Sco X-1, the brightest low mass X-ray binary, is likely to be a source for gravitational wave emission. In one mechanism, emission of a gravitational wave arrests the increase in spin frequency due to the accretion torque in a low mass X-ray binary.
Since the gravitational waveform is unknown, a detection method assuming no apriori knowledge of the signal is preferable. In this paper, we propose to search for a gravitational wave from Sco X-1 using a {{it source tracking}} method based on a coherent network analysis. In the method, we combine data from several interferometric gravitational wave detectors taking into account of the direction to Sco X-1, and reconstruct two polarization waveforms at the location of Sco X-1 in the sky as Sco X-1 is moving. The source tracking method opens up the possibility of searching for a wide variety of signals. We perform Monte Carlo simulations and show results for bursts, modeled, short duration periodic sources using a simple excess power and a matched filter method on the reconstructed signals.
Pulsar glitches are a potential source of gravitational waves for current and future interferometric gravitational wave detectors. Some pulsar glitch events were observed by radio and X-ray telescopes during the fifth LIGO science run. It is expected
that glitches from these same pulsars should also be seen in the future. We carried out Monte Carlo simulations to estimate the sensitivity of possible gravitational wave signals associated with a pulsar glitch using a coherent network analysis method. We show the detection efficiency and evaluate the reconstruction accuracy of gravitational waveforms using a matched filter analysis on the estimated gravitational waveforms from the coherent analysis algorithm.
Searches for gravitational wave bursts that are triggered by the observation of astronomical events require a different mode of analysis than all-sky, blind searches. For one, much more prior information is usually available in a triggered search whi
ch can and should be used in the analysis. Second, since the data volume is usually small in a triggered search, it is also possible to use computationally more expensive algorithms for tasks such as data pre-processing that can consume significant computing resources in a high data-volume un-triggered search. From the statistical point of view, the reduction in the parameter space search volume leads to higher sensitivity than an un-triggered search. We describe here a data analysis pipeline for triggered searches, called {tt RIDGE}, and present preliminary results for simulated noise and signals.
