Gravitational waves from binary neutron stars systems

The first observation of gravitational waves from a merger of binary neutron stars (BNS) along with measurements of electromagnetic counterpart has led the beginning of multi-messenger gravitational wave astronomy. In this thesis, we investigate various gravitational waveform models. These models are employed for extracting source properties from the gravitational wave signal from the BNS merger. We perform parameter estimation studies in order to deduce the systematics among these models. We employ different injection scenarios to understand the biases that occur due to differences in the physics included in different waveform models. We present the construction of hybrid waveforms and discuss their applications as a full waveform, e.g., for validation of other waveform models and to check the performance of the models by performing mismatch calculations and parameter estimation studies where hybrid waveforms used as a substitute for a real signal. Based on the systematics study, we show a few of the waveform models give biased esti- mates of the parameters for specific injection scenarios. We improve those models and present the results of the improved models. In the context of having an accurate yet fast-to-evaluate waveform model, we review reduced-order-modeling techniques and present its application for the multipolar TEOBResum model. Furthermore, to validate and tune analytical models, and to investigate the last few orbits near the merger and after the merger, numerical simulations are inevitable. We evaluate the performance of an initial data generating code, called new SGRID code for BNS systems. With the upcoming advance detectors, it is highly likely that events with extreme source properties will get observed. Therefore, in this thesis, we show preliminary results for numerical simulations of BNS mergers with high spins. We vary equation-of-states (EOSs) and spins to investigate the effects of spin and EOS on the dynamics and gravitational waves.