Plans for the LIGO TAMA joint search for gravitational wave bursts

Similar documents
arxiv:gr-qc/ v1 4 Dec 2003

Search for inspiralling neutron stars in LIGO S1 data

First upper limits from LIGO on gravitational wave bursts

Benefits of joint LIGO Virgo coincidence searches for burst and inspiral signals

arxiv:gr-qc/ v1 24 Sep 2006

ANALYSIS OF BURST SIGNALS IN LIGO DATA. Irena Zivkovic, Alan Weinstein

LASER INTERFEROMETER GRAVITATIONAL WAVE OBSERVATORY - LIGO - CALIFORNIA INSTITUTE OF TECHNOLOGY MASSACHUSETTS INSTITUTE OF TECHNOLOGY

LIGO Status and Advanced LIGO Plans. Barry C Barish OSTP 1-Dec-04

Searching for Gravitational Waves from Binary Inspirals with LIGO

arxiv: v1 [gr-qc] 25 Jul 2008

arxiv:gr-qc/ v1 10 Oct 2001

Searching for gravitational waves from binary inspirals with LIGO

Detecting the next Galactic supernova

LIGO Observational Results

Searching for Gravitational Waves from Coalescing Binary Systems

Status of LIGO. David Shoemaker LISA Symposium 13 July 2004 LIGO-G M

Results from LIGO Searches for Binary Inspiral Gravitational Waves

M.-A. Bizouard, F. Cavalier. GWDAW 8 Milwaukee 19/12/03

Search for the gravity wave signature of GRB030329/SN2003dh

The LIGO Project: a Status Report

Searching for gravitational waves. with LIGO detectors

Gravitational-Wave Data Analysis

Search for Gravitational Wave Transients. Florent Robinet On behalf of the LSC and Virgo Collaborations

Search for compact binary systems in LIGO data

The LSC glitch group: monitoring noise transients during the fifth LIGO science run

GRB-triggered searches for gravitational waves from compact binary inspirals in LIGO and Virgo data during S5/VSR1

Status and Prospects for LIGO

Applications of change point detection in Gravitational Wave Data Analysis. Soumya D. Mohanty AEI

Outline. 1. Basics of gravitational wave transient signal searches. 2. Reconstruction of signal properties

NEVER IGNORE A COINCIDENCE

Search for gravitational radiation with the Allegro and Explorer detectors

Gravity. Newtonian gravity: F = G M1 M2/r 2

Gravitational Wave Burst Source Direction Estimation Using Time and Amplitude Information

Gravitational Wave Burst Searches

arxiv:gr-qc/ v1 24 Dec 2004

Towards the first search for a stochastic background in LIGO data: applications of signal simulations

LIGO s continuing search for gravitational waves

Multiresolution techniques for the detection of gravitational-wave bursts

The LIGO Experiment Present and Future

Overview Ground-based Interferometers. Barry Barish Caltech Amaldi-6 20-June-05

GW150914: Observation of gravitational waves from a binary black hole merger

What can LIGO detect? Abstract

Chirplets pour la détection des ondes gravitationnelles

Figure 1: An example of the stretching and contracting of space-time caused by a gravitational wave.

Searches for Gravitational waves associated with Gamma-ray bursts

All-sky LIGO Search for Periodic Gravitational Waves in the Fourth Science Run (S4)

Tiffany Summerscales Ph.D. Department of Physics Andrews University Haughey Hall, 223 Berrien Springs, MI (269)

Data Analysis Pipeline: The Search for Gravitational Waves in Real life

An Aperçu about Gravitational Waves and Data Analysis

The Present Gravitational Wave Detection Effort

Gravitational Wave Astronomy

Gravitational Waves and LIGO

The direct detection of gravitational waves: The first discovery, and what the future might bring

Core-Collapse Supernova Science with Advanced LIGO and Virgo

arxiv:astro-ph/ Oct 2001

arxiv:gr-qc/ v1 16 Feb 2007

arxiv: v1 [astro-ph.im] 11 Jun 2016

Prospects for joint transient searches with LOFAR and the LSC/Virgo gravitational wave interferometers

GRAVITATIONAL WAVE SOURCES AND RATES FOR LISA

GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral

Astrophysical Stochastic Gravitational Waves. Jonah Kanner PHYS 798G March 27, 2007

Strategy for signal classification to improve data quality for Advanced Detectors gravitational-wave searches

Searching for gravitational waves from neutron stars

Data Characterization in Gravitational Waves

M.Alessandra Papa. Max Planck Inst. f. Gravitationsphysik and Univ. of Wisconsin- Milwaukee. ICGC Conference 2007, December Pune

Newtonian instantaneous action at a distance General Relativity information carried by gravitational radiation at the speed of light

Gravitational Wave Detection from the Ground Up

LIGO S2 Inspiral Hardware Injections

Probing the Universe for Gravitational Waves

Discovery of Gravita/onal Waves

An Introduction to Gravitational Waves

Searching for Intermediate Mass Black Holes mergers

Reconstruction of chirp mass in searches for gravitational wave transients

Gravity -- Studying the Fabric of the Universe

arxiv:gr-qc/ v1 24 Sep 2006

LASER INTERFEROMETER GRAVITATIONAL WAVE OBSERVATORY - LIGO - CALIFORNIA INSTITUTE OF TECHNOLOGY MASSACHUSETTS INSTITUTE OF TECHNOLOGY

LIGO Status and Plans. Barry Barish / Gary Sanders 13-May-02

LIGO Present and Future. Barry Barish Directory of the LIGO Laboratory

Joint search for gravitational waves and high-energy neutrinos with the ANTARES, LIGO and Virgo detectors

Gravitational Waves & Intermediate Mass Black Holes. Lee Samuel Finn Center for Gravitational Wave Physics

GRAVITATIONAL WAVE ASTRONOMY

Probing the Universe for Gravitational Waves

DEVELOPMENT OF A DMT MONITOR FOR STATISTICAL TRACKING OF GRAVITATIONAL-WAVE BURST TRIGGERS GENERATED FROM THE OMEGA PIPELINE

Long-term strategy on gravitational wave detection from European groups

Gravitational wave detection with Virgo and LIGO experiment - Case of the long bursts

Parameter estimation for signals from compact binary inspirals injected into LIGO data

LIGO Detection of Gravitational Waves. Dr. Stephen Ng

IGEC toolbox for coincidence search

Event Rates of Gravitational Waves from merging Intermediatemass

Using Non Harmonic Analysis (NHA) to reduce the influences of line noises for GW Observatory

Probing the Universe for Gravitational Waves

Gravitational Waves from Supernova Core Collapse: What could the Signal tell us?

GROUND NOISE STUDIES USING THE TAMA300 GRAVITATIONAL-WAVE DETECTOR AND RELATED HIGHLY-SENSITIVE INSTRUMENTS

Gravitational-wave astrophysics

Synergy with Gravitational Waves

J.R. Sanders. 201 Physics Building, Syracuse, NY

Search for gravitational wave bursts with the first science data from LIGO

THE NEW RUN OF EXPLORER AND NAUTILUS

A template bank to search for gravitational waves from inspiralling compact binaries: II. Phenomenological model

in2p , version 1-14 May 2012

Transcription:

INSTITUTE OF PHYSICS PUBLISHING Class. Quantum Grav. 21 (2004) S1801 S1807 CLASSICAL AND QUANTUM GRAVITY PII: S0264-9381(04)79467-8 Plans for the LIGO TAMA joint search for gravitational wave bursts Patrick J Sutton 1, Masaki Ando 2, Patrick Brady 3, Laura Cadonati 4, Alessandra Di Credico 5, Stephen Fairhurst 3, Lee Samuel Finn 6, Nobuyuki Kanda 7, Erik Katsavounidis 4, Sergey Klimenko 8, Albert Lazzarini 1, Szabolcs Marka 1, John W C McNabb 6, Saikat Ray Majumder 3, Peter R Saulson 5, Hideyuki Tagoshi 9, Hirotaka Takahashi 9,10,11, Ryutaro Takahashi 12, Daisuke Tatsumi 12, Yoshiki Tsunesada 12 and S E Whitcomb 1 1 LIGO Laboratory, California Institute of Technology, Pasadena, CA 91125, USA 2 Department of Physics, University of Tokyo, Hongo, Bunkyo-ku, Tokyo 113-0033, Japan 3 Department of Physics, University of Wisconsin-Milwaukee, Milwaukee, WI 53201, USA 4 LIGO Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139, USA 5 Department of Physics, Syracuse University, Syracuse, NY 13244, USA 6 Center for Gravitational Wave Physics, Department of Physics, The Pennsylvania State University, University Park, PA 16802, USA 7 Department of Physics, Graduate School of Science, Osaka City University, Sumiyoshi-ku, Osaka, Osaka 558-8585, Japan 8 Department of Physics, University of Florida, Gainesville, FL 32611, USA 9 Department of Earth and Space Science, Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043, Japan 10 Graduate School of Science and Technology, Niigata University, Niigata, Niigata 950-2181, Japan 11 Yukawa Institute for Theoretical Physics, Kyoto University, Kyoto, Kyoto 606-8502, Japan 12 National Astronomical Observatory of Japan, Mitaka, Tokyo 181-8588, Japan E-mail: psutton@ligo.caltech.edu Received 19 April 2004 Published 30 September 2004 Online at stacks.iop.org/cqg/21/s1801 doi:10.1088/0264-9381/21/20/023 Abstract 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. PACS number: 04.80.Nn 0264-9381/04/201801+07$30.00 2004 IOP Publishing Ltd Printed in the UK S1801

S1802 P J Sutton et al 1. Introduction At present several large-scale interferometric gravitational wave detectors are operating or are being commissioned: GEO [1], LIGO [2], TAMA [3] and Virgo [4]. In addition, numerous resonant mass detectors have been operating for a number of years [5, 6]. Cooperative analyses by multiple observatories could be valuable for making confident detections of gravitational waves and for extracting the most information from them. This is particularly true for gravitational wave bursts (GWBs) from systems such as core-collapse supernovae [8 11], black hole mergers [12, 13] and gamma-ray bursters [14], for which we have limited theoretical knowledge to guide us. Advantages of coincident observations include a decreased rate of false alarms from random noise fluctuations, the ability to locate a source on the sky, the ability to extract polarization information and better statistics on signal parameters [15 17]. Independent observations using different detector hardware and software also decrease the possibility of error or bias. The TAMA and LIGO Collaborations are currently conducting joint searches for gravitational waves in the LIGO Science Run 2 (S2) and TAMA Data Taking Run 8 (DT8) data, which were collected between 14 February and 14 April 2003 [2, 3]. Three classes of gravitational wave transients are being sought: (i) unmodelled GWBs from the gamma ray burst event GRB 030329 [18 21]; (ii) inspiral signals from galactic binary neutron star systems (see, for example, [22, 23]); and (iii) unmodelled GWBs without an electromagnetic trigger (see, for example, [24, 25]). In this paper we report on the planned untriggered search for unmodelled GWBs. This is the first four-interferometer coincidence analysis for GWBs 13, involving detectors at three sites. We will focus on some of the challenges facing this joint search, in particular the differences in the noise spectra and in the alignments of the LIGO and TAMA detectors. As we shall see, these motivate a coincidence search between LIGO and TAMA that involves a trade-off between the false alarm rate, the detection efficiency and the bandwidth. The result is a search for high-frequency GWBs (in the range [700 2000] Hz) characterized by a low false alarm rate. We adopt this conservative approach to detection by the LIGO TAMA detectors in order to gain experience sharing data and performing combined analyses. This search could form a prototype for more comprehensive collaborative analyses in the future. In section 2 we review the procedure used in the joint search for GWBs. In section 3 we examine some of the challenges facing the joint analysis and their implications. In section 4 we conclude with a few brief comments. 2. Untriggered bursts analysis The LIGO TAMA joint search for GWBs is structured in the style of the LIGO S1 GWB analysis [24] and the TAMA DT8 GWB analysis [25]. Our goal is to detect GWBs or place upper limits on the rate of GWBs detectable by our instruments. Since we aim to be sensitive to the widest variety of waveforms which contain significant energy in the frequency range of our detectors, we choose to make minimal assumptions about the GWB waveforms; in particular we do not use templates or matched filters. Instead, a variety of non-template based burst detection algorithms are employed to locate candidate GWBs [25 29]. The analysis procedure is straightforward: (i) Each collaboration independently generates lists of candidate GWBs. These events are characterized by (at minimum) a start and/or peak time, a duration and an amplitude or 13 See [5 7] for GWB searches involving networks of resonant mass detectors.

Plans for the LIGO TAMA joint search for gravitational wave bursts S1803 SNR. The LIGO events are also characterized by a frequency range, whereas the TAMA bursts detection scheme uses a fixed set of frequencies between 700 and 2000 Hz [25]. The event generation procedure may include vetoes of candidate events based on auxiliary data, such as from environmental monitors. In addition, events which are in coincidence in the three LIGO interferometers are subjected to a coherence test, in which the crosscorrelation of the data from each pair of LIGO interferometers is required to exceed a certain threshold [30]. (Further study is required before including TAMA events in this coherence test due to the different alignment of TAMA; see section 3.2.) (ii) The event lists from the four detectors are compared and candidate GWBs are selected by requiring coincident detection in each interferometer (LIGO and TAMA). (iii) The rate of accidental coincidences due to the background noise is estimated by repeating the coincidence and coherence tests after artificially shifting the time stamps of one or more lists of events. It is intended that the event generation thresholds in step (i) will be kept high enough so that the average background rate will be less than one event over the observation time (approximately 250 h). (iv) An excess of coincident events above the background will be investigated as a possible detection; in the absence of an excess an upper limit on the rate of detectable events will be set using the background estimate. (v) The efficiency of the LIGO TAMA network at detecting real GWBs will be estimated by adding various simulated signals (astrophysical and ad hoc) to the data in Monte Carlo experiments. It should be noted that the LIGO collaboration is also performing a separate coincidence search for GWBs using the LIGO S2 data alone below 1.1 khz. Candidate GWBs from the two analyses (if any) will be compared. 3. Challenges for LIGO TAMA A coincidence analysis involving LIGO and TAMA is more complicated than, for example, a coincidence analysis involving the LIGO detectors only. The main complications are that the LIGO and TAMA detectors have different noise spectra and different alignments. 3.1. Sensitivity versus frequency Figure 1 shows representative noise spectra from the LIGO Hanford 4 km detector and the TAMA detector during S2/DT8 (the other LIGO detectors are similar to the one shown). The individual detectors have maximum sensitivity at very different frequencies about 250 Hz for LIGO and 1300 Hz for TAMA. If we ignore differences in alignment, then requiring four-fold coincident detection means that the overall sensitivity of the network is limited by the least sensitive detector. This is because our coincident analysis requires a signal to be detected by all interferometers; hence, we choose to focus our search on the minimum of the noise envelope, which occurs around 1000 Hz. Specifically, we require that the frequency range of LIGO events overlaps [700, 2000] Hz; this is also the frequency range analysed in the production of TAMA events. Restricting the frequency range will reduce the rate of false alarms due to coincident noise fluctuations, while preserving sensitivity to real GWBs that are detectable by both LIGO and TAMA. Note that any low-frequency GWBs which are detectable by LIGO but not TAMA will be found in the LIGO collaboration s separate low-frequency search. Thus, there is no danger of

S1804 P J Sutton et al 10 17 H1 TAMA 10 18 Strain Noise (Hz 1/2 ) 10 19 10 20 10 21 10 22 10 2 10 3 Frequency (Hz) Figure 1. Representative amplitude noise spectra for the LIGO Hanford 4 km (H1) and TAMA detectors. Events from each detector are used in this analysis if they intersect the frequency range [700 2000] Hz, where each interferometer has approximately the same noise level. a detectable low-frequency GWB being missed due to our choice of frequency range for the LIGO TAMA coincidence. 3.2. Antenna patterns Figure 2 shows the antenna patterns ρ F+ 2 + F 2 (1) for the LIGO Hanford and TAMA detectors. (The LIGO Livingston antenna pattern is very similar to that of the LIGO Hanford detectors.) These detectors are strongly non-aligned in that the areas of maximum sensitivity for LIGO overlap minima of TAMA sensitivity, and vice versa. That is, LIGO and TAMA look with best sensitivity at different parts of the sky. This complicates LIGO TAMA coincidence; if we ignore differences in noise levels, then requiring four-fold coincident detection means that the overall sensitivity of the network to a given point on the sky is limited by the minimum of the antenna patterns. The impact of this alignment difference on the network sensitivity can be estimated via a simple Monte Carlo experiment. Let us assume each interferometer has the same sensitivity for some fiducial signal of interest. We model this detection efficiency as a simple sigmoid in the logarithm of the signal amplitude, as shown in figure 3. 14 We then generate linearly polarized signals of random amplitude and polarization angle, distributed isotropically across the sky. We project each signal onto the antenna pattern of each instrument, and compute the coincident detection efficiency for different combinations of interferometers 15. Figure 3 shows the resulting efficiencies as a function of the intrinsic signal amplitude. For example, 14 We use equation (5.10) of [24] with a = 0.1,b = 0. Choosing b = 0 is the case in which all four detectors have the same sensitivity; this is approximately true around 1 khz (see figure 1), and it is the simplest choice for illustrating the effect of the alignment separately from differences in the noise levels. Sigmoid widths of order from a 0.07 to a 0.15 are typical of the values found in the LIGO S1 analysis [24] and in the TAMA DT8 analysis. The network sensitivities depend only weakly on a. 15 See [31] for the efficiency of an alternative network detection algorithm.

Plans for the LIGO TAMA joint search for gravitational wave bursts S1805 Figure 2. Antenna patterns ρ (1) in Earth-based coordinates (see, for example, [32]). The upper plot is for the LIGO Hanford detectors, while the lower plot is for TAMA. Light areas denote sky directions of the highest sensitivity. The directions of LIGO s maximum sensitivity lie close to areas of TAMA s worst sensitivity and vice versa. the minimum signal amplitude required for the three (nearly co-aligned) LIGO detectors to have a 50% chance of simultaneous detection is approximately 4.7/2.9 1.6 times larger than the amplitude for 50% detection by a single interferometer (exemplified here by TAMA). The minimum amplitude for simultaneous detection by both LIGO and TAMA is approximately 2.6 times the single-detector amplitude, or 1.6 times the LIGO amplitude. The network efficiency is a function of the type of signal studied, and hence figure 3 is only illustrative of what is expected for LIGO TAMA. Nevertheless, the loss in efficiency when adding detectors to the network is generic when requiring coincident detection. Either collaboration could achieve better sky-averaged sensitivity by not requiring coincidence with the other. However, the loss in efficiency is accompanied by a drop in the false alarm rate. For example, typical false-alarm rates at the single-interferometer level in this analysis are in the range 0.1 1 Hz. Assuming a LIGO TAMA coincidence window of order 0.1 s, each additional detector added to the network will lower the coincident false alarm rate by a factor of 10 100. This false alarm reduction brings with it improved sensitivity to regions of the sky to which both LIGO and TAMA are simultaneously sensitive and is the principal scientific justification for our joint search 16. Another consequence of the difference in LIGO and TAMA alignments is that the detectors are sensitive to different polarization combinations of a gravitational wave. This effect could be particularly important when cross-correlating events between LIGO and TAMA. Coherence tests have been shown to be a powerful means of reducing coincidence false alarm rates in 16 An alternative approach would be to lower the thresholds for event generation to keep the network false alarm rate constant as detectors are added to the network. Whether the resulting gain in efficiency would be enough to offset the loss from adding differently aligned detectors depends on the relationship between the false alarm rate and efficiency for each detector.

S1806 P J Sutton et al 1 0.9 0.8 Detection Efficiency 0.7 0.6 0.5 0.4 0.3 0.2 Zenith 0.1 <TAMA> <LIGO> <LIGO TAMA> 0 10 1 10 0 10 1 10 2 Strain Amplitude (arbitrary units) Figure 3. Sample efficiency estimates for the LIGO TAMA network from Monte Carlo simulations. Each interferometer is assumed to have identical sensitivity to a fiducial source (represented by the zenith curve) when that source is optimally positioned and oriented with respect to the interferometer. The other curves show the efficiency to a population of such sources isotropically distributed over the sky and with random polarization. The TAMA curve shows the detection efficiency of the TAMA detector alone, the LIGO curve shows the efficiency for coincident detection by all three LIGO detectors and the LIGO TAMA curve shows the efficiency for coincidence detection by TAMA and all three LIGO interferometers. LIGO [30], and extending these tests between LIGO and TAMA could reduce the network false alarm rate further. This issue is under study. 4. Summary The LIGO and TAMA collaborations are conducting joint analyses of the LIGO S2 TAMA DT8 data collected during February April 2003. There are three analyses focusing on gravitational wave transients: unmodelled GWBs from GRB 030329, inspiral signals from galactic neutron star binaries and an untriggered search for unmodelled GWBs. We have reviewed the plans for the untriggered search for GWBs and some of the complications of comparing LIGO and TAMA events. We take a conservative approach to detection, requiring candidate GWBs to be seen in coincidence by all four interferometers at frequencies where all four have comparable sensitivity. While this is expected to result in lower sensitivity of the network compared to the individual interferometers, it is the most conservative approach with respect to false alarms. Acknowledgment The LIGO Laboratory operates under cooperative agreement with the National Science Foundation, grant PHY-0107417.

Plans for the LIGO TAMA joint search for gravitational wave bursts S1807 References [1] Willke B et al 2004 Class. Quantum Grav. 21 S417 [2] Sigg D 2004 Class. Quantum Grav. 21 S409 [3] Takahashi R 2004 Class. Quantum Grav. 21 S403 [4] Acernese F et al 2004 Class. Quantum Grav. 21 S385 [5] Allen Z A et al 2000 Phys. Rev. Lett. 85 5046 [6] Astone P et al 2003 Phys. Rev. D 68 022001 [7] Amaldi E et al 1989 Astron. Astrophys. 216 325 [8] Zwerger T and Muller E 1997 Astron. Astrophys. 320 209 [9] Dimmelmeier H, Font J and Muller E 2002 Astron. Astrophys. 388 917 [10] Dimmelmeier H, Font J and Muller E 2002 Astron. Astrophys. 393 523 [11] Ott C, Burrows A, Livne E and Walder R 2004 Astrophys. J. 600 834 [12] Flanagan E and Hughes S 1998 Phys. Rev. D 57 4535 [13] Flanagan E and Hughes S 1998 Phys. Rev. D 57 4566 [14] Mészáros P 2002 Ann. Rev. Astron. Astrophys. 40 137 [15] Finn L S 2001 Phys. Rev. D 63 102001 [16] Pai A, Dhurandhar S and Bose S 2001 Phys. Rev. D 64 042004 [17] Arnaud N et al 2003 Phys. Rev. D 68 102001 [18] Mészáros P 2003 Nature 423 809 [19] Price P A et al 2003 Nature 423 844 [20] Hjorth J et al 2003 Nature 423 847 [21] Mohanty S D, Marka S, Rahkola R, Mukherjee S, Leonor I, Frey R, Cannizzo J and Camp J 2004 Class. Quantum Grav. 21 S765 [22] Abbott B et al 2004 Analysis of LIGO data for gravitational waves from binary neutron stars Phys. Rev. D 69 122001 (Preprint gr-qc/0308069) [23] Takahashi H and Tagoshi H 2004 Class. Quantum Grav. 21 S769 [24] Abbott B et al 2004 Phys. Rev. D 69 102001 [25] Ando M et al 2004 Class. Quantum Grav. 21 S1679 [26] Sylvestre J 2002 Phys. Rev. D 66 102004 [27] Anderson W et al 2001 Phys. Rev. D 63 042003 [28] Klimenko S et al 2004 Class. Quantum Grav. 21 S1685 [29] McNabb J et al 2004 Class. Quantum Grav. 21 S1705 (Preprint gr-qc/0404123) [30] Cadonati L 2004 Class. Quantum Grav. 21 S1695 [31] Kanda N 2003 Class. Quantum Grav. 20 S761 [32] Althouse W et al 2001 Rev. Sci. Instrum. 72 3086