The standard noise model in gravitational wave (GW) data analysis assumes detector noise is stationary and Gaussian distributed, with a known power spectral density (PSD) that is usually estimated using clean off-source data. Real GW data often depart from these assumptions, and misspecified parametric models of the PSD could result in misleading inferences. We propose a Bayesian semiparametric approach to improve this. We use a nonparametric Bernstein polynomial prior on the PSD, with weights attained via a Dirichlet process distribution, and update this using the Whittle likelihood. Posterior samples are obtained using a blocked Metropolis-within-Gibbs sampler. We simultaneously estimate the reconstruction parameters of a rotating core collapse supernova GW burst that has been embedded in simulated Advanced LIGO noise. We also discuss an approach to deal with non-stationary data by breaking longer data streams into smaller and locally stationary components.
Using the latest numerical simulations of rotating stellar core collapse, we present a Bayesian framework to extract the physical information encoded in noisy gravitational wave signals. We fit Bayesian principal component regression models with known and unknown signal arrival times to reconstruct gravitational wave signals, and subsequently fit known astrophysical parameters on the posterior means of the principal component coefficients using a linear model. We predict the ratio of rotational kinetic energy to gravitational energy of the inner core at bounce by sampling from the posterior predictive distribution, and find that these predictions are generally very close to the true parameter values, with $90%$ credible intervals $sim 0.04$ and $sim 0.06$ wide for the known and unknown arrival time models respectively. Two supervised machine learning methods are implemented to classify precollapse differential rotation, and we find that these methods discriminate rapidly rotating progenitors particularly well. We also introduce a constrained optimization approach to model selection to find an optimal number of principal components in the signal reconstruction step. Using this approach, we select 14 principal components as the most parsimonious model.
Extreme mass ratio inspirals (EMRIs) are thought to be one of the most exciting gravitational wave sources to be detected with LISA. Due to their complicated nature and weak amplitudes the detection and parameter estimation of such sources is a challenging task. In this paper we present a statistical methodology based on Bayesian inference in which the estimation of parameters is carried out by advanced Markov chain Monte Carlo (MCMC) algorithms such as parallel tempering MCMC. We analysed high and medium mass EMRI systems that fall well inside the low frequency range of LISA. In the context of the Mock LISA Data Challenges, our investigation and results are also the first instance in which a fully Markovian algorithm is applied for EMRI searches. Results show that our algorithm worked well in recovering EMRI signals from different (simulated) LISA data sets having single and multiple EMRI sources and holds great promise for posterior computation under more realistic conditions. The search and estimation methods presented in this paper are general in their nature, and can be applied in any other scenario such as AdLIGO, AdVIRGO and Einstein Telescope with their respective response functions.
We introduce a signal processing model for signals in non-white noise, where the exact noise spectrum is a priori unknown. The model is based on a Students t distribution and constitutes a natural generalization of the widely used normal (Gaussian) model. This way, it allows for uncertainty in the noise spectrum, or more generally is also able to accommodate outliers (heavy-tailed noise) in the data. Examples are given pertaining to data from gravitational wave detectors.
We present a method that assesses the theoretical detection limit of a Bayesian Markov chain Monte Carlo search for a periodic gravitational wave signal emitted by a neutron star. Inverse probability yields an upper limit estimate for the strength when a signal could not be detected in an observed data set. The proposed method is based on Bayesian model comparison that automatically quantifies Occam s Razor. It limits the complexity of a model by favoring the most parsimonious model that explains the data. By comparing the model with a signal from a pulsar to the null model that assumes solely noise, we derive the detection probability and an estimate for the upper limit that a search, for example, for a narrow-band emission for SN1987a, might yield on data at the sensitivity of LIGO data for an observation time of one year.
We present in this paper a Bayesian parameter estimation method for the analysis of interferometric gravitational wave observations of an inspiral of binary compact objects using data recorded simultaneously by a network of several interferometers at different sites. We consider neutron star or black hole inspirals that are modeled to 3.5 post-Newtonian (PN) order in phase and 2.5 PN in amplitude. Inference is facilitated using Markov chain Monte Carlo methods that are adapted in order to efficiently explore the particular parameter space. Examples are shown to illustrate how and what information about the different parameters can be derived from the data. This study uses simulated signals and data with noise characteristics that are assumed to be defined by the LIGO and Virgo detectors operating at their design sensitivities. Nine parameters are estimated, including those associated with the binary system, plus its location on the sky. We explain how this technique will be part of a detection pipeline for binary systems of compact objects with masses up to $20 sunmass$, including cases where the ratio of the individual masses can be extreme.
