LIGO-G Z The Q Pipeline search for gravitational-wave bursts with LIGO Shourov K. Chatterji for the LIGO Scientific Collaboration APS Meeting Dallas, Texas April 25, 2006 APS

LIGO-G Z Overview of the problem For many potential sources of gravitational-waves (core collapse supernovae, binary black hole mergers, etc.), the waveform is not sufficiently well known to permit matched filtering Signals are expected to be near the noise floor of the LIGO detectors, which are subject to occasional transient non-stationarities Searches must be able to keep up with the LIGO data stream using limited computational resources Simple parameterization to describe unmodeled bursts Efficient algorithm to search data from multiple detectors for statistically significant signal energy that is consistent with the expected properties of gravitational radiation APS

LIGO-G Z Parameterization of unmodeled bursts Characteristic amplitude (|| h ||): Normalized waveform (  ): Time (  ), frequency (  ), bandwidth (  t ) and duration (  f ): Time-frequency uncertainty: Quality factor ( Q ) (aspect ratio): APS

LIGO-G Z Signal space and template placement Multiresolution basis of minimum uncertainty waveforms Overcomplete basis is desirable for detection Use matched filtering template placement formalism Tile the targeted signal space with the minimum number of tiles necessary to ensure no more than a given worst case energy loss due to mismatch Naturally yields multiresolution basis similar to discrete dyadic wavelet transform Logarithmic in frequency and Q, linear in time APS

LIGO-G Z Project onto basis of minimum uncertainty waveforms Alternative frequency domain formalism (heterodyne detector) allows efficient computation using the fast Fourier transform Frequency domain bi-square window has near minimum uncertainty with finite frequency domain support The Q transform APS

LIGO-G Z White noise statistics Whitening the data prior to Q transform analysis greatly simplifies the resulting statistics The result is equivalent to a matched filter search for minimum uncertainty waveforms in the whitened data The squared magnitude of Q transform coefficients are chi-squared distributed with 2 degrees of freedom Define the normalized tile energy For white noise, Z is just exponentially distributed Z also corresponds to a matched filter SNR for minimum uncertainty bursts of APS

LIGO-G Z Example Q transforms 20% loss, Q of 4 20% loss, Q of 8 1% loss, Q of 4 1% loss, Q of 8 APS

LIGO-G Z Preliminary look at LIGO data Understand the performance of the method on real data Hanford 2km and 4 km detectors Analyzed 44.4 days of good quality S5 data Hanford 2km, 4km, and Livingston 4km detectors Analyzed 27.9 days of good quality S5 data Searched in frequency from 90 Hz to 1024 Hz Searched in Q from 4 to 64 Single detector threshold: Z > 19 N detector threshold:  Z > 25.5N Tested for time-frequency coincidence between sites 15 ms coincidence window between sites 5 ms coincidence window between Hanford detectors Overlap in frequency APS

LIGO-G Z Preliminary look at LIGO data Accidental rate estimated by non-physical time shifts 100 time shift experiments from -50 to +50 seconds Hanford 2km, 4km, and Livingston 4km 6 events observed at non-zero time shifts Hanford 2km and 4km 1231 events observed at non-zero time shifts Expect additional events at zero time shift due to shared environment of Hanford detectors Consistency tests could also be applied between Hanford detectors Follow up coincident events by looking at auxiliary detector and environmental monitor data APS

LIGO-G Z Significance distribution of triggers

LIGO-G Z Q and frequency distribution of triggers Q and Frequency histogram of single detector triggers Normalized to the histogram of the Q and frequency of all of the measurement tiles H1 and L1 exhibit large trigger rates at low frequency and low Q H1 L1H2

LIGO-G Z Double coincident significance distribution No threshold

LIGO-G Z Double coincident significance distribution Z > 19 and  Z > 25.5

LIGO-G Z Double coincident significance distribution Z > 19 and  Z > 51 Frequency > 90 Hz

LIGO-G Z Triple coincident significance distribution No threshold

LIGO-G Z Triple coincident significance distribution Z > 19 and  Z > 76.5

LIGO-G Z Triple coincident significance distribution Z > 19 and  Z > 76.5 Frequency > 90 Hz

LIGO-G Z Double coincident trigger rate

LIGO-G Z Triple coincident trigger rate

LIGO-G Z Zero lag results Hanford 2km, 4km, and Livingston 4km 6 events observed at non-zero time shifts ~0 events expected at zero time shift 0 events observed at zero time shift Hanford 2km and 4km 1231 events observed at non-zero time shifts ~12 events expected at zero time shift due to accidental coincidence More events expected due to shared environment 78 events observed at zero time shift 34 events are not seen in the Livingston 4km detector Hanford consistency tests not yet applied

LIGO-G Z Example triple coincidence event 0 triple coincident events observed above threshold 9 triple coincidence events observed below threshold but above the online trigger production threshold All due to a population of low frequency glitches These glitches extend into our most sensitive region They seem to be a common feature of the data Suggest they become a focus of commissioning work H2H1L1

LIGO-G Z Sensitivity to inspiral hardware injections Study hardware injections since January 16th Provides rough estimate of sensitivity Applied same H1H2 thresholds 4 out of 4 1.4/1.4 M  injections detected at 2 Mpc 4 out of 5 1.4/1.4 M  injections detected at 5 Mpc 3 out of 4 10/10 M  injections detected at 40 Mpc Example QScan of a 1.4/1.4 M  injection at 5 Mpc: H1H2L1

LIGO-G Z Follow up of interesting events Q transform can also be applied to follow up events Used to identify statistically significant signal content in Gravitational-wave data Auxiliary detector data Environmental monitoring data Example spectrograms of a non-zero time shift event: APS H2H1

LIGO-G Z Consistency tests Require that the difference between the signals in the two Hanford detectors be consistent with noise APS H2 H1 H2 - H1

LIGO-G Z Outlook LIGO has reached its design sensitivity of an RMS strain of integrated over a 100 Hz band and is now collecting one year of coincident science data The LIGO detectors continue to undergo improvements in sensitivity Search algorithms for unmodeled gravitational-wave bursts are now running in real time on current data Many of these same tools are also being applied to identify and exclude anomalous detector behavior Follow-up tests of interesting events, simulated signals, and detector anomalies are currently under development Tests for consistency of candidate events in multiple detectors are currently under development APS

LIGO-G Z Future directions These results were produced with a simplified single detector Q Pipeline running under Onasys The coherent version Q Pipeline applied to the S2 H1H2 search and a follow-up consistency test will both be automated and incorporated into the online analysis The automated production of Q Scans for interesting events will also be incorporated into the online analysis A summer student will be working on a clustering extension to improve the detectability of bursts that are not well localized in the time-frequency plane The possibility of integrating fully coherent network analysis methods are also being considered A previous code review has been done for the single detector pipeline and more external review is needed

LIGO-G Z Documentation More information on the Q Pipeline is available at The technical details of the Q Pipeline (including validation and characterization studies on white noise and simulated LIGO data) are available at The code used for this study is available at (it is a simplified version of the code in lscsoft) More information on Q Scan is available at