Since 2002 Perimeter Institute has been recording seminars, conference talks, and public outreach events using video cameras installed in our lecture theatres. Perimeter now has 7 formal presentation spaces for its many scientific conferences, seminars, workshops and educational outreach activities, all with advanced audio-visual technical capabilities. Recordings of events in these areas are all available On-Demand from this Video Library and on Perimeter Institute Recorded Seminar Archive (PIRSA). PIRSA is a permanent, free, searchable, and citable archive of recorded seminars from relevant bodies in physics. This resource has been partially modelled after Cornell University's arXiv.org.
The dynamics of black-hole binaries is a very complex problem which has been solved only very recently through time-expensive numerical-relativity calculations. In spite of this mathematical complexity many results of these calculations can be accurately reproduced with phenomenological approaches based on test particles combined with Post-Newtonian theory and black-hole perturbation theory. In this talk I will focus on effective-one-body models, which have proved a useful and fast tool to accurately reproduce numerical-relativity waveforms.
The quest for gravitational waves from binary inspiral is performed via matched filtering and thus requires a detailed knowledge of the signal. For non-precessing binaries complete analytic waveforms exist from inspiral to merger and ring-down. Here we present complete waveforms for generically spinning equal mass systems.They have been constructed by bridging the gap between the analytically known inspiral phase described by spin Taylor (T4) approximants in the restricted waveform approximation and the ring-down phase.
Recently generated asymptotic expansions zanolin et al. arXiv:0912.0065 [gr-qc] showa frequentist approach to go beyond Fisher information assessments of the accuracy for maximum likelihood parameter estimations. In this talk we describe the application of these techniques to directional reconstruction fornumerical relativity waveforms.
Most searches with ground-based detectors for gravitational-wave signals from the inspirals of stellar-mass compact binaries use template based methods. Those work well for non-spinning systems but since the dimensionality of the parameter space of spinning waveforms is large a template bank search is not feasible. We describe Bayesian and Markov-chain Monte-Carlo methods for parameter estimation of spinning waveforms using hybrid spinning waveforms matching the ringdown from Numerical Relativity results. We compare those results when using post-Newtonian only waveforms.
Black hole-neutron star binary (BHNS) mergers are likely sources for detectable gravitational radiation and candidate engines for short-hard gamma-ray bursts. However, accurate modeling of these mergers requires fully general relativistic simulations, incorporating both relativistic hydrodynamics for the matter and Einstein's field equations for the (strong) gravitational fields. I will review techniques and results from recent fully general relativistic BHNS merger simulations.
The familiar post-Newtonian inspiral description of a binary neutron star system is sufficient for detection in current instruments. However, as we consider making astrophysical measurements using advanced detectors, the effects of matter and strong gravity on gravitational wave signals may become significant. I will review recent work modelling the waveforms produced by the inspiral and coalescence of binary neutron stars. In the mid-to-late inspiral this includes modifications to the post-Newtonian waveform models from tidal deformations.
A coherent multi-site search is expected to be more powerful than itscoincident counterpart in discriminating gravitational wave (GW) signals fromthe noise background. This is because the former tests the consistency of thesignals' amplitudes phases and time-delays across the sites with those expected from a real GW source. However the coherent statistic that is optimalin Gaussian noise is not guaranteed to perform as well in real data which arenon-Gaussian.
Check back for details on the next lecture in Perimeter's Public Lectures Series