% level for the derivative of the power spectrum. We find that, for Peacock & Dodds (1996) fitting formula, the corrections to the derivative of the power spectrum are independent of w a but changing with redshift. As a short term solution, a fitting form could be developed for w 0 , w a models based on this fact.

Eun-Joo "Sein" Ahn

The nature of ultra high energy cosmic rays (UHECRs) remains an enigma. UHECR detection rate is increasing with new generation detectors which will speed up the process of understanding these energetic particles. After a review on the field of UHECRs, we focus on air shower characterisation of primaries. We study both common primaries as well as more exotic possibilities. One such case is the TeV black hole (BH) creation which can happen in models of large extra dimensions. High energy neutrinos interacting with air molecules may form these objects in the Earth's atmosphere, and a good way of discriminating them from other backgrounds is through air shower studies. Full scale Monte Carlo simulations of air shower cascades are the best way to predict air shower characteristics. However, these are very computer-time consuming. The first interactions of an air shower is instructive as it gives information on how the shower will develop without the full scale simulation. The first interaction study is a far less time consuming method that is advantageous for testing new models with many parameters. Hadronic models are used to interact particles in simulation softwares, where low energy experimental data are extrapolated up to high energies using various models, such as minijets and pomerons. Here, we study the first interactions of well studied cosmic ray primaries - photon, proton, iron nucleus - are performed with SIBYLL 2.1, a hadronisation Monte Carlo which uses the minijet model. "Templates" from common primaries can then be used to compare with new primaries. The TeV BHs are used as an example of a new primary. Air shower simulations have been carried out, which showed that these neutrino induced BH air shower resembles a hadronic air shower. Using the first interaction templates, we show that BHs indeed resemble protons closely.

Jacqueline Chen

The spatial distribution of satellite galaxies around host galaxies can illuminate the relationship between satellites and dark matter subhalos and aid in developing and testing galaxy formation modes. The projected cross- correlation of bright and faint galaxies offers a promising avenue to putting constraints on the radial distribution of satellite galaxies. Previous efforts to constrain the distribution attempted to eliminate interlopers from the measured projected number density of satellites and found that the distribution is generally consistent with the expected dark matter halo profile of the parent hosts. The measured projected cross-correlation can be used to analyze contributions from satellites and interlopers together, using a halo occupation distribution (HOD) based analytic model for galaxy clustering. Tests on mock catalogs constructed from simulations show promise in this approach. Analysis of Sloan Digital Sky Survey (SDSS) data shows results generally consistent with interloper subtraction methods, although the radial distribution is poorly constrained with the current dataset and larger samples are required.