Regularized coherent network analysis pipeline for triggered - PowerPoint PPT Presentation

Regularized coherent network analysis pipeline for triggered searches Kazuhiro Hayama Center for Gravitational Wave Astronomy University of Texas at Brownsville Malik Rakhmanov, Shantanu Desai(Penn State) Soumya Mohanty(UTB) LIGO-G060653-00-0

Burst Triggered Search CGWA • Gamma ray burst, Neutrino burst, X-ray burst from transient astronomical events • The time and sky location of these events can be estimated by other astronomical observations such as HETE, SuperKamiokande, Chandra etc. Triggered search already going • GRB - Cross Correlation method ----- S. Mohanty’s talk • SGR - Excess Power method ----- L. Matone’s talk Our approach ----- regularized coherent network method Outline • Data Conditioning • Event Selection based on regularized coherent network analysis • Analysis • Detection Efficiency • Accuracy of waveform estimation LIGO-G060653-00-0

Data for Demonstration CGWA Data sky map likelihood sky map Conditioning post-processing 4 data at H1-H2-L1-GEO with same simulated detector noise(right figure). A burst signal is injected. ! 18 10 Band-pass filtered(64-2000Hz) signals Simulated line features: sinusoids strain noise spectrum(Hz ) -1/2 ! 20 ()10 ! 20 2 10 1 strain ! 22 10 0 ! 1 ! 24 10 ! 2 0 0"02 0"04 0"06 0"08 0"1 0"12 0"14 ! 26 time 10 2 3 10 10 frequency(Hz) LIGO-G060653-00-0

Regularized coherent network analysis pipeline CGWA Data sky map likelihood sky map Conditioning post-processing Time domain method Wavelet-based method band pass filtered at 64Hz-2000Hz ! 20 Time domain method 5 x 10 strain • Time domain noise floor 0 whitening S. Mukherjee CQG 21 (2004) ! 5 S1783 0 0.02 0.04 0.06 0.08 0.1 0.12 • Remove lines by Median Based after conditioning 0.5 Line Tracker S. Mohanty CQG 19 (2002) 1513 0 ! 0.5 0 0.02 0.04 0.06 0.08 0.1 0.12 LIGO-G060653-00-0 time(sec)

Regularized coherent network analysis pipeline CGWA Data sky map likelihood sky map Conditioning post-processing Time domain method Wavelet-based method band pass filtered at 64Hz-2000Hz +,-.*/,00*123456.*,4*6478 ! 200078 ! 20 Wavelet-based method ( )*10 • Select frequency region to strain 0 analyze by nulling around lines • In frequency region, spectrum is ! ( 0 0"02 0"04 0"06 0"08 0"1 0"12 estimated by wavelet de-noising. • Whiten data using estimated ,1456*9:-.242:-2-; 0"( spectrum 0 ! 0"( 0 0"02 0"04 0"06 0"08 0"1 0"12 LIGO-G060653-00-0 time(sec)

Regularized coherent network analysis pipeline CGWA Data sky map likelihood sky map Conditioning post-processing Detector output 1. Data is divided into chunks 2. generate skymap at each chunk rank defficiency Ill-posed problem R@source location One Solution-- Tikhonov regularization M. Rakhmanov CQG 23 (2006) S673 (h) signal included LIGO-G060653-00-0

Detection Efficiency CGWA Data sky map likelihood sky map Conditioning post-processing Receiver Operating Characteristic Curve Simulation SNR(H1,H2,L1,G1)= 1 (11.3,11.3, 13.9, 9.3) 5000 trials in which a burst is in the simulated noise 7etection ;ro<a<ility 0.8 (sampling rate = 4096Hz) 0.6 The burst: black hole (8.5, 8.5, 10.4, 7.0) merger 0.4 5000 trials in noise data 0.2 (7.3, 7.3, 9.0, 6.0) for 1 year observation, 0 1 in 876 triggers with confidence 95% ! 1 0 1 10 10 10 -1 (hour ) false alar/ rate 1123o5r6 LIGO-G060653-00-0

Effect of Data Conditioning CGWA ! 4 x 10 1 ! 80 3.4 ! 60 3.2 0.8 3 ! 40 Latitude (deg) detection probability Without 2.8 ! 20 0.6 2.6 0 DC 0.4 2.4 20 2.2 40 0.2 2 60 1.8 80 0 ! 4 ! $ ! 2 ! 1 0 10 10 10 10 10 ! 150 ! 100 ! 50 0 50 100 150 false alarm probability longitude (deg) 1 0.8 7etection 1ro3a3ility With 0.6 Up DC 0.4 0.2 0 ! 4 ! 3 ! 2 ! 1 0 10 10 10 10 10 false alar0 1ro3a3ility LIGO-G060653-00-0

sky map whitening CGWA Data sky map likelihood sky map Conditioning post-processing standard deviation:after standardized mean noise map standard deviation:noise map standard deviation:raw sky map Preliminary location of minimum after standardized > location of minimum before standarized KL basis component number LIGO-G060653-00-0

Accuracy of Waveform Estimation CGWA Data sky map likelihood sky map Conditioning post-processing signal gain=1 : corresponds to SNR(H1,H2,L1,L2)= (11.3,11.3, 13.9, 9.3) ! 20 2 x 10 blue:original green:reconstructed strain h+ 0 ! 2 0 0.02 0.04 0.06 0.08 0.1 0.12 time(sec) ! 20 5 x 10 hx 0 ! 5 0 0.02 0.04 0.06 0.08 0.1 0.12 LIGO-G060653-00-0

Accuracy of Waveform Estimation CGWA Data sky map likelihood sky map Conditioning post-processing To de-noise, wavelet-based waveform estimation method is used (red) Hayama, Fujimoto CQG 23 (2006) S9 ! 20 2 x 10 blue:original red:estimated strain h+ 0 ! 2 0 0.02 0.04 0.06 0.08 0.1 0.12 time(sec) ! 20 * x 10 hx 0 ! * 0 0.02 0.04 0.06 0.08 0.1 0.12 LIGO-G060653-00-0

Accuracy of Waveform Estimation CGWA Using H1-L1-VIRGO..... these detectors have comparable sensitivity Supernova signal at (112, -30) at 2kpc distant from Earth SNR(H1,L1,V)=(11.1, 14.0, 3.5) Dimmelmeier et al. A1B1G1_R ! 20 # )*10 blue:original red:estimated strain 0 ! # 0"0#6 0"0#8 0"06 0"062 0"064 0"066 0"068 0"0( time(sec) h waveform(duration=14msec) of the burst from 2kpc at (112,-30) can be estimated within MSE of 0.3 LIGO-G060653-00-0

Reconstruction of Inspiral signal CGWA Idea: Join continuous reconstructed h+,hx segments ------>> we can get arbitrary signal’s h+, hx time series. Example : Inspiral signal(1M-1M),1Mpc reconstructed hx x 10 ! 20 4 2 0 ! 2 ! 4 ! 6 0.2 0.4 0.6 0.8 Matched filter on h+, hx Theoretical SNR SNR (with rough spectrum estimation) H1:16.6 Hp:7.8 H2:16.6 Hc:3.2 L1:17.8 GEO:4.5 LIGO-G060653-00-0

Noise spectrum of reconstructed hx, h+ CGWA h+ hx ! 21 ! 21 10 10 strain noise spectrum(Hz ) strain noise spectrum(Hz ) -1/2 -1/2 ! 22 ! 22 10 10 ! 2$ ! 2$ 10 10 1 2 $ 4 1 2 $ 4 10 10 10 10 10 10 10 10 frequency (Hz) frequency (Hz)

Summary and future work CGWA Regularized coherent network analysis pipeline for triggered search has been developed From the pipeline, we get not only a detection statistic but also the reconstructed polarization waveforms. Using wavelet-based waveform estimation, we showed accuracy of estimated waveform We can get h+, hx time series for any given direction on the sky and can search for signals other than bursts.(e.g. template based search) In progress: application to real data LIGO-G060653-00-0

sky map whitening CGWA Data sky map likelihood sky map Conditioning post-processing Location of minimum R Standard deviation of skymap LIGO-G060653-00-0

Reconstruction of Inspiral signal CGWA Waveform 1 x 10 ! 21 SNR:8.65 0.( strain ! 20 5 x 10 0 ! 0.( ! 1 0.02 0.04 0.06 0.08 0.1 time(sec) strain 0 SNR H1:4.34 H2:8.65 L1:8.6 ! 5 GEO:1.83 0 0.02 0.04 0.06 0.08 0.1 0.12 time(sec)

Accuracy of Waveform Estimation CGWA Averaged mean square error normalized signal energy as a function of signal gain. # of trials at each SNR is 1000. signal gain=1 corresponds to SNR (H1,H2,L1,G1)= (11.3,11.3, 13.9, 9.3) h+ mean square error 1.5 1 0.5 0 0.5 1 1.5 2 signal gain hx mean square error 10 5 0 0.5 1 1.5 2 signal gain

Detection Efficiency CGWA Data Detection Efficiency Event Selection Conditioning Event Reconstruction Comparison of trigg and untrigg Detection probability vs False alarm probability ?? SNR(H1,H2,L1,G1)= 1 1 (11.3,11.3, 13.9, 9.3) Receiver Operating 7etection ;ro<a<ility 7,0,8094:-;.4<)<9*90= 0.8 0.8 Characteristic Curve 0.6 0.6 To obtain ROC, the numerical (8.5, 8.5, 10.4, 7.0) simulation consists of 5000 0.4 0.4 trials in which the burst is in simulated LIGO noise, and 0.2 0.2 (7.3, 7.3, 9.0, 6.0) 5000 trials in only noise. 0 0 ! 1 0 1 ! 1 0 1 10 10 10 10 10 10 false alar/ rate 1123o5r6 ()*+,-)*)./-.)0,-112345.6

Detection Efficiency CGWA Data Detection Efficiency Event Selection Conditioning Event Reconstruction Untriggered Search Detection probability vs SNR(H1,H2,L1,G1)= False alarm probability ?? 1 (11.3,11.3, 13.9, 9.3) Receiver Operating 7,0,8094:-;.4<)<9*90= 0.8 Characteristic Curve (8.5, 8.5, 10.4, 7.0) 0.6 To obtain ROC, the numerical simulation consists of 5000 0.4 trials in which the burst is in simulated LIGO noise, and 0.2 5000 trials in only noise. (7.3, 7.3, 9.0, 6.0) 0 ! 1 0 1 10 10 10 ()*+,-)*)./-.)0,-112345.6

Regularized coherent network analysis pipeline CGWA Data Detection Efficiency Event Selection Conditioning Event Reconstruction Statistical Property of R :tatisti7a- pr)pert8 )4 3 SNRs of signal at each detectors are '00 Noise Only (H1,H2,L1,G1)= ( 11.3,11.3, 13.9, 9.3 ) 500 4re5uen78 nu+9er Histograms are normalized by mean 400 Noise plus Signal and variance of R of noise only data 300 200 For event selection, Threshold of detection is decided to 100 satisfy adequate false alarm rate and 0 detection probability ! 30 ! 20 ! 10 0 10 n)r+a-i/ed 3

Regularized coherent network analysis pipeline Data Detection Efficiency Event Selection Conditioning Event Reconstruction Reconstructed sky maps around the true segment (Center)