arXiv subjects

Erik Katsavounidis

Publications and source records attributed to Erik Katsavounidis.

43 records · Page 3Linked to original sources

arXiv2017

Parameter estimation for gravitational-wave bursts with the BayesWave pipeline

We provide a comprehensive multi-aspect study on the performance of a pipeline used by the LIGO-Virgo Collaboration for estimating parameters of gravitational-wave bursts. We add simulated signals with four different morphologies (sine-Gaussians, Gaussians, white-noise bursts, and binary black hole signals) to simulated noise samples representing noise of the two Advanced LIGO detectors during their first observing run. We recover them with the BayesWave (BW) pipeline to study its accuracy in sky localization, waveform reconstruction, and estimation of model-independent waveform parameters. BW localizes sources with a level of accuracy comparable for all four morphologies, with the median separation of actual and estimated sky locations ranging from 25.1$^{\circ}$ to 30.3$^{\circ}$. This is a reasonable accuracy in the two-detector case, and is comparable to accuracies of other localization methods studied previously. As BW reconstructs generic transient signals with sine-Gaussian wavelets, it is unsurprising that BW performs the best in reconstructing sine-Gaussian and Gaussian waveforms. BW's accuracy in waveform reconstruction increases steeply with network signal-to-noise ratio (SNR$_{\rm net}$), reaching a $85\%$ and $95\%$ match between the reconstructed and actual waveform below SNR$_{\rm net} \approx 20$ and SNR$_{\rm net} \approx 50$, respectively, for all morphologies. BW's accuracy in estimating central moments of waveforms is only limited by statistical errors in the frequency domain, and is affected by systematic errors too in the time domain as BW cannot reconstruct low-amplitude parts of signals overwhelmed by noise. The figures of merit we introduce can be used in future characterizations of parameter estimation pipelines.

astro-ph.HE↗

arXiv2017

An information-theoretic approach to the gravitational-wave burst detection problem

The observational era of gravitational-wave astronomy began in the Fall of 2015 with the detection of GW150914. One potential type of detectable gravitational wave is short-duration gravitational-wave bursts, whose waveforms can be difficult to predict. We present the framework for a new detection algorithm for such burst events -- \textit{oLIB} -- that can be used in low-latency to identify gravitational-wave transients independently of other search algorithms. This algorithm consists of 1) an excess-power event generator based on the Q-transform -- \textit{Omicron} --, 2) coincidence of these events across a detector network, and 3) an analysis of the coincident events using a Markov chain Monte Carlo Bayesian evidence calculator -- \textit{LALInferenceBurst}. These steps compress the full data streams into a set of Bayes factors for each event; through this process, we use elements from information theory to minimize the amount of information regarding the signal-versus-noise hypothesis that is lost. We optimally extract this information using a likelihood-ratio test to estimate a detection significance for each event. Using representative archival LIGO data, we show that the algorithm can detect gravitational-wave burst events of astrophysical strength in realistic instrumental noise across different burst waveform morphologies. We also demonstrate that the combination of Bayes factors by means of a likelihood-ratio test can improve the detection efficiency of a gravitational-wave burst search. Finally, we show that oLIB's performance is robust against the choice of gravitational-wave populations used to model the likelihood-ratio test likelihoods.

gr-qc↗

arXiv2016

On similarity of binary black hole gravitational-wave skymaps: to observe or to wait?

Localization estimates for GW150914, the first binary black hole detected by the LIGO instruments, were shared with partner facilities for electromagnetic follow-up. While the source was a compact binary coalescence (CBC), it was first identified by algorithms that search for unmodeled signals, which produced the skymaps that directed electromagnetic observations. Later on, CBC specific algorithms produced refined versions, which showed significant differences. In this paper we show that those differences were not accidental and that CBC and unmodeled skymaps for binary black holes will frequently be different; we thus provide a way to determine whether to observe electromagnetically as promptly as possible (following a gravitational-wave detection), or to wait until CBC skymaps become available, should they not be available in low latency. We also show that, unsurprisingly, CBC algorithms can yield much smaller searched areas.

gr-qc↗

arXiv2016

Observational Selection Effects with Ground-based Gravitational Wave Detectors

Ground-based interferometers are not perfectly all-sky instruments, and it is important to account for their behavior when considering the distribution of detected events. In particular, the LIGO detectors are most sensitive to sources above North America and the Indian Ocean and, as the Earth rotates, the sensitive regions are swept across the sky. However, because the detectors do not acquire data uniformly over time, there is a net bias on detectable sources' right ascensions. Both LIGO detectors preferentially collect data during their local night; it is more than twice as likely to be local midnight than noon when both detectors are operating. We discuss these selection effects and how they impact LIGO's observations and electromagnetic follow-up. Beyond galactic foregrounds associated with seasonal variations, we find that equatorial observatories can access over $80\%$ of the localization probability, while mid-latitudes will access closer to $70\%$. Facilities located near the two LIGO sites can observe sources closer to their zenith than their analogs in the South, but the average observation will still be no closer than $44^\circ$ from zenith. We also find that observatories in Africa or the South Atlantic will wait systematically longer before they can begin observing compared to the rest of the world, although there is a preference for longitudes near the LIGOs. These effects, along with knowledge of the LIGO antenna pattern, can inform electromagnetic follow-up activities and optimization, including the possibility of directing observations even before gravitational-wave events occur.

astro-ph.HE↗

arXiv2014

Localization of short duration gravitational-wave transients with the early advanced LIGO and Virgo detectors

The Laser Interferometer Gravitational wave Observatory (LIGO) and Virgo, advanced ground-based gravitational-wave detectors, will begin collecting science data in 2015. With first detections expected to follow, it is important to quantify how well generic gravitational-wave transients can be localized on the sky. This is crucial for correctly identifying electromagnetic counterparts as well as understanding gravitational-wave physics and source populations. We present a study of sky localization capabilities for two search and parameter estimation algorithms: \emph{coherent WaveBurst}, a constrained likelihood algorithm operating in close to real-time, and \emph{LALInferenceBurst}, a Markov chain Monte Carlo parameter estimation algorithm developed to recover generic transient signals with latency of a few hours. Furthermore, we focus on the first few years of the advanced detector era, when we expect to only have two (2015) and later three (2016) operational detectors, all below design sensitivity. These detector configurations can produce significantly different sky localizations, which we quantify in detail. We observe a clear improvement in localization of the average detected signal when progressing from two-detector to three-detector networks, as expected. Although localization depends on the waveform morphology, approximately 50% of detected signals would be imaged after observing 100-200 deg$^2$ in 2015 and 60-110 deg$^2$ in 2016, although knowledge of the waveform can reduce this to as little as 22 deg$^2$. This is the first comprehensive study on sky localization capabilities for generic transients of the early network of advanced LIGO and Virgo detectors, including the early LIGO-only two-detector configuration.

astro-ph.HE↗

arXiv2013

Optimizing Vetoes for Gravitational-wave Transient Searches

Interferometric gravitational-wave detectors like LIGO, GEO600 and Virgo record a surplus of information above and beyond possible gravitational-wave events. These auxiliary channels capture information about the state of the detector and its surroundings which can be used to infer potential terrestrial noise sources of some gravitational-wave-like events. We present an algorithm addressing the ordering (or equivalently optimizing) of such information from auxiliary systems in gravitational-wave detectors to establish veto conditions in searches for gravitational-wave transients. The procedure was used to identify vetoes for searches for unmodelled transients by the LIGO and Virgo collaborations during their science runs from 2005 through 2007. In this work we present the details of the algorithm; we also use a limited amount of data from LIGO's past runs in order to examine the method, compare it with other methods, and identify its potential to characterize the instruments themselves. We examine the dependence of Receiver Operating Characteristic curves on the various parameters of the veto method and the implementation on real data. We find that the method robustly determines important auxiliary channels, ordering them by the apparent strength of their correlations to the gravitational-wave channel. This list can substantially reduce the background of noise events in the gravitational-wave data. In this way it can identify the source of glitches in the detector as well as assist in establishing confidence in the detection of gravitaional-wave transients.

astro-ph.IM↗

arXiv2004

Plans for the LIGO-TAMA Joint Search for Gravitational Wave Bursts

We describe the plans for a joint search for unmodelled gravitational wave bursts being carried out by the LIGO and TAMA collaborations using data collected during February-April 2003. We take a conservative approach to detection, requiring candidate gravitational wave bursts to be seen in coincidence by all four interferometers. We focus on some of the complications of performing this coincidence analysis, in particular the effects of the different alignments and noise spectra of the interferometers.

gr-qc↗