Numerical modelling of cosmic structure formation

In the current popular paradigm structure formation the Universe, the major matter component in the Universe is a yet undiscovered elementary particle that contributes overall more than 4 times the mass of all the baryons in the Universe. The structures in Universe seen today originates from small amplitude quantum fluctuations imprinted on an otherwise homogeneous universe during an epoch of inflationary expansion shortly after the Big Bang. These fluctuations grow over cosmic time due to gravitational instability with small-scale fluctuations being the first to become non-linear and collapse to form self-gravitating systems -- dark matter halos. Initially tiny, these halos merger to form bigger and bigger halos. The first generation stars, often referred to as ``Population-III'' stars are the first sources of light in our Universe after the glow from the Big Bang faded from the optical band. These stars are the origin of the first heavy elements. As it is still infeasible to observe the first generation stars directly, the properties of the first stars, e.g. what is a typical of mass and what is a mass range for the first generation stars, remain some of the biggest mysteries in Cosmology. Over the past decades, much progress has been achieved by using state of the art numerical simulation to follow the growth of the fluctuations imprinted during the inflationary epoch up to densities approaching that of stars. However the mass of the first stars is still poorly constrained. One of the biggest issues is to model the gas accretion around seeds of the first stars. To solve this problem, one needs to model, in a self-consistent way, the dynamics of the accreting gas, the evolution of the proto-star and the various radiative, chemical and mechanical feedback effects that are likely to be at play in the formation process. I will develop a sophisticated physical model considering all the relevant physics detailed above and integrate it into a simulation code to model gas accretion onto the proto-star in order to better obtain the mass of the first generation stars. Knowing the properties of the first stars, will enable us to study the assembly of Milky Way galaxies through the cosmic time in a self-consistent way--from the Big Bang to present day. This can be done by combining the so called semi-analytic modelling with state of the art N-body simulations. I will use unprecedentedly high resolution N-body simulations performed by the Virgo consortium which will resolve the structures capable of forming the first generations stars. Because the simulation itself follows only dark matter, one needs to model the baryonic physics with the semi-analytic techniques developed in recent years which has been applied successfully to explain a large body of observations. The existing semi-analytic codes will need to be extend to follow the baryon physics from the formation of the first stars to the assembly of the disk and bulge component of our Milky Way. While, all the above studies will be carried out within the current popular model. There exists less popular but nontheless plausible models which could explain some astrophysical phenomena where the standard popular model may not. One such candidate model supposes the dark matter is warm, that is it has a random velocity which is not negligible. As demonstrated by Gao and Theuns in their recent paper published in Science, the first object in a Warm Dark matter Universe is a filament in which gas will collpase to very high density then frgmentate into stars. This star formation scenario is radically different from what happend in CDM, and will potentailly lead to many interesting observational consquence to fausify Warm Dark Matter Model. As an important part of my planned research, I will try to identify more convincing observationl evidences in order to firmly constrain the nature of dark matter.